Endocrine Disrupting Chemicals in Food (Woodhead Publishing Series in Food Science, Technology and Nutrition)

Endocrine-disrupting chemicals in food

Related titles: Pesticide, veterinary and other residues in food (ISBN 978-1-85573-734-1) This wide-ranging text sums up the wealth of recent research on assessing the risks from pesticide, veterinary and other residues in food, and effective means for detecting and controlling them. Part I covers targeted and rapid methods for analysing residues in food together with the use of good agricultural practice and HACCP systems in managing them. Part II looks at veterinary residues, covering their safety, toxicology and detection. Part III examines pesticides, with chapters on surveillance and detection methods for fungicides and herbicides. In the final part, there are chapters summarising a wide range of other chemical residues in food, from xenoestrogens/endocrine disruptors and dietary estrogens, to polycyclic aromatic hydrocarbons, dioxins and polychlorinated biphenyls. Food chemical safety. Volume 1: Contaminants (ISBN 978-1-85573-462-3) Chemical contaminants in food, from pesticides and veterinary drug residues to contamination from food packaging, are a major concern for the food industry. Written by a distinguished international team of contributors, this authoritative collection describes the main chemical contaminants, their health implications, how they contaminate food products, methods for their detection and how such contaminants can be controlled. Chemical migration and food contact materials (ISBN 978-1-84569-029-8) Most food sold is packaged, and although packaging has many positive benefits, constituents from it may migrate into the food, possibly exposing consumers to chemicals which may be carcinogenic or genotoxic. As demand for pre-packaged food and ready meals increases, the potential for consumer exposure to these chemicals may also increase. This collection surveys key research in the area. Parts I and II review regulation of food contact materials and the latest developments in analysis of food contact materials and exposure estimation. Part III covers different food contact materials in detail and includes case studies of specific packaging types. Details of these books and a complete list of Woodhead’s titles can be obtained by: • •

visiting our website at www.woodheadpublishing.com contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 130; address: Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK)

Endocrine-disrupting chemicals in food Edited by Ian Shaw University of Canterbury, Christchurch, New Zealand

Oxford

Cambridge

New Delhi

Published by Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK www.woodheadpublishing.com Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi – 110002, India Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2009, Woodhead Publishing Limited and CRC Press LLC © 2009, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-218-6 (book) Woodhead Publishing ISBN 978-1-84569-574-3 (e-book) CRC Press ISBN 978-1-4200-7435-2 CRC Press order number: WP7435 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acidfree and elemental chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by SNP Best-set Typesetter Ltd., Hong Kong Printed by TJ International Limited, Padstow, Cornwall, UK

Contents

Contributor contact details ......................................................................... xiii Introduction .................................................................................................. xix Acknowledgements ...................................................................................... xxiii

Part I 1

Endocrine disruptors, health and behaviour

The effect of dietary endocrine disruptors on the developing fetus ............................................................................ I. Shaw, University of Canterbury and University of Auckland, New Zealand; B. Balakrishnan and M. D. Mitchell, University of Auckland, New Zealand 1.1 Introduction .............................................................................. 1.2 The effects of endocrine-disrupting chemicals on human development ................................................................ 1.3 The metabolism of endocrine disruptors .............................. 1.4 Endocrine disruptors in general ............................................ 1.5 The role of the endocrine system in fetal development ..... 1.6 Effects of endocrine-disrupting chemicals on sexual differentiation and congenital malformations of the developing fetus ............................................................ 1.7 Other effects of endocrine disruptors on the developing fetus ....................................................................... 1.8 Reproductive dysfunction in females .................................... 1.9 Endocrine disruptors and the placenta ................................. 1.10 Studies of placental transfer of xenoestrogens in humans ..................................................................................

3

3 5 7 11 16

16 20 23 23 25

vi

Contents 1.11 1.12 1.13 1.14

2

3

4

Conclusions ............................................................................... Future trends ............................................................................ Sources of further information and advice .......................... References ................................................................................

Human epidemiologic studies of exposure to endocrine-disrupting chemicals and altered hormone levels ........ J. D. Meeker, University of Michigan School of Public Health, USA 2.1 Introduction .............................................................................. 2.2 Persistent organochlorines ...................................................... 2.3 Non-persistent pesticides ........................................................ 2.4 Phthalates .................................................................................. 2.5 Metals ........................................................................................ 2.6 Other emerging compounds of concern ............................... 2.7 Future trends ............................................................................ 2.8 Sources of further information and advice ........................... 2.9 References ................................................................................ Epidemiological evidence on impaired reproductive function and cancer related to endocrine-disrupting chemicals .................. G. Toft, Aarhus University Hospital, Denmark; J. P. Bonde, Copenhagen University Hospital, Denmark 3.1 Introduction .............................................................................. 3.2 Methods ..................................................................................... 3.3 Reproductive abnormalities at birth ..................................... 3.4 Semen quality ........................................................................... 3.5 Menstrual cycle disturbances ................................................. 3.6 Endometriosis and fibroids ..................................................... 3.7 Time to pregnancy ................................................................... 3.8 Cancer studies .......................................................................... 3.9 Conclusions ............................................................................... 3.10 Future trends ............................................................................ 3.11 Sources of further information and advice .......................... 3.12 References ................................................................................ Nutritional phytoestrogens and bone health .................................. W. Wuttke, H. Jarry and D. Seidlová-Wuttke, Georg-August-Universität Göttingen, Germany 4.1 Introduction: trends in bone health ....................................... 4.2 Methods to study the effects of endocrine-disrupting chemicals on bone health ........................................................ 4.3 Effects of endocrine-disrupting chemicals on bone health ............................................................................... 4.4 Phytoestrogens and bone health ............................................

26 26 27 28

36

36 37 42 44 46 49 50 51 51

58

58 59 62 71 73 74 75 75 76 77 78 78 83

83 84 85 86

4.5 4.6 4.7

Part II

5

6

7

Contents

vii

Future trends ............................................................................ Sources of further information and advice ........................... References ................................................................................

96 96 97

Origin and analysis of endocrine disruptors in food products

Endocrine-disrupting chemicals: origins, fates and transmission into the food chain ............................................... L. Connolly, Queen’s University Belfast, UK 5.1 Introduction .............................................................................. 5.2 Natural endocrine-disrupting chemicals ............................... 5.3 Synthetic endocrine-disrupting industrial chemicals ........... 5.4 Fate of endocrine-disrupting chemicals and their transmission into the food chain .................................. 5.5 References ................................................................................. Surveillance of endocrine-disrupting chemicals in foods .............. M. Rose, Food and Environment Research Agency, UK 6.1 Introduction: importance of surveillance of endocrine-disrupting chemicals in food and the environment ....................................................................... 6.2 Environmental risk assessment versus dietary exposure estimates ................................................................... 6.3 Survey design ............................................................................ 6.4 Sampling .................................................................................... 6.5 Surveillance programmes ........................................................ 6.6 Dietary intake calculations and consumer exposure estimates ................................................................... 6.7 Monitoring time trends ........................................................... 6.8 Future trends ............................................................................ 6.9 References ................................................................................ 6.10 Appendix: check plan for sampling ....................................... Advances in chromatography coupled to mass spectrometry-related techniques for analysis of endocrine disruptors in food ............................................................. J.-P. Antignac, F. Courant and B. Le Bizec, Ecole Nationale Vétérinaire de Nantes (ENVN), France 7.1 Introduction .............................................................................. 7.2 Advances in gas chromatography – mass spectrometry-related techniques ............................................ 7.3 Case studies in gas chromatography – mass spectrometry-related techniques ............................................

103 103 104 108 116 121 126

126 127 129 131 135 139 142 142 143 147

149

149 152 153

viii

Contents 7.4 7.5 7.6 7.7 7.8

8

Biosensors for endocrine disruptors ................................................ E. Eltzov, A. Kushmaro and R. S. Marks, Ben-Gurion University of the Negev, Israel 8.1 Introduction .............................................................................. 8.2 General structure of biosensors ............................................. 8.3 Monitoring of specific endocrine-disrupting chemicals in food and environmental fields ........................................... 8.4 Future trends ............................................................................ 8.5 Acknowledgments .................................................................... 8.6 References ................................................................................

Part III

9

10

Advances in liquid chromatography – mass spectrometry-related techniques ............................................ Case studies in liquid chromatography – mass spectrometry-related techniques ............................................ Future trends ............................................................................ Sources of further information and advice .......................... References ................................................................................

162 164 171 174 174 183

183 186 193 199 200 200

Risk assessment of endocrine disruptors in food products

Exposure to endocrine-disrupting chemicals in food .................... B. M. Thomson, Institute of Environmental Science & Research Ltd, New Zealand 9.1 Introduction .............................................................................. 9.2 Selection of endocrine-disrupting chemicals ........................ 9.3 Exposure assessment methodologies .................................... 9.4 Exposure to total estrogenicity .............................................. 9.5 Exposure assessments for endocrine-disrupting chemicals ................................................................................... 9.6 Implication for the food industry .......................................... 9.7 Future trends ............................................................................ 9.8 Sources of further information and advice .......................... 9.9 References ................................................................................

211

Bioassays for the detection of hormonal activities ........................ T. F. H. Bovee and L. A. P. Hoogenboom, RIKILT-Institute of Food Safety, The Netherlands; B. M. Thomson, Institute of Environmental Science & Research Ltd, New Zealand 10.1 Introduction .............................................................................. 10.2 Compounds with hormonal activity ...................................... 10.3 In vivo bioassays for estrogens and androgens ................... 10.4 In vitro bioassays for estrogens and androgens ...................

259

211 213 214 215 218 249 249 250 251

259 262 263 264

Contents 10.5 10.6 10.7 10.8 10.9 11

12

13

In vitro bioassays to determine indirect effects on endogenous hormone levels ............................................. Ah-receptor assays .................................................................. Other hormonal bioassays ...................................................... Conclusions and future trends ............................................... References ................................................................................

Genetics, epigenetics and genomic technologies: importance and application to the study of endocrine-disrupting chemicals ......................................................... L. R. Ferguson and M. Philpott, The University of Auckland and Nutrigenomics, New Zealand 11.1 Introduction .............................................................................. 11.2 Genetic variability in susceptibility to endocrine-disrupting chemicals .............................................. 11.3 The potential of microarrays and related techniques for detection of effects of endocrine-disrupting chemicals ...... 11.4 Gene expression as a component of screening methods for the detection of endocrine-disrupting chemicals in food and environment ...................................... 11.5 Modulation of gene expression by endocrine disrupters through epigenetic mechanisms .......................... 11.6 Future trends ............................................................................ 11.7 Sources of further information and advice .......................... 11.8 Acknowledgements .................................................................. 11.9 References ................................................................................ Computer-aided methodologies to predict endocrine-disrupting potency of chemicals ........................................................................... A. Roncaglioni and E. Benfenati, Istituto di Ricerche Farmacologiche ‘Mario Negri’, Italy 12.1 Introduction .............................................................................. 12.2 In silico methods to predict the endocrine-disrupting potency of a chemical .............................................................. 12.3 Results and implications ......................................................... 12.4 Future trends ............................................................................ 12.5 Sources of further information and advice ........................... 12.6 Acknowledgement ................................................................... 12.7 References ................................................................................

ix

278 278 280 281 282

291

292 294 296

297 300 301 302 302 302 306

306 307 315 317 318 319 319

Endocrine disruptors in breast milk and the health-related issues of breastfeeding ....................................................................... 322 B. G. J. Heinzow, State Agency for Social Services SchleswigHolstein, Germany, and University of Notre Dame, Sydney School of Medicine Australia 13.1 Introduction .............................................................................. 322

x

Contents 13.2 Xenobiotics and transmission into human milk ................... 13.3 Nutritional phytoestrogens in human milk .......................... 13.4 Range and distribution of xenobiotic endocrine disruptors in human milk ........................................................ 13.5 Assessment of exposure .......................................................... 13.6 Risk assessment ........................................................................ 13.7 Current recommendations on breastfeeding ........................ 13.8 Conclusions ............................................................................... 13.9 References ................................................................................

14

Assessing the risks of endocrine-disrupting chemicals .................. A. Beronius, Karolinska Institutet, Sweden; C. Rudén, Royal Institute of Technology, Sweden; A. Hanberg, Karolinska Institutet, Sweden; J. Garai, University of Pecs, Hungary; and H. Håkansson, Karolinska Institutet, Sweden 14.1 Introduction .............................................................................. 14.2 The four model compounds ................................................... 14.3 Regulatory frameworks .......................................................... 14.4 Toxicity data requirements ..................................................... 14.5 Availability and scope of risk assessment guidelines .......... 14.6 Endocrine-disrupting chemical effects assessments ............ 14.7 Toxicological assumptions and principles in effect assessment ...................................................................... 14.8 Development of testing and assessment methods for endocrine-disrupting chemicals .............................................. 14.9 Conclusions ............................................................................... 14.10 References .................................................................................

Part IV

15

324 326 333 343 344 345 346 347 356

356 357 360 363 367 369 372 375 377 378

Examples of endocrine-disrupting chemicals associated with food and other consumer products

Dioxins, polychlorinated biphenyls and brominated flame retardants ................................................................................... L. A. P. Hoogenboom, RIKILT-Institute of Food Safety, The Netherlands 15.1 Introduction .............................................................................. 15.2 Dioxins and dioxin-like polychlorobiphenyls ...................... 15.3 Assessing the toxic effects of dioxins and dioxin-like polychlorobiphenyls ................................................................. 15.4 Analytical methods for dioxins and polychlorobiphenyls .... 15.5 Current exposure to dioxins and polychlorobiphenyls ....... 15.6 Brominated flame retardants ................................................. 15.7 Abbreviations ........................................................................... 15.8 References .................................................................................

383

383 384 388 392 394 395 399 400

Contents 16

17

18

xi

Bisphenol A.......................................................................................... J. E. Goodman and L. R. Rhomberg, Gradient Corporation, USA 16.1 Introduction .............................................................................. 16.2 Bisphenol A migration from packaging materials and containers into food and beverages ....................................... 16.3 Bisphenol A in humans ........................................................... 16.4 Mechanisms of action of bisphenol A ................................... 16.5 Bisphenol A risks to human health ....................................... 16.6 Positions of government bodies on potential human health risks of bisphenol A ..................................................... 16.7 Future trends ............................................................................ 16.8 Sources of further information and advice ........................... 16.9 Acknowledgment ..................................................................... 16.10 References ................................................................................

406

Phytoestrogens and phytosterols ...................................................... S. Hendrich, Iowa State University, USA 17.1 Introduction: phytoestrogens and phytosterols in food and endocrine disruption ........................................................ 17.2 Determining the adverse effects of phytoestrogens and phytosterols ....................................................................... 17.3 Assessing dietary intake of phytoestrogens and phytosterols ....................................................................... 17.4 Assessing the risks and benefits of phytoestrogens and phytosterols in food .......................................................... 17.5 Managing the risks of phytoestrogens and phytosterols in food ................................................................. 17.6 Future trends ............................................................................ 17.7 Sources of further information and advice ........................... 17.8 References ................................................................................

437

Pharmaceuticals ................................................................................... A. H. Piersma and M. Luijten, National Institute for Public Health and the Environment RIVM, The Netherlands; V. Popov and V. Tomenko, Wessex Institute of Technology, UK; M. Altstein, Agricultural Research Organization, Israel; F. Kagampang and H. Schlesinger, Analyst Research Laboratories Ltd, Israel 18.1 Introduction .............................................................................. 18.2 Classification of the mechanisms by which pharmaceuticals affect fecundity ........................................... 18.3 Exposure pathways of pharmaceutical products in food ....................................................................................... 18.4 Pharmaceutical products potentially affecting human fecundity and their assessed mechanism of action ..............

459

406 407 414 416 416 423 427 427 428 428

437 440 449 451 452 452 453 453

459 461 464 467

xii

Contents 18.5 18.6 18.7 18.8 18.9 18.10 18.11 18.12 18.13 18.14 18.15 18.16

19

20

21

Non-steroidal anti-inflammatory drugs ................................. Antipyretic drugs ..................................................................... Peroxisome proliferators ......................................................... Antihypertensive drugs ........................................................... Anticonvulsants ........................................................................ Serotonin reuptake inhibitors ................................................ Beta blockers ............................................................................ Steroid contraceptives ............................................................. Antibiotics ................................................................................. Risk assessment ........................................................................ Conclusions ............................................................................... References ................................................................................

467 470 471 472 473 475 477 479 494 497 503 504

Endocrine-active ultraviolet filters and cosmetics .......................... M. Schlumpf and W. Lichtensteiger, GREEN Tox, Switzerland 19.1 Chemicals used as ultraviolet filters ...................................... 19.2 Endocrine activity and developmental toxicity of ultraviolet filters ....................................................................... 19.3 Exposure to ultraviolet filters and other cosmetic ingredients ................................................................................. 19.4 Considerations of human risk ................................................ 19.5 Acknowledgments .................................................................... 19.6 References ................................................................................

519

Mechanisms of action of particular endocrine-disrupting chemicals .............................................................................................. F. Pakdel, O. Kah and B. Jégou, Université de Rennes 1, France 20.1 Introduction .............................................................................. 20.2 Nuclear receptor family: estrogen receptors ........................ 20.3 Estrogenic/anti-estrogenic potency of endocrine-disrupting chemicals .............................................. 20.4 Androgenic/anti-androgenic potency of endocrine-disrupting chemicals .............................................. 20.5 Dioxin-like potency of endocrine-disrupting chemicals ..... 20.6 Conclusions and future trends ............................................... 20.7 Acknowledgements ................................................................. 20.8 References ................................................................................

519 520 527 533 535 535

541

541 543 547 550 553 557 559 559

Epilogue ............................................................................................... 568 I. Shaw, University of Canterbury, New Zealand

Index ............................................................................................................. 571

Contributor contact details

(* = main contact)

Editor, Chapter 1 and Epilogue Professor I. Shaw* Pro-Vice-Chancellor College of Science University of Canterbury Christchurch 8020 New Zealand

Chapter 2 Dr J.D. Meeker Department of Environmental Health Sciences University of Michigan School of Public Health Ann Arbor MI 48109 USA

E-mail: [email protected] E-mail: [email protected]

Chapter 1 Professor M.D. Mitchell and B. Balakrishnan Liggins Institute University of Auckland Private Bag 92019 Auckland New Zealand E-mail: [email protected]

xiv

Contributor contact details

Chapter 3

Chapter 5

Dr G. Toft* Aarhus University Hospital Department of Occupational Medicine Noerrebrogade 44, Build 2C DK-8000 Århus C Denmark

Dr L. Connolly The Institute of Agri-Food and Land Use Queen’s University Belfast David Keir Building Stranmillis Road Belfast BT9 5AG Northern Ireland UK

Professor J.P. Bonde Department of Occupational Medicine Copenhagen University Hospital Bispebjerg Bakke 23 DK-2400 Copenhagen NV Denmark E-mail: [email protected]

Chapter 4 Professor W. Wuttke* Professor H. Jarry and Dr. D. Seidlová-Wuttke Department of Endocrinology Georg-August-Universität Göttingen Robert-Koch-Str. 40 37075 Göttingen Germany E-mail: ufkendo@med. uni-goettingen.de

E-mail: [email protected]

Chapter 6 Dr M. Rose Food and Environment Research Agency Sand Hutton York YO41 1LZ UK E-mail: [email protected]

Chapter 7 Dr J.-P. Antignac*, Dr Frédérique Courant and Professor Dr B. Le Bizec Laboratoire d’Etude des Résidus et Contaminants dans les Aliments (LABERCA) USC 2013 INRA, Ecole Nationale Vétérinaire de Nantes (ENVN), BP 50707, Nantes cedex 3 France E-mail: [email protected]

Contributor contact details

xv

Chapter 8

Chapter 10

E. Eltzov Unit of Environmental Engineering Faculty of Engineering Science Ben-Gurion University of the Negev Beer-Sheva Israel

Dr T.F.H. Bovee and Dr L.A.P. Hoogenboom* RIKILT-Institute of Food Safety, Wageningen UR Department of Safety & Health PO Box 230 6700 AE Wageningen The Netherlands

Dr. A. Kushmaro and Dr R.S. Marks* Department of Biotechnology Engineering Faculty of Engineering Science and National Institute for Biotechnology in the Negev Ben-Gurion University of the Negev Beer-Sheva Israel

E-mail: [email protected]

E-mail: [email protected]

Chapter 9 Dr B.M. Thomson Institute of Environmental Science & Research Ltd PO Box 29 181 Christchurch New Zealand 8540 E-mail: [email protected]. nz

Dr B.M. Thomson Institute of Environmental Science & Research Ltd PO Box 29 181 Christchurch New Zealand 8540

Chapter 11 Professor L.R. Ferguson* and Dr M. Philpott Discipline of Nutrition Faculty of Medical and Health Services The University of Auckland and Nutrigenomics Private Bag 92019 Auckland New Zealand E-mail: [email protected]

Chapter 12 Dr A. Roncaglioni and Dr E. Benfenati* Laboratory of Environmental Chemistry and Toxicology Istituto di Ricerche Farmacologiche ‘Mario Negri’ Via La Masa 19 20156 Milan Italy E-mail: [email protected]

xvi

Contributor contact details

Chapter 13

Chapter 15

Dr B.G.J. Heinzow State Agency for Social Services Schleswig-Holstein Brunswikerstrasse 4 D-24105 Kiel Germany

Dr L.A.P. Hoogenboom RIKILT-Institute of Food Safety, Wageningen UR Department of Safety and Health PO Box 230 6700 AE Wageningen The Netherlands

E-mail: birger.heinzow@lasd. landsh.de and University of Notre Dame Sydney School of Medicine Australia

Chapter 14 A. Beronius, A. Hanberg, H. Håkansson* Karolinska Institutet Institute of Environmental Medicine PO Box 210 SE-171 77 Stockholm Sweden E-mail: [email protected], [email protected], [email protected]

E-mail: [email protected]

Chapter 16 Dr Julie E. Goodman, PhD, DABT* and Dr Lorenze R. Rhomberg, PhD Gradient Corporation 20 University Road Cambridge MA 02138 USA E-mail: jgoodman@gradientcorp. com; lrhomberg@gradientcorp. com

Chapter 17

C. Rudén Royal Institute of Technology Sweden

Professor S. Hendrich 220 MacKay Iowa State University Ames, IA 50011-1123 USA

E-mail: [email protected];

E-mail: [email protected]

J. Garai University of Pecs Hungary

Contributor contact details

xvii

Chapter 18

Chapter 19

A.H. Piersma,* M. Luijten Laboratory for Health Protection Research National Institute for Public Health and the Environment RIVM Antonie van Leeuwenhoeklaan 9 PO Box 1 3720 BA Bilthoven The Netherlands

PD Dr M. Schlumpf and Professor W. Lichtensteiger* GREEN Tox Winterthurerstrasse 190 CH-8057 Zurich Switzerland E-mail: Walter.Lichtensteiger@ access.uzh.ch [email protected]

E-mail: [email protected] Dr Altstein Institute of Plant Protection The Volcani Center, Agricultural Research Organization Bet Dagan 50250 ISRAEL Dr Popov, Dr Tomenko Wessex Institute of Technology Ashurst Lodge Ashurst Southampton SO40 7AA UK Dr Schlesinger, Dr Kagampang Analyst Research Laboratories Ltd Rabin Park, 12 Hanada St Rehovot 76703, Israel

Chapter 20 Dr F. Pakdel* and O. Kah UMR CNRS 6026 – Interactions Cellulaires et Moléculaires IFR140 Université de Rennes 1 Campus de Beaulieu 35042 Rennes cedex France E-mail: [email protected] B. Jégou INSERM U625 GERHM IFR140 Université de Rennes 1 35042 Rennes cedex France

Editor’s dedication In memory of James Boyes 1977–2001 A student whose commitment and enthusiasm will never be realised.

Introduction I. Shaw, University of Canterbury, New Zealand

There is increasing concern worldwide about food and the effects it has on consumers. This concern must be set in perspective because a good proportion of the world does not have enough food, and the rest of the world has too much, which means that they can be selective about what they eat. Such selectivity leads to a thirst for knowledge that provides information upon which to base choices. The popular media has grasped the issues and given a chance will escalate them both to inform consumers, sell their newspapers and magazines and encourage people to watch the television or listen to the radio. It is arguable that the Salmonella in eggs saga in the UK in the mid-1980s began the UK public’s elevated interest in food safety issues and that this was cemented in place by the BSE debacle. Whatever the reason there is a hunger for food issues that is often out of proportion to their associated health risks. This desire to understand food has driven regulators to review the myriad facets of food and set limits for contaminants, additives, etc., and to regulate bacterial and viral contamination to make food safer. In turn this has fired the interest of scientists to look more closely at food, at what it contains and the effects of its components (whether natural or added) on health and well-being. The importance of endocrine-disrupting chemicals (EDCs) in this respect is only just being realised, and is on the verge of being taken seriously by regulators worldwide. To understand the significance of EDCs in a food context we must turn the clock back to 1994 when Guillette’s group (Guillette et al., 1994) noted that alligators in Lake Apopka in Florida, USA, had smaller penises than their counterparts in nearby Lake Woodruff. The explanation for this was that the Lake Apopka alligators were being exposed to xenoestrogenic pollutants (e.g. dichlorodiphenyltrichloroethane, DDT) originating from

xx

Introduction

agricultural runoff. A year later Sumpter’s group (Sumpter & Jobling, 1995) at Brunel University in the UK noticed that male fish caged near to sewage outfalls synthesised the egg protein vitellogenin. They thought that this might be due to estrogens originating from contraceptive pills; this was part of the explanation, but as time went by it became clear that the situation was far more complex and that there is a large number of chemical pollutants and natural chemicals that mimic hormones (in particular estrogen) and so fit and activate their receptors. The pharmacologically active doses of EDCs were found to be very low because of their analogy to the hormones which are often active at nanomolar levels. This means that dietary intakes of EDCs could result in pharmacological effect in consumers (Shaw & McCully, 2002; Thomson et al., 2003) at extremely low doses. Moreover since many EDCs act via common receptors (e.g. xenoestrogens via the estrogen receptor, ER) the effects of different EDC contaminants of food are at least additive. Sometimes this is forgotten by regulators who understandably prefer to produce toxicological exposure limits for individual compounds rather than classes of compounds – especially when the class is based on occupancy and activation of a receptor rather than some obvious structural analogy (e.g. consider the xenoestrogens bisphenol A and genistein which have very different molecular structures) or common mechanism of action. As time passed it became clear that there were many EDCs (particularly xenoestrogens) of industrial origin. Plasticisers were, and still are, in the spotlight in this respect. The fact that many pre-packaged foods are plastic wrapped and that plastic films are extensively used in home and commercial kitchens means that humans are likely to be exposed to EDC’s daily in their food. Because many EDCs are lipid soluble there was concern that they would migrate from plastic wraps into the food. This has been verified experimentally and led to a number of food-contact plastics being reformulated to remove EDCs (e.g. phthalates) from their composition. Some pesticides used in food production (both in horticulture and as veterinary medicines) were shown to have low-level direct EDC activity (e.g. DDT) whereas others (e.g. cypermethrin) are metabolised in the environment to xenoestrogens (McCarthy et al., 2006). Many such pesticides, particularly those with a high LgPow (i.e. lipophilic) are found as residues in food and are accumulated by consumers, which might explain why studies have shown that the children of female farm workers have a greater incidence of cryptorchidism than children of male farm workers or controls (Weidner et al., 1998). For the above reasons much interest has been focused on synthetic (e.g. plasticisers, pesticides) EDCs, but it soon became clear that there are many natural EDCs in food and that many of them have much higher unit estrogen activity than their synthetic counterparts. For example DDT has an estrogen equivalence (EEq) of about 10−6 whereas the soy phytoestrogen genistein has an EEq of 10−5. Interest refocused on high phytoestrogencontaining foods such as soy, particularly soy-based infant formulae.

Introduction

xxi

To some extent the ‘risk jury’ is still out and conjecture remains about the impact of EDCs in food on health and well-being. The negative impacts, such as precocious puberty in girls, breast cancer, reduced sperm count and the like, tend to be in the forefront of discussions, but we should balance these with the health benefits of some high phytoestrogen foods, for example in postmenopausal osteoporosis, xenoestrogen intake might be beneficial. Certainly we are nowhere near suggesting toxicologically acceptable doses in a food context. Indeed some tolerable daily intakes (TDIs) have been set using non-hormonal end points which puts into question their validity when assessing risks of hormone analogue EDCs. Thinking is now moving fast as important new information comes out of a significant body of international research. I think it is safe to say that it is accepted that EDCs have an impact on health and that there is a need to keep a close eye on exposure. So much so that even without regulations some companies are voluntarily changing their procedures to reduce environmental contamination with EDCs or exposure via food. Not only are we concerned about the effects of EDCs on the consumer, but also on the unborn child – an indirect consumer. The latter is assuming greater significance as it becomes clearer that chemical exposure in utero can influence gene expression and so determine the child’s future health and well-being (see Chapter 1). I hope this volume brings together the thinking across a broad spectrum of EDC issues relating to food, sets them in context and gives a steer towards the relevance to health and wellbeing. It is impossible to cover everything, or to get all of the key scientists from around the world to write chapters that make the text complete, so unavoidably some relevant issues are unaddressed, and there is inevitably some duplication. I hope this does not distract from the intent that we address the subject of EDCs in food in its very broadest context. I have tried to bring the thinking together in my epilogue which is a summation of what the authors of the chapters are saying, but, importantly, adds thinking from key areas that authors did not focus on (e.g. the role of EDCs in non-genotoxic carcinogenesis). One thing is certain, we will hear a great deal more about EDCs in the future and it is likely that foodborne EDCs, either as natural food components, contaminants or additives will be the subject of research and heated debate for many years to come. I hope that this text will be a source of good information for that debate.

References guillette lj jr, gross ts, masson gr, matter jm, percival hj & woodward ar (1994) developmental abnormalities of the gonad and sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environ. Health Perspect. 102, 680–688.

xxii

Introduction

mccarthy ar, thomson bm, shaw ic & abell ad (2006) Estrogenicity of pyrethroid insecticide metabolites. J. Environ. Monit. 8, 197–202. shaw ic and mccully s (2002) A review of the potential impact of dietary endocrine disrupters on the consumer. Int. J. Food Sci. Technol. 37, 471–476. sumpter jp and jobling s (1995) Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ. Health Perspect. 103 [Suppl 7], 173–178. thomson bm, cressey pj and shaw ic (2003) Dietary exposure to xenoestrogens in New Zealand. J Environ. Monit. 5, 229–235. weidner is, moller h, jensen tk & skakkerbaek ne (1998) Cryptorchidism and hypospadias in sons of gardeners and farmers. Environ. Health. Perspectives 106, 793–796.

Acknowledgements

I would like to thank the authors of the chapters and Sarah Whitworth, Lynsey Gathercole and Laura Pugh from Woodhead Publishing for their incredible patience when I missed nearly every deadline – it has been a joy to work with you all. Professor Ian Shaw Editor

1 The effect of dietary endocrine disruptors on the developing fetus I. Shaw, University of Canterbury and University of Auckland, New Zealand, B. Balakrishnan and M. D. Mitchell, University of Auckland, New Zealand

Abstract: It is accepted that both natural and synthetic endocrine disruptors are present in the food we eat, and that they are likely to have a pharmacological impact on the consumer. The magnitude of this impact is a cause of great controversy at present. Much of the debate has focused on the direct impact of endocrine disruptors on the consumer; this chapter speculates on possible impacts on the developing fetus that might lead to effects that manifest themselves much later in life. Key words: in utero effects, placental barrier, placental metabolism, developing fetus – effects on human development, protection of fetus.

1.1

Introduction

The first evidence of adverse effects of phytoestrogens in animal reproduction came over 60 years ago from Western Australia where male rams, feeding on clover pasture, became feminised and unable to breed. The cause of this phenomenon was eventually identified as the high levels of the isofavones genistein and diadzein present in clover (Bennets et al., 1946). Fifty years later Guillette et al. (1994) showed in their seminal work that alligators in Lake Apopka in the Florida Everglades have smaller penises than those from nearby Lake Woodruff, not because of direct effects of xenoestrogenic contaminants (e.g. dichlorodiphenyltrichloroethane, DDT) in Lake Apopka upon the alligators, but rather due to an effect of the contaminants on the developing egg. The resultant alligators displayed reduced testosterone levels, leading in turn to under-developed penises. This illustrates well the potential effects of in utero exposure to environmental endocrine-disrupting chemicals (EDCs) on adult reproductive function.

4

Endocrine-disrupting chemicals in food

A quite different form of long-term impact on reproduction following exposure to EDCs is the potential for estrogen mimics in food plants to signal to their consumers that in the following year there will be a bumper crop, and therefore that offspring are likely to survive because of the plentiful supply of food. There are many examples of this recently discovered phenomenon. For example the kakapo (Strigops habroptilus), a rare New Zealand flightless parrot, ovulates when rimu (Dacrydium cupressinum; a large rainforest podocarp tree) is likely to produce a good crop of seeds in the following year (Sutherland, 2002). Through this ingenious adaptation, the success of the offspring is supported by a plentiful supply of food in the season when the egg hatches. It is possible that the signal is the plant hormone gibberellic acid, which is produced at higher levels in the year preceding extensive fruiting. Gibberellic acid is a 17β-estradiol mimic (Fig. 1.1), and when consumed by the would-be mother kakapo might stimulate ovulation and egg-laying. There are other similar examples: for instance the grey squirrel (Sciurus carolinensis) eats the tips of pine trees. If the tips are high in gibberellic acid it is thought that this signals a good crop of pine cones the following season, and also stimulates the squirrels to reproduce (Richard Pharis, University of Calgary, USA, personal communication). The offspring are assured a good supply of seeds from the cones and so are more likely to thrive. This all makes good evolutionary sense and is a good example of exposure to endocrine disruptors affecting a future generation. Sumpter and Jobling (1995) commented that aquatic organisms live in ‘a sea of estrogens’. They were referring to the multifarious estrogen mimics that are present in sewage effluent and so find their way into rivers and streams; human consumers are as a result also exposed via drinking water that originates from these aquatic environments. In addition, there are estrogenic pesticide residues in our food (e.g. DDT), and myriad natural endocrine disruptors in the plants and animals we eat (Thomson et al., 2003). Sometimes we forget that the meat from female food animals and their milk all contain relatively high concentrations of 17β-estradiol (and other hormones). Therefore estrogenic chemicals are all around; we are exposed to them continuously.

OH O H CH2

OC HO CH3

Fig. 1.1

H

OH

O

Structure of gibberellic acid.

The effect of dietary endocrine disruptors on the developing fetus

5

The first potential effects of EDCs in humans were suggested by Carlsen et al. (1992), who reported a decline in semen quality during the preceding 50 years. The experimental design of this study has, however, been criticised (Pflieger-Bruss et al., 2004). Interestingly, semen quantity in bulls and other animals has remained unchanged over the corresponding 50-year period. It is now largely accepted that human sperm quality has declined over the past five or six decades, but that the magnitude of the decline is less than that suggested by Carlsen et al. in their 1992 paper (Pflieger-Bruss et al., 2004). This very brief review of the key work that has helped our understanding of environmental estrogens represents only the historical basis of what has become an important field of study in its own right. Since Sumpter’s work that led the thinking in the 1990s there have been thousands of publications on the individual compounds and their possible effects. It is not possible to review them fully here, but many are discussed elsewhere in this volume; xenoestrogens in the human food chain have been reviewed by Shaw et al. (2004). The mechanisms of effect of EDCs, particularly of the xenoestrogens, are less well understood. It is becoming increasingly evident that these can in some instances act across generations; parental exposure results in physiological changes in successive generations of their offspring. This chapter will focus on the possible mechanisms of resultant effects of maternal exposure to EDCs, and on the means by which the placenta protects the fetus.

1.2

The effects of endocrine-disrupting chemicals on human development

A number of compounds have been identified with estrogen modulating effects, with bisphenol A (BPA), p-nonylphenol, genistein and dioxin being a few examples of EDCs that are present ubiquitously in the environment and in food (Gierthy, 2002); studies on EDCs have now been extended to include compounds that modulate the actions of androgens, progestins, thyroid, hypothalamic and pituitary hormones (Gierthy, 2002). Over a span of decades research on wildlife epidemiology and animal studies have shown that fetal or prenatal exposure to EDCs might have adverse effects on the development and/or function of reproductive, nervous, cardiovascular and respiratory systems (Newbold et al., 2007). Long-term follow-up studies on human exposure to diethylstilbestrol (DES; Fig. 1.2) and subsequent observation of developmental abnormalities in their progenies strengthen this hypothesis (Giusti et al., 1995). Human development is a highly integrated process which spans gametogenesis through fertilisation, embryogenesis, maturation and senescence. Each of these processes is under strict hormonal control (Birnbaum, 1994, 1995). Autocrine, paracrine (such as growth factors) and endocrine (such as

6

Endocrine-disrupting chemicals in food HO CH3

H3C OH

Fig. 1.2

Structure of diethylstilbestrol (DES).

steroid) signals coordinate the direction of differentiation of tissues during critical periods in development (Bigsby et al., 1999). This involves endocrine regulation in all aspects of development; hence the adverse effects on development of EDCs, which can mimic, block or modulate natural hormones and other chemical messengers (Birnbaum, 1994, 1995). Gene activity during development can also be modified by environmental signals which act on developmental processes via mechanisms that include epigenetic regulation (Bigsby et al., 1999). Fetal development is associated with high rates of cellular proliferation and extensive differentiation balanced by apoptosis (cell death). For this reason it is prone to many environmental insults leading to mutagenic and epigenetic events that can predispose the developing child to cancer and related diseases (Birnbaum et al., 2003). The exquisite sensitivity of the developing fetus to estradiol was discussed by Howard Bern in a chapter entitled ‘Fragile fetus’ (Bern, 1992). Although questions remain unanswered around the impacts of exposure to EDCs at different times during gestation, it has become clear that individuals are more susceptible to the effects of EDC exposure during fetal and neonatal development than during adult life (Sweeney, 2002). This is because many biochemical systems are not fully developed in the fetus and so are susceptible to disruptions that might have long term effects. Examples include: • fetal endocrine feedback mechanisms (Sweeney, 2002); • the fetal immune system (Sweeney, 2002); • detoxifying enzyme activity and fetal hepatic metabolism (Miller, 1983). In addition EDCs tend to bind with low affinity to sex hormone binding globulins (SHBG) (Sweeny, 2002), hence enhancing their bioavailability. Also, changes in the physiology of the female during pregnancy might actually increase fetal exposure to EDCs. For example EDCs are hydrophobic and are likely to be sequestered in body fats; pregnancy puts extra energy demands on the system resulting in mobilisation of storage lipids, with the consequent release of ‘stored’ EDCs, thus making them available for placental transfer (Shaw, 2000).

The effect of dietary endocrine disruptors on the developing fetus

7

The important end effect of this potential biochemical interference is a change in the ratio of estrogen to testosterone (either directly, through an increase in levels of estrogen mimics, or indirectly through metabolic changes resulting in changed 17β-estradiol levels). This ratio is crucial for normal sexual development. Altered ratios in the fetus will irreversibly affect development (in the same way that the changed ratio of testosterone to estrogens in the Lake Apopka alligators resulted in reduced penis length (Guillette et al., 1994). Studies in children of DES/EDC-exposed mothers who had modified sexual development illustrate this well (Paris et al., 2006; Gupta, 2000a,b).

1.3

The metabolism of endocrine disruptors

When any chemical is ingested it is exposed to cellular detoxification systems. The body has sophisticated mechanisms for reducing the toxicity and increasing the water solubility of foreign compounds, in order to expedite their excretion from the body and minimise their impact upon the well-being of their consumer. In addition, mechanisms exist to sequester hormones as inactive analogues (sulphates or glucuronides) so that they do not affect non-target cells. Metabolic processes can either reduce (detoxify) or increase (activate) pharmacological activity; see Timbrell (1991) and Shaw and Chadwick (1998) for detailed accounts of cellular and environmental foreign compound metabolism. The basic mechanism of detoxification comprises two phases: in Phase I, the chemical is oxidised (e.g. hydroxylation) by the cytochrome P450 enzyme complex found in most tissues, but predominantly in the liver. This oxidation both (usually) reduces the toxicity of the foreign chemical, and also adds a chemical group to facilitate subsequent conjugation. In Phase II, the parent compound (if it has the appropriate chemical groups) or the product of Phase I metabolism is conjugated with a highly water-soluble moiety (e.g. glucuronic acid), to facilitate its excretion in the water-based excretory systems (predominantly bile and urine). Endocrine disruptors are subjected to the two-phase metabolic system as are other compounds foreign to the body. They are absorbed from the gut and transported via the hepatic portal system to the liver, where the relatively high activity of the cytochrome P450 system might catalyse oxididation of the molecule. Many endocrine disruptors already have the appropriate chemical groups (e.g. —OH) to go straight into Phase II; whether Phase I metabolism is required or not, the resulting molecule is chemically prepared for excretion and so its impact upon the body is minimised. If the molecule is large (molecular weight >500 daltons in humans, 250 daltons in rats) it will be secreted in bile into the intestine and be excreted from the body in the faeces. If it is a smaller molecule it will be released

8

Endocrine-disrupting chemicals in food

into the circulatory system and eventually excreted via the renal system in urine. The biliary excretion route is not necessarily final because the gut microflora excretes deconjugating enzymes (e.g. β-glucuronidase) that can release the parent compound or Phase I metabolite for reabsorption. This is the process of biliary recirculation. It means that the body is exposed to the parent compound or its Phase I metabolite repeatedly. On each cycle a proportion of the compound is not re-absorbed and so the circulating levels gradually diminish.

1.3.1 Metabolic activation As mentioned above the ‘detoxification’ system does not always reduce pharmacological activity. It can also increase the impact of a foreign compound on the body. For example the potent carcinogen benzo[A]pyrene acts via its cytochrome P450 epoxy reactive metabolite, which interacts with DNA to result in cellular transformation. Some EDC precursors might act in a similar way. For example the estrogenic non-ionic detergent nonyl phenol is likely to act via its alkyl oxidation product (hydroxynonyl phenol), which can be produced by bacteria in the environment (e.g. in the sediments of waterways) or in the body by (for example) hepatic cytochrome P450 (Zalko et al., 2003). It has recently been shown (McCarthy et al., 2006) that the pyrethroid insecticides – particularly Cypermethrin – are either metabolised or degrade naturally to generate estrogenic metabolites which represent another source of environmental xenoestrogens.

1.3.2

Protection of the developing fetus from maternal dietary contaminants Even before the advent of synthetic chemicals the fetus was potentially exposed to many harmful natural food components (e.g. isothiocyanates in brassicas). For this reason an armoury of protection developed to minimise the fetus’s exposure to toxic insults from its mother’s dietary components. This is loosely referred to as the placental fetal barrier and represents an apparently impenetrable filter that only lets through nutrients, etc. that will benefit the developing embryo and fetus. Phase I and Phase II metabolism constitute important elements of the placental fetal barrier. It is important that the developing offspring is not exposed to high levels of maternal estrogens, because this would result in feminisation of the fetus. For this reason the placenta has a very active conjugation system that, akin to Phase II metabolism, conjugates maternal estrogens with glucuronic acid or sulphuric acid, making them both inactive and highly water-soluble. This process means that the estrogens stay on the maternal side of the placenta and so have no pharmacological effect on the developing embryo or fetus.

The effect of dietary endocrine disruptors on the developing fetus

9

The placental barrier is driven by the metabolic processes that minimise passage of potentially harmful chemicals. In order to be effective the system has to ‘recognise’ potentially pharmacological undesirables. This is biochemically simple for the small array of maternally endogenous pharmacologically active chemicals such as 17β-estradiol, but it might not be effective for the enormous array of dietary contaminants that could find their way into the maternal circulatory system and so present a risk to the embryo or fetus. The cytochrome P450 system is an effective mechanism to deal with a very broad array of structurally unrelated chemicals because of its multivariant active site. Molecules that already have appropriate groups for conjugation (e.g. —OH), or those that might be oxidised by cytochrome P450 in order to add the site for conjugation, are not likely to cross the placental barrier. However some molecules are not oxidised by the cytochrome P450 system and do not have appropriate chemical groups to facilitate conjugation. Such molecules might cross the placental barrier and thus present a pharmacological risk to the embryo or fetus. Some of these molecules might be EDCs. An example is the potent xenoestrogen BPA – a chemical used in the protective lacquers used in the canning of some foods (e.g. canned tomatoes). This will be discussed in more detail later in this chapter.

1.3.3 Metabolism in the placenta The relatively rapid placental transfer of EDCs has been reported in humans (Todaka et al., 2005; Engel et al., 2006; Balakrishnan et al., 2008). On the other hand placental accumulation of hydrophobic molecules such as 2,3,7,8-tetrachloro dibenz dioxin (TCDD) occurs (Chao et al., 2007); this might be due to binding to placental arylhydrocarbon (AHC) receptors (Hakkola et al., 1997). Compounds such as TCDD are therefore slowly transferred across the placental barrier to the embryo/fetus (Takahashi and Oishi, 2000). Compounds that are more hydrophobic are transferred more quickly, attaining fetal concentrations approaching those in the maternal circulatory system (Takahashi and Oishi, 2000). Compounds that bind avidly to placental proteins are less likely to cross over to the embryo/fetus. Dioxins fall into this category and thus while concentrations on the maternal side of the placental barrier might be high, levels remain low or undetectable in cord blood (Chao et al., 2007). These principles are very likely to apply to the vast array of molecularly diverse EDCs.

1.3.4 Bisphenol A BPA is a good example of a non-steroidal xenoestrogen that has a significant structure–activity relationship (SAR) with 17β-estradiol. In this case the SAR arises not because BPA has a steroid nucleus as part of its molecular structure, but rather because its functional groups (—OH) are spatially

10

Endocrine-disrupting chemicals in food

separated in such a way that they interact with the estrogen receptor (ER) in a manner analogous to the hydroxyls of 17β-estradiol itself (see Chapter 16). BPA is metabolised in mammals to a series of phenols, which are in turn conjugated with glucuronic acid and excreted in urine as highly watersoluble glucuronides (Knaak and Sullivan, 1966; Sheftel, 1995). In the rat 28% of an orally administered BPA dose was excreted in the urine and 56% in the faeces (Groshart et al., 2001 and references therein). The latter could be due either to poor absorption from the gut, or to biliary excretion of metabolites. The placenta has the metabolic apparatus to carry out these reactions; specific cytochrome P450 activities such as P450arom (Li et al., 2004) have been detected in placental tissue, and P450-catalysed metabolism of xenobiotics in placental tissues have been demonstrated (e.g. carbamezapine is metabolised in isolated placental tissue to 10-hydroxycarbamezapine, a classic P450-catalysed metabolic transformation; Myllynen et al., 1998). BPA is reasonably hydrophilic (log octanol/water partition coefficient (lgPow) = 3.32) and rapidly crosses the placenta (Takahashi and Oishi, 2000). Having entered fetal tissue it remains there longer than in maternal tissues (Domoradzki et al., 2003). Interestingly the pharmacokinetics of BPA at higher doses are non-linear, most likely due to metabolic saturation (Domoradzki et al., 2003). Indeed fetal microsomes prepared from rats showed sluggish glucuronyltransferase activity (Matsumoto et al., 2003), which points to the potential for metabolic activation. Administration of radioactively labelled BPA to pregnant mice showed that about 4% of the dose crossed the placenta to the offspring (Zalko et al., 2003; Rubin et al., 2006). This clearly shows that passage across the placental barrier occurs, but does not identify the molecular form of the radioactivity (i.e. it might be a non-estrogenic metabolite that is actually transferred). So the question remains, does the placenta protect the developing embryo or fetus from endocrine-disrupting xenobiotics? It has the enzymes necessary to do this, but is this armoury used to protect against xenobiotics or just the mother’s circulating estrogens? Perhaps the best way to address this is to look at fetal responses to endocrine disruptors administered to pregnant females. A study has been performed in rats (Yoshino et al., 2004), in which BPA was administered prenatally to females (at the time of pairing for 18 days) and effects on the immune system of the offspring observed (BPA is not only a xenoestrogen, but is also immunoactive). The 8-week-old progeny were challenged with an antigen (hen egg lysosyme; HEL) and the immune response measured. It was found that maternal exposure to BPA resulted in up-regulation of the offspring’s immune response. This result suggests that BPA crosses the placenta to the fetus in a pharmacologically active form. This in turn suggests that the metabolic apparatus in the placenta does not effectively protect the fetus from exposure to this particular EDC. The experiment did not measure endocrine activity of BPA on the offspring, but if the compound was present it is likely to have had an estrogenic effect.

The effect of dietary endocrine disruptors on the developing fetus

1.4

11

Endocrine disruptors in general

It is not possible to generalise about the potential impact of this structurally diverse group of chemicals. However it is likely that EDCs that are structural analogues of steroids will be ‘dealt with’ by the placenta’s mechanism for protecting the embryo and fetus from maternal estrogens, and that the non-steroidal EDCs, particularly the xenoestrogens (e.g. BPA), might cross the placental barrier to the embryo or fetus and so exert their pharmacological effect. Recent work postulated a hierarchy of estrogenic impact of dietary xenoestrogens on human consumers (Table 1.1; from Thomson et al., 2003). This was based on both estimated dietary exposure (from dietary surveys) and estimated total circulating estrogenic activity (based on absorption efficiency and relative estrogenicity (estrogen equivalents – EQ) to 17β-estradiol). All of the dietary xenoestrogens in Table 1.1 are non-steroids. Some have a greater structural similarity to steroids than others. For example, BPA has a molecular structure very different from the steroid nucleus, whereas genistein has a flavonoid nucleus which is more akin to the steroids (Fig. 1.3). It is much more likely that a flavonoid will fit the active site of a steroid metabolising enzyme than, for example, BPA. Based on this theory, and the evidence of pharmacological activity in the offspring of exposed rat dams, the xenoestrogen to which we have the greatest exposure (i.e. BPA) is likely to escape the fetal/placental protection mechanism and so impact upon the hormone balance of the embryo/fetus. The same is likely to apply to the other non-steroid, non-flavonoid xenoestrogens listed in Table 1.1.

Table 1.1 Proposed hierarchy of human exposure to xenoestrogens (Thomson et al., 2003) Xenoestrogen BPA Genistein Alkyl phenols Kaempferol Phloretin Diadzein Quercetin Enterolactone Endosulfan Polychlorinated biphenyls Butylated hydroxyanisol DDT + metabolites Enterodiol

Estimated blood EQ (mg/L) 1.6 × 1.5 × 8.4 × 2.8 × 2.8 × 2.2 × 1.9 × 7.8 × 4.5 × 2.7 × 1.4 × 3.2 × 4.0 ×

10−3 10−3 10−4 10−4 10−4 10−4 10−5 10−6 10−6 10−6 10−6 10−7 10−8

12

Endocrine-disrupting chemicals in food OH

17 β-Estradiol HO OH HO

O

Genistein HO

OH Bisphenol A

O

OH

Fig. 1.3 Structures of 17β-estradiol, and of the non-steroidal xenoestrogens BPA and genistein.

1.4.1 Pharmacologically relevant doses of xenoestrogens Conventional toxicology relies upon measures of exposure that equate to pharmacological or physiological adverse effects, such as the no observable effect level (NOEL) the highest dose at which there is no measurable effect and no observable adverse effect level (NOAEL) – the highest dose at which there is no measurable adverse effect. However, hormones and hormone mimics often do not comply with these simplistic dose-related rules, and EDCs are no exception to this. They are effectively hormones and can have disproportionate dose/effect relationships just as the true hormones can. For example in rats the NOAEL for genistein is 50 mg/kg body weight/day, but the NOEL is only 5 mg/kg body weight/day. The latter is due to a hormonal end point, the former to a conventional toxicological end point (Michael McClain et al., 2006). Recent studies have confirmed that EDCs can exert significant actions at levels well below those prescribed through traditional toxicological investigation. The results of such a study, in which the effects of estradiol were monitored over a wide dose range in the human mammary cancer cell line MCF-7, are represented in Fig. 1.4. Estradiol was cytotoxic at high doses (above 1 μm) but was found to stimulate proliferation at low doses (100 pm–1 μm), an action that is not reflected in the available NOEL data. Low-dose effects vom Saal et al. (1998) has advocated the significance of low-dose effects of EDCs on development, having observed that only a small increase in the

DNA per well (% of control)

The effect of dietary endocrine disruptors on the developing fetus

13

350 300 Low-dose range

250 200

High-dose range

150 Control

100 50

0 10−15 10−14 10−13 10−12 10−11 10−10 10−9 10−8 10−7 Concentration (M) 1

10 100 fg/mL (ppq)

1

10 100 pg/mL (ppt)

1

10−6

10 100 ng/mL (ppb)

10−5

1

10−4 10−3

10 100 μg/mL (ppm)

Concentration (mass/mL)

Fig. 1.4 Bimodal dose–response effect observed in estrogen receptor positive MCF-7 cells treated with estradiol (modified from Welshons et al., 2003, with permission from Environmental Health Perspectives).

circulating estradiol concentration can affect prostate weight in experimental animals. Welshons et al. (2003) suggested that the low-dose effects of EDCs are mediated through the ER while a different mechanism (e.g. conventional cellular toxicity) is involved in high-dose effects. Prenatal and postnatal exposure to low but environmentally relevant concentrations of estrogenic chemicals have resulted in a range of developmental effects such as sex reversal in turtles (Sheehan et al., 1999), reduced testes sizes and sertoli cell numbers in rats (Atanassova et al., 1999), enhanced induction of cytochrome P-450 1A activity (DeLong and Rice, 1997), increased anogenital distances (Gupta, 2000a), increased preputial gland size (vom Saal et al., 1998), and increased reproductive tract organ sizes (vom Saal et al., 1998; Timms et al., 2005) in mice. vom Saal et al. (1997) have also shown that it is low doses of BPA or estrogens which cause the increased prostatic bud number, cell proliferation and adult prostate size observed in mice. Clearly the mechanism of action of EDCs is not simple toxicity, but is likely to be via hormonal routes that might be initiated at very low exposure concentrations. We must look carefully at toxicological NOEL doses from which acceptable daily intakes (ADI) might be calculated if they are not based on hormonal end points; ADIs calculated from non-hormonal end points for suspected EDCs will be very misleading and might give an impression that the EDC dose necessary to have an affect in humans is much higher than it really is. Indeed, low doses of estrogens have been observed to cause an inverted manifestation of most of the effects observed at high dose

14

Endocrine-disrupting chemicals in food

ranges (Putz et al., 2001). Low-dose exposure to estrogens may alter hepatic steroid enzyme expression in rats, which Putz hypothesised to be responsible for the observed organ-specific responses (Putz et al., 2001). Prins et al. (2007) hypothesised that environmentally relevant doses of BPA increase the sensitivity of the prostate gland to carcinogenesis following adult insults such as elevated circulating estrogens. Clearly exposure to environmental EDCs occurs, but whether the dose is pharmacologically relevant is the key question. It must also be remembered that most estrogenic EDCs act via occupancy of the ER, and so additive effects with other estrogens are very likely. To be relevant, in vitro and animal studies with EDCs must be carried out at doses that are likely to result from environmental exposure (including exposure from dietary sources); thus for example the environmentally relevant dose of BPA translated into in vitro studies has been estimated at 0.23 ng/ml in tissue culture media (Wozniak et al., 2005). In a recent study it was suggested that an estrogenic EDC would be biologically active in the fetus if the resulting total estrogen equivalent activity was only 0.1 pg/ml (0.37 pm estradiol concentration equivalent) above the endogenous level (vom Saal et al., 1997). This is interesting because it suggests that very low doses may have a pharmacological effect on the fetus. We should exercise caution when extrapolating results from animal studies to humans because biologically active hormone levels differ greatly among species. For example the biologically active dose of ethinylestradiol in rats is 100–400 times that for women (Putz et al., 2001). There have been a vast number of studies on a broad array of EDCs in many species of animal. Most have shown biological effects at environmentally relevant doses (vom Saal et al., 1997; Sheehan et al., 1999; Murray et al., 2007; and many other studies). Dioxins (e.g. TCDD) are EDCs with very long half-lives in humans of some 4–7 years (Michalek et al., 2002). TCDD is detectable at 10 parts per trillion in background plasma and adipose tissue of general populations (Sara Mariasole et al., 2006); at such a low level it is non-toxic in a conventional sense, but may adversely affect fetal growth and development (Sara Mariasole et al., 2006). High-dose effects As discussed above, the adverse effects of ECDs differ very much depending on the dose. Recent research has shown that high doses of exogenous natural or synthetic hormones, for example estrogens administered during early postnatal life can advance puberty in females, but delay its onset in males (Putz et al., 2001). It has been hypothesised that high intra-uterine exposure to estrogen might predispose the offspring to mammary cancer (Ekbom et al., 1992). In human epidemiological studies twin pregnancy has been used as an indicator of high estrogen exposure and pre-eclampsia as an indicator of low exposure (Ekbom et al., 1992; Murray et al., 2007). Interestingly there is a high correlation between breast cancer and twin

The effect of dietary endocrine disruptors on the developing fetus

15

pregnancy, and a low correlation between breast cancer and pre-eclampsia (Ekbom et al., 1992). It is clearly important to distinguish between high dose (i.e. conventional) toxicological effects and low dose (i.e. hormonal) effects. For example, high doses of BPA disrupt placental function and therefore lead to reproductive disorders (i.e. teratogenicity) by adversely affecting the developing fetus (Lee et al., 2005). This is not an in utero EDC effect even though it results in an effect on the offspring. The adverse effects of low and high doses of various EDCs on various developmental stages are described in Table 1.2.

Table 1.2

Effects of exposure to EDCs during critical periods of development

Chemical

Class

Deleterious effects of exposure

1

Diethylstilbestrol

Estrogen agonist

2

BPA

Xenoestrogen

1. Incidence of clear cell cervical adenocarcinoma in human (in utero exposure in humans) 2. Incidence of preterm birth and 2nd trimester fetal loss in humans (Kaufman et al., 2000) 3. Testicular dysgenesis syndrome (humans)? 1. Reproductive abnormalities in experimental animals (vom Saal et al., 1997) 2. Ambiguous genitalia in humans? (Paris et al., 2006) 3. Recurrent miscarriages in humans (Sugiura-Ogasawara et al., 2005) 4. Alters sexual differentiation of brain and behaviour 5. Prostatic interepithelial neoplasia in rodents (Prins et al., 2007) 6. Stimulation of mammary growth and ductal mammary carcinoma in rodents (Murray et al., 2007; Durando et al., 2007) 7. Alters immune functions in rodents 8. Meiotic aneuploidy in female mice (Hunt et al., 2003) 9. Implantation failure in mice (Takai et al., 2000; Berger et al., 2007) 10. Low birth weight, hypogonadotropism, dampened LH surge in female lambs (Savabieasfahani et al., 2006) 11. Modulates drug efflux mechanism of placenta (Jin and Audus, 2005)

16

Endocrine-disrupting chemicals in food

Table 1.2

Continued

Chemical

Class

3

4-NP

Xenoestrogen

4

Dioxin (TCDD)

Anti-estrogen

5

DDT

Xenoestrogen

6

Vinclozolin

Anti-androgen

1.5

Deleterious effects of exposure 1. Reproductive abnormalities in rodents (Lee, 1998) 2. Delayed testes descent in rodents (Lee, 1998) 1. Premature reproductive senescence in female rats (Shi et al., 2007) 2. Miscarriage in monkeys (Wilbur, 1984) 3. Incidence of preterm birth in humans (Revich et al., 2001) 1. Micropenis in alligators (Guillette et al., 1994) 2. Preterm birth in humans (Longnecker et al., 2001) 1. Developmental abnormalities in various systems in rodents (Anway et al., 2006) 2. Transgenerational effects (Anway et al., 2005, 2006)

The role of the endocrine system in fetal development

As discussed above, the endocrine system plays a major role in different stages of pregnancy and throughout fetal development, growth and parturition. There is a complex interplay between maternal, fetal and placental hormones during fetal development. For this reason if endogenous hormones are perturbed by EDCs it is likely that fetal development will be affected.

1.6

Effects of endocrine-disrupting chemicals on sexual differentiation and congenital malformations of the developing fetus

1.6.1 Sexual differentiation in the male fetus Sexual differentiation is hormonally controlled; the presence of androgen is required for male development and its absence for female development. Changes in hormone levels or the ratio of hormones can disrupt sexual differentiation – clearly EDCs that mimic important hormones (e.g. 17βestradiol) will perturb the levels that are required for normal development. This might explain the increased prevalence of hypospadias, cryptorchidism and micropenis (Toppari and Skakkebaek, 1998).

The effect of dietary endocrine disruptors on the developing fetus

17

The formation of the penis, scrotum and accessory sex glands are under the influence of steroid hormones secreted during the hormonal phase of testicular development (Basrur, 2006). Testosterone and the androgen dihydrotestosterone are the two major hormones involved in the above process (Basrur, 2006). In the male conceptus the androgen receptor is expressed as early as 12–20 weeks of gestation (Basrur, 2006). Most human data on perturbation of male sexual differentiation come from follow-up studies of the sons born to mothers dosed with the potent synthetic estrogen DES (Sultan et al., 2001) as part of fertility enhancement treatment; these children have an increased incidence of poor testicular development and cryptorchidism (Basrur, 2006). It has been reported that 20.8% of the males exposed to DES in utero had epididymal cysts, 4.4% had hypospadias, 11.4% presented with cryptorchidism and hypoplastic testes and 1.5% had micropenis (Sultan et al., 2001). A similar range of changes were also observed in DES-exposed mice (Sultan et al., 2001). The first trimester of gestation is the most susceptible period in terms of fetal sex differentiation (Sultan et al., 2001). Boys exposed to DES in utero at this time had an increased incidence of several structural and functional genital abnormalities, such as epididymal cysts, meatal stenosis, hypospadias and testicular abnormalities including cryptorchidism, hypoplastic testis and capsular degeneration (Brevini et al., 2005). In a neonatal screening programme of ambiguous genitalia, Paris et al. (2006) identified three male newborns with male pseudohermaphroditism whose mothers had been exposed to EDCs during gestation. Based on these findings it was hypothesised that the pre- and neo-natal exposure of these children to EDCs had perturbed sexual differentiation leading to genital ambiguity (Paris et al., 2006).

1.6.2 Sexual differentiation in the female fetus Several genes responsible for morphogenesis of the fetal ovary have been identified (Basrur, 2006). Gonadogenesis in the female fetus is under the control of maternal female hormones (Basrur, 2006) which act by upregulation of these genes (e.g. DAX1, WNT4) and down-regulation of the genes involved in male sexual differentiation (e.g. SOX9, WT1 and SF1). CYP19 is a gene involved in the conversion of androgens to estrogens; it is expressed in somatic cells of the fetal ovary during early gestation, thus controlling the synthesis of estrogens during this important time of sex determination in the developing child. Clearly, hormone ratios are crucial to the normal course of events; the introduction of EDCs at this stage would be likely to interfere with sexual differentiation. Little work has been reported in this complex area, but the aneuploidogenic potential of BPA has been demonstrated (Hunt et al., 2003; Dash et al., 2006), and Shi et al. (2007) observed premature reproductive senescence following chronic in utero exposure to TCDD. Whether these are true EDC effects is uncertain.

18

Endocrine-disrupting chemicals in food

1.6.3 Congenital malformations The higher incidence of male reproductive abnormalities such as hypospadias, cryptorchidism and testicular germ cell cancer coupled with a fall in sperm count have been observed globally from the mid-twentieth century onwards (West et al., 2005). This condition in humans, labelled testicular dysgenesis syndrome, is believed to be due to exposure to EDCs. This is now largely accepted to be related to the widespread dissemination of hormonally active chemicals and of phytoestrogens in soya-based foods, although it remains difficult to show a positive correlation between exposure and the defects (West et al., 2005). The issue of dose and effect is brought into question when one considers the huge doses of phytoestrogens that people from Asia receive in their diets. However, this might be explained by their development of resistance to phytoestrogens owing to their consumption over many generations. This resistance could take the form of increased metabolism with consequent deactivation of phytoestrogens. This argument can be taken further by considering that western populations might manifest a genetic predisposition to the effects of phytoestrogens (West et al., 2005). The ability of babies to absorb genistein and diadzein can result in their estrogenic content being increased by more than a thousand-fold compared with endogenous levels (Setchell et al., 1987; Irvine et al., 1998; West et al., 2005). West et al. (2005) observed that dietary genistein during the intrauterine or neonatal period interrupts the differentiation of round to elongated spermatids, thus compromising the offspring’s sperm production. In addition Fas receptors (which when activated initiate a cascade, leading to apoptosis of the cell) in rodent germ cells and Fas ligands in sertoli cells were found to be up-regulated when exposed to phthalates, possibly leading to apoptosis and reduction in sertoli and germ cell numbers (Atanassova et al., 1999; Kumi-Diaka et al., 1999). Nair and Shaha (2003) demonstrated a similar mechanism for DES. Clearly in utero exposure to xenoestrogens is able to significantly perturb spermatogenesis. Long-term monitoring of lactose-intolerant children receiving soya-based infant formula will provide invaluable data to test the above hypothesis in years to come (Irvine et al., 1998).

1.6.4 Cryptorchidism Less than 5% of male neonates suffer from cryptorchidism (Toppari et al., 1996). Testicular descent is controlled by hormones; the first phase is migration to the groin, controlled by insulin-like (Insl-3) factor. Insl-3 is produced by Leidig cells and its synthesis is inhibited by estrogens. Between 5 and 10% of cryptorchidism cases are due to problems in the first phase of descent. The second phase involves migration of the testes to the scrotal sac and is thought to be influenced by androgens; failure at this stage results in the testes descending only to the inguinal region (Werler, 2007). The involve-

The effect of dietary endocrine disruptors on the developing fetus

19

ment of xenoestrogens in cryptorchidism has been known since work by Gill and Stillman (e.g. Gill et al., 1979; Stillman, 1982) on DES in the 1970s and 1980s. The involvement of xenoestrogens in reproductive developmental abnormalities was reviewed more generally by Toppari and Skakkebaek (1998), who demonstrated a correlation between exposure and effect. Studies in rats and mice have supported these findings in humans (Ivell and Hartung, 2003). Perhaps the best evidence of EDCs resulting in cryptorchidism is a 40year cohort study carried out in East and West Berlin maternity hospitals, which showed a strong correlation between cryptorchidism and DDT exposure (Ivell and Hartung et al., 2003). This is a particularly good study because of the differential exposure to DDT between the former East and West Germany. Possibly because of differences between the regulatory systems of the two countries, East Germans have very much higher body burdens of DDT and its metabolites than do West Germans. This is illustrated well by human milk levels of DDT in women from the two countries: East Berlin = 2.3 mg/kg; West Berlin = 0.8 mg/kg (Burke et al., 2003). This study was followed by a 25-year cohort study in Berlin following the banning of DDT which showed a reduction in the incidence of cryptorchidism. This study clinched the epidemiological evidence for DDT causing cryptorchidism. Experimental studies of EDC exposure coupled with epidemiological observation of increased prevalence of cryptorchidism and hypospadias in EDC-contaminated areas strongly support the hypothesis that EDCs cause cryptorchidism (Ivell and Hartung, 2003).

1.6.5 Hypospadia Hypospadia is a developmental anomaly defined as the displacement of the urethral meatus from the tip of the glans penis to the ventral side of the phallus, scrotum or perineum; treatment usually involves surgical reconstruction (Pierik et al., 2002). It occurs in 0.02–0.4% of live births (Pierik et al., 2002), and is thought to be related to the effects of accumulated antiandrogens on the developing fetus at 6–14 weeks of gestation (Basrur, 2006). Exposure to EDCs also leads to hypospadias (Basrur, 2006), which clearly is not unexpected since xenoestrogens would be expected to have the same pharmacological effect as anti-androgens or reduced testosterone levels. Since the differentiation of the genital tubercle into male external genitalia has been noted to require the signalling factor protein FGF 10 in mice, it is possible that estrogen mimics or anti-androgenic compounds in the environment contributes to this malformation by disrupting FGF 10 signalling (Basrur, 2006). The dominance in expression patterns of estrogen receptors over androgen receptors was reported in the penile tissues of neonates with hypospadias (Celayir et al., 2007). This may be one of the possible mechanisms for endocrine disruptive action of xenoestrogens in

20

Endocrine-disrupting chemicals in food

the development of hypospadias. A higher incidence of hypospadias has been observed both in male children of in vitro fertilization (IVF) mothers (Ericson and Kallen, 2001) possibly due to a hormone imbalance caused by the procedure, and in regions where maternal exposure to EDCs is higher than normal (Sharpe and Skakkebaek, 1993).

1.7

Other effects of endocrine disruptors on the developing fetus

EDCs may cause a vast array of developmental defects in the fetus; such effects might vary from miscarriage (i.e. fetal loss) to developmental abnormalities in different systems in the fetus depending upon the time of exposure.

1.7.1 Fetal loss In a large follow-up study on the reproductive performance of daughters prenatally exposed to DES, it was shown that there was a high incidence of preterm birth and second trimester fetal loss (Kaufman et al., 2000). In addition a positive correlation has been shown between recurrent spontaneous abortion and high serum BPA levels (Sugiura-Ogasawara et al., 2005). Hunt et al. (2003) demonstrated an induction of meiotic aneuploidy in mice following environmentally relevant doses of BPA. It has also been reported that BPA inhibits implantation at high doses (Berger et al., 2007), and at environmentally relevant concentrations it has been shown to affect the in vitro development of mice embryos in a dose-dependent manner (Takai et al., 2000). BPA also affects postnatal weaning weight; since it is known that weaning weight is linked to a propensity to adult disease, BPA exposure in utero might therefore influence disease susceptibility in later life (Durando et al., 2007); indeed there is mounting evidence that gene regulation in utero affects genetically determined disease in later life. TCDD has also been linked to miscarriage in monkeys exposed to a single dose during pregnancy (Wilbur, 1984).

1.7.2 Preterm birth There is some evidence that EDCs might result in preterm births, but it is scant and often contradictory. A large cohort study showed an association between serum DDE, but not DDT, and preterm birth (Longnecker et al., 2001); on the other hand Farhang et al. (2005) did not find any relationship between first trimester DDE exposure and preterm delivery. Saxena et al. (1981) reported an association between placental concentration of organochloride pesticides and preterm birth in India. An Italian study found no association between TCDD exposure and preterm delivery (Eskenazi et al.,

The effect of dietary endocrine disruptors on the developing fetus

21

2003), while a significant correlation was observed in a highly contaminated area in Russia (Revich et al., 2001). Links between phthalates and premature delivery have been suggested (Latini et al., 2003, 2005). Phthalates and their metabolites have been shown to increase the expression of peroxysome proliferator activated receptor-γ (PPAR-γ receptor; a signalling molecule associated with inflammatory conditions) in a mouse placental trophoblast cell line (Xu et al., 2005), and it is hypothesised that they can induce an inflammatory response via the PPAR-γ pathway and thus decrease the period of gestation (Xu et al., 2005).

1.7.3 Cardiovascular development There have been many studies that extol the cardiovascular-protecting benefits of soy products in adults. These findings cannot simply be extrapolated to the unborn child, because little is known about the passage of soy xenoestrogens across the placental barrier to the fetus. However it has been reported that feeding genistein and daidzein to pregnant rats resulted in offspring with shorter cardiac myocytes, thus conferring some degree of cardio-protection (Messina et al., 2006).

1.7.4 Neuroendocrine development There has been an increase in the prevalence of neurodevelopmental and other developmental defects in humans since the introduction of synthetic endocrine disruptors such as BPA and DDT (Colborn, 2004); the question is of course whether a causal relationship exists. There is some evidence of a connection: for example Kabuto et al. (2004) showed that exposure to BPA resulted in underdeveloped brain, kidney and testes in fetuses and neonates exposed in utero or via milk. Similarly it has been suggested that an increased incidence of learning disabilities, attention deficit hyperactivity disorder (ADHD), childhood cancers and juvenile diabetes during the mid-1990s may be due to in utero exposure to EDCs (Colborn, 2004). While there is no definitive experimental evidence for this hypothesis, it is known that neuroendocrine developmental defects occur in animal models after exposure to EDCs (Petersen et al., 2006). Indeed AhR antagonists such as dioxins accumulate in the developing fetus, and might have irreversible effects on neuronal development if their accumulation spans the neuronal development stage (Petersen et al., 2006). Recent studies have shown a positive correlation between prenatal exposure to EDCs and both postnatal neurological underdevelopment and poor maternal nursing in rodents (Palanza et al., 2002) and humans (Jacobson and Jacobson, 2002). Interestingly exposure of dams to BPA during gestation resulted in increased levels of dopamine and serotonin metabolites in the brains of both the dams and their female offspring (Honma et al., 2006);

22

Endocrine-disrupting chemicals in food

this presents a plausible biochemical explanantion of the developmental changes observed, and is likely to be due to the xenoestrogenic effects of BPA.

1.7.5 Intergenerational effects The daughters of DES-treated women have an increased incidence of adenocarcinoma (Hatch et al., 2001) and if animal studies can be extrapolated may be transmitted to their granddaughters. Clearly there is a genetic or epigenetic change initiated by DES that is passed down the generations. This is thought to involve epigenetic regulation (Ruden et al., 2005). This finding in humans has been supported by animal experiments (Birnbaum and Fenton, 2003). In contrast, no intergenerational effects on reproduction or development were found in multigenerational studies with low doses of BPA (Ema et al., 2001; Tyl et al., 2002). Recently Anway et al. (2006) reported transgenerational effects following maternal exposure to the anti-androgenic EDC Vinclozolin (a fungicide used in horticulture) in a four-generation study. In this study the effects included prostate, kidney, testicular and immune system functional disorders, as well as an increased incidence of mammary carcinoma (Anway et al., 2006). These effects were proposed to be mediated through epigenetic mechanisms which also resulted in modified sexual differentiation (Anway et al., 2005, 2006).

1.7.6 Neonatal morbidity There is conflicting evidence concerning the effects of 4-nonylphend (4-NP) on the developing child. A study in neonatal rats at a dose of 20.8 mg/kg/day showed decreased reproductive organ weight and delayed testes descent (Lee, 1998), whereas a similar study (Odum and Ashby, 2000) did not confirm the findings. De Jager et al. (1999) showed reduced testicular mass, sperm count and adversely affected seminiferous tubules in 4-NP-treated rats (de Jager et al., 1999; Pflieger-Bruss et al., 2004). Postnatal exposure of rats to BPA on days 21–35 resulted in suppressed serum Luteinising hormone (LH), testosterone and estrogen levels; the latter may be due to inhibition of aromatase activity in Leydig cells (PfliegerBruss et al., 2004). In addition, low doses of BPA caused reduced sperm count in both rats and mice (Pflieger-Bruss et al., 2004). Marmoset monkeys (Callithrix jacchus) exposed to a single BPA dose displayed impaired spermatogenesis and reduced Leydig cell 3β-hydroxysteroid hydroxygenase activity (Rune et al., 1991). Interestingly Swan et al. (2007) reported a correlation between the amount of beef consumed by pregnant women and oligospermia in their sons. It was speculated that this was related to xenoestrogens in beef.

The effect of dietary endocrine disruptors on the developing fetus

1.8

23

Reproductive dysfunction in females

An accidental observation by Hunt et al. (2003) showed a correlation between BPA exposure and aneuploidy in developing mouse oocytes. In addition it has been reported that women with recurrent spontaneous abortion had significantly higher levels of serum BPA than healthy women from the same city in Japan (Sugiura-Ogasawara et al., 2005). These interesting data do not allow the establishment of a cause and effect relationship; however, there are physiological and biochemical reasons that might explain the observed effects: BPA increases progesterone receptor expression in the hypothalamus, which in turn alters hypothalamic mechanisms and affects the onset of estrus and the receptivity of the uterus (Funabashi et al., 2003). The finding that prenatal exposure to BPA in sheep caused a reduction in their lambs’ birth weights, hypergonadotropism, delayed breeding season and dampened LH surge in the female lambs (Savabieasfahani et al., 2006), and that exposure to 4-NP increased uterine weight and accelerated the vaginal opening in pre-pubertal rats (Bandiera, 2006) adds further weight to the strengthening argument that exposure to EDCs upsets reproductive function.

1.9

Endocrine disruptors and the placenta

The placenta protects the developing embryo and fetus from environmental chemical insults and prevents maternal sex hormones influencing cellular sex hormone-mediated biochemistry. The placenta’s biochemical processes recognise endogenous hormones (e.g. 17β-estradiol) and conjugate them both to prevent their passage across the placental barrier and to reduce their biological activity. Other multifunctional mechanisms (e.g. cytochrome P450 mixed function oxidase complexes) also exist to detoxify exogenous environmental chemical insults. Whether these systems recognise potential embryo or fetal effectors that might harm the developing child is the key issue. It appears that the placenta is geared to recognise endogenous but perhaps not exogenous chemicals. This is logical since the fetus is developing in a female biochemical environment that it must be protected against from a hormonal point of view. In terms of potential impact on the developing child xenoestrogens or estrogen mimics are the issue, because many of them have molecular structures very different from 17β-estradiol even though they still bind to and activate the estrogen receptor. Xenoestrogens occur in the diet and therefore preventing their access to the fetus or detoxifying them is important in order to minimise their impact on the developing child. It is possible that the placenta recognises xenoestrogens with a molecular structure similar to 17β-estradiol (e.g. genistein from soya) and bars their

24

Endocrine-disrupting chemicals in food

passage across the placental barrier, while not recognising those diverse nonestrogen-like estrogen mimics (e.g. BPA). The latter are likely to interfere with fetal development and therefore we should be aware of the implications of exposure during pregnancy. This is informed speculation at the moment and is the subject of a significant research project in our laboratory.

1.9.1 Placental transporters and endocrine-disrupting chemicals Many studies have reported that estrogens and progesterone can influence the function and expression of multi-drug resistance (MDR) proteins in the placenta. These proteins are important because they are responsible for the active transport of xenobiotics from the fetal to the maternal side of the placenta, and are thus likely to be involved in maintaining the placental barrier and protecting the developing fetus. Steroid hormones modulate the activity of these transporters: for example progesterone modulates the MDR protein P-glycoprotein (P-gp) in the human placenta in a dose-dependent manner (Jin and Audus, 2005). On the other hand 17β-estradiol reverses breast cancer resistance protein (BCRP)mediated drug resistance (Imai et al., 2002), and anti-estrogens (e.g. tamoxifen) reverse P-gp-mediated drug resistance (Jin and Audus, 2005). Previously it was reported that BPA is a substrate for P-gp in the intestine (Yoshikawa et al., 2002). Recently BPA has been shown to regulate the drug efflux activity of P-gp in a human trophoblast cell line (BeWo, a choriocarcinoma cell line; Jin and Audus, 2005). It is hypothesised that a direct interaction between BPA and P-gp stimulates the drug efflux mechanism in this trophoblast model. Further it has been shown that BPA inhibits P-gp-mediated drug efflux mechanisms and that 17β-estradiol additively enhances this effect, suggesting an interaction of the two molecules at the same P-gp binding site (Jin and Audus, 2005).

1.9.2 Effects on placental hormones TCDD decreases the production of progesterone in placental cells (Augustowska et al., 2003), and TCDD administered to mice increases levels of lipid metabolites in their amniotic fluid (Hassoun et al., 1995). Both observations might be explained by reduced levels of mitochondrial enzymes (CYP450sec) that convert cholesterol to pregnenolone or pregnenolone to progesterone. Cholesterol is a precursor of the steroid hormones and is synthesized from lipids, so reduced cholesterol synthesis would increase lipid metabolites while reducing steroid hormone synthesis. Phytoestrogens inhibit progesterone production from choriocarcinoma cell lines (Plessow et al., 2003), but the effects on human chorionic gonadotrophin (hCG) production varied depending on the phytoestrogen and the dose used (Matscheski et al., 2006). Similarly dose-dependent inhibition of hCG was reported in human trophoblasts treated with phytoestrogens

The effect of dietary endocrine disruptors on the developing fetus

25

(Jeschke et al., 2005); therefore it appears that the inhibitory effect of phytoestrogens is mediated by inhibiting hCG production. In addition phytoestrogens induce expression of ERα and progesterone receptor (PR) in cultured cells (JEG-3), which might represent another anti-proliferative mechanism – if receptor expression is increased a greater ligand concentration would be necessary to induce cell proliferation (Matscheski et al., 2006). 4-NP is a more avid inducer of hCG than 17β-estradiol in first trimester placental explants (Bechi et al., 2006). The enhanced activity of 4-NP may be due to the formation of more stable and potent metabolites (Bechi et al., 2006); indeed, it is likely that hydroxynonylphenol is a cellular metabolite of 4-NP and that it would be a better ligand for the ER than would 4-NP per se. Clearly there is evidence that EDCs can affect placental hormone secretions and that such changes could have a profound effect on the developing child. The interrelationships between EDC exposure and hormone effects are complex and occur via a multitude of mechanisms. There is much work to be done in this area before we begin to understand the complexity of effects.

1.9.3

Effects of xenobiotic and steroid metabolizing enzymes on the placenta Aromatase (CYP-19) catalyses the aromatisation of androgens into estrogens. Studies in isolated cells (JEG-3) have shown that placental aromatase activity is inhibited by TCDD in a dose-dependent manner (Drenth et al., 1998). BPA also inhibits aromatase activity by interacting directly with the aromatase enzymatic complex (Nativelle-Serpentini et al., 2003). These findings are very important because they point to a non-receptorbased mechanism for the effects of EDCs on cellular sex hormone responses. This might be an important mechanism of action of some EDCs.

1.10

Studies of placental transfer of xenoestrogens in humans

We studied the placental transfer of genistein in ex vivo perfused human placentae (Shin, 2004). From an initial concentration of genistein in the maternal perfusate of 200 μg/L, we observed a decline in the concentration of genistein in the maternal circuit over a period of 3 hours, with a concomitant increase in the concentration of the same in the fetal side. The fetal concentration of genistein was approximately 12.5% within 3 hours of initiating perfusion. We used 17β-estradiol as a control at a concentration of 1 μg/L; levels of unconjugated 17β-estradiol in the fetal compartment were minimal throughout the course of the experiment. In an ongoing study, 10–20% of initial levels of BPA in the maternal compartment (10 μg/L)

26

Endocrine-disrupting chemicals in food

could be detected in the fetal compartment after 3 hours of perfusion. Our studies thus indicate that EDCs can transfer across the human placenta (Balakrishnan et al., 2008).

1.11

Conclusions

Our understanding of the implications of a mother’s exposure to EDCs on the future life of her child is in its extreme infancy. Most scientists working in the field agree that there are effects and that there is the potential for such effects to be profound. It is clear from this chapter that there is a very great deal of evidence for effects of EDCs on all facets of pregnancy and the neonatal child. It is equally clear that there are myriad mechanisms for these effects. The interrelationships between the EDCs and their mechanisms of action will determine the impact on the child – these complex interactions are only now being probed, and will not be fully understood for many years. Suffice to say that as we gain understanding of the effects of EDCs it will become increasingly possible through early intervention to minimise negative impacts on growth and development, either by modifying the mother’s exposure or by influencing industry to cut potent EDCs from their manufacturing processes. Canada is the first country to tackle this issue by initiating a ban on the use of BPA plastics in the manufacture of baby bottles in order to reduce children’s exposure to this potent EDC. This is a wise and forward-looking move. In closing this chapter it is interesting to reflect on the considerable increase in the use of IVF techniques to achieve pregnancy, particularly in the developed world. IVF embryos spend their first 4–8 cell stages in plastic bottles floating in a buffer very likely to contain leached EDCs – this the subject of a joint research project between the authors of this chapter at the Universities of Auckland and Canterbury, New Zealand.

1.12

Future trends

As discussed above this research is in its infancy. Indeed this chapter is among the first syntheses of information on the subject. There is a very long way to go and a very great deal of research to do before we understand the mechanisms of action of EDCs on the unborn child and the risks they pose to children’s health. Perhaps the most important research area is investigating cause–effect relationships in humans – this requires large epidemiological studies to investigate exposure and effects in a highly controlled way in order to reduce other variables. If we establish genuine cause–effect relationships this will give a significant impetus to work on individual ECD exposures and effects in in vitro, in vivo and ultimately in human studies.

The effect of dietary endocrine disruptors on the developing fetus

27

Finally work aimed at understanding the mechanisms of effect of EDCs on the unborn child, in particular the role of the placenta in protecting the fetus, is of crucial importance. There is little point in studying the theoretical mechanisms of EDC interactions with embryonic or fetal cells (i.e. in vivo or ex vivo experiments) if the placenta adequately protects against their impact. We have much to learn at both a fundamental biochemical level and at the higher physiological level before any of us can even pretend to understand. There are exciting times ahead!

1.13

Sources of other information and advice

1. National Toxicology Programme (NTP) – Centre for Evaluation of Risks to human reproduction (CERHR): Draft from NTP panel on low-dose effects of BPA is available at this site. http://cerhr.niehs.nih.gov/chemicals/bisphenol/bisphenol.html 2. World Health Organization: http://www.who.int/ipcs/publications/endocrine_disruptors/endocrine_ disruptors/en/ 3. Global Endocrine Disruptor Research Inventory (GEDRI) http://oaspub.epa.gov/gedri/pack_edri.All_Page 4. Our Stolen Future http://www.ourstolenfuture.org/ 5. Dioxins: introduction to health issues in New Zealand http://www.moh.govt.nz/dioxins 6. WWF toxic chemicals http://worldwildlife.org/toxics/ 7. Dioxin in New Zealand food supply http://www.nzfsa.govt.nz/consumers/food-safety-topics/chemicals-infood/dioxins/index.htm 8. EPA–NCEA–Exposure and Human Health Reassessment of 2,3,7,8tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds National Academy Sciences (NAS) Review Draft http://www.epa.gov/ncea/pdfs/dioxin/nas-review/ 9. Centre for health effects of environmental contamination (Endocrine Disruptors and Pharmaceutically Active Chemicals in Drinking Water Workshop) http://www.cheec.uiowa.edu/conferences/edc_2000/index.html 10. Dioxin home page (a site with updated information on all aspects of dioxin) http://www.ejnet.org/dioxin/ 11. Human toxome project: mapping the pollution in people (environmental working group). Biomonitoring methods to test blood, urine, breast milk and other tissues for industrial chemicals that enter the human body http://www.ewg.org/sites/humantoxome/

28

Endocrine-disrupting chemicals in food

12. About soy products and toxicity http://www.soyonlineservice.co.nz/index.htm 13. Chemical exposure (a collection of articles from Environment Health Perspectives) http://www.ehponline.org/topic/chemexp.html

1.14

References

anway m d, cupp a s, uzumcu m, skinner m k (2005), ‘Epigenetic transgenerational actions of endocrine disruptors and male fertility’, Science, 308 (5727), 1466–1469. anway m d, leathers c, skinner m k (2006), ‘Endocrine disruptor vinclozolin induced epigenetic transgenerational adult-onset disease’, Endocrinology, 147 (12), 5515–5523. atanassova n, mckinnell c, walker m, turner k j, fisher j s, morley m, millar m r, groome n p, sharpe r m (1999), ‘Permanent effects of neonatal estrogen exposure in rats on reproductive hormone levels, Sertoli cell number, and the efficiency of spermatogenesis in adulthood’, Endocrinology, 140 (11), 5364–5373. augustowska k, gregoraszczuk e l, milewicz t, krzysiek j, grochowalski a, chrzaszcz r (2003), ‘Effects of dioxin (2,3,7,8-TCDD) and PCDDs/PCDFs congeners mixture on steroidogenesis in human placenta tissue culture’, Endocr Regul, 37 (1), 11–19. balakrishnan b, henare k, thorstensen e, mitchell m d (2008), ‘Transfer of BPA across the human placenta’, Reprod Sci, 15 (2), 141A-142A. bandiera s m (2006), ‘Reproductive and endocrine effects of p-nonylphenol and mthoxychlor: a review’, Immunol Endo Metabol Agents Med Chem, 6 (1), 15–26. basrur p k (2006), ‘Disrupted sex differentiation and feminization of man and domestic animals’, Environ Res, 100 (1), 18–38. bechi n, ietta f, romagnoli r, focardi s, corsi i, buffi c, paulesu l (2006), ‘Estrogen-like response to p-nonylphenol in human first trimester placenta and BeWo choriocarcinoma cells’, Toxicol Sci, 93 (1), 75–81. bennets h w, underwood e j, shier f l (1946), ‘A specific breeding problem of sheep on subterranean clover pastures in Western Australia’, Aus Vet J, 22, 2–13. berger r g, hancock t, decatanzaro d (2007), ‘Influence of oral and subcutaneous bisphenol-A on intrauterine implantation of fertilized ova in inseminated female mice’, Reprod Toxicol, 23 (2), 138–144. bern h (1992), ‘The fragile fetus’ in Chemically-induced Alterations in Sexual and Functional Development: The Wildlife/Human Connection, Colborn T and Clement C, Princeton Scientific Publishing Co., Princeton, NJ, 9–15. bigsby r, chapin r e, daston g p, davis b j, gorski j, gray l e, howdeshell k l, zoeller r t, vom saal f s (1999), ‘Evaluating the effects of endocrine disruptors on endocrine function during development’, Environ Health Perspect, 107 Suppl 4, 613–618. birnbaum l s (1994), ‘Endocrine effects of prenatal exposure to PCBs, dioxins, and other xenobiotics: implications for policy and future research’, Environ Health Perspect, 102 (8), 676–679. birnbaum l s (1995), ‘Developmental effects of dioxins and related endocrine disrupting chemicals’, Toxicol Lett, 82–83, 743–750. birnbaum l s, fenton s e (2003), ‘Cancer and developmental exposure to endocrine disruptors’, Environ Health Perspect, 111 (4), 389–394.

The effect of dietary endocrine disruptors on the developing fetus

29

brevini t a, zanetto s b, cillo f (2005), ‘Effects of endocrine disruptors on developmental and reproductive functions’, Curr Drug Targets Immune Endocr Metabol Disord, 5 (1), 1–10. burke e r, holden a j, shaw i c, suharyanto f x, sihombing g (2003), ‘Organochlorine pesticide residues in human milk from primiparous women in Indonesia’, Bull Environ Contam Toxicol, 71 (1), 148–155. carlsen e, giwercman a, keiding n, skakkebaek n e (1992), ‘Evidence for decreasing quality of semen during past 50 years’, BMJ, 305 (6854), 609–613. celayir s, elicevik m, tireli g, dervisoglu s, sander s (2007), ‘Expression of estrogen and androgen receptors in children with hypospadias: preliminary report’, Arch Androl, 53 (2), 83–85. chao h r, wang s l, lin l y, lee w j, papke o (2007), ‘Placental transfer of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in Taiwanese mothers in relation to menstrual cycle characteristics’, Food Chem Toxicol, 45 (2), 259–265. colborn t (2004), ‘Neurodevelopment and endocrine disruption’, Environ Health Perspect, 112 (9), 944–949. dash c, marcus m, terry p d (2006), ‘BPA: do recent studies of health effects among humans inform the long-standing debate?’ Mutat Res, 613 (2–3), 68–75. de jager c, bornman m s, van der horst g (1999), ‘The effect of p-nonylphenol, an environmental toxicant with oestrogenic properties, on fertility potential in adult male rats’, Andrologia, 31 (2), 99–106. delong g t, rice c d (1997), ‘Tributyltin potentiates 3,3′,4,4′,5-pentachlorobiphenylinduced cytochrome P-4501A-related activity’, J Toxicol Environ Health, 51 (2), 131–148. domoradzki j y, pottenger l h, thornton c m, hansen s c, card t l, markham d a, dryzga m d, shiotsuka r n, waechter j m, jr. (2003), ‘Metabolism and pharmacokinetics of BPA (BPA) and the embryo-fetal distribution of BPA and BPAmonoglucuronide in CD Sprague-Dawley rats at three gestational stages’, Toxicol Sci, 76 (1), 21–34. drenth h j, bouwman c a, seinen w, van den berg m (1998), ‘Effects of some persistent halogenated environmental contaminants on aromatase (CYP19) activity in the human choriocarcinoma cell line JEG-3’, Toxicol Appl Pharmacol, 148 (1), 50–55. durando m, kass l, piva j, sonnenschein c, soto a m, luque e h, munoz-de-toro m (2007), ‘Prenatal BPA exposure induces preneoplastic lesions in the mammary gland in Wistar rats’, Environ Health Perspect, 115 (1), 80–86. ekbom a, trichopoulos d, adami h o, hsieh c c, lan s j (1992), ‘Evidence of prenatal influences on breast cancer risk’, Lancet, 340 (8826), 1015–1018. ema m, fujii s, furukawa m, kiguchi m, ikka t, harazono a (2001), ‘Rat twogeneration reproductive toxicity study of BPA’, Reprod Toxicol, 15 (5), 505–523. engel s m, levy b, liu z, kaplan d, wolff m s (2006), ‘Xenobiotic phenols in early pregnancy amniotic fluid’, Reprod Toxicol, 21 (1), 110–112. ericson a, kallen b (2001), ‘Congenital malformations in infants born after IVF: a population-based study’, Hum Reprod, 16 (3), 504–509. eskenazi b, mocarelli p, warner m, chee w y, gerthoux p m, samuels s, needham l l, patterson d g, jr (2003), ‘Maternal serum dioxin levels and birth outcomes in women of Seveso, Italy’, Environ Health Perspect, 111 (7), 947–953. farhang l, weintraub j m, petreas m, eskenazi b, bhatia r (2005), ‘Association of DDT and DDE with birth weight and length of gestation in the Child Health and Development Studies, 1959–1967’, Am J Epidemiol, 162 (8), 717–725. funabashi t, sano a, mitsushima d, kimura f (2003), ‘BPA increases progesterone receptor immunoreactivity in the hypothalamus in a dose-dependent manner and affects sexual behaviour in adult ovariectomized rats’, J Neuroendocrinol, 15 (2), 134–140.

30

Endocrine-disrupting chemicals in food

gierthy j f (2002), ‘Testing for endocrine disruption: how much is enough?’ Toxicol Sci, 68 (1), 1–3. gill w b, schumacher g f, bibbo m, straus f h 2nd, schoenberg h w (1979), ‘Association of diethylstilbestrol exposure in utero with cryptorchidism, testicular hypoplasia and semen abnormalities’, J Urology, 122 (1), 36–39. giusti r m, iwamoto k, hatch e e (1995), ‘Diethylstilbestrol revisited: a review of the long-term health effects’, Ann Intern Med, 122 (10), 778–788. groshart c p, okkerman p c, pijnenburg a m c m (2001), Chemical study on Bisphenol-A, RIKZ Report 2001.027, Nat. Inst. Coast. and Marine Management (RIKZ), The Hague, The Netherlands. guillette l j, jr, gross t s, masson g r, matter j m, percival h f, woodward a r (1994), ‘Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida’, Environ Health Perspect, 102 (8), 680–688. gupta c (2000a), ‘Reproductive malformation of the male offspring following maternal exposure to estrogenic chemicals’, Proc Soc Exp Biol Med, 224 (2), 61–68. gupta c (2000b), ‘The role of estrogen receptor, androgen receptor and growth factors in diethylstilbestrol-induced programming of prostate differentiation’, Urol Res, 28 (4), 223–229. hakkola j, pasanen m, pelkonen o, hukkanen j, evisalmi s, anttila s, rane a, mantyla m, purkunen r, saarikoski s, tooming m, raunio h (1997), ‘Expression of CYP1B1 in human adult and fetal tissues and differential inducibility of CYP1B1 and CYP1A1 by Ah receptor ligands in human placenta and cultured cells’, Carcinogenesis, 18 (2), 391–397. hassoun e a, bagchi d, stohs s j (1995), ‘Evidence of 2,3,7,8-tetrachlorodibenzo-pdioxin (TCDD)-induced tissue damage in fetal and placental tissues and changes in amniotic fluid lipid metabolites of pregnant CF1 mice’, Toxicol Lett, 76 (3), 245–250. hatch e e, herbst a l, hoover r n, noller k l, adam e, kaufman r h, palmer j r, titus-ernstoff l, hyer m, hartge p, robboy s j (2001), ‘Incidence of squamous neoplasia of the cervix and vagina in women exposed prenatally to diethylstilbestrol (United States)’, Cancer Causes Control, 12 (9), 837–845. honma t, miyagawa m, suda m, wang r s, kobayashi k, sekiguchi s (2006), ‘Effects of perinatal exposure to BPA on brain neurotransmitters in female rat offspring’, Ind Health, 44 (3), 510–524. hunt p a, koehler k e, susiarjo m, hodges c a, ilagan a, voigt r c, thomas s, thomas b f, hassold t j (2003), ‘BPA exposure causes meiotic aneuploidy in the female mouse’, Curr Biol, 13 (7), 546–553. imai y, tsukahara s, ishikawa e, tsuruo t, sugimoto y (2002), ‘Estrone and 17betaestradiol reverse breast cancer resistance protein-mediated multidrug resistance’, Jpn J Cancer Res, 93 (3), 231–235. irvine c h, shand n, fitzpatrick m g, alexander s l (1998), ‘Daily intake and urinary excretion of genistein and daidzein by infants fed soy- or dairy-based infant formulas’, Am J Clin Nutr, 68 (6 Suppl), 1462S-1465S. ivell r, hartung s (2003), ‘The molecular basis of cryptorchidism’, Mol Hum Reprod, 9 (4), 175–181. jacobson j l, jacobson s w (2002), ‘Association of prenatal exposure to an environmental contaminant with intellectual function in childhood’, J Toxicol Clin Toxicol, 40 (4), 467–475. jeschke u, briese v, richter d u, bruer g, plessow d, waldschlager j, mylonas i, friese k (2005), ‘Effects of phytoestrogens genistein and daidzein on production of human chorionic gonadotropin in term trophoblast cells in vitro’, Gynecol Endocrinol, 21 (3), 180–184.

The effect of dietary endocrine disruptors on the developing fetus

31

jin h, audus k l (2005), ‘Effect of BPA on drug efflux in BeWo, a human trophoblast-like cell line’, Placenta, 26 Suppl A, S96-S103. kabuto h, amakawa m, shishibori t (2004), ‘Exposure to BPA during embryonic/ fetal life and infancy increases oxidative injury and causes underdevelopment of the brain and testis in mice’, Life Sci, 74 (24), 2931–2940. kaufman r h, adam e, hatch e e, noller k, herbst a l, palmer j r, hoover r n (2000), ‘Continued follow-up of pregnancy outcomes in diethylstilbestrol-exposed offspring’, Obstet Gynecol, 96 (4), 483–489. knaak j b, sullivan l j (1966), ‘Metabolism of BPA in the rat’, Toxicol Appl Pharmacol, 8 (2), 175–184. kumi-diaka j, nguyen v, butler a (1999), ‘Cytotoxic potential of the phytochemical genistein isoflavone (4′,5′,7-trihydroxyisoflavone) and certain environmental chemical compounds on testicular cells’, Biol Cell, 91 (7), 515–523. latini g, de felice c, presta g, del vecchio a, paris i, ruggieri f, mazzeo p (2003), ‘In utero exposure to di-(2-ethylhexyl)phthalate and duration of human pregnancy’, Environ Health Perspect, 111 (14), 1783–1785. latini g, massaro m, de felice c (2005), ‘Prenatal exposure to phthalates and intrauterine inflammation: a unifying hypothesis’, Toxicol Sci, 85 (1), 743. lee c k, kim s h, moon d h, kim j h, son b c, kim d h, lee c h, kim h d, kim j w, kim j e, lee c u (2005), ‘Effects of BPA on the placental function and reproduction in rats’, J Prev Med Pub Health, 38 (3), 330–336. lee p c (1998), ‘Disruption of male reproductive tract development by administration of the xenoestrogen, nonylphenol, to male newborn rats’, Endocrine, 9 (1), 105–111. li x, strauss l, makela s, streng t, huhtaniemi i, santti r, poutanen m (2004), ‘Multiple structural and functional abnormalities in the p450 aromatase expressing transgenic male mice are ameliorated by a p450 aromatase inhibitor’, Am J Pathol, 164 (3), 1039–1048. longnecker m p, klebanoff m a, zhou h, brock j w (2001), ‘Association between maternal serum concentration of the DDT metabolite DDE and preterm and small-for-gestational-age babies at birth’, Lancet, 358 (9276), 110–114. matscheski a, richter d u, hartmann a m, effmert u, jeschke u, kupka m s, abarzua s, briese v, ruth w, kragl u, piechulla b (2006), ‘Effects of phytoestrogen extracts isolated from rye, green and yellow pea seeds on hormone production and proliferation of trophoblast tumor cells Jeg3’, Horm Res, 65 (6), 276–288. matsumoto a, kunugita n, kitagawa k, isse t, oyama t, foureman g l, morita m, kawamoto t (2003), ‘BPA levels in human urine’, Environ Health Perspect, 111 (1), 101–104. mccarthy a r, thomson b m, shaw i c, abell a d (2006), ‘Estrogenicity of pyrethroid insecticide metabolites’, J Environ Monit, 8 (1), 197–202. messina m, mccaskill-stevens w, lampe j w (2006), ‘Addressing the soy and breast cancer relationship: review, commentary, and workshop proceedings’, J Natl Cancer Inst, 98 (18), 1275–1284. michalek j e, pirkle j l, needham l l, patterson d g, jr., caudill s p, tripathi r c, mocarelli p (2002), ‘Pharmacokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso adults and veterans of operation Ranch Hand’, J Expo Anal Environ Epidemiol, 12 (1), 44–53. michael mcclain r, wolz e, davidovich a, pfannkuch f, edwards j a, bausch j (2006), ‘Acute, subchronic and chronic safety studies with genistein in rats’, Food Chem Toxicol, 44 (1), 56–80. miller r k (1983), ‘Perinatal toxicology: its recognition and fundamentals’, Am J Ind Med, 4 (1–2), 205–244. murray t j, maffini m v, ucci a a, sonnenschein c, soto a m (2007), ‘Induction of mammary gland ductal hyperplasias and carcinoma in situ following fetal BPA exposure’, Reprod Toxicol, 23 (3), 383–390.

32

Endocrine-disrupting chemicals in food

myllynen p, penimaki p, raunio h, vahangas k (1998), ‘Microsomal metabolism of carbamazepine and oxcarbazepine in liver and placenta’, Human Exp Toxicol, 17, 668–678. nair r, shaha c (2003), ‘Diethylstilbestrol induces rat spermatogenic cell apoptosis in vivo through increased expression of spermatogenic cell Fas/FasL system’, J Biol Chem, 278 (8), 6470–6481. nativelle-serpentini c, richard s, seralini g e, sourdaine p (2003), ‘Aromatase activity modulation by lindane and bisphenol-A in human placental JEG-3 and transfected kidney E293 cells’, Toxicol In Vitro, 17 (4), 413–422. newbold r r, padilla-banks e, snyder r j, phillips t m, jefferson w n (2007), ‘Developmental exposure to endocrine disruptors and the obesity epidemic’, Reprod Toxicol, 23 (3), 290–296. odum j, ashby j (2000), ‘Neonatal exposure of male rats to nonylphenol has no effect on the reproductive tract’, Toxicol Sci, 56 (2), 400–404. palanza p l, howdeshell k l, parmigiani s, vom saal f s (2002), ‘Exposure to a low dose of BPA during fetal life or in adulthood alters maternal behavior in mice’, Environ Health Perspect, 110 (Suppl 3), 415–422. paris f, jeandel c, servant n, sultan c (2006), ‘Increased serum estrogenic bioactivity in three male newborns with ambiguous genitalia: a potential consequence of prenatal exposure to environmental endocrine disruptors’, Environ Res, 100 (1), 39–43. petersen s l, krishnan s, hudgens e d (2006), ‘The aryl hydrocarbon receptor pathway and sexual differentiation of neuroendocrine functions’, Endocrinology, 147 (6 Suppl), S33–42. pflieger-bruss s, schuppe h c, schill w b (2004), ‘The male reproductive system and its susceptibility to endocrine disrupting chemicals’, Andrologia, 36 (6), 337–345. pierik f h, burdorf a, nijman j m, de muinck keizer-schrama s m, juttmann r e, weber r f (2002), ‘A high hypospadias rate in The Netherlands’, Hum Reprod, 17 (4), 1112–1115. plessow d, waldschlager j, richter d u, jeschke u, bruer g, briese v, friese k (2003), ‘Effects of phytoestrogens on the trophoblast tumour cell lines BeWo and Jeg3’, Anticancer Res, 23 (2A), 1081–1086. prins g s, birch l, tang w y, ho s m (2007), ‘Developmental estrogen exposures predispose to prostate carcinogenesis with aging’, Reprod Toxicol, 23 (3), 374–382. putz o, schwartz c b, leblanc g a, cooper r l, prins g s (2001), ‘Neonatal low- and high-dose exposure to estradiol benzoate in the male rat: II. Effects on male puberty and the reproductive tract’, Biol Reprod, 65 (5), 1506–1517. revich b, aksel e, ushakova t, ivanova i, zhuchenko n, klyuev n, brodsky b, sotskov y (2001), ‘Dioxin exposure and public health in Chapaevsk, Russia’, Chemosphere, 43 (4–7), 951–966. rubin b s, lenkowski j r, schaeberle c m, vandenberg l n, ronsheim p m, soto a m (2006), ‘Evidence of altered brain sexual differentiation in mice exposed perinatally to low, environmentally relevant levels of BPA’, Endocrinology, 147 (8), 3681–3691. ruden d m, xiao l, garfinkel m d, lu x (2005), ‘Hsp90 and environmental impacts on epigenetic states: a model for the trans-generational effects of diethylstibesterol on uterine development and cancer’, Hum Mol Genet, 14 Spec No 1, R149–155. rune g m, desouza p, krowke r, merker h j, neubert d (1991), ‘Morphological and histochemical effects of 2,3,7,8-tetrachlorodibenzo-p dioxin (TCDD) on marmoset (Callithrix jacchus) testes’, Arch Androl, 26 (3), 143–154. sara mariasole g, lifang h, pier alberto b, andrea b (2006), ‘Dioxin effects on neonatal and infant thyroid function: routes of perinatal exposure, mechanisms

The effect of dietary endocrine disruptors on the developing fetus

33

of action and evidence from epidemiology studies’, Internat Arch Occup Env Health, V79 (5), 396–404. savabieasfahani m, kannan k, astapova o, evans n p, padmanabhan v (2006), ‘Developmental programming: differential effects of prenatal exposure to bisphenol-A or methoxychlor on reproductive function’, Endocrinology, 147 (12), 5956–5966. saxena m c, siddiqui m k, seth t d, krishna murti c r, bhargava a k, kutty d (1981), ‘Organochlorine pesticides in specimens from women undergoing spontaneous abortion, premature of full-term delivery’, J Anal Toxicol, 5 (1), 6–9. setchell k d, welsh m b, lim c k (1987), ‘High-performance liquid chromatographic analysis of phytoestrogens in soy protein preparations with ultraviolet, electrochemical and thermospray mass spectrometric detection’, J Chromatogr, 386, 315–323. sharpe r m, skakkebaek n e (1993), ‘Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract?’ Lancet, 341 (8857), 1392–1395. shaw i c (2000), ‘Pesticides’ residues in food’, J Env Monit, 2, 34N-37N. shaw i c, chadwick j (1998), Principles of Environmental Toxicology, Taylor & Francis Ltd, London. shaw i c, thomson b m, cressey p j (2004), Xenoestrogens’ in Pesticides, Veterinary and other Residues in Food, Ed. D Watson, Woodhead Publishing Ltd, Cambridge, Chapter 17. sheehan d m, willingham e, gaylor d, bergeron j m, crews d (1999), ‘No threshold dose for estradiol-induced sex reversal of turtle embryos: how little is too much?’ Environ Health Perspect, 107 (2), 155–159. sheftel v o (1995), Handbook of Toxic Properties of Monomers and Additives, CRC Press, New York. shi z, valdez k e, ting a y, franczak a, gum s l, petroff b k (2007), ‘Ovarian endocrine disruption underlies premature reproductive senescence following environmentally relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzo-p-dioxin’, Biol Reprod, 76 (2), 198–202. shin m (2004), ‘The passage of estrogens and other factors across the human placenta and actions of cellular communication and functions’, Masters Thesis, University of Auckland. stillman r j (1982), ‘In utero exposure to diethylstilbestrol: adverse effects on the reproductive tract and reproductive performance and male and female offspring’, Am J Obstet Gynecol 142 (7), 905–921. sugiura-ogasawara m, ozaki y, sonta s, makino t, suzumori k (2005), ‘Exposure to BPA is associated with recurrent miscarriage’, Hum Reprod, 20 (8), 2325–2329. sultan c, balaguer p, terouanne b, georget v, paris f, jeandel c, lumbroso s, nicolas j-c (2001), ‘Environmental xenoestrogens, antiandrogens and disorders of male sexual differentiation’, Mol Cell Endocrinol, 178 (1–2), 99–105. sumpter j p, jobling s (1995), ‘Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment’, Environ Health Perspect, 103 (Suppl 7), 173–178. sutherland w j (2002), ‘Conservation biology: science, sex and the kakapo’, Nature, 419 (6904), 265–266. swan s h, liu f, overstreet j w, brazil c, skakkebaek n e (2007), ‘Semen quality of fertile US males in relation to their mothers’ beef consumption during pregnancy’, Hum Reprod, 22 (6), 1497–1502. sweeney t (2002), ‘Is exposure to endocrine disrupting compounds during fetal/ post-natal development affecting the reproductive potential of farm animals?’ Domest Anim Endocrinol, 23 (1–2), 203–209.

34

Endocrine-disrupting chemicals in food

takahashi o, oishi s (2000), ‘Disposition of orally administered 2,2-bis(4hydroxyphenyl)propane (BPA) in pregnant rats and the placental transfer to fetuses’, Environ Health Perspect, 108 (10), 931–935. takai y, tsutsumi o, ikezuki y, hiroi h, osuga y, momoeda m, yano t, taketani y (2000), ‘Estrogen receptor-mediated effects of a xenoestrogen, BPA, on preimplantation mouse embryos’, Biochem Biophys Res Commun, 270 (3), 918–921. thomson b m, cressey p j, shaw i c (2003), ‘Dietary exposure to xenoestrogens in New Zealand’, J Environ Monit, 5 (2), 229–235. timbrell j a (1991), Principles of Biochemical Toxicology, Taylor & Francis Ltd, London. timms b g, howdeshell k l, barton l, bradley s, richter c a, vom saal f s (2005), ‘Estrogenic chemicals in plastic and oral contraceptives disrupt development of the fetal mouse prostate and urethra’, Proc Natl Acad Sci USA, 102 (19), 7014–7019. todaka e, sakurai k, fukata h, miyagawa h, uzuki m, omori m, osada h, ikezuki y, tsutsumi o, iguchi t, mori c (2005), ‘Fetal exposure to phytoestrogens–the difference in phytoestrogen status between mother and fetus’, Environ Res, 99 (2), 195–203. toppari j, skakkebaek n e (1998), ‘Sexual differentiation and environmental endocrine disrupters’, Baillieres Clin Endocrinol Metab, 12 (1), 143–156. toppari j, larsen j c, christiansen p, giwercman a, grandjean p, guillette l j, jr., jegou b, jensen t k, jouannet p, keiding n, leffers h, mclachlan j a, meyer o, muller j, rajpert-de meyts e, scheike t, sharpe r, sumpter j, skakkebaek n e (1996), ‘Male reproductive health and environmental xenoestrogens’, Environ Health Perspect, 104 (Suppl 4), 741–803. tyl r w, myers c b, marr m c, thomas b f, keimowitz a r, brine d r, veselica m m, fail p a, chang t y, seely j c, joiner r l, butala j h, dimond s s, cagen s z, shiotsuka r n, stropp g d, waechter j m (2002), ‘Three-generation reproductive toxicity study of dietary BPA in CD Sprague-Dawley rats’, Toxicol Sci, 68 (1), 121–146. vom saal f s, timms b g, montano m m, palanza p, thayer k a, nagel s c, dhar m d, ganjam v k, parmigiani s, welshons w v (1997), ‘Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses’, Proc Natl Acad Sci USA, 94 (5), 2056–2061. vom saal f s, cooke p s, buchanan d l, palanza p, thayer k a, nagel s c, parmigiani s, welshons w v (1998), ‘A physiologically based approach to the study of BPA and other estrogenic chemicals on the size of reproductive organs, daily sperm production, and behavior’, Toxicol Ind Health, 14 (1–2), 239–260. welshons w v, thayer k a, judy b m, taylor j a, curran e m, vom saal f s (2003), ‘Large effects from small exposures. I. Mechanisms for endocrine-disrupting chemicals with estrogenic activity’, Environ Health Perspect, 111 (8), 994–1006. werler m (2007), ‘Maternal smoking and undescended testes: reaching a tipping point’, Epidemiology, 18 (2), 197–198. west m c, anderson l, mcclure n, lewis s e (2005), ‘Dietary oestrogens and male fertility potential’, Hum Fertil (Camb), 8 (3), 197–207. wilbur p m (1984), ‘Fetotoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) for rhesus macaques (Macaca mulatta)’, Am J Primatol, 6 (1), 41–47. wozniak a l, bulayeva n n, watson c s (2005), ‘Xenoestrogens at picomolar to nanomolar concentrations trigger membrane estrogen receptor-alpha-mediated Ca2+ fluxes and prolactin release in GH3/B6 pituitary tumor cells’, Environ Health Perspect, 113 (4), 431–439. xu y, cook t j, knipp g t (2005), ‘Effects of di-(2-ethylhexyl)-phthalate (DEHP) and its metabolites on fatty acid homeostasis regulating proteins in rat placental HRP1 trophoblast cells’, Toxicol Sci, 84 (2), 287–300.

The effect of dietary endocrine disruptors on the developing fetus

35

yoshikawa y, hayashi a, inai m, matsushita a, shibata n, takada k (2002), ‘Permeability characteristics of endocrine-disrupting chemicals using an in vitro cell culture model, Caco-2 cells’, Curr Drug Metab, 3 (5), 551–557. yoshino s, yamaki k, li x, sai t, yanagisawa r, takano h, taneda s, hayashi h, mori y (2004), ‘Prenatal exposure to BPA up-regulates immune responses, including T helper 1 and T helper 2 responses, in mice’, Immunology, 112 (3), 489–495. zalko d, soto a m, dolo l, dorio c, rathahao e, debrauwer l, faure r, cravedi j p (2003), ‘Biotransformations of BPA in a mammalian model: answers and new questions raised by low-dose metabolic fate studies in pregnant CD1 mice’, Environ Health Perspect, 111 (3), 309–319.

2 Human epidemiologic studies of exposure to endocrine-disrupting chemicals and altered hormone levels J. D. Meeker, University of Michigan School of Public Health, USA

Abstract: There is growing evidence that some chemicals which are commonly encountered by humans in the environment can alter endocrine function, potentially leading to reproductive, developmental, and other disorders. This chapter reviews findings from human studies that have been conducted to date on hormone levels in relation to exposure to phthalates, polychlorinated biphenyls, pesticides, metals, and other potential endocrine-disrupting compounds (EDCs). The chapter also discusses future trends and research needs in identifying risks associated with human exposure to known or suspected EDCs. Key words: bisphenol A, endocrine disruption, metals, PBDE, PCB, pesticides, phthalates.

2.1

Introduction

There is scientific, governmental, and public concern over the potential adverse human health risks of exposure to environmental endocrinedisrupting compounds (EDCs). In males, environmental or occupational exposures to EDCs may be associated with or lead to declined reproductive capacity or possibly increased risk of testicular or prostate cancer (Fleming et al., 1999; Pfleiger-Bruss et al., 2004; Toft et al., 2004). In fact, a number of studies have suggested the use of reproductive hormone levels (folliclestimulating hormone (FSH) and/or inhibin B) as a surrogate measure for semen quality or fecundity in epidemiologic studies (Jensen et al., 1997; Uhler et al., 2003; Mabeck et al., 2005), although other recent studies suggest hormone levels may lack sufficient ability to predict poor semen quality (Dhooge et al., 2007; Meeker et al., 2007a). Hormone alterations in females resulting from environmental or occupational exposure may represent

Human epidemiologic studies

37

increased risk for endometriosis, reproductive and other endocrine-related cancers, or impaired oocyte competence, ovarian function, or menstrual cycling (Nicolopoulou-Stamati and Pitsos, 2001; Pocar et al., 2003; Windham et al., 2005). Effects of early life exposures to EDCs remain unclear, though it has been suggested that fetal or childhood exposure may lead to altered sex differentiation (Toppari and Skakkebaek, 1998), effects on neurological and reproductive development (Tilson, 1998; Teilmann et al., 2002; Colborn, 2004, 2006; Swan et al., 2005) and increased risk of reproductive problems or cancer later in life (Damgaard et al., 2002; Aksglaede et al., 2006; Main et al., 2006a). For a more in-depth discussion of clinical end points related to EDC exposure and hormone disruption, see Chapter 3 and other chapters from Part 1 of this text. Exposure to EDCs may cause altered hormone levels through a number of biological mechanisms alone or in combination at different levels of the hypothalamic–pituitary–gonad/thyroid axis, ranging from effects on hormone receptors to effects on hormone synthesis, secretion, or metabolism (Boas et al., 2006; Bretveld et al., 2006). The purpose of this chapter is not to discuss the various biological pathways or the thousands of animal and in vitro studies that have been conducted on EDCs and potential EDCs, but rather to review the existing epidemiologic literature on human exposure to these compounds and circulating hormone levels. While the health impacts of sub-clinical alterations in circulating hormone levels remain unclear, there is a limited but growing body of evidence for these changes to be associated with environmental and occupational exposure to potential EDCs. In addition, altered circulating hormone levels may cause, accompany, or be a result of other exposure-related disorders. Because such a large number of people are exposed to background levels of a number of proven or suspected EDCs, even seemingly subtle epidemiologic associations may result in large increases reproductive and other endocrine-related disease among populations and thus should be of great public health concern. This chapter is meant as an introductory review of human studies conducted in this area to date. The reader is directed to the individual references for additional study detail.

2.2

Persistent organochlorines

Organochlorines have been fairly well-studied in humans and may be associated with a number of adverse health outcomes related to endocrine disruption (Toft et al., 2004). Polychlorinated biphenyls (PCBs) are a class of synthetic, persistent, lipophilic, halogenated aromatic compounds that were widely used in industrial and consumer products for decades before their production was banned in the late 1970s. PCBs were used in cutting oils, lubricants and as electrical insulators. Organochlorine pesticides (OCs) were introduced in the 1940s for their effectiveness against a variety of

38

Endocrine-disrupting chemicals in food

insects. However, owing to their environmental persistence and potential negative impact on human health, they were banned for most uses in developed nations during the 1970s and 1980s. Two of the OCs for which there is convincing evidence of continued human exposure include dichlorodiphenyltrichloroethane (DDT) and hexachlorobenzene (HCB). DDT was commonly used in agriculture and against mosquitos to prevent malaria, but its use has been prohibited by most nations beginning in the 1970s. DDT continues to be used in some developing countries for malaria control as is evidenced by ongoing elevated exposure in human populations (Burke et al., 2003; Tanabe and Kunisue, 2007). A major environmental and biological metabolite of DDT, dichlorodiphenyldichloroethylene (DDE), is also of concern for its toxicity in laboratory studies and because it is more persistent than DDT (ATSDR, 2002a). HCB, a fungicide used to pretreat grain, is also still produced in several countries despite a ban on its use in most industrialized nations in the 1970s and 1980s (ATSDR, 2002b; Barber et al., 2005). As a result of their extensive use and persistence, OCs remain ubiquitous environmental contaminants because of their very long environmental half-life. They are distributed worldwide and have been measured in food, air, water, house dust, soil, aquatic and marine sediments, fish, and wildlife. Furthermore, they are biologically concentrated and stored in human adipose tissue. Thus, the bioavailability of OCs is reduced under normal circumstances but increased during fat mobilization (e.g. weight loss), and are readily passed to infants of breastfeeding mothers. The general population is exposed primarily through ingestion of contaminated foods (e.g., fish, meat, and dairy products), as organochlorines can bioaccumulate up the food chain. Measurable levels of several of these compounds continue to be found in the majority of blood samples from the general population (CDC, 2005; Minh et al., 2006; Thomas et al., 2006).

2.2.1 Reproductive hormones There are a number of recent studies of reproductive hormone levels in men with high levels of p,p′-DDE associated with DDT application, but results have been inconsistent (Ayotte et al., 2001; Martin et al., 2002; Cocco et al., 2004; Dalvie et al., 2004). A small study among 24 young men from Chiapas, Mexico, where DDT was being used found a positive association between p,p′-DDE and sex hormone binding globulin (SHBG) and an inverse association between p,p′-DDE and free testosterone (Ayotte et al., 2001). Conversely, a study of 50 men working near a malaria control center in South Africa reported positive associations between DDT or its metabolites and testosterone and estradiol (Dalvie et al., 2004). A positive association between DDT metabolites and estradiol was also reported in a study of adult men from northern Thailand who had relatively high exposures (Asawasinsopon et al., 2006a), though no associations were observed

Human epidemiologic studies

39

between DDT exposure markers and testosterone, FSH, or luteinizing hormone (LH). Other studies have reported no significant findings. A study of 137 North Carolina farmers, who had much lower p,p′-DDE levels than men in the Mexican and South African studies, found no association with testosterone levels (Martin et al., 2002). A retrospective study of men in an Italian anti-malarial campaign in the late 1940s also found no evidence of an association between past exposure to DDT and sex hormone levels later in life (Cocco et al., 2004). Several studies have been conducted among populations with OC exposure primarily through fish consumption. A recent US study of male sportcaught fish consumers found significant inverse associations between serum PCB concentrations and SHBG-bound testosterone but not total or free testosterone, suggesting PCB exposure may affect steroid binding (Turyk et al., 2006). Similar results were reported in an earlier study of an overlapping but larger population of men who had fished in the Great Lakes (Persky et al., 2001). In Sweden, there was no association between PCB 153 or p,p′-DDE and reproductive hormones among 195 male fisherman (Rignell-Hydbom et al., 2004), but a study of young men from the general Swedish population with lower background exposure levels found an inverse association between PCB 153 and free testosterone (Richthoff et al., 2003). A study of 110 Latvian and Swedish men who consumed varying amounts of fatty fish from the Baltic Sea reported no significant associations between PCBs, HCB or DDT metabolites, and FSH, LH, prolactin, or testosterone (Hagmar et al., 2001a). Among an Inuit population in Greenland a positive association between serum levels of PCB 153 and LH was reported, but these results were inconsistent with other European cohorts included in the study (Giwercman et al., 2006). When all cohorts were combined there were positive associations between PCB 153 and SHBG and between p,p′-DDE and FSH (Giwercman et al., 2006). Studies of female reproductive hormones in relation to OC exposure are rare compared with studies among males, probably because of issues accounting for the high temporal variability of reproductive hormone levels in women. Two recent studies have suggested DDT exposure is associated with adverse effects on ovarian function and menstrual cycling. A US study of 50 Southeast Asian immigrant women reported that increased DDT and/or p,p′-DDE in serum was associated with significantly shorter menstrual cycle (luteal phase) length as well as declined luteal-phase progesterone metabolite levels in urine (Windham et al., 2005). Serum PCB levels were not associated with the outcomes measured in the study. In 287 newly married Chinese women attempting to become pregnant, serum p,p′-DDE was associated with decreased levels of estrogen metabolites in the peri-ovulation phase and decreased levels of progesterone metabolite levels in the luteal phase (Perry et al., 2006). The timing of these alterations may be detrimental to proper ovulation and early pregnancy maintenance.

40

Endocrine-disrupting chemicals in food

2.2.2 Thyroid hormones There is much animal data on the potential for PCBs, p,p′-DDE and HCB to alter thyroid hormones (ATSDR, 2000, 2002a,b). Human studies on PCBs and thyroid hormones are rather numerous, while studies of DDT, DDE, and/or HCB and thyroid hormones are more limited. Although numerous, findings from human studies on PCBs and thyroid hormone levels are difficult to interpret as study populations have differed greatly and results have not been consistent across studies (Hagmar, 2003). A recent study among 341 male partners in couples presenting to a Massachusetts infertility clinic reported an inverse association between serum PCBs and total T3 when adjusting for potential covariates, including serum p,p′-DDE concentration (Meeker et al., 2007b). Serum OC distributions in the men were similar to levels found among the general population (CDC, 2005). Consistent with those results, a small study among 16 obese men that underwent a 15-week weight-loss program reported that T3 levels were significantly and inversely related to plasma concentrations of several PCB congeners (Pelletier et al., 2002). The associations for PCBs remained after adjusting for confounding by weight loss, as weight loss is independently associated with decreased T3 and with increased concentrations of lipophilic compounds, such as OCs, that are released from adipose tissue during fat mobilization, which leads to increased concentrations in circulating blood. Conversely, a study of adult male Baltic Sea fish consumers reported no associations between PCBs and thyroid hormones (Hagmar et al., 2001a), while a study of 178 US men who had fished in the Great Lakes reported an inverse association between PCBs and total T4 but no association with total T3 (Persky et al., 2001). A more recent study of male sport-caught fish consumers found inverse associations between serum PCB concentrations and T3, T4, and thyroid-stimulating hormone (TSH) (Turyk et al., 2006). As mentioned, OCs bioaccumulate up the food chain so consumption of fish from contaminated waters is currently a major source of human OC exposure, and populations nearer to polluted waterways that frequently eat locally caught fish have higher OC exposures than those found among the general population (Langer et al., 2007). Relationships between PCBs and thyroid status in children, women, or pregnant women may not be comparable to results in men, but several epidemiologic studies have been carried out in these populations. Consistent with results from the recent Massachusetts study was a Canadian study of 149 pregnant women that found a significant inverse relationship between low environmental levels of PCB 153 and ΣPCBs in plasma and total T3, but no associations with free T4 and TSH (Takser et al., 2005). Likewise, inverse associations between PCBs and T3 levels were also previously observed in the wives of Swedish fishermen (Hagmar et al., 2001b) and in German schoolchildren (Osius et al., 1999), while another study among pregnant Dutch women reported an inverse association with T3 that was also accompanied by an inverse association with T4 in the women and a

Human epidemiologic studies

41

positive association with TSH in the infants following birth (KoopmanEsseboom et al., 1994). Other studies have reported an inverse association between PCBs and T4 and/or a positive association between PCBs and TSH, with no association between PCBs and total T3 (Emmett et al., 1988; Schell et al., 2004; Wang et al., 2005), while another reported no associations between PCBs and thyroid hormones among 160 neonates from North Carolina from 1978 to 1982 (Longnecker et al., 2000). Inconsistent results have been reported for the association between p,p′DDE and thyroid hormones. In the Massachusetts male study associations between p,p′-DDE and increased free T4 and total T3, and decreased TSH, were observed (Meeker et al., 2007b). Conversely, a study of male sportcaught fish consumers found a suggestive inverse association between p,p′DDE and T4 (Turyk et al., 2006), while two other studies of male fishermen found no associations between p,p′-DDE and thyroid hormones (Hagmar et al., 2001a; Persky et al., 2001). Takser et al. (2005) found an inverse association between p,p′-DDE and total T3, with no associations between p,p′DDE and free T4 or TSH in pregnant women. A recent Thai study of OCs and thyroid hormones in cord serum from 39 infants found a significant inverse association between p,p′-DDE and total T4, suggesting a potential role for DDT and its metabolites to adversely affect fetal or infant neurodevelopment (Asawasinsopon et al., 2006b). A small number of studies have explored the relationship between HCB and thyroid hormones, though many of them overlap with studies of PCBs and/or DDE. No association with thyroid hormones was reported in one study among Swedish fishermen (Hagmar et al., 2001a), while several other human studies have reported statistically significant relationships. Similar to their results for PCBs, HCB was inversely associated with total T3 among Massachusetts men when taking into account serum p,p′-DDE concentrations (Meeker et al., 2007b). Two other studies, also mentioned earlier for their PCB results, likewise reported inverse associations between plasma HCB levels and total T3 in obese men following participation in a weight loss program (Pelletier et al., 2002) and among pregnant women (Takser et al., 2005). A study among 608 adults in Spain living near an OC factory that produced HCB found a significant inverse association between serum HCB levels and total T4 (Sala et al., 2001). Total T3 levels were not measured in the study, and they found no associations between HCB and free T4 or TSH. Using preliminary data from 66 men in the New York State Angler Cohort Study, Bloom et al. (2003) also reported an inverse relationship between serum HCB levels and total T4. The study did not measure free T4, total T3, or TSH. Taken together, these studies suggest that there is a relationship between OC exposure and altered hormone levels in humans. However, the inconsistent nature of the specific results underscores the need for future welldesigned studies among men, women, and children that are large enough to provide adequate statistical power to test these associations while taking

42

Endocrine-disrupting chemicals in food

into account important confounding variables such as age, smoking and body mass index (BMI), among others. Also needed is more in-depth exploration into the potential that multiple mechanisms of action are taking place simultaneously since levels of OC pesticides, individual PCB congeners, and various OC metabolites in blood and tissue are often times highly correlated with one another.

2.3

Non-persistent pesticides

The term ‘non-persistent pesticides’ (also commonly called ‘contemporaryuse pesticides’) refers to chemical mixtures that are currently available to control insects (insecticides), weeds (herbicides), fungi (fungicides), or other pests (e.g. rodenticides), as opposed to pesticides that have been banned from use in most countries (e.g. many of the formerly popular OC pesticides such as DDT). Some common classes of non-persistent pesticides in use today include organophosphates, carbamates, and pyrethroids. Though environmentally non-persistent, owing to the extensive use of pest control in these various settings a majority of the general population is exposed to some of the more widely used pesticides at low levels. Human studies on non-persistent pesticide exposure and reproductive hormones are limited. A study among Danish farmers found that traditional farmers, who were presumably more highly exposed to pesticides, had a lower testosterone/SHBG ratio (free androgen index) than organic farmers (Larsen et al., 1999). In a prospective follow-up study of 67 professional pesticide applicators and 125 comparison subjects, Straube and co-workers (1999) found a significant increase in LH and a significant decrease in testosterone and estradiol in the pesticide applicators. They also found non-significant increases in FSH and prolactin. Among Chinese factory workers exposed to the organophosphates (OPs) parathion and methamidophos, Padungtod et al. (1998) found that exposure was associated with increased serum LH and decreased serum testosterone. A more recent study of pesticide applicators in Peru who had sprayed OPs also reported significantly declined testosterone and LH levels compared with controls, as well as a significantly increased testosterone : LH ratio (Yucra et al., 2006). A small number of epidemiological studies have also explored the relationship between exposure to non-persistent pesticides and thyroid hormone levels. A study of Mexican pesticide applicators highly exposed to ethylenebis(dithiocarbamate) (EBDC) fungicides reported an increase in TSH associated with exposure, but no decline in T4 (Steenland et al., 1997). Zaidi and co-workers (2000) also found increased TSH among pesticide formulators in India exposed to a number of compounds, along with significantly decreased total T3 levels and suggestively decreased T4, compared with controls. A more recent study reported only minor disturbances in thyroid

Human epidemiologic studies

43

hormone levels in Danish greenhouse workers, with no clear patterns in relation to the exposure metrics used (Toft et al., 2006). Results from these studies suggest a possible association between insecticide exposure and male endocrine function, most notably decreased testosterone. However, the majority of these studies had small numbers of subjects which limited statistical power. Also, the non-specific nature of the exposure assessments in these studies somewhat limit their interpretation. Researchers are now beginning to utilize urinary and serum biomarkers of pesticide exposure to explore associations of specific pesticides with adverse impacts on health, though these studies remain limited in number. Urinary markers of OP exposure were significantly associated with decreased FSH and LH, and suggestively associated with decreased estradiol, in Mexican agricultural workers (Recio et al., 2005). However, no associations with testosterone were reported. Using urinary biomarker data representative of low environmental levels of non-persistent pesticides commonly encountered among the general population, the Massachusetts infertility clinic study investigated associations between urinary concentrations of 3,5,6-trichloro-2-pyridinol (TCPY) and 1-naphthol (1N), metabolites of chlorpyrifos and carbaryl, respectively, and serum reproductive hormone levels in men (Meeker et al., 2006a). After adjusting for age, BMI, smoking, and time of blood sample, there was a statistically significant inverse association between TCPY and testosterone. An interquartile range increase in TCPY was associated with a 25 ng/dL (95% confidence interval −40, −10) decline in testosterone concentration. The association appeared to be dose-dependent when exposure was divided into quintiles. The highest TCPY quintile was associated with a testosterone decline of 83 ng/dL (−128, −39) compared with the lowest TCPY quintile (Fig. 2.1). This was equal to a 20% reduction in testosterone among men in the highest TCPY quintile. An inverse association between TCPY and free androgen index (a ratio of testosterone to SHBG), and a suggestive inverse association between TCPY and LH were also observed. Based on results from an animal study that reported an association between chlorpyrifos exposure and thyroxine (T4) levels (Rawlings et al., 1998), the association between TCPY and serum thyroid hormones was also explored in the Massachusetts male population (Meeker et al., 2006b). In multiple linear regression analyses, for the median levels of free T4 (1.26 ng/ dL) and TSH (1.44 μIU/ml) an interquartile range increase in TCPY was associated with a statistically significant 2.5% (−4.8, 0.0%) decline in free T4 and a 9% (0.0, 18%) increase in TSH. There was no association between TCPY and levels of total T3. In these studies urinary 1N concentrations were associated with some of the same hormone levels, but the associations were not as strong as for TCPY and they did not demonstrate the same dose-dependent pattern (Meeker et al., 2006a,b). The suggestive results from these studies for a relationship between exposure to non-persistent pesticides or their metabolites and reductions in free T4 and testosterone

Endocrine-disrupting chemicals in food Change in testosterone level (ng/dL)

44

150

75

0

−75

−150

Q1

Q2

Q3

Q4

Q5

TCPY quintiles

Fig. 2.1 Regression coefficients (diamonds) and associated 95% confidence intervals for a change in testosterone level associated with increasing quintiles of specific gravity-adjusted TCPY in 268 men, adjusted for age, BMI, ln-transformed SHBG, smoking, and time of day that blood sample was collected (Meeker et al., 2006a).

need to be further tested in other human populations. Mechanisms that could more clearly explain these relationships should also be investigated in future animal and in vitro studies.

2.4

Phthalates

The diesters of 1,2-benzenedicarboxylic acid (phthalic acid), commonly known as phthalates, are a group of synthetic chemicals with a wide spectrum of industrial applications. High molecular weight phthalates (e.g., di(2ethylhexyl) phthalate [DEHP], di-isononyl phthalate [DiNP], di-n-octyl phthalate [DnOP]), are primarily used as plasticizers in the manufacture of flexible vinyl which, in turn, is used in consumer products, flooring and wall coverings, food contact applications, and medical devices (ATSDR, 1997, 2002c; David et al., 2001). Manufacturers use low molecular weight phthalates (e.g., diethyl phthalate [DEP] and dibutyl phthalate [DBP]) in personal-care products (e.g., perfumes, lotions, cosmetics), as solvents and plasticizers for cellulose acetate, and in making lacquers, varnishes, and coatings, including those used to provide timed releases in some pharmaceuticals (David et al., 2001; ATSDR, 1995, 2001). Very few human studies have investigated associations between exposure to phthalates and endogenous hormone levels. In a study of workers producing PVC flooring with high exposure to DEHP and DBP, urinary

Human epidemiologic studies

45

concentrations of metabolites of these phthalates were inversely associated with free testosterone levels (Pan et al., 2006). A report on 295 men from the Massachusetts infertility clinic study found a suggestive inverse association between urinary mono(2-ethylhexyl) phthalate (MEHP; a metabolite of DEHP) and testosterone, along with a positive association between urinary mono-n-butyl phthalate (MBP; a urinary metabolite of DBP) and inhibin B (a glycoprotein hormone produced by the gonads that has an inhibitory effect on pituitary FSH production), and an inverse association between urinary monobenzyl phthalate (MBzP; a metabolite of dibenzyl phthalate) and FSH (Duty et al., 2005). However, the significant results for MBP and MBzP and hormone levels were in patterns inconsistent with the authors’ hypotheses. It is interesting to note that although MEHP concentrations in the Massachusetts study were several orders of magnitude lower than those measured in the exposed Chinese workers (Pan et al., 2006), the evidence for decreased testosterone in relation to DEHP/MEHP was consistent between the two studies. On the other hand, a study of 234 young Swedish men found an inverse association between urinary monoethyl phthalate (MEP; a metabolite of DEP) and LH but no association between MEP, MEHP, or other phthalate metabolites in urine and FSH, testosterone, estradiol, or inhibin B (Jonsson et al., 2005). A relationship between phthalates and hormone levels in infants has also been reported (Main et al., 2006b; Lottrup et al., 2006). Within a Danish/ Finnish cohort on cryptorchidism, Main and coworkers (2006b) analyzed breast milk samples for phthalate metabolites and measured reproductive hormone levels in 3-month-old boys. There were positive associations between MEP, monomethyl phthalate (MMP), and MPB with LH : FAI ratio, which is a measure of Leydig cell function. There were also positive associations between MEP, MBP, and SHBG and between mono-isononyl phthalate (MiNP) and LH, and an inverse association between MBP and free testosterone. These results supported earlier findings where a study among US infants found an inverse association between MBP and anogenital separation, which is thought to be a sensitive marker for androgen activity (Swan et al., 2005). The potential for phthalates to affect thyroid function has been demonstrated in animal studies but human studies are limited to a single recent investigation within the Massachusetts male infertility clinic study (Meeker et al., 2007c). Among 408 men with phthalate metabolite concentrations measured in urine and serum thyroid hormone measures, MEHP was inversely associated with free T4 (Fig. 2.2) and total T3 but was not associated with TSH. The inverse association between MEHP and free T4 became stronger when also taking into account concentrations of oxidative DEHP metabolites, which were positively associated with free T4. These results may reflect metabolic susceptibility to the adverse effects of MEHP among individuals who less efficiently oxidize DEHP and/or MEHP (Meeker et al., 2007c).

46

Endocrine-disrupting chemicals in food

Change in free T4 level (ng/dL)

0.2

0.1

0

−0.1

−0.2 Q1

Q2

Q3

Q4

Q5

MEHP quintiles

Fig. 2.2 Regression coefficients (diamonds) and associated 95% confidence intervals for a change in free T4 level associated with increasing quintiles of specific gravity-adjusted MEHP in 408 men, adjusted for age, BMI, smoking, and time of day that blood sample was collected (Meeker et al., 2007c).

2.5

Metals

High exposures to heavy metals are common in a number of occupations such as welding and work in smelters or foundries, though there are relatively few studies assessing the relationship between exposure to metals and hormone levels in these work environments. The general population is exposed to metals at trace concentrations through intake of contaminated food and water or contact with contaminated soil, dust, or air. A number of metals are reproductive toxins and suspected endocrine disruptors, though the biological mechanisms through which they may impact the human endocrine system remain unclear. As with the organic compounds previously discussed, even though exposure is prevalent, human studies of exposure to metals and altered hormone levels to date are quite limited.

2.5.1 Cadmium Cadmium exposure among humans is pervasive, and can result from occupational and/or environmental sources. Environmental sources of cadmium include contaminated ambient air and soil from industrial pollution and other synthetic or natural combustion sources such as cigarette smoke and volcanic activity. Cadmium ingested in foods and drinking water also plays an important role in aggregate human exposure, and is of most concern for long-term exposure among much of the general population since it is a cumulative toxicant (ATSDR, 1999). Cadmium accumulates up the food

Human epidemiologic studies

47

chain and is present in food grown in contaminated soils (e.g. agricultural areas using phosphate fertilizers or sewage sludge) and can be introduced in food processing, preparation, or storage. Cadmium can be measured in virtually any food sample, but food products with the highest cadmium levels include grain and cereal products, potatoes, leafy and root vegetables, organ meats (liver and kidney) and shellfish (Gartrell et al., 1986; ATSDR, 1999). Cadmium is recognized as an endocrine disruptor but the mechanisms involved are not well understood (Henson and Chedrese, 2004). Human, animal, and in vitro study results on cadmium’s effect on hormone levels have not been consistent, and there have not been many human epidemiologic studies on cadmium and hormone levels to date. A recent Nigerian study by Akinloye et al. (2006) reported a positive correlation between cadmium and FSH levels in seminal plasma but no associations with LH or testosterone. Among 166 Chinese men occupationally exposed to cadmium, urinary cadmium concentrations were associated with increased testosterone and LH after adjusting for age, smoking, and alcohol consumption (Zeng et al., 2002). Another Chinese study of 263 men from areas with little, moderate or heavy smelter pollution found that the percentage of men with abnormally high testosterone was associated with increased urinary cadmium concentrations (Zeng et al., 2004). However, the study did not find associations between cadmium and testosterone, FSH or LH in multiple linear regression. A study of 149 male Croatian industrial workers found blood cadmium levels were significantly associated with increased testosterone levels, while there were suggestive associations between blood cadmium and increased LH and reduced prolactin (Telisman et al., 2000). Using multiple linear regression, a more recent study among 123 Croatian men with no specific occupational exposure to metals reported that blood cadmium concentrations were associated with increased serum testosterone, FSH, and estradiol, but decreased serum prolactin, after adjusting for several potential confounders (Jurasovic et al., 2004). Similarly, a study of 164 postmenopausal Japanese women reporting for breast cancer screening also found a significant positive association between urinary cadmium concentration and serum testosterone levels in multivariate analysis (Nagata et al., 2005). Thus, though fairly limited in number, there has been some consistency with regard to the positive relationship between cadmium exposure and testosterone levels in human studies conducted to date.

2.5.2 Lead Lead is one the most well-studied occupational and environmental contaminants among human populations. Although lead exposure levels have been declining in industrialized nations for the past few decades, health effects from low exposure levels remain a concern. Several studies have

48

Endocrine-disrupting chemicals in food

investigated the relationship between exposure to lead and hormone levels in humans, but results have been conflicting. A Belgian study of male lead smelter workers found a significant positive association between blood lead levels and inhibin B (Mahmoud et al., 2005). Two other studies of workers exposed to lead reported positive associations between exposure and FSH and/or LH (McGregor and Mason, 1990; Ng et al., 1991), but additional studies have observed limited or no evidence for associations between occupational lead exposure and reproductive hormones (Alexander et al., 1996; Telisman et al., 2000; Erfurth et al., 2001). Thyroid function in relation to lead exposure has been tested in a number of studies with inconsistent findings (see Dundar et al., 2006). Most recently, exposure was associated with increased TSH (with no changes in T3 or T4) in adult men (Singh et al., 2000) but associated with decreased T4 (with no changes in T3 or TSH) in adolescent males working as auto mechanics (Dundar et al., 2006). The lack of consistent findings across studies may suggest that lead is not acting directly as an endocrine disruptor, but rather that the observed alterations in hormone levels may have stemmed from other lead-related health effects or may have been caused by other unmeasured but correlated exposures.

2.5.3 Manganese, chromium, and other metals Studies of exposure to metals other than cadmium or lead and altered hormone levels are more limited, and findings again have been inconsistent. Among male welders, one European study found suggestive evidence for an inverse association between exposure to stainless steel welding fume and testosterone levels, and a positive dose-related relationship between mild steel welding fume exposure and FSH (Bonde, 1990). Conversely, another European study found no associations between welding and FSH, LH or testosterone levels (Hjollund et al., 1998). A Korean study of male welders reported higher levels of FSH, LH, and thyroid-stimulating hormonereleasing hormone (TRH) compared with age-matched office workers (Kim et al., 2007). More detailed manganese exposure measures revealed welders with higher manganese exposure had significantly higher TRH levels compared with welders with lower manganese exposure and office workers in a dose-dependent manner (Kim et al., 2007). Manganese exposure was also associated with significantly higher serum prolactin levels among male manganese alloy production workers (Ellingsen et al., 2003). Exposure to hexavalent chromium from stainless steel welding was associated with increased FSH in one study (Li et al., 1999) but not in another (Bonde and Ernst, 1992). Finally, mercury, a transition metal and pervasive environmental contaminant, was recently found to be associated with increased estradiol levels in a small residential population in Cambodia (Agusa et al., 2007), which was in agreement with a previous study among women with repeated miscarriages (Gerhard et al., 1998). However, slight to no relationships were reported in earlier studies of occupational exposure to mercury

Human epidemiologic studies

49

and reproductive or thyroid hormone levels (Erfurth et al., 1990; McGregor and Mason, 1991; Barregard et al., 1994).

2.6

Other emerging compounds of concern

There are other classes of chemicals to which humans are widely exposed that require further study as to their relation with altered hormone levels in humans. These chemicals, for which experimental studies have shown endocrine-disrupting potential, include perchlorate, brominated flame retardants such as polybrominated diphenyl ethers (PBDE), bisphenol A (BPA), alkylphenols such as 4-nonylphenol, and fluorinated organic compounds such as perfluorooctane octanoate (PFOA) and perfluorooctane sulfonate (PFOS). Perchlorate is an inorganic anion that is a common contaminant in drinking water and foods. Brominated flame retardants such as PBDEs are used in electronics, furniture, and other consumer products, and the primary exposure route for PBDEs in most humans is through food consumption owing to their ability to bioaccumulate. BPA, an estrogenic compound, is used in the manufacture of polycarbonate plastics and epoxy resins and is often found in food containers and in the lining of food cans, while alkylphenols are used as surface-active agents in cleaning/washing agents, paints, and cosmetics. The perfluorinated compounds are used to make fabrics stainresistant/water repellant and in coatings on cookware and other products. Although widespread human exposure to these chemicals has been demonstrated and animal studies suggest endocrine-disrupting properties, the health effects data in humans remain severely limited. Perchlorate was associated with increased TSH in one ecologic study of infants (Brechner et al., 2000) but not in another (Li et al., 2000). A recent study of adolescents and adults participating in the US National Health and Nutrition Examination Survey (NHANES) found an inverse association between urinary perchlorate concentrations and T4 and a positive association between perchlorate concentration and TSH in women but not in men (Blount et al., 2006). The results for T4 differed when stratifying women by urinary iodide level. A number of brominated flame retardants are considered pervasive environmental contaminants of concern but only two human studies have explored associations between exposure and hormone alterations. Plasma PBDE 47 was inversely associated with TSH in a study of adult males from Sweden and Latvia (Hagmar et al., 2001a), whereas a longitudinal study among a small number of PBDE-exposed workers (n = 11) reported no associations between exposure to sum PBDE or specific congeners and T3, T4, or TSH (Julander et al., 2005). Only two small human studies were located that have investigated exposure to BPA and hormone levels, where statistically significant positive correlations were found between BPA concentrations in serum and circulating total and free testosterone levels in both men and women (Takeuchi and Tsutsumi, 2002; Takeuchi et al., 2004).

50

Endocrine-disrupting chemicals in food

Given the widespread human exposure to these compounds and the increasing concern for their endocrine-disrupting potential, there is a clear need for more epidemiological research.

2.7

Future trends

Although the epidemiological data on these historic, contemporary, and emerging environmental contaminants suggest that there may be associations with altered circulating hormone levels, the quantity and quality of the data available for the different types of compounds varied. For example, although there are hundreds of different pesticides currently in use worldwide, limited human data existed on hormone levels for only a select few. Also, for some of the more studied chemicals, such as PCBs, the data across studies were not consistent. This may be due to small study sizes and lack of statistical power or differences in study design, study populations, exposure levels and sources, multiple/competing physiologic mechanisms, analytical approaches, and potential confounding variables considered in the statistical analysis (age, BMI, season, etc.). The limited human data, and in certain instances inconsistent data across studies, highlight the need for further epidemiological research on these classes of chemicals. Most studies to date have been cross-sectional in nature. Future longitudinal studies are needed to explore the temporal relationship between exposure to EDCs and circulating hormone levels (i.e. causality). Owing to the complex nature of the endocrine system, studies should also explore ratios between relevant hormones in addition to individual hormone levels (e.g. LH : testosterone ratio in males as a marker for Leydig cell function), which may help provide clues for biological mechanisms of xenobiotic activity in humans. Prior to carrying out epidemiological studies, experimental structure–activity relationship (SAR)-based approaches that are performed in a logical manner should be utilized to prioritize candidate EDCs that need to be investigated in human populations (Devillers et al., 2006). A future challenge to understanding the relationship between endocrinedisrupting compounds and hormone levels includes the changes in exposure levels among populations over time due to the ever-changing patterns of production and use of these compounds. Another challenge is to understand how simultaneous co-exposures to these chemicals may affect endocrine function. It is well known that humans are exposed to all of these compounds simultaneously, as well as to many other chemicals. However, most studies to date have addressed only single chemicals or classes of chemicals, and there are limited data on the interactions between chemicals within a class or across classes. Chemicals may interact additively, multiplicatively, or antagonistically in what is commonly referred to as the ‘cocktail effect’. The human health risks of exposure to chemical mixtures are much understudied. Despite these challenges, evolving and innovative technologies designed to improve

Human epidemiologic studies

51

the assessment of human exposure and hormone levels should provide enhanced opportunities for improving our understanding of the relationship between environmental chemicals and endocrine function. Innovations include improved biomarkers of exposure, more sophisticated statistical methods that deal with multiple exposures simultaneously, and sensitive new measures of intermediate alterations in human endocrine function. More information is required on biological mechanisms of EDCs in humans as well as the potential consequences of hormone level changes on the order of those observed in human studies. To date, most hormone alterations attributed to environmental and occupational exposures have been considered subclinical. However, much remains unknown as to whether hormone changes currently considered subclinical may be associated with increased risk of adverse systemic effects in the long term. Furthermore, although seemingly subtle, small changes in hormone levels resulting from exposure may be of public health importance when considering the prevalence of exposure to EDCs among entire populations. Finally, human research is needed on potential latent and transgenerational effects (e.g., epigenetic modifications) of exposure to EDCs as well as genetic, metabolic, demographic, or environmental characteristics owing to increased individual susceptibility for adverse health effects following exposure.

2.8

Sources of further information and advice

US Agency for Toxic Substances and Disease Registry (ATSDR), Toxicological Profiles: http://www.atsdr.cdc.gov/toxpro2.html US Centers for Disease Control and Prevention (CDC), National Report on Human Exposure to Environmental Chemicals: http://www.cdc.gov/ exposurereport/ US National Library of Medicine (NLM), Medline (PubMed) Search: http:// www.pubmed.gov

2.9

References

agusa t, kunito t, iwata h, monirith i, chamnan c, tana ts, subramanian a and tanabe s (2007), ‘Mercury in hair and blood from residents of Phnom Penh (Cambodia) and possible effect on serum hormone levels’, Chemosphere, 68 (3), 590–6. akinloye o, arowojolu ao, shittu ob and anetor ji (2006), ‘Cadmium toxicity: a possible cause of male infertility in Nigeria’, Reprod Biol, 6 (1), 17–30. aksglaede l, juul a, leffers h, skakkebaek ne and andersson am (2006), ‘The sensitivity of the child to sex steroids: possible impact of exogenous estrogens’, Hum Reprod Update, 12 (4), 341–9. alexander bh, checkoway h, van netten c, muller ch, ewers tg, kaufman jd, mueller ba, vaughan tl and faustman em (1996), ‘Semen quality of men employed at a lead smelter’, Occup Environ Med, 53 (6), 411–16.

52

Endocrine-disrupting chemicals in food

asawasinsopon r, prapamontol t, prakobvitayakit o, vaneesorn y, mangklabruks a and hock b (2006a), ‘Plasma levels of DDT and their association with reproductive hormones in adult men from northern Thailand’, Sci Total Environ, 355 (1–3), 98–105. asawasinsopon r, prapamontol t, prakobvitayakit o, vaneesorn y, mangklabruks a and hock b (2006b), ‘The association between organochlorine and thyroid hormone levels in cord serum: a study from northern Thailand’, Environ Int, 32 (4), 554–9. atsdr (1995), Toxicological Profile for Diethyl phthalate (DEP), Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (1997), Toxicological Profile for Di-n-octyl phthalate (DNOP), Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (1999), Toxicological Profile for Cadmium, Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (2000), Toxicological Profile for Polychlorinated Biphenyls (PCBs), Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (2001), Toxicological Profile for Di-n-butyl phthalate (DBP), Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (2002a), Toxicological Profile for DDT, DDE and DDD, Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (2002b), Toxicological Profile for Hexachlorobenzene, Agency for Toxic Substances and Disease Registry, Atlanta, GA. atsdr (2002c), Toxicological Profile for Di(2-ethylhexyl)phthalate (DEHP), Agency for Toxic Substances and Disease Registry, Atlanta, GA. ayotte p, giroux s, dewailly e, hernandez avila m, farias p, danis r and villanueva diaz c (2001), ‘DDT spraying for malaria control and reproductive function in Mexican men’, Epidemiology, 12 (3), 366–7. barber jl, sweetman aj, van wijk d and jones kc (2005), ‘Hexachlorobenzene in the global environment: emissions, levels, distribution, trends and processes’, Sci Total Environ, 349 (1–3), 1–44. barregard l, lindstedt g, schutz a and sallsten g (1994), ‘Endocrine function in mercury exposed chloralkali workers’, Occup Environ Med, 51 (8), 536–40. bloom ms, weiner jm, vena je and beehler gp (2003), ‘Exploring associations between serum levels of select organochlorines and thfyroxine in a sample of New York state sportsmen: the New York State Angler Cohort Study’, Environ Res, 93 (1), 52–66. blount bc, pirkle jl, osterloh jd, valentin-blasini l and caldwell kl (2006), ‘Urinary perchlorate and thyroid hormone levels in adolescent and adult men and women living in the United States’, Environ Health Perspect, 114 (12), 1865–71. boas m, feldt-rasmussen u, skakkebaek ne and main km (2006), ‘Environmental chemicals and thyroid function’, Eur J Endocrinol, 154 (5), 599–611. bonde jp (1990), ‘Semen quality and sex hormones among mild steel and stainless steel welders: a cross sectional study’, Br J Ind Med, 47 (8), 508–14. bonde jp and ernst e (1992), ‘Sex hormones and semen quality in welders exposed to hexavalent chromium’, Hum Exp Toxicol, 11 (4), 259–63. brechner rj, parkhurst gd, humble wo, brown mb and herman wh (2000), ‘Ammonium perchlorate contamination of Colorado River drinking water is associated with abnormal thyroid function in newborns in Arizona’, J Occup Environ Med, 42 (8), 777–82. bretveld rw, thomas cm, scheepers pt, zielhuis ga and roeleveld n (2006), ‘Pesticide exposure: the hormonal function of the female reproductive system disrupted?’, Reprod Biol Endocrinol, 4 (30), 1–14. burke er, holden aj, shaw ic, suharyanto fx and sihombing g (2003), ‘Organochlorine pesticide residues in human milk from primiparous women in Indonesia’, Environ Contam Toxicol, 71, 148–54.

Human epidemiologic studies

53

cdc (2005), Third National Report on Human Exposure to Environmental Chemicals, Centers for Disease Control and Prevention, Washington, DC. cocco p, loviselli a, fadda d, ibba a, melis m, oppo a, serra s, taberlet a, tocco mg and flore c (2004), ‘Serum sex hormones in men occupationally exposed to dichloro-diphenyl-trichloro ethane (DDT) as young adults’, J Endocrinol, 182 (3), 391–7. colborn t (2004), ‘Neurodevelopment and endocrine disruption’, Environ Health Perspect, 112 (9), 944–9. colborn t (2006), ‘A case for revisiting the safety of pesticides: a closer look at neurodevelopment’, Environ Health Perspect, 114 (1), 10–17. dalvie ma, myers je, lou thompson m, dyer s, robins tg, omar s, riebow j, molekwa j, kruger p and millar r (2004), ‘The hormonal effects of long-term DDT exposure on malaria vector-control workers in Limpopo Province, South Africa’, Environ Res, 96 (1), 9–19. damgaard in, main km, toppari j and skakkebaek ne (2002), ‘Impact of exposure to endocrine disrupters in utero and in childhood on adult reproduction’, Best Pract Res Clin Endocrinol Metab, 16 (2), 289–309. david rm, mckee rh, butala jh, barter ra and kayser m (2001), ‘Esters of aromatic mono-, di-, and tricarboxylic acids, aromatic diacids, and di-, tri-, or polyalcohols’, in Patty’s Toxicology (Eds, Bingham E, Cohrssen B and Powell CH), John Wiley and Sons, New York, pp. 635–932. devillers j, marchand-geneste n, carpy a and porcher jm (2006), ‘SAR and QSAR modeling of endocrine disruptors’, SAR QSAR Environ Res, 17 (4), 393–412. dhooge w, van larebeke n, comhaire f and kaufman jm (2007), ‘Regional variations in semen quality of community-dwelling young men from flanders are not paralleled by hormonal indices of testicular function’, J Androl, 28 (3), 435–43. dundar b, oktem f, arslan mk, delibas n, baykal b, arslan c, gultepe m and ilhan ie (2006), ‘The effect of long-term low-dose lead exposure on thyroid function in adolescents’, Environ Res, 101 (1), 140–5. duty sm, calafat am, silva mj, ryan l and hauser r (2005), ‘Phthalate exposure and reproductive hormones in adult men’, Hum Reprod, 20 (3), 604–10. ellingsen dg, haug e, gaarder pi, bast-pettersen r and thomassen y (2003), ‘Endocrine and immunologic markers in manganese alloy production workers’, Scand J Work Environ Health, 29 (3), 230–8. emmett ea, maroni m, jefferys j, schmith j, levin bk and alvares a (1988), ‘Studies of transformer repair workers exposed to PCBs: II. Results of clinical laboratory investigations’, Am J Ind Med, 14 (1), 47–62. erfurth em, schutz a, nilsson a, barregard l and skerfving s (1990), ‘Normal pituitary hormone response to thyrotrophin and gonadotrophin releasing hormones in subjects exposed to elemental mercury vapour’, Br J Ind Med, 47 (9), 639–44. erfurth em, gerhardsson l, nilsson a, rylander l, schutz a, skerfving s and borjesson j (2001), ‘Effects of lead on the endocrine system in lead smelter workers’, Arch Environ Health, 56 (5), 449–55. fleming le, bean ja, rudolph m and hamilton k (1999), ‘Cancer incidence in a cohort of licensed pesticide applicators in Florida’, J Occup Environ Med, 41 (4), 279–88. gartrell mj, craun jc, podrebarac ds and gunderson el (1986), ‘Pesticides, selected elements, and other chemicals in adult total diet samples, October 1980–March 1982’, J Assoc Off Anal Chem, 69 (1), 146–59. gerhard i, waibel s, daniel v and runnebaum b (1998), ‘Impact of heavy metals on hormonal and immunological factors in women with repeated miscarriages’, Hum Reprod Update, 4 (3), 301–9. giwercman ah, rignell-hydbom a, toft g, rylander l, hagmar l, lindh c, pedersen hs, ludwicki jk, lesovoy v, shvets m, spano m, manicardi gc, bizzaro d, bonefeld-jorgensen ec and bonde jp (2006), ‘Reproductive hormone levels in

54

Endocrine-disrupting chemicals in food

men exposed to persistent organohalogen pollutants: a study of Inuit and three European cohorts’, Environ Health Perspect, 114 (9), 1348–53. hagmar l (2003), ‘Polychlorinated biphenyls and thyroid status in humans: a review’, Thyroid, 13 (11), 1021–8. hagmar l, bjork j, sjodin a, bergman a and erfurth em (2001a), ‘Plasma levels of persistent organohalogens and hormone levels in adult male humans’, Arch Environ Health, 56 (2), 138–43. hagmar l, rylander l, dyremark e, klasson-wehler e and erfurth em (2001b), ‘Plasma concentrations of persistent organochlorines in relation to thyrotropin and thyroid hormone levels in women’, Int Arch Occup Environ Health, 74 (3), 184–8. henson mc and chedrese pj (2004), ‘Endocrine disruption by cadmium, a common environmental toxicant with paradoxical effects on reproduction’, Exp Biol Med (Maywood), 229 (5), 383–92. hjollund nh, bonde jp, jensen tk, ernst e, henriksen tb, kolstad ha, giwercman a, skakkebaek ne and olsen j (1998), ‘Semen quality and sex hormones with reference to metal welding’, Reprod Toxicol, 12 (2), 91–5. jensen tk, andersson am, hjollund nh, scheike t, kolstad h, giwercman a, henriksen tb, ernst e, bonde jp, olsen j, mcneilly a, groome np and skakkebaek ne (1997), ‘Inhibin B as a serum marker of spermatogenesis: correlation to differences in sperm concentration and follicle-stimulating hormone levels. A study of 349 Danish men’, J Clin Endocrinol Metab, 82 (12), 4059–63. jonsson ba, richthoff j, rylander l, giwercman a and hagmar l (2005), ‘Urinary phthalate metabolites and biomarkers of reproductive function in young men’, Epidemiology, 16 (4), 487–93. julander a, karlsson m, hagstrom k, ohlson cg, engwall m, bryngelsson il, westberg h and van bavel b (2005), ‘Polybrominated diphenyl ethers – plasma levels and thyroid status of workers at an electronic recycling facility’, Int Arch Occup Environ Health, 78 (7), 584–92. jurasovic j, cvitkovic p, pizent a, colak b and telisman s (2004), ‘Semen quality and reproductive endocrine function with regard to blood cadmium in Croatian male subjects’, Biometals, 17 (6), 735–43. kim ea, cheong hk, joo kd, shin jh, lee js, choi sb, kim mo, lee iu j and kang dm (2007), ‘Effect of manganese exposure on the neuroendocrine system in welders’, Neurotoxicology, 28 (2), 263–9. koopman-esseboom c, morse dc, weisglas-kuperus n, lutkeschipholt ij, van der paauw cg, tuinstra lg, brouwer a and sauer pj (1994), ‘Effects of dioxins and polychlorinated biphenyls on thyroid hormone status of pregnant women and their infants’, Pediatr Res, 36 (4), 468–73. langer p, kocan a, tajtakova m, petrik j, chovancova j, drobna b, jursa s, radikova z, koska j, ksinantova l, huckova m, imrich r, wimmerova s, gasperikova d, shishiba y, trnovec t, sebokova e and klimes i (2007), ‘Fish from industrially polluted freshwater as the main source of organochlorinated pollutants and increased frequency of thyroid disorders and dysglycemia’, Chemosphere, 67 (9), S379–385. larsen sb, spano m, giwercman a and bonde jp (1999), ‘Semen quality and sex hormones among organic and traditional Danish farmers. ASCLEPIOS Study Group’, Occup Environ Med, 56 (2), 139–44. li fx, byrd dm, deyhle gm, sesser de, skeels mr, katkowsky sr and lamm sh (2000), ‘Neonatal thyroid-stimulating hormone level and perchlorate in drinking water’, Teratology, 62 (6), 429–31. li h, chen q, li s, xu y, yao w and chen c (1999), ‘Studies on male reproductive toxicity caused by hexavalent chromium [English abstract]’, Zhonghua Yu Fang Yi Xue Za Zhi, 33 (6), 351–3. longnecker mp, gladen bc, patterson dg, jr and rogan wj (2000), ‘Polychlorinated biphenyl (PCB) exposure in relation to thyroid hormone levels in neonates’, Epidemiology, 11 (3), 249–54.

Human epidemiologic studies

55

lottrup g, andersson am, leffers h, mortensen gk, toppari j, skakkebaek ne and main km (2006), ‘Possible impact of phthalates on infant reproductive health’, Int J Androl, 29 (1), 172–80; discussion 181–5. mabeck lm, jensen ms, toft g, thulstrup m, andersson m, jensen tk, giwercman a, olsen j and bonde jp (2005), ‘Fecundability according to male serum inhibin B – a prospective study among first pregnancy planners’, Hum Reprod, 20 (10), 2909–15. mahmoud a, kiss p, vanhoorne m, de bacquer d and comhaire f (2005), ‘Is inhibin B involved in the toxic effect of lead on male reproduction?’ Int J Androl, 28 (3), 150–5. main km, toppari j and skakkebaek ne (2006a), ‘Gonadal development and reproductive hormones in infant boys’, Eur J Endocrinol, 155 (suppl 1), S51–S57. main km, mortensen gk, kaleva mm, boisen ka, damgaard in, chellakooty m, schmidt im, suomi am, virtanen he, petersen dv, andersson am, toppari j and skakkebaek ne (2006b), ‘Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age’, Environ Health Perspect, 114 (2), 270–6. martin sa, jr., harlow sd, sowers mf, longnecker mp, garabrant d, shore dl and sandler dp (2002), ‘DDT metabolite and androgens in African-American farmers’, Epidemiology, 13 (4), 454–8. mcgregor aj and mason hj (1990), ‘Chronic occupational lead exposure and testicular endocrine function’, Hum Exp Toxicol, 9 (6), 371–6. mcgregor aj and mason hj (1991), ‘Occupational mercury vapour exposure and testicular, pituitary and thyroid endocrine function’, Hum Exp Toxicol, 10 (3), 199–203. meeker jd, ryan l, barr db and hauser r (2006a), ‘Exposure to nonpersistent insecticides and male reproductive hormones’, Epidemiology, 17 (1), 61–8. meeker jd, barr db and hauser r (2006b), ‘Thyroid hormones in relation to urinary metabolites of non-persistent insecticides in men of reproductive age’, Reprod Toxicol, 22 (3), 437–42. meeker jd, godfrey-bailey l and hauser r (2007a), ‘Relationships between serum hormone levels and semen quality among men from an infertility clinic’, J Androl, 28 (3), 397–406. meeker jd, altshul l and hauser r (2007b), ‘Serum PCBs, p,p′-DDE and HCB predict thyroid hormone levels in men’, Environ Res, 104 (2), 296–304. meeker jd, calafat am and hauser r (2007c), ‘Di-(2-ethylhexyl) phthalate metabolites may alter thyroid hormone levels in men’, Environ Health Perspect, 115 (7), 1029–34. minh tb, watanabe m, kajiwara n, iwata h, takahashi s, subramanian a, tanabe s, watanabe s, yamada t and hata j (2006). ‘Human blood monitoring program in Japan: contamination and bioaccumulation of persistent organochlorines in Japanese residents’, Arch Environ Contam Toxicol, 51 (2), 296–313. nagata c, nagao y, shibuya c, kashiki y and shimizu h (2005), ‘Urinary cadmium and serum levels of estrogens and androgens in postmenopausal Japanese women’, Cancer Epidemiol Biomarkers Prev, 14 (3), 705–8. ng tp, goh hh, ng yl, ong hy, ong cn, chia ks, chia se and jeyaratnam j (1991), ‘Male endocrine functions in workers with moderate exposure to lead’, Br J Ind Med, 48 (7), 485–91. nicolopoulou-stamati p and pitsos ma (2001), ‘The impact of endocrine disrupters on the female reproductive system’, Hum Reprod Update, 7 (3), 323–30. osius n, karmaus w, kruse h and witten j (1999), ‘Exposure to polychlorinated biphenyls and levels of thyroid hormones in children’, Environ Health Perspect, 107 (10), 843–9. padungtod c, lasley bl, christiani dc, ryan lm and xu x (1998),‘Reproductive hormone profile among pesticide factory workers’, J Occup Environ Med, 40 (12), 1038–47. pan g, hanaoka t, yoshimura m, zhang s, wang p, tsukino h, inoue k, nakazawa h, tsugane s and takahashi k (2006), ‘Decreased serum free testosterone in

56

Endocrine-disrupting chemicals in food

workers exposed to high levels of di-n-butyl phthalate (DBP) and di-2-ethylhexyl phthalate (DEHP): a cross-sectional study in China’, Environ Health Perspect, 114 (11), 1643–8. pelletier c, doucet e, imbeault p and tremblay a (2002), ‘Associations between weight loss-induced changes in plasma organochlorine concentrations, serum T(3) concentration, and resting metabolic rate’, Toxicol Sci, 67 (1), 46–51. perry mj, ouyang f, korrick sa, venners sa, chen c, xu x, lasley bl and wang x (2006), ‘A prospective study of serum DDT and progesterone and estrogen levels across the menstrual cycle in nulliparous women of reproductive age’, Am J Epidemiol, 164 (11), 1056–64. persky v, turyk m, anderson ha, hanrahan lp, falk c, steenport dn, chatterton r, jr and freels s (2001), ‘The effects of PCB exposure and fish consumption on endogenous hormones’, Environ Health Perspect, 109 (12), 1275–83. pflieger-bruss s, schuppe hc and schill wb (2004), ‘The male reproductive system and its susceptibility to endocrine disrupting chemicals’, Andrologia, 36 (6), 337–45. pocar p, brevini ta, fischer b and gandolfi f (2003), ‘The impact of endocrine disruptors on oocyte competence’, Reproduction, 125 (3), 313–25. rawlings nc, cook sj and waldbillig d (1998), ‘Effects of the pesticides carbofuran, chlorpyrifos, dimethoate, lindane, triallate, trifluralin, 2,4-D, and pentachlorophenol on the metabolic endocrine and reproductive endocrine system in ewes’, J Toxicol Environ Health A, 54 (1), 21–36. recio r, ocampo-gomez g, moran-martinez j, borja-aburto v, lopez-cervante m, uribe m, torres-sanchez l and cebrian me (2005), ‘Pesticide exposure alters follicle-stimulating hormone levels in Mexican agricultural workers’, Environ Health Perspect, 113 (9), 1160–3. richthoff j, rylander l, jonsson ba, akesson h, hagmar l, nilsson-ehle p, stridsberg m and giwercman a (2003), ‘Serum levels of 2,2′,4,4′,5,5′-hexachlorobiphenyl (CB-153) in relation to markers of reproductive function in young males from the general Swedish population’, Environ Health Perspect, 111 (4), 409–13. rignell-hydbom a, rylander l, giwercman a, jonsson ba, nilsson-ehle p and hagmar l (2004), ‘Exposure to CB-153 and p,p′-DDE and male reproductive function’, Hum Reprod, 19 (9), 2066–75. sala m, sunyer j, herrero c, to-figueras j and grimalt j (2001), ‘Association between serum concentrations of hexachlorobenzene and polychlorobiphenyls with thyroid hormone and liver enzymes in a sample of the general population’, Occup Environ Med, 58 (3), 172–7. schell lm, gallo mv, decaprio ap, hubicki l, denham m and ravenscroft j (2004), ‘Thyroid function in relation to burden of PCBs, p,p′-DDE, HCB, mirex and lead among Akwesasne Mohawk youth: a preliminary study’, Environ Toxicol Pharmacol, 18 (2), 91–9. singh b, chandran v, bandhu hk, mittal br, bhattacharya a, jindal sk and varma s (2000), ‘Impact of lead exposure on pituitary–thyroid axis in humans’, Biometals, 13 (2), 187–92. steenland k, cedillo l, tucker j, hines c, sorensen k, deddens j and cruz v (1997), ‘Thyroid hormones and cytogenetic outcomes in backpack sprayers using ethylenebis(dithiocarbamate) (EBDC) fungicides in Mexico’, Environ Health Perspect, 105 (10), 1126–30. straube e, straube w, kruger e, bradatsch m, jacob-meisel m and rose hj (1999), ‘Disruption of male sex hormones with regard to pesticides: pathophysiological and regulatory aspects’, Toxicol Lett, 107 (1–3), 225–31. swan sh, main km, liu f, stewart sl, kruse rl, calafat am, mao cs, redmon jb, ternand cl, sullivan s and teague jl (2005), ‘Decrease in anogenital distance among male infants with prenatal phthalate exposure’, Environ Health Perspect, 113 (8), 1056–61.

Human epidemiologic studies

57

takeuchi t and tsutsumi o (2002), ‘Serum bisphenol a concentrations showed gender differences, possibly linked to androgen levels’, Biochem Biophys Res Commun, 291 (1), 76–8. takeuchi t, tsutsumi o, ikezuki y, takai y and taketani y (2004), ‘Positive relationship between androgen and the endocrine disruptor, bisphenol A, in normal women and women with ovarian dysfunction’, Endocr J, 51 (2), 165–9. takser l, mergler d, baldwin m, de grosbois s, smargiassi a and lafond j (2005), ‘Thyroid hormones in pregnancy in relation to environmental exposure to organochlorine compounds and mercury’, Environ Health Perspect, 113 (8), 1039–45. tanabe s and kunisue t (2007), ‘Persistent organic pollutants in human breast milk from Asian countries’, Environ Pollut, 146 (2), 400–13. teilmann g, juul a, skakkebaek ne and toppari j (2002), ‘Putative effects of endocrine disrupters on pubertal development in the human’, Best Pract Res Clin Endocrinol Metab, 16 (1), 105–21. telisman s, cvitkovic p, jurasovic j, pizent a, gavella m and rocic b (2000), ‘Semen quality and reproductive endocrine function in relation to biomarkers of lead, cadmium, zinc, and copper in men’, Environ Health Perspect, 108 (1), 45–53. thomas go, wilkinson m, hodson s and jones kc (2006). ‘Organohalogen chemicals in human blood from the United Kingdom’, Environ Pollut, 141 (1), 30–41. tilson ha (1998), ‘Developmental neurotoxicology of endocrine disruptors and pesticides: identification of information gaps and research needs’, Environ Health Perspect, 106 Suppl 3, 807–11. toft g, hagmar l, giwercman a and bonde jp (2004), ‘Epidemiological evidence on reproductive effects of persistent organochlorines in humans’, Reprod Toxicol, 19 (1), 5–26. toft g, flyvbjerg a and bonde jp (2006), ‘Thyroid function in Danish greenhouse workers’, Environ Health, 5, 32. toppari j and skakkebaek ne (1998), ‘Sexual differentiation and environmental endocrine disrupters’, Baillieres Clin Endocrinol Metab, 12 (1), 143–56. turyk me, anderson ha, freels s, chatterton r, jr, needham ll, patterson dg, jr, steenport dn, knobeloch l, imm p and persky vw (2006), ‘Associations of organochlorines with endogenous hormones in male Great Lakes fish consumers and nonconsumers’, Environ Res, 102 (3), 299–307. uhler ml, zinaman mj, brown cc and clegg ed (2003), ‘Relationship between sperm characteristics and hormonal parameters in normal couples’, Fertil Steril, 79 Suppl 3, 1535–42. wang sl, su ph, jong sb, guo yl, chou wl and papke o (2005), ‘In utero exposure to dioxins and polychlorinated biphenyls and its relations to thyroid function and growth hormone in newborns’, Environ Health Perspect, 113 (11), 1645–50. windham gc, lee d, mitchell p, anderson m, petreas m and lasley b (2005), ‘Exposure to organochlorine compounds and effects on ovarian function’, Epidemiology, 16 (2), 182–90. yucra s, rubio j, gasco m, gonzales c, steenland k and gonzales gf (2006), ‘Semen quality and reproductive sex hormone levels in Peruvian pesticide sprayers’, Int J Occup Environ Health, 12 (4), 355–61. zaidi ss, bhatnagar vk, gandhi sj, shah mp, kulkarni pk and saiyed hn (2000), ‘Assessment of thyroid function in pesticide formulators’, Hum Exp Toxicol, 19 (9), 497–501. zeng x, lin t, zhou y and kong q (2002), ‘Alterations of serum hormone levels in male workers occupationally exposed to cadmium’, J Toxicol Environ Health A, 65 (7), 513–21. zeng x, jin t, buchet jp, jiang x, kong q, ye t, bernard a and nordberg gf (2004), ‘Impact of cadmium exposure on male sex hormones: a population-based study in China’, Environ Res, 96 (3), 338–44.

3 Epidemiological evidence on impaired reproductive function and cancer related to endocrine-disrupting chemicals G. Toft, Aarhus University Hospital, Denmark; J. P. Bonde, Copenhagen University Hospital, Denmark

Abstract: This present chapter set out to give an overview of the epidemiological evidence on impaired reproductive health and cancer related to endocrinedisrupting chemicals in food, with special focus on persistent organochlorine compounds and phthalates. The outcomes evaluated includes: reproductive abnormalities at birth, semen quality, menstrual cycle disturbances, endometriosis, fibroids, time to pregnancy and cancer studies. We conclude that there are suggestive epidemiological data indicating adverse reproductive effects following exposure to endocrine disrupting compounds in food but the evidence is still limited. Key words: polychlorinated biphenyls (PCB), phthalate, cryptorchidism, semen quality, time to pregnancy.

3.1

Introduction

In order to evaluate whether endocrine-disrupting compounds from dietary exposure are affecting human reproductive health, epidemiological studies are needed to assess the direct influence on human populations. However, when performing epidemiological studies several pitfalls need to be carefully evaluated to ensure that the results are presenting unbiased estimates of the effects (Smith, 2001; Smith and Ebrahim, 2002). Humans are simultaneously exposed to complex mixtures of endocrine-disrupting compounds. This complicates the evaluation of single compounds, which may be affected by interaction with other compounds. On the other hand, the enormous complexity of the interaction issue that will never be completely resolved in the laboratory emphasizes the need to perform studies on humans in the real environment. In addition, variation in susceptibility between geno-

Epidemiological evidence on impaired reproductive function

59

types may further complicate the prediction of effects after exposure to endocrine-disrupting compounds – but when gene–environment interaction is demonstrated this may greatly help making causal inferences (Susser, 1998). From animal studies, it is known that certain narrow time windows in the fetal and perinatal period are the most sensitive to reproductive disturbances of endocrine disruptors – but how can we relate fetal exposure to adult reproductive health in humans some 20–30 years later? This question has been approached by studying direct intake of hormones during pregnancy or conditions with deviant hormonal levels (twin pregnancies, preeclampsia, nausea) as surrogates for environmental endocrine disruption and from studies of migration. Only very few studies are available that explicitly address the relation between actual measurable xenobiotic exposure in the fetal period and neonatal outcomes of importance for fertility later in life. More often exposures in the adult period have been related to male and female fertility in cross-sectional studies. In the present review we set out to give an overview of the epidemiological evidence on impaired reproductive health and cancer related to endocrine-disrupting chemicals in food.

3.2

Methods

3.2.1 Exposure selection The exposures to be included in the present review include a selection of chemical compounds from the large number of compounds present in small amounts in food. The compounds were selected to include the compounds with known endocrine-disrupting effects found in the highest concentration in food. Furthermore, known persistence in the environment and bioaccumulation in humans was a criteria to select compounds that are found in high concentrations in human blood or tissue samples. Finally, epidemiological studies are available based on only a limited number of compounds, reducing the number of compounds we can include in the present study. We ended up with three major groups of exposures: 1. Persistent organochlorine compounds. The most abundant persistent organochlorine pollutants in human serum samples are polychlorinated biphenyls (PCB), hexachlorobenzene (HCB), dichlorodiphenyltrichloroethane (DDT) and its major degradation product dichlorodiphenyldichloroethene (p,p′-DDE). These compounds have been used in electrical equipments and as pesticides since the 1940s. The compounds are highly persistent in human tissues (half-life 5–10 years), and although the compounds have not been produced and used in Western countries since the 1970s they are still among the compounds detected in the highest concentration in dietary products and in humans from all over

60

Endocrine-disrupting chemicals in food

the world. The compounds are especially found in high concentrations in animals in the top of the marine food chain. Polychlorinated dibenzodioxins (PCDDs) commonly named ‘dioxins’ are also persistent organochlorines. Although these compounds have not been intentionally produced in large volumes they are contaminants of pesticides and produced in small amounts in combustion processes. In addition, unintentional food contamination accidents in Taiwan and Japan adds to the knowledge on the potential adverse effects of dioxins on humans (Aoki, 2001). 2. Phthalates are another group of ubiquitous contaminant in food. Phthalates are used as a softener in plasticware, and in cosmetics. The human contamination with phthalates mainly originates from food with diary products, fat, oil and grain as the main sources, whereas the remaining comes from ingested dust, dermal contact and inhalation. The degradation of phthalates in the body is fast compared with the above-described persistent organohalogen compounds, but owing to continuous exposure, the compounds can be detected in more than 95% of adults in the US and Europe (Koch et al., 2006; Silva et al., 2006). The estimated daily intake of di(z-ethylhexyl) phthalate (DEHP) based on excreted urinary phthalate metabolites was evaluated in a German study and found to be on average below the EU tolerable daily intake (TDI) and US Environmental Protection Agency (EPA) reference dose (RfD) among the 85 adults and 254 children investigated, but about 10% of the population exceeded the RfD. Up to 20-fold overstepping of the RfD in the group of highest exposed children gives reason for concern (Koch et al., 2006). 3. We also intended to include studies of bisphenol A, which is used as a primary monomer in polycarbonate plastic and epoxy resins, in the present review, but no epidemiological studies on the potential adverse human reproductive health effects of bisphenol A were available. Table 3.1 presents the structure formula and summarises the (anti)estrogenic and (anti)androgenic effects demonstrated of the included chemicals. In addition the natural sex hormones estradiol and dihydrotestosterone are presented in the figure for compairson of structural similarities.

3.2.2 Outcome selection The outcomes of interest we selected were: hypospadias, cryptorchidism, anogenital distance, semen quality, menstrual cycle, endometriosis, fibroids, time to pregnancy or cancer of the testis, prostate, uterus, ovaries or cervix. Several studies on breast cancer have been performed and recent reviews indicates that adult exposure to PCBs and DDE are not consistently associ-

Epidemiological evidence on impaired reproductive function

61

Table 3.1 Structural formula of the most potent natural estrogen and androgen in humans and classes of selected endocrine-disrupting compounds Name

Structure formula CH3 OH

17β-estradiol

Estrogenica

Androgenica

Reference

+

+

Bolger et al. (1998), Yeh et al. (1999)

+

+

Bolger et al. (1998)

−

0

Bonefeld-Jorgensen et al. (2001)

+

−

Bolger et al. (1998), Sohoni and Sumpter (1998)

0

−

Gray et al. (2006)

−

−

Hombach-Klonisch et al. (2006), Mably et al. (1992)

0

−

Gray et al. (2006)

+

−

Harris et al. (1997), Gray et al. (2006)

HO CH3 OH

Dihydrotestosterone

CH3

O

PCB-153

Cl

Cl Cl

Cl Cl

Cl Cl

p,p′-DDT Cl H

CCl3

Cl

p,p′-DDE Cl

Cl Cl

TCDD

DEHP

Cl

O

Cl

Cl

O

Cl

O O O O

DBP

O O O O

a

Indication of whether the selected compounds have endocrine activity assessed in in vitro assays: + = agonistic effect; − = antagonistic effect; 0 = no (anti)estrogenic or (anti)androgenic effects.

ated with breast cancer, although specific genotypes may have increased vulnerability (Negri et al., 2003; Lopez-Cervantes et al., 2004). Furthermore breast cancer usually appears after the reproductive period and is not directly affecting fertility and this outcome is therefore not included in the present review. We ended using the following search string in Pub Med: (PCB or DDE or DDT or HCB or dioxin or phthalate or bisphenol A) and (hypospadias or cryptorchidism or semen or menstrual or endometriosis or fibroids or pregnancy or cancer) in English, last 5 years (11 June 2007), and limited to human studies. This gave a total of 759 hits (113 reviews). All of the studies were evaluated for presence of the selected exposures and outcomes.

62

Endocrine-disrupting chemicals in food

Several studies included only an estimation of the exposure level, based on consumption of contaminated compounds or presence in contaminated areas without actual measurement of exposure. These studies were excluded from the present review, to avoid exposure misclassification. Information on studies dating more than five years back in time were mainly obtained from reviews and are only briefly summarized in this chapter. A total of 35 studies met the criteria for inclusion in the present review. The selected studies are presented in Table 3.2, including an evaluation of the magnitude of the effects in the separate studies.

3.3

Reproductive abnormalities at birth

The possible effects on chryptochidism and hypospadias of in utero exposure to DDT and p,p′-DDE have been studied in two American nested case-referent studies: the Child Health and Development Study, including 75 cases of cryptorchidism, 66 cases of hypospadias and 283 controls (Bhatia et al., 2005), and the Collaborative Perinatal Project including 219 cases of cryptorchidism, 199 cases of hypospadias and 552 controls (Longnecker et al., 2002). Both studies included pregnancies in the period from 1959 to 1966, and they both found slightly but not significantly increased risk of cryptorchidism and hypospadias with odds ratios in the range of 1.3 at the highest maternal DDE exposure level compared to the lowest. These observations are of particular interest because p,p′-DDE as opposed to many PCB congeners which exhibit anti-androgenic activity in in vitro experiments (Kelce et al., 1995) and because p,p′-DDE and other anti-androgens produce similar abnormalities of the male reproductive development in laboratory animals (Gray et al., 2001). In a recent Danish study, breast-milk contamination with persistent pesticides was used as a marker of in utero exposure. It should be noted that the correlation of breast milk to cord blood concentration of organochlorine pesticides is much weaker than the correlation of maternal serum and cord blood concentrations (Jaraczewska et al., 2006). The study on breast milk contamination reported a higher sum of the eight most abundant persistent pesticides among the 62 mothers giving birth to cryptorchid boys compared with 68 mothers giving birth to healthy boys (Damgaard et al., 2006). In the same study population phtahalate contamination of breast milk was not associated to cryptorchidism, but some of the measured phthalate metabolites seemed to be related to the level of reproductive hormones in the offspring (increased sex hormone binding globulin (SHBG), and Luteinizing hormone (LH)/free testosterone ratio and decreased free testosterone) (Main et al., 2006). Based on the small and for most of the studies non-statistically significant effects observed in these studies it remains inconclusive whether DDE or other persistent pesticides as well as phthalate at environmental exposure levels increase the risk of hypospadias and cryptorchidism.

p,p′-DDE (serum)

p,p′-DDE (serum) DDT (serum)

27 organochlorine pesticides (breast milk)

Phthalate monoesters (breast milk)

Phthalate methabolites (maternal urine)

Nested casecontrol

Nested casecontrol

Nested casecontrol

Crosssectional

Bhatia et al. (2005)

Damgaard et al. (2006)

Main et al. (2006)

Swan et al. (2005)

Exposure

Nested casecontrol

Type of study

Anogenital distance/body weight (AGI)

Cryptorchidism

Cryptorchidism

Cryptorchidism Hypospadias Cryptorchidism Hypospadias

Cryptorchidism Hypospadias Polythelia

Outcome

134

62 cases 68 controls

219 cases 199 cases 167 cases 519 controls 75 cases 66 cases 283 controls 75 cases 66 cases 62 cases 68 controls

N

(0.5;3.5) (0.5;3.0) (0.4;2.3) (0.3;1.9)

Chance pr log unit increase ß: −0.095 (−0.165;−0.025)

Higher concentration of 16/21 measurable pesticides in cases. For sum of 8 most abundant pesticides p = 0.03 Similar level in cases and controls: p > 0.4

OR: 1.3 OR: 1.2 OR: 1.0 OR: 0.8

OR: 1.3 (0.7;2.4) OR: 1.2 (0.6;2.4) OR: 1.9 (0.9;4.0)

Results (odds ratio for high vs. low exposure, regression coefficients or p values for group comparisons).

Same study found correlations between phthalates and reproductive hormones

Blood samples collected in 1959–1967

Blood samples collected in 1959–1966

Notes

Overview of the epidemiological studies on reproductive effects of endocrine-disrupting compounds 2002–2007

Longnecker et al. (2002)

Study

Table 3.2

+

0

0 0 +

(+) 0

(+) 0 (+)

Evaluationa

Crosssectional

Multicentre crosssectional

Crosssectional

Toft et al. (2006)

Jager et al. (2006)

Type of study

Continued

Longnecker et al. (2007)

Study

Table 3.2

DDE

DDE

PCB

p,p′-DDE

Exposure

Sperm decondensation

Morphology

Concentration Motility Morphology Concentration Motility Morphology Semen quality: Concentration Motility

Anogenital distance Stretched penis length Semen quality:

Outcome

116 men

798 men

781

N

ß: −4.7 (−26.3;16.9) ß: -0.09 (-0.175;-0.007) ß: −0.006 (−0.044;0.032) ß: 0.036 (0.001;0.070)

Change pr log unit increase ß: 0.08 (−0.01;0.17) ß: -3.6 (-5.6;-1.7) ß: −0.01 (−0.07;0.05) ß: 0.09 (−0.002;0.18) ß: -2.8 (-4.8;-0.7) ß: −0.02 (−0.08;0.05)

0.020 (−0.034;0.073)

Change pr μg/g 0.029 (−0.024;0.082)

Results (odds ratio for high vs. low exposure, regression coefficients or p values for group comparisons).

(−) + 0 (−) + 0 0 + 0 +

+ specific effects on tail defect.

(−) 0

Evaluationa

The study included only fertile men.

Notes

Crosssectional

Crosssectional

Crosssectional

Crosssectional

AneckHahn et al. (2007)

Dalvie et al. (2004)

Hauser et al. (2006)

Jonsson et al. (2005)

Mono ethyl phthalate

Phthalic acid

Mono butyl phthalate

DDT

Motility (CASA)

DDE

Concentration Motility Concentration Motility (CASA)

Concentration Motility Morphology Semen quality:

Semen quality:

Semen quality: Total sperm count

Morphology

Motility (CASA)

Total count

Morphology

Semen quality: Total count

DDT

234 men

443 men

60 men

311 men

Risk of low sperm count. OR 3.3 (1.2;8.5) OR 1.8 (1.1;3.2) OR 0.8 (0.4;1.6) Difference high vs low tertile −4.1 (−27;18) -9.4 (-15;-3.7) 5.0 (−15;25) 8.8 (0.8;17)

ß: -3.7 SE 1.7

ß: −0.001 (−0.007;0.005) ß: -0.049 (-0.072; 0.027) ß: 0.0002 (−0.0006;0.0009) ß: -0.0003 (-0.0006; -0.000004) ß: -0.016 (-0.028;0.007) ß: 0.00006 (−0.0003;0.0004)

No effect of other phthalate metabolites (+ effects on LH and testis volume)

No significant effect on motility or morphology. No effect of other phthalate metabolites

High level exposure – residential indoor spraying

0 − 0 +

+ + (+)

+

0 + 0 + + 0

crosssectional

Multicentre crosssectional

Crosssectional

Zhang et al. (2006)

Cooper et al. (2005)

Ouyang et al. (2005) Windham et al. (2005)

Crosssectional

Type of study

Continued

Study

Table 3.2

DDT DDE PCB

DDT

PCB

DDE

Di-2-ethylhexyl phthalate

Di-n-butyl phthalate

Di-ethyl phthalate

Exposure

Luteral phase Cycle length

Short cycle (<21 days)

Semen quality: Concentration Rate of malformation Vitality Concentration Rate of malformation Vitality Concentration Rate of malformation Vitality Menstrual cycle: Cycle length Irregularity Cycle length Irregularity

Outcome

50

466

2324 women

52 men

N

High vs. low exposure -1.5 (-2.7;-0.3) -1.4 (-2.6;-0.2) −0.5 (−1.8;0.9)

High vs. low exposure OR 2.8 (1.1;7.1)

−0.31 P for trend 0.35 0.27 0.02 0.13

−0.25 −0.23 0.36

−0.13 −0.26 0.29

Correlations −0.25 0.19

Results (odds ratio for high vs. low exposure, regression coefficients or p values for group comparisons).

+ + 0

+

0 0 + (+)

0 0 0 0 0 0 0 + (−)

Not adjusted

Restrictions to primiparous showed significant effects on irregularity

Evaluationa

Notes

Crosssectional

Crosssectional

Casecontrol

Casecontrol

Chao et al. (2007)

Chen et al. (2005)

Louis et al. (2005)

Porpora et al. (2006) Fierens et al. (2003)

Cohort

Follow-up

Eskenazi et al. (2002)

PCDD/F pg TEQ/g fat 12 PCB markers ng/g fat

Total PCBs

Total PCBs Estrogenic PCBs Antiestrogenic PCBs

DDE

Dioxin TEQ PCB TEQ PCB 128 + 153 + 180 DDT

TCDD postmenarcheal

TCDD premenarcheal

Laparoscopic confirmed endometriosis Self-reported endometriosis

Cycle length Menstrual duration Cycle length Menstrual duration Laparoscopic confirmed endometriosis

Cycle length Menstrual duration Cycle length Menstrual duration Length of menstrual cycle < vs. > = 33 days

Conc. case (CI); control (CI) 26.2 (18.2;37.7); 25.6 (24.3;28.9) 294 (215;401); 372 (351;403)

142 women 10 endometriosis cases

High vs. low exposure OR: 5.3 (1.3;23)

High vs. low exposure OR: 1.44 (0.40;5.15) OR: 1.01 (0.29;3.57) OR: 3.30 (0.87;12.46)

40 case 40 control

32 cases 52 control

Only 10 cases

0

0

+

(+) 0 (+)

0 0 0 0

ß: −0.08 SE 0.92 ß: −0.45 SE 0.37

47 ß: 0.42 SE 0.38 ß: −0.15 SE 0.16

(+) + +

(+) (+) 0 0

P for difference 0.08 0.006 0.002

ß: −0.03 (−0.61;0.54) ß: 0.16 (−0.18;0.50)

ß: 0.93 (−0.01;1.86) ß: 0.18 (−0.15;0.51)

119

167

134

Cohort

Eskenazi et al. (2002b) Cobellis et al. (2003)

Case-control

Casecontrol

Cohort

Reddy et al. (2006)

Luisi et al. (2006)

Eskenazi et al. (2007)

Casecontrol

Casecontrol

Type of study

Continued

Heilier et al. (2005)

Study

Table 3.2

Dioxin-like PCBs

DEHP MEHP TCDD

PCB 29 Phthalates (DEHP)

DEPH MEHP

TCDD

Adenomyotic nodules Endometriosis

Uterine fibroids

Uterine fibromatosis

Laparoscopic confirmed endometriosis Laparoscopic confirmed endometriosis Laparoscopic confirmed endometriosis

Adenomyotic nodules

Endometriosis

Outcome

PCDD/F

Exposure

251 cases 956 controls

15 cases 20 control

24 control 85 cases 135 control

55 cases

19 cases 277 controls

21 controls

adenomyotic nodules

25

endometriosis

25

N

P for difference in conc 0.003 0.003 High vs. low exposure OR: 0.62 (0.44;0.89)

P for difference in conc 0.005 0.12 P for difference in conc <0.05 <0.05

high vs. low exposure OR: 2.1 (0.5;8.0)

Conc. case (CI); control (CI) 20.9 (18.1;24.0); 15.1 (13.1;18.4) 26.0 (21.9;30.8); 15.1 (13.1;18.4) 11.0 (9.1;13.3); 8.5 (6.9;10.5) 12.4 (10.3;14.9); 8.5 (6.9;10.5)

Results (odds ratio for high vs. low exposure, regression coefficients or p values for group comparisons).

Other PCBs and phthalates measured with similar results

Notes

− − −

+ +

+ (−)

(+)

+

(+)

+

+

Evaluationa

Crosssectional

Casecontrol

Casecontrol

Law et al. (2005)

Hardell et al. (2006b)

Hardell et al. (2006a) Ritchie et al. (2003)

Casecontrol

Multicentre crosssectional

Axmon et al. (2006)

PCB 153 HCB DDE PCB DDE

PCB HCB DDE

PCB DDE

DDE

PCB

Prostate cancer

Prostate cancer

Testis cancer

Fecundity ratio Greenland Sweden Poland Ukraine Greenland Sweden Poland Ukraine Fecundity OR

58 cases 99 control

45 cases 10 control

34 cases 20 control

390 women

1505 women

OR: 3.15 (1.04;9.54) OR: 2.39 (0.81;7.09) OR: 2.30 (0.77;6.85) OR: 1.67 (0.66;4.22) OR: 1.08 (0.47;2.50)

OR: 3.8 (1.4;10) OR: 4.4 (1.7;12) OR: 1.3 (0.5;5.0)

High vs. low exposure OR: 0.82 (0.60;1.10) OR: 1.09 (0.70;1.71) OR: 1.20 (0.78;1.86) OR: 0.79 (0.46;1.38) OR: 0.75 (0.54;1.05) OR: 1.02 (0.78;1.34) OR: 0.99 (0.47;2.08) OR: 1.11 (0.56;2.19) OR: 0.65 (0.36;1.18) OR: 0.65 (0.32;1.15) Non lipid adjusted values presented. Attenuated after lipid adjustment. Associations to mothers POP level presented No association to own POP level Increased OR for high PSA group. Increased OR with PCB 180 or oxychlordane exposure but decreased with dieldrin exposure

+ (+) (+) (+) 0

+ + (+)

(+) 0 0 0 0 0 0 0 (+) (+)

Casecontrol

Hardell et al. (2004)

PCB HCB DDE

Cumulative TCDD

Exposure

Prostate cancer Herbicide applications Comparison group Endometrial cancer

Outcome

76 cases 39 control

2516 men (140 persons with PC)

N

OR: 0.9 (0.4;2.3) OR: 0.8 (0.3;2.1) OR: 1.9 (0.8;4.8)

OR: 1.10 (0.69;1.75)

High vs. low exposure OR: 1.32 (0.75;2.34)

Results (odds ratio for high vs. low exposure, regression coefficients or p values for group comparisons).

Increased risk in comparison group with long service.

Notes

Significant associations are marked in bold. a Evaluated by the following criteria: + = statistical significant confirmation of adverse effects. (+) = non-significant adverse effect related to exposure (p < 0.20), OR > 1.3 or <1.7; ß dif. from 0 > one of CI low or high dif from 0. (−) = non-significant beneficial effects related to exposure (p < 0.20), OR > 1.3 or <1.7; ß dif. from 0 > one of CI low or high dif from 0. 0 = no association (as indicated above). CASA, computer-assisted semen analysis.

Follow-up

Type of study

Continued

Pavuk et al. (2006)

Study

Table 3.2

0 0 (+)

0

(+)

Evaluationa

Epidemiological evidence on impaired reproductive function

71

Another outcome indicating feminization of males is decreased anogenital distance, since females have a shorter anogenital distance than males. This outcome was in animal studies more sensitive to exposure to environmental anti-androgens than hypospadias and cryptrochidism (Gray et al., 2001). Thus, although the clinical consequences of reduced anogenital distance is probably limited, it may be a sensitive marker of endocrine disturbances in humans. So far two studies on anogenital distance have been performed in human populations. In one study in the US, 134 boys were examined and it was found that anogenital distance/body weight (but not the anogenital distance by itself) was inversely related to the concentration of four phthalate metabolites sampled from their mothers during pregnancy (Swan et al., 2005). In a large study among 781 mother–child pairs in Chiapas, Mexico, of which 29% reported living in DDT-sprayed homes, no indication of any association between p,p′-DDE exposure and anogenital distance or penile length was observed, indicating that even high exposure to p,p′-DDE seemed not to disturb these outcomes in humans (Longnecker et al., 2007).

3.4

Semen quality

Since a publication about 15 years ago indicating that human semen quality has decreased during the last 50 years (Carlsen et al., 1992) several studies have been performed to try to elucidate the possible causes of decreased semen quality, including studies of endocrine-disrupting compounds. During the last years new studies have added to the previous conflicting results on the effects of organochlorines on semen quality reviewed in Toft et al. (2004). The new studies includes a multicenter study of semen quality conducted in Greenland, Sweden, Poland and Ukraine, including measurement of PCB-153 and p,p′-DDE on a total of 798 men (Toft et al., 2006). In that study it was found that sperm cell motility decreased consistently with increasing PCB levels across the four populations, but p,p′-DDE exposure was not associated with sperm motility. However, sperm concentration and morphology was unrelated to PCB or p,p′-DDE exposure within the exposure levels that can be experienced by European and Arctic populations. In populations with present or recent use of DDT, the exposure to this compound is several fold higher. In the largest study including 311 South African men using DDT indoors to protect against malaria, the authors found that the total sperm count and sperm cell motility was negatively associated with DDT or DDE exposure (Aneck-Hahn et al., 2007). Also two smaller studies indicate that semen quality may be affected in Mexican and South African populations exposed to higher DDT levels. In the Mexican study of 116 men, sperm cell motility, morphology and sperm chromatin condensation was negatively affected in the highest exposure group (Jager et al., 2006),

72

Endocrine-disrupting chemicals in food

whereas decreased sperm concentration, but no other effects on semen quality or sexual function was observed among 60 South African working in a malaria control centre using DDT (Dalvie et al., 2004). A recent study that examined interactive effects of androgen receptor gene polymorphism on effects on semen quality of PCBs and p,p′-DDE found that the effect on sperm count of PCB and on sperm chromatin integrity of p,p’-DDE was only observed among men with short CAG repeat length (Giwercman et al., 2007). This study did not reveal the mechanism of which the different CAG repeat length modifies the effect of PCB and DDE on sperm count and chromatin integrity. However, it can be hypothesized that the different susceptibility among men with differences in the androgen receptor genotype is due to differences in the three-dimensional structure of the androgen receptor, which affects the binding of PCB and DDE to the androgen receptor. Taken together, the rather limited number of studies so far indicate consistently that sperm cell motility is dose-dependently disrupted by PCBs at environmental concentrations. Furthermore, although less consistently, DDT or one or more of its metabolites may impair different aspects of semen quality at high exposure levels as encountered in occupational settings or with indoor residential use and perhaps the most susceptible in a population may be affected at lower exposure levels. What effects, if any, are attributable to endocrine disruption or are conveyed through other mechanisms is not known. Another environmental exposure that has been related to semen quality is the phthalates. A study among 463 infertility patients at the Massachusetts General Hospital, Boston, USA, indicated that increasing urinary concentrations of monobutyl phthalate was associated with increased risk of low sperm concentration and motility, but no association with other phthalate monoesters and their oxidative metabolites was found (Hauser et al., 2006). In a Swedish study among 234 young men, another phthalate metabolite: monoethyl phthalate was associated to reduced semen motility, whereas the phthalate ester concentration was positively associated to sperm motility and testis volume (Jonsson et al., 2005). In a smaller Chinese study phthalates were measured in semen samples from 52 men, and related to semen quality (Zhang et al., 2006). An increased rate of abnormal sperm cells and decreased sperm density and semen volume was found with increasing concentrations of all the measured phthalates (diethyl phthalate, di-n-butyl phthalate and di-2-ethylhexyl phthalate). However, the results lack adjustment for potential confounders such as abstinence time and age, and only a few of the associations were statistically significant, leaving large uncertainties in the question about whether phthalate exposure might affect semen quality. In summary, the evidence at this stage is inadequate to refute or corroborate effect of phthalates on semen quality in humans. There is a complete lack of studies that link pre- and perinatal exposure levels with semen quality after puberty.

Epidemiological evidence on impaired reproductive function

3.5

73

Menstrual cycle disturbances

The potential effect of organochlorines on menstrual cycles has been assessed in several recent studies (Eskenazi et al., 2002b; Chen et al., 2005; Cooper et al., 2005; Ouyang et al., 2005; Windham et al., 2005; Chao et al., 2007). Unfortunately the same definitions of menstrual cycles disturbances were not used in these studies, so it is not possible to make combined estimates of the effect. The largest study included 2314 pregnant women participating in the Collaborative Perinatal Project, where data on menstrual cycle and a blood sample were collected in the 1960s in the US and in 1997–1999 11 PCB congeners and DDE were measured in the stored serum samples from pregnant women, and associated with pre-pregnancy menstrual characteristics (Cooper et al., 2005). The main results indicated longer menstrual cycle lengths and increased cycle irregularity (defined as selfreported irregularity or reporting of skipping of periods or difference between shortest and longest cycle greater than 7 days) with increasing exposure to PCB or DDE. However, the results did not suggest any association of PCB or DDE to bleeding duration, risk of heavy bleeding or dysmenorrhea. On the other hand, a Chinese study on 466 textile workers not occupationally exposed to DDT indicates that the risk of short cycles (<21 days) but not long cycles (>40 days) increased with increasing DDT exposure (Ouyang et al., 2005). A similar effect was found among 50 Southeast Asian immigrants residing in the US and who were mainly exposed by past exposure before immigration. They had decreased average length of the cycle with increasing DDT or DDE exposure (Windham et al., 2005). With the use of measurement of progesterone and estradiol methabolites on daily urine samples collected from these women it was indicated that the luteal phase of the cycle was particularly affected. Dioxin exposure tends to increase menstrual cycle length. In particular, women exposed to a high concentration of dioxin after the Seveso accident in the premenarchal age, experienced long-term effects on menstrual cycle length 20 years after the accident (Eskenazi et al., 2002b). However, progesterone, estradiol and ovarian morphology assessed by ultrasonography was not associated with 2,3,7,8-tetrachloro-dibenz-dioxin (TCDD) level among women from the same population (Warner et al., 2007). In a recent Taiwanese study among 119 mothers, higher concentrations of PCB were found in women with long (>33 days) cycles before pregnancy (Chao et al., 2007). In summary, in recent studies organochlorines have been associated with shorter as well as longer menstrual cycle length. Different compounds may have opposite effects. The cross-sectional nature of these studies does not allow for cause–effect relationships to be established. Also studies performed more than five years ago report somewhat conflicting results on the effects of organochlorines on menstrual cycles (Toft et al., 2004). However, in most previous studies organochlorine exposure was estimated

74

Endocrine-disrupting chemicals in food

from fish consumption, and the recent cohort studies with measurement of organochlorines are probably more accurate as regards to exposure classification.

3.6

Endometriosis and fibroids

In addition to menstrual cycle disturbances other gynaecological disorders such as endometriosis and fibroids are prevalent diseases. About 15% of women of reproductive age develop endometriosis (Porpora et al., 2006) and about 70% of women have experienced fibroids by the age of 50 years (Baird et al., 2003). The presence of these diseases can affect the fecundability of the women. Several case control studies have evaluated whether PCBs, dioxins or phthalates may be associated with these hormonedependent diseases. The general picture indicates that PCBs, dioxins and phthalates may increase the risk of endometriosis, although only nonsignificant associations were found in some of the studies (Eskenazi et al., 2002a; Cobellis et al., 2003; Fierens et al., 2003; Heilier et al., 2005; Louis et al., 2005; Porpora et al., 2006), whereas phthalates and dioxins seemed to be protective against the occurrence of fibroids (Luisi et al., 2006; Eskenazi et al., 2007). Only one study on endometriosis included more than 55 cases. In this Indian study of 85 cases including infertility patients with endometriosis confirmed by laparoscopy and 135 controls having laparoscopic sterilization without detection of endometriosis (Reddy et al., 2006). A dose–response association indicating the lowest concentration in the controls and increasing levels of both PCBs and phthalates with increasing severity of endometriosis was observed in this study. However, very high detection levels of some of the compounds cast doubt on the analytical quality, and in this, as in all other case control studies, the selection of the control group may cause bias if the control group differs from the case group in other population characteristics as, for example, dietary intake of contaminants, which may affect the exposure level. Although the presently performed studies may have several weaknesses, the indication of adverse effects of PCBs, dioxins or phthalates on endometriosis in six out of seven studies raises the concern about adverse effects and calls for large cohort studies to appropriately evaluate the potential adverse effects. The effect of phthalates on uterine fibroids has been studied only in a smaller study with 15 cases and 20 controls (Luisi et al., 2006) and thus the results indicating protective effects of phthalates can be regarded only as preliminary. However, similar protective effects of dioxins on fibroids was found in a large and well-controlled study of the Seveso cohort including 251 cases and 951 non-cases (Eskenazi et al., 2007). The authors explain the apparently protective effect of dioxins as an anti-estrogenic effect that might inhibit the formation of these estrogen sensitive fibroids.

Epidemiological evidence on impaired reproductive function

3.7

75

Time to pregnancy

The most comprehensive study of potential adverse effects of PCB and DDE on male or female time to pregnancy was performed among in total 2269 women and 1172 men from Greenland, Sweden, Poland and Ukraine. The study suggested that exposure to PCB or DDE was associated with decreased fecundability among both males and females from the high PCBexposed population of Greenland (adjusted odds ratio (OR) 0.8; confidence interval (CI) 0.6;1.1 and OR 0.8 CI: 0.5;1.2 respectively), but not in any of the other populations (Axmon et al., 2006). An American study on 390 women in the 1960s similarly suggested a weak reduction of female fecundity at the highest exposure level of PCB (OR 0.7 CI 0.4;1.2) and DDE (OR 0.7 CI 0.3;1.3) (Law et al., 2005). From both of these studies it was observed that several other lifestyle and medical factors such as use of birth control and contraceptive behaviour were likely to be of greater importance for time to pregnancy than organochlorine exposure, but a small additional effect of organochlorines on fecundity cannot be ruled out.

3.8

Cancer studies

Only a few studies have assessed the risk of male or female cancers in the reproductive organs after exposure to endocrine-disrupting compounds. One of the reasons may be that the most susceptible period is during rapid growth of the child (from fetus to prepuberty), but the onset of the cancers is often much later in life, making a long follow-up time necessary to appropriately address this problem (Birnbaum and Fenton, 2003). From studies of humans exposed to the syntetic estrogenic compound diethylstilbestrol (DES) during the fetal period, it is, however, known that this compound increases the risk of vaginal and cervical adenocarcinoma in females whereas the effects on male testis cancer is less consistent (Veurink et al., 2005). However, when Storgaard et al. (2006) reviewed the consequences on the male reproductive system of increased estrogen level in utero, by natural or artificial causes, they found that elevated risk of testis cancer seemed to be the only outcome that was likely to be affected by elevated estrogen levels in utero. In one study the exposure to organochlorine compounds seemed not to be associated with an elevated testis cancer risk among young men, but the measured organochlorine level in their mothers was related to an increased risk of testis cancer (Hardell et al., 2006b). The study was associated with several uncertainties, since the blood samples for organochlorine analysis were measured at the time when the sons were adult in the 44 case mothers and 45 controls included in the study. The transfer of organochlorines to the sons took place in utero or during lactation, some 30 years earlier and although the compounds are persistent, with a half-life in the range of 10 years, the present-day exposure may not be a very good estimate of previ-

76

Endocrine-disrupting chemicals in food

ous exposure, owing to individual differences in excretion and exposure to organochlorines after giving birth, which is not necessarily correlated to previous exposure. However, the topic certainly warrants further studies. The risk of prostate cancer in relation to organochlorine exposure has also been studied by Hardell et al. (2006a), who found an elevated risk of prostate cancer at high exposure levels of all the organochlorines studied, but the risk was only significantly elevated at high PCB-153 and transchlordane levels in this limited study including 58 cases and 20 controls. When stratifying on prostate specific antigen (PSA) level, it seemed that the elevated risk was mainly found among subjects with high PSA level. A previous pilot study by Ritchie et al. (2003) found some associations of organochlorine exposure (PCB-180 and oxychlordane) and increased risk of prostate cancer, but apparently protective effects of dieldrin and no dose–response associations, suggesting that the observed associations among the 58 cases and 99 controls included in the study may have been chance findings. Also, in a study of Vietnam war veterans, there was no clear associations between high TCDD exposure and prostate cancer risk, when comparing with a control group with service in Southeast Asia without direct exposure to contaminated herbicides (Pavuk et al., 2006). However, the veterans serving before 1969 and people serving for longer periods experienced an increased risk of prostate cancer based on a limited number of cases. Thus, the risk of organochlorine exposure on prostate cancer risks is still not fully elucidated. In women, one case control study indicated elevated risk (although not statistically significant) of endometrial cancer at high DDE exposure level, but no associations with a number of other organochlorines measured (Hardell et al., 2004). Also other cancers in the reproductive system, including uterine and ovary, may be related to exposure to endocrine-disrupting compounds based on animal studies, but so far the human epidemiological evidence is limited (Birnbaum and Fenton, 2003).

3.9

Conclusions

Although some associations between exposures and outcomes were observed in the presented studies, caution should be taken in making causal inferences since the studies are all observational and thus the exposures may be related to the outcomes due to confounding factors that the studies were not able to control for. When all the studied outcomes in Table 3.2 are evaluated together, 68% of the listed studied associations indicated no statistical significant associations. However, the studies including confirmation of adverse effects largely outnumbered the studies showing protective effects (29 vs. 4%). Also studies indicating non-significant adverse associations were much more prevalent than studies indicating beneficial effects (20 vs. 5%). The large number of

Epidemiological evidence on impaired reproductive function

77

studies reporting statistical significant associations and fewer studies reporting non-significant positive associations point to concern about publication bias (Smith, 2001). However, the majority of studies reporting nonsignificant adverse effects compared with studies reporting non-significant beneficial effects is less likely to be influenced by publication bias, since there is no reason to believe authors finding non-significant adverse effects should be markedly more likely to try to publish these results compared with non-significant beneficial effects. In conclusion, the weight of the evidence points to occurrence of adverse reproductive effects in human populations after exposure to endocrine-disrupting compounds in food, although the effects on most of the studied outcomes seems to be weak.

3.10

Future trends

From the present summary of the recent epidemiological evidence on impaired reproductive health and cancer related to endocrine-disrupting chemicals, it is clear that the effects of the present level of endocrinedisrupting compounds is probably not causing great harm to human reproductive health. However, most of the presented studies focused on adult exposure and immediate effects on reproductive outcomes, whereas exposure during the fetal period may be much more sensitive to disturbances caused by endocrine-disrupting compounds (Sharpe and Skakkebæk, 1993). Follow-up studies on mother–child cohorts are underway, but the children in most of the established cohorts are still young, limiting the possibility of studing effects on reproduction that appear as adults and cancers that usually appear at even older ages. Another factor that needs to be considered in future studies is differences in susceptibility to disturbances of the reproductive function by persons with different genotypes. A recent study indicated that sperm counts were reduced among males with short (<20) CAG repeat length of the androgen receptor and higher than median PCB exposure, whereas persons with longer CAG repeat lengths were unaffected by PCB exposure (Giwercman et al., 2007). Several other polymorphisms in genes involved in metabolism of endocrine-disrupting compounds may increase the susceptibility of individuals carrying these genes, including polymorphisms of both Phase I enzymes (cytocrome P450, N-acetyltransferase, glutathione S-transferase) and Phase II enzymes (paraoxonase) (Cummings and Kavlock, 2004). The human evidence of the modifying effects of genotypes on the effect of endocrine-disrupting compounds are, however, very limited, but studies of other exposures (e.g. tobacco or alcohol) clearly indicate that genetic differences may produce markedly different risks of getting the disease, and abstaining from studying genetic polymorphism may hide significant risks on subgroups carrying the susceptible genes.

78

3.11

Endocrine-disrupting chemicals in food

Sources of further information and advice

3.11.1 Information on large research projects mentioned above The Seveso study. Follow up of consequences on male and female health after accidental dioxin exposure. Contact person: mocarelli@uds. unimib.it The study for future families. Multicentre study of semen quality and prenatal development in the USA. More info: http://www.fcm.missouri.edu/ research-currentProjects-futureFamilies2.htm. Contact person: Shanna Swan, [email protected]. The INUENDO project, A multicentre study of couple fertility in relation to organochlorine exposure. More info: www.inuendo.dk. Contact person, Jens Peter Bonde/Gunnar Toft, [email protected]

3.11.2 Core review papers for further information on the subject The human epidemiological evidence of estrogenic exposure in utero and effects on male offspring is evaluated by Storgaard et al. (2006). They used known conditions of elevated natural estrogens to supplement the sparse studies of exposure to medications or other estrogenic chemicals in utero in humans. The consequences of accidental exposure to dioxins and PCBs after consumption of contaminated cooking oil in Japan and Taiwan is described in Aoki (2001). The epidemiological evidence of persistent organochlorines on human reproductive health was reviewed by Toft et al. (2004). Only the main results from this previous publication are summarized in this chapter. The influence of gene–environment interactions on reproduction and development has been evaluated by Cummings and Kavlock (2004).

3.12

References

aneck-hahn nh, schulenburg gw, bornman ms, farias p and de jager c (2007) Impaired semen quality associated with environmental DDT exposure in young men living in a malaria area in the Limpopo Province, South Africa. J Androl 28, 423–434. aoki y (2001) Polychlorinated biphenyls, polychloronated dibenzo-p-dioxins, and polychlorinated dibenzofurans as endocrine disrupters – what we have learned from Yusho disease. Environ Res 86, 2–11. axmon a, thulstrup am, rignell-hydbom a, pedersen hs, zvyezday v, ludwicki jk, jonsson ba, toft g, bonde jp and hagmar l (2006) Time to pregnancy as a function of male and female serum concentrations of 2,2′4,4′5,5′-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p′-DDE). Hum Reprod 21, 657–665. baird dd, dunson db, hill mc, cousins d and schectman jm (2003) High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstetrics Gynecol 188, 100–107.

Epidemiological evidence on impaired reproductive function

79

bhatia r, shiau r, petreas m, weintraub jm, farhang l and eskenazi b (2005) Organochlorine pesticides and male genital anomalies in the child health and development studies. Environ Health Perspect 113, 220–224. birnbaum ls and fenton se (2003) Cancer and developmental exposure to endocrine disruptors. Environ Health Perspect 111, 389–394. bolger r, wiese te, ervin k, nestich s and checovich w (1998) Rapid screening of environmental chemicals for estrogen receptor binding capacity. Environ Health Perspect 106, 551–557. bonefeld-jorgensen ec, andersen hr, rasmussen th and vinggaard am (2001) Effect of highly bioaccumulated polychlorinated biphenyl congeners on estrogen and androgen receptor activity. Toxicology 158, 141–153. carlsen e, giwercman a, keiding n and skakkebaek ne (1992) Evidence for decreasing quality of semen during past 50 years. BMJ 305, 609–613. chao hr, wang sl, lin ly, lee wj and papke o (2007) Placental transfer of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in Taiwanese mothers in relation to menstrual cycle characteristics. Food Chem Toxicol 45, 259–265. chen a, zhang j, zhou l, gao es, chen l, rogan wj and wolff ms (2005) DDT serum concentration and menstruation among young Chinese women. Environ Res 99, 397–402. cobellis l, latini g, de felice c, razzi s, paris i, ruggieri f, mazzeo p and petraglia f (2003) High plasma concentrations of di-(2-ethylhexyl)-phthalate in women with endometriosis. Hum Reprod 18, 1512–1515. cooper gs, klebanoff ma, promislow j, brock jw and longnecker mp (2005) Polychlorinated biphenyls and menstrual cycle characteristics. Epidemiology 16, 191–200. cummings am and kavlock rj (2004) Gene–environment interactions: a review of effects on reproduction and development. Crit Rev Toxicol 34, 461–485. dalvie ma, myers je, thompson ml, robins tg, dyer s, riebow j, molekwa j, jeebhay m, millar r and kruger p (2004) The long-term effects of DDT exposure on semen, fertility, and sexual function of malaria vector-control workers in Limpopo Province, South Africa. Environ Res 96, 1–8. damgaard in, skakkebaek ne, toppari j, virtanen he, shen h, schramm kw, petersen jh, jensen tk and main km (2006) Persistent pesticides in human breast milk and cryptorchidism. Environ Health Perspect 114, 1133–1138. eskenazi b, mocarelli p, warner m, samuels s, vercellini p, olive d, needham ll, patterson dg, jr., brambilla p, gavoni n et al. (2002a) Serum dioxin concentrations and endometriosis: a cohort study in Seveso, Italy. Environ Health Perspect 110, 629–634. eskenazi b, warner m, mocarelli p, samuels s, needham ll, patterson dg, jr., lippman s, vercellini p, gerthoux pm, brambilla p et al. (2002b) Serum dioxin concentrations and menstrual cycle characteristics. Am J Epidemiol 156, 383–392. eskenazi b, warner m, samuels s, young j, gerthoux pm, needham l, patterson d, olive d, gavoni n, vercellini p et al. (2007) Serum dioxin concentrations and risk of uterine leiomyoma in the Seveso women’s health study. Am J Epidemiol 166, 79–87. fierens s, mairesse h, heilier jf, de burbure c, focant jf, eppe g, de pauw e and bernard a (2003) Dioxin/polychlorinated biphenyl body burden, diabetes and endometriosis: findings in a population-based study in Belgium. Biomarkers 8, 529–534. giwercman a, rylander l, rignell-hydbom a, jonsson ba, pedersen hs, ludwicki jk, lesovoy v, zvyezday v, spano m, manicardi gc et al. (2007) Androgen receptor gene CAG repeat length as a modifier of the association between persistent organohalogen pollutant exposure markers and semen characteristics. Pharmacogenet Genomics 17, 391–401.

80

Endocrine-disrupting chemicals in food

gray le, ostby j, furr j, wolf cj, lambright c, parks l, veeramachaneni dn, wilson v, price m, hotchkiss a et al. (2001) Effects of environmental antiandrogens on reproductive development in experimental animals. Hum Reprod Update 7, 248–264. gray le, jr., wilson vs, stoker t, lambright c, furr j, noriega n, howdeshell k, ankley gt and guillette l (2006) Adverse effects of environmental antiandrogens and androgens on reproductive development in mammals. Int J Androl 29, 96–104. hardell l, van bavel b, lindstrom g, bjornfoth h, orgum p, carlberg m, sorensen cs and graflund m (2004) Adipose tissue concentrations of p,p′-DDE and the risk for endometrial cancer. Gynecol Oncol 95, 706–711. hardell l, andersson so, carlberg m, bohr l, van bavel b, lindstrom g, bjornfoth h and ginman c (2006a) Adipose tissue concentrations of persistent organic pollutants and the risk of prostate cancer. J Occup Environ Med 48, 700–707. hardell l, bavel b, lindstrom g, eriksson m and carlberg m (2006b) In utero exposure to persistent organic pollutants in relation to testicular cancer risk. Int J Androl 29, 228–234. harris ca, henttu p, parker mg and sumpter jp (1997) The estrogenic activity of phthalate esters in vitro. Environ Health Perspect 105, 802–811. hauser r, meeker jd, duty s, silva mj and calafat am (2006) Altered semen quality in relation to urinary concentrations of phthalate monoester and oxidative metabolites. Epidemiology 17, 682–691. heilier jf, nackers f, verougstraete v, tonglet r, lison d and donnez j (2005) Increased dioxin-like compounds in the serum of women with peritoneal endometriosis and deep endometriotic (adenomyotic) nodules. Fertil Steril 84, 305–312. hombach-klonisch s, pocar p, kauffold j and klonisch t (2006) Dioxin exerts anti-estrogenic actions in a novel dioxin-responsive telomerase-immortalized epithelial cell line of the porcine oviduct (TERT-OPEC). Toxicol Sci 90, 519–528. jager c, farias p, barraza-villarreal a, avila mh, ayotte p, dewailly e, dombrowski c, rousseau f, sanchez vd and bailey jl (2006) Reduced seminal parameters associated with environmental DDT exposure and p,p′-DDE concentrations in men in Chiapas, Mexico: a cross-sectional study. J Androl 27, 16–27. jaraczewska k, lulek j, covaci a, voorspoels s, kaluba-skotarczak a, drews k and schepens p (2006) Distribution of polychlorinated biphenyls, organochlorine pesticides and polybrominated diphenyl ethers in human umbilical cord serum, maternal serum and milk from Wielkopolska region, Poland. Sci Total Environ 372, 20–31. jonsson ba, richthoff j, rylander l, giwercman a and hagmar l (2005) Urinary phthalate metabolites and biomarkers of reproductive function in young men. Epidemiology 16, 487–493. kelce wr, stone cr, laws sc, gray le, kemppainen ja and wilson em (1995) Persistent DDT metabolite p,p′-DDE is a potent androgen receptor antagonist. Nature 375, 581–585. koch hm, preuss r and angerer j (2006) Di(2-ethylhexyl)phthalate (DEHP): human metabolism and internal exposure – an update and latest results. Int J Androl 29, 155–165. law dc, klebanoff ma, brock jw, dunson db and longnecker mp (2005) Maternal serum levels of polychlorinated biphenyls and 1,1-dichloro-2,2-bis(pchlorophenyl)ethylene (DDE) and time to pregnancy. Am J Epidemiol 162, 523–532. longnecker mp, klebanoff ma, brock jw, zhou h, gray ka, needham ll and wilcox aj (2002) Maternal serum level of 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene and risk of cryptorchidism, hypospadias, and polythelia among male offspring. Am J Epidemiol 155, 313–322.

Epidemiological evidence on impaired reproductive function

81

longnecker mp, gladen bc, cupul-uicab la, romano-riquer sp, weber jp, chapin re and hernandez-avila m (2007) In utero exposure to the antiandrogen 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene (DDE) in relation to anogenital distance in male newborns from Chiapas, Mexico. Am J Epidemiol 165, 1015–1022. lopez-cervantes m, torres-sanchez l, tobias a and lopez-carrillo l (2004) Dichlorodiphenyldichloroethane burden and breast cancer risk: a meta-analysis of the epidemiologic evidence. Environ Health Perspect 112, 207–214. louis gm, weiner jm, whitcomb bw, sperrazza r, schisterman ef, lobdell dt, crickard k, greizerstein h and kostyniak pj (2005) Environmental PCB exposure and risk of endometriosis. Hum Reprod 20, 279–285. luisi s, latini g, de felice c, sanseverino f, di pasquale d, mazzeo p and petraglia f (2006) Low serum concentrations of di-(2-ethylhexyl)phthalate in women with uterine fibromatosis. Gynecol Endocrinol 22, 92–95. mably ta, moore rw and peterson re (1992) In utero and lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. Effects on androgenic status. Toxicol Appl Pharmacol 114, 97–107. main km, mortensen gk, kaleva mm, boisen ka, damgaard in, chellakooty m, schmidt im, suomi am, virtanen he, petersen dv et al. (2006) Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age. Environ Health Perspect 114, 270–276. negri e, bosetti c, fattore e and la vecchia c (2003) Environmental exposure to polychlorinated biphenyls (PCBs) and breast cancer: a systematic review of the epidemiological evidence. Eur J Cancer Prev 12, 509–516. ouyang f, perry mj, venners sa, chen c, wang b, yang f, fang z, zang t, wang l, xu x et al. (2005) Serum DDT, age at menarche, and abnormal menstrual cycle length. Occup Environ Med 62, 878–884. pavuk m, michalek je and ketchum ns (2006) Prostate cancer in US Air Force veterans of the Vietnam war. J Expo Sci Environ Epidemiol 16, 184–190. porpora mg, ingelido am, di domenico a, ferro a, crobu m, pallante d, cardelli m, cosmi ev and de felip e (2006) Increased levels of polychlorobiphenyls in Italian women with endometriosis. Chemosphere 63, 1361–1367. reddy bs, rozati r, reddy s, kodampur s, reddy p and reddy r (2006) High plasma concentrations of polychlorinated biphenyls and phthalate esters in women with endometriosis: a prospective case control study. Fertil Steril 85, 775–779. ritchie jm, vial sl, fuortes lj, guo h, reedy ve and smith em (2003) Organochlorines and risk of prostate cancer. J Occup Environ Med 45, 692–702. sharpe rm and skakkebæk ne (1993) Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? The Lancet 341, 1392–1395. silva mj, reidy ja, preau jl, samandar e, needham ll and calafat am (2006) Measurement of eight urinary metabolites of di(2-ethylhexyl) phthalate as biomarkers for human exposure assessment. Biomarkers 11, 1–13. smith gd (2001) Reflections on the limitations to epidemiology. J Clin Epidemiol 54, 325–331. smith gd and ebrahim s (2002) Data dredging, bias, or confounding. BMJ 325, 1437–1438. sohoni p and sumpter jp (1998) Several environmental oestrogens are also antiandrogens. J Endocrinol 158, 327–339. storgaard l, bonde jp and olsen j (2006) Male reproductive disorders in humans and prenatal indicators of estrogen exposure A review of published epidemiological studies. Reprod Toxicol 21, 4–15. susser m (1998) Does risk factor epidemiology put epidemiology at risk? Peering into the future. J Epidemiol Community Health 52, 608–611.

82

Endocrine-disrupting chemicals in food

swan sh, main km, liu f, stewart sl, kruse rl, calafat am, mao cs, redmon jb, ternand cl, sullivan s et al. (2005) Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ Health Perspect 113, 1056–1061. toft g, hagmar l, giwercman a and bonde jp (2004) Epidemiological evidence on reproductive effects of persistent organochlorines in humans. Reprod Toxicol 19, 5–26. toft g, rignell-hydbom a, tyrkiel e, shvets m, giwercman a, lindh ch, pedersen hs, ludwicki jk, lesovoy v, hagmar l et al. (2006) Semen quality and exposure to persistent organochlorine pollutants. Epidemiology 17, 450–458. veurink m, koster m and berg lt (2005) The history of DES, lessons to be learned. Pharm World Sci 27, 139–143. warner m, eskenazi b, olive dl, samuels s, quick-miles s, vercellini p, gerthoux pm, needham l, patterson dg and mocarelli p (2007) Serum dioxin concentrations and quality of ovarian function in women of Seveso. Environ Health Perspect 115, 336–340. windham gc, lee d, mitchell p, anderson m, petreas m and lasley b (2005) Exposure to organochlorine compounds and effects on ovarian function. Epidemiology 16, 182–190. yeh s, chang hc, miyamoto h, takatera h, rahman m, kang hy, thin th, lin hk and chang c (1999) Differential induction of the androgen receptor transcriptional activity by selective androgen receptor coactivators. Keio J Med 48, 87–92. zhang yh, zheng lx and chen bh (2006) Phthalate exposure and human semen quality in Shanghai: a cross-sectional study. Biomed Environ Sci 19, 205–209.

4 Nutritional phytoestrogens and bone health W. Wuttke, H. Jarry and D. Seidlová-Wuttke, Georg-August-Universität Göttingen, Germany

Abstract: There is no doubt about the osteoprotective power of estradiol-17β and the hormone and estrogen therary (HT/ET) estrogens. Along the lines of animal experimental findings, a number of clinical studies indicate a weak osteoprotective effect of isoflavones. There are, however, a number of reports that did not find any significant osteoprotective effect of isoflavones and several authors point out that the positive effects of isoflavones on the bone are small and warrant further investigation. However, some slight osteoprotective effect may be exerted. Both estradiol and soy isoflavones appear to improve symptoms of osteoarthritis. Key words: soy, red clover, isoflavones, bone.

4.1

Introduction: trends in bone health

Bone modelling occurs by resorption of old bone by osteoclasts, which transmit their activity message to osteoblasts, resulting in new bone formation. Under several conditions the equilibrium between bone resorption and bone formation is disturbed, which may result in excessive bone formation (osteopetrosis) or bone resorption (osteoporosis). The major cause of osteoporosis is estrogen deficiency in the postmenopausal state. Hence, the increase in life expectancy has resulted in an increased number of osteoporotic diseases, including fractures, which are of great socio-economic concern. Postmenopausal osteoporosis is best prevented by treatment with 17βestradiol (E2) or conjugated estrogens used in preparations for hormone replacement therapy (HRT). Owing to the alarming results of the Heart and Estrogen/progestin Replacement Studies (HERS I and II; Grady et al., 1998, 2002) and the Women’s Health Initiative (WHI; Rossouw et al., 2002)

84

Endocrine-disrupting chemicals in food

in which increased risk for the development of cardiovascular diseases, such as heart attacks and strokes, and of mammary cancers were reported, physicians, patients and particularly drug companies are currently seeking for alternatives for HRT. E2 is the most active endogenous estrogen produced primarily in ovaries, to a lesser degree in testes, and locally in a number of aromatase-producing organs which utilize circulating testosterone as a precursor for aromatization and thereby for E2 production. Hence, relatively high concentrations of E2 may be produced locally in organs not only of females but also of males, including the bone (Simpson, 2002). Hence, estrogens may play an important role in maintaining the homeostasis of bones, and their effects appear to be mediated via the estrogen receptor (ER) of the α subtype (ERα), in both females and males. This notion is substantiated by data derived from different animal models: • Ovariectomy and orchidectomy of many species, including rats and mice results in severe osteoporosis which can be largely prevented by treatment with E2 or other estrogenic substances (Seidlova-Wuttke et al., 2005). • Mice with deleted ERα and men with similar natural mutations have low bone mineral density and never establish peak bone masses observed in genetically intact animals (Couse and Korach, 1999; Grundberg et al., 2007). • Likewise experimental deletion of the aromatase gene or natural mutations in men, hence the inability to aromatize androgens to estrogens, results in reduced bone size and bone mineral density (Perez et al., 2006; Grundberg et al., 2007). Similar results were reported in animals treated with aromatase inhibitors (Vanderschueren et al., 1997). From these data it is clear that estrogens are important for the maintenance of bone homeostasis not only in female experimental animals or human beings but also in males. Selective estrogen receptor modulators (SERMs) with anti-estrogenic effects in the mammary gland, no estrogenic effects in the uterus, but antiosteoporotic (i.e. estrogenic effects) in the bone have been promoted. They do, however, have anti-estrogenic effects in the hypothalamus and thereby increased psychosomatic climacteric complaints, such as hot flushes, galloping heart attacks and mood swings (Lewis and Jordan, 2005).

4.2

Methods to study the effects of endocrine-disrupting chemicals on bone health

Numerous studies have been performed utilizing either immortalized bone cancer derived cell lines or primary cells of animal or human bone origin. A number of endocrine-disrupting chemicals (EDCs) including phytoestrogens

Nutritional phytoestrogens and bone health

85

have been tested in these systems and mostly the EDCs with estrogenic actions proved to have beneficial effects in the bone. Similarly numerous animal experiments were performed and the gonadectomized rat is a most frequently used model. Ovariectomized or orchidectomized rats develop severe osteoporosis within a few months. In particular, the metaphysis of the tibia was shown to be particularly sensitive to withdrawal of sex steroids. Limited experiments have been performed with gene-manipulated animals such as ERα, ERβ or aromatase knockout mice. Studies in monkeys are scarce and limited information is also available for exposure to EDCs and their effects in bone (patho-) physiology.

4.3

Effects of endocrine-disrupting chemicals on bone health

Taken collectively it appears that primarily EDCs with estrogenic effects may have beneficial effects in the bone; in particular, EDCs that activate the ERα were shown to be effective in this respect.

CH 3

OH

H H

H HO Estradiol-17β

O

O

OH

OH

HO Daidzein

Fig. 4.1 The structure of estradiol-17β (E2) and of isoflavones (given here is the structure of daidzein), are similar. They bind to both estrogen receptor subtypes and to activate their transcriptional properties.

86

4.4

Endocrine-disrupting chemicals in food

Phytoestrogens and bone health

Another vigorously publicized alternative to synthetic SERMS are naturally occurring estrogens produced by plants, so-called phytoestrogens. Phytoestrogens belong to three main classes: • flavonoids; • lignans; • coumestans. The isoflavones are the most important phytoestrogen flavonoids. These substances are widely produced by many plants but their production is particularly high in soy (Glycine max) and red clover (Trifolium partense). Their chemical structure (Fig. 4.1) resembles the structure of E2. Isoflavones are present in these plants in a free form and as glycosylates. The major isoflavones in soy are genistein and daidzein. Both can be deglycosylated by the gut flora and in addition daidzein can be reduced to equol in the gut of experimental rats and of approximately 25–30% of human beings (Atkinson et al., 2005; Muthyala et al., 2004). This conversion is shown in Fig. 4.2. Isoflavones are primarily promoted as alternative for OH

O

Daidzein HO

O OH

O

Dihydro-daidzein

HO

O

OH

OH

Tetrahydro-daidzein HO

O OH

OH O

HO

O Equol

HO

CH3 OH

O-Desmethyl-angolensin (O-DMA)

Fig. 4.2 Bacterial degradation of daidzein may result in formation of equol and ODMA: all rats but not all human beings are able to this biological transformation.

Nutritional phytoestrogens and bone health

87

hormone replacement therapy. They are currently of major concern as it is disputed whether they have SERM activities with beneficial estrogenic effects in the bone to prevent osteoporosis and climacteric complaints but with no adverse estrogenic effects in the uterus or in the mammary gland. This question is not yet totally answered (see below) but it appears clear from the literature that some isoflavones share antiosteoporotic effects with other estrogens. Phytoestrogens have estrogenic activity because they bind to the ERs of both types (ERα and ERβ). In experiments utilizing selective ERα or ERβ agonists it was convincingly shown that the bone responds primarily to ERα agonists (Harris et al., 2002; Hillisch et al., 2004) as treatment of ovariectomized rats and ERα but ERβ agonists do not prevent osteoporosis. Hence, if isoflavones exert estrogenic anti-osteoporotic effects in bone this must be mediated via the ERα. The fact that isoflavones act on bone via the ERα gives cause for concern about potential adverse effects in other organs. As mentioned above, of concern are estrogenic effects in the uterus where they may result in endometrial hyperplasia and eventually to endometrial cancer. This is prevented by additional administration of progestins. Treatment of ovariectomized rats with genistein, daidzein or equol increased uterine weights. Since only ERα, not ERβ, agonists stimulate uterine growth (Hillisch et al., 2004), this is a clear indication that the effects of the isoflavones are ERα mediated (Fig. 4.3). An increased incidence of endometrial hyperplasia following such treatment was indeed reported recently: in a 5-year study the daily administration of 150 mg of isoflavones (40–45% genistein, 40–45% daidzein, 10–20% glycetin) to postmenopausal women resulted in endometrial hyperplasia in 3.37% of the subjects whereas no such pathological features were seen in the placebo-treated postmenopausal women (Unfer et al., 2004). Also the mammary gland is of some concern because estrogens stimulate proliferation and thereby lobulo alveolar growth. Such proliferative effects were recently shown to occur also in the mammary glands of ovariectomized rats treated with genistein, daidzein or equol (Fig. 4.4). High proliferation activity leads to a higher density of the proliferating tissue and high breast density was shown to be a risk factor for the development of breast cancer (Maskarinec et al., 2001, 2004). Higher breast densities as determined mammographically were also seen in women with a high isoflavone intake and this was interpreted as a significant risk factor for breast cancer in these women (Maskarinec et al., 2001). Despite this concern, in an interventional study, long-term treatment appeared to be devoid of adverse effects on the mammary gland (Maskarinec et al., 2004). There are a number of investigations published utilizing phytoestrogens as drugs or food additives for the treatment of osteoporosis. Numerous cell biological experiments demonstrated convincingly that isoflavones may stimulate osteoblast activity, though to a lesser degree than E2 (Setchell et al., 2003).

88

Endocrine-disrupting chemicals in food 0.6

* *

0.5

Uterine weight (g)

*p<0.05 vs ovx, sf 0.4

0.3 *

0.2

* 0.1

x+ E2 ov hi x+ gh ge ni st ei ov n x+ lo w ge ni st ei n ov hi x+ gh da id ze ov in x+ lo w da id ze in hi gh ov x+ eq uo ll ow ov x+ eq uo lh ig h

lo w

ov

ov x+ E2

ov

xsf

0.0

Fig. 4.3 Effects of E2, genistein, daidzein and equol on uterine weights. Note the clear estrogenic effects of the higher dose = 15.5 mg/rat/day (lower dose = 1.9 mg/ rat/day and the clear estrogenic, i.e. stimulatory, effects of all the isoflavones on uterine weight. Doses of the isoflavones were: genistein low: 1.9 mg/animal/day, high: 15.5 mg/animal/day; daidzein low: 4.5 mg/animal/day, high: 15.5 mg/animal/day; equol low: 0.9 mg/animal/day, high: 15.8 mg/animal/day; estradiol 17β low: 0.05 mg/ animal/day, high: 0.25 mg/animal/day. ovx = ovarietomized; sf = soy-free.

In animal experiments highly controversial data have been published concerning beneficial effects of phytoestrogens on bone. This controversial literature has been reviewed recently by Reinwald and Weaver (2006). The ovariectomized rat is a widely used model is a recommended by the US Food and Drug Administration (FDA) to study the effects of drug interventions on bone. Our own results from a recently completed 3-month study with ovariectomized rats indicated a weak osteoprotective effect of some of the isoflavones tested. Genistein was effective in this respect, but less so than daidzein (Fig. 4.5). In our study ovariectomized rats received the phytoestrogens with their daily food, and bone mineral density and computercalculated fragility of the bones were estimated. The body weight adjusted strain strengths index (SSI), an index for bone stability, for all of the phytoestrogen-treated animals did not yield a significant improvement of bone stability in comparison with the ovariectomized control animals. In

Nutritional phytoestrogens and bone health

89

60 Amount of PCNA positive cells/ 500 cells

*

*p<0.05 vs ovx, sf

50 * 40

*

30

*

20

10

hi ov gh x+ ge ni st ei ov n x+ lo w ge ni st ei n ov hi x+ gh da id ze in ov lo x+ w da id ze in hi gh ov x+ eq uo ll ow ov x+ eq uo lh ig h

ov x+ E2

ov

x+

E2

lo w

ov xsf

0

Fig. 4.4 Effects of E2, genistein, daidzein and equol on PCNA positivity in mamma structure. The protein expression of the proliferating cell nuclear antigen (PCNA) was largely stimulated by E2 to a lesser degree, though statistically significant, also by genistein, daidzein and equol.

each case bone mineral density of the metaphysis of the tibia as well as bodyweight adjusted bone stability were significantly improved by treatment with 17β-estradiol. There is evidence that daidzein, a major component of soy extracts, can be metabolized by the gut flora of about 25–30% of human beings to Equol and O-desmethyl angolensin (O-DMA) (Frankenfeld et al., 2005) (Fig. 4.2). It was suggested that only those postmenopausal women who are able to metabolize daidzein to O-DMA may benefit from a high soy isoflavone intake. In a later study it was shown that soy intake exerted only bone protective effects in postmenopausal O-DMA – but not equol–producers (Frankenfeld et al., 2006). In the later study total leg and head bone mineral density were higher in O-DMA producers in comparison to O-DMA non-producers. Surprisingly spinal bone mineral density was 20% lower in the equol producers than in the non-producers. Other experiments performed in rats indicated that Equol does have osteoprotective effects (Mathey et al., 2007) as administration of 110 mg Equol/kg body weight to ovariectomized rats largely prevented osteoporosis. These observations

90

Endocrine-disrupting chemicals in food 250 *

* *p<0.05 vs ovx, sf

Cencellous density (mg/cm3)

200 * * *

150

*

*

100

50

ig lh uo

uo ov

x+

eq

eq x+ ov

da ov

x+

h

ow ll

hi in ze id

id da x+ ov

gh

w lo ze

ei st ni

ge x+ ov

in

hi n

n ei st ni ge

x+ ov

gh

w lo

gh hi E2 x+

ov

ov

x+

E2

ov

x-

lo

sf

w

0

Fig. 4.5 Effects of estradiol 17β, genistein, daidzein and equol on bone mineral density (BMD) of the metaphysis of the tibia. Note the clear bone-sparing effects of estradiol-17β and the higher dose of daidzein and equol whereas genistein, at a dose which stimulated uterine weights (see Fig. 4.2), was ineffective to partially prevent osteoporosis. The bodyweight-related strain strength index (SSI/100 g bodyweight) was highest in the E2 treated animals and no significant differences compared with the ovariectomized control rats were seen in the other animals.

are in contrast to our animal experimental data in which we showed that ovariectomized rats had only a minor benefit from high equol intake (Fig. 4.4). Similarly it was shown (Devareddy et al., 2006) that administration of isoflavone-enriched soy proteins to ovariectomized rats did not restore bone loss that had a reduced bone density, i.e. that were osteopenic. However, tibial microstructure properties were improved. In an attempt to modulate the isoflavone bioavailability the intestinal microflora of ovariectomized rats was modified by short-term administration of fructooligosaccarides or live Lactobacillus casei (Mathey et al., 2007). The effect of daidzein was improved by the treatment with intestinal microflora modulators. Both treatments were suggested to increase the intestinal metabolism of daidzein to equol. In this interesting approach genistein, daidzein and equol – when given separately – had a bone protective effect. Puerarin, a glycosylated daidzein from Pueraria lobata (Kudzu), also appears to have low activity

Nutritional phytoestrogens and bone health

91

against osteoporosis. It was therefore not surprising to observe small antiosteoporotic effects similar to daidzein or equol. The closest animal model to humans to study diseases such as osteoporosis is undoubtedly the monkey. A study has been performed in ovariectomized monkeys (Macaca fascicularis) in which treatment with soy phytoestrogens had no osteoprotective effect (Register et al., 2003). The monkeys received either a soy protein isolated which provided the daily dose of 35–40 mg of isoflavones (~66% genistein, 33% daidzein and 1% glycetein), conjugated equine estrogens (~0.2 mg/day) or alcohol extracted soy over a period of 3 years. Lumbar spine bone mineral content (BMC) and bone mineral density (BMD) as well as whole body mineral content of the whole body yielded no significant difference between the controls and the isoflavone-fed animals. Both groups lost BMD and BMC over the investigation period while the conjugated equine estrogens had a significant bone protective effect.

4.4.1 Studies in humans The following is a review of studies in postmenopausal women. The results of interventional, double-blind studies are summarized in Table 4.1. It is difficult to summarize the results of the interventional studies because they are contradictory – it is not possible to explain this at present. In addition to the interventional studies a number of cohort studies also yielded conflicting results. Previous studies showed that soy extract or isoflavone intake and serum isoflavone levels did not correlate with bone mineral densities; the authors concluded therefore that their data do not support the association of BMD with soy or isoflavone intake or serum isoflavone levels (Nagata et al., 2002). In another study the opposite was found (Somekawa et al., 2001). In this study it was concluded that an estimated daily dose of 54 mg/70 kg of isoflavones by postmenopausal with normal body mass index was associated with increased bone mass in postmenopausal women and might therefore be useful in preventing hypo-estrogenic effects on the bone (Somekawa et al., 2001). In another large cohort study more than 20 000 Chinese postmenopausal women in China were evaluated for their average daily intake of soy protein or isoflavones and a significant reduction of fracture risks was related to high soy/isoflavone intake (Zhang et al., 2005). Overall, the available evidence from animal studies and clinical trials suggests that soy, red clover or isoflavones may have some minor boneprotecting effects in postmenopausal women. It is, however, questionable whether these positive effects justify the use of isoflavone-containing preparations because of the ERα-mediated risks on the uterus and mammary gland. In addition the isoflavone-containing preparations are food additives and therefore out of any medical regulatory control which would assess such risks as part of the approvals process. Many patients wrongly believe

Lumbar spine BMC and BMD significantly higher in ISP90 group

Postmenopausal women (n = 66) equally distributed over 3 groups, 6 months

Perimenopausal women (n = 69) with folliclestimulating hormone (FSH) >30 IU/L 17 postmenopausal women

Two doses of isoflavones (ISP56 and ISP90 meaning not specified vs casein/non fat dry milk in various food items

Soy protein isolate containing isflavone

Soy isolates containing 0.13, 1.00 or 2.01 isoflavones over 3 months

Potter et al. (1998)

Alekel et al. (2000)

Wangen et al. (2000)

No significant effects on serum surrogate parameters of osteoblast or osteoclast activity

Minor but significant loss of BMD and BMC in lumbar spine in low isoflavone group but not in isoflavone high group, intermediate loss in isoflavone low group

Gain of vertebral BMD (after 1 year significant) in ipriflavone group Significant loss of BMD in placebo group after 2 years. Significantly lower serum osteocalcin and urinary hydroxyproline

Postmenopausal women (n = 198) with t-score 1 standard deviations below age matched normal population, 2 years

Synthetic isoflavone: ipriflavone 3 × 200 mg/day vs. placebo, both groups received additional Ca2+ (1 g/day)

Agnusdei et al. (1997)

Main outcome

Test preparations

Number of evaluated test subjects and test duration

ISP90 group had significantly lower BMC and BMD at beginning of study in comparison to ISP56 and control group

Special remarks

-

+

+

+

Outcome in short

Double-blind placebo-controlled clinical studies with soy/red clover/isoflavone preparations and the postmenopausal bone

Reference

Table 4.1

Postmenopausal Brazilian women (n = 40) of Japanese origin, 10 weeks Postmenopausal women (n = 203), distributed evenly over 3 groups

Roasted soybeans containing 37.3 mg isoflavones or sesame diet

Soy germ extract containing 40 or 80 mg isoflavones vs corn starch, both supplemented with Vit.D and Ca2+

Soy protein with 52 or 96 mg isoflavones/day vs soy without isoflavones Double blind

Yamori et al. (2002)

Chen et al. (2003)

Gallagher et al. (2004)

Postmenopausal women (n = 65) equally distributed over 3 groups 15 months

Postmenopausal women (n = 90) equally distributed over 3 groups, 1 year

Pure genistein (54 mg/ day) vs 17ß-estradiolnorethisterone acetate (1 mg and 0.5 mg respectively HRT) vs placebo

Morabito et al. (2002)

No significant differences in spine or femoral BMD. Significantly higher BMD in trochanter at 9 and 15 months in placebo group

No significant effects on BMD or BMC of whole body, L1–L4 or on hip structures

Significant reduction of bone resorption markers, no change in bone stiffness as measured by ultrasound sonography

Increased BMD in femoral neck and lumbar spine in comparison to pretreated in Genistein and HRT groups Slight decrease in placebo patients. Significantly higher serum osteocalcin in genistein and lower values in HRT group. Bone resorption markers decreased in both serum groups

+

Intake of isoflavone verified by measurement in serum, soy protein without isoflavones more effective than with isoflavones

(+)

Urinary excretions of isoflavones increased in verum group

Favourable and significant effects in slim women with history of menopause >4 years and low BMC at beginning of study

+

Genistein and E2 intake verified by measurement in serum

Test preparations

Isoflavone (110 mg/day) containing capsules vs placebo capsules, double blind, crossover

Soy protein containing 99 mg isoflavones in milk protein, double blind, 1 year

Four-armed study 1. Soy milk containing 76 mg isoflavones + variety of minerals and vitamins. 2. Transdermal progesterone (540 mg/3 weeks) followed by 1 week break 3. Soy milk (isoflavone poor) + transdermal testosterone (540 mg/3 weeks) 4. Isoflavone poor soy milk, double blind

Harkness et al. (2004)

KreijkampKaspers et al. (2004)

LydekingOlsen et al. (2004)

Continued

Reference

Table 4.1

Postmenopausal women (n = 89, distributed evenly over 4 groups) not older than 75 years, two years

Postmenopausal women (n = 175, distributed evenly over 2 groups)

Postmenopausal women (n = 19) 10 patients first 6 months verum then placebo. Other 9 patients first placebo, then verum

Number of evaluated test subjects and test duration

Significant reduction of BMD in control and soy + progesterone groups but not in progesterone at soy alone groups

Isoflavones Effective to prevent demineralisation but not in presence of progesterone

Surprising reduction of osteocalcin

Slightly increased BMC of protein vs groups no effects in hip BMC or BMD significantly less reduction of osteocalcin in placebo groups Significant reduction of hydroxyprolin No difference in total hip, trochanter and lumbar spines

Special remarks

Main outcome

+

−

(+)

Outcome in short

128 postmenopausal women distributed evenly over 4 groups

Postmenopausal women (n = 13) were pretreated with radioactively labelled 41. After 200 days bone turnover was determined during a 50 day pre- and a 50 day treatment period

47 mg isoflavones ± bodily exercise vs placbo ± exercise, double blind over 24 weeks

Isoflavones: 0, 97.5 or 135.5 mg/day in baked products or beverages

Wu et al. (2006)

Cheong et al. (2007)

+ = positive effects. (+) = positive effects in bone markers. − = no effects.

Postmenopausal women (n = 87, distributed evenly over 2 groups) younger than 65 years, 1 year

Soy food containing 60 mg isoflavones vs isoflavone-free food, double blind

Arjmandi et al. (2005)

Isoflavones intake was assured by measurement of serum concentration

Isoflavones intake was verified by measurements in the serum

No effects on biochemical markers of bone turnover no on bone turnover as assessed by urinary 41 Ca/Ca ratios

Isoflavones ineffective

No significant effects on serum surrogate parameters a of osteoblast or osteoclast activity, no effect on BMD of body, lumbar spine, total hip, femoral neck or trochanter

No significant difference in whole body or lumbar BMD or BMG in serum surrogate parameters of osteoblast or osteoclast activities between the 2 treatment groups

-

-

-

96

Endocrine-disrupting chemicals in food

that ‘little helps little, more helps more’ and therefore might take high doses in the mistaken view that they will have a greater benefit – such doses are likely to have adverse effects on the uterus (Unfer et al., 2004).

4.5

Future trends

In the future it will be necessary to study the potential adverse effects of phytoestrogens on the mammary gland and uterus and weigh up these risks against benefits such as increasing bone density and effects on serum. It is the authors’ view that the estrogenic effects of soy, red clover and/or isoflavones share many of effects of estradiol-17β even though their action is manifold less potent. In a recent review (Wuttke et al., 2007) the risks and benefits of phytoestrogens have been considered; it was concluded that they share the properties of 17β-estradiol, but at significantly lower activity, but more importantly they share the adverse estrogenic effects. This, of course, means that their safe use in, for example, the treatment of osteoporosis would be very much in question. Animal experiments and studies were published in which children were investigated that moved prior to puberty or after reaching sexual maturity from Far East Asian countries to the US (Wuttke et al., 2007). These migration studies have shown that there are beneficial effects of phytoestrogens on the mammary gland of girls exposed during puberty. This peripubertal exposure appears to result in a higher degree of differentiation of the mammary gland (Lamartiniere, 2002) and tissues with greater differentiation are always less prone to malignancy. This might explain the ‘Japanese phenomenon’, i.e. Japanese women have a lower incidence of mammary cancer; however, when they migrate into countries with Western habits they often change their lifestyle and cover their protein needs with meat rather than with phytoestrogen-containing soy products. Therefore Japanese women growing up in the US have a mammary cancer incidence similar to the white Caucasian US population. It is arguable that in countries with low isoflavone intake girls should be exposed to these phytoestrogens prior to, and during, puberty which might result in a reduced mammary cancer incidence 30–40 years later.

4.6

Sources of further information and advice

The interested reader will find more detailed information in several reviews that have been published recently (Lewis et al., 2005; Reinwald and Weaver 2006; Wuttke et al., 2007) and in the internet under ‘Estrogenic effects of isoflavones’.

Nutritional phytoestrogens and bone health

4.7

97

References

agnusdei d et al. A double blind, placebo-controlled trial of ipriflavone for prevention of postmenopausal spinal bone loss. Calcif Tissue Int 1997; 61 (2): 142–7. alekel dl et al. Isoflavone-rich soy protein isolate attenuates bone loss in the lumbar spine of perimenopausal women. Am J Clin Nutr 2000; 72 (3): 844–52. arjmandi bh et al. One year soy protein supplementation has positive effects on bone formation markers but not bone density in postmenopausal women. Nutr J 2005; 4: 8. atkinson c et al. Gut bacterial metabolism of the soy isoflavone daidzein: exploring the relevance to human health. Exp Biol Med (Maywood) 2005; 230 (3): 155–70. chen x et al. Effects of genistein on expression of bone markers during MC3T3-E1 osteoblastic cell differentiation. J Nutr Biochem 2003; 14 (6): 342–9. cheong jm et al. Soy isoflavones do not affect bone resorption in postmenopausal women: a dose-response study using a novel approach with 41Ca. J Clin Endocrinol Metab 2007; 92 (2): 577–82. couse jf, korach ks. Reproductive phenotypes in the estrogen receptor-alpha knockout mouse. Ann Endocrinol (Paris) 1999; 60 (2): 143–8. devareddy l et al. Soy moderately improves microstructural properties without affecting bone mass in an ovariectomized rat model of osteoporosis. Bone 2006; 38 (5): 686–93. frankenfeld cl et al. High concordance of daidzein-metabolizing phenotypes in individuals measured 1 to 3 years apart. Br J Nutr 2005; 94 (6): 873–6. frankenfeld cl et al. Postmenopausal bone mineral density in relation to soy isoflavone-metabolizing phenotypes. Maturitas 2006; 53 (3): 315–24. gallagher jc et al. The effect of soy protein isolate on bone metabolism. Menopause 2004; 11 (3): 290–8. grady d et al. Heart and Estrogen/progestin Replacement Study (HERS): design, methods, and baseline characteristics. Control Clin Trials 1998; 19 (4): 314–35. grady d et al. Cardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study follow-up (HERS II). JAMA 2002; 288 (1): 49–57. grundberg e et al. The impact of estradiol on bone mineral density is modulated by the specific estrogen receptor-alpha cofactor retinoblastoma-interacting zinc finger protein-1 insertion/deletion polymorphism. J Clin Endocrinol Metab 2007; 92 (6): 2300–6. harkness ls et al. Decreased bone resorption with soy isoflavone supplementation in postmenopausal women. J Womens Health (Larchmt) 2004; 13 (9): 1000–7. harris ha et al. Characterization of the biological roles of the estrogen receptors, ERalpha and ERbeta, in estrogen target tissues in vivo through the use of an ERalpha-selective ligand. Endocrinology 2002; 143 (11): 4172–7. hillisch a et al. Dissecting physiological roles of estrogen receptor alpha and beta with potent selective ligands from structure-based design. Mol Endocrinol 2004; 18 (7): 1599–609. kreijkamp-kaspers s et al. Effect of soy protein containing isoflavones on cognitive function, bone mineral density, and plasma lipids in postmenopausal women: a randomized controlled trial. JAMA 2004; 292 (1): 65–74. lamartiniere ca. Timing of exposure and mammary cancer risk. J Mammary Gland Biol Neoplasia 2002; 7 (1): 67–76. lewis js, jordan vc. Selective estrogen receptor modulators (SERMs): mechanisms of anticarcinogenesis and drug resistance. Mutat Res 2005; 591 (1–2): 247–63. lydeking-olsen e et al. Soymilk or progesterone for prevention of bone loss – a 2 year randomized, placebo-controlled trial. Eur J Nutr 2004; 43 (4): 246–57.

98

Endocrine-disrupting chemicals in food

maskarinec g, meng l. An investigation of soy intake and mammographic characteristics in Hawaii. Breast Cancer Res 2001; 3 (2): 134–41. maskarinec g et al. A 2-year soy intervention in premenopausal women does not change mammographic densities. J Nutr 2004; 134 (11): 3089–94. mathey j et al. Modulation of soy isoflavones bioavailability and subsequent effects on bone health in ovariectomized rats: the case for equol. Osteoporos Int 2007; 18 (5): 671–9. morabito n et al. Effects of genistein and hormone-replacement therapy on bone loss in early postmenopausal women: a randomized double-blind placebocontrolled study. J Bone Miner Res 2002; 17 (10): 1904–12. muthyala rs et al. Equol, a natural estrogenic metabolite from soy isoflavones: convenient preparation and resolution of R- and S-equols and their differing binding and biological activity through estrogen receptors alpha and beta. Bioorg Med Chem 2004; 12 (6): 1559–67. nagata c et al. Soy product intake and serum isoflavonoid and estradiol concentrations in relation to bone mineral density in postmenopausal Japanese women. Osteoporos Int 2002; 13 (3): 200–4. perez ea et al. Effect of letrozole versus placebo on bone mineral density in women with primary breast cancer completing 5 or more years of adjuvant tamoxifen: a companion study to NCIC CTG MA.17. J Clin Oncol 2006; 24 (22): 3629–35. potter sm et al. Soy protein and isoflavones: their effects on blood lipids and bone density in postmenopausal women. Am J Clin Nutr 1998; 68 (6 Suppl): 1375S–9S. register tc et al. Soy phytoestrogens do not prevent bone loss in postmenopausal monkeys. J Clin Endocrinol Metab 2003; 88 (9): 4362–70. reinwald s, weaver cm. Soy isoflavones and bone health: a double-edged sword? J Nat Prod 2006; 69 (3): 450–9. rossouw je et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 2002; 288 (3): 321–33. seidlova-wuttke d et al. Effects of estradiol-17beta, testosterone and a black cohosh preparation on bone and prostate in orchidectomized rats. Maturitas 2005; 51 (2): 177–86. setchell kd, lydeking-olsen e. Dietary phytoestrogens and their effect on bone: evidence from in vitro and in vivo, human observational, and dietary intervention studies. Am J Clin Nutr 2003; 78 (3 Suppl): 593S–609S. simpson er. Aromatization of androgens in women: current concepts and findings. Fertil Steril 2002; 77 Suppl 4: S6–10. somekawa y et al. Soy intake related to menopausal symptoms, serum lipids, and bone mineral density in postmenopausal Japanese women. Obstet Gynecol 2001; 97 (1): 109–15. unfer v et al. Endometrial effects of long-term treatment with phytoestrogens: a randomized, double-blind, placebo-controlled study. Fertil Steril 2004; 82 (1): 145–8, quiz 265. vanderschueren d et al. Aromatase inhibition impairs skeletal modeling and decreases bone mineral density in growing male rats. Endocrinology 1997; 138 (6): 2301–7. wangen ke et al. Effects of soy isoflavones on markers of bone turnover in premenopausal and postmenopausal women. J Clin Endocrinol Metab 2000; 85 (9): 3043–8. wu j et al. Cooperative effects of isoflavones and exercise on bone and lipid metabolism in postmenopausal Japanese women: a randomized placebo-controlled trial. Metabolism 2006; 55 (4): 423–33.

Nutritional phytoestrogens and bone health

99

wuttke w et al. Isoflavones-safe food additives or dangerous drugs? Ageing Res Rev 2007; 6 (2): 15–88. yamori y et al. Soybean isoflavones reduce postmenopausal bone resorption in female Japanese immigrants in Brazil: a ten-week study. J Am Coll Nutr 2002; 21 (6): 560–3. zhang x et al. Prospective cohort study of soy food consumption and risk of bone fracture among postmenopausal women. Arch Intern Med 2005; 165 (16): 1890–5.

5 Endocrine-disrupting chemicals: origins, fates and transmission into the food chain L. Connolly, Queen’s University Belfast, UK

Abstract: It is widely accepted that food is one of the most important exposure routes for both natural and synthetic chemicals with endocrine disrupting properties. The release of endocrine-disrupting chemicals (EDCs) into the environment can result in their subsequent inclusion within the food chain and their eventual ingestion by humans and wildlife. EDCs can originate from many sources; their behaviour and fate are major factors in our exposure to such compounds through the food chain. Key words: endocrine-disrupting chemicals, origins, transmission into the food chain, fates.

5.1

Introduction

It is widely accepted that food and diet are the most important exposure routes for both natural and synthetic chemicals with endocrine-disrupting properties. For many of the chemicals mentioned in this chapter, the primary route of exposure is through food. For example, exposure to persistent organochlorine chemicals (OCs) in developed countries where they are banned must almost exclusively come from contaminated food. Chemicals such as these have long environmental half-lives and can contaminate food through residues which can remain in soils from times when they were used legally. Unfortunately the problem is exacerbated by an ongoing illicit use market. The vast majority of natural endocrine-disrupting chemicals (EDCs) in our diet come from the endogenous plant chemicals, phytoestrogens but natural sources also include endogenous animal hormones. Synthetic chemicals have become an integral part of our modern world and lifestyles. Not

104

Endocrine-disrupting chemicals in food

only are they used in agriculture and industry, they are also found in a wide range of consumer products such as furniture, electrical appliances, clothing, toiletries and detergents. Their use has certainly enhanced our standard of living, but at what cost? Many of these chemicals can persist in the environment and hence gain entry to the food chain. This chapter aims to firstly address in detail the types and origins of both natural and synthetic EDCs present in the food chain. In the context of diet, water is an essential constituent and as such will be included in the category of food in this chapter. The harmful effects of EDCs in food present health risks. Exposure to such chemicals via the diet may lead to detrimental health effects such as infertility and cancer in both humans and wildlife. EDCs such as phytoestrogens may be metabolised and excreted by meat-producing animals prior to consumption by humans, so reducing the risk to the consumer. However, humans are particularly exposed to persistent chemicals such as dichlorodiphenyltrichloroethane (DDT) which are susceptible to bioaccumulation in the environment and biomagnification via the food chain. While relatively low concentrations of natural hormones result in biological activity, EDCs must be consumed at relatively high concentrations to produce a corresponding effect on the human endocrine system because their hormonal activites are usually five or six orders of magnitude lower than 17βestradiol. However, a combination of EDCs at lower concentrations may result in additive, enhanced or low-level cocktail effects, thus posing a greater risk to the consumer. Finally this chapter will discuss the transmission and fate of EDCs in the food chain.

5.2

Natural endocrine-disrupting chemicals

EDCs can originate from many sources. These sources can be broadly broken down into synthetic (e.g. pesticides) and natural (e.g. phytoestrogens) compounds. While modern-day living is introducing more and more new synthetic chemicals into our lives, natural sources such as phytoestrogens have always existed. Table 5.1 lists some of the most important natural sources of EDCs present in the food chain.

5.2.1 Natural steroid hormones To date environmental compounds with endocrine-disrupting activity have attracted the most research attention. Studies investigating the exposure of consumers to naturally occurring steroids with endocrine-disrupting potential in food of animal origin are extremely limited (Andersson and Skakkebaek, 1999). Animal food products such as cheese, meat, milk and eggs have been shown to contain endogenous hormones such as estrogens, progesterone and testosterone (Hartmann et al., 1998; Courant et al., 2007, 2008). The presence of natural steroid hormones in food, estrogens in par-

Endocrine-disrupting chemicals

105

Table 5.1 Examples of the most important natural sources of EDCs found in the food chain Substance category Natural hormones

Phytoestrogens

Mycotoxins

Occurrence in foods

Examples

Sources/information

Estrogens 17β-Estradiol Estrone Estriol Progestagens Progesterone Androgens Testosterone Lignans Enterodiol Enterolactone Isoflavones Genistein Daidzein/equol Coumestrol Zearalonone

Produced naturally in the body. Present in foods produced by animals or through release in human and animal slurry practices.

Meat, milk, eggs and cheese. Water.

Produced naturally in some plants. Equol, a gut bacterial fermentation product of daidzein, may be the most estrogenic isoflavone. A natural fungal toxin produced by Fusarium spp. with estrogenic properties.

Soya, legumes, beans, sprouts, alfalfa, cabbage, spinach, soybean, grains and hops. Cereal-based foods Corn Wheat

ticular, has become an issue of growing concern among scientists. The potential risk is high because natural estrogens are up to 100 000 times more potent than xenoestrogens such as estrogen-mimicking pesticides. Milk is one food that has raised concerns due to its hormone content. Estrogen levels are particularly high in dairy cows managed under modern practices whereby the cow is usually pregnant in order to maintain a high milk yield. The more progressed the pregnancy the higher the progesterone levels also (Gyawu and Pope, 1983; Malekinejad et al., 2006). As a result the major source of animal-derived estrogens in the human diet is milk and dairy products accounting for 60–70% of estrogens consumed (Hartmann et al., 1998). It could be argued that humans have been drinking cows’ milk for thousands of years without any obvious harm. However, milk production practices have drastically changed from the low production levels of 100 years ago to today’s high production. The level of natural estrogens in modern milk has increased as a result and is becoming a new health concern (Ganmaa et al., 2001). The consumption of modern milk, containing increased levels of natural estrogens, is now being linked to cancers such as prostate (Qin et al., 2004). Estrone sulphate, a derivative of estrone with a long biological half-life, is present naturally at high levels in milk from pregnant cows (Henderson et al., 1994). Estrone sulphate is, in effect, a biological estrogen pool because it can be converted to the more active estradiol. Estradiol is the major human estrogen having a critical impact on

106

Endocrine-disrupting chemicals in food

the reproductive system and on sexual function. The natural estrogens estrone and estradiol have also been detected in eggs (Courant et al., 2007; Hartmann et al., 1998), raising concerns for critical populations such as prepubertal children. The level of natural hormones present in agriculturally derived foods may also be enhanced by the misuse of hormones for veterinary purposes or as growth promoters (Scippo et al., 1993). This will be discussed later in Section 5.3.2.

5.2.2 Phytoestrogens Plants produce estrogen-like substances termed phytoestrogens. Phytoestrogens include lignans, isoflavones and coumestrol. Lignans are associated with dietary fibres and are found in, for example, whole grains. The main estrogenic lignan derivatives, enterodiol and enterolactone, are produced by gut bacterial action on lignan precursors. The isoflavones genistein and daidzein are constituents of many foodstuffs, including beans, sprouts, cabbage, spinach, soybean, grains and hops. However, isoflavones are present in the greatest amounts in soybeans. Glycitein is a minor isoflavone, but present in amounts similar to the predominant daidzein in soygerm, which may be used as a food ingredient. The isoflavone equol is a gut bacterial fermentation product of daidzein formed by about one-third of humans studied (Hendrich, 2004). Equol may be the most potent of the isoflavones in estrogenicity terms, at least by some criteria (Hendrich, 2004). A third phytoestrogen, coumestrol, is found in a limited range of foods such as alfalfa sprouts. These naturally occurring chemicals have many structural similarities to 17b-estradiol and are more potent estrogens in vitro than many of the synthetic chemicals tested to date. There is much current interest and debate regarding the significance of human dietary exposure to phytoestrogens, particularly with respect to infant diets (Bhatia et al., 2008). There is evidence that hormonal disturbance may have its greatest effect during the early stages of life because the developing fetus has underdeveloped feedback mechanisms that regulate the endocrine system in the adult. This disturbance continues into the first few months after birth until prepubertal stage. The inclusion of infant soy milk formula, containing isoflavones, has therefore caused some concern (Tuohy, 2003). It has been speculated that animals may become adapted to dietary phytoestrogens, that is animals can metabolise and excrete them, through co-evolution with plants as a major food source, while exposure to synthetic chemicals is much more recent (i.e. post-industrial revolution – mid-nineteenth century) for this to have occurred. Phytoestrogens do not bioaccumulate, they can be conjugated and excreted like true estrogens because of their molecular structural analogy. Other xenoestrogens such as DDT do not share the steroid nucleus analogy, are extremely hydrophobic,

Endocrine-disrupting chemicals

107

are not metabolised and therefore will bioaccumulate. Concerns to date have been focused on synthetic EDCs because of their potential to persist and bioaccumulate. However, the relative significance of exposure to natural and xenoestrogens remains to be clarified. Nilsson (2000) discusses the research trends on EDCs in the food chain and environment within the context of significant exposure and risk assessment of the relevant potency of natural versus synthetic compounds.

CH3 OH H H

17-β-Estradiol is the main oestrogen in humans

H

HO OH

CH3

O O

HO

O

Zearalenone O

OR

General structure of a paraben (para-hydroxybenzoate) R is an alkyl group, e.g. methyl (CH3), propyl (C3H7), butyl (C4H9) OH

H3C O

CH3

H3C

O

CH3 CH3 OH 2-tert-butyl-4-hydroxyanisole

CH3 CH3 OH

CH3

3-tert-butyl-4-hydroxyanisole

Butylated hydroxyanisole (BHA) is a mixture of two isomeric organic compounds, 2-tert-butyl-4-hydroxyanisole and 3-tert-butyl-4-hydroxyanisole. It is prepared from 4-methoxyphenol and isobutylene. It is a waxy solid that exhibits antioxidant properties. E number of this food additive is E320.

Fig. 5.1 Comparison of the molecular structure of zearalenone (an estrogenic mycotoxin), parabens and hydroxylated butanol (food additives with weak estrogenic activity) and 17-β-estradiol, the main estrogen in humans.

108

Endocrine-disrupting chemicals in food

5.2.3 Mycotoxins Mycotoxins are the toxic metabolites of fungi which can contaminate grain crops in temperate climates during harvesting or storage. Foods based on these grains may be contaminated by a variety of mycotoxins (Council for Agricultural Science and Technology (CAST), 2003). However to date only one mycotoxin, zearalenone, has been identified as an EDC (Mirocha et al., 1971; Loomis and Thomas, 1999). Zearalenone and its derivatives are common corn and soy contaminants produced by Fusarium spp. Surveys in various countries such as China (Li et al., 1999), Canada (Roscoe et al., 2008) and Germany (Schollenberger et al., 2007) have reported zearalenone contamination as high as 214 μg/kg in foodstuffs such as soy and at frequencies greater than 20% in breakfast cereals. Structural analogy of zearalenone with 17β-estradiol (Fig. 5.1) is thought to confer its estrogenic activity (Shaw et al., 2004). As these toxins are capable of producing estrogenic effects, the European Union has set an advisory level of 50 μg/kg (Commission Regulation (EC) No. 856/2005). There is an additional concern that because of the pharmacological properties of this fungal toxin it has been used in some parts of the world as a feed additive to enhance feed efficiency and muscle deposition in livestock (Welshons et al., 1990).

5.3

Synthetic endocrine-disrupting industrial chemicals

Human activity particularly during and after the Industrial Revolution has introduced thousands of new chemicals into our lives; some have entered our food chain. While there are lifestyle and therapeutic advantages of many of these chemicals, some have undesirable endocrine-disrupting effects and still more have yet to be investigated. These chemicals include pharmaceuticals, agricultural and industrial chemicals and food additives. Table 5.2 summarises some of the most important sources of synthetic EDCs which have entered the human food chain.

5.3.1 Pharmaceuticals Pharmaceuticals such as birth control pills, painkillers, antidepressants, cancer treatments and tranquilisers have been developed to promote human health and well-being. However, many of these compounds and their metabolites may eventually find their way into the environment, predominantly via excretion and inclusion in sewage. In 1994, Purdom and colleagues reported that sewage treatment plant effluent was estrogenic to fish. Initial attempts to identify the cause of feminisation of fish exposed to sewage treatment plant effluents focused on synthetic organic chemicals with known estrogenic effects, such as plasticisers and surfactants (Sumpter, 1995). However recent research suggests that natural estrogens and a common synthetic birth control

Drugs with hormonal activity may be excreted or disposed of by humans into sewage systems. Groundwater destined for drinking water may become contaminated. Synthetic hormones used illegally as growth promoters. Residues can remain in animal food products or be released to the environment through animal slurry.

Chemicals used in farming/crop production. Some fat-soluble persistent organic pollutants such as DDT have been banned for at least 30 years. Carbendazium is a metabolite of Benomyl.

Contraceptives Ethinyl estradiol Antibiotics Painkillers Cancer drugs Anti-depressants

Estrogenic Diethylstilbestrol Hexestrol Androgenic Trenbolone Progestagenic Medroxyprogesterone Glucocorticoid Dexamethasone

Insecticides DDT Aldrin Fungicides Benomyl/Carbendazium Vinclozolin Hexachlorobenzene Herbicide Atrazine 2,4 Dichlorophen-oxyacetic acid Nematocide Aldicarb DBCP

Pharmaceuticals

Agricultural chemicals (veterinary drugs)

Agricultural chemicals (pesticides)

Sources/information

Examples

Fish, meat and dairy products. Fruit and vegetables. Cereals. Drinking water.

Food-producing animals and their food products. Drinking water.

Drinking water

Occurrence in foods

Examples of known or suspected endocrine-disrupting synthetic chemicals and their occurrence in food types

Substance category

Table 5.2

Polychlorinated biphenyls (PCBs)

Industrial chemicals

Antifungal agent, preservative and used in foods to prolong shelf-life. Weak estrogenic mimickers. Phenolic compounds that are added to food to preserve fat.

Parabens

Food additive chemicals

Butylated hydroxyanisole

By-products of combustion of many materials, herbicide production and the paper industry. Fat soluble.

Dioxins and furans

Chemicals used in the production of electrical goods, banned since the 1980s. Widespread, persistent environmental pollutants. Fat soluble. Non-ionic surfactants used in detergents, paints, herbicides, pesticides, fertilisers and plastics. Breakdown products such as nonylphenol and octylphenol, are found in sewage and industrial effluents. Widely used as plasticisers for PVC. Common environmental pollutants. A component of polycarbonate plastics and epoxy resins used to line food cans and water pipes. Synthetic brominated chemicals used as flame retardants in plastics foams and textiles. Fluorine containing chemicals with unique properties to make materials stain and stick resistant. Persistent contaminants. Used in microwave popcorn bags, pizza boxes and non-stick cookware.

Sources/information

Industrial by-waste chemicals

Perfluorinated compounds (PFCs)

Brominated flame retardants (BFRs)

Bisphenol A (BPA)

Phthalates

Alkylphenol polyethoxylates (APEs)

Examples

Continued

Substance category

Table 5.2

Long-life foods containing fat.

Long-life foods.

Meat, eggs and dairy products.

Foods which are packaged in greaseresistant packaging, pizza, chips or cooked in non-stick cookware.

Drinking water. Fish.

Drinks and foods stored in plastic packaging. Canned food and drinking water.

Drinking water. Fish.

Fish, meat and dairy products.

Occurrence in foods

Endocrine-disrupting chemicals

111

pharmaceutical, ethinyl estradiol, are the most potent estrogens in sewage effluent (Snyder et al., 2001; Muller et al., 2008). Ethinyl estradiol is excreted together with natural hormones in women’s urine. In fact, researchers have demonstrated that ethinyl estradiol can induce endocrine-disruptive effects in fish at concentrations present in some municipal sewage effluents (Purdom et al., 1994; Jobling et al., 1998). This led Mittwoch and her colleagues (1993) to refer to human sexual development in a ‘sea of estrogens’. This raises concerns regarding the contamination of groundwater destined for use as drinking water and the entry of such compounds into the food chain. Scientists are concerned whether the hormones being discharged into water courses are being effectively removed from drinking water supplies (Bodzek and Dudziak, 2006). Various drugs with endocrinological functions or side-effects are commonly found, but the estrogenic hormones originating from natural and oral contraceptives are of particular concern. Ethinyl estradiol has been highlighted as an important contributor to water pollution and has been implicated in a UK study as a major estrogenic substance contaminating domestic sewage effluents (Brighty, 1997).

5.3.2 Veterinary medicines Veterinary medicines which may occur as residues in food are assessed and regulated by the appropriate bodies: the Committee for Medicinal Products for Veterinary Use (CVMP) for the European Union and the Joint Expert Committee on Food Additives (JECFA) for the international food standards programme. Violative levels of prescribed drug residues in animal products can be avoided by adhering to recommended withdrawal periods between dosing and slaughter. However, failure to do so and the use of banned drugs gives rise to the potential for residues entering the human food chain. Such food contaminants may produce toxicological, pharmacological, microbiological, immunological and endocrine disruption hazards. There is a lack of reports in the scientific literature describing cases of acute toxicity through the presence of pharmacologically active drug residues in animal products implying that residue levels are so low that direct pharmacologic effects are rare. However, the potential for such food contamination presenting an unacceptable risk exists. One such isolated case occurred in Spain when human food poisoning by clenbuterol residues in veal liver was reported in 1991 (Pulce et al., 1991). A number of cases of hormonal contamination of foods have been reported with serious effects for the consumer. The synthetic hormone diethylstilbestrol (DES), once used as an agricultural growth promoter (banned in the EU in 1981) was detected several times in 1980 in baby food made with veal (Jones and Knifton, 1982). The baby food was manufactured from French cows treated with DES. This reportedly led to various deformities in

112

Endocrine-disrupting chemicals in food

infants, e.g. babies growing breasts. For example, an outbreak of premature breast development before the age of 8 years occurred in a school in northern Italy (Fara et al., 1979) and a similar event was reported in Puerto Rico (Bongiovanni, 1983). In both cases the cause of precocious sexual development was attributed to meat containing estrogenic substances. The increased use of veterinary drugs and the associated risks to human health have led to an abundance of legislation controlling their use. Bovine growth hormone (BGH), also known as bovine somatotropin (BST), is a protein hormone, banned in Europe, but is given to cows to improve milk production in the US. Although BGH may be good for the economies of dairy farmers it may not be good for the consumer – this is the subject of great debate, indeed BGH has been implicated as a possible breast cancer risk. However, other studies suggest that it is not a risk because it does not survive digestion as might be expected for a simple peptide.

5.3.3 Pesticides Pesticides have been used to control the destructive effects of insects, weeds, fungi, worms and higher animals throughout the world in agriculture and public health situations for many years. Concern is growing because many of these chemicals, such as the OCs, are persistent organic pollutants (POPs) and as a result they can bioaccumulate in the environment, the food chain and ultimately the fat of wildlife and humans. Owing to these concerns many OCs such as DDT and chlorinated cyclohexanes (e.g. γhexachlorocyclohehane; lindane; γ-HCH) have been banned from use in Europe and some even worldwide. However, such chemicals are very persistent and can be found in food many years after they were last used in the area of the food’s origin. Pyrethroids are a group of non-OC synthetic pesticides commonly used around the world in forestry and food crops. The pyrethroids were regarded as safe because of their favourable insect : mammalian toxicity ratio, their rapid metabolism and environmental degradation. As a result, pyrethroid residues in food are rare. However, concerns have recently been raised about the estrogenic activity of their metabolites the hydroxyphenols and environmentally degraded forms (McCarthy et al., 2006). These breakdown products are not normally monitored as part of pesticide residues surveillance programmes in food and their impact on human health cannot be assessed without detailed residue data. Although people can be exposed to these chemicals in several ways, such as contaminated air, water or through the skin in countries where they are still used, the primary route of exposure to humans in developed countries is through contaminated food. The main food types which have been found to be contaminated typically include higher fat content foods such as oily fish, fatty meats and dairy products. A number of studies which report the contamination of the food chain by pesticides are listed in Table 5.3.

Table 5.3 Examples of studies which report the contamination of the food chain by pesticides and other EDCs Food type

Contaminant

Information

Reference

Various supermarket food items

Various POPs

Analysis including fruit, vegetables, cereals, baked goods, fish and animal products (meat, butter, cheese, fat and eggs).

Schafer and Kegley (2002)

Butter

Various POPs: PCDD/Fs PCBs Dioxin-like PCBs HCB DDT

Survey of butter worldwide. Background contamination found in most samples (throughout Europe) but below EU action and maximum levels.

Weiss et al. (2005) Malisch and Dilara (2006)

Honey

Organohalogens Organophosphates Pyrethoids Organonitrogens Organochlorine pesticides

Organohalogens and organophosphorous detected in Brazilian honey. Spanish and Portuguese honey had detectable levels of organochlorines.

Rissato et al. (2007) Blasco et al. (2004)

Fruit and vegetables

Organochlorine pesticides HCH, HCB, DDT, Aldrin, Dieldrin

Detected in Nigerian fruits, vegetables and tubers. None was detected above the FAO’s maximum residue limits.

Adeyeye and Osibanjo (1999)

Salmon

Organochlorine pesticides

Comparison of farmed and wild salmon. Farmed salmon has much higher levels of organochlorines in the US and particularly Europe.

Hites et al. (2004)

Shellfish

Organochlorine pesticides PCBs

Detected in shellfish and fish in China.

Yang et al. (2006)

Pork

Organochlorine pesticides HCH DDT

Detected in pork from Romania.

Covaci et al. (2004)

Meat and fish

Organochlorine pesticides HCH DDT

Croatia – imported and domestic. Noted a decline in chlorinated hydrocarbons of domestic origin meat and fish in comparison to a study ten years previously.

Kipcic et al. (2002)

Milk

Organochlorine pesticides HCH DDT DDE DDD

Detected OCPs in 206 out of 305 samples collected in India. 2001 survey of milk in Beijing supermarkets detected OCPs.

Nag and Raikwar (2008) Zhong et al. (2003)

114

Endocrine-disrupting chemicals in food

The EU has set maximum levels for pesticide residues such as DDT and hexachlorobenzene (HCB) in foodstuffs of animal origin (including meat, meat products, offal, dairy products, and eggs) (Commission Regulation (EC) No 396/2005). As a result of legislation limiting or banning dangerous pesticides, compounds such as p,p′-DDT, and its metabolites, p,p′-DDE and p,p′-TDE, have been reported in decreasing concentrations in various food studies (Nakata et al., 2002; Kipcic et al., 2002; Matsumoto et al., 2006).

5.3.4 Industrial chemicals Many industrial chemicals such as polychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs), brominated flame retardants (BFRs) and alkylphenol polyethoxylates (APEs) are of concern because they are widespread environmental pollutants with estrogenic properties. Dioxin is a generic name for the PCDDs, PCDFs and PCBs. Dioxins are one of the most dangerous pollutants known to humankind; they are known carcinogens, supressors of the immune system, cause damage to the liver, cause birth defects and exposure to them may even result in death. Many of these contaminants are fat soluble, can bioaccumulate in the food chain and can be found in fish, meat and dairy products. The groups causing the most concern to human health, PCBs, PCDDs and PCDFs, are 3 of the 12 compounds included in the banned ‘dirty dozen list’ of POPs as agreed by the Stockholm convention (UNEP, 2001; Table 5.4). As a result of such measures levels of PCBs, PCDDs and PCDFs in food and humans have started to decline. However, other persistent endocrine-disrupting compounds such as BFRs are increasing causing concern about possible future effects. The dioxins PCDDs and PCDFs are not intentionally produced but are by-products of various industrial processes such as bleaching, herbicide production and incineration. Dioxins get into the food chain, water supply and the atmosphere mainly from the incineration of municipal and industrial waste. It is impossible to burn waste without producing dioxins unless

Table 5.4 Biomagnification of DDT in the aquatic food web, data taken from Woodwell et al. (1967) Food web

DDT concentration (ppm)

Biomagnification factor

Water Plankton Shrimp Insects Eel Tern Cormorant

0.000 005 0.04 0.16 0.3 0.28 4.75 26.4

1 800 3 200 6 000 5 600 95 000 520 000

Endocrine-disrupting chemicals

115

the temperature is extremely high. These toxins can be carried through the air to areas many miles from their source where they can be deposited in water and soil, and directly onto food crops. Deposition of airborne dioxins onto plant and soil surfaces followed by ingestion of this contaminated vegetation by food animals is considered the primary pathway by which dioxins enter the food chain (Fries, 1995). Fish can become contaminated due to airborne dioxin deposition into water and also from contaminated soil or industrial waste washed into rivers and lakes, leading to high sediment concentrations. Several incidents of dioxins in the food chain have been reported, such as their presence in the milk of cows that had grazed on contaminated grass (Liem et al., 1991). A much larger and recent incident occurred in Belgium in 1999 when chickens and their food products were contaminated with dioxins as a result of using animal feed that was produced with contaminated fat (Bernard et al., 1999, 2002; Schoeters and Hoogenboom, 2006). PCBs are highly toxic chlorinated dibenzofurans and have previously been used in oils for coolants and insulators in transformers and capcitors. However, owing to their neurotoxic nature and widespread persistence in the environment they were banned within the US and Europe in the 1980s but continued to be produced in other countries such as North Korea and Russia. Their existence is extended because the typical lifespan of a transformer is at least 10 years and so many PCB-containing transformers are still in use today. In 1979 an incident of PCB contamination of the food chain by accidental release from a broken transformer illustrated the need for vigilance over the fate of these contaminants (Drotman et al., 1983). Oil from the broken transformer leaked into a waste reservoir which held waste products from a pig processing area. The waste products were further processed into meatmeal to be used as protein supplements in animal feeds. Chickens fed on contaminated meal resulted in the widespread distribution of chicken and egg products contaminated with PCBs throughout the US, Canada and Japan. Human milk sample analysis showed that PCB absorption had occurred following egg consumption. Other incidents of PCB contamination have been reported worldwide (Lonky et al., 1996; Hsu et al., 1984). BFRs such as polybrominated diphenyl ethers (PBDEs) have been used as fire prevention chemicals in plastic, foam and textiles since the 1970s. PBDEs are widespread, persistent and bioaccumulating environmental contaminants. They are also EDCs and cause neurodevelopmental effects. Concerns are growing because studies have shown PBDEs to be increasing in wildlife in areas such as the Arctic (Ikonomou et al., 2002) and in human milk (Mieronyte et al., 1999). As a result of these concerns the EU introduced Council Directive 2003/11/EC, banning a number of PBDEs. APEs are non-ionic surfactants used in detergents, paints, herbicides, pesticides, fertilisers and plastics. They have been shown to be estrogenic (Hong et al., 2008) and can enter the food chain via discharge into water systems (Meier et al., 2007).

116

Endocrine-disrupting chemicals in food

5.3.5 Food packaging/contact chemicals Many materials used in food packaging or which come into contact with food contain EDCs that can migrate into the packaged food. Phalate esters are weakly estrogenic plasticisers (Jobling et al., 1995) used to soften plastics destined for material such as cling film and plastic wrappers. Studies have found that phthalates have entered the food chain (Jarosova, 2006) and are present in food samples (MAFF UK, 1996). Bisphenol A is a component of polycarbonate plastics and epoxy resins used in resin linings of food cans and water pipes. Perfluorinated compounds (PFCs) can be found in nonstick Teflon cookware and are also used in fast-food packaging for their stain-resistant properties. These compounds are covered elsewhere within this book in more detail.

5.3.6 Food additive chemicals Food additives include numerous chemical compounds which can be used in food processing as colour enhancers, preservatives, antioxidants, sweeteners, emulsifiers, thickeners and gelling agents. However, a number of these food additives have endocrine-disrupting properties. Paraben, an antifungal agent and preservative which has been used in foods to prolong shelf-life, has been shown to have weak estrogenic activity (Okubo et al., 2001). Butylated hydroxyanisole is a phenolic antioxidant compound used to preserve fat and is weakly estrogenic (Jobling et al., 1995). Like zearalenone, parabens and butylated hydroxyanisole show structural similarities to 17βestradiol (Fig. 5.1). For specific food additive chemicals, data requirements are generally similar to those required for agricultural veterinary medicines and their use in Europe is controlled by Council Directive 89/107/EEC which sets out the basis for controls on food additives authorised for use in foodstuffs intended for human consumption.

5.4

Fate of endocrine-disrupting chemicals and their transmission into the food chain

Our environment can be contaminated by EDCs originating from many sources. Domestic sewage waste containing excreted medicines (e.g. ethinyl estradiol), household cleaners (e.g. alkyphenols), natural hormones (e.g. estradiol) and plasticisers (e.g. phthalates) can contaminate waterways destined for drinking water. Domestic solid waste can leach from landfill, releasing phthalates from plastics and PCBs from electrical equipment. Industrial chemicals can be released via landfill sites, spills, combustion processes, incineration and industrial effluent. Agricultural practices may introduce contaminants such as veterinary drugs, natural hormones and pesticides through the use of veterinary medicines, slurry spreading and pesticide

Endocrine-disrupting chemicals

117

spraying respectively. Our extensive transport systems can contribute contaminants via fuel emissions from the combustion of fossil fuels. In recent years, new chemicals such as BFRs (used for fire prevention in plastic goods such as TVs) and PFCs (used for stain-proof and non-stick coatings) have added to the concerns over persistent contaminants. These newer synthetic chemicals can escape into the environment at many points including during the manufacturing process, during the use of products containing them or as the product degenerates. As a result the air and dust in our environment can contain such chemicals through their release from furniture, textiles and electrical appliances. The presence of synthetic chemicals originating from all of these important sources of environmental contamination has resulted in ecosystems throughout the world being contaminated. Some of these chemicals (e.g. veterinary drugs) may be short-lived and can be controlled to a large extent through legislation. However, other chemicals such as DDT have been banned for decades (apart from its use as an agent to control the transmission of malaria) and can still be found throughout the environment due to its persistence and the existence of a robust black market for it in Asia (Shaw, 1999; Shaw et al., 2000). The route of human exposure to EDCs can be relatively direct, such as the absorption of phthalates found in personal care products through the skin or the inhalation of fire retardants released from consumer products into indoor air and dust (Harrad et al., 2004). However, for the chemicals discussed in this chapter, particularly those which are persistent and bioaccumulative, the most important exposure route is through the ingestion of food. The release of chemicals into the environment (soil, water and air), their subsequent inclusion in the food chain and their eventual ingestion by humans and wildlife is a chain of events leading to our exposure to EDCs. Contaminants such as DDT and PCBs can enter plants and animals at the bottom of the food chain. Biomagnification can occur as the animals at the bottom of the food chain are then consumed by animals higher up. When we consume food from animals and their food products such as meat, dairy products and fish, the contaminants in these foods gain entry to our bodies. Table 5.5 outlines the biomagnification of DDT in the aquatic food web. Water contains low levels of DDT, plankton filter the water, shrimps eat the plankton, insects eat the shrimps, eels eat the insects, terns eat the eels and finally the higher animal, the cormorant eats the terns. This process magnifies the concentration of DDT in water (0.000 05 ppm) by a factor of 520 000 to a concentration in the cormorant of around 26.4 ppm (Woodwell et al., 1967). Aquatic wildlife is especially sensitive to waterborne endocrine disruptors to which they are constantly exposed. Some surface waters, rivers and lakes, and the sea are very vulnerable to contamination by waste and airborne pollutants. Sex hormones can appear in soil, surface water and groundwater as a result of manure application. Groundwater is vulnerable to pollution by chemicals carried by rainwater, leaching from waste sites

118

Endocrine-disrupting chemicals in food Table 5.5 The dirty dozen; a list of the 12 banned POPs as agreed by the Stockholm Convention, 2001 Persistent organic pollutant 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Aldrin Chlordane DDT Dieldrin Dioxins Endrin Furan Heptachlor Hexachlorobenzene (HCB) Mirex Polychlorinated biphenyls (PCBs) Toxaphene

Consumption – by humans

Processing and cooking – additives

Animal food products – meat, milk, cheese

Transport – emission fumes

Fig. 5.2

Crops – grains, fruit, vegetables, phytoestrogens

Storage and packaging – plastics, tins, greaseproof

Wildlife – fish

Water ways – drinking water

Agricultural practices Industrial chemicals – slurry, pesticides, – spills, waste, veterinary drugs incineration

Waste – domestic, sewage, landfill

Contamination of the food chain from various sources by EDCs.

or from wastewater carrying industrial, agricultural or domestic effluent. Treatment of drinking water may remove some, but not all, of these contaminants. Some polycarbonate or metal pipes that are lined with epoxy resin lacquers may release bisphenol A. The many sources of environmental contamination can all contribute to the contamination of our food as depicted in Fig. 5.2. PCBs, phthalates,

Endocrine-disrupting chemicals

119

pesticides and medicines can enter the food chain via water systems, wildlife, crops and animal food products. Further processing, food additives (e.g. parabens), plastics in packaging (e.g. phthalates), food cans (e.g. bisphenol A) and greaseproof packaging (e.g. PFCs) can further add to the contamination profile of food resulting in the exposure of humans to a cocktail of EDCs in the diet. Tolerances for individual chemicals are established without consideration of the effect of combinational low-level exposures, also termed the ‘low-level cocktail effect’; this is particularly important as EDCs are ligands for the same receptors. There is the added concern of long-term low-level exposure, especially to the developing fetus and young children (see Chapter 1). It is important to understand the fate and behaviour of chemicals in the environment as part of the process of assessing the risks to humans. Based on a chemical’s inherent physicochemical properties, it is possible to predict with some degree of certainty which environmental compartment it is likely to reside in and to what extent it is likely to be bioavailable and accumulate through the food chain and so end up as a component of the diet. EDCs that are lipophilic and poorly biodegraded may persist in the environment for many years, bioaccumulating (through storage and buildup) in the fat of animals – these are termed POPs. Most POPs are products and by-products of industry and are of relatively recent origin such as industrial chemicals, pesticides and unwanted by-products such as dioxins. Examples include PCBs, DDT and BFRs (Schecter et al., 2004). The food chain is especially vulnerable to contamination by POPs because they are hydrophobic and therefore bioaccumulate. The problem is worse in colder climates because POPs biodegrade more slowly at lower temperatures. They can travel through air and water, resulting in traces of these chemicals being found in most human beings and all kinds of wildlife, even in isolated parts of the globe (WWF-UK, 2003) – indeed it has been suggested that every cell in every living being contains at least one DDT molecule. Many of these chemicals have been detected in both children and adults. Of particular concern is the fact that in some cases they have been found in higher concentrations in children. As discussed in other chapters in this book there is also growing concern over possible links between EDCs and human health effects such as cancer, infertility, cardiovascular disease, birth defects, diabetes and obesity and disruption of infant brain development. POPs are banned in most countries because of their harmful effects. The United Nations has called them ‘a serious threat to human health’ but despite this they are still being used as pesticides in many countries. POPs create a risk that cannot be managed due to their persistence, so the most appropriate way to deal with them is to eliminate them from entry to the environment – even then it will take many years for their environmental contamination levels to reduce significantly. Consequently the Stockholm

120

Endocrine-disrupting chemicals in food

Convention (UNEP, 2001) was agreed to reduce or eliminate the emission of 12 POPs (listed in Table 5.4) into the environment. All aspects of the environment can become contaminated by POPs but food is the main way that POPs enter our bodies. As they are soluble in fat and not easily broken down in the body, they accumulate in the fatty tissues – as one animal eats another, the level of POPs in fat biomagnifies. This means that the highest levels are found in predator animals at the top of food chains such as polar bears, seals, toothed whales, birds of prey and humans. As a result POPs will accumulate especially in foods of a high fat content and so the greatest risk of exposure comes from eating foods such as fatty meat, dairy products and oily fish from contaminated waters. For this reason milk and dairy products are good indicators for the contamination of POPs along the food chain. For the same reason human milk is a good indicator of exposure to POPs (Shaw et al., 2000; Burke et al., 2003). The use of milk and milk products as indicators of contamination has been recognised by the United Nations Environment Programme (UNEP) which recommends milk as an appropriate matrix to indicate the presence of POPs (UNEP, 2001). In 2006 the World Wildlife Fund (WWF, 2006) commissioned a report, ‘Chain of contamination: the food link’, to explore the dietary route of exposure to chemicals. This report describes the preliminary analysis of 27 food items, selected from seven EU countries, for the presence of synthetic chemical contaminants. The food items were analysed for chemicals that had previously been detected in the WWF’s biomonitoring studies (WWFUK, 2003) (PCBs, DDT, BFRs, PFCs, phthalates and artificial musks) and in other biomonitoring and indoor air and dust studies (organotins and alkylphenols). All the food items were found to be contaminated with chemicals, many of which had already been detected in the preceding studies in wildlife and humans. This emphasises the signifigance of diet as a direct means of exposing humans to chemical contaminants. The report highlights the widespread contamination of food with synthetic chemicals and concludes that the situation is of a global nature. The biomonitoring and food results presented by the WWF’s surveys (WWF-UK, 2003; WWF, 2006) have determined that although we are exposed to and consequently contaminated with a cocktail of potentially hazardous chemicals, it is not possible to assess the true risk to the consumer due to the lack of adequate health and safety data. It is unlikely that chemical levels in food pose an immediate health risk but concerns remain over the long-term, low-level and cocktail effect that they present to humans, particularly the developing fetus, infants and young children. Food safety legislation is one of the most important preventative measures designed to protect the consumer from dietary exposure to hazardous substances. The ultimate, and somewhat idealistic and optimistic, solution for reducing our exposure lies in continuing to reduce or even end the emission of hazardous substances into the environment.

Endocrine-disrupting chemicals

5.5

121

References

adeyeye a, osibanjo o. (1999). Residues of organochlorine pesticides in fruits, vegetables and tubers from Nigerian markets. The Science of the Total Environment, 231(2–3): 227–233. andersson am, skakkebaek ne. (1999). Exposure to exogenous estrogens in food: possible impact on human development and health. European Journal of Endocrinology, 140: 477–485. bernard a, hermans c, broeckaert f, poorter g de, cock ade, hoins g. (1999). Food contamination by PCBs and dioxins; an isolated episode in Belgium is unlikely to have affected human health. Nature, 401: 231–232. bernard a, broeckaert f, poorter g de, cock ade, hermans c, saegerman c, hoins g. (2002). The Belgian PCB/Dioxin incident: analysis of the food chain contamination and health risk evaluation. Environmental Research Section A, 88: 1–18. bhatia j, greer f, and the committee on nutrition. (2008). Use of soy protein-based formulas in infant feeding. Pediatrics, 121(5): 1062–1068. blasco c, lino cm, pico y, pena a, font g, silvara mi. (2004). Determination of organochlorine pesticide residues in honey from the central zone of Portugal and the Valencian community of Spain. Journal of Chromatography A, 1049(1–2): 155–160. bodzek m, dudziak m. (2006). Elimination of steroidal sex hormones by conventional water treatment and membrane processes. Desalination, 198: 24–32. bongiovanni am. (1983). An epidemic of premature thelarche in Puerto-Rico, Journal of Paediatrics, 103: 245–246. brighty g. (1997). ‘The Environment Protection Agency project 490’, The identification and assessment of oestrogenic substances in sewage treatment works effluents, R&D publication, The Environmental Agency, London. burke er, holden aj, shaw ic, suharyanto fx, sihombing g. (2003). Organochlorine pesticide residues in human milk from primiparous women in Indonesia. Environmental Contamination and Toxicology, 71: 148–154. commission regulation of 23 February 2005, (EC) No 396/2005, on maximum residue levels of pesticides in or on food and feed of plant and animal origin and amending council directive 91/414/EEC. Official Journal of the European Union, L70, 16.3.2005, 1–16. commission regulation of 6 June 2005, (EC) No 856/2005, amending regulation (EC) No 466/2001 as regards Fusarium toxins. Official Journal of the European Union, L143, 7.6.2005, 3–8. council for agricultural science and technology (cast). (2003). Mycotoxins: risks in plant, animal and human systems. Task Force Report, no.139. courant f, antignac jp, maume d, monteau f, andre f, le bizec b. (2007). Determination of naturally occurring oestrogens and andorgens in retail samples of milk and eggs. Food Additives and Contaminants, 24(12): 1358–1366. courant f, antignac jp, laille l, monteau f, andre f, le bizec b. (2008). Exposure assessment of prepubertal children to steroid endocrine disruptors. 2. Determination of steroid hormones in milk, egg and meat samples. Journal of Agricultural and Food Chemistry, 56: 3176–3184. council directive of 21 December 1988, 89/107/EEC, on the approximation of the laws of the Member States concerning food additives authorized for use in foodstuffs intended for human consumption. Official Journal of the European Communities, L40, 11.2.89, 27–33. council directive of 6 February 2003, 2003/11/EC, on the marketing and use of certain dangerous substances and preparations (pentabromodiphenyl ether, octabromodiphenyl ether). Official Journal of the European Communities, L42, 15.2.2003, 45–46.

122

Endocrine-disrupting chemicals in food

covaci a, gheorghe a, schepens p. (2004). Distribution of organochlorine pesticides, polychlorinated biphenyls and [alpha]-HCH enantiomers in pork tissues. Chemosphere, 56(8): 757–766. drotman dp, baxter pj, liddle ja, brokoff cd, skinner md. (1983). Contamination of the food chain by polychlorinated biphenyls from a broken transformer. American Journal of Public Health, 73: 290–292. fara gm, del corvo gd, bernuzzi s, bigatello a, dipietro c, scaglioni s, chiumello g. (1979). Epidemic of breast enlargement in an Italian school. The Lancet, 313: 295–297. fries gf. (1995). Transport of organic environmental contaminants to animal products. Reviews of Environmental Contamination and Toxicology, 141: 71–109. ganmaa d, wang py, qin lq, hoshi k, sato a. (2001). Is milk responsible for male reproductive disorders? Medical Hypotheses, 57(4): 510–514. gyawn p, pope gs. (1983). Oestrogens in milk. Journal of Steroid Biochemistry, 19(1): 877–882. harrad s, wijesekera r, hunter s, halliwell c, baker r. (2004). Preliminary assessment of U.K. human dietary and inhalation exposure to polybrominated diphenyl ethers. Environmental Science Technology, 38(8): 2345–2350. hartmann s, lacorn m, steinhart h. (1998). Natural occurrence of steroid hormones in food. Food Chemistry, 62(1): 7–20. henderson km, karanikolas m, kenealy l, macmillan kl. (1994). Concentrations of oestrone sulphate during pregnancy in milk from Jersey and Friesian dairy cows differing in milk yield and composition. New Zealand Veterinary Journal, 42: 89–92. hendrich s. (2004). ‘Dietary estrogens’, in Watson DH (ed) Pesticide, Veterinary and Other Residues in Food, Woodhead Publishing, Cambridge. hites ra, foran ja, carpenter do, hamilton mc, knuth ba, schwager sj. (2004). Global assessment of organic contaminants in farmed salmon. Science, 303: 226–229. hong s, xiao-lin x, jian hua q, xia h, yu-bang w, li-chun x, xin-ru w. (2008). 4Alkylphenols and related chemicals show similar effect on the function of human and rat estrogen receptor α in reporter gene assay. Chemosphere, 71: 582–588. hsu st, ma ci, hsu sk, wu ss, hsu nh, yeh c. (1984). Discovery and epidemiology of PCB poisoning in Japan. American Journal of Independent Medicine, 5: 71–79. ikonomou mg, rayne s, addison rf. (2002). Exponential increases of the brominated flame retardants, polybrominated diphenyl ethers, in the Canadian Arctic from 1981 to 2000. Environmental Science and Technology, 36(9): 1886–1892. jarasova a. (2006). Phthalic acid esters (PAEs) in the food chain. Czech Journal of Food Science, 24: 223–231. jobling s, reynolds t, white r, parker mg, and sumpter jp. (1995). A variety of environmentally persistent chemicals, including some phthalate plasticisers, are weakly estrogenic. Environmental Health Perspectives, 103(7): 582–587. jobling s, nolan m, tyler cr, brighty g, sumpter jp. (1998).Widespread sexual disruption in wild fish. Environmental Science Technology, 32: 2498–2506. jones de, knifton a. (1982). Oestrogen therapy without stilbenes. Veterinary Record, 110: 441–443. kipcic d, vukusic j, sebecic b. (2002). Monitoring of chlorinated hydrocarbon pollution of meat and fish in Croatia. Food Technology and Biotechnology, 40(1): 39–47. li f, luo x, yoshizawa t. (1999). Mycotoxins (tricothecenes, zearalenone and fumonsins) in cereals associated with human red-mold intoxications stored since 1989 and 1991 in China. Natural Toxins, 7: 93–97. liem akd, hoogerbrugge r, kootstra pr, van der velde eg, de jong apjm. (1991). Occurrence of dioxins in cow’s milk in the vicinity of municipal waste incinerators and a metal reclamation plant in the Netherlands. Chemosphere, 23: 1975–1984.

Endocrine-disrupting chemicals

123

lonky e, reihman j, darvill t, mather j, daly h. (1996). Neonatal behaviour assessment scale performance in humans influenced by maternal consumption of environmentally contaminated Lake Ontario fish. Journal of Great Lakes Research, 22: 198–212. loomis ak, thomas p. (1999). Binding characteristics of estrogen receptor (ER) in Atlantic croaker (Micro pogonias undulatis) testis: different affinity for estrogens and xenobiotics from that of hepatic ER. Biology of Reproduction, 61: 51–60. maff uk. (1996). Phthalates in food. Food surveillance information sheet – number 82. Department of Health Scottish Executive, MAFF UK. malekinejad h, scherpenisse p, bergwerff a. (2006). Naturally occurring estrogens in processed milk and in raw milk (from gestated cows). Journal of Agricultural and Food Chemistry, 54: 9785–9791. malish r, dilara p. (2006). PCDD/Fs and PCBs in butter samples from new European Union member states and a candidate country: Analytical quality control, results and certain PCB-specific aspects. Chemosphere, 67(9): S79-S89. matsumoto h, kuwabara k, murakami y, murata h. (2006). Survey of PCB and organochlorine pesticide residues in meats and processed meat products collected in Osaka, Japan. Shokuhin Eiseigaku Zasshi, 47(3): 127–135. mccarthy ar, thomson bm, shaw ic, abell ad. (2006). Estrogenicity of pyrethoid insecticide metabolites. Journal of Environmental Monitoring, 8: 197–202. meier s, andersen te, norberg b, thorsen a, taranger gl, kjesbu os, dale r, morton hc, klungsoyr j, svardal a. (2007). Effects of alkyphenols on the reproductive system of Atlantic cod (Gadus morhua). Aquatic Toxicology, 81: 207–218. mieronyte d, noren k, bergman a. (1999). Analysis of polybrominated diphenyl ethers in Swedish human milk. A time-related trend study, 1972–1997. Journal of Toxicology and Environmental Health, 58: 329–341. mirocha cj, christensen cm, nelson gh. (1971). ‘F-2 (Zearalenone) estrogenic mycotoxin from Fusarium’ in Microbial Toxins, Vol. VII, in Kadis S, Ciegler A, Ajl SJ (eds). Academic Press, New York, pp 107–138. mittwoch u, burgess amc, baker pj. (1993). Male sexual development in a sea of estrogens. Lancet, 342: 123–124. muller m, rabenoellina f, balaguer p, patureau d, lemenach k, budzinski h, bardcelo d, de alda ml, kuster m, delgenes jp, hernandez-raquet g. (2008). Chemical and biological analysis of endocrine-disrupting hormones and estrogenic activity in an advanced sewage treatment plant. Environmental Toxicology and Chemistry, 27(8): 1649–1658. nag sk, raikwar mk. (2008). Organic pesticide residues in bovine milk. Bulletin of Environment Toxicology, 80: 5–9. nakata h, kawazoe m, arizona k, abe s, kitano t, shimada h, li w, ding x. (2002). Organochlorine pesticides and polychlorinated biphenyl residues in foodstuffs and human tissues from China: status of contamination, historical trend, and human dietary exposure. Archives of Environmental Contamination and Toxicology, 43(4): 473–480. nilsson r. (2000). Endocrine modulators in the food chain and environment. Toxicologic Pathology, 28(3): 420–431. okubo t, yokoyama y, kano k, kano i. (2001). ER-dependent estrogenic activity of parabens assessed by proliferation of human breast cancer MCF-7 cells and expression of ERα and PR. Food and Chemical Toxicology, 39: 1225–1232. pulce c, lamaison d, keck g, bostvironnois c, nicholas j, descotes j. (1991). Collective human food poisonings by clenbuterol residues in veal liver. Veterinary and Human Toxicology, 33(5): 480–481. purdom ce, hardiman pa, bye vj, eno nc, tyler cr, sumpter jp. (1994). Estrogenic effects of effluents from sewage treatment works. Chemistry and Ecology, 8: 275–285.

124

Endocrine-disrupting chemicals in food

qin l, wang p, kaneko t, hoshi k, sato a. (2004). Estrogen: one of the risk factors in milk for prostate cancer. Medical Hypothesis, 62: 133–142. rissato sr, galhiane ms, dealmeida mv, gerenutti m, apon bm. (2007). Multi-residue determination of pesticides in honey samples by gas chromatography-mass spectrometry and application in environmental contamination. Food Chemistry, 101(4): 1719–1726. roscoe v, lombaert ga, huzel v, neumann g, meleition j, kitchen d, kotello s, krakalovich t, trelka r, scott pm. (2008). Mycotoxins in breakfast cereals from the Canadian retail market: a three year survey. Food Additives and Contaminants, 25(3): 347–355. schafer ks, kegley se. (2002). Persistent toxic chemicals in the US food supply. Journal of Epidemiology and Community Health, 56(11): 813–817. schecter a, papke o, tung kc, staskal d, birnbaum l. (2004). Polybrominated diphenyl ethers contamination of United States food. Environmental Science Technology, 38(20): 5306–5311. schoeters g, hoogenboom r. (2006) Contamination of free-range chicken eggs with dioxins and dioxin-like polychlorinated biphenyls. Molecular Nutrition and Food Research, 50: 908–914. scippo ml, gaspar p, degand g, brose f, maghuin-rogister g. (1993). Control of illegal administration of natural steroid hormones in urine and tissues of veal calves and in plasma of bulls. Analytica Chimica Acta, 275: 57–74. schollenberger m, muller hm, rufle m, terry-jarra h, suchy s, plank s, drochner w. (2007). Natural occurrence of Fusarium toxins in soy food marketed in Germany. International Journal of Food Microbiology, 113: 142–146. shaw i, thomson b, cressey p. (2004). ‘Xenoestrogens’, in Watson DH (ed) Pesticide, Veterinary and Other Residues in Food, Woodhead Publishing, Cambridge. shaw ic. (1999). Regulation and monitoring of pesticides in Indonesia. Pesticide News, 44: 10. shaw ic, burke e, suharyanto f, sihombing f. (2000). Residues of p,p′-DDT and hexachlorobenzene in human milk from Indonesian women. Environmental Science and Pollution Research, 7: 75–77. snyder sa, villeneuve dl, snyder em, glesy p. (2001). Identification and quantification of estrogen receptor agonists in wastewater effluents. Environmental Science Technology, 35(18): 3620–3625. sumpter jp. (1995). Feminized responses in fish to environmental estrogens. Toxicological Letters, 82–83: 737–742. tuohy pg. (2003). Soy infant formula and phytoestrogens. Journal of Paediatric Child Health, 39: 401–405. unep. (2001). The Stockholm Convention on persistent organic pollutants. United Nations Environmental Programme. http://chm.pops.int/ Accessed: 27 September 2008. weiss j, papke o, bergman a. (2005). A worldwide survey of polychlorinated dibenzop-dioxins, dibenzofurans, and related contaminants in butter. Ambio, 34(8): 589–597. welshons wv, rottinghaus ge, nonneman dj, dolan-timpe m, ross pf. (1990). A sensitive bioassay for detection of dietary estrogens in animal feeds. Journal of Veterinary Diagnostic Investigation, 2: 268–273. woodwell gm, worster cfj, isaacson pa. (1967). DDT residues in an east coast estuary: a case of biological concentration of a persistent insecticide. Science, 156: 821. wwf-uk. (2003). ContamiNATION, the results of WWF’s biomonitoring survey. http://www.wwf.org.uk/filelibrary/pdf/biomonitoringresults.pdf Accessed: 21 May 2008.

Endocrine-disrupting chemicals

125

wwf. (2006). Chain of contamination. The food link. http://www.wwf.org.uk/ filelibrary/pdf/contamination.pdf Accessed: 20 May 2008. yang n, matsuda m, kawano m, wakimoto t. (2006). PCBs and organochlorine pesticides (OCPs) in edible fish and shellfish from China. Chemosphere, 63(8): 1342–1352. zhong w, xu d, chai z, mao x. (2003). 2001 survey of organochlorine pesticides in retail milk from Beijing, P. R. China. Food Additives and Contaminants, 20(3): 254–258.

6 Surveillance of endocrine-disrupting chemicals in foods M. Rose, Food and Environment Research Agency, UK

Abstract: Surveillance of foods for endocrine-disrupting chemicals is essential for the risk assessment process, specifically to enable dietary intake estimates to be made. Key aspects of surveillance including survey design, sampling and the process of making consumer exposure estimates are covered, and a summary of studies resulting in estimates of dietary exposure is included. Consideration is given to dealing with ‘at-risk’ groups, high-level consumers, exposure to mixtures, time trends and future research activities. Key words: dietary exposure estimates, survey design, sampling, risk assessment, time trends.

6.1

Introduction: importance of surveillance of endocrinedisrupting chemicals in food and the environment

When endocrine disrupters enter the aquatic and/or terrestrial food chain, they have the potential to alter the normal function of the hormonal systems of humans and wildlife. These compounds can be naturally occurring or synthetic in origin and cover several chemical classes of compound, including natural products (coumestrol, genistein; Cornwell et al., 2004), some pesticides (fungicides, some herbicides including atrazine and insecticides including dieldrin, toxaphene, endosulfan, phenylphenol, DDT and metabolites, methoxychlor, vinclozolin; Beeson et al., 1999; Turusov et al., 2002), some medicines and veterinary medicines (e.g. hydroxyflutamide, nilutamide, tamoxifen, diethylstilbestrol, oral contraceptives such as ethynylestradiol; Skegg, 1995; Menard et al., 2000), commercial and/or industrial chemicals such as bisphenol A (European Commission, 2002), alkylphenols (e.g. nonylphenol; Fernandes et al., 2003), polychlorinated dioxins, furans and biphenyls (PCDD/Fs and PCBs; Startin and Rose, 2003), brominated

Surveillance of endocrine-disrupting chemicals in foods

127

flame retardants (BFRs; D’Silva et al., 2004) and brominated dioxins (PBDD/Fs; Fernandes et al., 2008), phthalate plasticisers (Latini, polycyclic aromatic hydrocarbons, 2005), (PAHs; Isobe et al., 2007) and some trace elements e.g. arsenic (Davey et al., 2007). Endocrine disruptors can be classed according to their mode of action: estrogens and anti-estrogens mimic or block the female hormone estradiol, whereas androgens and anti-androgens mimic or block the male hormone testosterone. They can all act at very low concentrations to alter embryonic development and sex differentiation. Much of the research done to date has focused on the effects of estrogens, which have a feminising effect, and less has been directed towards the androgens and anti-androgens (see Chapter 20).

6.2

Environmental risk assessment versus dietary exposure estimates

Estrogenic compounds contained in sewage effluent can be discharged into waterways, where they can have endocrine-disrupting effects on fish and aquatic wildlife (Sumpter and Jobling, 1995). There is also often a downstream intake of water which, after treatment, can be used as drinking water for humans. As a result of the effects widely reported in fish, many surveys have been conducted with respect to water quality, and in turn, some attention has been paid to the surveillance of fish. To investigate whether endocrine-disrupting effects (specifically estrogenic effects) were occurring in the marine environment, laboratory studies and extensive field surveys of river systems have been carried out using sensitive species of fish for monitoring purposes, e.g. the euryhaline flounder Platichthys flesus. Results of such studies carried out in the UK for example, have confirmed that several estuaries, particularly the Tyne, Tees and Mersey, were severely contaminated with estrogens (Matthiessen et al., 1998; Allen et al., 1999a,b). Effects can be manifested as vitellogenin (VTG) induction in males and the presence of intersex testes. With the exception of research on the impacts of tributyltin oxide (TBTO), research on endocrine-disrupting effects in marine invertebrates in the UK is limited. In the Firth of Forth, Scotland, there have been reports of a high prevalence of intersex harpacticoid copepods in the vicinity of sewage treatment works outfalls (Moore and Stevenson, 1991, 1994), but conclusive causal relationships are very difficult, if not impossible, to establish. Such surveys are designed to help with environmental risk assessment, and often species of fish and geographical location of sampling sites are not chosen with a direct objective to help with dietary exposure estimates – for example species of fish widely consumed is not an important consideration in designing the surveys. Also, the monitoring of a bio-indicator such as VTG measures total biological effect rather than concentrations of individual

128

Endocrine-disrupting chemicals in food

chemicals. This may be more relevant when looking at total capacity for endocrine disruption, but has the disadvantage that other possible more significant toxic effects (e.g. dioxins) will be missed, and it will not be known which classes of compounds are causing the problem. Other effects based bioassays are also available, e.g. the YES yeast-based estrogen assay and the CALUX ER assay (Lemmen et al., 2002). While all such assays have the advantage of covering a range of chemicals with similar modes of action, and are usually rapid and relatively easy to implement, they all have similar drawbacks to those mentioned above for VTG monitoring. Many of the aspects relating to the surveillance of foods for specific classes of endocrine-disrupting chemicals, are common to factors relating to many other classes of chemicals in foods. As such, many of the specific details discussed below are more widely applicable to other classes of chemicals that may be present in foods either as natural toxins, or as a result of contamination, as additives or as residues of compounds applied during production. Surveys may be carried out by regulatory authorities, the food industry, or others, such as consumer organisations, pressure groups or academics who may be interested in chemicals in food with a potential impact on health (food safety) or chemicals that would not normally be expected to be found although they have no known health impact (food quality). For industry, surveys of products manufactured can demonstrate due diligence in their production methods and show that every precaution is taken to ensure that products are safe and of good quality. Surveys can help to protect and inform consumers, judge the effectiveness of regulation and/or production methods, monitor trends and to assess risks by (i) protecting consumer safety (in this case with respect to exposure to endocrine disrupters from the diet), (ii) allowing consumers to make informed choices (from an ethical, environmental or health perspective), (iii) informing authorities in respect of need for policy position or regulation, (iv) assessing the effectiveness of current legislation and Codes of Practice, etc. as implemented, (v) monitoring trends, both in terms of concentrations and geographical location and over time, and (vi) in enabling consumer exposure assessments and dietary intake calculations to be made. Such exposure assessments, when robust, form an integral part of the risk assessment process (Fig. 6.1) (see Chapter 14).

Risk assessment • Hazard identification • Hazard characterisation • Exposure assessment (surveillance) • Risk characterisation

Fig. 6.1 The key stages in the risk assessment process (FAO/WHO, 1995).

Surveillance of endocrine-disrupting chemicals in foods

6.3

129

Survey design

Surveys can be relatively simple and directed towards particular food types for specific compounds of concern, for example phthalates in plasticwrapped food, or dioxins in fish, or they can be more complex and directed towards gathering data for population exposure estimates and can take the form of, for example, total diet studies, duplicate diet studies or market basket surveys. The total diet study (Harries et al., 1969; Buss and Lindsay, 1978; Peattie et al., 1983) categorises food into groups such as milk, eggs, fish, etc. based on their relevant importance in the diet. Individual samples of retail food from these groups are purchased at regular intervals from locations within an area (typically a country) selected at random but weighted according to population density. They are prepared for consumption, and combined to form single composite samples for each of the defined food groups. Such an analysis of a year’s combined samples ensures that a large, diverse base of individual samples sourced from across the region of study are included, and that they represent average dietary habits. This has been done for example with dioxins and PCBs (FSA, 2003; Fernandes et al., 2004). Although the total diet study approach as described here for the UK is used in some other countries/regions, e.g. the Basque region in Northern Spain (Urieta et al., 1991), most countries use market basket approach as described below, although to add confusion this is sometimes referred to as a total diet study. A duplicate diet study involves volunteers who collect a replicate sample of all food consumed over a given period. The samples from each individual are made into a single pooled sample, and these can be used to look at intake variation between individuals. Care has to be taken with such studies because the behaviour of some volunteers can change for the duration of the study and the diets collected do not always represent a typical diet. This approach was used for example in a study where the levels of phytoestrogens, inorganic trace elements, natural toxicants and nitrates were measured in summer and winter vegetarian duplicate diets (Clarke et al., 2003). There is another approach to estimating dietary intake by using biomarkers. This is where the compounds of interest are monitored in some form of biological tissue taken from the exposed population. This can include urine, where typically a collection is made over a given time period, and, assuming the compound or a known metabolite is excreted at a concentration that can relate to intake (either quantitatively or by factoring in a known half-life, etc.), then exposure can be estimated. This approach was used in a study to monitor exposure to phthalates using urinary biomarkers (Anderson et al., 2001). For some lipophilic compounds with long body residence half-lives, exposure can estimated from the measurement of human milk samples. One example of an ambitious study, the SUREmilk project funded by the UK Department of Health (DH), the Food Standards

130

Endocrine-disrupting chemicals in food

Agency (FSA), the Department of Environment, Food and Rural Affairs (DEFRA) and the Health and Safety Executive (HSE) (Surveillance of Residues in human milk: Pilot studies to explore alternative methods for the recruitment, collection, storage and management of an archive of breast milk samples; http://www.food.gov.uk/multimedia/pdfs/suremilkcontents. pdf), measured concentrations of phthalates, PCBs/organochlorines (OCs), metals and dioxins (Woolridge et al., 2004). Market basket surveys are where shoppers buy samples that reflect the commodities widely consumed by a population, but reflect a spot check rather than a robust time-averaged estimate of intake (Freijer et al., 2001; Baumann et al., 2002; Bakker et al., 2003). When planning a survey, consideration should be given to the reasons behind the survey. The questions to ask and points listed below focus on surveys conducted in order to conduct consumer exposure assessments and dietary intake calculations as listed below. Different considerations may be useful for environmental risk assessment and environmental quality monitoring. • Objectives: Why is the survey being conducted? If the survey contains quantitative measurements, what statistics can be generated from the data? Consider how many samples or measurements will be needed to confirm or detect differences (e.g. time trends, geographical locations, food and product types, brands, population groups and sub-groups, e.g. children, vegetarians) and ensure that sub-populations are assigned large enough sample sizes to detect any differences that may be of interest or importance. • Sample units: At what level should the sample be taken? Should samples be combined into a composite for analysis? For example an estimate of exposure to chemicals in milk can be obtained from taking packs of milk from different supermarkets and other retail outlets, and combining to reflect market share. Results from this composite would reflect average concentrations, and exposure can be calculated by using estimates of consumption (typically average and high (97.5 percentile). However, if differences between milk produced at different farms are to be measured, then sample units may be taken from the milk bulk tanks from the farms and analysed individually. If differences in milk from individual cows are important, then samples need to be taken on this basis. Is the compound(s) being surveyed likely to be distributed heterogeneously? Is any information available on its likely distribution within a single sample, among samples of the same batch, or among batches? Does the sample size or number of replicates for each brand need to reflect this potential variability, or does a large sample need to be taken and blended prior to analysis? • Sample size: It is important to decide how many sample units need to be selected. Are usable results relating to parts of the overall sample

Surveillance of endocrine-disrupting chemicals in foods

•

•

•

•

•

131

needed as well as overall results? For example, are results needed on a regional basis or just a national basis; are results needed for several different types of food? What size of sample is required, e.g. weight/volume/ number of individual units? Geographical coverage: A decision on geographic coverage is required for the survey to ensure that it is adequately representative for the intended use of the data. A characteristic may be temporally variable; statistical advice may be needed with respect to the sampling period required so that the results are not affected by seasonal or short-term phenomena. Consideration should be given to whether a food may be imported or home/locally produced. Is there regional variability among products? In certain circumstances it may be acceptable only to sample in some areas, or to adjust sampling by region to reflect differences in regional consumption habits. Will local practices need to be taken into account? Local knowledge can often provide useful advice at the planning stage. Choice of sampling locations: The range of retailers to be covered needs consideration. In retail surveys, samples should as far as possible be selected from a range of major and smaller retail outlets as well as independent retail outlets, and should be chosen and weighted to reflect market share, unless weighting is needed to ensure a statistically adequate number of minority categories are included. Timing: When is the sample to be taken? Is the product available only at certain times of the year? Do products tend to originate from different places at different times of the year? The level of analytes found may vary with the season. Origin of samples: Will imported food be included? If so, when and where will samples be taken? Consignments may enter the UK at different places and come from different countries at different times of the year, or may cease at certain times of the year. Previous or existing knowledge: Make use of available information, for example market share data, pilot study, archived survey data, scientific literature, a basis for sampling can be taken from existing data sets such as lists of food retailers held by local authorities.

6.4

Sampling

6.4.1 Sample purchase A plan should be produced containing details from the survey design above. It should clearly state the details required; e.g.: • numbers of samples – it may also be a good idea to include contingency samples where the food is fragile and easily damaged in transit (e.g. eggs or samples in glass containers);

132

Endocrine-disrupting chemicals in food

• sample size, volume and/or weight; • whether the sample can or should be made up from separate packs/ sub-samples with the same batch number; • product types; • countries of origin; • type of outlet to purchase from; • market share – may need to state numbers of samples from specific stores to ensure market share is properly reflected; • the timeframe over which samples are to be obtained; • avoidance of cross- or other contamination; • how to store the samples prior to packaging and transportation; • the timing of sample purchase and supply to the analytical laboratory; • if the samples are unpackaged, e.g. milk to be used in the manufacture of a milk product, the intended use should be stated, e.g. cheese manufacture; • sample logging details; and • whether the store should be told that a sample has been taken and why – this is especially important if the samples are taken for regulatory purposes that could result in follow-up action or if the results and locations from which samples were taken are to be made public.

6.4.2 Recording sample details Exact sample details should be recorded, in order to ensure the existence of an audit trail and to provide details that may be required in any followup action. Details to consider should include: • unique number allocated to sample when purchased, ideally to link with potential details to be maintained by the receiving laboratory in order to maintain traceability; • product type and name (sample purchasers should be prompted to record those product attributes that might materially affect the product in a relevant way – for example, whether sampled dairy products are UHT-treated, pasteurised or unpasteurised may be relevant in a microbiological survey); • date of purchase; • place of purchase; • sample condition at time of purchase; • brand (i.e. supermarket own brand or other brand name); • ingredients; • nutrition information; • description of product; • size of packet purchased; • packaging type, e.g. vacuum packed, modified atmosphere, multipack, etc.;

Surveillance of endocrine-disrupting chemicals in foods

133

• suggested serving size for food; • the name and/or business name and an address or registered office of either or both of: • the manufacturer or packer, or • the retailer, • ‘best before’ or ‘use by’ date, • batch/lot code, • country of origin, as described on the retail pack, • any deviations from the sampling instructions document.

6.4.3 Sample handling It is important that samples are properly packaged between taking the sample and arrival at the analytical laboratory to ensure that there is an adequate audit trail (in the case of formal samples) and in order to avoid cross- or other contamination; damage during transportation; deterioration of samples of products; loss of unstable contaminants or growth and/or changes to the microorganisms present in the sample due to temperature changes. This generality is important for all surveys, but particularly for environmental contaminants such as some of the endocrine disrupters. An assessment of the stability of samples and analytes, and, where necessary, of sample temperature during transit before the main collection phase should be undertaken. Necessary arrangements should be made to ensure that storage between sample purchase, transport and subsequent storage at the analytical laboratory prior to analysis does not prejudice the reliability of the final results. Receipt and storage of the samples must be considered and chosen to avoid misleading results when the sample is analysed. This may involve keeping the samples at ambient conditions, freeze dried (lyophilised) or frozen at a specified temperature, typically −20 °C or −70 °C. Steps should be taken to ensure the integrity of the samples if there is any likelihood that they may arrive at the laboratory outside working hours; consideration as to how to store samples received in these circumstances so as to maintain and protect the stability of the samples must be given. The condition of samples on receipt at the analytical laboratory should be recorded. The process of collection and preparation of samples is represented in Fig. 6.2.

6.4.4 Sample preparation Samples should be thoroughly homogenised before analysis. Where formal samples are involved that may result in follow-up action, there may be legal requirements involved. These may include, for example, the subdivision of a sample. One sub-sample should be used for the required analysis/analyses and one for confirmation as necessary. The other two sub-samples should be stored under suitable conditions, as determined by stability studies, one

Sub-sample for reference/courts

Storage (at conditions required for sample and analyte stability)

Sub-sample for supplier

Sub-sample for supplier

Yes

Storage (at conditions required for sample and analyte stability)

Sub-sample for reference/courts

Are there separate screening and confirmation methods?

Confirmation

Analysis

Sub-sample for confirmatory analysis

Analysis

Sub-sample for analysis

Fig. 6.2 Collection and preparation of samples for monitoring endocrine-disrupting chemicals in food (modified from FSA, 2006).

Confirmation

Analysis

Sub-sample for analysis

No

Homogenisation of sample and preparation of composites where required

Storage (at conditions required for sample and analyte stability)

Sample

Surveillance of endocrine-disrupting chemicals in foods

135

for analysis by the company, retailer, etc., if required, and one for retention in case the result is disputed.

6.5

Surveillance programmes

6.5.1 Current surveillance efforts and findings Surveillance efforts for human dietary exposure are rarely targeted towards endocrine disrupters per se, but more usually towards the specific compounds or classes of compounds within this group. In other words groups of compounds are often targeted for surveillance because of other toxic properties or wider public interest (e.g. dioxins, pesticides). As such it may be viewed that the monitoring for endocrine disrupters is more by accident than design. This is contrary to surveillance of endocrine disrupters as indicators of aquatic or environmental quality where total endocrinedisrupting activity is frequently measured (see earlier). Most surveillance for endocrine-disrupting compounds has been carried out in Western Europe and in North America. It is clear from many reports that levels of dioxins in foods and in the population are decreasing, although this decline is slowing down and even levelling out (e.g. Furst, 2001). Several review articles have described levels of these compounds in food and in the general population (e.g. Startin and Rose, 2003). Similar trends are seen in data-sets from various places around the globe such as New Zealand (Bates et al., 1999), the US (Anderson et al., 1998), Norway (Johansen et al., 1996), Sweden (Norén and Meironyté, 2000) and Canada (Dewailly et al., 1996). Some of these studies also included data for pesticides and other compounds. PBDE concentrations are found to be considerably higher in North America than in Europe and the rest of the world and levels have been shown to be rising in recent years although there is some evidence that levels may be levelling off as restrictions on production come into force throughout the world. Continued use of deca-brominated diphenyl ether (BDE) (209) mean that this generalisation does not hold for this congener, although there is less data available for BDE 209 which is more difficult to measure (D’Silva et al., 2004). Concentrations of dichlorodiphenyl-trichloroethane (DDT) and metabolites are decreasing in the Northern Hemisphere, but still prevail at relatively high concentrations in some populations (Norén and Meironyté, 2000). Levels of dieldrin followed the same pattern. In Mexico City, however, DDT and metabolites in human milk are still relatively high (Torres-Arreola et al., 1999), and even higher levels are found of DDT, lindane and hexachlorobenzene where these chlorinated insecticides are still used. Phthalates are used primarily as plasticisers in materials, including those used for food packaging, and can be incorporated at up to 40% in some plastics. Various estimates exist of exposure to phthalates (Blount et al.,

136

Endocrine-disrupting chemicals in food

2000a,b). There have been numerous studies looking into exposure via the diet as a result of the use of food packaging containing phthalates (e.g. Anderson et al., 2001). Exposure to atrazine has been assessed primarily by measuring metabolites in urine or less frequently in blood (e.g. Barr et al., 1999). Above are just some examples of assessments that have been made of exposure to endocrine disrupters. This is discussed in more detail in other chapters of this book and is intended here only to give examples of work that has been done and different approaches that have been taken and to give an indication of the diversity of considerations that need to be given for the different classes of chemical compounds that fall into the endocrinedisrupter category. There is a vast difference in the amount of data available according to chemical class. This is partly due to the way these compounds are regulated. Pesticides and veterinary medicines are widely surveyed in statutory monitoring programmes designed to ensure good practice in the farming and food production industries and also to ensure food safety. Similarly there are limits within Europe for dioxins and PCBs in food and there is a monitoring scheme whereby each Member State is obliged to conduct an annual survey. Different limits exist in terms of maximum limits, which if exceeded should result in food being withdrawn from sale, action limits, which if exceeded should trigger investigation by the authorities to identify sources and to take action in order to remove or control the problem and target levels, which are long-term goals with values set to ensure that the population is not exposed at levels that could exceed the acceptable daily intake (ADI). These aspects and a wider discussion of regulatory limits in force for different classes of compounds including pesticides, veterinary medicines, etc. and covering different parts of the world are covered elsewhere within this book (see Chapter 18). In addition to surveillance to ensure regulatory compliance and enforcement of regulations, other surveys are conducted on an ad hoc basis to assess dietary exposure. These can be for the same compounds monitored for regulatory purposes or for compounds not included in such programmes, for example the phytoestrogens. Such surveys can include surveys of particular food types for compounds of concern, for example phthalates in plastic wrapped food, or can take the form of total diet studies, duplicate diet studies, biomarkers of intake or market basket surveys as discussed earlier. In summary, there is no cohesive programme of monitoring for endocrine disrupters and data need to be assimilated from the various surveillance schemes that are in place for different compounds and in different locations. A future need could be identified to coordinate such activity into a cohesive and harmonised programme. A summary of exposure levels and established tolerable daily intakes and reference doses for endocrine-disrupting compounds is given in Table 6.1, but more details about levels of endocrine disrupters in foods and their

Bisphenol

•

•

Leaching of the chemical from cans, plastic bottles and dental products

•

Bisphenol A intake (mg/kg bw/day): 1.2 in Japan and 1.6 in USA

• Total DDT intake in Japan (mg/kg bw/day): 0.006

• TDI: 10 mg/kg bw/day (ESCSFd) • RfD: 50 mg/kg bw/day (USEPAe)

• TDI: 0.5 mg/kg bw/day (USEPAe, RIVMf)

Mainly though foods, e.g., leafy and root vegetables, farm meat, fish and poultry Occupational exposures through inhalation and dermal contact

•

•

•

DDT/DDE

•

TDI: 2 pg TEQ/kg bw/day (COT, 2001). Tolerable weekly Intake of 14 pg WHO-TEQ/kg bodyweight (SCF, 2001); 70 pg WHO-TEQ/kg bw/month (JECFA, 2001)

•

Mainly though foods especially milk (breast and dairy), fish, and other meats Occupational exposure occurs mainly via the inhalation and dermal routes

•

PCBs/dioxins

Dioxinc intake in Japan (pg TEQ/kg bw/day): 1.49 in 2002 (0.52 and 0.97 for PCDD/Fs and PCBs, respectively) Dioxin (exc. PCBs) intake in Germany: approximately 2 in the late 1980s, approximately 1 in 1994–1995 Dioxinc intake in UK: 7.2 in 1982; 0.9 in 2001 (Fernandes et al., 2004)

TDIa or RfDb

Estimated exposure levels

Exposure sources

Exposure source and level of endocrine-disrupting compounds with regulation levels (adapted from Yang et al., 2006)

Endocrinedisrupting chemical

Table 6.1

NP intake in Germany (mg/ day): Adult, breast-milk-fed infant and formula-fed infants, 7.5, 0.2 and 1.4, respectively

Isoflavone intakes (mg isoflavone aglucone/day): 0.76 (USA), 1.6 (UK), 14.88 (Korea), 17 (Australia), 25.4 (China), 61.4 (Singapore), 70 (Japan).

•

•

Food and water from fields spread with sewage sludge containing alkylphenols Foods with aseptic packing Occupational exposures through inhalation and dermal contact

Mainly though foods Isoflavone: legumes, lentils, soybean Coumestans: young sprouting legumes Lignans: cereals, linseed, fruit and vegetables

•

• •

f

e

d

c

b

a

•

•

• •

TDI, tolerable daily intake. RfD, reference dose. Dioxins, PCDDs, PCDFs and coplanar-PCBs. ESCSF, European Commission Scientific Committee on Food. USEPA, US Environmental Protection Agency. RIVM, National Institute of Public Health and the Environment, the Netherlands.

Phytoestrogens

Alkylphenols

DEHP intake (mg/kg bw/day): 3–30 Particularly 8.9–9.1 (infant), 19 (toddler), 14 (child), 8.2 (teen), 5.8 (adult) in Canada

•

DEP: dermal and inhalative exposures DEHP: oral exposures

•

Phthalates

•

Estimated exposure levels

Exposure sources

Continued

Endocrinedisrupting chemical

Table 6.1

• TDI: 5 mg/kg bw/day for NP • 13 mg/kg bw/day for NPE

• TDI: 37 mg/kg bw/day for DEHP (ESCSFd) • RfD: 20 mg/kg bw/day for DEHP (USEPAe)

TDIa or RfDb

Surveillance of endocrine-disrupting chemicals in foods

139

effects on consumers can be found in the chapters discussing particular compound classes elsewhere in this volume. 6.5.2

Interpreting the results of surveillance programmes – follow-up action As mentioned above surveys are carried out for a variety of purposes, including compliance with statutory limits. These limits are often based on ADIs for a particular chemical which is established based on toxicological studies. It is very unlikely that any endocrine-disrupting effects are used when establishing the ADI; this will almost always be derived from no or lowest observable adverse effect level (NOAEL or LOAEL) associated with more conventional toxicological end points rather than a hormonal effect. Where they are a statutory obligation and legal enforceable limits exist, action should be taken where limits are exceeded. This regularly happens with, for example, pesticide residues or with residues of veterinary medicines. In terms of food safety and risk assessment, this action may be disproportionate across the class of endocrine-disrupting compounds, since some of the limits are derived from the maximum that should be present if good agricultural practice is followed rather than from a toxicological or health protection viewpoint. Environmental contaminants are regulated from a toxicological view but surveillance programmes are much less extensive. There is also the added complication with the three tiers of levels for dioxins (PCDD/Fs) and dioxin-like PCBs as discussed above; these being (i) maximum limits at which products should be removed from sale; (ii) action limits which trigger further investigation to identify and control the source of contamination; and (iii) target limits which are long-term goals designed to ensure that the population is unlikely to exceed any ADI and values that should become the maximum limits at some point in the future when pollution control measures have all taken effect and environmental levels of these compounds have reduced sufficiently to make these values workable. There is some dispute about some endocrine-disrupting compounds such as the phytoestrogens as to whether they have a net beneficial or harmful effect or whether this can depend upon the status of the individual with respect to age, sex, etc. (see Chapter 17). Surveys for these compounds are not performed on a regulatory basis, but available data can be used with epidemiological evidence to help with this debate, although we are still at a stage where much more data are needed to develop firm conclusions.

6.6

Dietary intake calculations and consumer exposure estimates

The role of surveillance in risk assessment from the point of view of gathering data for dietary intake calculations and consumer exposure estimates was discussed earlier, and is subject of greater discussion elsewhere in this

140

Endocrine-disrupting chemicals in food

book. It is usual practice to look at average (or, e.g., high 97.5 percentile) concentrations of compounds in food, to combine this with average (or high) consumption data and to calculate dietary intake in a deterministic fashion. This will typically be done for average and high-level consumers in order to formulate advice and to take into account the precautionary principle. While a comprehensive definition of the precautionary principle was never formally adopted by the EU, a working definition and implementation strategy for the EU context has been proposed (Fisher et al., 2006): Where, following an assessment of available scientific information, there are reasonable grounds for concern for the possibility of adverse effects but scientific uncertainty persists, provisional risk management measures based on a broad cost/benefit analysis whereby priority will be given to human health and the environment, necessary to ensure the chosen high level of protection in the Community and proportionate to this level of protection, may be adopted, pending further scientific information for a more comprehensive risk assessment, without having to wait until the reality and seriousness of those adverse effects become fully apparent.

There are also concerns about the application of the precautionary principle and this has been widely debated (Goldstein and Carruth, 2004; Sunstein, 2005). This relatively simple approach does not include uncertainty and does not take into account the true distributions of the levels of chemicals found, or consumption patterns. It also cannot include uncertainty in the toxicological assessment for the chemicals measured. There are a limited number of examples in the literature where probabilistic methods have been used to include distributions and uncertainty (Harris, 2000; Smith et al., 2002), and there is a move towards the use of these methods more generally (Ferrier et al., 2006).

6.6.1 ‘At risk’ groups When designing surveys, it is important to give consideration with respect to ‘at risk’ groups. This can include children, where (i) there are still critical development windows that may be encountered and (ii) the relative intake of chemicals on a body weight basis will be high because of low body weight and high energy requirement. Pregnant or nursing mothers may be considered at risk because of the fetal or infant exposure to chemicals resulting from the mothers’ food intake. The elderly may be particularly susceptible to certain diseases, or have less resilience to illness. Those on special diets, for example vegetarians, will have a different intake of chemicals from the average consumer. Also there are groups such as those who grow their own fruit and vegetables, or those living in certain communities, e.g. in a fishing village where diets may differ significantly from the norm. Certain geographical locations can contain high levels of organic pollutants. These can

Surveillance of endocrine-disrupting chemicals in foods

141

include areas close to industry or the source of pollutant, but also the arctic regions where these compounds can be transported over time due to atmospheric transport associated with global movement. As such Inuit or Lap communities living in or close to the Arctic where there are high levels of persistent organic pollutants combined with a diet high in marine fat may need special advice or protection. Food produced from the flood plains of contaminated rivers may contain elevated levels of some contaminants (Lake et al., 2005). When making exposure intake estimates, it is important to consider these groups, and where there is a chance that there may be higher than normal intakes for a particular group of the community, revised estimates should be made in line with the precautionary principle.

6.6.2 Mixtures Humans are exposed to chemicals, including endocrine disrupters, as mixtures, either as a result of consuming a varied diet, or because a variety of compounds may be found in a single food type. An important aspect of identifying the causes of endocrine disruption is to assess whether, and how, chemicals implicated as causative agents may act together when present in mixtures. The importance of considering such combination effects has been shown with mixtures of non-estrogenic aquatic toxicants, and with estrogenic agents. These studies suggested that hazards existing in mixed exposure situations (as usually occurs) may go undetected when assessments focus exclusively on the effects of single chemicals (Payne et al., 2001; Silva et al., 2002; Rajapakse et al., 2004). Different classes of endocrine disrupters exhibit various modes of action and it is unclear as to whether effects may be additive, synergistic or antagonistic even when two similar compounds or classes of compounds are present. For example, estradiol binds to the estrogen receptor and has been shown to be estrogenic in its mode of action. In fish, vitellogenin is produced following exposure of the fish to estrogenic compounds. It is known that exposure can lead to reproductive health effects. Compounds with antiandrogenic activity such as linuron bind to the androgen receptor but no vitellogenin is produced. There is nevertheless resulting feminisation in terms of reproductive health effects. Exposure to both compounds have similar effect, i.e. it results in feminisation, but only one is via an estrogenic mechanism of action. Both estrogenic effects and anti-androgenic effects and any other endocrine-disrupting behaviours are important to consider when the topic of endocrine disrupters is taken as a whole. The question remains whether we can tell which substances will interact and with which endocrine systems. It is unlikely that we can predict all potential interactions for a single substance; chemicals can behave in a ‘promiscuous’ manner, that is they exhibit multiple modes of action (Yang et al., 2006). The steroid ethinyl estradiol is a potent estrogen and a weak anti-androgen, i.e. the same chemical with different modes of action. Hormones have complex,

142

Endocrine-disrupting chemicals in food

concentration dependent but non-linear responses. The cross-reactivity between androgens and estrogens could be a function of concentration and could also be related to the structural similarity between molecules. There is a need to understand the molecular mechanisms of disruption from the gene to the response. An approach may be to use molecular technology such as microarray and to identify up/down-regulation of genes, or proteomics for the analysis of gene products. It is important to link genetranscript-protein information to organ and whole animal health effects.

6.7

Monitoring time trends

Surveys can be repeated at intervals designed to monitor changes of intake or exposure over time. It is important to note that food choices and diet may also vary with time on a population basis, and these changes should be factored in to any such survey. It is not always a case of simply repeating a survey after a given period, although this may be the best way to monitor if levels of a compound have changed over time in a particular food group due to changes in production methods, etc. Again, it is important to consider the questions that need to be answered and the objectives of the study in detail when planning the survey. Changes in legislation such as the introduction of limits can have an effect on production processes and monitoring that is undertaken as industry tries to comply. This can have an unexpected effect on typical concentrations at particular time points if new processes bring down concentrations of particular substances, or if particular food samples with high levels of compounds are selectively removed from the market. In addition to gathering data to ensure regulatory compliance, monitoring can be used to generate consumer exposure estimates and dietary intake calculations (e.g. Thomson et al., 2003).

6.8

Future trends

Many of the surveys reported in the literature on endocrine disrupters have been conducted because of other aspects of toxicology and concern relating to those chemicals (e.g. dioxins, pesticides) and there is little concerted or directed effort to the endocrine disrupters per se. There is currently only scarce information connecting exposures to endocrine-disrupting compounds with health outcomes in wildlife and even less for humans. Predominantly, research efforts have to date focused on compounds that persist and bioaccumulate in organisms and their environment. Only recently have efforts been directed at exposure studies of less persistent compounds and in the development of biologically-based assays, which enable more direct assessments of total exposure to endocrine-active compounds. Given the

Surveillance of endocrine-disrupting chemicals in foods

143

dynamic nature of the endocrine system, future efforts in the study of these compounds need more focus on the timing, frequency and duration of exposure to these chemicals, and to the effects that may result from exposure to mixtures. There is a need to develop improved methodologies for assessing dose– response relationships at environmentally relevant concentrations, rather than the levels often used in experiments that bear no relevance to the concentrations of these chemicals to which most individuals are exposed. It would also be useful to have more specific and sensitive biomarkers for detecting endocrine-mediated effects in individuals and populations. More monitoring over the longer term is needed in order to provide baseline data on population status. More international collaboration and cooperative research are required to assess the exposure and effects of endocrine disrupters on a more global basis. Monitoring of trends in relevant human health outcomes would be useful in order to provide information that is comparable across regions and over time. We should continue to identify new chemicals (persistent and non-persistent, naturally occurring and anthropogenic) that are the most likely candidates for endocrine-disrupting activity in populations at environmentally relevant concentrations.

6.9

References

allen y.t., scott a.p., matthiessen p., haworth s., thain, j.e. and feist, s.w. (1999a). Survey of estrogenic activity in United Kingdom estuarine and coastal waters and its effects on gonadal development of the flounder. Platichthys Flesus. Environ. Toxicol. Chem., 18: 1791–1800. allen y., matthiessen p, scott a.p., haworth s., feist s. and thain j.e. (1999b). The extent of oestrogenic contamination in the UK estuarine and marine environments – further surveys of flounder. Sci. Total Environ., 233: 5–20. anderson, h.a., falk, c., hanrahan, l., olson, j., burse, v.w., needham, l., paschal, d., patterson, d.j. and hill, r.h.j. (1998). Profiles of Great Lakes critical pollutants: a sentinel analysis of human blood and urine. The Great Lakes Consortium. Environ. Health Perspect., 106: 279–289. anderson, w.a.c., castle, l., scotter, m.j., massey, r.c. and springall, c. (2001). A biomarker approach to measuring human dietary exposure to certain phthalate diesters. Food Additives Contaminants, 18(12): 1068–1074. bakker, m.i., baars, a.j., baumann, r.a., boon, p.e. and hoogerbrugge, r. (2003). Indicator PCBs in foodstuffs: occurrence and dietary intake in the Netherlands at the end of the 20th century. RIVM report 639102025/2003 www.rivm.nl. barr, d.b., barr, j.r., driskell, w.j., hill, r.h.j., ashley, d.l., needham, l.l., head, s.l. and sampson, e.j. (1999). Strategies for biological monitoring of exposure for contemporary-use pesticides. Toxicol. Indu. Health, 15: 168–179. bates, m.n., buckland, s.j., ellis, e.k., garrett, n., needham, l.l., patterson, d.g., jr, turner, w.e., russell, d., wilson, n. and duncan, a. (1999). PCDDs and PCDFs in the serum of the nonoccupationally exposed New Zealand population. Organohalogen Compounds, 44: 17–21. baumann, r.a., den boer, a.c., groenemeijer, g.s., den hartog, r.s., hijman, w.c., liem, a.k.d., marsman, j.a. and hoogerbrugge, r. (2002). Dioxinen en

144

Endocrine-disrupting chemicals in food

dioxineachtige PCBs in Nederlandse consumptiemelk: trendonderzoek 1997–2001 [Dioxins and dioxin-like PCBs in Dutch milk for consumption: recent developments and research 1997–2001]. RIVM rapport 639102024/2002. www.rivm.nl. beeson, m.d., driskell, w.j. and barr, d.b. (1999). Isotope dilution high-performance liquid chromatography/tandem mass spectrometry method for quantifying urinary metabolites of atrazine, malathion, and 2,4-dichlorophenoxyacetic acid. Anal. Chem., 71: 3526–3530. blount, b.c., milgram, k.e., silva, m.j., malek, n.a., reidy, j.a., needham, l.l. and brock, j.w. (2000a). Quantitative detection of eight phthalate metabolites in human urine using HPLC-APCIMS/MS. Anal. Chem., 72: 4127–4134. blount, b.c., silva, m.j., caudill, s.p., needham, l.l., pirkle, j.l., sampson, e.j., lucier, g.w., jackson, r.j. and brock, j.w. (2000b). Levels of seven urinary phthalate metabolites in a human reference population. Environ. Health Perspect., 108: 979–982. buss, d.h. and lindsay, d.g. (1978). Reorganization of the UK total diet study for monitoring minor constituents of food. Food Cosmetics Toxicol., 16: 597–600. clarke, d.b., barnes, k.a., castle, l., rose, m., wilson, l., baxter, m.j., price, k. and dupont, s. (2003). Levels of phytoestrogens, inorganic trace-elements, natural toxicants and nitrate in vegetarian duplicate diets. Food Chem., 81: 287–300. cornwell, t., cohick, w. and raskin, i. (2004). Dietary phytoestrogens and health. Phytochemistry, 65: 996–1016. cot (2001). Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment. Statement on the Tolerable Daily Intake for dioxins and dioxinlike polychlorinated biphenyls. Dept of Health. Available at: www.food.gov.uk/ multimedia/pdfs/cot-diox-full. davey j.c., bodwell j.e., gosse, j.a. and hamilton, j.w. (2007). Arsenic as an endocrine disruptor: effects of arsenic on estrogen receptor-mediated gene expression in vivo and in cell culture. Toxicol. Sci., 98(1): 75–86. dewailly, e., ayotte, p., laliberté, c., weber, j.-p., gingras, s. and nantel, a.j. (1996). Polychlorinated biphenyl (PCB) and dichlorodiphenyl dichloroethylene (DDE) concentrations in the breast milk of women in Quebec. Am. J. Public Health, 86: 1241–1246. d’silva, k., fernandes, a. and rose, m. (2004). Brominated organic micro-pollutants – igniting the flame retardant issue. Crit. Rev. Environ. Sci. Technol., 34: 141–207. european commission (2002). Opinion of the Scientific Committee on Food: bisphenol A. http://europa.eu.int/comm./food/fs/sc/scf/out128_en.pdf. fao/who (1995). Application of Risk Analysis to Food Standards Issues, a Joint FAO/WHO Expert Consultation, Geneva, Switzerland, 13–17 March 1995. Available at: http://www.who.int/foodsafety/publications/micro/en/march1995.pdf. fernandes, a.r., costley, c.t. and rose, m. (2003). Determination of 4-octylphenol and 4-nonylphenol congeners in composite foods. Food Additives Contaminants, 20(9): 846–852. fernandes, a., gallani, b., gem, m., white, s. and rose, m. (2004). Trends in the dioxin and PCB content of the UK diet. Organohalogen Compounds, 66: 2053–2060. fernandes, a., dicks, p., mortimer, d., gem, m., smith, f., driffield, m., white, s. and rose, m. (2008). Brominated and chlorinated dioxins, PCBSs and brominated flame retardants in Scottish shellfish: methodology, occurrence and human dietary exposure. Mol. Nutr. Food Res., 52: 238–249. ferrier, h., shaw, g., nieuwenhuijsen, m., boobis, a. and elliott, p. (2006). Assessment of uncertainty in a probabilistic model of consumer exposure to pesticide residues in food. Food Add. Contam., 23: 601–615.

Surveillance of endocrine-disrupting chemicals in foods

145

fisher, e., jones, j. and von schomberg, r. (eds) (2006). Implementing the Precautionary Principle: Perspectives and Prospects, Cheltenham, UK and Northampton, MA, US: Edward Elgar. freijer, j.i., hoogerbrugge, r., van klaveren, j.d., traag, w.a., hoogenboom, l.a.p. and liem, a.k.d. (2001). Dioxins and dioxin-like PCBs in foodstuffs: occurrence and dietary intake in The Netherlands at the end of the 20th century. RIVM report 639102 022. Available at: www.rivm.nl. fsa (2003). Dioxins and Dioxin-like PCBs in the UK Diet: 2001 total diet study samples, Food Surveillance Information Sheet No. 38/03, Food Standards Agency, London. Available at: http://www.food.gov.uk/multimedia/pdfs/fsis38_ 2003.pdf. fsa (2006). Food Standards Agency Guidelines for Undertaking Surveys, November London, FSA. furst, p. (2001). Organochlorine pesticides, dioxins, PCBs, and polybrominated diphenylethers in human milk from Germany in the course of time. Organohalogen Compounds, 52: 185–188. goldstein, b.d. and carruth, r.s. (2004). Implications of the Precautionary Principle: is it a threat to science? Int. J. Occupational Med. Environ. Health, 17(1): 153–161. harries, j.m., jones, c.m. and tatton, j.o’g. (1969). Pesticide residues in the total diet oin England and Wales, 1966–1967. I – Organisation of a total diet study. J. Sci. Food and Agric., 20: 242–249. harris, c.a. (2000). How the variability issue was uncovered: the history of the UK residue variability findings. Food Add. Contami., 17: 491–495. isobe, t., takada, h., kanai, m., tsutsumi, s., isobe, k.o., boonyatumanond, r. and zakaria, m.p. (2007). Distribution of polycyclic aromatic hydrocarbons (PAHs) and phenolic endocrine disrupting chemicals in South and Southeast Asian mussels. Environ. Monitoring Assess., 135: 423–440. jecfa (2001). Joint FAO/WHO Expert Committee on Food Additives, Fiftyseventh meeting. Summary and Conclusions. Available at www.who.int/pcs/jecfa/ Summary57.pdf. johansen, h.r., alexander, j., rossland, o.j., planting, s., lovik, m., gaarder, p.i., gdynia, w., bjerve, k.s. and becher, g. (1996). PCDDs, PCDFs, and PCBs in human blood in relation to consumption of crabs from a contaminated fjord area in Norway. Environ. Health Perspecti., 104: 756–764. lake, i., foxall, c., lovett, a., fernandes, a., dowding, a., white, s. and rose, m., (2005). Effects of river flooding on PCDD/F and PCB levels in cows’ milk, soil, and grass. Environ. Sci. Technol., 39(23): 9033–9038. latini, g. (2005). Monitoring phthalate exposure in humans. Clin. Chim. Acta., 361: 20–29. lemmen, j.g., van den brink, c.e., legler, j., van der saag, p.t. and van der burg, b. (2002). Detection of estrogenic activity of steroids present during mammalian gestation using ERα and ERβ specific assays. J. Endocrinol., 174: 435–446. matthiessen, p., allen, y.t., allchin, c.r., feist, s.w., kirby, m.f., law, r.j., scott, a.p., thain, j.e. and thomas, k.v. (1998). Oestrogenic Endocrine Disruption in Flounder (Platichthys flesus) from United Kingdom Estuarine and Marine Waters. CEFAS Science Series Technical Report No. 107. menard, s., tagliabue, e., campiglio, m. and balsari, a. (2000). Re: Estimation of tamoxifen’s efficacy for preventing the formation and growth of breast tumors. J. Nat. Cancer Institute, 92: 943–944. moore, c.g. and stevenson, j.n. (1991). The occurrence of intersexuality in harpacticoid copepods and its relationship with pollution. Mar. Poll. Bull., 22: 72–74.

146

Endocrine-disrupting chemicals in food

moore, c.g. and stevenson, j.n. (1994). Intersexuality in benthic harpacticoid copepods in the Firth of Forth, Scotland. J. Nat. Hist., 28: 1213–1230. norén, k. and meironyté, d. (2000). Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere, 40: 1111–1123. payne, j., scholze, m. and kortenkamp, a. (2001). Mixtures of four organochlorines enhance human breast cancer cell proliferation. Environ. Health Perspectives, 109(4): 391–397. peattie, m.e., buss, d.h., lindsay, d.g. and smart, g.q. (1983). Reorganisation of the British Total Diet Study for Monitoring Food Constituents from 1981. Food Chem. Toxicol., 21: 503–507. rajapakse, n., silva, e., scholze, m. and kortenkamp, a. (2004). Deviation from additivity with estrogenic mixtures containing 4-nonylphenol and 4-tertoctylphenol detected in the E-SCREEN assay. Environ. Sci. Technol., 38(23): 6343–6352. scf (2001). Opinion of the Scientific Committee on Food, on the Risk Assessment of Dioxins and Dioxin-like PCBs in Food. 30 May 2001. Available at http://europa. eu.int/comm/food/fs/sc/scf/out90_en.pdf. silva, e., rajapakse, n. and kortenkamp, a. (2002). Something from ‘nothing’ – eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ. Sci. Technol., 36(8): 1751–1756. skegg, d. (1995). Risk assessment issues in breast cancer. Mutation Res., 333: 51–58. smith, g., hart, a., rose, m., macarthur, r., fernandes, a., white, s. and moore, d. (2002). Intake estimation of polychlorinated dibenzo-p-dioxins, dibenzofurans (PCDD/Fs) and polychlorinated biphenyls from salmon: the inclusion of uncertainty. Food Add. Contam., 19(8): 770–778. startin, j. and rose, m. (2003). Dioxins and dioxin-like PCBs in food. Chapter 3 in 2nd edition of Dioxins and Health, edited by A Schecter and T Gasiewicz. New York, Wiley. sumpter, j.p. and jobling, s. (1995). Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ. Health Perspect., 103(suppl 7): 173–178. sunstein, c.r. (2005). Laws of Fear: Beyond the Precautionary Principle. New York, Cambridge University Press. thomson, b., cressey, p.j. and shaw, i.c. (2003). Dietary exposure to xenoestrogens in New Zealand. J. Environ. Monitoring., 5: 229–235. torres-arreola, l., lopez-carrillo, l., torres-sanchez, l., cebrian, m., rueda, c., reyes, r. and lopez-cervantes, m. (1999). Levels of dichloro-diphenyltriachloroethane (DDT) metabolites in maternal milk and their determinant factors. Arch. Environ. Health, 54: 124–129. turusov, v., rakitsky, v. and tomatis, l. (2002). Dichlorodiphenyltrichloroethane (DDT): ubiquity, persistence, and risks. Environ. Health Perspectives, 110: 125–128. urieta, i., jalón, m., garcia, j. and galdeano, l. (1991). Food surveillance in the Basque Country (Spain) I. The design of a total diet study. Food Additives Contaminants, 13: 29–52. woolridge, m., hay, a. and renfrew, m. (2004). Final Report: SUREmilk study – Surveillance of Residues in human milk: pilot studies to explore alternative methods for the recruitment, collection, storage and management of an archive of breast milk samples. University of Leeds. Available at: http://www.food.gov.uk/ multimedia/pdfs/suremilkcontents.pdf and http://www.food.gov.uk/multimedia/ pdfs/suremilkmain.pdf. yang, m., park, m.s. and lee, h.s. (2006). ‘Endocrine disrupting chemicals: human exposure and health risks’. J. Environ. Sci. Health, Part C, 24(2): 183–224.

Surveillance of endocrine-disrupting chemicals in foods

6.10 Appendix: check plan for sampling (1) Obtaining a representative sample What is sampled? When is sampled? Where is sampled? How is sampled? Special equipment needed? Who samples? How many samples? Which size per sample? How are samples coded? Is a composite sample needed? Are samples directly contained for transport? Which containers are used? Can contamination and losses at containment be avoided? In situ analysis necessary? In situ sample cleaning needed? Geographical and meteorological data needed? Which are recorded and how?

(2) Sub-sampling and repacking In situ sub-sampling needed? In situ sub(sampling) repacking needed? Which containers are used? Are containers clean? Can contamination and losses be avoided? Are containers coded? In situ chemical preservation needed? In situ physical preservation needed? Which data are recorded and how?

(3) Transport and sample preparation Deadline for sample transport? Storage room available? Immediate preparation needed? Special conditions in lab needed? Is sample cleaning needed? Is sample division needed? Is sample reduction needed? Is drying needed? Is crushing needed? Is milling needed?

147

148

Endocrine-disrupting chemicals in food

Is sieving needed? Is mixing needed? Can contamination, losses and change in sample composition be avoided in all these steps? Is immediate analysis needed? Is the laboratory sample stable? Which data are recorded and how?

7 Advances in chromatography coupled to mass spectrometry-related techniques for analysis of endocrine disruptors in food J.-P. Antignac, F. Courant and B. Le Bizec, Ecole Nationale Vétérinaire de Nantes (ENVN), France

Abstract: Analytical methods dedicated to endocrine-disrupting chemicals (EDCs) in food require high performances both in terms of sample preparation and measurement. Focused on confirmatory methods based on chromatography coupled to mass spectrometry (MS), this chapter presents the current state-of-the art in the field from real cases studies dealing with various classes of estrogenic EDCs. It also illustrates potential benefits associated to the more recent innovations of these MS-coupled techniques including ultra-resolutive systems in liquid (fast-LC, UPLCTM) and gas chromatography (two-dimensional GC × GC), new types of ionisation techniques (photoionisation) and mass analysers (linear ion trap, OrbitrapTM), or emerging technologies (bioassay-MS coupling, metabolomics). Key words: endocrine disruptors, mass spectrometry, steroid hormones, persistent organic pollutants, xenoestrogens.

7.1

Introduction

7.1.1 Analysing endocrine-disrupting chemicals in food: the challenge The presence of various classes of endocrine-disrupting chemicals (EDCs) in the environment has been recognised for decades. Persistent organic pollutants (POPs), such as dioxins, polychlorobiphenyls and brominated flame retardants, have been measured in soil, sediment and air samples. Various other EDCs, including steroid hormones, pesticides, phthalates and alkylphenols, have been measured in sea, river and wastewater. Specific EDCs, for instance those that derive from naturally occurring phytoestro-

150

Endocrine-disrupting chemicals in food

gens or anthropogenic phytosanitary products, have been monitored in plant species for a relatively long period of time. In contrast, the identification of EDCs in food, especially in food products of animal origin, is a more recent development. At least two technical reasons help to explain this situation. One is that residual concentration levels of the target EDCs in these products appear to be at least an order of magnitude lower (10 to 1000 times or more) than in the environmental matrices. Hence there is a need for more efficient analytical methods to determine the presence of these pollutants at trace (μg g−1) or ultra-trace (ng kg−1) levels. The need for improved sensitivity is also strengthened by the current tendency to investigate not only acute toxicity but also low-dose and mixture effects of EDCs. The second technical difficulty relates to the complexity of most biological samples, far higher than that observed for many environmental matrices. The presence of numerous compounds in food samples (other exogenous compounds, endogenous substances, metabolites, etc.), with chemical structures sometimes very close to the compound of interest, make it much more difficult to analyse EDCs in food than in most environmental matrices. Analytical methods aimed at exploring EDC levels in food matrices therefore have to be extremely rigorous, both in terms of sample preparation procedures and measurement technique. This chapter focuses on confirmatory techniques based on chromatography coupled to mass spectrometry for the detection of EDCs. The sample preparation problem, which in most cases is a key factor in the efficiency of the method, is also described and examples discussed.

7.1.2 Measuring endocrine-disrupting chemicals in food: the mass spectrometry advantage Of the existing measurement methods, mass spectrometry (MS) occupies a particular place, primarily because of its specificity. Indeed, alternative methods for the targeted detection and quantification of EDCs in food (UV, fluorimetry and radioimmunological methods) may suffer from a lack of specificity and the risk of interference or ‘noise’ from other compounds. Mass spectrometry reduces this problem to a large extent, by providing one (and usually more) highly specific diagnostic signal(s), which lead to a more dependable and unambiguous interpretation. This advantage is recognised at the regulatory level, MS being considered worldwide as the most efficient confirmatory method for identifying and quantifying many classes of residues and contaminants. The second benefit of MS is undoubtedly its extremely high sensitivity, which in most cases is fully compatible with the expected concentration levels of EDCs in food (i.e. μg kg−1 or ng g−1). The possibility of developing extended multi-residue methods also represents a great advantage for MS-based techniques. Thus, the simultaneous monitoring of several tens

Advances in chromatography

151

(very commonly) or even hundreds (extreme cases, already proposed for instance for pesticides) of target compounds using the same MS equipment is achievable without any major problems. Nevertheless, it must be underlined that sample preparation is critical, and the procedure used must ensure the simultaneous presence of all monitored analytes in the final extract(s) injected into the MS system. Another benefit associated with MS is its suitability for determining chemical structure. In this area, the combined use of multi-stage mass spectrometry (MSn) for collecting structural information, and high-resolution mass spectrometry (HRMS) to get the exact masses of the observed ions (with possible corresponding raw formulae), could be of great interest. Possible applications might include the investigation of potential unknown EDCs revealed by biological tests performed on specific sample fractions, or the identification of EDC metabolites of interest. The final argument in favour of MS-based techniques is that numerous significant advances and innovations have been observed in this field during the last few years. Regarding chromatography, the development of ultraresolutive systems both in liquid chromatography (fast-LC, UPLCTM) and gas chromatography (two-dimensional GC × GC) clearly offers new analytical perspectives. In MS, improved and some totally new types of mass analysers (linear ion trap, OrbitrapTM), as well as the relative democratisation of time-of-flight (TOF) instruments, are powerful tools for performing multi-stage and high-resolution MS. In terms of ionisation, the introduction of new techniques, such as atmospheric pressure photoionisation (APPI), also contributes to the development and application of LC-MS related techniques. The main objective of this chapter is to illustrate to what extent these MS-coupled techniques, associated with their more recent developments and innovations, may be exploited in the field of EDC analysis in food. This topic will be covered in Sections 7.2 and 7.4, dealing with the GC-MS and LC-MS systems, respectively.

7.1.3

Measuring both the cause and the effect of endocrine disruption: the ultimate perspective One recent trend in MS concerns the development of global (i.e. nonspecific, untargeted) profiling and/or imaging strategies as innovative tools in response to the growing concern for integrated approaches (so-called ‘system biology’). In this frame, mass spectrometry-based metabolomics may be considered as a new method for large-scale and high-throughput sample characterisation. In the context of endocrine disruption, metabolomics is opening new perspectives, not only for detecting multiple EDCs in food, but also for assessing the potential physiological impacts of these EDCs in living organisms. This advanced use of MS techniques will be covered in Section 7.6. Other MS applications, dealing with the on-line coupling of bioassays and MS measurements, will also be discussed.

152

Endocrine-disrupting chemicals in food

7.2

Advances in gas chromatography – mass spectrometry-related techniques

7.2.1

Main application field of gas chromatography – multi-stage mass spectrometry-related techniques: which target endocrine-disrupting chemicals? GC is the separation method of choice for volatile and non-polar compounds. It is commonly used to analyse many POPs such as dioxins (polychlorinated dibenzo-p-dioxins/polychlorinated dibenzofurans, PCDD/ PCDF), polychlorobiphenyls (PCB) and polybromodiphenylethers (PBDE). The very high chromatographic resolution typically obtained in GC, due to the existence of numerous isobaric congeners in each family, represents an incomparable advantage. The extremely low concentration levels typically expected for many POPs in food products often requires the use of HRMS, with resolution equal or greater than 10 000, and using electromagnetic sector or TOF instruments to detect the target analytes. However, tandem or multi-stage MS with triple quadrupole or ion trap instruments may be used in some cases as an alternative to HRMS. A second, widely used application for GC-MS concerns moderately polar substances, pending the introduction of a chemical derivatisation step. In particular, GC-MS-related techniques are particularly suitable for measuring steroid hormones, as detailed in Section 7.3.1. Numerous GC-MSbased methods have been also proposed for the analysis of various classes of (mainly halogenated) pesticides (Stan et al., 2000; Amendola et al., 2002; Martinez Vidal et al., 2002; Arrebola et al., 2003; de Koning et al., 2003; Cajka and Hajslova, 2004; Hong et al., 2004; Zambonin et al., 2004; Patel et al., 2005; Fernandez Moreno et al., 2006; Garrido Frenich et al., 2006; Papadakis et al., 2006; Walorczyk and Gnusowski, 2006; Frenich et al., 2007). Depending on the expected target concentrations and required performances in terms of sensitivity and specificity, the detection may be achieved using GC-MS (simple quadrupole), GC-MSn (triple quadrupole or ion trap), or GCHRMS (TOF, electromagnetic sectors). 7.2.2

Recent advances in gas chromatography – multi-stage mass spectrometry-related techniques: what benefit for endocrine-disrupting chemical analysis? In terms of chromatographic separation, one incontestable advance of the last few years concerns the introduction of comprehensive two-dimensional chromatographic systems (GC × GC). By combining two different GC capillary columns (usually fitted with complementary stationary phases), these systems enable extremely efficient separations for complex mixtures, and clearly facilitate the interpretation of diagnostic ion chromatograms for compounds which were previously difficult to resolve. One practical example illustrating this advantage is detailed in Section 7.3.2, concerning the analysis of POPs.

Advances in chromatography

153

Regarding MS measurement, the introduction and relative democratisation of GC-TOF instruments is a significant advance. Indeed, these mediumresolution mass analysers (R = 10 000 to 15 000) have to some extent replaced the electromagnetic sector instruments historically used in dioxin and PCB analysis in food (Focant et al., 2003), and they are also used for pesticides (de Koning et al., 2003; Zrostlikova et al., 2003; Cajka and Hajslova, 2004) and alkylphenols (Moeder et al., 2006). Compared with older systems, the major advantages of TOF include ease of daily management and maintenance, an unlimited mass range and a rapid scanning speed. Moreover, some major fundamental and technological advances have occurred in TOFMS in recent years, and this trend appears to be continuing. All these recent improvements have led to high-speed GC-TOFMS that offers new and promising perspectives.

7.3

Case studies in gas chromatography – mass spectrometry-related techniques

7.3.1 Case study A: steroid hormones Steroid hormones are considered as reference substances in the context of endocrine disruption, as they are generally the most biologically active compounds. The activity of most estrogenic EDCs is usually expressed by comparing it with the activity of estradiol, with values commonly found between 102 and 105 times lower. The identification of natural (estradiol = E2, estone = E1, estriol = E3) and synthetic (ethinylestradiol = EE2) estrogens in food today represents a new application, even though these compounds have been studied for many years in other domains. In particular, numerous analytical methods and improvements originated from antidoping (Saugy et al., 2000) and chemical food safety laboratories in charge of the control of anabolic steroids as growth promoters in cattle (Le Bizec et al., 2004). In these two specific areas, GC-MS/MS and GC-HRMS after derivatisation remain the current approaches of choice for measuring steroid hormones. N-Methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) and/or trimethyliodosilane (TMIS) are the more commonly employed derivatisation reagents, leading to the transformation of both phenolic and aliphatic hydroxyl groups into trimethylsilyl (TMS) derivatives (Shareef et al., 2006). N-Methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide (MTBSTFA) has also been reported as a derivatisation reagent (Zuo et al., 2007). To increase the electron capture properties of estrogens, halogenated derivatisation reagents such as pentafluorobenzylbromide (PFBBr) can be used to good effect (Xiao et al., 2001; Courant et al., 2007), especially when combined with the negative chemical ionisation mode (NCI). As this reagent only protects phenolic hydroxyl groups, a second silylation reaction is usually employed in order to derivatise the remaining aliphatic hydroxyl groups (Fig. 7.1).

154

Endocrine-disrupting chemicals in food

(a)

(b) + PFB (pentafluorobenzylbromide)

H

H

(

H

HO

O

CH3

TMS

OH

CH3

+ BSTFA N,O-bis-(trimethylsilyl) trifluoro-acetamide

H F H

F

H

O

)

F

Estradiol

F Estradiol PFB,TMS-derivative

F

(c) GC-EI-MS, full scan 524

181

(d)

393

434 129

73

393 CH3 O

253 -PFB

TMS

H

F H

F

253

H

O

129

F

F F 181 107 50

100

150

434

340

159

225 200

281 250

313

300

365 350

400

450

m/z

500

(e) GC-NCI-MS, full scan 100

(f)

343.1

253 CH3 O

-TMSOH

TMS

H

343 F H

F

H

O

%

F

F F 181

344.5

345.6 253.2 205.6 331.6 181.0 160.4 311.5 251.5 254.5 269.4271.5281.4 213.2 224.9231.3 141.4146.1 193.7196.5 329.6 149.5 162.2 177.5 237.6 295.4 308.3 339.0 0 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320 330 340

347.5 350

m/z

Fig. 7.1 Synthesis and mass spectrometric behaviour of pentafluorobenzylbromide, trimethylsilylated (PFB, TMS) derivative of estradiol after electron impact (EI) or negative chemical ionisation (NCI).

Advances in chromatography

155

GC-MS with electron ionisation (EI) on single quadrupole instruments was historically the method most widely used to measure steroid TMSderivatives in biological matrices. At a typical 70 eV energy, the main resulting diagnostic signals produced in this mode correspond to the molecular ions [M]+· (appearing at m/z = 416 for E2, m/z = 414 for E1, m/z = 504 for E3, and m/z = 425 for EE2). Single ion monitoring (SIM) acquisition mode is then commonly employed to detect these ions with maximum sensitivity. This approach is suitable for relatively concentrated samples (limits of detection typically achieved are in the 0.1–1.0 μg kg−1 range), or when the limited sensitivity can be balanced by a pre-concentration step (Yang et al., 2006). However, this level of performances is usually insufficient for reliable measurement in most food products, in which estrogens are expected to be present at levels of around 0.01–0.1 ppb. Another important limitation of low-resolution GC-MS is the lack of structural information and signal specificity, these two factors largely preventing the unambiguous identification of the target analytes as well as their accurate quantification. Multidimensional mass spectrometry (GC-MS/MS with triple quadrupole or GC-MSn with ion trap) is therefore usually preferred for the difficult exercise of measuring estrogens at trace or ultra-trace levels in food (Kelly 2000; Impens et al., 2002; Fuh et al., 2004). Indeed, the introduction of tandem MS has revolutionised the field of residue analysis by permitting noticeable gains in specificity. When EI is used as the ionisation technique, the diagnostic signal more commonly monitored for E2 TMS derivative is then the single reaction monitoring (SRM) transition 416 > 285. This fragmentation of the molecular ion (precursor ion, m/z = 416) in the collision cell to produce the fragment measured (product ion, m/z = 285) leads to a significant decrease in background noise visible on the final ion chromatogram. The limits of sensitivity obtained may then be decreased by a factor of around five compared with that achieved in simple MS, i.e. reaching the 0.05 to 0.1 μg kg−1 concentration range. Moreover, the possibility of measuring more than one diagnostic signal per target analyte, i.e. monitoring two or three product ions in multiple reaction monitoring (MRM) mode, meets the regulatory requirements in terms of unambiguous identification and confirmation (2002/657/EC decision for Europe). This specificity advantage is also highly beneficial for quantification performance. When PFBBr is used as a derivatisation reagent, GC-NCI-MSn may also be a powerful approach for measuring the estrogens PFB,TMS-derivatives. Indeed, NCI ionisation, associated with improved electron capture properties provided by the introduction of halogenated atoms in the molecular structure, leads to a sensible gain in term of signal specificity (i.e. reaching the 0.01–0.05 μg kg−1 range). In this case, the observed base peak corresponds to the ion [M-PFB]− (appearing at m/z = 343 for E2, m/z = 269 for E1, m/z = 431 for E3, and m/z = 367 for EE2), which concentrates the diagnostic signal of the target analytes under a single MS peak, with resulting improved sensitivity (Fig. 7.1). On the other hand, this ionic species exhibits

156

Endocrine-disrupting chemicals in food

a very high stability in the gas phase, leading to significantly reduced fragmentation in MS/MS at conventional collision energies (i.e. 20–40 eV). Thus, the diagnostic signal commonly monitored in this mode is the pseudoMRM transition [M-PFB]− > [M-PFB]−. As a consequence, this approach may be considered extremely suitable for high-sensitivity screening purpose and quantification, but not really for unambiguous identification purposes (Fig. 7.2). An alternative to multidimensional MS for improving signal specificity may be to use medium or high-resolution equipment. In this case, interference in the expected ion chromatogram is removed by monitoring the exact masses of the target analytes. However, it must be noted that, because the elemental composition of steroids is based on CxHyOz chemical structures, no natural ‘mass defect’ is available for really efficient HRMS. In other words, the advantage of HRMS is quite limited when TMS derivatives are monitored after EI ionisation, the resulting performance being in this case quite similar to that obtained with GC-MS/MS. HRMS may, however, became extremely powerful when PFB,TMS derivatives are monitored after NCI ionisation. In this case, the mass defect induced by the introduction of fluorine atoms leads to a significant mass clean-up on the expected diagnostic chromatograms, with immediate increased facility in terms of interpretation at ultra-low concentration levels. This HRMS measurement may be achieved both on TOF or electromagnetic sector instruments (Fig. 7.2). In conclusion, there are various efficient analytical methods based on GC-MS techniques for measuring estrogenic steroids in food. Depending on the expected concentration levels of the target, both GC-MS/MS (triple quadrupole, ion trap) and GC-HRMS (electromagnetic sector, TOF) after EI or NCI ionisation may be used. The final performances of these methods are affected by the nature of the analysed matrix and the efficiency of the sample preparation procedure. Measuring estrogenic steroid hormones at the ng kg−1 (ppt) level demands extremely powerful purification of the sample extracts whatever measurement technique is used. Sample preparation is not the main topic of this chapter, but there are numerous procedures for extracting and purifying steroid hormones from complex biological matrices. Generally, at least two or three successive steps are the minimum, with combinations of liquid/liquid or liquid/solid (depending on the nature of the sample) extraction and complementary reverse and normal solid phase extraction (SPE) purification (Marchand et al., 2000; Blasco et al., 2007). An additional enzymatic hydrolysis step is included for food products of animal origin (milk, egg, meat) in order to deconjugate potential glucuronide and/or sulphate metabolites. This general scheme is usually suitable for water and other medium complex matrices such as muscle or milk. For more complex and difficult matrices, such as liver or eggs, a final stage of purification using semi-preparative HPLC may be highly beneficial.

Advances in chromatography

157

GC-NCI-MS/MS (qQq, SRM, 343>343) T6 050406019 100

19.14 2451 37446

α -E2 18.58 1333 21009

β -E2 E2

Egg (yolk) sample a-E2 = 150 ng kg-1 (ppt) b-E2 = 240 ng kg-1 (ppt)

%

3

Time 18.00 18.20 18.40 18.60 18.80 19.00 19.20 19.40 19.60 19.80 20.00 20.20

GC-NCI-HRMS (EB, R=5000, SIM, 343.21) 18334 β -E2

1.5 α -E2

1.0 0.5

Time 13.7 13.8 13.9 14.0 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 15.00 (min)

GC-NCI-MS/MS (qQq, SRM, 343>343) 13d 130106013 100

21.79 3311 43866

α -E2 β -E2 22.52 781 10420

Milk sample a-E2 = 40 ng kg-1 (ppt) b-E2 = 20 ng kg-1 (ppt)

%

4

Time 21.20 21.40 21.60 21.80 22.00 22.20 22.40 22.60 22.80 23.00 23.20 23.40

GC-NCI-HRMS (TOF, R=5000, SIM, 343.21) 100 000 95 000 90 000 85 000 80 000 75 000 70 000 65 000 60 000 55 000 50 000 45 000 40 000 35 000 30 000 25 000 20 000 15 000 10 000 5 000 0

α -E2

β -E2 E2 17.10

17.20

17.30

17.40

17.50

17.60

17.70

17.80

17.90

18.00

Fig. 7.2 Diagnostic ion chromatograms obtained for 17α-estradiol (α-E2) and 17β-estradiol (β-E2) PFB,TMS derivatives after negative chemical ionisation (NCI) for real egg and milk samples and acquisition on various GC-MS/MS (triple quadrupole, QqQ) and GC-HRMS (electromagnetic sector, BE, or time-of-flight, TOF) instruments.

158

Endocrine-disrupting chemicals in food

In parallel with these historical GC-MS approaches, a recent trend in measuring steroid hormones has been the use of LC-MS methods. LC-MS/ MS with triple quadrupole or ion trap (Lopez de Alda et al., 2003; RodriguezMozaz et al., 2004; Shao et al., 2005b; Xu et al., 2006; Blasco et al., 2007; Van Poucke et al., 2007), and LC-HRMS with TOF, OrbitrapTM or Fourier transform ion cyclotron resonance (FTICR) (Nielen et al., 2007) have been used. The shift is partially explained by recent technical progress in this field. The efficiency of some ionisation interfaces has been notably improved, e.g., atmospheric pressure chemical ionisation (APCI) and especially APPI (Singh et al., 2000; Guo et al., 2006), along with overall optical and electronic performance (ion transmission and detection). Another advantage of LC-MS should be the possibility of directly measuring highly polar steroid glucuronide and/or sulphate phase II metabolites (Isobe et al., 2003; Antignac et al., 2005b; Saudan et al., 2006; Strahm et al., 2007). The exact determination of the chemical forms of target estrogenic substances in samples may be highly relevant in terms of bioavailability or toxicology. However, LC-MS techniques remain particularly prone to ion suppression and other matrix effects that may affect measurement stability and quantification, especially at very low concentrations. In this case, achieving very high purification of the injected extracts may be even more crucial than for GC-MS.

7.3.2 Case study B: halogenated persistent organic pollutants POPs are lipophilic chemicals that originate from human activities, and include various classes of environmental contaminants such as dioxins (PCDD/PCDF), PCBs, PBDEs and polyaromatic hydrocarbons (PAH). PCBs have been recognised as EDCs for a long time. Indeed, the first scientific evidence of endocrine disruption induced by environmental chemical pollutants was obtained from wildlife observations (feminisation, abnormal mortality of offspring, etc.) in relation to contamination with PCB. Historically, the preferred MS method for measuring POPs in food products is GC-HRMS on electromagnetic sector instruments. Owing to their very low polarity, no derivatisation reaction is needed prior to GC analysis. The most commonly used ionisation technique for POPs is EI, generally at a lower energy level than that typically used for many other small organic molecules, i.e. 35–45 eV rather than 70 eV. This specificity ensures the desired limited fragmentation of the target analytes, which helps avoid interference phenomenon between the different monitored congeners when chlorine atoms are lost. In this mode, the more commonly monitored diagnostic ions correspond to the molecular ions [M]+· as well as their isotopic contribution, which is increased due to the presence of halogen atoms (35/37Cl or 79/81Br). This conventional GC-HRMS approach still represents the current standard for a majority of laboratories undertaking dioxin and PCB analysis in

Advances in chromatography

159

food. The method is largely recognised at the regulatory level as extremely efficient and powerful in terms of sensitivity, specificity and accurate quantification. The limits of detection generally achieved for most PCBs measured in food products such as fish, meat, milk or fat, are in the 0.1 to 1 ng kg−1 (ppt) range. However, it must be stressed that sample preparation is again a crucial element in POP analysis, mainly because of the very high difficulty in managing lipophilic extracts. This huge analytical challenge is solved by using preliminary Soxhlet or accelerated solvent extraction (ASE), followed by a relatively laborious multi-stage purification on activated silica, fluorite and/or charcoal columns, this procedure commonly requiring about three days of treatment. More recently, the same approach was extended to other POPs, some of which are also considered to be EDCs. Brominated flame retardants such as PBDE are examples of emerging pollutants characterised by a chemical structure naturally well adapted to GC-HRMS measurement (Covaci et al., 2003; Cariou et al., 2005). In this case however, external contamination phenomenon for several PBDE congeners add another source of analytical difficulty, and impose very drastic quality control constraints (Papke et al., 2004). There are numerous efficient methods based on GC-HRMS for a wide range of EDCs belonging to the extended family of halogenated POPs. For all these compounds, the presence of halogenated atoms in the chemical structure clearly justifies the use of HRMS, owing to the resulting mass defect and final efficient clean-up visible on the diagnostic ion chromatograms. To some extent, TOF instruments may replace electromagnetic sector equipment to advantage in these methods (Focant et al., 2003, 2005), based on the qualities of this type of mass analyser, as already discussed. GC-MS/MS with triple quadrupole or GC-MSn with ion trap instruments may also represent alternative measurement systems for POP-related EDCs in food (Derouiche et al., 2007). However, owing to the usually relatively poor and/or non-specific fragmentation of these compounds, HRMS is still considered as the reference approach in this field. Despite their global suitability, conventional GC-MS approaches do have some drawbacks, including for example the limited separation power of GC, peak co-elutions and limited acquisition rates. To overcome these limitations, fractionation may be included in the analytical protocol to efficiently separate analytes into sub-classes. After fractionation, several parallel injections, analyses and data processing steps have to be performed separately before the results are recombined to produce a final report. But the increasing number of analytes of interest (numerous congeners of dioxins, PCB, PBDE, toxaphenes, and more) increases the number of fractionation processes. To eliminate laborious multiple fractionations, a more versatile analytical tool is needed for multi-group analytical procedures. A major recent advance in the field of GC-MS techniques is undoubtedly the introduction of a new generation of chromatographic separation. This

160

Endocrine-disrupting chemicals in food

method, comprehensive GC × GC, has many advantages over conventional, one-dimensional GC. The most interesting are increased peak capacity, increased sensitivity and selectivity, independent retention processes in the two dimensions, and identification of each substance by two independent retention times. General reviews of GC × GC technology and applications are available elsewhere (Marriott and Shellic, 2002; Ong and Marriott, 2002), so only a few main features will be mentioned here. The two coupled GC columns have to be linked to some interface or modulator that is capable of either sampling or collecting the effluent from the first column and periodically introducing it to the second column. The interface must perform this task of sampling/collecting followed by injection at a rate that allows the original first dimension separation to be preserved. Primary columns typically used in these systems are generally 15–30 m × 0.25 mm, with a film thickness in the range of 0.25–1.0 μm. These columns allow for the generation of peak widths in the first dimension of the order of 10–20 s, which are required for typical modulation periods, in the range of 3–6 s. The first dimension columns typically have a non-polar stationary phase, either a 100% polydimethylsiloxane phase, or a 95/5 methyl/phenylsiloxane phase. The second dimension separation must be very fast and performed with a stationary phase that is different from that used in the primary column. Typical dimension ranges for secondary columns are 0.5–1.5 m in length ×0.1 mm. Thinner film thicknesses are generally used (i.e. 0.1–0.25 μm) to increase the separation efficiency but not the retention strength. The direct consequence of such a high-resolution system is the multitude of chromatographic peaks to monitor, each of them presenting an extremely narrow width, therefore, the MS detector must have a very high scanning rate. TOF-MS appears to be perfectly suited to accommodate the fast measurement of multiple compounds (van Deursen et al., 2000; Dallüge et al., 2002). This non-mass-scanning device allows collection of all ions at the same time, offering valuable comprehensive mass analysis. Additionally, because all ion fragments represent the same time point on the chromatographic peak, there is no concentration bias, unlike scanning mass spectrometers. TOF, therefore, also provides spectral continuity over the entire GC peak. This important feature allows mass spectral deconvolution of overlapping peaks if the fragmentation pattern is different, which reduces the chromatographic resolution requirements and decreases the time taken for analysis. Deconvoluted ion current (DIC) can thus be used to solve chromatographic co-elution problems, and TOF-MS therefore acts as an analyte separation tool (Fig. 7.3). These new GC × GC-TOFMS approaches clearly represent a valuable analytical tool for efficient identification and quantification of complex mixtures of EDCs belonging to various classes of POP (Fig. 7.3). In particular, this technique has been proven useful for the separation and unambiguous measurement of several PCB, PBB and brominated dioxins and furans

404+186+188+221+256+220+222+290+292+294+254+256+324+326+328+210+288+290+360+362+322+324+326+359+394+396+358+360+362+383+335+397+428+429+430+432+392+394+386+427+429+431+463+464+465+468+424+426+428

(a) BDE-153 BDE-85 CB-157 CB-156

2nd Dimension

2.6

1.6

CB-110 Dieldrin

CB-167 CB-128 CB-158 CB-138 4,4’-DDT BDE-28 CB-118 CB-105

BB-153 CB-189 BDE-99 BDE-100 BDE-47 BDE-66 BDE-154 CB-170 CB-194 Mirex

CB-209

CB-87 CB-99

CB-180 CB-172 CB-196, CB-203 CB-201 CB-101 CB-177 CB-66 CB-183 CB-74, Heptachlor epoxide CB-187 CB-178 β-HCH CB-44 DDE CB-153 t-Nonachlor CB-146 γ-HCH CB-28 Oxychlordane 2,4’-DDT CB-49 CB-149 CB-52 CB-18 CB-151 HCB

0.6 952

1452

1952 1st Dimension

(b)

2452

2952

(c)

90 000

1

1.55

2nd Dimension

80 000 70 000 60 000 1

2

50 000

2

1.45

1.35

2

40 000

1 1

30 000 20 000

CB-206

CB-195

1 2

1.25 2

Time 1.15 (seconds) 1788 1792 1792 1796 1796 1800 1800 1780 2 0 2 0 2 0 2 Time (seconds) 237+261+263+265+343+345+347+380+316+318

1800

Fig. 7.3 (a) Comprehensive GC × GC–TOFMS chromatogram obtained for 38 polychlorobisphenyls (PCB) + 11 persistent pesticides + 8 polybromodiphenylethers (PBDE) + 1 polybromobiphenyl (PBB). (b,c) A closer look at the region of the chromatogram where DDE (1) and dieldrin (2) elute. The chromatogram has been reconstructed using the sum of the characteristic ions of the two species. (b) The cluster corresponding to the ‘slices’ that can be recombined to produce the GC × GC contour plot shown in (b) and (c). Reproduced from Focant J-F, Sjodin A, Patterson DG. Qualitative evaluation of thermal desorption-programmable temperature vaporization-comprehensive two-dimensional gas chromatographytime-of-flight mass spectrometry for the analysis of selected halogenated contaminants. Journal of Chromatography A 2003;1019:143–156.

162

Endocrine-disrupting chemicals in food

that may co-elute using standard methods (Focant et al., 2003, 2004a,b, 2005). Comprehensive GC × GC has also been used to analyse pesticides (Zrostlikova et al., 2003; Khummueng et al., 2006) and nonylphenols (Moeder et al., 2006).

7.4 7.4.1

Advances in liquid chromatography – mass spectrometry-related techniques

Main application field of liquid chromatography – multi-stage mass spectrometry-related techniques: which target endocrine-disrupting chemicals? LC is the separation method of choice for relatively to extremely polar or ionic species (small molecules). LC-MS techniques are widely applicable in EDC analysis. Examples of LC-MS methods have been described for naturally occurring compounds such as phytoestrogens and mycotoxins (Section 7.5.1), as well as for various environmental contaminants including pesticides and phthalates (Section 7.5.2), alkylphenols (Benomar et al., 2001; Jahnke et al., 2004; Carabias-Martinez et al., 2006; Shao et al., 2005a, 2007) and other surfactants (Gonzalez et al., 2007). For these classes of substances, important advantages include the direct solubility compatibility with mobile phases typically used for HPLC systems coupled with MS (usually reversed phase systems with aqueous/alcoholic mixtures), and the fact that derivatisation steps are not required. Recent innovations in LC-MS techniques mean that this approach may be also envisaged for more lipophilic compounds. The introduction of photo-ionisation (APPI) is a typical example of a recent technology that offers new and promising perspectives for LC-MS, because of its efficiency with a range of compounds inaccessible to conventional atmospheric pressure ionisation techniques, i.e., electrospray ionisation (ESI) and APCI (Robb et al., 2000; Hanold et al., 2004). In particular, LC-APPI-MS/MS has been proposed as an alternative to GC-EI-MS/MS and/or GC-EIHRMS for measuring steroid hormones (Guo et al., 2006), PCB and PBDE (Debrauwer et al., 2005). In the latter case, the new LC-MS approach considerably reduces the thermal degradation phenomenon observed in GC for several highly brominated PBDE congeners (octa-, nona- and decabrominated diphenyl ether (BDE)). However, it must be stressed that the sensitivity achieved to date using this technique remains noticeably inferior to that obtained using conventional GC-MS approaches. Further work will help determine the suitability of this technique for measuring EDCs in food. Another application where LC-MS techniques can overcome GC-MS limitations is in the analysis of macromolecules (carbohydrates, protein, nucleic acids). While not directly related to EDC analysis, this possibility may be of interest for specific investigations. Detection of EDC–receptor

Advances in chromatography

163

complexes, or characterisation of EDC-DNA adducts (Embrechts et al., 2003), are examples of such innovative applications.

7.4.2

Recent advances in liquid chromatography – multi-stage mass spectrometry techniques: what benefit for endocrine-disrupting chemical analysis? Recent trends in LC separation are towards ultra-resolutive systems, matching the move towards comprehensive GC × GC. The main technological parameter for advanced LC separation efficiency is the reduction in particle size of the stationary phase. Following basic LC theory, the number of theoretical plates is much higher for small particle sizes (i.e. <2.0 μm) than for more conventional particle sizes (i.e. 3.5–5.0 μm). One reason for the lack of commercially available new generation HPLC columns until recently was the difficulty of production on an industrial scale. Two crucial elements are the repeatability of the desired distribution of particle diameter, and the reproducibility of the column filling process especially for HPLC columns with reduced internal diameters (i.e. 1.0–2.1 mm). But these industrial challenges have been largely solved in the past few years, and different systems are now available that permit LC resolutions quite similar to those provided by GC, i.e. characterised by chromatographic peak widths of only a few seconds, instead of the 15–30 seconds commonly obtained with conventional systems. The quality of separation achieved with the new systems greatly improves LC efficiency and is of special interest for analysing chemical residues in complex biological matrices, because of the reduced risk of co-elution with interfering compounds. Another direct consequence of the reduced peak width is the immediate gain in terms of sensitivity, which is highly beneficial for analysing EDCs that occur at trace or ultra-trace concentrations. Last but not least, ultra-resolutive LC reduces the ion suppression commonly observed in conventional HPLC, which negatively impacts signal repeatability and quantification (King et al., 2000; Annesley, 2003; Avery, 2003; Mei et al., 2003; Antignac et al., 2005a). Structural elucidation studies of unknown EDCs and metabolites may also benefit from this advance, for instance due to the increased possibility of separating isomer compounds. The past few years have been without doubt the most exiting and profitable period in terms of LC-MS development since this coupling technique was first introduced. Significant advances include ionisation interfaces and mass analyser devices. The new possibilities offered by APPI are discussed later in this chapter. Some real improvements have also been introduced in more conventional API interfaces, i.e. ESI and APCI. Innovations in terms of source geometry, the simultaneous combination of different ionisation modes, i.e. ESI/APCI or APCI/APPI (Syage et al., 2004), and efficient mass calibration based on continuous or pulsed introduction of appropriate reference solutions, are some examples of optimisations. These technical

164

Endocrine-disrupting chemicals in food

and technological advances have significantly raised sensitivity and mass accuracy in the latest generation of LC-MS systems, mainly through improved efficiency in terms of ionisation, transmission and focalisation. With regard to new mass analyser features, the introduction and/or relative democratisation of optimised low resolution (linear 2D ion trap), medium resolution (TOF), and totally new ultra-high resolution (OrbitrapTM) mass filters have already been mentioned. The availability of hybrid systems combining the advantages of two different types of mass analyser (Q-TOF, LTQ-OrbitrapTM) should also be noted as a recent advance with potential interest for analysing residues and contaminants (including EDCs) in food (Bobeldijk et al., 2001; Hernandez et al., 2005). However, it must be recognised that these extremely costly instruments are as yet confined to a small number of specialised laboratories, and used for instance in the emerging field of ‘omics’ technologies, for instance metabolomic and MS imaging, which will be discussed in the final section of this chapter.

7.5 7.5.1

Case studies in liquid chromatography – mass spectrometry-related techniques

Case study C: naturally occurring contaminants (phytoestrogens, mycoestrogens, etc.) Among the large and increasing number of compounds recognised (or suspected) to be EDCs, naturally occurring substances such as phytoestrogens remain poorly investigated compared to molecules of industrial origin such as POPs, pesticides, bisphenol A and alkylphenols. However, while numerous papers have described the potential positive effects of phytoestrogens, for instance on menopause symptoms or cardiovascular diseases, others have highlighted their estrogenic properties and possible adverse effects on health, especially in the case of exposure at critical stages of development. Indeed, the chemical structures of phytoestrogens show marked similarities to the structure of estradiol, explaining their potential interaction with natural estrogen receptors (ERα and ERβ). Most of the existing methods for analysing phytoestrogens focus on vegetable-based matrices (Wang et al., 2002; Wilkinson et al., 2002; Wu et al., 2004; Antonelli et al., 2005). But owing to the significantly lower concentrations expected in food products of animal origin, analytical methods capable of detecting phytoestrogens in this second type of matrix are more scarce. The sensitivity and specificity of mass spectrometry coupling techniques presents a clear advantage. For instance, some LC-MS methods have been described for measuring these compounds in soy milk, bovine milk (Antignac et al., 2004) and breast milk. This focus on the milk matrix is fully justified by the particular vulnerability of new-born and young children to endocrine disruption.

Advances in chromatography

165

Owing to their chemical structures and associated physicochemical properties, phytoestrogen compounds (belonging to the isoflavones, lignans, coumestans or natural stilbens families) appear well suited to LC-MSn measurement. Indeed, their relatively high polarity (presence of two or more phenolic hydroxyl groups on the structure) leads to easier ionisation in either ESI or APCI mode. Thus, the base peak usually observed for phytoestrogens in LC-API-MS corresponds to an intense pseudo-molecular ion, i.e. [M+H]+ or [M—H]− in positive and negative ionisation modes, respectively. This property of phytoestrogens to be efficiently ionised in both positive and negative modes is remarkably helpful for confirmatory analysis (the number of possible diagnostic signals is increased), and specificity/sensitivity (background noise and interfering peaks are always reduced in the negative mode, so the signal to noise ratio observed on the target analyte ion chromatograms is usually increased in this mode). Phytoestrogens behave very interestingly in their MS/MS fragmentation pathways, and this can be exploited in LC-MSn (Antignac et al., 2003). The tandem mass spectra obtained in positive ionisation mode reveals the formation of a relatively high number of product ions, which corresponds to three main fragmentation pathways (Fig. 7.4). The first of these involves the cleavage of the hydroxyl (—OH) and/or methoxyl (—CH3OH) groups. The second corresponds to the cleavage of the carbonyl (ketone) groups with rearrangement leading to the loss of carbon monoxide (CO). The third and last was observed in particularly with isoflavones, and corresponds to typical retro-Diels–Alder (RDA) rearrangements linked to the favourable double bond configuration of the A and B rings. A significant number of product ions are also observed in the negative ionisation mode which correspond either to the same fragmentation pathways already observed in positive mode (cleavage of the ketone groups and RDA reactions) or to complementary fragmentations (Fig. 7.4). The large number of fragment ions produced after both positive and negative ionisation permits the development of efficient LC-MSn measurement methods based on MRM acquisition mode, as illustrated in Fig. 7.5. The possibility of recording more than 10 MRM transitions for each analyte greatly helps in identifying the target phytoestrogens, and is of special benefit in very complex food matrices (for example cereals or mixed vegetable-based baby food). It also helps in discriminating between potential isomer compounds, for example those belonging to the wide isoflavone family. In terms of equipment, this targeted identification and quantification approach for phytoestrogens may be achieved using either triple quadrupole or ion trap (3D or 2D) devices. However, other acquisition modes associated with a triple quadrupole, such as neutral loss scanning (for example, based on the CO loss) or precursor ion scanning (for example, based on RDA fragments), may also be useful, for example in identifying some phytoestrogen phase I metabolites with potentially estrogenic properties.

(a)

Daidzein: R=H, MW=254 - Genistein: R=OH, MW=270 OH + [H]+

R O

−H2O R O HO

+ [H]+ O R=H, m/z 237

Retro−diels−alder rearrangement R=H, m/z 255 O (RDAR) −CO R O H OH C R O H + [H]+ and + [H]+ O + HO III + [H] O R=H, m/z 227 R=H, m/z 137 m/z 119 HO HO

−CO R HO

(b)

−CO

H

H + [H]+ R=H, m/z 199 O

III

H + [H]+ H m/z 91

LC-ESI(+)-MS/MS, daughter scan [255] PE001-A20 7 (0.387) 100

Daughters of 255ES+ 1.44e5

152

[M+H]+

137

254

181

127

%

198

255

199

153

197

119 106 115 133 121 104 108 141145 125 138 148 111

183 227 169 157 212 225 228 237 242 252 162165 171175 187 191195 200 209 218 234 239 249 257 259

m/z

10 0 10 5 11 0 11 5 12 0 12 5 13 0 13 5 14 0 14 5 15 0 15 5 16 0 16 5 17 0 17 5 18 0 18 5 19 0 19 5 20 0 20 5 21 0 21 5 22 0 22 5 23 0 23 5 24 0 24 5 25 0 25 5 26 0

0

(c)

LC-ESI(–)MS/MS, daughter scan [253] PE001-B48 7 (0.387)

Daughters of 253ES7.10e3 [M–H]–

132

100

253

223 208 133 252

% 195

135 179 196 224 180 250 117 167 181 197 169 178 105 116 225 154 141 143 249 183 193 127130 206 159 151 171 165 255 257 177 210216221 232235241 109 121 199 139 145

104

m/z

10 0 10 5 11 0 11 5 12 0 12 5 13 0 13 5 14 0 14 5 15 0 15 5 16 0 16 5 17 0 17 5 18 0 18 5 19 0 19 5 20 0 20 5 21 0 21 5 22 0 22 5 23 0 23 5 24 0 24 5 25 0 25 5 26 0

0

Fig. 7.4 Main fragmentation pathways of isoflavone phytoestrogens after positive electrospray ionisation (a), and typical tandem mass spectrum observed for daidzein after positive (b) and negative (c) electrospray ionisation (ESI). The numerous and complementary product ions produced in ESI(+) and ESI(−) permit a very high confidence level to be reached for unambiguous identification in complex biological matrices.

Daidzein-d3 (I.S.) 2: Daidzeine-d3 (EI) 181205029 Sm (Mn, 1x1) 6.73 100 4168 24883

Daidzein (2.4 ppb)

F1 256 > 211 2.56e4 Area, height

4: Daidzeine 181205029 Sm (Mn, 1x1) 100

6.78 371 2262

%

9 181205029 Sm (Mn, 1x1) 6.78 100 338 2075

%

Genistein (4.5 ppb)

F1 253 > 208 6.59e3 Area, height

5.00

7.50

10.00

Time

Equol (189.9 ppb)

F1 253 > 91 2.80e3 Area, height

F1 241 > 121 4.80e4 Area, height

%

F1 241 > 119 4.25e4 Area, height

%

1

7.50

10.00

Time

Glycitein (28.7 ppb) 9: Glyciteine 181205029 Sm (Mn, 1x1) 7.07 100 2042 11508

7.50

10.00

Time

7: Formononetine 181205029 Sm (Mn, 1x1) 8.95 100 9821 61728

F1 267 > 252 6.25e4 Area, height

1 181205029 Sm (Mn, 1x1) 8.99 100 2373 14465

1 5.00

F1 252 > 223 1.51e4 Area, height

5.00

7.50

10.00

Time

Time

10.00

181205029 Sm (Mn, 1x1) 100

F1 283 > 268 1.22e4 Area, height

10.01 516 3157

5 181205029 Sm (Mn, 1x1) 100 10.05 177 1012

F1 268 > 239 1.90e3 Area, height

%

7.50

10.00

12.50

Time

30 7.50

181205029 Sm (Mn, 1x1) 6.97 100 33899 200684

F1 297 > 253 2.01e5 Area, height

10.00

12.50

Time

Coumestrol (2.9 ppb) 14: Coumestrol 181205029 Sm (Mn, 1x1) 8.95 100 345 1951

F1 267 > 211 2.68e3 Area, height

%

0 181205029 Sm (Mn, 1x1) 6.97 100 24015 146206

F1 297 > 107 1.48e5 Area, height

%

%

7.50

Biochanin A (2.5 ppb)

Enterolactone (32.1 ppb)

F1 268 > 239 1.90e3 Area, height

5.00

%

%

5 181205029 Sm (Mn, 1x1) 7.11 100 256 1237

57

8: BiochanineA

11: Enterolactone F1 283 > 268 1.22e4 Area, height

%

32

5.00

%

5.00

F1 269 > 63 1.07e3 Area, height

%

%

1 181205029 Sm (Mn, 1x1) 7.21 100 6759 41740

48 181205029 Sm (Mn, 1x1) 7.88 100 81 407

Formononetin(6.3 ppb)

6: Equol 181205029 Sm (Mn, 1x1) 7.21 100 7659 47255

22

F1 269 > 133 1.26e3 Area, height

%

%

2

5: Genisteine 181205029 Sm (Mn, 1x1) 7.88 100 114 584

23 181205029 Sm (Mn, 1x1) 8.95 100 278 1549

F1 267 > 239 2.26e3 Area, height

%

0

Time 5.00

7.50

10.00

26 5.00

7.50

10.00

Time 12.50

Fig. 7.5 Example of typical diagnostic ion chromatograms obtained in LC-ESI(−)MS/MS (Micromass Waters QuattroLC triple quadrupole instrument) for eight phytoestrogen compounds in a bovine milk sample. Sample preparation procedure includes one liquid/liquid extraction (aceton/acetate buffer pH = 5.2) followed by an enzymatic hydrolysis (Helix pomatia) and a purification on two successive SPE cartridges (C18 and SiOH).

168

Endocrine-disrupting chemicals in food

Sample preparation is again of crucial importance in optimising LC-MSn measurement. For most food products of animal origin (milk, meat), the combination of liquid/liquid or liquid/solid extraction (using a solvent appropriate to the relatively difficult solubility of phytoestrogens, for instance acetone or methanol) with a two-step purification on reverse (C18) and normal (SiOH) SPE cartridges may be used. A preliminary enzymatic hydrolysis step (β-glucuronidase) may be added for biological matrices where the phytoestrogens are in conjugated forms (bovine milk, for example). For vegetable-based material, chemical hydrolysis (acidic conditions) may also be necessary to improve extraction recovery (strong interaction between the target analytes and the matrix), as well as to cleave the phytoestrogen precursors typically encountered in these kinds of products (glucosylated forms with acetyl and/or malonyl moieties). Mycoestrogens are substances belonging to the very large mycotoxin family. These compounds are microbial agents produced by fungi such as the Aspergillus, Penicillium and Fusarium species, and may contaminate feed and food (in particular cereal-based products). Among the numerous subclasses of mycotoxins, the trichothecene and zearalenone groups represent the main sources of molecules recognised as estrogenic EDCs. Trichothecenes have a hydroxylated tetracyclic sesquiterpene-based skeleton, and can be divided in four subtypes A, B, C and D, according to the nature of various substitutive groups. Within this family, molecules with a measurable estrogenic potency correspond to the more hydroxylated compounds, such as nivalenol (NIV) and deoxynivalenol (DON), which are type-B trichothecenes. Zearalenone (ZON) belongs to the resorcilic acid lactone (RALs) chemical group. After ingestion by an animal, ZON is metabolised into α-zearalenol (α-ZOL), β-zearalenol (β-ZOL), α-zearalanol (α-ZAL), βzearalanol (β-ZAL) and zearalanone (ZAN). Zearalenone-related compounds are involved in various reproductive disorders, including hyperestrogenism, pseudopregancy, enlarged mammary glands and abnormal lactation, in animal species such as domestic animals and pigs. From an analytical point of view, mycoestrogens are commonly measured in food and feed using LC-MS/MS triple quadrupole or ion trap equipment (Razzazi-Fazeli et al., 1999; Rundberget and Wilkins, 2002; van Bennekom et al., 2002; Berthiller et al., 2005; Sforza et al., 2005; Zollner and Mayer-Helm, 2006; Ren et al., 2007). Negative ionisation (preferably with APCI, but also ESI) is usually preferred for type-B trichothecenes and zearalenone compounds. Under these conditions, the main ion produced in the source for NIV, DON and ZON corresponds to the deprotonated molecule [M—H]−. In some cases, the loss of formaldehyde by internal collision induced dissociation (CID) may also be observed (leading to the ion [M— H—CH2O]−). When formic or acetic acid are present in the mobile phase, some [M+HCOO]− or [M+CH3COO]− adducts are also formed. When tandem MS is used to fragment these precursor ions, one or more product ions can be monitored for each target analyte without major difficulty.

Advances in chromatography

169

However, it should be mentioned that several authors have reported significant repeatability and quantification problems associated with the determination of mycoestrogen levels in complex food matrices with LC-MS/MS, owing to the existence of strong matrix effects (ion suppression phenomenon). In response to this classical source of analytical problems, the efficiency of sample preparation in achieving a fit-for-purpose purification of the extracts is again relevant. The latest generation of ultra-resolutive LC systems may help to minimise this problem.

7.5.2

Case study D: entropic contaminants (phytosanitary products, phthalates, etc.) Among the various classes of environmental chemicals, pesticide analysis has been of significant concern in recent years. Indeed, some phytosanitary products have been recognised as EDCs in specific exposure conditions. Historically, dichlorodiphenyltrichloroethane (DDT) and its main metabolite dichlorodiphenyldichloroethylene (DDE) were among the first insecticides identified as severe EDCs in wildlife. Herbicides such as alachlor and atrasine were also recognised as EDCs, as well as several fungicides, including vinclozolin. As mentioned above (Section 7.2.1), GC-MS or GC-MS/MS analysis is traditionally used, since many of the first generation pesticides were identifiable by GC. However, with increased industrial production and use of non-volatile, thermally labile and/or polar compounds, LC-MS techniques are nowadays the analytical approach of choice for analysing a wide range of phytosanitary products, as attested by the extended literature on this topic (Taylor et al., 2002; Ferrer and Thurman, 2003; Mol et al., 2003; Garrido Frenich et al., 2004; Granby et al., 2004; Ortelli et al., 2004; Sannino et al., 2004; Blasco et al., 2005; Ferrer et al., 2005; Hernandez et al., 2005, 2006; Garcia-Reyes et al., 2007; Hiemstra and de Kok, 2007; Pirard et al., 2007; Pizzutti et al., 2007; Soler and Pico, 2007; Venkateswarlu et al., 2007). The suitability of ESI and/or APCI ionisation techniques has been demonstrated for a large number of these substances (Thurman et al., 2001), and targeted MS measurement of the resulting diagnostic ions with simple quadrupole (SIM acquisition mode), triple quadrupole or ion trap equipments (SRM or MRM acquisition modes) does not represent any major difficulty (Blasco et al., 2004). As observed for mycotoxins, however, sample preparation is somewhat more problematic, as reported by many authors involved in analysing pesticides in complex biological samples. Some disturbing matrix effects have been reported. Optimising sample preparation may benefit from some of the recent technological advances already mentioned. For example, the use of fast-LC systems (Fig. 7.6) may reduce significantly the co-elution of target analytes with interfering compounds (Kovalczuk et al., 2006; Leandro et al., 2006, 2007). Another recent development is to use the mass defect which appears for most halogenated pesticides, using medium to high-resolution TOF instruments (Bobeldijk et al.,

170

Endocrine-disrupting chemicals in food

100

17

10−12

1

Used gradient (% methanol)

8,9 5+6 2

7

3 13

%

14,15 4

0

1.00

2.00

3.00

4.00

16

5.00

6.00

7.00

8.00

9.00 10.00 11.00

Time

Fig. 7.6 UPLC–MS/MS chromatogram of apple crude extract spiked with 17 (semi)polar pesticides (conc. 0.02 mg − 1 of each) based on the quantifying MS/MS transitions: 1 carbendazim, 2 thiabendazole, 3 carbofuran, 4 carbaryl, 5 linuron, 6 methiocarb, 7 epoxiconazole, 8 flusilazole, 9 diflubenzuron, 10 tebuconazole, 11 imazalil, 12 propiconazole, 13 triflumuron, 14 bitertanol, 15 prochloraz, 16 teflubenzuron, 17 flufenoxuron. Reproduced from Kovalczuk T, Jech M, Poustka J, Hajslova J. Ultra-performance liquid chromatography–tandem mass spectrometry: a novel challenge in multiresidue pesticide analysis in food. Analytica Chimica Acta 577: 8–17 (2006).

2001; Ferrer and Thurman, 2003; Ferrer et al., 2005). In spite of the numerous methods in this field (mainly still based on single stage GC-MS or LCMS), pesticide analysis in food has high potential for future progress, both in terms of sample preparation and MS measurement using improved LC resolution combined with tandem and/or high-resolution MS. These new approaches remain extremely costly at present, but a few, recently reported ultra-performance liquid chromatography (UPLC)-TOFMS methods enabling several tens (and even hundreds) of compounds to be measured in only a few minutes confirm that this technique is under development. Diesters of phthalic acid, commonly known as phthalates, are another category of environmental pollutants for which the question of endocrine disruption is growing. This class of chemicals is produced and used industrially on a very large scale, and released into the environment at a significant rate. These substances have many commercial uses, including as solvents, additives and plasticisers, so they appear in a vast range of consumables, such as personal care products (e.g. perfumes, lotions, cosmetics), paints, industrial plastics, and certain medical devices and pharmaceuticals. Phthalates are characterised by a bi-carboxylated benzoic structure. The main

Advances in chromatography

171

compounds most commonly encountered are benzylbutylphthalate (BBP), dibutylphthalate (DBP), diethylphthalate (DEP), di-2-ethylhexylphthalate (DEHP) and di-2,6-isononylphthalate (DINP). Some GC-MS approaches have been proposed for analysing phthalates in environmental matrices such as sediments or wastewater (Penalver et al., 2000; Ballesteros et al., 2006; Aparicio et al., 2007). The number of existing methods dedicated to the determination of phthalates in food products is noticeably scarcer (Serodio and Nogueira, 2000; Feng et al., 2005). The current method of choice for measuring these molecules is LC-MS/MS (Koch et al., 2003; Silva et al., 2003; Calafat et al., 2004; Takatori et al., 2004; Ito et al., 2005; Cantero et al., 2006). Negative ionisation modes (using either ESI or APCI) are mostly employed, but the positive mode has also been used, for example for DEHP. In the negative mode, the precursor ion usually selected corresponds to the pseudo-molecular species [M–H]−. For most published methods, one single main product ion is monitored. It is important to note that, owing to close structural similarities between the different phthalate congeners, some product ions resulting from the fragmentation of the pseudo-molecular ions are the same for different monitored compounds (for instance those appearing at m/z = 77 or m/z = 121). If the HPLC separation is insufficient, this could constrain unambiguous identification and accurate quantification. Another huge difficulty in the field of phthalate analysis is linked to the existence of significant external contamination due to the ubiquity of these substances (presence into the environment and in many laboratory consumables and materials). This is particularly challenging and requires drastic quality control precautions to avoid sample contamination.

7.6 7.6.1

Future trends

Coupling bioassays with mass spectrometry: the missing link between biology and chemistry The direct coupling of bioassays with mass spectrometry is probably one of the more exciting prospects in analytical chemistry in general, and in EDC research in particular. Although a considerable amount of scientific activity (number of research projects, publications and funding) has focused on endocrine disruption over the last decade, there is still a dichotomy between physicochemical and biological methods. Studies on the unambiguous identification and quantification of specific EDCs from various matrices have been carried out, while other excellent work has focused on the biological effects associated with EDCs. These disassociated approaches may constrain the assessment of chemical hazards associated with the presence of multiple EDCs in food in more detail. For instance, MS detection of new metabolites from a recognised EDC in a product immediately poses questions about its eventual biological activity, and hence potential effects on

172

Endocrine-disrupting chemicals in food

human health. On the other hand, the measurement of significant hormonal activity in a sample extract can never be directly associated with a given compound (or even a compound family), owing to the relatively low specificity of these kinds of tests. Of course the offline coupling of such different approaches is possible, but this procedure (sample fractionation, biological testing of each collected fraction, and subsequent MS measurement for these fractions) remains an extremely laborious task. A current development involves the direct (online and automated) coupling of bioassays, to detect sample fractions presenting a biological activity, and MS measurement, to confirm the nature of the substances responsible for the activity. This kind of equipment is still at the prototype stage, but the promise of such new technology is tremendous. The possibility of high throughput and characterisation of food products, in terms of chemical and microbial contamination associated with undesired biological activity and/ or toxicity, undoubtedly represents the near future of monitoring as well as new perspectives for risk assessment. Little literature is available as yet on this very new topic (Krone et al., 1997; Nelson et al., 1997; Grote et al., 2005). It is notably developed in the framework of an integrated project of the European 6th framework programme, BIOCOP. The purpose of this project is to develop ‘new strategies to screen multiple contaminants in food’ based on emerging technologies, and the development of a directly coupled bioassay and MS, using a biosensor based on surface plasmon resonance (SPR) is a part of this project. Two types of interface between SPR-based screening assays and MSbased identification are under investigation. In one, the sample is applied to an LC column and the effluent split between two identical 96-well plates. One of these plates is subjected to SPR assay for bioactivity, generating a ‘biogram’, which is used to identify the relevant samples of the duplicate plate for MS analysis. In the other approach, the applicability of an automated serial SPR biosensor/ESI TOF-MS coupling to identify compounds captured on the sensor surface is being explored. The accurate mass capability of TOF-MS allows the identification of unknown substances. Chemical substance databases can be used to check whether the unknown substance has previously been identified as a product of metabolism, an unmarketed drug, or a synthetic intermediate in any patent or scientific paper. If not, a chemical structure for the unknown compound can be postulated using data from QTOF-MS/MS experiments.

7.6.2

Mass spectrometric fingerprinting: the global metabolomic approach The so-called ‘omics’ techniques (transcriptomics proteomics and metabolomics) have become well known in recent years, and routinely applied in biology. More recently, some of them are also emerging in the fields of chemistry and analytical chemistry. The idea is to explore ‘life complexity’

Advances in chromatography

173

using unrestrictive descriptive methodologies. These approaches are all based on the generation of large sets of descriptors characterising the biological system (cell, tissue, organ or entire organism) under consideration. In metabolomics, the signals measured correspond to chemical substances (so-called metabolites) accessible to the analysis, which are present in the sample as a result of the complex mechanisms of transcription, translation and regulation. The general principle of metabolomics is to characterise a biological sample by a kind of fingerprint. From an analytical point of view, the technique most widely used for this purpose historically was nuclear magnetic resonance (NMR). However, the use of MS in this field is increasing. Indeed, MS presents some incomparable advantages for at least two crucial reasons. First, MS has very high sensitivity, which is useful for measuring species with low abundance (minor but potentially informative metabolites). Secondly, the specificity of MS (through high-resolution and/or MSn techniques) permits the elucidation of the chemical structure of potential metabolites (biomarkers) of interest. The latest hybrid instruments, combining conventional quadrupole or linear ion trap with high to ultra-high resolution mass filters (such as Q-TOF, LTQ-Orbitrap or LTQ-FTICR), are particularly suitable for metabolomic investigations. Metabolomics has been gaining popularity due to its application across diverse fields related to medical sciences, e.g. functional genomics, toxicology, nutritional science and disease diagnosis (Watkins and German, 2002). But in addition to its significance in understanding biology, metabolomics has also provided a new way of characterising biological samples, and this particular use may be of special interest in the area covered by this book. Thus metabolomics might be envisaged as a new method for characterising food products. We can envisage, for example, building reference libraries of large sets of MS metabolomic fingerprints collected from typical food products, i.e. well-characterised products obtained under standardised conditions of production and free of any chemical residues and contaminants. This system could provide new labels for traceability, quality and/or production origin. It could also provide new controls, based on high throughput and rapid collection of the MS metabolomic fingerprints for any sample collected for monitoring purposes, in order to reveal potential deviations and the possible presence of unwanted EDCs. An even more exciting use of MS is in ultimate imaging technologies that permit full cartography of the analysed samples, useful for example in drug and pharmaceutical distribution (Rubakhin et al., 2005; Hsieh et al., 2007; Prideaux et al., 2007; Reyzer and Capriol, 2007). Of course the difficulties and unsolved problems in this area are still significant, and achieving the desired goal will take time. Some critical analytical parameters will have to be harmonised and standardised, new validation procedures will need to be invented, and this approach ultimately accepted at the regulatory level. However, the exponential rise in the

174

Endocrine-disrupting chemicals in food

number of studies dealing with metabolomics in the past few years, the rapid development of extended metabolomic databases worldwide, and continuous technological progress in the field of MS-coupled techniques, suggest that this dream has some chance of becoming reality, as already confirmed by some recent papers (Bajad and Shulaef, 2007; Lommen et al., 2007).

7.7

Sources of further information and advice

BIOCOP, ‘New strategies for screening multiple contaminants in food’. Integrated Project of the 6th European Framework Program. http://www. biocop.org Gelpi E. Advances in mass spectrometry, volume 15. Chichester, John Wiley & Sons Ltd. 2001. Keith LH, Jones-Lepp TL, and Needham LL. Analysis of environmental endocrine disruptors. Washington, DC, Heidelburg., American Chemical Society, 2000. Metzler M. Endocrine disruptors, part II. Springer-Verlag Telos, 2002. Niessen WMA, Gross ML, Caprioli RM. The encyclopaedia of mass spectrometry, volume 8: hyphenated methods. Oxford, Elsevier Ltd, 2006.

7.8

References

amendola l, botre f, carollo a s, longo d and zoccolillo l (2002), ‘Analysis of organophosphorus pesticides by gas chromatography-mass spectrometry with negative chemical ionization: a study on the ionization conditions’, Analytica Chimica Acta, 461, 97–108. annesley t m (2003), ‘Ion suppression in mass spectrometry’, Clinical Chemistry, 49, 1041–1044. antignac j-p, cariou r, cravedi j-p, bizec b l and andre f (2003), ‘Identification of phytoestrogens in bovine milk using liquid chromatography-electrospray tandem mass spectrometry’, Rapid Communications in Mass Spectrometry, 17, 1256–1264. antignac j-p, cariou r, bizec b l and andre f (2004), ‘New data regarding phytoestrogens content in bovine milk’, Food Chemistry, 87, 275–281. antignac j-p, brosseaud a, gaudin i, andré f and le bizec b (2005a), ‘Analytical strategies for the direct mass spectrometric analysis of steroid and corticosteroid urinary phase II metabolites’, Steroids, 70, 205–216. antignac j-p, de wasch k, monteau f, de brabander h, andré f and le bizec b (2005b), ‘The ion suppression phenomenon in liquid chromatography-mass spectrometry and its consequences in the field of residue analysis’, Analytica Chimica Acta, 529, 129–136. antonelli m l, faberi a, pastorini e, samperi r and lagana a (2005), ‘Simultaneous quantitation of free and conjugated phytoestrogens in Leguminosae by liquid chromatography–tandem mass spectrometry’, Talanta, 66, 1025–1033. aparicio i, santos j l and alonso e (2007), ‘Simultaneous sonication-assisted extraction, and determination by gas chromatography-mass spectrometry, of

Advances in chromatography

175

di-(2-ethylhexyl)phthalate, nonylphenol, nonylphenol ethoxylates and polychlorinated biphenyls in sludge from wastewater treatment plants’, Analytica Chimica Acta, 584, 455–461. arrebola f j, martinez vidal j l, mateu-sanchez m and alvarez-castellon f j (2003), ‘Determination of 81 multiclass pesticides in fresh foodstuffs by a single injection analysis using gas chromatography-chemical ionization and electron ionization tandem mass spectrometry’, Analytica Chimica Acta, 484, 167–180. avery m j (2003), ‘Quantitative characterization of differential ion suppression on liquid chromatography/atmospheric pressure ionization mass spectrometric bioanalytical methods’, Rapid Communication in Mass Spectrometry, 17, 197–201. bajad s and shulaev v (2007), ‘Highly-parallel metabolomics approaches using LCMS2 for pharmaceutical and environmental analysis’, TrAC Trends in Analytical Chemistry, 26, 625–636. ballesteros o, zafra a, navalon a and vilchez j l (2006), ‘Sensitive gas chromatographic-mass spectrometric method for the determination of phthalate esters, alkylphenols, bisphenol A and their chlorinated derivatives in wastewater samples’, Journal of Chromatography A, 1121, 154–162. benomar s h, clench m r and allen d w (2001), ‘The analysis of alkylphenol ethoxysulphonate surfactants by high-performance liquid chromatography, liquid chromatography-electrospray ionisation-mass spectrometry and matrix-assisted laser desorption ionisation-mass spectrometry’, Analytica Chimica Acta, 445, 255–267. berthiller f, schuhmacher r, buttinger g and krska r (2005), ‘Rapid simultaneous determination of major type A- and B-trichothecenes as well as zearalenone in maize by high performance liquid chromatography-tandem mass spectrometry’, Journal of Chromatography A, 1062, 209–216. blasco c, font g and pico y (2004), ‘Multiple-stage mass spectrometric analysis of six pesticides in oranges by liquid chromatography–atmospheric pressure chemical ionization-ion trap mass spectrometry’, Journal of Chromatography A, 1043, 231–238. blasco c, font g and pico y (2005), ‘Analysis of pesticides in fruits by pressurized liquid extraction and liquid chromatography-ion trap-triple stage mass spectrometry’, Journal of Chromatography A, 1098, 37–43. blasco c, van poucke c and van peteghem c (2007), ‘Analysis of meat samples for anabolic steroids residues by liquid chromatography/tandem mass spectrometry’, Journal of Chromatography A, 1154, 230–239. bobeldijk i, vissers j p c, kearney g, major h and van leerdam j a (2001), ‘Screening and identification of unknown contaminants in water with liquid chromatography and quadrupole-orthogonal acceleration-time-of-flight tandem mass spectrometry’, Journal of Chromatography A, 929, 63–74. cajka t and hajslova j (2004), ‘Gas chromatography-high-resolution time-of-flight mass spectrometry in pesticide residue analysis: advantages and limitations’, Journal of Chromatography A, 1058, 251–261. calafat a m, slakman a r, silva m j, herbert a r and needham l l (2004), ‘Automated solid phase extraction and quantitative analysis of human milk for 13 phthalate metabolites’, Journal of Chromatography B, 805, 49–56. cantero m, rubio s and perez-bendito d (2006), ‘Determination of alkylphenols and alkylphenol carboxylates in wastewater and river samples by hemimicelle-based extraction and liquid chromatography-ion trap mass spectrometry’, Journal of Chromatography A, 1120, 260–267. carabias-martinez r, rodriguez-gonzalo e and revilla-ruiz p (2006), ‘Determination of endocrine-disrupting compounds in cereals by pressurized liquid extraction and liquid chromatography-mass spectrometry: study of background contamination’, Journal of Chromatography A, 1137, 207–215.

176

Endocrine-disrupting chemicals in food

cariou r, antignac j-p, marchand p, berrebi a, zalko d, le bizec b and andré f (2005), ‘Analytical strategy for the multi-residue analysis of brominated flame retardants in various biological matrices from unique reduced size sample’, Journal of Chromatography A, 1100, 144–152. courant f, antignac j-p, maume d, monteau f, andersson a-m, skakkebaek n, andré f and le bizec b (2007), ‘Exposure assessment of prepubertal children to steroid endocrine disrupters: 1 Analytical strategy for estrogens measurement in plasma at ultra-trace level’, Analytica Chimica Acta, 586, 105–114. covaci a, voorspoels s and de boer j (2003), ‘Determination of brominated flame retardants, with emphasis on polybrominated diphenyl ethers (PBDEs) in environmental and human samples – a review’, Environment International, 29, 735–756. dallüge j, vreuls r, beens j and brinkman u a t (2002), ‘Optimisation and characterisation of comprehensive two-dimensional gas chromatography with time-offlight mass spectrometric detection (GC × GC-TOFMS)’, Journal of Separation Science, 25, 201–214. de koning s, lach g, linkerhagner m, loscher r, horst tablack p and brinkman u a t (2003), ‘Trace-level determination of pesticides in food using difficult matrix introduction-gas chromatography-time-of-flight mass spectrometry’, Journal of Chromatography A, 1008, 247–252. debrauwer l, riu a, jouahri m, rathahao e, jouanin i, antignac j-p, cariou r, le bizec b and zalko d (2005), ‘Probing new approaches using atmospheric pressure photo ionization for the analysis of brominated flame retardants and their related degradation products by liquid chromatography–mass spectrometry’, Journal of Chromatography A, 1082, 98–109. derouiche a, driss m r, morizur j-p and taphanel m-h (2007), ‘Simultaneous analysis of polychlorinated biphenyls and organochlorine pesticides in water by headspace solid-phase microextraction with gas chromatography–tandem mass spectrometry’, Journal of Chromatography A, 1138, 231–243. embrechts j, lemière f and van dongen w (2003), ‘Detection of estrogen DNAadducts in human breast tumor tissue and healthy tissue by combined nano-LC nano-ES tandem mass spectrometry’, Journal of American Society for Mass Spectrometry, 14, 482–491. feng y-l, zhu j and sensenstein r (2005), ‘Development of a headspace solid-phase microextraction method combined with gas chromatography mass spectrometry for the determination of phthalate esters in cow milk’, Analytica Chimica Acta, 538, 41–48. fernandez moreno j l, arrebola liebanas f j, garrido frenich a and martinez vidal j l (2006), ‘Evaluation of different sample treatments for determining pesticide residues in fat vegetable matrices like avocado by low-pressure gas chromatography–tandem mass spectrometry’, Journal of Chromatography A, 1111, 97–105. ferrer i and thurman e m (2003), ‘Liquid chromatography/time-of-flight/mass spectrometry (LC/TOF/MS) for the analysis of emerging contaminants’, TrAC Trends in Analytical Chemistry, 22, 750–756. ferrer i, garcia-reyes j f and fernandez-alba a (2005), ‘Identification and quantitation of pesticides in vegetables by liquid chromatography time-of-flight mass spectrometry’, TrAC Trends in Analytical Chemistry, 24, 671–682. focant j-f, sjodin a and patterson d g (2003), ‘Qualitative evaluation of thermal desorption-programmable temperature vaporization-comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry for the analysis of selected halogenated contaminants’, Journal of Chromatography A, 1019, 143–156. focant j-f, sjodin a and patterson d g (2004a), ‘Improved separation of the 209 polychlorinated biphenyl congeners using comprehensive two-dimensional gas

Advances in chromatography

177

chromatography-time-of-flight mass spectrometry’, Journal of Chromatography A, 1040, 227–238. focant j-f, sjodin a, turner w and patterson d g (2004b), ‘Measurement of selected polybrominated diphenyl ethers, polybrominated and polychlorinated biphenyls, and organochlorine pesticides in human serum and milk using comprehensive two-dimensional gas chromatography isotope dilution time-of-flight mass spectrometry’, Analytical Chemistry, 76, 6613–6620. focant j-f, eppe g and scippo m-l (2005), ‘Comprehensive two-dimensional gas chromatography with isotope dilution time-of-flight mass spectrometry for the measurement of dioxins and polychlorobiphenyls in foodstuffs. Comparison with other techniques’, Journal of Chromatography A, 1086, 45–60. frenich a g, bolanos p p and vidal j l m (2007), ‘Multiresidue analysis of pesticides in animal liver by gas chromatography using triple quadrupole tandem mass spectrometry’, Journal of Chromatography A, 1153, 194–202. fuh m-r, huang s-y and lin t-y (2004), ‘Determination of residual anabolic steroid in meat by gas chromatography-ion trap-mass spectrometer’, Talanta, 64, 408–414. garcia-reyes j f, ferrer c, gomez-ramos m j, fernandez-alba a r, garcia-reyes j f and molina-diaz a (2007), ‘Determination of pesticide residues in olive oil and olives’, TrAC Trends in Analytical Chemistry, 26, 239–251. garrido frenich a, martinez vidal j l, lopez lopez t, cortes aguado s and martinez salvador i (2004), ‘Monitoring multi-class pesticide residues in fresh fruits and vegetables by liquid chromatography with tandem mass spectrometry’, Journal of Chromatography A, 1048, 199–206. garrido frenich a, martinez vidal j l, cruz sicilia a d, gonzalez rodriguez m j and plaza bolanos p (2006), ‘Multiresidue analysis of organochlorine and organophosphorus pesticides in muscle of chicken, pork and lamb by gas chromatographytriple quadrupole mass spectrometry’, Analytica Chimica Acta, 558, 42–52. gonzalez s, barcelo d and petrovic m (2007), ‘Advanced liquid chromatographymass spectrometry (LC-MS) methods applied to wastewater removal and the fate of surfactants in the environment’, TrAC Trends in Analytical Chemistry, 26, 116–124. granby k, andersen j h and christensen h b (2004), ‘Analysis of pesticides in fruit, vegetables and cereals using methanolic extraction and detection by liquid chromatography–tandem mass spectrometry’, Analytica Chimica Acta, 520, 165–176. grote j, n d, gedig e and koenig s (2005), ‘Surface plasmo resonance/mass spectrometry interface’, Analytical Chemistry, 77, 1157–1162. guo t, taylor r l, singh r j and soldin s j (2006), ‘Simultaneous determination of 12 steroids by isotope dilution liquid chromatography–photospray ionization tandem mass spectrometry’, Clinica Chimica Acta, 372, 76–82. hanold k, fischer s, cormia p, miller c and syage j (2004), ‘Atmospheric pressure photoionization. 1. general properties for LC/MS’, Analytical Chemistry, 76, 2842–2851. hernandez f, pozo o j, sancho j v, lopez f j, marin j m and ibanez m (2005), ‘Strategies for quantification and confirmation of multi-class polar pesticides and transformation products in water by LC-MS2 using triple quadrupole and hybrid quadrupole time-of-flight analyzers’, TrAC Trends in Analytical Chemistry, 24, 596–612. hernandez f, pozo o j, sancho j v, bijlsma l, barreda m and pitarch e (2006), ‘Multiresidue liquid chromatography tandem mass spectrometry determination of 52 non gas chromatography-amenable pesticides and metabolites in different food commodities’, Journal of Chromatography A, 1109, 242–252. hiemstra m and de kok a (2007), ‘Comprehensive multi-residue method for the target analysis of pesticides in crops using liquid chromatography–tandem mass spectrometry’, Journal of Chromatography A, 1154, 3–25.

178

Endocrine-disrupting chemicals in food

hong j, kim h-y, kim d-g, seo j and kim k-j (2004), ‘Rapid determination of chlorinated pesticides in fish by freezing-lipid filtration, solid-phase extraction and gas chromatography-mass spectrometry’, Journal of Chromatography A, 1038, 27–35. hsieh y, chen j and korfmacher w a (2007), ‘Mapping pharmaceuticals in tissues using MALDI imaging mass spectrometry’, Journal of Pharmacological and Toxicological Methods, 55, 193–200. impens s, de wasch k, cornelis m and de brabander h f (2002), ‘Analysis on residues of estrogens, gestagens and androgens in kidney fat and meat with gas chromatography–tandem mass spectrometry’, Journal of Chromatography A, 970, 235–247. isobe t, shiraishi h, yasuda m, shinoda a, suzuki h and morita m (2003), ‘Determination of estrogens and their conjugates in water using solid-phase extraction followed by liquid chromatography–tandem mass spectrometry’, Journal of Chromatography A, 984, 195–202. ito r, seshimo f, miura n, kawaguchi m, saito k and nakazawa h (2005), ‘Highthroughput determination of mono- and di(2-ethylhexyl)phthalate migration from PVC tubing to drugs using liquid chromatography–tandem mass spectrometry’, Journal of Pharmaceutical and Biomedical Analysis, 39, 1036–1041. jahnke a, gandrass j and ruck w (2004), ‘Simultaneous determination of alkylphenol ethoxylates and their biotransformation products by liquid chromatography/ electrospray ionisation tandem mass spectrometry’, Journal of Chromatography A, 1035, 115–122. kelly c (2000), ‘Analysis of steroids in environmental water samples using solidphase extraction and ion-trap gas chromatography-mass spectrometry and gas chromatography–tandem mass spectrometry’, Journal of Chromatography A, 872, 309–314. khummueng w, trenerry c, rose g and marriott p j (2006), ‘Application of comprehensive two-dimensional gas chromatography with nitrogen-selective detection for the analysis of fungicide residues in vegetable samples’, Journal of Chromatography A, 1131, 203–214. king r, bonfiglio r, fernadez-metzler c, miller-stein c and olah t (2000), ‘Mechanistic investigation of ionization suppression in electrospray ionization’, Journal of the American Society for Mass Spectrometry, 11, 942–950. koch h m, gonzalez-reche l m and angerer j (2003), ‘On-line clean-up by multidimensional liquid chromatography–electrospray ionization tandem mass spectrometry for high throughput quantification of primary and secondary phthalate metabolites in human urine’, Journal of Chromatography B, 784, 169–182. kovalczuk t, jech m, poustka j and hajslova j (2006), ‘Ultra-performance liquid chromatography–tandem mass spectrometry: a novel challenge in multiresidue pesticide analysis in food’, Analytica Chimica Acta, 577, 8–17. krone j, nelson r, dogruel d, william p and granzow r (1997), ‘BIA/MS: interfacing biomolecular interaction analysis with mass spectrometry’, Analytical Biochemistry, 244, 124–132. le bizec b, maume d, marchand p, monteau f, bichon e and andré f (2004), ‘Review: control of anabolic steroid in breeding animals: mass spectrometry, a powerful analytical tool’, Chromatographia, 59, S3–S11. leandro c c, hancock p, fussell r j and keely b j (2006), ‘Comparison of ultraperformance liquid chromatography and high-performance liquid chromatography for the determination of priority pesticides in baby foods by tandem quadrupole mass spectrometry’, Journal of Chromatography A, 1103, 94–101. leandro c c, hancock p, fussell r j and keely b j (2007), ‘Ultra-performance liquid chromatography for the determination of pesticide residues in foods by tandem

Advances in chromatography

179

quadrupole mass spectrometry with polarity switching’, Journal of Chromatography A, 1144, 161–169. lommen a, van der weg g, van engelen m c, bor g, hoogenboom l a p and nielen m w f (2007), ‘An untargeted metabolomics approach to contaminant analysis: Pinpointing potential unknown compounds’, Analytica Chimica Acta, 584, 43–49. lopez de alda m j, diaz-cruz s, petrovic m and barcelo d (2003), ‘Liquid chromatography-(tandem) mass spectrometry of selected emerging pollutants (steroid sex hormones, drugs and alkylphenolic surfactants) in the aquatic environment’, Journal of Chromatography A, 1000, 503–526. marchand p, le bizec b, gade c, monteau f and andre f (2000), ‘Ultra trace detection of a wide range of anabolic steroids in meat by gas chromatography coupled to mass spectrometry’, Journal of Chromatography A, 867, 219–233. mariott p and shellie r (2002), ‘Principles and applications of comprehensive twodimensional gas chromatography’, TrAC Trends in Analytical Chemistry, 21, 573–5583. martinez vidal j l, arrebola f j and mateu-sanchez m (2002), ‘Application of gas chromatography–tandem mass spectrometry to the analysis of pesticides in fruits and vegetables’, Journal of Chromatography A, 959, 203–213. mei h y h nardo c, xu x, wang s, ng k and korfmacher w a (2003), ‘Investigation of matrix effects in bioanalytical high-performance liquid chromatography/ tandem mass spectrometric assays: application to drug discovery’, Rapid Communication in Mass Spectrometry, 17, 97–103. moeder m, martin c, schlosser d, harynuk j and gorecki t (2006), ‘Separation of technical 4-nonylphenols and their biodegradation products by comprehensive two-dimensional gas chromatography coupled to time-of-flight mass spectrometry’, Journal of Chromatography A, 1107, 233–239. mol h g j, van dam r c j and steijger o m (2003), ‘Determination of polar organophosphorus pesticides in vegetables and fruits using liquid chromatography with tandem mass spectrometry: selection of extraction solvent’, Journal of Chromatography A, 1015, 119–127. nelson r, krone j and jansson o (1997), ‘Surface plasmon resonance biomolecular interaction analysis mass spectrometry. 1. Chip-based analysis’, Analytical Chemistry, 69, 4363–4368. nielen m w f, van engelen m c, zuiderent r and ramaker r (2007), ‘Screening and confirmation criteria for hormone residue analysis using liquid chromatography accurate mass time-of-flight, Fourier transform ion cyclotron resonance and orbitrap mass spectrometry techniques’, Analytica Chimica Acta, 586, 122–129. ong r and marriott p j (2002), ‘Review of basic concepts in comprehensive twodimensional gas chromatography’, Journal of Chromatographic Science, 40, 276–291. ortelli d, edder p and corvi c (2004), ‘Multiresidue analysis of 74 pesticides in fruits and vegetables by liquid chromatography–electrospray–tandem mass spectrometry’, Analytica Chimica Acta, 520, 33–45. papadakis e n, vryzas z and papadopoulou-mourkidou e (2006), ‘Rapid method for the determination of 16 organochlorine pesticides in sesame seeds by microwave-assisted extraction and analysis of extracts by gas chromatography–mass spectrometry’, Journal of Chromatography A, 1127, 6–11. papke o, furst p and herrmann t (2004), ‘Determination of polybrominated diphenylethers (PBDEs) in biological tissues with special emphasis on QC/QA measures’, Talanta, 63, 1203–1211. patel k, fussell r j, hetmanski m, goodall d m and keely b j (2005), ‘Evaluation of gas chromatography–tandem quadrupole mass spectrometry for the determination of organochlorine pesticides in fats and oils’, Journal of Chromatography A, 1068, 289–296.

180

Endocrine-disrupting chemicals in food

penalver a, pocurull e, borrull f and marce r m (2000), ‘Determination of phthalate esters in water samples by solid-phase microextraction and gas chromatography with mass spectrometric detection’, Journal of Chromatography A, 872, 191–201. pirard c, widart j, nguyen b k, deleuze c, heudt l, haubruge e, de pauw e and focant j-f (2007), ‘Development and validation of a multi-residue method for pesticide determination in honey using on-column liquid–liquid extraction and liquid chromatography–tandem mass spectrometry’, Journal of Chromatography A, 1152, 116–123. pizzutti i r, de kok a, zanella r, adaime m b, hiemstra m, wickert c and prestes o d (2007), ‘Method validation for the analysis of 169 pesticides in soya grain, without clean up, by liquid chromatography–tandem mass spectrometry using positive and negative electrospray ionization’, Journal of Chromatography A, 1142, 123–136. prideaux b, atkinson s j, carolan v a, morton j and clench m r (2007), ‘Sample preparation and data interpretation procedures for the examination of xenobiotic compounds in skin by indirect imaging MALDI-MS’, International Journal of Mass Spectrometry, 260, 243–251. razzazi-fazeli e, böhm j and luf w (1999), ‘Determination of nivalenol and deoxynivalenol in wheat using liquid chromatography-mass spectrometry with negative ion atmospheric pressure chemical ionisation’, Journal of Chromatography A, 854, 45–55. ren y, zhang y, shao s, cai z, feng l, pan h and wang z (2007), ‘Simultaneous determination of multi-component mycotoxin contaminants in foods and feeds by ultra-performance liquid chromatography tandem mass spectrometry’, Journal of Chromatography A, 1143, 48–64. reyzer m l and caprioli r m (2007), ‘MALDI-MS-based imaging of small molecules and proteins in tissues’, Current Opinion in Chemical Biology, 11, 29–35. robb d, covey t and bruins a (2000), ‘Atmospheric pressure photoionisation and ionisation method for liquid chromatography-mass spectrometry’, Analytical Chemistry, 72, 3653–3659. rodriguez-mozaz s, lopez de alda m j and barcelo d (2004), ‘Monitoring of estrogens, pesticides and bisphenol A in natural waters and drinking water treatment plants by solid-phase extraction–liquid chromatography–mass spectrometry’, Journal of Chromatography A, 1045, 85–92. rubakhin s s, jurchen j c, monroe e b and sweedler j v (2005), ‘Imaging mass spectrometry: fundamentals and applications to drug discovery’, Drug Discovery Today, 10, 823–837. rundberget t and wilkins a l (2002), ‘Determination of Penicillium mycotoxins in foods and feeds using liquid chromatography–mass spectrometry’, Journal of Chromatography A, 964, 189–197. sannino a, bolzoni l and bandini m (2004), ‘Application of liquid chromatography with electrospray tandem mass spectrometry to the determination of a new generation of pesticides in processed fruits and vegetables’, Journal of Chromatography A, 1036, 161–169. saudan c, entenza j m, baume n, mangin p and saugy m (2006), ‘Short-term stability of testosterone and epitestosterone conjugates in urine samples: Quantification by liquid chromatography–linear ion trap mass spectrometry’, Journal of Chromatography B, 844, 168–174. saugy m, cardis c, robinson n and schweizer c (2000), ‘Test methods: anabolics’, Best Practice & Research Clinical Endocrinology & Metabolism, 14, 111–133. serodio p and nogueira j m f (2006), ‘Considerations on ultra-trace analysis of phthalates in drinking water’, Water Research, 40, 2572–2582.

Advances in chromatography

181

sforza s, dall’asta c and marchelli r (2005), ‘Recent advances in mycotoxin determination in food and feed by hyphenated chromatographic techniques/mass spectrometry’, Mass Spectrometry Review, 25, 54–76. shao b, han h, hu j, zhao j, wu g, xue y, ma y and zhang s (2005a), ‘Determination of alkylphenol and bisphenol A in beverages using liquid chromatography/electrospray ionization tandem mass spectrometry’, Analytica Chimica Acta, 530, 245–252. shao b, zhao r, meng j, xue y, wu g, hu j and tu x (2005b), ‘Simultaneous determination of residual hormonal chemicals in meat, kidney, liver tissues and milk by liquid chromatography–tandem mass spectrometry’, Analytica Chimica Acta, 548, 41–50. shao b, han h, tu x and huang l (2007), ‘analysis of alkylphenol and bisphenol A in eggs and milk by matrix solid phase dispersion extraction and liquid chromatography with tandem mass spectrometry’, Journal of Chromatography B, 850, 412–416. shareef a, angove m j and wells j d (2006), ‘Optimization of silylation using N-methyl-N-(trimethylsilyl)-trifluoroacetamide, N,O-bis-(trimethylsilyl)trifluoroacetamide and N-(tert-butyldimethylsilyl)-N-methyltrifluoroacetamide for the determination of the estrogens estrone and 17[alpha]-ethinylestradiol by gas chromatography–mass spectrometry’, Journal of Chromatography A, 1108, 121–128. silva m j, malek n a, hodge c c, reidy j a, kato k, barr d b, needham l l and brock j w (2003), ‘Improved quantitative detection of 11 urinary phthalate metabolites in humans using liquid chromatography–atmospheric pressure chemical ionization tandem mass spectrometry’, Journal of Chromatography B, 789, 393–404. singh g, gutierrez a, xu k and blair i (2000), ‘Liquid chromatography–electron capture atmospheric-pressure chemical ionization mass spectrometry: analysis of pentafluorobenzyl derivatives of biomolecules and drugs in the attomole range’, Analytical Chemistry, 72, 3007–3013. soler c and pico y (2007), ‘Recent trends in liquid chromatography-tandem mass spectrometry to determine pesticides and their metabolites in food’, TrAC Trends in Analytical Chemistry, 26, 103–115. stan h-j (2000), ‘Pesticide residue analysis in foodstuffs applying capillary gas chromatography with mass spectrometric detection: state-of-the-art use of modified DFG-multimethod S19 and automated data evaluation’, Journal of Chromatography A, 892, 347–377. strahm e, saudan c, sottas p-e, mangin p and saugy m (2007), ‘Direct detection and quantification of 19-norandrosterone sulfate in human urine by liquid chromatography–linear ion trap mass spectrometry’, Journal of Chromatography B, 852, 491–496. syage j, hanold k, lynn t, horner j and thakur r (2004), ‘Atmospheric pressure photoionization. 2. Dual source ionization’, Journal of Chromatography A, 1050, 137–149. takatori s, kitagawa y, kitagawa m, nakazawa h and hori s (2004), ‘Determination of di(2-ethylhexyl)phthalate and mono(2-ethylhexyl)phthalate in human serum using liquid chromatography-tandem mass spectrometry’, Journal of Chromatography B, 804, 397–401. taylor m j, hunter k, hunter k b, lindsay d and le bouhellec s (2002), ‘Multiresidue method for rapid screening and confirmation of pesticides in crude extracts of fruits and vegetables using isocratic liquid chromatography with electrospray tandem mass spectrometry’, Journal of Chromatography A, 982, 225–236. thurman e m, ferrer i and barcelo d (2001), ‘Choosing between atmospheric pressure chemical ionization and electrospray ionization interfaces for the HPLC-MS analysis of pesticides’, Analytical Chemistry, 73, 5441–5449.

182

Endocrine-disrupting chemicals in food

van bennekom e o, brouwer l, laurant e h m, hooijerink h and nielen m w f (2002), ‘Confirmatory analysis method for zeranol, its metabolites and related mycotoxins in urine by liquid chromatography–negative ion electrospray tandem mass spectrometry’, Analytica Chimica Acta, 473, 151–160. van deursen m, beens j, janssen h-g, leclerq and cramers c a (2000), ‘Evaluation of time of flight mass spectrometric detection for fast gas chromatography’, Journal of Chromatography A, 878, 205–213. van poucke c, detavernier c, van cauwenberghe r and van peteghem c (2007), ‘Determination of anabolic steroids in dietary supplements by liquid chromatography–tandem mass spectrometry’, Analytica Chimica Acta, 586, 35–42. venkateswarlu p, rama mohan k, ravi kumar c and seshaiah k (in press) ‘Monitoring of multi-class pesticide residues in fresh grape samples using liquid chromatography with electrospray tandem mass spectrometry’, Food Chemistry, 105, 1760–1766. walorczyk s and gnusowski b (2006), ‘Fast and sensitive determination of pesticide residues in vegetables using low-pressure gas chromatography with a triple quadrupole mass spectrometer’, Journal of Chromatography A, 1128, 236–243. wang c-c, prasain j k and barnes s (2002), ‘Review of the methods used in the determination of phytoestrogens’, Journal of Chromatography B, 777, 3–28. watkins s m and german j b (2002), ‘Toward the implementation of metabolomic assessments of human health and nutrition’, Current Opinion in Biotechnology, 13, 512–516. wilkinson a p, wahala k and williamson g (2002), ‘Identification and quantification of polyphenol phytoestrogens in foods and human biological fluids’, Journal of Chromatography B, 777, 93–109. wu q, wang m and simon j e (2004), ‘Analytical methods to determine phytoestrogenic compounds’, Journal of Chromatography B, 812, 325–355. xiao x-y, mccalley d v and mcevoy j (2001), ‘Analysis of estrogens in river water and effluents using solid-phase extraction and gas chromatography-negative chemical ionisation mass spectrometry of the pentafluorobenzoyl derivatives’, Journal of Chromatography A, 923, 195–204. xu c l, chu x g, peng c f, jin z y and wang l y (2006), ‘Development of a faster determination of 10 anabolic steroids residues in animal muscle tissues by liquid chromatography tandem mass spectrometry’, Journal of Pharmaceutical and Biomedical Analysis, 41, 616–621. yang l, luan t and lan c (2006), ‘Solid-phase microextraction with on-fiber silylation for simultaneous determinations of endocrine disrupting chemicals and steroid hormones by gas chromatography–mass spectrometry’, Journal of Chromatography A, 1104, 23–32. zambonin c g, quinto m, de vietro n and palmisano f (2004), ‘Solid-phase microextraction–gas chromatography mass spectrometry: A fast and simple screening method for the assessment of organophosphorus pesticides residues in wine and fruit juices’, Food Chemistry, 86, 269–274. zollner p and mayer-helm b (2006), ‘Trace mycotoxin analysis in complex biological and food matrices by liquid chromatography–atmospheric pressure ionisation mass spectrometry’, Journal of Chromatography A, 1136, 123–169. zrostlikova j, hajslova j and cajka t (2003), ‘Evaluation of two-dimensional gas chromatography-time-of-flight mass spectrometry for the determination of multiple pesticide residues in fruit’, Journal of Chromatography A, 1019, 173–186. zuo y, zhang k and lin y (2007), ‘Microwave-accelerated derivatization for the simultaneous gas chromatographic-mass spectrometric analysis of natural and synthetic estrogenic steroids’, Journal of Chromatography A, 1148, 211–218.

8 Biosensors for endocrine disruptors E. Eltzov, A. Kushmaro and R. S. Marks, Ben-Gurion University of the Negev, Israel

Abstract: Endocrine-disrupting compounds (EDCs) have become a serious problem due to their potential to mimic or antagonize the actions of endogenous hormones at the molecular level. Accumulation of these compounds in mammalian and plant tissues and exposure to humans through the food chains turn them into a real health risk. Owing to the chemical diversity of EDCs, there is a growing demand for new, fast and reliable methods for EDC detection. Biosensors that will answer these criteria, are usually built from two major components, the biorecognition element (enzymes, DNA, proteins, whole cells, etc.) intimately attached to a transducer (SPR, optic fiber, etc.) using a variety of techniques (adsorption, covalent binding, trapping, etc.). Even though each year some new compounds are added to the list of EDCs and in spite of the great progress in the field of the biosensors, there is still a serious lack of biosensors capable of high-resolution detection of these compounds . Key words: biosensors, endocrine-disrupting compounds, bioluminescence, fiber optic.

8.1

Introduction

The vast developments in industry have caused a negative impact on the environment. Each year numerous new compounds, with unknown effects on human health, have been developed and eventually found their way into the environment. Many of these synthetic chemicals possess estrogenic activity and have been classified as ‘endocrine-disrupting compounds’ (EDCs), including polycarbonate by-products, surfactants such as octylphenol, insecticides such as dichlorodiphenol trichloroethane (DDT) and its metabolites, endosulfan, phthalate plasticisers, polychlorinated biphenyls (PCBs), dioxins, alkylphenols (APs), bisphenol A, parabens, polycyclic aromatic hydrocarbons (PAHs), etc. Some of these compounds have been included in a priority list of the Water Framework Directive (WFD) 2000/60/ EC [1]. In addition, biodegradation products of many known compounds

184

Endocrine-disrupting chemicals in food

also produce chemicals that have estrogenic activity, such as the partial degradation of the fire-retardant tetrabromobisphenal A (TBBPA), which are used at a relatively high concentration in many applications including electronic equipment, textiles plastic and cars, and produce bisphenol-A as a metabolite [2]. Another synthetic EDC, mainly found in marine and aquatic environments, is the antifoulant agent tributyltin. This chemical is known to accumulate in the environment and in the food chain and is of concern to human health [3]. The activities of these chemicals have reportedly been mediated by their potential mimic or antagonistic capability on the actions of endogenous hormones at the molecular level. EDCs are defined not by their chemical nature but by their biological effect. Therefore many various compounds with dissimilar chemical structures may be classified as EDCs [4]. Exposure to these chemicals can cause reproductive abnormalities and feminization of wildlife [5–7] and possible reproductive disorders in humans [8–10]. In some cases EDCs are long-lived and can accumulate in the tissues of plants and animals [11]. Human exposure to EDCs is a critical concern. The EDC group compound nonylphenol (NP) has been found in over 60 different food products in Germany [12] and importantly has been found to be a potential cause of decrease in the sperm count in the human population in Tokyo Bay, Japan [13]. Potential routes of exposure to EDCs for terrestrial and aquatic wildlife are by contact with contaminated surface waters. EDCs can enter the surface water by a variety of mechanisms including direct discharge of industrial and domestic wastewaters, discharge of sewage treatment plant (STP) effluents, agricultural drains to streams and rivers and overland flow after rainfall events [1]. Since the chemical structure of the EDCs varies considerably, evaluation of the health risks have to be based on biological and not chemical effects [14]. The traditional techniques for detection of EDCs, such as highperformance liquid chromatography (HPLC), gas chromatography (GCL) or enzyme-linked immunosorbent assay (ELISA) [15] enable the detection of a single compound or a group of structurally related compounds at any given time [16]. Many EDCs have effects at very low concentrations, such as nanograms per litre levels for estradiol. There is therefore a need to develop analytical methods applicable to mixtures and to trace levels. It is also important to integrate and correlate chemical analytical data with the actual endocrine-disrupting effects. Thus there is an urgent demand for new analytical devices that can provide such specialized detection in a reliable and rapid way [17]. In this sense, biosensors seem to be a preferable choice. The definition of a biosensor accepted in the literature is a selfcontained bionic integrated device, including a biological recognition element (enzyme, antibody, receptors or microorganisms) that can respond in a concentration-dependent manner to a biochemical species [18] (Fig. 8.1). Food contaminants and environmental pollutants require the same

Biosensors for endocrine disruptors

185

Immobilization layer S P

Enzyme DNA hybridization

Output signal

Microorganism

Sample

Receptor Immunoreaction Transducing element

Biorecognition components Interface

Fig. 8.1

Schematic representation of biosensor.

sensitivity, limit of detection and stability as medical applications; however, such parameters as the volume of the sample, matrix complexity and requirement for on-site continuous monitoring make developing these biosensors much more complicated. Biosensors offer some advantages for EDC analysis when comparing with conventional analytical methods. They can be cheap and simple to use (e.g. glucose biosensors), and are frequently able to measure EDCs in complex matrices with minimal sample preparation [19]. Other advantages offered by biosensors over conventional analytical techniques include the possibility of miniaturization and portability, which permits their use as field devices working on-site. The main disadvantage of traditional methods in assessing effects of contaminants is the lack of possibility of measuring biological effects of tested compounds for toxicity, cytotoxicity, genotoxicity or endocrine-disrupting effects unless complex bioassays are used, such as fish, daphnia tech. In some cases this information may be more significant than the chemical composition itself. Table 8.1 shows the most important characteristic differences between biosensors and traditional methods employed in the sensing of various environmental pollutants especially EDCs. Unlike biosensors, bioassays or bioanalytical systems require additional processing steps, such as reagent addition [20]. Furthermore the assay design in bioassays or bioanalytical systems is permanently fixed in the construction of the device. In general, biosensors allow discovery, detection and biologically effect (toxicity) prediction of various EDCs.

186

Endocrine-disrupting chemicals in food

Table 8.1 methods

Characteristic comparison between biosensors and traditional

Traditional techniques

Biosensors

Comparatively high sample volume Long period of analysis

Small sample volume Fast monitoring time, real time monitoring, high throughput Availability of portable detection systems, on-site monitoring Cheap Low biological systems stability Determination of toxicity (biological effect) and some time identity of compound Possibility to overlook compounds with no biological effects User friendly Minimal waste generation; need to use genetically modified organisms Most methods in prototype stage

Limited portability Expensive equipment No biological stability qualification Determination of chemical structure of tested compounds Determination of all chemical compounds in sample Qualified personal required Generation of chemical or organic solvents waste Most of methods commercially available

8.2

General structure of biosensors

The characteristic biosensor structure may be divided into three major parts: biorecognition components, interface (immobilization techniques) and transducing elements (Fig. 8.1).

8.2.1 Biorecognition components The biological recognition elements of a biosensor interact selectively with the target analyte(s), assuring the selectivity of the sensors. These elements can be classified into five main classes: whole cell, nucleic acids, immunochemical, enzymatic and non-enzymatic receptors [21]. Enzyme-based biosensors are the most common and well-developed group of sensors in the environmental and food applications. Enzymatic biosensors depend on the turn-over number of the enzyme and provide a significant amplification system for the sensitive detection of the analyte (substrate). Enzymes are often chosen as bioreceptors based on their specific binding capabilities as well as their catalytic activity. In biocatalytic recognition mechanisms, the detection is amplified by a reaction catalyzed by macromolecules called biocatalysts [22]. In food contaminant analysis, both the catalytic conversion of the substrate and the dose-dependent inhibition of an enzyme reaction are important determinants of the contaminant concentration. The most common way to measure enzymatic activities is through the use of electrochemical transducers with four main strategies used for monitoring

Biosensors for endocrine disruptors

187

different pollutants. In the first strategy the corresponding signal decreases from its initial value due to the consumption of the detectable compound, e.g. oxygen depleted by oxidase [23]. In the second strategy, recycling the enzyme products correspond to the signal increment, e.g., H+, CO2, NH3, produced by oxidoreductase [24]. The third strategy makes use of the detection of the state of the biocatalyst redox active center, cofactor, prosthetic group evolution in the presence of tested analyte. The last method exploits the direct electron transfer between the active site of the enzyme and transducer [25]. The acetycholinesterases (AChEs)-based enzymatic biosensors are good examples for enzyme-based biosensors. Owing to the rise in insecticides in our environment and the associated health problems, AChE biosensors have gained enormous attention [26, 27]. Although enzyme-based biosensors provide significant advantages in certain pollution monitoring tasks, they have some inherent limits, such as a lack of versatility and inherent sensitivity to matrix inhibition. In contrast to enzymatic biosensors, antibody-based biosensors have much more versatility, because of their ability to bind to structurally related groups with very wide scopes of affinity. All interactions in immunosensors take place on the solid–liquid interface, since either antibodies or antigens are immobilized on a solid support [28]. The huge benefit of immunosensors is the ability, through development and fixation of new recombinant antibodies, to control their affinity and selectivity. There are, however, several limitations to the use of immunosensors for EDCs monitoring applications. The limitations include the number of specialized components needed (antibodies, substrates, etc.) and the fact that each target compound is an independent biosensor device unless arrays of electrodes of fiber optics are used [29, 30]. Another group of biorecognition elements are the so-called protein-based elements. Recent advances in the development of receptor-based biosensors have focused on the EDC field. Receptor-based biosensor (non-catalyric proteins) systems can be used for wide ranges of structurally dissimilar compounds which share the same mechanisms of toxicity. For example, endocrine receptor biosensors have been developed using the optical sensor BIAcore (plasmon resonance) sensor platform [31]. The benefit of this system is its ability to sense and characterize a wide range of estrogenic compounds. Whole-cell-based biosensors (Fig. 8.2) use various cell types (bacteria, yeast, algae and tissue culture cells) which limit their specific detection of chemicals but provide the biological effect of a specific pollutant with the understood limitation in terms of mixed EDCs in a sample providing either antagonistic or synergistic activity. Another advantage of these systems is their tolerance of various assay conditions and the possibility for preparation of almost unlimited test quantities. Genetic engineered bacteria or yeast cells that bear the lux or luc gene operon, expressing luminescent proteins such as the green fluorescence protein (GFP), have been investigated in

188

Endocrine-disrupting chemicals in food Sample Soil [40] Water [19] Analyte EDCs [19] Heavy metals [38] Organic compounds [39] Antibiotics [37]

Immobilization matrices

Photodetector

Alginate [34–36] Silane [54]

Fiber optic [34–36] Optic devices [37]

Biosensor

Data processing by computer Reporter genes (bioluminescence and fluorescence) fused to specific stress promoters

Plasmid

Genome Luciferase fluorescent molecule

Luciferase fluorescent molecule

Fig. 8.2 Concept of a basic whole-cell-based biosensor. (a) The fiber–probe interface. (b) The polymer layers with the approximate thickness at around 80–100 μm.

recent years [32–40]. Generally there are two main assay types in wholecell-based biosensor systems: constitutive and inducible. In the constitutive type, the toxic and harmful conditions affect the naturally high basal expression of the reporter gene; damage then leads to the reduction in the measurements of the signal. In the inducible method the expression of the reporter signal begins with exposure of the biosensor to the analyte. A genetically engineered bioluminescent yeast-based bioassay has been developed and found suitable for the screening of a large number of compounds. The developed yeast androgen bioassay is based on recombinant Saccharomyces cerevisiae cells that can detect androgenic and anti-androgenic compounds in aqueous solutions without the need to break cells or for washing steps before the measurements [19, 41]. Another group of whole-cell biosen-

Biosensors for endocrine disruptors

189

sors are based on mammalian cell lines. The advantage of mammalian cell biosensors over other biorecognition components is that their response to the toxic compounds may resemble more that of the human response than whole bacteria do [33], taking into account the lack of a 3D structure which dampens toxicity. The putative advantages are that, once engineered, these cells can be reproduced at low cost and there is no need for enzymic cofactor regeneration or introduction of expensive substrates [42, 43]. However, these biosensors often suffer from problems of long response time, low sensitivity, and complex sensor assembly. There are various methods for the detection of toxic compounds by whole-tissue biosensors. For example detection of atrazine through a biocatalytic layer on whole plant tissue that was physically attached to a detecting electrode surface with a support membranes was developed [44]. An additional example is the on-site estimation of water quality by human hepatoblastoma Hep G2 cells [45]. Nucleic acids have been incorporated into a wide range of biosensors in environmental and food contamination fields. While most nucleic acid biosensors are based on complementary strands of RNA or DNA molecules, detection of EDCs can be done by monitoring the affinity small compounds to DNA [46]. The final group of biorecognition elements is biomimetic molecules. The biomimetic-based biosensor is based on recognition elements that are not biologically derived. Because of the non-biological nature of the sensing molecule, there is a division of opinions for its relevancy for use as a biosensor. However, owing to their ability to mimic the biological activity of antibodies, DNA, receptors, etc., they have been tried such as molecularly imprinted polymer-based sensors (MIPs) [47] for the extraction of organic compounds (drugs, toxins, pesticides, etc.) [48].

8.2.2 Interface techniques The immobilization strategy depends on the bioreceptor that will be linked to a given transducer. Some conditions that must be considered are: (1) maintaining biological activity after immobilization, (2) proximity of the biological layer to the transducer, (3) stability and durability of the biological layer and (4) sensing specificity of the biological component to a specific analyte [49] and for some uses the possible future reuse of the biomaterial [50]. The principal methods of immobilization are adsorption, cross-linking, covalent binding entrapment, sol–gel entrapment and Langmuir–Blodgett (LMB) deposition self-assembled biomembranes, and bulk modifications [49]. Absorption is the oldest and simplest immobilization method. It makes use of the physical adsorption of the biorecognition elements to the matrices. The forces exploited in those interaction applications are the hydrophobic, electrostatic, and van der Waals attractive forces. The main advantages

190

Endocrine-disrupting chemicals in food

of the adsorption method are that there is no need to chemically modify the biological components and the possibility of regeneration of the matrix membrane. However, owing to various environmental parameters (pH, ionic strength, temperature and molecular deformation), there is the possibility of the loss of biorecognition [51]. Covalent binding provides strong binding of enzymes, proteins and other biorecognition elements to the transducer surface through a choice of functional groups (amino, carboxyl, sulfhydryl, hydroxyl, imidazole, phenolic, thiol, threonine and indole groups). The main disadvantage of this method is that it requires some form of chemical reaction. Moreover, covalent binding may alter the conformational structure of the receptor or active center of the enzyme, resulting in major loss of activity. This bond is so strong that no leakage of the enzymes will occur, even in the presence of substrate or solution of high ionic strength, that the enzyme is immobilized into either a membrane matrix or directly onto the surface of the transducer [49, 52, 53]. Covalent binding techniques are much more efficient methods for enzymes and antibodies than whole-cell fixation [50, 54] however the fixation of the bioreporter cells has been shown to work [54]. Cross-linking makes use of the enhanced stability of the adsorbed enzymes or proteins that are covalently bound to the support. Despite several advantages, such as the stability and simplicity, there are a number of disadvantages associated with this method: difficulties in controlling the reaction, requirements of large amounts of the biorecognition components, lack in the rigidity of formed layer, difficulties in diffusion due to the large diffusional barriers [49], and loss of the activity of the immobilized layers due to the distortion of molecular conformation during cross-linking [55]. Cross-linking techniques may also be used for the immobilization of cells [50] and enzymes [56]. An alternative method for the immobilization of biological molecules is encapsulation in ceramics, glasses, and other inorganic materials using the sol–gel method [57, 58]. The physical rigidity, chemical inertness and thermal stability make this method attractive for biosensor application [59]. Owing to the optical transparency of the porous matrices, the chemical interactions of the entrapped biorecognition elements allow them to be seen. The possibility of carrying out the process at room temperature, the ability of the matrix to retain the protein conformation and reactivity and, finally, synthesis conditions that are not severe enough to denature most of the immobilized biomolecules also contribute to attractiveness of this method for biosensor applications [60]. Entrapment is another useful technique when viable cells are used [61]. In addition to leakage of the biorecognition elements that occurs during use, which may result in a loss of activity, the diffusion barrier that is created from trapping procedures could affect permeability of the membrane [62] and even interact with certain analyte affecting measurement, which could result in lower sensitivity and detection limits [63]. The use of semipermeable membranes which stop the leakage of the sensing part from the

Biosensors for endocrine disruptors

191

biosensor and are still able permit free passage of analyte may solve this problem. An additional technique for fixation of biorecognition elements is the Langmuir–Blodgett (LMB) method which allows for the formation of model membrane structures by transferring amphiphilic molecules (e.g. fatty acids, phospholipids) residing at the air–water interface to a solid substrate. The substrate can pass through the monolayer-covered liquid surface. This technique immobilized enzymes and protein (e.g. immunoglobulins) films onto piezoelectric, electrochemical and optical supports, with the possibility of fabricating multilayer structures and controlling the density of biological molecules [49]. It is well known that surface-active agents, commonly known as surfactants, can form several self-aggregated structures, namely, micelles, reverse micelles and liquid crystalline phases in aqueous solutions that can generate hydrophobic domains to solvate and solubilize non-polar species. The use of aqueous surfactant media provides a route for the formation of a highly compact, well-packed, ordered and dense selfassembled monolayers (SAM) [64]. Their surface properties allow the control of protein adsorption and cell adhesion [65] as well as their orientation during immobilization [66]. Another self-assembled system is the bilayer membrane (BLM) which mimics natural biological membranes in various biosensor applications [67] and provides a natural environment for the embedding of proteins, pigments and other membrane constituents with little denaturation. The key to the successful construction of BLM-based sensors is the ability to embed functional molecules into the lipid bilayer environment which is hydrophobic, liquid-like and self-organizing [68]. A further method used in biosensor applications is the incorporation of the biorecognition element within the bulk of the entire electrode material. This allows for the integration of various additional components, high stability and close proximity of the biorecognition elements as applied in carbon paste electrodes [69] and screen-printed electrodes [70].

8.2.3 Transducing elements Biosensors can be classified not only by their biorecognition elements but also by the transducing methods they employ (Fig. 8.3). There are four major groups: electrochemical, optical, mass sensitive and thermal biosensors. Optical transducers offer the largest number of possible detection strategies and may use techniques such as UV–vis absorption, bioluminescence, chemiluminescence, fluorescence, phosphorescence, reflectance, scattering and refractive index changes produced by the interaction of the receptor with the target analyte [22, 49]. Optical techniques have certain advantages over other methods in that they are simple, flexible and allow for multichannel and remote sensing. Fiber-optic biosensors use optical fibers as the transduction element, and rely exclusively on optical transduction mechanisms for detecting target biomolecules [71]. Fiber optics enable

192

Endocrine-disrupting chemicals in food Biosensors

Analyte

Bioreceptor

Immobilization

Transduction

Transducer

Enzymes

Antibody

Adsorption

Optical

Gold

Antibodies

Enzyme

Entrapment

Electrochemical

ITO

Microorganisms

DNA

Cross-linking

Piezoelectric

Silica

Antigens

Cell

Covalent bonding

Acoustic

Carbon

Gases

Biomimetic

Affinity interactions

Calorimetric

Hologram

Ions

Antigen

Other

Platinum

Proteins Other

Fig. 8.3

Biocomponents and transducers employed in construction of biosensors.

optical spectroscopy to be performed on sites inaccessible to conventional spectroscopy, over large distances, or even at several spots along fiber Bragg grating. Fiber optics, in being optical waveguides, enable evanescent wave spectroscopy, are now available with transmissions over a wide spectral range; however, plastic fiber optics limit transmissivity due to background fluorescence [72]. Surface plasmon resonance (SPR) works as a direct optical transducer without the head for labeling [73] in antigen–antibody-binding events, thus avoiding a separation step to remove free labels from unbound [74, 75]. Electrochemical transducers exhibit high sensitivity and compatibility with modern microfabrication technologies, are portable, disposable, require minimal power, and are independent of sample turbidity or optical pathway [76]. The electrochemical biosensor can be classified in conductimetric, amperometric, impedimetric and potentiometric methodologies. Heat changes that occur in chemical reactions catalyzed by enzymes and microorganisms may be monitored over time by biosensors based on thermal transducer elements [77]. The main disadvantages of this method are the loss of heat during signal measurement due to the irradiation, conduction

Biosensors for endocrine disruptors

193

and convention of heat from non-flawless adiabatic sensing system and the general problem of isolating the system from external temperatures. Only a few EDC applications have been described up to now such as the detection of pesticides [78] or various food contaminants [79] in masssensitive techniques the biorecognition element being immobilized on the surface of an oscillating crystal that resonates at a specific frequency, with the frequency of oscillation depending on the electrical frequency applied to crystal as well as the mass of the crystal. The interactions of the analyte with the biorecognition elements sitting on the crystal cause mass changes that change the frequencies which can be analyzed to the measurable signal. Most of the biosensors based on this method make use of piezoelectric materials. Piezoelectric sensors are much simpler and cheaper than the SPR systems, but are sensitive to environmental temperature changes [80–82]. Microcantilever biosensors are another group of mass-sensitive biosensors that are based on the bending of microfabricated silicon cantilevers. In addition to the mass change due to various surface interactions, the microcantilevers also undergo bending if the molecular adsorption is confined to a single surface of a microcantilever [83]. This cantilever bending is due to a differential surface stress caused by the forces involved in the adsorption process and is amplified by making the cantilever surfaces chemically different. Lack of specificity, is overcome by using the extremely selective biochemical reactions such as receptor–ligand, antibody–antigen or enzyme–substrate reactions [83]. This method, although miniaturized, requires large-scale sophisticated instrumentation which prevents its use for now as a routine tool.

8.3

Monitoring of specific endocrine-disrupting chemicals in food and environmental fields

It is now recognized that many different EDCs have been released into the environment in large quantities since World War II. Evidence already exists that a number of EDCs have reached detectable concentrations in aquatic food sources that can lead to substantial functional deficits in animals that consume this food, and therefore may then be hazardous to human health [84]. Table 8.2 summarizes some examples of biosensor applications for the detection of EDCs.

8.3.1 Pesticides The Environmental Protection Agency (EPA) defines a pesticide as ‘any substance or mixture of substances intended for preventing, destroying, repelling, or lessening the damage of any pest’. The Codex Alimentarius Commission of the United Nations’ Food and Agriculture Organization (FAO) and the World Health Organization (WHO) has established maximal

PCB

Phenol

Entrapment

Naphthalene

Phenol

Enzyme tyrosinase

Screen-printing

PCB Surfactant

Antibodies

Covalent bonding

4,4bis-(4-ydroxyphenyl) valeric acid (BVA) BPA

Entrapment

Screen-printing Entrapment

Covalent bonding

Sol–gel

Estrogen receptors

Biotine–Avidine

17β-Estradiol

DNA Pseudomonas and Achromobacter Enzyme tyrosinase

Antibodies

Sphingomonas sp. B1 or Pseudomonas fluorescens WW4 DNA

Estrogen receptors

Covalent bonding

Naphthalene and phenanthrene Aroclors and trichlorophenoxybutyrate PCB Sodium dodecyl sulfate

PAH

Estrogen receptor Estrogen receptor Bisphenol A Bisphenol A PAH

Enzyme AChE Enzyme AChE AChE OPH OPH Enzyme tyrosinase Antibodies

Steroids

LB film Sol–gel Screen-printing Covalent bonding Covalent bonding Cross-linking

Biorecognition element

Covalent bonding

Organophosphorus Carbaryl and dichlorvos Malaoxon and paraoxon Organophosphorus Paraoxon Atrazine

Pesticides Pesticides Pesticides Pesticides Pesticides Pesticides

Immobilization technique

Estrone, isoproturon and atrazine Estrogen

Analyte

Examples of different techniques for detection EDCs

EDC group

Table 8.2

Electrochemical (amperometric)

Electrodes Electrodes

Optical (fluorescence) Fiber optic

Electrochemical (amperometric) Electrochemical (amperometric)

SPR

Piezoelectric

Electrodes

Fiber optic Fiber optic Electrodes Microcantilever Optical Electrodes (amperometric) Fiber optic

Transducer

[114]

[153]

0–10 mg L−1

0.47 nA nm−1

[152]

1.2 nA mg−1 L−1

[159] [167]

[147]

0.1–12 μM

0.2 mg−1 L−1 0.2 kg m−3

[143]

0.4 μg L−1

[158]

[138]

7.8 nmol L−1

1 ppm

[137]

[129]

[91] [92] [94] [98] [99] [104]

Reference

0.155, 0.046, and 0.084 mμg/l 10–9 m

Up to 2 ppm 108 and 5.2 g L−1 0.6 and 0.5 μg L−1 10−7 m 20−6 m 5 μmol dm−3

Detection range

Biosensors for endocrine disruptors

195

residue limits for pesticides in a variety of foods [85]. The presence of pollutants with adverse effects on human androgen receptors has in fact been previously reported as a result of intensive farming and agricultural effluents [86, 87]. Moreover animal studies have shown that prenatal exposure to some pesticides such as methoxychlor or phthalates can reduce spermatogenesis [88, 89]. The use of chlornitrofen (2,4,6-trichlorophenyl-4′nitrophenyl ether; CNP) is a good example of a human healthcare threat due to indiscreet use of the pesticides. CNP was widely used in large quantities as a herbicide to control various weeds in rice fields in Japan during the period 1965–1994. This herbicide was produced and used mostly in Japan [90]. Several studies reported unusually high levels of CNP residue in freshwater fish and shellfish during the application period [91, 92]. Thus, immoderate use of herbicides leads to high accumulations of CNP in fish and shellfish in lakes and seas surrounding areas of rice cultivation. CNP entered the food chain and there is a suspicion that the high mortality rates due to biliary tract cancer in the areas where the pesticides were widely used may be related to CMP concentrations [93]. Over the last decade, AChE biosensors have emerged as an ultra-sensitive and rapid technique for toxicity analysis in environmental monitoring, food and quality control. The many AChE biosensor applications have been developed for the detection of organophosphorus (OP), carbamate, and many other groups of pesticides [94]. These enzymes can be incorporated into various transducer and interface techniques, such as the fiber optic measurements in the color change of the substrate [95] or color and pH changes tracing [96, 97] and other possible variations [94]. Despite their high sensitivity, AChE-based biosensors are inhibited by neurotoxins (OP pesticides, carbamate pesticides, etc.) and therefore cannot be used for quantitation of either an individual or a class of pesticides. Genetic engineering helped develop new enzymes for certain analytes or families. For example, biosensors with three different genetically engineered Drosophila melanogaster AChEs were created and tested for the quantitative detection of paraoxon and carbofuran pesticides [98]. A highly sensitive amperometric engineered D. melanogaster AChE-based biosensor has been developed for the determination of various pesticides (e.g., carbaryl, carbofuran and pirimicard) [99]. The comprehensive profile of different expression systems for the production of recombinant AChEs was reviewed by Schulze et al. [100]. Another popular group of enzymes applied in the pesticide biosensor field are the organophosphorus hydrolase (OPH) enzymes or other OPdegrading enzymes. OP compounds are potent cholinesterase inhibitors, accounting for their widespread use as pesticides (paraoxon or soman) and chemical warfare agents. OPH is an ideal biocatalyst because of its broad substrate specificity, stability over broad pH and temperature ranges and lack of requirement for expensive cofactors [101]. There are numerous optical and electrochemical methods [78, 102–105] used for detecting and identifying OP agents.

196

Endocrine-disrupting chemicals in food

A further group of biosensors, based on the inhibition of tyrosinase [106], has been used to detect various pesticides, e.g. cyanide [107], diethyldithiocarbamates [108], hydrazines [109], and diazinon and dichlorvos [110]. In some cases, immunosensors are preferable because they are extremely sensitive, inexpensive, easy to perform, and do not require sophisticated instruments. An amperometric immunosensor based on alkaline phosphatase (AP) inhibition were tested in the detection of 2,4-dichlorophenoxyacetic acid (2,4-D) in water [111]. In another application, the ‘River Analyzer’ (RIANA) immunosensor has been applied for the determination of pesticides such as atrazine, simazine, isoproturon, 2,4-D, alachlor, estrone and paraquat in natural waters [112–114]. The atrazine pesticide may be tracked by optical immunosensors based on the evanescent wave principle [115] or by piezoelectric immunosensors [116–118], by plant tissue electrode [44] and finally by an electrochemical method [119]. Other piezoelectric or optical immunosensor applications were used to measure 2,4-D [120], estrone [121], triazine [122] and parathion [123] pesticides. The herbicides such as atrazine, simazine, isoproturon, diuron, phenylurea and triazine may inhibit photosynthetic electron flow by blocking the photosystem II (PSII) quinone-binding site and therefore, owing to change in the chlorophyll fluorescence, are able to track the toxic compounds [124, 125].

8.3.2 Steroids Among the various compounds considered as emerging pollutants, steroid sex hormones and synthetic steroids are of particular concern, both because of the volume of these substances used and because of their activity as endocrine disruptors at very low concentrations. Increasing concentrations of these EDCs in an organism could result in disruption of the natural endocrine system. Only limited knowledge is available about these EDCs effect in long-term and low-level exposure on human and wildlife health. For example, each year in the US, 98 million cattle are raised, and hormones are used to increase their weight in 63% of them. Of the six hormones used in bulk-inducing cocktails in the US various combinations of up to three hormones, estradiol, progesterone and testosterone are natural hormones, and zeranol, trenbolone acetate and melengestrol acetate are synthetic hormones. The meat of these animals with residues of the various EDCs continues up the food chain to humans. Furthermore animal excreta contaminate the soil and water with EDC near the farms [126] and these may continue into the watershed. The concentrations of these contaminants in water may be very low [127, 128], yet many publications acknowledge the dangerous consequences of their presence in the aquatic environment [129– 131]. There are only a few biosensor applications for the detection of steroids in different matrixes [132]. One good example is the use of an optical immunosensor for detection of estrone and other organic pollutants in real water samples [19, 133].

Biosensors for endocrine disruptors

197

8.3.3 Estrogen receptor biosensors As is mentioned before, EDCs are a class of emerging contaminants that are not defined by their chemical nature but by their biological effect. Therefore the EDC determinations require the development of methods based on the monitoring of the biological effect, rather than on chemical analysis. EDCs interfere with endogenous hormone systems, and many of them can bind to the natural estrogen receptor (ER) as agonists or antagonists. This binding ability of ER can be applied for creating biosensors with natural receptors for testing chemicals with potential environmental toxicity. The main advantage of these biosensors is their simplicity and ability for detection of major groups of EDCs particularly those that directly affect the ERs [134, 135]. For the evaluation of estrogens and xenoestrogens in water samples human ER was used in an SPR-based biosensor (BIAcore) [136–138]. This system can be used to estimate the estrogenic potential of chemicals, for drinking water control and environmental monitoring. Additional optical biosensors based on recombinant cells expressing human ER were developed recently and may be used in water quality monitoring [139, 140]. Other methods have been used in ER-based biosensors applications, e.g. electrochemical [106, 141] and piezoelectric [142, 143]. A biosensor with genetically modified S. cerevisiae yeast cells entrapped in alginate hydrogel matrices, has been developed and tested for EDC determination [19]. The advantage of this application is the possibility of in-field determination of various EDCs in water samples. Another wholecell biosensor uses the recombinant fluorescent Chinese hamster ovary cell line to monitor various toxicants, especially EDCs, in diverse aqueous environments [144]. In some cases, detection of biomarkers, such as vitellogenin, will indicate the presence of endocrine substances (preferable over direct detection of EDCs [134]) as exposure of EDCs to fish will produce high amounts of vitellogenin easily monitored by biosensors [145, 146].

8.3.4 Bisphenol A Bisphenol A (BPA) is a chemical used in the production of epoxy resin polycarbonate plastics or degradation products of TBBPA. It may be found in food and drink packages, baby bottles and lacquer coating of certain metal products. Due to its EDC activity biosensors have been developed, the majority of them using antibodies (immunosensors) measured by SPR [147], total internal reflection fluorescence (TIRF) [148], or piezoelectric effect [149]. Very recently the SPR technique has been applied to the measurement of BPA through the use of transport proteins. Some amperometric biosensors with enzymatic biorecognition elements have been realized for the detection of BPA [106, 150] as well as different tyrosinase carbon paste modified electrodes [151].

198

Endocrine-disrupting chemicals in food

8.3.5 Polycyclic aromatic hydrocarbons PAHs are ubiquitous contaminants and primarily occur as a result of incomplete combustion processes [152]. The carcinogenic effect of some PAHs are well known, and some have been identified as potential environmental endocrine disruptors. PAHs have two modes of action, by blocking the activation of estrogen receptors or by induction of Ah-responsive genes that result in a broad spectrum of antiestrogenic responses [153, 154]. Two different techniques, immobilization of the recombinant bioluminescent bacterial cells and use of a non-toxic biosurfactant, were combined to develop an in situ toxicity biosensor system for phenanthrene detection in soil [155]. Naphthalene and phenanthrene are the most water-soluble PAHs, so they are priority pollutants in aqueous solutions. For those reasons amperometric biosensors for naphthalene were developed using either Sphingomonas sp. B1 or Pseudomonas fluorescens WW4 cells immobilized within a polyurethane-based hydrogel [156]. Another biosensor application for the detection of naphthalene and phenanthrene use DNA as a biorecognition element [157]. The PAH Benzo[a]pyrene (BaP) was also monitored using a fiber optic fluoroimmunosensor, thorough laser excitation of fluorescent BaP [158]. 8.3.6 Polychlorinated compounds PCBs have been used in many different products, including electrical equipment, surface coatings, inks and adhesives. PCBs may be released into the environment, for instance when waste that contains PCBs is incinerated or stored in landfills. Because of the possible impacts on human health and the environment, the use and production of PCBs are now banned or severely restricted in many countries, but about 10% of the PCBs produced since 1929 persist in the environment today. The majority of studies found that PCBs have estrogenic activity [159–161]. A fiber optic immunosensor, consisting of a quartz fiber coated with partially purified polyclonal antiPCB antibodies (Abs), was used to detect PCBs [162]. Various other applications have been developed for the detection of PCBs, e.g. DNA biosensor based on chronopotentiometric [163] and electrochemical [164] methods, SPR system [165], the BIAcore 2000, in combination with an immunoassay, screen-printed electrodes based on the use of AP as a label in an indirect ELISA format [166] and whole-cell biosensors based on the reporter strain Ralstonia eutropha ENV307 [167]. Recently an additional chemical group within polychlorinated compounds, called dioxins, have been included in the lists of EDCs [165] and for which immunoassays [168, 169] and SPR methods were devised [165]. 8.3.7 Surfactants Among EDCs surfactants represent an increasingly vast range of organic compounds, with marked endocrine effects, which are found in large

Biosensors for endocrine disruptors

199

quantities in the water cycle. Methods based on the change of the oxygen electrode current as a result of decrease in dissolved oxygen levels in solution due to consumption of surfactant by bacterial strains Pseudomonas and Achromobacter have been described [170, 171]. Other surfactants used in variety of industrial applications and which demonstrate endocrine-disrupting properties of their metabolic products (APs) and carboxylic derivatives (APECs) called alkylphenol ethoxylates (APEOs) [130, 172]. A capillarybased flow immunoassay, using glucose dehydrogenase (DH) as the label marker, was developed and tested on both APEOs and APs [173]. 8.3.8 Phenols The determination of phenol derivates is very important in food, medical, and both environmental ground- and surface water. Some of them pose a danger to the human health because of their inherent toxicity. Some amperometric techniques were developed for the detection of phenol derivates in environmental samples. For example biosensors were developed with either an immobilized hygrogel on a graphite electrode (e.g., biosensor based on tyrosinase – a polyphenol oxidase with a relatively wide selectivity for phenolic compounds) [118] or are bioelectroanalytical electrode based on simple and reproducible multiwalled carbon nanotube–Nafion–Tyr nanobiocomposite film coated with tyrosinase [174]. 8.3.9 Tributyltin Tributyltin (TBT) compounds are a subgroup of the trialkyl organotin family of compounds. They have been extensively used as biocides in many fields such as wood treatment and preservation, antifouling of boats (in marine paints), antifungal action in textiles and industrial water systems (cooling tower and refrigeration water systems), wood pulp and paper mill systems, and breweries. The high toxicity of TBT, toxic at nanogram per liter levels, led to the settling of oyster spat and caused oyster shell thickening and growth inhibition. The main source of organotin intake for humans is contaminated fish and seafood. A recent study demonstrated that exposure of rats to organotins severely affected pregnancy outcome and perinatal survival of rats offspring [175]. A biosensor based on genetically modified bacteria, with a specific TBT-sensitive DH1 chromosome fused with luxAB, were developed and found suitable for on-line and in situ TBTs measurements in water [176].

8.4

Future trends

The field of biosensors for potential food and environmental applications has seen great advancement in the past decade in areas such as the development of new immobilization processes, enzymatic and microorganism

200

Endocrine-disrupting chemicals in food

genetic modifications, and development and improvement of transducer techniques. The use of synthetic material with biomimetic-based functions has increased. In the future, new techniques based on biomimetic recognition elements will be a part of the evolving biosensor field owing to the potential of biomimetic elements to overcome some of the shortfalls associated with biological components, primarily poor stability and higher cost of production. Another field that achieved great advances is whole-cell biosensors. Development of genetic techniques will allow creation of a bettercharacterized bioreporter organism with improved reversibility and the possibility of detection of a more diverse groups of contaminants. Furthermore, genetic modification techniques are not only suitable for whole-cell biosensors but also show potential for improving the enzymatic specificity and variability. The main disadvantage of many current biosensors is the deficiency for multi-analyte detection. Thus the future will likely focus on the construction of multi-arrays based on immobilized enzymes, DNA, or other biorecognition elements that will allow the detection of hundreds of totally separate compounds. In spite of the vast technology improvement of biosensors, they are still complicated and only a few types have enjoyed successful marketing.

8.5

Acknowledgments

R.M. thanks the Israel Ministry of Science for financial support (#34137).

8.6

References

1. petrovic, m., et al., Analysis and environmental levels of endocrine-disrupting compounds in freshwater sediments. Trac Trends Anal Chem, 2001. 20(11): p. 637–48. 2. brenner, a., et al., Biodegradability of tetrabromobisphenol A and tribromophenol by activated sludge. Ecotoxicology, 2006. 15(4): p. 399–402. 3. antizar-ladislao, b., Environmental levels, toxicity and human exposure to tributyltin (TBT)-contaminated marine environment. A review. Environ Int, 2007. 34(2): 292–308. 4. sonnenschein, c. and a.m. soto, An updated review of environmental estrogen and androgen mimics and antagonists. J Steroid Biochem Mol Biol, 1998. 65(1– 6): p. 143–50. 5. kuhl, a.j. and m. brouwer, Antiestrogens inhibit xenoestrogen-induced brain aromatase activity but do not prevent xenoestrogen-induced feminization in Japanese medaka (Oryzias latipes). Environ Health Perspect, 2006. 114(4): p. 500–6. 6. fry, d.m. and c.k. toone, DDT-induced feminization of gull embryos. Science, 1981. 213(4510): p. 922–4. 7. facemire, c.f., t.s. gross, and l.j. guillette, jr, Reproductive impairment in the Florida panther: nature or nurture? Environ Health Perspect, 1995. 103 Suppl 4: p. 79–86.

Biosensors for endocrine disruptors

201

8. Norgil damgaard, i., et al., Impact of exposure to endocrine disrupters in utero and in childhood on adult reproduction. Best Pract Res Clin Endocrinol Metab, 2002. 16(2): p. 289–309. 9. carlsen, e., et al., Evidence for decreasing quality of semen during past 50 years. BMJ, 1992. 305(6854): p. 609–13. 10. giwercman, a., et al., Evidence for increasing incidence of abnormalities of the human testis: a review. Environ Health Perspect, 1993. 101 Suppl 2: p. 65–71. 11. rhind, s.m., Endocrine disrupting compounds and farm animals: their properties, actions and routes of exposure. Domest Anim Endocrinol, 2002. 23(1–2): p. 179–87. 12. guenther, k., et al., Endocrine disrupting nonylphenols are ubiquitous in food. Environ Sci Technol, 2002. 36(8): p. 1676–80. 13. isobe, t., et al., Distribution and behavior of nonylphenol, octylphenol, and nonylphenol monoethoxylate in Tokyo metropolitan area: their association with aquatic particles and sedimentary distributions. Environ Sci Technol, 2001. 35(6): p. 1041–9. 14. houtman, c.j., et al., Toxicological profiling of sediments using in vitro bioassays, with emphasis on endocrine disruption. Environ Toxicol Chem, 2004. 23(1): p. 32–40. 15. de meulenaer, b., et al., Development of an enzyme-linked immunosorbent assay for bisphenol a using chicken immunoglobulins. J Agric Food Chem, 2002. 50(19): p. 5273–82. 16. cliquet, p., et al., Generation of group-specific antibodies against sulfonamides. J Agric Food Chem, 2003. 51(20): p. 5835–42. 17. barcelo, d. and a. kettrup, Endocrine disruptors. Anal Bioanal Chem, 2004. 378(3): p. 547–8. 18. rodriguez-mozaz, s., et al., Biosensors for environmental monitoring – A global perspective. Talanta, 2005. 65(2): p. 291–7. 19. fine, t., et al., Luminescent yeast cells entrapped in hydrogels for estrogenic endocrine disrupting chemical biodetection. Biosensors Bioelectronics, 2006. 21(12): p. 2263–9. 20. patel, p.d., (Bio)sensors for measurement of analytes implicated in food safety: a review. Trac Trends Analytical Chem, 2002. 21(2): p. 96–115. 21. marks, r.s., et al., Handbook of biosensors and biochips. 2007: John Wiley & Sons, Ltd. 22. vo-dinh, t. and b. cullum, Biosensors and biochips: advances in biological and medical diagnostics. Fresenius J Anal Chem, 2000. 366(6–7): p. 540–51. 23. mehrvar, m. and m. abdi, Recent developments, characteristics, and potential applications of electrochemical biosensors. Anal Sci, 2004. 20(8): p. 1113–26. 24. sirkar, k., a. revzin, and m.v. pishko, Glucose and lactate biosensors based on redox polymer/oxidoreductase nanocomposite thin films. Anal Chem, 2000. 72(13): p. 2930–6. 25. thevenot, d.r., et al., Electrochemical biosensors: recommended definitions and classification. Biosens Bioelectron, 2001. 16(1–2): p. 121–31. 26. waibel, m., et al., Screen-printed bienzymatic sensor based on sol–gel immobilized Nippostrongylus brasiliensis acetylcholinesterase and a cytochrome P450 BM-3 (CYP102-A1) mutant. Biosens Bioelectron, 2006. 21(7): p. 1132–40. 27. bucur, b., et al., Biosensors based on highly sensitive acetylcholinesterases for enhanced carbamate insecticide detection. Anal Chim Acta, 2006. 562(1): p. 115–21. 28. herrmann, s., et al., Chemiluminescent optical fiber immunosensor for the detection of anti-West Nile virus IgG. Talanta, 2005. 66(1): p. 6–14. 29. torres, c.m., y. pico, and j. manes, Determination of pesticide residues in fruit and vegetables. J Chromatogr A, 1996. 754(1–2): p. 301–31.

202

Endocrine-disrupting chemicals in food

30. yang, s.s., a.i. goldsmith, and i. smetena, Recent advances in the residue analysis of N-methylcarbamate pesticides. J Chromatogr A, 1996. 754(1–2): p. 3–16. 31. seifert, m., s. haindl, and b. hock, Development of an enzyme linked receptor assay (ELRA) for estrogens and xenoestrogens. Anal Chim Acta, 1999. 386(3): p. 191–9. 32. gu, m.b., r.j. mitchell, and b.c. kim, Whole-cell-based biosensors for environmental biomonitoring and application. Adv Biochem Eng Biotechnol, 2004. 87: p. 269–305. 33. bousse, l., Whole cell biosensors. Sensors Actuators B: Chem, 1996. 34(1–3): p. 270–5. 34. polyak, b., et al., Optical fiber bioluminescent whole-cell microbial biosensors to genotoxicants. Water Sci Technol, 2000. 42(1–2): p. 305–311. 35. polyak, b., et al., Bioluminescent whole cell optical fiber sensor to genotoxicants: system optimization. Sensors Actuators B – Chem, 2001. 74(1–3): p. 18–26. 36. polyak, b., s. geresh, and r.s. marks, Synthesis and characterization of a biotin–alginate conjugate and its application in a biosensor construction. Biomacromolecules, 2004. 5(2): p. 389–96. 37. eltzov, e., et al., Detection of sub-inhibitory antibiotic concentrations via luminescent sensing bacteria and prediction of their mode of action. Sensors Actuators B: Chem, 2008. 129(2): p. 685–92. 38. hakkila, k., et al., Detection of bioavailable heavy metals in EILATox-Oregon samples using whole-cell luminescent bacterial sensors in suspension or immobilized onto fibre-optic tips. J Appl Toxicol, 2004. 24(5): p. 333–42. 39. pancrazio, j.j., et al., EILATox-Oregon biomonitoring workshop: summary and observations. J Appl Toxicol, 2004. 24(5): p. 317–21. 40. ivask, a., et al., Fibre-optic bacterial biosensors and their application for the analysis of bioavailable Hg and As in soils and sediments from Aznalcollar mining area in Spain. Biosens Bioelectron, 2007. 22(7): p. 1396–402. 41. michelini, e., et al., A new recombinant cell-based bioluminescent assay for sensitive androgen-like compound detection. Biosens Bioelectron, 2005. 20(11): p. 2261–7. 42. sidwell, j.s. and g.a. rechnitz, Progress and challenges for biosensor using plant tissue materials. Biosensors, 1986. 2(4): p. 221–233. 43. wijesuriya, d.c. and g.a. rechnitz, Biosensors based on plant and animal tissues. Biosens Bioelectron, 1993. 8(3–4): p. 155–60. 44. mazzeia, f., et al., Plant tissue electrode for the determination of atrazine. Anal Chim Acta, 1995. 316(1): p. 79–82. 45. shoji, r., et al., Development of a rapid and sensitive bioassay device using human cells immobilized in macroporous microcarriers for the on-site evaluation of environmental waters. Appl Microbiol Biotechnol, 2000. 54(3): p. 432–8. 46. wang, j., et al., DNA electrochemical biosensors for environmental monitoring. A review. Anal Chim Acta, 1997. 347(1–2): p. 1–8. 47. meir, a., r.s. marks, and m.N. stojanovic, Aptameric biosensors, in Handbook of biosensors and biochips, R.S. Marks, et al., Editors. 2007, John Wiley & Sons, Ltd. p. 217–32. 48. pichon, v., Selective sample treatment using molecularly imprinted polymers. J Chromatogr A, 2007. 1152(1–2): p. 41–53. 49. collings, a.f. and f. caruso, Biosensors: recent advances. Rep Prog Phys, 1997. 60(11): p. 1397–445. 50. d’souza, s.f., Microbial biosensors. Biosens Bioelectron, 2001. 16(6): p. 337–53. 51. mello, l.d. and l.t. kubota, Review of the use of biosensors as analytical tools in the food and drink industries. Food Chem, 2002. 77(2): p. 237–56.

Biosensors for endocrine disruptors

203

52. starodub, n.f., i.m. samodumova, and v.n. starodub, Usage of organosilanes for integration of enzymes and immunocomponents with electrochemical and optical transducers. Sensors Actuators B: Chem, 1995. 24(1–3): p. 173–76. 53. scouten, w.h., j.h.t. luong, and r.s. brown, Enzyme or protein immobilization techniques for applications in biosensor design. Trends Biotechnol, 1995. 13(5): p. 178–85. 54. premkumar, j.r., et al., Antibody-based immobilization of bioluminescent bacterial sensor cells. Talanta, 2001. 55(5): p. 1029–38. 55. kuswandi, b., r. andres, and r. Narayanaswamy, Optical fibre biosensors based on immobilised enzymes. Analyst, 2001. 126(8): p. 1469–91. 56. tyagi, r., r. batra, and m.n. gupta, Amorphous enzyme aggregates: stability toward heat and aqueous-organic cosolvent mixtures. Enzyme Microbial Technol, 1999. 24(5–6): p. 348–54. 57. avnir, d., et al., Enzymes and other proteins entrapped in sol–gel materials. Chem Mat, 1994. 6(10): p. 1605–14. 58. wang, j., p.v.a. pamidi, and d.r. zanette, Self-assembled silica gel networks. J Am Chem Soc, 1998. 120(23): p. 5852–3. 59. dong, s. and x. chen, Some new aspects in biosensors. Rev Mol Biotechnol, 2002. 82(4): p. 303–23. 60. tess, m.e. and j.a. cox, Chemical and biochemical sensors based on advances in materials chemistry. J Pharmaceut Biomed Anal, 1999. 19(1–2): p. 55–68. 61. kawabata, n., s. nishimura, and t. yoshimura, New method of immobilization of microbial-cells by capture on the surface of insoluble pyridinium-type resin. Biotechnol Bioeng, 1990. 35(10): p. 1000–5. 62. keusgen, m., Biosensors: new approaches in drug discovery. Naturwissenschaften, 2002. 89(10): p. 433–44. 63. lei, y., w. chen, and a. mulchandani, Microbial biosensors. Anal Chim Acta, 2006. 568(1–2): p. 200–10. 64. ganesh, v., et al., Self-assembled monolayers (SAMs) of alkoxycyanobiphenyl thiols on gold surface using a lyotropic liquid crystalline medium. Electrochim Acta, 2007. 52(9): p. 2987–97. 65. arimaa, y. and h. iwata, Effect of wettability and surface functional groups on protein adsorption and cell adhesion using well-defined mixed self-assembled monolayers. Biomaterials, 2007. 28(20): p. 3074–82. 66. wink, t., et al., Self-assembled monolayers for biosensors. Analyst, 1997. 122(4): p. R43–R50. 67. ottova, a., et al., Self-assembled BLMs: biomembrane models and biosensor applications. Supramolecular Sci, 1997. 4(1–2): p. 101–12. 68. ye, j.s., et al., Nanostructured platinum-lipid bilayer composite as biosensor. Bioelectrochemistry, 2003. 59(1–2): p. 65–72. 69. kozana, j.v.b., et al., Biosensing hydrogen peroxide utilizing carbon paste electrodes containing peroxidases naturally immobilized on coconut (Cocus nucifera L.) fibers. Anal Chim Acta, 2007. 591(2): p. 200–7. 70. rawson, f.j., et al., Fabrication and characterisation of novel screen-printed tubular microband electrodes, and their application to the measurement of hydrogen peroxide. Electrochim Acta, 2007. doi:10.1016/j.electacta.2007.05.062. 71. leunga, a., p.m. shankarb, and r. mutharasan, A review of fiber-optic biosensors. Sensors Actuators B: Chem, 2007. 125(2): 688–703. 72. wolfbeis, o.s., Fiber-optic chemical sensors and biosensors. Anal Chem, 2006. 78(12): p. 3859–74. 73. d’orazio, p., Biosensors in clinical chemistry. Clin Chim Acta, 2003. 334(1–2): p. 41–69. 74. ziegler, c. and w. gopel, Biosensor development. Curr Opin Chem Biol, 1998. 2(5): p. 585–91.

204

Endocrine-disrupting chemicals in food

75. paddle, b.m., Biosensors for chemical and biological agents of defence interest. Biosens Bioelectron, 1996. 11(11): p. 1079–113. 76. wang, j., Electrochemical nucleic acid biosensors. Anal Chim Acta, 2002. 469(1): p. 63–71. 77. ramanathan, k. and b. danielsson, Principles and applications of thermal biosensors. Biosens Bioelectron, 2001. 16(6): p. 417–23. 78. zheng, y.h., et al., Detection of dichlorvos residue by flow injection calorimetric biosensor based on immobilized chicken liver esterase. J Food Eng, 2006. 74(1): p. 24–9. 79. ramanathan, k., et al., The development and applications of thermal biosensors for bioprocess monitoring. Trends Biotechnol, 1999. 17(12): p. 499–505. 80. leonard, p., et al., Advances in biosensors for detection of pathogens in food and water. Enzyme Microbial Technol, 2003. 32(1): p. 3–13. 81. deng, t., et al., A novel immunosensor based on self-assembled chitosan/ alginate multilayers for the detection of factor B. Sensors Actuators B: Chem, 2004. 99(1): p. 123–9. 82. wang, j., m. jiang, and e. palecek, Real-time monitoring of enzymatic cleavage of nucleic acids using a quartz crystal microbalance. Bioelectrochem Bioenerg, 1999. 48(2): p. 477–80. 83. hansen, k.m. and t. thundat, Microcantilever biosensors. Methods, 2005. 37(1): p. 57–64. 84. colborn, t., f.s.v. saal, and a.m. soto, Developmental effects of endocrinedisrupting chemicals in wildlife and humans. Environ Health Perspect, 1993. 101(5): p. 378–84. 85. ahmed, f.e., c.f. moffat, and k.j. whittle, Environmental contaminants in food, Sheffield Academic Press, 1999. p. 500. 86. lemaire, g., et al., Effect of organochlorine pesticides on human androgen receptor activation in vitro. Toxicol Appl Pharmacol, 2004. 196(2): p. 235–46. 87. jenkins, r.l., et al., Androstenedione and progesterone in the sediment of a river receiving paper mill effluent. Toxicol Sci, 2003. 73(1): p. 53–9. 88. hotchkiss, a.k., et al., Androgens and environmental antiandrogens affect reproductive development and play behavior in the Sprague-Dawley rat. Environ Health Perspect, 2002. 110 Suppl 3: p. 435–9. 89. moore, r.w., et al., Abnormalities of sexual development in male rats with in utero and lactational exposure to the antiandrogenic plasticizer di(2-ethylhexyl) phthalate. Environ Health Perspect, 2001. 109(3): p. 229–37. 90. sakurai, t., Dioxins in aquatic sediment and soil in the Kanto region of Japan: major sources and their contributions. Environ Sci Technol, 2003. 37(14): p. 3133–40. 91. yamagishi, t. and k. akiyama, 1,3,5-Trichloro-2-(4-nitrophenoxy)benzene in fish, shellfish, and seawater in Tokyo Bay, 1977–1979. Arch Environ Contam Toxicol, 1981. 10(5): p. 627–35. 92. ohyama, t., et al., 1,3,5-Trichloro-2-(4-nitrophenoxy)benzene (CNP) in water, sediments, and shellfish of the Ishikari River. Bull Environ Contam Toxicol, 1986. 37(3): p. 344–9. 93. yamamoto, m., et al., Ecological correlation between the use of agricultural chemicals and biliary tract cancers in Japan. Acta Med Biolo, 1987. 35: p. 63–8. 94. andreescua, s. and j.l. marty, Twenty years research in cholinesterase biosensors: From basic research to practical applications. Biomol Eng, 2006. 23(1): p. 1–15. 95. choi, j.w., et al., Optical organophosphorus biosensor consisting of acetylcholinesterase/viologen hetero Langmuir-Blodgett film. Biosens Bioelectron, 2001. 16(9–12): p. 937–43.

Biosensors for endocrine disruptors

205

96. andreou, v.g. and y.d. clonis, A portable fiber-optic pesticide biosensor based on immobilized cholinesterase and sol–gel entrapped bromcresol purple for in-field use. Biosens Bioelectron, 2002. 17(1–2): p. 61–9. 97. andres, r.t. and r. narayanaswamy, Fibre-optic pesticide biosensor based on covalently immobilized acetylcholinesterase and thymol blue. Talanta, 1997. 44(8): p. 1335–52. 98. bachmann, t.t., et al., Improved multianalyte detection of organophosphates and carbamates with disposable multielectrode biosensors using recombinant mutants of Drosophila acetylcholinesterase and artificial neural networks. Biosens Bioelectron, 2000. 15(3–4): p. 193–201. 99. bucur, b., et al., Biosensors based on highly sensitive acetylcholinesterases for enhanced carbamate insecticides detection. Anal Chim Acta, 2006. 562(1): p. 115–21. 100. schulze, h., et al., Development, validation, and application of an acetylcholinesterase-biosensor test for the direct detection of insecticide residues in infant food. Biosens Bioelectron, 2002. 17(11–12): p. 1095–105. 101. mcdaniel, c.s., et al., Enzyme-based additives for paints and coatings. Prog Organic Coatings, 2006. 55(2): p. 182–8. 102. karnati, c., et al., Organophosphorus hydrolase multilayer modified microcantilevers for organophosphorus detection. Biosens Bioelectron, 2007. 22(11): p. 2636–42. 103. simoniana, a.l., et al., Nanoparticle-based optical biosensors for the direct detection of organophosphate chemical warfare agents and pesticides. Anal Chim Acta, 2005. 534(1): p. 69–77. 104. daunert, s., et al., Calmodulin-mediated reversible immobilization of enzymes. Colloids Surf B Biointerfaces, 2007. 58(1): p. 20–7. 105. deo, r.p., et al., Determination of organophosphate pesticides at a carbon nanotube/organophosphorus hydrolase electrochemical biosensor. Anal Chim Acta, 2005. 530(2): p. 185–9. 106. dempsey, e., d. diamond, and a. collier, Development of a biosensor for endocrine disrupting compounds based on tyrosinase entrapped within a poly(thionine) film. Biosens Bioelectron, 2004. 20(2): p. 367–77. 107. smit, m.h. and g.a. rechnitz, Toxin detection using a tyrosinase-coupled oxygenelectrode. Anal Chem, 1993. 65(4): p. 380–5. 108. mcardle, f.a. and k.c. persaud, Development of an enzyme-based biosensor for atrazine detection. Analyst, 1993. 118(4): p. 419–23. 109. wang, j. and l. chen, Hydrazine detection using a tyrosinase-based inhibition biosensor. Anal Chem, 1995. 67(20): p. 3824–7. 110. everett, w.r. and g.a. rechnitz, Mediated bioelectrocatalytic determination of organophosphorus pesticides with a tyrosinase-based oxygen biosensor. Anal Chem, 1998. 70(4): p. 807–10. 111. wilmer, m., et al., Amperometric immunosensor for the detection of 2,4dichlorophenoxyacetic acid (2,4-D) in water. Anal Lett, 1997. 30(3): p. 515–25. 112. mallat, e., et al., Part per trillion level determination of isoproturon in certified and estuarine water samples with a direct optical immunosensor. Anal Chim Acta, 2001. 426(2): p. 209–16. 113. mallat, e., et al., Fast determination of paraquat residues in water by an optical immunosensor and validation using capillary electrophoresis-ultraviolet detection. Anal Chim Acta, 2001. 427(2): p. 165–71. 114. mallat, e., et al., River analyzer for chlorotriazines with a direct optical immunosensor. Environ Sci Technol, 1999. 33(6): p. 965–71. 115. schipper, e.f., et al., New detection method for atrazine pesticides with the optical waveguide Mach-Zehnder immunosensor. Anal Chim Acta, 1997. 341(2–3): p. 171–6.

206

Endocrine-disrupting chemicals in food

116. minunni, m., p. skládal, and m. mascini, A piezoelectric crystal biosensor for detection of atrazine. Life Chem Rep, 1994. 10: p. 391–8. 117. yokoyama, k., et al., Highly sensitive quartz-crystal immunosensors for multisample detection of herbicides. Anal Chim Acta, 1995. 304(2): p. 139–45. 118. parellada, j., et al., Amperometric immunosensors and enzyme electrodes for environmental applications. Anal Chim Acta, 1998. 362(1): p. 47–57. 119. zaccoa, e., et al., Electrochemical biosensing of pesticide residues based on affinity biocomposite platforms. Biosens Bioelectron, 2007. 22(8): p. 1707–15. 120. minunni, m., p. skladal, and m. mascini, A piezoelectric quartz-crystal biosensor as a direct affinity sensor. Anal Lett, 1994. 27(8): p. 1475–87. 121. tschmelak, j., g. proll, and g. gauglitz, Optical biosensor for pharmaceuticals, antibiotics, hormones, endocrine disrupting chemicals and pesticides in water: Assay optimization process for estrone as example. Talanta, 2005. 65(2): p. 313–23. 122. schipper, e.f., et al., The waveguide Mach-Zender interferometer as atrazine sensor. Anal Chem, 1998. 70(6): p. 1192–7. 123. Ngehngwainbi, j., et al., Parathion antibodies on piezoelectric-crystals. J Amer Chem Soc, 1986. 108(18): p. 5444–7. 124. koblizek, m., et al., A biosensor for the detection of triazine and phenylurea herbicides designed using Photosystem II coupled to a screen-printed electrode. Biotechnol Bioeng, 2002. 78(1): p. 110–16. 125. vedrine, c., et al., Optical whole-cell biosensor using Chlorella vulgaris designed for monitoring herbicides. Biosens Bioelectron, 2003. 18(4): p. 457–63. 126. lange, i.g., et al., Sex hormones originating from different livestock production systems: fate and potential disrupting activity in the environment. Anal Chim Acta, 2002. 473(1–2): p. 27–37. 127. ternes, t.a., p. kreckel, and j. mueller, Behaviour and occurrence of estrogens in municipal sewage treatment plants – II. Aerobic batch experiments with activated sludge. Sci Total Environ, 1999. 225(1–2): p. 91–9. 128. belfroid, a.c., et al., Analysis and occurrence of estrogenic hormones and their glucuronides in surface water and waste water in The Netherlands. Sci Total Environ, 1999. 225(1–2): p. 101–8. 129. larsson, d.g.j., et al., Ethinyloestradiol – an undesired fish contraceptive? Aquatic Toxicol, 1999. 45(2–3): p. 91–7. 130. purdom, c., et al., Estrogenic effects of effluents from sewage treatment works. Chem Ecol, 1994. 8(4): p. 275–85. 131. hansen, p.d., et al., Vitellogenin – a biomarker for endocrine disruptors. Trac Trends Anal Chem, 1998. 17(7): p. 448–51. 132. de alda, m.j.l. and d. barcelo, Review of analytical methods for the determination of estrogens and progestogens in waste waters. Fresenius J Anal Chem, 2001. 371(4): p. 437–47. 133. rodriguez-mozaz, s., et al., Simultaneous multi-analyte determination of estrone, isoproturon and atrazine in natural waters by the RIver ANAlyser (RIANA), an optical immunosensor. Biosens Bioelectron, 2004. 19(7): p. 633–40. 134. oosterkamp, a.j., et al., Novel monitoring strategies for xenoestrogens. Trac Trends Anal Chem, 1997. 16(10): p. 544–53. 135. kroger, s., s. piletsky, and a.p.f. turner, Biosensors for marine pollution research, monitoring and control. Marine Poll Bull, 2002. 45(1–12): p. 24–34. 136. usami, m., k. mitsunaga, and y. ohno, Estrogen receptor binding assay of chemicals with a surface plasmon resonance biosensor. J Steroid Biochem Mol Biol, 2002. 81(1): p. 47–55. 137. hock, b., m. seifert, and k. kramer, Engineering receptors and antibodies for biosensors. Biosens Bioelectron, 2002. 17(3): p. 239–49.

Biosensors for endocrine disruptors

207

138. seifert, m., s. haindl, and b. hock, Development of an enzyme linked receptor assay (ELRA) for estrogens and xenoestrogens. Anal Chim Acta, 1999. 386(3): p. 191–9. 139. hahn, t., et al., A novel estrogen sensor based on recombinant Arxula adeninivorans cells. Biosens Bioelectron, 2006. 21(11): p. 2078–85. 140. wozei, e., s.w. hermanowicz, and h.y.n. holman, Developing a biosensor for estrogens in water samples. Study of the real-time response of live cells of the estrogen-sensitive yeast strain RMY/ER-ERE using fluorescence microscopy. Biosens Bioelectron, 2006. 21(8): p. 1654–8. 141. murata, m., et al., Novel biosensor for the rapid measurement of estrogen based on a ligand–receptor interaction. Anal Sci, 2001. 17(3): p. 387–90. 142. mo, z.h., x.h. long, and w.l. fu, A new sandwich-type assay of estrogen using piezoelectric biosensor immobilized with estrogen response element. Anal Commun, 1999. 36(7): p. 281–3. 143. sesay, a.m. and d.c. cullen, Detection of hormone mimics in water using a miniaturised SPR sensor. Environ Monitoring Assessment, 2001. 70(1–2): p. 83–92. 144. kim, e.j., et al., Application of recombinant fluorescent mammalian cells as a toxicity biosensor. Water Sci Technol, 2002. 46(3): p. 51–6. 145. darain, f., et al., Development of an immunosensor for the detection of vitellogenin using impedance spectroscopy. Biosens Bioelectron, 2004. 19(10): p. 1245–52. 146. darain, f., et al., A separation-free amperometric immunosensor for vitellogenin based on screen-printed carbon arrays modified with a conductive polymer. Biosens Bioelectron, 2005. 20(9): p. 1780–7. 147. marchesini, g.r., et al., Biosensor immunoassays for the detection of bisphenol A. Anal Chim Acta, 2005. 528(1): p. 37–45. 148. rodriguez-mozaz, s., m.l. de alda, and d. barcelo, Analysis of bisphenol A in natural waters by means of an optical immunosensor. Water Res, 2005. 39(20): p. 5071–9. 149. park, j.w., et al., Piezoelectric immunosensor for bisphenol A based on signal enhancing step with 2-methacrolyloxyethyl phosphorylcholine polymeric nanoparticle. Analyst, 2006. 131(1): p. 155–62. 150. andreescu, s. and o.a. sadik, Correlation of analyte structures with biosensor responses using the detection of phenolic estrogens as a model. Anal Chem, 2004. 76(3): p. 552–60. 151. mitaa, d.g., et al., Enzymatic determination of BPA by means of tyrosinase immobilized on different carbon carriers. Biosens Bioelectron, 2007. doi:10.1016/j.bios.2007.03.010. 152. ashley, j.t.f. and j.e. baker, Hydrophobic organic contaminants in surficial sediments of Baltimore Harbor: inventories and sources. Environ Toxicol Chem, 1999. 18(5): p. 838–49. 153. vinggaard, a.m., c. hnida, and j.c. larsen, Environmental polycyclic aromatic hydrocarbons affect androgen receptor activation in vitro. Toxicology, 2000. 145(2–3): p. 173–83. 154. arcaro, k.f., et al., Antiestrogenicity of environmental polycyclic aromatic hydrocarbons in human breast cancer cells. Toxicology, 1999. 133(2–3): p. 115–27. 155. gu, m.b. and s.t. chang, Soil biosensor for the detection of PAH toxicity using an immobilized recombinant bacterium and a biosurfactant. Biosens Bioelectron, 2001. 16(9–12): p. 667–74. 156. konig, a., et al., Microbial sensors for naphthalene using Sphingomonas sp B1 or Pseudomonas fluorescens WW4. Appl Microbiol Biotechnol, 1996. 45(6): p. 844–50.

208

Endocrine-disrupting chemicals in food

157. doong, r.a., h.m. shih, and s.h. lee, Sol-gel-derived array DNA biosensor for the detection of polycyclic aromatic hydrocarbons in water and biological samples. Sensors Actuators B Chem, 2005. 111: p. 323–30. 158. alarie, j.p., et al., Fluorescence monitoring of a benzo[a]pyrene metabolite using a regenerable immunochemical-based fiberoptic sensor. Anal Chim Acta, 1990. 236(2): p. 237–44. 159. brouwer, a., et al., Characterization of potential endocrine-related health effects at low-dose levels of exposure to PCBs. Environ Health Perspect, 1999. 107: p. 639–49. 160. gierthy, j.f., k.f. arcaro, and m. floyd, Assessment of PCB estrogenicity in a human breast cancer cell line. Chemosphere, 1997. 34(5–7): p. 1495–505. 161. connor, k., et al., Hydroxylated polychlorinated biphenyls (PCBs) as estrogens and antiestrogens: structure–activity relationships. Toxicol Appl Pharmacol, 1997. 145(1): p. 111–23. 162. zhao, c.q., et al., Fiber optic immunosensor for polychlorinated-biphenyls. J Agric Food Chem, 1995. 43(8): p. 2308–15. 163. marrazza, g., i. chianella, and m. mascini, Disposable DNA electrochemical biosensors for environmental monitoring. Anal Chim Acta, 1999. 387(3): p. 297–307. 164. mascini, m., Affinity electrochemical biosensors for pollution control. Pure Appl Chem, 2001. 73(1): p. 23–30. 165. shimomura, m., et al., Simple and rapid detection method using surface plasmon resonance for dioxins, polychlorinated biphenylx and atrazine. Anal Chim Acta, 2001. 434(2): p. 223–30. 166. delcarlo, m., et al., Disposable screen-printed electrodes for the immunochemical detection of polychlorinated biphenyls. Anal Chim Acta, 1997. 342(2–3): p. 189–97. 167. layton, a.c., et al., Construction of a bioluminescent reporter strain to detect polychlorinated biphenyls. Appl Environ Microbiol, 1998. 64(12): p. 5023–6. 168. stanker, l.h., et al., Monoclonal-antibodies for dioxin–antibody characterization and assay development. Toxicology, 1987. 45(3): p. 229–43. 169. lee, n., c.k. holtzapple, and l.h. stanker, Cloning, expression, and characterization of recombinant Fab antibodies against dioxin. J Agric Food Chem, 1998. 46(8): p. 3381–8. 170. reshetilov, a.n., et al., The amperometric biosensor for detection of sodium dodecyl sulfate. Anal Chim Acta, 1997. 347(1–2): p. 19–26. 171. taranova, l., et al., Bacteria-degraders as the base of an amperometric biosensor for detection of anionic surfactants. Biosens Bioelectron, 2002. 17(8): p. 635–40. 172. sole, m., et al., Estrogenicity determination in sewage treatment plants and surface waters from the Catalonian area (NE Spain). Environ Sci Technol, 2000. 34(24): p. 5076–83. 173. rose, a., et al., GDH biosensor based off-line capillary immunoassay for alkylphenols and their ethoxylates. Biosens Bioelectron, 2002. 17(11–12): p. 1033–43. 174. tsai, y.c. and c.c. chiu, Amperometric biosensors based on multiwalled carbon nanotube–Nafion–tyrosinase nanobiocomposites for the determination of phenolic compounds. Sensors Actuators B: Chem, 2007. doi:10.1016/ j.snb.2007.01.032. 175. grote, k., et al., Effects of in utero and lactational exposure to triphenyltin chloride on pregnancy outcome and postnatal development in rat offspring. Toxicology, 2007. 238(2–3): p. 177–85. 176. thouand, g., et al., Development of a biosensor for on-line detection of tributyltin with a recombinant bioluminescent Escherichia coli strain. Appl Microbiol Biotechnol, 2003. 62(2–3): p. 218–25.

9 Exposure to endocrine-disrupting chemicals in food B. Thomson, Institute of Environmental Science & Research Ltd, New Zealand

Abstract: Food is a major route of exposure to endocrine-disrupting chemicals due to their natural occurrence in some food plants, past and present agricultural practices, industrial applications and atmospheric deposition into the food chain. This chapter presents information on human exposure to phytoestrogens, zearalenone, pyrethroid and organochlorine pesticides, alkyl phenols, bisphenol A, phthalates, polybrominated biphenyls and polycyclic aromatic hydrocarbons from food. Key words: dietary exposure, endocrine-disrupting chemicals, phytoestrogens, pesticides, industrial chemicals, environmental contaminants.

9.1

Introduction

Food is a major route of exposure to endocrine-disrupting chemicals (EDCs) due to their natural occurrence in some food plants, past and present agricultural practices, industrial applications and atmospheric deposition into the food chain. EDCs in food comprise a diverse range of chemicals (Fig. 9.1) in a wide variety of foods and humans are thus exposed to a cocktail of EDCs from the diet. Because different EDCs have different potencies, some EDCs will be of more relevance to human health than others. Assessing exposure to natural and anthropogenic EDCs from food is key to evaluating human risk and thence guiding priorities for risk management and future research. External exposure is a combination of concentration of the chemical in food and consumption of that food and will therefore vary with individual and cultural preferences. Since timing of exposure to EDCs may well be important to assessing potential effects, it is useful to assess exposure for different population groups including males, females, adolescents and the very young.

212

Endocrine-disrupting chemicals in food OH O

OH HO

OH

OH OH

O

O

O

HO

OH O

Coumestrol

Quercetin

Genistein

HO OH OH

OH

OH O

O

OH O

Cl

CH3

Cl C

Cl

O O

HO OH Enterodiol

Cl

Cl

Zearalenone

p,p′-DDT OH

O

OH O O

Br

Br

Br Br

Br Br

Nonylphenol

Br O

CH3 H3C

O Dibutyl phthalate

Br

HO

H3C

Br

Bisphenol A

Cl

O

Cl

Cl

O

Cl

Cl Cl Cl

Cl Cl

Cl

Br

PBDE #209

Tetrachlorodibenzo-p-dioxin

PCB #153

CH3 OH

H3C CH3 O Cl Benzo(a)pyrene

Fig. 9.1

O

O CN HO

Fenvalerate

17 β-Estradiol

Structural diversity of selected EDCs found in food, including, for comparison, the endogenous hormone 17β-estradiol.

Exposure to endocrine-disrupting chemicals in food

213

Exposure assessments are of limited usefulness for ranking priorities because they do not include consideration of potency. Since EDCs can potentially disrupt the endocrine system via several mechanisms it is challenging indeed to rank the relative significance of different EDCs to inform and guide risk mitigation strategies. However, there is some acceptability for assessing total exposure of those EDCs that may act via the classical pathway of genomic activation or suppression of the estrogen receptor (ER), to a common effect, in terms of estrogen equivalents. For this pathway, exposure to many of the EDCs is standardised relative to the estrogenic response of the female hormone 17-β estradiol (Fig. 9.1). In this chapter exposure assessments for individual EDCs in food are collated from the scientific literature and combined estrogenicity is derived for both a Western and an Asian diet.

9.2

Selection of endocrine-disrupting chemicals

Chemicals for inclusion as potential EDCs were identified from the scientific literature. Chemicals with the potential to interfere with the endocrine system based on in vitro and/or in vivo evidence and that have been measured in human foods include the phytoestrogens, a mycotoxin, industrial chemicals and environmental contaminants. EDCs for which dietary exposure data are available from the scientific literature are shown in Table 9.1. This list is not exhaustive and will undoubtedly grow as more chemicals are

Table 9.1 EDCs known to occur in food, or produced from food constituents Phytoestrogens

Mycotoxin Pesticides

Industrial chemicals

Environmental contaminants

Coumestrol Flavonoids: chalcones (phloretin) flavanones (naringenin) flavones (luteolin, apigenin) flavonols (quercetin, kaempferol) Isoflavones: genistein, daidzein Lignans: enterolactone, enterodiol Zearalenone Dieldrin DDT/DDE Endosulfan Synthetic pyrethroids Alkyl phenols Bisphenol A Phthalates PBDEs Dioxins PAH PCBs and OH-PCBs

214

Endocrine-disrupting chemicals in food

identified with potential endocrine-disrupting properties and as more human exposure data is published.

9.3

Exposure assessment methodologies

Exposure assessment is a measure of the amount of hazard to which a person or organism is exposed, or the likelihood of being exposed, and is a means of assessing risk, where risk = hazard × exposure. Exposure assessment may be a measure of external intake from air, food and water, or internal dose assessed from blood, serum or urine levels, the latter accounting for absorption, metabolism and excretion. External exposure allows identification of the food source and an understanding of the impact of different diets on exposure. For a foodborne hazard, exposure requires knowledge of the levels of the chemical in foods and amounts of the relevant foods consumed. Various study designs were adopted to measure the concentrations of EDCs used for the exposure assessments cited in this chapter, namely: • total diet studies, representative of the majority of foods consumed by the group of interest; • targeted studies of selected foods only; • duplicate diet studies where duplicate meals were collected and analysed. Food consumption information was determined by either one of a variety of approaches: • Per capita consumption from food balance sheets that estimate average food consumption from national data of food production, less exports, plus imports, minus wastage. Food balance sheets are thought to overestimate consumption of most commodities and therefore represent high percentile consumers and a worst case scenario. • Food diary records, where respondents recorded all food consumed over a specified period. • 24-hour diet recalls, where a sample of respondents recalled all food consumed in a defined 24-hour period. • Food frequency questionnaires where respondents recalled how often a food was consumed. Ideally, concentration and consumption data are derived for the population group of interest. However, for EDCs this information is often not available and in some cases, international concentration data were combined with consumption data for consumers from a different country to give an estimate of exposure (e.g. Thomson et al., 2003; Johannot and Somerset, 2005; Rossi et al., 2006).

Exposure to endocrine-disrupting chemicals in food

215

Exposure estimates for the selected EDCs in food were retrieved from the published scientific literature, or rarely, from a personal communication. For consistency across EDCs, mean or median exposures have been presented. Where exposure data were reported as mg or μg/day, they were normalised to body weight. Where body weight of the target population was not given, the following assumptions were applied for a Western adult = 70 kg, a Western adult male = 75 kg, a Western adult female = 65 kg, an Asian adult = 65 kg, an Asian adult male (70 kg) and an Asian adult female (60 kg). Although these weights are higher than 60 kg applied by the WHO (1997), I consider them to be closer to actual weights where these were reported in Tables 9.2 to 9.15 and therefore result in an exposure assessment closer to reality.

9.4

Exposure to total estrogenicity

9.4.1 Exposure to estrogenic endocrine-disrupting chemicals The upper mean exposure for each potentially estrogenic EDC from Tables 9.2 to 9.15 was selected as the basis for assessing relative contributions to total estrogenicity. Where data were available, values representative of both a ‘Western’ and an ‘Asian’ diet were selected, because of the significant differences in exposure to phytoestrogens between the two diets. Where there were no data for an ‘Asian’ diet, the ‘Western’ diet value was applied. Intake data from the traditional diet of Greenland that incorporates seal and whale blubber were considered atypical of general diets and therefore not included in the selection of mean exposures for total estrogenicity. These selections are biased to a worst case scenario.

9.4.2 Relative estrogenic potency Many of the EDCs including genistein, daidzein, enterolactone, enterodiol, coumestrol, quercetin, kaempferol, luteolin, apigenin, naringenin, phloretin, zearalenone, dieldrin, dichlorophenyltrichloroethane (DDT), endosulfan, synthetic pyrethroids, alkyl phenols, bisphenol A, polycyclic aromatic hydrocarbons (PAH), polychlorinated biphenyls (PCBs) and phthalates have demonstrated estrogenic activity. Estrogenic potency may be measured by a range of whole animal, in vivo, and in vitro, assays (Fig. 9.2, see also chapter 10 Fig. 10.2). In the absence of sufficient data on in vivo effects, relative estrogenicity based on in vitro cell proliferation assays is considered the best option as a measure of dose –response for two reasons. Firstly, this assay represents a higher level of biological complexity than either competitive binding or gene expression assays. Secondly, cell proliferation assays based on the MCF-7 breast cancer cell line are one of the most widely employed assays of estrogenicity, providing a common basis for diverse environmental estrogens (Soto et al., 1995; Breinholt and Larsen, 1998).

216

Endocrine-disrupting chemicals in food Cell membrane Nucleus Receptorbinding assay

Transcription

Receptor-dependent gene expression assay

Cell proliferaion assay

Protein synthesis

Cell division

Fig. 9.2 A simplified schematic of the classical mechanism of an estrogen-responsive effect showing the locations of commonly used in vitro assays (from Thomson et al., 2003, and kindly drawn by Matt Walters).

For internal consistency, results from Breinholt and Larsen (1998) were used as a basis for relative potency in this chapter. Where these data were not available, and where different studies derived different estimates of estrogenic potency for the same chemical, the highest estimate (worst case) was used for the assessment of exposure to estrogenicity. In the absence of cell proliferation data, estrogenic potency was based on the structure– activity relationship (e.g. isoliquiritigenin compared with apigenin) or on gene expression results (e.g. phthalates and benzo[a]pyrene (B[a]P)). These

Exposure to endocrine-disrupting chemicals in food

217

worst case assumptions follow the ‘precautionary principle’ (Burger, 2003), in recognition of the uncertainty of the potential endocrine effect. Given the public aversion to added chemicals in food and recognising that there are undoubtedly estrogenic compounds that we do not yet know about, there are merits in a precautionary approach. Results from different assay methods cannot be compared directly. However, it is arguable that relative estrogenicity (EQ), standardised to 17β-estradiol, can be compared for estrogenic compounds acting via the same receptor mechanism, in a similar way to the toxicity assessments for dioxins and dibenzofurans (van den Berg et al., 1998). 9.4.3 Cumulative exposure to estrogenicity Assuming additivity of effect (Gaido et al., 1997; Payne et al., 2000; Silva et al., 2002; Rajapakse et al., 2002) exposure to total estrogenicity in the diet may be estimated by summing the EQ for individual estrogenic EDC using the formula: EQ = ∑ ( EC i × EPi ). where ECi is exposure to chemical i, and EPi is its estrogenic potency relative to a standard (17β-estradiol) (Safe, 1995; Thomson et al., 2003; Huang et al., 2007). Additivity is undoubtedly a simplification because it does not take into account multiple mechanisms, interactions between different compounds of differences of effects in different target cells. But it is a pragmatic basis for estimating risk and highlighting priorities for action. 9.4.4 Blood levels of estrogenic endocrine-disrupting chemicals Any potential pharmacological effect of an EDC depends not on dietary exposure, but rather on the circulating concentration in the blood. An estimate of theoretical serum levels of EDCs corresponding to estimated dietary exposures was made by assuming average serum volumes of 3.3, 2.5 and 2.9 litres for males, females, and males and females combined, 100% absorption and that all estrogenicity was associated with the serum component of blood as applied previously (Shaw and McCully, 2002; Thomson et al., 2003). A blood level, accounting for absorption, distribution, metabolism and excretion, was extrapolated from the external dietary exposure estimate by adjusting for the ratio of actual (A) serum (or plasma) level to theoretical (T) serum level from published data. Information on actual serum levels used, in order of priority, was: • studies relating dietary intake of the environmental estrogen to actual steady-state levels of the same chemical in serum for the same study group; • studies reporting serum levels of environmental estrogens for the general population from the same country as the intake assessment;

218

Endocrine-disrupting chemicals in food

• studies reporting serum levels of environmental estrogens from a region of similar diet to the intake data. Hence an A : T ratio was determined. Where no appropriate serum data exist, an A : T ratio of 1 was assigned. This factor assumes the same ratio of actual to theoretical concentrations between individuals and population groups and is undoubtedly a simplification. Nevertheless, this conversion to an internal dose is considered an improvement over previous exposure estimates.

9.5 Exposure assessments for endocrine-disrupting chemicals 9.5.1

Phytoestrogens

Coumestans Coumestans are phytoestrogens that are structurally similar to isoflavones. One coumestan, coumestrol (Fig. 9.1) is estrogenic and therefore a potential EDC. Coumestrol has been reported in alfalfa sprouts, chickpeas, split peas, lima beans, pinto beans and soybean sprouts. Exposure estimates ranging from <0.01 to 0.5 μg/kg bw/day have been reported for three Western diets (Table 9.2). For Asian diets less information is available. A single assessment of various Korean population groups reported a dietary exposure of 7 μg/kg body weight (bw)/day for both adults and toddlers, with the greatest contributor to this intake being soybean sprouts. Flavonoids Flavonoids are a broad group of compounds including chalcones, flavones, flavonols and flavanones (Fig. 9.3). Estrogenicity, hence endocrine-disrupting

Table 9.2

Dietary exposure estimates for coumestrol

Country

Population group

Korea

Adults, 30–49 yrs, 56 kg Toddler, 1–2 yrs, 14.5 kg Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Adult women, 50–69 yrs, 70 kg Adult women, postmenopausal, 65 kg1

New Zealand

The Netherlands USA 1

Mean exposure (μg/kg bw/day)

Body weight assumed, intake reported as mg/day.

7.1 6.9 0.2 0.4 0.5 0.4 <0.01 0.2

Reference Surh et al. (2006) Thomson et al. (2003), Thomson (2005) Keinan Boker et al. (2002) De Kleijn et al. (2001)

O

R'''

R'

O

Flavones Apigenin Luteolin

R

O

R'' R'

O

Isoflavones Genistein Daidzein Glycetin Biochanin A Formononetin

R

O

R''''

R'

O

OH

Flavonols Quercetin Kaempferol Myricetin

R

O R'''

R'' R'

R

Flavanones Naringenin Hesperetin Eriodictyol

O

O

R''', R''''

R''

General structures of different flavonoid groups with sites of hydroxylation (R, RI, RII, RIII) or methoxylation (RIIII) and major dietary members of each group.

Chalcones Isoliquiritigenin Phloretin

R

Fig. 9.3

R'

R''

220

Endocrine-disrupting chemicals in food

potential, has been reported for apigenin, isoliquiritigenin, kaempferol, luteolin, naringenin, phloretin and quercetin based on receptor binding, gene expression and cell proliferation assays. The potential cancer protective properties of flavonoids and publication of dietary databases (e.g. USDA, 2002, 2003; Kyle and Duthie, 2006; Ritchie et al., 2006) has spurred recent epidemiological studies of the association between dietary intake of flavonoids and health outcomes. Where possible, exposure to individual chemicals is presented but in a number of studies, individual flavonoids have been grouped into sub-classes, particularly flavones, flavonols and flavanones, and the grouped exposures are shown (Table 9.3). The major dietary sources of these flavonoids are: apigenin from celery, kaempferol from tea and broccoli, luteolin from celery and red pepper, naringenin from grapefruit, oranges and fruit juice, quercetin from tea, apples and onions.

Table 9.3 Country

Dietary exposure estimates for flavonoids Population group

Mean exposure (μg/kg bw/day)

Reference

Chalcones-phloretin New Adult male, 80 kg Zealand Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg

14–461 14–49 23–97 15–72

Thomson et al. (2003), Thomson (2005)

Isoliquiritigenin New Adult male, 80 kg Zealand Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg

29 35 33 35

Thomson et al. (2003), Thomson (2005)

31 14 34 5473 36 28 41 28 141

Johannot and Somerset (2005)

Flavanones-naringenin Australia Adult, 70 kg2 Young adult, 19–24 yrs, 65 kg2 Child, 8–11 yrs, 40 kg2 Italy Adult, 19–74 yrs, 70 kg2 New Adult male, 80 kg Zealand Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Scotland, Adult, 16–79 yrs, 70 kg2 UK USA Adult men, 75 kg2 Adult women 65 kg2 Flavones-luteolin Australia Adult, 70 kg2 Young adult, 19–24 yrs, 65 kg2 Child, 8–11 yrs, 40 kg2

1873 2023 1.1 0.8 1

Rossi et al. (2006) Thomson et al. (2003), Thomson (2005) Theodoratou et al. (2007) Chun et al. (2007)

Johannot and Somerset (2005)

Exposure to endocrine-disrupting chemicals in food

221

Table 9.3 Continued Mean exposure (μg/kg bw/day)

Country

Population group

Italy New Zealand

Adult, 19–74 yrs, 70 kg2 Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Adult, 16–79 yrs, 70 kg2

7.14 1.3 2.1 1.4 1.5 164

Adult men, 75 kg2 Adult women 65 kg2

184 264

Scotland USA

Apigenin Australia New Zealand

Adult, 70 kg2 Young adult, 19–24 yrs, 65 kg2 Child, 8–11 yrs, 40 kg2 Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg

Flavonols-quercetin Australia Adult, 70 kg2 Young adult, 19–24 yrs, 65 kg2 Child, 8–11 yrs, 40 kg2 Italy Adult, 19–74 yrs, 70 kg2 New Adult male, 80 kg Zealand Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg USA Adult men, 75 kg2 Adult women 65 kg2 Kaempferol Australia Adult, 70 kg2 Young adult, 19–24 yrs, 65 kg2 Child, 8–11 yrs, 40 kg2 New Adult male, 80 kg Zealand Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Scotland, Adult, 16–79 yrs, 70 kg2 UK 1 2 3 4 5 6

0.6 1.8 3 3.9 8.3 4.4 4.8

Reference Rossi et al. (2006) Thomson et al. (2003), Thomson (2005) Theodoratou et al. (2007) Chun et al. (2007)

Johannot and Somerset (2005) Thomson et al. (2003), Thomson (2005)

179 99 91 3085 109 149 97 145 1605 1865

Johannot and Somerset (2005)

80 37 12 44 77 26 46 1396

Johannot and Somerset (2005)

Rossi et al. (2006) Thomson et al. (2003), Thomson (2005) Chun et al. (2007)

Thomson et al. (2003), Thomson (2005) Theodoratou et al. (2007)

Exposure based on minimum and maximum concentration values respectively. Body weight assumed, intake reported as mg/day. Flavanones, includes sum of eriodictyol, hesperetin, and naringenin. Flavones, includes sum of luteolin and apigenin. Flavonols, includes sum of quercetin, kaempferol, myricetin, isorhamnetin. Calculated from flavonols minus contribution from quercetin.

222

Endocrine-disrupting chemicals in food

Isoliquiritigenin occurs in liquorice and soybeans but concentration data are limited to 9600 mg/kg in liquorice-root extract powder (Weinberg et al.,1993). Phloretin occurs solely in apples (Burda et al., 1990; Andrade et al., 1998). Dietary exposure to the flavonoids varies widely from 0.6 μg/ kg bw/day of apigenin to 547 μg/kg bw/day of flavanones for an Italian adult, the latter from consumption of citrus fruits. There is a marked lack of data from Asian populations. Isoflavones Included in this group are the compounds genistein, daidzein, biochanin A and formononetin (COT, 2003) (Fig. 9.3) with exposure from commonly consumed foods dominated by the greater prevalence of genistein and daidzein. Although found in many plant tissues, highest levels are observed in legumes, particularly soybeans, with more variable levels in oilseeds and nuts. Isoflavones are often present as glucoside conjugates in plants and foods (e.g. genistin and daidzin) and since these are readily converted to the aglycone by gut microflora, intake of both the glycoside and aglycone must be considered for a realistic exposure assessment. Exposure to genistein and daidzein is highly variable, spanning three orders of magnitude globally (Table 9.4). Exposure to genistein from an Asian diet is in the order of 400–500 μg/kg bw/day and for daidzein is approximately 250 μg/kg bw/ day. This is contrasted with lower exposure from a Western diet. Based on current data it is difficult to assign an average level of exposure for either genistein or daidzein for a Western diet. Estimates of genistein for an average Western consumer in the UK range from 3 to 32 μg/kg bw/day and the US range from 5 to 23 μg/kg bw/day. The higher estimates are also substantiated by the estimates from New Zealand. A Western vegetarian or ‘soya eater’ has a higher exposure than the average omnivorous consumer of 70–100 μg/kg bw/day of genistein and 60–70 μg/kg bw/day of daidzein (Kirk et al., 1999; Clarke et al., 2003; Thomson, 2005). A number of studies report the greatest contributors to dietary intake of genistein and daidzein in a Western diet are from the use of soya flour, soya milk and soya protein in processed foods (van Erp-Baart et al., 2003; Clarke and Lloyd, 2004; Mulligan et al., 2007). At least part of the variability seen in the estimates of Western diets is a reflection of the limitations of the databases used to derive intakes where concentration data of processed foods is, or may be, outdated and/or lacking. Lignans Lignans are components of lignin that give strength and stiffness to plant cell walls and protect the plant from microbial attack. Lignans are a class of phytoestrogens that exist as minor constituents in many cereals, vegetables and fruit with linseed (flaxseed) the richest known source. In humans, gut microflora convert plant lignans to enterolactone and enterodiol

Table 9.4 Dietary exposure estimates for the isoflavones genistein and daizein Country

Population group

Genistein China

Adult men, 40–74 yrs, 70 kg1 Finland Adult men, 75 kg1 Adult women, 65 kg1 Ireland Adult men, 75 kg1 Adult women, 65 kg1 Italy Adult men, 75 kg1 Adult women, 65 kg1 Japan Adult women, 60 kg1 Adult women, 60 kg1 Adult men, 45–74 yrs, 70 kg1 Korea Adults, 30–49 yrs, 56 kg Toddler, 1–2 yrs, 14.5 kg New Zealand Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg The Adult men, 75 kg1 Netherlands Adult women, 65 kg1 Adult women, 50–69 yrs,70 kg UK Adult men, 75 kg1 Adult women, 65 kg1 Adult, vegetarian, 69.2 kg Adult, 70 kg1

USA

Daidzein China Finland

Adult women, 65 kg1 Adult men, 81 kg – nonsoya eaters Adult men, 78 kg – soya eaters Adult women, 69 kg – non-soya eaters Adult women, 65 kg – soya eaters Adult, soya consumers,70 kg1 Adult women, postmenopausal, 65 kg1 Women, Caucasian, 65 kg1 Women, African American, 65 kg1 Women, Japanese, 56.5 kg Women, Chinese, 58.4 kg Adult men, 40–74 yrs, 70 kg1 Adult men, 75 kg1 Adult women, 65 kg1

Mean exposure Reference (μg/kg bw/day) 287

Lee et al. (2007)

7.4 6.2 5.8 4.7 4.6 4.1 390 508 3132

Valsta et al. (2003)

148 359 24 18 24 45–1033 7.4 7.4 2.3 5.6 5.5 115 32 3.7 4.3

van Erp-Baart et al. (2003) van Erp-Baart et al. (2003) Uehara et al. (2000) Arai et al. (2000) Kurahashi et al. (2007) Surh et al. (2006) Thomson et al. (2003)

van Erp-Baart et al. (2003) Keinan Boker et al. (2002) van Erp-Baart et al. (2003) Clarke et al. (2003) Clarke and Lloyd (2004) Grace et al. (2004) Mulligan et al. (2007)

51 3.9 67 100 5.2 15

Kirk et al. (1999) De Kleijn et al. (2001) Greendale et al. (2002)

4 191 100 212 4.6 4.1

Lee et al. (2007) Valsta et al. (2003)

Table 9.4 Continued Country

Population group

Ireland

Adult men, 75 kg1

Adult women, 65 kg1 Adult men, 75 kg1 Adult women, 65 kg1 Adult, 19–74 yrs, 70 kg1 Japan Adult women, 60 kg1 Adult women, 60 kg1 Adult men, 45–74 yrs, 70 kg1 Korea Adults, 30–49 yrs, 56 kg Toddler, 1–2 yrs, 14.5 kg New Zealand Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg The Adult men, 75 kg1 Netherlands Adult women, 65 kg1 Scotland Adult, 16–79 yrs, 70 kg1 Italy

UK

USA

1 2 3 4

Adult women, 50–69 yrs, 70 kg Adult men, 75 kg1 Adult women, 65 kg1 Adult, vegetarian, 69.2 kg Adult, 70 kg1 Adult women, 65 kg1 Adult men, 81 kg – nonsoya eaters Adult men, 78 kg – soya eaters Adult women, 69 kg – non-soya eaters Adult women, 65 kg – soya eaters Adult, 70 kg1, soya consumers Adult women, postmenopausal, 65 kg1 Women, Caucasian, 65 kg1 Women, African American, 65 kg1 Women, Japanese, 56.5 kg Women, Chinese, 58.4 kg Adult men, 75 kg1 Adult women 65 kg1

Mean exposure Reference (μg/kg bw/day) 5.6 4.8 3.8 3.4 0.44 243 278 1972 395 552 15 13 16 70 6.1 5.4 8.34 2.1 4.6 4.4 60 12 3.0 3.6

van Erp-Baart et al. (2003) van Erp-Baart et al. (2003) Rossi et al. (2006) Uehara et al. (2000) Arai et al. (2000) Kurahashi et al. (2007) Surh et al. (2006) Thomson et al. (2003), Thomson (2005) van Erp-Baart et al. (2003) Theodoratou et al. (2007) Keinan Boker et al. (2002) van Erp-Baart et al. (2003) Clarke et al. (2003) Clarke and Lloyd (2004) Grace et al. (2004) Mulligan et al. (2007)

39 3.2 53 57 5.2 8

Kirk et al. (1999) De Kleijn et al. (2001) Greendale et al. (2002)

2 133 52 154 204

Body weight assumed, intake reported as mg/day. Mid-point of 2nd and 3rd quartile intakes. Based on maximum and minimum concentration values. Total isoflavones.

Chun et al. (2007)

Exposure to endocrine-disrupting chemicals in food CH3

CH3

CH3

O

O

HO

O

O OH

HO

HO OH

OH O OH CH3

O OH CH3

Lariciresinol

225

OH OH O OH CH3

Isolariciresinol

Secoisolariciresinol

CH3 O HO

OH

HO OH OH

O O

O

O OH CH3 Matairesinol

O

OH Enterodiol

OH Enterolactone

Fig. 9.4 Structures and names of the principal lignans identified in food: lariciresinol, isolariciresinol, secoisolariciresinol and matairesinol and the metabolites enterodiol and enterolactone.

(Adlercreutz and Mazur, 1997) and since balance studies in humans have indicated that most of the lignan precursors remain unidentified, levels of enterolactone and enterodiol probably give a more realistic measure of exposure to lignans in food than estimates based on individual plant lignans. Structures of the principal lignans identified in food: lariciresinol, isolariciresinol, matairesinol and secoisolariciresinol and the metabolites enterodiol and enterolactone are shown in Fig. 9.4. Limited exposure estimates to lignans in food are shown in Table 9.5 with adult exposures to enterolactone ranging from 0.4 to 3.7 μg/kg bw/day and enterodiol ranging from a low of 0.1 in the UK to 9.6 in the US. For a New Zealand diet the major contributing foods were carrots, bread, potato and beans (Thomson et al., 2003). For a group of healthy postmenopausal Caucasian women in the US, the main source of lignans was fruits (de Kleijn et al., 2001).

9.5.2 Mycotoxins The only mycotoxin associated with EDC effects is zearalenone. Zearalenone (Fig. 9.1) is a secondary fungal metabolite, produced by several Fusarium species that proliferate in poorly stored grains where the grain is stored

226

Endocrine-disrupting chemicals in food

Table 9.5 enterdiol

Dietary exposure estimates for the lignans enterolactone and

Country Enterolactone Finland New Zealand

The Netherlands UK USA Enterodiol Finland New Zealand

The Netherlands UK USA

1 2 3 4 5

Population group

Adult men, 75 kg1 Adult women, 65 kg1 Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Adult women, 50–69 yrs, 70 kg Adult women, 65 kg1 Adult women, postmenopausal, 65 kg1 Adult men, 75 kg1 Adult women, 65 kg1 Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Adult women, 50–69 yrs, 70 kg Adult women, 65 kg1 Adult women, postmenopausal, 65 kg1 Adult women, 44 yrs, 65 kg1 Adult women, 63 yrs, 65 kg1

Mean exposure (μg/kg bw/day) 0.6 0.5 3.5 2.9 3.7 6.9 1.02 1.2 0.42

3.3 8.83,4 1.9 1.5 2.3 3.5 13.23 0.1 9.63 7.85

Reference

Valsta et al. (2003) Thomson et al. (2003), Thomson (2005) Keinan Boker et al. (2002) Grace et al. (2004) De Kleijn et al. (2001) Valsta et al. (2003) Thomson et al. (2003), Thomson (2005) Keinan Boker et al. (2002) Grace et al. (2004) De Kleijn et al. (2001) McCann et al. (2004)

8.85

Bodyweight assumed, intake reported as mg/day. Expressed as the precursor, matairesinol. Expressed as precursor, secoisolariciresinol. Exposure highly variable (standard deviation = 40 × mean). Total lignan precursors.

moist. Zearalenone occurs in both temperate and warm regions of the world. Exposure estimates range from 0.007 μg/kg bw/day in Norway, based on individual dietary records to 0.06 μg/kg bw/day for a Middle Eastern regional diet (Table 9.6). Major contributing food groups are maize, corn, rice and wheat products.

Exposure to endocrine-disrupting chemicals in food

227

Table 9.6 Dietary exposure estimates for zearalenone Country Zearalenone, t-TDI = African diet European diet Far Eastern diet Latin American diet Middle Eastern diet Austria Canada Denmark Finland France

France Italy The Netherlands New Zealand1

Norway Norway Portugal UK

USA

Population group 0.2 or 0.5 μg/kg bw/day Adult, 60 kg Adult, 60 kg Adult, 60 kg Adult, 60 kg Adult, 60 kg Adult, 75 kg Adult, 60 kg Adult, 60 kg (assumed) Adult, 77.1 kg Adult, 66.4 kg Male, 73.9 kg Female, 60.1 kg Child, 3–15 yrs, 31.6 kg Adult, 15+ years Child, 3–14 yrs Adults, 70 kg Adults, 65.8 kg Child, 1–4 yrs, 13.8 kg Child, 1–6 yrs, 17.1 kg Adult male, 82 kg Adult female, 70 kg Young male, 19–24 yrs, 78 kg Vegetarian female, 19– 40 yrs, 70 kg Child, 23 kg Male, 81 kg Female, 66 kg Infant, 6 months, 8 kg Adult, 60 kg (assumed) Total population, 65 kg Adult male, 70.1 kg Adult female, 70.1 kg Infant, 6–12 mths, 8.7 kg Child, 4–6 yrs, 20.5 kg Adult, 60 kg (assumed)

Mean exposure (μg/kg bw/day) 0.041 0.025 0.056 0.036 0.059 0.028 <0.02 0.02 0.027 0.027 0.029 0.024 0.042 0.033 0.066 0.0008 0.021 0.046 0.050 0.012 0.011 0.015

Reference

JECFA (2000) JECFA (2000) JECFA (2000) JECFA (2000) JECFA (2000) SCOOP (2003) JECFA (2000) EC (2000) SCOOP (2003) SCOOP (2003)

Leblanc et al. (2005) SCOOP (2003) SCOOP (2003) Thomson et al. (2003), Thomson (2005)

0.032 0.04 0.008 0.007 0.012–1.51 0.02 0.004 0.014 0.013 0.050 0.055 0.03

SCOOP (2003) EC (2000) SCOOP (2003) SCOOP (2003)

EC (2000)

t-TDI = temporary tolerable daily intake. RD = Reference Dose. 1 Based on maize-based foods only.

9.5.3 Pesticides Organochlorine pesticides (dichlorodiphenyldichloroethane/ dichlorodiphenyldichloroethene, dieldrin, endosulfan) Since the publication of Silent Spring by Rachel Carson in 1962 much attention has been given globally to the levels of a group of persistent

228

Endocrine-disrupting chemicals in food Cl

Cl Cl

C

Cl

Cl

Cl

Cl

C

Cl

Cl

Cl o,p′-DDT

p,p′-DDT Cl

C

Cl

Cl

Cl

p,p′ -TDE

Dieldrin

Cl

Cl Cl

Cl

Cl

Cl Cl

Cl

Cl

p,p′-DDE

O Cl

C

Cl

Cl Cl

O Cl Cl

S

O

O Cl Endosulfan

Fig. 9.5 Structures and names of the organochlorine pesticides p,p′ and o,p′-DDT, dieldrin and endosulfan and the DDT metabolites p,p′-DDE p,p′-TDE.

organochlorine pesticides in the environment. At least three of these are potential EDCs, namely DDT and its metabolite dichlorodiphenyldichloroethene (DDE), dieldrin and endosulfan (Fig. 9.5). Recent estimates of dietary exposure to DDT and its metabolite DDE, dieldrin and endosulfan are shown in Table 9.7. For adults, total DDT exposure, including metabolites DDE and/or DDD, ranged from 0.0005 μg/ kg bw/day in Australia to 0.46 μg/kg bw/day from the consumption of traditional foods in Greenland where seal blubber was the dominating source (Johansen et al., 2004). Exposure to organochlorine pesticides has been declining globally with cessation of organochlorine (OC) pesticide use since the 1970s (IPCS, 2002; Vannoort and Thomson, 2005). The lower exposure of vegetarians to DDT compared with non-vegetarians (Battu et al., 2005) reflects the importance of animal products and accumulation of DDT up the food chain. Breast milk has been extensively studied and the most recent estimates for breastfed infants showed infant exposure to DDT/ DDE ranged from 1.2 to 7.0 μg/kg bw/day with the highest exposure levels seen in Vietnam. The reader is referred to the references cited in these papers and to the IPCS global assessment (IPCS, 2002) for further past data. Clearly, breastfed infants are more highly exposed to DDT than adults. Dietary exposure estimates for dieldrin and total endosulfan are lower and less prevalent than for DDT. Synthetic pyrethroids Pyrethroids are a group of synthetic derivatives of pyrethrin that are currently used on field crops, including cereals, root crops and grass, on animals including cattle and sheep, and domestically, for insect control. Included in this class of compounds are allethrin, bifenthrin, cyfluthrin, cyhalothrin,

Exposure to endocrine-disrupting chemicals in food

229

Table 9.7 Current dietary exposure estimates for organochlorine pesticides (DDT/DDE, dieldrin and endosulfan) Country

Population group

DDT/DDE

ADI = 2 μg/kg bw/day PTDI = 10 μg/kg bw/day TDI = 20 μg/kg bw/day Adult males, 25–34 yrs, 82 kg Adult females, 25–34 yrs, 66 kg Toddler, 2 yrs, 14 kg All ages, 1–65+ Child, 1–4 yrs, 14.4 kg Adults

Australia

Canada Greenland

India

Indonesia New Zealand

Adults, 18–60+, 70 kg1 , spring Adults, 18–60+, 70 kg1, fall Adults, 60 kg – vegetarian, 2001 Adults, 60 kg – nonvegetarian, 2001 Adults, 60 kg – vegetarian, 2002 Adults, 60 kg – nonvegetarian, 2002 Infant – breastfed, 5 kg

Poland

Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Infant – breastfed

Sweden

Adult, 17–79 yrs, 73.7 kg

Taiwan

Infant – breastfed, 5 kg

Thailand

Adult, 50 kg

USA

Adult males, 25–30 yrs

Vietnam

Infant–breastfed, 5 kg

Dieldrin

TDI = 0.1 μg/kg bw/day PTDI = 0.1μg/kg bw/day

Mean exposure (μg/kg bw/day)

DDT total: 0.0006

Reference FSANZ (2003) IPCS (2006) WHO (1984) FSANZ (2003)

DDT total: 0.0005 DDT total: 0.0010 DDT total: 0.006 DDT total: 0.017 DDE: 0.12 DDT: 0.0185 DDT total: 0.46 DDT total: 0.36 DDT total: 0.04

Rawn et al. (2004) Deutch et al. (2006) Johansen et al. (2004) Battu et al. (2005)

DDT total: 0.23 DDT total: 0.14 DDT total: 0.45 DDE: 1.1–4.7 DDT: 0.1–0.4 DDT total: 0.0216 DDT total: 0.0170 DDT total: 0.0511 DDE: 3.60 DDT: 0.28 DDE: 0.003 DDT: 0.004 DDE: 1.32 DDT: 0.1 DDT total: 0.005 DDE: 0.006 DDT: 0.0004 DDE: 6.3 DDT: 0.7

Sudaryanto et al. (2006) Vannoort and Thomson (2005) Syrwin´ska and Lulek (2007) Darnerud et al. (2006) Chao et al. (2006) Vongbuddhapitak et al. (2002) Egan (2005)3 Minh et al. (2004) Johansen et al. (2004) IPCS (2006)

230

Endocrine-disrupting chemicals in food

Table 9.7 Continued Mean exposure (μg/kg bw/day)

Reference

ND

FSANZ (2003)

Thailand

Adult males, 25–34 yrs, 82 kg Adult females, 25–34 yrs, 66 kg Toddler, 2 yrs, 14 kg All ages, 1–65+ Child, 1–4 yrs, 14.4 kg Adults, 18–60+, 70 kg1, spring Adults, 18–60+, 70 kg1, fall Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Adult, 50 kg

USA

Adult males, 25–30 yrs

0.0013

Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) Egan (2005)2

0.0023

FSANZ (2003), IPCS (2006) FSANZ (2003)

Country

Population group

Australia

Canada Greenland New Zealand

Endosulfan, ADI = 6 μg/kg bw/day total Australia Adult males, 25–34 yrs, 82 kg Adult females, 25–34 yrs, 66 kg Toddler, 2 yrs, 14 kg Canada All ages, 1–65+ Child, 1–4 yrs, 14.4 kg India Adults, 60 kg – vegetarian, 2001 Adults, 60 kg – nonvegetarian, 2001 Adults, 60 kg – vegetarian, 2002 Adults, 60 kg – nonvegetarian, 2002 New Adult male, 25+ yrs Zealand Adult female, 25+ yrs Toddler, 1–3 yrs Thailand Adult, 50 kg USA

Adult males, 25–30 yrs

ND ND 0.001 0.003 0.11 0.10 0.0015 0.0012 0.0036 ND

Rawn et al. (2004) Johansen et al. (2004)

0.0026 0.0033 0.017 0.032 ND

Rawn et al. (2004) Battu et al. (2005)

ND ND ND 0.0375 0.0404 0.0391 0.028 0.0131

Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) Egan (2005)3

ADI = Acceptable daily intake PTDI = Provisional tolerable daily intake TDI = Tolerable daily intake ND = not detected in any food analysed. 1 Bodyweight assumed, intake reported as mg/day. 2 Assumes 5 kg body weight and 700 g/day consumption of breast milk from Minh et al. (2004). 3 Egan K, Centre for Foods Safety and Nutrition, Personal communication, 2005.

Exposure to endocrine-disrupting chemicals in food H3C CH3 CH 3 CH2 H3C O H 3C

H O

O

Cl

H3C CH 3 O O

F3C

Allethrin

F 3C

O O CN

O

Cl

O CN

O O CN

O

H3C CH3 O

H 3C

H3C CH 3 O O

Cyfluthrin

O

Br Br

O

Permethrin

Cl O

Sumithrin/phenothrin

F3C

H3C CH NH

O

O CN Deltmethrin

O

O

Fenvalerate

H 3C

O CN

CH3

Cypermethrin

Cl Cl

O

H3C CH3

Cl Cl

Cyhalothrin H3C CH3 O

F

O

Bifenthrin

H3C CH3

Cl

H3C CH3 Cl Cl

231

O H3C CH 3 H3COOC O CH3 H3C Pyrethrin

3

O O CN

CH2

OH O

Taufluvalinate

HO

O

Permethrin metabolite

Fig. 9.6 Chemical structures and names of synthetic pyrethroids that may occur in food through their use for insect control in field crops, animals and for domestic use and an estrogenic metabolite of permthrin.

cypermethrin, deltamethrin, fenvalerate, permethrin, pyrethrin, sumithrin and taufluvalinate (Davies, 1985), most of which are registered for use in New Zealand (NZFSA, 2004). The chemical structures of pyrethroid insecticides are shown in Fig. 9.6 and structural analogies to 17β-estradiol are not obvious. However, in mammals, metabolism is via alcohol intermediates that are more similar in structure to 17β-estradiol. Some of the metabolites are also estrogenic (McCarthy et al., 2005). Limited estimates of dietary exposure to pyrethroid pesticides have been published (Table 9.8). Vongbuddhapitak and co-workers (2002) reported the pyrethroids were found only in vegetables. No pyrethroid pesticides were detected in 46 male vegetarian and non-vegetarian diets in India (Battu et al., 2005). Dietary exposure for five individual pyrethroid compounds ranged from 0.0002 to 0.0476 μg/kg bw/day across four countries. No pyrethroid residues were detected in 155 beef and sheep fat samples tested in Ireland (NFRD, 1999). To date, there are no assessments of exposure that have considered a contribution of effect from metabolites of synthetic pyrethroid compounds.

232

Endocrine-disrupting chemicals in food

Table 9.8

Dietary exposure estimates for synthetic pyrethroid pesticides Mean exposure (μg/kg bw/day)

Country

Population group

Bifenthrin

ADI = 20 μg/kg bw/day = 10 μg/kg bw/day Adult males, 25–34 yrs, 82 kg Adult females, 25–34 yrs, 66 kg Toddler, 2 yrs, 14 kg Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs

0.0058 0.0004 0.0004 0.0008

Thailand

ADI = 0–5 μg/kg bw/day Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Adult, 50 kg, (1996)

0.0002 0.0002 0.0028 0.004

USA

Adult males, 25–30 yrs

0.0004

Cypermethrin New Zealand

ADI = 50 μg/kg bw/day Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Adult, 50 kg, (1996)

0.0017 0.0014 0.0183 0.005

Thailand

ADI = 20 μg/kg bw/day Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Adult, 50 kg, 1996

0.0004 0.0003 0.0034 ND

Permethrin

ADI = 0–50 μg/kg bw/day

Australia

0.0146

Thailand

Adult males, 25–34 yrs, 82 kg Adult females, 25–34 yrs, 66 kg Toddler, 2 yrs, 14 kg Adult male, 25+ yrs Adult female, 25+ yrs Toddler, 1–3 yrs Adult, 50 kg, (1996)

USA

Adult males, 25–30 yrs

0.0476

Australia

New Zealand

Cyhalothrin New Zealand

Thailand Fenvalerate New Zealand

New Zealand

0.0030

Reference IPCS (2006) FSANZ (2003) FSANZ (2003)

0.0032 Vannoort and Thomson (2005) IPCS (2006) Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) Egan (2005)1 IPCS (2006) Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) IPCS (2006) Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) IPCS (2006), FSANZ (2003) FSANZ (2003)

0.0149 0.0205 0.0057 0.0058 0.0075 0.004

Vannoort and Thomson (2005) Vongbuddhapitak et al. (2002) Egan (2005)1

ADI = acceptable daily intake. 1 Egan K, Centre for Foods Safety and Nutrition, US FDA, personal communication, 2005.

Exposure to endocrine-disrupting chemicals in food

233

9.5.4 Industrial chemicals Alkyl phenols Alkyl phenols are a group of degradation products derived from alkylphenol ethoxylates, which are non-ionic surfactants used widely in industrial detergents, paints, herbicides, household products, plastics and as process aids in pulp and paper production and textile manufacturing (Guenther et al., 2002). Occurrence in a wide variety of German foods suggests multiple entry points into the food chain such as residues of cleaning detergents or use as excipients in pesticide formulations. Nonylphenol- and octylphenolethoxylates (Fig. 9.7) account for approximately 80% and 20% of industrial alkyl phenol ethoxylate use respectively (Yang et al., 2006). Limited data exist on the concentration of alkyl phenols in foods from Germany (Guenther et al., 2002) and duplicate diet samples from the UK (Fernandes et al., 2003). Exposure to nonylphenol for a German adult was estimated to be 0.11 μg/kg bw/day assuming a 70 kg body weight. Exposures for infants exclusively fed with breast milk or infant formula were 0.04 and 0.31 μg/kg bw/day respectively for a 4.5 kg infant. Estimated nonylphenol exposures for New Zealand adults, based on German concentration data, ranged from 0.04 to 0.07 μg/kg bw/day with the greatest contributions to intake from apples, tomatoes, milk, bread and butter. Although exposure estimates were not reported for the UK samples, concentrations of nonylphenol were above average for two vegetarian and one ethnic Indian duplicate diet compared with the 47 normal diets (Table 9.9).

Table 9.9 Dietary exposure estimates for nonylphenol Country

Population group

Germany

Adult, 70 kg Infant, 0–6 mth, 4.5 kg – breastfed Infant, 0–6 mth, 4.5 kg – infant formula Adult male, 80 kg Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg

New Zealand

Mean exposure (μg/kg bw/day) 0.11 0.04

Reference Guenther et al. (2002)

0.31 0.04 0.05 0.07 0.05

Thomson (2005)

Bisphenol A Bisphenol A (2,2-bis(4-hydroxyphenyl)propane) (Fig. 9.1) is a synthetic chemical manufactured from phenol and acetone and used in the plastics industry especially for the production of polycarbonate and epoxy resins (EC, 2002). Consumers are exposed to bisphenol A from food as a result

234

Endocrine-disrupting chemicals in food

Table 9.10

Dietary exposure estimates for bisphenol A Mean exposure (μg/kg bw/day)

Country

Population group

Bisphenol A

t-TDI = 10 μg/kg bw/day RD = 50 μg/kg bw/day Adult Adult, 60 kg Children, 8.8 kg Infants, 0–4 mths, 4.5 kg Adult, 75 kg

11 0.372 0.852 1.63 0.008

Adult, 75 kg Child, 1–5 yrs, 17 kg

0.084 0.124

Europe

New Zealand USA

Reference EC (2002) EPA (1993) Bolt et al. (2001) EC (2002), EU (2003) Thomson and Grounds (2005) NAP (1999) Wilson et al. (2007)

t-TDI = temporary tolerable daily intake. RD = reference dose. 1 A maximum exposure estimate. 2 For a 97.5th percentile (high) consumer and a mean concentration value of 20 μg/kg. 3 Based on a worst case migration from polycarbonate bottles of 10 μg/kg and consumption of 0.7l per day. 4 Median exposure of 300 children.

of migration from the epoxy resin lacquers applied to food containing vessels, especially food cans and leaching from polycarbonate bottles. Exposure estimates vary from 0.008 to 1 μg/kg bw/day for an adult (Table 9.10), reflecting differences in concentration data and limitations of the exposure modelling. The exposure of 0.008 μg/kg bw/day in New Zealand (Thomson and Grounds, 2005) is based on scenarios for 4400 individuals with concentration data for individual foods and is therefore more robust than those from the UK (EC, 2002) and USA (NAP, 1999) that assume an average concentration or migration value and a population consumption amount. The UK estimate of 0.4 μg/kg bw/day is derived from the 97.5th percentile consumer and is therefore a high rather than a mean estimate. The risk of exposure on the health of the consumer is discussed in Chapter 16. Phthalates Phthalates are used primarily as a softener for products made with polyvinyl chloride (PVC). Certain plastics may contain up to 40% phthalate by weight (IPCS, 2002a). In the food industry phthalate compounds are most commonly used for plastic packaging materials. Because the phthalates are not covalently bound within the plastic matrix they can migrate from the product into food and the environment. Chemical structures of the phthalates included in the limited exposure estimates reported, namely di-2ethylhexyl phthalate (DEHP), dibutyl phthalate (DBP), benzyl butyl phthalate (BBP), di-isononyl phthalate (DINP) and di-isodecyl phthalate (DIDP) are shown in Fig. 9.7.

Exposure to endocrine-disrupting chemicals in food O

235

O O

OR' OR'' O Phthalate ester

O Benzyl butyl phthalate (BBP) O

O

Di-2-ethylhexyl phthalate (DEHP)

O Dibutyl phthalate (DBP) O O O

O O

O O O

O O

O O

O Di-isononyl phthalate (DINP)

O Di-isodecyl phthalate (DIDP)

Fig. 9.7 General chemical structure of a phthalate ester and specific phthalate esters known to occur in food or children’s toys.

In all samples of hospital duplicate diets in Japan, DEHP was present at the highest concentration (Table 9.11; Tsumura et al., 2001). DINP is the principal phthalate in soft toys, apparent by the high exposure for very young children in Denmark (but not the US) (Müller et al., 2003; Babich et al., 2004). The difference in results between these two studies suggests a difference in the composition of the plastics or a major difference in child behaviours between Denmark and the USA – the latter reason seems less likely. Polybrominated diphenyl ethers Polybrominated diphenyl ethers (PBDEs) are synthetic chemicals added to a wide variety of consumer products such as plastics, polyurethane foam and textiles to improve their fire resistance. Three main commercial products, namely pentabromodiphenyl ether (pentaBDE), octabromodiphenyl ether (octaBDE) and decabromodiphenyl ether (decaBDE) are produced, each product containing distinctive combinations of the 209 theoretically possible congeners. Higher congener numbers indicate a higher degree of bromination. The general chemical structure for PBDEs and the specific congeners that are most commonly measured in food are shown in Fig. 9.8. Although decaBDE accounted for almost 80% of the total world annual usage of 70 000 tonnes in 2001, very few studies of dietary exposure have included the decaBDE congener, 209 (FAO/WHO, 2006). Exposure estimates from food are shown in Table 9.12. Using these data, the FAO/WHO recently estimated regional dietary exposures of 0.0022 μg/kg bw/day for a

Table 9.11

Dietary exposure estimates for phthalates Mean exposure (μg/kg bw/day)

Country

Population group

DEHP Denmark

Japan

TDI = 50 μg/kg bw/day Adults Children, 1–6 yrs Children, 7–14 yrs Adult, 70 kg Adult, 65 kg1

4.5 26 11 2.7–4.3 8.0

Sweden

Adult

5.6

United Kingdom

Adult Newborn infants Infant, 6 mths

2.5 13.3 (1998) 35 (1996) 7.7 (1998) 23 (1996)

DBP Denmark

TDI = 10 μg/kg bw/day Adults

1.8

Sweden

Children, 1–6 yrs Children, 7–14 yrs Adult, 70 kg Adult

8 3.5 1.8–4.1 2.7

UK

Adult Newborn infants Infant, 6 mths

0.2 2.4 (1998) 14 (1996) 1.4 (1998) 9.3 (1996)

BBP Denmark

TDI = 500 μg/kg bw/day Adults Children, 1–6 yrs Children, 7–14 yrs Adult, 70 kg Adult, 65 kg1

0.97 5.9 2.4 0.3–0.4 0.07

UK

Adult Newborn infants Infant, 6 mths

0.1 0.2 (1998) 8.7 (1996) 0.1 (1998) 5.6 (1996)

DINP Denmark Japan

TDI = 150 μg/kg bw/day Adults Children, 1–6 yrs Children, 7–14 yrs Adult, 65 kg1

5.1 63.42 10 1.0

UK USA

Adult Children, 1–2 yrs

<0.17 0.08

DIDP Denmark

TDI = 150 μg/kg bw/day Adults Children, 1–6 yrs Children, 7–14 yrs Adult

2.9 53.42 6.8 <0.17

Japan

UK

Reference EFSA (2005a) Müller et al. (2003) EFSA (2005a) Tsumara et al. (2001) Franco et al. (2007) MAFF (1996) MAFF (1998) EFSA (2005b) Müller et al. (2003) EFSA (2005b) Franco et al. (2007) MAFF (1996) MAFF (1998) EFSA (2005c) Müller et al. (2003) EFSA (2005c) Tsumara et al. (2001) MAFF (1996) MAFF (1998) EFSA (2005d) Müller et al. (2003) Tsumara et al. (2001) EFSA (2005d) Babich et al. (2004) EFSA (2005e) Müller et al. (2003) EFSA (2005e)

TDI = tolerable daily intake, DEHP = di-2-ethylhexyl phthalate, DBP = dibutyl phthalate, BBP = benzyl butyl phthalate, DINP = di-isononyl phthalate, DIDP = di-isodecyl phthalate. 1 Body weight assumed, intake reported as μg/day. 2 Mainly from toys.

Exposure to endocrine-disrupting chemicals in food Brx 3'

Bry 2'

4'

2

3

6'

6

Br

Br

Br Br Br Br

BDE-153 (2,2′,4,4′,5,5′-hexaBDE)

Br

O Br

Br

O

Br BDE-99 (2,2′,4,4′,5-pentaBDE)

BDE-47 (2,2′,4,4′-tetraBDE)

Br Br O

Br

Br

O

5

Polybrominated diphenyl ethers (PBDE)

Br

Br Br

Br

Br 4

O 5'

237

Br Br Br

BDE-154 (2,2′,4,4′,5,6′-hexaBDE)

Br

Br

Br O

Br Br

Br Br Br BDE-209 (2,2′,4,4′,5,5′,6,6′-decaBDE) Br

Fig. 9.8 General chemical structure of polybrominated diphenyl ethers (PBDEs) and four of the individual congeners that are commonly measured in food (BDE-47, BDE-99, BDE-153 and BDE-154). BDE-209 is the most widely used PBDE but is less often measured in food.

European diet and 0.0036 μg/kg bw/day for a North American diet. Fish and shellfish were the main contributors to total intakes of PBDEs in the European countries and Japan while meats, poultry and products derived from these foods were the major contributors to the total intakes of PBDEs in Canada and the US (for references see Table 9.12). Exposure of PBDEs for breastfed infants in North America may be in the order of 100 times higher than breastfed infants in Germany. Analysis of archived samples collected over the past 30–40 years has demonstrated significant increases in concentrations of PBDEs in environmental samples and in some samples from humans in Europe and North America (FAO/WHO, 2006).

9.5.5 Environmental contaminants Dioxins Dioxins are persistent organochlorine contaminants widely dispersed in the environment that accumulate in fatty foods. The term ‘dioxins’ is commonly used to refer to a group of 75 polychlorinated dibenzo-p-dioxin (PCDD), 135 polychlorinated dibenzofuran (PCDF) congeners and a number of PCBs that have similar chemical structures and properties (Fig. 9.9). In the 1998 WHO evaluation, 29 congeners were considered to have similar ‘dioxin-like’ toxicity and were assigned toxicity equivalent factors (TEFs) for the calculation of toxic equivalents (TEQs) (van den Berg et al., 1998). The TEQ approach is generally used to express the sum of exposure to combinations of different dioxins, taking into account the differing toxicities of individual congeners relative to the most toxic dioxin

238

Endocrine-disrupting chemicals in food

Table 9.12

Dietary exposure estimates for PBDEs

Country Canada

Population group

Finland

Per capita, 70 kg1 Adult, 70 kg1

Japan

Adult, 65 kg1

Spain

Per capita, 70 kg1

Sweden

Adult female, 65 kg1 Adult male, 75 kg1 Adult, 17–79 yrs, 73.3 kg Adult, 70 kg1

The Netherlands UK

Adult, 70 kg1

USA

Adult male

Per capita, 1–70+ yrs, 53 kg 1 2

PBDE congeners included

Mean exposure (μg/kg bw/day)

28, 47, 99, 100, 153, 154, 183 47, 99, 100, 153, 154 47, 49, 66, 99, 100, 119, 153, 154 Tetra-, penta-, hexa-, hepta-, octaBDE 47, 99, 100, 153, 154

0.0004–0.0006 0.0006 0.0010–0.0017

Reference FAO/WHO (2006) FAO/WHO (2006) FAO/WHO (2006)

0.0012

FAO/WHO (2006)

0.0006

FAO/WHO (2006)

0.0006 47, 99, 100, 153, 154

0.0007

28, 47, 71, 77, 99, 100, 153, 154, 183, 190, 209 47, 99, 100, 153, 154 17, 28, 47, 66, 77, 85, 99, 100, 138, 153, 154, 183, 209 28, 33, 47, 85, 99, 100, 153, 154, 183

0.0002

0.0013 0.0020

0.0084–0.0252

Darnerud et al. (2006) FAO/WHO (2006) FAO/WHO (2006) FAO/WHO (2006) Huwe and Larsen (2005)2

Body weight assumed, intake reported as ng/day. Lean to high fat meat consumption only.

2,3,7,8-tetrachloro-dibenz-dioxin (TCDD). PCDDs and PCDFs are generated as unwanted by-products of manufacturing such as bleaching pulp and paper and the preparation of some pesticides as well as combustion processes, such as forest fires, volcanic eruptions and burning chlorinated plastics. PCBs have entered the environment from their commercial use, for example, as electrical insulators in power transformers. Food consumption accounts for over 90% of exposure to dioxins, with products of animal origin and fish making the greatest contribution to this exposure. Dioxins are the most widely quantified environmental contaminants and Liem et al. (2000) have published a concise summary of the global situation up to 2000.

Exposure to endocrine-disrupting chemicals in food Cly

Clx 1 2

O

Clx

Cly 1

9 8

9

2

8

3

3 4

O

7

Polychlorinated dibenzo-p-dioxins (PCDDs) 2,3,7,8-TCDD 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDD OCDD

239

4

O

7 6

Polychlorinated dibenzofurans (PCDFs) 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-hxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDF

Fig. 9.9 General chemical structure of dioxins (PCDDs) and dibenzofurans (PCDFs) and identification of those that have similar ‘dioxin’-like toxicities.

A compilation of more recent exposure estimates is shown in Table 9.13. The dietary exposure of dioxins (including the dioxin-like PCBs) by the general population of industrialised countries ranges from 0.1 to 7 WHOTEQ pg/kg bw/day. Infants and young children have a relatively higher intake because of their small body weight and may exceed the tolerable daily intake (TDI) (Bocio and Domingo, 2005; Fattore et al., 2006; Weijs et al., 2006). As a result of deliberate efforts to reduce the release of dioxins to the environment, human exposure has fallen over time in all countries where data are available (Liem et al., 2000; FSA, 2003; Baars et al., 2004; Bocio and Domingo, 2005). The risk of exposure to dioxins, on the health of the consumer is discussed in Chapter 2. Polychlorinated biphenyls PCBs were produced commercially from the 1930s to the 1970s and used in a wide range of applications, such as inks, flame-retardants, paints, electrical insulators, heat transfer systems and hydraulic fluids. Twelve of the 209 possible PCB congeners are ‘dioxin-like’ with assigned TEF, namely IUPAC numbers 77, 81, 105, 114, 118, 123, 126, 156, 157, 167, 169 and 189. In some cases, dietary exposure to non-dioxin-like PCBs is expressed as a summation of exposure to seven ‘indicator PCBs’ considered representative of all non-dioxin like PCBs found in food (IUPAC numbers 28, 52, 101, 118. 138, 153 and 180) (Fig. 9.10)

240

Endocrine-disrupting chemicals in food

Table 9.13 Recent dietary exposure estimates for dioxins and dioxin-like PCBs Country

Population group

Dioxins + DL-PCBs

TDI = 2 pg WHOTEQs /kg bw/day = 4 pg WHO-TEQs /kg bw/day Adult male, 30–44 yrs Adult female, 30–44 yrs Toddler, 2–4 yrs

Australia

Egypt

Adult, 60 kg

Italy

Adults, 13–94 yrs Children, 7–12 yrs Infants, nonbreastfed, 0–6 mths Adult, 50 kg

Japan

Korea Spain

Sweden Taiwan The Netherlands

UK

Per capita, 1–70+ yrs Adult, 60 kg, marine foods only Adult, 70 kg

Mean exposure (WHO TEQ/kg bw/day)1

SCF (2001)

PCDD/F/DL-PCBs: 0.1–0.5 PCDD/F/DL-PCBs: 0.1 PCDD/F/DL-PCBs: 0.2–1.2 PCDD/F/DL-PCBs: 6.0–6.7 PCDD/F/DL-PCBs: 2.28 PCDD/F/DL-PCBs: 3.37 PCDD/F/DL-PCBs: 5.34 PCDD/F/DL-PCBs: 1.552 PCDD/F/DL-PCBs: 1.49 PCDD/F/DL-PCBs: 0.68 PCDD/F/DL-PCBs: 3.22

Adult male, 70 kg

PCDD/Fs: 0.91

Adult female, 55 kg Child, 4–9 yrs, 24 kg Adult, 17–79 yrs, 73.7 kg

PCDD/Fs: 1.16 PCDD/Fs: 2.17 PCDD/F/DL-PCBs: 1.3

Adult male, 64.8 kg Adult female, 56.3 kg Per capita, 1–97 yrs

PCDD/F/DL-PCBs: 1.49 PCDD/F/DL-PCBs: 1.32

Infants, nonbreastfed, 5 mths Infants, nonbreastfed, 13 mths Adults Children, 4–18 yrs

PCDD/F/DL-PCBs: 1.1

Toddlers, 1.5–4.5 yrs

Reference

PCDD/F/DL-PCBs: 6.8

Mato et al. (2007) FSANZ (2004)

Loutfy et al. (2006) Fattore et al. (2006)

Sasamoto et al. (2006) Mato et al. (2007) Moon and Ok (2006)3 Fernández et al. (2004) Bocio and Domingo (2005) Darnerud et al. (2006) Hsu et al. (2007) Baars et al. (2004) Weijs et al. (2006)

PCDD/F/DL-PCBs: 2.3 PCDD/F/DL-PCBs: 0.9 PCDD/F/DL-PCBs: 0.7–1.8 PCDD/F/DL-PCBs: 1.7–2.2

FSA (2003)

Table 9.13 Continued Mean exposure (WHO TEQ/kg bw/day)1

Country

Population group

USA

Adult, 53 kg

PCDD/F/DL-PCBs: 0.1–0.3

Per capita

PCDD/Fs: 0.2–0.45

Children, 2 yrs

PCDD/Fs: 0.7–1.25

Reference Huwe and Larsen (2005)4 Charnley and Doull (2005)

TDI = tolerable daily intake, PCDD = polychlorinated dibenzo-p-dioxins, PCDFs = dibenzofurans, DL-PCBs = dioxin like polychlorinated biphenyls. 1 Expressed as WHO toxic equivalents (van den Berg et al., 1998). 2 2004 data only shown. 3 Seafood only included in assessment. 4 Meat only included in the assessment. 5 2002 data only shown.

Cly

Clx 3'

2'

2

Cl 3

4'

Cl

4 5'

6'

6

Cl

Cl Cl

PCB-126

Cl

Cl

Cl

Cl

PCB-169

Cl

PCB-118

PCB-28 Cl Cl

Cl Cl

Cl Cl

Cl Cl

Cl

Cl

PCB-52

PCB-101

Cl Cl

Cl Cl

Cl

Cl Cl PCB-153

Cl

Cl

Cl

Cl

Cl Cl

Cl

Cl

Cl

PCB-77

Cl

Cl

Cl

5

PCB Cl

Cl

Cl

Cl

Cl Cl

Cl PCB-138

Cl Cl

Cl Cl

Cl PCB-180

Fig. 9.10 General chemical structure of polychlorinated biphenyls (PCBs). Ten possible chlorine atoms leads to 209 congeners. The co-planar non-ortho, dioxin-like PCBs, 77,126 and 169, mono-ortho 118 and ‘indicator’ congeners, 28, 52, 101, 138, 153 and 180, are also shown.

242

Endocrine-disrupting chemicals in food

A summation of recent data for exposure to PCBs on a μg/kg bw/day basis is given in Table 9.14. Despite variations in the number of congeners included in the estimates, dietary exposure of PCBs by the general population of industrialised countries spans a narrow range from 0.001 to 0.008 μg/ kg bw/day, with levels about 100 times higher for the unique diet of Greenland. Fish is a major contributor to PCB exposure (Buckland et al., 1998; Baars et al., 2004; Darnerud et al., 2006) with meat and cereals significant sources in some countries (Baars et al., 2004; Turci et al., 2006). Exposure to PCBs from food has decreased in the same way as exposure to dioxins (SCF, 2001; JECFA, 2002; Baars et al., 2004). Polycyclic aromatic hydrocarbons PAHs are a group of more than 100 compounds composed of two or more fused aromatic rings and formed during the incomplete combustion of organic matter. Structures of the 16 US Environment Protecting Agency (EPA) priority pollutant PAH are shown in Fig. 9.11. PAHs are ubiquitous environmental contaminants that enter the food chain from atmospheric deposition on to plants and soil and as a result of smoking or cooking at high temperatures (e.g. barbequing/charbroiling). The most studied, the most carcinogenic and weakly estrogenic PAH, benzo[a]pyrene (B[a]P), is sometimes analysed as a biomarker for total PAH. Dietary exposure assessments for ‘total PAHs’ and B[a]P are shown in Table 9.15. In the most recent assessment cereals accounted for the greatest contribution to PAH exposure from food followed by meat and meat products, dairy products, fish and shellfish and oils and fats (Ibánˇez et al., 2005).

9.5.6 Discussion of exposure assessments Current estimates of external exposure to EDCs from food range from a low of picogram equivalents of the unwanted environmental contaminant dioxins to half milligram intakes of the naturally occurring isoflavones and flavanones. Data for environmental contaminants are much more prevalent than for naturally occurring compounds, reflective perhaps of political pressures and involuntary risk aversion to synthetic chemicals. Real risks of synthetic versus naturally occurring EDCs are yet to be confirmed. Only mean exposure data have been presented in this chapter but wide ranges in intake are observed, particularly for the phytoestrogens that are associated with particular plants compared with industrial chemicals and environmental contaminants that are generally spread across a wide range of foods, including those of animal origin. Where comparative data exist, Asian diets are higher by an order of magnitude in exposure to the phytoestrogens coumestrol and isoflavones, genistein and daidzein, largely because of cultural differences in soy consumption. The seemingly high DDT levels observed in 46 diets from India

Adult male, 80 kg

New Zealand

Per capita, 1–97 yrs

0.0056

0.0007 0.008

0.001 0.001

Baars et al. (2004)

Darnerud et al. (2006)

Thomson (2005)

Koizumi et al. (2005)

0.0033 0.001

Johansen et al. (2004) Turci et al. (2006)

Reference

0.33 0.004

Mean exposure (μg/kg bw/day)

Of a possible 209 congers, ‘dioxin-like’ PCBs are IUPAC numbers 77, 81, 105, 114, 118, 123, 126, 156, 157, 167, 169, and 189 and ‘indicator PCBs’ are IUPAC numbers 28, 52, 101, 118, 138, 153 and 180. 2 Bodyweight assumed, intake reported as mg/day. 3 1995 data.

1

The Netherlands

28, 31, 52, 66, 74, 77, 101, 105, 110, 114, 118, 126, 128, 138, 149, 153, 156–8, 167, 169, 170, 180 28, 52, 101, 118, 138, 153, 180

Adult females, 55 kg

Japan

Sweden

28, 31, 52, 101, 105, 138, 153, 156, 180 18, 21, 28, 30, 31, 37, 44, 47, 49, 52, 61, 66, 70, 74, 77, 81, 97, 99, 101, 105, 110, 114, 118, 119, 123, 126, 128, 138, 143, 146, 149, 151, 153, 156, 157, 158, 167–172, 177, 178, 180, 183, 187, 190, 193–196, 199, 201, 202, 206–209 74, 99, 118, 138, 146, 153, 156, 163, 170, 180, 182, 52, 77, 99, 101, 105, 114, 118, 123, 126, 138, 153, 156, 157, 167, 169, 170, 180, 183, 187, 189, 194, 202, 206

Adults, 18–60+,70 kg2 Adult, 65 kg

Greenland Italy

Adult female, 65 kg Young male, 19–24 yrs, 70 kg Vegetarian female, 65 kg Adult, 17–79 yrs, 73.7 kg

PCBs included

Population group

Dietary exposure estimates for total PCBs1

Country

Table 9.14

244

Endocrine-disrupting chemicals in food

Naphthalene

Acenaphthylene

Acenaphthene

Fluorene

Fluoranthene

Pyrene

Phenanthrene

Anthracene

Benz[a]anthracene

Chrysene

Benzo[k]fluoranthene

Benzo[b]fluoranthene

Benzo[a]pyrene

Indeno[1,2,3-cd]pyrene

Dibenz[a,h]anthracene

Benzo[ghi]perylene

Fig. 9.11

Chemical structures of the 16 EPA priority pollutant polycyclic aromatic hydrocarbons (PAH) that are commonly analysed in food.

may well not reflect exposures for other regions of Asia, but rather reflect local usage patterns of DDT, also observed in breast milk studies from Indonesia (Burke et al., 2003). Exposure of infants and children to EDCs is generally higher than adults due to a combination of low body weight, diet and behaviours, such as mouthing (teething) toys, and of concern since this population group is at a developmental stage. Adult female exposure is important as a predictor of in utero exposure (see also Chapter 1) and may be the most critical period for exposure because of the establishment of the foetal endocrine system.

9.5.7 Exposure to estrogenic EDCs The use of EQs to assess total estrogenicity from food is a simplification of a complex system where different EDCs may act by different mechanisms to produce a multitude of effects. However, it is considered helpful to provide an estimate of risk and to inform priorities for action (Rudel, 1997; Huang et al., 2007). The data in Table 9.16 show relative estrogenic potencies ranging from 1 × 10−7 to 1 × 10−2 compared with the endogenous hormone 17β-estradiol (E2), for those EDCs with demonstrated in vitro estrogenic activity. Four compounds, bisphenol A, phthalates (BBP and DBP), synthetic pyrethroids and zearalenone are noticeably more estrogenic than the other EDCs and might therefore result in a greater estrogenic effect than exposure to other EDCs. Of course it must be remembered that the exposures shown in Table 9.16 relate only to exposure from food. Total exposure

Exposure to endocrine-disrupting chemicals in food

245

Table 9.15 Dietary exposure estimates for PAH and benzo(a)pyrene (B[a]P) Mean exposure Total PAH (No. PAH B(a)P included in (μg/kg bw/day) assessment) (μg/kg bw/day)

Country

Population group

Finland Italy

Adult, 70 kg1 Adult, 70 kg1

0.57 (16) 0.04 (9)

0.004 0.002

New Zealand

Adult, 70 kg

0.07 (16)

0.003

Norway

Adult men, 75 kg1 Adult women, 65 kg1 Adult men, 75 kg1 Adult women, 65 kg1 Adult men, 75 kg1 Adult women, 65 kg1 Adult, 70 kg1

0.10 (16)

0.0003

0.08

0.0003

0.11 (16)

0.002

0.10

0.001

0.13 (16)

0.002

0.11

0.002

0.05

0.003

Adult, 70 kg1 Young male, 70 kg1 Adult male, 19–50 yrs, 75 kg1 Adult, 70 kg1

0.07 0.07–0.25 (17)

0.001 0.002–0.004

NR

0.0006–0.0009

Adult, 70 kg1

NR

0.001

Spain

UK The Netherlands USA

1

0.04 (9)

Reference

EC (2004) Lodovici et al. (1995) Thomson and Lake (1994) EC (2004)

Falco et al. (2003) Ibánˇez et al. (2005) Dennis et al. (1983) EC (2002) De Vos et al. (1990) Menzie et al. (1992) Kazerouni et al. (2001) Sinha et al. (2005)

Body weight assumed, intake reported as μg/day.

including inhalation and dermal absorption will add to the estrogenic load. Ideally internal exposure would be derived from serum levels but this does not distinguish food from other sources and currently there are very limited data on human serum levels for representative population groups. The low A : T ratios of the naturally occurring phytoestrogens (of the order of 1%) are indicative of extensive metabolism by gut microflora and/or in the liver, excretion and elimination. A similarly low A : T ratio is observed for the synthetic pyrethroid pesticides that are rapidly metabolised by

246

Endocrine-disrupting chemicals in food

esterases and eliminated by the kidneys. By contrast, the synthetic estrogenic compounds show variable A : T ratios, most probably reflective of the limited serum data available but also perhaps indicative of less extensive metabolism and bioaccumulation. Bioaccumulation is a feature of lipidsoluble PCBs and explains the high A : T ratio for this group of EDCs. The lipid-soluble compounds are likely to be sequestered in fat and so perhaps less active until released during fat mobilisation such as the time of breastfeeding. The lack of serum data for zearalenone is an important data gap, because of the high estrogenic potency of this compound. Metabolism might also increase estrogenicity by converting a nonestrogenic compound to a product that is estrogenic. We know this is the case for PCBs and synthetic pyrethroids. Table 9.16 Summary of upper mean adult exposures to EDCs and estimate of total estrogenicity from food

EDC

Relative estrogenic potency × 106

Phytoestrogens Apigenin Coumestrol Daidzein Enterodiol Enterolactone Genistein Isoliquiritigenin Kaempferol Luteolin Naringenin Phloretin Quercetin Mycotoxins Zearalenone

1504 3005 1104 0.16 16 2604 257 704 604 804 254 18

Intake (μg/kg bw/day)1 Western diet

13 0.5 16 9.6 3.7 32 35 139 13 141 97 179

Asian diet

XEQ (μg/l) A:T ratio1

Western diet2

Asian diet3

NR 7.1 278 NR NR 508 NR NR NR NR NR NR

0.01 0.01 0.005 0.005 0.005 0.01 0.01 0.001 0.01 0.003 0.006 0.003

4.7E-04 3.6E-05 2.1E-04 1.9E-07 7.1E-06 2.0E-03 2.1E-04 2.3E-04 1.9E-04 8.2E-04 3.5E-04 1.3E-04

4.4E-04 4.8E-04 3.4E-03 1.7E-07 6.6E-06 3.0E-02 2.0E-04 2.2E-04 1.7E-04 7.6E-04 3.3E-04 1.2E-04

10,0006,9

0.056

0.033

110

1.4E-02

7.4E-03

Pesticides Dieldrin ΣDDT12 Endosulfan Synthetic pyrethroids13

19 19 19 100014

0.0015 0.004 0.0404 0.0032

NR 0.45 NR 0.009

111 0.4 170 0.02

3.6E-08 3.9E-08 1.7E-04 1.5E-06

3.4E-08 4.0E-06 1.5E-04 4.0E-06

Industrial chemicals Bisphenol-A Nonyl phenol PBDEs Phthalates (DBP, BBP)16

100015 309 NA 1

0.084 0.11 0.0252 3.7

NR NR NR NR

1 10 NA 1

2.0E-03 8.0E-04 NA 8.9E-02

1.9E-03 7.4E-04 NA 8.3E-02

Exposure to endocrine-disrupting chemicals in food

247

Table 9.16 Continued

EDC

B[a]P Dioxins PCBs Total XEQ ng/l

Relative estrogenic potency × 106 0.318 NA 19

Intake (μg/kg bw/day)1 Western diet 0.004 0.000 006 8 0.008

Asian diet NR NR 0.003

XEQ (μg/l) A:T ratio1 1 NA 100

Western diet2

Asian diet3

2.9E-08 NA 1.9E-05 110

2.7E-08 NA 6.7E-06 130

NA = not applicable, NR = no result, B[a]P = benzo[a]pyrene, A : T = measured serum level/ theoretical serum level, based on food intake/serum volume (2.9l), XEQ = intake × REP × 10−6 × A : T × body weight. 1 Thomson et al. (2003), Thomson (2005). 2 Based on 70 kg body weight. 3 Based on 65 kg body weight. 4 Breinholt and Larsen (1998). 5 Mayr et al. (1992). 6 Welshons et al. (1987). 7 Assigned the same estrogenicity as phloretin on the basis of structural similarity and limited gene expression data. 8 Le Bail et al. (1998). 9 Soto et al. (1995). 10 Assumed in the absence of serum data. 11 Ranges from 0.5 to 6.4 based serum levels from Buckland et al. (2001). 12 Excluding non-estrogenic DDE. 13 Excluding non-estrogenic permethrin. 14 Go et al. (1999). 15 Steinmetz et al. (1997). 16 Includes estrogenic DBP and BBP. 17 Zacharewski et al. (1998). 18 Vondrácˇek et al. (2002).

Adult exposure to total estrogenicity for the EDCs was estimated to be 110 ng/l of serum and 130 ng/l for Western and Asian diets respectively (Table 9.16). In the context of normal serum levels of E2 ranging from 10 to 50 ng/l for human males, and 20–350 ng/l for non-pregnant premenopausal women during the menstrual cycle (Greenspan and Gardener, 2001), the contribution from food would be expected to have a pharmacological effect. This total exposure from food is higher than has been estimated previously due to the inclusion of phthalates (Thomson et al., 2003) and combines upper mean exposure estimates across different nations. However, it is highly unlikely that upper mean exposures would apply for all chemicals across any one population group and it is acknowledged that a worst case is presented. On the other hand, there are undoubtedly chemicals with potential endocrine effects that have not been included, or even discovered yet! When the various estrogenic EDCs are ranked, exposure

Fig. 9.12 10.0

0.0 Isoliquiritigenin

Daidzein

Kaempferol

Phloretin

Apigenin

Nonyl phenol

Naringenin

Genistein

Bisphenol A

∑DDT2 B(a)P1

B(a)P1

Synthetic pyrethroids3

Dieldrin

Enterodiol

∑DDT2

Dieldrin

Synthetic pyrethroids3

Enterolactone

Enterodiol

PCBs Enterolactone

PCBs

Coumestrol

20.0

Quercetin

30.0 Quercetin

40.0

Endosulfan

Asian diet

Luteolin

50.0 Endosulfan

60.0

Luteolin

70.0

Isoliquiritigenin

Kaempferol

Phthalates Zearalenone

Contribution to estrogenicity (%) 70.0

Phloretin

Apigenin

Coumestrol

Nonyl phenol

Naringenin

Bisphenol A

Daidzein

Zearalenone

Genistein

Phthalates

Contribution to estrogenicity (%)

248 Endocrine-disrupting chemicals in food

90.0

80.0

60.0 Western diet

50.0

40.0

30.0

20.0

10.0

0.0

Contribution of estrogenic EDCs to total estrogenicity for average Western and Asian diets.

to total estrogenicity is clearly dominated by the phthalates (DBP and BBP) for both Asian and Western diets (Fig. 9.12). Such a ranking rightly draws attention to the robustness of the contributing data. The dietary exposure data are limited to three countries and are

Exposure to endocrine-disrupting chemicals in food

249

dominated by the work of Franco and colleagues from Sweden (2007). More importantly, the estrogenicity is based on in vitro assays undertaken by Zacharewski and co-workers (1998) who found that of eight phthalates tested, only BBP, DBP and DHP exhibited estrogenicity in some assays but none of the eight phthalate esters showed estrogenic effects in two in vivo assays. These results serve to raise caution in assessing the potential hazard of chemicals based solely upon the results of in vitro experiments. Clearly the possible role of phthalates as xenoestrogens warrants clarification.

9.6

Implication for the food industry

Humans are exposed, via the consumption of food, to an array of chemicals that may disrupt the endocrine system. Therefore the food industry will be, and is, involved in voluntary or regulatory controls for some of these chemicals. For example there are limits on the allowable migration of bisphenol A into the contents of canned food. Exposure may be as low as picograms per kilogram body weight per day and it is extremely difficult to prove, or disprove, that exposure to a chemical from food results in an effect that may not be observed for 10–30 years. The current lack of evidence of human health effects does not equate to lack of effect but more likely to the lack of being able to prove, or disprove, an effect. Such proof can be accumulated only from a combination of epidemiological, toxicological and mechanistic studies. In addition, the scientific community is challenged to prove cause and effect from exposure to low doses of mixtures of chemicals. Only when an EDC is sufficiently characterised as causing an adverse health effect will limits possibly be imposed on the food industry for other EDCs. Many contributing foods (e.g. fish, cheese, vegetables, soy) have beneficial features that need to be taken into account as part of risk management. Similarly, some EDCs, in particular the phytoestrogens, have claimed beneficial effects that also need to be included in risk management strategies for the food industry. It would be a pity to draw negative conclusions about foods with beneficial characteristics simply because they contain estrogenic compounds. Similarly, in some circumstances estrogenic components of food per se might be beneficial. A responsible food industry will be responsive to emerging information as this becomes available to ensure the safety of the food supply with respect to EDCs and improved public health for all and will make appropriate risk/benefit decisions.

9.7

Future trends

Knowledge of the level of exposure to EDCs in food quantifies the external human dose to this group of compounds. This knowledge is now being applied to epidemiological studies of associations between dietary exposures

250

Endocrine-disrupting chemicals in food

and risk of colorectal, ovarian and prostate cancer in addition to heart disease (Rossi et al., 2006; Gates et al., 2007; Kurahashi et al., 2007; Mursu et al., 2007; Theodoratou et al., 2007). Exposure information is key for the design of physiologically relevant molecular approaches that offer exciting potential to investigate the hazard characterisation of individual EDCs and EDCs in combination, using in vitro and in vivo systems. An understanding of mechanistic pathways at a molecular level is one piece of evidence to be considered in conjunction with whole animal and human epidemiological studies to clarify the risk of EDCs for humans. There is an immediate need, and opportunity, to explore the health effects of phytoestrogens. In general Asians eat much more soy and have lower rates of hormonal cancers, so is soy more good than bad, or do different population groups respond differently because of genetic susceptibilities? Soy is a relatively recent and increasing component of a Western diet – is this a good thing or not? The problem is, of course, proving a cause and effect relationship between dietary components and physiological effects on the consumer. The list of EDCs in food will grow as more chemicals are identified as having potential EDC effects. This list will include metabolites. Further data on both external and internal dose will be published. The role of endogenous hormones that may be present in dairy foods is gaining some attention. What role do they play in the suite of EDCs? At-risk sub-populations will be identified from mechanistic and epidemiological studies and targeted for risk management. The higher exposure of children, and particularly exposure in utero (see Chapter 1), is an area of needed research. Does imprinting for conditions, such as cancer in later life, occur as a result of exposure from food that the mother has consumed? Exposure assessments give us information about how much of the wide array of chemicals that are potential EDCs we are consuming in our diet. Evidence to date suggests the amounts are sufficient to cause a pharmacological effect for humans. However, at this time we are far from having a clear understanding of which compounds are, or are not, a health issue for humans.

9.8

Sources of further information and advice

Detail of specific investigations can be gained from the references cited. The following two international collaborative publications provide excellent background to the overall area of EDCs in the environment and impact for human health, up to 2003: International Programme on Chemical Safety, 2002a Global assessment of the state-of-the science of endocrine disruptors. Damstra T, Barlow S, Bergman A, Kavlock R, Van der Kraak G. (Eds.) Geneva,WHO.

Exposure to endocrine-disrupting chemicals in food

251

SCOPE/IUPAC, 2003, Implications of endocrine active substances for humans and wildlife – a SCOPE/IUPAC project. Pure and Applied Chemistry, 2003 75(11–12), 1619–2611.

9.9

References

adlercreutz h and mazur w, Phytoestrogens and Western diseases. Ann Med; 1997, 29, 95–120. andrade pb, carvalho arf, seabra rm, ferreira ma, A previous study of phenolic profiles of quince, pear and apple purees by HPLC diode array detection for the evaluation of quince puree genuineness, J Agric Food Chem; 1998, 46: 968–972. arai y, watanabe s, kimira m, shimoi k, mochizuki r, kinae n, Dietary intake of flavonols, flavones and isoflavones by Japanese women and the inverse correlation between quercetin intake and plasma LDL cholesterol concentration. J Nutr; 2000, 130, 2243–2250. baars aj, bakker mi, baumann ra, boon pe, freijer ji, hoogenboom lap, hoogerbrugge r, van klaven jd, liem akd, traag wa, de vries j, Dioxins, dioxin-like and non-dioxin like PCBs in foodstuffs: occurrence and dietary intake in The Netherlands, Toxicol Lett; 2004, 151, 51–61. babich ma, chen sb, greene ma, kiss ct, porter wk, smith tp, wind ml, zamula ww, Risk assessment of oral exposure to diiononyl phthalate from children’s products, Regul Toxicol Pharmacol; 2004, 40(2), 151–167. battu rs, singh b, kang bk, joia bs, Risk assessment through dietary intake of total diet contaminated with pesticide residues in Punjab, India, 1999–2002, Ecotoxicol Environ Saf; 2005, 62, 132–139. bocio a, domingo jl, Daily intake of polychlorinated dibenzo-p-dioxins/polychlorinated dibenzofurans (PCDD/PCDFs) in foodstuffs consumed in Tarrogona, Spain: a review of recent studies (2001–2003) on human PCDD/PCDF exposure through the diet, Environ Res; 2005, 97, 1–9. bolt hm, janning p, michna h, degen gh, Comparative assessment of endocrine modulators with oestrogenic activity: 1. Definition of a hygiene-based margin of safety (HBMOS) for xeno-oestrogens against the background of European developments, Arch Toxicol; 2001, 74(11), 649–662. breinholt v, larsen jc, Detection of weak estrogenic flavonoids using a recombinant yeast strain and a modified MCF7 cell proliferation assay. Chem Res Toxicol; 1998, 11, 622–629. buckland s, scobie s, heslop v, Organochlorines in New Zealand: Concentrations of PCDDs, PCDFs and PCBs in retail foods and an assessment of dietary intake for New Zealanders, 1998, Ministry for the Environment, Wellington, New Zealand. buckland s, bates m, garrelt n, ellis h, van maanen t, Concentrations of selected organochlorines in the serum of non-occupationally exposed New Zealand population, 2001, Ministry for the Environment, Wellington, New Zealand. burda s, oleszek w, lee cy, Phenolic compounds and their changes in apples during maturation and cold storage, J Agric Food Chem; 1990, 38, 945–948. burger j, Making decisions in the 21st century: Scientific data, weight of evidence, and the precautionary principle. Pure Applied Chem; 2003, 75(11–12), 2505–2513. burke er, holden aj, shaw ic, suharyanto fx, sihombing g, Organochlorine pesticide residues in human milk from primiparous women in Indonesia. Environ Contam Toxicol; 2003, 71, 148–154. carson r, 1962, Silent Spring. Hamish Hamilton Ltd, London.

252

Endocrine-disrupting chemicals in food

chao h-r, wang s-l, lin t-c, chung x-h, Levels of organochlorine pesticides in human milk from central Taiwan, Chemosphere; 2006, 62, 1774–1785. charnley g, doull j, Human exposure to dioxins from food, 1999–2002, Food Chem Toxicol; 2005, 43, 671–679. chun ok, chung sj, song wo, Estimated dietary flavonoid intake and major food sources of US adults, J Nutr; 2007, 137, 1244–1252. clarke db, lloyd as, Dietary exposure estimates of isoflavones from the 1998 UK Total Diet Study. Food Addit Contam; 2004, 21(4), 305–316. clarke db, barnes ka, castle l, rose m, wilson la, baxter mj, price kr, dupont s, Levels of phytoestrogens, inorganic trace-elements, natural toxicants and nitrate in vegetarian duplicate diets. Food Chem; 2003, 81, 287–300. cot (committee on toxicity), Committee on toxicity of chemicals in food, consumer products and the environment. Phytoestrogens and health; 2003, London, UK: Crown Copyright. Available at http://www.food.gov.uk/science/ouradvisors/ toxicity/COTwg/wg_phyto/ darnerud po, atuma s, aune m, bjerselius r, glynn a, petersson grawé k, becker w, Dietary intake estimations of organohalogen contaminants (dioxins, PCB, PBDE and chlorinated pesticides, e.g. DDT) based on Swedish market basket data, Food Chem Toxicol; 2006, 44, 1597–1606. davies jh, The pyrethroids: an historical introduction. In Leahey JP (Ed) The Pyrethroid Insecticides, 1985. London: Taylor & Francis, 1–41. de kleijn mjj, van der schouw yt, wilson pwf, adlerctreutz h, mazur w, grobbee de, jacques pf, Intake of dietary phytoestrogens is low in postmenopausal women in the United States: The Framington study, J Nutr; 2001, 131, 1826–1832. dennis mj, massey rc, mcweeny dj, knowles me, watson d, Analysis of polycyclic aromatic hydrocarbons in UK total diets, Food Chem; 1983, 21, 569–574. deutch b, dyerberg j, sloth pedersen h, asmund g, møller p, hansen jc, Dietary composition and contaminants in north Greenland, in the 1970s and 2004, Sci Total Environ; 2006, 370, 372–381. de vos rh, van dokkum hw, schouten a, jong-berkhout p, Polycyclic aromatic hydrocarbons in Dutch total diet samples (1984–1986), Food Chem Toxicol; 1990, 28, 263–268. ec (european commission scientific committee on food), 2000, Opinion of the Scientific Committee on Food on Fusarium Toxins. Part 2: Zearalenone. Available from http://europa.eu.int/comm/food/fs/sc/scf/out65_en.pdf ec (european commission scientific committee on food), 2002, Opinion of the Scientific Committee on Food on Bisphenol A. Available from http://ec.europa. eu.int/comm/food/fs/sc/scf/out128_en.pdf ec (european commission), Reports on tasks for scientific cooperation (SCOOP). Collection of occurrence data on polcyclic aromatic hydrocarbons in food. Report of experts participating in task 3.2.12, 2004, Brussels, European Commission. efsa (european food safety authority), Opinion of the AFC panel related to bis(2-ethylhexyl) phthalate (DEHP) for use in food contact materials, EFSA J; 2005a, 243, 1–20. Available from http://www.efsa.europa.eu/en/science/afc/afc_ opinions/1142.html efsa (european food safety authority), Opinion of the AFC panel related to dibutylphtalate (DBP), EFSA J; 2005b, 242, 1–17. Available from http://www.efsa. europa.eu/en/science/afc/afc_opinions/1141.html efsa (european food safety authority), Opinion of the AFC panel related to butylbenzylphthalate (BBP) for use in food contact materials, EFSA J; 2005c, 241, 1–14. Available from http://www.efsa.europa.eu/en/science/afc/afc_opinions/1140. html efsa (european food safety authority), Opinion of the AFC panel related to diisononylphthalate (DINP) for use in food contact materials, EFSA J; 2005d, 244,

Exposure to endocrine-disrupting chemicals in food

253

1–18. Available from http://www.efsa.europa.eu/en/science/afc/afc_opinions/1144. html efsa (european food safety authority), Opinion of the AFC panel related to diisodecylphthalate (DIDP) for use in food contact materials, EFSA J; 2005e, 243, 1–20. Available from http://www.efsa.europa.eu/en/science/afc/afc_opinions/1143. html egan k, Centre for Foods Safety and Nutrition, US FDA, personal communication, 2005. epa (us environmental protection agency), 1993, Bisphenol A, CASRN 80-05-7, IRIS, Integrated Risk Information System, on-line, 1993. Available at http://www. epa.gov/iris/subst/0356.htm eu (european union), European Union risk assessment report. 4,4′-isopropylidenediphenol (bispehenol A), 2003, Luxembourg, European Commission. falco g, domingo jl, llobet jm, teixido a, casas c, müller l, Polycyclic aromatic hydrocarbons in foods: human exposure through the diet in Catalonia, Spain, J Food Prot; 2003, 66, 2325–2331. fao/who, Evaluation of certain food contaminants, Sixty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives, 2006, Geneva, World Health Organization. fattore e, fanelli r, turrini a, di domenico a, Current dietary exposure to polychlorodibenzo-p-dioxins, polychlorodibenzofurans, and dioxin-like polychlorobiphenyls in Italy, Mol Nutr Food Res; 2006, 50, 915–921. fernandes ar, costley ct, rose m, Determination of 4-octylphenol and 4nonylphenol congeners in composite foods, Food Addit Contam; 2003, 20, 846–852. fernández ma, gómara b, bordajandi lr, herrero l, abad e, abalos m, rivera j, gonzález mj, Dietary intakes of polychlorinated dibenzo-p-dioxins, dibenzofurans and dioxin-like polychlorinated biphenyls in Spain, Food Addit Contam; 2004, 21(10), 983–991. franco a, prevedouros k, alli r, cousins it, Comparison and analysis of different approaches for estimating the human exposure to phthalate esters, Environ Int; 2007, 33(3), 283–291. fsa (foods standards agency), Dioxins and dioxin-like PCBs in the UK diet: 2001 total diet study samples 38/03, 2003, London, Food Standards Agency. http://www. food.gov.uk/science/surveillance/ fsanz (food standards australia new zealand), The 20th Australian Total Diet Survey, 2003, Canberra, Food Standards Australia New Zealand, http: //www. foodstandards.gov.au fsanz (food standards australia new zealand), Dioxins in food, dietary exposure assessment and risk characterisation, Technical report series No.27, 2004, Canberra, Food Standards Australia New Zealand, http: //www.foodstandards.gov.au gaido kw, mcdonnell dp, korach ks, safe sh, Estrogenic activity of chemical mixtures: Is there synergism? CIIT Activities; 1997, 17(2), 1–7. gates ma, tworoger ss, hecht jl, de vivo i, rosner b, hankinson se, A prospective study of dietary flavonoid intake and incidence of epithelial ovarian cancer, Int J Cancer; 2007, 121(10), 2225–2232. go v, garey j, wolff ms, pogo bgt, Estrogenic potential of certain pyrethroid compounds in the MCF-7 human breast carcinoma cell line, Environ Health Perspect; 1999, 107, 173–177. grace pb, taylor ji, low y-l, luben rn, mulligan aa, botting np, dowsett m, welch aa, khaw k-t, wareham nj, day ne, bingham sa, Phytoestrogen concentrations in serum and spot urine as biomarkers for dietary phytoestrogen intake and their relation to breast cancer risk in European prospective investigation of cancer and nutrition-Norfolk, Cancer Epidemiol Biomarkers & Prev; 2004, 13(5), 698–708.

254

Endocrine-disrupting chemicals in food

greendale ga, fitzgerald g, huang m-h, sternfeld b, gold e, seeman t, sherman s, sowers m, Dietary soy isoflavones and bone mineral density: results from the Study of Women’s Health Across the Nation, Am J Epidemiol; 2002, 155, 746–754. greenspan f, gardner d, 2001, Ovaries, in Basic & Clinical Endocrinology. New York, Lange Medical Books/McGraw-Hill. guenther k, heinke v, thiele b, kleist e, prast h, raecker t, Endocrine disrupting nonylphenols are ubiquitous in food, Environ Sci Technol; 2002, 36,1676–1680. hsu m-s, hsu k-y, wang s-m, chou u, chen s-y, huang n-c, liao c-y, yu t-p, ling y-c, A total diet study to estimate PCDD/Fs and dioxin-like PCBs intake from food in Taiwan, Chemosphere; 2007, 67, S65–S70. huang y-w, phillips jr, hunter ld, Human exposure to medicinal, dietary, and environmental estrogens, Toxicol Environ Chem; 2007, 89(1), 141–160. huwe jk, larsen gl, Polychlorinated dioxins, furans, and biphenyls, and polybrominated diphenyl ethers in a US meat market basket and estimates of dietary intake, Environ Sci Technol; 2005, 39, 5606–5611. ibánˇ ez r, agudo a, berenguer a, jakszyn p, josé tormo m, josé sanchéz m, quirós jr, pera g, navarro c, martinez c, larranˇ aga n, dorronsoro m, dolores chirlaque m, barricarte a, ardanaz e, amiana p, gonalez ca, Dietary intake of polcyclic aromatic hydrocarbons in a Spanish population, J Food Prot; 2005, 68(10), 2190–2195. ipcs (international programme on chemical safety), Global assessment of the state-of-the science of endocrine disruptors, Damstra T, Barlow S, Bergman A, Kavlock R, Van der Kraak G. (Eds.), 2002a, Geneva, WHO. ipcs (international programme on chemical safety), Inventory of IPCS and other WHO pesticide evaluations and summary of toxicological evaluations performed by the joint meeting on pesticide residues (JMPR), evaluations through 2006, United Nations Environment Programme, International Labour Organisation and the WHO, 2006, Geneva, Switzerland, http://www.who.int/pcs/ publications/jmpr/jmpr_pesticide/en/ jecfa (joint fao/who expert committee on food additives), Zearalenone. Safety evaluation of certain food additives, WHO Food Additive Series 44, 2000, Geneva, WHO. jecfa (joint fao/who expert committee on food additives), Polychlorinated dibenzodioxins, polychlorinated dibenzofurans, and coplanar polychlorinated biphenyls. Safety evaluation of certain food additives and contaminants, WHO Food Additive Series, vol 48. 2002, Geneva, WHO. johannot l, somerset sm, Age-related variations in flavonoid intake and sources in the Australian population, Public Health Nutr; 2005, 9(8), 1045–1054. johansen p, muir d, asmund g, riget f, Human exposure to contaminants in the traditional Greenland diet, Sci Total Environ; 2004, 331, 189–206. kazerouni n, sinha r, hsu c-h, greenburg a, rothman n, Analysis of 200 food items for benz(a)pyrene and estimation of its intake in an epidemiologic study, Food Chem Toxicol; 2001, 39, 423–436. keinan boker l, van der schouw yt, de kleijn mjj, jacques pf, grobbee d, peeters phm, Intake of dietary phytoestrogens by Dutch Women, J Nutr; 2002, 132(6), 1319–1328. kirk p, patterson re, lampe j, Development of a soy food frequency questionnaire to estimate isoflavone consumption in US adults, J Am Diet Assoc; 1999, 99, 558–563. koizumi a, yoshinaga t, harada k et al., Assessment of human exposure to polychlorinated biphenyls and polybrominated diphenyl ethers in Japan using archived samples from the early 1980s and mid-1990s, Environ Res; 2005, 99, 31–39.

Exposure to endocrine-disrupting chemicals in food

255

kurahashi n, iwasaki m, sasazuki s, otani t, inoue m, tsugane s, Soy product and isoflavone consumption un relation to prostate cancer in Japanese men, Cancer Epidemiol Biomarkers Prev; 2007, 16(3), 538–545. kyle jam, duthie gg, Flavonoids in foods, 2006, in Flavonoids: Chemistry, biochemistry and applications, Boca Raton, CRC Press. le bail jc, varnat f, nicolas jc, habrioux g, Estrogenic and antiprolferative activities on MCF-7 human breast cancer cells by flavonoids, Cancer Lett; 1998, 130, 209–216. leblanc jc, tard a, volatier jl, verger p, Estimated dietary exposure to principal food mycotoxins from the first French total diet study, Food Addit Contam; 2005, 22(7), 652–672. lee s, wen w, xiang y, barnes s, liu s, cai d, zheng w, shu xo, Assessment of dietary isoflavone intake among middle-aged Chinese, J Nutr; 2007, 137, 1011–1016. liem akd, fürst p, rappe c, Exposure of populations to dioxins and related compounds, Food Addit Contam; 2000, 17, 241–259. lodovici m, dolara p, casalini s, ciappellano s, testolin g, Polycyclic aromatic hydrocarbon contamination in the Italian diet. Food Addit Contam; 1995, 12, 703–713. loutfy n, fuerhacker m, tundo p, raccanelli s, el dien ag, tawfic ahmed m, Dietary intake of dioxins and dioxin-like PCBs, due to the consumption of dairy products, fish/seafood and meat from Ismailia, Egypt, Sci Tot Environ; 2006, 370, 1–8. maff uk, Phthalates in food, Food surveillance information sheet No.82, 1996, available from http://archive.food.gov.uk/maff/archive/food/infsheet/1996/no82/82phthal.htm maff uk, Phthalates in infant formulae-follow-up survey, Food surveillance information sheet No.168 1998, available from http://archive.food.gov.uk/maff/archive/ food/infsheet/1998/no168/168phtha.htm mato y, suzuki n, katatani n, kadokami k, nakano t, nakayama s, sekii h, komoto s, miyake s, morita m, Human intake of PCDDs, PCDFs, and dioxin like PCBs in Japan, 2001 and 2002, Chemosphere; 2007, 67, S247–S255. mayr u, butsch a, schneider s, Validation of two in vitro test systems for estrogenic activities with zearalenone, phytoestrogens and cereal extracts, Toxicology; 1992, 74, 135–149. mccann se, muti p, vito d, edge sb, trevisan m, freudenheim jl, Dietary lignan intakes and risk of pre- and postmenopausal breast cancer, Int J Cancer; 2004, 111, 440–443. mccarthy ar, thomson bm, shaw ic, abell ad. (2005), Estrogenicity of pyrethroid insecticide metabolites. J Environ Mon, 8, 197–202. menzie ca, potocki bb, santodonato j, Exposure to carcinogenic PAHs in the environment, Environ Sci Tech; 1992, 26, 1278–1284. minh nh, someya m, minh tb, kunisue t, iwata h, watanabe m, tanabe s, viet ph, tuyen bc, Persistent organochlorine residues in human breast milk from Hanoi and Hochiminh city, Vietnam: contamination, accumulation kinetics and risk assessment for infants, Environ Pollut; 2004, 129, 431–441. moon h-b, ok g, Dietary intake of PCDDs, PCDFs and dioxin-like PCBs, due to the consumption of various marine organisms from Korea, Chemosphere; 2006, 62, 1142–1152. müller ak, nielson e, ladefoged o, Human exposure to selected phthalates in Denmark, 2003, Søborg, Denmark, The Danish Veterinary and Food Administration. Available from http://www.foedevarestyrelsen.dk/Fdir/Publications/2003015/ Rapport.pdf mulligan aa, welch aa, mctaggaert aa, bhaniani a, bingham sa, Intakes and sources of soya foods and isoflavones in a UK population cohort study (EPICNorfolk), Eur J Clin Nutr; 2007, 61, 248–254.

256

Endocrine-disrupting chemicals in food

mursu j, nurmi t, tuomainen t, salonen jt, voutilainen s, The intake of flavonoids and carotid atherosclerosis: the Kuopio ischaemic heart disease risk factor study, Br J Nutr; 2007, 28(4), 814–818. nap (national academy press), Hormonally active agents in the environment, 1999, available from http://www.nap.edu/openbook/0309064198/html nfrd (national food residue database of ireland), Pyrethroid anthelmintics in beef and sheep fat, 1999. 1999, available from http://nfrd.teagasc.ie/study_main. asp?caller=9&StudyID=63 nffsa, Registered Veterinary medicines and plant compounds, 2004, New Zealand Food Safety Authority, available from http://www.nzfsa.govt.nz/acvm/registers_ lists payne j, rajapakse, wilkins m, kortenkamp a, Prediction and assessment of the effects of mixtures of four xenoestrogens, Environ Health Perspect; 2000, 108, 983–987. rajapakse n, silva e, kortenkamp a, Combining xenoestrogens at levels below individual no-observed-effect-concentrations dramatically enhances steroid hormone action, Environ Health Perspect; 2002, 110(9), 917–920. rawn dfk, cao x-l, doucet j, davies dj, sun w-f, dabeka rw, newsome wh, Canadian Total Diet Study in 1998: pesticide levels in foods from Whitehorse, Yukon, Canada, and corresponding dietary intake estimates, Food Addit Contam; 2004, 21, 232–250. ritchie mr, cummings jh, morton ms, michael sc, bolton-smith c, riches ac, A newly constructed and validated isoflavone database for the assessment of total genistein and daidzein intake, Br J Nutr; 2006, 95, 204–213. rossi m, negri e, talamini r, bosetti c, parpinel m, gnagnarella p, franceschi s, dal maso l, montella m, giacosa a, la vecchia c, Flavonoids and colorectal cancer in Italy, Cancer Epidemiol Biomarkers Prev; 2006, 15(8), 1555–1558. rudel r, Predicting health effects of exposure to compounds with estrogenic activity: methodological issues, Environ Health Perspect; 1997, 105(3), 655–663. safe sh, Environmental and dietary estrogens and human health: is there a problem? Environ Health Perspect; 1995, 103, 346–351. sasamoto t, ushio f, kikutani n, saitoh y, yamaki y, hashimoto t, horii s, nakagawa j, ibe a, Estimation of 1999–2004 dietary daily intake of PCDDs, PCDFs and dioxin-like PCBs by a total diet study in metropolitan Tokyo, Japan, Chemosphere; 2006, 64, 634–641. scf (scientific committee on food), Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food, Update based on new scientific information available since the adoption of the SCF opinion of 22nd November 2000, 2001, Brussels, European Commission, http://ec.europa.eu/food/ fs/sc/scf/out90_en.pdf scoop (european commission scientific co-operation), Collection of occurrence data of Fusarium toxins in food and assessment of dietary intake by the population of EU Member States. Reports on tasks for scientific co-operation. Report of experts participating in Task 3.2.10, 2003, available at http://europa.eu.int/comm/food/fs/ scoop/task3210.pdf shaw ic, mccully s, The potential impact of dietary endocrine disrupters on the consumer, Int J Food Sci Technol; 2002, 36, 1–6. silva e, rajapakse n, kortenkamp a, Something from ‘nothing’-eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects, Environ Sci Technol; 2002, 36, 1751–1756. sinha r, kulldorff m, unter mj, strickland p, rothman n, Dietary benzo(a)pyrene intake and risk of colorectal adenoma, Cancer Epidemiol Biomarkers Pre; 2005, 14(8), 2030–2034. soto am, sonnenschein c, chung kl, fernandez mf, olea n, olea-serrano f, The E-SCREEN assay as a tool to identify estrogens: an update on estrogenic

Exposure to endocrine-disrupting chemicals in food

257

environmental pollutants. Environ Health Perspect; 1995, 103(Supplement 7), 113–122. steinmetz r, brown ng, allen dl, bigsby rm, ben-jonathan n, The environmental estrogen bisphenol A stimulates prolactin release in vitro and in vivo, Endocrinology; 1997, 138(5), 1780–1786. sudaryanto a, kunisue t, kajiwara n, iwata h, adibroto ta, hartono p, tanabe s, Specific accumulation of organochlorines in human breast milk from Indonesia: levels, distribution, accumulation kinetics and infant health risk, Environ Pollut; 2006, 139, 107–117. surh j, kim m-j, koh e, kim y-kl, kwon h, Estimated intakes of isoflavones and coumestrol in Korean population, Int J Food Sci Nutr; 2006, 57(5/6), 325–344. szyrwin´ska k, lulek j, Exposure to specific polychlorinated biphenyls and some chlorinated pesticides via breast milk in Poland, Chemosphere; 2007, 66, 1895–1903. theodoratou e, kyle j, cetnarskyj r, farrington s, tenesa a, barnetson r, porteous m, dunlop m, campbell h, Dietary flavonoids and the risk of colorectal cancer, Cancer Epidemiol Biomarkers Prev; 2007, 16(4), 684–693. thomson b, 2005, Human health implications of exposure to xenoestrogens from food, PhD thesis, Christchurch, New Zealand, University of Canterbury. Available at http://library.canterbury.ac.nz/etd/adt-NZCU20070712.141358 thomson bm, grounds pr, Bisphenol A in canned foods in New Zealand – an exposure assessment, Food Addit and Contam; 2005, 22(1), 65–72. thomson bm, lake rj, Cancer risk of selected ingested carcinogens, ESR Client report FW94/1, 1994, Christchurch (New Zealand), ESR. thomson bm, cressey pt, shaw ic, Dietary exposure to xenoestrogens in New Zealand, J Environ Mon; 2003, 5, 229–235s. tsumura y, ishimitsu s, saito h, kobayashi y, tonogai y, Eleven phthalate esters and di(2-ethylhexyl)adipate in one-week duplicate diet samples obtained from hospitals and their estimated daily intake, Food Addit Contam; 2001, 15(5), 449–460. turci r, turconi g, comizzoli s, roggi c, minoia c, Assessment of dietary intake of polychlorinated biphenyls from a total diet study conducted in Pavia, Northern Italy, Food Addit Contam; 2006, 23(9), 919–938. uehara m, arai y, watanabe s, adlercreutz h, Comparison of plasma and urinary phytoestrogens in Japanese and Finnish women by time-resolved fluoroimmunoassay, Biofactors; 2000, 12(1–4), 217–225. usda (united states department of agriculture), Iowa State University database on the isoflavone content of food, release 1.3, 2002, 2002, Beltsville (Maryland), USDA. usda (united states department of agriculture), USDA database for the flavonoid content of selected foods, 2003, Beltsville (Maryland), USDA. valsta lm, kilkkinen a, mazur w, nurmi t, lampi a, ovaskainen m, korhonen t, adlercreutz h, pietinen p, Phyto-oestrogen database of foods and average intake in Finland, Br J Nutr; 2003, 89(1), S31–S38. van den berg m, birnbaum l, bosveld atc, brunström b, cook p, feeley m, giesy jp, hanberg a, hasegawa r, kennedy sw, kubiak t, larsen jc, van leeuwen fxr, liem akd, nolt c, peterson re, poellinger l, safe s, schrenk d, tillitt d, tysklind m, younes m, waern f, zacharewski t, Toxic equivalency factors (TEFs) for PCBs, PCDDs and PCDFs for humans and wildlife, Environ Health Perpect; 1998, 106, 775–792. van erp-baart m-aj, brants ham, kiely m, mulligan a, turrini a, sermoneta c, kilkkinen a, valsta lm, Isoflavone intake in four different European countries: the VENUS approach, Br J Nutr; 2003, 89(1), S25–S30.

258

Endocrine-disrupting chemicals in food

vannoort r, thomson b, 2003/04 New Zealand Total Diet Survey, Agricultural compound residues, selected contaminants and nutrients, 2005, Wellington, New Zealand, New Zealand Food Safety Authority. vondrácˇek j, kozubik a, machala m, Modulation of estrogen receptor dependent reporter construct activation and Go/G1-S-phase transition by polycyclic aromatic hydrocarbons in human breast carcinoma MCF-7 cells, Toxicol Sci; 2002, 70, 193–201. vongbuddhapitak a, atisook k, thoophom g, sungwaranond b, lertreungdej y, suntudrob j, kaewklapanyachareon l, Dietary exposure of Thais to pesticides during 1989–1996, JAOAC Int; 2002, 85, 134–140. weijs pjm, bakker mi, korver kr, van goor ghanaviztchi k, van wijnen jh, Dioxin and dioxin-like PCB exposure of non-breastfed Dutch infants, Chemosphere; 2006, 64, 1521–1525. weinberg ds, manier ml, richardson md, haibach fg, Identification and quantification of isoflavonoid and triterpenoid compliance markers in a licorice-root extract powder, J Agric Food Chem; 1993, 41, 42–47. welshons wv, murphy cs, koch r, calaf g, jordan vc, Stimulation of breast cancer cells in vitro by the environmental estrogen enterolactone and the phytoestrogen equol, Breast Cancer Res Treatment; 1987, 10, 169–175. who, Pesticide residues in food DDT, Joint meeting of the Food and Agricultural Organization of the United Nations and World Health Organization, 1984, Geneva, Agricultural Organization of the United Nations and World Health Organization. who, Guidelines for predicting dietary intake of pesticide residues (revised), 1997, WHO/FSF/FOS/97.7, Geneva, Programme of Food Safety and Food Aid World Health Organization. wilson nk, chuang jc, morgan mk, lordo ra, sheldon ls, An observational study of the potential exposures of preschool children to pentochlorophenol, bisphenolA, and nonylphenol at home and daycare, Environ Res; 2007, 103, 9–20. yang m, park ms, lee hs, Endocrine-disrupting chemicals: human exposure and health risks, J Environ Sci Health Part C; 2006, 24, 183–224. zacharewski tr, meek md, clemons jh, wu zf, fielden mr, matthews jb, Examination of the in vitro and in vivo estrogenic activities of eight commercial phthalate esters, Toxicol Sci; 1998, 46, 282–293.

10 Bioassays for the detection of hormonal activities T. F. H. Bovee and L. A. P. Hoogenboom, RIKILT-Institute of Food Safety, The Netherlands; B.M. Thomson, Environmental Science & Research Ltd, New Zealand

Abstract: This chapter describes and discusses the major available bioassays for the detection of potential endocrine-disrupting activities, ranging from in vivo bioassays to in vitro cell proliferation assays, receptor binding assays, and transcription activation assays with mammalian and yeast cells. The overview focuses on (anti)estrogenic, (anti)androgenic and dioxin-like activities. In addition, we discuss the role of metabolism and use of specific cell lines as bioassays to determine indirect effects of compounds on endogenous hormone levels. Various assays are available for detection of androgens and estrogens, each with its own specific advantages and disadvantages. Their combined use may offer valuable insight in the mechanism of action. For detecting Ah-receptor agonists, such as dioxins and dioxin-like PCBs, the CALUX bioassays are currently the most suitable tests. At the same time, this assay demonstrates the importance of potential cross-talk between compounds, in particular that between corticosteroids and AhR agonists. Overall, it is clear that knowledge about the advantages and limitations of the assay system used is crucial for understanding the mechanism of action of different compounds. Key words: AhR-agonists, androgens, cross-talk, estrogens, food safety, mammalian cells, metabolism, yeast cells.

10.1

Introduction

Food for human or animal consumption may contain residues of different xenobiotics, such as environmental contaminants, pesticides, substances migrating from packaging materials, processing contaminants, veterinary drugs and growth-promoting agents, but also natural toxins, like myco- and phycotoxins. For many of these compounds maximum residue limits (MRLs) have been established to ensure good agricultural practices and monitoring programmes are required to check food items for the presence of residues

260

Endocrine-disrupting chemicals in food

that may endanger the consumer’s health. In the case of potent bioactive compounds with low tolerable intakes, or mixtures of compounds, such as polychlorinated aromatic hydrocarbons, the analytical methods are often laborious and expensive. As a result, monitoring is usually performed on only a limited number of samples. Furthermore, analytical methods are most often dedicated to specific compounds and will not detect novel hazards. Historically, humankind has relied on bioassays to determine the safety of food and the environment. In classic and medieval times, food tasters were employed to ensure that food was free of poisons and miners used canaries to detect the possible presence of toxic gases in mining tunnels. With increasing knowledge about the responsible toxicants and technical improvements in analytical chemistry, monitoring now relies heavily on chemical methods aimed at the detection of individual compounds utilising their physical and chemical properties. Nowadays, the use of in vivo animal bioassays is restricted to hazard characterisation of new compounds and testing for the presence of a number of specific substances. Bioassays with mice and rats are still the prescribed way to detect paralytic (PSP) and diarrheic shellfish poisons (DSP) in shellfish (Hungerford, 1995; LeDoux and Hall, 2000), and the neurotoxins produced by Clostridium botulinum (Galey et al., 2000). Fish assays are still widely used for testing the quality of drinking water and in the second Gulf War the US Army used small animals to detect toxins in the field. Current developments in the field of shellfish poisons are, however, directed towards abolishing the animal assays by development of analytical multimethods that can detect all the different toxins. However, since part of the toxins is still not identified and standards for others are not available, also in vitro methods are being developed based on specific biological properties of these compounds. Despite the rapid improvements in analytical chemistry we start to realise that these methods may not be sufficient to deal with the often very complex mixtures of chemicals or ever-changing chemical structures of toxicants present in the environment or as residues in our food chain. Changes in chemical structures or the occurrence of toxicants in ecosystems might even be accelerated by changes of the climate (Tester, 1994; Peperzak, 2005). Moreover, in the area of antibiotics and growth promoters, there is a tendency to switch to novel unknown compounds to circumvent control on the illegal application of such substances in food-producing animals. To screen for the presence of antibiotics and antibacterial drugs in food, bioassays based on bacteria are widely used (Nouws et al., 1988, 1998), and are so reliable that in many cases confirmation of the chemical identity of the responsible substances is no longer required. Cell-lines have been used to detect compounds with Ah receptor agonist activity and similarly for estrogenic and androgenic activity. Recent advances in cell biology and in

Bioassays for the detection of hormonal activities

261

particular biotechnology have allowed the development of a new generation of bioassays, based on the possibilities of introducing specific receptors and reporter gene constructs into cellular systems. The DR CALUX® assay, a bioassay used for the detection of dioxins and dioxin-like polychlorinated

Human population

Whole animal

Tissue culture

Cell based

Structure–activity R1

O

R2

Fig. 10.1 Spectrum of assays from human studies to structure–activity relationships (kindly drawn by Matt Walters).

262

Endocrine-disrupting chemicals in food

biphenyls (dl-PCBs), is one of the first examples of such an in vitro transcription activation assay system and was validated for food, feed and environmental samples (Murk et al., 1996; Bovee et al., 1998; Hoogenboom et al., 2004). This bioassay was extensively used in a number of incidents with dioxins in the food chain and is officially recognised by the EU (EC 2002a, b). Recently, the use of the assay resulted in the discovery of dioxins in fat derived from a gelatine production plant, due to the use of contaminated hydrochloric acid (Hoogenboom et al., 2007). This demonstrates the value of high-throughput bioassays as early warning systems. Moreover, bioassays are indispensable in order to elicit the presence of unknown and new emerging pollutants. In this chapter it is shown that the use of bioassays can overcome most of the analytical problems mentioned above and offer a number of additional advantages. Figure 10.1 shows the spectrum of bioassays ranging from studies of the human population, through in vivo animal studies, in vitro tissue culture and cell-based assays to structure–activity relationships based on the chemical features of the toxin of interest. While recognising the value of in vivo assays, this chapter will focus on in vitro bioassays for estrogenic and androgenic hormonal activities and Ah receptor agonists, which are all key effects of endocrine-disrupting chemicals.

10.2

Compounds with hormonal activity

Many environmental contaminants, pesticides, substances migrating from packaging materials, processing contaminants, veterinary drugs, growthpromoting agents and even dioxins and PCBs have been shown to possess hormonal or endocrine-disrupting activities. The 2002 incident in the Netherlands and Belgium involving medroxyprogesterone 17-acetate (MPA) demonstrated that even pharmaceutical waste may end up in the food chain (Van Leengoed et al., 2002). In that case a yeast estrogen bioassay indicated that the animal feed was not only contaminated with MPA, but also with 17β-estradiol, explaining the problems with the pregnant pigs in breeding farms (Bovee et al., 2006). Pain-killing drugs such as paracetamol and non-steroidal anti-inflammatory drugs (NSAIDs), antidepressants (SSRIs), birth control pills and tablets for hormone replacement therapy (HRT) are widely prescribed and it was waste from the latter that was responsible for the MPA incident. Birth control and HRT tablets contain large amounts of compounds that have strong hormonal activities. Cheap, fast and robust bioassays as early warning systems for hormonal or endocrine-disrupting activities are thus very valuable for cost effective monitoring programs. Moreover, bioassays are invaluable in the different stages of drug development. Specific anti-estrogens and anti-androgens are important to treat breast and prostate tumours via mediation by the human estrogen receptor

Bioassays for the detection of hormonal activities

263

α (hERα) or the human androgen receptor (hAR) respectively. Specific agonists for the hERα are interesting for contraception and to treat climacteric complaints in females, while specific agonists for the hAR might be used to treat libido loss, impotence or muscle weakness in males. Sections 10.3–10.5 deal primarily with bioassays that were designed to detect estrogen or androgen activity. In addition, it is shown that these bioassays are also capable of detecting anti-estrogenic and anti-androgenic activity. Prohormones can also be detected, depending on the metabolic activity of the cell type that is used, or in combination with an enzymatic activation system. However, the detection of prohormones is not within the scope of this chapter.

10.3

In vivo bioassays for estrogens and androgens

In vivo hormonal bioassays have the advantage of determining an overall biological effect by allowing for interactions between cells and between different components of the endocrine system and thence, in effect, the whole animal. In addition, they take account of absorption, disposition, metabolism and excretion of the test chemicals. Assays usually employ immature, ovariectomised adult female, or male animals to minimise contributions from endogenous estrogens and androgens. The earliest and best-known in vivo bioassay used to determine estrogenic potency is that published by Allen and Doisy (1923). A spayed female mouse is injected with the test material and the appearance of cornified cells on vaginal smears or the increase of the uterus weight constitutes a positive reaction. In this way Allen and Doisy discovered the existence and effect of 17β-estradiol. Within a period of 15 years the other hormones that influence the uterus were discovered and the interrelationship between them became clear. Allen also proved that estrogens cause the onset of puberty in immature animals and demonstrated that the hormonal mechanisms of primates, including humans, are similar to those of rodents. However, there are clear differences in hormone metabolism and the effects of hormones between different species (Coecke et al., 2006). Since then, a diverse array of in vivo assays of estrogenic potency have been reported in the literature, including the following: • Measures of cell proliferation in the rodent female genital tract. Examples include: measurement of uterine epithelial height and vaginal epithelial thickness in ovariectomised mice (Ulrich et al., 2000), and measurement of uterine weight in weanling mice (Shelby et al., 1996). • Measures of other estrogen-related effects on the female genital tract, including: increases in uterine vascular permeability (Milligan et al., 1998), and increases in glycogen production in immature rat uteri (Bitman and Cecil, 1970).

264

Endocrine-disrupting chemicals in food

• Measures of testicular changes or sperm quality in male rodents, including: reduction in testicular size, reduction in ventral prostate weight, and reduction in sperm production in male rats (Sharpe et al., 1995). • Induction of vitellogenin in fish. Induction of the protein, vitellogenin, in male and female fish after exposure to estrogenic compounds has been carried out for a number of fish species, including: rainbow trout (Sumpter, 1985), carp (Tyler and Sumpter, 1990) and winter flounder (Pereira et al., 1992). • Sex determination in turtles. The sex of offspring of the red-eared slider turtle, as for other reptiles, is determined by the temperature of incubation. The ability of xenoestrogens applied to the eggshell to overcome this effect was used as an assay of estrogenic activity (Bergeron et al., 1994). The Herschberger assay (Herschberger et al., 1953) is an in vivo test used to determine androgenic potency and is conducted in castrated adult male animals, most often rats. The identification of androgens and anti-androgens is based on changes in the weights of five androgen-responsive tissues: ventral prostate, paired seminal vesicles and coagulating glands, the levator ani and bulbocavernosus muscles, the penis gland, and paired Cowper’s or bulbo-urethral glands. The Organization for Economic Co-operation and Development (OECD) recently completed phase 2 of an international programme to validate the rodent Herschberger bioassay; the next validation phase will employ coded substances (Owens et al., 2007). The advantages of these in vivo bioassays are that they determine the overall biological effect of a compound. The protocols of the Allen–Doisy and Herschberger assays turned out to be reproducible and transferable across laboratories. Although in vivo bioassays are considered as the gold standard, these tests are labour intensive, expensive, have a relatively poor sensitivity, give a modest response, and require the use of animals. Therefore they are not suited to large-scale screening of chemicals and samples.

10.4

In vitro bioassays for estrogens and androgens

There is a whole range of different in vitro bioassays to determine the estrogenic and androgenic activities of compounds. As illustrated in Fig. 10.2 (see also chapter 9 Fig. 9.2), these vary from receptor binding assays to receptor-dependent gene expression assays and cell proliferation assays, corresponding to different steps in the estrogen or androgen receptordependent pathway. Their major features, advantages and disadvantages are discussed below. Theoretically, these in vitro bioassays have the drawback that they cannot determine the overall biological effect of a compound, but only allow testing of compounds showing their hormone activity exclusively via the receptor. However, the main mechanism of action of all well-known

Bioassays for the detection of hormonal activities

265

Cell membrane Nucleus

Transcription

Protein synthesis

Cell division

Fig. 10.2 A simplified schematic of the classical mechanism of an estrogen responsive effect showing the locations of commonly used in vitro assays (from Thomson et al., 2003, and kindly drawn by Matt Walters).

hormones involves the binding to their cognate receptor, confirmed by many studies performed with knock-out animals.

10.4.1 Receptor binding assays Competitive ligand binding assays are rapid and easy to perform, but they determine only the strength of the binding of a substance to the receptor and not activation of the receptor. Thus, they cannot distinguish between

266

Endocrine-disrupting chemicals in food

receptor agonists and antagonists. Moreover, competitive binding assays using cytosol preparations suffer from cross-talk of other nuclear receptors present in the homogenate. Only binding assays using recombinant receptor protein do not suffer from this possible artefact (Waller et al., 1996). In addition, competitive binding assays are mostly performed with radioactive ligands and there is still a lot of work to optimise fluorescence-based methods to replace the use of radiolabels.

10.4.2 Transcription activation assays Alternatively, receptor-based transcription activation assays can be used to detect compounds having affinity for a given receptor (Garcia-Reyero et al., 2001; Mueller, 2002). In contrast to receptor binding assays, receptor–reporter gene bioassays also include the transactivation step and can distinguish between receptor agonists and antagonists. A large number of reporter–receptor gene assays have been developed, using both yeast and mammalian cells. In vitro transcription activation assays for estrogens Receptor–reporter gene transcription activation assays based on yeast cells were developed in order to improve the speed, ease and high-throughput character and also to eliminate the cross-talk from other receptor types that could lead to false negative screening results in the E-screen, a cell proliferation assay based on human breast cancer cells (see Section 10.4.3). Until recently, most yeast estrogen bioassays were based on an extrachromosomal reporter construct with β-galactosidase as a substrate-based reporter protein (see Table 10.1). The most well-known assays are the yeast estrogen bioassay developed by Pham et al. (1992) and the yeast estrogen screen (YES) developed by Routledge and Sumpter (1996). There is only a small difference between these two assays. In the YES the hERα is expressed continuously while in the assay of Pham et al., 0.05 mm of CuSO4 has to be added to induce the expression of hERα by the CUP1 metallothionein promoter. Several other yeast assays were subsequently developed, in most cases using β-galactosidase as the reporter protein. A major disadvantage of this reporter, but also an alternative protein, luciferase, is the fact that the enzyme needs to be released from the cells following exposure, which in the case of yeast cells, is not very easy and may introduce large variation in the results (Bovee et al., 2004a). The development of a yeast estrogen bioassay based on the increased expression of yeast-enhanced green fluorescent protein (yEGFP) should therefore be regarded as an important improvement, since this protein can be measured in intact cells. Furthermore, both the hERα gene receptor construct and the estrogen responsive element driven reporter construct (ERE-yEGFP) were stably introduced in the yeast genome (Bovee et al., 2004a). A similar yeast estrogen bioassay expressing hERβ, and a yeast estrogen bioassay expressing

Bioassays for the detection of hormonal activities

267

Table 10.1 In vitro bioassays for estrogenic activities Type/name

Cells

Modifications

Ref

Proliferation assay E-screen

MCF-7

None

Soto et al. (1992)

Mammalian reporter gene assays MVLN MCF-7 ERE-CAT ER CALUX T47D ERE-Luc None CHO RO-Luc Lumicell BG-1 ERE-Luc HeLa HELNα hERα, ERE-Luc HeLa HELNβ hERβ, ERE-Luc MMV-Luc MCF-7 MAR-Vit-Luc ER CALUX new U2-OS hERα, ERE-Luc

Pons et al. (1990) Legler et al. (1999) Schoonen et al. (2000) Rogers and Denison (2000) Paris et al. (2002) Paris et al. (2002) Willemsen et al. (2004) Sonneveld et al. (2005)

Yeast reporter gene assays None Yeast

Pham et al. (1992)

None None YES

Yeast Yeast Yeast

hERα (CUP1 MT), ERE-lacZ hERα, ERE-lacZ hERα, ERE-lacZ hERα, ERE-lacZ

REA (ERα) REA (ERβ) None None

Yeast Yeast Yeast Yeast

hERα, ERE-yEGFP hERβ, ERE-yEGFP hERβ, ERE-Luc hERα, ERE-Luc

Petit et al. (1995) Arnold et al. (1996) Routledge and Sumpter (1996) Bovee et al. (2004a) Bovee et al. (2004b) Leskinen et al. (2005) Leskinen et al. (2005)

both hERα and hERβ were developed (Bovee et al., 2004b). Overall, yeast estrogen bioassays have proven to be highly valuable for determining the estrogenic potency of compounds and environmental samples (Gaido et al., 1997; Breithofer et al., 1998; Graumann et al., 1999; Rehmann et al., 1999; Le Guevel and Pakdel, 2001; Morito et al., 2001; Witters et al., 2001; Legler et al., 2002). In parallel with yeast assays, in vitro transcription activation bioassays based on mammalian cells were developed for studying hormonal mechanisms and testing compounds for their hormonal activities. One of the first assays was the MVLN assay developed by Pons et al. (1990) (Table 10.1). Subsequent assays similarly used breast carcinoma cells already expressing the estrogen receptor, thus requiring only the introduction of an EREreporter gene construct. Examples are the ER CALUX (Chemical Activated LUciferase gene eXpression) assay (Legler et al., 1999) and the MMV-Luc assay (Willemsen et al., 2004). Other investigators have used Chinese hamster (CHO) and human ovarian cell-lines (BG-1) (Rogers and Denison, 2000; Schoonen et al., 2000). As described below, an important disadvantage of mammalian cell-line-based assays is the presence of additional receptors that can interfere with the specific response of the cells. To

268

Endocrine-disrupting chemicals in food

2000

1700

Fluorescence

1400

1100

800

500

200 1

10

100

1000

10 000

100 000 1 000 000 1e+07

1e+08

Concentration [pM] E2

E2-benz

z-lenone

genistein

E1

DES

EE2

estriol

Fig. 10.3 Dose-related response for different estrogenic compounds in yeast cells expressing the hERα-receptor and yEGFP in response to the binding (Bovee et al., 2004b).

overcome this problem, a new ER CALUX assay was developed, based on an osteosarcoma (U2-OS) cell-line that apparently does not express any steroid receptors (Sonneveld et al., 2005). Similar to the yeast assays, both the reporter construct and the receptor were introduced into these cells. An important issue is the validation of assays based on their ability to detect well-known agonists but also antagonists. Figure 10.3 shows the response of a number of estrogenic compounds in the yeast estrogen bioassay expressing yEGFP as a reporter protein, displaying the typical S-shape curves also observed in other assays. In general, both mammalian assays and yeast assays are able to detect known estrogens, but there are some differences with compounds that have anti-estrogenic properties. Specific properties of yeast and mammalian cell-based transcription activation assays for estrogens Yeast screens have failed to show the antagonistic activities of tamoxifen and 4-hydroxytamoxifen, with both compounds behaving as estrogen receptor agonists in yeast-based assays (Legler et al., 2002). Since it was long assumed that tamoxifen and 4-hydroxytamoxifen were classical and pure anti-estrogens, the outcome of the yeast assays were regarded as an artefact (NIH, 2003). Moreover, some compounds, for example, o,p′-DDT, fenarimol,

Bioassays for the detection of hormonal activities

269

p,p′-DDE, and benzo(a)pyrene (B[a]P) were not active in yeast, whereas they demonstrated partial agonist activity in different mammalian cell reporter gene systems (Charles et al., 2000). B[a]P for instance demonstrated partial agonist activity in transiently transfected human MCF-7 cells (Fang et al., 2000). Possible explanations for these observations are that certain compounds are not able to pass the yeast cell wall or are pumped out of the yeast cells by transporters before binding to the receptor (Egner et al., 1998; Tutulan-Cunita et al., 2005). However, the differences observed with tamoxifen and B[a]P can be explained otherwise and will be discussed in detail. Because of the apparent lack of antagonist activity for tamoxifen in the yeast assay, transcription activation assays based on mammalian cell-lines were considered to be more specific and more relevant to humans. In most of the breast cancer cell-lines, tamoxifen and 4-hydroxytamoxifen showed the presumed and expected antagonist activity. Moreover, transcription activation assays based on mammalian, or more particularly human, celllines were shown to be more sensitive than yeast-based assays, and might also be able to identify compounds that require human metabolism for activation into their active state (NIH, 2003). However, by now it is clear that both tamoxifen and 4-hydroxytamoxifen are not pure antagonists, but selective estrogen receptor modulators (SERMs) showing antagonistic effects in breast tissue, but agonist activity on the endometrium and bone. After 5 years of treatment, most breast carcinomas become resistant to tamoxifen and prolonged use of tamoxifen is associated with an increased risk of endometrial cancer (Juretic et al., 2006; Leblanc et al., 2007). This is in line with the fact that tamoxifen shows weak estrogenic properties in the mouse uterotrophic assay, still the gold standard for testing estrogenicity (Sonneveld et al., 2006a). The outcome in the yeastbased assays, in which tamoxifen acts as an agonist, can thus not be regarded as an artefact. Moreover, the pure anti-estrogens ICI 182780 and RU 58668 are more potent than tamoxifen in inhibiting the growth of breast cancer cells and are also anti-estrogenic on the uterus. Neither compound is associated with an increased risk of endometrial cancer (Dauvois et al., 1992; Elkak and Mokbel, 2001; Maillard et al., 2006). When tested in the yeast estrogen bioassay, ICI 182780 and RU 58668 did not give an estrogenic response and when these compounds were tested by co-administration of 17β-estradiol they inhibited the signal caused by 17β-estradiol (Fig. 10.4). However, in general, compared with their mammalian counterparts, yeast estrogen bioassays show relative modest responses upon exposure to these antiestrogens (Bovee et al., 2007b). Variability of effects observed for tamoxifen have been extensively investigated and studies showed that even different breast cell-lines gave different responses, agonistic or antagonistic, with tamoxifen and other related compounds (Watanabe et al., 1997; McInerney et al., 1998; Yoon

270

Endocrine-disrupting chemicals in food

Fluorescence (hERα)

700

500

17β-E2 ICI 182780 RU 58668

300

1.5 nM E2 + ICI 182780 1.5 nM E2 + RU 58668 100

−100 0.01

0.1

1

10

100

1000 10 000 100 000

Concentration [nM]

Fig. 10.4 Response of the pure anti-estrogens ICI 182780 and RU58668 in the absence or presence of 1.5 nm 17β-estradiol in yeast expressing hERα (Bovee et al., 2007b). Included is a dose–response curve for estradiol. The anti-estrogens show no response by themselves but inhibit the response obtained with estradiol.

et al., 2001; Jordan, 2003). The fact that some compounds can act both as estrogens and anti-estrogens is rather confusing when dealing with screening assays for testing on this type of properties. Moreover, breast cells often not only express hERα, but also hERβ. This co-expression of hERβ has a great influence on how cells respond to estrogens. In general, the hERβ decreases the response of hERα and seems to act as a kind of regulator (repressor) of hERα (Gustafsson, 1999; Bovee et al., 2004b). Of course this may have physiological importance for some tissues, but it complicates the interpretation of the results, especially if researchers are not aware that the ERβ might be expressed in their cell-line. Differences between data obtained with endometrial, breast and yeast cells may be explained by different mechanisms that underlie endocrine responses (Gibson et al., 1991; Jordan, 2003), different organ and species sensitivities (e.g. assembling and presence of both co-activators and corepressors, amounts of receptors), limitations of the assay techniques (e.g. selective uptake or solubility of the test compounds), or differences in metabolism. In addition, some mammalian cell-lines expressing endogenous receptors might also contain receptor isoforms or even mutated receptor types that differ from those in normal tissues. The drawbacks with mammalian and human breast cancer cell-lines in particular, are now being recognised and were the major reason why the improved ERα CALUX test was based on a human bone osteosarcoma

Bioassays for the detection of hormonal activities

271

epithelial cell (Sonneveld et al., 2005, 2006a). This U2-OS cell-line is thought to express no, or only very low levels of the endogenous estrogen, androgen, progesterone and glucocorticoid receptor, although other investigators have reported findings that indicated the presence of androgen and progesterone receptors (Orwoll et al., 1991). According to Petz et al. (2004), the endogenous progesterone receptor was only expressed in U2-OS when cells were transfected with an ERα expression vector. Thus, compared with other mammalian cell-lines, the possible artefacts due to cross-talk from other receptor types are reduced to a minimum, but not fully eliminated, in this U2-OS cell-line. Based on the response with various hormones, this new ERα CALUX reporter gene assay is both sensitive and specific and the correlations found between the data of this ERα CALUX, another reporter gene assay based on a CHO cell-line, an in vitro receptor binding assay, and the Allen–Doisy test were good (Sonneveld et al., 2006a). Only the relative potencies of estrone and estriol were rather low compared with the other test systems. In vitro transcription activation bioassays for androgens Of all endocrine disruptors, environmental contaminants with estrogenic potency are the most studied (Jobling et al., 1995). However, recent studies show a crucial involvement of the androgen receptor in abnormal sex development. The presence of pollutants with adverse effects on human androgen receptor has been reported from paper-mill effluents and as a result of intensive farming (Jenkins et al., 2003; Lemaire et al., 2004). Disorders that have been related to androgenic or anti-androgenic exposure include testicular cancer, hypospadias, cryptorchidism and poor sperm quality (Pike et al., 1993; Skakkebæk et al., 2006). Very recently prepubertal gynecomastia was linked to both estrogenic and anti-androgenic effects of lavender and tea tree oil (Henley et al., 2007). In comparison with estrogens, only a few receptor–reporter gene bioassays for testing androgenicity have been developed (Table 10.2). However, especially in the case of androgens, the lack of known endogenous receptors in yeast is a big advantage compared with mammalian cell-lines, as androgen-responsive elements (AREs) can also be activated via the progesterone and glucocorticoid receptor (PR and GR). The glucocorticoid receptor is normally expressed in all mammalian cell types. To avoid the potential cross-talk in mammalian cell-lines, many efforts were made to construct an ARE that is specific and not inducible by the progesterone or glucocorticoid receptor (Haelens et al., 2003; Shaffer et al., 2004; Brodie and McEwan, 2005). However, up to now such an ARE does not seem to exist. So far, this has resulted in cell-lines that are not specific for androgens but also respond to progesterone or glucocorticoids, like the CHO cell-line and the T47-D human breast carcinoma cell-line. The high interference of natural glucocorticosteroids and dexamethasone with the glucocorticoid receptor, and of progestagens with the progesterone receptor can be

272

Endocrine-disrupting chemicals in food

Table 10.2

In vitro bioassays for androgenic activities

Type/name

Cells

Modifications

Ref

Proliferation assay A-screen

MCF-7

hAR

Szelei et al. (1997)

hAR, MMTV-Luc MMTV-Luc hAR, MMTV-Luc hAR, ARE-LUC

Schoonen et al. (2000) Willemsen et al. (2002) Willemsen et al. (2004) Sonneveld et al. (2006a)

hAR (CUP1 MT), ARE-lacZ hAR, ARE-Luc hAR, ARE-yEGFP

Gaido et al. (1997)

Mammalian reporter None TM-Luc TARM-Luc AR-CALUX new

gene assays CHO T47D T47D U2-OS

Yeast reporter gene assays Yeast androgen Yeast bioassay none Yeast RAA Yeast

Michelini et al. (2005) Bovee et al. (2007a)

suppressed by the addition of onapristone, a non-dimerising PR and GR antagonist, making the androgen test based on the CHO-AR cell-line more specific for androgens (Schoonen et al., 1999). The AR-LUX in the T47-D human breast carcinoma cell-line responds to androgens, but also to progesterone and the synthetic glucocorticoid dexamethasone. Moreover, this cell-line has a relatively high EC50 value of 115 nm for 5α-dihydrotestosterone (5α-DHT) and shows no response to 17β-testosterone (Blankvoort et al., 2001). The TM-Luc T47-D human breast cell-line constructed by Willemsen was transfected with the MMTV-Luc reporter plasmid, known to contain ARE sequences but also glucocorticoid and progesterone response element (PRE/GRE) sequences. It showed maximum responses with 100 nm 17β-trenbolone, an androgenic compound, but also 100 nm progesterone. Dexamethasone did not elicit a significant response, but was only tested at 1 nm (Willemsen et al., 2005). The latest AR CALUX test, like the new ERα CALUX, is also based on the U2-OS cell-line that expresses no or only very low levels of the endogenous estrogen, androgen and glucocorticoid receptor. This AR CALUX cell-line has been shown to be specific and sensitive, the EC50 for 17β-testosterone being about 1 nm. Dexamethasone gives a response, but the maximum response obtained with this compound is only 8% of that obtained with androgens. A relative androgenic potency (RAP) of 0.000 was reported for 17β-estradiol, but this compound did show a full dose– response in the μm range. In addition, the gestagens progesterone and MPA were also able to give a response. Correlations found between the activities of a number of compounds of this AR CALUX, another reporter gene assay based on a CHO cell-line, an in vitro receptor binding assay, and the in vivo Herschberger test were good (Sonneveld et al., 2006a).

Bioassays for the detection of hormonal activities

273

1100

Fluorescence

900 700 500 300 100 −100 0.1

1

10

100

1000

10 000

100 000 1 000 000

Concentration [nM] 17β-T

DHT

Prog

Dex

17β-E2

Bold

Fig. 10.5 Dose–response for the androgens 17β-testosterone, dihydrotestosterone (DHT) and boldenone, the estrogen 17β-estradiol, the gestagen progesterone and the glucocorticoid dexamethasone in yeast cells expressing the human androgen receptor (Bovee et al., 2008).

Similar results were found with a new yeast androgen bioassay based on the expression of yEGFP (Fig. 10.5). Dexamethasone did not show a response, but 17β-estradiol gave a full dose–response and also progesterone and MPA gave a clear response in this yeast androgen bioassay (Bovee et al., 2007a). However, responses obtained with both the yeast androgen and the AR CALUX bioassay are in line with the known androgenic effects of 17β-estradiol, progesterone and MPA. This new yeast androgen bioassay showed relative androgenic potencies (RAPs) of compounds that were in line with (quantitative) structure–activity relationship ((Q)SAR) findings, confirming the importance of the 17β-OH group, the 5α-steroidal framework and the 3-keto group. This yeast androgen assay is highly specific, simple and reliable and appears to be one of the best tools to check and optimise QSAR models for predicting activities of different isomers and designer steroids (Bovee et al., 2007b). In addition, this assay is sensitive to the anti-androgenic properties of flutamide and some brominated flame retardants (BFRs), as shown in Fig. 10.6 by the ability to inhibit the response obtained with 5α-DHT, the most potent endogenous androgen (Bovee et al., 2007a; Canton et al., 2007). These results demonstrate that the yeast androgen biosensor is suitable for detecting compounds with both agonistic and antagonistic characteristics.

274

Endocrine-disrupting chemicals in food 140

% of DHT-induced response

120 100 80 60 40 20 0 1

10

100

1000

10 000

100 000

1 000 000

Concentration [nM] DHT+Flut

DHT+TBP

DHT+4OH-BDE17

DHT+BDE39

Fig. 10.6 Effect of the anti-androgen flutamide, tetrabromobisphenol, 4-hydroxyBDE 47 and BDE 39 on the response obtained with 5α-DHT in yeast-cells expressing the human androgen receptor (Bovee et al., 2007a).

Using this assay, the androgenic potency of the new designer steroid tetrahydrogestrinone (THG) could easily be demonstrated and detected in spiked human urine samples (Nielen et al., 2006). Prohormones with an androgenic mode of action, e.g. dehydroepiandrosterone (DHEA), are not active in this yeast androgen bioassay and require metabolic activation before they can be detected (Rijk et al., 2006). Compared with other yeast androgen bioassays, this new androgen bioassay showed a similar limit of detection and dynamic range. The measurement of the fluorescence (yEGFP), however, can be followed as a function of incubation time and is thus easier, quicker and cheaper than the measurement of the β-galactosidase or luciferase activity, which requires cell wall disruption and/or the addition of expensive substrates (Gaido et al., 1997; Lee et al., 2003; Michelini et al., 2005).

10.4.3 Cell proliferation assays Cell proliferation is a process further down the mechanistic pathway than transcription (see Fig. 10.2). The E-screen was one of the first in vitro bioassays used to determine the estrogenic activity of compounds and extracts of food samples (Soto et al., 1995). This cell proliferation assay uses ERpositive, estrogen responsive MCF-7 breast cancer cells that show increased

Bioassays for the detection of hormonal activities

275

proliferation upon exposure to ER agonists (Soto et al., 1995, 1998, 2006). The use of this assay resulted in the coincidental discovery of the estrogenic properties of nonylphenols (Soto et al., 1991). The test is also capable of detecting anti-estrogenic activity by incubating test compounds in the presence of 17β-estradiol. The assay is per definition relevant for the higher vertebrates as it measures an end point of estrogenicity in human breast cancer cells. However, most cell-lines, including MCF-7, also express androgen, progesterone, glucocorticoid and retinoid receptors. This may compromise the suitability of the assay if extracts also contain substances that are able to bind to other receptors, as it has been shown that androgens, progestins and glucocorticoids can antagonise estradiol-induced cell proliferation. Moreover, compared with endometrial cells, breast cancer cells often respond very differently to estrogens and anti-estrogens and as mentioned above there are also great differences between different breast cancer celllines. Furthermore, proliferative responses can only be shown after a number of days, resulting in a test that is not very rapid (Ramamoorthy et al., 1997; Zacharewski, 1997). An older proliferation assay used to study androgen action was based on a human LNCa-FGC prostate cancer cell-line, but this cell-line turned out to have a point mutation in the androgen-binding domain, thus enhancing estrogen binding. In parallel to the E-screen, the group of Soto also developed an A-screen for detecting androgenic activity based on cell proliferation (Szelej et al., 1997; Körner et al., 2004; Soto et al., 2004, 2006). For this test, the MCF7-cells were modified and transfected with the hAR (MCF7-AR1 cells). In response to AR agonists, the cells produce APRIN, an endogenous inhibitor of cell proliferation, and as a result the cells stop growing. The mechanism behind this effect is poorly understood and rather complex. The assay is a kind of reversed E-screen that measures androgenic activity indirectly. Moreover, it is rather difficult to understand the relation between this test and the male tissues examined in the in vivo assays such as the Herschberger test. Corticosteroids also show an effect in this assay but at relatively high concentrations (Szelei et al., 1997). For detection of anti-androgenic activity, compounds are tested in the presence of an androgen. Elimination of the inhibition of proliferation points to anti-androgenic properties of the test compound. Just like the E-screen, the test has some drawbacks due to possible cross-talk and the duration of the assay.

10.4.4 Metabolism of test chemicals The importance of metabolism has been underestimated for a long time and there are many indications that metabolism is an important source of differences between different assay types, both in vitro and in vivo. This includes both activation of a compound into an active metabolite, but certainly also the degradation of hormonally active compounds, especially

276

Endocrine-disrupting chemicals in food

Metabolism (activation/deactivation) Uptake

Uptake

RE

Gene

Degradation

Protein

Fig. 10.7 Simplified scheme showing the mechanism leading to a cellular response after binding of a compound to a receptor, followed by its transport to the nucleus and binding to a response element (RE) in the DNA. Compounds may be metabolised by the cells resulting in activation or deactivation. Alternatively, the receptor/ligand complex may not be transported to the nucleus but degraded instead (antagonist).

in vivo. A schematic of the potential influence of metabolism of any endocrine-disrupting chemical is shown in Fig. 10.7. In principle it seems ideal if the metabolism is taken into account when testing compounds for their potential effects in humans. However, in practice this may not always be the case, especially when interested in the active component or performing QSAR studies. It is not necessarily an advantage that certain mammalian cells may be able to identify compounds that require metabolic activation. Some mammalian cells are able to convert estrone into 17β-estradiol and vice versa. This conversion is ascribed to 17β-hydroxysteroid dehydrogenase 3 (17β-HSD) and this conversion explains the high relative estrogenic potency (REP) of estrone in the ER CALUX test with T47-D breast cancer cells, i.e. a reported REP for estrone of 1.0 (Hoogenboom et al., 2001). In yeast, the reported REP of 0.2 for estrone (Bovee et al., 2004b) correlates much better with reported binding affinities and the REP in the Allen–Doisy test. Also the inactivity of B[a]P in the yeast estrogen bioassays can easily be explained by differences in metabolism between yeast and mammalian cells. It was shown that the 3hydroxy and 9-hydroxy metabolites of B[a]P, and not B[a]P itself, are responsible for the activation in transiently transfected MCF-7 cells (Charles et al., 2000; Fertuck et al., 2001). Moreover, B[a]P does not elicit uterotrophic effects in vivo, suggesting that neither B[a]P nor its hydroxyl metabolites could induce estrogenic effects in the mouse uterus (Fertuck et al., 2001), or that the hydroxy metabolites are simply not formed in sufficient quantities.

Bioassays for the detection of hormonal activities

277

Yeast generally lacks the metabolic capacity of mammalian cells. It has, for example, been demonstrated that the Saccharomyces cerevisiae strain used as a host in the yeast androgen bioassay, expressing yEGFP, does not posses any aromatase, 5α-reductase, 3α-HSD or 17β-HSD activity. Owing to this absence and the lack of cross-talk from endogenous hormone receptors, yeast-based assays are probably the best tool for optimising QSAR model approaches (Bovee et al., 2008).

10.4.5 Practical application of hormone bioassays Thus, both mammalian cell- and yeast-based bioassays are suited to detect both agonist and antagonist activity. On the other hand, cell-lines have been shown to be more sensitive than yeast-based assays, and to some extent may be able to identify compounds that require human metabolism for activation into their active state. However, the latter is not necessarily an advantage and yeast-based assays have several other advantages. These include low costs, easy handling, lack of known endogenous receptors that may compete with the receptor activity under investigation, and the use of media that are devoid of steroids. Yeast cells are also very robust and the yeast estrogen bioassay based on the expression of yEGFP is the first hormonal bioassay that was fully validated according to the international criteria that were put forward in EC Council Decision 2002/657. The test is accredited in the Netherlands (ISO 17025) for both calf urine and animal feed (Bovee et al., 2005, 2006). Klein et al. (1994) used a yeast estrogen bioassay to measure the very low levels of non-conjugated estrogens in blood of prepubertal children, which at that stage was not possible with chemical-analytical and immunological assays. The detection limit was <0.02 pg/ml estradiol equivalents, levels in boys varied between <0.02 and 0.7 pg/ml, levels in girls between <0.02 and 2.2 pg/ml. Recently, Courant et al. (2007) were able to confirm with gas chromatography/high-resolution mass spectrometry (GC/HRMS) that the estrogen levels in young children were indeed much lower than thus far assumed. This was based on the fact that in most samples, the 17βestradiol levels were below the LOQ of 2 pg/ml, which is still much higher than the detection limit of the bioassay. Overall it can be concluded that in order to investigate the hormonal properties of different compounds and their mechanism of action, there is no superior assay system. The only way to unravel the mechanism of action is to be aware of the limitations of the different assay techniques, to fully understand the mechanisms in the assay itself, including the metabolism, and then to compare the outcomes from the different assay types. Different assay types are thus complementary and best used in combination to elicit the mechanism of action of agonists, partial agonists, pure antagonists and their metabolites and will be very useful to identify new metabolites (Nielen et al., 2006; Rijk et al., 2006; Van Liempd et al., 2006).

278

10.5

Endocrine-disrupting chemicals in food

In vitro bioassays to determine indirect effects on endogenous hormone levels

The in vitro transcription activation bioassays described above can be used to determine the estrogenic, anti-estrogenic, androgenic and anti-androgenic potency of compounds based on their ability to bind specifically to one or more of the nuclear receptors (NR). Although the main mechanism of action of all well-known hormones involves the binding to their cognate receptor, some compounds have been shown to exert hormone-like effects without binding to a receptor. These compounds may, for example, elicit their effects by changing the endogenous hormone levels. The H295R steroidogenesis assay described by Hecker and Giesy (2008) is able to identify compounds that alter the endogenous hormone levels. The H295R cell-line is derived from the human adrenal cortex and is unique as it expresses all the key enzymes for steroidogenesis. The assay is performed under standard cell culture conditions in 24-well plates exposed for 48 h. At the end of the exposure period, the medium is removed from each well and the hormone contents are analysed. This assay is promising and is suited to detect the activity of compounds that do not exert their effect by binding to a cognate NR. Similarly, tests have been developed to detect potential effects through the aromatase enzyme (Sanderson et al., 2000). This enzyme plays a crucial role in the metabolism of androgens into estrogens. Various compounds have been shown to either inhibit or induce the activity of this enzyme, thus disturbing the equilibrium in the animal. Heneweer et al. (2004) used such an assay to test compounds for their potential to inhibit or induce this enzyme in human adrenocorticocarcinoma H295R and rat R2C Leydig cells.

10.6

Ah-receptor assays

The arylhydrocarbon or Ah-receptor is located in the cytosol, as opposed to the nucleus. Its migration to the nucleus upon ligand binding and its activation of a number of specific genes show strong similarities with that of the steroid hormone receptors. The general presence of this pathway in nearly all tissues and organs of many different species suggests that it must somehow be involved in the regulation of important processes in the life cycle. To date, however, the natural ligand of this receptor pathway has not been characterised. The receptor received its name based on its affinity for polycyclic aromatic hydrocarbons (PAHs) such as B[a]P. In addition, a number of chlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls (PCBs) are also able to bind to the receptor and activate the pathway (Poland and Knutson, 1982). In fact, most if not all, adverse effects of these environmental contaminants are thought to be regulated by this receptor.

Bioassays for the detection of hormonal activities

279

These effects include formation of liver tumours, but also more subtle effects on the immune system, learning ability and reproductive systems. As a result dioxins are considered endocrine disruptors. The exact mechanism behind these effects remains to be elucidated. At the gene level, the most well-known effects are the increased expression and activity of a number of enzymes involved in the biotransformation of both endogenous compounds and xenobiotics. Examples are cytochromes P450 1A1, 1A2, 1B1 and Phase II enzymes such as glutathione-S-transferase, UDP-glucuronyltransferase and NAD(P)H quinone reductase (NQO1). In fact, the most classic bioassays for the detection of AhR agonists are based on the increased hydroxylation of arylhydrocarbons (such as B[a]P) or deethylation of 7ethoxy-resorufin, either by liver homogenates or liver cells after exposure to the test compounds (Sawyer et al., 1984; Schrenk et al., 1991; Kennedy et al., 1993; Behnisch et al., 2002). These so-called AHH- and EROD assays were used to determine the relative potencies of the various different dioxins and dioxin-like PCBs. Their use for screening of food and environmental samples has been explored (Tillit et al., 1991), but the actual application in this area appears to be rather limited. A potential disadvantage of the EROD assay is the inhibition of the enzyme activity by certain agonists, thus leading to a false-negative result. Bergamottin, a furocoumarin present in grapefruit juice, is the best-known example, but other compounds, including certain PCBs, may be capable of competitive inhibition of the EROD activity (Naderi-Kalali et al., 2005). The latter may be relevant when testing mixtures or test samples. For this reason, it was decided to develop a new assay by introducing a reporter gene under the control of a dioxin-responsive element into mouse and rat hepatoma cell-lines (Aarts et al., 1995; Denison et al., 2004; Han et al., 2004). These so-called CALUX assays were subsequently validated and used during several incidents with dioxins and PCBs in the feed and food chain (Bovee et al., 1998; Hoogenboom et al., 1999, 2004, 2007). In principle, many different compounds are able to cause a response in the test, including polyaromatic hydrocarbons (Bovee et al., 1996; Machala et al., 2001) and brominated dioxins (Behnisch et al., 2003), but also many natural compounds. Selectivity for dioxins and dioxin-like PCBs is obtained by inclusion of various clean-up steps, the most important one being a clean-up on acid silica. Nevertheless, certain types of samples give high false-positive rates and suggest the presence of unknown agonists, which may be relevant to the consumer. The assay has been accredited by a number of laboratories and is officially recognised by the EU as a screening tool for dioxins and dl-PCBs (EC 2002a, b). A rather remarkable fact is the cross-talk between corticosteroids and AhR agonists in the rat hepatoma cells used for both the EROD and CALUX assays. Addition of the synthetic corticosteroid dexamethasone results in an increase in the response to TCDD (Wiebel and Cikryt, 1990; Hoogenboom et al., 1999; Lai et al., 2004). In the CALUX assay, dexamethasone itself also

280

Endocrine-disrupting chemicals in food

causes a dose-related response. The role of the GR was demonstrated by the fact that onapristone, a non-dimerising GR antagonist, completely suppressed the response caused by dexamethasone itself and the effect on the TCDD response. The effect of dexamethasone alone was shown to be an artefact caused by the presence of GRE/PREs in the original construct used in the first cell-lines. Exclusion of these GREs (Sonneveld et al., 2006b) removed this effect of dexamethasone but did not result in the abolishment of the cross-talk between dexamethasone and TCDD, suggesting another pathway, which is also active in the case of the EROD assay. Sonneveld et al. (2007) propose an up-regulation of the Ah-receptor levels by TCDD in rat and mouse cells. When testing samples, this cross-talk is only relevant when using a less stringent clean-up procedure, since acid silica will remove corticosteroids from the extracts. However, when testing pure compounds one should be aware of this potential artefact. Whether the interaction between AhR agonists and corticosteroids is also relevant for the toxicity of dioxins for animals and humans remains to be investigated. There are indications that the effect is restricted to rodents (Sonneveld et al., 2007), but Celander et al. (1997) observed it also in porcine and human endothelial cells. The first CALUX-assays used an unstable luciferase as reporter protein. The luciferase produced during the first hours of incubation is degraded by the cells. This turned out to be an advantage when dealing with Ah-receptor agonists that are metabolised by the cells, including e.g. B[a]P (Bovee et al., 1996; Hoogenboom et al., 1999). As a result the CALUX-assay was able to distinguish metabolically stable Ah-receptor agonists (e.g. dioxins) from unstable agonists (e.g. B[a]P) by varying the exposure time. Nevertheless, new assays were developed with a stable luciferase (Han et al., 2004; Denison et al., 2004). In addition, assays using green fluorescence protein as reporter were introduced (so-called CAFLUX). In practice these assays are suitable alternatives for dioxins and dioxin-like PCBs since the clean-up with acid silica, possibly extended with activated carbon, will remove most of the unstable agonists. A potential disadvantage is the gradual increase in the background response, due to the continuous production of low amounts of the reporter protein in response to low levels of AhR agonists in the culture medium.

10.7

Other hormonal bioassays

The group of McDonnell O’Malley not only developed yeast assays for estrogens and androgens, but also constructed a yeast transformant that contained the human progesterone receptor along with the appropriate steroid responsive elements upstream of the beta-galactosidase reporter gene (Mak et al., 1989; Gaido et al., 1997). A similar yeast progesterone bioassay was developed by Berg et al. (2000). This assay was miniaturised

Bioassays for the detection of hormonal activities

281

into 384- and 1536-well plate format and turned out to be very useful in modern drug discovery facilities. Li et al. (2006) also developed a recombinant hPR-gene yeast and used it for testing drinking water and Chatterjee et al. (2008) developed a recombinant hPR yeast expressing yEGFP as a marker protein. There are a number of bioassays based on mammalian celllines for progesterone-like steroids and glucocorticoids (Sonneveld et al., 2006a). Yeast systems for the detection of glucocorticoids are not yet available. The β2-adrenergic receptor is the receptor type that is responsive to anabolic agents that are illegally used in sports doping and meat production. Agents such as clenbuterol, mabuterol, salmeterol and ractopamine are the best known examples (Meenagh et al., 2001). Development of a specific reporter gene assay for β-agonists appears to be extremely difficult and is probably impossible. Receptors for β-agonists are located on cell membranes and upon binding of an agonist, these 7-transmembrane receptors activate an adenylate cyclase via a G-protein. The subsequent formation of the intracellular adenosine 3′,5′-cyclic monophosphate (cAMP) signalling molecule is not specific enough and can be influenced by many cell processes. In 2003, Nielen et al. detected by accident a new β-agonist with an immunoassay, designed for clenbuterol (Nielen et al., 2003). The β-adrenergic potency was confirmed by use of a radioreceptor assay. Up to now, these kinds of binding assay have been the most suitable tests for detecting β-agonists in food samples (Meenagh et al., 2001, 2002; Nielen et al., 2003).

10.8

Conclusions and future trends

It is evident that the need for bioassays has resulted in the development of a large number of different assays, and that there are still new developments to improve these assays. In particular the large number of compounds that still require investigation stress the need for high-throughput and easy-touse methods which accurately predict the potential hormonal properties of the compounds in vivo. It should be stressed that the actual properties of suspected compounds require further evaluation, including animal testing. In particular the kinetics of compounds and potential degradation is normally not covered by in vitro assays. Similar is true for the activation and this may result in false-negative results. Agonists or antagonists that require metabolic activation for conversion to a bioactive state may not be detected. In contrast, compounds that are metabolically inactivated by mammalian cells may persist and be especially potent in yeast assays. Future research should focus on the inclusion of such parameters in the test strategies. A clear example is the use of liver homogenates in the assays in order to detect so-called prohormones. However, degradation may be at least as important as activation of compounds, since many false-positive results may be generated. When choosing a particular kind of assay, it is evident that current

282

Endocrine-disrupting chemicals in food

yeast assays offer an equally good alternative as mammalian cell assays and in general are easier to use in routine laboratories with microbiological experience. Another important area is the application for detecting hormone-like compounds in the environment or food-chain. Both mammalian cell- and yeast-based assays have been used for this purpose. A potential drawback of yeast-based assays is their lower sensitivity but in general this is sufficient for routine application in many areas. On the other hand, mammalian cells are less robust and often suffer from cross-talk between different types of hormonal activity. In the end, both types of assays have their merit and deserve further exploration on their potential applications. Future work should also focus on the development of strategies and methods for the identification of unknown substances with (anti)hormonal activity and the inclusion of biotransformation by use of e.g. pure enzymes, liver homogenates or liver slices (Van der Kerkhof et al., 2007).

10.9

References

aarts j m m j g, denison m s, cox m a, schalk m a c, garrison p m, tullis k, de haan l h j and brouwer a (1995) ‘Species-specific antagonism of Ah receptor action by 2,2′,5,5′-tetrachloro- and 2,2′,3,3′,4,4′-hexachlorobiphenyl’, Eur J Pharmacol Environ Toxicol, 293 (4), 463–474. allen a and doisy e a (1923) ‘An ovarian hormone. Preliminary report on its location, extraction and partial purification and action in test animals’, J Am Med Assoc, 81, 819–821. arnold s f, collins b m, robinson m k, guillette jr l j and mclachlan j a (1996) ‘Differential interaction of natural and synthetic estrogens with extracellular binding proteins in a yeast estrogen screen’, Steroids, 61, 642–646. behnisch p a, hosoe k, brouwer a and sakai s (2002) ‘Screening of dioxin-like toxicity equivalents for various matrices with wildtype and recombinant rat hepatoma H4IIE cells’, Toxicol Sci, 69 (1), 125–130. behnisch p a, hosoe k and sakai s (2003) ‘Brominated dioxin-like compounds: in vitro assessment in comparison to classical dioxin-like compounds and other polyaromatic compounds’, Environ Int, 29 (6), 861–877. berg m, hundís k, thiericke r, moore t and posten c (2000) ‘Miniaturization of a functional transcription assay in yeast (human progesterone receptor) in the 384and 1536-well plate format’, J Biomol Screening, 5, 71–76. bergeron j m, crews d and mclachlan j a (1994) ‘PCBs as environmental oestrogens. Turtle sex determination as a biomarker of environmental contamination’, Environ Health Perspect, 102, 780–781. bitman j and cecil h c (1970) ‘Estrogenic activity of DDT analogs and polychlorinated biphenyls’, J Agric Food Chemi, 18, 1108–1112. blankvoort b m j, de groene e m, van meeteren-kreikamp a p, witkamp r f, rodenburg r j t and aarts j m m j g (2001) ‘Development of an androgen reporter gene assay (AR-LUX) utilising a human cell line with an endogenously regulated androgen receptor’, Anal Biochem, 298 (1), 93–102. bovee t f h, hoogenboom l a p, traag w a, zuidema t, horstman j h j, aarts j m m j g, murk t j, brouwer a, denison m s and kuiper h a (1996) ‘Biological screening of Ah receptor agonist activity in butter fat and coconut oil by

Bioassays for the detection of hormonal activities

283

means of chemical-activated luciferase expression in a genetically engineered cell line (CALUX)’, Organohalogen Compounds, 27, 303–308. bovee t f h , hoogenboom l a p, hamers a r m, traag w a, zuidema t, aarts j m m j g, brouwer a and kuiper h a (1998) ‘Validation and use of the CALUX-bioassay for the determination of dioxins and PCBs in bovine milk’, Food Add Contam, 15 (8), 863–875. bovee t f h, helsdingen j r, koks p d, kuiper h a, hoogenboom l a p and keijer j (2004a) ‘Development of a rapid yeast estrogen bioassay, based on the expressing of green fluorescent protein’, Gene, 325, 187–200. bovee t f h, helsdingen j r, rietjens i m c m, keijer j and hoogenboom l a p (2004b) ‘Rapid yeast estrogen bioassays stably expressing human estrogen receptors α and β, and green fluorescent protein: a comparison of different compounds with both receptor types’, J Steroid Biochem Mol Biol, 91 (3), 99–109. bovee t f h, heskamp h h, hamers a r m, hoogenboom l a p and nielen m w f (2005) ‘Validation of a rapid yeast estrogen bioassay, based on the expression of green fluorescent protein, for the screening of estrogenic activity in calf urine’, Anal Chim Acta, 529 (1–2), 57–64. bovee t f h, bor g, heskamp h h, hoogenboom l a p and nielen m w f (2006) ‘Validation and application of a robust yeast estrogen bioassay for the screening of estrogenic activity in animal feed’, Food Add Contam, 23 (6), 556–568. bovee t f h, helsdingen j r, hamers a r m, van duursen, m b m, nielen m w f and hoogenboom l a p (2007a) ‘A new highly specific and robust yeast androgen bioassay for the detection of agonists and antagonists’, Anal Bioanul Chem, 389, 1549–1588. bovee t f h, hamers, a r m, schoonen w g e j, jorge bento m and penijnburg a a c m (2006b) ‘Screening of synthetic and plant-derived compounds for (anti-) estrogenic and (anti-) androgenic activities’, Anal Bioanul Chem, 390, 1111–1119. bovee t f h, lommerse j p m, peijnenburg a a c m, antunes fernandes e and nielen m w f (2008) ‘A new highly androgen specific yeast biosensor, enabling optimisation of (Q)SAR model approaches’, J Steroid Biochem Mol Biol, 108, 121–131. breithofer a, graumann k, scicchitano m s, karathanasis s k, butt t r and jungbauer a (1998) ‘Regulation of human estrogen receptor by phytoestrogens in yeast and human cells’, J Steroid Biochem Mol Biol, 67, 421–429. brodie j and mcewan i j (2005) ‘Intra-domain communication between the Nterminal and DNA-binding domains of the androgen receptor: modulation of androgen response element DNA binding’, J Mol Endocrinol, 34, 603–615. canton r f, bovee t f h, daamen f, van duursen m and van den berg m (2007) ‘In vitro antiandrogenicity of PBDEs, HBCD, TBP and hydroxylated and methoxylated PBDEs based on a yeast assay’, Short abstract, NST Annual Meeting 2007, Wageningen. celander m, weisbrod r and stegeman j j (1997) ‘Glucocorticoid potentiation of cytochrome P4501A1 induction by 2,3,7,8-tetrachlorodibenzo-p-dioxin in porcine and human endothelial cells in culture’, Biochem Biophys Res Comm, 232 (3), 749–753. charles g d, bartels m j, zacharewski t r, gollapudi b b, freshour n l and carney e w (2000) ‘Activity of benzo[a]pyrene and its hydroxylated metabolites in an estrogen receptor α reporter gene assay’, Tox Sci, 55, 320–326. chatterjee s, kumar v, majumder c b and roy p (2008) ‘Screening of some antiprogestin endocrine disruptors using a recombinant yeast-based in vitro bioassay’, Toxicology in Vitro, 22, 788–798. coecke s et al. (2006) ‘Metabolism: A bottleneck in in vitro toxicological test development’, The report and recommendations of ECVAM Workshop 54. ATLA 34, 49–84.

284

Endocrine-disrupting chemicals in food

courant f, antignac j-p, maume d, monteau f, andersson a-m, skakkebæk n, andre f and le bizec b (2007) ‘Exposure assessment of prbubertal children to steroid endocrine disrupters; 1. Analytical strategy for estrogens measurement in plasma at ultra-trace level’, Anal Chim Acta, 586 (1–2), 105–114. dauvois s, danielian p s, white r and parker m g (1992) ‘Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover’, Proc Natl Acad Sci, 89, 4037–4041. denison m s, zhao b, baston d s, clark g c, murata h and han d (2004) ‘Recombinant cell bioassay systems for the detection and relative quantitation of halogenated dioxins and related chemicals’, Talanta, 63, 1123–1133. ec (2002a) ‘Commission Directive 2002/69/EC of 26 July 2002 laying down the sampling methods and the methods of analysis for the official control of dioxins and the determination of dioxin-like PCBs in foodstuffs’, Official J Eur Comm, L209, 5–14. ec (2002b) ‘Commission Directive 2002/70/EC of 26 July 2002 establishing requirements for the determination of levels of dioxins and dioxin-like PCBs in feedingstuffs’, Official J Eur Comm, L209, 15–21. egner r, rosenthal f e, kralli a, sanglard d and kuchler k (1998) ‘Genetic separation of FK506 susceptibility and drug transport in the yeast Pdr5 ATP-binding cassette multidrug resistance transporter’, Mol Biol Cell, 9, 523–543. elkak a e and mokbel k (2001) ‘Pure antiestrogens and breast cancer’, Curr Med Res Opin, 17, 282–289. fang h, tong w, perkins r, soto a m, prechtl n v and sheehan d m (2000) ‘Quantitative comparison of in vitro assays for estrogenic activities’, Environ Health Perspect, 108, 723–729. fertuck k c, matthews j b and zacharewski t r (2001) ‘Hydroxylated benzo[a]pyrene metabolites are responsible for in vitro estrogen receptor-mediated gene expression induced by benzo[a]pyrene, but do not elicit uterotrophic effects in vivo’, Tox Sci, 59, 231–240. gaido k w, leonard l s, lovell s, gould j c, babaï d, portier c j and mcdonnell d p (1997) ‘Evaluation of chemicals with endocrine modulating activity in a yeastbased steroid hormone receptor gene transcription assay’, Toxicol Appl Pharm, 143, 205–212. galey f d, terra r, walker r, adaska j, etchebarne m a, puschner b, fisher e, whitlock r h, rocke t, willoughby d and tor e (2000) ‘Type C botulism in dairy cattle from feed contaminated with a dead cat’, J Vet Diagn Investig, 12, 204–209. garcia-reyero n, grau e, castillo m, lopez de alda m j, barcelo d and pina b (2001) ‘Monitoring of endocrine disruptors in surface waters by the yeast recombinant assay’, Environ Toxicol Chem, 20, 1152–1158. gibson m k, nemmers l a, beckman jr w c, davis v l, curtis s w and korach k s (1991) ‘The mechanism of ICI 164,384 antiestrogenicity involves rapid loss of estrogen receptor in uterine tissue’, Endocrinology, 129, 2000–2010. graumann k, breithofer a and jungbauer a (1999) ‘Monitoring of estrogen mimics by a recombinant yeast assay: synergy between natural and synthetic compounds?’, Sci Total Env, 225, 69–79. gustafsson j a (1999) ‘Estrogen receptor β – a new dimension in estrogen mechanism of action’, J Endocrinol, 163, 379–383. haelens a, verrijdt g, callewaert l, christiaens v, schauwaers k, peeters b, rombauts w and claessens f (2003) ‘DNA recognition by the androgen receptor: evidence for an alternative DNA-dependent dimerization, and an active role of sequences flanking the response element on transactivation’, Biochem J, 369, 141–151. han d, nagy s r and denison m s (2004) ‘Comparison of recombinant cell bioassays for the detection of Ah receptor agonists’, Biofactors, 20, 11–22.

Bioassays for the detection of hormonal activities

285

hecker h and giesy j p (2008) ‘Novel trends in endocrine disruptor testing: the H295R steroidogenesis assay to identify inducers and inhibitors of hormone production’, Anal Bioanal Chem, 390, 287–291. heneweer m, van den berg m and sanderson j t (2004) ‘A comparison of human H295R and rat R2C cell lines as in vitro screening tools for effects on aromatase’, Toxicol Lett, 146, 183–194. henley d v, lipson n, korach k s and bloch c a (2007) ‘Prepuberal gynecomastia linked to lavender and tea tree oils’, New Eng J Med, 356, 479–485. herschberger l g, shiply e g and meyer r k (1953) ‘Myotrophic activity of 19nortestosterone and other steroids determined by modified levator and muscle method’, Proc Soc Exp Biol Med, 83, 175–180. hoogenboom l a p, hamers a r m and bovee t f h (1999) ‘Bioassays for the detection of growth-promoting agents, veterinary drugs and environmental contaminants in food’, The Analyst, 124, 79–85. hoogenboom l a p, de haan l, hooijerink d, bor g, murk a j and brouwer a (2001) ‘Estrogenic activity of estradiol and its metabolites in the ER-CALUX assay with human T47D breast cells’, Acta Pathol Microbiol Immunol Scand, 109, 101–107. hoogenboom l a p, bovee t f h, portier l, bor g, weg g van der, onstenk c and traag w a (2004) ‘The German bakery waste incident; use of a combined approach of screening and confirmation for dioxins in feed and food’, Talanta, 63 (5), 1249–1253. hoogenboom l a p, van eijkeren j c h, zeilmaker m j, mengelers m j b, herbes r, immerzeel j and traag w a (2007) ‘A novel source for dioxins present in recycled fat from gelatin production’, Chemosphere, 68, 814–823. hungerford j m (1995). AOAC Official Method 959.08. Paralytic shellfish poison. Official Methods of Analysis, 16th edition (Arlington, VA: AOAC International), Chapter 35.1.37. jenkins r l, wilson e m, angus r a, howell w m and kirk m (2003) ‘Androstenedione and progesterone in the sediment of a river receiving paper mill effluent’, Toxicol Sci, 73, 53–59. jobling s, reynolds t, white r, parker m g and sumpter j p (1995) ‘A variety of environmentally persistent chemicals, including some phthalate plasticizers, are weakly estrogenic’, Environ Health Perspect, 103, 582–587. jordan v c (2003) ‘Antiestrogens and selective estrogen receptor modulators as multifunctional medicines’, J Medicinal Chem, 46, 883–908. juretic a, saric n, bisof v and basic-coretic m (2006) ‘Fulvestrant: a new agent in endocrine treatment for breast cancer’, Lijec Vjesn, 128, 31–36. kennedy s w, lorenzen a, james c a and collins b t (1993) ‘Ethoxyresorufin Odeethylase and porphyrin analysis in chicken embryo hepatocyte cultures with a fluorescence multiwell plate reader’, Anal Biochem, 211, 102–112. klein k o, baron j, colli m j, mcdonnell d p and cutler g b (1994) ‘Estrogen levels in childhood determined by an ultrasensitive recombinant cell bioassay’, J Clin Invest, 94, 2475–2480. körner w, vinggaard a m, térouanne b, ma r, wieloch c, schlumpf m, sultan c and soto a m (2004) ‘Interlaboratory comparison of four in vitro assays for assessing androgenic and antiandrogenic activity of environmental chemicals’, Environ Health Perspect, 112 (6), 695–702. lai k p, wong m h and wong c k c (2004) ‘Modulation of AhR-mediated CYP1A1 mRNA and EROD activities by 17 beta-estradiol and dexamethasone in TCDDinduced H411E cells’, Toxicol Sci, 78 (1), 41–49. leblanc k, sexton e, parent s, belanger g, dery m c, boucher v and asselin e (2007) ‘Effects of 4-hydroxytamoxifen, raloxifene and ICI 182780 on survival of uterine cancer cell lines in the presence and absence of exogenous estrogens’, Int J Oncol, 30, 477–487.

286

Endocrine-disrupting chemicals in food

ledoux m and hall s (2000) ‘Proficiency testing of eight French laboratories in using the AOAC mouse bioassay for paralytic shellfish poisoning: interlaboratory collaborative study’, J AOAC Int, 83, 305–310. lee h j, lee y s, kwon h b and lee k (2003) ‘Novel yeast bioassay system for detection of androgenic and antiandrogenic compounds’, In Vitro Toxicol, 17, 237–244. legler j, van den brink c e, brouwer a, murk a j, van der saag p t, vethaak d a and van der burg b (1999) ‘Development of a stably transfected estrogen receptor-mediated luciferase reporter gene assay in the human T47D breast cancer cell line’, Toxicol Sci, 48, 55–66. legler j, dennekamp m, vethaak a d, brouwer a, koeman j h, van der burg b and murk a j (2002) ‘Detection of estrogenic activity in sediment-associated compounds using in vitro reporter gene assays’, Sci Total Environ, 293, 69–83. le guevel r and pakdel f (2001) ‘Streamlined beta-galactosidase assay for analysis of recombinant yeast response to estrogens’, Biotechniques, 30, 1000–1004. lemaire g, terouanne b, mauvais p, michel s and rahmani r (2004) ‘Effect of organochlorine pesticides on human androgen receptor activation in vitro’, Toxicol Appl Pharmacol, 196, 235–246. leskinen p, michelini e, picard d, karp m and virta m (2005) ‘Bioluminescent yeast assays for detecting estrogenic and androgenic activity in different matrices’, Chemosphere, 61, 259–266. li j, cui q, ma m, rao k f and wang z j (2006) ‘Recombinant hPR gene yeast for assessing in drinking water’, Huan Jing Ke Xue, 27, 2463–2466. machala m, vondracek j, blaha l, ciganek m and neca j (2001) ‘Aryl hydrocarbon receptor-mediated activity of mutagenic polycyclic aromatic hydrocarbons determined using in vitro reporter gene assay’, Mutation Res, 497 (1–2), 49–62. maillard s, gauduchon j, marsaud v, gouilleux f, connault e, opolon p, fattal e, sola b and renoir j m (2006) ‘Improved antitumoral properties of pure antiestrogen RU 58668-loaded liposomes in multiple myeloma’, J Steroid Biochem Mol Biol, 100, 67–78. mak p, mcdonnell d p, weigel n l, schrader w t and o’malley b w (1989) ‘Expression of functional chicken oviduct progesterone receptors in yeast’, J Biol Chem, 264, 21613–21618. mcinerney e m, weis k e, sun j, mosselman s and katzenellenbogen b s (1998) ‘Transcription activation by the human estrogen receptor subtype β (ERβ) studied with ERβ and ERα receptor chimeras’, Endocrinology, 139, 4513–4522. meenagh s a, elliot c t, buick r k, izeboud c a and witkamp r f (2001) ‘The preparation, solubilisation and binding characteristics of a β2-adrenoceptor isolated from transfected Chinese hamster cells’, The Analyst, 126, 491–494. meenagh s a, mcevoy j d g and elliott c t (2002) ‘Determination of carazolol residues in porcine tissue by radioreceptor assay’, Anal Chim Acta, 462, 149–156. michelini e, leskinen p, virta m, karp m and roda a (2005) ‘A new recombinant cell-based bioluminescent assay for sensitive androgen-like compound detection’, Biosens Bioelectronics, 20, 2261–2267. milligan s r, balasubramanian a v and kalita j c (1998) ‘Relative potency of xenobiotic estrogens in an acute in vivo mammalian assay’, Environ Health Perspect, 106, 23–26. morito k, hirose t, kinjo j, hirakawa t, okawa m, nohara t, ogawa s, inoue s, muramatsu m and masamune y (2001) ‘Interaction of phytoestrogens with estrogen receptors α and β’, Biol Pharm Bull, 24, 351–356. mueller s o (2002) ‘Overview of in vitro tools to assess the estrogenic and antiestrogenic activity of phytoestrogens’, J Chromatogr B, 777, 155–165. murk a j, legler l, denison m s, giesy j p, van de guchte c and brouwer a (1996) ‘Chemical-activated luciferase gene expression (CALUX): a novel in vitro

Bioassays for the detection of hormonal activities

287

bioassay for Ah receptor active compounds in sediments and pore water’, Fundam Appl Toxicol, 33, 149–160. naderi-kalali b, allameh a, rasaee m j, bach h j, behechti a, doods k, kettrup a and schramm k w (2005) ‘Suppressive effects of caraway (Carum carvi) extracts on 2,3,7,8-tetrachloro-dibenzo-p-dioxin-dependent gene expression of cytochrome P450 1A1 in the rat H4IIE cells’, Toxicol In Vitro, 19 (3), 373–377. nielen m w f, elliott c t, boyd s a, courtheyn d, essers m l, hooijerink h h, van bennekom e o and fuchs r e m (2003) ‘Identification of an unknown betaagonist in feed by liquid chromatography/bioassay/quadrupole time-of-flight tandem mass spectrometry with accurate mass measurement’, Rapid Comm Mass Spect, 17 (14), 1633–1641. nielen m w f, bovee t f h, van engelen m c, rutgers p, hamers a r m, van rhijn j a and hoogenboom l a p (2006) ‘Urine testing for designer steroids by liquid chromatography with androgen bioassay detection and electrospray quadrupole time-of-flight mass spectrometry identification’, Anal Chem, 78, 424–431. nih (2003) ICCVAM Evaluation of In Vitro Methods for Detecting Potential Endocrine Disruptors. Publication No: 03-4503. nouws j f m, broex n j g, den hartog j m p and driesens f (1988) ‘The new Dutch kidney test’, Arch Lebensmittelhyg, 39, 135–138. nouws j f m, loeffen g, schouten j, van egmond h, keukens h and stegeman h (1998) ‘Testing of raw milk for tetracycline residues’, J Dairy Sci, 81, 2341–2345. orwoll e s, stribrska l, ramsey e e and keenan e j (1991) ‘Androgen receptors in osteoblast-like cell lines’, Calcif Tissue Int, 49, 183–187. owens w, gray l e, zeiger e, walker m, yamasaki k, ashby j and jacob e (2007) ‘The OECD program to validate the rat Herschberger bioassay to screen compounds for in vivo androgen and antiandrogen responses: phase 2 dose–response studies’, Environ Health Perspect, 115, 671–678. paris f, balaguer p, térouanne b, servant n, lacoste c, cravedi j-p, nicolas j-c and sultan c (2002) ‘Phenylphenols, biphenols, bisphenol-A and 4-tertoctylphenol exhibit alpha and beta estrogen activities and antiandrogen activity in reporter cell lines’, Mol Cell Endocrinol, 193, 43–49. peperzak l (2005) ‘Future increase in harmful algal blooms in the Noth Sea due to climate change’, Water Sci Technol, 51, 31–36. pereira j j, ziskowski j, marcaldo-allen r, kuropat c, luedke d and gould e (1992) ‘Vitellogenin in winter flounder (Pleuronectes americanus) from Long Island Sound and Boston harbour’, Estuaries, 15, 289–297. petit f, valotaire y and pakdel f (1995) ‘Differential functional activities of rainbow trout and human estrogen receptors expressed in the yeast Saccharomyces cerevisiae’, Eur J Biochem, 233, 584–592. petz l n, ziegler y s, schultz j r and nardulli a m (2004) ‘Fos and Jun inhibit estrogen-induced transcription of the human progesterone receptor gene through an activator protein-1 site’, Mol Endocrinol, 18, 521–532. pham t a, hwung y p, santiso m d, mcdonnell d p and o’malley b w (1992) ‘Liganddependent and independent function of the transcriptivation regions of the human estrogen receptor in yeast’, Mol Endocrinol, 6, 1043–1050. pike m c, spicer d v, dahmoush l and press m f (1993) ‘Estrogens, progesterons, normal breast cell proliferation, and breast cancer risk’, Epidemiol Rev, 15, 17–35. poland a and knutson j c (1982) ‘2,3,7,8-tetrachlorodibenzo-p-dioxin and related halogenated aromatic hydrocarbons: examination of the mechanism of toxicity’, Ann Rev Pharmacol Toxicol, 22, 517–542. pons m, gagne d, nicolas j c and mehtali m (1990) ‘A new cellular model of response to estrogens: a bioluminescent test to characterize (anti) estrogen molecules’, Biotechniques, 9, 450–459.

288

Endocrine-disrupting chemicals in food

ramamoorthy k, wang f, chen i, norris j d, mcdonnell d p, leonard l s, gaido k w, bocchinfuso w p, korach k s and safe s (1997) ‘Estrogenic activity of a dieldrin/toxaphene mixture in the mouse uterus, MCF-7 human breast cancer cells, and yeast-based estrogen receptor assays: no apparent synergism’, Endocrinology, 138, 1520–1527. rehmann k, schramm k and kettrup a a (1999) ‘Applicability of a yeast oestrogen screen for the detection of oestrogen-like activities in environmental samples’, Chemosphere, 38, 3303–3312. rijk j c w, bovee t f h, groot m j, peijnenburg a a c m and nielen m w f (2006) ‘Metabolic activation and TOFMS identification of prohormones using liver S9 in combination with a highly androgen-specific yeast bioassay’, Short abstract, Fifth Int Symp Hormone Vet Residue Anal, Antwerpen. rogers j m and denison m s (2000) ‘Recombinant cell bioassays for endocrine disruptors: development of a stably transfected human ovarian cell line for the detection of estrogenic and anti-estrogenic chemicals’, In Vitro Molec Toxic, 13 (1), 67–82. routledge e j and sumpter j p (1996) ‘Estrogenic activity of surfactants and some of their degradation products using a recombinant yeast screen’, Environ Toxicol Chem, 15, 241–248. sanderson j t, seinen w, giesy j p and van den berg m (2000) ‘2-Chloro-striazine herbicides induce aromatase (CYP19) activity in H295R human adrenocortical carcinoma cells: a novel mechanism for estrogenicity?’, Toxicol Sci, 54, 121–127. sawyer t w, vatcher a d and safe s (1984) ‘Comparative aryl hydrocarbon hydroxylase induction of commercial PCBs in Wistar rats and hepatoma H4IIE cells in culture’, Chemosphere, 13, 695–701. schoonen w g, vermeulen g j, deckers g h, verbost p m and kloosterboer h j (1999) ‘Antiprogestins: their mechanisms of action and the consequences for compound selection by in vitro and in vivo studies’, Curr Topics Steroid Res, 2, 15–54. schoonen w g, deckers g, de gooyer m e, de ries r and kloosterboer h j (2000) ‘Hormonal properties of norethisterone, 7α-methyl-noresthisterone and their derivates’, J Steroid Biochem Mol Biol, 74, 213–222. schrenk d, lipp h-p, wiesmüller t, hagenmaier h and bock k w (1991) ‘Assessment of biological activities of mixtures of polychlorinated dibenzo-p-dioxins: comparison between defined mixtures and their constituents’, Arch Toxicol, 65, 114–118. shaffer p l, jivan a, dollins d e, cleassens f and gewirth d t (2004) ‘Structural basis of androgen receptor binding to selective androgen response elements’, Proc Natl Acad Sci, 101, 4758–4763. sharpe r m, fisher j s, millar m m, jobling s and sumpter j p (1995) ‘Gestational and lactational exposure to xenoestrogens results in reduced testicular size and sperm production’, Environ Health Perspect, 103 (12), 1136–1143. shelby m d, newbold r r, tully d b, chae k and davis v l (1996) ‘Assessing environmental chemicals for estrogenicity using a combination of in vitro and in vivo assays’, Environ Health Perspect, 104 (12), 1296–1300. skakkebæk n e, jørgensen n, main k m, rajpert-demeyts e, leffers h, andersson a, juul a, carlsen e, mortensen g k, jensen t k and toppari j (2006) ‘Is human fecundity declining?’, Int J Andrology, 29, 2–11. sonneveld e, jansen h j, riteco j a c, brouwer a and van der burg b (2005) ‘Development of androgen- and estrogen-responsive bioassays, members of a panel of human cell line-based highly selective steroid-responsive bioassays’, Toxicol Sci, 83, 136–148. sonneveld e, riteco j a c, jansen h j, pieterse b, brouwer a, schoonen w g and van der burg b (2006a) ‘Comparison of in vitro and in vivo screening models for androgenic and estrogenic activities’, Toxicol Sci, 89, 173–187.

Bioassays for the detection of hormonal activities

289

sonneveld e, jansen h j, man s, jonas a, brouwer a and burg b van der (2006b) ‘Glucocorticoid mediated expression of dioxin target genes in rat H4IIE cells but not in human HEPG2 and T47D cells’, Organohal Comp, 68, 564–567. sonneveld e, jonas a, meijer o c, brouwer a and van der burg b (2007) ‘Glucocorticoid-enhanced expression of dioxin target genes through regulation of the rat aryl hydrocarbon receptor’, Toxicol Sci, 99, 455–469. soto a m, justicia h, wray j w and sonnenschein c (1991) ‘p-Nonyl-phenol: an estrogenic xenobiotic released from “modified” polysterene’, Environ Health Perspect, 92, 167–173. soto a m, lin t-m, justicia h, silvia r m and sonnenschein c m (1992) ‘An “in culture” bioassay to assess the estrogenicity of xenobiotics’, in Colborn T & Clement C (eds.) Chemically Induced Alterations in Sexual Development: The Wildlife/Human Connection. Princeton, NJ: Princeton Scientific Publishing, pp. 295–309. soto a m, sonnenschein c, chung k l, fernandez m f, olea n and olea-serrano f (1995) ‘The E-screen assay as a tool to identify estrogens: an update on estrogenic environmental pollutants’, Environ Health Perspect, 103 (suppl. 7), 113–122. soto a m, lin t m, justicia h, silvia r m and sonnenschein c (1998) ‘An “in culture” bioassay to assess the estrogenicity of xenobiotics (E-SCREEN)’, J Clean Techn, Environm Toxicol Occup Med, 7 (3), 331–343. soto a m, calabro j m, prechtl n v, yau a y, orlando e f, daxenberger a, kolok a s, guilette l j, le bizec b, lange i g and sonnenschein c (2004) Androgenic and estrogenic activity in water bodies receiving cattle feedlot effluent in Eastern Nebraska, USA’, Environ Health Perspect, 112 (3), 346–352. soto a m, maffini m v, schaeberle c m and sonnenschein c (2006) ‘Strengths and weaknesses of in vitro assays for estrogenic and androgenic activity’, Best Practice Res Clin Endocrin Metab, 20, 15–33. sumpter j p (1985) ‘The purification, radioimmunoassay and plasma levels of vitellogenin from the rainbow trout, Salmo gairdneri’, in: Lofts B, Holmes WN (Eds), Current Trends in Comparative Endocrinology. University Press: Hong Kong; pp 355–357. szelei j, jimenez j, soto a m, luizzi m f and sonnenschein c (1997) ‘Androgeninduced inhibition of proliferation in human breast cancer MCF7 cells transfected with androgen receptor’, Endocrinology, 138 (4), 1406–1412. tester p a (1994) ‘Harmful marine phytoplankton and shellfish toxicity. Potential consequences of climate change’, Ann N Y Acad Sci, 740, 69–76. thomson b, cressey pj and shaw ic (2003) ‘Dietary exposure to xenoestrogens in New Zealand’, J Environ Monit, 5, 229–235. tillitt d e, giesy j p and ankley g t (1991) ‘Characterization of the H4IIE rat hepatoma cell bioassay as a tool for assessing toxic potency of planar halogenated hydrocarbons in environmental samples’, Environ Sci Technol, 25, 87–92. tutulan-cunita a c, mikoshi m, mizunuma m, hirata d and miyakawa t (2005) ‘Mutational analysis of the yeast multidrug resistance ABC transporter Pdr5p with altered drug specificity’, Genes to Cells, 10, 409–420. tyler c r and sumpter j p (1990) ‘The development of a radioimmunoassay for carp, Cyprinus carpio, vitellogenin’, Fish Physiol Biochemi, 8, 129–140. ulrich e m, caparell-grant a, jung s-h, hites r a and bigsby r m (2000) ‘Environmentally relevant xenoestrogen tissue concentrations correlated to biological responses in mice’, Environ Health Perspect, 108 (10), 973–977. van der kerkhof e g, de graaf i a, de jager m h and groothuis g m (2007) ‘Induction of phase I and II drug metabolism in rat small intestine and colon in vitro’, Drug Metab Dispos, 35, 898–907. van leengoed l, kluivers m, herbes r, langendijk p, stephany r, van den berg m, seinen w, grinwis g, van der lugt j, meulders f, geudeke t and verheijden j

290

Endocrine-disrupting chemicals in food

(2002) ‘The weakest link: Medroxyprogesterone acetate in pig feed’, Tijdschrift Diergeneeskunde, 127, 516–519. van liempd s m, kool j, niessen w m a, van elswijk d e, irth h and vermeulen n p e (2006) ‘On-line formation, separation, and estrogen receptor receptor affinity screening of cytochrome P450-derived metabolites of selective estrogen receptor modulators’, Drug Metabol Dispos, 34, 1640–1649. waller c l, juma b w, gray l e and kelce w r (1996) ‘Three-dimensional quantitative structure–activity relationships for androgen receptor ligands’, Toxicol Appl Pharmacol, 137, 219–227. watanabe t, inoue s, ogawa s, ishii y, hiroi h, ikeda k, orimo a and muramatsu m (1997) ‘Agonistic effect of tamoxifen is dependent on cell type, ERE-promoter context, and estrogen receptor subtype: functional difference between estrogen receptors α and β’, Biochem Biophys Res Commun, 236, 140–145. wiebel f j and cikryt p (1990) ‘Dexamethasone-mediated potentiation of P450IA1 induction in H4IIEC3/T hepatoma cells is dependent on a time-consuming process and associated with induction of the Ah receptor’, Chem-Biol Interact, 76, 307–320. willemsen p, scippo m, maghuin-rogister g, martial j a and muller m (2002) ‘Use of specific bioluminescent cell lines for the detection of steroid hormone (ant)agonists in meat producing animals’, Anal Chim Acta, 473, 119–126. willemsen p, scippo m, kausel g, figueroa j, maghuin-rogister g, martial j a and muller m (2004) ‘Use of reporter cell lines for detection of endocrine-disrupter activity’, Anal Bioanal Chem, 378, 655–663. willemsen p, scippo m, maghuin-rogister g, martial j a and muller m (2005) ‘Enhancement of steroid receptor-mediated transcription for the development of highly responsive bioassays’, Anal Bioanal Chem, 382, 894–905. witters h e, vangenechten c and berckmans p (2001) ‘Detection of estrogenic activity in Flemish surface waters using an in vitro recombinant assay with yeast cells’, Water Sci Technol, 43, 117–123. yoon k, pallaroni l, stoner m, gaido k and safe s (2001) ‘Differential activation of wild-type and variant forms of estrogen receptor α by synthetic and natural estrogenic compounds using a promoter containing three estrogen-responsive elements’, J Steroid Biochem Mol Biol, 78, 25–32. zacharewski t (1997) ‘In vitro bioassays for assessing estrogenic substances’, Environment Sci Technol, 31, 613–623.

11 Genetics, epigenetics and genomic technologies: importance and application to the study of endocrine-disrupting chemicals L. R. Ferguson and M. Philpott, The University of Auckland and Nutrigenomics, New Zealand

Abstract: Early observations on endocrine-disrupting chemicals (EDCs) associated these with significant changes in the nature and diversity of animal populations, through environmental exposures. In human populations, dietary exposures to EDCs may be even more important. Most EDCs are classified as non-genotoxic and appear to interact with hormone receptors to disrupt normal differentiation of sex organs, leading to sexual dimorphism and undermasculinisation. Both animal and human populations show variations in susceptibility to endocrine disruption, which have been related to the frequencies of specific variant single nucleotide polymorphisms (SNPs) and trinucleotide repeats. Microarray technologies, especially using animal models, permit measurement of multiple effects of EDCs on the expression of genes in a tissueand time-specific manner. Hierarchical gene network analyses reveal impacts on interlinking networks of genes, increasing the activity of some and decreasing that of others. Such approaches will continue to be important to understand the mechanism of new and novel EDCs, but are not appropriate to mass screening. Knowledge of the genes whose expression is impacted has been used to develop reporter gene assays to permit high numbers of chemicals to be interrogated for EDC capability. While such changes in gene expression may sometimes reflect classic mutational changes, there is increasing evidence to suggest that the major mechanism is epigenetic, via DNA methylation changes. The association with transgenerational effects suggests that new strategies of study, such as ChIP-chip arrays that combine chromatin immunoprecipitation with microarrays, may become an important approach to future understanding of the importance and mechanistic basis of this important group of chemicals. Key words: endocrine-disrupting chemicals, genetics, epigenetics, genomics, microarrays.

292

11.1

Endocrine-disrupting chemicals in food

Introduction

Evidence is accumulating for an adverse environmental impact of endocrine-disrupting chemicals (EDCs) that can mimic or inhibit endogenous hormones. In a recent review, McLachlan et al. (2006) describe the development of female characteristics by male fish in detergent-contaminated water, sex reversal of turtles through exposure to polychlorinated biphenyls (PCBs), multiple ovary formation in male frogs exposed to a common herbicide, and pseudohermaphroditic offspring produced by polar bears, and seals in contaminated water that show an excess of uterine fibroids. Many of the types of chemicals associated with such events are industrial contaminants such as pesticides and plasticisers (Manson and Carr, 2003; Henley and Korach, 2006). While most of the chemicals of interest exhibit estrogenic effects, a few of them are anti-estrogenic or anti-androgenic. It is becoming clear that those same chemicals that are affecting wildlife are also capable of significantly affecting human health (Guillette et al., 1994; Sumpter and Jobling, 1995; Toppari et al., 1996; Rudel, 1997; Shaw and McCully, 2002; Thomson et al., 2003; Trubo, 2005). Natural phytoestrogens found in food plants and/or in herbal supplements, for example genistein (Fig. 11.1) in soy, also have similar effects (Knight and Eden, 1995; Wang et al., 1996; Henley and Korach, 2006). Dietary exposure may be at least as important or more so to humans as compared with environmental exposure to EDCs. Estrogen is of considerable importance in women’s health, in playing a key role on both puberty and menopause. In women, EDCs are likely to affect breast growth and lactation, and may play a role in uterine diseases such as fibroids and endometriosis. In men, they have been suggested to increase the incidence of syndromes associated with male undermasculinization, such as hypospadias and cryptorchidism. This may be particularly relevant in genetically susceptible individuals (Trubo, 2005). Early observational studies of EDCs utilised population-based descriptions (animal or human), or long-term animal studies to demonstrate reproductive effects. This group of chemicals was not considered as genetic toxicants, since they failed to trigger positive responses in standard batteries of genetic toxicology tests (Rakitsky et al., 2000). However, as the mechanistic basis of endocrine disruption has become better established, it is realised that such chemicals interact with the genome in more subtle ways than had been recognised. The flexibility and power of new genomic technologies are becoming of increasing importance in detailing the mechanistic basis of individual susceptibility to such compounds, and their effects on the expression of genes. Furthermore, demonstration of potential transgenerational effects has provided new insights into the long-term implications of epigenetic effects of these compounds (Anway et al., 2005). This chapter will consider the application of genomic technologies and the manner in which this is increasing our understanding of the mechanistic

Genetics, epigenetics and genomic technologies

293

OH

OH

O

HO Estrodiol (estrogen)

Testosterone (androgen) OH

OH

OH

O

HO

O

O Genistein

DHT

CI CI

F

H N

CI

F F + O– N O

CI

CI DDT

Flutamide OH

OH

HO HO DES

BPA

OH R HO HO

OP (R=CH3), NP (R=CH2CH3)

Ethinylestradiol CI CI

O CI CI

O

N O

O

O Methoxychlor

Fig. 11.1

CI

Vinclozolin

Structures of the most common estrogen (estrodiol) and androgen (testosterone) and various EDCs mentioned in the text.

294

Endocrine-disrupting chemicals in food

basis and long-term implications of EDCs. The major focus will be on human health, but animal examples will be used where necessary to illustrate a point.

11.2

Genetic variability in susceptibility to endocrinedisrupting chemicals

Both animal and human populations vary considerably in their susceptibility to EDCs, even between individuals who are otherwise closely related. For example, 20-hydroxyecdysone (20E) is a phytoecdysteroid that has shown endocrine disruption and deterrent effects in peach-potato aphid (Myzus persicae Sulzer), a major pest in these species (Malausa et al., 2006). These authors compared the reproductive viability of six clones of M. persicae exposed to various concentrations of 20E, and estimated the ability of offspring from these clones to detect 20E in choice experiments. Even within this small number of clones, they demonstrated strong variability in response that apparently had a genetic basis (although the nature of this basis was not tested). While the immediate findings of this particular study were extrapolated to raise questions about the sustainability of control methods using 20E, they also raise the more general issue of the genetic basis for such effects. A candidate gene approach to susceptibility loci has been driven by increased understanding of the molecular biology and biochemical events underlying sex development in the male. An orderly sequence of developmental events is coordinated by an interplay of genetic and hormonal controlling elements, that are time- and concentration-dependent (Hughes et al., 2006). Once a testis is formed, differentiation of the internal and external male genitalia is androgen dependent. These stages define the three main categories of syndromes of sex reversal – defects in formation of the testis, defects in production of androgens, and defects in their action. Varying degrees of male undermasculinisation are associated with single nucleotide polymorphism (SNP) variants in several of the genes involved in male development. An increasing series of studies is associating adverse events in human populations with the combination of high frequencies of such variant SNPs, coupled with exposure to EDCs. An increased prevalence of undermasculinised external genitalia in several countries, including Japan, has been associated with increases in EDCs such as dioxins (Soneda et al., 2005). The frequencies of the Arg554Lys polymorphism of the gene for aryl hydrocarbon receptor (AHR) and the Pro185Ala polymorphism of the gene for aryl hydrocarbon receptor repressor (AHRR), were compared in 73 boys with micropenis and 80 control males. Although the allele and genotype frequencies of the AHR polymorphism were comparable between the two groups, those of the AHRR polymorphism were significantly different. The authors suggested that the

Genetics, epigenetics and genomic technologies

295

AHRR Pro185Ala polymorphism might represent a susceptibility locus for the development of micropenis in response to dioxins. Cryptorchidism affects 2% to 5% of full-term male neonates in Western countries (Toppari and Kaleva, 1999), and its prevalence has increased during the past few decades in several countries (Yoshida et al., 2005). El Houate et al. (2007) have associated the risk of this disease in Morocco with certain variant SNPs involving the Insulin-like factor 3 gene. The variants associated in the studies by Yoshida et al. (2005), were more complex. These authors considered a Japanese population group, and associated risk with a specific haplotype of the gene for estrogen receptor-α (ESRα) that mediates the estrogenic effects of EDCs. They identified a haplotype block that spanned an approximately 50-kb region, encompassing single nucleotide polymorphisms 10–14 in the 3′ region of ESRα. They reported that the frequency of this particular haplotype block was higher in the patients than in the control males. Not only did heterozygotes for this haplotype show significantly enhanced risk, the authors identified homozygotes only in the patient but not the control group. Other genomic variants have also been reported. For example, Radpour et al. (2007) associated certain trinucleotide repeats (long polyglycine tracts or GGN repeats) in exon 1 of the androgen receptor gene with cryptorchidism and penile hypospadias in Iranian subjects. Hypospadias, part of the testicular dysgenesis syndrome, is the second most frequent congenital anomaly seen in newborn males. This congenital abnormality is characterised by altered development of the urethra, foreskin and ventral surface of the penis (Wilkin et al., 1979; Manson and Carr, 2003). Case-control studies as well as data from consanguineous families have identified risk factors in the form of allelic variants in genes controlling androgen action and metabolism that are important for normal male sexual differentiation. For example, Beleza-Meireles et al. (2006) associated polymorphisms of estrogen receptor-β (ERβ) gene with the risk of hypospadias, while a later study by this group (Beleza-Meireles et al., 2007) further suggested a co-chaperone of the androgen receptor, FKBP52, as a candidate gene for hypospadias. It seems likely that the variant alleles predispose individuals carrying them to adverse effects of exposure to EDCs. The combination of genetic susceptibility and EDC exposure would appear to result in some sort of threshold risk being passed, leading to progeny carrying this birth defect. Results from laboratory animal models suggest a number of environmental chemicals that could be causal, such as persistent organohalogen pollutants (POPs) (Manson and Carr, 2003). Given this, it might have been predicted that the incidence of hypospadias would be high among Greenlanders, since their consumption of contaminated sea mammals make them one of the most POP-exposed human populations known (Giwercman et al., 2006). These authors noted, however, that of 11 076 boys born in Greenland between 1982 and 2002, only two cases of hypospadias were observed

296

Endocrine-disrupting chemicals in food

(incidence 0.02%; 95% CI: 0.002–0.06), as compared with more commonly reported rates between 0.2 and 0.7% of the population. The function of the androgen receptor is regulated by polymorphic repeats of CAG and GGN trinucleotide bases (Giwercman et al., 2006), and these authors reported that GGN = 23 is less frequent in patients with hypospadias than in the general population. Thus the observation that 85% of the Greenland population were carriers of GGN = 23 suggests this polymorphic variant could be helping to protect this population group from environmental insult, despite high exposure to POPs.

11.3

The potential of microarrays and related techniques for detection of effects of endocrine-disrupting chemicals

Several authors (e.g. Iguchi, 2006) point to the power of microarray methodologies (Ahmed, 2006; Jayapal and Melendez, 2006) to provide a mechanistic insight into receptor-mediated EDC effects in various animal species, including humans. The field of ecotoxicogenomics utilises information from transcriptomic analysis of gene expression, typically measured using microarrays, protein expression through proteomics, metabolite profiling (metabolonomics). The data are most powerfully used to identify patterns of gene expression in different tissues or organs of experimental animals exposed to those chemicals. The data need to be correlated through strong bioinformatic capabilities. Similar approaches can be used for nutrigenomics (Ferguson, 2006), and this field may be equally important for considering EDC effects on humans. Such approaches are already being used effectively, as shown by the following examples. There is a high incidence of precocious puberty in foreign children who have migrated to Belgium, and the plasma of this group appears to contain exceptionally high levels of a long-lasting dichlorodiphenyltrichloroethane (DDT, Fig. 11.1) residue (Parent et al., 2005), which is known to be estrogenic (Bitman et al., 1968). This observation has been used as evidence to suggest a potential role of EDCs in the early onset of puberty. Parent et al. (2005) exposed immature female rats to DDT, and were able to demonstrate experimentally an early onset of puberty in first generation offspring. They used a case-control methodology to compare the gene expression profile of hypothalamic hamartoma (a non-cancerous disorganised mass growing at the same rate as the surrounding tissue, but with the potential to disrupt normal function of the host organ) associated with precocious puberty with those of normal animals. The data enabled identification of gene networks responsible for both hamartoma-dependent sexual precocity, and the onset of normal human puberty. Tabuchi et al. (2006) studied the mechanism of action of bisphenol A (BPA) in the induction of cell injury in mouse testicular Sertoli TTE3 cells.

Genetics, epigenetics and genomic technologies

297

Genome-wide microarrays were used to analyse 22 690 probe sets. The 661 probe sets that were down-regulated and 604 probe sets that were upregulated by >2-fold were subjected to hierarchical gene network analyses, in order to identify nine important gene clusters. Two significant genetic networks that were down-regulated were associated with cell growth and proliferation, and the cell cycle. Two significant genetic networks that were upregulated included many basic-region leucine zipper transcription factors associated with cell death and DNA replication, recombination, and repair. Where specific questions are being asked of a single gene or group of genes, gene-specific real-time polymerase chain reaction (PCR) may be an appropriate method. For example, Hanafy et al. (2006) considered the effects of diethylstilbestrol (DES) and ethinylestradiol (Fig. 11.1) on gene transcription of very low-density apolipoprotein II (apoVLDL gene) in the liver of Japanese quail. This is one of the constituents of yolk in avian eggs, and it appears that liver-specific expression of the gene occurs in mature female birds during the egg-laying period. Although such gene expression does not normally occur in immature male birds, it can be stimulated by exogenous estrogens, and quantified using. Hanafy et al. (2006) were able to show that a single injection of two estrogenic compounds, DES and ethinylestradiol, up-regulated the expression of apoVLDL mRNA in the liver of 3-week-old, immature male quail. Ashworth et al. (2006) asked a more specific question as to timing of expression of porcine endometrial prostaglandin synthase during the estrous cycle and early pregnancy, and disruption of pregnancy by EDCs. Their rationale was that porcine trophoblast attachment to the uterine surface is associated with increased conceptus and endometrial production of prostaglandins. Although secretion of estrogen by the conceptus on Day 12 of gestation is important for the establishment of pregnancy; early exposure to estrogen-mimicking EDCs may lead to embryonic mortality. Their studies established the temporal and spatial pattern of endometrial prostaglandinendoperoxide synthase 1 (PTGS1) and PTGS2 expression during the estrous cycle and early pregnancy. Their results led them to suggest a new mechanism in which EDC disruption of pregnancy causes embryonic loss when it occurs during the implantation phase in the pig.

11.4

Gene expression as a component of screening methods for the detection of endocrine-disrupting chemicals in food and environment

Regulations in the US now require the testing of pesticides used in food crops and drinking water contaminants, for estrogenicity and other hormonal activities. In 1998 the Endocrine Disrupter Screening and Testing Advisory Committee (EDSTAC) recommending a two-tiered screening

298

Endocrine-disrupting chemicals in food

and testing strategy to identify potential EDCs (United States Environmental Protection Agency, 1998). The recommended initial stage provides a battery of tests designed to identify EDCs through their interactions with various hormonal systems, typically using in vitro assays, such as receptor binding and reporter gene assays. The group further recommended that all chemicals currently produced at or above 10 000 pounds (4536 kg) per year (approximately 15 000 chemicals) should be subjected to the first tier of screening. In a 1997 review, Ashby et al. commented on the challenge posed by EDCs, and again in 2000, Ashby made a strong plea for toxicologist to recognise the novel nature of the events being tested, the need to ask precise questions and to design experiments to answer those questions equivocally. Reporter assays that monitor effects of chemicals on expression of selected genes are effectively being used at an experimental level. For example, in studying the role of aryl hydrocarbon receptor (AhR) in female reproduction, Baba et al. (2005) used an in vitro reporter gene assay in association with in vivo chromatin immunoprecipitation assays. They found that AhR cooperates with an orphan nuclear receptor, Ad4BP/SF-1, thereby activating the expression of ovarian P450 aromatase (Cyp19), a key enzyme in estrogen synthesis, at least in ovarian granulosa cells. Administration to female mice of 9,10-dimethyl-1,2-benzanthracene (DMBA), an AhR ligand, leads to induction of ovarian Cyp19 gene expression, irrespective of the phase of the estrus cycle (Fig. 11.2). Similar reporter gene assays have been and continue to be developed, and suggested for use as more general screening tools. Xu et al. (2005) developed a human androgen receptor (hAR) assay, using the African monkey kidney cell line CV-1, transiently transfected with a reporter gene construct and a hAR expression plasmid. Positive and negative controls for the assay were the known AR agonist 5α-dihydrotestosterone (DHT), and AR antagonist flutamide (Fig. 11.1). DHT induced AR-mediated transcriptional activity in a concentration-dependent manner, while flutamide inhibited it (Kojima et al., 2003). BPA, 4-octylphenol (OP) and 4-nonylphenol (NP) (Fig. 11.2) were investigated for their agonistic and antagonistic activities using this assay. BPA showed significant inhibitory effects on the transcriptional activity induced by DHT, while OP and NP acted similarly as AR antagonists, albeit to a lesser extent. Satoh et al. (2004) developed a genetically engineered stable cell line, the AR-EcoScreenTM, from Chinese hamster ovary cells. These stably express a plasmid encoding an androgen receptor response element (derived from the rat prostate C3 gene) fused to a luciferase gene, along with a plasmid encoding the androgen receptor cDNA sequence (Kojima et al., 2003). Thus, they express the AR, along with an AR-responsive luciferase gene reporter, expression of which can be determined by the addition of luciferin with the subsequent measurement of light emission allowing quantification of AR activation. The assay was initially validated using 40 known

Genetics, epigenetics and genomic technologies

Estrogen

OH

299

EDCs

HO

XAP2

Arnt Hsp90

AhR

XAP2

p23

Hsp90

AhR

Ad4BP/SF-1

p23 XAP2

CYP19

Hsp90

Arnt AhR

p23

CYP19 expression

Ad4BP/SF-1

Fig. 11.2 EDC signalling through AhR. EDCs, including dioxin, 3methylcholanthrene and DMBA, bind to the AhR complex, triggering translocation of the complex into the nucleus. Arnt and Ad4BP/SF-1 are recruited to the complex which then binds to xenobiotic response element and Ad4 motifs in the promoter of the CYP19 gene, driving transcription. The resulting CYP19 protein is a key enzyme in estrogen synthesis, thus these EDCs lead to estrogen production via AhR signalling.

AR agonists and antagonists, with adequate performance. This was followed by a pilot study that considered 253 industrial chemicals for AR agonist and antagonist activities, and identified some novel activities. The ease and low cost of such assays was suggested to be important for wider screening purposes. Given the number of such screens now available, it may be appropriate to go back to the strong pleas made by Ashby and co-workers in their 1997 and 2000 papers. For example, a considerable number of the steps that were involved in the early development and validation of genotoxicity assays may not have been performed in the development of some of the screens that continue to appear for testing EDCs. Examples may include metabolic activation in in vitro assays, timing of exposure in in vivo assays, and a heirachical strategy for test prioritisation. Although cell lines utilised in reporter assays do not necessarily have the ability to metabolise chemicals, the experimental protocols do not always add an activation system. For in vivo assays, the detailed composition of the diet may itself profoundly affect the expression of regulatory pathways that impact estrogen disruption. Whereas

300

Endocrine-disrupting chemicals in food

classic toxicology utilises an indicator of acute toxic effects to provide a clear measure of a maximum tolerated dose, some of the genomically based studies utilise very much lower dose levels for which a negative response may simply reflect inappropriate dosages.

11.5

Modulation of gene expression by endocrine disrupters through epigenetic mechanisms

Ho et al. (2006) used a rodent model to associate susceptibility to prostate carcinogenesis with developmental exposure to estradiol and BPA. They provided transient exposure to low, environmentally relevant doses of BPA or estradiol in early stages of development, and demonstrated a significant increase in prostate gland susceptibility to adult-onset precancerous lesions and hormonal carcinogenesis. They found permanent alterations in the DNA methylation patterns of multiple cell signalling genes, suggesting an epigenetic basis for estrogen imprinting. For example, the phosphodiesterase type 4 variant 4 (PDE4D4) is an enzyme responsible for cyclic AMP breakdown. The authors identified a specific methylation cluster in the 5′flanking CpG island of this gene, that was gradually hypermethylated as a result of ageing in normal prostates, resulting in gradual loss of gene expression. However, after neonatal estradiol or BPA exposure, early and prolonged hypomethylation at this site led to continued, elevated expression of the PDE4D4 gene (Fig. 11.3). These changes in methylation pattern appeared before histopathologic changes of the gland were apparent. Thus, Ho et al. (2006), suggested that PDE4D4 could provide a candidate molecular marker for prostate cancer risk assessment by EDCs.

(a)

(c)

(b) PDE4D4

PDE4D4 gene expression

Ageing

PDE4D4

Estradiol / BPA

Reduced PDE4D4 gene expression

PDE4D4 Increased PDE4D4 gene expression

Fig. 11.3 Epigenetic regulation of the PDE4D4 gene by estradiol or EDCs in prostate tissue. (a) At birth, CpGs in the promoter region of the PDE4D4 gene are largely unmethylated (open circles) and the PDE4D4 gene is expressed. (b) With ageing, CpGs in the promoter are methylated (filled circles), which impairs gene expression by chromatin remodelling and prevention of transcription factor binding. (c) Estradiol or the EDC BPA can inhibit this methylation, resulting in PDE4D4 expression in adults, a hallmark of an immature prostate.

Genetics, epigenetics and genomic technologies

301

The synthetic estrogen DES is also a potent perinatal EDC, that has been extensively studied in rodents (McLachlan et al., 1975; Newbold et al., 2006). A mouse model has been used to document effects of exposure to DES during critical periods of reproductive tract differentiation, revealing multiple mechanisms playing a role in its carcinogenic and toxic effects. Analysis of the murine uterus following this exposure reveals altered gene expression pathways that include an estrogen-regulated component. Similarly, when developmental exposure to other EDCs are tested at equal estrogenic doses, this results in a similar increased incidence of uterine neoplasia. Perhaps even more importantly, this increased tumour susceptibility is passed on from the maternal lineage to subsequent generations of male and female descendants (transgenerational effects) involving both genetic and epigenetic events (Anway et al., 2005; Anway and Skinner, 2006; Crews and McLachlan, 2006). Anway et al. (2005) reported studies that confirmed trangenerational effects of two EDCs, the anti-androgenic fungicide, vinclozolin, or the estrogenic methoxychlor (Fig. 11.1) during transient exposure of a gestating female rat. Vinclozin acts transiently at the time of embryonic sex determination, to promote decreased spermatogenic capacity, as both reduced cell numbers and viability, and reduced fertility in F1 generation males. They also observed enhanced disease susceptibility in older animals. These phenotypic characteristics were transferred through the male germ line to all subsequent generations analysed (F1–F4). Transgenerational effects of environmental or dietary toxins require changes in the germ line, involving either a chromosomal or epigenetic alteration. In the studies described above, phenotypic effects were related to altered DNA methylation patterns in the germ line. The authors stressed, however, that this could be a secondary effect rather than a causal relationship, and more detailed molecular studies will be needed to confirm or disprove the role of epigenetics in transgenerational events.

11.6

Future trends

Deeper knowledge of mechanisms of action has revealed new and even more serious concerns about environmental and/or human exposures to EDCs than initially supposed. The ability of such chemicals to reprogramme the germ line and promote a transgenerational disease state has profound implications for evolutionary biology, since it will provide a far more rapid diversity in animal or human populations than has been seen in the past. It also has serious implications for disease aetiology, since it implies that behaviour that might have been thought appropriate to modify risk may not be sufficient to overcome transgenerational effects. Responding to these newer observations may require a paradigm shift in testing rationale and strategies, and regulatory thinking. For example, given that the effects of

302

Endocrine-disrupting chemicals in food

EDCs on methylation patterns of genes are proving to be important, it may be appropriate to utilise some of the newer technologies that probe the interaction between DNA methylation events, cis-acting elements and protein factors in the regulation of gene expression (Ferguson et al., 2007). ChIP-chip combines chromatin immunoprecipitation with microarrays, to permit genome-wide interrogation of the interactions between DNA and proteins. The application of these powerful new technologies will provide an essential tool to fully explore the manner in which EDCs affect the expression of genes. Only through a complete understanding of the mechanism by which these are working will it be possible to learn the most potent strategies to intervene in the action of EDCs.

11.7

Sources of further information and advice

A historic view of this field can be found in Ashby et al. (1997) and Ashby (2000). Reviews of epigenetics and transgeerational effects are provided by Anway and Skinner (2006), Crews and McLachlan (2006) and McLachlan et al. (2006).

11.8

Acknowledgements

Nutrigenomics New Zealand is a collaboration between AgResearch Ltd., Crop & Food Research, HortResearch and The University of Auckland, with funding through the Foundation for Research Science and Technology.

11.9

References

ahmed, f. e. (2006). ‘Microarray RNA transcriptional profiling: part I. Platforms, experimental design and standardization.’ Expert Review of Molecular Diagnostics 6(4): 535–50. anway, m. d. and m. k. skinner (2006). ‘Epigenetic transgenerational actions of endocrine disruptors.’ Endocrinology 147(6 Suppl): S43–9. anway, m. d., a. s. cupp, et al. (2005). ‘Epigenetic transgenerational actions of endocrine disruptors and male fertility.’ Science 308(5727): 1466–9. ashby, j. (2000). ‘The Gerhard Zbinden memorial lecture. Are environmental chemicals affecting the integrity or expression of the human genome?’ Toxicology Letters 112–113: 3–8. ashby, j., e. houthoff, et al. (1997). ‘The challenge posed by endocrine-disrupting chemicals.’ Environmental Health Perspectives 105(2): 164–9. ashworth, m. d., j. w. ross, et al. (2006). ‘Expression of porcine endometrial prostaglandin synthase during the estrous cycle and early pregnancy, and following endocrine disruption of pregnancy.’ Biology of Reproduction 74(6): 1007–15.

Genetics, epigenetics and genomic technologies

303

baba, t., j. mimura, et al. (2005). ‘Intrinsic function of the aryl hydrocarbon (dioxin) receptor as a key factor in female reproduction.’ Molecular & Cellular Biology 25(22): 10040–51. beleza-meireles, a., d. omrani, et al. (2006). ‘Polymorphisms of estrogen receptor beta gene are associated with hypospadias.’ Journal of Endocrinological Investigation 29(1): 5–10. beleza-meireles, a., m. barbaro, et al. (2007). ‘Studies of a co-chaperone of the androgen receptor, FKBP52, as candidate for hypospadias.’ Reproductive Biology & Endocrinology 5: 8. bitman, j., h. c. cecil, et al. (1968). ‘Estrogenic activity of o,p′-DDT in the mammalian uterus and avian oviduct.’ Science 162(851): 371–2. crews, d. and j. a. mclachlan (2006). ‘Epigenetics, evolution, endocrine disruption, health, and disease.’ Endocrinology 147(6 Suppl): S4–10. el houate, b., h. rouba, et al. (2007). ‘Novel mutations involving the INSL3 gene associated with cryptorchidism.’ Journal of Urology 177(5): 1947–51. ferguson, l. r. (2006). ‘Nutrigenomics: integrating genomic approaches into nutrition research.’ Molecular Diagnosis & Therapy 10(2): 101–8. ferguson, l. r., m. p. g. barnett, et al. (2007). ‘Influence of nutritional factors on genetic and epigenetic events relevant to chronic diseases.’ Agro Food Industry High-tech 3: 49–51. giwercman, y. l., k. e. kleist, et al. (2006). ‘Remarkably low incidence of hypospadias in Greenland despite high exposure to endocrine disrupters; possible protective effect of androgen receptor genotype.’ Pharmacogenetics & Genomics 16(5): 375–7. guillette, l. j., jr, t. s. gross, et al. (1994). ‘Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida.’ Environmental Health Perspectives 102(8): 680–8. hanafy, a. m., t. sasanami, et al. (2006). ‘Effects of diethylstilbestrol and ethinylestradiol on gene transcription of very low-density apolipoprotein II in the liver of Japanese quail, Coturnix japonica.’ Environmental Toxicology & Chemistry 25(5): 1354–9. henley, d. v. and k. s. korach (2006). ‘Endocrine-disrupting chemicals use distinct mechanisms of action to modulate endocrine system function.’ Endocrinology 147(6 Suppl): S25–32. ho, s. m., w. y. tang, et al. (2006). ‘Developmental exposure to estradiol and bisphenol A increases susceptibility to prostate carcinogenesis and epigenetically regulates phosphodiesterase type 4 variant 4.’ Cancer Research 66(11): 5624–32. hughes, i. a., h. martin, et al. (2006). ‘Genetic mechanisms of fetal male undermasculinization: a background to the role of endocrine disruptors.’ Environmental Research 100(1): 44–9. iguchi, t. (2006). ‘[Importance of development of ecotoxicogenomics in understanding molecular mechanisms of chemicals in developing animals].’ Nippon Eiseigaku Zasshi – Japanese Journal of Hygiene 61(1): 11–18. jayapal, m. and a. j. melendez (2006). ‘DNA microarray technology for target identification and validation.’ Clinical & Experimental Pharmacology & Physiology 33(5–6): 496–503. knight, d. c. and j. a. eden (1995). ‘Phytoestrogens – a short review.’ Maturitas 22(3): 167–75. kojima, h., m. iida, et al. (2003). ‘Effects of a diphenyl ether-type herbicide, chlornitrofen, and its amino derivative on androgen and estrogen receptor activities.’ Environmental Health Perspectives 111(4): 497–502.

304

Endocrine-disrupting chemicals in food

malausa, t., m. salles, et al. (2006). ‘Within-species variability of the response to 20-hydroxyecdysone in peach-potato aphid (Myzus persicae Sulzer).’ Journal of Insect Physiology 52(5): 480–6. manson, j. m. and m. c. carr (2003). ‘Molecular epidemiology of hypospadias: review of genetic and environmental risk factors.’ Birth Defects Research 67(10): 825–36. mclachlan, j. a., r. r. newbold, et al. (1975). ‘Reproductive tract lesions in male mice exposed prenatally to diethylstilbestrol.’ Science 190(4218): 991–2. mclachlan, j. a., e. simpson, et al. (2006). ‘Endocrine disrupters and female reproductive health.’ Best Practice & Research Clinical Endocrinology & Metabolism 20(1): 63–75. newbold, r. r., e. padilla-banks, et al. (2006). ‘Adverse effects of the model environmental estrogen diethylstilbestrol are transmitted to subsequent generations.’ Endocrinology 147(6 Suppl): S11–17. parent, a. s., g. rasier, et al. (2005). ‘Early onset of puberty: tracking genetic and environmental factors.’ Hormone Research 64(Suppl 2): 41–7. radpour, r., m. rezaee, et al. (2007). ‘Association of long polyglycine tracts (GGN repeats) in exon 1 of the androgen receptor gene with cryptorchidism and penile hypospadias in Iranian patients.’ Journal of Andrology 28(1): 164–9. rakitsky, v. n., v. a. koblyakov, et al. (2000). ‘Nongenotoxic (epigenetic) carcinogens: pesticides as an example. A critical review.’ Teratogenesis, Carcinogenesis, & Mutagenesis 20(4): 229–40. rudel, r. (1997). ‘Predicting health effects of exposures to compounds with estrogenic activity: methodological issues.’ Environmental Health Perspectives 105(Suppl 3): 655–63. satoh, k., k. ohyama, et al. (2004). ‘Study on anti-androgenic effects of bisphenol a diglycidyl ether (BADGE), bisphenol F diglycidyl ether (BFDGE) and their derivatives using cells stably transfected with human androgen receptor, AREcoScreen.’ Food & Chemical Toxicology 42(6): 983–93. shaw, i. c. and s. mccully (2002). ‘The potential impact of dietary endocrine disrupters on the consumer.’ The International Journal of Food Science and Technology 37: 6. soneda, s., m. fukami, et al. (2005). ‘Association of micropenis with Pro185Ala polymorphism of the gene for aryl hydrocarbon receptor repressor involved in dioxin signaling.’ Endocrine Journal 52(1): 83–8. sumpter, j. p. and s. jobling (1995). ‘Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment.’ Environmental Health Perspectives 103(Suppl 7): 173–8. tabuchi, y., i. takasaki, et al. (2006). ‘Identification of genetic networks involved in the cell injury accompanying endoplasmic reticulum stress induced by bisphenol A in testicular Sertoli cells.’ Biochemical & Biophysical Research Communications 345(3): 1044–50. thomson, b. m., p. j. cressey, et al. (2003). ‘Dietary exposure to xenoestrogens in New Zealand.’ Journal of Environmental Monitoring 5(2): 229–35. toppari, j. and m. kaleva (1999). ‘Maldescendus testis.’ Hormone Research 51(6): 261–9. toppari, j., j. c. larsen, et al. (1996). ‘Male reproductive health and environmental xenoestrogens.’ Environmental Health Perspectives 104(Suppl 4): 741–803. trubo, r. (2005). ‘Endocrine-disrupting chemicals probed as potential pathways to illness.’ Journal of the American Medical Association 294(3): 291–3. wang, t. t., n. sathyamoorthy, et al. (1996). ‘Molecular effects of genistein on estrogen receptor mediated pathways.’ Carcinogenesis 17(2): 271–5. wilkin, p., j. o. metcalfe, et al. (1979). ‘Hypospadias: a review.’ Canadian Journal of Surgery 22(6): 532–7.

Genetics, epigenetics and genomic technologies

305

xu, l. c., h. sun, et al. (2005). ‘Evaluation of androgen receptor transcriptional activities of bisphenol A, octylphenol and nonylphenol in vitro.’ Toxicology 216(2–3): 197–203. united states environmental protection agency (1998). ‘Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) Final Report (http: www. epa.gov/endo/pubs/edspoverview/finalrpt.htm), accessed October 2008. yoshida, r., m. fukami, et al. (2005). ‘Association of cryptorchidism with a specific haplotype of the estrogen receptor alpha gene: implication for the susceptibility to estrogenic environmental endocrine disruptors.’ Journal of Clinical Endocrinology & Metabolism 90(8): 4716–21.

12 Computer-aided methodologies to predict endocrine-disrupting potency of chemicals A. Roncaglioni and E. Benfenati, Istituto di Ricerche Farmacologiche ‘Mario Negri’, Italy

Abstract: This chapter focuses on a series of techniques, based on computational studies of chemical compounds, which can play an important role in predicting the endocrine-disrupting potency of chemicals. There are many advantages related to the application of computational methods: they are cheaper and faster than experimental studies, and do not require animal use. For these reasons computeraided methodologies have been applied to endocrine disrupters to fill the data gap and lack of knowledge existing on this emerging issue. The major computational techniques used to model the endocrine-disrupting potency of chemical compounds will be introduced (quantitative structure–activity relationship – QSAR – 3D-QSAR and virtual docking) along with practical examples of the results obtained so far in this field. Key words: quantitative structure–activity relationship, 3D-QSAR, docking, endocrine disrupters, estrogen receptor.

12.1

Introduction

A series of techniques based on computational studies (termed in silico [derived from the Latin in silicon] methods) of chemical compounds can play an important role in predicting the endocrine-disrupting potency of chemicals. There are a number of advantages related to the application of these techniques; they are cheaper and faster than experimental studies, so they can be employed to prioritise compounds which can benefit the most from experimental testing; they can also be applied in a pre-synthesis stage to show where less harmful solutions can be produced by synthesis. These methods are already accepted and used in the USA, Canada, Japan and some EU countries at a national level to replace experimental

The endocrine-disrupting potency of a chemical

307

testing in a regulatory context for a series of end points, but they are receiving more and more attention nowadays due to their application as part of the innovative REACH EC Regulation No. 1907/2006 on the regulation of industrial chemicals, which supports an increased use of in silico and other non-animal testing methods in order to reduce costs and the use of animals for experimental testing to assess the safety of chemicals. For all these reasons, in silico tools have also been applied to endocrine disrupters (EDs) to search for a solution to fill the data gap that exists for this emerging issue. The assumption behind these methods is that the chemical structure of the compounds suspected to be EDs can implicitly encode the characteristics responsible for their effects. Computational chemistry methods investigate this hypothesis by exploring this idea from different perspectives. In particular three methods and their respective outcomes will be presented here: quantitative structure–activity relationships (QSAR), 3D-QSAR and the virtual docking approach.

12.2

In silico methods to predict the endocrine-disrupting potency of a chemical

From a theoretical point of view, computational chemistry methods deal with the chemical composition of molecules, in particular by computing the chemical properties that are important for describing a specific phenomenon. Each method has a distinct perspective: QSAR deals with small molecular structures, ignoring the biological side of the interaction, 3D-QSAR focuses its attention on the comparison of molecular spatial organization, and virtual docking includes the receptor side of the interaction in its analysis.

12.2.1 Quantitative structure–activity relationship Since the nineteenth century, many attempts have been made to investigate whether some properties of chemical compounds could be described on the basis of their chemical structure (Kubinyi, 2002). This finding was better formalised in the 1970s by Corvin Hansch who described the fundamentals of modern QSAR (Schultz et al., 2003). Hansch’s paradigm was based on the observation that, for congeneric series of chemicals, some observed properties were dependent upon an equation describing substituents’ contributions to the parent compound with regard to hydrophobic, electronic and steric effects (Hansch, 1969). McFarland then provided a biological interpretation of this equation by relating hydrophobic characteristics to the compound’s ability to penetrate the biosystem, and steric and electrostatic characteristics to its capacity for interacting with the target site of the reaction (McFarland, 1970).

308

Endocrine-disrupting chemicals in food

Molecular descriptors The first requirement for a QSAR analysis is to obtain a matrix of descriptors, starting from the chemical structure of the compounds. Recently, to remove the requirement for congeneric series, a wide variety of molecular descriptors have been suggested by a number of researchers. Molecular descriptors are numerical parameters for evaluating chemical structures, representing some of their specific characteristics. Although a wide variety of descriptors exist, and other sources describe in more detail their particular characteristics (Todeschini and Consonni, 2000), brief descriptions of the most common classes of descriptors commonly utilised in QSAR are reported in Table 12.1. One of the aspects of all these descriptors that merits specific attention is the level of structural information used to calculate the descriptors. It is common for molecular structures to be stored in several file formats, which support different levels of information. For instance SMILES (Simplified Molecular Input Line Entry System) notation (Daylight, 2007) is a character string format often used to encode large databases of molecules since only one line of ASCII characters, representing atom and bond symbols, is required to describe a single molecule. This condensed way of representing chemical structures is able to maintain information about the connectivity and bond types between atoms. Some molecular descriptors, e.g. topological or constitutional, can be calculated from even this very simplified molecular representation. Other types of descriptors, e.g. geometrical, require the chemical structure to be known in its three-dimensional (3D) shape. However, each molecule exists in multiple conformations whose stability and occurrence may vary. The commonly accepted strategy for coping with the conformation issue is to use the most energetically stable conformation as a reference structure to calculate the molecular descriptors, even if it does not necessarily represent the bioactive conformation adopted by a molecule to interact with the biological environment. To obtain the minimum energetic conformation, the entire conformational space has to be sampled using a conformational search technique and the global minimum of the structure is then identified by optimising the geometry through common force field, semi-empirical or ab initio methods (Leach, 2001). Of course, it is more complex to calculate 3D descriptors, but on the other hand they contain valuable information about important molecular characteristics that cannot be properly described on the basis of the bi-dimensional structure alone. Modelling step The principal step in any QSAR analysis is to find the best function for describing a data set made up of n compounds, characterised by m molecular descriptors, to a given property A, whose values for the compounds under investigation are known, as shown in Figure 12.1. This step includes two sub-tasks: (1) descriptor selection to include only relevant variables and

The endocrine-disrupting potency of a chemical

309

Table 12.1 A brief description of the major classes of descriptor types Descriptor types

Category description

Constitutional

They reflect only the molecular composition of the compound, without using connectivity or geometry of the structure. Topological indices describe the atomic connectivity in the molecule. Molecular descriptors calculated as information content of molecules. They reflect the characteristics of charge distribution in the molecule. These descriptors require 3D-coordinates of the atom placement for a stable conformer of a given molecule and on this basis describe molecular shape and volume. They involve descriptors related to energy evaluation of the molecule, based on the results of quantummechanical calculations that permitted to obtain the optimised chemical structure (e.g. LUMO– lowest unoccupied molecular orbital–energy). They include a variety of atom-centred fragments, counts for functional groups and also fingerprint type descriptors constituted by a string of bit encoding for the presence or absence of some predefined fragments. They can include several kinds of experimentally calculated physicochemical properties (e.g. boiling point) but often they are computed by specific software. They reflect some properties of chemical compounds and among them lipophilicity is the most widely used. It is a partition coefficient referring to the partitioning equilibrium of a substance between octanol and the aqueous phase.

Topological Information indexes Electrostatic Geometrical

Quantum-mechanic

Fragments-based descriptors

Physicochemical properties

(2) the application of specific algorithms to find the relationship between the variables and the target property. Thousands of descriptors can easily be computed on molecular structures, but their inclusion in a model can be questionable. Many descriptors can contain redundant information (i.e. they are correlated) or simply introduce noise (i.e. the characteristics they describe are not correlated with the activity to be modelled). Furthermore, including a large number of variables increases the occurrence of chance correlation, leading to the problem of over-fitting. This aspect will be discussed further in the next section. The selection of variables can be hypothesis-driven, including a limited number of variables considered by a human expert to be relevant for modelling the end point, or statistically driven, using mathematical algorithms to search for the most significant solutions. In the latter case, a series of methods (Guyon and Elisseeff, 2003) including multivariate data analysis can be used. For instance, principal components analysis (PCA) can be used

310

Endocrine-disrupting chemicals in food

H 3C

N

O OH 3D

2D

H3C

N O OH

D(n,m)

Activity Compounds (1,..., n)

Compounds (1,..., n)

Descriptors (1,..., m)

A

A = f (D(n,m))

Fig. 12.1 A schematic view of the QSAR process. Molecular structures are encoded in different descriptors describing different properties of the molecule (bidimensional or three-dimensional ones, the latter requiring molecular geometries being optimized). Each descriptor is a column in the matrix having as many rows as the compounds constituting the data set. A QSAR model is searching the best relationship between some of the descriptors in the D matrix and the activities contained in the A column.

to represent many descriptors in a few orthogonal latent variables (forming the principal components) made up from linear combinations of the original descriptors. Variables can be also be iteratively included or excluded using a stepwise approach. If there are too many initial variables, however, these methods do not efficiently explore all possible combinations and more sophisticated tools are required. Nowadays, genetic algorithms (GA) are often considered to be some of the most promising algorithms. They are based on the Darwinian evolutionary theory. Each variable is encoded in a string of binary numbers to indicate the presence or absence of that variable in the model. The best individuals in a population of models are crossed over, merged, mutated, then iteratively evaluated against a fitness function which gives a statistical evaluation of the model’s performances (Devillers, 1996a).

The endocrine-disrupting potency of a chemical

311

Several algorithms can be used to derive those mathematical functions that are better suited for describing the data trends. Preference can be given to different methods, based on the specific data distribution to be described. In particular, the type of model may vary depending on the property to be modelled: for continuous variables, regression methods are used, while for categorical ones (i.e. activity classes), several classification methods are available (Burkard, 2004). With the increased speed in computational calculation over the last few decades, the variety of algorithms for developing new models has rapidly increased and more sophisticated non-linear methods have been introduced, such as neural networks (NN) (Devillers, 1996b). NN identify the best variables related to the property to be modelled, using a series of ‘attempts’ in which these variables are combined in a non-linear way. Validation procedures Whatever the technique chosen to model a specific data set, one of the most important issues in the QSAR field is model validation. The function derived from the data set matrix (commonly known as the training set) can describe the trend of the activity under investigation quite well, and this is referred to as the fitting abilities of the model. By increasing the number of initial variables or increasing the algorithm complexity, it becomes possible to obtain a function that more precisely describes the data distribution (i.e. has a lower error). However, using many variables can create the particular risk known as the problem of over-fitting. A graphical representation of the over-fitting problem, shown in Fig. 12.2, can better explain this situation. Imagine that two different models can be derived from a training set of six compounds. The linear model reaches a lower accuracy in the fitting, indicated by the squared correlation coefficient R2, compared with the polynomial model. If the two models are now used to predict activity values for two new test compounds, it appears that the most complex model, with a better fitting, also has higher errors in predicting activity for these compounds. This happens because the increased precision in the fitting does not produce a more specific description of the data trend, since the noise or variability that each data set commonly contains is also being modelled. The essential purpose of validation procedures is to ensure that besides having a reasonable goodness-of-fit, the model is statistically robust, which is achieved by using internal and external validation methods (Hawkins, 2004). Internal validation relies on a statistical evaluation of the model’s performances, based on using perturbed training sets for assessing robustness and chance correlation. The most common method is the cross-validation procedure, where one or more compounds are iteratively excluded from the training set and predicted from the model developed with the reduced set. The cross-validated performances thus obtained need to be very similar to those obtained from fitting.

312

Endocrine-disrupting chemicals in food

Experimental activity

10 Training set Test compounds

8

y = 0.979x + 0.344

6

R 2 = 0.956 4

y = –0.062 x 4 + 1.293 x 3 – 9.472 x 2 + 29.24 x – 27.37 R 2 = 0.999

2 0 0

2

4 6 8 Descriptor D1

10

Fig. 12.2 Graphical representation of the problem of over-fitting. The model described with the dotted line although less precise in describing the training set trend, is more predictive on the test compounds than the more complex model indicated with the continuous line, affected by over-fitting.

There is still a debate about whether internal validation procedures are enough to assess model performances or if they are, at least, too optimistic a measure (Golbraikh and Tropsha, 2002). In general, they are acceptable if the data set is made up of a smallish number of compounds. For a data set of hundreds of compounds, an external test set or at least the leavemany-out approach is preferable. External validation uses new compounds, not used to develop the model, as a blind test set. When using this approach, one should ensure that these new compounds are not randomly chosen but selected as being representative of the training set’s chemical composition (Golbraikh et al., 2003; Tropsha et al., 2003). Overall, the validation operation should ensure that the model is statistically significant, reliable and robust to noise and data perturbation, and maintains its validity when the relationship is extrapolated to compounds sharing similarity with the training data, at least within a defined chemical space. The validity of a model is therefore judged on the basis of a number of factors, which rely on a wide variety of statistical parameters.

12.2.2 3D quantitative structure–activity relationship The 3D-QSAR method is an extension of the classical QSAR approach and is distinguished by the type of descriptors used to develop the model. Molecular features are based on the concept of molecular interaction fields (MIF) (Cruciani, 2006), and CoMFA (comparative molecular field analysis) is the most popular method used (Cramer et al., 1988). Field-based descriptors are calculated by placing the molecule within a lattice and calculating its interaction energies – mainly steric and electrostatic ones – with a probe (e.g. a sp3 carbon atom) at each point of the grid as shown in Fig. 12.3.

The endocrine-disrupting potency of a chemical

St1 St2

Stm El1 El2

313

Elm

Mol 1 Mol 2 Mol 1

Mol n

Fig. 12.3 3D-QSAR descriptors are calculated by placing appropriately overlapped molecules within a lattice and calculating with a probe the steric and electrostatic field contribution of the molecule in that point. Each matrix column contains the contribution to the field for each molecule in a specific grid point.

The descriptor matrix thus obtained is then used as in standard QSAR analysis. Owing to the large number of descriptors calculated, the method most commonly used is partial least squares (PLS), a regression method able to handle collinear variables, which generates ‘latent variables’, orthogonal linear combinations of the original set of descriptors. An essential task in 3D-QSAR is the alignment of the molecules forming the data set, because the position of a molecule in the grid influences the field values. For this reason, various methodologies of superposition can be adopted based on a common skeleton superposition, evaluating docking or crystallographic information, or starting from a pharmacophoric hypothesis. The results of the modelling step also provide a graphical representation of the areas surrounding the molecule that exhibit a positive or negative steric or electrostatic contribution in modulating the target properties in a regression map. This information cannot be obtained with QSAR methods, which deal with the chemical entity as a single input, without local information on the contributions of parts of the molecule. Compared with QSAR methods, 3D-QSAR has a more detailed description of the chemical structure, can handle chirality (typically not evaluated in QSAR) and the results can be related to specific parts of the molecule which exhibit some sterical or electrostatic behaviour. However, 3D-QSAR models typically work on compounds with a similar skeleton, and are more affected by manual optimisation. The issue of validation discussed previously applies to 3D-QSAR models too. In particular, there is a need to demonstrate the statistical robustness of the model, showing its prediction capability.

314

Endocrine-disrupting chemicals in food

12.2.3 Virtual docking Virtual docking computationally explore the potential matching of two molecules, usually a protein and another macromolecule (protein or DNA) or a small molecule (ligand). In the context of EDs this method is particularly useful to model the interaction of small molecular ligands with the binding site of nuclear receptors. Virtual docking relies on the 3D structure of the protein that has to be solved with X-ray crystallography or in alternative with NMR spectroscopy or by homology modelling. Then, a series of poses have to be selected among all possible combinations of protein-ligand complexes by exploring the conformational space. For these poses, which describe the relative positions of the ligand and receptor, the complementarity between the two molecules (see Fig. 12.4) is explored in terms of their matching not only from a steric point of view but also when considering electrostatic or other energy interactions (e.g. solvation effect). This evaluation is done using scoring functions that are approximate estimations of the binding energy of the ligand–receptor complex (Oprea and Marshall, 1998). Although scoring functions can produce a fast estimation of the interaction energy, they are not accurate enough to be quantitatively correlated with the binding energy obtained experimentally in vitro. However it is now feasible to have a larger number of ligands exhibiting binding among the highest ranked complexes. In this way it is possible to evaluate an enrichment compared with the pool of ligands tested virtually (Warren et al., 2006). This approach is often used in the drug discovery process to identify new lead compounds.

Fig. 12.4 Mutual recognition process between a protein and some possible ligands. The precision of the matching is a measure of the binding strength. Arrows indicate binding.

The endocrine-disrupting potency of a chemical

315

Compared with QSAR methods, docking studies are much more closely related to the biochemical process they want to address. Vice versa, QSAR models can implicitly include information resulting from multiple processes including, for instance, metabolism.

12.3

Results and implications

The endocrine-disrupting process includes a number of possible pathways and action mechanisms. The possibility of investigating this issue with in silico methods is closely related to the mechanistic knowledge gained in related research fields, since the modelling step itself relies on the available experimental data. For this reason the majority, if not all, of the studies on EDs on a broad scale were focused on chemical compounds directly interacting with receptors, in particular nuclear receptors (NRs). Owing to the availability of a larger amount of data for the estrogen receptor (ER), this was one of the most studied targets. A deeper understanding of ER-mediated processes is also available from a number of experimental findings in multiple sectors, since it has been considered an interesting target from a pharmaceutical point of view in the treatment of estrogendependent tumours, e.g. breast cancer. As computational chemistry methods are a valid complement to others in drug discovery, a number of studies have explored these methods, especially for pharmaceutical purposes. In particular, QSAR and 3D-QSAR were used to investigate activity trends in groups of compounds sharing a common moiety (Gao et al., 1999; Salum et al., 2007). Similarly QSAR was used to study some characteristics such as the selectivity for the alpha or beta subtype for specific therapeutic classes (Mukherjee et al., 2005). Docking finds an application in protein-based drug design to explore the most favourable protein ligands interactions for the refinement of lead drug compounds (Sarkhel et al., 2003). Virtual screening has also been tested to evaluate the enrichment factor in large databases of potential drugs (Bissantz et al., 2000). In this context, docking is effective in drug design, since it gives better performances than the random selection of possible hits, as indicated by the enrichment factor. As soon as endocrine-disrupting chemicals became an emerging and alarming issue, attention on in silico methods able to prioritise chemical compounds needing deeper in vivo and in vitro investigations definitely increased. In addition, among the chemical groups requiring special regulatory attention, the REACH legislation mentions those affecting the endocrine and reproductive systems. Perhaps a better structured example of the in silico approach to the study of EDs is the US Environmental Protection Agency (US EPA) endocrine-disrupting screening programme (Tong et al.,

316

Endocrine-disrupting chemicals in food

2003). It focused on the development of an in silico pre-screening tool, to be applied first in a two-tiered approach. A data set of 200 chemical contaminants of heterogeneous nature was experimentally assessed in vitro for their binding activity against ER. Based on this data set, a four-phase system of priority setting was developed, relying on a filtering approach (for the first stage), a collection of binary classification models to discriminate active or inactive compounds (in the second phase), leading in the third phase to a quantitative prediction of the binding with a CoMFA model. The last phase, following on hierarchically from the previous ones, considered more specifically some chemical characteristics responsible for the activity in a rule-based way (Tong et al., 2003). Even if the entire architecture is rather complex, it is clear that its goal is to reduce false negatives, as expected from a pre-screening method, without needing to pass all compounds to the higher tier of in vitro or in vivo testing. The combination of different modelling methods enables a wide variety of ways to encode the chemical content (structural alerts, 2D descriptors, CoMFA descriptors) to be explored, without missing any relevant structural information. This approach, or at least some of the individual models constituting the complete framework, has also been validated and verified with other data about ER activity available in the literature. It is interesting to note that the models developed for regulatory purposes have a target that is the opposite to that of the previously mentioned models used for drug screening. Industry definitely wants to avoid false positives, i.e. compounds which are predicted as active but whose results prove them not to be good candidates for drugs. Vice versa, regulators want to avoid evaluating toxic compounds as safe (false negative predictions). Although lacking the comprehensive view elaborated by Tong et al., many other studies have addressed the issue of EDs. As a rule, classical QSAR has often been compared to 3D-QSAR (Waller 2004; Brown et al., 2005). A general trend that was observed in these cases was that, by using rather large (several hundreds of compounds) and chemically heterogeneous data sets, results based on a classical QSAR approach gave similar performances to those from using 3D-QSAR. Moreover, the inclusion of descriptors computed on the bi-dimensional structure alone or relying on chemical fragments also seemed to be enough to obtain models with comparable performances. Some authors also explored the use of multiple, energetically reasonable conformations for each chemical to overcome the limitations of using a single 3D global minimum as a reference structure to calculate descriptors (Schmieder et al., 2003). The distribution in the population of values for the descriptors calculated on the active compounds was then compared with that obtained for the inactive compounds and used to derive a classification model. Some attempts have been made to improve the regulatory acceptance of in silico models, particularly by starting to address the question of defining the limits for the application of the models in terms of chemical space

The endocrine-disrupting potency of a chemical

317

coverage and comparing them with existing inventories of industrial chemical compounds (Tong et al., 2004; Netzeva et al., 2006; Liu et al., 2007). Virtual screening based on docking methods has been employed less often in the analysis of EDs. One reason could be that compared with its use in drug design, where the enrichment factor demonstrates better performances compared with random prediction, this type of outcome cannot be satisfactory enough if used as pre-screening tool, leading to too many false negatives. Yoon and Welsh (2004) explored the opportunity of using docking as a prioritisation tool with a large set of heterogeneous compounds, investigating the enrichment factors. Interestingly, to account for receptor flexibility, a subset of receptor crystal structures was used in parallel in the docking process and this approach increased the precision gained compared with the use of a single complex. Even if computationally quite demanding, it allows very large databases of compounds to be screened once the method is set up. A similar approach has been used to screen an existing database of food contaminants, which are or may be present in food. A combination of receptor conformations was used, which improved the overall results (Lo Piparo et al., 2006). An interesting advantage is that some in silico methods (for instance docking studies) in some cases clearly relate to a given biochemical mechanism, which can be explored and understood in a convincing way. However, these techniques reflect the basis of the data they use. This should be considered in the case of endocrine-disrupting chemicals. Since, so far, the data are mainly on partial specific steps of the overall path which leads to the endocrine disruption effect, these techniques are necessarily partial.

12.4

Future trends

The demand for fast, inexpensive, animal-friendly methods to explore the properties of chemical compounds is continuously increasing. Although, in the past, in silico approaches were often considered to lead to less authoritative outcomes compared with other methods, the recent efforts to improve model reliability and transparency have gradually modified this feeling and attracted the attention of those searching for alternatives. A further stimulus was produced by the interest in these tools from a regulatory point of view like that of REACH or the 7th Amendment to the Cosmetics Directive (Council Directive 2003/15/EC), requiring the animal testing of cosmetic ingredients to be abandoned in a few years. A number of directions can be predicted for the better exploitation of in silico models applied in the specific context of EDs. To have a more comprehensive view of the problem, a battery of models investigating the remaining NRs has to be developed. In the literature, at the moment the androgen receptor (AR) is the second most studied target, although some research groups have projects to extend this approach to other receptors

318

Endocrine-disrupting chemicals in food

to deal with the interaction of exogenous substances with other targets, causing potentially dangerous effects (Vedani et al., 2005). A critical step for advances in this direction would be to access reliable and numerous experimental data assessing the binding for the different NRs. This is the primary drawback of environmentally oriented studies compared to pharmacological ones, since resources are more limited while the chemical diversity of the environmentally relevant chemical space is quite broad. To modify this trend and efficiently exploit the use of in silico methods, crossfertilisation with other experts in a multidisciplinary approach will be essential. In particular, it is now recognised that different approaches to assess toxicity show a high degree of complementarities which, if further explored, can lead to improvements. A rationale for their integration can lead to the development of intelligent testing strategies, reducing the need for animal testing. However, the use of binding activity data or the results of other in vitro assays enable studies of specific mechanisms of endocrine disruption, but are a long way from providing a complete explanation for the biologically relevant effects that can be observed in a live animal or even at the population level. The advantage of the QSAR method is that when experimentally determined in vivo data become available, in silico models can be used to investigate these end points, having an increasing complexity but also a more realistic outcome (e.g. in vivo uterotrophic assay). The modelling of global toxic end points has already obtained reasonable results in other fields, such as the study of acute toxicity or the carcinogenic process (Cronin and Livingstone, 2004). More effort now needs to be applied to QSAR, and especially to 3DQSAR and virtual docking, to extend their degree of application by moving from models being developed for pure scientific investigation, to something more directly useful for practical application and more user-oriented in the field of regulation.

12.5

Sources of further information and advice

Besides the scientific literature quoted in the references, nowadays the World Wide Web is surely the easiest source for accessing the most up-todate information on a wide range of topics related to the in silico field. For those interested in the innovations introduced by REACH legislation, the EU website is the most up-to-date source of information (http://ec. europa.eu/environment/chemicals/reach/reach_intro.htm; http://ec.europa. eu/enterprise/reach/index_en.htm). The European Chemical Bureau (http://ecb.jrc.europa.eu), part of the European Research body, is particularly active in the field of computational toxicology and its application to REACH legislation. Several documents and some free software dealing with a broad range of topics related to the use of QSAR for regulatory purposes are freely accessible from this site.

The endocrine-disrupting potency of a chemical

319

The Computational Chemistry List (http://www.ccl.net/) is the most authoritative on-line resource for anyone interested in computational chemistry. Various on-line resources are devoted to molecular descriptors (www.moleculardescriptors.eu, http://qsar.sourceforge.net/dicts/qsardescriptors/index.xhtml) including websites where it is possible to compute some of them for free (www.vcclab.org). The branch of chemistry devoted to the elaboration and modelling of raw data is commonly known as chemometrics. A freely available electronic manual explaining a wide range of algorithms to derive mathematical models and statistical principles is available on the internet (http://www. statsoft.com/textbook/stathome.html). Along with commercial software, some open-source programs implementing statistical algorithms are available, such as the R Project for Statistical Computing (http://www.r-project. org/index.html) and the Waikato Environment for Knowledge Analysis (WEKA, http://www.cs.waikato.ac.nz/ml/weka/). Among the resources that are more focused on the in silico treatment of EDs, more information on the US EPA endocrine-disruptors screening programme, the priority setting strategies and the developed QSAR models can be found on the Center for Toxicoinformatics of the FDA’s National Center for Toxicological Research website, particularly in the Endocrine Disruptor Knowledge Base (http://edkb.fda.gov/index.html) or in the Distributed Structure-Searchable Toxicity (DSSTox) Public Database Network (http://www.epa.gov/ncct/dsstox/sdf_nctrer.html). The Danish (Q)SAR Database (http://130.226.165.14/index.html) also contains predicted values for some EDs’ related end points such as ER and AR binding affinity.

12.6

Acknowledgement

We thank the European Commission for funding (project CASCADE NoE, FOOD-CT-2004-506319).

12.7

References

bissantz c, folkers g, rognan d (2000), ‘Protein-based virtual screening of chemical databases. 1. Evaluation of different docking/scoring combinations’, J Med Chem, 43 (25), 4759–4767. doi: 10.1021/jm001044l brown n, mckay b, gasteiger j (2005), ‘Fingal: a novel approach to geometric fingerprinting and a comparative study of its application to 3D-QSAR modelling’, QSAR Comb Sci, 24 (4), 480–484. doi: 10.1002/qsar.200430923 burkard u (2004), ‘Methods for data analysis’, in Gasteiger J and Engel T, Chemoinformatics, Weinheim, Wiley-VCH, 439–485. doi: 10.1002/3527601643 cramer r d, patterson d e, bunce j d (1988), ‘Comparative molecular field analysis (CoMFA). 1. Effect of shape on binding of steroids to carrier proteins’, J Am Chem Soc, 110, 5959–5967. doi: 10.1021/ja00226a005

320

Endocrine-disrupting chemicals in food

cronin m t d and livingstone d j (2004), Predicting Chemical Toxicity and Fate, Boca Raton, FL, CRC Press. cruciani g (2006), Molecular Interaction Fields: Applications in Drug Discovery and ADME Prediction, Weinheim, Wiley-VCH. doi: 10.1002/3527607676 daylight (2007) Daylight Theory Manual Version 4.9, ‘3. SMILES – A Simplified Chemical Language’, 5–19. Daylight Chemical Information Systems, Inc., http:// www.daylight.com/dayhtml/doc/theory/index.pdf. devillers j (1996a), Genetic Algorithms in Molecular Modeling (Principles of QSAR and Drug Design), London, Academic Press. devillers j (1996b), Neural Networks in QSAR and Drug Design, London, Academic Press. gao h, katzenellenbogen j a, garg r, hansch c (1999), ‘Comparative QSAR analysis of estrogen receptor ligands’, Chem Rev, 99 (3),723–744. doi: 10.1021/cr980018g golbraikh a and tropsha a (2002), ‘Beware of q2!’, J Mol Graph Model, 20, 269–276. doi:10.1016/S1093-3263(01)00123-1 golbraikh a, shen m, xiao z, xiao y d, lee k h, tropsha a (2003), ‘Rational selection of training and test sets for the development of validated QSAR models’, J Comput Aid Mol Des, 17, 241–253.doi: 10.1023/A:1025386326946 guyon i and elisseeff a (2003), ‘An introduction to variable and feature selection’, J Mach Learn Res, 3, 1157–1182. hansch c (1969), ‘Quantitative approach to biochemical structure–activity relationships’, Accounts Chem Res, 2, 232–239. doi: 10.1021/ar50020a002 hawkins d m (2004), ‘The problem of overfitting’, J Chem Inf Model, 44, 1–12. doi: 10.1021/ci0342472 kubinyi h (2002), ‘From narcosis to hyperspace: the history of QSAR’, Quant Struct Act Relat, 21, 348–356. leach a r (2001), Molecular Modelling: Principles and Applications, Upper Saddle River, NJ, Prentice Hall. liu h, papa e, walker j d, gramatica p (2007), ‘In silico screening of estrogen-like chemicals based on different nonlinear classification models’, J Mol Graph Model, 26, 135–144. doi:10.1016/j.jmgm.2007.01.003 lo piparo e, koehler k, benfenati e (2006), ‘Virtual screening of food contaminant to identify potential endocrine disruptors: handling estrogen receptor flexibility’, Proceedings of the XVI European Symposium on Quantitative Structure–Activity Relationships and Molecular Modelling, 10–17 September 2006, Mediterranean Sea/Italy, 291–292. mcfarland j w (1970), ‘Parabolic relation between drug potency and hydrophobicity’, J Med Chem, 13, 1192–1196. doi: 10.1021/jm00300a040 mukherjee s, saha a, roy k (2005), ‘QSAR of estrogen receptor modulators: exploring selectivity requirements for ER(alpha) versus ER(beta) binding of tetrahydroisoquinoline derivatives using E-state and physicochemical parameters’, Bioorg Med Chem Lett, 15 (4), 957–961. doi:10.1016/j.bmcl.2004.12.048 netzeva t i, gallegos saliner a, worth a p (2006), ‘Comparison of the applicability domain of a quantitative structure–activity relationship for estrogenicity with a large chemical inventory’, Environ Toxicol Chem, 25, 1223–1230. doi: 10.1897/05–367R. oprea t i and marshall g r (1998), ‘Receptor-based prediction of binding affinities’, Perspect Drug Discovery Design, 9–11, 35–61. doi: 10.1023/A:1027299602978 salum l d b, polikarpov i, andricopulo a d (2007), ‘Structural and chemical basis for enhanced affinity and potency for a large series of estrogen receptor ligands: 2D and 3D QSAR studies’, J Mol Graph Model, 26, 434–442. doi:10.1016/j. jmgm.2007.02.001

The endocrine-disrupting potency of a chemical

321

sarkhel s, sharon a, trivedi v, maulik p r, singh m m, venugopalan p, ray s (2003), ‘Structure-based drug design: synthesis, crystal structure, biological evaluation and docking studies of mono- and bis-benzo[b]oxepines as non-steroidal estrogens’, Bioorgan Med Chem, 11 (23), 5025–5033. doi: 10.1016/j.bmc.2003.08.026 schmieder p k, ankley g, mekenyan o, walker j d, bradbury s (2003), ‘Quantitative structure–activity relationship models for prediction of estrogen receptor binding affinity of structurally diverse chemicals’, Environ Toxicol Chem, 22 (8), 1844– 1854. doi: 10.1897/01-345 schultz t w, cronin m t d, walker j d, aptula a o (2003), ‘Quantitative structureactivity relationships (QSARs) in toxicology: a historical perspective’, J Mol Struc-THEOCHEM, 622, 1–22. doi:10.1016/S0166-1280(02)00614-0 todeschini r and consonni v (2000), Handbook of Molecular Descriptors, Weinheim, Wiley-VCH. tong w, fang h, hong h, xie q, perkins r, anson j, sheehan d m (2003), ‘Regulatory application of SAR/QSAR for priority setting of endocrine disruptors: A perspective’, Pure Appl Chem, 75 (11–12) 2375–2388. doi:10.1351/pac200375112375 tong w, xie q, hong h, shi l, fang h, perkins r (2004), ‘Assessment of prediction confidence and domain extrapolation of two structure–activity relationship models for predicting estrogen receptor binding activity’, Environ Health Persp, 112, 1249–1254. doi:10.1289/txg.7125 tropsha a, gramatica p, gombar v k (2003), ‘The importance of being earnest: validation is the absolute essential for successful application and interpretation of QSPR models’, QSAR Comb Sci, 22, 69–77. doi: 10.1002/qsar.200390007 vedani a, dobler m, lill m a (2005), ‘Virtual test kits for predicting harmful effects triggered by drugs and chemicals mediated by specific proteins’, Altex, 22 (3), 123–134. waller c l (2004), ‘A comparative QSAR study using CoMFA, HQSAR, and FRED/SKEYS paradigms for estrogen receptor binding affinities of structurally diverse compounds’, J Chem Inf Model, 44 (2), 758–765. doi: 10.1021/ci0342526 warren g l, andrews c w, capelli a m, clarke b, lalonde j, lambert m h, lindvall m, nevins n, semus s f, senger s, tedesco g, wall i d, woolven j m, peishoff c e, head m s (2006), ‘A critical assessment of docking programs and scoring functions’, J Med Chem, 49, 5912–5931. doi: 10.1021/jm050362n yoon s and welsh w j (2004), ‘Identification of a minimal subset of receptor conformations for improved multiple conformation docking and two-step scoring’, J Chem Inf Model, 44 (1), 88–96. doi: 10.1021/ci0341619

13 Endocrine disruptors in human milk and the health-related issues of breastfeeding B. G. J. Heinzow, State Agency for Social Services Schleswig-Holstein, Germany, and University of Notre Dame, Sydney School of Medicine, Australia

Abstract: Breastfeeding has an essential role in a healthy development of the newborn, but human milk has been compromised by pollutants from our environment. The main contaminants of human milk with endocrine-disrupting properties, xenobiotics and phytoestrogens, are presented together with their level of exposure. In most countries persistent organic pollutants (POPs) are declining; however some compounds such as polybrominated diphenyl-ether (PBDE) have shown an increase in human milk and new ones, such as sunscreens, have been identified only recently in breast milk. Risk assessment of xenobiotics in human milk is presented and concluded that, despite the presence of contaminants, breastfeeding ensures the best possible health as well as the best developmental and psychosocial outcomes for the infant. Continued environmental monitoring and biomonitoring based on harmonised protocols is necessary to ensure that known exposures to endocrine-disruptive compounds (EDCs) in human milk do not increase unexpectedly or new ones develop unnoticed. Key words: breastfeeding, human milk, risk assessment, biomonitoring, exposure, endocrine-disruptive compounds, EDC, PCDD, PCDF, PCB, organochlorine pesticide, POPs, pharmaceuticals and personal care products (PPCP), PFC, pesticide, phthalate.

13.1

Introduction

Fetal and early postnatal development constitutes the most vulnerable stage of human life in regard to adverse effects of environmental hazards. Breastfeeding, both by psychosocial factors and optimal balanced nutritional and other physiological constituents, has an essential and beneficial

Endocrine disruptors in human milk

323

role in a healthy development of the newborn. Every kind of milk has its own biological specificity. Human milk is the natural and superior food for infants containing the optimal composition to meet their nutritional needs in early life and providing associated immunological, psychological and economic advantages. Evidence for the health advantages of breastfeeding and scientific evidence to support this practice has continued to increase. The World Health Organization (WHO) and other health and professional bodies have placed emphasis on promoting and increasing the rate and length of breastfeeding (Slusser, 2004; AAP, 2005; WHO, 2006). On a population basis, exclusive breastfeeding for six months is the recommended feeding mode for the vast majority of infants, followed by continued breastfeeding with appropriate complementary foods for up to two years or beyond. Yet human milk has been unintentionally compromised by pollutants from our environment, as a result of living in a ‘modern’ industrialised world (Landrigan et al., 2002). Thus the contamination of human milk especially with endocrine-disrupting compounds (EDC) has raised concern for public and environmental health (Rogan and Ragan, 2003; Massart et al., 2005; Nickerson, 2006). However, the mere presence of an environmental chemical in human milk does not necessarily indicate that a serious health risk exists for breastfed infants. The presence of contaminants in human milk must be seen as a hazard. A hazard exists where a situation (e.g. a xenobiotic) has a built-in ability to cause an adverse effect. Risk, on the other hand, is the chance that such effects will occur. The risk can be high or negligible and depends mainly on exposure. Referring to the precautionary principle it has been sometimes advocated to stop breastfeeding. However, breastfeeding provides proven and important benefits to mother and infant. In the assessment one must weigh the many benefits of breastfeeding for mothers and infants against the risk of exposing infants to chemicals (and drugs) present in breast milk (Howard and Lawrence, 2001). Numerous epidemiological studies have demonstrated that human milk and the practice of breastfeeding confer significant, measurable health benefits to infants and to nursing mothers (Boersma and Lanting, 2000; Eskenazi et al., 2006). Therefore, it must be emphasised that the accumulated data overwhelmingly support the positive health value of breastfeeding infants, and should be encouraged strongly as the optimal feeding choice. In most countries the general downward trend in the levels of persistent organic pollutants (POPs) indicates a continuing decline in exposure, as measures to reduce emissions have been implemented. The decline in northern Germany is depicted as an example in Fig. 13.1. Analysis of human milk, maternal blood and adipose tissue are all relevant matrices for assessment of body burdens for persistent organic pollutants. Biomonitoring of human milk data can provide information on

324

Endocrine-disrupting chemicals in food 1.4

PCB DDT HCB β-HCH

POP [μg/g lipid]

1.2 1.0 0.8 0.6 0.4 0.2

06

04

02

00

98

96

94

92

90

88

86

0 Year

Fig. 13.1 Time trend of persistent organochlorine pollutants in breast milk from Germany (PCB, polychlorinated biphenyl; DDT, dichlorodiphenyltrichloroethane; HCB, hexachlorobenzene; β-HCH, β-hexachlorocyclohexane).

the exposure of the mother as well as the infant (Hooper, 1999). Furthermore, such information provides guidance on the need for measures to reduce levels of these substances in food, which is the main source of exposure for most people. Better understanding of our exposure to harmful environmental chemicals will help to further control such chemicals by eliminating or reducing emissions or by limiting their presence in the food supply (Hooper and McDonald, 2000). This is the main reason to test mothers’ milk by finding out body burden and types of pollutants.

13.2

Xenobiotics and transmission into human milk

The lactation period can be considered an extended part of pregnancy. This is because the mother can supply the nutrient needs to the infant through her human milk, much as she did so during pregnancy through the placenta. It is well recognised that drugs and other xenobiotics are transferred from the mother’s circulation into human milk (Scialli, 1992; Ito and Lee, 2003; Anderson, 2006). The underlying mechanism of transmission is based mainly on passive diffusion and the dose an infant receives during breastfeeding depends upon the plasma concentration of the xenobiotics of the mother, the amount excreted into milk, and the daily volume of milk ingested (Begg et al., 2002). Alveolar cells of the mammary glands are the interface between maternal blood and milk, the substrates for milk production and xenobiotics are brought to the alveoli by the bloodstream. Free (unbound) substances are transferred from plasma to milk by diffusion and less commonly through aqueous pores. Factors that influence the passive diffusion rate and transfer into milk (Table 13.1) are the size of the molecule, lipid and water solubility (log Pow), ionisation (pKa), pH of substrate, and finally plasma protein binding. Binding

Endocrine disruptors in human milk

325

Table 13.1 Factors influencing transfer of xenobiotics into breast milk Transfer by diffusion

In favour

Molecular weight (MW) Lipophilicity (log Pow) Plasma protein binding Ionisation pKa Carrier transport

Low (<200) High (>4) Low (<85%) Cations Not known

of a xenobiotic to plasma proteins limits its transfer since only the unbound ‘free fraction’ is in equilibrium accross membranes. The passage also depends upon the size of the molecule (molecular weight, MW). In general for substances with a MW below 200, transport is not limited, whereas substances above MW 800 rarely pass. The degree of drug ionisation depends upon the pH of plasma (pH = 7.4) and milk (pH = 6.8). Thus weak bases become more ionized in milk, which has a lower pH and the ionized compound will therefore be at higher concentration in milk. In general weakly alkaline compounds are in equal or higher levels in milk, whereas weakly acidic compounds occur at higher concentration in plasma. Milk composition changes over time (colostrum, transitional milk and mature milk) and during a single suckling (foremilk, hindmilk). The first postpartum week colostrum has a high protein (10%) and a low fat (1%) content, mature milk however has a low protein (1%) and a high fat (4%) level. Because of the high protein content in colostrum, chemicals with high protein binding may be more likely to pass into human milk at this time. The fat content at any feeding increases over time, the foremilk is low in fat (1–2%) and the hindmilk (4–6%) is about five times richer in fat content towards the end of a feeding. Carrier-mediated active transport also exists for specific compounds and has been recognised for some drugs as a cation transporter (Cimetidine) and anion transporter (Benzylpenicillin). It is unknown whether this affects other xenobiotics and EDCs. Presently, there is no appropriate model to predict precisely milk concentrations of EDCs in humans from the plasma concentration in the mother, unless the milk to plasma ratio (M/P ratio) has been determined. The M/P ratio is the concentration of a xenobiotic in the milk versus the concentration in the maternal plasma at the same time. This information is not available for EDCs. Data on M/P concentration ratios exist only for some medications and show large inter- and intra-individual variability. Lipophilicity, protein binding in milk and plasma, fat partitioning (log Pow), and ionization (pKa) have been used to predict the extent of drug

326

Endocrine-disrupting chemicals in food

accumulation in milk (Atkinson and Begg, 1990; Notarianni et al., 1995; Agatonovic-Kustrin et al., 2002; Begg et al., 2002; Clewell and Gearhart, 2002; Larsen et al., 2003; Katritzky et al., 2005; Zhao et al., 2006). Protein binding seems to be the most important single predictor. If no data are available on the transfer of a xenobiotic into breast milk knowledge of protein binding properties and lipophilicity can be used to get a rough estimate of the exposure characteristics (Anderson, 2006). In general for compounds >85% protein bound, infant exposure does not result in clinically relevant amounts (Anderson, 2006), as for most xenobiotics (drugs) with high protein binding amounts in milk will not exceed 1% of maternal dose. However, lipophilic and non-ionised compounds are found at higher concentrations in milk than in blood, because of the high lipid content of milk (∼4%) compared with plasma (0.7%). Thus especially for lipophilic and bioaccumulating chemicals exposure of the infant can be expected to be be high and relevant, because the postnatal exposure via milk is much higher than the prenatal exposure via the placenta. The amount of a xenobiotic (∼dose) the infant will receive can be calculated by the concentration in milk and the daily milk intake (Neville et al., 1988). Milk intake on a body weight basis is quite low immediately after birth, gradually increasing and levelling off at about 5–10 days of age. Peak average daily lipid intakes occur during the next several weeks and gradually decrease over time. The 0–6 month average daily lipid intake rate is 4.5 g/kg body weight (Dewey et al., 1991). As a rule of the thumb a milk intake of 150 mL/kg bw can be used to calculate mik intake for an infant. For more precise exposure estimates data from German (AGLMB, 1995) or US (EPA, 2006) exposure factor handbooks are recommended.

13.3

Nutritional phytoestrogens in human milk

Phytoestrogens do occur in human milk when the lactating woman is consuming a diet with, for example, high isoflavone content. This is typical for women in some Asian countries, where soyfoods are regularly consumed (Adlerkreutz et al., 1999). Women in Western countries may produce milk containing the isoflavones genistein and daidzein between 4 and ∼20 μg/L, although there is large variation between individuals depending upon nutritional intake (Morton et al., 1996; Franke and Custer, 1996; Franke et al., 1998, 2006). A typical infant weighing 7 kg and consuming 0.8 L of breastmilk per day would have an upper range of isoflavone consumption of ∼4 μg/kg body weight per day. This can be compared with exposure of formula-fed babies. Soy-based infant formulas – in which milk protein is substituted by the protein from soybeans – have been available since 1962, and are the most widely used alternative to milk-based formulas (Borgert, 2003). Babies who drink soy formula receive significant amounts of

Endocrine disruptors in human milk Table 13.2

Contaminants with suspected endocrine activity in human milk

Organochlorines

Mean [ng/g lipid]

Chlordane Oxychlordane cis-Nonachlor trans-Nonachlor Chlordanes Oxychlordane

72.9 7.2 ± 3.4 3.7 ± 1.7 18.8 ± 8.6 2.5 4.1

p,p-DDE p,p-DDT Σ–DDT Σ–DDT

21.4 2.93 193.6 747

p,p-DDE Σ-DDT p,p-DDE

490 14960 760 ± 1460

p,p-DDT

70 ± 30

p,p-DDE p,p-DDT p,p-DDE p,p-DDE

327

130

Range [ng/g lipid]

4.0

p,p-DDT p,p-DDE Σ-DDT

190 2200 166

p,p-DDT p,p-DDE p,p-DDT

19 228 240

p,p-DDE Σ–DDT p,p-DDE

1720 2040 634

Σ–DDT

169

p,p-DDT

4.35

Dieldrin

4.65 4 5 ± 3.6 3.1

Reference

Japan Japan

Azuma (1999) Nakai et al. (2004) Nakai et al. (2004) Nakai et al. (2004) Minh et al. (2004) Damgaard et al. (2006)

Vietnam Sweden

119–3207

23.4–894

6.2 ± 3.5 142.3 ± 73.5 83.8

p,p-DDT

Country

Egypt

Saleh et al. (1996)

Japan Germany East Ghana Thailand Indonesia, rural Indonesia, rural Sweden

Azuma (1999) MSG-SA (2001)

Japan Denmark

Vietnam 24–2300

UK Taiwan China, Beijing Poland Germany West

Ntow (2001) Stuetz et al. (2001) Burke et al. (2003) Burke et al. (2003) Darnerud et al. (2004) Nakai et al. (2004) Nakai et al. (2004) Damgaard et al. (2006) Damgaard et al. (2006) Minh et al. (2004) Minh et al. (2004) Kalantzi et al. (2004) Chao et al. (2007) Chao et al. (2007) Yu et al. (2006) Yu et al. (2006) Yu et al. (2006) Jaraczewska et al. (2006) Vieth (2007) Vieth (2007)

1–34

Japan Germany East Japan Denmark

Azuma (1999) MSG-SA (2001) Nakai et al. (2004) Damgaard et al. (2006)

328

Endocrine-disrupting chemicals in food

Table 13.2 Continued Organochlorines

Mean [ng/g lipid]

α-Endosulfan

HCB

Range [ng/g lipid]

Country

Reference

11.4 ng/mL

Spain

6.7 ng/mL

Denmark

Cerrillo et al. (2005) Damgaard et al. (2006)

40 43

9–227

129 17.1 ± 10.1 4.2 6.8 28–880 29

8.42 127.6 59

α-HCH γ-HCH Σ-HCH

3 6 408

β-HCH β-HCH Σ-HCH

83.4 ± 55.1 69 16.6

Σ-HCH

1180

β-HCH

12.3

β-HCH Σ-HCH

1.2

γ-HCH

0.8

β-HCH

24

HE

6.2 3

HE cis-HE HE Heptachlor

10–262

4.0 2.3

Egypt Japan Germany East

1–14 1–62

1.2–1500

Russia, Arctic Japan Vietnam UK China, Beijing Denmark

100–37 000

3.7 ± 1.4 2.2

Russia Germany West Denmark Finland

12.3 8.45 γ-HCH Σ-HCH β-HCH

Ghana Germany East Russia, Arctic Japan Vietnam Denmark

Taiwan Russia

Germany West 1–41

Japan Germany East Japan Denmark Taiwan

Ntow (2001) MSG-SA (2001) Polder et al. (2003) Nakai et al. (2004) Minh et al. (2004) Damgaard et al. (2006) Tsydenova et al. (2007) Vieth (2007) Shen et al. (2007) Shen et al. (2007) Saleh et al. (1996) Azuma (1999) MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) Polder et al. (2003) Nakai et al. (2004) Minh et al. (2004) Kalantzi et al. (2004) Yu et al. (2006) Damgaard et al. (2006) Chao et al. (2007) Tsydenova et al. (2007) Tsydenova et al. (2007) Vieth (2007) Azuma (1999) MSG-SA (2001) Nakai et al. (2004) Damgaard et al. (2006) Chao et al. (2007) Chao et al. (2007)

Endocrine disruptors in human milk

329

Table 13.2 Continued Organochlorines

Mean [ng/g lipid]

Σ-Toxaphene (4)

Range [ng/g lipid] 7–24

13 Triclosan

0–2100

Dioxins + dl-PCB

Mean TEQ[pg/g lipid]

PCDD/F PCB PCDD

12.2 10 10.0

PCDF

7.8

dl-PCB

9.9

Range TEQ[pg/g lipid]

15.2

7.1–40.3

PCDD/F

29.4

16–52

PCDD/F + PCB

40.8

PCDD/F + PCB dl-PCB PCB mono-otho

20.84 5.64 4.03 ± 2.6

0.7–16.7

PCB non-ortho

3.2 ± 1.5

1.46–5.5

PCDD/F

6.82 ± 3.02

1.86–11.1

PCDD/F PCDD/F * PCB PCDD/F

14.9 25.6 13.9

PCDD/F + PCB

23.7

PCDD/F PCDD/F + PCB

7.4 9

6–15.2

27.27

3–78.7

15.2

7.1–40.3

PCDD PCDF dl-PCB

14.7 ± 9.4 8.7 4.9 4.98 9.92

Reference

Germany Russia, Arctic

Skopp et al. (2002) Polder et al. (2003)

US

Dayan (2007)

Country

Reference

Japan

Azuma (1999) Azuma (1999) Nakagawa et al. (1999) Nakagawa et al. (1999) Nakagawa et al. (1999) MSG-SA (2001)

Japan

PCDD/F

PCDD/F dl-PCB

Country

Germany East Belgium

Korea Greece

Japan Japan

4.21–52.8

Taiwan Australia Germany West Germany East Taiwan Spain Germany

Focant et al. (2002) Focant et al. (2002) Yang et al. (2002) Yang et al. (2002) Costopolou et al. (2006) Costopolou et al. (2006) Costopolou et al. (2006) Tajimi et al. (2004) Tajimi et al. (2004) Nakatani et al. (2005) Nakatani et al. (2005) Chao et al. (2007) Harden et al. (2007) Wittsiepe et al. (2007) Wittsiepe et al. (2007) Hsu et al. (2007) Schuhmacher et al. (2007) Raab et al. (2007) Raab et al. (2007) Raab et al. (2007)

330 Table 13.2

Endocrine-disrupting chemicals in food Continued

PCB

Mean [ng/g lipid]

Range [ng/g lipid]

Country

Reference MSG-SA (2001)

Σ-PCB

252

Σ-PCB

458

Σ-PCB s 15

190

77–547

Germany East Russia, Arctic Sweden

Σ-PCB 15

150

26–530

UK

PCB 28

2.86

0.25–30.7

Sweden

PCB 153

62

11.4–186

Σ-PCB sum 11 Σ-PCB 18 Σ-PCB 15 Σ-PCB 6

64–1013

1800

Faroe Islands (1999) Poland

133 63.9 156 ± 40.9

Σ-PCB 16

240

Σ-PCP 82 Total PCB

147 270

OH–PCB sum 12

3 pg/g milk

PBDE

Mean [ng/g lipid]

38.4–660

49–415 0.1–5

Japan Greece, Athens Italy, Rome US, CAN Germany West Sweden

Reference Akutsu et al. (2003) Guvenius et al. (2003) Lind et al. (2003) Schecter et al. (2003) Kalantzi et al. (2004) Fängström et al. (2005)

0.56–3.97

Japan

Σ-PBDE 11

2.14

0.56–7.7

Sweden Sweden US

Σ-PBDE 5 Σ-PBDE 13

4.0 73.9

n.d.–28.2 6.2–419

Σ-PBDE 15

6.6

0.3–69

UK

4.7–13

Faroe Islands (1999) China Germany Japan Poland

Σ-PBDE 12 Σ-PBDE 14

3.75 1.56 2 96

Guvenius et al. (2003)

Country

1.64

Σ-PBDE Σ-PBDE 8 Σ-PBDE 13 Σ-PBDE 11

Jaraczewska et al. (2006) Inoue et al. (2006) Costopolou et al. (2006) Ingelido et al. (2007) She et al. (2007) Vieth (2007)

Range [ng/g lipid]

Σ-PBDE 16

Σ-PBDE 11

Polder et al. (2003) Guvenius et al. (2003) Kalantzi et al. (2004) Darnerud et al. (2004) Darnerud et al. (2004) Fängström et al. (2005)

1.5–17 0.85–24.6 0.8–8.4 6–321 0.46–1.7

US Russia

Bi et al. (2006) Fürst (2006) Inoue et al. (2006) Jaraczewska et al. (2006) She et al. (2007) Tsydenova et al. (2007)

Endocrine disruptors in human milk Table 13.2

331

Continued

PBDE

Mean [ng/g lipid]

Σ-PBDE 11

4.1

Σ-PBDE 12 Σ-PBDE

11.1 ± 3.2 2.2

Σ-PBDE

Range [ng/g lipid]

Country

Reference

Italy, Rome

Ingelido et al. (2007) Toms et al. (2007) Schuhmacher et al. (2007) Sudaryanto et al. (2007)

6.1–18

Australia Spain

0.49–13

Indonesia

Phthalates

Mean [ng/mL]

Range [ng/mL]

Country

Reference

mMP mEP mBP mBzP mEHP miNP mMP mEP mBP mBzP mEHP miNP DEP DBP DEHP DEHP DBP DEP

0.09 0.97 12 13 13 89 0.1 0.93 4.3 0.9 9.5 101 0.14 ng/g milk] 0.51 109 222 0.87 0.31

0.01–0.37 0.25–41.4 2.4–123 0.4–26 4.0–1410 28–230 0.1–5.5 0.07–33.6 0.6–10900 0.2–14 1.5–191 27–469 −8.1 n.d.–0.11 2920 156–398 0.62–1.2 –

Finland

Main Main Main Main Main

Denmark

Main et al. (2006) Main et al. (2006) Main et al. (2006) Main et al. (2006) Main et al. (2006) Main et al. (2006) Zhu et al. (2006) Zhu et al. (2006) Zhu et al. (2006) Zhu et al. (2006) Zhu et al. (2006) Zhu et al. (2006)

Phenols

Mean [ng/mL]

Range [ng/mL]

Country

Reference

<0.09–0.7

Japan n = 4 Japan Japan colostrum Japan Germany

Otaka et al. (2003) Sun et al. (2004) Kuruto-Niwa et al. (2007) Otaka et al. (2003) Butte and Fooken (1990)

Range [ng/mL]

Country

Reference

0.45–3.6 0.47–2.1

China

So et al. (2006) So et al. (2006) Kärrman et al. (2006)

Bisphenol A

0.61 ± 0.3 3.4 ± 0.13

<0.5–1.4

Nonylphenol Pentachlorophenol

1.5 ng/mL

Fluororanics

Mean [ng/mL]

PFOS PFOA PFC

1–7

0.34

Canada

Sweden

et et et et et

al. al. al. al. al.

(2006) (2006) (2006) (2006) (2006)

332

Endocrine-disrupting chemicals in food

Table 13.2 Continued Musk

Mean [ng/g lipid]

Musk xylene

25 ± 8.7

Musk ketone HHCB AHTN Musk xylene

11 ± 4.2 49 ± 36 26 ± 18 100

Musk xylene

41

Musk ketone

10

Range [ng/g lipid]

10–1220

16–189

AHTV

8–58

Musk xylene

23.6 ± 15.6

Musk ketone HHCB AHTN Musk xylene

17 ± 6.1 180 ± 110 20 ± 9.8 6

Musk ketone Musk mosken Musk ambrette HHCB

5 – – 60

<2–83 <0.2–3 4

Musk xylene

12

4–83

Musl ketone Galaxolid HHCB Tonalid AHTN Celestolid Phantolid Traseolid Musk xylene

6 6 22 2 3 2 30 ± 36

1–62 <2–132 <2–150 <1–14 <1–20 <1–5 <2–150

Musk ketone

74 ± 66

<2–238

HHCB

220 ± 212

<5–917

AHTN

47 ± 36

<5–144

28.6 and 88

Reference

Germany, Hesse

Ott et al. (1999)

Germany, Bavaria Germany

HHCB

HHCB-lactone

Country

Denmark

<1–48

Germany, Bavaria

Germany, Bavaria Germany East

US, California

only in 2 samples

Ott et al. (1999) Ott et al. (1999) Ott et al. (1999) Liebl and Ehrenstorfer (1993) Rimkus et al. (1994) Rimkus et al. (1994) Rimkus and Wolf (1996) Rimkus and Wolf (1996) Duedahl-Olesen et al. (2005)

Liebl et al. (2000) Liebl et Liebl et Liebl et Liebl et

al. al. al. al.

(2000) (2000) (2000) (2000)

MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) MSG-SA (2001) Reiner et al. (2007) Reiner et al. (2007) Reiner et al. (2007) Reiner et al. (2007) Reiner et al. (2007)

Endocrine disruptors in human milk

333

phytoestrogens in the form of isoflavones. Studies by Franke and by Setchell (Franke, 1997; Setchell et al., 1997) show that infants fed soy-containing formulas are exposed to much higher concentrations of isoflavones than are breast-fed infants. Soy-based formulas contain at least 10-fold the amount of phytoestrogens found in human milk; however, the difference may be as much as 1000-fold (Setchell and Welsh, 1987; Nguyenle et al., 1995; Morton, et al., 1996; Murphy et al., 1997; Knight et al., 1998). Isoflavone contents in formula from recent studies were mostly in the range of 20–40 mg/L reconstituted formula, which is equivalent to a dose of 2 mg/kg day. In human milk isoflavones are found as glucuronide conjugates (Franke and Custer, 1996), whereas they are present as glycosidic conjugates in soy milk (Coward et al., 1993). It is not known whether these compositional differences may influence bioavailability (Setchell et al., 1997).

13.4

Range and distribution of xenobiotic endocrine disruptors in human milk

For four decades the conventional organochlorine (OC) pollutants have captured most of our attention. Numerous human milk samples from women all over the world have been analysed for POPs, especially chlororganic pesticides (dichlorodiphenyltrichloroethane, DDT, mirex, α-, β-, γand δ-hexachlorocyclohexane, alpha- and gamma-chlordane, oxychlordane, transnonachlor, p, p′-DDT and some analogues, heptachlor epoxide, dieldrin and octachlorostyrene, polychlorinated biphenyls, benzenes, naphthalenes, and terphenyls). Systematic investigations of organochlorine compounds in human milk from women living in the Stockholm region started in 1967 (Norén and Meironyté, 2000). The WHO has collected and evaluated information on levels of POPs in human milk since 1976. Over the period 1987–2003, it has coordinated three international studies of human milk to assess the levels and trends of polychlorinated dibenzodioxins, polychlorinated dibenzofurans and dioxin-like polychlorinated biphenyls. During the course of the past 20–30 years the levels of banned organochlorine compounds in human milk have decreased to various extents; half-lives were estimated between 4 and 17 years (Norén and Meironyté, 2000). Data from New Zealand can be taken as an example to illustrate this time course, which has been observed in other countries as well. In a replication of a 1987–1988 study, milk samples of New Zealand women collected in 1998–1999 were analysed for polychlorinated dibenzo-p-dioxins and dibenzofurans, polychlorinated biphenyl (PCB) congeners and organochlorine pesticides. Levels of most compounds measured in the study had declined approximately 70% from their 1987–1988 values, indicating that regulatory measures to reduce exposures have been effective (Bates et al., 2002).

334

Endocrine-disrupting chemicals in food

18 000 16 000 14 000 12 000 10 000 8000 6000 4000 2000 0

C

an ad Ke a ny a G US er A m N any Z no rt Br h a Tu zil r k N Z ey so ut h In Th di ai a l Vi and e Zi tna m ba m b M we ex i Jo co rd S an Af ric a

Total DDT [μg/g lipid]

Continental, regional and even provincial differences in residue levels are large. Considerable variation in levels of the analysed OCs exists also between individuals in the same area. Human milk from mothers nursing their second or third child (multiparous) show a different profile from that of mothers in their first lactation (primiparous). Also maternal age and diet (e.g. fish and meat versus vegetarian food) can have an influence on levels of lipophilic milk residues (Mes et al., 1993; Schade and Heinzow, 1998). Although for most persistent compounds a decline has been observed over the past, some compounds which are highly lipophilic, such as polybrominated diphenyl-ethers (PBDEs) and polycyclic musks showed a steady increase in human milk until recently (Lundén and Norén, 1998; Norén and Meironyté, 2000; Reiner et al., 2007). PBDEs originate from their use as flame retardants. Heterocyclic or polycyclic musks (PCM) are artificial musks that contain more than one ring in the molecular structure and are used for odoriferous compositions. In Table 13.2 data from the most recent studies on contaminants in breast milk – preferably dating back less than ten years – are compiled. Most data exist for OCs, especially the persistent pesticides, DDT, PCBs, dioxins and furans. For these compounds (POPs) an overview of human milk contamination worldwide is possible. For other contaminants only sparse information is available on only a limited number of chemicals, from small cohorts in few geographic locations. Depending upon industrialisation and use of agrochemicals and climatic factors (e.g. evaporation and condensation in hot and cold climates), as can be seen from Figs 13.2–13.4, countries display a typical pattern of contaminants which is a fingerprint for that geographical location (Raum et al., 1998; Polder et al., 2003; Costopoulou et al., 2006). If physicochemical characteristics (e.g. lipophilicity) and the exposure pathway are identical (for example: enrichment in the food chain, biomagnified food

Fig. 13.2

Mean concentrations of total DDT (μg/g lipid) in human milk from different countries in the eighties (adopted from Smith, 1999).

Endocrine disruptors in human milk

335

(a) Fiji Brazil Philippines Australia Bulgaria Croatia Hungary New Zealand USA Norway Greece Ireland Czech Republic Hong Kong Romania Slovakia Russia Finland Sweden Ukraine Spain Germany Italy Luxembourg Belgium Netherlands Egypt 0

5

10

15

20

25

PCDD/F [pg TEQ/g lipid]

Fig. 13.3 Polychlorodibenzdioxins/furans (PCDD/F), dl-PCB and standard PCB in human milk from different countries (adopted from Costopoulou et al., 2006).

336

Endocrine-disrupting chemicals in food (b) Fiji Brazil Philippines Australia Bulgaria Croatia Hungary New Zealand USA Norway Greece Ireland Czech Republic Hong Kong Romania Slovakia Russia Finland Sweden Ukraine Spain Germany Italy Luxembourg Belgium Netherlands Egypt 0

5

10

15

dl-PCB [pg TEQ/g lipid]

Fig. 13.3 Continued

20

25

Endocrine disruptors in human milk (c) Fiji Brazil Philippines Australia Bulgaria Croatia Hungary New Zealand USA Norway Greece Ireland Czech Republic Hong Kong Romania Slovakia Russia Finland Sweden Ukraine Spain Germany Italy Luxembourg Belgium Netherlands Egypt 0

100

200

300

400

PCB [ng/g lipid]

Fig. 13.3 Continued

500

600

337

338

Endocrine-disrupting chemicals in food

Total PBDE [ng/g lipid]

80 70 60 50 40 30 20 10

U SA

It Be aly lg iu m N No et rw he a rla y nd M s ex ic o Fa ro e Is UK la nd C s an ad a

Vi

et

na m Ja pa Sw n ed Fi en nl G and er m an y

0

Fig. 13.4 Mean concentrations of total PBDE (pg/g lipid) in human milk from different countries between 1998 and 2003.

contaminant – exposure via fish), often a close correlation can be found between different lipophilic compounds, although the primary source of the pollution (e.g. agriculture and industry) might be different. Since for DDT and PCBs the main exposure pathway is usually animal fat, concordant higher levels in milk are related to age of the mother and her dietary intake of animal fat. If an unusual high value of PCBs is found in the milk of a mother together with low Σ-DDT (total DDT = sum of DDT and metabolites) levels, this individual biomonitoring result strongly indicates a specific exposure for PCBs, for example from indoor air contamination or occupation. If otherwise Σ-DDT is abnomally elevated this might point at personal exposure, e.g. from vector control. Thus biomonitoring information on several lipophilic compounds (exposure-fingerprint) rather than on one single compound should be obtained when studying human milk to provide guidance on indentification of the common and the specific exposure sources. The following concise reviews of the contamination of persistent organic pollutants in human human milk are recommended for further information: Jensen and Slorach (1991), Fürst et al. (1994), Sonawane (1995), Smith (1999), LaKind et al. (2004, 2005), LaKind (2007), Solomon and Weiss (2002), Jaga and Dharmani (2003), Fürst (2006), and Tanabe and Kunisue (2007).

13.4.1

Dichlorodiphenyltrichloroethane and other organochlorine pesticides With the POP-convention the persistent chlorinated pesticides particularly aldrin, chlordane, DDT, dieldrin, endrin, heptachlor, hexachlorobenzene

Endocrine disruptors in human milk

339

(HCB), mirex and toxaphene, are targets for global reduction and/or elimination. Levels in human milk are usually low, except total DDT (Σ-DDT). Although banned in most countries, DDT is still used in many parts of the world designated for application in malaria control (Smith, 1999). DDT was banned mainly for ecological reasons and not because of human toxicity, but subsequent research has shown that exposure to DDT at amounts used for malaria control might cause preterm birth and early weaning, abrogating the benefit of reducing infant mortality from malaria. DDT might be useful in controlling malaria, but the evidence of its adverse effects on human health needs appropriate research on whether it achieves a favourable balance of risk versus benefit (Rogan and Chen, 2005). Commercial DDT is a mixture of several closely related compounds, with p,p′-DDT generally comprising >75% and o,p′-DDT 15% of the formulation, Following use in agriculture or for vector pest control, degradation or metabolism of the parent compound (DDT) occurs, hence environmental samples mainly contain the metabolites DDE (dichlorodiphenylchloroethane) and DDD (dichlorodiphenyldichloroethane) unless samples are taken soon after DDT’s use. DDE is DDT’s main metabolite and also the most persistent form of the chemical and accumulates in fish and other organisms depending upon the trophic level and magnification through the food chain. The proportion of DDT and DDE found in human tissues can be used as an indication of the length of time since last use. In areas where DDT exposure has been recent, the DDE/DDT ratio is low, while in areas where substantial time since exposure has passed, the DDE/DDT ratio is higher. Following bans and restrictions a continuous decline in human milk contamination was observed and the longer DDT was restricted, the lower the average levels of Σ-DDT in human milk are in that country. Although studies have been conducted in many countries, data are not always representative and might differ in other regions of that country (Burke et al., 2003). DDT levels are usually much lower in developed nations than in developing ones. It is difficult to make comparisons between countries because studies were not done at the same point in time. For comparison human milk levels from different countries from similar time periods in the 1980s are shown in Fig. 13.2. Σ-DDT levels have also declined in developing countries (Table 13.2).

13.4.2 Polychlorinated biphenyls PCBs were manufactured mainly as cooling and insulating fluids for industrial transformers and capacitors, as hydraulic fluids, and also as additives to paints, caulking agents and electrical insulation. PCB production was banned in the 1970s in most countries but use in ‘closed systems’ (e.g. capacitors) continued for longer periods. Large exposure differences exist between countries. Higher levels in human milk are found in industrial regions, and if fish taken from polluted rivers and lakes is consumed.

340

Endocrine-disrupting chemicals in food

Data from different research groups are not always comparable, since the number of congeners analysed and the calculation of total PCBs differ (Longnecker et al., 2003). Usually the six congeners PCB 28, 52, 101, 134, 153 and 180 are referred to as standard congeners. Some PCBs show dioxinlike properties. These planar non-ortho and mono-ortho PCBs are termed dl-PCB and their summarised toxic equivalents (TEQ) contribute in industrialized countries as much as polychlorinated dioxins and furans to the total TEQ-body burden.

13.4.3 Polychlorinated dioxins and furans Polychlorinated dioxins and furans – often referred to as ‘dioxins’ – result mainly from combustion and waste incineration of chlorinated materials (i.e. PVC) and from industrial processes (e.g. metal reclamation, production of pentachlorophenol, pulp and paper bleaching), and like PCBs are found mainly in industrialised countries (Fig. 13.3). Restrictions on releases of dioxins (polychlorodibenzodioxins, PCDDs) and dibenzofurans (polychlorodibenzofurans, PCDFs) have been introduced. As a consequence concentrations in dietary sources for humans have declined over the past two decades. This has led to decreased concentrations in human milk (Fürst, 2006; LaKind, 2007).

13.4.4 Polybrominated diphenyl ethers PBDEs are flame retardants. In 1998 Swedish scientists noticed for the first time substances related to penta brominated diphenyl-ethers (pentaBDE) were accumulating in human milk. The levels reflect the environmental contamination and background exposure of the general population. In contrast to organochlorine compounds, the concentrations of PBDEs have increased during the period 1972–1997, indicating a doubling of the levels by five years (Meironyté et al., 1999; Norén and Meironyté, 2000; Hooper and Jianwen, 2002; Akutsu et al., 2003; Sjödin et al., 2005). Geographical distribution shows that concentrations of PBDEs are relatively uniform and the levels are of the same order in Asia and Europe, but are one or two orders of magnitude higher in North America (Fig. 13.4). Although the number of congeners analysed by different research groups for calculation of total polybrominated diphenyl ethers (Σ-PBDE) may differ, the congener pattern is generally dominated by the most abundant congener BDE-47.

13.4.5 Non-persistent pesticides Few data are available on levels of non-persistent pesticides in human milk (Anderson and Wolff, 2000). Most non-persistent pesticides are soluble in water and therefore may partition to the water fraction of human milk. However some non-persistent organophosphates (OPs) such as malathion

Endocrine disruptors in human milk

341

(log Pow = 4.5) could partition into the lipid fraction of human milk and possibly other non-persistent pesticides such as herbicides and fungicides may also be found in human milk, although available data are extremely limited. The dose an infant might receive would therefore be low and probably insignificant in toxicological terms.

13.4.6 Phthalates Data on human milk contamination with phthalate monoesters and metabolites is very limited, and because of substantial day-to-day variation individual exposure assessment based on one sample is imprecise. All phthalate monoesters can be found in human milk with large variations namely, monomethyl phthalate (mMP), monoethyl phthalate (mEP), mono-n-butyl phthalate (mBP), monobenzyl phthalate (mBzP), mono-(2-ethylhexyl) phthalate (mEHP), and monoisononyl phthalate (mNP) (Calafat et al., 2004). By comparison, levels in consumer milk and infant formula are lower (Mortensen et al., 2005). Utilising the latest monitoring data, median daily intake estimates for adults for diethyl-hexyl-phthalate (DEHP) are 5.6 μg/kg body weight/day (Franco et al., 2007) For breast-fed infants the mean daily intake over the first 6-month period was estimated at 25 μg d−1 for DEHP (Zhu et al., 2006). This is equal to the current acceptable daily intake (ADI) and a matter of concern. Whether or not phthalates at this exposure level might have adverse effects in infants and later in life is unknown. More systematic identification of endocrine disrupting properties of phthalate exposure during fetal development and infancy is of high priority.

13.4.7 Pharmaceuticals and personal care products Pharmaceuticals and personal care products (PPCP) comprise a diverse collection of thousands of chemical substances, including prescription and over-the-counter therapeutic drugs, fragrances, cosmetics, sunscreen agents, nutraceuticals, and biopharmaceuticals. Several frequently used UV filter (sunscreens) possess estrogenic activity in vitro and in vivo, in the range of other known xenoestrogens (Schlumpf et al., 2001). However no published data on human milk levels are available. Considering the widespread use of PPCP, occurrence of these chemicals – especially of 4-methylbenzylidene camphor (4-MBC) – in human milk should be investigated more extensively, because this compound is a potent inhibitor of the pituitary–thyroid– axis in rats and might cause effects comparable to hypothyroidism (Hamann et al., 2006). (Please see Chapter 19 for a more in-depth discussion of this topic.) Synthetic lipophilic musk fragrances are used in many consumer products, including perfumes and deodorants, and to scent laundry detergents. Commonly used nitro musk are musk xylene (MX) and musk ketone (MK),

342

Endocrine-disrupting chemicals in food

and common polycyclic musks are Galaxolide (HHCB, 1,3,4,6,7,8hexahydro-4,6,6,7,8,8-hexamethylcylopenta(γ)-2-benzo-pyran), Tonalide (Musk Plus, AHTN, 6-acetyl-1,1,2,4,4,7-hexamethyltetraline), Phantolide, Celestolide (Crysolide) and Traesolide. Few studies provide data on the occurrence of synthetic musks in human milk samples, a recent study showed that synthetic musks can presently be found in most of human milk samples from California (Reiner et al., 2007). The concentrations of HHCB (Galaxolide) were higher than the concentrations of the other synthetic musks. Interestingly, the mean concentration of HHCB (220 ng/g, lipid weight) was five times greater than the concentrations reported 10 years ago for human milk samples collected in Germany and Denmark (DuedahlOlesen et al., 2005; Fromme et al., 2007). Maternal age was not correlated with the concentrations of MX, MK, Galaxolide or Tonalide. Based on the residue patterns and accumulation features, it can be concluded that the exposure characteristics for synthetic musks are different from those of POPs, and that the major source of exposure to synthetic musks is probably via dermal absorption or by inhalation (Käfferelein and Angerer, 2001).

13.4.8 Persistent perfluorinated chemicals Perfluorooctanesulfonyl fluoride-based compounds have been used in a wide variety of consumer products, such as carpets, upholstery, garments and textiles. These compounds degrade to perfluorooctanesulfonate (PFOS), a lipophobic but persistent metabolite that accumulates in humans and wildlife. Only limited data exist on lactation as an exposure source of persistent perfluorinated chemicals (PFCs) for infants. PFOS and perfluorohexanesulfonate (PFHxS) were found in similar concentrations in Europe and the US, while perfluorooctanesulfonamide (PFOSA), perfluorooctanoic acid (PFOA), and perfluorononanoic acid (PFNA) were detected less frequently. Although total PFC in human milk is on average 1% of the corresponding serum level, lactation is a relevant source of exposure for infants because of persistence and biaccumulation in the organism.

13.4.9 Phenol compounds Bisphenol A (BPA), a primary monomer in polycarbonate plastic and epoxy resins, and nonylphenol (NP), a degradation product of the non-ionic surfactant alkylphenol polyethoxylates, occur in products to which humans are widely exposed. Unfortunately there is still no systematic survey on human milk contamination since the analysis of BPA in fat-matrix is a challenging task. In Japan in three human milk samples provided by different volunteers BPA was analysed and detected in two samples at low concentrations of 0.65 and 0.70 ng/g (Otaka et al., 2003). In 20 human milk samples from the US, BPA, o-phenylphenol (OPP) and 2-hydroxy-4-methoxybenzophenone

Endocrine disruptors in human milk

343

(BP-3) were detected in more than 60% of the samples tested. The free (unconjugated) species of these compounds appear to be most prevalent in milk (Ye et al., 2006). BPA might be of specific interest because low protein binding might favour estrogen-receptor availability (Nagel et al., 1998). Analysis of more samples is required to better understand the general contamination level of human milk.

13.5

Assessment of exposure

Measuring the amount of chemicals in the body by examining blood, urine, body tissue or human milk is known as biomonitoring. The level of chemicals in a person’s body is referred to as chemical body burden. Biomonitoring studies using human milk have shown contamination with a multitude of chemicals and proven the reliability of human milk as a marker of human exposures to toxic chemicals. But few countries have national human milk monitoring programmes similar to Germany (BfR, 2000), Sweden (Norén and Meironyté 2000) and New Zealand (Bates et al., 2002). Researchers and health authorities may use human milk sampling as a measure of community-wide contamination because it is a rapid, sensitive and less invasive method than drawing blood or obtaining a fat biopsy. The interest in establishing such a system is not based on concerns about hazards per se to breastfed infants, but to publicise the need to clean up the environment or to provide a simple, low cost means of tracking the impact of POPs in the human food chain and other pathways of exposure. Using individual milk data, exposure assessment can be done for the mother. For persistent contaminants human milk fat levels mirror the body burden of the mother since the partitioning ratio between adipose and human milk lipid is close to 1 : 1, but might reach 3.5 : 1 for the highly chlorinated PCB and PCDD/F congeners (Wittsiepe et al., 2007). For other non-lipophilic compounds, especially water-soluble and protein-bound xenobiotics, human milk concentration can be used to estimate dose (mg/kg body weight) for the infant. However, it is often difficult to draw conclusions about national and international trends, because of the many factors affecting measured levels, and because of limitations in the way data are generated and reported (Sim and McNeill, 1992; LaKind et al., 2001): • No accepted and standardised method for conducting breast-milk monitoring studies has been established. • The data are not representative for the country, especially if there is a demarcation between agricultural and urban zones. With very small sample sizes, it is difficult to draw conclusions about an entire population. • Differences in measurement methodology and data reporting (i.e. total DDT and PCBs) complicate data comparisons.

344

Endocrine-disrupting chemicals in food

• The selection of study participants may introduce a bias by selecting women with potentially high exposure to the chemical of interest, or voluntary participation favouring women with higher education and social class. To fully understand the extent of contamination and to keep human milk pure, harmonised national and international breast milk biomonitoring programmes for constant vigilance are essential and needed. To overcome the limitations outlined above, appropriate proposals have been made (Needham et al., 2002; La Kind et al., 2004) and should be followed as guidance for future surveys.

13.6

Risk assessment

In terms of the health of the mother, there are little data that can be used to link levels of most environmental chemicals in human milk to a particular health outcome in the mother herself (Khanjani and Sim, 2006, 2007). Any chemical stored in human body fat can potentially be transferred to the newborn infant during breastfeeding. Thus women pass on hazardous environmental contaminants and naturally occurring substances during pregnancy and nursing. In general, the period of highest susceptibility to adverse effects from environmental exposures is thought to be the in utero period (Jacobson and Jacobson, 1996). Depending on the parity and age of the mother, the breastfed infant may then receive close to an adult body burden at the earliest stage of life. Estimates for the average total daily intake of various POPs from all media indicate that breastfed infants in particular have relatively much greater exposure to these contaminants than do adults. For PCBs and dioxins this exceeds many times the amount that adults receive, and is thus above accepted for tolerable daily intakes (Brouwer et al., 1998). While the exposure through human milk is much greater than during development in the womb, in utero exposure is relatively more significant due to the greater vulnerability of the fetal brain and central nervous system to the toxic effects of contaminants such as PCBs and dioxins. Extremely high DDT exposure has been associated with shortened duration of lactation and difficulty producing breast milk (Gladen and Rogan, 1995; Lopez-Carrillo et al., 1996). Barely no data exist linking exposure via human milk with specific health outcomes in the infant, and so far only in very rare and extreme situations involving high levels of contamination like the Yusho incident have effects on infants occurred through human milk consumption. Usually, risks are estimated based on dose, rather than on residues of environmental chemicals in the body which reflect the cumulative dose. The relationship between dose and human tissue levels is complex and data on the latter are often not available.

Endocrine disruptors in human milk

345

For the infant, information on dose is available because the infant is only exposed to environmental chemicals via milk as a ‘dose’ from which risk estimates can be derived. However, the traditional risk assessment approach is not designed to consider the benefits to the infant associated with breastfeeding and is complicated by the relatively short-term exposures to the infant from breastfeeding. Additionally it is difficult to separate the influence of breastfeeding on health outcomes from in utero exposures. Thus, the concept of ‘risk assessment’ as it applies to human milk biomonitoring is not straightforward, and methodologies for undertaking this type of assessment have not yet been fully developed (LaKind et al., 2005). Despite these caveats the breastfed infant is the reference against which all alternative feeding methods must be measured with regard to growth, health, development, and all other short- and long-term outcomes. Studies show that breastfeeding counteracts some of the negative effects of exposure to environmental contaminants in utero. Compared with formula-fed infants, neurological insults to infants attributed to maternal exposure during fetal development were attenuated by breastfeeding (Dorea, 2006). Epidemiologic research shows that human milk and breastfeeding of infants provide advantages with regard to general health, growth and development, while significantly decreasing risk for a large number of acute and chronic diseases (Wegienka et al., 2006; Ferguson and Molfese, 2007). A variety of fucosylated oligosaccharides, specific to human milk, form part of the innate immune system. The presence of leptin, adiponectin, ghrelin, insulin and insulin-like growth factor (IGF-I) in human milk has been suggested to modulate the infant’s growth, appetite, and regulation of energy balance (Bernt and Walker, 1999; Grummer-Strawn and Mei, 2004; Miralles et al., 2006; Weyermann et al., 2006). Breastfeeding is a guarantor of health benefits, whereas its alternative (infant formula) is a predictor of some health limitations (Dorea, 2006). Use of formula milk as a substitute for human milk is not a recommended solution to contaminants in human milk. Bottlefed infants are also likely to be exposed to endocrine disruptors (e.g. BPA) leaching from the bottle or phytoestrogens from soy-based formula. Few, if any, adverse effects have been documented as being associated solely with consumption of human milk containing background levels of environmental chemicals, and none have been clinically or epidemiologically demonstrated. However, especially for phytoestrogens, BPA and phthalates, and for combined exposures, studies are urgently needed to investigate subtle effects that could occur during development or that could surface later in life (Badger et al., 2002; Massart et al., 2005).

13.7

Current recommendations on breastfeeding

Unfortunately, it is possible that the concerns expressed by scientist may be sensationalised or misinterpreted by interest groups and the media and

346

Endocrine-disrupting chemicals in food

cause undue concern among breastfeeding mothers. The world literature is replete with studies on the short- and long-term health benefits of breastfeeding to both mother and infant. These benefits include protection against acute illnesses, long-term health protection, positive psychological relationship and even savings in healthcare costs (Lawrence, 2000). Although economic, cultural and political pressures often confound decisions about infant feeding, the WHO and many other professional health organisations adhere to the position that breastfeeding ensures the best possible health as well as the best developmental and psychosocial outcomes for the infant. Currently, the WHO recommends that a child be breastfed for at least two years: ‘as a global public health recommendation, infants should be exclusively breastfed for the first six months of life to achieve optimal growth, development and health. Thereafter, to meet their evolving nutritional requirements, infants should receive nutritionally adequate and safe complementary foods while breastfeeding continues up to two years or beyond.’ (WHO, 2006)

The American Academy of Pediatrics explicitly recommends that infants receive human milk as the sole source of nourishment through the first six months of age, then receive human milk as the sole source of milk through 12 months, and support for breastfeeding as long as mutually desired by mother and child (AAP, 2005; Shaikh and Chantry, 2006). Mothers concerned about exposure to environmental contaminants should be advised of the proven benefits of breastfeeding and the possible limitations of formula feeding. Women with average environmental exposure do not need to worry about having their human milk screened for endocrine-disrupting compounds and other pollutants. For the very rare cases of known poisonous exposures, diagnostic testing of human milk may be indicated (Byczkowski et al., 1994). Because non-commercial fish and wildlife ingestion can be a very significant environmental source of pollutants (Falk et al., 1999) health professionals should advise women at childbearing age and especially pregnant and nursing women to follow fish and wildlife consumption guidelines – such as the US advisories (EPA, 1998, http://www.epa.gov/ost/fish).

13.8

Conclusions

Human milk is a reliable marker of human exposures to toxic chemicals and as such a tool for monitoring community health. The presence of contaminants in human milk has broad public health implications throughout the world. However, to date systematic research has been conducted only on few contaminants. Monitoring human milk should be continued and based on protocols for sampling and analysis, and concepts for consideration in interpretation and

Endocrine disruptors in human milk

347

communication of study results, as developed by the WHO and by the Technical Workshop on Human Milk Surveillance and Research for Environmental Chemicals in the US (Needham et al., 2002; LaKind et al., 2004, 2005; Barr et al., 2005; Berlin et al., 2005). Since robust analytical methods exist only for some endocrine disruptors and environmental contaminats in human milk, a process for prioritising the compounds and development of analytical methods is necessary. Risk communication must inform correctly and effectively about the situation of xenobiotics in human milk and must likewise counteract the likelihood of negatively impacting already low breastfeeding rates. Continued environmental monitoring and biomonitoring is necessary to ensure that known exposures to EDCs in human milk do not increase unexpectedly or new ones develop unnoticed.

13.9

References

aap (american academy of pediatrics) (2005), AAP Policy Statement of Breastfeeding and the Use of Human Milk. Pediatrics 115:496–506 adlercreutz h, yamad t, wahala k, watanabe s (1999), Maternal and neonatal phytoestrogens in Japanese women during birth. Am J Obstet Gynecol 180:737–43 agatonovic-kustrin s, ling lh, tham sy, alany rg (2002), Molecular descriptors that influence the amount of drugs transfer into human breast milk. J Pharm Biomed Anal 29(1–2):103–19 aglmb (arbeitsgemeinschaft der leitenden medizinalbeamtinnen und – beamten der länder. ausschuss fuer umwelthygiene) (1995), Standards zur Expositionsschätzung. Behörde für Arbeit, Gesundheit und Soziales, Hamburg akutsu k, kitagawa m, nakazawa h, makino t, iwazaki k, oda h, hori s (2003), Time-trend (1973–2000) of polybrominated diphenyl ethers in Japanese mother’s milk. Chemosphere 53(6):645–54 anderson gd (2006), Using pharmacokinetics to predict the effects of pregnancy and maternal–infant transfer of drugs during lactation. Expert Opinion Drug Metab Toxicol 2:947–60 anderson ha, wolff ms (2000), Environmental contaminants in human milk. J Expo Anal Environ Epidemiol 10:755–60 atkinson hc, begg ej (1990), Prediction of drug distribution into human milk from physicochemical characteristics. Clin Pharmacokin 18(2):151–67 azuma k (1999), Scientific study of Health and Welfare Ministry ‘Research for Dioxins in Breast Milk’ brief of the results. http://www.mhw.go.jp/houdou/ 1108/h0802-1_18.html badger tm, ronis mj, hakkak r, rowlands jc, korourian s (2002), The health consequences of early soy consumption. J Nutr 132(3):559S–65S barr db, wang ry, needham ll (2005), Biologic monitoring of exposure to environmental chemicals throughout the life stages: requirements and issues for consideration for the National Children’s Study. Environ Health Perspect 113:1083–91 bates mn, thomson b, garrett n (2002), Reduction in organochlorine levels in the milk of New Zealand women, Arch Environ Health 57:591–8 begg ej, duffull sb, hackett lp, ilett kf (2002), Studying drugs in human milk: time to unify the approach. J Hum Lact 18(4):323–32

348

Endocrine-disrupting chemicals in food

berlin cm, lakind js, fenton se, wang ry, bates mn, brent rl, condon m, crase bl, dourson ml, ettinger as, foos b, furst p, giacoia gp, goldstein da, haynes sg, hench kd, kacew s, koren g, lawrence ra, mason a, mcdiarmid ma, moy g, needham ll, paul im, pugh lc, qian z, salamone l, selevan sg, sonawane b, tarzian aj, rose tully m, uhl k (2005), Conclusions and recommendations of the expert panel: technical workshop on human milk surveillance and biomonitoring for environmental chemicals in the United States. J Toxicol Environ Health A 68(20):1825–31 bernt km, walker wa (1999), Human milk as a carrier of biochemical messages. Acta Paediatr Suppl 88:27–41 bfr (bundesinstitut für risikobewertung) (2000), Trends der Rückstandsgehalte in Frauenmilch der Bundesrepublik Deutschland – Aufbau der Frauenmilch – und Dioxin-Humandatenbank am BgVV, Berlin. http://www.bfr.bund.de/cd/424 bi x, qu w, sheng g, zhang w, mai b, chen d, yu l, fu j (2006), Polybrominated diphenyl ethers in South China maternal and foetal blood and breast milk. Environ Pollut 144(3):1024–30 boersma er, lanting ci (2000), Environmental exposure to polychlorinated biphenyls (PCBs) and dioxins. Consequences for longterm neurological and cognitive development of the child lactation. Adv Exp Med Biol 478:271–87 borgert cj, lakind js, witorsch rj (2003), A critical review of methods for comparing estrogenic activity of endogenous and exogenous chemicals in human milk and infant formula. Environ Health Perspect 111(8):1020–36 brouwer a, ahlborg ug, van leeuwen fx, feeley mm (1998), Report of the WHO working group on the assessment of health risks for human infants from exposure to PCDDs, PCDFs and PCBs. Chemosphere 37(9–12):1627–43 burke er, holden aj, shaw ic (2003), A method to determine residue levels of persistent organochlorine pesticides in human milk from Indonesian women. Chemosphere 50(4):529–35 butte w, fooken c (1990), Simultaneous determination of pentachlorophenol and neutral organochlorine compounds in human milk. Fresenius J Anal Chem 336:511–14 byczkowski jz, gearhart jm, fisher jw (1994), Occupational exposure of infants to toxic chemicals via breast milk. Nutrition 10(1):43–8 calafat am, slakman ar, silva mj, herbert ar, needham ll (2004), Automated solid phase extraction and quantitative analysis of human milk for 13 phthalate metabolites. J Chromatogr B Analyt Technol Biomed Life Sci 805(1):49–56 cerrillo i, granada a, lopez-espinosa mj, olmos b, jimenez m, cano a, olea n, fatima olea-serrano m (2005), Endosulfan and its metabolites in fertile women, placenta, cord blood, and human milk. Environ Res 98(2):233–9 chao hr, wang sl, lee wj, wang yf, päpke o (2007), Levels of polybrominated diphenyl ethers (PBDEs) in breast milk from central Taiwan and their relation to infant birth outcome and maternal menstruation effects. Environ Int 33(2):239–45 clewell ra, gearhart jm (2002), Pharmacokinetics of toxic chemicals in breast milk: use of PBPK models to predict infant exposure. Environ Health Perspect 110(6):A333–7 costopoulou d, vassiliadou i, papadopoulos a, makropoulos v, leondiadis l (2006), Levels of dioxins, furans and PCBs in human serum and milk of people living in Greece. Chemosphere 65(9):1462–9 coward l, barnes nc, setchell kdr, barnes s (1993), Genistein, daidzein, and their B-glycoside conjugates: antitumor isoflavones in soybean foods from American and Asian diets. J Agric Food Chem 41:1961–7

Endocrine disruptors in human milk

349

damgaard in, skakkebæk ne, toppari j, virtanen he, shen h, schramm kw, petersen jh, jensen tk, main km, study group nc (2006), Persistent pesticides in human breast milk and cryptorchidism. Environ Health Perspect 114(7):1133–8 darnerud po, lignell s, atuma s, aune m, cnattingius s, glynn a (2004), Levels and trends of PCB 28, PCB 153 and DDE in breast milk from primiparae women in Uppsala County, Sweden. Organohalogen Compounds 66:2884–90 dayan ad (2007), Risk assessment of triclosan [Irgasan] in human breast milk. Food Chem Toxicol 45(1):125–9 dewey kg, heinig mj, nommsen la, lonnerdal b (1991), Adequacy of energy intake among breast-fed infants in the DARLING study: relationships to growth velocity, morbidity, and activity levels. J Pediatr 119:538–47 dorea jg (2006), Maternal exposure to endocrine-active substances and breastfeeding. Am J Perinatol 23(5):305–12 duedahl-olesen l, cederberg t, pedersen kh, hojgard a (2005), Synthetic musk fragrances in trout from Danish fish farms and human milk. Chemosphere 61(3):422–31 epa (1998), Update: National listing of wildlife advisories. Environmental Protection Agency Washington, DC epa (2006), Child-specific Exposure Factors Handbook (EPA/600R-06/096A). http:// cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=56747 eskenazi b, marks ar, bradman a, fenster l, johnson c, barr db, jewell np (2006), In utero exposure to DDT and DDE and neurodevelopment among young Mexican American children. Pediatrics 118:233–41 falk c, hanranhan l, anderson ha, kanarek ms, draheim l, needham l, patterson d (1999), Body burden levels of dioxin, furans, and PCBs among frequent consumers of great lakes sport fish. Environ Res 80:S19–25 fängström b, strid a, grandjean p, weihe p, bergman a (2005), A retrospective study of PBDEs and PCBs in human milk from the Faroe Islands. Environ Health 14:4–12 ferguson m, molfese pj (2007), Breast-fed infants process speech differently from bottle-fed infants: evidence from neuroelectrophysiology. Dev Neuropsychol 31(3):337–47 focant jf, pirard c, thielen c, de pauw e (2002), Levels and profiles of PCDDs, PCDFs and cPCBs in Belgian breast milk. Estimation of infant intake. Chemosphere 48(8):763–70 franke a (1997), Isoflavone content of breast milk and soy formulas: benefits and risks. Clin Chem 43(5):850–1 franke aa, custer lj (1996), Daidzein and genistein concentrations in human milk after soy consumption. Clin Chem 42:955–64 franke aa, custer lj, tanaka y (1998), Isoflavones in human breast milk and other biological fluids. Am J Clin Nutr 68:1466S–73S franke aa, halm bm, custer lj, tatsumura y, hebshi s (2006), Isoflavones in breastfed infants after mothers consume soy. Am J Clin Nutr 84(2):406–13 franco a, prevedouros k, alli r, cousins it (2007), Comparison and analysis of different approaches for estimating the human exposure to phthalate esters. Environ Int 33(3):283–91 fromme h, bolte g, koch hm, angerer j, boehmer s, drexler h, mayer r, liebl b (2007), Occurrence and daily variation of phthalate metabolites in the urine of an adult population. Int J Hyg Environ Health 210(1):21–33 fürst p (2006), Dioxins, polychlorinated biphenyls and other organohalogen compounds in human milk. Levels, correlations, trends and exposure through breastfeeding. Mol Nutr Food Res 50(10):922–33

350

Endocrine-disrupting chemicals in food

fürst p, fürst c, wilmers k (1994), Human milk as a bioindicator for body burden of PCDDs, PCDFs, organochlorine pesticides, and PCBs. Environ Health Perspect J 102:187–93 gladen bc, rogan wj (1995), DDE and shortened duration of lactation in a northern Mexican town. Am J Public Health 85(4):504–8 grummer-strawn lm, mei z (2004), Does breastfeeding protect against pediatric overweight? Analysis of longitudinal data from the centers for disease control and prevention pediatric nutrition surveillance system. Pediatrics 113: 81–6 guvenius dm, aronsson a, ekman-ordeberg g, bergman a, noren k (2003), Human prenatal and postnatal exposure to polybrominated diphenyl ethers, polychlorinated biphenyls, polychlorobiphenylols, and pentachlorophenol. Environ Health Perspect 111(9):1235–41 hamann ih, schmutzler c, kirschmeyer p, jarry h, koehrle j (2006), 4Methylbenzylidene-camphor (4MBC) pituitary effects comparable to hypothyroidism. Endocrine Abstracts 11 OC60 harden fa, toms lm, symons r, fürst p, berry y, muller jf (2007), Evaluation of dioxin-like chemicals in pooled human milk samples collected in Australia. Chemosphere 67(9):S325–33 hooper k (1999), Breast Milk Monitoring Programs (BMMPs): world-wide early warning system for polyhalogenated POPs and for targeting studies in children’s environmental health. Environ Health Perspect 107(6):429–30 hooper, k, jianwen s (2002), Lessons from the polybrominated diphenyl ethers (PBDEs): precautionary principle, primary prevention, and the value of community-based body-burden monitoring using breast milk. Environ Health Perspect 111:109–14 hooper k, mcdonald ta (2000), The PBDEs: an emerging environmental challenge and another reason for breast-milk monitoring programs. Environ Health Perspect 108(5):387–92 howard cr, lawrence ra (2001), Xenobiotics and breastfeeding. Pediatr Clin North Am 48(2):485–504 hsu jf, guo yl, liu ch, hu sc, wang jn, liao pc (2007), A comparison of PCDD/ PCDFs exposure in infants via formula milk or breast milk feeding. Chemosphere 66(2):311–19 ingelido am, ballard t, dellatte e, di domenico a, ferri f, fulgenzi ar, herrmann t, iacovella n, miniero r, paepke o, porpora mg, de felip e (2007), Polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) in milk from Italian women living in Rome and Venice. Chemosphere 67:S301–6 inoue k, harada k, takenaka k, uehara s, kono m, shimizu t, takasuga t, senthilkumar k, yamashita f, koizumi a (2006), Levels and concentration ratios of polychlorinated biphenyls and polybrominated diphenyl ethers in serum and breast milk in Japanese mothers. Environ Health Perspect 114(8):1179–85 ito s, lee a (2003), Drug excretion into breast milk – overview. Adv Drug Deliv Rev 55(5):617–27 jacobson jl, jacobson sw (1996), Intellectual impairment in children exposed to polychlorinated biphenyls in utero. N Engl J Med 335:783–9 jaga k, dharmani c (2003), Global surveillance of DDT and DDE levels in human tissues. Int J Occup Med Environ Health 16(1):7–20. Corrected and republished in: Int J Occup Med Environ Health 19(1):83 (2006) jaraczewska k, lulek j, covaci a, voorspoels s, kaluba-skotarczak a, drews k, schepens p (2006), Distribution of polychlorinated biphenyls, organochlorine pesticides and polybrominated diphenyl ethers in human umbilical cord serum, maternal serum and milk from Wielkopolska region, Poland. Sci Total Environ 372(1):20–31

Endocrine disruptors in human milk

351

jensen aa, slorach sa (1991), Chemical Contaminants in Human Milk. Boca Raton, Ann Arbor, Boston: CRC Press, Inc. käfferlein hu, angerer j (2001), Trends in the musk xylene concentrations in plasma samples from the general population from 1992/1993 to 1998 and the relevance of dermal uptake. Int Arch Occup Environ Health 4(7):470–6 kärrman a, ericson i, van bavel b, darnerud po, aune m, glynn a, lignell s, lindstrom g (2006), Exposure of perfluorinated chemicals through lactation: levels of matched human milk and serum and a temporal trend, 1996–2004, in Sweden. Environ Health Perspect 115(2):226–30 kalantzi oi, martin fl, thomas go, alcock re, tang hr, drury sc, carmichael pl, nicholson jk, jones kc (2004), Different levels of polybrominated diphenyl ethers (PBDEs), and chlorinated compounds in breast milk from two U.K. regions. Environ Health Perspect 112(10):1085–91 katritzky ar, dobchev da, hur e, fara dc, karelson m (2005), QSAR treatment of drugs transfer into human breast milk. Bioorg Med Chem 13(5):1623–32 khanjani n, sim mr (2006), Maternal contamination with dichlorodiphenyltrichloroethane and reproductive outcomes in an Australian population. Environ Res 101(3):373–9 khanjani n, sim mr (2007), Maternal contamination with PCBs and reproductive outcomes in an Australian population. J Expo Sci Environ Epidemiol 17(2):191–5 knight dc, eden ja, huang jl, waring ma (1998), Isoflavone content of infant foods and formulas. J Paed Child Health 34 (2): 135–8 kuruto-niwa r, tateoka y, usuki y, nozawa r (2007), Measurement of bisphenol A concentrations in human colostrum. Chemosphere 66(6):1160–4 lakind js (2007), Recent global trends and physiologic origins of dioxins and furans in human milk. J Expo Sci Environ Epidemiol 17:510–24 lakind js, berlin cm, naiman dq (2001), Infant exposure to chemicals in breast milk in the United States: what we need to learn from a breast milk monitoring program. Environ Health Perspect 109(1):75–88 lakind js, amina wilkins a, berlin cm (2004), Environmental chemicals in human milk: a review of levels, infant exposures and health, and guidance for future research. Toxicol Appl Pharmacol 198(2):184–208 lakind js, brent rl, dourson ml, kacew s, koren g, sonawane b, tarzian aj, uhl k (2005), Human milk biomonitoring data: interpretation and risk assessment issues. J Toxicol Environ Health A 68(20):1713–69 landrigan pj, sonawane b, mattison d, mccally m, garg a (2002), Chemical contaminants in breast milk and their impacts on children’s health: an overview. Environ Health Perspect 110:A313–15 larsen la, ito s, koren g (2003), Prediction of milk/plasma concentration ratio of drugs. Ann Pharmacother 37:1299–306 lawrence ra (2000), Breastfeeding: benefits, risks and alternatives. Curr Opin Obstet Gynecol 12(6):519–24 liebl b, ehrenstorfer s (1993), Nitro-musk compounds in breast milk. Gesundheitswesen 55(10):527–32 liebl b, mayer r, ommer s, sonnichsen c, koletzko b (2000), Transition of nitro musks and polycyclic musks into human milk. Adv Exp Med Biol 478:289–305 lind y, darnerud po, atuma s, aune m, becker w, bjerselius r, cnattingius s, glynn a (2003), Polybrominated diphenyl ethers in breast milk from Uppsala County, Sweden. Environ Res 93(2):186–94 longnecker mp, wolff ms, gladen bc, brock jw, grandjean p, jacobson jl, korrick sa, rogan wj, weisglas-kuperus n, hertz-picciotto i, ayotte p, stewart p, winneke g, charles mj, jacobson sw, dewailly e, boersma er, altshul lm,

352

Endocrine-disrupting chemicals in food

heinzow b, pagano jj, jensen aa (2003), Comparison of polychlorinated biphenyl levels across studies of human neurodevelopment. Environ Health Perspect 111(1):65–70 lopez-carrillo l, torres-arreola l, torres-sanchez l, espinosa-torres f, jimenez c, cebrian m, waliszewski s, saldate o (1996), Is DDT use a public health problem in Mexico? Environ Health Perspect 104(6):584–8 lundén a, norén k (1998), Polychlorinated naphthalenes and other organochlorine contaminants in Swedish human milk, 1972–1992. Arch Environ Contam Toxicol 34(4):414–23 main km, mortensen gk, kaleva mm, boisen ka, damgaard in, chellakooty m, schmidt im, suomi am, virtanen he, petersen dv, andersson am, toppari j, skakkebaek ne (2006), Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age. Environ Health Perspect 114(2):270–6 massart f, harrell jc, federico g, saggese g (2005), Human breast milk and xenoestrogen exposure: a possible impact on human health. J Perinatol 25(4):282–8 meironyté d, norén k, bergman a (1999), Analysis of polybrominated diphenyl ethers in Swedish human milk. A time-related trend study, 1972–1997. J Toxicol Environ Health A 58(6):329–41 mes j, davies dj, doucet j, weber d, mcmullen e (1993), Levels of chlorinated hydrocarbon residues in Canadian human breast milk and their relationship to some characteristics of the donors. Food Addit Contam 10(4):429–41 mgs-sa (ministerium für gesundheit und soziales des landes sachsen-anhalt) (2001), Umweltmedizinische Untersuchungen im Landkreis Bitterfeld. Untersuchungen auf Chlororganische Schadstoffe, Nitromoschusverbindungen und Polychlorierte Dibenzo-p-dioxine und – furane in Muttermilch sowie in Seren der Wöchnerinnen. http://www.sachsen-anhalt.de/LPSA/fileadmin/Files/ muttermilc_studie_01.pdf minh nh, someya m, minh tb, kunisue t, iwata h, watanabe m, tanabe s, viet ph, tuyen bc (2004), Persistent organochlorine residues in human breast milk from Hanoi and Hochiminh City, Vietnam: contamination, accumulation kinetics and risk assessment for infants. Environ Pollut 129(3):431–41 miralles o, sanchez j, palou a, pico c (2006), A physiological role of breast milk leptin in body weight control in developing infants. Obesity 14(8):1371–7 mortensen gk, main km, andersson am, leffers h, skakkebaek ne (2005), Determination of phthalate monoesters in human milk, consumer milk, and infant formula by tandem mass spectrometry (LC-MS-MS). Anal Bioanal Chem 382(4):1084–92 morton ms, leung ssf, davies dp, griffiths k, evans baj (1996), The determination of isoflavonoids and lignans in human breast milk from British and Chinese women by gas chromatography–mass spectrometry. Proc., 2nd Int Symp. on the Role of Soy in Preventing and Treating Chronic Disease (15–18 Sept 1996, Belgium):50–1 murphy pa, song t, buseman g, barua k (1997), Isoflavones in Soy-based infant formulas. J Agric Food Chem 45(12):4635–38 nagel sc, vom saal fs, welshons wv (1998), The effective free fraction of estradiol and xenoestrogens in human serum measured by whole cell uptake assays: physiology of delivery modifies estrogenic activity. Proc Soc Exp Biol Med 217(3):300–9 nakagawa r, hirakawa h, iida t, matsueda t, nagayama j (1999), Maternal body burden of organochlorine pesticides and dioxins. AOAC Int 82(3):716–24 nakai k, suzuki k, oka t, murata k, sakamoto m, okamura k, hosokawa t, sakai t, nakamura t, saito y, kurokawa n, kameo s, satoh h (2004), The Tohoku Study

Endocrine disruptors in human milk

353

of Child Development: A cohort study of effects of perinatal exposures to methylmercury and environmentally persistent organic pollutants on neurobehavioral development in Japanese children. Tohoku J Exp Med 202(3):227–37 nakatani t, okazaki k, ogaki s, itano k, fujita t, kuroda k, endo g (2005), Polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, and coplanar polychlorinated biphenyls in human milk in Osaka City, Japan. Arch Environ Contam Toxicol 49(1):131–40 needham ll, ryan jj, furst p (2002), Technical Workshop on Human Milk Surveillance and Research on Environmental Chemicals in the United States. Guidelines for analysis of human milk for environmental chemicals. J Toxicol Environ Health A 65:1893–908 neville mc, keller r, seacat j, lutes v, neifert m, casey c, allen j, archer p (1988), Studies in human lactation: milk volumes in lactating women during the onset of lactation and full lactation. Am J Clin Nutr 48:1375–86 nguyenle t, wang e, cheung ap (1995), An investigation on the extraction and concentration of isoflavones in soy-based products. J Pharm Biomed Anal 14:221–32 nickerson k (2006), Environmental contaminants in breast milk. J Midwifery Women’s Health 51(1):26–34 norén k, meironyté d (2000), Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere 40:1111–23 notarianni lj, belk d, aird sa, bennett pn (1995), An in vitro technique for the rapid determination of drug entry into breast milk. Br J Clin Pharmacol 40:333–7 ntow wj (2001), Organochlorine pesticides in water, sediment, crops, and human fluids in a farming community in Ghana. Arch Environ Contam Toxicol 40(4):557–63 otaka h, yasuhara a, morita m (2003), Determination of bisphenol A and 4nonylphenol in human milk using alkaline digestion and cleanup by solid-phase extraction. Anal Sci 19:1663–6 ott m, failing k, lang u, schubring c, gent hj, georgii s, brunn h (1999), Contamination of human milk in Middle Hesse, Germany – a cross-sectional study on the changing levels of chlorinated pesticides, PCB congeners and recent levels of nitro musks. Chemosphere 38:13–32 polder a, odland jo, tkachev a, fa¸reid s, savinova tn, skaare ju (2003), Geographic variation of chlorinated pesticides, toxaphenes and PCBs in human milk from sub-arctic and arctic locations in Russia. Sci Total Environ 306(1–3):179–95 raab u, schwegler u, preiss u, albrecht m, fromme h (2007), Bavarian breast milk survey – pilot study and future developments. Int J Hyg Environ Health 210(3–4):341–4 raum e, seidler a, schlaud m, knoll a, wessling h, kurtz k, schwartz fw, robra bp (1998), Contamination of human breast milk with organochlorine residues: a comparison between East and West Germany through sentinel practice networks. Epidemiol Community Health 52:50S–55S reiner j, wong mc, arcaro kf, kannan k (2007), Synthetic Musk Fragrances in Human Milk from the United States. Environ Sci Technol 41(11):3815–20 rimkus g, wolf m (1996), Determination of polycyclic musk fragrances in human fat and human milk by GC-EI-MS. 18th International Symp. Capillary Chromatography 2:957–66 rimkus g, rimkus b, wolf m (1994), Nitro musks in human adipose tissue and breast milk. Chemosphere 28(2):421–32

354

Endocrine-disrupting chemicals in food

rogan wj, chen a (2005), Health risks and benefits fo bis(4-chlorophenyl)-1,1,1trichloroethane (DDT). Lancet 336:736–7 rogan wj, ragan nb (2003), Evidence of effects of environmental chemicals on the endocrine system in children. Pediatrics 112:247–52 saleh m, kamel a, ragab a, el-baroty g, el-sebae ak (1996), Regional distribution of organochlorine insecticide residues in human milk from Egypt. J Environ Sci Health B 31(2):241–55 schade g, heinzow b (1998), Organochlorine pesticides and polychlorinated biphenyls in human milk of mothers living in northern Germany: current extent of contamination, time trend from 1986 to 1997 and factors that influence the levels of contamination. Sci Tot Environ 215:31–9 schecter a, pavuk m, papke o, ryan jj, birnbaum l, rosen r (2003), Polybrominated diphenyl ethers (PBDEs) in U.S. mothers’ milk. Environ Health Perspect 11(14):1723–9 schlumpf m, cotton b, conscience m, haller v, steinmann b, lichtensteiger w (2001), In vitro and in vivo estrogenicity of UV screens. Environ Health Perspect 109:239–44 schuhmacher m, kiviranta h, vartiainen t, domingo jl (2007), Concentrations of polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) in milk of women from Catalonia, Spain. Chemosphere 67:S295–300 scialli ar (1992), Social Guide to Reproductive and Developmental Toxicology. Boca Raton: CRC Press setchell kdr, welsh mb (1987), High-performance liquid chromatographic analysis of phytoestrogens in soy protein preparations with ultraviolet, electrochemical and thermospray mass spectrometric detection. J Chromatogr 386:315–23 setchell kdr, zimmer-nechemias l, cai j, heubi je (1997), Exposure of infants to phytoestrogens from soy-based infant formula. Lancet 350:23–7 shaikh u, chantry c (2006), Reflections on the American Academy of Pediatrics’ 2005 Policy Statement on ‘Breastfeeding and the Use of Human Milk’. J Hum Lact 22(1):108–10 she j, holden a, sharp m, tanner m, williams-derry c, hooper k (2007), Polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in breast milk from the Pacific Northwest. Chemosphere 67(9):S307–17 shen h, main km, virtanen he, damggard in, haavisto am, kaleva m, boisen ka, schmidt im, chellakooty m, skakkebaek ne, toppari j, schramm kw (2007), From mother to child: investigation of prenatal and postnatal exposure to persistent bioaccumulating toxicants using breast milk and placenta biomonitoring. Chemosphere 67(9):S256–62 sim mr, mcneil jj (1992), Monitoring chemical exposure using breast milk: a methodological review. Am J Epidemiol 136(1):1–11 sjödin a, lakind js, patterson dg, needham ll, wang r, paul im (2005), Current concentrations and changes in concentrations of PBDEs, persistent pesticides, and PCBs in human milk. Organohalogen Compounds 73:1745–8 skopp s, oehme m, fürst p (2002), Enantiomer ratios, patterns and levels of toxaphene congeners in human milk from Germany. J Environ Monit 4(3):389–94 slusser wm (2004), More evidence in support of AAP recommendations on breastfeeding. AAP Grand Rounds 11:30–1 smith d (1999), Worldwide trends in DDT levels in human milk. Int J Epidemiol 28:179–88 so mk, yamashita n, taniyasu s, jiang q, giesy jp, chen k, lam pk (2006), Health risks in infants associated with exposure to perfluorinated compounds in human breast milk from Zhoushan, China. Environ Sci Technol 40:2924–9 solomon gm, weiss pm (2002), Chemical contaminants in breast milk: time trends and regional variability. Environ Health Perspect 110(6):A339–47

Endocrine disruptors in human milk

355

sonawane br (1995), Chemical contaminants in human milk: an overview. Environ Health Perspect 103(Suppl 6):197–205 stuetz w, prapamontol t, erhardt jg, classen hg (2001), Organochlorine pesticide residues in human milk of a Hmong hill tribe living in Northern Thailand. Sci Total Environ 273:53–60 sudaryanto a, kajiwara n, takahashi s, muawanah, tanabe s (2007), Geographical distribution and accumulation features of PBDEs in human breast milk from Indonesia. Environ Pollut 151(1):130–8 sun y, irie m, kishikawa n, wada m, kuroda n, nakashima k (2004), Determination of bisphenol A in human breast milk by HPLC with column-switching and fluorescence detection. Biomed Chromatogr 18(8):501–7 tajimi m, watanabe m, oki i, ojima t, nakamura y (2004), PCDDs, PCDFs and coPCBs in human breast milk samples collected in Tokyo, Japan. Acta Paediatr 93(8):1098–102 tanabe s, kunisue t (2007), Persistent organic pollutants in human breast milk from Asian countries. Environ Pollut 146(2):400–13 tsydenova ov, sudaryanto a, kajiwara n, kunisue t, batoev vb, tanabe s (2007), Organohalogen compounds in human breast milk from Republic of Buryatia, Russia. Environ Pollut 146(1):225–32 toms lm, harden fa, symons rk, burniston d, furst p, muller jf (2007), Polybrominated diphenyl ethers (PBDEs) in human milk from Australia. Chemosphere 68(5):797–803 vieth b (2007), German Commission on Human Biomonitoring, personal communication. wegienka g, ownby dr, havstad s, williams lk, johnson cc (2006), Breastfeeding history and childhood allergic status in a prospective birth cohort. Ann Allergy Asthma Immunol 97(1):78–83 weyermann m, beermann c, brenner h, rothenbacher d (2006), Adiponectin and leptin in maternal serum, cord blood, and breast milk. Clin Chem 52: 2095–102 who (2006), Planning Guide for National Implementation of the Global Strategy for Infant and Young Child Feeding (working draft). Geneva, World Health Organization wittsiepe j, fürst p, schrey p, lemm f, kraft m, eberwein g, winneke g, wilhelm m (2007), PCDD/F and dioxin-like PCB in human blood and milk from German mothers. Chemosphere 67(9):S286–94 yang yh, chang ys, kim bh, shin dc, ikonomou mg (2002), Congener-distribution patterns and risk assessment of polychlorinated biphenyls, dibenzo-p-dioxins and dibenzofurans in Korean human milk. Chemosphere 47(10):1087–95 ye x, kuklenyik z, needham ll, calafat am (2006), Measuring environmental phenols and chlorinated organic chemicals in breast milk using automated on-line column-switching-high performance liquid chromatography-isotope dilution tandem mass spectrometry. Chromatogr B Analyt Technol Biomed Life Sci 831(1–2):110–15 yu hf, zhao xd, zhao jh, zhu zq, zhao z (2006), Continuous surveillance of organochlorine pesticides in human milk from 1983 to 1998 in Beijing, China. Int J Environ Health Res 16(1):21–6 zhao c, zhang h, zhang x, zhang r, luan f, liu m, hu z, fan b (2006), Prediction of milk/plasma drug concentration (M/P) ratio using support vector machine (SVM) method. Pharm Res 23(1):41–8 zhu j, phillips sp, feng yl, yang x (2006), Phthalate esters in human milk: concentration variations over a 6-month postpartum time. Environ Sci Technol 40(17):5276–81

14 Assessing the risks of endocrinedisrupting chemicals A. Beronius, Karolinska Institutet, Sweden; C. Rudén, Royal Institute of Technology, Sweden; A. Hanberg, Karolinska Institutet, Sweden; J. Garai, University of Pecs, Hungary; and H. Håkansson, Karolinska Institutet, Sweden

Abstract: In this chapter, four European regulatory frameworks are compared concerning how they cope with identifying and risk assessing endocrine-disrupting chemicals. This comparison is performed by using four example chemicals, bisphenol A, dioxins, ethinyl estradiol and vinclozolin, representing four different regulatory categories. It is concluded that within these regulations, there are no general test requirements that specifically enable identification of endocrine disruptors, and there seem to be little conformity in the risk assessment processes even though the route of exposure and the mode of toxicity are similar. Further development of risk assessment guidance for endocrine disruptors requires continued cooperation between experimental scientists, risk assessors, regulators and industry. Key words: Chemicals legislation, risk assessment, hazard identification, test requirements, EDC.

14.1

Introduction

The European Commission (EC) has expressed its concern regarding the concentrations of endocrine-modulating compounds in our environment, and the environmental and health risks that the resulting exposures may pose (EC, 1999). Chemicals with the capacity to modulate the endocrine system are often referred to as endocrine-disrupting chemicals (EDCs), and the purpose of this chapter is to investigate how the regulatory frameworks in Europe cope with the task of identifying and risk assessing chemicals with this property by using four different EDCs as examples. There are currently no generally agreed upon regulatory procedures to identify and risk assess substances with EDC characteristics. The lack of

Assessing the risks of endocrine-disrupting chemicals

357

procedures includes what end points are crucial to investigate, and criteria for data interpretation. Within the European Union (EU), different legislations apply depending on the intended use of a chemical; there are, for example, separate rules for industrial chemicals, plant protection products and pharmaceuticals. As a result the regulatory risk assessment process, as well as underlying policies, criteria and requirements, may differ for chemicals with a similar mode-of-action, such as EDCs. To address this lack of regulatory coordination we use four EDCs as examples: bisphenol A (BPA), dioxins, ethinyl estradiol (EE) and vinclozolin. These chemicals represent four different chemical categories, namely existing industrial chemicals, environmental pollutants in food, i.e. contaminants originating from industrial processes, pharmaceuticals and existing active substances in plant protection products. The respective regulations of these chemicals are described and compared in terms of their scope and purpose, toxicity testing requirements and risk assessment guidelines. In addition, actual risk assessment documents for each compound are described and compared. The analyses presented in this chapter are limited to EU chemicals legislation. It should be noted that other rules or agreements might also be applicable to these chemicals, including international conventions and national legislations. These are however not under scrutiny here. The EU legislation for existing (and new) industrial chemicals was replaced by the REACH legislation on 1 June 2007. REACH stands for Registration, Evaluation, Authorization and Restriction of Chemicals. Since the evaluations in this chapter were performed in retrospect, scrutinizing risk assessments already made, it is not possible to include an industrial chemical risk assessed within the REACH system at this point. To evaluate the actual outcome of the new legislation is important, but can only be made some years after its implementation.

14.2

The four model compounds

In this section our model EDCs are briefly introduced, including information about their mechanism of toxicity, and the main sources of consumer exposure (see Table 14.1 for a summary).

14.2.1 Bisphenol A BPA is a high production-volume chemical mainly used in the manufacture of polycarbonate plastics and epoxy resins. Consumer exposure occurs by release of the BPA monomer (Fig. 14.1) from polycarbonate plastics, from certain dental sealants and from BPAcontaining linings of many food and beverage cans (EU, 2003). BPA is an

358

Endocrine-disrupting chemicals in food

Table 14.1 An overview of the four model compounds, their mechanism of action for toxicity and main sources of consumer exposure Bisphenol A

Dioxins

Ethinyl estradiol

Vinclozolin

Regulatory group

Existing industrial chemical

Environmental pollutant in food

Pharmaceutical

Mechanism of toxicity

Estrogen receptor agonist Food in contact with plastic products

Ah-receptor agonist

Estrogen receptor agonist Therapeutic (oral contraceptives)

Existing active substance in plant protection products Androgen receptor antagonist Residues in vegetables and fruit

Main source of human exposure

Food of animal origin

OH

HO

Fig. 14.1 The structure of bisphenol A (4,4′-dihydroxy-2,2-diphenylpropane).

CI

O

CI

CI

O

CI

Fig. 14.2 The structure of 2,3,7,8-TCDD, a dioxin.

estrogen agonist, exerting its toxicity by binding to and activating the estrogen receptor (ER) (EU, 2003).

14.2.2 Dioxins Dioxins are a group of planar, polyhalogenated hydrocarbons, which include dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) (Fig. 14.2) (SCF, 2000). Dioxins and furans are not deliberately produced; they are formed as by-products of reactions such as the combustion of organics, in pulp and paper production and in other industrial processes.

Assessing the risks of endocrine-disrupting chemicals

359

The dioxins are very resistant to both environmental and biological degradation. Hence they persist in the environment and may enter the food chain and bioaccumulate (SCF, 2000). As a result the main route of human exposure to dioxins is via food. About 90% of human exposure comes from foods of animal origin (SCF, 2000). The dioxins bind to and activate the cellular arylhydrocarbon (Ah) receptor, which results in the transcription of a number of genes and consequent cellular responses, such as the induction of cytochrome P450 enzymes. Dioxins have several endocrine-disrupting properties and the best characterized to date are their potential to act as anti-estrogens and their potential to interfere with thyroid hormone and retinoid systems. The individual compounds in this group of chemicals differ greatly in their ability to bind to the Ah-receptor and hence in their endocrine-disrupting potency.

14.2.3 Ethinyl estradiol EE (Fig. 14.3) is the synthetic form of the endogenous estrogen, estradiol. EE has been used as the estrogen ingredient in pharmaceuticals since the 1960s. The most well-known pharmaceutical use of EE is as oral contraceptives (EPC, 2005). Some pregnancies occur in spite of the use of oral contraceptives. Exposure of the human fetus to high doses of EE in utero is therefore possible and may raise concerns for adverse developmental effects in these cases. In this chapter, EE is considered a therapeutic agent, not as a contaminant in food, and the exposure conditions are therefore different from the other compounds, for which exposure via the diet is specifically considered. However, EE was included as a model compound because of its mechanism of action (estrogen receptor agonist), which may readily be compared with the other compounds, and because it belongs to a separate regulatory group (pharmaceuticals).

14.2.4 Vinclozolin Vinclozolin has been extensively used in the EU as a fungicide on crops such as grapes, berries, stone fruits and lettuce (EC, 1997). Use of CH3 OH

CH3

H H

H

HO

Fig. 14.3 The structure of ethinyl estradiol, a synthetic form of the endogenous hormone estradiol.

360

Endocrine-disrupting chemicals in food (b) Cl (a) Cl

O O N

Cl

N H O

O

CH3 O CH

Cl

Cl

O N H

Cl

COOH CH2

CH3 OH C

H CH2

Fig. 14.4 The structure of vinclozolin (a) and two of its endocrine-disrupting metabolites (b).

vinclozolin is banned within the EU since 2006. Vinclozolin is readily degraded in the environment and its two main metabolites (Fig. 14.4) are potent androgen antagonists (EC, 1997; Kelce et al., 1994). When administered to rats, vinclozolin is quickly metabolized into these two antiandrogenic metabolites (Kelce et al., 1994). Vinclozolin may disturb the function and development of tissues that are sensitive to testosterone.

14.3

Regulatory frameworks

In this section the regulatory frameworks relevant to the model substances are briefly introduced. Table 14.2 summarizes the main components of the different regulatory frameworks for the four model compounds. 14.3.1 Priority existing industrial substances As already mentioned, the industrial substances are as of 1 June 2007 regulated within the new European legislation, REACH. REACH replaces the previous rules for existing industrial chemicals (Council Regulation No 793/93/EEC, Commission Regulation No 1488/94/EC). In addition, REACH also replaces the rules for new industrial chemicals which are not discussed here. Still, for reasons explained above, the REACH system is not included in this evaluation. The industrial chemicals were divided into ‘new’ and ‘existing’ substances. Existing substances are those which were marketed in Europe in September 1981, when the European Inventory of Existing Chemical Substances (EINECS) was closed. The new substances were those introduced on the market after this time, and thus not included in EINECS. There are about 100 000 substances regulated as existing industrial chemicals and 141 of these were prioritized for risk assessment before the legisla-

Assessing the risks of endocrine-disrupting chemicals

361

Table 14.2 The regulatory frameworks and division of responsibilities between different European authorities and Member States for the four categories of compounds Priority existing industrial substances

Environmental pollutants in food

Active substances in pharmaceuticals

Legislation regulating risk assessment

Dir 67/548 Reg 793/93 Reg 1488/94

None

Dir 2001/83 Dir 2003/63 Dir 2004/27 Reg 726/2004

Time of implementation Purpose of the risk assessment

1993

Not applicable

2001

Scientific basis for health risk management European Commission

Derivation of a European TDI

Scientific basis for marketing authorization

European Food Safety Authority (EFSA)* Reactive

European Medicines Agency (EMEA) Proactive

Authority responsible for risk assessment Nature of risk assessment

Reactive

Existing active substances in plant protection products (stage 1) Dir 67/548 Dir 91/414 as amended by Dir 94/79 and Dir 97/57 Reg 3600/92 1991 Scientific basis for the authorization for use European Commission Reactive

* Before 2002 the Scientific Committee on Food of the Directorate-General for Health and Consumer Protection (DG SANCO) was responsible for the data collection and risk assessment of environmental pollutants.

tion ceased to be in force. Risk assessment of the industrial substances was carried out in order to evaluate if estimated exposure levels are acceptable in the light of existing hazard data, or if restrictions of manufacture and/or use are needed. This overall purpose of risk assessment is similar in REACH.

14.3.2 Environmental pollutants in food ‘Environmental pollutants in food’ are defined here as substances that are inadvertently formed during industrial processes and thereafter end up in the food chain. These chemicals have no intended use and no manufacturer and they are therefore not covered by the European chemicals legislation. However, when such a chemical has been identified as a contaminant in the food chain, other regulations will become applicable. EU Regulation 178/2002 establishes the European Food Safety Authority (EFSA), which, since 2002, has been responsible for the risk assessment of environmental pollutants in food, including these ‘orphan’ contaminants of industrial

362

Endocrine-disrupting chemicals in food

process origin. Before 2002 this responsibility belonged to the Health and Consumer Protection Directorate General (DG SANCO). Risk assessment of these compounds is carried out on a case-by-case basis with the aim to estimate consumer exposure and to derive guidance values, such as acceptable or tolerable daily intakes (ADI/TDI). Dioxins are well-known chemicals in this group.

14.3.3 Active substances in pharmaceuticals The safety assessment of human pharmaceuticals intended for the EU market is regulated by Directive 2001/83/EC (European legislation of medicinal products reaches back to 1965 and has been frequently and substantially amended in different directives. – Directive 2001/83/EC is the result of compiling these into one single text), Commission Directive 2003/63/EC, Directive 2004/27/EC and Regulation 726/2004/EC. The risk assessment of pharmaceuticals is referred to as a ‘safety assessment’ in the legislation. The toxicological safety assessment of the active substances in a pharmaceutical product is carried out in conjunction with the marketing authorization process for that product. Since 2004 the central authority responsible for marketing authorizations of pharmaceutical products in the EU is the European Medicines Agency (EMEA).

14.3.4 Existing active substances in plant protection products (stage 1) A plant protection product is an active substance or preparation containing one or more active substances used to protect plants or plant products from animals, other plants or microorganisms (Council Directive 91/414/EEC). The risk assessment of plant protection products is regulated by five directives and one regulation (Council Directive 67/548/EEC, Coucil Directive 91/414/EEC, Commission Directive 93/71/EEC, Commission Directive 94/79/EC, Council Directive 97/57/EC, Commission Regulation 3600/92/EEC). The active substances in plant protection products are divided into ‘new’ and ‘existing’ substances according to Directive 91/414. Existing substances are those that were already in use on the European market in 1993 when Directive 91/414 was implemented, and new substances are those introduced after 1993. According to Directive 91/414, evaluation of the existing active substances is made in order to decide whether their use should be allowed to continue in the EU. Approved active substances are added to Annex I of the directive. Altogether there are 984 existing substances to be evaluated. In accordance with Regulation 3600/92 the substances to be evaluated have been prioritized based on health and/or environmental concern, the presence of residues in food, data gaps, and the agricultural/economic importance of the product. Furthermore, chemicals having similar biological and chemical properties are grouped together. The first evaluation stage includes

Assessing the risks of endocrine-disrupting chemicals

363

the substances of highest concern and covers 90 substances, including vinclozolin.

14.3.5 Comparison The purpose of risk assessment is different in the four cases studied here: to provide a scientific basis for determining whether or not estimated exposure levels can be accepted in light of effects data, to establish guidance values, such as an ADI or TDI, to prove that the product is safe for human use, or safe to be included in plant protection products used in the EU. In all cases, except for pharmaceuticals, the risk assessment is reactive rather than proactive, i.e. a decision that a risk assessment has to be made is taken after the chemical has become commercially available and concerns regarding adverse health or environmental effects have been raised. Pharmaceuticals, on the other hand are assessed as part of the process to approve them for marketing and use. The situation would also be different for new industrial chemicals and plant protection substances, where the assessment is made before the product is approved for release onto the EU market.

14.4

Toxicity data requirements

In this section the data requirements for chemicals belonging to these four regulatory groups are briefly described. Table 14.3 summarizes the toxicity data requirements for each of the four regulatory frameworks investigated.

14.4.1 Priority existing industrial substances The data requirements of the 141 priority existing substances were specified in Council Directive 67/548. For these chemicals, the manufacturer was required to provide data at least corresponding to the so-called ‘base-set’. Base-set data included the following tests: acute toxicity by at least two routes of exposure, eye and skin irritation, skin corrosivity, skin sensitization, a 28-day toxicity study, mutagenicity in at least two in vitro tests, and finally a screening test for reproductive toxicity. (There were no toxicity data requirements for the almost 100 000 non-prioritized existing industrial substances.)

14.4.2 Environmental pollutants in food For the chemicals defined here as environmental industrial pollutants in food, e.g. contaminants originating from industrial processes, no ‘manufacturer’ can be identified to be held responsible for the production of data and consequently there are no legislated test requirements.

Required but no further criteria stated in legislation

28-day study

Repeated dose toxicity

Toxicokinetic behaviour to the extent that can be derived from base set and other available data At least two routes of exposure: orally and by inhalation or percutaneously Eye and skin

Skin sensitization

Corrosiveness/ irritation/local tolerance

Absorption, distribution, excretion, metabolism Acute toxicity/ single dose

Priority existing industrial substances

Not required

Not required

Short-term + long-term studies in at least two mammals (one non-rodent)

Required but no further criteria stated in legislation 90-day study in rat and dog

Administered orally, percutaneously and via inhalation (if relevant) Eye and skin

In at least two mammalian species and using at least two different routes of administration All observations made in acute tests (as above). In addition local tolerance tests in sites of the body that may come in contact with the substance Required in at least one test system for chemicals applied to the skin

Not required

Not required

Tests in one mammal (usually rat)

Existing active substances in plant protection products (stage 1)

Pharmacokinetics

Active substances in pharmaceuticals

Not required

Environmental pollutants in food

Table 14.3 Toxicity data required (as stated in legislation) from the manufacturer to be used as basis for risk assessment

Not required

Not required

Neurotoxicity

Human data

* If test is positive further testing must be carried out.

Screening test

One bacterial gene mutation test and one other in vitro test capable of detecting chromosomal aberrations* Not required

Toxicity to reproduction

Chronic toxicity/ carcinogenicity

Mutagenicity/ genotoxicity

Not required

Not required

Not required

Not required

Not required

Yes, clinical trials

If justified by chemical analogy, mutagenicity or other tox-test results, or if likely to be administered to patient over an extended period of time Embryo/foetal toxicity in two mammals (one nonrodent) and peri-/postnatal toxicity in one mammal. Reproductive toxicity by “appropriate” tests Not required

Obligatory for any new substance (case-by-case basis)

Required if the substance shows similarities to organophosphates Not required

2-generation study in rat and teratogenicity in rat and rabbit

One 2-year study in rat and one carcinogenicity study in mouse

In vitro mutagenicity tests*

366

Endocrine-disrupting chemicals in food

14.4.3 Active substances in pharmaceuticals The data requirements for active substances in pharmaceuticals are specified in Commission Directive 2003/63 (amending Directive 2001/83), and include information on pharmacokinetics, acute toxicity in at least two mammalian species and via two routes of exposure, local tolerance, a shortterm repeated dose test lasting 2 or 4 weeks and one long-term test, the duration of which depends on the conditions of clinical use. These tests should be carried out in two mammalian species, one of which has to be a non-rodent. Mutagenicity tests are also required, but requirements will depend on the state of scientific knowledge. Carcinogenicity should be evaluated if considered relevant according to available knowledge including results from initial testing, or if the product is proposed to be used for an extended period of time. Reproductive toxicity should be investigated using “appropriate” tests. Embryo/fetal toxicity testing is required in two mammalian species and perinatal toxicity in one. In addition to those non-clinical data, information on the efficacy and toxicity of the substance is also required from clinical trials on humans. Directive 2001/83 (Article 10 (a) (ii)) states that if the manufacturer demonstrates that a substance has a so-called ‘well established use, with recognized efficacy and an acceptable level of safety’ then there are no requirements to provide any non-clinical toxicological data for that substance (EMEA, 2005). Whether this is applicable for individual substances is determined on a case-by-case basis.

14.4.4 Existing active substances in plant protection products (stage 1) Toxicity data requirements for plant protection products are laid down in Council Directive 91/414 as amended by Commission Directives 93/71 and 94/79. Tests of absorption, tissue distribution, excretion and metabolic pathways are required, and acute toxicity should be tested at least orally and percutaneously. Furthermore, tests for eye and skin irritation, skin sensitization and in vitro mutagenicity, as well as 90-day repeated dose toxicity studies in both rat and dog, a 2-year chronic toxicity study in rat, and one carcinogenicity study in mouse, are required. One two-generation study in rat and teratogenicity studies in rat and rabbit are required to investigate reproductive toxicity. If the substance shows structural similarities to organophosphorus compounds, then its potential to cause delayed neurotoxicity after acute exposure should also be evaluated.

14.4.5 Comparison There are no legislated data requirements for environmental pollutants in food. Data requirements for prioritized existing industrial substances are limited, while for pharmaceuticals and plant protection products the data

Assessing the risks of endocrine-disrupting chemicals

367

requirements are extensive and include long-term and multigenerational testing. While there is a systematic review system in place for prioritized industrial chemicals, pharmaceuticals and plant protection products, there are no such systems in place for environmental pollutants. Besides the standardized, regulatory required test data, additional information may become available via research published in the open scientific literature. Obviously, such data rarely influence the pre-marketing assessment of pharmaceuticals, while for environmental pollutants and existing industrial chemicals, such sources can in some cases be an important and even dominating part of the scientific basis for risk assessment. When such data were obtained by experiments not performed according to currently accepted guidelines or good laboratory practice (GLP), the quality of the data needs to be assessed on a case-by-case basis. Data considered to have sufficient quality and relevance can be used for risk assessment and the manufacturer is then not obliged to carry out new tests. This practice is also applicable to pharmaceuticals even if the substance has a ‘well established use, with recognized efficacy and an acceptable level of safety’ (EMEA, 2005). Human data are only required for pharmaceutical substances.

14.5

Availability and scope of risk assessment guidelines

Risk assessment is a process performed with the overall purpose to protect human health and the environment. Health risk assessment is conducted in several steps: hazard assessment, exposure assessment and risk characterization. The hazard assessment entails a thorough review of the existing data on the substance. In parallel, the exposure assessment aims at assembling and evaluating the available information on human exposure. In the risk characterization, the hazard and exposure data are combined to reach a conclusion regarding the size and nature of risk. Guidance documents for risk assessment are issued to specify the legislation, and the purpose of issuing such guidelines is to further clarify what is required of the risk assessor and to promote predictability in the risk assessment process (EC, 2003).

14.5.1 Priority existing industrial substances An extensive Technical Guidance Document (EC, 2003) has been issued by the EC, Directorate-General of the Joint Research Centre. These guidelines specify predefined criteria for conducting risk assessments of existing and new industrial chemicals and biocides, but the reliance on expert judgement is still an integral part of the risk assessment process (EC, 2003). For example, if a study was not conducted according to GLP it is up to the risk assessor to judge the quality and relevance of such data on a case-by-case basis, as

368

Endocrine-disrupting chemicals in food

well as if there is a need to conduct new tests. Furthermore, a no observed adverse effect level (NOAEL) should be determined on which to base the risk characterization. Deriving the NOAEL entails determining if observed effects should be considered adverse or not and this decision may consequently be dependent on the expert judgement of the risk assessor. 14.5.2 Environmental pollutants in food As there is no legislation stating any requirements or criteria for the risk assessment of environmental pollutants there are consequently no guidelines for the risk assessment process available to the risk assessor. As a result the decision on what data to include, how to evaluate the data in terms of quality and what effects to focus on in the assessment is based on case-by-case expert judgement. 14.5.3 Active substances in pharmaceuticals Guidelines for evaluation of non-clinical and clinical data on pharmaceuticals are available from the EMEA (EMEA, 2003). These guidelines are issued in accordance with the International Conference on Harmonization (ICH) in order to harmonize the application process for registration of pharmaceutical products on the European, US and Japanese markets. For the purposes of the European market, these guidelines support the requirements and criteria stated in Commission Directive 2003/63. The guidance document is intended for the experts commissioned by the manufacturer to conduct the safety assessment. The purpose of these guidelines is to assist the authors in the preparation of the reports in an ‘acceptable format’ and they cover mainly what should be included in the report, i.e. which effects and aspects of toxicity to be considered. However, the expert is responsible for highlighting the most important findings. The guidelines do not state any criteria about which information to consider or how to evaluate the data. These issues are determined by expert judgement. 14.5.4 Existing active substances in plant protection products (stage 1) Guidelines for carrying out the risk assessment of active substances in plant protection products have been issued by the EC, Directorate-General for Agriculture (EC, 1998b). These guidelines outline general criteria for the layout, subject matter, terminology, and units of measurements, but state that the authority responsible for risk assessment is ‘required to use expert judgement in preparing the documentation concerned’ (EC, 1998b). 14.5.5 Comparison Guidelines for conducting risk assessments are available for three of the four investigated chemical groups. As there is no legislation covering the

Assessing the risks of endocrine-disrupting chemicals

369

requirements for risk assessment of environmental pollutants in food there are no guidelines for this process. The guidance document for priority existing industrial substances, the TGD (EC, 2003), is significantly more comprehensive and detailed than for any of the other compounds. One important aspect of the guidelines is to what extent the risk assessment process should make use of predefined criteria and how much that is left to case-by-case judgements. Reliance on predefined criteria contributes to making the risk assessment process predictable, but at the same time less flexible. A system based on case-by-case judgements on the other hand, makes the process more dependent on the knowledge, views and experiences of the person(s) conducting the assessment. Expert judgement is required in all four cases: however, for risk assessment of priority existing industrial substances the criteria are relatively detailed. For environmental pollutants, expert judgement is the primary basis for effects assessment. Expert judgement has also been given significant importance in assessments of pharmaceutical substances and of active substances in plant protection products. The availability and scope of guidelines are summarized in Table 14.4.

14.6

Endocrine-disrupting chemical effects assessments

To identify actual differences in the evaluation of scientific data and hazard assessment, a recent European risk assessment report, conducted according to EU regulations and directives, were identified for each of the four substances. The contents of each of these reports were thereafter scrutinized and the principles and criteria underpinning the assessments in them were compared. The risk assessment of BPA (EU, 2003) was attained from the website of the European Chemicals Bureau (ECB), the assessment of dioxins (SCF, 2000, 2001) was available from the website of the DirectorateGeneral for Health and Consumer Protection (DG SANCO) and the vinclozolin documentation (EC, 1997, 1998a) was provided by the Swedish Chemicals Agency. Concerning EE, it became apparent that the constituents in pharmaceutical products are not evaluated separately but as part of the evaluation for the product as a whole. The non-clinical and clinical documentation for an oral contraceptive, Yasmin® (Schering, 2000a, b), was therefore used. The following paragraphs briefly describe the risk assessors’ conclusions.

14.6.1 Bisphenol A All the effects stipulated in Regulation 1488/94 have been evaluated in this risk assessment (EU, 2003). The overall conclusion of the risk assessment of BPA was that there is mainly a concern for liver and reproductive toxicity.

370

Endocrine-disrupting chemicals in food

Table 14.4 The availability and scope of guidelines issued to assist the risk assessor in conducting the risk assessments Priority existing industrial substances

Environmental pollutants in food

Active substances in pharmaceuticals

Existing active substances in plant protection products (stage 1)

Guidelines available? Guidelines issued by

Yes

No

Yes

Yes

European Commission – DG JRC

–

Legislation implemented by guidelines Who is the guidance for?

Reg 793/93

–

European Medicines Agency (EMEA) Dir 2003/63

European Commission – DG Agriculture Dir 91/414

Risk assessor and manufacturer 311**

–

Risk assessor*

Risk assessor

–

113**

94

Criteria and expert judgement

Expert judgement

Expert judgement

Expert judgement

Number of pages Effects assessment mainly subject to criteria or expert judgement?

* The format of the report assessing the safety of a pharmaceutical substance is a safety assessment, ‘risk assessor’ might therefore not be a correct term for the expert conducting the report. ** Includes only guidelines for carrying out human health risk assessment.

Concern for eye and respiratory tract irritation was considered relevant for occupational exposure during the production of products containing BPA, such as polycarbonate plastics and epoxy resins. The assessment also indicates that developmental toxicity could be a critical effect but the risk assessors could not agree on the relevance of these findings to humans. It was concluded that further testing is needed before a robust conclusion can be drawn. Studies on endocrine-modulating activity were also evaluated in the risk assessment, and the risk assessors concluded that BPA can act weakly estrogenic by binding to nuclear estrogen receptors. The critical effect identified in the European risk assessment (EU, 2003) was reproductive toxicity with a NOAEL of 50 mg/kg/day. The NOAEL was based on a reduction in litter size observed in rats exposed to 500 mg/kg/day in a multi-generation study.

Assessing the risks of endocrine-disrupting chemicals

371

14.6.2 Dioxins The risk assessment of dioxins made by the Scientific Committee on Food (SCF) in 2000 (SCF, 2000) and updated in 2001 (SCF, 2001) was based on a previous assessment carried out by the World Health Organization (WHO-ECEH/IPCS, 2000). Since there is no legislation to regulate what effects should be evaluated for environmental pollutants, the focus of SCF was directed towards effects indicated as critical in the WHO assessment (WHO-ECEH/IPCS, 2000) and in the general scientific discussions (SCF, 2000, 2001). Therefore, the assessment was limited to carcinogenicity and toxicity to reproduction. In the SCF assessment, reproductive toxicity was extensively evaluated and it was concluded that developmental toxicity is of high concern. The SCF assessment also included an evaluation of data on endocrinemodulating activity, concluding that dioxins are Ah-receptor agonists interfering with the function of estrogen receptors, often resulting in antiestrogenic effects. The SCF identified perturbed development of the male reproductive tract as the critical effect of dioxin exposure. A NOAEL could however not be derived from the scientific material available to the risk assessors, therefore a lowest observed adverse effect level (LOAEL) of 25 ng/kg body weight (body burden) was used to derive a TDI of 2 pg TEQ/kg body weight.

14.6.3 Ethinyl estradiol There is no separate risk assessment of EE available from the Swedish Medical Products Agency (MPA). For medical products the results from non-clinical and clinical trials of the entire product as well as active constituents are reviewed and evaluated. Therefore the assessment of the oral contraceptive Yasmin® was used for this analysis. The report on Yasmin® contains only a very brief summary of the safety assessment of EE, and no critical effect for the substance was explicitly identified. The safety assessment states that EE is non-mutagenic but carcinogenic in rodents. The carcinogenic potential was however concluded to be due to a rodent-specific effect on endocrine regulation and was not deemed, by the assessor, to be relevant to humans. EE reduces fertility in rats and rabbits by inhibiting the fertilization of eggs and implantation of blastocysts, which is the desired pharmacological effect for oral contraceptives. In the toxicity studies summarized in the safety assessment embryo lethality was induced in rabbits and rats at high EE doses but there was no indication of teratogenicity. In humans, the use of high EE doses are limited to so-called ‘morning-after pills’, which are administered as a combination of EE and progestogen, that also have the purpose to prevent implantation. However, according to the EE safety assessment, interruption of pregnancy after implantation has occurred is not possible even at very high doses in humans.

372

Endocrine-disrupting chemicals in food

The developmental toxicity of EE was not evaluated at all in the safety assessment. It was concluded by the safety assessor that successful use in humans since the 1960s supports the use of EE as the active ingredient in oral contraceptives.

14.6.4 Vinclozolin The initial risk assessment of vinclozolin (EC, 1997) was amended in 1998 (EC, 1998a), mainly due to a re-evaluation of the reproductive toxicity of this compound. In reproductive toxicity studies adverse effects on development, in particular feminization of males, was shown. Relatively high doses were tested and maternal toxicity occurred in many cases. The anti-androgen properties of two of the main metabolites of vinclozolin, were established in endocrine-modulating tests. The evaluation of vinclozolin furthermore concluded that the substance is a non-genotoxic carcinogen in rodents. The critical effect of vinclozolin was established as non-neoplastic changes in liver and adrenals in male rats due to the anti-androgen effects of the substance with a NOAEL of 1.2 mg/kg/day.

14.6.5 Comparison Table 14.5 summarizes the results of the hazard assessments for the four compounds. It is noteworthy that endocrine-modulating activity has been evaluated in all four cases even though this is not required in the legislation for any of these compounds. It should also be noted that all the model compounds modulate steroid hormone signalling pathways and that this is compatible with the observed adverse effects of BPA, dioxins and vinclozolin as well as the therapeutic effects of EE. The above comparison also shows how the scope of the risk assessments differs between these groups of compounds. This mirrors the requirements stated in the EU Directives and Regulations for each chemical category. The evaluation of dioxins, for which there are no legislated requirements, was focused on the effects deemed the most sensitive in previous evaluations, such as the WHO evaluation from 1998, i.e. carcinogenicity and reproductive toxicity. In contrast the assessments of the other compounds have a wider scope, determined by their respective legislation, and aimed at hazard identification.

14.7

Toxicological assumptions and principles in effect assessment

In this section two important toxicological concepts, mode of action and dose–response relationships, are discussed in relation to the model compounds.

Feminization of male offspring and adverse effects on spermatogenesis –

Adverse effects on fertility, no consensus regarding developmental toxicity –

Estrogenic

Human carcinogen

Anti-estrogenic

Not genotoxic

–

Estrogenic

Carcinogenic in rodents Adverse effects on fertility, no developmental toxicity CNS effects at very high doses

Not genotoxic

Low

–

–

– Not covered by the reviewed risk assessment documents. * This is not a complete evaluation but draws upon previous evaluations where most effects have been assessed.

Endocrine-modulating activity

Neurotoxicity

Chronic toxicity/ carcinogenicity Toxicity to reproduction

Mutagenicity/genotoxicity

Repeated dose toxicity

Very low –

EE (Schering, 2000a,b)

– –

Dioxins (SCF, 2000, 2001)*

Low Irritating to eyes and respiratory tract Limited (more information needed) Multinuclear giant hepatocytes in mice observed Aneugenic in vitro, not in vivo No carcinogenic potential

BPA (EU, 2003)

Anti-androgenic

–

Carcinogenic in rodents Feminization of male offspring

Not genotoxic

Low

Very low Not irritating to skin or eye Yes

Vinclozolin (EC, 1997, 1998a)

Conclusions regarding the effects assessed for the four model compounds in each risk assessment document

Acute toxicity/single-dose Corrosiveness/irritation/ local tolerance Skin sensitization

Table 14.5

374

Endocrine-disrupting chemicals in food

14.7.1 Mode of action The relevance of the critical mode-of-action, i.e. the endocrine-modulating potential, to humans has been regarded differently in the risk assessments of the four compounds. In the BPA risk assessment (EU, 2003) the estrogenic activity was evaluated based on in vitro and in vivo data. A marked strain difference in rats was indicated by the data, but no explicit comment was made in this risk assessment report regarding the relevance of the different results to humans. Nevertheless, the health risk assessment of BPA was based on estrogenic effects seen in a study on rats (Tyl et al., 2002) and it was thus assumed by the risk assessors that this effect has human relevance. However, results in subsequent studies indicate that the rat strain used in the proposed pivotal study may be less sensitive to estrogenic effects than humans (Yamasaki et al., 2002). The conclusion in the dioxin risk assessment (SCF, 2000, 2001) was that humans are ‘less sensitive than responsive rodent strains’ to adverse effects caused by Ah-receptor activation and that rodents and humans show ‘comparable sensitivity’ to the induction of CYP 1A1 and 1A2. As a result the NOAEL was extrapolated from rats to humans without the use of uncertainty factors to account for any interspecies or interindividual variability in sensitivity. Regarding EE, the conclusion of the pharmacodynamics assessment (Schering, 2000a) was that the estrogenic effects seen in test animals are qualitatively similar to effects in humans. However, the tumorigenic potential of EE observed in mice and rats was concluded to be of no relevance to humans as it seemed to be ‘related to the endocrine activity of the compound which cannot be directly extrapolated to humans due to profound interspecies differences in sensitivity of target organs and endocrine regulation’. In the vinclozolin assessment (EC, 1997) similarities between this chemical and the anti-androgenic commercial drug flutamide used to treat prostate cancer were discussed. The conclusion was that since this drug is efficient in humans, the anti-androgenic mechanism of toxicity of the structurally similar vinclozolin should be regarded as relevant to humans. However, results from clinical trials on flutamide led the risk assessors to conclude that humans are less sensitive to perturbations in hormone balance than rodents and that the mechanism causing neoplastic changes in rats is species-specific and does not pose a hazard to humans.

14.7.2 Dose–response relationships For BPA and vinclozolin very high doses, compared to estimated consumer exposure, were tested in the studies used in the risk assessments. In contrast, in the dioxin assessment low doses, i.e. only about 10 times the estimated consumer exposure based on body burden comparisons, were used. EE was tested in therapeutically relevant doses. The reasoning behind the selection of doses used in toxicity testing was usually that effects at low doses can be predicted by effects seen at high doses (EU, 2003; EC, 1997). However, in

Assessing the risks of endocrine-disrupting chemicals

375

the case of EE, the reasoning was different. It was recognized that, for endocrine-modulating compounds the effects seen at doses higher than clinically recommended cannot be extrapolated to lower doses and contribute very little useful information regarding the risk assessment for humans (Schering, 2000a). It is generally assumed in toxicology that the dose–response curve is monotonic, i.e. above a certain concentration (a ‘threshold’) increasing dose leads to increasing response. This is one of the assumptions on which risk assessments are traditionally based. However, regarding EDCs, nonmonotonic, i.e. bell-shaped or (inverted) U-shaped, dose–response relationships are often discussed (Gray et al., 1997; vom Saal et al., 1997; Gupta, 2000; Sheehan, 2000; Rubin et al., 2001; Timms et al., 2005). In addition, it may be argued that since these substances mimic endogenous hormones no threshold can be assumed: since endogenous hormones already occur at concentrations sufficient to cause an effect the threshold is already exceeded (Barlow, 1999; Sheehan et al., 1999, Sheehan, 2000). The shape of the dose–response curve and potential thresholds for toxicity were not explicitly discussed in the risk assessment reports studied. However, both the shape of the curve and assumed thresholds can be inferred from the proposed NOAEL/LOAEL values and how these values have been used in the risk characterization. Simply, the setting of a NOAEL indicates that a threshold for effect was assumed. The application of assessment factors to derive health based guidance values, as in the case of dioxins and vinclozolin, indicates that a monotonic dose–response relationship was assumed. In addition, the calculation of a margin of safety (MOS), as for BPA, indicates that the dose–response curve was assumed to be monotonic. A linear extrapolation from high to low dose can only be made if the dose–response curve is assumed to be monotonic, i.e. effects are assumed to be qualitatively similar at high and low doses, and only change quantitatively in a proportional manner. 14.7.3 Comparison Table 14.6 summarizes some of the toxicological assumptions and principles used in the hazard assessments for the four model compounds. Significant differences are noted regarding the doses tested and the assumed shape of the dose–response curve. The traditional toxicological principle of extrapolating from high to low doses was made for BPA, dioxins and vinclozolin but for EE it was recognized that, for endocrine-modulating compounds effects at high doses may be qualitatively different from effects at low doses.

14.8

Development of testing and assessment methods for endocrine-disrupting chemicals

The Organization for Economic Cooperation and Development (OECD) has set up a task force on Endocrine Testing and Assessment (EDTA) to

376

Endocrine-disrupting chemicals in food

Table 14.6 A summary of the toxicological principles and assumptions on which effects assessment is based in the four risk assessment documents BPA (EU, 2003) Mode of action relevant to humans?

Not clearly stated

Dose-range evaluated in risk assessment

1000–10 000 times estimated consumer exposure High doses predict the effects at low doses There is a threshold for toxicity The dose– response relationship is monotonic

Dose-selection assumptions Assumptions regarding the shape of the dose–response curve

Dioxins (SCF, 2000, 2001)

EE (Schering, 2000a, b)

Vinclozolin (EC, 1997, 1998a)

Yes – stated that humans are as sensitive or less sensitive than responsive rodent strains About 10 times estimated consumer exposure

Yes – stated that humans are less sensitive than rodents to the tumorigenic potential of EE Therapeutic levels

Yes – stated that humans are less sensitive than test animals

Lowest doses tested are evaluated

Effects at high doses cannot predict effects at low doses No assumptions of a threshold are made Compounds with hormone-like properties do not have a monotonic dose–response curve

There is a threshold for toxicity The dose– response relationship is monotonic

1000–100 000 times estimated consumer exposure High doses predict the effects at low doses There is a threshold for toxicity The dose– response relationship is monotonic

facilitate and harmonize the development and completion of projects that are included in the OECD Test Guidelines Programme rolling work plan, and that relate to endocrine disrupters testing and assessment. Updated versions of some of the standard OECD Test Guidelines as well as other tests and assays are currently being considered and/or developed to include end points relevant to evaluating EDC toxicity. However, the process of validation is slow. In 2002 the EDTA Task Force agreed on a conceptual framework for the testing and assessment of potential endocrine disruptors. This framework can be considered as a ‘toolbox’ with tests organized into five compartments, or levels, corresponding to different levels of biological complexity. Level one entails sorting and prioritization based on existing information regarding physical and chemical properties, exposure and hazard. Level two includes in vitro assays to provide mechanistic data, e.g. receptor binding affinity and transcriptional activation. Level three consists

Assessing the risks of endocrine-disrupting chemicals

377

of in vivo assays to provide data about single endocrine mechanisms and effects, e.g. the uterotrophic and Hershberger assays. The Test Guideline on the uterotrophic bioassay has been adopted by the OECD Council on 16 October 2007 and is now available on the OECD public website as Test Guideline 440. The validation of the Hershberger bioassay by the OECD has been performed, a validation report is under development and the adoption of the related Test Guideline is expected in 2009. At level four more comprehensive in vivo assays are used to provide data about multiple endocrine mechanisms and effects. The revised 28-day toxicity study (OECD Test Guideline 407) is included at this level. The validation of parameters suitable to potentially detect endocrine activity of test substances has recently been completed and the revised TG 407 is expected to be adopted in 2008. At level five further in vivo tests are meant to provide data on effects from endocrine and other mechanisms. At this level the effects of perinatal exposure is also evaluated. A new Test Guideline on an Extended One-Generation Reproductive Toxicity Study is currently being developed. It would contain modules which cover, in particular, endocrine end points. The purpose of this conceptual framework is not to work as a complete testing scheme and chemicals can be entered and exited at all levels of the framework depending on the information needs for hazard and risk assessment purposes. Further tests and assays will be added to the framework as they are validated in the future.

14.9

Conclusions

There is currently no generally agreed procedure under any of the mentioned legislations for chemicals control that directly specify how substances with EDC characteristics are to be identified or risk assessed, what end points are crucial to investigate, or how the results of such investigations are to be interpreted. Technical guidelines have been introduced to promote conformity in the risk assessment process. However, different guidelines, issued by different authorities, apply to different regulatory groups of chemicals. Indeed, there is little conformity in the risk assessment processes between the groups of chemicals discussed in this chapter, even though the primary consumer exposure scenario (oral exposure) and the mode of toxicity (endocrine modulation) are similar. The purposes of using these chemicals, as well as the resulting exposure scenarios, differ. Plant protection products are deliberately sprayed on fruits and vegetables in large quantities to control pest organisms and increase crop yield. Pharmaceuticals, on the other hand, are deliberately administered to patients with the purpose of improving the well-being of that patient. Both pesticides and pharmaceuticals are thus produced with the purpose of interacting in a specific manner with biological organisms; pesticides in order to be highly toxic to particular organisms, and

378

Endocrine-disrupting chemicals in food

pharmaceuticals to exert a specific pharmacological effect in humans. This intentional use contrasts with the unintentional and widespread exposure of consumers to low concentrations of industrial substances and environmental pollutants that inadvertently end up in food. In our view, the abovementioned differences in uses and exposures motivate some of the identified differences in legislation such as the degree of flexibility in testing and risk assessment. However, the identified differences in the interpretation and evaluation of data seem not to be well motivated by policy or by science. Generally, for the four model compounds, there are no requirements for testing end points that specifically enables discovery of endocrine disruption. Standardized screening tests with international regulatory acceptance are in fact not available at present. It is acknowledged that EDCs can affect humans and animals at low exposure levels and that responses to EDCs are in many cases complex, activating a range of different molecular events, e.g. receptor-agonism/antagonism and enzyme induction, in multiple hormone systems. As a result, regulatory testing for these effects, and evaluating the results is complicated and test methods, assumptions and criteria for data interpretation commonly used for general toxicity and carcinogenicity, might not be directly applicable. Work to develop and validate test methods to screen for and evaluate EDCs is currently being conducted by the OECD-EDTA. In our view further development of risk assessment guidance, i.e. how to interpret data from experiments using doses significantly higher than the expected human exposures, which assumptions regarding the shape of the dose–response curve that can be made, and principles for extrapolating data from experimental species to humans, should also be aimed at. This is supported by an international agreement that there is reason to move in this direction, based on a general acceptance that EDCs present a risk to humans and the environment alike. It is our opinion that the success of the work towards improving the legislation for EDCs requires continued close cooperation between experimental scientists, risk assessors, regulators and industry.

14.10 References barlow sm. (1999) Dilemmas facing regulatory and advisory bodies dealing with conflicting results. Chemosphere 39:1287–1292. commission directive 2003/63/EC of 25 June 2003 amending Directive 2001/83/EC of the European Parliament and of the Council on the Community code relating to medicinal products for human use. Offi J Communities L159, 27/06/2003, 0046–0094. commission directive 93/71/EEC of 27 July 1993 amending Council Directive 91/414/EEC concerning the placing of plant protection products on the market. Offi J Communities L221, 31/08/1993, 0027–0036. commission directive 94/79/EC of 21 December 1994 amending Council Directive 91/414/EEC concerning the placing of plant protection products on the market. Offi J Communities L354, 31/12/1994, 0016–0031.

Assessing the risks of endocrine-disrupting chemicals

379

commission regulation (EC) No 1488/94 of 28 June 1994 laying down the principles for the assessment of risks to man and the environment of existing substances in accordance with Council Regulation (EEC) No 793/93. Offi J Communities L161, 29/06/1994, 0003–0011. commission regulation (EEC) No 3600/92 of 11 December 1992 laying down the detailed rules for the implementation of the first stage of the programme of work referred to in Article 8 (2) of Council Directive 91/414/EEC concerning the placing of plant protection products on the market. Offi J Communities L366, 15/12/1992, 0010–0016. council directive 67/548/EEC of 27 June 1967 on the approximation of laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances. Offi J Communities 196, 16/08/1967, 0001–0098. council directive 91/414/EEC of 15 July 1991 concerning the placing of plant protection products on the market. Offi J Communities L230, 19/08/1991, 0001–0032. council directive 97/57/EC of 22 September 1997 establishing Annex VI to Directive 91/414/EEC concerning the placing of plant protection products on the market. Offi J Communities L265, 27/09/1997, 0087–0109. council regulation (EEC) No 793/93 of 23 March 1993 on the evaluation and control of the risks of existing substances. Offi J Communities L084, 05/04/1993, 0001–0075. Directive 2001/83/EC of the European Parliament and of the Council of 6 November 2001 on the Community code relating to medicinal products for human use. Offi J Communities L311, 28/11/2001, 0067–0128. Directive 2004/27/EC of the European Parliament and of the Council of 31 March 2004 amending Directive 2001/83/EC on the Community code relating to medicinal products for human use. Offi J Communities L136, 30/04/2004, 0034–0057. ec (1997) European Commission Peer Review Programme, ECCO meetings, Vinclozolin, Volume 3 Annex B. ec (1998a) European Commission Peer Review Programme, ECCO meetings, Vinclozolin, Addendum to EC, 1997b – toxicology and metabolism. ec (1998b) Guidelines and criteria for the evaluation of dossiers and for the preparation of reports to the European Commission by Rapporteur Member States relating to the proposed inclusion of active substances in Annex I of Directive 91/414/EEC. Available on-line at http://europa.eu.int/comm/food/plant/ protection/evaluation/ ec (1999) Communication from the Commission to the Council and the European Parliament No. 706. Community strategy for endocrine disrupters. ec (2003) Technical guidance document on risk assessment, Part I. Available on-line at http://ecb.jrc.it/existing-chemicals/ emea (2003) CPMP/ICH/2887/99 Common technical document for the registration of pharmaceuticals for human use. Available on-line at http://www.emea.eu.int emea (2005) CPMP/SWP/799/95 Guideline on the non-clinical documentation for mixed marketing authorisation applications. Available on-line at http://www. emea.eu.int epc (2005) Reproduktiv hälsa i ett folkhälsoperspektiv. Report no 2005-112-5. Swedish National Board of Health and Safety, Centre for Epidemiology. Available on-line at http://www.socialstyrelsen.se/ eu (2003) Bisphenol-A: European Union Risk Assessment Report (CAS no. 80-057). Available on-line at: http://ecb.jrc.it/ gray, jr le, ostby js and kelce wr. (1997) A dose–response analysis of the reproductive effects of a single gestational dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male Long Evans hooded rat offspring. Toxicol Appl Pharmacol 146:11–20.

380

Endocrine-disrupting chemicals in food

gupta c. (2000) Reproductive malformation of the male offspring following maternal exposure to estrogenic chemicals. Proc Soc Exp Biol Med 224:61–68. kelce wr, monosson e, gamcsik mp, laews sc and gray, jr le. (1994) Environmental hormone disruptors: evidence that vinclozolin developmental toxicity is mediated by androgenic metabolites. Toxicol Appl Pharmacol 126:276–285. Regulation (EC) No 178/2002 of the European Parliament and the Council of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety. Offi J Communities L31, 1/2/2002, 1–24. Regulation (EC) No 726/2004 of the European Parliament and of the Council of 31 March 2004 laying down Community procedures for the authorisation and supervision of medicinal products for human and veterinary use and establishing a European Medicines Agency. Offi J Communities L136, 30/04/2004, 0001–0033. rubin bs, murray mk, damassa da, king jc and soto am. (2001) Perinatal exposure to low doses of Bisphenol A affects body weight, patterns of estrous cyclicity, and plasma LH levels. Environ Health Perspect 109:675–680. scf (2000) Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food. Available on-line at http://europa.eu. int/comm/food/fs/sc/scf/out78_en.pdf scf (2001) Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food. Up-date based on new scientific information. Available on-line at http://europa.eu.int/comm/food/fs/sc/scf/out90_en.pdf sheehan dm. (2000) Activity of environmentally low doses of endocrine disruptors and the bisphenol-A controversy: initial results confirmed. Proc Soc Exp Biol Med 224:57–60. sheehan dm, willingham e, gaylor d, bergeron jm and crews d. (1999) No threshold for estradiol-induced sex reversal of turtle embryos: how little is too much? Environ Health Perspect 107:155–159. schering. (2000a) Pharmacological-toxicological expert report to the application for drug marketing approval of Yasmin®. Provided by the Swedish MPA, February 2006. schering. (2000b) Expert report on the clinical documentation for Yasmin®. Provided by the Swedish MPA, February 2006. timms bg, howdeshell kl, barton l, bradley s, richter ca and vom saal fs. (2005) Estrogenic chemicals in plastic and oral contraceptives disrupt development of the fetal mouse prostate and urethra. Proc Natl Acad Sci 102:7014–7019. tyl rw, myers cb, marr mc, thomas bf, keimowitz ar, brine dr, veselica mm, fail pa, chang ty, seely jc, joiner rl, butala jh, dimond ss, cagen sz, shiotsuka rn, stropp gd and waechter jm. (2002) Three-generation reproductive toxicity study of dietary bisphenol A in CD Sprague-Dawley rats. Toxicol Sci 68:121–146. vom saal fs, timms bg, montano mm, palanza p, thayer ka, nagel sc, dhar md, ganjam vk, parmigiani s and welshons wv (1997) Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses. Proc Natl Acad Sci 94:2056–2061. who-eceh/ipcs (2000) Assessement of the health risk of dioxins: re-evaluation of the tolerable daily intake (TDI). Food Addit Contam 17:223–369. yamasaki k, sawaki m, shuji n, imatanaka n and takatasuku m (2002) Subacute oral toxicity study of ethynylestradiol and bisphenol A, based on the draft protocol for the ‘Enhanced OECD test guideline no. 407’. Arch Toxicol 76:65–74.

15 Dioxins, polychlorinated biphenyls and brominated flame retardants L. A. P. Hoogenboom, RIKILT-Institute of Food Safety, The Netherlands

Abstract: During the past decades various incidents with dioxins and dioxin-like PCBs have drawn the attention of the general public. In addition to occupational exposure, some of these compounds accumulate in the food chain and cause a number of adverse effects in animals and humans. This includes effects on the sexual development and the thyroid, allowing their classification as endocrine disruptors. Although it is evident that the arylhydrocarbon (Ah) receptor plays a crucial role in the toxic effects, the exact mechanism is still unclear. Based on the effects on sperm counts in the male offspring of treated dams, an exposure limit has been set corresponding to about 2 pg TEQ/kg body weight/day. At present the margin between the actual exposure and the exposure limit is small and even non-existent for part of the population. As a result the EU has set strict limits for feed and food in order to further reduce the exposure. Current levels are declining. However, at the same time a number of new compounds have been shown to be persistent like the brominated flame retardants. These compounds too have been shown to affect the thyroid of animals. At present no exposure limit has been established but margins between exposure of human consumers and effect levels in animals appear to be large, although these calculations do not take into account potential differences in kinetics. In addition the production and incineration of polybromodiphenylethers (PBDEs) may result in the formation of bromodioxins with similar adverse effects as their chlorinated counterparts. Key words: dioxins, PCBs, flame retardants, endocrine disruption, feed and food incidents, exposure.

15.1

Introduction

Environmental contaminants may affect human health, in particularly when present in food. At present dioxins and other persistent organic pollutants make up a group of compounds that causes great concern. This is based not

384

Endocrine-disrupting chemicals in food

only on the persistence and accumulation of these compounds in our food chain, but also on the small to non-existent margin between the actual exposure of humans to these compounds and the toxicological guidelines on safe human intakes. Effects caused by dioxins and related compounds in animals and humans show that they should be regarded as endocrine disruptors. The arylhydrocarbon (Ah) receptor, ubiquitously present in mammalian cells but with no obvious physiological role, plays a major role in the effects of dioxins. However, the exact mechanisms behind the effects remain to be elucidated. Owing to various measures, levels of polychlorobiphenyls (PCBs) and dioxins in food and humans have started to decline during the past decades. However, more recently it was shown that levels of other persistent compounds, in particular brominated flame retardants have started to increase in food and human samples, causing concern about possible future effects. Although the environment may contribute to the exposure in particular cases, the major part comes from food. Products of animal origin, including fish, contribute most to the intake. This chapter presents the current situation on the toxicity of these compounds, the tools to detect their presence in the food chain and current exposure levels.

15.2

Dioxins and dioxin-like polychlorobiphenyls

15.2.1 Incidents with dioxins and polychlorobiphenyls ‘Dioxins’ is a generic name for two groups of compounds, the polychlorinated dibenzofurans (PCDFs) and the polychlorinated dibenzo-p-dioxins (PCDDs) (Fig. 15.1). These compounds are characterized by a planar structure and a varying number of chlorine atoms. In practice the 2,3,7,8substituted congeners are more or less metabolically resistant and as a result accumulate in biological systems. In addition to dioxins, it has been shown that 12 out of the 209 PCB-congeners may also have a planar structure and resemble dioxins in their stability and biological effects. In particular PCB 126, a non-ortho-PCB, contributes significantly to the overall exposure to dioxins and dioxin-like PCBs. Dioxins were first identified as a possible threat to the food chain during the early 1960s. Large numbers of chickens died from a disease called chicken edema disease after consuming contaminated feed (Sanger et al., 1958; Schmittle et al., 1958; Higginbotham et al., 1968; Firestone, 1973). The feed contained fat prepared from cowhides that had been treated with chlorinated phenols. It took about ten years to identify dioxins as the causative factor in chicken edema disease, which can partly be explained by the fact that these compounds were new to researchers on environmental contaminants, but also by the very great toxic potency and hence the low levels required for toxic effects. Other major incidents such as the Yu-Sho and Yu-Cheng rice oil contaminations in Japan and Taiwan in 1968 and

Dioxins, polychlorinated biphenyls and brominated flame retardants CIy

9

PCDD

7 CIy

2 3

O

6

CIx

1

O

8

4

9

CIx

1

8 PCDF

2

7 CIy 3

3

O

6

PCB

385

2

4 2′

4

3′ CI

x

4′ 5

6

6′

5′

Fig. 15.1 Structures of polychlorinated dibenzo-p-dioxins (PCDD) dibenzofurans (PCDF) and polychlorobiphenyls (PCB). The most toxic congener is 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD). The other toxic six PCDD and ten PCDF congeners contain at least four chlorine atoms and always at the 2, 3, 7 and 8 positions. The 12 toxic PCBs contain four or more chlorine atoms and none (nonortho) or only one (mono-ortho) of them at the 2, 2′, 6, or 6′ position.

1979 respectively, and the Seveso incident in Italy in 1976 confirmed the considerable toxicity of these compounds at relatively low levels. In these incidents, chloracne of the skin were the first overt signs of the exposure of the victims. The two rice oil incidents were caused by leakage of PCB oil, used as heat transfer fluid, into the oil used for human consumption (Kuratsune et al., 1972; Hsu et al., 1985). Only later did it become apparent that the PCB oil contained dioxin-like compounds, like the dioxin-like PCBs but also PCDFs, which might be responsible for most of the observed effects. In the late 1980s another important source for dioxins was detected: the incineration of municipal waste (Olie et al., 1977). In several countries including the Netherlands it was shown that this resulted in elevated levels of dioxins in soil and grass and as a result in milk from dairy cows (Liem et al., 1991). At least in the Netherlands, large investments were made to reduce the emission of dioxins from municipal waste incinerators (MWIs) with the result that dioxin levels in milk decreased considerably over time. More recent incidents with PCBs and dioxins have shown that these compounds still pose a major threat to the food chain. In the US, dioxins were detected in ball clay used in feed for chickens and cat fish (Hayward et al., 1999; Ferrario et al., 2000). In Europe, citrus pulp used for animal feed was shown to be contaminated with PCDDs, resulting in moderately increased levels in milk (Malisch, 2000). The source was the mixing of the pulp with lime from an industrial plant, which turned out to be contaminated with high levels of dioxins. A much larger incident in terms of impact, was the chicken crisis in Belgium involving an estimated 200–300 kg PCB oil that

386

Endocrine-disrupting chemicals in food

contaminated 60 tonnes of fat used for the production of animal feed (Bernard et al., 1999; Hoogenboom et al., 1999; Larebeke et al., 2001). Again chickens were affected most, showing decreased egg hatching and symptoms resembling chicken edema disease. The incident had strong political consequences in Belgium and, at least in Europe, resulted in strict regulations on dioxins (EC, 2001, 2002). As a result of increased monitoring several novel sources were elucidated like kaolinic clay, cholin chloride (Llerena et al., 2003), sequestered minerals, dried grass, dried bakery meal (Hoogenboom et al., 2004), potato peels containing contaminated kaolinic clay (Hoogenboom et al., 2005) and waste fat resulting from the production of gelatin with contaminated hydrochloric acid (Hoogenboom et al., 2007). In all these cases the incidents started in products used as animal feed and were traced back to recycling of fat or other ingredients.

15.2.2 Sources and production Sources of dioxins Dioxins are produced as by-products in the synthesis of certain chemicals, like the wood-preserving agent pentachlorophenol and the herbicide 2,4,5trichlorophenoxyacetic acid (2,4,5-T; Agent Orange). However, they may also be present in heated PCB oil and formed during incineration of plastics and other waste. Each source may produce its own specific congener pattern. Figure 15.2 shows the congener patterns of dioxins in the most recent food and feed incidents, starting with the Brazilian citrus pulp incident in 1998.

Relative contribution to TEQ (%)

90 80 70 60 50

PCB feed 1999 Citrus pulp 1998 Kaolinite 1999 Cholin chloride 2000 Dried bakery waste 2004 Gelatin fat 2006

40 30 20 10

2, 3

,7

2, ,8-T 1, 3,7 CD 2, ,8 3, -T F 2, 7,8 CD 3, 4, Pe D 1, 7,8 CD 2, F 3 P 1, ,7, eCD 82, P F 3 1, ,4,7 eC D 2, 3, ,8-H D 2, 6,7 xC 3, , 4, 8-H DF 1, 6,7 xC 2, , 3 8-H DF 1, ,7,8 xC 2, 3, ,9-H DF 1, 4,7 xC 2, , 3, 8-H DF 1, 6,7 xC , 2, D 3, 8-H D 7, 1, xC 8 2, ,9 D 3, D 1, 4,6 HxC 2, ,7 D 3 ,8 D 1, ,4,7 -H p 2, , 3, 8,9 CD 4, 6, -Hp F 7, 8- CD H F pC D D O C D O F C D D

0

Fig. 15.2 Dioxin congener patterns of a number of contaminations of the food chain, expressed as the relative contribution to the TEQ level.

Dioxins, polychlorinated biphenyls and brominated flame retardants

387

The patterns are presented as the relative contributions of each congener to the toxic equivalent (TEQ) level. The patterns are very typical for the source and can actually be used to identify a source. The use of lime originating from a poly(vinylchloride) (PVC) production plant for neutralization and water removal of citrus pulp led to another typical pattern (Malisch, 2000). In the case of the PCB contamination of Belgian feed, oxidation resulted primarily in the formation of PCDFs. In the case of the choline chloride for example, the contamination was traced back to the mixing with pentachlorophenol (PCP)-treated wood, showing the typical pattern of higher chlorinated dioxins (Llerena et al., 2003). In 2003, the use of waste wood for drying of bakery waste used in animal feed resulted in another incident. The contaminated bakery waste showed a pattern resembling that of burned painted wood (Wunderli et al., 2000; Hoogenboom et al., 2004). At the end of 2004, the highest dioxin levels ever measured in milk from two farms in the Netherlands was traced back to the use of kaolinic clay for selection of potatoes based on density. The clay ended up in the peels that were fed to cows on two neighboring farms. Early 2006, fat used for animal feed was shown to contain levels up to 400 ng TEQ/kg with a very typical pattern that could thus far not be traced back to the source. Contaminated hydrochloric acid was used for the production of gelatin from pig bones and the fat from the bones was used for feed production (Hoogenboom et al., 2007). However, the actual source for the contamination of the hydrochloric acid has not been disclosed. PCBs PCBs are a group of 209 different congeners, which have been produced as technical mixtures with names such as Aroclor and Kaneclor. They have been used in large amounts as heat transfer fluids, hydraulic lubricants and dielectric fluids for capacitors and transformers (Safe, 1994). Although these uses are now generally prohibited, the main potential problems are in the disposal of old electrical equipment and the persistence of PCB residues in the environment. Twelve of the PCB congeners have a planar structure and similar properties to dioxins. Other PCBs have been shown to affect brain development (Schmidt, 1999) and appear to be responsible for the tumor promotion effects of these mixtures (Plas et al., 2001). In addition, metabolites of non-dioxin-like PCBs interfere with the homeostasis of vitamin A and thyroid hormones (Safe, 1994). In practice, the consumer will be exposed to a mixture of dioxins, dioxin-like PCBs and other PCBs, and the toxicology of these mixtures is very complex. EFSA reviewed the toxicity of nondioxin-like PCBs and concluded that the most sensitive effects, the tumor promoting effects, are likely to be caused by the dioxin-like compounds in the tested mixtures (EFSA, 2005). As a result, no exposure limits for nondioxin-like PCBs have been established thus far and no harmonized food limits in, for example, the EU. However, there are limits for PCBs in food and feed in various countries. Furthermore, high levels of PCBs point to the

388

Endocrine-disrupting chemicals in food

presence of dioxin-like PCBs and PCDFs, although this relationship may vary due to the different PCB-mixtures used in the past (EFSA, 2005). In order to avoid difficulties in analyzing all 209 congeners, it is customary to analyze only six or seven so-called indicator PCBs that represent the different technical mixtures. These include PCBs 28, 53, 101, 138, 153 and 180 (sum of 6) and sometimes PCB 118 (sum of 7), which is a dioxin-like mono-ortho PCB. It is important to realize that the total amount of PCBs in a technical mixture may be three to four times higher than the amounts of these indicator substances. This relative contribution may change in animal-derived products due to selective absorption, metabolism, carryover and accumulation of the different congeners. In particular the higher chlorinated PCBs, like 138, 153 and 180 tend to accumulate.

15.3

Assessing the toxic effects of dioxins and dioxin-like polychlorobiphenyls

15.3.1 Effects in humans and animals Dioxins and dioxin-like PCBs have clear endocrine-disrupting properties. They bind to the so-called Ah receptor present in mammalian cells, thus resulting in the transcription of a large number of genes. These genes include those encoding for biotransformation enzymes like cytochromes P450 1A1, 1A2 and 1B1, UDP-glucuronyl-transferase and glutathione-S-transferase 2α, as well as genes involved in the growth regulation of cells (Guo et al., 2004). In laboratory animals, exposure to 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD) results in liver tumors in female rats, and at lower levels of exposure in effects on the immune and reproductive systems, as well as impaired learning (WHO, 2000). Another typical effect is endometriosis, a symptom where cells from the endometrium grow in sites outside the uterine cavity. This effect has first been observed in rhesus monkeys exposed for 4 years to doses of 5 and 25 ng TCDD/kg in the diet (Rier et al., 1993). The effect only became apparent after an additional 10 years on clean feed. Effects observed in humans, exposed accidentally to these compounds, are chloracne, neurodevelopmental delays and neurobehavioural effects, and an increased risk of diabetes and certain cancer types such as soft tissue sarcomas and liver cancer (WHO, 2000). From studies with knock-out animals, it is clear that the Ah-receptor plays a major role in the adverse effects of dioxins. The exact mechanism, however, remains to be elucidated. The increased activity of several biotransformation enzymes appears to play an intermediate role, since the metabolism of certain hormones may be increased and thus their elimination from the body. This results, for example, in decreased thyroid hormone levels (Giacomini et al., 2006). The effect is further exaggerated by the competition between dioxins and PCBs, and the thyroid hormones T3 and T4 for the carrier protein transthyretin (TTR). As a result the levels of

Dioxins, polychlorinated biphenyls and brominated flame retardants

389

bound hormones decrease and the degradation of the hormone is further increased. Feed-back regulation results in elevated levels of thyroidstimulating hormone (TSH) and subsequently effects on the thyroid gland, including tumors (Knerr and Schrenk, 2006). TTR is also involved in the transport of retinoids and TCDD has been shown to interfere also with the homeostasis of these compounds (Murphy et al., 2007). Brouillette and Quirion (2008) showed, for example, that TCDD causes a deficit in the memory function of female mice, which could be counteracted by the addition of retinoic acid, but also estradiol. Other studies showed a direct effect on the sex organs. In sex glands of male and female rats, TCDD causes decreased levels of 17 alpha-hydroxylase/17,20-lyase cytochrome P450 expression, most likely due to an interference with the effect of hormones like hCG and LH (Moran et al., 2003; Fukuzawa et al., 2004). The latter study actually showed decreased testosterone levels in the testis of 12-weekold mice, an effect that was not observed in Ah receptor knock-outs. As a result, the production of androgens and estrogens are decreased, resulting in lower plasma levels. However, at lower TCDD doses Haavisto et al. (2001) observed an increased testosterone production in young male rats exposed in utero, indicating that timing and dose are very important determinants in the actual effects. Decreased levels of thyroid hormones may also play a role in the effect on testosterone production by the testis, since hypothyroidism is known to affect the Leydig cells in the testis. Exposure of rats and mice to TCDD results in tumors in liver, thyroid, lung, skin, oral cavity and other sites (Knerr and Schrenk, 2006). Observations that liver tumors primarily occur in female rats have suggested a role of estrogens in this effect, possible through the increased formation of reactive catecholestrogens (Knerr and Schrenk, 2006). In particular cytochrome P450 1B1 could play an important role in this process. The interaction between estrogens and dioxins is complex and appears to occur also at various other levels. This interaction requires further studies. Another important issue is the possibly long-lasting effects of perinatal exposure to dioxins. Fenton et al. (2002) observed that in utero exposure of rats to TCDD resulted in a higher incidence of dimethylbenzanthracene (DMBA) induced mammary tumors in adult rats, suggesting an imprinting effect of TCDD. Jenkins et al. (2007) observed persistent changes in mammary gland development of rats following in utero exposure to TCDD, which points in the same direction. Nayyar et al. (2007) observed in mice exposed in utero a similar uterine phenotype at adult age as observed in women with endometriosis. Another important target for dioxins is the immune system, since TCDD has been shown to decrease the immune response. Again the exact mechanism remains to be resolved. The relevance of the effects observed in animals for humans, especially those exposed to low levels of dioxins and PCBs, remains to be shown. TTR, for example, is not as important in humans as in rodents, since thyroid binding globulin (TBG) is the more

390

Endocrine-disrupting chemicals in food

important carrier protein for thyroid hormones. TCDD has only a weak affinity for TBG. However, the vast amount of data clearly points to the possibility that exposure to dioxins results in a number of subtle effects related to disturbance of hormonal processes. Based on the effects in animals, the World Health Organization (WHO) and the EU Scientific Committee on Food (SCF) have set very low exposure limits, being respectively a tolerable daily intake (TDI) of 2 pg TEQ/kg body weight (bw)/day (WHO, 2000) and provisional tolerable weekly intake (pTWI) of 14 pg TEQ/kg bw/week (SCF, 2001). The SCF based the limit on a reduced sperm count of male rats exposed to TCDD in utero (Faqi et al., 1998). A similar effect was observed in several previous studies but could not be reproduced in more recent studies (Bell et al., 2007a, b). However, in the latter study, chronic dosing of the mother resulted in a delayed puberty in the male pups, even in the lowest dose group. Overall, it is clear that the levels in the body, required for adverse effects, are in a similar range. In deriving the exposure limit, the WHO and SCF took into account the levels in the various tissues (internal dose) but also kinetic differences between rats and humans. The expression of limits on a weekly basis acknowledges the fact that the toxic effects are normally not a result of a single exposure, but the systematic increase of body burdens due to continuous intake of these compounds and their storage in body fat. The toxicity of dioxins and PCBs is thought to be jointly due to their effects at low levels, their lipophilic nature and their resistance to metabolic degradation, resulting in accumulation in the body. Eventually this may result in body burdens that are higher than safe levels, thus resulting in the adverse effects described above. The exposure limit should prevent that body burden levels eventually reach levels that may cause adverse effects. One of the few mechanisms leading to a significant reduction in body burden levels is the excretion of these compounds in breast milk. Similarly in food-producing animals, dioxins can be transferred into milk and eggs.

15.3.2 The toxic equivalents principle In practice the exposure limits include the 17 toxic 2,3,7,8-chlorinated PCDDs and PCDFs as well as 12 planar PCBs. This is also true for the food and feed limits set by the EU. The dioxin-like PCBs have been included since November 2006 (EC, 2006a, b). Levels of dioxins and dioxin-like PCBs are normally expressed in TEQs, which refers to the use of the so-called TEQ principle. Most of the toxicological data have been obtained from studies with TCDD. Other dioxins and dioxin-like PCBs are not equally toxic, partly owing to different kinetics. At the same time there is strong evidence that the effects of the different congeners are additive. In order to deal with this particular problem, the so-called TEQ approach has been introduced. The different congeners have been assigned a so-called toxic equivalency factor

Dioxins, polychlorinated biphenyls and brominated flame retardants

391

(TEF) value, which is a weighted value, based on the differential effects in various experiments. TEF values are consensus factors agreed upon by international scientists involved in the field. The last time these TEF values were evaluated was during a WHO workshop in 2005. These values are shown in Table 15.1, together with the previous set published in 1998 (Van den Berg et al., 1998, 2006). In principle, a compound should have the following properties to obtain a TEF value: • A compound must show a structural relationship to the PCDDs and PCDFs.

Table 15.1 TEF values as assigned by the WHO in 1998 and 2006 Compound

TEFs 1998

TEFs 2006

PCDDs 2,3,7,8-TCDD 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDD OCDD

1 1 0.1 0.1 0.1 0.01 0.000 1

1 1 0.1 0.1 0.1 0.01 0.000 3

PCDFs 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDF

0.1 0.05 0.5 0.1 0.1 0.1 0.1 0.01 0.01 0.000 1

0.1 0.03 0.3 0.1 0.1 0.1 0.1 0.01 0.01 0.000 3

non-ortho PCBs 3,4-3′4′ TeCB (77) 3,4,5-4′ TeCB (81) 3,4-3′4′5′ HxCB (126) 3,4,5-3′4′5′ HxCB (169)

0.000 1 0.000 1 0.1 0.01

0.000 1 0.000 3 0.1 0.03

mono-ortho PCBs 2,3,4-3′,4′ PeCB (105) 2,3,4,5-4′ PeCB (114) 2,4,5-3′,4′ PeCB (118) 3,4,5-2′,4′ PeCB (123) 2,3,4,5-3′,4′ HxCB (156) 2,3,4-3′,4′,5′ HxCB (157) 2,4,5-3′,4′,5′ HxCB (167) 2,3,4,5-3′,4′,5′ HpCB (189)

0.000 1 0.000 5 0.000 1 0.000 1 0.000 5 0.000 5 0.000 1 0.0000 1

0.000 03 0.000 03 0.000 03 0.000 03 0.000 03 0.000 03 0.000 03 0.000 03

392

Endocrine-disrupting chemicals in food

• A compound must bind to the Ah receptor. • A compound must elicit Ah receptor-mediated biochemical and toxic responses. • A compound must be persistent and accumulate in the food chain. Many other compounds have been shown to have affinity for the Ah receptor, including polyaromatic hydrocarbons and a number of plant secondary metabolites. However, these compounds failed to pass the other criteria primarily due to their rapid degradation in the body. Other compounds like the polybrominated dioxins and biphenyls might be considered for a TEF value, but adequate studies are missing for the assignment of a TEF value. An important issue is whether the effects of different dioxins and dioxinlike PCBs, but also other Ah receptor agonists are actually additive or whether there may also be antagonistic activity. Furthermore, Safe (1995, 1998) was one of the first to point out the often very high levels of Ah receptor agonists in certain foods of plant origin, which may be consumed on a daily basis. However, there is a substantial difference between these compounds, and dioxins and PCBs, since these plant-derived compounds are often metabolized completely before entering the bloodstream. Certain flavonoids are rapidly conjugated in the gut wall, thus losing their biological activity. Further studies are required to reveal which naturally occurring compounds may actually retain their activity in the body. In addition, it is thought that the toxicity of dioxins is due not to the occasional intake of a relatively high level, but the increase in body burden above a certain threshold, thus resulting in a continuous stimulation of the Ah receptor pathway. It remains to be elucidated whether natural Ah receptor agonist may reach effective concentrations and cause the related effects. It was agreed to evaluate the TEF values every five years, depending on new data from toxicological studies. As a result the TEQ levels in food may change considerably if the TEF value of a congener that contributes significantly is changed. The last revision of the TEFs in practice results in a decrease of the levels of about 15% (Van den Berg et al., 2006), but food items containing high levels of mono-ortho PCBs may show an even larger decrease. Since the EU based its food and feed limits on the background levels determined with the 1998 TEFs, these old TEFs will be used in the EU for determining the TEQ levels until the next revision of the food and feed limits (EC, 2006a, b).

15.4

Analytical methods for dioxins and polychlorobiphenyls

Because of the low limits for food and feed, dioxins can at present only be analyzed by high-resolution, gas chromatography–mass spectroscopy (GC/

Dioxins, polychlorinated biphenyls and brominated flame retardants

393

MS) (HRGC/HRMS). In addition, an extensive clean-up is required starting with the extraction of fat and its subsequent removal by gel permeation chromatography (GPC) or acid silica. Pesticides and other lipophilic substances need to be removed by subsequent clean-up steps. Eventually dioxins and PCBs are separated on an activated carbon column. In order to account for possible recovery losses in the different column and evaporation steps, 13C-labeled standards are added at the first step. Individual dioxins are identified by retention time and mass, and quantified. Using the TEF values, the levels are transferred into TEQs and summed into figures for dioxins, dioxin-like PCBs and the total sum. The levels are normally reported as lower bound and/or upper bound levels, thereby dealing with the levels of non-detected congeners. When reported as upper bound levels, as required by current legislation in the EU, the levels of non-detected congeners are set equal to the detection limits. The intensive clean-up and associated equipment is very costly. Furthermore, sample throughput is relatively small, causing further problems during situations as during the Belgian dioxin crisis in 1999. The whole clean-up procedure has now been automated, thus eliminating the removal of the solvents between each step. Nevertheless, there is a major need for cheap and high-throughput screening methods that can be used to select potentially contaminated samples and, equally important, to clear negative samples. Initial assays were based on the induction of cytochrome P450 enzymes in hepatoma cells or freshly isolated hepatocytes. These cells contain the different steps of the Ah receptor pathway. Elevated levels of these enzymes are detected by the increased metabolism of ethoxyresorufin into resorufin (EROD activity). However, the assay is potentially sensitive for false-negative results, since the enzyme has a number of inhibitors. Therefore novel assays, such as CALUX, were developed based on cells with reporter genes such as that encoding for the firefly enzyme luciferase. Following exposure to dioxin-like compounds, the cells start to synthesize luciferase which can easily be detected by a light-producing assay. Compared with HRGC/HRMS, the assay requires only a very simple clean-up, usually based on acid silica. However, the assay does not allow the use of internal standards. As a result the clean-up must be simple and highly reproducible, and proper control samples must be included in each test series to control for recovery and possible contaminants in the chemicals used for clean-up. Under these conditions the assay has been proven its usefulness for controlling dioxins (Hoogenboom, 2002; Hoogenboom et al., 2004, 2007). The EU has set guidelines for the performance of analytical methods within a laboratory, including screening assays. HRGC/HRMS was established as the reference method. The analysis of the so-called indicator PCBs and, in fact, also the monoortho PCBs does not require the same sensitivity due to the much higher levels. In practice, GC with either electron capture detection (ECD) or MS/MS detection is used. As a result this analysis is much cheaper and more

394

Endocrine-disrupting chemicals in food

analytical capacity is in general available. In specific cases, like during the Belgian dioxin/PCB crisis in 1999 and possibly also in the case of, for example, eels from polluted rivers, indicator PCB analysis might actually be used as a screening tool for samples (Hoogenboom et al., 2006). The major problem is that this might result in a more general application and that dioxins from other non-PCB-like sources are overlooked.

15.5

Current exposure to dioxins and polychlorobiphenyls

Several studies have shown that at present, the exposure of part of the population in Western countries to dioxins and dioxin-like compounds exceeds the exposure limits set by the WHO and SCF (Liem et al., 2000). A recent study in the Netherlands showed that the median of the lifelong intake of dioxins and dioxin-like PCBs by the Dutch population at present is 1.2 pg TEQ/kg bw. Dioxins and PCBs contributed to a similar extent to the exposure. The 90th percentile was 1.9 pg TEQ/kg bw and it was estimated that 8% of the population exceeded the current exposure limits (Baars et al., 2004). These intakes should be compared with the exposure limit (pTWI) of 14 pg TEQ/kg bw/week or on average 2 pg TEQ/kg bw/day. In the same study, the median and 90th percentile for the estimated intake figures for indicator PCBs were respectively 5.6 and 10.4 ng/kg bw/day. Dairy products, meat, oils and fats, fish and vegetable products contributed 27, 23, 17, 16 and 13% respectively to the intake of dioxin-like compounds. Similar figures were obtained for the indicator PCBs, although fish showed a 10% higher contribution. Slightly higher intake levels for dioxins only were obtained by Llobet et al. (2003) for the Catalonia region in Spain, showing an average intake of 1.4 pg TEQ/kg bw/day. The higher intake might be explained by the higher consumption of fish products that contributed for 31% to the intake. In a recent study from Japan the relative contribution of fish products was even higher, being more than 50% of the average daily intake of respectively 1.6 and 3.2 pg TEQ/kg bw/day for dioxins and the sum of dioxins and dioxin-like PCBs (Tsutsumi et al., 2001). It is important to note that in general the intake levels have decreased during the last decades, due to various measures to control the major sources. In the Netherlands the intake of dioxins and dioxin-like PCBs decreased from about 8 and 4 pg TEQ/kg bw/day in 1978 and 1984 to respectively 2.2 and 1.2 pg TEQ/kg bw/day in 1990 and 1999 (Baars et al., 2004). This decreased intake is also reflected in the levels detected in human milk, which also showed a marked decrease in the various countries. Apart from incidental high exposures, the major impact of an incident like the one in Belgium in 1999 is probably that these efforts are temporarily diminished. Consumers with certain preferences may have an increased intake of dioxins and PCBs. It has been shown for example, that eel from contami-

Dioxins, polychlorinated biphenyls and brominated flame retardants

395

nated rivers may contain relatively high levels of dioxins and dioxin-like PCBs (Hoogenboom et al., 2006; Leeuwen et al., 2007). As a result, the frequent consumption of such eel may result in elevated exposure. Unfortunately, the same may be true for people consuming eggs from free-range chickens, especially when the hens spend a great deal of their time outside (Kijlstra et al., 2007). These eggs may contain levels up to 10 pg TEQ/g fat or more and therefore contribute to a significant exposure.

15.5.1 Limits in food and feed In order to further reduce the intake of the population, the EU has developed a strategy for further reducing exposure. This includes the establishment of residue limits for dioxins in food products: 1 pg TEQ/g fat for pork, 2 pg TEQ/g fat for poultry, and 3 pg TEQ/g fat for beef, milk and eggs (EC, 2001, 2006a). In addition limits have been set for animal feed (0.75 ng TEQ/ kg) and their ingredients (0.75–6 ng/kg TEQ) (EC, 2002, 2006b). Since November 2006, planar dioxin-like PCBs are included in the EU food and feed limits. As a result there are limits for dioxins and the sum of dioxins and dl-PCBs for each food and feed item. In addition, the EU has set action limits for dioxins and dl-PCBs, being about 60% of the tolerance limits. Samples exceeding the action limit require follow-up to determine the source of the pollution. Eventually, this approach should result in a further decrease of the levels and guarantee that the no consumers exceed the exposure limit or pTWI of 14 pg TEQ/kg bw/week.

15.6

Brominated flame retardants

Flame retardants are compounds used to decrease the flammability of plastics and thereby the risk of fire. Several classes of compounds have been or are being used, like the polybrominated biphenyls (PBBs), the polybrominated diphenylethers (PBDEs), tetrabromobisphenol A (TBBP-A) and hexabromocyclododecane (HBCD) (Fig. 15.3). PBBs are no longer produced and have been replaced by other compounds. PBDEs have been produced in the Netherlands, France, the UK, the US and Japan, primarily as the tetra-, penta-, octa- and deca-mixtures (de Boer, 2000). The pentamixture contains primarily tetra- and penta-BDEs, the octa-mixture heptaand octa-BDEs and the decamixture deca-PBE. High quantities are added to plastics such as polyurethane foam. TBBP-A is used as an additive in electronic equipment (circuit boards). The use of the tetra and penta-PBDE mixtures have been forbidden in Europe since 2003 but the deca-mixture is still used. PBBs were involved in one of the most tragic accidents in foodproducing animals, being the Michigan incident, which occurred in the autumn of 1973 (Carter, 1976). The feed additive Nutrimaster (magnesium

396

Endocrine-disrupting chemicals in food 2′

2

Bry 3

4′

4 5

6′

6

Br

3′ Br x

C

HO

5′

Br

PBB

Br

CH3

OH

CH3

Br Tetrabromobisphenol A Br

Br Bry

2

3 4

2′

O 6

Brx

Br

Br

3′ 4′

6′

5

5′ PBDE Br

Br HBCD

Fig. 15.3 Structures of the most important polybrominated flame retardants. In the case of the PBDEs, congeners numbered 47, 99, 100, 153, 154 and 209 correspond to respectively 2,4,2′,4′-tetraBDE, 2,4,5,2′,4′-pentaBDE, 2,4,6,2′,4′-pentaBDE, 2,4,5,2′,4′,5-hexaBDE, 2,4,5,2′,4′,6′-hexaBDE and 2,3,4,5,6,2′,3′,4′,5′,6′-decaBDE.

oxide) was accidentally substituted with FireMaster FF-1, a flame-retardant based on PBBs. Total PBB intakes as high as 250 g per cow in about 16 days were estimated, initially resulting in symptoms such as body weight loss, a decreased milk production and a very typical effect on hoof growth. Fries (1985) estimated that up to 250 kg of FireMaster was fed to livestock, of which 125 kg was eliminated through the faeces and 94 kg ended up in human foods before regulation. Elevated levels of PBBs can still be measured in the milk of women in the affected area. In general, the use of PBBs has stopped and these compounds do not appear to be a major concern any more. However, this may not be the case for the PBDEs and HBCD.

15.6.1 Toxicity of flame retardants In general the data on the toxicity of the PBDEs is still limited (Darnerud et al., 2001). The decabrominated compound has been tested extensively, showing that the compound is not mutagenic or genotoxic (Hardy, 2002). There are, however, indications for weak carcinogenic effects based on tumors in livers of male mice. Preliminary reports on the potential binding of the non-planar compounds to the Ah receptor (Meerts et al., 1998; Brown et al., 2004) have now been attributed to impurities. This is shown by more recent data from studies with purified standards (Peters et al., 2006). This contradiction may be explained by the presence of brominated dioxins in the test compounds, but at the same time there are strong indications that such impurities are also present in the technical products used in practice.

Dioxins, polychlorinated biphenyls and brominated flame retardants

397

Furthermore, they may be formed during combustion of PBDEs (Sakai et al., 2001). Although data on the toxicity of bromodioxins are rather limited, Birnbaum et al. (2003) concluded that they are similarly toxic as their chlorinated counterparts. Based on the criteria described above, including their mechanism of action, they should be included in the TEQ principle and may in practice be added to the dioxins and dioxin-like PCBs present in food. Whether these compounds would actually contribute to a significant extent to the TEQ levels in, for example, food is unclear since very limited data are available on their occurrence in food. More data are available on the PBDEs and several recent studies have estimated the current exposure to these compounds (see below). In the case of PBDEs, there are indications for an increased sensitivity of the fetus. As in the case of non-dioxin-like PCBs, exposure of young animals to the pentabromo congener BDE-99 (0.8 and 12 mg/kg bw/day) resulted in decreased learning and behavioral disturbance at a later age (Viberg et al., 2007). Comparable effects, but to a lesser extent, were observed for the tetrabromo congener BDE-47, whereas TBBP-A showed no effects. Effects on behavior have also been reported for the decabromo congener BDE-209 (Viberg et al., 2007). Similar to certain non-dioxin-like PCBs, PBDEs can affect the homeostasis of thyroid hormones. The no observable adverse effect level (NOAEL) for this effect of the pentabromoBDE is around 1 mg/kg bw/day. Using a factor of 10, a NOAEL of 0.1 mg/ kg bw was calculated by Darnerud et al. (2001). However, in order to extrapolate the NOAEL from animal studies to humans, it is essential to take into account potential species differences in the kinetics, as in the case of dioxins and PCBs. Bakker et al. (2008) used the very limited data about kinetics and calculated a human NOAEL for BDE-99 of 0.2–0.3 ng/kg bw/ day. This limit was based on the effects of BDE-99 on reduced sperm counts in rats, following exposure of the dams (Kuriyama et al., 2005) with a lowest observable adverse effect level (LOAEL) of 60 μg/kg bw. An uncertainty factor of 95 was used to extrapolate to a NOAEL and to account for interand intraspecies differences. Further studies are required to deal with the various uncertainties in this approach. The toxicity of HBCD has been investigated by the producers and is summarized in an Environmental Protection Agency (EPA) report (ACC, 2001; Hardy, 2001). HBCD consists of three stereo isomers (α,β,γ) which contribute 6, 8 and 80% respectively. The metabolism in rats is very extensive and more than 80% of a radiolabeled substance was excreted during the first 72 hours. The alpha congener was the relatively most persistent congener in the tissues. This was confirmed in a recent study with rats by Van der Ven et al. (2006). There are no indications that HBCD is mutagenic, genotoxic or carcinogenic. Exposure of rats for 28 or 90 days resulted in reversible effects on liver size, as well as histological effects on liver and thyroid. There was also a decrease in thyroid hormone (T4) levels in the blood. The LOAEL was around 300 mg/kg bw/day, the NOAEL at

398

Endocrine-disrupting chemicals in food

100 mg/kg bw/day. Within the EU-FIRE project, the toxicity of HBCD was re-evaluated. A 28-day rat trial was performed with a dose-range varying between 0.3 and 200 mg/kg bw/day. Again, effects on increased liver and thyroid weight, as well as decreased total T4 levels in blood were observed. Furthermore, the activity of T4-UGT, involved in the glucuronidation of T4 was increased at the higher dose levels. Effects were only clear in female rats. Groups were too small to observe statistical significance for the effects, although a NOAEL of 10 mg/kg bw for the increased thyroid and liver weight seemed obvious from the data. However, using a benchmark dose approach, the authors calculated a benchmark dose low (BMDL) of 1.6 mg/ kg bw/day based on the lower 5% confidence interval for a 10% increase of thyroid weight. This BMDL may be regarded as a NOAEL. In this study, this level of exposure resulted in an HBCD level in liver of 40 μg/g lipid. As in the case of PCBs and dioxins, the effects on the thyroid could be caused by the induction of liver enzymes, resulting in an increased degradation of T4 and a feed back regulation on the thyroid. Germer et al. (2006) showed that the exposure to HBCD induced drug-metabolizing enzymes probably via the constitutive-androstane receptor/pregnane-X receptor (CAR/PXR) signaling pathway. The induction of cytochrome P450s and co-regulated enzymes of phase II of drug metabolism could affect the homeostasis of endogenous substrates such as steroid and thyroid hormones and as such underly the observed effects. Hamers et al. (2006) tested a number of different BFRs and their hydroxylated metabolites in a battery of in vitro tests, developed for the detection of hormonal and antihormonal effects. They also included the competition with transthyretin, the transport protein for thyroid hormones. It was concluded that several BFRs have endocrine-disrupting potencies that have not been described before, especially in vivo. This may be due to pharmacokinetic aspects but further investigations seem appropriate to exclude these subtle effects.

15.6.2 Analysis of flame retardants Most of the BFRs such as the PBBs and PBDEs can be analysed by GC using ECD or MS detection. In particular the BDE 209 may cause problems and requires specific measures. HBCD, being a mixture of three different congeners, can be analysed by liquid chromatography – mass spectrometry (LC/MS). Bromodioxins, like their chlorinated counterparts, require HRGC/ HRMS analysis. However, these compounds are also detected in the CALUX screening assay (Brown et al., 2004).

15.6.3 Exposure to flame retardants Darnerud et al. (2001) estimated the average exposure in Sweden at 51 ng/ day based on market basket samples. This study was based on the congeners

Dioxins, polychlorinated biphenyls and brominated flame retardants

399

47, 99, 100, 153 and 154. Using the same congeners, Ryan and Patry (2001), Lind et al. (2002) and Wijesekera et al. (2002) calculated intakes for the Canadian, Swedish and UK population of respectively 44, 41 and 91 ng/day. Bocio et al. (2003) calculated from the total diet study mentioned above for dioxins (Llobet et al., 2003), a slightly higher intake for the Catalonian population being around 90 ng/day. An even higher intake was calculated by Winter-Sorkina et al. (2003) for the Dutch population, being around 213 ng/day or 3 ng/kg bw/day. However, a more recent study by Bakker et al. (2008) calculated a median lifelong exposure of 0.8 ng/kg bw/day. In these studies it was shown that fish, pork, dairy and beef contributed most to the intake. Schecter et al. (2008) analyzed food from the US market and showed that in particular the levels of PBDEs, and as a consequence the exposure, is much higher than in Europe. This is consistent with human milk levels, also being an order of magnitude higher in the US. PBDEs have also been detected in human blood and milk in the lower ng/g lipid levels, thus being much lower than levels of PCBs. Although levels have increased over the last decades, there are some indications that at least for PBDEs these levels have reached their maximum and started to decline (Sjödin et al., 2003). Winter-Sorkina et al. (2003) also estimated the average daily intake of HBCD in the Netherlands and came to a figure of 190 ng/ day. This figure was similar to the one estimated by Lind et al. (2002) for the average daily intake in Sweden, being around 162 ng/day with a 95th percentile of 332 ng/day. Based on the NOAEL of 1.6–100 mg/kg bw/day mentioned above, it is clear that the margin between the current average exposure around 3 ng/kg bw/day and the toxicological relevant intake is several orders of magnitude.

15.7

Abbreviations

2,4,5-T: Ah: bw: EROD: GC-ECD: HBCD: HpCDD: HpCDF: HxCB: HxCDD: HxCDF: LOAEL: MWI: NOAEL: OCDD:

2,4,5-trichlorophenoxyacetic acid arythydrocarbon body weight ethoxyresorufin-O-deethylase gas chromatography–electron capture detection hexabromocyclododecane heptachlorodibenzo-p-dioxin heptachlorobenzofuran hexachlorobiphenyl hexachlorodibenzo-p-dioxin hexachlorobenzofuran lowest observable adverse effect level municipal waste incinerator no observable adverse effect level octachlorodibenzo-p-dioxin

400

Endocrine-disrupting chemicals in food

OCDF: PBB: PBDEs: PCBs: PCDD: PCDF: PCP: PeCB: PeCDD: PeCDF: SCF: TCDD: TCDF: TeCB: TEF: TEQ: TSH: WHO:

15.8

octachlorobenzofuran polybromobiphenyls polybromodiphenylethers polychlorobiphenyls polychlorinated dibenzo-p-dioxins polychlorinated dibenzofurans pentachlorophenol pentachlorobiphenyl pentachlorodibenzo-p-dioxin pentachlorobenzofuran Scientific Committee on Food 2,3,7,8-tetrachlorodibenzo-p-dioxin 2,3,7,8-tetrachlorobenzofuran tetrachlorobiphenyl toxic equivalency factor toxic equivalents thyroid-stimulating hormone World Health Organization

References

acc (2001) ‘Data summary and test plan for hexabromocyclododecane (HBCD)’, at www.epa.gov/chemrtk/cyclodod/c13459.pdf. baars a j, bakker m i, baumann r a, boon p e, freijer j i, hoogenboom l a p, hoogerbrugge r, klaveren j d, liem a k d, traag w a and de vries j (2004) ‘Dioxins, dioxin-like PCBs and non-dioxin-like PCBs in foodstuffs: occurrence and dietary intake in the Netherlands’, Toxicol. Lett., 151, 51–61. bakker m i, de winter-sorkina r, de mul a, boon p e, van donkersgoed g, van klaveren j d, baumann b a, hijman w c, van leeuwen s p j, de boer j and zeilmaker m j (2008) ‘Dietary intake and risk evaluation of polybrominated diphenyl ethers in The Netherlands’, Mol. Nutr. Food Res., 52, 204–216. bell d r, clode s, fan m q f, fernandes a, foster p m d, jiang t, loizou g, macnicoll a, miller b g, rose m, tran l and white s (2007a) ‘Toxicity of 2,3,7,8tetrachlorodibenzo-p-dioxin in the developing male Wistar(Han) rat. I: no decrease in epididymal sperm after a single acute dose’, Toxicol. Sci., 99, 214–223. bell d r, clode s, fan m q f, fernandes a, foster p m d, jiang t, loizou g, macnicoll a, miller b g, rose m, tran l and white s (2007b) ‘Toxicity of 2,3,7,8tetrachlorodibenzo-p-dioxin in the developing male Wistar(Han) rat. II: Chronic dosing causes developmental delay’, Toxicol. Sci., 99, 224–233. bernard a, hermans c, broeckaert f, poorter g de, cock a de and hoins g (1999) ‘Food contamination by PCBs and dioxins; an isolated episode in Belgium is unlikely to have affected public health’, Nature, 401, 231–232. birnbaum l s, staskal d f and diliberto j j (2003) ‘Health effects of polybrominated dibenzo-p-dioxins (PBDDs) and dibenzofurans (PBDFs)’, Environment Int., 29, 855–860. bocio a, llobet j m, domingo j l, corbella j, teixidoa a and casas c (2003) ‘Polybrominated diphenyl ethers (PBDEs) in foodstuffs: human exposure through the diet’, J. Agric. Food Chem., 51, 3191–3195.

Dioxins, polychlorinated biphenyls and brominated flame retardants

401

de boer j (2000) ‘Polybrominated biphenyls and diphenylethers. New types of persistent halogenated compounds’, J. Paasvirta, 3 (k), 62–95 brouillette j and quirion r (2008) ‘The common environmental pollutant dioxininduced memory deficits by altering estrogen pathways and a major route of retinol transport involving transthyretin’, Neurotoxicology, 29, 318–327. brown d j, van overmeire i, goeyens l, denison m s, de vito m j, clark g c (2004) ‘Analysis of Ah receptor pathway activation by brominated flame retardants’, Chemosphere, 55, 1509–1518. carter l j (1976) ‘Michigan’s PBB incident: chemical mix-up leads to disaster’, Science, 192, 240–243. darnerud p o, eriksen g s, joa hannesson t, larsen p b and viluksela m (2001) ‘Polybrominated diphenyl ethers: occurrence, dietary exposure, and toxicology’, Environ. Health Persp. Suppl., 109, 49–68. ec (2001) ‘Council Regulation (EC) No 2375/2001 of 29 November 2001 amending Commission Regulation (EC) No 466/2001 setting maximum levels for certain contaminants in foodstuffs’, Off. J. European Communities, L321, 1–5. ec (2002) ‘Council Directive 2001/102/EC of 27 November 2001 amending Directive a999/29/EC on the undesirable substances and products in animal nutrition’, Off. J. European Communities, L6, 45–49. ec (2006a) ‘Commission Regulation (EC) No 199/2006, of 3 February 2006 amending Regulation (EC) No 466/2001 setting maximum levels for certain contaminants in foodstuffs as regards dioxins and dioxin-like PCBs’, Off. J. European Union, L32, 34–38. ec (2006b) ‘Commission Directive 2006/13/EC of 3 February 2006 amending Annexes I and II to Directive 2002/32/EC of the European Parliament and of the Council on undesirable substances in animal feed as regards dioxins and dioxinlike PCBs’, Off. J. European Union, L32, 44–53. efsa (2005) Opinion of the scientific panel on contaminants in the food chain on a request from the Commission related to the presence of non-dioxin-like polychlorinated biphenyls (PCB) in feed and food. The EFSA J., 284, 1–137. faqi a s, dalsenter p r, merker h j and chahoud i (1998) ‘Reproductive toxicity and tissue concentrations of low doses of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male offspring rats exposed throughout pregnancy and lactation’, Toxicol. Appl. Pharmacol., 150, 383–392. fenton s e, hamm j t, birnbaum l s and youngblood g l (2002) ‘Persistent abnormalities in the rat mammary gland following gestational and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)’, Toxicol. Sci., 67, 63–74. ferrario j p, byrne c j and cleverly d h (2000), ‘2,3,7,8±Dibenzo-p-dioxins in mined clay products from the United States: evidence for possible natural origin’, Environ. Sci. Technol., 34, 4524–4532. firestone d (1973) ‘Etiology of chick edema disease’, Environ. Health Perspect., 5, 59–66. fries g f (1985) ‘The PBB episode in Michigan: an overall appraisal’, CRC Crit. Rev. Toxicol., 16, 105–156. fukuzawa n h, ohsaka s, wu q, sakaue m, fujii-kuriyama y, baba t and tohyama c (2004) ‘Testicular cytochrome P450scc and LHR as possible targets of 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD) in the mouse’, Mol. Cell. Endocrinol., 221, 87–96. germer s, piersma a h, van der ven l, kamyschnikow a, fery y, schmitz h j and schrenk d (2006), ‘Subacute effects of the brominated flame retardants hexabromocyclododecane and tetrabromobisphenol A on hepatic cytochrome P450 levels in rats’, Toxicology, 218, 229–236. giacomini s m, hou l f, bertazzi p a and baccarelli a (2006) ‘Dioxin effects on neonatal and infant thyroid function: routes of perinatal exposure, mechanisms

402

Endocrine-disrupting chemicals in food

of action and evidence from epidemiology studies’, Int. Arch. Occupational Environ. Health, 79, 396–404. guo j, sartot m, karayalas, medvedovic m, kann s, puga a, ryan p and tomlinson c g (2004) ‘Expression of genes in the TGF-α signalling pathway is significantly deregulated in smooth muscle cells from aorta of aryl hydrocarbon receptor knockout mice’, Toxicol. Appl. Pharmocol., 1994, 79–89. haavisto t, nurmela k, pohjanvirta r, huuskonen h, el-gehani f, paranko j (2001) ‘Prenatal testosterone and luteinizing hormone levels in male rats exposed during pregnancy to 2,3,7,8-tetrachlorodibenzo-p-dioxin and diethylstilbestrol’, Mol. Cell. Endocrinol., 178, 169–179. hamers t, kamstra j h, sonneveld e, murk a j, kester m h, andersson p l, legler j and brouwer a (2006), ‘In vitro profiling of the endocrine-disrupting potency of brominated flame retardants’, Toxicol. Sci., 92, 157–173. hardy m l (2001) ‘Regulatory status and environmental properties of brominated flame retardants undergoing risk assessment in the EU: DBDPO, OBDPO, PeBDPO and HBCD’, Polym. Degrad. Stability 64, 545–556. hardy m l (2002) ‘The toxicology of the three commercial polybrominated diphenyl oxide (ether) flame retardants’, Chemosphere, 46, 757–777. hayward d g, nortrup d, gardner a and clower m (1999) ‘Elevated TCDD in chicken eggs and farm-raised catfish fed a diet with ball clay from a Southern United States mine’, Environ. Res. Section A, 81, 248–256. higgenbotham g r, huang a, firestone d, verrett j, ress j and campbell a d (1968) ‘Chemical and toxicological evaluations of isolated and synthetic chloro derivatives of dibenzo-p-dioxin’, Nature, 220, 702–703. hoogenboom l a p (2002) ‘The combined use of the CALUX bioassay and the HRGC/HRMS method for the detection of novel dioxin sources and new dioxinlike compounds’, Environ. Sci. Pollution Res., 9, 9–11. hoogenboom l a p, traag w a and mengelers m j b (1999), ‘The Belgian dioxin crisis; involvement of RIKILT’, poster presented at Dioxin ’99 in Venice, Italy. hoogenboom l a p, bovee t f h, portier l, bor g, weg g van der, onstenk c and traag w a (2004) ‘The German bakery waste incident; use of a combined approach of screening and confirmation for dioxins in feed and food’, Talanta, 63 (5), 1249–1253. hoogenboom l a p, zeilmaker m, kan k, mengelers m j b, eijkeren j van, and traag w a (2005) ‘Kaolinic clay derived dioxins in potato by-products’, Organohalogen Compounds, 67, 1470–1473. hoogenboom l a p, bovee t h g, traag w a, hoogerbrugge r, baumann b, portier l, van der weg g and de vries j (2006) ‘The use of the DR CALUX® bioassay and indicator PCBs for screening of elevated levels of dioxins and dioxin-like PCBs in eel’, Mol. Nutr. Food Res., 50, 945–957. hoogenboom l a p, van eijkeren j c h, zeilmaker m j, mengelers m j b, herbes r, immerzeel j and traag w a (2007) ‘A novel source for dioxins present in recycled fat from gelatin production’, Chemosphere, 68, 814–823. hsu s t, ma c i, hsu s k, wu s s, hsu n h and yeh c (1985) ‘Discovery and epidemiology of PCB poisoning in Taiwan: a four-year follow-up’, Environ. Health Perspect. 59, 5–10. jenkins s, rowell c, wang j and lamartiniere c a (2007) ‘Prenatal TCDD exposure predisposes for mammary cancer in rats’, Reproductive Toxicol., 23, 391–396. kijlstra a, traag w a, hoogenboom l a p (2007) ‘Effect of flock size on dioxin levels in eggs from chickens kept outside’, Poultry Sci., 86, 2042–2048. knerr s and schrenk d (2006) ‘Carcinogenicity of 2,3,7,8-tetrachlorodibenzo-pdioxin in experimental models’, Mol. Nutr. Food Res., 50, 897–907. kuratsune m, yoshimura t, matsuzaka j and yamaguchi a (1972) ‘Epidemiologic study on Yusho, a poisoning caused by ingestion of rice oil contaminated with a

Dioxins, polychlorinated biphenyls and brominated flame retardants

403

commercial brand of polychlorinated biphenyls’, Environ. Health Perspect., 1, 119–128. kuriyama s n, talsness c e, grote k and chahoud i (2005) ‘Developmental exposure to low-dose PBDE-99: effects on male fertility and neurobehavior in rat offspring’, Environ. Health Perspect., 113, 149–154. larebeke n van, hens l, schepens p, covaci a, bayens j, everaert k, bernheim j l, vlietinck r and poorter g de (2001) ‘The Belgian PCB and dioxin incident of January–June 1999: exposure data and potential impact on health’, Environ. Health Perspect., 109, 265–273. leeuwen s p j van, leonards p e g, traag w a, hoogenboom l a p and boer j de (2007) Polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls in fish from the Netherlands: concentrations, profiles and comparison with DR CALUX® bioassay results. Anal. Bioanal. Chem., 389, 321–333. liem a k d, hoogerbrugge r, kootstra p r, van der velde e g and de jong a p j m (1991) ‘Occurrence of dioxins in cow’s milk in the vicinity of municipal waste incinerators and a metal reclamation plant in the Netherlands’, Chemosphere, 23, 1975–1984. liem a k d, fürst p and rappe c (2000) ‘Exposure of populations to dioxins and related compounds’, Food Additives Contam., 4, 241–259. lind y, aune m, atuma s, becker w, bjerselius r, glyn a and darnerud p o (2002) ‘Food intake of the brominated flame retardants PBDEs and HBCD in Sweden’, Organohalogen Comp., 58, 181–184. llerena j j, abad e, caixach j and rivera j (2003) ‘An episode of dioxin contamination in feedingstuff: the choline chloride case’, Chemosphere, 53, 679–683. llobet j m, domingo j l, bocio a, casas c, teixidó a and müller l (2003) ‘Human exposure to dioxins through the diet in Catalonia, Spain: carcinogenic and noncarcinogenic risk’, Chemosphere, 50, 1193–1200. malisch r (2000) ‘Increase of the PCDD/F-contamination of milk, butter and meat samples by use of contaminated citrus pulp’, Chemosphere, 40, 1041–1053. meerts i, luijks e, marsh g, jakobsson e, bergman a and brouwer a (1998) ‘Polybrominated diphenylethers (PCBEs) as Ah-receptor agonists and antagonists’, Organohalogen Comp., 37, 147–150. moran f m, vandevoort c a, overstreet j w, lasley b l and conley a j (2003) ‘Molecular target of endocrine disruption in human luteinizing granulosa cells by 2,3,7,8-tetrachlorodibenzo-p-dioxin: inhibition of estradiol secretion due to decreased 17 alpha-hydroxylase/17,20-lyase cytochrome P450 expression’, Endocrinology, 144, 467–473. murphy k a, quadro l and white l a (2007) ‘The intersection between the aryl hydrocarbon receptor (AHR)- and retinoic acid-signaling pathways’, Vitamin A, 75, 33–67. nayyar t, bruner-tran k l, piestrzeniewicz-ulanska d and osteen k g (2007) ‘Developmental exposure of mice to TCDD elicits a similar uterine phenotype in adult animals as observed in women with endometriosis’, Reproductive Toxicol., 23, 326–336. olie k, vermeulen p l and hutzinger o (1977) ‘Chlorodibenzo-p-dioxins and chlorodibenzofurans are trace components of fly ash and flue gas of some municipal incinerators in the Netherlands’, Chemosphere, 8, 455–459. peters a k, nijmeijer s, gradin k, backlund m, bergman å, poellinger l, denison m s and berg m van den (2006) ‘Interactions of polybrominated diphenyl ethers with the aryl hydrocarbon receptor pathway’, Toxicol. Sci., 92, 133–142. plas s a van der, sundberg h, berg h van den, scheu g, wester p, jensen s, bergman a, de boer j, koeman j h and brouwer a (2001) ‘Contribution of planar (0–1 ortho) and non-planar (2 ± 4 ortho) fractions of Arochlor 1260 to the induction of altered

404

Endocrine-disrupting chemicals in food

hepatic foci in female Spague-Dawley rats’, Toxicol. Appl. Pharmacol., 169, 255–268. rier s e, martin d c, bowman r e, dmowski w p and becker j l (1993) ‘Endometriosis in rhesus monkeys (Macata mulatta) following chronic exposure to 2,3,7,8 tetrachlorodibenzo-p-dioxin’, Fundamental Appl. Toxicol., 21, 433–441. ryan j k and patry b (2001) ‘Body burdens and food exposure in Canada for polybrominated diphenyl ethers (BDEs)’, Organohalogen Comp., 51, 226–229. safe s h (1994), ‘Polychlorinated biphenyls (PCBs): environmental impact, biochemical and toxic responses, and implications for risk assessment’, Crit. Rev. Toxicol., 24, 87–149. safe s h (1995) ‘Human dietary intake of arylhydrocarbon (Ah) receptor agonists: mass balance estimates of exodioxins and endodioxins and implications for health assessment’, Organohalogen Comp., 26, 7–13. safe s h (1998) ‘Development validation and problems with the toxic equivalency factor approach for risk assessment of dioxins and related compounds’, J. Anim. Sci., 76, 134–141. sakai s, watanabe j, honda y, takatsuki h, aoki i, futamatsu m and shiozaki k (2001) ‘Combustion of brominated flame retardants and behavior of its byproducts’, Chemosphere, 42, 519–531. sanger v l, scott l, hamdy a, gale c and pounden w d (1958) ‘Alimentary toxemia in chickens’, J. Am. Vet. Med. Assoc., 133, 172–176. scf (2001) ‘Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food. Scientific Committee on Food report CS/ CNTM/DIOXIN/20 final’, http://europa.eu.int/comm/food/fs/sc/scf/out90_en.pdf. schecter a, harris t r, shah n, musumba a and päpke o (2008) ‘Brominated flame retardants in US food’. Mol. Nutr. Food Res., 52, 266–272. schmidt c w (1999) ‘Poisoning young minds’, Environ. Health Perspect., 107, 302–307. schmittle s c, edwards h m and morris d (1958) ‘A disorder of chickens probably due to a toxic feed-preliminary report’, J. Am. Vet. Med. Assoc., 132, 216–219. sjödin a, patterson d g and bergman a (2003) ‘A review on human exposure to brominated flame retardants-particularly polybrominated diphenyl ethers’, Environ. Int., 29, 829–839. tsutsimi t, yanagi t, nakamura m, kono y, uchibe h, iida t, hori t, nakagawa r, tobiishi k, matsuda r, sasaki k and toyoda m (2001) ‘Uptake of daily intake of PCDDs, PCDFs, and dioxin-like PCBs from food in Japan’, Chemosphere, 45, 1129–1137. van den berg m, birnbaum l s, bosveld a t c, brunström b, cook ph, feeley m, giesy j p, hanberg a, hasegawa r, kennedy s w, kubiak t, larsen j c, leeuwen f x r van, liem a k d, nolt c, peterson r e, poellinger, l, safe, s, schrenk, d, tillitt, d, tysklind, m, younes, m., wærn, f and zacharewski, t (1998) ‘Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife’, Environ. Health Perspect., 106, 775–792. van den berg m, birnbaum l, denison m, devito m, farland w, feeley m, foedler h, hakanson h, hanberg a, haws l, rose m, safe s, schrenk d, tohyama c, tritscher a, tuomisto j, tysklind m, walker n and peterson r e (2006) ‘The 2005 World Health Organisation reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds’, Toxicol. Sci., 93, 223–241. van der ven l t, verhoef a, van de kuil t, slob w, leonards p e, visser t j, hamers t, herlin m, hakansson h, olausson h, piersma a h and vos j g (2006), ‘A 28-day oral dose toxicity study enhanced to detect endocrine effects of hexabromocyclododecane in wistar rats’, Toxicol. Sci., 94, 281–292. viberg h, fredriksson a and eriksson p (2007) ‘Changes in spontaneous behaviour and altered response to nicotine in the adult rat, after neonatal exposure to the

Dioxins, polychlorinated biphenyls and brominated flame retardants

405

brominated fire retardant, decabrominated diphenyl ether (PBDE 209)’, Neurotoxicology, 28, 136–142. who (2000) ‘Consultation on assessment of the health risk of dioxins; re-evaluation of the tolerable daily intake (TDI): executive summary’, Food Additives Contaminants, 17, 223–240. wijesekera r, halliwell c, hunter s and harrad s a (2002) ‘A preliminary assessment of UK human exposure to polybrominated diphenyl ethers (PBDEs)’, Organohalogen Comp., 55, 239–242. winter-sorkina r de, bakker m i, donkersgoed g van and klaveren j d van (2003) ‘Dietary intake of brominated flame retardants by the Dutch population’, RIVMreport 310305001/2003, Bilthoven, The Netherlands. wunderli s, zennegg m, dolezal i s, gujer e, moser u, wolfensberger m, hasler p, noger d, studer c and karlaganis g (2000) ‘Determination of polychlorinated dibenzo-p-dioxins and dibenzo-furans in solid residues from wood combustion by HRGC/HRMS’, Chemosphere, 40, 641–649.

16 Bisphenol A J. E. Goodman and L. R. Rhomberg, Gradient Corporation, USA

Abstract: Bisphenol A (BPA) is a chemical used in the manufacture of polycarbonate plastic, epoxy resins, and a number of commonly used products. This chapter discusses the routes by which BPA can enter food, the levels of BPA that have been measured in food, and levels of human BPA intake. It also reviews possible mechanisms of BPA toxicity and the potential for the low levels of BPA, to which people are exposed, to affect health. Finally, the positions of government bodies in their scientific reviews of the potential health risks of BPA are also discussed. Key words: bisphenol A, migration into food, biomonitoring, endocrine-active agent, plastic.

16.1

Introduction

Bisphenol A (BPA) is generated by the condensation of phenol and acetone, which is catalyzed by an acid or alkaline compound (Fig. 16.1a; EU, 2003). Like estradiol (Fig. 16.1b), it can bind to the estrogen receptor (ER) and exhibit both estrogenic and anti-androgenic properties (Kitamura et al., 2005; Xu et al., 2005; Sun et al., 2006). These properties are influenced primarily by the 4-hydroxyl group on the A-phenyl ring and the hydrophobic moiety at the 2-position of the propane moiety (Kitamura et al., 2005). BPA’s most prominent use is in the manufacture of polycarbonate plastic and epoxy resins, but it is also used in the production and processing of polyvinylchloride (PVC) and modified polyamide, and the manufacture of carbonless and thermal paper, wood filler, adhesives, printing inks, surface coatings, polyols/polyurethane, brake fluid, dental resin-based composites and sealants, flame retardants, paints, and tires (EU, 2003; EFSA, 2006). Many of the products manufactured with BPA come into contact with food (EU, 2003; EFSA, 2006). These products manufactured from BPA can

Bisphenol A

407

(a) HO

OH

(b) CH3

OH

H H

H

HO

Fig. 16.1

(a) Bisphenol A and (b) 17β-estradiol.

contain trace amounts of residual monomer and additional BPA may be generated during the breakdown of polymer (EU, 2003), so small amounts of BPA can migrate into food in contact with these products. This chapter begins with a description of routes by which BPA can enter food, followed by a discussion of levels of BPA that have been measured in food and of levels of human BPA intake levels. This is followed by a description of possible mechanisms of BPA toxicity and a review of the potential effects of BPA on human health. Lastly, the positions of government bodies, including those in the European Union, Japan, Canada, and the United States, on potential health risks of BPA are discussed.

16.2

Bisphenol A migration from packaging materials and containers into food and beverages

Although BPA can be released into the environment during the production, processing, and use of BPA-containing materials (EU, 2003), levels in environmental samples are generally very low or undetectable. This is because BPA has low volatility and a short half-life in the atmosphere, is rapidly biodegraded in water, and is not expected to be stable, mobile, or bioavailable from soils based on biodegradation tests (Staples et al., 1998; EU, 2003). There is little evidence that humans are exposed to environmental BPA. Rather, most human exposure occurs via residues contained in food or beverages that have been in contact with polycarbonate plastic or with containers lined with epoxy resins. Minor sources of human exposure to BPA include resin-based composites and sealants used in dentistry, and the use of epoxy-phenolic resin-based paints, wood filler, adhesives, printing inks, and thermal paper (EFSA, 2006).

408

Endocrine-disrupting chemicals in food

16.2.1 Bisphenol A products with food-contact uses The two main products manufactured with BPA that come into contact with food are polycarbonate plastic and epoxy resins. Polycarbonate plastic is clear, has low reactivity, and is shatter-resistant. It is used in water bottles, baby bottles, tableware and food storage containers (EU, 2003). BPA is also a starting material for most types of cross-linked epoxy resins, which are used in the manufacturing, packaging, and transporting of food (Goodson et al., 2002; Arvanitoyannis and Bosnea, 2004). Epoxy resins provide a protective lining on the interior surface of food and beverage cans to prevent the cans’ contents from contacting and reacting with the metal walls, ensuring sterility. Although not normally present in PVC organosol coatings, BPA diglycidyl ether (BADGE) may be used as an additive to scavenge hydrogen chloride in these coatings, which can result in the presence of BPA residues (Goodson et al., 2002). The use of BPA in polycarbonate water pipes and epoxy resins used as a surface coating on residential water storage tanks and wine vats may also lead to food contact (EFSA, 2006). There are few studies on exposure from these sources, but estimates have been made (EFSA, 2006). For example, the mean BPA concentration in 46 of 59 wine samples was 0.58 μg/L, while levels in the remaining 13 samples were below the limits of quantitation. Based on this study, potential residues of BPA in wine have been estimated to be at the same levels as those found in canned beverages. One study examined BPA migration into water from resins on pipes, and reported a range of 0.1–1734 μg BPA/m2 coating, but these coatings were prepared in such a way that the results were not applicable to practical use. As noted above, virtually all BPA reacts during the manufacturing process to create polymers, so the amount of free, unreacted BPA remaining in the finished products is very small. The residual content of BPA in polycarbonate is generally less than 50 mg/kg, and is usually less than 10 mg/kg (EU, 2003). For example, the concentration of residual BPA in polycarbonate plastic used for food contact has been reported to range from 7 to 58 mg/kg (Biles et al., 1997a). The mean concentration of BPA residues in polycarbonate baby bottles from Singapore was 28.1 mg/kg (Wong et al., 2005). Residual BPA present in food packaging has the ability to partially migrate into the packaged food. Some, but not all, studies suggest that this migration is influenced by pH, salt, oil, and glucose in the food (Brotons et al., 1995; Kang et al., 2003; Arvanitoyannis and Bosnea, 2004; Thomson and Grounds, 2005). Food components, particularly fat, can migrate into plastics and increase mobility of plastic components, enhancing its migration into contained food. Migration can also increase if coated cans are heated under pressure (Brotons et al., 1995). Therefore, migration of BPA from packaging is greatest in fatty food or foods packed under elevated temperature and pressure.

Bisphenol A

409

16.2.2 Resulting bisphenol A concentrations in food and beverages Several studies have measured levels of BPA in food or beverages that have come in contact with cans, polycarbonate plastic and other types of packaging (summarized in Tables 16.1, 16.2, and 16.3). Overall, despite contact with packaging and containers, the resulting concentrations of BPA in food and beverages, as experienced by the consumer, are very low. This fact, in addition to the low toxicity of BPA, which will be discussed in the following sections, means that the risk to humans from exposure to BPA in their food and beverages is low. Reported BPA levels in canned foods range from undetectable to several hundred μg/kg, with most levels <100 μg/kg (Table 16.1; Goodson et al., 2002; Imanaka et al., 2001a; FSA, 2001; EFSA, 2006). The highest BPA levels were measured in canned meat – this is both fatty and packed under elevated temperature and pressure and so would be expected to have higher levels of BPA residues. Goodson et al. (2004), in their research on the effects of heat-processing or damaging cans containing food, concluded that there was no appreciable difference in the BPA levels in food in normal cans or in heated/damaged cans. Miyamoto and Kotake (2006) summarized BPA concentrations in food, as measured in other studies. For these Japanese studies, the values measured ranged from 0 to 602 μg/kg, with the maximum level detected, once again, in the ‘meat and eggs’ food group (Miyamoto and Kotake, 2006). BPA levels also ranged from undetectable to several hundred μg/L in canned beverages and other liquids, with the highest levels in canned coconut cream (Table 16.1; Goodson et al., 2002; Brenn-Struckhofva and Cichna-Markl, 2006; Howe et al., 1998; Braunrath et al., 2005; Kang and Kondo, 2002; Inoue et al., 2003; Brotons et al., 1995; Thomson and Grounds, 2005). Concentrations of BPA in undiluted infant formula packaged in BPA-based epoxy coated cans ranged from 0.1 to 13 μg/L (Biles et al., 1997b), while BPA levels ranged from <1.7 to <5.1 μg/L in milk and other dairy beverages stored in polycarbonate plastic (Casajuana and Lacorte, 2004; Maragou et al., 2006). Levels of human BPA intake are discussed below. Reported BPA levels typically range from undetectable levels to ≤50 μg/ L in liquids in polycarbonate bottles (Table 16.2; EU, 2003; EFSA, 2006). For example, the amount of BPA detected in water stored in polycarbonate bottles for up to 39 weeks ranged from undetectable levels to 5 ng/L (Biles et al., 1997a). When hot water was stored in washed, polycarbonate baby bottles for 1 hour (a ‘worst-case’ scenario, since conditions were considered to be ‘more severe than real use’), mean BPA levels ranged from 6.7 to 8.4 μg/L (Brede et al., 2003). In another ‘worst-case’ scenario, several children’s products were examined that had been manufactured with polycarbonate containing over 500 mg/kg BPA in Japan (Kawamura et al., 1998). BPA was measured in water, ethanol, n-heptane or acetic acid that had been in contact with children’s tableware for 30 to 60 minutes at 25, 60 or 95 °C,

2.1 10 8

2 2 5

Braunrath et al. (2005) Imanaka et al. (2001a) Imanaka et al. (2001b)

<20 ND3

8 20

Fish Thomson and Grounds (2005) Goodson et al. (2002)

<10 19 5 <10 tr2

43 11 192

109 44

<10 38 24 <10 7.3

48 35 5.59 13.0 95.3 75

9 8.5 <0.20 11.9 <10 2.3

16 3 4 5 4

24

Maximum (ng/g or ng/mL)

<10

Fruits Thomson and Grounds (2005) Goodson et al. (2002) Braunrath et al. (2005) Yoshida et al. (2001) Imanaka et al. (2001a)

Minimum (ng/g or ng/mL)

27 1 15 6 8 2 9 14

Number of samples

Levels of bisphenol A detected in canned food and beverages1

Vegetables, legumes Thomson and Grounds (2005) Braunrath and Cichna (2005) Goodson et al. (2002) Braunrath et al. (2005) Munguia-Lopez et al. (2002) Goodson et al. (2004) Yoshida et al. (2001) Imanaka et al. (2001a)

Food group

Table 16.1

Canada, USA (Alaska), Thailand Portugal, USA, Seychelles, Canada, South Africa, Thailand, Denmark Mauritius, Morocco Japan Japan

New Zealand, Australia, South Africa, Philippines, China Greece, Italy Indonesia, South Africa, Thailand China, Australia, Thailand South Africa, Japan

New Zealand, Australia, Italy, Thailand, Spain Austria UK, the Netherlands, Belgium, Italy, Canada Germany, Italy Mexico UK Japan, USA, Indonesia, China USA, Japan, China

Country

5 8 8

Sugar, sweets Thomson and Grounds (2005) Goodson et al. (2002) Goodson et al. (2004)

Beverages Thomson and Grounds (2005) Braunrath and Cichna (2005) Goodson et al. (2002) Braunrath et al. (2005) Kang and Kondo (2002) Brenn-Struckhofova and Cichna-Markl (2006) Howe et al. (1998) Horie et al. (1999) Kawamura et al. (1998)

6 10 4 8

Meat, poultry Thomson and Grounds (2005) Goodson et al. (2002) Goodson et al. (2004) Imanaka et al. (2001a)

UK, Unknown Austria Japan Austria

<7 3.4 84 2.1 ND7 212.1 213

ND3 0.1 14.1 ND6 ND7 ND8 ND9

16 7 3 59 3 80 47

USA Japan Japan

New Zealand

<10

New Zealand, Singapore, Australia UK, Unknown UK

<10

<20 14 77.3

Australia, New Zealand, USA the Netherlands, Denmark, Brazil UK Japan

New Zealand, Italy UK UK

New Zealand Japan

Australia, New Zealand UK, the Netherlands Austria, Thailand UK Japan

4

<10 ND3 10.5

98 422 25.7 602

<10 90.7 0.041

<10 67.3 ND5

4 2 7 10 16 8.6 17

21 86

<20 53 37.6 22.0 11

<10 ND4

<10 ND3 9.6 18.5 N/A

4 2

4 24 6 2 1

Sauces Thomson and Grounds (2005) Imanaka et al. (2001a) Pasta, grains Thomson and Grounds (2005) Goodson et al. (2004) FSA (2001)

Soups Thomson and Grounds (2005) Goodson et al. (2002) Braunrath et al. (2005) Goodson et al. (2004) Imanaka et al. (2001a)

Continued

USA

UK, France, Unknown Taiwan Japan Greece

Japan, USA, Indonesia, China, Australia, Thailand Brazil, Spain, France Turkey, USA

Thailand, Samoa, Sri Lanka Thailand Japan, China, Spain, Argentina, Canada, Hungary, USA, Mexico, New Zealand, Australia, Italy, Vietnam

Country

2

Data from heat-processed or damaged food cans not included. Imanaka et al. (2001a) define ‘tr’ as 0.5≤ tr < 1.5 ng/g. 3 Goodson et al. (2002) LOD for BPA is 2 ng/g. 4 LOD for Imanaka et al. (2001a) is <0.5 ng/g. 5 LOQ for FSA (2001) is 2 ng/g. 6 LOQ for Brenn-Struckhofova and Cichna-Markl (2006) is 0.2 ng/ml. 7 LOD for Howe et al. (1998) is 5 ppb. 8 Detection limit for Horie et al. (1999) was <2 ng/ml. 9 Detection limit for Kawamura et al. (1998) was <2 ng/ml. 10 Type of plastic used for storage in Inoue et al. (2003) not specified. About 90% of the honey samples had originally been stored in epoxy-resin coated metal drums. 11 Limit of detection for Inoue et al. (2003) was <2.0 ng/g. 12 Detection limit not specified for Brotons et al. (1995). 13 Data for Brotons et al. (1995) is in μg/kg, based on the weight of the can.

1

13.2

0.1

6 5 3 7 1 14

Milk, dairy, infant formula Goodson et al. (2002) Kuo and Ding (2004) Kang and Kondo (2003) Maragou et al. (2006) Casajuana and Lacorte (2004) Biles et al. (1997a) ND3 113 43 15.2

<5 7613

<5 ND12

14 10

Yoshida et al. (2001) Brotons et al. (1995) ND3 44 21 <1.7

33.3

ND11

Maximum (ng/g or ng/mL) 192

Minimum (ng/g or ng/mL) <20

Number of samples

3 1 107

Other liquids Thomson and Grounds (2005) Braunrath et al. (2005) Inoue et al. (2003)10

Food group

Table 16.1

Bisphenol A

413

Table 16.2 Levels of bisphenol A detected in milk stored in polycarbonate plastic containers Food group

Number of samples

Minimum (ng/g)

Maximum (ng/g)

Country

4 2

0.99 <1.7

2.64 <5.1

Greece

Milk, dairy Casajuana and Lacorte (2004) Maragou et al. (2006)

Table 16.3 Levels of bisphenol A detected in foods and beverages in unknown/ no containers Food group

Number of samples

Minimum (ng/g or ng/mL)

Maximum (ng/g or ng/mL)

Country

No container Fish Kitano et al. (2003) Basheer et al. (2004) Imanaka et al. (2001b) Kumano et al. (2000)

61 6 25 80

ND1 13.3 ND2 <0.001

58 213.1 31 0.01

Japan Singapore Japan Japan

Milk, dairy Kang and Kondo (2003)

37

<1

<3

Japan

Milk, dairy Imanaka et al. (2001b)

13

ND2

9

Japan

Pasta, grains Imanaka et al. (2001b)

3

ND2

ND2

Japan

Fruits Imanaka et al. (2001b)

9

ND2

ND2

Japan

Vegetables Imanaka et al. (2001b)

11

ND2

ND2

Japan

Unknown container

1 2

LOQ for Kitano et al. (2003) is 5 ng/g. LOD for Imanaka et al. (2001b) is 0.4 ng/g.

and levels were found to be low (≤40 μg/kg). Migration was even lower (<5 μg/kg) for all other products tested under conditions that included boiling water, microwaving, and washing, leading the authors to conclude that, for everyday use of products currently on the market (e.g. nursing bottles and children’s tableware), only infinitesimally small levels of BPA will migrate.

414

16.3

Endocrine-disrupting chemicals in food

Bisphenol A in humans

16.3.1 Metabolism and pharmacokinetics of bisphenol A in humans Since humans efficiently convert BPA into rapidly excreted, toxicologically inactive forms, BPA does not accumulate after ingestion. In humans, BPA undergoes first pass metabolism in the liver and intestinal wall, where it is directly converted to one major metabolite, a glucuronide, as well as two minor metabolites, a sulfate and a glucuronide/sulfate diconjugate (Pritchett et al., 2002; Inoue et al., 2003). Essentially all of the BPA that is absorbed after oral exposure is metabolized in the liver and intestine and is then eliminated, so it never reaches the general circulation as free BPA. The glucuronide has no known biological or estrogenic activity (Matthews et al., 2001) and is generally eliminated from the body within one day (Völkel et al., 2002, 2005). The minor sulfate metabolite is also nonestrogenic (Shimizu et al., 2002). Studies conducted in humans have shown that, to the extent that free BPA would be present in human blood, approximately 95% is bound to albumin and unable to enter tissues where it could exert an estrogenic effect (Teeguarden et al., 2005). A few studies have examined whether BPA can cross the placenta in humans (Ikezuki et al., 2002; Schönfelder et al., 2002, 2004; Fujimaki et al., 2004; Shi et al., 2004; Engel et al., 2006). Most of these studies found either no or extremely low BPA levels in fetal samples. Two studies reported BPA levels several orders of magnitude higher than the other studies (Schönfelder et al., 2004; Shi et al., 2004), and their results have been challenged owing to their inconsistencies with extensive and well-documented urine data (Goodman et al., 2006). Some of the studies used an enzymelinked immunosorbent assay (ELISA), which cannot distinguish between the metabolized (conjugated) and the active (unconjugated) forms of BPA, and cross-reacts with other chemicals, as well. Thus, in these studies, the total levels of BPA reported in fetal samples are likely to over-represent levels of the active form of BPA.

16.3.2

Bisphenol A exposure estimates based on measurements in human samples Human exposure to BPA can be estimated based on measurements of BPA in serum, plasma, and urine. Since BPA is excreted rapidly in urine (within 24 hours and with a half-life of approximately 4 hours), measurements of total BPA excreted in urine over a defined time interval (such as 24 hours) are good indicators of total BPA daily exposure (Völkel et al., 2005). Although not as rigorous as sampling over a defined interval, spot urine samples can be used to determine daily intakes if one assumes that BPA exposure and excretion are relatively constant (Goodman et al., 2006). In general, urine biomonitoring results are consistent across studies, and indicate median total BPA urinary concentrations in the general

Bisphenol A

415

population and in pregnant women of about 1 to 2 ng/mL (Matsumoto et al., 2003; Tsukioka et al., 2004; Arakawa et al., 2004; Fujimaki et al., 2004; Calafat et al., 2005; Kawaguchi et al., 2005). Typical daily intake values inferred from BPA measured in 24-hour or spot urine samples are about 0.04–0.08 μg/kg/day (i.e., 20–60 ng/kg/day) for adults (Arakawa et al., 2004; Fujimaki et al., 2004; Tsukioka et al., 2004; Calafat et al., 2005; Fukata et al., 2006; Miyamoto and Kotake, 2006). Systemically available free BPA levels (i.e., not conjugated) are actually much lower than the total daily intake because of the elimination of BPA following first pass metabolism in the form of inactive, non-estrogenic metabolites (Völkel et al., 2002, 2005). The few studies that have measured free BPA in urine found that it was detected at very low levels, if at all, and represented only a small fraction of the total (i.e., conjugated + free) urinary BPA (Völkel et al., 2002; Kim et al., 2003; Fukata and Mori, 2004; Tsukioka et al., 2004; Ye et al., 2005). It should be noted that the free BPA detected in urine may have been overestimated because some hydrolysis of the BPA-glucuronide is likely to have occurred during sample collection, handling, storage, and/ or analysis (Ye et al., 2007).

16.3.3 Bisphenol A exposure estimates based on food intake Several studies have estimated BPA exposure from food based on levels measured in food (Goodson et al., 2002; Howe et al., 1998; Thomson and Grounds, 2005; EFSA, 2006). The most recent assessment was conducted by the European Food Safety Authority (EFSA). Based on conservative assumptions, the EFSA determined that infants were exposed to BPA at a range of 0.2 to 13 μg/kg/day. They estimated exposure in young children to be 5.3 μg/kg/day and exposure in adults to be 1.5 μg/kg/day. These values are based on migration from packaging into food; they do not account for microwave heating or potential migration of BPA into drinking water from the use of polycarbonate and epoxy resins in water pipes and water storage tanks. EFSA stated that, because conservative assumptions were made for all other sources of exposure and summed together, this still lead to a conservative estimate (EFSA, 2006). Human exposure estimates calculated from food intake are much higher than those based on measurements in human urine. For example, the EFSA exposure estimate for adults based on food intake (1.5 μg/kg/day) is 25 to 75 times higher than the daily intake values for adults inferred from BPA urine levels (about 0.04 to 0.08 μg/kg/day). This inconsistency is most likely due to the conservative assumptions for food intake. BPA levels in urine are a more reliable biomarker of human exposure. It is noteworthy that even based on food intake calculations, exposure estimates are less than 30% of the tolerable daily intake (TDI) established by EFSA in all population groups considered (EFSA, 2006), and actual measured exposures are at most a few percent of the TDI.

416

16.4

Endocrine-disrupting chemicals in food

Mechanisms of action of bisphenol A

BPA is generally thought to exert its effects by binding the ER and mimicking or disrupting the effects of the body’s natural estrogens, such as estradiol (Fig. 16.1). The affinity of the ligand for the hormone receptor is a measure of its ability to act as a hormone mimic and potentially disrupt hormonal action. BPA has a very weak affinity for the nuclear ERα, of about 1/104 to 1/105 that of human 17β-estradiol (Nagel et al., 1997; Bolger et al., 1998; Welshons et al., 1999; Blair et al., 2000), while its relative binding affinity for ERβ is approximately 35-fold less than that of 17β-estradiol (SeidlovaWuttke et al., 2005). It should be noted that the ability to bind the ER is not a direct indication of potential adverse impacts on health, reproduction, and development. Endocrine activity, such as effects on hormones or hormone receptors, is not a toxic end point in itself, but a mechanism of potential toxicity. Furthermore, there is a threshold level of ER binding that must be exceeded to produce effects. A developing research area is the investigation of potential ‘nontraditional’ modes of action for BPA and other estrogens (that is, effects attributable to processes other than nuclear ER binding). For example, some researchers are investigating BPA’s effects on the estrogen-dependent hippocampal synapse formation in rats (MacLusky et al., 2005). Another group is studying whether BPA can affect insulin resistance in pancreatic cells in mice (Alonso-Magdalena et al., 2006). This work is preliminary, however, and it remains to be seen whether their findings will be replicated in other studies. The overall significance of these non-traditional modes of action in the context of adverse impacts on health, reproduction, and development in humans at typical exposure levels has yet to be determined. If any of these modes of action were to result in adverse effects, one would have expected these effects to have been observed in animal studies. Yet, as discussed below, there are no consistent, reproducible adverse effects attributable to low doses of BPA. The lack of clear responses in vivo following these low doses suggests that non-traditional modes of action have minimal, if any, consequences.

16.5

Bisphenol A risks to human health

As discussed above, under normal circumstances humans are exposed to very low levels of BPA (∼0.04–0.08 μg/kg/day in adults based on urine biomonitoring). To determine whether BPA causes adverse health effects in humans, it is necessary to determine (1) whether any health effects can be caused by BPA, and, if so, what types, and (2) what doses are required to cause these effects. Specifically, the question is whether any effects occur at the very low human exposure levels. In a 103-week rat study conducted by the US National Toxicology Program (NTP, 1982), a lowest observed adverse effect level (LOAEL) of

Bisphenol A

417

50 mg/kg/day was established based on decreases in body weight. In Europe, the Scientific Committee for Food (SCF) based their safety assessment on a no observed adverse effect level (NOAEL) of 5 mg/kg/day in a threegeneration rat study for liver toxicity, and not for a hormonal effect (SCF, 2002). In these and other studies, hormonal effects (as reflected in uterotrophic assays) were only noted at much higher doses (Ashby and Odum, 2004). Thus, overall, it was thought that hormone activity would not be a result of low doses. In 1997, scientists from Fredrick vom Saal’s laboratory reported that young mice that had been exposed to low levels of BPA in utero (through oral dosing of pregnant mice) had increased prostate weights (Nagel et al., 1997). This led to the ‘low-dose hypothesis’ – that exceedingly low doses of BPA could lead to developmental and reproductive health effects that might not be observed at higher doses. This hypothesis led to well over 100 studies published on the topic of possible low-dose BPA effects (well below the previously recognized no-effect doses), including attempts to directly replicate vom Saal’s study design (Ashby et al., 1999; Cagen et al., 1999; Gupta, 2000; Tyl et al., 2008). With the exception of a minority view (vom Saal and Hughes, 2005; vom Saal et al., 2007), there is a clear international consensus about safe dose levels and the enormous gap between demonstrably toxic levels for any end point and human exposure levels (discussed in Section 16.6). The studies that claim to show adverse effects of very low exposures to BPA are rodent studies. Studies in humans have shown significant pharmacokinetic differences compared with rodents, which leads to lower internal doses of BPA in humans at similar oral intake levels (Teeguarden et al., 2005; Völkel et al., 2002). In humans, BPA is entirely excreted in urine, while in rodents there is extensive biliary elimination (Völkel et al., 2002, 2005). In rodents, bile is excreted into the intestines, where bacterial action can de-conjugate BPA, and the free BPA can then be re-absorbed (enterohepatic recirculation). Since this enterohepatic recirculation does not occur in humans, they have much less bioavailable BPA for a given oral intake compared to rodents (Teeguarden et al., 2005; Völkel et al., 2002, 2005). Elimination of BPA in humans (half-life of about 4 hours) is faster than in rodents relative to the usual physiological pace of elimination in these species (Völkel et al., 2005). In addition, humans may be less susceptible to BPA than rodents because circulating endogenous hormone levels are much higher in humans than in rodents during pregnancy. That is, if a rodent and a human received the same dose of BPA (normalized to body weight), the BPA concentration would pale in comparison with endogenous estrogens in humans, but not in rodents (Witorsch, 2002). Together, these factors indicate that a rodent BPA dose is equivalent to a much higher exposure in humans. In the sections that follow, we summarize the results of animal studies examining reproductive and developmental effects, neurological effects, and carcinogenic and mutagenic effects, followed by a summary of health effects reported in human studies.

418

Endocrine-disrupting chemicals in food

16.5.1 Reproductive and developmental effects As discussed above, BPA is estrogenic at very high doses. This fact and the vom Saal prostate result have led to the ‘low-dose hypothesis’ and extensive investigations of reproductive and developmental effects at doses much lower than the no-effect levels observed in previous animal studies. These studies have been reviewed extensively (Gray et al., 2004; vom Saal and Hughes, 2005; Goodman et al., 2006, 2008; vom Saal et al., 2007). Because humans are exposed to quite low doses of BPA (∼0.04–0.08 μg/kg/day in adults based on urine biomonitoring), animal studies of low doses (e.g., ≤5 mg/kg/day) are more relevant for determining risks to human health than studies with higher animal doses. In addition, studies with oral exposure routes are more applicable to determining human risks than studies using non-oral exposures both because this is how humans are exposed, and because this route of exposure results in much lower tissue levels of BPA owing to first-pass metabolism to inactive metabolites in the liver and intestine (Völkel et al., 2002, 2005). Based on over 130 reproductive and developmental studies (and several dozen supporting studies) conducted in animals with doses ≤5 mg/kg/day, and an examination of the whole body of evidence, endpoint by endpoint, there is an overwhelming preponderance of lack-of-effect findings compared with findings of effect, over a wide variety of reproductive and developmental toxicity end points (Gray et al., 2004; Goodman et al., 2006, 2008). The most robust BPA studies are the multigeneration studies that have been conducted with Sprague-Dawley rats (Ema et al., 2001; Tyl et al., 2002) and CD-1 Swiss mice (Tyl et al., 2008). These studies examined a wide variety of hormonally sensitive end points, a large number of animals, and oral doses ranging from ng-to-g/kg/day. Many endpoints were examined in these studies, including weights of reproductive organs, the morphology and cytology of these organs, sperm characteristics, the onset of puberty, and fertility. Certain effects were observed at high doses, and a NOAEL of 5 mg/kg/day has been established, although not based on reproductive effects, which occur only at the highest tested doses in these studies (EFSA, 2006). In these multigeneration studies, effects at lower doses were observed only sporadically (for example, in only one generation, with no dose–response pattern) and are not likely to be dose related. Among all BPA low-dose animal studies, there are some studies that report responses at some low doses, but no marked or consistently repeatable effects are observed for any health endpoint. Reported effects are not consistent between rats and mice, and there are no consistent patterns among dose groups and evaluation times. For example, although two mouse studies reported an increase in prostate weight (Nagel et al., 1997; Gupta, 2000), subsequent mouse studies failed to replicate these findings (e.g., Ashby et al., 1999; Cagen et al., 1999; Toyama et al., 2004; Tyl et al., 2008), and rat studies examining changes in prostate weight are mostly negative. Furthermore, the reported effects considered together seem to lack any

Bisphenol A

419

common pattern consistent with a hormonal mode of action. For example, Honma et al. (2002) reported an increase in anogenital distance of offspring when dams were exposed to BPA via injection, but one would predict an estrogenic compound to cause a decrease in anogenital distance. Importantly, no impairment of fertility, mating success, or development has been observed in multigenerational studies that examined comparable doses (Tyl et al., 2002, 2008; Ema et al., 2001). One would expect that if any of the reported histological changes had toxicological significance, they would be observed in multigenerational studies. In general, for no endpoint is there a pattern of an effect large enough to constitute an unambiguous response, consistent enough across studies to constitute a repeatable finding, and with a pattern across dose groups or time that supports the claim of modulation of the magnitude of effect with dose (including non-monotonic patterns). Instead, the pattern actually observed is one in which studies reporting effects are typically not corroborated, with several studies finding no effects at similar dose ranges using similar routes of exposure (Gray et al., 2004; Goodman et al., 2006, 2008). The question of repeatability and consistency among the studies on reproductive and developmental endpoints has been reviewed systematically by Gray et al. (2004) and Goodman et al. (2006, 2008). As noted in the scientific literature on reproductive and developmental effects by international and government regulatory bodies, the evidence on the low-dose hypothesis is not sufficiently credible to warrant restrictions on human exposures to BPA.

16.5.2 Neurological effects Natural hormones during fetal and childhood development have influences on brain development and subsequent behavior, including sexual behavior and sexual differences in such behavioral aspects as aggression and other social interactions. Because of BPA’s weak estrogenicity, some investigators have examined whether BPA exposures during fetal development can affect subsequent behavior of the pups (e.g., Schantz and Widholm, 2001; Kubo et al., 2003). There is no consistent evidence that low doses of BPA (e.g., ≤5 mg/kg/ day) cause adverse effects on behavioral endpoints, including (1) ontogeny of sensory/motor behaviors and reflexes (e.g., Ema et al., 2001; Palanza et al., 2002); (2) frequently measured gross behaviors of ambulation, rearing, grooming, and defecation (e.g., Farabollini et al., 1999; Ema et al., 2001; Kubo et al., 2001, 2003; Aloisi et al., 2002; Negishi et al., 2003, 2004); (3) complex cognitive behaviors (Morris water maze, Biel maze, active avoidance, passive avoidance, and operant behavior) (e.g., Ema et al., 2001; Kubo et al., 2001; Carr et al., 2003; Negishi et al., 2003, 2004); (4) complex emotional behaviors such as impulsivity, anxiety, and pain perception (e.g., Farabollini et al., 1999; Sashihara et al., 2001; Aloisi et al., 2002; Adriani et al., 2003); and (5) social behaviors such as play, aggression, maternal behavior, and sexual behavior

420

Endocrine-disrupting chemicals in food

(Farabollini et al., 1999, 2002; Dessi-Fulgheri et al., 2002; Palanza et al., 2002; Kawai et al., 2003; Kubo et al., 2003). Some studies suggest that low BPA doses cause masculinization of female behavior or feminization of male behavior on individual measures, but there has been no convincing or consistent pattern of effect that supports this conclusion (e.g., Farabollini et al., 1999, 2002; Kubo et al., 2001; DessiFulgheri et al., 2002; Adriani et al., 2003). While there is preliminary evidence that BPA may potentiate behavioral effects of drugs that are known to increase catecholamine levels in the synapse (Suzuki et al., 2003), there are conflicting reports on the effects of lower BPA doses on behavioral effects following pharmacologic challenge to drugs that increase levels of dopamine and norepinephrine in the synapse (Adriani et al., 2003; Negishi et al., 2004; Mizuo et al., 2004). One laboratory reported that perinatal exposure to low doses of BPA reduced the sexual difference in size of the locus coeruleus, which is a nucleus in the midbrain related to the central noradrenergic system, but noted no effects on the sexually dimorphic nucleus of the preoptic area in the hypothalamus (Kubo et al., 2001, 2003). This finding has not been confirmed in other studies. Taken together, these preliminary pharmacologic and morphologic effects suggest a potential developmental effect of low BPA doses on the structure and receptor function of the brain. It is unclear whether this represents an adverse effect, however, and these findings have not been confirmed in other experiments.

16.5.3 Immunological effects Because there are natural interactions between the immune system and estrogens, several investigators have studied whether BPA can alter immune function. The data are currently insufficient for drawing firm conclusions about possible immunological effects at relatively low doses (such as those experienced by humans). Several in vivo and ex vivo studies have been conducted on the effects on the BPA on the immune system (e.g., Byun et al., 2001, 2005; Pyo and Byun, 2001; Sawai et al., 2003; Sugita-Konishi et al., 2003; Yoshino et al., 2003, 2004). In mice, suppression of immunological responses after acute or shortterm in vivo exposure has generally been observed only at doses above 100 mg/kg, although there have been some exceptions (e.g., Byun et al., 2001, 2005; Pyo and Byun, 2001). Lee et al. (2003) found that 25 mg/kg BPA, administered via intraperitoneal injection during the antigen sensitization phase, produced an increase in serum IgE levels in KLH-primed mice, which was associated with an increase in IL-4 production. Other studies have reported effects on Th1 and/or Th2 responses and immune cell proliferation in mice at BPA doses ≤5 mg/kg/day (Youn et al., 2002; Sawai et al., 2003; Sugita-Konishi et al., 2003; Tian et al., 2003; Yamashita et al., 2003; Yoshino et al., 2003, 2004). It should be noted, however, that the quality and

Bisphenol A

421

utility of most of these studies for determining human risk is mixed. Many of the in vivo studies do not use estradiol or another estrogen receptor agonist to assess whether effects have a hormonal component (Youn et al., 2002; Sawai et al., 2003; Sugita-Konishi et al., 2003; Tian et al., 2003). In addition, some studies present methodological concerns (e.g., Youn et al., 2002; Sugita-Konishi et al., 2003; Yamashita et al., 2003), some administered only a single dose of BPA (e.g., Sawai et al., 2003; Yamashita et al., 2003), some are conducted in strains of mice that are bred to be susceptible to immunological effects (e.g., Tian et al., 2003), and some have only examined nonoral doses (e.g., Sugita-Konishi et al., 2003). In light of several multigenerational studies reporting no pathology indicative of immune system dysfunction at low doses (Ema et al., 2001; Tyl et al., 2002, 2008), the findings reported in these studies are unlikely to be indicative of potential human risks.

16.5.4 Carcinogenic and mutagenic effects Because BPA is estrogenic, several researchers have examined whether it could also be carcinogenic via a hormonal mode of action. BPA was first tested for carcinogenicity by the US NTP in a lifetime feeding bioassay using F344 rats and B6C3F1 mice. The study included doses up to 600 mg/ kg/day in the food of male mice, up to 1300 mg/kg/day in the food of female mice, up to 148 mg/kg/day in the food of male rats, and up to 135 mg/kg/day in the food of female rats (Haighton et al., 2002). All dosing began at a few weeks of age and continued for 2 years. The NTP report concluded that ‘Under the conditions of this bioassay, there was no convincing evidence that BPA was carcinogenic for F344 rats or B6C3F1 mice of either sex’ (NTP, 1982). Based on the available data, none of the prominent government and international organizations classify BPA as a carcinogen. The NTP, in its latest (11th) edition of the NTP Report on Carcinogens (NTP, 2005), does not list BPA as a carcinogen, nor does the International Agency for Research on Cancer (IARC), a division of the World Health Organization, nor the US EPA in its Integrated Risk Information System (IRIS; US EPA, 1993). The European Union also concludes that BPA does not have carcinogenic potential, based on the results of the NTP study, discussed above (EU, 2003). Recently, a panel of prominent scientists conducted a weight-of-evidence evaluation of the potential carcinogenicity of BPA, considering several factors: the available bioassays; BPA’s ability to cause cancer precursor effects of concern (especially hormone-mimetic effects); an assessment of BPA mutagenicity; and its metabolism in animals and humans (Haighton et al., 2002). The panel concluded ‘that BPA is not likely to be carcinogenic to humans,’ citing the NTP bioassay as providing ‘no substantive evidence to indicate that BPA is carcinogenic to rodents.’ They noted the lack of toxicities associated with potential carcinogenic action (such as pre-neoplastic lesions) in subchronic BPA toxicity studies, the lack of

422

Endocrine-disrupting chemicals in food

genotoxicity of BPA, the lack of oxidative metabolism that might produce reactive products, and the rapid glucuronidation and excretion of orally administered BPA. They concluded that ‘a large number of standard, validated, genotoxicity assays . . . have demonstrated a lack of mutagenic and clastogenic potential for BPA’ and noted that ‘the overall weight of evidence for mutagenicity/genotoxicity studies indicates that BPA is unlikely to be genotoxic to humans.’ Importantly, they noted that the lifetime carcinogenicity bioassays in rats and mice ‘have not provided indication of tumorigenic activity, proliferative cell growth, or otherwise adverse effects on endocrine-associated tissues. As a result, in vitro and in vivo assays that have reported estrogen-like activity cannot be interpreted to indicate carcinogenic potential’ (Haighton et al., 2002). Since the Haighton report was published, a few additional non-oral BPA studies have examined the potential precursor effects of carcinogenicity of BPA. One study found that fetal BPA exposure (2.5–1000 μg/kg/day via an osmotic pump implanted in the dam) could induce preneoplastic and neoplastic lesions in the mammary gland (Murray et al., 2006) and another found that perinatal BPA exposure (25 or 250 ng/kg/day, also via a pump) led to persistent alterations in mammary gland development (Munoz-deToro et al., 2005). A third study reported that low-dose (10 μg/kg) BPA exposures via Silastic capsule implants could lead to changes in the prostate epigenome, which the authors suggested could promote prostate cancer development with aging (Ho et al., 2006). Because none of these studies used oral exposure routes, their applicability to humans is questionable. Also, although these results are suggestive of possible carcinogenic potential, none of these studies noted the development of cancer, and BPA was not found to be carcinogenic in the earlier long-term bioassays or more recent multigeneration studies (Ema et al., 2001; Tyl et al., 2002, 2008). Thus, these recent findings, suggested to be possible cancer precursors, are not necessarily supportive with BPA being carcinogenic.

16.5.5 Health effects in human studies There have been few studies focused on the possible health effects of BPA in humans. Thus far, Sugiura-Ogasawara et al. (2005) examined miscarriages, Takeuchi and Tstsumi (2002) and Takeuchi et al. (2004) studied polycystic ovarian syndrome and obesity, Hiroi et al. (2004) evaluated endometrial hyperplasia and endometrial cancer, Yamada et al. (2002) examined chromosomal abnormalities in fetuses, and Yang et al. (2006) studied reproductive parameters (number of children and ratio of sons to daughters), self-reported endocrine-related symptoms, and chromosomal abnormalities. All these studies have major methodological shortcomings, which makes their results difficult to interpret. All studies were based on one BPA measurement per person. Because of the short half-life of BPA, the degree of

Bisphenol A

423

exposure over the putative period of causation of the observed effects is unknown. It is also doubtful that BPA measurements in all of these studies (except that of Yang et al., 2006, who used high-pressure liquid chromatography, HPLC) were accurate owing to the use of an ELISA. ELISA has been shown to cross-react with other substances, including BPAglucuronide and phytoestrogens, and is likely to overestimate BPA levels (Fukata et al., 2006). Most of the studies were conducted with a small number of subjects (ranging from 6 to 48 per treatment group), and confounders and effect modifiers were not always accounted for, as discussed below. Sugiura-Ogasawara et al. (2005) compared miscarriages in cases and controls, but it is unclear if the controls were trying to get pregnant. They also used parametric statistics, but BPA levels in cases were not normally distributed, and median BPA levels in cases and controls were equal. Takeuchi and Tstsumi (2002) and Takeuchi et al. (2004) examined BPA levels in healthy men and women and in women with polycystic ovarian syndrome (PCOS). It is unclear if subjects in these two studies were the same individuals. The authors reported BPA levels were correlated with women who were obese, with PCOS, and with several hormone levels, but it is unclear whether BPA levels were causal or a result of these conditions or hormones. Hiroi et al. (2004) examined the association of BPA with endometrial hyperplasia in premenopausal women and endometrial cancer in postmenopausal women. Women with cancer in the study were older and had lower gravidity, parity, and height and weight than other study subjects, again making results difficult to interpret. Yamada et al. (2002) reported lower BPA concentrations in the serum of women carrying normal fetuses than in those carrying fetuses with chromosomal abnormalities, but levels in the amniotic fluid from karyotypically normal and abnormal fetuses were not statistically different. Yang et al. (2006), who reported no association between BPA exposure and endocrine-related symptoms, chromosomal abnormalities, or reproductive parameters (number of children and ratio of sons to daughters), had more study subjects (172) and used a more reliable method for determining BPA levels. Because these authors also had only one urine measurement per study subject, however, it is not clear how meaningful their results are. Taken together, the human studies conducted thus far do not support or refute the hypothesis that BPA exposure leads to adverse reproductive or developmental health effects in humans.

16.6

Positions of government bodies on potential human health risks of bisphenol A

Several government bodies have undertaken reviews in response to claims of low-dose BPA toxicity (e.g., ≤5 mg/kg/day). (As explained above, humans are exposed to much lower levels of BPA, around 0.04–0.08 μg/kg/day in adults based on urine biomonitoring.) Thus far, no such review has

424

Endocrine-disrupting chemicals in food

concluded that adverse effects that are reported to occur below the traditionally defined no-effect level are sufficiently plausible to constitute a health threat. With the exception of Health Canada (which nonetheless concluded that current exposures are likely to be without harm), these reviewers concluded that the weight of evidence does not warrant precautionary government action. These government agencies include the European Commission Joint Research Centre (ECJRC), the European Scientific Committee on Food (SCF), the European Food Safety Authority (EFSA), the European Union Scientific Committee on Toxicity, Ecotoxicity and the Environment (CSTEE), the Japanese Ministry of Economy, Trade, and Industry (METI), the Japanese Ministry of Environment, the US Food and Drug Administration (US FDA), Health Canada, the US NTP, and NTP’s Center for the Evaluation of Risks to Human Reproduction (CERHR).

16.6.1 Europe The use of BPA in products intended to come in contact with food was first evaluated in Europe in 1986 by the SCF. The SCF calculated a TDI of 0.05 mg/kg, based on a NOAEL of 25 mg/kg/day for effects on body weight in a 90-day rat study, and an uncertainty factor of 500 (SCF, 2002). In 2002, SCF conducted a comprehensive review of toxicity and exposure information focused on food contact applications of BPA. They derived a temporary TDI of 0.01 mg/kg/day, based on an overall NOAEL of 5 mg/kg/day in a three-generation rat study (for which liver toxicity, not a hormonal effect, was the most sensitive response) and a 500-fold uncertainty factor (10 for interspecies differences, 10 for inter-individual variability, and 5 for uncertainties in the database on reproductive and developmental toxicity); they concluded that worst-case human exposures to BPA are well below the temporary TDI (SCF, 2002). In 2006, the EFSA reported that this NOAEL was still valid and further supported by a two-generation BPA study in mice (EFSA, 2006). In addition, they reported that the database had been strengthened to the point that the database uncertainty factor could be eliminated. They established a final TDI of 0.05 mg/kg/day, based on a 100fold uncertainty factor. They stated: The Panel considered that low-dose effects of BPA in rodents have not been demonstrated in a robust and reproducible way, such that they could be used as pivotal studies for risk assessment. Moreover, the species differences in toxicokinetics, whereby BPA as parent compound is less bioavailable in humans than in rodents, raise considerable doubts about the relevance of any low-dose observations in rodents for humans. The likely high sensitivity of the mouse to oestrogens raises further doubts about the value of that particular species as a model for risk assessment of BPA in humans (EFSA, 2006).

An EU Risk Assessment Report (RAR) in 2003 concluded that, based on a thorough review of the evidence, no risk management actions need be

Bisphenol A

425

proposed for polycarbonate plastic or epoxy resin products. They established an overall NOAEL of 50 mg/kg/day based on the multi-generation study of BPA in rats (EU, 2003). The EU Scientific Committee on Toxicity, Ecotoxicity and the Environment, an independent scientific committee, affirmed the key conclusions of the EU risk assessment in their detailed opinion (CSTEE, 2002). In 2008, the EU released an addendum to the 2003 RAR (EU, 2008). It stated: The worst case combined exposure would be for someone exposed via the environment near to a BPA production plant, and who is also exposed via food contact materials (oral exposure from canned food and canned beverages and from polycarbonate tableware and storage containers). Given the very large margins of safety, there are no concerns for repeated dose toxicity and reproductive toxicity . . . There is at present no need for further information and/or testing or for risk reduction measures beyond those which are being applied already.

16.6.2 Japan The Japanese National Institute of Advanced Industrial Science and Technology, which is a public research organization affiliated with the Japanese METI, concluded in 2005 (AIST, 2005) that, based on a thorough review of safety information, it was unlikely that humans, including infants and children, were at unacceptable risk from possible exposures to BPA in air, food, consumer products and water, and no adjustment to allowable exposures was deemed necessary for claimed low-dose effects (Miyamoto and Kotake, 2006). A comprehensive risk assessment conducted by this organization established a NOAEL of 50 mg/kg/day for reproductive and developmental toxicity based on the results of a multigeneration study in rats (Nakanishi et al., 2005). The report further concluded ‘An additional uncertainty due to the low-dose effects was not incorporated because the findings in the low-dose studies were not robust, while those in negative studies were consistent’. The Japanese Ministry of Environment (MOE) concluded in 2005 that no clear endocrine-disrupting effects were found at low doses and that no regulatory action is required to manage BPA risks (MOE, 2005). This was based on their own low-dose tests of BPA, including a comprehensive reproductive test in laboratory animals.

16.6.3 Canada In 2008, Health Canada concluded that BPA exposure to newborns and infants (through polycarbonate baby bottles when they are exposed to high temperatures and through the migration of BPA from cans into infant formula) is below levels that may pose a risk, but that the gap between exposure and potential effects is not large enough. As a result of

426

Endocrine-disrupting chemicals in food

this screening assessment, in April 2008, the government of Canada began action to consider banning the importation, sale, and advertising of polycarbonate baby bottles containing BPA, to develop stringent migration targets for BPA in infant formula cans, and to list BPA under Schedule 1 of the Canadian Environmental Protection Act (Health Canada, 2008).

16.6.4 United States The US FDA has approved polycarbonate resins, epoxy resins, and other resins made with BPA for use in contact with food, including can coatings, based on ‘a reasonable certainty that the materials would not be harmful under their intended conditions of use’ (Pauli/FDA, 2005). In response to a 2005 request from the California legislature, US FDA acknowledged that they were aware of several reports suggesting estrogenic activity at very low doses, but that ‘other reports appear to dispute any reason to expect harm at the low exposures that humans experience.’ US FDA concluded that ‘considering all the evidence, including measurements by FDA chemists of levels in canned foods and migrating from baby bottles, US FDA sees no reason at this time to ban or otherwise restrict the uses now in practice,’ and that ‘the current uses with food are safe’ (Pauli/FDA, 2005). A scientific panel organized in 2001 by the US NTP determined that ‘low-dose’ effects of BPA were not conclusively established as a general or reproducible finding (NTP, 2001). They conducted a peer review of the scientific evidence on low-dose reproductive and developmental effects for several chemicals. The subpanel focused specifically on BPA concluded: ‘There is credible evidence that low doses of BPA can cause effects on specific endpoints. However, due to the inability of other credible studies in several different laboratories to observe low dose effects of BPA, and the consistency of these negative studies, the subpanel [was] not persuaded that a low dose effect of BPA has been conclusively established as a general or reproducible finding’ (as discussed in Melnick et al., 2002).

In other words, because the majority of credible studies found no evidence for effect, these studies outweighed the few studies that found evidence for effect. The NTP’s CERHR released a report in 2007, detailing studies to date of reproductive and developmental effects of BPA, and applying a ranking of the utility of each study for determining risks to humans (CERHR, 2007). In April 2008, NTP issued a draft brief on BPA, in which it stated: The NTP concurs with the conclusion of the CERHR Expert Panel on Bisphenol A that there is some concern for neural and behavioral effects in fetuses, infants, and children at current human exposures. The NTP also has some concern for bisphenol A exposure in these populations based on effects in the prostate gland, mammary gland, and an earlier age for puberty in females. (NTP, 2008)

Bisphenol A

427

NTP had negligible concern that exposure of pregnant women to BPA would result in fetal/neonatal mortality, birth defects, reduced birth weight or growth of offspring, or that BPA would cause reproductive effects in non-occupationally exposed adults. NTP also had minimal concern for reproductive effects in workers exposed to higher levels in occupational settings.

16.7

Future trends

Given the large number of existing studies on reproductive and developmental effects of BPA, including several robust multigeneration studies, it is unlikely that more studies in this area will shed new light on this topic, although further biomonitoring data may help to refine current estimates of human exposure. Even if certain effects reported from single laboratories (e.g., histological and sperm effects) and non-traditional modes of action are investigated in future studies to determine if they constitute reproducible low-dose effects, their consequences on health must be considered in light of the multitude of studies that reported no effects. In view of the wide lack of acceptance of the low-dose hypothesis by many scientists and government agencies, one could hope that the scientific controversy about possible human health impacts of existing low exposures to BPA would be resolved. Even the precautionary approach employed in Europe did not lead the EFSA to express concern about current BPA uses. Not only did the EFSA conclude that the case for effects at low dose was unproven but that the case against such effects was sufficiently robust. Of course, it is always important to be open to new data and new interpretations, and periodic review of the state of the science on BPA is necessary. There continues to be public concern about BPA uses in consumer products, however, based, in our view, on attendance to isolated results of particular studies instead of evaluation and weighing of all the evidence – including the considerable body of countervailing evidence.

16.8

Sources of further information and advice

Many comprehensive BPA reviews are readily available on the internet for the interested reader. Some of the key BPA reviews written by government agencies can be found at: • European Union Risk Assessment Report and Addendum – http://ecb. jrc.it/DOCUMENTS/Existing-Chemicals/RISK_ASSESSMENT/ REPORT/bisphenolareport325.pdf http://ecb.jrc.it/documents/ExistingChemicals/RISK_ASSESSMENT/ADDENDUM/bisphenola_add_325. pdf

428

Endocrine-disrupting chemicals in food

• Opinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food on a request from the Commission related to 2,2-BIS(4HYDROXYPHENYL)PROPANE (Bisphenol A) – http://www.efsa. europa.eu/en/science/afc/afc_opinions/bisphenol_a.html • Center for the Evaluation of Risks to Human Reproduction(CERHR) Chemicals – Bisphenol A http://cerhr.niehs.nih.gov/chemicals/ bisphenol/bisphenol.html • Japanese National Institute of Advanced Industrial Science and Technology (AIST) – Bisphenol A Risk Assessment Document (English summary) – http://unit.aist.go.jp/crm/mainmenu/e_1-10.html • Health Canada – Draft Screening Assessment for Phenol, 4,4′(1-methylethylidene)bis-(80-05-7) – http://www.ec.gc.ca/substances/ese/ eng/challenge/batch2/batch2_80-05-7_en.pdf Links to additional comprehensive reviews by independent researchers and government agencies can be found at the following industry-sponsored website: • http://www.bisphenol-a.org, a website sponsored by the Polycarbonate/ BPA Global Group, a group organized regionally at the American Chemistry Council, Plastics Europe, and the Japan Chemical Industry Association. Lastly, numerous reviews have been published in the peer-reviewed, scientific literature. For example, Gray et al. (2004) and Goodman et al. (2006, 2008) have conducted comprehensive weight-of-the evidence reviews of the reproductive and developmental effects of low doses of BPA. As summarized above, these reviews provide extensive information on studies conducted through mid-2006 and demonstrate the overwhelming preponderance of lack-of-effect findings. Other recent BPA reviews include Tsai (2006) and Haighton et al. (2002). Details about the minority opinion (concerns about effects at low levels of BPA exposure) can be found in vom Saal and Hughes (2005) and vom Saal et al. (2007).

16.9

Acknowledgment

These authors have worked with the American Chemistry Council in the past on bisphenol A, but this chapter was written independently.

16.10

References

adriani, w; della seta, w; dessi-fulgheri, f; farabollini, f; laviola, g. 2003. ‘Altered profiles of spontaneous novelty seeking, impulsive behavior, and response to D-amphetamine in rats perinatally exposed to bisphenol A.’ Environ. Health Perspect. 111(4):395–401.

Bisphenol A

429

aist (japanese national institute of advanced industrial science and technology). 2005. ‘Human Risk Assessment and Ecological Risk Assessment for Bisphenol A.’ Research Center for Chemical Risk Management. November. aloisi, am; seta, dd; rendo, c; ceccarelli, i; scaramuzzino, a; farabollini, f. 2002. ‘Exposure to the estrogenic pollutant bisphenol A affects pain behavior induced by subcutaneous formalin injection in male and female rats.’ Brain Res. 937:1–7. alonso-magdalena, p; morimoto, s; ripoll, c; fuentes, e; nadal, a. 2006. ‘The estrogenic effect of bisphenol-A disrupts the pancreatic beta-cell function in vivo and induces insulin resistance.’ Environ. Health Perspect. 114(1):106–112. arakawa, c; fujimaki, k; yoshinaga, j; imai, h; serizawa, s; shiraishi, h. 2004. ‘Daily urinary excretion of bisphenol A.’ Environ. Health Prev. Med. 9:22–26. arvanitoyannis, is; bosnea, l. 2004. ‘Migration of substances from food packaging materials to foods.’ Crit. Rev. Food Sci. Nutr. 44:63–76. ashby, j; odum, j. 2004. ‘Gene expression changes in the immature rat uterus: effects of uterotrophic and sub-uterotrophic doses of bisphenol A.’ Toxicol. Sci. 82:458–467. ashby, j; tinwell, h; haseman, j. 1999. ‘Lack of effects for low dose levels of bisphenol A and diethylstilbestrol on the prostate gland of CF1 mice exposed in utero.’ Regul. Toxicol. Pharmacol. 30:156–166. basheer, c; lee, hk; tan, ks. 2004. ‘Endocrine disrupting alklphenols and bisphenolA in coastal waters and supermarket seafood from Singapore.’ Mar. Pollut. Bull. 48:1145–1167. biles, je; mcneal, tp; begley, th; hollifield, tc. 1997a. ‘Determination of bisphenolA in reusable polycarbonate food-contact plastics and migration to foodsimulating liquids.’ J. Agric. Food Chem. 45:3541–3544. biles, je; mcneal, tp; begley, th. 1997b. ‘Determination of bisphenol-A migrating from epoxy can coatings to infant formula liquid concentrates.’ J. Agric. Food Chem. 45:4697–4700. blair, rm; fang, h; branham, ws; hass, bs; dial, sl; moland, cl; tong, w; shi, l; perkins, r; sheehan, dm. 2000. ‘Estrogen receptor relative binding affinities of 188 natural and xenochemicals: Structural diversity of ligands.’ Toxicol. Sci. 54:138–153. bolger, r; wiese, te; ervin, k; nestich, s; checovich, w. 1998. ‘Rapid screening of environmental chemicals for estrogen receptor.’ Environ. Health Perspect. 106:551-557. Accessed on 4 May, 2006 at http://www.ehponline.org/members/1998/ 106p551-557bolger/bolger-full.html. braunrath, r; cichna, m. 2005. ‘Sample preparation including sol-gel immunoaffinity chromatography for determination of bisphenol A in canned beverages, fruits and vegetables.’ J. Chromatogr. 1062:189–198. braunrath, r; podlipna, d; padlesak, s; cichna-markl, m. 2005. ‘Determination of bisphenol A in canned foods by immunoaffinity chromatography, HPLC, and fluorescence detection.’ J. Agric. Food Chem. 53(23):8911–8917. brede, c; fjeldal, p; skjevrak, i; herikstad, h. 2003. ‘Increased migration levels of bisphenol A from polycarbonate baby bottles after dishwashing, boiling and brushing.’ Food Additives Contamin. 20(7):684–689. brenn-struckhofva, z; cichna-markl, m. 2006. ‘Determination of bisphenol A in wine by sol-gel immunoaffinity chromatography, HPLC and fluorescence detection.’ Food Additives Contamin. 23(11):1227–1235. brotons, ja; olea-serrano, mf; villalobos, m; pedraza, v; olea, n. 1995. ‘Xenoestrogens released from lacquer coatings in food cans.’ Environ. Health Perspect. 103(6):608–612. byun, j-a; heo, y; hyun, sm; chang, jh; kim, yo; pyo, m-y. 2001. ‘Immunomodulatory effects of bisphenol A, an endocrine disruptor, on macrophage activation and splenic cellularity in mice.’ Toxicologist 60(1):29.

430

Endocrine-disrupting chemicals in food

byun, ja; heo, y; kim, yo; pyo, my. 2005. ‘Bisphenol A-induced downregulation of murine macrophage activities in vitro and ex vivo.’ Environ. Toxicol. Pharmacol. 19(1):19–24. cagen, sz; waechter, jm; jr., dimond, ss; breslin, wj; butala, jh; jekat, fw; joiner, rl; shiotsuka, rn; veenstra, ge; harris, lr. 1999. ‘Normal reproductive organ development in CF-1 mice following prenatal exposure to bisphenol A.’ Toxicol. Sci. 50(1):36–44. calafat, am; kuklenyik, z; reidy, ja; caudill, sp; ekong, j; needham, l. 2005. ‘Urinary concentrations of bisphenol A and 4-nonylphenol in a human reference population.’ Environ. Health Perspect. 113:391–395. carr, rl; bertasi, fr; betancourt, am; bowers, sd; gandy, bs; ryan, pl; willard, st. 2003. ‘Effects of neonatal rat bisphenol A exposure on performance in the Morris water maze.’ J. Toxicol. Environ. Health Part A 66:2077–2088. casajuana, n; lacorte, s. 2004. ‘New methodology for the determination of phthalate esters, bisphenol A, bisphenol A diglycidyl ether, and nonylphenol in commercial whole milk samples.’ J. Agric. Food Chem. 52:3702–3707. cerhr (center for the evaluation of risks to human reproduction). 2007. ‘NTPCERHR Report on the Reproductive and Developmental Toxicity of Bisphenol A (Interim Draft).’ National Toxicology Program, US Dept. of Health and Human Services, April. cstee (scientific committee on toxicity, ecotoxicity and the environment). 2002. ‘Opinion on the results of the Risk Assessment of: Bisphenol A; Human Health Part.’ European Commission. 22 May. dessi-fulgheri, f; porrini, s; farabollini, f. 2002. ‘Effects of perinatal exposure to bisphenol A on play behaviour of female and male juvenile rats.’ Environ. Health Perspect. 110(3):403–407. efsa (european food safety association). 2006. ‘Opinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food on a request from the Commission related to 2,2-bis(4-hydroxyphenyl)propane (bisphenol A).’ The EFSA Journal 428. Accessed at http://www.efsa.europa.eu/ en/science/afc/afc_opinions.html. ema, m; fujii, s; furukawa, m; kiguchi, m; ikka, t; harazono, a. 2001. ‘A two-generation reproductive toxicity study of bisphenol A.’ Reprod. Toxicol. 15:505–523. engel, sm; levy, b; liu, z; kaplan, d; wolff, ms. 2006. ‘Xenobiotic phenols in early pregnancy amniotic fluid.’ Reprod. Toxicol. 21(1):110–112. eu. 2003. European Union Risk Assessment Report. Bisphenol A, CAS No: 80-05-7. Institute for Health and Consumer Protection, European Chemicals Bureau, European Commission Joint Research Centre, 3rd Priority List, Luxembourg: Office for Official Publications of the European Communities. eu. 2008. Updated European Risk Assessment Report for 4,4′-isopropylidenediphenol (Bisphenol-A) (CAS No. 80-05-7) (EINECS No. 201-245-8) (Final). Office for Official Publications of the European Communities (Luxembourg). Accessed on 16 July, 2008 at http://ecb.jrc.it/documents/Existing-Chemicals/ RISK_ASSESSMENT/ADDENDUM/bisphenola_add_325.pdf 392p. farabollini, f; porrini, s; dessi-fulgheri, f. 1999. ‘Perinatal exposure to the estrogenic pollutant bisphenol A affects behavior in male and female rats.’ Pharmacol. Biochem. Behav. 64(4):687–694. farabollini, f; porrini, s; seta, dd; bianchi, f; dessi-fulgheri, f. 2002. ‘Effects of perinatal exposure to bisphenol A on sociosexual behavior of female and male rats.’ Environ. Health Perspect. 110(Suppl. 3):409–413. fsa (food standards agency). 2001. ‘Survey of bisphenols in canned foods (Number 13/01)’. Accessed at http://www.food.gov.uk/science/surveillance/fsis2001/bisphenols. fujimaki, k; arakawa, c; yoshinaga, j; watanabe, c; serizawa, s; imai, h; shiraishi, h; mizumoto, y. 2004. ‘Estimation of intake level of bisphenol A in Japanese preg-

Bisphenol A

431

nant women based on measurement of urinary excretion level of the metabolite.’ Japan. J. Hyg. 59(4):403–408. fukata, h; mori, c. 2004. ‘Current situation of human exposure to BPA.’ In The 7th Annual Meeting of Japan Society of Endocrine Disrupters Research, 14–15 December. fukata, h; miyagawa, h; yamazaki, n; mori, c. 2006. ‘Comparison of ELISA and LC-MS based methodologies for the exposure assessment of bisphenol A.’ Toxicol. Mechan. Meth. 16(8):427–430. goodman, je; mcconnell, ee; sipes, ig; witorsch, rj; slayton, tm; yu, cj; lewis, as; rhomberg, lr. 2006. ‘Updated weight of the evidence evaluation of reproductive and developmental effects of low doses of bisphenol A.’ Crit. Rev. Toxicol. 36:387–457. goodman, je; witorsch, rj; mcconnell, ee; sipes, ig; slayton, tm; yu, cj; franz, am; rhomberg, lr. 2008. ‘Weight-of-evidence evaluation of reproductive and developmental effects of low doses of bisphenol A.’ Crit. Rev. Toxicol., doi 10.1080/10408440802157839. goodson, a; summerfield, w; cooper, i. 2002. ‘Survey of bisphenol A and bisphenol F in canned foods.’ Food Addit. Contam. 19:796–802. goodson, a; robin, h; summerfield, w; cooper, i. 2004. ‘Migration of bisphenol A from can coatings – effects of damage, storage conditions and heating.’ Food Addit. Contam. 21:1015–1026. gray, gm; cohen, jt; cunha, g; hughes, c; mcconnell, ee; rhomberg, l; sipes, ig; mattison, d. 2004. ‘Weight of the evidence evaluation of low-dose reproductive and developmental effects of bisphenol A.’ Human Ecol. Risk Assess. 10:875–921. gupta, c. 2000. ‘Reproductive malformation of the male offspring following maternal exposure to estrogenic chemicals.’ Proc. Soc. Exp. Biol. Med. 224:61–68. haighton, la; hlywka, jj; doull, j; kroes, r; lynch, bs; munro, ic. 2002. ‘An evaluation of the possible carcinogenicity of bisphenol A to humans.’ Regul. Toxicol. Pharmacol. 35:238–254. health canada. 2008. ‘Draft Screening Assessment for Phenol, 4,4′-(1methylethylidene)bis-(bisphenol A),’ Chemical Abstracts Service Registry Number 80-05-07. April. hiroi, h; tsutsumi, o; takeuchi, t; momoeda, m; ikezuki, y; okamura, a; yokota, h; taketani, y. 2004. ‘Differences in serum bisphenol A concentrations in premenopausal normal women and women with endometrial hyperplasia.’ Endocrine J. 51:595–600. ho, sm; tang, wy; belmonte de frausto, j; prins, gs. 2006. ‘Developmental exposure to estradiol and bisphenol A increases susceptibility to prostate carcinogenesis and epigenetically regulates phosphodiesterase type 4 variant 4.’ Cancer Res. 66:5624–5632. honma, s; suzuki, a; buchanan, dl; katsu, y; watanabe, h; iguchi, t. 2002. ‘Low dose effect of in utero exposure bisphenol A and diethylstilbestrol on female mouse reproduction.’ Reprod. Toxicol. 16:117–122. horie, m; yoshida, t; ishii, r; kobayashi, s; nakazawa, h. 1999. ‘Determination of bisphenol A in canned drinks by LC/MS.’ Bunseki Kagaku 48(6):579–587. howe, sr; borodinsky, l; lyon, rs. 1998. ‘Potential exposure to bisphenol A from food-contact use of epoxy coated cans.’ J. Coatings Technol. 15(3):370–375. ikezuki, y; tsutsumi, o; takai, y; kamei, y; taketani, y. 2002. ‘Determination of bisphenol A concentrations in human biological fluids reveals significant early prenatal exposure.’ Hum. Reprod. 17(11):2839–2841. imanaka, m; hino, s; kadota, m; utsugi, j. 2001a. Okayama Prefecture Inst. Health Environ. Res. Annual Rep. 25:64.

432

Endocrine-disrupting chemicals in food

imanaka, m; sasaki, k; nemoto, s; ueda, e; murakami, e; miyata, d; tongoai, y. 2001b. ‘Determination of bisphenol A in foods using GC/MS [Japanese].’ Shokuhin Eiseigaku Zasshi 42:71–78. inoue, h; yuki, g; yokota, h; kato, s. 2003. ‘Bisphenol A glucuronidation and absorption in rat intestine.’ Drug Metab. Dispos. 31(1):140–144. kang, jh; kondo, f. 2002. ‘Bisphenol A migration from cans containing coffee and caffeine.’ Food Addit. Contam. 19:886–890. kang, jh; kondo, f. 2003. ‘Determination of bisphenol A in milk and dairy products by high-performance liquid chromatography with fluorescence detection.’ J. Food Prot. 66(8):1439–1443. kang, jh; kito, k; kondo, f. 2003. ‘Factors influencing the migration of bisphenol A from cans.’ J. Food Prot. 66:1444–1447. kawaguchi, m; sakui, n; okanouchi, n; ito, r; asito, k; izumi, s; makino, t; nakazawa, h. 2005. ‘Stir bar sorptive extraction with in situ derivatization and thermal desorption-gas chromatography–mass spectrometry for measurement of phenolic xenoestrogens in human urine samples.’ J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 820:49–57. kawai, k; nozaki, t; nishikata, h; aou, s; takii, m; kubo, c. 2003. ‘Aggressive behavior and serum testosterone concentration during the maturation process of male mice: the effects of fetal exposure to bisphenol A.’ Environ. Health. Perspect. 111(2):175–178. kawamura, y; koyama, y; takeda, y; yamada, t. 1998. ‘Migration of bisphenol A from polycarbonate products.’ J. Food Hygiene 99:206–212. kim, p; lee, a; hwang, s. 2003. ‘The bisphenol A: modulator of pregnancy in rats.’ Kor. J. Env. Health Soc. 29(4):27–34. kitamura, s; suzuki, t; sanoh, s; kohta, r; jinno, n; sugihara, k; yoshihara, s; fujimoto, n; watanabe, h; ohta, s. 2005. ‘Comparative study of the endocrine-disrupting activity of bisphenol A and 19 related compounds.’ Toxicol. Sci. 84:249–259. kitano, r; shiratani, y; ohfuji, m; ogawa, t; takemae, m. 2003. ‘Survey of the bisphenol A concentration in fish and shellfish.’ Kyoto-fu Hoken Kankyo Kenkyusho Nenpo 48:93–96. kubo, k; arai, o; ogata, r; omura, m; hori, t; aou, s. 2001. ‘Exposure to bisphenol A during the fetal and suckling periods disrupts sexual differentiation of the locus coeruleus and of behavior in the rat.’ Neurosci. Lett. 304:73–76. kubo, k; arai, o; omura, m; watanate, r; ogata, r; aou, s. 2003. ‘Low dose effects of bisphenol A on sexual differentiation of the brain and behavior in rats.’ Neurosci. Res. 45(3):345–356. kumano, m; honda, t; yatsunami, m. 2000. ‘Organic tin compounds and bisphenol A in fish.’ Nagasaki Prefecture Inst. Publ. Health Environ. Sci. Ann. Rep. 46:86. kuo, h; ding, w. 2004. ‘Trace determination of bisphenol A and phytoestrogens in infant formula powders by gas chromatography–mass spectrometry.’ J. Chromatog. A 1027:67–74. lee, hj; chattopadhyay, s; gong, e; ahn, rs; lee, k. 2003. ‘Antiandrogenic effects of bisphenol A and nonylphenol on the function of androgen receptor.’ Toxicolog. Sci. 75:40–46. maclusky, nj; hajszan, t; leranth, c. 2005. ‘The environmental estrogen bisphenol A inhibits estradiol-induced hippocampal synaptogenesis.’ Environ. Health Perspect. 113:675–679. maragou, nc; lampi, en; thomaidis, ns; koupparis, ma. 2006. ‘Determination of bisphenol A in milk by solid phase extraction and liquid chromatography-mass spectrometry.’ J. Chromatogr. 1129:165–173. matsumoto, a; kunugita, n; kitagawa. k; isse, t; oyama, t; foureman, gl; morita, m; kawamoto, t. 2003. ‘Bisphenol A levels in human urine.’ Environ. Health Perspect. 111(1):101–104.

Bisphenol A

433

matthews, jb; twomey, k; zacharewski, tr. 2001. ‘In vitro and in vivo interactions of bisphenol A and its metabolite, bisphenol A glucuronide, with estrogen receptors a and b.’ Chem. Res. Toxicol. 14(2):149–157. melnick, r; lucier, g; wolfe, m; hall, r; stancel, g; prins, g; gallo, m; reuhl, k; ho, sm; brown, t; moore, j; leakey, j; haseman, j; kohn, m. 2002. ‘Summary of the National Toxicology Program’s report of the endocrine disruptors low-dose peer review.’ Environ. Health Perspect. 110:427–431. miyamoto, k; kotake, m. 2006. ‘Estimation of daily bisphenol A intake of Japanese individuals with emphasis on uncertainty and variability.’ Environ. Sci. 13:15–29. mizuo, k; narita, m; miyagawa, k; narita, m; okuno, e; suzuki, t. 2004. ‘Prenatal and neonatal exposure to bisphenol-A affects the morphine-induced rewarding effect and hyperlocomotion in mice.’ Neurosci. Lett. 356:95–98. moe (japanese ministry of environment). 2005. ‘MOE’s Perspectives on Endocrine Disrupting Effects of Substances.’ March. munguia-lopez, em; gonzalez-leon, a; peralta, e; vargas-requena, c; soto-valdez, h. 2005. ‘Migration of bisphenol A (BPA) from can coatings to jalapeno peppers and an acid food stimulant.’ J. Agric. Food Chem. 50:7299–7302. munoz-de-toro, m; markey, c; wadia, pr; luque, eh; rubin, bs; sonnenschein, c; soto, am. 2005. ‘Perinatal exposure to bisphenol A alters peripubertal mammary gland development in mice.’ Endocrinology 146(9):4138–4147. murray, tj; maffini, mv; ucci, aa; sonnenschein, c; soto, am. 2006. ‘Induction of mammary gland ductal hyperplasias and carcinoma in situ following fetal bisphenol A exposure.’ Reprod. Toxicol. doi:10.1016/j.reprotox.2006.10.002. nagel, sc; vom saal, fs; thayer, ka; dhar, mg; boechler, m; welshons, wv. 1997. ‘Relative binding affinity-serum modified access (RBA-SMA) assay predicts the relative in vivo bioactivity of the xenoestrogens bisphenol A and octylphenol.’ Environ. Health Perspect. 105(1):70–76. nakanishi, j; miyamoto, k; kawasaki, h. 2005. ‘Bisphenol A Risk Assessment Document (Summary).’ New Energy and Industrial Technology Development Organization; Research Center for Chemical Risk Management; National Institute of Advanced Industrial Science and Technology. AIST Risk Assessment Document Series No. 6. negishi, t; kawasaki, k; takatori, a; ishii, y; kyuwa, s; kuroda, y; yoshikawa, y. 2003. ‘Effects of perinatal exposure to bisphenol A on the behavior of offspring in F344 rats.’ Environ. Toxicol. Pharmacol. 14:99–108. negishi, t; kawasaki, k; suzaki, s; maeda, h; ishii, y; kyuwa, s; kuroda, y; yoshikawa, y. 2004. ‘Behavioral alterations in response to fear-provoking stimuli and tranylcypromine induced by perinatal exposure to bisphenol A and nonylphenol in male rats.’ Environ. Health. Perspect. 112(11):1159–1164. ntp (national toxicology program) 1982. ‘NTP Technical Report on the Carcinogenesis Bioassay of Bisphenol A (Cas No. 80-05-7) in F344 Rats and B6C3F Mice (Feed Study).’ US Public Health Service, National Institutes of Health. NTP-8035; NIH Publication No. 82-1771. ntp (national toxicology program). 2001. ‘Report of the Endocrine Disruptors Low Dose Peer Review.’ August. ntp (national toxicology program). 2005. ‘Eleventh Report on Carcinogens.’ Accessed on 8 April 2005 at http://ntp-server.niehs.nih.gov/ntp/roc/toc11.html, 6p. ntp (national toxicology program) 2008. ‘Draft NTP Brief on Bisphenol A [CAS NO. 80-05-7].’ Accessed on 17 July 2008 at http://cerhr.niehs.nih.gov/chemicals/ bisphenol/BPADraftBriefVF_04_14_08.pdf 69p. palanza, p; howdeshell, kl; parmigiani, s; vom saal, fs. 2002. ‘Exposure to a low dose of bisphenol A during fetal life or in adulthood alters maternal behavior in mice.’ Env. Health Perspect. 110(S3):415–422.

434

Endocrine-disrupting chemicals in food

pauli, gh/fda. 2005. Private communication to G. Aghazarian. pritchett, jj; kuester, rk; sipes, ig. 2002. ‘Metabolism of bisphenol A in primary cultured hepatocytes from mice, rats, and humans.’ Drug Metab. Dispos. 30(11):1180–1185. pyo, my; byun, j. 2001. ‘Effects of acute oral administration of bisphenol A on the immune function in mice.’ Yakhak Hoeji. 45(1):55–63. sashihara, k; ohgushi, a; ando, r; yamashita, t; takagi, t; nakanishi, t; yoshimatsu, t; furuse, m. 2001. ‘Effects of central administration of bisphenol A on behaviors and growth in chicks.’ J. Poult. Sci. 38(4):275–281. sawai, c; anderson, k; walser-kuntz, d. 2003. ‘Effect of bisphenol A on murine immune function: modulation of interferon-gamma, IgG2a, and disease symptoms in NZBxZW F2 mice.’ Environ. Health Perspect 111(16):1883–1887. scf (scientific committee on food). 2002. ‘Opinion of the Scientific Committee on Food on Bisphenol A.’ European Commission. 17 April. schantz, sl; widholm, jj. 2001. ‘Cognitive effects of endocrine-disrupting chemicals in animals.’ Environ. Health Perspect 109(12):1197–1206. schönfelder, g; wittfoht, w; hopp, h; talsness, ce; paul, m; chahoud, i. 2002. ‘Parent bisphenol A accumulation in the human maternal-fetal-placental unit.’ Environ. Health Perspect. 110(11):703–707. schönfelder, g; friedrich, k; paul, m; cahoud, i. 2004. ‘Developmental effects of prenatal exposure to bisphenol A on the uterus of rat offspring.’ Neoplasia 6(5):584–594. seidlova-wuttke, d; jarry, h; christoffel, j; rimoldi, g; wuttke, w. 2005. ‘Effects of bisphenol-A (BPA), dibutylphtalate (DBP), benzophenone-2 (BP2), procymidone (Proc), and linurone (lin) on fat tissue, a variety of hormones and metabolic parameters: a 3 month comparison with effects of estradiol (E2) in ovariectomized (ovx) rats.’ Toxicology 213(1–2):13–24. shi, j; yang, s; xiao, g; zheng, l; zhou, z. 2004. ‘Detection of serum bisphenol A level in general population.’ J. Environ. Occup. Med. 21(3):190–193. shimizu, m; ohta, k; matsumoto, y; fukuoka, m; ohno, y; ozawa, s. 2002. ‘Sulfation of bisphenol A abolished its estrogenicity based on proliferation and gene expression in human breast cancer MCF-7 cells.’ Toxicol. In Vitro 16:549–556. staples, ca; dorn, pb; klecka, gm; o’block, st; harris, lr. 1998. ‘A review of the environmental fate, effects, and exposures of bisphenol A.’ Chemosphere 36(10):2149–2173. sugita-konishi, ya; shimura, s; nishikawa, t; sunaga, f; naito, h; suzuki, y. 2003. ‘Effect of bisphenol A on non-specific immunodefenses against non-pathogenic Escherichia coli.’ Toxicol. Lett. 136(3):217–227. sugiura-ogasawara, m; ozaki, y; sonta, s; makino, t; suzumori, k. 2005. ‘Exposure to bisphenol A is associated with recurrent miscarriage.’ Hum. Reprod. 20(8):2325–2329. sun, h; xu, l-c; chen, j-f; song, l; wang, x-r. 2006. ‘Effect of bisphenol A, tetrachlorobisphenol A and pentachlorophenol on the transcriptional activities of androgen receptor-mediated reporter gene.’ Food Chem. Toxicol. 44:1916–1921. suzuki, t; mizuo, k; nakazawa, h; funae, y; fushiki, s; fukushima, s; shirai, t; narita, m. 2003. ‘Prenatal and neonatal exposure to bisphenol-A enhances the central dopamine D1 receptor-mediated action in mice: enhancement of the methamphetamine-induced abuse state.’ Neuroscience 117(3):639–644. takeuchi, t; tsutsumi, o. 2002. ‘Serum bisphenol A concentrations showed gender differences, possibly linked to androgen levels.’ Biochem. Biophys. Res. Comm. 291:76–78. takeuchi, t; tsutsumi, o; ikezuki, y; takai, y; taketani, y. 2004. ‘Positive relationship between androgen and the endocrine disruptor, bisphenol A, in normal women and women with ovarian dysfunction.’ Endocrine J. 51(2):165–169.

Bisphenol A

435

teeguarden, jg; waechter, jm; clewell, hj; covington, tr; barton, ha. 2005. ‘Evaluation of oral and intravenous route pharmacokinetics, plasma protein binding, and uterine tissue dose metrics of bisphenol A: A physiologically based pharmacokinetic approach.’ Toxicol. Sci. 85:823–838. thomson, bm; grounds, pr. 2005. ‘Bisphenol A in canned foods in New Zealand: an exposure assessment.’ Food Addit. Contam. 22:65–72. tian, x; takamoto, m; sugane, k. 2003. ‘Bisphenol A promotes IL-4 production by Th2 cells.’ Int. Arch. Allergy Immunol. 132(3):240–247. toyama, y; suzuki-toyota, f; maekawa, m; ito, c; toshimori, k. 2004. ‘Adverse effects of bisphenol A to spermiogenesis in mice and rats.’ Arch. Histol. Cytol. 67(4):373–381. tsai, w.-t. 2006. ‘Human health risk on environmental exposure to bisphenol-A: A review.’ J. Environ. Sci. Health, Part C: Environ. Carcinogen. Ecotoxicol. Rev. 24(2):225–255. tsukioka, t; terasawa, j; sata, s; hatayama, y; makino, t; nakazawa, h. 2004. ‘Development of analytical method for determining trace amounts of BPA in urine samples and estimation of exposure to BPA.’ J. Environ. Chem. 14(14):57–63. tyl, rw; myers, cb; thomas, bf; keimowitz, ar; brine, dr; veselica, mm; fail, pa; chang, ty; seely, jc; joiner, rl; butala, jh; dimond, ss; cagen, sz; shiotsuka, rn; stropp, gd; waechter, jm. 2002. ‘Three-generation reproductive toxicity study of dietary bisphenol A in CD Sprague-Dawley rats.’ Toxicol. Sci. 68:121–146. tyl, rw; myers, cb; marr, mc; sloan, cs; castillo, np; veselica, mm; seely, jc; dimond, ss; van miller, jp; shiotsuka, rn; beyer, d; hentges, sd; waechter, jm. 2008. ‘Twogeneration reproductive toxicity study of dietary bisphenol A (BPA) in CD-1® (Swiss) mice.’ Tox. Sci. 104(2):362–384. us epa. 1993. ‘Bisphenol A (CASRN 80-05-7).’ Integrated Risk Information System (IRIS). Accessed on 3 February 2006 at http://www.epa.gov/iris/subst/0356.htm. völkel, w; colnot, t; csanady, ga; filser, jg; dekant, w. 2002. ‘Metabolism and kinetics of bisphenol A in humans at low doses following oral administration.’ Chem. Res. Toxicol. 15(10):1281–1287. völkel, w; bittner, n; dekant, w. 2005. ‘Quantitation of Bisphenol A and Bisphenol A Glucuronide in Biological Samples by HLPC-MS/MS.’ American Society for Pharmacology and Experimental Therapeutics. August. vom saal, fs; hughes, c. 2005. ‘An extensive new literature concerning low-dose effects of bisphenol A shows the need for a new risk assessment.’ Environ. Health Perspect. doi:10.1289/ehp.7713. vom saal, fs; akingbemi, bt; belcher, sm; birnbaum, ls; crain, da; eriksen, m; farabollini, f; guillette, lj; hauser, r; heindel, jj; ho, sm; hunt, pa; iguchi, t; jobling, s; kanno, j; keri, ra; knudsen, ke; laufer, h; leblanc, ga; marcus, m; mclachlan, ja; myers, jp; nadal, a; newbold, rr; olea, n; prins, gs; richter, ca; rubin, bs; sonnenschein, c; soto, am; talsness, ce; vandenbergh, jg; vandenberg, ln; walser-kuntz, dr; watson, cs; welshons, wv; wetherill, y; zoeller, rt. 2007. ‘Chapel Hill Bisphenol A Expert Panel consensus statement: integration of mechanisms, effects in animals and potential to impact human health at current levels of exposure.’ Reprod. Toxicol. 24(2):131–138. welshons, wv; nagel, sc; thayer, ka; judy, bm; vom saal, fs. 1999. ‘Low-dose bioactivity of xenoestrogens in animals: fetal exposure to low dose methoxychlor and other xenoestrogens increases adult prostate size in mice.’ Toxicol. Ind. Health 15:12–25. witorsch, rj. 2002. ‘Low-dose in utero effects of xenoestrogens in mice and their relevance to humans: an analytical review of the literature.’ Food Chem. Toxicol. 40(7):905–912.

436

Endocrine-disrupting chemicals in food

wong, ko; leo, lw; seah, hl. 2005. ‘Dietary exposure assessment of infants to bisphenol A from the use of polycarbonate baby milk bottles.’ Food. Addit. Contam. 22:280–288. xu, l-c; sun, h; chen, j-f; bian, q; qian, j; song, l; wang, x-r. 2005. ‘Evaluation of androgen receptor transcriptional activities of bisphenol A, octylphenol and nonylphenol in vitro.’ Toxicology 216:197–203. yamada, h; furuta, i; kato, e; kataoka, s; usuki, y; kobashi, g; sata, f; kishi, r; fujimoto, s. 2002. ‘Maternal serum and amniotic fluid bisphenol A concentrations in the early second trimester.’ Reprod. Toxicol. 16:735–739. yamashita, u; sugiura, t; yoshida, y; kuroda, e. 2003. ‘Effect of endocrine disrupters on thymocytes in vitro.’ J UOEH 25:161–170. yang, m; kim, sy; chang, ss; lee, is; kawamoto, t. 2006. ‘Urinary concentrations of bisphenol A in relation to biomarkers of sensitivity and effect and endocrinerelated health effects.’ Environ. Mol. Mutagen. 47:571–578. ye, x; kuklenyik, z; needham, ll; calafat, am. 2005. ‘Quantification of urinary conjugates of bisphenol A, 2, 5-dichlorophenol, and 2-hydroxy-4-methoxybenzophenone in humans by online solid phase extraction-high performance liquid chromatography-tandem mass spectrometry.’ Anal. Bioanal. Chem. 383(4):638–644. ye, x; bishop, am; reidy, ja; needham, ll; calafat, am. 2007. ‘Temporal stability of the conjugated species of bisphenol A, parabens, and other environmental phenols in human urine.’ J. Expo. Sci. Environ. Epidemiol. 17(6):567–572. yoshida, t; horie, m; hoshino, y; nakazawa, h. 2001. ‘Determination of bisphenol A in canned vegetables and fruit by high performance liquid chromatography.’ Food Addit. Contam. 718:69–75. yoshino, s; yamaki, k; yanagisawa, r; takano, h; hayashi, h; mori, y. 2003. ‘Effects of bisphenol A on antigen-specific antibody production, proliferative responses of lymphoid cells, and TH1 and TH2 immune responses in mice.’ Br. J. Pharmacol. 138(7):1271–1276. yoshino, s; yamaki, k; li, x; sai, t; yanagisawa, r; takano, h; taneda, s; hayashi, h; mori, y. 2004. ‘Prenatal exposure to bisphenol A up-regulates immune responses, including T helper 1 and T helper 2 responses, in mice.’ Immunology 112:489–495. youn, jy; park, hy; lee, jw; jung, io; choi, kh; kim, k; cho, kh. 2002. ‘Evaluation of the immune response following exposure of mice to bisphenol A: induction of Th1 cytokine and prolactin by BPA exposure in the mouse spleen cells.’ Arch. Pharmacal Res. 25(6):946–953.

17 Phytoestrogens and phytosterols S. Hendrich, Iowa State University, USA

Abstract: This chapter compares the estrogenic and related toxic effects of isoflavones, coumestrol, lignans, zearalenone (a fungal estrogen) and phytosterols, considered to be the major naturally occurring potential estrogens in the human diet. Estrogenicity and toxic dose/responses are compared among these compounds and across sexes and species, and key experimental approaches in vitro and in vivo are critiqued. Dietary exposures to these various compounds are estimated, and compared with extrapolated toxic doses considered to be most relevant to humans. Risk management for naturally occurring dietary estrogens is discussed. Key words: phytoestrogens, isoflavone, coumestrol, isoflavone, daidzein, genistein, lignan, enterodiol, enterolactone, zearalenone, phytosterol, toxicity, estrogenicity.

17.1

Introduction: phytoestrogens and phytosterols in food and endocrine disruption

Recent concern about dietary components that may interfere with endogenous human hormone function has focused attention on phytoestrogens, plant components with estrogen-like structures and functions, i.e. estrogen mimics or xenoestrogens. These phytoestrogenic components include flavonoids, especially of the subclass isoflavones (Fig. 17.1), which are nearly exclusively found in soybeans, soy foods and soy protein-derived food ingredients. Coumestrol is another potent phytoestrogen found in fairly large amounts in legume sprouts (Fig. 17.1). The lignans secoisolariciresinol and matairesinol, and their respective gut microbial metabolites enterodiol and enterolactone, are phytoestrogens mainly found in flaxseed (Fig. 17.1). Zearalenone, a fungal metabolite from Fusarium spp., is a common mycotoxin associated with grains, and may therefore also be considered to be a phytoestrogen (Fig. 17.1). Phytosterols (Fig. 17.2) may alter aspects of sex steroid hormone function in diverse species, such as zebrafish (Brachidanio

438

Endocrine-disrupting chemicals in food O HO

CH2OH

HO O

CH2OH Enterolactone

Enterodiol

OH

OH

CH3

O

OH

H

HO O

O HO Zearalenone

Coumestrol O

HO

O

O

HO

HO O

R2 R1 Equol

O

O

HO

Isoflavone

R1

Daidzein Genistein Glycitein

H H H OH OCH3 H

HO

R2 N O OH

17-β-Estradiol OH

Tamoxifen

Fig. 17.1 Major dietary phytoestrogens and key phytoestrogen metabolites, compared with endogenous estrogen, 17β-estradiol and the anti-estrogenic drug, tamoxifen.

Phytoestrogens and phytosterols

439

O MeO

Glucos Glucos

MeO O

Fig. 17.2

Secoisolariciresinol diglucoside, major dietary precursor of lignans.

rerio; Nakari and Erkomaa, 2003) and voles (Microtus agrestis; Nieminen et al., 2003), and are therefore functionally definable as phytoestrogens as well. A well-explained risk/benefit assessment of these major phytoestrogens is the principal aim of this review. It is important to consider both potentially positive and negative impacts of the phytoestrogens on human health. On the positive side, most of the phytoestrogens have been shown to have various health-beneficial effects. For example, isoflavones may inhibit mammary carcinogenesis when administered early in the life cycle of rodents (Whitsett and Lamartiniere, 2006). Isoflavones may also be needed for the cholesterol-lowering effects of soy protein (Crouse et al., 1999), although this remains controversial (Sacks et al., 2006). Hamsters statistically clustered as high apparent absorbers of isoflavones (n = 11) experienced plasma cholesterol lowering after ingesting soy protein for 4 weeks, whereas low absorbers of isoflavones (n = 27) did not show significantly lower plasma cholesterol (Zhong et al., 2006). Lignans inhibited mammary carcinogenesis in rodents dosed with cancer-initiating compounds (Chen et al., 2003), and may also lower plasma cholesterol, as seen experimentally in rabbits (Prasad, 2005). In a rodent model, the zearalenone metabolite, α-zearalenol, lowered plasma cholesterol (Dai et al., 2004). Phytosterols are well established to lower human blood cholesterol contents (Ostlund, 2007) and are commonly used as food additives for that purpose in margarine-like spreads. But, on the negative side, are the possible sex hormonedisrupting effects of phytoestrogens and phytosterols fully understood, and have the ramifications of increasing use of these substances been explored sufficiently? I will examine the existing evidence to clarify these questions and propose courses of action to minimize any adverse effects on human health.

440

17.2

Endocrine-disrupting chemicals in food

Determining the adverse effects of phytoestrogens and phytosterols

Phytoestrogens are substances in plants that act like estrogens produced endogenously in animals. 17β-Estradiol (Fig. 17.1) is the major endogenous estrogen that binds to estrogen receptors and triggers extensive changes in gene expression that lead to sexual maturation in females, including increased expression of estrogen receptors. Estrogen equivalency may be defined as relative ability to displace 17β-estradiol from binding to cytosolic fractions that contain presumptive estrogen receptors (Mueller, 2002). The mouse uterine growth assay is a well-accepted standard for determination of estrogenicity, in which weanling mice (∼21 days old) are dosed with test substances for a few days and uterine weight measured (Diel et al., 2002). Estrogenicity may be defined also as the ability to stimulate estrogen synthesis (aromatase) activity; ability to displace endogenous estrogens from sex hormone binding globulin (SHBG); or selective estrogen receptor modulator (SERM) action that induces estrogen-like changes in gene expression in tissues. Defining estrogenic activity is complicated by the presence of at least two estrogen receptors (ER), α and β. ER-β has different patterns of tissue expression and relative binding affinity of various estrogens compared with ER-α (Kuiper et al., 1996). Estrogen-related receptors (ERR-α, -γ for example) have also been identified that modulate estrogen responses seemingly independent of estrogen binding, but ligands or modulators of ERRs cannot be ruled out at this early stage in understanding their actions (Horard and Vanacker, 2003). This review focuses on phytoestrogens that cause uterine enlargement and that modulate gene expression by ER-α and/or β.

17.2.1 Isoflavones All three soybean isoflavones, genistein, daidzein and glycitein (Fig. 17.1), were estrogenic in the mouse uterine growth assay, with genistein and genistin (genistein glucoside) being equally potent on a molar basis, and about fourfold more potent than daidzin (Farmakalidis et al., 1985) when given to immature CD-1 mice for 4 days. Genistein given as above (3 mg/kg/day for 4 days) was threefold less potent than the same dose of glycitein in weanling B6D2F1 mice (Song et al., 1999). In a rat uterine growth assay, when 20 diets varying in daidzein and genistein content, an estimated lowest observed uterotrophic dose of genistein was estimated as 40 mg/kg body weight (150 μmol/kg) was derived, and an advisory level of 325 mg genistein/kg background diet (or 1.2 mmol/kg) was proposed as a level below which the dietary isoflavone content would not interfere with uterotrophic assays of suspected environmental estrogens (Owens et al., 2003). Beyond this classical estrogenicity assay, numerous other studies have explored possible adverse effects of isoflavones in vitro and in vivo.

Phytoestrogens and phytosterols

441

In vitro studies Ovarian germ cell numbers from 18-day chicken embryos were significantly increased in culture by 0.4 and 4 μmol daidzein/L, but not by 0.04 μmol daidzein/L (Liu et al., 2006). In these cells, proliferating cell nuclear antigen (PCNA) labeling index as measured immunochemically was increased by 4 μmol daidzein, an effect blocked by tamoxifen (Fig. 17.1), supporting an estrogenic mechanism of action of daidzein. Genistein at 0.01–1 μmol/L significantly stimulated growth of MtT/E-2 rat pituitary tumor cell line (Fujimoto and Honda, 2003); 250 mg genistein/kg diet stimulated growth of these tumor cells inoculated into ovariectomized rats, but 25 mg genistein/ kg did not have that effect. Genistein (12.5 μmol/L) inhibited gap junction intercellular communication between mouse Leydig TM3 cells (cell models for testicular carcinogenesis) while not inhibiting cell growth. By the use of selective inhibitors, this effect was shown to be mediated by estrogen receptors and protein kinase C (Iwase et al., 2006). These studies indicate widely varying potency of isoflavones in their cell proliferative effects. But relevance of these studies to human isoflavone exposures and human cells is unclear. By yeast estrogen screen (YES) assay, equol, a major daidzein metabolite (Fig. 17.1, 7–17 μg/g manure) was shown to be the major contributor to estrogenicity from hog manure stored for several months, but runoff water from drainage tiles in a manure-treated farm field did not show the presence of equol. In this study, equol had 50% inhibitory concentrations of 5.3 μmol/L for goldfish sex steroid binding protein, 4.5 μmol/L for goldfish androgen receptor and 2.3 μmol/L for rainbow trout estrogen receptor (Burnison et al., 2003). This study suggests that manure spills deserve further attention for endocrine-disrupting effects of phytoestrogens, but that the use of manure as fertilizer does not pose an endocrine-disrupting risk from phytoestrogens.

In vivo studies Single high dose studies in 30 adult men (Busby et al., 2002) or 24 postmenopausal women (Bloedon et al., 2002) given up to 16 mg total isoflavones/kg body weight showed no significant adverse effects of isoflavones. A safety review by Munro et al. (2003) that summarized data from acute, subchronic and chronic studies in humans and animals concluded that human dietary contents of isoflavones posed no significant safety concerns. Interference of dietary isoflavones with human reproductive development may be worth further examination, but such studies remain to be done in primate or other animal models that may be more relevant to humans than are rodent models. However, in male ICR mice fed 1 mg genistein/mouse from days 1–6 after birth, at 12 weeks of age, no morphological changes in testes were found, but expression of androgen and estrogen receptor genes in testes was down-regulated (Adachi

442

Endocrine-disrupting chemicals in food

et al., 2004). Female Sprague-Dawley rats injected subcutaneously with four doses of 250 μg genistein on postnatal days 1 and 2 showed decreased anteroventral periventricular neurons expressing ERα, indicating defeminization of development of this part of the hypothalamus (Patisaul et al., 2006). Groups of six New Zealand white male rabbits given 0.1 mg daidzein per day (∼0.03 mg/kg) from ∼6 to 9 months of age showed 60% lower serum testosterone than controls. Organ bath studies of corpus carvenosum showed increased contractila response after daidzein or 17β-estradiol treatment, suggesting erectile impairment due to a human equivalent dose of ∼2 mg daidzein per person per day (Srilatha and Adaikan, 2004). In Wistar rats (30 per sex per treatment) fed 5, 50 or 500 mg genistein/kg body weight per day for 52 weeks, estrogenic effects including prostatic inflammation and ovarian atrophy were observed at 50 or 500 mg genistein/kg. This would suggest a no observable adverse effect level (NOAEL) for genistein was determined to be 5 mg/kg per day (McClain et al., 2006). Extrapolating this to humans, with a sevenfold lesser surface area than rats, the human NOAEL would be ∼0.7 mg/kg per day for long-term isoflavone intake, or ∼50 mg genistein for a 70 kg person, which is at the high end of genistein exposure in soy-consuming regions (e.g., Japan) and more than 50-fold greater than isoflavone intakes in countries such as the US for individuals who do not consume soy food or soy protein as a significant food ingredient (Lee, 2006). Both the long-term data in rats and short-term prenatal or postnatal exposure data in rodent models support the safety of current human isoflavone intakes. But further study regarding isoflavones and erectile dysfunction is suggested. Isoflavones have been evaluated in several interesting models for their effects in diets for livestock or pets. Mature male medaka fish (Oryzias latipes) fed a diet containing 0.22 mmol (∼60 mg) genistein and 0.15 mmol (∼40 mg) daidzein/kg showed increased hepatic vitellogenin after 28 days, compared with two other diets containing either non-detectable or >20-fold less isoflavone concentrations. Vitellogenin is a proposed biomarker of environmental estrogen exposure, but although it was increased by an isoflavone-containing diet, there was no impairment in fertility or fecundity in paired medaka fed this diet, compared with the two control diets (Inudo et al., 2004). Vasotocin neurons are involved with regulation of reproductive behavior in birds. In Japanese quail, eggs injected with 10 or 100 μg genistein three days after fertilization produced males that had normal copulatory behavior at 7 weeks of age, although 100 μg genistein decreased vasotocin neuronal area, whereas 1000 μg genistein injected into 3-day-old quail eggs significantly suppressed 7-week-old male copulatory behavior (Panzica et al., 2005). This isoflavone dose effect cannot be extrapolated to mammals, but the transfer of isoflavones from poultry diets to eggs deserves study. In groups of three beagles/sex dosed for 4 weeks or four beagles/sex

Phytoestrogens and phytosterols

443

dosed for 52 weeks, genistein at up to 500 mg/kg showed minor effects to increase uterine weight and decrease testis weight, but these effects were so minimal as to support a no-observed adverse effect level of >500 mg genistein/kg (McClain et al., 2005).

17.2.2 Coumestrol Doses of 10–270 mg/kg coumestrol fed to immature mice for five days significantly increased uterine weight, indicating estrogen receptor mediated effects of this compound (Galey et al., 1993). Effects of phytoestrogens may be mediated by mechanisms beyond ERs, depending on cell type. For example, coumestrol (10 nmol/L) caused rapid transient phosphorylation of extracellular-regulated kinases (ERKs) in the GH3/B6/F10 rat pituitary tumor cell line (cells that highly express ER-α) which was inhibited early by Ag 14, an epidermal growth factor receptor inhibitor and inhibited throughout the timecourse studied by PP2, a Src kinase inhibitor, whereas an ER inhibitor did not block coumestrol’s activation of ERK (Bulayeva and Watson, 2004). In this same cell model, 1 pmol/L coumestrol rapidly stimulated calcium-dependent prolactin secretion, suggesting another possibly highly sensitive mechanism for endocrine disruption (Wozniak et al., 2005). Coumestrol is often used as an experimental model phytoestrogen. Its effects in comparison with other phytoestrogens will be summarized in Section 17.2.6.

17.2.3 Lignans Feeding 5 or 10% flaxseed or an amount of secoisolariciresinol diglucoside (SDG, Fig. 17.2, that is metabolized to the phytoestrogenic lignans, enterodiol and enterolactone (Fig. 17.1)) equivalent to that found in a diet containing 5% flaxseed (100 mg SDG/kg diet) to rats during pregnancy caused significantly increased uterine weight in female offspring at 21 days of age, supporting estrogenic effects of lignans (Tou et al., 1998). Neither male nor female offspring of female Sprague-Dawley rats fed 10% flaxseed or the equivalent amount of SDG (177 mg/kg diet) from birth to 132 days of age differed from controls fed a semipurified AIN-93 diet in reproductive organ weights or other indicators of reproductive capability (anogenital distance, estrus cycling, histopathology of reproductive organs; Ward et al., 2001). Male offspring of rats fed 20 or 40% flaxseed during pregnancy did not have impaired spermatogenic anatomy, even when fed 20 or 40% flaxseed from birth to 70 days of age after maternal flaxseed exposure (Sprando et al., 2000). In addition to assessing effects of phytoestrogens on various aspects of reproductive function, genotoxicity testing of phytoestrogens may indicate

444

Endocrine-disrupting chemicals in food

if these compounds interfere with reproduction or other aspects of function by non-hormonal mechanisms. This is important in assessing their toxicity overall. Genotoxicity of lignans was assessed in V79 cells, and the compounds showed no such effects at up to 100 μmol/L (their limit of solubility) as reviewed by Niemeyer and Metzler (2002, Table 2). Lignans may be of particular concern for reproductive development if exposures occur in utero, but extrapolation to humans of this finding in rats will require additional studies.

17.2.4 Zearalenone Zearalenone (Fig. 17.1) at 25–45 mg/kg diet fed to 18-day-old B6C3F1 mice for six days caused significant uterine hypertrophy (Lemke et al., 2001). Diets naturally contaminated with 0.42 mg zearalenone/kg and 3.9 mg deoxynivalenol/kg caused increased uterine weight in pigs fed these diets from day 70 of age until slaughter, compared with controls. No abnormalities in histology were observed (Doll et al., 2004). The estrogenicity of zearalenone and related effects have been further examined in vitro and in vivo. When zearalenone and α- and β-zearalenol (hydroxyls substituted for the zearalenone ketone (Fig. 17.1), in α or β configuration, respectively) were compared in E-screen assays (MCF-7 cell proliferative response), α-zearalenol was 100–200 times more potent than zearalenone (EC50 of 0.3–1 nmol/L), which was 2–5 times more potent than was βzearalenone (Minervini et al., 2005). Similar relative potencies of these three compounds were noted across several different estrogen-screening cell systems (MCF-7 cells, Ishikawa cells, yeast expressing ER, as reviewed by Minervini et al., 2005). Zearalenone was compared for human and trout ER binding, using homogenates from MCF-7 human mammary tumor cells and rainbow trout liver. Zearalenone concentrations causing 50% inhibition of estrogen binding to ER in both tissues was ∼0.20 μmol/L, whereas concentration causing 50% increase in MCF-7 cell growth (EC50) was 0.4 nmol/L. Zearalenone at 0.5 μmol/L stimulated 50% increase in vitellogenin content of cultured trout hepatocytes (Olsen et al., 2005). Thus, in some respects zearalenone is similar across species, but its efficacy may also vary widely depending on cell type studied. Zearalenone’s main cytochrome P450-derived metabolites, α- and β-zearalenol (7.5–30 μmol/ L), inhibited progesterone and expression of its synthetic enzymes in follicle-stimulating hormone-treated porcine ovarian granulosa cells (Tiemann et al., 2003), showing another site and mechanism of action of this compound. High concentrations of zearalenone and α-zearalenol (125–250 μmol/L) effectively killed or damaged motility of boar sperm (Tsakmakidis et al., 2006). Atlantic croaker fish (Micropogonias undulates) sperm motility stimulation by progestin was blocked by 0.1 μmol zearalenone/L (Thomas and Doughty, 2004). Furthermore, in human hepatoma Hep G2 cells transfected with human pregnane X-receptor (PXR) gene,

Phytoestrogens and phytosterols

445

zearalenone (EC50,1.5 μmol/L) induced cytochrome P450 3A4, the prototypical gene product responsive to PXR, supporting zearalenone as a potent PXR activating ligand (Ding et al., 2006). This may have significance for food–drug interactions, in cases of zearalenone contamination of grains. In vivo endocrine disruption by zearalenone has been further investigated recently. Sprague-Dawley rats (n = 27/treatment) dosed from day 6–19 of gestation with 1, 2, 4 or 8 mg zearalenone/kg showed significant maternal and fetal weight suppression and lengthened anogenital distance relative to body weight (masculinization) in both male and female fetuses compared with controls. Zearalenone at 4 and 8 mg/kg increased early fetal deaths significantly, and 8 mg zearalenone/kg caused significantly increased fetal resorptions. A NOAEL for prenatal zearalenone exposure was determined to be <1 mg/kg (Collins et al., 2006). Within two days after dosing, 10-week-old Sprague-Dawley rats given one dose of 5 mg zearalenone/kg showed spermatic apoptosis by TUNEL analysis (Kim et al., 2003). The zearalenone doses found above to disrupt reproductive development and function are many orders of magnitude above mean human dietary intakes of 20–30 ng/kg per day and the provisional maximum tolerable daily intake (PMTDI) of 0.5 μg/kg body weight established by the Joint FAO/WHO Expert Commission on Food Additives (Zinedine et al., 2007). Although the above short-term studies may model the rare event of an undiscovered outbreak of zearalenone contamination, single or short term in utero highdose exposures to zearalenone would be quite unusual, and not relevant to typical human dietary exposures.

17.2.5 Phytosterols These compounds (Fig. 17.3) have not been evaluated in the mouse uterine growth assay unlike other phytoestrogens (Table 17.1), or in genotoxicity studies analogous to those used for other phytoestrogens (Table 17.2), to my knowledge. Phytosterols may substitute for cholesterol in mammalian cells (Xu et al., 2005), although small amounts of cholesterol seem to be required for cell growth. Phytosterols may undergo oxidation as does cholesterol, and phytosterol oxidation products have been investigated for their toxicity. C57Bl/6 mouse macrophage cell line was cultured with 200 μg/mL phytosterol or cholesterol oxides; both showed relatively similar significant increases in lactate dehydrogenase leakage and decreases in cell viability (Adcox et al., 2001); phytosterol oxides functioned similarly to cholesterol oxides. When phytosterol oxides were compared with corresponding cholesterol oxides in human cell lines U937 macrophage, CaCo-2 colon cancer, or Hep G2 hepatoma cells, the cholesterol oxide was generally cytotoxic at lower doses (30 μmol/L) than the phytosterol oxides (60–120 μmol/L). Alpha-tocopherol protected against cholesterol oxide toxicity but not against phytosterol oxides, which may have been partly due to the higher

446

Endocrine-disrupting chemicals in food β-Sitosterol

CH3 H3C CH3

CH3

CH3 H CH3

H

H HO Campesterol H3C CH3 CH3

CH3 H CH3

H

H HO CH3

Stigmasterol H3C CH3 CH3

CH3 H CH3

H

H HO

Fig. 17.3

Major dietary phytosterols.

concentration of phytosterol oxides used (Ryan et al., 2005). The concentrations of phytosterol oxides shown to exert cytotoxicity are seemingly well above those likely from human dietary intakes of phytosterols, given the relatively limited bioavailability of these compounds (13% absorption of campesterol, 4% absorption of β-sitosterol and stigmasterol observed in rats given radiolabeled compounds; Sanders et al., 2000). These general cytotoxic effects of phytosterols may not be related to their estrogenicity, but that remains to be determined. A two-generation reproductive toxicity study of a mixture of the main human dietary phytosterol esters in Wistar rats showed no adverse effects at up to 8.1% by weight of diet or >1.5 g/kg body weight (WaalkensBerendsen et al., 1999). However, reproductive toxicity studies in wildlife species showed various effects of phytosterols. Eyed grayling (Thymallus

Table 17.1 Uterine growth effects of phytoestrogens Uterotrophic dose μmol/kg body weight per day

Phytoestrogen

Species

Coumestrol Daidzin

Mouse Mouse

7 (5 days) 44 (4 days)

Genistein/genistin

Mouse

11 (4 days)

Glycitein Genistein and daidzein Lignans: secoisolariciresinol diglucoside Zearalenone

Mouse Rat

12 (4 days) 150 (3 days)

Rat Mouse

7 (to dams) 16 (6 days)

Reference Galey et al. (1993) Farmakalidis et al. (1985) Farmakalidis et al. (1985) Song et al. (1999) Owens et al. (2003) Tou et al. (1998) Lemke et al. (2001)

Note: Studies were performed in weanling animals with compounds administered orally (by diet or gavage) for 3–6 days, with the exception of Tou et al., 1998, in which lignan was fed to dams during pregnancy. Uterine weights were statistically compared with control animals fed semipurified diets or standard rodent chow.

Table 17.2 Comparative genotoxicity of phytoestrogens Phytoestrogen

Toxic dose

Genotoxicity test

Reference

Coumestrol

5 μmol/L

Hypoxanthine guanine phosphoribosyl transferase (HPRT) mutation test, V79 cells Single strand breaks, V79 cells Micronuclei, V79 cells Micronuclei, L5178Y cells Micronuclei, V79 cells

Kulling and Metzler (1997)

25 μmol/L Daidzein

10 μmol/L Not toxic at 100 μmol/L

Genistein

1.5 nmol/L

Glycitein Matairesinol, secoisolariciresinol, enterodiol, enterolactone Zearalenone

30 nmol/L 5 μmol/L

Micronuclei, L5178Y cells Comet assay Micronuclei, V79 cells

Not toxic at 150 μmol/L Not toxic at 100 μmol/L

Micronuclei, L5178 cells HPRT/micronuclei, V79 cells

Schmitt et al. (2001) Kulling et al. (2002) Boos and Stopper (2000) Virgilio et al. (2004) Stopper et al. (2005) Kulling et al. (1998)

5 μmol/L

Micronuclei, Vero monkey kidney cells Bovine oocyte sister chromatid exchange

Ouanes et al. (2003) Lioi et al. (2004)

0.1 μmol/L

448

Endocrine-disrupting chemicals in food

thymallus) eggs dosed with 1, 10 and 50 μg/L of a phytosterol mix containing 75% β-sitosterol showed significantly decreased time to hatching at all phytosterol doses, with no apparent disruption of thyroid hormone status or histopathology (Honkanen et al., 2005). Postmetamorphic female but not male clawed frogs (Xenopus laevis) exposed to 30 μg/L phytosterol mix (80% β-sitosterol) for 14 days showed significantly decreased plasma T3 (Koponen et al., 2004). But the relation between thyroid hormone status and estrogenicity was not examined. Zebrafish (Brachidanio rerio) exposed to phytosterol mixtures containing mostly β-sitosterol and at 10 or 20 μg/L for three generations showed increased plasma vitellogenin in both sexes in the first two generations, and in males in the third generation. Sex ratio was 70% male in the F1 generation but 20% male in the F2 generation (Nakari and Erkomaa, 2003). Groups of four to six field voles (Clethrionomys glareolus) of each sex fed 5 and 50 mg/kg phytosterol mix (89% βsitosterol) for 14 days showed increased serum testosterone and estradiol in males at 5 mg/kg (Nieminen et al., 2003). Twenty-three pairs of tundra voles (Microtus oeconomus) were fed 5 mg/kg phytosterol mix (89% βsitosterol) for ∼6 months, through reproduction, and their offspring continued on the phytosterol diet for ∼6 months. Reproduction was successful for 85% of phytosterol-fed voles but for only 60% of controls. Plasma and testicular testosterone was significantly less in phytosterol-fed male parents, but their offspring males had greater testicular testosterone than control males (Nieminen et al., 2004). At a minimum, the studies in voles suggest a need for human epidemiology regarding effects of use of phytosterolcontaining foods on sex hormonal status and reproductive function, because human phytosterol doses from such foods could exceed 5 mg/kg body weight.

17.2.6 Comparative studies Phytoestrogens have not been directly compared across all of the major classes for relative potency in the mouse uterine growth assay, nor have their minimally effective doses been determined, but their relative molar doses showing statistically significant mouse or rat uterine hypertrophy are summarized in Table 17.1, which may be a beginning for assessing the relative estrogenicity of these compounds. These compounds seem to have relatively similar uterotrophic doses within species but not across species, to the extent that this has been studied. But some differences in action have been noted among these compounds. CD-1 mice at 15 days of age were given four daily subcutaneous doses of 10 mg/kg of genistein or zearalenone, tested along with several other environmental estrogens. Both phytoestrogens accelerated puberty onset (vaginal opening) significantly compared with control, but zearalenone-treated mice remained anovulatory at eight weeks of age, unlike all other treatments including controls. No morphological abnormalities were observed in any treatment group (Nikaido et al., 2005).

Phytoestrogens and phytosterols

449

In a set of in vitro assays comparing phytoestrogens, MVLN and HGELN cell lines expressing ER-α were compared with MCF-7 cells (E-screen) for the ability of genistein, coumestrol and β-sitosterol to activate ER-α dependent luciferase reporter gene expression (MVLN or HGELN cells) or cell proliferation (MCF-7 cells). The three phytoestrogens had similar EC50 values for all three assays, ranging from 38 to 55 nmol/L, with coumestrol (Fig. 17.1) being slightly more potent, and β-sitosterol (Fig. 17.3) slightly less potent than the other phytoestrogens, and all 1000–10 000 fold less potent than 17-β-estradiol. Each phytoestrogen was also tested for competitive binding to purified human ER-α and -β. β-sitosterol and coumestrol were 1000-fold less potent in binding to ER-α than was 17-βestradiol, whereas genistein (EC50 of 35 μmol/L) was 10 000-fold less potent than estradiol. Genistein was most potent in binding to ER-β (EC50 of 2 μmol/L), 50% more potent than coumestrol and twice as potent as βsitosterol, and only ∼30-fold less potent than 17-β-estradiol (Gutendorf and Westendorf, 2001). ER-β mediated effects should be a focus of further studies of phytoestrogens. Comparative genotoxicity data, again, not in studies directly comparing all phytoestrogens, but across studies using relatively similar methodologies, indicate genotoxicity for coumestrol, zearalenone and genistein, and not for the other phytoestrogens, absent data on phytosterols (Table 17.2). Additional comparative studies are clearly needed, especially with nutritionally relevant and molar equivalent doses of the phytoestrogens. Effects of prenatal phytoestrogen exposures in human relevant models are one of the most pressing areas of concern.

17.3

Assessing dietary intake of phytoestrogens and phytosterols

Human intakes of phytoestrogens vary with world region, cultural and individual dietary preferences, and medical or preventive measures being adopted by individuals, especially for isoflavones, lignans and phytosterols. The task of determining human dietary phytoestrogen intake will always be uncertain because it is not possible to precisely determine human food intakes outside a laboratory setting, and foods naturally vary by genetic variety and environment in their chemical contents. But, combining assessment of food intake, as is done regularly in many countries by the best methods practically available, with phytoestrogen contents of foods from databases is useful. We can be fairly certain that the major dietary sources of phytoestrogens are known, and that appropriate analytical techniques are available to quantify these substances. Databases or references exist and are continuously being updated. For example the USDA/Iowa State University database on isoflavones and coumestrol is readily accessible (http://www.nal.usda.gov/fnic/foodcomp/Data/isoflav/isoflav.html). Milder

450

Endocrine-disrupting chemicals in food

Table 17.3

Estimated human dietary intakes of phytoestrogens

Phytoestrogen class

Daily intake (μg/kg1)

Isoflavones

<14

Isoflavones Isoflavones Coumestrol

290–1140 33–42 Negligible2

Coumestrol Lignans Lignans Phytosterols Zearalenone

0.4 ∼14 5–6 4070 0.020

Zearalenone Zearalenone

0.030 0.01

Country, conditions

Reference

US (in absence of soy food intake) Japan New Zealand US

Lee (2006)

New Zealand Netherlands New Zealand Netherlands Denmark, Canada, Norway US New Zealand

Lee (2006) Thomson et al. (2003) USDA/Iowa State University Isoflavone Database (2002) Thomson et al. (2003) Milder et al. (2006) Thomson et al. (2003) Normén et al. (2001) Zinedine et al. (2007) Zinedine et al. (2007) Thomson et al. (2003)

1

Based on estimated adult average body weight of 70 kg. Coumestrol content of 100 g portions of various foods: soy sprouts (38.6 mg), alfalfa sprouts mixed with clover sprouts (466 mg), clover sprouts (28 mg), alfalfa sprouts (0 or 4.7 mg), split peas (8 mg), pinto beans (3.6 mg). 2

et al. (2005, 2006) have quantified lignans in foods and estimated their intake in human populations in the Netherlands. Normen et al. (2001) quantified phytosterols and their intake in the Netherlands as well. Combining food analysis with various methods of assessing food intake (e.g., food frequency questionnaires, repeated food recalls) has yielded relatively recent estimates of human phytoestrogen intake (Table 17.3). In the case of zearalenone, the extent of surveillance and action levels for dietary contamination and the extent of fungal contamination outbreaks determine human intake. There are many technological, economic and political barriers to adequately assessing human intake of zearalenone, but the FAO/WHO is addressing these issues. There may be some concern about carryover (i.e. resulting in food residues) of phytoestrogens from animal feeds into meat, milk products or eggs. But this is likely to be minimal, given the relatively rapid metabolism of these compounds and relative lack of storage in body tissues. Goyaarts et al. (2007) measured carryover of zearalenone fed to pigs at 56 μg/kg and concluded that while carryover was measurable, it was minimal. Direct intake of plant food sources of phytoestrogens accounts for virtually all human phytoestrogen intake. The data to date indicate that phytosterols are the most predominant phytoestrogens in human diets by far. Dietary supplementation with isoflavones or lignans would increase intake of those phytoestrogens by 10–100 fold (10–100 mg/day), depending on the product consumed. Coumestrol

Phytoestrogens and phytosterols

451

intake could rival phytosterol intake only in the unusual circumstance of individuals consuming very large quantities of legume sprouts.

17.4

Assessing the risks and benefits of phytoestrogens and phytosterols in food

Applying toxic dose extrapolations from animal studies to human exposures is a field that has benefited recently from advances in statistical science. Components of variability have been compared across hundreds of studies and more statistically robust extrapolation factors across species, to study time courses, and to account for inter-individual variability may now be applied (Kodell and Gaylor, 1999). The NOAEL dose shown in animal studies may be used to derive a human maximal tolerated daily intake (MTDI) by dividing by 46, a statistically confident combination of interspecies and inter-individual variability factors (uncertainty factors), or by dividing by 184 (combined variability factors for interspecies, inter-individual and for extrapolating from a lowest observed adverse effect level (LOAEL) to an NOAEL; Kodell and Gaylor, 1999). These data are summarized in Table 17.4, but also included is the Joint Expert Commission on Food Additives’ recommendation for zearalenone (Zinedine et al., 2007). Note that only for isoflavone intake in Japan, and presumably in some other Asian countries is the phytoestrogen intake significantly greater than the MTDI (Table 17.4). Given the long history of soy food intake in these countries Table 17.4 Maximum tolerable daily intake (MTDI) levels for humans extrapolated from animal no observed adverse effect levels (NOAEL) of phytoestrogens; comparison with estimated intakes Phytoestrogen

NOAEL in animal studies

Isoflavones

5 mg/kg

Coumestrol

4 mg/kg (LOAEL)* 100 mg/kg (LOAEL)* 1.5 g/kg

Lignans Phytosterols Zearalenone

Human MTDI (μg/kg) 110 22* 540* 3300 0.5

MTDI/human estimated daily intake 7.9 (US) 0.096 (Japan) Not determined, intake negligible 38.6 1.23 16.7

Reference McClain et al. (2006) Galey et al. (1993) Tou et al. (1998) WaalkensBerendsen et al. (1999) Zinedine et al. (2007)

* Human MTDI is extrapolated from lowest observed adverse effect levels (LOAEL) in animal studies, an uncertainty factor of 184 is applied in those cases (Kodell and Gaylor, 1999).

452

Endocrine-disrupting chemicals in food

and the lack of studies linking isoflavone intake with reproductive function in these populations, the animal model used a basis for dose extrapolation is at this point questionable. The beneficial effects of phytoestrogens are not well established in humans, with the exception of the phytosterols (Ostlund, 2007), although this is an active field of investigation. It would seem at this point, that for most phytoestrogens, there is some latitude for increasing human intake without exceeding the very roughly estimated MTDIs proposed by this review.

17.5

Managing the risks of phytoestrogens and phytosterols in food

Because of the potential and diverse health benefits of phytoestrogens and the relatively lax regulatory climate in some countries such as the US, phytoestrogen supplementation or addition as food ingredients is ongoing. Continued monitoring of phytoestrogen content of human diets would be prudent, and special attention to further investigation of effects of phytoestrogens during pregnancy seems warranted based on animal data. Several studies have shown the ability of feed additive binders (e.g., clays, carbohydrates) to prevent zearalenone toxicity in livestock species (e.g., Chowdury et al., 2005; Abbès et al., 2006). For serious zearalenone outbreaks in grains, some of these binders deserve further study for inclusion in human foods. It may be a choice between starvation and consumption of contaminated grain; such mycotoxin binders might prove helpful. Public health will best be supported by adopting policies that put scientific data at the center of decisions regarding risks and benefits of phytoestrogens. Much more research ought to be supported, given promising health benefits and some significant aspects of risk to reproductive development.

17.6

Future trends

Studying phytoestrogens in vitro and in vivo in diet-relevant doses and taking into account the amounts of these compounds likely to be bioavailable is imperative, but quite challenging. Much work remains to be done to understand the bioavailability and metabolism of these compounds, which is a prerequisite for bioactivity. Inter-individual variability in metabolism of phytoestrogens has been a focus of research in my laboratory, and has been demonstrated to be a primary determinant of isoflavone efficacy in cholesterol lowering (Zhong et al., 2006). The identification of human gut microbes that cause beneficial metabolism of isoflavones, lignans and possibly phytosterols may be an interesting research frontier (Turner et al., 2003). Mammalian metabolism of phytoestrogens seemingly involves mostly glucuronidation of hydroxyl groups, as we have shown for isoflavones (Zhang

Phytoestrogens and phytosterols

453

et al., 2003). The glucuronide metabolites were tenfold or less toxic than parent isoflavones (Zhang et al., 1999). But such metabolites are not well studied for phytoestrogens, and there may be novel microbial metabolites to examine as well. Beyond this, a major trend in life sciences should be followed with respect to phytoestrogens, and that is to examine human and animal genetic variations in responsiveness. Identifying human or human-relevant genetic polymorphisms altering responsiveness to phytoestrogens, probably focusing on estrogen receptor and estrogen receptor-related pathways and identifying additional pathways of action of these compounds could yield important insights into their associated health benefits and risks.

17.7

Sources of further information and advice

Recent reviews provide additional information and perspectives on phytoestrogens (Miller et al., 2004; Krebs et al., 2004; Hendrich and Murphy, 2006) and zearalenone in particular (Zinedine et al., 2007). The above cited references and database on phytoestrogens in foods may also be useful. But risk and benefit assessments remain to be completed for these substances. From standpoints of chronic disease protection and early influences on reproductive development, phytoestrogens deserve further and more systematic study.

17.8

References

abbès s, ben salah-abbès j,ouanes z, zohra houas z,othman o, hassen bacha h, abdel-wahhab ma, oueslati r. 2006. Preventive role of phyllosilicate clay on the immunological and biochemical toxicity of zearalenone in Balb/c mice. Int Immunopharmacol 6: 1251–8. adachi t, ono y, koh, k-b, takashima k, tainaka h, matsuno y, nakagawa s, todaka e, sakurai k, fukata h, iguchit, komiyama m, mori c. 2004. Long-term alteration of gene expression without morphological change in testis after neonatal exposure to genistein in mice: toxicogenomic analysis using cDNA microarray. Fd Chem Toxicol 42: 445–52. adcox c, boyd l, oerhl l, allen j, fenner g. 2001. Comparative effects of phytosterol oxides and cholesterol oxides in cultured macrophage-derived cell lines. J Ag Fd Chem 49: 2090–5. bloedon lt, jeffcoat ar, lopaczynski w, schell mj, black tm, dix kj, thomas bf, albright c, busby mg, crowell ja, zeisel sh. 2002. Safety and pharmacokinetics of purified soy isoflavones: single-dose administration to postmenopausal women. Am J Clin Nutr 76: 1126–37. boos g, stopper h. 2000. Genotoxicity of several clinically used topoisomerase II inhibitors. Toxicol Lett 116: 7–16. bulayeva nn, watson cs. 2004. Xenoestrogen-induced ERK-1 and ERK-2 activation via multiple membrane-initiated signaling pathways. Environ Health Perspect 112: 1481–7.

454

Endocrine-disrupting chemicals in food

burnison bk, hartmann a, lister a, servos mr, ternes t, van der kraak g. 2003. A toxicity identification evaluation approach to studying estrogenic substances in hog manure and agricultural runoff. Environ Toxicol Chem 10: 2243–50. busby mg, jeffcoat ar, bloedon lt, koch ma, black t, dix kj, heizer wd, thomas bf, hill jm, crowell ja, zeisel sh. 2002. Clinical characteristics and pharmacokinetics of purified soy isoflavones: single-dose administration to healthy men. Am J Clin Nutr 75: 126–36. chen j, tan kp, ward we, thompson lu. 2003. Exposure to flaxseed or its purified lignan during suckling inhibits chemically induced rat mammary tumorigenesis. Exp Biol Med 228: 951–8. chowdhury sr, smith tk, boermans hj, woodward b. 2005. Effects of feed-borne Fusarium mycotoxins on hematology and immunology of laying hens. Poultry Sci 84: 1841–50. collins tfx, sprando rl, black tn, nicholas olejnik n, eppley rm, alam hz, rorie j, ruggles di. 2006. Effects of zearalenone on in utero development in rats. Fd Chem Toxicol 44: 1455–65. crouse jr iii, morgan t, terry jg, ellis j, vitolins m, burke gl. 1999. A randomized trial comparing the effect of casein with that of soy protein containing varying amounts of isoflavones on plasma concentrations of lipids and lipoproteins. Arch Intern Med 159: 2070–6. dai s, duan j, lu y, zhang y, cheng j, ren j, zhao x, wu y, yu y, zuo p, wu y, ge q. 2004. Phytoestrogen α-zearalenol inhibits atherogenesis and improves lipid profile in ovariectomized cholesterol-fed rabbits. Endocrine 25: 121–9. diel p, schmidt s, vollmer g. 2002. In vivo test sytems for the quantitative and qualitative analysis of the biological activity of phytoestrogens. J Chromatog B 777: 191–202. ding x, lichti k, staudinger j. 2006. The mycoestrogen zearalenone induces CYP3A through activation of the pregnane X receptor. Tox Sci 91: 448–55. döll s, dänicke s, schnurrbusch u. 2004. The effect of increasing concentrations of Fusarium toxins in piglet diets on histological parameters of the uterus and vagina. Arch Anim Nutr 58: 413–7. farmakalidis e, hathcock jn, murphy pa. 1985. Oestrogenic potency of genistin and daidzin in mice. Food Chem Toxicol 23: 741–5. fujimoto n, honda h. 2003. Effects of environmental estrogenic compounds on growth of a transplanted estrogen responsive pituitary tumor cell line in rats. Fd Chem Toxicol 41: 1711–17. galey fd, mendez le, whitehead we, holstege dm, plumlee kh, johnson b. 1993. Estrogenic activity in forages: diagnostic use of the classical mouse uterine bioassay. J Vet Diagn Invest 5: 603–8. goyarts t, dänicke s, valenta h, ueberschär kh. 2007. Carry-over of Fusarium toxins (deoxynivalenol and zearalenone) from naturally contaminated wheat to pigs. Food Addit Contam 24: 369–80. gutendorf b, westendorf j. 2001. Comparison of an array of in vitro assays for the assessment of the estrogenic potential of natural and synthetic estrogens, phytoestrogens and xenoestrogens. Toxicology 166: 79–89. hendrich s, murphy pa. 2006. Isoflavones: source and metabolism. In: Handbook of Nutraceuticals and Functional Foods, 2nd edition, ed. Wildman, R., CRC Press LLC, Boca Raton, FL, p. 23–54. honkanen jo, kostamo a, kukkone jvk. 2005. Toxicity of a phytosterol mixture to grayling (Thymallus thymallus) during early developmental stages. Arch Environ Contam 48: 391–6. horard b, vanacker jm. 2003. Estrogen receptor-related receptors: orphan receptors desperately seeking a ligand. J Mol Endocrin 31: 349–57.

Phytoestrogens and phytosterols

455

inudo m, ishibashi h, matsumura n, matsuoka m, mori t, taniyama s, kadokami k, koga m, shinohara r, hutchinson th, iguchi t, arizono k. 2004. Effect of estrogenic activity, and phytoestrogen and organochlorine pesticide contents in an experimental fish diet on reproduction and hepatic vitellogenin production in medaka (Oryzias latipes). Comp Med 54: 673–80. iwase y, fukata h, mori c. 2006. Estrogenic compounds inhibit gap junctional intercellular communication in mouse Leydig TM3 cells. Toxicol Appl Pharmacol 212: 237–46. kim i-h, son h-y, cho s-w, ha c-s, kang b-h. 2003. Zearalenone induces male germ cell apoptosis in rats. Toxicol Lett 138: 185–93. kodell rl, gaylor dw. 1999. Combining uncertainty factors in deriving human exposure levels of noncarcinogenic toxicants. Ann NY Acad Sci, 188–95. koponen ps, nieminen p, mustonen a-m, kukhonen jvk. 2004. Post metamorphic Xenopus Laevis shows decreased plasma triiodothyronine concentrations and phosphorylase activity due to subacute phytosterol exposure. Chemosphere 57: 1683–9. krebs ee, ensrud ke, macdonald r, wilt tj. 2004. Phytoestrogens for treatment of menopausal symptoms: a systematic review. Obstet Gynecol 104: 824–36. kuiper ggjm, enmark e, pelto-huikko m, nilsson s, gustafsson j-å. 1996. Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 93: 5925–30. kulling se, metzler m. 1997. Induction of micronuclei, DNA strand breaks and HPRT mutations in cultured Chinese hamster V79 cells by the phytoestrogen coumoestrol. Food Chem Toxicol 35: 605–13. kulling s, jacobs e, pfeiffer e, metzler m. 1998. Studies on the genotoxicity of the mammalian lignans enterolactone and enterodiol and their metabolic precursors at various endpoints in vitro. Mut Res 416: 115–24. kulling se, lehmann l, metzler m. 2002. Oxidative metabolism and genotoxic potential of major isoflavone phytoestrogens. J Chromatogr B 777: 211–18. lee n. 2006. Phytoestrogens as bioactive ingredients in functional foods: Canadian regulatory update. J AOAC Int 89: 1135–7. lemke sl, mayura k, reeves wr, wang n, fickey c, phillips td. 2001. Investigation of organophilic montmorillonite clay inclusion in zearalenone-contaminated diets using the mouse uterine weight bioassay. J Toxicol Environ Health A 62: 243–58. lioi mb, santoro a, barbieri r, salzano s, ursini mv. 2004. Ochratoxin A and zearalenone: a comparative study on genotoxic effects and cell death induced in bovine lymphocytes. Mut Res 557: 19–27. liu h, zhang c, zeng w. 2006. Estrogenic and antioxidant effects of a phytoestrogen daidzein on ovarian germ cells in embryonic chickens. Domest An Endocrinol 31: 258–68. mcclain rm, wolz e, davidovich a, pfannkuch f, bausch j. 2005. Subchronic and chronic safety studies with genistein in dogs. Fd Chem Toxicol 43: 1461–82. mcclain rm, wolz e, davidovich a, pfannkuch f, edwards ja, bausch j. 2006. Acute, subchronic and chronic safety studies with genistein in rats. Fd Chem Toxicol 44: 56–80. milder iej, arts icw, van de putte b, venema dp, hollman pch. 2005. Lignan contents of Dutch plant foods: a database including lariciresinol, pinoresinol, secoisolariciresinol and matairesinol. Br J Nutr 93: 393–402. milder iej, feskens ejm, arts icw, bas bueno-de-mesquita h, hollman pch, kromhout d. 2006. Intakes of 4 dietary lignans and cause-specific and all-cause mortality in the Zutphen Elderly Study. Am J Clin Nutr 84: 400–5.

456

Endocrine-disrupting chemicals in food

miller kp, borgeest c, greenfeld c, tomic d, flaws ja. 2004. In utero effects of chemicals on reproductive tissues in females. Toxicol Appl Pharmacol 198: 111–31. minervini f, giannoccaro a, cavallini a, visconti a. 2005. Investigations on cellular proliferation induced by zearalenone and its derivatives in relation to the estrogenic parameters. Toxicol Lett 159: 272–83. mueller so. 2002. Overview of in vitro tools to assess the estrogenic and antiestrogenic activity of phytoestrogens. J Chromatog B 777: 155–65. munro ic, harwood m, hlywka jj, stephen am, doull j, flamm wg, adlercreutz h. 2003. Soy isoflavones: a safety review. Nutr Rev 61: 1–33. nakari t, erkomaa k. 2003. Effects of phytosterols on zebrafish reproduction in multigeneration test. Environ Poll 123: 267–73. niemeyer hb, metzler m. 2002. Oxidative metabolites and genotoxic potential of mammalian and plant lignans in vitro. J Chromatog B 777: 321–7. nieminen p, mustonen a-m, lindstrom-seppa p, karkkainen v, mussalo-rauhamaa h, kukkonen jvk. 2003. Phytosterols affect endocrinology and metabolism of the field vole (Microtus agrestis). Exp Biol Med 228: 188–93. nieminen p, mustonen a-m, paivalainen p, kukkonen jvk. 2004. Reproduction of the tundra vole (Microtus oeconomus) with dietary phytosterol supplement. Fd Chem Toxicol 42: 945–51. nikaido y, danbara n, tsujita-kyutoku m, yuri t, uehara n, tsubura a. 2005. Effects of prepubertal exposure to xenoestrogen on development of estrogen target organs in female CD-1 mice. In Vivo 19: 487–94. normén al, brants ham, voorrips le, andersson ha, van den brandt pa, goldbohm ra. 2001. Plant sterol intakes and colorectal cancer risk in the Netherlands Cohort Study on Diet and Cancer. Am J Clin Nutr 74: 141–8. olsen cm, meussen-elholm etm, hongslo jk, stenersen j, tollefsen k-e. 2005. Estrogenic effects of environmental chemicals: an interspecies comparison. Comp Biochem Physiol C 141: 267–74. ostlund re jr. 2007. Phytosterols, cholesterol absorption and healthy diets. Lipids 42: 41–5. ouanes z, abid s, ayed i, anane r, mobio t, creppy ee, bacha h. 2003. Induction of micronuclei by zearalenone in Vero monkey kidney cells and in bone marrow cells of mice: protective effect of Vitamin E. Mut Res 538: 63–70. owens w, ashby j, odum j, onyon l. 2003. The OECD program to validate the rat uterotrophic bioassay. Phase 2: dietary phytoestrogen analyses. Environ Health Perspect 111: 1559–67. panzica g, mura e, pessatti m, viglietti-panzica c. 2005. Early embryonic administration of xenoestrogens alters vasotocin system and male sexual behavior of the Japanese quail. Domest Anim Endocrinol 29: 436–45. patisaul hb, fortino ae, polston ek. 2006. Neonatal genistein or bisphenol-A exposure alters sexual differentiation of the AVPV. Neurotoxicol Teratol 28: 111–8. prasad k. 2005. Hypocholesterolemic and antiatherosclerotic effect of flax lignan complex isolated from flaxseed. Atherosclerosis 179: 269–75. ryan e, chopra j, mccarthy f, maguire ar, o’brien nm. 2005. Qualitative and quantitative comparison of the cytotoxic and apoptotic potential of phytosterol oxidation products with their corresponding cholesterol oxidation products. Br J Nutr 94: 443–51. sacks fm, lichtenstein a, van horn l, harris w, kris-etherton p, winston m. soy protein, isoflavones, and cardiovascular health. 2006. An American Heart Association Science Advisory for Professionals from the Nutrition Committee. Circulation 113: 1034–44.

Phytoestrogens and phytosterols

457

sanders dj, minter hj, howes d, hepburn pa. 2000. The safety evaluation of phytosterol esters. Part 6. The comparative absorption and tissue distribution of phytosterols in the rat. Fd Chem Toxicol 38: 485–91. schmitt e, dekant w, stopper h. 2001. Estrogenic and genotoxic potential of phytoestrogens promoted as ‘natural’ alternatives to estrogen-replacement therapy. Naunyn-Schmiedeberg’s Arch Pharmacol 363: R145. song tt, hendrich s, murphy pa. 1999. Estrogenic activity of glycitein, a soy isoflavone. J Agric Food Chem 47: 1607–10. sprando rl, collins tf, wiesenfeld p, babu us, rees c, black t, olejnik n, rorie j. 2000. Testing the potential of flaxseed to affect spermatogenesis: morphometry. Food Chem Toxicol 38: 887–92. srilatha b, adaikan pg. 2004. Estrogen and phytoestrogen predispose to erectile dysfunction: Do ER-α and ER-β in the cavernosum play a role? Urology 63: 382–6. stopper h, schmitt e, kobras k. 2005. Genotoxicity of phytoestrogens. Mut Res 574: 139–55. thomas p, doughty k. 2004. Disruption of rapid, nongenomic steroid actions by environmental chemicals: interference with progestin stimulation of sperm motility in Atlantic croaker. Environ Sci Technol 38: 6328–32. thomson b, cressey pj, shaw ic. 2003. Dietary exposure to xenoestrogens in New Zealand. J Environ Monit 5: 229–35. tiemann u, tomek w, schneider f, vanselow j. 2003. Effects of the mycotoxins αand β-zearalenol on regulation of progesterone synthesis in cultured granulosa cells from porcine ovaries. Reprod Toxicol 17: 673–81. tou jc, chen j, thompson lu. 1998. Flaxseed and its lignan precursor, secoisolariciresinol diglycoside, affect pregnancy outcome and reproductive development in rats. J Nutr 128: 1861–8. tsakmakidis ia, lymberopoulos ag, alexopoulos c, boscos cm, kyriakis sc. 2006. In vitro effect of zearalenone and α-zearalenol on boar sperm characteristics and acrosome reaction. Reprod Domest Anim 41: 394–401. turner nj, thomson b, shaw i. 2003. Bioactive isoflavones in functional foods: the importance of gut microflora on bioavailability. Nutr Rev 61: 203–14. usda-iowa state university database on the isoflavone content of foods, release 1.3. 2002. http://www.nal.usda.gov/fnic/foodcomp/Data/isoflav/isoflav. html. virgilio al, iwami k, watjen w, kahl r, degen gh. 2004. Genotoxicity of the isoflavones genistein, daidzein and equol in V79 cells. Toxicol Lett 151: 151–62. waalkens-berendsen dh, wolterbeek apm, wijnands mvw, richold m, hepburn pa. 1999. Safety evaluation of phytosterol esters. Part 3. Two-generation reproduction study in rats with phytosterol esters – a novel functional food. Food Chem Toxicol 37: 683–96. ward we, chen j, thompson lu. 2001. Exposure to flaxseed or its purified lignan during suckling only or continuously does not alter reproductive indices in male or female offspring. J Toxicol Environ Health A 64: 567–77. whitsett tg jr, lamartiniere ca. 2006. Genistein and resveratrol: mammary cancer chemoprevention and mechanisms of action in the rat. Expert Rev Anticancer Ther 6: 1699–706. wozniak al, bulayeva nn, watson cs. 2005. Xenoestrogens at picomolar to nanomolar concentrations trigger membrane estrogen receptor-α-mediated Ca2+ fluxes and prolactin release in GH3/B6 pituitary tumor cells. Environ Health Perspect 113: 431–9. xu f, rychnovsky sd, belani jd, hobbs hh, cohen jc, rawson rb. 2005. Dual roles for cholesterol in mammalian cells. Proc Natl Acad Sci USA 102: 14551–6.

458

Endocrine-disrupting chemicals in food

zhang y, song tt, cunnick je, murphy pa, hendrich s. 1999. Daidzein and genistein glucuronides in vitro are weakly estrogenic and activate human natural killer cells in nutritionally relevant concentrations. J Nutr 129: 399–405. zhang y, murphy pa, hendrich s. 2003. Glucuronides are the main isoflavone metabolites in women. J Nutr 133: 399–404. zhong y, renouf m, lee s-o, hauck cc, murphy pa, hendrich s. 2006. High urinary isoflavone excretion phenotype decreases plasma cholesterol in golden Syrian hamsters fed soy protein. J Nutr 136: 2773–8. zinedine a, soriano jm, molto jc, manes j. 2007. Review on the toxicity, occurrence, metabolism, detoxification, regulations and intake of zearalenone: an oestrogenic mycotoxin. Fd Chem Toxicol 45: 1–18.

18 Pharmaceuticals A. H. Piersma and M. Luijten, National Institute for Public Health and the Environment RIVM, The Netherlands; V. Popov and V. Tomenko, Wessex Institute of Technology, UK; M. Altstein, Agricultural Research Organization, Israel; F. Kagampang and H. Schlesinger, Analyst Research Laboratories Ltd, Israel

Abstract: An inventory is presented of pharmaceuticals with a potential to affect human fecundity via exposure through the human food chain. Pharmaceuticals are reviewed in particular with respect to their mechanism of action, especially in view of endocrine disruption, their use pattern and their detection and persistence in the environment as important indicators of possible human exposure via the food chain. Sections on mechanisms of toxicity to fertility and on nutritional exposure pathways are followed by an extensive review of a series of relevant classes of pharmaceuticals. Finally, risk assessment approaches are reviewed. This work has been conducted as the first phase in the EU-FP6 project Food & Fecundity (F&F) as part of the identification of pharmaceutical compounds of interest for further analysis in food matrices. Key words: fecundity, fertility, pharmaceuticals, endocrine disruption, environmental exposure.

18.1

Introduction

The volume of pharmaceuticals consumed in the European Union (EU) is large, with the number of different substances used estimated as approximately 3000 in human medicine (Fent et al., 2006), e.g. analgesics, anti-inflammatory drugs, antibiotics, contraceptives. A large number of pharmaceutical products (PP) are also used in veterinary medicine, e.g. antibiotics and anti-inflammatory drugs. Possible accumulation of these pharmaceuticals in the environment has not been of great concern in the past. However, with the improvement of the accuracy and sensitivity of the detection methods the awareness of the presence of pharmaceutically active compounds in the environment has increased. The EU-FP6 project ‘Food and Fecundity’ was aimed at the identification of possibly hazardous

460

Endocrine-disrupting chemicals in food

concentrations of pharmaceutical residues in the food chain, specifically with respect to adverse effects on fertility. As a first step in the project, an inventory of relevant pharmaceuticals was made. The prime relevant characteristic of pharmaceuticals of interest is their known effect on fertility, through an endocrine mechanism of action. We have therefore reviewed pharmaceuticals for causing infertility, sexual dysfunction in men or women, and altered libido. In Section 18.2, a detailed description of disorders falling into each category is given together with evidence reported in the literature. In Section 18.3 exposure pathways of pharmaceuticals in food are identified. The link between three media (soil, surface water and groundwater) is shown, indicating possible cross-contamination. The classification of main pathways is given together with detailed description of how dispersion of pharmaceuticals in the environment may occur. Finally, the importance of water as PPs transportation media is stipulated. Sections 18.4–18.13 provide an overview of pharmaceuticals which are suspected of affecting human fecundity. The classes of interest identified are as follows: non-steroidal anti-inflammatory drugs, NSAIDs (ibuprofen, naproxen, diclofenac and indomethacin), antipyretic drugs (acetaminophen), peroxisome proliferators (clofibrate and gemfibrozil), antihypertensive drugs (methyldopa), anticonvulsants (carbamazepine, valproic acid, phenobarbital and phenytoin), selective seratonin reuptake inhibitors (SSRIs) (fluoxetine hydrochloride, fluvoxamine maleate and sertraline), beta-blockers (propranolol, metoprolol and atenolol), progestins (ethynodiol diacetate, norethindrone and levonorgestrel) and estrogens (17α-ethynylestradiol), antibiotics (sulfamethoxazole–trimethoprim combination, tetracycline, doxycycline, minocycline and erythromycin). The key factors which affected inclusion of PPs in the list were: evidence of adverse effects related to fertility and fecundity, production volume, presence and persistence in the environment, and ability to reach target populations through relevant exposure pathways. The results of the findings are included in the description of PPs. The risk assessment process for compounds present in food is discussed in Section 18.14, starting from the classical components: hazard identification, dose–response assessment, exposure assessment and risk characterisation. Special attention is given to approaches currently accepted by regulatory agencies. Additionally, owing to limitations of adopted techniques, a probabilistic assessment is described for both dose–response and exposure assessment, which allows replacing point estimates with the range of plausible values together with their uncertainties. Finally, (quantitative) structure–activity relationships models are discussed. These models act as a supportive tool in the risk assessment process and are potentially capable of reducing the number of animal tests. There are significant gaps in the knowledge of whether and how pharmaceutical residues reach effective exposure levels and affect human fecun-

Pharmaceuticals

461

dity owing to a lack of dedicated research. However, the growing amount of literature devoted to assessment of PPs in relation to fecundity indicates increased concern and recognition of this class of chemicals as being able to cause problems. There are ample indications of endocrine-modulating effects of these compounds. On the other hand, critical data on many aspects of PPs’ mechanism of action, production volumes, environmental concentrations and persistence are missing. This inventory is a compilation of our knowledge of PPs, and priotitises compounds of concern for further study of concentrations in food matrices as a basis for a better informed risk assessment.

18.2

Classification of the mechanisms by which pharmaceuticals affect fecundity

18.2.1 Infertility Infertility is one element of a spectrum of reproductive disorders that includes miscarriage, congenital abnormality, premature delivery and stillbirth (Gnoth et al., 2005; Waghmarae, 1972). Infertility, defined as the failure to conceive after two years of unprotected intercourse, is fairly common, affecting about 15% of all couples at some time during their reproductive lives (Fernandez et al., 1991; Kolettis, 2003). It is generally detected only when a couple is actively trying to conceive. It can be difficult to draw firm conclusions about trends in infertility rates but the high number of patients currently attending fertility clinics suggests a growing problem. Causes of infertility in women include failure of ovulation, tubal damage, endometriosis and hostile cervical mucus (Olive et al., 2003; Wardle et al., 1985; Zawar et al., 2003). In men, sperm defects, coital factors such as impotence or retrograde ejaculation, and hypogonadism may be implicated (Boyd, 1988; Oehninger & Alexander, 1991). In as many as 30% of cases, a cause cannot be found (Tadokoro et al., 1997). Drugs and other toxins may be responsible in a number of cases, but, in general, the effects of drugs on fertility have been poorly studied. The activity of the gonads (testes or ovaries) is regulated by the pituitary gonadotrophins, follicle-stimulating hormone (FSH) and luteinising hormone (LH) (Knobil, 1988a). Secretion of both hormones is controlled by gonadotrophin-releasing hormone (GnRH) from the hypothalamus (Knobil, 1988b, 1990; Weiner, 1996). FSH regulates the development of Sertoli cells (which are involved in sperm maturation) in the testes, and the Graafian follicle in females. LH controls formation of the corpus luteum in females and testosterone production by the Leydig cells in males. Both FSH and LH regulate estrogen production and ovulation. Decreased amounts of FSH and/or LH reaching the testes can inhibit spermatogenesis. About 30% of infertile women have anovulatory infertility (Baird, 1979). They may be present with amenorrhoea (primary or secondary),

462

Endocrine-disrupting chemicals in food

oligomenorrhoea (infrequent or irregular periods) or occasionally with regular menstrual cycles but low or undetectable serum progesterone concentrations in the putative luteal phase. Secondary amenorrhoea is defined as the absence of menstruation for at least six months in a woman with previously normal and regular menses (Marti, 1991). Hyperprolactinaemia is a common finding in women with amenorrhoea or oligomenorrhoea (Godo, 1984; Judd et al., 1976; Molitch, 1992); this can be drug-induced. Drugs known to increase prolactin include methyldopa (Arze et al., 1981), metoclopramide (Anderson et al., 1981; Rossi et al., 2002), cimetidine (Gonzales-Villapando et al., 1980), phenothiazines (Yarkoni et al., 1978) and oestrogens (Furuhjelm et al., 1980). Amenorrhoea is also associated with high-dose corticosteroids (Turkington & MacIndoe, 1972), danazol (Dmowski, 1988) and isoniazid (Klein et al., 1976).

18.2.2 Sexual dysfunction Sexual function may be divided into three categories reflecting the sexual response cycle: (1) libido or sexual desire; (2) arousal, including erectile function in men and lubrication in women; and (3) release. Drugs can affect one or more areas of the response cycle. Understanding of the sexual response remains incomplete but there is evidence of dopaminergic, adrenergic, muscarinic and serotonergic involvement. In general, increase in sexual behaviour by dopamine (Giuliano & Allard, 2001) and inhibition by serotonin (Barnes et al., 1979) have been reported. Libido is influenced by reproductive hormones and by the emotional and physical health of the individual. Testosterone is necessary for normal sexual arousal, probably in both men and women, and in men testosterone deficiency is associated with impotence (Buvat, 2003).

18.2.3 Sexual dysfunction in men The aetiology of erectile dysfunction is often vascular but other contributory factors include drug therapy, endocrine disease and neurological dysfunction (Hafez & Hafez, 2005). Male sexual function depends on the coordination of neurogenic, hormonal and psychological mechanisms and disruption of one or more of these may result in erectile dysfunction. About 25% of cases of erectile dysfunction are believed to be drug-induced (Keene & Davies, 1999; Sidi et al., 1986). The classes of drugs most frequently implicated are antihypertensives (Della et al., 2003; Dusing, 2005; Kloner, 2003), antidepressants (Labbate et al., 2003; Rosen and Marin, 2003; Rudkin et al., 2004), antipsychotics (Segraves, 1988), (Compton & Miller, 2001) and anti-epileptics (Smaldone et al., 2004). Ejaculation is achieved via stimulation of alpha-adrenergic receptors, leading to contraction of the smooth muscle of the prostate, seminal vesicles and vas deferens. Disorders of ejaculation comprise ejaculatory failure and retrograde ejaculation in

Pharmaceuticals

463

which semen passes into the bladder. A number of drugs have also been implicated in these disorders. High rates of erectile dysfunction and ejaculatory failure are associated with the older adrenergic blockers reserpine (Cameron et al., 1996; Dail et al., 1987) and guanethidine (Moss & Procci, 1982), which are no longer used. Clonidine (Beeley, 1984; Hedlund & Andersson, 1985) and methyldopa (Melman et al., 1984; Newman & Salerno, 1974) have also caused loss of libido, erectile dysfunction and ejaculatory failure. The alpha-adrenergic blockers indoramin (Holmes & Sorkin, 1986; Pentland et al., 1981) and prazosin (Hedlund & Andersson, 1989; Smith & Talbert, 1986) can cause ejaculatory failure and retrograde ejaculation. The incidence of sexual dysfunction in men taking diuretics is between two and six times higher than in men taking placebo (Chang et al., 1991). Thiazides may cause reduced libido, erectile dysfunction and problems with ejaculation (Joseph & Schuna, 1990; Muller et al., 1991). The underlying mechanism is unclear as thiazides lack significant hormonal, autonomic or central nervous system effects but a direct effect on smooth muscle is thought to be responsible. Erectile dysfunction is well documented with propranolol and can occur with other beta-blockers (Bathen, 1978; Frances & Kaplan, 1982; Silvestri et al., 2003). The problem is more likely with lipid soluble beta-blockers but has also been reported with atenolol (Morrissette et al., 1993; Silvestri et al., 2003) and with ophthalmic timolol (Fraunfelder & Meyer, 1985; Katz, 1986). Reduced perfusion pressure caused by a drop in blood pressure or a direct effect on smooth muscle may be responsible. Calcium channel blockers seem to cause fewer problems with sexual function than diuretics or beta-blockers although there are several published case reports of erectile dysfunction (Fovaeus et al., 1987; Sparwasser et al., 1998).

18.2.4 Sexual dysfunction in women In women, sexual dysfunction has not been thoroughly investigated and the underlying mechanisms are not fully understood. Most reported problems relate to orgasm dysfunction, reduced lubrication or loss of libido. Thioridazine has been known since 1961 to inhibit ejaculation in men but it was not until 20 years later that the first report of inhibition of female orgasm was published (Shen & Park, 1982; Shen & Sata, 1983, 1990). Failure to achieve orgasm is one of the most common sexual adverse effects of psychotropic drugs in women. This problem has been described with antidepressants (tricyclics or TCAs, monoamine oxidase inhibitors or MAOIs, and SSRIs). Such effects have also been reported with MAOIs (Lesko et al., 1982; Moss, 1983; Pohl, 1983; Shen & Mallya, 1983), TCAs (Cohen & Bartlik, 1998), clozapine (Hummer et al., 1999), risperidone (Kelly & Conley, 2006; Wirshing et al., 2002) and the antihypertensives clonidine and methyldopa (Beeley, 1984; Smith & Talbert, 1986).

464

Endocrine-disrupting chemicals in food

18.2.5 Altered libido Loss of libido or sexual desire is frequently attributed to medication in both men and women. In women, loss of libido is the commonest reported form of sexual dysfunction; it is extremely difficult to quantify and manage. Changes in desire may be due to illness, stress or fatigue, or may be druginduced. In controlled studies women have rarely been questioned about the effect of medication on sexual function and therefore most reports of altered libido are anecdotal or case reports. Several antihypertensives, including clonidine and methyldopa, reduce female libido (Beeley, 1984; Cavalier, 1995; Meston et al., 1997), Studies of both men and women taking methyldopa report an incidence of decreased libido ranging from 7 to 14% (Chowdhury, 1987; Weiss, 1991). Spironolactone has anti-androgenic effects and is clearly linked with decreased libido (Cuttler et al., 1979; Mantero & Lucarelli, 2000). Psychotropic drugs affect sexual desire in men and women by several possible mechanisms, including sedation, effects on central or peripheral neurotransmitters, or effects on hormones (e.g., prolactin) (Clayton & Shen, 1998). Antidepressants have been reported to decrease sexual desire (Rosen et al., 1999). MAOIs, particularly phenelzine, are frequently implicated (Gupta et al., 1999; Warneke, 1994). The SSRIs have all been reported to decrease libido, possibly as a consequence of an indirect effect on dopamine; the incidence in men and women may be as high as 40% (Kanalay & Berman, 2002; Meston, 2004; Montejo-Gonzales et al., 1997; Rosen et al., 1999). In general, rates of sexual dysfunction appear to be greatest with the SSRIs, followed by MAOIs then TCAs. Rates of sexual dysfunction appear to be similar for all the SSRIs and it is not known if switching between them will diminish sexual side effects. Case reports of decreased libido with anxiolytics have been published; centrally mediated sedation and muscle relaxation are thought to be responsible. Cimetidine has been reported to cause loss of libido, possibly because of its anti-androgen activity (Biron, 1979; Pierce, 1983; Webster, 1979). This is likely to be dose-related. The influence of testosterone on libido is well recognised and any drug that reduces serum testosterone may lead to a loss of sexual desire. In men, this includes drugs such as estrogens (Matuszkiewics-Rowinska et al., 1999), anti-androgens (Bancroft et al., 1974; Holzbeierlein et al., 2004) and gonadorelin analogues (Falkson et al., 1991; Holzbeierlein et al., 2003; Kher & Kalla, 1996).

18.3

Exposure pathways of pharmaceutical products in food

The main exposure pathways for pharmaceuticals ending in human food chain are shown in Fig. 18.1. In Fig. 18.1 the links between soil, surface water and groundwater are emphasised, indicating that any contaminant in one of these media may eventually contaminate the other two. There are three

Pharmaceuticals

Pharmceutical products Humans 5 Human excreta Sewage 9

1

Sewage treatment plant 11 12

10

Landfill effluent treatment

Surface water and sediment

4

Wastewater

Landfill

Effluent

Leaching 14

Fish

3

7

13

2

Production

6 Not used discarded

8

Veterinary medicinal products 19 Poultry and Leakage cattle

465

16

Poultry litter and cattle manure

Sewage 17 sludge

Agricultural use

Leaching dissipation run-off

18

Meat, eggs and dairy products

Fruits and vegetables

20

Evaporation

15

Soil

Groundwater

Air

Water for domestic use

Fig. 18.1 Exposure pathways of pharmaceutical products that may end up in human food chain.

main pathways through which dispersion of pharmaceuticals in the environment may occur: (i) the local pathway in the vicinity of the production line; (ii) the distribution/dispersion through farmed animals; (iii) the distribution/dispersion due to consumption by patients. Pathways (ii) and (iii) are named distribution/dispersion because initially the distribution of the pharmaceuticals takes place to patients/farmed animals and in the next stage the dispersion in the environment occurs through human excreta or poultry litter or cattle manure. The local pathways would depend on the level of security implemented at the production line and would be represented by the main pathways consisting of (4) in Fig. 18.1 in the form of leakage in the soil, groundwater, surface water or air and by (3) in the form of wastewater which is discharged directly or after pre-treatment into the sewage treatment system. Another pathway would be related to disposal, incineration or treatment of waste products, which is not shown in the figure. In this study we presume that the local exposure pathways are negligible due to proper safety measures implemented at the production site. The second exposure pathway is through veterinary medicinal products administered to food-producing animals, in Fig. 18.1 represented by (2). This

466

Endocrine-disrupting chemicals in food

link provides a quite short and direct link between the drug and supply food chain for human consumption. As can be seen in the figure, after the veterinary medicinal product has been administered to the animal it can end up in the meat products, eggs and through milk into dairy products. In the EU Directive 2001/82/EC of the European Parliament and of the Council on the Community code relating to veterinary medicinal products amended by Directive 2004/28/EC of the European Parliament regulate the use of veterinary medicinal product in the EU and if these directives are followed no adverse effects should occur in consumers. The main cause for concern is the possibility that some substances are used by farmers without proper control, in which case the residues of the drugs may end up in food for human consumption. This pathway can lead to a less direct link to human and environmental exposure through animal excreta which may contaminate the soil, surface waters and groundwater around farms. In some cases poultry litter and cattle manure are used in agriculture and it may be that in some cases the residues and metabolites end up in agricultural products for human consumption. However, more research in this area is required. There is a possibility that the drug residues are dispersed due to leaching and run-off, while dispersion through air due to evaporation from fields where poultry litter and cattle manure are applied is likely to be negligible. The third pathway, denoted as route 1 in Fig. 18.1, considered in this study is through pharmaceutical products consumed by patients. It is presumed that there are two main routes for the drugs to end up in the environment. The first one is by patients discarding the unused drugs in domestic waste which further may end up in landfills (route 6). Leaching from landfills may contaminate the soil and groundwater, though the soil would represent a filter which would reduce the amount of drug residues that would reach groundwater. However, in some cases the groundwater table may rise to the bottom of the landfill, establishing a direct link to the leachate. The landfill effluent after treatment may be discarded in surface waters or soil from where it can contaminate groundwater. Also, in some cases the landfill effluent may be discharged in the local sewage system or directly into the sewage treatment plant for further treatment. After the passage through the sewage treatment plant when part of the drug residual would be removed, it may be discharged in surface waters or soil. Also, it may be present in the sewage sludge which may further be used in agriculture. Such application is regulated within the EC by the Council Directive 86/278/EEC on the protection of the environment, and in particular of the soil, when sewage sludge is used in agriculture. From the above considerations it is evident that water is a very important medium for transport of PPs, since the PP residues can potentially end up in the sewage system and sewage treatment works through different pathways and from there can be dispersed in the environment. Also, water is essential for sustaining life on our planet and therefore any contaminant

Pharmaceuticals

467

ending up in surface water or groundwater can have a large impact on the safety of humans and environment.

18.4

Pharmaceutical products potentially affecting human fecundity and their mechanism of action

One of the objectives of this work was to create a prioritisation list of PPs bearing a potential of affecting human fecundity by entering the food chain. The list is based on an extensive literature search while considering the following criteria: • Does the available data indicate existence of a mechanism of action with an effect on fecundity? • Is the production volume sufficiently large to cause concern? • Has the PP been detected in food and/or environment? • Is the PP sufficiently persistent in the environment? In this chapter pharmaceutical compounds which have been selected according to the above selection criteria are evaluated in detail, prior to further investigation in the EU-FP6 project Food & Fecundity (F&F). The compounds investigated belong to the following groups of pharmaceutical products: NSAIDs, antipyretic drugs, peroxisome proliferators, antihypertensive drugs, anticonvulsants, SSRIs, beta-blockers, steroid contraceptives and antibiotics.

18.5

Non-steroidal anti-inflammatory drugs

The production volume of NSAIDs is high and NSAIDs are prescribed in high amounts. In addition, they have been detected in environmental samples, albeit at low concentrations. Mechanistically, NSAIDs may play a role in at least one type of female infertility involving disruption of sex hormone homeostasis. Prostaglandin inhibition appears to increase the incidence of luteinised unruptured follicle syndrome, a condition in which normal ovarian follicular development is followed by an elevation of serum progesterone compatible with ovulation, but the cycle remains anovulatory because the follicular wall remains unruptured (Killick & Elstein, 1987; Marik & Hulka, 1978). Rat and rabbit studies have reported ovulation inhibition in association with the prostaglandin inhibitor, indomethacin (Armstrong & Grinwich, 1972; O’Grady et al., 1972; Espey et al., 1982). The currently available animal data have raised an as-yet unresolved dispute about the possible fertility effects of NSAIDs. In women, ultrasound scans of follicular development have been used to show a fivefold increase in the incidence of this syndrome in the presence of some NSAIDs (Killick & Elstein, 1987). The prolonged use of NSAIDs, which may occur in the

468

Endocrine-disrupting chemicals in food

Table 18.1 The consumption of NSAIDs in the Netherlands in defined daily dose (DDD) Pharmaceutical Ibuprofen Naproxen Indomethacin Diclofenac

2000

2001

2002

2003

2004

29 093 500 36 432 400 4 404 800 54 018 300

30 426 100 36 715 000 3 934 600 53 731 200

31 550 800 35 531 500 3 574 500 52 343 500

31 926 100 33 585 600 3 316 100 51 529 900

24 287 400 29 445 500 2 890 100 48 853 700

treatment of chronic pain and inflammation of rheumatological conditions, is most likely to be associated with this antifertility effect. Similar findings have been reported for both COX-1 and COX- 2 NSAIDs (Killick & Elstein, 1987; Pall et al., 2001). In conclusion, the low environmental levels argue against an actual risk of NSAID residues for human health. The high production and use, in addition to possible fertility effects with a mechanistic plausibility argue towards the opposite. NSAIDs may be considered part of PCPP (pharmaceuticals and personal care products), which is a large and very varied group of chemicals for which it is not possible to make general statements on their relevance for F&F. The consumption of NSAIDs in the Netherlands in defined daily dose (DDDs) was estimated as given in Table 18.1.

18.5.1 Ibuprofen Ibuprofen may inhibit follicular collapse, but this effect is only seen in a small group of study subjects (Uhler et al., 2001). Ibuprofen has been detected in the environment: • • • • • • • • • • • • • •

STP effluent, Italy, 0.121 μg/L (Zuccato et al., 2005) STP effluent, Finland, 0.004–0.064 μg/L (Lindqvist et al., 2005) STW effluent, UK, median 3.086 μg/L (Ashton et al., 2004) STP effluent, Källby, Sweden, 0.15 μg/L (Bendz et al., 2005) STP effluents, survey, 0.05–7.11 μg/L (Andreozzi et al., 2003) German rivers, <0.005–0.139 μg/L (Halling-Sørensen et al., 1998) Effluent sedimentation tank, up to 12 μg/L (Halling-Sørensen et al., 1998) River Elbe, up to 0.024 μg/L (Wiegel et al., 2004) STP effluent, California, USA, 0.007–0.037 μg/L (Gross et al., 2004) Santa Ana River, California, USA, 0.013–0.151 μg/L (Gross et al., 2004) STP effluent, Switzerland, up to 2.2 μg/L (Tauxe-Wuersch et al., 2005) STW effluent, UK, 1.979–4.239 μg/L (Roberts & Thomas, 2006) River Tyne, UK, 0.144–2.370 μg/L (Roberts & Thomas, 2006) STP effluents, Canada, 0.079–1.885 μg/L (Metcalfe et al., 2003)

Pharmaceuticals

469

Ibuprofen has been detected in German drinking water at 0.003 μg/L (Webb et al., 2003). 18.5.2 Naproxen Naproxen significantly reduced ovulatory efficiency and progesterone (PG) production both in vivo and in vitro in human chorionic gonadotropin (hCG)-treated rabbits (Zanagnolo et al., 1996). Naproxen has been detected in the environment: • Sewage treatment plant (STP) effluent, Finland 0.017–0.057 μg/L (Lindqvist et al., 2005) • River Elbe, up to 0.032 μg/L (Wiegel et al., 2004) • STP effluent, Källby, Sweden, 0.25 μg/L (Bendz et al., 2005) • STP effluents survey, 1.12–5.22 μg/L (Andreozzi et al., 2003) • STP effluent, California, USA, 0–0.089 μg/L (Gross et al., 2004) • Santa Ana River, California, USA, 0–0.022 μg/L (Gross et al., 2004) • STP effluents, Canada, 0.021–0.524 μg/L (Metcalfe et al., 2003) 18.5.3 Diclofenac Diclofenac inhibits ovulation in the rat and rabbit (Armstrong & Grinwich, 1972; Espey, 1983; O’Grady et al., 1972). Diclofenac delays implantation in the rat (Carp et al., 1988). Diclofenac has been detected in the environment: • • • • • • • • • • • •

STP effluent, France, 0.25–0.41 μg/L (Ferrari et al., 2003) STP effluent, Greece, 0.89 μg/L (Ferrari et al., 2003) STP effluent, Italy, 0.47–5.45 μg/L (Ferrari et al., 2003) German rivers, 0.015–0.49 μg/L (Halling-Sørensen et al., 1998) STW effluent, UK, median 0.424 μg/L (Ashton et al., 2004) STW effluent, UK, 0.261–0.598 μg/L (Roberts & Thomas, 2006) STP effluent, Finland, 0.011–0.040 μg/L (Lindqvist et al., 2005) STP effluent, Källby, Sweden, 0.12 μg/L (Bendz et al., 2005) STP effluents survey, 0.68–5.45 μg/L (Andreozzi et al., 2003) STP effluent, Switzerland, up to 1.9 μg/L (Tauxe-Wuersch et al., 2005) River Elbe, up to 0.033 μg/L (Wiegel et al., 2004) STP effluents, Canada, 0.005–0.359 μg/L (Metcalfe et al., 2003)

Diclofenac has been detected in German drinking water at 0.006 μg/L (Webb et al., 2003). 18.5.4 Indomethacin Administration of indomethacin has been demonstrated to induce delayed follicular rupture or luteinised unruptured follicle (LUF) in previously ovulating women (Stone et al., 2002). Indomethacin affects fertility: it is

470

Endocrine-disrupting chemicals in food

concluded that the antifertility effect of indomethacin at the time of implantation is exerted by reducing progesterone concentrations in plasma and uterine fluid, possibly affecting steroidogenesis, and by reducing the percentage of albumin in plasma and in uterine fluid, probably by increasing renal excretion of albumin. These effects of indomethacin provide an environment within the uterus that would not support embryo implantation and development (El Banna et al., 1993). Indomethacin been detected in the environment: • German rivers, up to 0.121 μg/L (Halling-Sørensen et al., 1998) • STP effluents, Canada, 0.010–0.378 μg/L (Metcalfe et al., 2003) 18.5.5 Conclusions on non-steroidal anti-inflammatory drugs There is little evidence for an adverse effect on fecundity by ibuprofen or naproxen. In addition, the removal efficiency during sewage treatment for both compounds is higher than 90%. Both diclofenac and indomethacin affect fecundity at least partly through an endocrine-disrupting mechanism and have been detected in the environment. Diclofenac has been detected in STP effluents and even in drinking water. In addition, it is used in higher amounts than indomethacin. Based on these data, diclofenac would be first and indomethacin second in possible risk.

18.6

Antipyretic drugs

18.6.1 Acetaminophen or paracetamol Acetaminophen or paracetamol is a non-opiate, non-salicylate analgesic and antipyretic drug. It is present in more than 850 over-the-counter and prescription formulas (Prescott, 2000). In humans acetaminophens can significantly lower basal levels of gonadotrophin and estradiol (Cramer et al., 1998) and can therefore be considered as possible endocrine disrupters. Several in vivo animal studies suggest that acetaminophen may also alter some hormone-regulated processes in reproductive tissues. It was reported to reduce the reproductive capacity, testicular weight and spermatogenesis of mice (Reel et al., 1992) and reduced estradiol-induced uterine peroxidase activity and nuclear progesterone receptor protein in immature mice (Patel & Rosengren, 2001). It has been detected in surface water monitoring studies at concentrations of up to 10 μg/L (Boxall, 2004; Daughton & Ternes, 1999; Lam et al., 2004). 18.6.2 Conclusions on acetaminophen Considering the amount of acetaminophen in use nowadays, presence in the environment and the evidence of its effects on fecundity, this compound is relevant for further investigation.

Pharmaceuticals

471

Table 18.2 The consumption of lipid regulators in the Netherlands in defined daily dose (DDD) Pharmaceutical Clofibrate Gemfibrozil

18.7

2000

2001

2002

2003

2004

89 050 6 202 900

83 277 6 090 700

72 237 5 727 000

59 263 5 462 000

53 325 5 058 300

Peroxisome proliferators

Exposure to some peroxisome proliferators leads to toxic effects on sex organ function, possibly by alterations of steroid hormone metabolism. This mechanism marks these drugs as possible endocrine disruptors. Two examples of widely used peroxisome proliferators are the lipid regulators clofibrate and gemfibrozil. The consumption of lipid regulators in the Netherlands in DDD was estimated as given in Table 18.2.

18.7.1 Clofibrate Clofibrate affects hCG and progesterone concentrations (Hashimoto et al., 2004). Clofibrate has a selective stimulatory effect on the hormonal action of estradiol in the mammary gland but not in the uterus (Xu et al., 2001). The clinical significance of these findings is unknown; however, according to the manufacturer (Ayerst Laboratories, New York), clofibrate use has been associated with impotence and decreased libido in men. Clofibrate has been reported to be uterotrophic to immature female rats (Chandra et al., 1982), but others could not confirm these findings (Ashby et al., 1997). Clofibrate has been detected in the environment: • STP effluent, Italy, 0–0.68 μg/L (Ferrari et al., 2003) • STP effluent, Sweden, 0.46 μg/L (Ferrari et al., 2003) • River water, various locations, up to 1.75 μg/L (Halling-Sørensen et al., 1998) • STP effluent, Switzerland, 0.020–0.025 μg/L (Tauxe-Wuersch et al., 2005) • STW effluent, UK, up to 0.044 μg/L (Roberts & Thomas, 2006) • STP effluents, Canada, 0.002–0.044 μg/L (Metcalfe et al., 2003) Clofibrate has been detected in groundwater and surface water up to concentrations 0.100 μg/L (Stolker et al., 2004). Clofibrate has been detected in German drinking water 0.070 μg/L (Webb et al., 2003) and in drinking water in the concentration of 0.025–0.100 μg/L (Stolker et al., 2004).

18.7.2 Gemfibrozil Exposure to environmental levels of gemfibrozil leads to bioconcentration of the drug in plasma and a reduction of plasma testosterone levels

472

Endocrine-disrupting chemicals in food

(Mimeault et al., 2005). Gemfibrozil affects hCG and progesterone concentrations (Hashimoto et al., 2004). Male rats given about 17 times the average daily human dose of gemfibrozil showed inconsistent and equivocal lower rates of fertility relative to the concurrent controls (FitzGerald et al., 1987). Gemfibrozil is occasionally associated with impotence and decreased libido (Bain et al., 1990; Pizarro et al., 1990). In vitro studies using rat tissues have reported that gemfibrozil and other inducers of hepatic peroxisome proliferation may alter the steroidogenic function of Leydig cells (Liu et al., 1996). Exposure to gemfibrozil results in decreased expression of enzymes that inactivate estradiol. The reported increased expression of aromatase may explain why male rats exposed to gemfibrozil have higher serum estradiol levels. These higher estradiol levels in male rats have been thought to be mechanistically linked to Leydig cell hyperplasia and adenomas (Corton et al., 1997). Gemfibrozil has been detected in the environment: River Elbe, up to 0.027 μg/L (Wiegel et al., 2004) STP effluent, Källby, Sweden, 0.18 μg/L (Bendz et al., 2005) STP effluent, California, USA, 0.015–0.065 μg/L (Gross et al., 2004) Santa Ana River, California, USA, 0.001–0.059 μg/L (Gross et al., 2004) • STP effluents survey, 0.84–4.76 μg/L (Andreozzi et al., 2003) • STP effluents, Canada, 0.005–1.493 μg/L (Metcalfe et al., 2003)

• • • •

18.7.3 Conclusions on peroxisome proliferators Both clofibrate and gemfibrozil are candidates for further study in view of their endocrine-mediated mechanism of action as well as their environmental detection.

18.8

Antihypertensive drugs

18.8.1 Methyldopa (Aldomet) Methyldopa is a drug that is used to treat high blood pressure. It works by relaxing the blood vessels so that blood can flow more easily through the body. Methyldopa decreased sperm count, sperm motility, the number of late spermatids and the male fertility index when given to male rats at 200 and 400 mg/kg/day (3.3 and 6.7 times the maximum daily human dose when compared on the basis of body weight; 0.5 and 1 times the maximum daily human dose when compared on the basis of body surface area) (Weiss, 1991). Methyldopa appears in breast milk (Beardmore et al., 2002; White et al., 1985). Methyldopa interferes with sex hormone homeostasis via an increase in prolactin levels. Elevated prolactin serum concentrations inhibit gonadotropin secretion and sex steroid synthesis. Because prolactin con-

Pharmaceuticals

473

centrations higher than 60 μg/L are associated with anovulation, women with hyperprolactinemia typically present with menstrual irregularities such as oligomenorrhea or amenorrhea and infertility. In addition, approximately 40–70% of women with hyperprolactinemia will have galactorrhea (Arze et al., 1981). Hyperprolactinemia in men, although rare, may cause decreased libido, erectile dysfunction, infertility, galactorrhe, or gynecomastia (Ou et al., 1991; Molitch, 1992). Methyldopa has been detected in the environment: • River Lee, 17.5 μg/L (Richardson & Bowron, 1985)

18.8.2 Conclusions on methyldopa Methyldopa is relevant for further analysis, based on its endocrine modulation, reported effects on fecundity and presence in the environment.

18.9

Anticonvulsants

The consumption of anticonvulsants in the Netherlands in DDD was estimated as given in Table 18.3.

18.9.1 Carbamazepine Carbamazepine affects sex hormone homeostasis through increases in serum sex hormone-binding globulin (SHBG) concentrations in both men and women with epilepsy. Over time, the increase in serum SHBG levels leads to reduced bioactivity of testosterone and estradiol, which may result in reduced potency in men and menstrual disorders in some women, and thus to reduced fertility (Isojarvi et al., 2005). Use of carbamazepine is associated with changes in serum sex-hormone levels and sperm abnormalities in men with epilepsy (Isojarvi et al., 2004; Mikkonen et al., 2004). However, Roste et al. (2003) could not demonstrate any significant changes in semen quality. Male rats fed carbamazepine for 30–60 days had decreased

Table 18.3 The consumption of anticonvulsants in the Netherlands in defined daily dose (DDD) Pharmaceutical

2000

2001

2002

2003

2004

Carbamazepine Valproic acid Phenobarbital Phenytoin

10 352 500 10 877 000 3 278 100 6 288 200

10 216 900 11 378 900 3 240 900 5 893 700

10 175 100 12 017 200 3 034 300 5 550 700

10 134 200 12 648 200 2 962 700 5 338 900

10 028 900 13 301 600 2 972 100 5 175 200

474

Endocrine-disrupting chemicals in food

testicular weight, sperm cell concentration, live sperm, and percentage of progressively motile spermatozoa (Soliman et al., 1999). Carbamazepine is highly persistent in the environment: STP effluent, Berlin, Germany, >1 μg/L (Zuehlke et al., 2004) Several STP effluents, Italy, 0.3 μg/L (Zuccato et al., 2005) STP effluent, Källby, Sweden, >1 μg/L (Bendz et al., 2005) River Elbe, Germany, >1 μg/L (Wiegel et al., 2004) STP effluent, France, 0.98–1.2 μg/L (Wiegel et al., 2004) STP effluent, Greece, 1.03 μg/L (Wiegel et al., 2004) STP effluent, Italy, 0.3–0.5 μg/L (Wiegel et al., 2004) STP effluent, Sweden, 0.87 μg/L (Wiegel et al., 2004) STP effluents survey, 0.87–1.2 μg/L (Andreozzi et al., 2003) STP effluent, Peterborough, Canada, 0.251 μg/L (Miao & Metcalfe, 2003) • STP effluents, Canada, 0.007–0.126 μg/L (Metcalfe et al., 2003)

• • • • • • • • • •

Carbamazepine has been detected in groundwater up to concentrations of 1.1 μg/L (Heberer, 2002; Stolker et al., 2004). Carbamazepine has been detected in drinking water in the concentration of 0.030 μg/L (Heberer, 2002; Webb et al., 2003) and up to 0.025 μg/L (Stolker et al., 2004).

18.9.2 Valproic acid Valproic acid medication is possibly endocrine disrupting as it may modulate serum androgen concentrations and it reduces serum FSH levels in men with epilepsy. In women, use of valproic acid appears to be associated with a frequent occurrence of reproductive endocrine disorders characterised by polycystic changes in the ovaries, high serum testosterone concentrations (hyperandrogenism) and menstrual disorders (Isojarvi et al., 2005). Use of valproic acid is associated with changes in serum sex-hormone levels, sperm abnormalities and a lower testicular size/body mass index (BMI) ratio in men with epilepsy (Mikkonen et al., 2004; Roste et al., 2003). Valproic acid has not been detected in the environment so far.

18.9.3 Phenobarbital Phenobarbital increases serum SHBG concentrations in both men and women with epilepsy which influences sex hormone homeostasis. Over time, the increase in serum SHBG levels leads to reduced bioactivity of testosterone and estradiol, which may result in reduced potency in men and menstrual disorders in some women, and thus to reduced fertility (Isojarvi et al., 2004). Phenobarbital inhibits the biological clock control of ovulation in hamsters (Alleva & Alleva, 1995). Phenobarbital delays ovulation and affects oocyte function in the rodent (Stoker et al., 2001). Phenobarbital has not been detected in the environment so far.

Pharmaceuticals

475

18.9.4 Phenytoin Phenytoin is an endocrine modulator as it increases serum SHBG concentrations in both men and women with epilepsy. Over time, the increase in serum SHBG levels leads to reduced bioactivity of testosterone and estradiol, which may result in reduced potency in men and menstrual disorders in some women, and thus to reduced fertility (Isojarvi et al., 2004). Phenytoin inhibits both the first ovulation and uterine development in gonadotropin-primed immature rats (Tamura et al., 2000). Phenytoin has not been detected in the environment so far.

18.9.5 Conclusions on anticonvulsants Several anticonvulsants affect fecundity through an endocrine-disrupting mechanism and are consumed in large quantities, but only carbamazepine is highly persistent and has been detected in STP effluents, in groundwater and even in drinking water.

18.10 Serotonin reuptake inhibitors SSRIs are a class of antidepressants. They act within the brain to increase the amount of the neurotransmitter, serotonin (5-hydroxytryptamine or 5HTP), in the synaptic gap by inhibiting its reuptake.

18.10.1 Fluoxetine hydrochloride (Prozac) There was a significant increase in the incidence of sexual dysfunction (i.e. delayed orgasm or ejaculation, impotence) in humans taking fluoxetine. Sexual dysfunction was positively correlated with dose. Individuals experienced substantial improvement in sexual function when the dose was diminished or the drug was withdrawn. Men showed more incidence of sexual dysfunction than women, but women’s sexual dysfunction was more intense than men’s (Gregorian et al., 2002; Hu et al., 2004; Montejo-Gonzalez et al., 1997; Montgomery et al., 2002). Fluoxetine affects sex hormone homeostasis through the elevation of prolactin levels, and a modest elevation persists during administration; however, possibly associated clinical manifestations (e.g. galactorrhea and breast enlargement) were observed (Ficicioglu et al., 1995; Haddad & Wieck, 2004; Jorgensen et al., 1996; Masala et al., 1979; Meltzer et al., 1982). Decreased ovary weight, and corpora luteal depletion and uterine atrophy were observed in females receiving fluoxetine alone (Cortes et al., 1978; Fell et al., 2004, 2005). In rat reproduction studies, there is an increase in stillborn pups, a decrease in pup weight and an increase in pup deaths following maternal exposure to fluoxetine during gestation and during both gestation and lactation (Nulman & Koren, 1996; Stanford & Patton, 1993). The effect of

476

Endocrine-disrupting chemicals in food

fluoxetine on labour and delivery in humans is unknown. Fluoxetine crosses the placenta; therefore, there is a possibility that fluoxetine may have adverse effects on the newborn (Gentile, 2005; Heikkine et al., 2002; Hendrick et al., 2003; Morisson et al., 2005; Pohland et al., 1989). In humans, the relatively slow elimination of fluoxetine (elimination half-life of 1–3 days after acute administration and 4–6 days after chronic administration) and its active metabolite, norfluoxetine (elimination half-life of 4–6 days after acute and chronic administration), leads to significant accumulation of these active species in chronic use and delayed attainment of steady state, even when a fixed dose is used. After 30 days of dosing at 40 mg/day, plasma concentrations of fluoxetine in the range of 91–302 ng/ml and norfluoxetine in the range of 72–258 ng/ml have been observed. Plasma concentrations of fluoxetine were higher than those predicted by single-dose studies, because fluoxetine’s metabolism is not proportional to dose. Norfluoxetine, however, appears to have linear pharmacokinetics. Its mean terminal half-life after a single dose was 8.6 days and after multiple dosing was 9.3 days. Steady state levels after prolonged dosing are similar to levels seen at 4–5 weeks. The long elimination half-lives of fluoxetine and norfluoxetine ensure that, even when dosing is stopped, active drug substance will persist in the body for weeks (primarily depending on individual characteristics, previous dosing regimen, and length of previous therapy at discontinuation). This is of potential consequence when drug discontinuation is required or when drugs are prescribed that might interact with fluoxetine and norfluoxetine following the discontinuation of fluoxetine (Johnson et al., 2005; Wilens et al., 2002; Young & Ashton, 1996). Fluoxetine and its metabolite norfluoxetine were detected at levels greater than 0.1 ng/g in all tissues examined from fish residing in a municipal effluent-dominated stream in North Texas, USA (Brooks et al., 2005). Fluoxetine was detected in most STP effluents and some surface water samples in the lower Great Lakes (Lake Ontario and Lake Erie), at sites near the two STPs for the city of Windsor (ON, Canada), and at sites in Hamilton Harbour (ON, Canada) (Metcalfe et al., 2003). According to a BBC report (BBC, 2004) traces of the fluoxetine can be found in the drinking water according to the UK Environment Agency.

18.10.2 Fluvoxamine maleate (Luvox) Similar to fluoxetine, fluvoxamine also cause a significant increase in the incidence of sexual dysfunction (i.e. delayed orgasm or ejaculation, impotence) in humans. Sexual dysfunction was positively correlated with dose. Individuals experienced substantial improvement in sexual function when the dose was diminished or the drug was withdrawn. Men showed more incidence of sexual dysfunction than women, but women’s sexual dysfunction was more intense than men’s (Dorevitch & Davis, 1994; Gregorian et al., 2002; Montejo-Gonzalez et al., 1997; Montgomery et al., 2002). In

Pharmaceuticals

477

another study in humans, the incidence of sexual dysfunction during fluvoxamine therapy in healthy volunteers is 35% (Nafziger et al., 1999). Fluvoxamine has an elimination half-life of 15 hours in patients with normal hepatic function. In patients with cirrhosis and the elderly, there may be as much as a 40–50% reduction in clearance and dosing should be adjusted accordingly.

18.10.3 Sertraline (Zoloft) A decrease in fertility was seen in one of two rat studies at a dose of 80 mg/ kg (four times the maximum recommended human dose (MRHD) on a mg/m2 basis) (Davies and Klowe 1998). When female rats received sertraline during the last third of gestation and throughout lactation, there was an increase in the number of stillborn pups and in the number of pups dying during the first four days after birth. Pup body weights were also decreased during the first four days after birth. These effects occurred at a dose of 20 mg/kg (1 times (i.e. the same as) the MRHD on a mg/m2 basis). The no effect dose for rat pup mortality was 10 mg/kg (half the MRHD on a mg/m2 basis). The decrease in pup survival was shown to be due to in utero exposure to sertraline. The clinical significance of these effects is unknown. There are no adequate and well-controlled studies in pregnant women. Sertraline and its metabolite desmethylsertraline were detected at levels greater than 0.1 ng/g in all tissues examined from fish residing in a municipal effluentdominated stream in North Texas, USA (Brooks et al., 2005).

18.10.4 Conclusions on serotonin reuptake inhibitors In previous studies fluoxetine hydrochloride showed clear effects on fecundity. Although environmental levels of fluoxetine hydrochloride may be low, the mere fact that its half-life is very long (days instead of hours, as it is for most pharmaceuticals) means that it can persist in the body for weeks and accumulate due to prolonged exposure. Fluvoxamine’s half-life is also relatively long and clearly has an effect on fecundity but no data on environmental levels are available. Data on sertraline also shows possible effect on fecundity but there is no data on environmental levels in the European setting. For fluoxetine, an endocrine mechanism of fecundity effects has been indicated. For both other compounds the situation is less clear, but in view of their common primary mechanism of action further study of this class of compounds is warranted.

18.11

Beta-blockers

The consumption of beta blockers in the Netherlands in DDD was estimated as given in Table 18.4.

478

Endocrine-disrupting chemicals in food

Table 18.4 The consumption of beta blockers in the Netherlands in defined daily dose (DDD) Pharmaceutical Propranolol Metoprolol Atenolol

2000

2001

2002

2003

2004

9 300 300 80 386 700 57 244 000

9 092 100 86 849 600 60 216 300

8 967 400 92 654 400 62 832 400

9 025 200 98 935 200 66 683 700

9 030 500 108 001 000 70 251 900

18.11.1 Propranolol Propranolol has been identified as having endocrine-disrupting potential as it affects both total and free testosterone (Rosen et al., 1988; el Sayed et al., 1998). Propranolol induces a significant decrease in percent of progressive motility of sperm, a significant increase in sperm head and tail abnormalities, and histopathological alterations in testis, epididymis and seminal vesicles (el Sayed et al., 1998). Studies on the binding of propranolol in rat Leydig cell cultures suggest that propranolol is capable of inhibiting testosterone synthesis in the testis (Tinajero et al., 1993). Propranolol acts in vitro as a spermicide for human sperm at a concentration of about 2 × 10−3 m (Zipper et al., 1982). Propranolol has been detected in the environment: • • • •

STP effluent, Källby, Sweden, 0.03 μg/L (Bendz et al., 2005) STP effluents survey, 0.01–0.09 μg/L (Andreozzi et al., 2003) STW effluent, UK, 0.195–0.373 μg/L (Roberts & Thomas, 2006) STW effluent, UK, median 0.076 μg/L (Ashton et al., 2004)

18.11.2 Metoprolol Metoprolol affects both total and free testosterone levels (Rosen et al., 1988; el Sayed et al., 1998). Metoprolol induces a significant decrease in percent of progressive motility of sperm, a significant increase in sperm head and tail abnormalities, and histopathological alterations in testis, epididymis and seminal vesicles (el Sayed et al., 1998). Metoprolol has been detected in surface water up to concentrations of 0.100 μg/L (Stolker et al., 2004). Metoprolol has been detected in the environment: • STP effluent, Källby, Sweden, 0.19 μg/L (Bendz et al., 2005) • STP effluents survey, 0.08–0.39 μg/L (Andreozzi et al., 2003) • River Saale, >0.100 μg/L (Wiegel et al., 2004)

18.11.3 Atenolol Atenolol affects both total and free testosterone levels (Rosen et al., 1988; el Sayed et al., 1998). Atenolol induces a significant decrease in percent of progressive motility of sperm, a significant increase in sperm head and tail abnormalities, and histopathological alterations in testis, epididymis and

Pharmaceuticals

479

seminal vesicles (el Sayed et al., 1998). Atenolol causes a significant reduction in testosterone release by rat Leydig cells (Fogari et al., 2002; Khan et al., 2004). Atenolol has been detected in the environment: • STP effluent, Källby, Sweden, 0.16 μg/L (Bendz et al., 2005) • Several STP effluents, Italy, 0.466 μg/L (Zuccato et al., 2005)

18.11.4 Conclusions on beta-blockers All three beta-blockers are endocrine modulators as they affect testosterone levels and spermatogenesis. However, metoprolol and atenolol are consumed in far higher quantities than propranolol. In addition, all three compounds have been detected in the environment.

18.12

Steroid contraceptives

General Table 18.5 is a summary of the steroids currently used as contraceptives. Table 18.6 is a summary of the combined oral contraceptive preparations available on the market. Pharmacological effects of progestins in oral contraceptives A number of pharmacological effects contribute to the contraceptive effects of progestins. These include inhibiting ovulation by suppressing the function

Table 18.5

Contraceptive progestins

Class compound

Name

19 Nor-testosterone progestins Estranes Norethindrone Norethindrone acetate Ethynodiol diacetate Gonanes Levonorgestrel Norgestrel Norgestimate Desogestrel Gestodene Pregnane progestins Megestrol acetate Medroxyprogesterone acetate

Relative progestational activity (arbitrary units)

Relative androgenic activity (arbitrary units)

1 1.2 1.4 5.3 2.6 1.3 9 12.6

1 1.6 0.6 8.3 4.2 1.9 3.4 8.6

0.4 0.3

0 0

480

Endocrine-disrupting chemicals in food

Table 18.6 Available combination oral contraceptives

50 mcg estrogen

Name

Progestin (mg)

Type of estrogen (mcg)

Ogestrel/Ovral Necon/Nelova/ Norethin/Norinyl/ Ortho-Novum 1/50 Ovcon 50 Norlestrin 1/50

Norgestrel (0.5) Norethindrone (1.0)

EE (50) Mestranol (50)

Norethindrone (1.0) Norethindrone acetate (1.0) Ethynodiol diacetate (1.0)

EE (50) EE (50)

Demulen 50/Zovia 1/50 <50 mcg estrogen plus monophasic

Lo-Ovral/ Low-Ogestrel Ovcon 35 Desogen/Ortho-cept Levlen/Levora/ Nordette Ortho-Cyclen Necon/Nelova/ Norinyl/Norethrin/ Ortho-Novum 1/35 Mircette Brevicon/Modicon/ Necon/Nelova 0.5/35 Loestrin 1.5/30 Alesse/Levlite Loestrin 1/20 Demulen/Zovia 1/35

<50 mcg estrogen plus multiphasic

Ortho-Novum 7/7/7 Tri-Levlen/Triphasil/ Trivora Jenest Necon/Nelova/ Ortho-Novum 10/11 Ortho Tri-Cyclen Tri-Norinyl Estrostep

EE (50)

Norgestrel (0.3)

EE (30)

Norethindrone (0.4) Desogestrel (0.15) Levonorgestrel (0.15)

EE (35) EE (30) EE (30)

Norgestimate (0.25) Norethindrone (1.0)

EE (35) EE (35)

Desogestrel (0.15) Norethindrone (0.5)

EE (20) EE (35)

Norethindrone acetate (1.5) Levonorgestrel (0.1) Norethindrone acetate (1.0) Ethynodiol diacetate (1.0)

EE (30)

Norethindrone (0.5, 0.75, 1.0) Levonorgestrel (0.05, 0.075, 0.125) Norethindrone (0.5, 1.0) Norethindrone (0.5, 1.0) Norgestimate (0.18, 0.215, 0.250) Norethindrone (0.5, 1.0, 0.5) Norethindrone acetate (1.0, 1.0, 1.0)

EE (20) EE (20) EE (35) EE (35, 35, 35) EE (30, 40, 30) EE (35, 35) EE (35, 35) EE (35, 35, 35) EE (35, 35) EE (20, 30, 35)

Pharmaceuticals

481

of the hypothalamic–pituitary–ovarian (HPO) axis; modifying the subsequent pituitary surge of LH and FSH; slowing transport of the ovum through the Fallopian tubes, which limits the time available for fertilisation; thickening cervical mucus, which impedes sperm transit; and inhibiting the activation of spermatic enzymes required for ovum penetration (capacitation). Thus, the primary mechanism of oral contraceptives defines these compounds as endocrine disruptors. Family tree of contraceptive progestins Synthetic progestins used in oral contraceptives can be classified as those that are structurally related to progesterone or testosterone. Progestins structurally related to progesterone include progesterone itself and medroxyprogesterone acetate compounds that have 21 carbons. Progestins structurally related to testosterone are structural derivatives of testosterone and are not synthesised from testosterone. Removal of the methyl group from the testosterone molecule produces norethindrone, a compound with high progestational activity, high oral activity, and almost no androgenicity. Adding an additional methyl group forms an ethyl group and produces the compound norgestrel, which has even greater progestational activity than norethindrone. Norgestrel is synthesised chemically into dextro-norgestrel, an inactive form, and levonorgestrel, the active form. Another classification of progestins uses the terms gonane or estrane and is based on the number of carbons: gonanes have 17 carbons, and estranes have 18 carbons. A family tree of contraceptive progestins is presented in Table 18.7. Biologically active forms of progestins When assessing a contraceptive progestin, several factors need to be considered. The first consideration is whether the progestin is in active form or needs to be converted. Some progestins are prodrugs that must be converted to biologically active forms. The next is the progestin’s affinity for human tissues, including inhibition of ovulation and binding affinity to human receptors. The third consideration is the pharmacokinetic profile, including half-life and bioavailability. The clinical relevance of animal data compared with human data should also be assessed. Five estrane progestins are in commercial use. Three of these – norethindrone acetate, ethynodiol

Table 18.7

Family tree of contraceptive progestins

Gonanes (levonorgestrel family)

Estranes (norethindrone family)

Levonorgestrel Desogestrel Norgestimate Gestodene

Norethindrone Norethindrone acetate Ethynodiol acetate Lynestrenol

482

Endocrine-disrupting chemicals in food

diacetate, and lynestrenol – are prodrugs. Before these three can exert progestational activity, they must undergo biochemical conversion to norethindrone, their biologically active form (Stanczyk & Roy, 1990). Levonorgestrel and gestodene are gonane progestins that are active in their current forms. Desogestrel and norgestimate are prodrugs that must undergo biochemical conversion in the liver. Desogestrel is transformed to 3keto-desogestrel, which is its only active form, whereas norgestimate is converted to levonorgestrel and levonorgestrel-3-oxime, which are its active forms (Stanczyk, 1997). Bioavailability of progestins The extent to which a contraceptive progestin enters the circulation without undergoing hepatic metabolism determines its bioavailability. There is a great deal of interindividual variability in the bioavailability of contraceptive progestins. The range goes from gestodene (>90%) and levonorgestrel (∼90%) to the metabolites produced by norgestimate (<25%). Norethindrone and 3-keto-desogestrel (active form of desogestrel) are in the intermediate range at approximately 64% and 62%, respectively (Back et al., 1978, 1981; Orme et al., 1991; Stanczyk & Roy, 1990). Serum half-lives Serum half-lives of contraceptive progestins are not absolute values, but change depending on whether women receive progestin only or oral contraceptives which additionally contains an estrogenic compound. Levonorgestrel has been shown to have the longest half-life of 15 hours. Both 3-keto-desogestrel (the active form of desogestrel) and gestodene have half-lives of 12 hours, and the half-life of norethindrone is 7 hours (Fotherby & Caldwell, 1994). All progestins were given in combination with ethinyl estradiol (30–35 μg). Plasma levels Plasma levels of norethindrone (1000 μg dose) and levonorgestrel (150 μg dose) after a single oral dose indicate that a considerably higher level of norethindrone (about 14 ng/mL) occurs within the first hour as compared with levonorgestrel (about 2 ng/mL). However, levels of norethindrone fall precipitously to undetectable levels – below 1 ng/mL at 24 hours compared to levonorgestrel, which is still detectable at 48 hours (Stanczyk, 1994). Relative binding affinities for human uterine progesterone receptor In vitro studies of uterine progesterone receptor binding of progestins give a range of relative binding affinities (RBAs), depending on the species studied, various study parameters and compounds used for comparison. Compounds such as levonorgestrel (LNG) have a very high affinity for the human uterine progesterone receptor, as does 3-keto-desogestrel (3-

Pharmaceuticals

483

keto-DSG), levonorgestrel-17-acetate (LNG-17-acetate) and gestodene (GSD). Two prodrugs, desogestrel (DSG) and norgestimate (NGM), do not bind to the human uterine progesterone receptor. Among the norgestimate metabolites, levonorgestrel-3-oxime (LNG-3-oxime) has a very low RBA for human uterine progestin receptors, even though serum levels may be high. LNG-17-acetate, however, has substantial progestational activity, but is barely detectable in serum following administration of norgestimate (Juchem et al., 1993). Dose/ovulation inhibition dependence In studies looking at various progestins combined with 30–35 μg ethinylestradiol (EE), the progestin dose needed for ovulation inhibition varied widely from high doses for norethindrone (approximately 400 μg per day) and norgestimate (200 μg per day), to levonorgestrel and desogestrel (60 μg per day), to smaller doses for gestodene (approximately 30 μg per day) (Teichmann, 1996). Effect of oral contraceptives on sex hormone binding globulin/testosterone The results are presented (Van der Vange et al., 1990) of a study comparing seven oral contraceptives with regard to their effect on SHBG, total testosterone (total T), and free testosterone (free T). The oral contraceptives all contained EE (30–40 μg per dose) but different types and doses of progestin. In this study, the increases in SHBG were extremely variable, and total T varied to a lesser degree. (One oral contraceptive, CPA 2000 CPA μg/EE 35 μg, actually caused total T to increase.) Despite these variations, all the oral contraceptives reduced free T to a similar degree. A decrease in free T is considered the most important factor when evaluating the effect of oral contraceptives on acne and other androgenic conditions. Oral contraceptive effects on androgens The effects of two 20 μg EE oral contraceptives, LNG 100 μg and norethindrone acetate (NETA) 1000 μg, were compared (Thorneycroft et al., 1999) on androgen levels and acne lesion counts. Patients were evaluated at baseline and during cycle 3 (days 17 to 21) for androgen and SHBG levels, acne lesion count and weight. Results demonstrated that, among the 41 evaluable women at the end of the study, there were statistically significant reductions in all measured androgen levels. At the end of three cycles, both 20 μg EE formulations decreased androgens and increased SHBG from baseline, although the oral contraceptive with NETA increased the mean SHBG more than the oral contraceptive with LNG. Compared with the formulation consisting of EE and LNG, the formulation consisting of EE and NETA was associated with two times greater relative increase in SHBG. At the same time, the formulations had equivalent decreases in bioavailable testosterone.

484

Endocrine-disrupting chemicals in food

Oral contraceptive changes in biochemical markers of androgenicity Changes in biochemical markers of androgenicity were studied in 58 young women (>14 years old) randomised to placebo (n = 29) or a low-dose oral contraceptive, EE 20 μg/LNG 100 μg (n = 29). Mean percentage changes from baseline were determined at the end of cycles 4 and 6. Statistically significant (P < 0.05) reductions were noted in 3α-androstanediol glucuronide (3α-diol G), as well as marked reductions in the treatment group in androstenedione (A), androsterone glucuronide (AG), dihydrotestosterone (DHT), total testosterone (TT) and dehydroepiandrosterone sulphate (DHEAS), although the reductions were not statistically significant. Statistically significant reductions in the oral contraceptive group were observed for A, AG and 3α-diol G vs. increases with placebo. The oral contraceptive significantly decreased androgen levels in ovarian (A, TT) and peripheral (3α-diol G) compartments as compared to placebo (Stanczyk et al., 2000). Adverse effects of progestins Adverse effects of progestins are reviewed for individual compounds below.

18.12.1 Ethynodiol diacetate Following oral administration of ethynodiol diacetate plus mestranol to mice, increased incidences of pituitary tumours were observed in animals of each sex. Ethynodiol diacetate plus ethinyloestradiol was tested for carcinogenicity by oral administration to mice and rats. In mice, it induced increased incidences of pituitary tumours in animals of each sex and of malignant tumours of connective tissues of the uterus. In rats, malignant mammary tumours were produced in animals of each sex (IARC Monographs, 1979a).

18.12.2 Norethindrone, norethindrone acetate Aneuploidy was observed in oocytes of mice treated with high doses of norethindrone acetate. In a test for dominant lethal mutations in which female mice were exposed orally to norethindrone acetate, no increase was seen in one strain of mice, and a second strain showed an increase only when females were mated within two weeks of treatment. However, the compound did not induce aneuploidy or chromosomal aberrations in cultured human lymphocytes. Neither norethindrone nor its acetate was mutagenic to bacteria (IARC Monographs, suppl. 1987). Norethindrone and its acetate were tested by oral administration in mice and rats, and by subcutaneous implantation in mice. In mice, norethindrone and its acetate increased the incidence of benign liver-cell tumours in males; norethindrone increased the incidence of pituitary tumours in females and

Pharmaceuticals

485

produced granulosa-cell tumours in the ovaries of females. Norethindrone increased the incidence of benign liver-cell tumours and benign and malignant mammary tumours in male rats (IARC Monographs, 1979b). Rats fed 3–4 mg/kg body weight (bw) per day norethindrone acetate (about 100 times the daily human dose) for two years had an increased incidence of neoplastic nodules of the liver; an increase in the incidence of uterine polyps was seen in females (Schardein, 1980). In rats given weekly intramuscular injections for 104 weeks of norethindrone enanthate at doses of 10, 30 and 100 mg/kg bw (20, 60 and 200 times the daily human contraceptive dose), there was a dose-related increase in pituitary gland tumours in males, whereas in females no effect on pituitary glands was observed with the lowest dose and a reduction in pituitary tumours was observed with the highest dose. Benign mammary tumours were observed in males at all doses, but there was little effect in females; the incidence of malignant mammary tumours was greatly increased in both males and females given the two higher dose levels and was dose-related. A dose-related increase in the incidence of liver tumours was also seen in animals of each sex (El Etreby & Neumann, 1980).

18.12.3 Levonorgestrel The combination of T plus LNG suppressed sperm production much more than T alone (Bebb et al., 1996). Some 67% of the T plus LNG group achieved azoospermia (33% for T alone group). Severe oligospermia developed in 94% of the T plus LNG group compared with the 61% T alone group. T plus LNG also suppressed sperm production more rapidly than T alone. Time to azoospermia was 9.9 ± 1.0 vs. 15.3 ± 1.9 weeks in the T plus LNG and T alone groups, respectively (mean ± SEM; P < 0.05). Consumption of Levonorgestrel The consumption of defined daily doses of estrogens with progestogens in fixed proportions in the Netherlands in 2000–2004 is estimated as given in Table 18.8. The table shows that the use of Levonorgestrel in contraceptives in Netherlands is higher than the use of other progestogens. This combined with longer half-life implies a possibility of higher environmental concentrations.

18.12.4 Conclusions on progestins From the above-compiled data on progestins, Levonorgestrel seem to be the most relevant compound for further analysis, since in experimental studies it showed significant effects on fecundity. Though there are no reported results on presence in the environment, we believe that this is mainly due to lack of attempts for detection of this compound in the environment. We must take into account that the amount of Levonorgestrel

486

Endocrine-disrupting chemicals in food

Table 18.8 The consumption of defined daily doses (DDD) of estrogens with progestogens in fixed proportions in the Netherlands in 2000–2004 Pharmaceutical Estrogen with lynestrenol (Ministat ®) Estrogen with norethisteron (Neocon ®) Estrogen with levonorgestrel (Microgynon ®) Estrogen with desogestrel (Marvelon ®) Estrogen with gestodeen (Meliane ®) Estrogen with norgestimaat (Cilest ®) Estrogen with drospirenon (Yasmin ®) Estrogen with norelgestromine (Evra ®) Total

2000

2001

2002

2003

17 925 900

15 678 100

14 017 600

13 563 700

914 490

5 501 200

5 458 900

5 424 400

5 471 800

596 630

254 781 700

275 320 400

298 341 600

318 354 500

82 111 800

139 903 000

125 862 600

104 173 300

94 577 100

5 761 400

59 303 500

59 994 700

51 045 000

47 089 300

5 214 500

5 327 500

5 057 000

4 440 000

4 118 900

345 290

5 072 400

9 349 400

9 809 000

3 216 800

109 740

71 101

493 094 040

98 232 011

482 742 800

492 444 100

486 791 300

2004

used is high, which suggests a high likelihood for presence in the environment, especially wastewater treatment plants, and that the half-life is relatively long, which suggests a possibility for bioaccumulation.

18.12.5 Ethynylestradiol The synthetic analogue of 17β-estradiol, 17α-ethynylestradiol (EE2), is a potent xenoestrogen and the most widely used estrogenic component of modern oral contraceptive preparations in the world. Around 95–98% of plasma EE2 is bound, virtually all to albumin (Akpoviroro et al., 1981). The affinity to the hormone-binding proteins in serum, including SHBG, albumin and α-fetoprotein strongly influence the in vivo estrogenicity of a compound. Only the unbound fraction, less than 5% of the total, is considered biologically active (Orme et al., 1983). It is estimated that about 98% of the endogenous 17β-estradiol is bound to binding proteins, especially SHBG, resulting in only a small percentage available to the cells (Ben Rafael et al., 1986).

Pharmaceuticals

487

The major site of metabolism for EE2 is the liver, with the two major metabolic pathways being 2-hydroxylation and 16β-hydroxylation. These pathways result in a number of metabolites, which are then conjugated with glucuronide and/or sulphate, and are considered to be biologically inactive (IARC Monographs, 1979a). However, a major portion of EE2 is conjugated directly with glucuronic acid and excreted in the urine. In contrast to the metabolites of natural estrogens, a significant proportion of the metabolites of EE2 are excreted by the faecal route. In radiolabelled studies, the ratio of faecal/urine radioactivity has been reported to be about 4 : 6, and the total recovery of radioactivity from both sources is about 90%. One study reported that about 30% is excreted in the faeces, of which one-third is excreted as the unchanged form (which may be a result of deconjugation in the colon). The remainder is excreted in the urine mainly as the EE2 glucuronide conjugate. Other glucuronide (and to a lesser extent sulphate) conjugates include: 2-hydroxyoestradiol; 2-methoxyethynyloestradiol and 3-methoxy-2-hydroxy-ethynylestradiol. It has been reported that only 1% of unchanged EE is excreted in the urine, although a higher value of 16% has also been reported. De-ethynylated estrogens (e.g. estrone, 17β-estradiol and estriol) only account for 1–2% of the dose in women (Orme et al., 1983). Effects of 17a-ethynylestradiol related to fecundity EE2 is used as an oral contraceptive in humans and as such its effects on fecundity in humans is well known. The effects of EE2 on animals has been investigated in several studies and some results are given below. These studies may give some indication of what the no observed effect concentration (NOEC) and lowest observed effect concentration (LOEC) in humans might be. The relative estrogenic potency (REP) of EE2 compared with 17β-estradiol and estimated using in vitro assays has been stated to be between 0.5 and 5.71 (Tanaka et al. (2001), REPEE2 = 0.5; Korner et al. (2001), REPEE2 = 0.91; Gutendorf and Westendorf (2001), REPEE2 = 1.25, 1.25 and 5.71). The effects of exposure to EE2 upon the reproductive success of a marine fish was investigated recently (Robinson et al., 2003). Sand goby (Pomatoschistus minutus) were exposed for seven months to EE2 or a sewage effluent containing known xeno-estrogens (alkylphenol polyethoxylates) and bred using within treatment crosses. Nominal exposure concentrations were 6 ng/L EE2, 0.3 or 0.03% v/v sewage effluent. At the end of the breeding trials, expression of hepatic zona radiata protein (Zrp) and vitellogenin (Vtg) mRNA were determined. Exposure to 6 ng/L EE2 induced Zrp and Vtg mRNA expression in male and female sand goby, impaired male maturation and reproductive behaviour, reduced female fecundity and reduced egg fertility. As a consequence, fertile egg production of the EE2-exposed population was reduced by 90%. Exposure to sewage effluent (0.3% v/v) increased adult mortality and female Zrp and Vtg

488

Endocrine-disrupting chemicals in food

mRNA expression, but did not induce male vitellogenesis. Exposure to EE2 and 0.3% v/v sewage effluent impaired development of the male urogenital papilla. Fish exposed to 0.03% v/v sewage effluent produced more fertile eggs than those exposed to 0.3% effluent, or those receiving no effluent. In another study (Berg et al., 1999) two synthetic estrogens, diethylstilbestrol (DES) and EE2, were injected into the yolks of embryonated eggs. At a dose as low as 2 ng EE2/g egg, all male embryos became feminised, containing ovary-like tissue in the left testis. The extent of feminisation of the testes was determined by measuring the relative area of the ovary-like component. Persistent Mullerian ducts oviducts in male embryos, and malformations of the Mullerian ducts in females occurred at 2 ng EE/g egg and higher doses. DES was approximately one-third to one-tenth as potent as EE2. The morphological changes studied were dose-dependent, indicating that they are useful as test end points for estrogenic activity. Feminisation of the left testis in males proved to be the most sensitive end point. Papoulias et al. (1999) evaluated the effects of a model environmental estrogen, EE2, on the Japanese medaka (Oryzias latipes, a freshwater fish) using a nano-injection exposure. Gonad histopathology indicated that a single injection of 0.5–2.5 ng EE2/egg can cause phenotypic sex-reversal of genetic males to females. Sex-reversed males had female-typical duct development and the secondary sex characteristics were generally consistent with phenotype. No instances of gonadal intersexes were observed. EE2 also appeared to reduce growth but not condition (weight-at-length) and exposed genetic females appeared to have a higher incidence of atretic follicles relative to controls. The results suggested that EE2 may influence sexual differentiation and development. Scholz and Gutzeit (2000) exposed freshly hatched Japanese medaka (Oryzias latipes) for two months to nominal EE2 concentrations of 0, 1, 10 or 100 ng/L under semi-static conditions. The exposure period was followed by a six week recovery period in order to detect long-lasting effects on sexual differentiation. Sex ratio, gonadal growth, spawning, fecundity, histology as well as ovarian gene expression of aromatase was monitored. Growth was unaffected in all treatment groups. At 100 ng/L, all genetically male medaka were sex reversed and had developed an ovary. At lower test concentrations, no alteration of testicular structure was detected (including testis–ova or ovarian-like structures) and male fertility appeared to be unchanged. In genetic females, significantly reduced ovarian weight was observed at 10 and 100 ng/L as well as a significantly decreased egg production rate. There was a 80% reduction in egg production at 10 ng/L and complete inhibition occurred at the highest test concentration, likely to be caused by the absence of males. Aromatase, which is normally only expressed in ovaries, was also detectable in testis of genetic males exposed to 10 ng/L. A full life-cycle study with fathead minnow (Pimephales promelas) revealed a variety of effects on survival, growth, gross development, gonad development, sex determination and reproductive maturity (Länge et al.,

Pharmaceuticals

489

2001). Newly fertilised embryos (<24 h old) were exposed to nominal concentrations of 0.2, 1, 4, 16 and 64 ng/L in flow-through conditions at 25 ± 1 °C for 305 days (four days pre-hatch and 301 days post-hatch). Exposure concentrations were confirmed by radioimmunoassay analysis and ranged from 58 to 84% with mean measured values ≥70%. Hatching success of embryos was not significantly different from controls at any exposure concentration (NOEC >64 ng/L). Larval growth was reduced at 16 ng/L and a NOEC of 4 ng/L identified at day 28. In addition, juvenile fish growth was reduced when sampled at days 28 and 56 and NOEC and LOEC values of 1 and 4 ng/L were reported, respectively. Gross morphological changes were seen in fish at test concentrations of 16 and 64 ng/L. No males (with appropriate secondary sexual characteristics and territorial behaviour) were seen after 172 days post-hatch at a concentration of 4 ng/L or above. Histology of exposed fish at 56 days post-hatch revealed a female : male sex ratio of 84:5 (with ova–testes in 11% of fish) at a concentration of 4.0 ng/L. No significant effects were seen at lower test concentrations. After 172 days post-hatch, no testicular tissue was observed in any fish exposed to 4 ng/L. Thus the NOEC and LOEC values based on gonad histology were 1 and 4 ng/L, respectively. There are several other studies reported in the open literature where similar results to the above are reported (Metcalfe et al., 2001; Nash et al., 2004; Wenzel et al., 2001; Zillioux et al., 2001). Production volume and use of 17a-ethylnylestradiol EE2 can be used in human medicine to treat various gynaecological disorders and post-menopausal breast cancer. However, its largest use is in oral contraceptives, when it is usually administered in combination with a synthetic progestin. Its concentration in the contraceptive pill ranges from 20 to 50 μg, with 35 μg most commonly prescribed (Archand-Hoy et al., 1998). An annual use of 0.029 tonnes of EE2 has been estimated in the UK (Webb, 2000). By comparison, it has been estimated that 0.088 tonnes of oral contraceptives (EE2 and mestranol) are used annually in the US (Archand-Hoy et al., 1998). Persistence of 17a-ethylnylestradiol in the environment Synthetic estrogens (EE2 and mestranol) are more resistant to microbial degradation than natural steroids (estradiol, estrone, estriol). The data on the physicochemical properties of EE2 and its environmental fate (Table 18.9) indicate that the compound is relatively persistent in the aquatic environment. It is likely that adsorption of EE2 to soil is a major removal process (log Koc = 3.8). It has been reported that EE2 is highly stable and is not sufficiently eliminated during biological treatment of the wastewater (Ternes et al.,

490

Endocrine-disrupting chemicals in food

Table 18.9 Physicochemical properties and environmental fate data Physicochemical property

Value

Physical state at ambient temperature Water solubility Octanol–water partition coefficient (log Kow) Organic carbon water partition coefficient (log Koc)

Solid 4.7–19 mg/L1 3.62–4.7 3.8 (4.5)

Type of degradation Aquatic–abiotic Aquatic–biotic Terrestrial

Sorption is the major removal process with photolysis being of lower importance and volatilisation being negligible A number of laboratory studies have indicated that EE2 is relatively persistent No data are available on the persistence of EE2 in soil though it is likely that adsorption to soil is a major removal process

1999). Several studies have examined the persistence of EE2 in rivers (Jurgens et al., 2002), activated sewage and wastewater effluent. After the primary treatment of the sewage, the mean percent of remaining EE2 is equal to 75%. After the secondary treatment, approximately 65% remaining EE2 was detected (Tabak et al., 1981). These values correspond to the study of EE2 removal rates in French STPs (mean removal rate – 40%, Cargouet et al., 2004). In this study, it was estimated that EE2 accounted for 35–50% of the estimated estrogenic activity in rivers. Close persistency values were observed by Kuch and Ballschmiter, 2000 (43%). The low removal rate of EE2 can be explained by its slow microbial degradation during the treatment process (Tabak et al., 1981; Ternes et al., 1999). In addition, EE2 concentrations in the STPs can be increased by the partial conversion of other drugs into this molecule (Kuhnz et al., 1997). Additionally, reports from laboratory biodegradation studies (Desbrow et al., 1998) indicated that EE2 was highly stable and persistent in activated sludge, with no detectable degradation occurring after 120 h of treatment. The solubility of EE2 in pure water and sewage treatment water was reported to be 4.2 and 4.7 mg/L, respectively, which was three-fold less soluble than natural steroidal estrogens. This fact is believed to contribute to the increased resistance of EE2 to biodegradation as compared with natural steroidal estrogens. According to Ying et al., 2002, EE2 was principally persistent under selected aerobic conditions. Comparatively, 70–80% of added estradiol (E2) was mineralised to CO2 within 24 h by biosolids from wastewater treatment

Pharmaceuticals

491

plants, whereas the mineralisation of EE2 was 25–75-fold less. EE2 was also reported to be degraded completely within six days by nitrifying activated sludge and resulted in the formation of hydrophilic compounds. EE2 was found to be microbially degraded (Colucci and Topp, 2001). According to the studies, the dissipation half-life of EE2 ranged from 7.7 days at 4 °C to 3 days at 30 °C. Exposure routes There are several exposure routes that may lead to contamination of food or water with EE2. If we neglect routes 3 and 4 (see Fig. 18.1) related to effects which would exist just in the vicinity of the production plant and should be negligible if the necessary precautions are in operation at the plant, then the most likely exposure route would be route 1 in Fig. 18.1. After the human usage part of the EE2 would either be discarded and end up in landfills (route 5) or would end up in sewage through human excreta (route 6) and from there would enter an STP. EE2 can be further transported from a landfill as effluent through the landfill effluent treatment system, and from there into the sanitary sewage or STP (route 8), or could be released into surface waters or land (route 13), depending on the level of treatment applied to it. Landfill leachate can percolate the containment system and pollute soil and groundwater (route 15); however, this exposure route should not represent a significant threat to the environment in a well-designed and maintained landfill and therefore will not be considered further. The sewage sludge from a STP, among other options like incineration for example, may be disposed into a landfill, route 11, or could be used in agriculture, route 16. The EU Directive which regulates the use of sewage sludge in agriculture is 86/278/EEC. However, this Directive does not mention endocrine disrupters such as EE2 or E2, and therefore tests for such EDs are not required. From the agricultural fields the endocrine-disrupting chemicals (EDCs) may be transported to surface waters, soil and groundwater by leaching, dissipation and run-off, route (18). If EE2 is transported to surface waters it may end up in the food chain by bioacumulation in fish or as water for domestic use, and if it reaches groundwater it may further be used as tap water for human consumption. Once sewage sludge is applied to agricultural fields, EE2 may end up in plants through plant uptake. However, Directive 86/278/EEC instructs that sludge must not be applied to soil in which fruit and vegetable crops are growing or grown, or less than 10 months before fruit and vegetable crops are to be harvested. Grazing animals must not be allowed access to grassland or forage land less than three weeks after the application of sludge. Unless direct measurements show otherwise, the risks from EE2 being present in the air due to evaporation from landfills, STPs and agricultural fields where sewage sludge is applied, will be considered to be negligible.

492

Endocrine-disrupting chemicals in food

Evidence for presence of 17a-ethylnylestradiol in the environment Several studies were examining the presence of EE2 in STPs in raw sewage as well as effluent. In a study by Stumpf et al. (1996) EE2 was detected in all 20 STPs investigated above the quantification level of 1 ng/L and in 15 effluents >10 ng/L. The median concentration of EE2 was 17 ng/L and the maximum 62 ng/L. Similar results have been found in the UK where concentrations in effluents were up to 7 ng/L EE2 (Aherne and Briggs, 1989; Desbrow et al., 1998). Belfroid et al. (1999) reported that EE2 was detected at one occasion in three and two STP effluent samples in Netherlands, respectively. The data also showed that concentrations of all hormones were higher in domestic effluents than in industrial effluents. Ternes et al. (1999) reported that in the raw sewage of the Brazilian STP of Penha/Rio de Janeiro, the natural estrogens 17β-estradiol and estrone were detected with average concentrations of 0.021 and 0.040 μg/L, respectively. In the German municipal STP close to FrankfurtrMain the raw sewage was contaminated by E2 and estrone with average concentrations of 0.015 and 0.027 μg/L, respectively. The evaluated removal rates were much lower than those obtained in the Brazilian STP. For instance, the loads of estrone and EE2 were not appreciably reduced while passing through the German STP. Considering the standard deviation no elimination rate could be evaluated. The differences between the absolute removal rates of the German and Brazilian STP might be caused by the low temperatures in the German sampling period with −2 °C on average compared to above 20 °C in Rio de Janeiro. E2 and 16α-hydroxyestrone were eliminated with a higher efficiency than EE2 and estrone. In German STPs median values could be evaluated for EE2 in the range of 1 ng/L (detection limit). In comparison, the concentrations of EE2 were higher in Canadian effluents compared with those determined in the German STP effluents (median: 9 ng/L). In a study performed by Larsson et al. (1999), the effluent from STP in Sweden was analysed. The results revealed significant levels of estrogenic substances in sewage effluent water (4.5 ng/L for EE2). The steroids were mainly present in unconjugated form. Since humans primarily excrete both natural estrogens and EE2 as conjugates (Ranney, 1977), these results suggest that deconjugation (activation) occurs within the sewage system, and/or that the conjugates are more rapidly degraded. The ratio between EE2 and natural estrogens in the water is higher than the theoretical ratio based on human secretion rates of natural and synthetic estrogens (von Rathner and Sonneborn, 1979), indicating a faster degradation of the natural estrogens. Larsson et al. (1999) reported that the bile of fish caged downstream of the STP contained estrogenic substances at concentrations 104– 106 times higher than water levels. The estimated EE2 concentration in the creek was 1.5 ng/L during the experiment, taking into account the flow rate through the STP and the dilution in the creek, while the EE2 concentration

Pharmaceuticals

493

in bile from caged juvenile rainbow trout exposed to diluted sewage effluent water for four weeks showed concentrations of EE2 of approximately 1 μg/g bile. This shows that exposure to different environmental estrogens results in accumulation of prominent amounts of these substances. A recent study by Cargouet et al. (2004) showed that the concentration of EE2 in STP influents in Paris area has a mean value ranging from 4.9 to 7.1 ng/L, which represents 11–15% of the total detected steroids. All these results show that EE2 is present in the raw sewage as well as effluent of STPs irrespective of the country were the tests are performed. Once released in the surface waters/rivers the effluent is diluted and depending on the extent of dilution EE2 may be detectable or not. In the UK, immunoassay detection revealed the presence of EE2 in rivers in concentrations below 5 ng/L in September 1982 and 2–15 ng/L in August 1987 (Aherne and Briggs, 1989). In Germany, in the Ruhr district, EE2 has been detected in surface water in concentrations between <1 and 4 ng/L (Stumpf et al., 1996). Caged fish held downstream of most STW produced vitellogenin, indicating the presence of estrogenic substances (Harries et al., 1996, 1997; Purdom et al., 1994). The nature of the inducer(s) was, however, not clearly elucidated. Hohenblum et al. (2004) monitored surface waters in Austria for EE2 and some other compounds, and found EE2 in four samples with maximum concentration of 0.33 ng/L. Vogel et al. (2003) studied continuous infiltration experiments over a period of two years and run-off experiments in order to investigate the behaviour of EDCs in agricultural soils after sewage sludge application. In infiltration experiments transport of EE2 towards lower soil layers was observed. They did not detect considerable EE2 concentrations in the leachate, leading them to the conclusion that adsorption to the soil matrix and/or biodegradation prevent in some cases a direct EE2 transport to groundwater. However, since the experimental conditions were very specific (groundwater table >90 cm below ground surface, high soil organic matter) infiltration of EE2 to the groundwater under certain conditions cannot be ruled out. Recent studies have shown that disposal of animal manure to agricultural land could lead to movement of estrogenic steroids into surface and groundwater (Peterson et al., 2001). Peterson et al. (2001) measured 17β-estradiol concentrations ranging from 6 to 66 ng/L in mantled karst aquifers in northwest Arkansas. The observed 17β-estradiol concentration trends imitated the changes in stage over the recharge event. The contamination was associated with poultry litter and cattle manure waste applied on the area. Hohenblum et al. (2004) detected EE2 in one sample of 112 tested, though 17β-estradiol was detected in about half the samples. This study supports the findings of Vogel et al. (2003) that the adsorption to the soil matrix and/or biodegradation prevent a direct EE2 transport to

494

Endocrine-disrupting chemicals in food

groundwater. The maximum concentration of E2 found was 0.79 ng/L while EE2 was 094 ng/L.

18.12.6 Conclusions on 17a-ethylnylestradiol There is awareness of the importance of investigating the presence of EE2 in the environment, and many of the previous studies included EE2 as the main representative of the synthetic steroids. As can be seen from the above studies, evidence exists of the presence of EE2 in the environment. EE2 was found basically in every medium where an attempt was made for its detection, i.e. raw sewage, STP effluent, rivers and even groundwater. We are not aware of any attempt being made to detect EE2 in landfills, landfill effluent or agricultural fields where sewage sludge is applied. The detection of EE2 is difficult since its concentrations in the environment are usually on the detection limit, but this certainly does not mean that the risk from EE2 is negligible. EE2 can cause changes in animals in very low concentrations. Chronic exposure under laboratory conditions, including studies of chronic exposure over two complete generations, to as little as 1 ng/L EE2 (below the limits of chemical detection for most effluents) was sufficient to sex reverse male zebrafish and 1.5 ng/L stimulated vitellogenesis in juvenile fish (Orn et al., 2003; Hahlbeck et al., 2004). Bioaccumulation, as shown by Larsson et al. (1999), can increase the EE2 concentrations by several orders of magnitude. One should also not forget that EE2 concentrations in the STPs can be increased by the partial conversion of other drugs into this molecule (Kuhnz et al., 1997). Finally, several studies have shown that EE2 is more persistent in the environment than the natural estrogens. All the above facts make EE2 a compound of major interest for further study.

18.13

Antibiotics

The consumption of antibiotics in the Netherlands in DDD is estimated as given in Table 18.10.

18.13.1 Sulfamethoxazole–trimethoprim combination Trimethoprim is a folic acid antagonist. As such, it can cause abnormal embryo development in experimental animals (Helm et al., 1976). A role of trimethoprim therapy in human birth defects has not been established. Treatment with a sulfamethoxazole/trimethoprim combination causes a drop in sperm concentration between 7 and 88%. Possible mechanisms for this effect is folate deprivation of spermatogenic cells through the inhibitory action of trimethoprim on dihydrofolate reductase (Murdia et al., 1978). A decrease in sperm concentration and total number of sperm has been

Pharmaceuticals

495

Table 18.10 The consumption of antibiotics in the Netherlands in defined daily dose (DDD) Pharmaceutical Trimethoprim Sulfamethoxazole/ trimethoprim Doxycycline Tetracycline Minocycline Erythromycin

2000

2001

2002

2003

2004

1 592 200 2 403 900

1 575 300 2 284 800

1 530 200 2 206 300

1 554 100 2 141 900

1 503 100 2 070 700

12 059 200 1 239 660 1 355 100 602 910

11 825 300 1 086 542 1 412 200 609 710

11 382 000 945 461 1 372 700 547 020

10 813 900 887 715 1 373 900 489 480

10 507 400 870 980 1 504 500 478 670

reported after treatment of rams with sulfamethoxazole–trimethoprim combination (Tanyildizi & Bozkurt, 2003). Treatment with a sulfamethoxazole–trimethoprim combination leads to a significant impairment of spermatogenesis (Crotty et al., 1995). In vitro analysis of human sperm function by Hargreaves et al. (1998) has shown that the combination with trimethoprim increases the sensitivity of spermatozoa to the drug approximately 10-fold. Trimethoprim and sulfamethoxazole are highly persistent (Bendz et al., 2005). Sulfamethoxazole has been detected in the environment: • • • • • • • • • • •

STP effluent Källby, Sweden, 0.02 μg/L (Bendz et al., 2005) STP effluents survey, 0.05–0.09 μg/L (Andreozzi et al., 2003) STW effluent, UK, median <0.050 μg/L (Ashton et al., 2004) River Elbe, up to 0.070 μg/L (Wiegel et al., 2004) Several STP effluents, Italy, 0.13 μg/L (Zuccato et al., 2005) River water, about 1.0 μg/L (Halling-Sørenson et al., 1998) STP effluents Wisconsin, USA, 0.05–0.37 μg/L (Karthikeyan & Meyer, 2006) STP effluent, Germany, median 0.40 μg/L (Hirsch et al., 1999) Surface water, Germany, median 0.03 μg/L (Hirsch et al., 1999) STP effluent, Canada, median 0.243 μg/L (Miao et al., 2004) STP effluents, Sweden, 0.135–0.304 μg/L (Lindberg et al., 2005)

Trimethoprim has been detected in the environment: STP effluent Källby, Sweden, 0.04 μg/L (Bendz et al., 2005) STP effluents survey, 0.04–0.13 μg/L (Andreozzi et al., 2003) STW effluent, UK, median 0.070 μg/L (Ashton et al., 2004) STW effluent, UK, 0.218–0.322 μg/L (Roberts & Thomas, 2006) River Tyne, UK, 0.004–0.019 μg/L (Roberts & Thomas, 2006) River Elbe, up to 0.040 μg/L (Wiegel et al., 2004) STP effluents, Wisconsin, USA, 0.05–0.55 μg/L (Karthikeyan & Meyer, 2006) • STP effluent, Germany, median 0.32 μg/L (Hirsch et al., 1999)

• • • • • • •

496

Endocrine-disrupting chemicals in food

• STP effluents, Canada, 0.009–0.194 μg/L (Metcalfe et al., 2003) • STP effluents, Sweden, 0.066–1.34 μg/L (Lindberg et al., 2005)

18.13.2 Tetracycline Tetracycline appears to be relatively non-toxic to spermatogenesis (Kushniruk, 1976; Timmermans, 1974). It has significant effects on sperm movement. Effects have been seen at concentrations as low as 2.5 mg/ml, well within those achieved following therapeutic doses of the antibiotic (Hargreaves et al., 1998). Tetracyclines are rapidly metabolised and moreover form relatively stable complexes with metal cations (Miao et al., 2004). Another source classifies tetracycline as non-degradable (Halling-Sørenson et al., 1998). Nevertheless, it has been detected in STP effluents. Tetracycline has been detected in the environment: • STP effluents, Wisconsin, USA, 0.05–0.37 μg/L (Karthikeyan & Meyer, 2006) • STP effluent, Canada, median 0.151 μg/L (Miao et al., 2004) • River water, about 1 μg/L (Halling-Sørenson et al., 1998)

18.13.3 Doxycycline Doxycycline decreases hyperactivation of cryopreserved human sperm (King et al., 1997). Doxycycline has been detected in the environment: • STP effluents, Canada, 0.04 μg/L (Miao et al., 2004) • STP effluents, Wisconsin, USA, 0.05 μg/L (Karthikeyan & Meyer, 2006) • STP effluents, Sweden, up to 915 ng/L (Lindberg et al., 2005) • Sweden sewage sludge: some samples had 1.5 mg/kg dry weight (Lindberg et al., 2005)

18.13.4 Minocycline Minocycline has been shown to be toxic to sperm (Schlegel et al., 1991). Minocycline may interfere with oral contraception, causing breakthrough bleeding (De Groot et al., 1990). No data were available on persistence and environmental fate.

18.13.5 Erythromycin Erythromycin had significant effects on rapid movement of sperm at concentrations >100 μg/ml (Hargreaves et al., 1998). Erythromycin application in pregnancy is associated with an increase in cardiac malformations in infants (Kallen & Olausson, 2003; Kallen et al., 2005). It may inhibit hepatic degradation of carbamazepine and theophylline (Blagg & Gleckman, 1981;

Pharmaceuticals

497

Mitch, 1989). It has a prolonged stability with a half-life of over one year (Zuccato et al., 2005) and has been detected in the environment: • STP effluents, Italy, 47 ng/L (Zuccato et al., 2005) • STP effluents, Canada, 80 ng/L (Miao et al., 2004) • STP effluents, Wisconsin USA 20 ng/L (Karthikeyan & Meyer, 2006) River water, around 1 μg/L (Halling-Sørenson et al., 1998) • STP effluents, Germany, median level 2.5 μg/L(Hirsch et al., 1999) • Surface waters, Germany, median level 150 ng/L (Hirsch et al., 1999) • STP effluent, UK, up to 290 ng/l (Roberts & Thomas, 2006) • River water, UK, up to 70 ng/L (Roberts & Thomas, 2006) • River water, Germany, up to 70 ng/L (Wiegel et al., 2004) • STW effluent, UK, mean 109 ng/L (Ashton et al., 2004)

18.13.6 Conclusions on antibiotics The sulfamethoxazole–trimethoprim combination appears to be the most important pharmaceutical in this group in view of mechanism of action, persistence and environmental exposure. Second in this group is erythromycin, for which the mechanism of action is less clear but the effects are relevant for fertility. In addition, this compound is stable in the environment and has been detected through many environmental studies. These compounds share significant effects on spermatogenesis and sperm function, which warrant their inclusion in this compilation, although the mechanism of action is less clear. Further study is needed as to the possible causation of these effects through endocrine mechanisms. Although doxycycline has the highest usage pattern, limited information on effects on fertility and on persistence precludes conclusions on the priority of studying this compound.

18.14

Risk assessment

According to the European Chemicals Bureau guidance (European Chemicals Bureau, 2003) the assessment of compounds is based on four main components: hazard identification, dose–response assessment, exposure assessment and risk characterization. A detailed description of these components is given below. Additionally, risk assessment can be extended and supported with the help of (quantitative) structure–activity relationships (QSARs), which assess compounds from the point of view of their structural properties. Moreover, if specific restrictions on data collected so far are met, QSARs may be useful in reducing the number of animal tests. In the frame of the current study, the hazard identification step includes selection of compounds which bear the intrinsic potential to disrupt the

498

Endocrine-disrupting chemicals in food

human endocrine system and cause fertility and fecundity problems in target populations. The key factors, which influenced the selection, were identified as follows (Luijten et al., 2005): • Potential and subsequent evidence of adverse effects related to fertility and fecundity. • Production volume. • Presence in the environment (STP effluent, rivers, groundwaters, soil, etc.). • Persistence in the environment. • Ability to reach the target populations through relevant exposure pathways.

18.14.1 Dose–response assessment The objective of the dose–response assessment is to analyse results of in vivo/in vitro studies for subsequent estimation of so-called threshold doses/ concentrations, which do not produce adverse effects on species/cells being tested. In the risk characterisation step these results are extrapolated using uncertainty factors to obtain safe doses/concentrations of the compound for target populations. There are several approaches for estimation of such threshold values, which depend on the acceptance by regulatory agencies, data available and subpopulations of interest: • no (low) observable adverse effect level (N(L)OAEL); • benchmark dose (BMD); • probabilistic analysis (PA). Each approach has inherent advantages and disadvantages. Firstly, among currently accepted approaches BMD is more accurate in estimation of ‘safe’ doses. Secondly, the BMD approach can be extended by applying probabilistic analysis, which is based on resampling and combines ranges of plausible dose estimates together with their uncertainties. Additionally, by applying distributions of uncertainty factors instead of point estimates, it is possible to estimate risks for different subpopulations (general/sensitive). Finally, it is possible to apply these approaches in a consecutive manner. No (low) observable effect level The N(L)OAEL represents the highest experimental dose for which no adverse effects have been documented (Crump et al., 1995). This approach is currently recognised and accepted by all regulatory agencies both in EU and USA. The principal procedure for calculation of N(L)OAEL is NOSTASOT dose (Crump et al., 1995). Although simple and straightforward, the N(L)OAEL has many limitations, such as:

Pharmaceuticals

499

• the selection of a ‘safe’ dose is limited to the set of experimental doses; • N(L)OAEL varies with the number of species being tested; • the slope of the dose–response plays little role in determining N(L)OAEL; • if no ‘safe’ dose was determined, a new set of studies should be carried out, which is both time and resource consuming. In case of failure to determine a N(L)OAEL from the initial studies, another option is to use LOAEL (lowest observable adverse effect level) and additionally introducing another uncertainty factor (usually 10). Benchmark dose The BMD (Crump, 1984) is the statistical lower confidence limit for a dose that produces a predetermined change in response rate of an adverse effect (benchmark response, BMR) compared with background. Unlike N(L)OAEL, BMD takes into account the whole dose–response information by fitting the mathematical model to dose–response data. Therefore, slopes are taken into account, which decreases the uncertainty of the resulting ‘safe’ doses. The sequence of steps for determination of BMD is the following: 1. Fit a mathematical model to the data (Crump, 1984): using maximum likelihood procedures, the predefined model (polynomial, Weibull, etc.) is being fitted to the set of dose–response pairs. 2. Definition of BMR: define the change in response rate, specific to given study (typical values are 1%, 5% and 10%) and with the help of fitted model determine the corresponding dose (this dose is the point estimate which is the basis of confidence limits calculation) (Crump et al., 1995). 3. Determination of BMD (Cox & Lindley 1974): BMD is defined to be the lower confidence limit of the dose obtained on step 2. In most cases, 95% lower limit is sufficient. The BMD is currently accepted by US Environmental Protection Agency and is increasingly recognised as the more accurate approach than the NOAEL. Probabilistic analysis PA (Slob & Pieters, 1998) represents ‘safe’ dose in terms of distribution, thus combining the range of plausible values together with their uncertainties. The basic idea is to replace the point estimate obtained by, for example, the BMD approach by the set of values, generated according to some predefined distribution model (usually log-normal). Therefore, the first steps are similar to BMD approach. For generation of the set of values some resampling technique can be used (Monte Carlo or Latin hypercube) (Vose,

500

Endocrine-disrupting chemicals in food

2000). In the risk characterisation step, the resulting distribution can be combined with distributions of uncertainty factors to obtain uncertainty distribution of ‘safe’ human dose.

18.14.2 Exposure assessment The objective of exposure assessment is to quantify the doses of the compounds, identified in hazard identification step, which are taken by target populations. The results of exposure assessment are then compared with dose–response assessment results in the risk characterisation step. The core of exposure assessment includes identification of relevant pathways of exposure (Luijten et al., 2005) and estimation of total chemical intake with respect to these pathways. Estimation of total chemical intake is based on both chemical concentrations in food and consumption patterns. The exposure assessment methods can be subdivided into three classes: • screening tools; • tools based on specific data; • confirmatory methods. Tools which are based on specific data are especially useful in risk characterisation step for comparison with results of dose–response analyses. There are three currently applied techniques to combine chemical concentration and consumption patterns (Kroes et al., 2002): • Point estimates: assume single (best guess) estimates for both concentration and consumption. • Simple distributions: a method that employs distributions of consumption variables but uses a fixed value for the concentration. The results are more informative than those of the point estimates because they take account of the variability that exists in food consumption patterns. • Probabilistic analysis: variables are described in terms of distributions to characterise their variabilities and/or uncertainties. The method takes account of all the possible values that each variable could take and weights each possible outcome by the probability of its occurrence. Provided that data are adequate and models are selected properly, probabilistic assessment should provide the most realistic estimates of exposure. When data are adequate, it is preferable to apply simple distributions and probabilistic analysis for exposure assessment because they both take into account the probabilistic nature of consumption patterns. The comparison of the results, obtained with the help of these two techniques, will determine the most suitable approach for risk characterisation. Finally, since PA replaces point estimates with distributions, this provides additional

Pharmaceuticals

501

data for comparison with results of dose–response assessment (also represented by distributions). 18.14.3 Risk characterisation The final stage of risk assessment process combines results obtained from previous steps. The main objective is to compare results of dose–response assessment and exposure assessment in order to identify the strategy for eliminating/reducing the risk. Estimation of reference dose Reference dose (RfD) (Environmental Protection Agency, 1993) is the dose of the chemical, which is regarded to be safe for target population. RfD is the result of extrapolation of ‘safe’ doses estimated for species being tested during dose–response assessment. In order to extrapolate doses to sensitive humans uncertainty factors are applied: RfD =

N(L)OAEL UF1 × UF2 × . . . × UFn

(18.1)

where UFi corresponds to ith uncertainty factor. The most widely applied uncertainty factors include (Crump et al., 1995): • H (interhuman): describes variation in sensitivity in target population (default 10). • A (animal to man): accounts for the uncertainty in extrapolating animal data to humans (default 10). • S (subchronic to chronic): accounts for the uncertainty in extrapolating from subchronic to chronic NOAELs (default 10). The value of the denominator in equation. (18.1) should not exceed 104. More advanced approaches, such as probabilistic analysis, assume distributions of uncertainty factors instead of point estimates. This gives more flexible results provided there are enough data for determination of such distributions (but default log-normal distributions are considered to be plausible) (Vermeire et al., 2001). Comparison of dose–response data and exposure data The final stage of risk characterisation is to compare dose–response and exposure data and draw a conclusion on further actions. The comparison is based on evaluation of the margin of safety (MOS) (European Chemicals Bureau, 2003). The possible outcomes of the comparison and therefore of the whole risk assessment process are the following: • Need for further information and/or testing. • At present no need for further information and/or testing and no need for risk reduction measures. • Need for limiting the risk.

502

Endocrine-disrupting chemicals in food

18.14.4 Quantitative structure–activity relationships QSARs are estimation methods developed and used for prediction of specific effects/properties of chemicals which are based on the structure of the substance. QSAR models have been created for a range of end points, including several toxicological and ecotoxicological end points and physicochemical/fate parameters (European Chemicals Bureau, 2003). In case of exposure assessment, in the absence of experimental data, e.g. if it is not possible to obtain reliable measured data, specific physicochemical/fate parameters may be derived by applying QSARs. For the risk characterisation step, if the comparison of exposure and dose–response steps is inconclusive, QSARs may serve as a supporting tool in taking decisions. Additionally, QSARs may also be used to optimise the testing strategies. It should be noted that estimates resulting from QSARs cannot be the only basis of risk assessment for a given compound, since QSARs are an estimation method and therefore there is a certain probability that the estimate is poor. Instead, QSARs should be seen as a complementary tool, which evaluated together with dose–response and exposure assessment can provide a more complete understanding of the characteristics of the substance. Furthermore, the result of QSARs should be evaluated for consistency in the light of available experimental data and validated estimates from other end points. The development of a QSAR is based on the assumption that chemical substances which reach and interact with a target site by the same mechanism do so because of their similar chemical properties. Since different mechanisms of interaction usually will depend on different properties, QSARs must be generally developed for each mode of action. Some QSARs are developed using quantitative data in order to predict a quantitative parameter. There are two types of predictive methods: • formalised methods; • expert judgement. Formalised methods are methods which can be subjected to validation, e.g. applied by one assessor and are both reproducible and transparent to other assessors. They are based on mathematical computations and/or fixed rules. Critical evaluation of the models should be carried out, including the evaluation of the appropriateness and validity of the descriptor variables, the evaluation of the form of the models and the methods used to construct the models. These models should be applied critically acknowledging the limitations of the model, such as which compounds are within the domain of the model. Consequently, the specific information concerning the model which is used should be made available to the other assessors in order to ensure transparency and reproducibility (European Chemicals Bureau, 2003). Methods based on expert judgement rely on the expert’s experience and intuition. They are generally non-quantitative methods based on structural similarity and/or analogues. These methods should be used with caution, as

Pharmaceuticals

503

they rely on the judgement of the individual assessor and may not be reproducible by the others. For a QSAR to be used for the risk assessment process, it is necessary that the end point estimated is compatible with an end point used in the risk assessment. If such compatibility exists, then the QSAR can be used for such purposes as: • assisting in data evaluation; • contributing to the decision-making process on whether further testing is necessary to clarify an end point of concern and, if further testing is needed, to optimise the testing strategies, where appropriate; • establishing parameters (descriptors) which are necessary to conduct the exposure and/or effects assessment; • identifying effects which may be of potential concern on which test data are not available. Validated QSARs are not currently available for human health-related toxicity end points. Instead, expert judgement is used in the light of data on close structural analogy and/or the presence of ‘structural alerts’ (i.e. fragments associated with affects) in the substance. However, recently the techniques have been developed aiming to incorporate advanced classification schemes in order to categorise compounds on the basis of numerical representation of their chemical structure (Asikainen et al., 2006).

18.15

Conclusions

As part of the Food & Fecundity EC FP6 project, a prioritisation list has been created of pharmaceutical compounds with the potential to enter the environment and the human food chain and with suspected or proven ability to affect human fecundity through an endocrine mechanism of action. The list is based on an extensive literature search while considering the following criteria: 1. Do the available data indicate existence of an endocrine mechanism of action with an effect on fecundity? 2. Is the production volume sufficiently large to cause concern? 3. Has the PP been detected in food and/or environment? 4. Is the PP sufficiently persistent in the environment? The chapter also discusses the endocrine and alternative mechanisms by which the drugs can affect human fecundity in men and women. An overview of possible pathways by which pharmaceutical products can enter the environment and human food chain are discussed, identifying water as the major media for transport and dispersion of pharmaceutical products in the environment and therefore providing potential to also enter the human food chain. The currently available data are too limited to allow for

504

Endocrine-disrupting chemicals in food

definitive conclusions about human risks due to environmental exposures to endocrine-active drug residues. On the other hand, this study shows that such exposures are actually occurring, although seemingly below thresholds for human concern. However, given trends of increasing production and use of the drugs involved, both environmental and human exposure levels are likely to increase as well. When estimating the related actual risks for human health, additional consideration should be given to the fact that exposures to a range of pharmaceutical residues are likely to occur simultaneously, increasing the chance for combined exposures above thresholds of significant endocrine effects. The increasing amount of emerging new data also illustrates the enhanced awareness about hazards and anticipated possible risks of endocrine-disrupting pharmaceutical residues for human fecundity. It is concluded that for a series of pharmaceuticals, current data on production, use pattern and environmental fate warrant further study on possible human exposure and health risks. The risk assessment paradigm is reviewed and will be applied to forthcoming data on concentrations of the selected compounds in drinking water and foodstuffs. These analyses will enable informed conclusions about current risks of human exposure to pharmaceutical residues via the food chain.

18.16

References

aherne gw and briggs r (1989) The relevance of the presence of certain synthetic steroids in the aquatic environment. J Pharm Pharmacol 41, 735–736. akpoviroro jo, mangalam m, jenkins n and fotherby k (1981) Binding of the contraceptive steroids medroxyprogesterone acetate and ethynyloestradiol in blood of various species. J. Steroid Biochem., 14, 493–498. alleva jj and alleva fr (1995) Synergistic inhibition by benzodiazepine and barbiturate drugs of the clock control of ovulation in hamsters. Chronobiol. Int., 12, 1–7. andersen op, hansen p and madsen h (1981) Hyperprolactinemic amenorrhea induced by metoclopamid (PrimperanR). Acta Obstet. Gynecol. Scand., 60, 341–342. andreozzi r, raffaele m and nicklas p (2003) Pharmaceuticals in STP effluents and their solar photodegradation in aquatic environment. Chemosphere, 50, 1319–1330. archand-hoy ld, nimrod ac and benson wh (1998) Endocrine-modulating substances in the environment: estrogenic effects of pharmaceutical products. Int. J. Toxicol. 17, 139–158. armstrong dt and grinwich dl (1972) Blockade of spontaneous and LH-induced ovulation in rats by indomethacin, an inhibitor of prostaglandin biosynthesis. I. Prostaglandins, 1, 21–28. arze rs, ramos jm, rashid hu and kerr dn (1981) Amenorrhoea, galactorrhoea, and hyperprolactinaemia induced by methyldopa. Br. Med. J. (Clin. Res. Ed.), 283, 194. ashby j, lefevre pa, odum j, tinwell h, kennedy sj, beresford n and sumpter jp (1997) Failure to confirm estrogenic activity for benzoic acid and clofibrate: impli-

Pharmaceuticals

505

cations for lists of endocrine-disrupting agents. Regul. Toxicol. Pharmacol., 26, 96–101. ashton d, hilton m and thomas kv (2004) Investigating the environmental transport of human pharmaceuticals to streams in the United Kingdom. Sci. Total. Environ., 333, 167–184. asikainen a, kolehmainen m, ruuskanen j and tuppurainen k (2006) Structurebased classification of active and inactive estrogenic compounds by decision tree, LVQ and kNN methods. Chemosphere, 62(4), 658–673. back dj, breckenridge am, crawford fe, et al. (1978) Kinetics of norethindrone in women. II. Single-dose kinetics. Clin. Pharmacol. Ther., 24, 448–453. back dj, bates m, breckenridge jm, et al. (1981) The pharmacokinetics of levonorgestrel and ethinylestradiol in women – studies with Ovran and Ovranette. Contraception, 23, 229–239. bain sc, lemon m and jones af (1990) Gemfibrozil-induced impotence. Lancet, 336, 1389. baird dt (1979) Endocrinology of female infertility. Br. Med. Bull., 35, 193–198. bancroft j, tennent g, loucas k and cass j (1974) The control of deviant sexual behaviour by drugs. I. Behavioural changes following oestrogens and anti-androgens. Br. J. Psychiatry, 125, 310–315. barnes tr, bamber rw and watson jp (1979) Psychotropic drugs and sexual behaviour. Br. J. Hosp. Med., 21, 594–600. bathen j (1978) Propranolol erectile dysfunction relieved. Ann. Intern. Med., 88, 716–717. bbc (2004) Prozac ‘found in drinking water’ (http://news.bbc.co.uk/1/hi/health/ 3545684.stm). beardmore ks, morris jm and gallery ed (2002) Excretion of antihypertensive medication into human breast milk: a systematic review. Hypertens. Pregnancy, 21, 85–95. bebb ra, anawalt bd, christensen rb, paulsen ca, bremner wj and matsumoto am (1996) Combined administration of levonorgestrel and testosterone induces more rapid and effective suppression of spermatogenesis than testosterone alone: a promising male contraceptive approach. J. Clin. Endocrinol. Metab., 81(2), 757–762. beeley l (1984) Drug-induced sexual dysfunction and infertility. Adverse Drug React Acute Poisoning Rev., 3, 23–42. belfroid ac, van der horst a, vethaak ad, schäfer aj, rijs gb, wegener j and cofino wp (1999) Analysis and occurrence of estrogenic hormones and their glucuronides in surface water and waste water in The Netherlands. Sci. Total Environ., 225(1–2), 101–108. ben rafael z, mastroianni l, jr, meloni f, lee ms and flickinger gl (1986) Total estradiol, free estradiol, sex hormone-binding globulin, and the fraction of estradiol bound to sex hormone-binding globulin in human follicular fluid. J. Clin. Endocrinol. Metab., 63, 1106–1111. bendz d, paxeus na, ginn tr and loge fj (2005) Occurrence and fate of pharmaceutically active compounds in the environment, a case study: Hoje River in Sweden. J. Hazard. Mater., 122, 195–204. berg c, halldin k, fridolfsson a-k, brandt i and brunstro b (1999) The avian egg as a test system for endocrine disrupters: effects of diethylstilbestrol and ethynylestradiol on sex organ development. The Science of the Total Environment, 233, 57–66. biron p (1979) Diminished libido with cimetidine therapy. Can. Med. Assoc. J., 121, 404–405. blagg na and gleckman ra (1981) Erythromycin. new indications and toxicities. Postgrad. Med., 69(3), 59–68.

506

Endocrine-disrupting chemicals in food

boyd as (1988) Varicoceles and male infertility. Am. Fam. Physician, 37, 252–258. boxall ab (2004) The environmental side effects of medication. EMBO Rep., 5, 1110–1116. brooks bw, chambliss ck, stanley jk, ramirez a, banks ke, johnson rd and lewis rj (2005) Determination of select antidepressants in fish from an effluentdominated stream. Environ. Toxicol. Chem., 24, 464–469. buvat j (2003) Androgen therapy with dehydroepiandrosterone. World J. Urol., 21, 346–355. cameron st, critchley ho, thong kj, buckley ch, williams ar and baird dt (1996) Effects of daily low dose mifepristone on endometrial maturation and proliferation. Hum. Reprod., 11, 2518–2526. cargouet m, perdiz d, mouatassim-souali a, tamisier-karolak s and levi y (2004) Assessment of river contamination by estrogenic compounds in Paris area (France). Science of the Total Environment, 324, 55–66. carp hj, fein a and nebel l (1988) Effect of diclofenac on implantation and embryonic development in the rat. Eur. J. Obstet. Gynecol. Reprod. Biol., 28, 273–277. cavalier m (1995) Sexually speaking . . . the dance. Alaska Med., 37, 113. chandra d, thulasimany m and biswas ar (1982) Uterotropic effect of clofibrate and phenylbutazone in immature female rats. Indian J. Physiol. Pharmacol., 26, 147–151. chang sw, fine r, siegel d, chesney m, black d and hulley sb (1991) The impact of diuretic therapy on reported sexual function. Arch. Intern. Med., 151, 2402–2408. chowdhury ar (1987) Effect of pharmacological agents on male reproduction. Adv. Contracept. Deliv. Syst., 3, 347–352. clayton do and shen ww (1998) Psychotropic drug-induced sexual function disorders: diagnosis, incidence and management. Drug Saf., 19, 299–312. cohen aj and bartlik b (1998) Ginkgo biloba for antidepressant-induced sexual dysfunction. J. Sex Marital Ther., 24, 139–143. colucci ms and topp e (2001) Persistence of estrogenic hormones in agricultural soils: II. 17a-Ethynylestradiol. J. Environ. Qual., 30, 2077–2080. compton mt and miller ah (2001) Priapism associated with conventional and atypical antipsychotic medications: a review. J. Clin. Psychiatry, 62, 362–366. cortes a, merino g and contreras e (1978) Effect of fluoxetine (Lilly II0I40) on serotonin, noradrenaline and acetylcholine in isolated organs. Cell. Mol. Biol. Incl. Cyto. Enzymol., 23, 179–182. corton jc, bocos c, moreno es, merritt a, cattley rc and gustafsson ja (1997) Peroxisome proliferators alter the expression of estrogen-metabolizing enzymes. Biochimie, 79, 151–162. cox dr and lindley dv (1974) Theoretical Statistics. Chapman & Hall, London. cramer dw, liberman rf, hornstein md, mcshane p, powers d, li ey and barbieri r (1998) Basal hormone levels in women who use acetaminophen for menstrual pain. Fertil. Steril., 70, 371–373. crotty kl, may r, kulvicki a, kumar d and neal de, jr (1995) The effect of antimicrobial therapy on testicular aspirate flow cytometry. J. Urol., 153, 835–838. crump k (1984) A new method for determining allowable daily intakes. Fundamental Appl. Toxicol., 4, 854–871. crump k, allen b and faustman e (1995) The Use of Benchmark Dose Approach in Health Risk Assessment. EPA/630/R-94/007. cutler gb jr, sauer ma and loriaux dl (1979) SC 25152: a potent mineralocorticoid antagonist with decreased antiandrogenic activity relative to spironolactone. J. Pharmacol Exp. Ther., 209, 144–146.

Pharmaceuticals

507

dail wg, mcguffee l, minorsky n and little s (1987) Responses of smooth muscle strips from penile erectile tissue to drugs and transmural nerve stimulation. J. Auton. Pharmacol., 7, 287–293. daughton cg and ternes ta (1999) Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Perspect., 107, Suppl 6, 907–938. davies ts and klowe wm (1998) Preclinical toxicological evaluation of sertraline hydrochloride. Drug Chem Toxicol., 21(4), 521–537. de groot ac de, eshuis h and stricker bh (1990) In effectiviteit van orale anticonceptic tijdens gebruik van minocycline (in Dutch). Ned Tijdschr Geneeskd. 134(25), 1227–1229. della ca, pfiffner d, meier b and hess om (2003) Sexual activity in hypertensive men. J. Hum. Hypertens., 17, 515–521. desbrow c, routledge ej, brighty gc, sumpter jp and waldock m (1998) Identification ofestrogenic chemicals in STW effluent. 1. Chemical fractionation and in vitro biological screening. Environ. Sci. Technol., 32(11), 1549–1558. dmowski wp (1988) Danazol-induced pseudomenopause in the management of endometriosis. Clin. Obstet. Gynecol., 31, 829–839. dorevitch a and davis h (1994) Fluvoxamine-associated sexual dysfunction. Ann. Pharmacother., 28, 872–874. dusing r (2005) Sexual dysfunction in male patients with hypertension: influence of antihypertensive drugs. Drugs 65, 773–786. el banna aa, al gahtani sj and sedrani sh (1993) Effect of indomethacin administered during the normal time of implantation on the progesterone and protein composition of plasma and uterine flushings of rabbits. J. Reprod. Fertil., 97, 277–285. el etreby mf and neumann f (1980) Influence of sex steroids and steroid antagonists on hormone-dependent tumors in experimental animals. In: Iacobelli S, King RJB, Lindner HR and Lippman ME, eds, Hormones and Cancer, New York, Raven Press, pp. 321–336. el sayed mg, el sayed mt, elazab abd es, hafeiz mh, el komy aa and hassan e (1998) Effects of some beta-adrenergic blockers on male fertility parameters in rats. Dtsch. Tierarztl. Wochenschr., 105, 10–12. environmental protection agency (1993) Reference Dose (RfD): Description and Use in Health Risk Assessments. US EPA http://www.epa.gov/iris/rfd.htm espey ll (1983) Comparison of the effect of nonsteroidal and steroidal antiinflammatory agents on prostaglandin production during ovulation in the rabbit. Prostaglandins, 26, 71–78. espey ll, stein vi and dumitrescu j (1982) Survey of antiinflammatory agents and related drugs as inhibitors of ovulation in the rabbit. Fertil. Steril., 38, 238–247. european chemicals bureau (2003) European Commission Technical Guidance Document in Support of Commission Directive 93/67/EEC on Risk Assessment for New Notified Substances and Commission Regulation No. 1488/94 on Risk Assessment for Existing Substances, Office for Official Publications of the European Communities: Luxembourg, 1996 (and as revised, 2003). falkson ci, falkson g and falkson hc (1991) Long term follow-up of patients with advanced prostatic cancer treated with nasal buserelin. Ann. Oncol., 2, 303–304. fell mj, marshall km, williams j and neill jc (2004) Effects of the atypical antipsychotic olanzapine on reproductive function and weight gain in female rats. J. Psychopharmacol., 18, 149–155. fell mj, gibson r, mcdermott e, sisodia g, marshall km and neill jc (2005) Investigation into the effects of the novel antipsychotic ziprasidone on weight gain and reproductive function in female rats. Behav. Brain Res., 160, 338–343.

508

Endocrine-disrupting chemicals in food

fent k, weston aa and caminada d (2006) Ecotoxicology of human pharmaceuticals. Aquatic Toxicol., 76, 122–159. ferrari b, paxeus n, lo gr, pollio a and garric j (2003) Ecotoxicological impact of pharmaceuticals found in treated wastewaters: study of carbamazepine, clofibric acid, and diclofenac. Ecotoxicol. Environ. Saf., 55, 359–370. fernandez pr, kessler ag and rawlins rg (1991) Modern approaches to the treatment of human infertility through assisted reproduction. P. R. Health Sci. J., 10, 75–81. ficicioglu c, tekin hi, arioglu pf and okar i (1995) A murine model of adenomyosis: the effects of hyperprolactinemia induced by fluoxetine hydrochloride, a selective serotonin reuptake inhibitor, on adenomyosis induction in Wistar albino rats. Acta Eur. Fertil., 26, 75–79. fitzgerald je, petrere ja and de la iglesia fa (1987) Experimental studies on reproduction with the lipid-regulating agent gemfibrozil. Fundam. Appl. Toxicol., 8, 454–464. fogari r, preti p, derosa g, marasi g, zoppi a, rinaldi a and mugellini a (2002) Effect of antihypertensive treatment with valsartan or atenolol on sexual activity and plasma testosterone in hypertensive men. Eur. J. Clin. Pharmacol., 58, 177–180. fotherby k and caldwell ads (1994) New progestogens in oral contraception. Contraception, 49, 1–31. frances a and kaplan hs (1982) Patient on hypertensive drug has increasing erectile problems. Hosp. Community Psychiatry, 33, 431–432. fraunfelder ft and meyer sm (1985) Sexual dysfunction secondary to topical ophthalmic timolol. JAMA, 253, 3092–3093. fovaeus m, andersson ke and hedlund h (1987) Effects of some calcium channel blockers on isolated human penile erectile tissues. J. Urol., 138, 1267–1272. furuhjelm m, rydner t and carlstrom k (1980) Hyperprolactinemia in cases of infertility and amenorrhea. Acta. Obstet. Gynecol. Scand., 59, 137–141. gentile s (2005) The safety of newer antidepressants in pregnancy and breastfeeding. Drug Saf., 28, 137–152. giuliano f and allard j (2001) Dopamine and sexual function. Int. J. Impot. Res., 13 Suppl 3, S18–S28. gnoth c, godehardt e, frank-herrmann p, friol k, tigges j and freundl g (2005) Definition and prevalence of subfertility and infertility. Hum. Reprod., 20, 1144–1147. godo g (1984) Hyperprolactinaemia and female infertility. Acta. Med. Hung., 41, 185–193. gonzalez-villapando c, szabo m and frohman la (1980) Central nervous systemmediated stimulation of prolactin secretion by cimetidine, a histamine H2-receptor antagonist: impaired responsiveness in patients with prolactin-secreting tumors and idiopathic hyperprolactinemia. J. Clin. Endocrinol. Metab., 51, 1417–1424. gregorian rs, golden ka, bahce a, goodman c, kwong wj and khan zm (2002) Antidepressant-induced sexual dysfunction. Ann. Pharmacother., 36, 1577–1589. gross b, montgomery-brown j, naumann a and reinhard m (2004) Occurrence and fate of pharmaceuticals and alkylphenol ethoxylate metabolites in an effluentdominated river and wetland. Environ. Toxicol. Chem., 23, 2074–2083. gupta s, masand p, ashton ak and berry sl (1999) Phenelzine-induced sexual dysfunction treated with sildenafil. J. Sex Marital Ther., 25, 131–135. gutendorf b and westendorf j (2001) Comparison of an array of in vitro assays for the assessment of the estrogenic potential of natural and synthetic estrogens, phytoestrogens and xenoestrogens. Toxicology, 166, 79–89.

Pharmaceuticals

509

haddad pm and wieck a (2004) Antipsychotic-induced hyperprolactinaemia: mechanisms, clinical features and management. Drugs, 64, 2291–2314. hafez es and hafez sd (2005) Erectile dysfunction: anatomical parameters, etiology, diagnosis, and therapy. Arch. Androl., 51, 15–31. hahlbeck e, katsiadaki i, mayer i, adolfsson-erici m, james j and bengtsson be (2004) The juvenile three-spined stickleback (Gasterosteus aculeatus L.) as a model organism for endocrine disruption II–kidney hypertrophy, vitellogenin and spiggin induction. Aquat. Toxicol., 70(4), 311–326. halling-sørensen b, nors nielsen s, lanzky pf, ingerslev f, holten lützhøft hc and jørgensen se (1998) Occurrence, fate and effects of pharmaceutical substances in the environment – a review. Chemosphere, 36, 357–393. hargreaves ca, rogers s, hills f, rahman f, howell rj and homa st (1998) Effects of co-trimoxazole, erythromycin, amoxycillin, tetracycline and chloroquine on sperm function in vitro. Hum. Reprod., 13, 1878–1886. harries je, sheahan da, jobling s, matthiessen p, neall p, routledge ej, rycroft r, sumpter jp and tylor t (1996) A survey of estrogenic activity in United Kingdom inland water. Environ. Toxicol. Chem., 15, 1993–2002. harries je, sheahan da, jobling s, matthiessen p, neall p, sumpter jp, tylor t and zaman n (1997) Estrogenic activity in five United Kingdom rivers detected by vitellogenesis in caged male trout. Environ. Toxicol. Chem., 16, 534–542. hashimoto f, oguchi y, morita m, matsuoka k, takeda s, kimura m and hayashi h (2004) PPARalpha agonists clofibrate and gemfibrozil inhibit cell growth, downregulate hCG and up-regulate progesterone secretions in immortalized human trophoblast cells. Biochem. Pharmacol., 68, 313–321. heberer t (2002) Occurrence, fate, and removal of pharmaceutical residues in the aquatic environment: a review of recent research data. Toxicol. Lett., 131, 5–17. hedlund h and andersson ke (1985) Comparison of the responses to drugs acting on adrenoreceptors and muscarinic receptors in human isolated corpus cavernosum and cavernous artery. J. Auton. Pharmacol., 5, 81–88. hedlund h and andersson ke (1989) Effects of prazosin in men with symptoms of bladder neck obstruction and a non-hyperplastic prostate. Scand. J. Urol. Nephrol., 23, 251–254. heikkine t, ekblad u and laine k (2002) Transplacental transfer of citalopram, fluoxetine and their primary demethylated metabolites in isolated perfused human placenta. Brit. J. Obst. Gynecol., 109, 1003–1008. helm f, kretzschmar r, leuschner f and neumann w (1976) [Investigations on the effect of the combination sulfamoxole/trimethoprim on fertility and fetal development in rats and rabbits (author’s tranls)]. Arzneimittelforschung., 26, 643–651. hendrick v, stowe zn, altshuler ll, hwang s, lee e and haynes d (2003) Placental passage of antidepressant medications. Am. J. Psychiatry, 160, 993–996. hirsch r, ternes t, haberer k and kratz kl (1999) Occurrence of antibiotics in the aquatic environment. Sci. Total Environ., 225, 109–118. hohenblum p, gans o, moche w, scharf s and lorbeer g (2004) Monitoring of selected estrogenic hormones and industrial chemicals in groundwaters and surface waters in Austria. Sci. Total Environ., 333, 185–193. holmes b and sorkin em (1986) Indoramin. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in hypertension and related vascular, cardiovascular and airway diseases. Drugs, 31, 467–499. holzbeierlein jm, castle ep and thrasher jb (2003) Complications of androgen-deprivation therapy for prostate cancer. Clin. Prostate Cancer, 2, 147–152.

510

Endocrine-disrupting chemicals in food

holzbeierlein jm, mclaughlin md and thrasher jb (2004) Complications of androgen deprivation therapy for prostate cancer. Curr. Opin. Urol., 14, 177–183. hu xh, bull sa, hunkeler em, ming e, lee jy, fireman b and markson le (2004) Incidence and duration of side effects and those rated as bothersome with selective serotonin reuptake inhibitor treatment for depression: patient report versus physician estimate. J. Clin. Psychiatry, 65, 959–965. hummer m, kemmler g, kurz m, kurzthaler i, oberbauer h and fleischhacker ww (1999) Sexual disturbances during clozapine and haloperidol treatment for schizophrenia. Am. J. Psychiatry, 156, 631–633. IARC Monographs (1979a) 21, 233–255, 257–278, 387–395. IARC Monographs (1979b) 21, 441–460. IARC Monographs (1987) Suppl. 6: 427–429. isojarvi ji, lofgren e, juntunen ks, pakarinen aj, paivansalo m, rautakorpi i and tuomivaara l (2004) Effect of epilepsy and antiepileptic drugs on male reproductive health. Neurology, 62, 247–253. isojarvi ji, tauboll e and herzog ag (2005) Effect of antiepileptic drugs on reproductive endocrine function in individuals with epilepsy. CNS Drugs, 19, 207–223. johnson dj, sanderson h, brain ra, wilson cj, bestari kj and solomon kr (2005) Exposure assessment and microcosm fate of selected selective serotonin reuptake inhibitors. Regul. Toxicol. Pharmacol., 42, 313–323. jorgensen h, knigge u, kjaer a and warberg j (1996) Interactions of histaminergic and serotonergic neurons in the hypothalamic regulation of prolactin and ACTH secretion. Neuroendocrinology, 64, 329–336. joseph jc and schuna aa (1990) Management of hypertension in the diabetic patient. Clin. Pharm., 9, 864–873. juchem m, pollow k, elger w, hoffmann g and möbus v (1993) Receptor binding of norgestimate – a new orally active synthetic progestational compound. Contraception, 47, 283–294. judd sj, lazarus l and compton p (1976) Hyperprolactinaemia: its physiology and investigation. Med. J. Aust., 2, 23–25. jurgens md, holthaus kie, johnson ac, smith jjl, hetheridge m and williams rj (2002) The potential for estradiol and ethinylestradiol degradation in English rivers. Environ. Toxicol. Chem., 21(3), 480–488. khan ua, aslam m and saeed sa (2004) Effect of beta adrenergic antagonist on the production of testosterone by rat’s Leydig cells. J. Ayub. Med. Coll. Abbottabad., 16, 26–28. kallen baj and olausson po (2003) Maternal drug use in early pregnancy and infant cardiovascular defect. Reprod. Toxicol., 17, 255–261. kallen baj, olausson po and danielsson br (2005) Is erythromycin teratogenic in humans? Reprod. Toxicol., 20, 209–214. kanaly ka and berman jr (2002) Sexual side effects of SSRI medications: potential treatment strategies for SSRI-induced female sexual dysfunction. Curr. Womens Health Rep., 2, 409–416. katz im (1986) Sexual dysfunction and ocular timolol. JAMA, 255, 37–38. karthikeyan kg and meyer mt (2006) Occurrence of antibiotics in wastewater treatment facilities in Wisconsin, USA. Sci. Total Environ., 361(1–3), 196–207. keene lc and davies ph (1999) Drug-related erectile dysfunction. Adverse Drug React. Toxicol. Rev., 18, 5–24. kelly dl and conley rr (2006) A randomized double-blind 12-week study of quetiapine, risperidone or fluphenazine on sexual functioning in people with schizophrenia. Psychoneuroendocrinology, 31(3), 340–346.

Pharmaceuticals

511

kher r and kalla nr (1996) Antifertility effects of an LHRH agonist in male mice. Contraception, 53, 299–306. killick s and elstein m (1987) Pharmacologic production of luteinized unruptured follicles by prostaglandin synthetase inhibitors. Fertil. Steril., 47, 773–777. king k, chan pj, patton wc and king a (1997) Antibiotics: effect on cryopreservedthawed human sperm motility in vitro. Fertil. Steril., 67(6), 1146–1151. klein ta, richmond ja and mishell dr, jr (1976) Pelvic tuberculosis. Obstet. Gynecol., 48, 99–104. kloner ra (2003) Erectile dysfunction in the cardiac patient. Curr. Urol. Rep., 4, 466–471. knobil e (1988a) The neuroendocrine control of ovulation. Hum. Reprod., 3, 469–472. knobil e (1988b) The hypothalamic gonadotrophic hormone releasing hormone (GnRH) pulse generator in the rhesus monkey and its neuroendocrine control. Hum. Reprod., 3, 29–31. knobil e (1990) The GnRH pulse generator. Am. J. Obstet. Gynecol., 163, 1721–1727. kolettis pn (2003) Evaluation of the subfertile man. Am. Fam. Physician, 67, 2165–2172. korner w, spengler p, bolz u, schuller w, hanf v and metzger jw (2001) Substances with estrogenic activity in effluents of sewage treatment plants in southwestern Germany. 2. Biological analysis. Environ. Chem. Toxicol., 20, 2142. kroes r, muller d, lambe j, loewik mrh, van klaveren j, kleiner j, massey r, mayer s, urieta i, verger p and visconti a (2002) Assessment of intake from the diet. Food Chem. Toxicol., 40, 327–385. kushniruk ii (1976) [Effect of several antibacterial preparations on the concentration of nucleic acids in spermatogenic epithelial cells]. Tsitol. Genet., 10, 342–344. kuch hm and ballschmiter k (2000) Determination of endogenous and exogenous estrogens in effluents from sewage treatment plants at the ng/L-level. Fresenius J. Anal. Chem., 366, 392–395. kuhnz w, heuner a, humpel m, seifert w and michaelis k (1997) In vivo conversion of norethisterone and norethisterone acetate to ethinyl etradiol in postmenopausal women. Contraception, 56, 379–385. labbate la, croft ha and oleshansky ma (2003) Antidepressant-related erectile dysfunction: management via avoidance, switching antidepressants, antidotes, and adaptation. J. Clin. Psychiatry, 64 Suppl 10, 11–19. lam mw, young cj, brain ra, johnson dj, hanson ma, wilson cj, richards sm, solomon kr and mabury sa (2004) Aquatic persistence of eight pharmaceuticals in a microcosm study. Environ. Toxicol. Chem., 23, 1431–1440. länge r, hutchinson th, croudace cp, siegmund f, schweinfurth h, hampe p, panter gh and sumpter jp (2001) Effects of the synthetic estrogen 17α-ethinylestradiol on the life-cycle of the fathead minnow (Pimephales promelas). Environ. Toxicol. Chem., 20, 1216–1227. larsson dgj, adolfsson-erici m, parkkonen j, pettersson m, berg ah, olsson p-e and forlin l (1999) Ethinyloestradiol – an undesired fish contraceptive? Aquatic Toxicol., 45, 91–97. lesko lm, stotland nl and segraves rt (1982) Three cases of female anorgasmia associated with MAOIs. Am. J. Psychiatry, 139, 1353–1354. lindberg rh, wennberg p, johansson mi, tysklind m and andersson ba (2005) Screening of human antibiotic substances and determination of weekly mass flows in five sewage treatment plants in Sweden. Environ. Sci. Technol., 39, 3421–3429.

512

Endocrine-disrupting chemicals in food

lindqvist n, tuhkanen t and kronberg l (2005) Occurrence of acidic pharmaceuticals in raw and treated sewages and in receiving waters. Water Res., 39, 2219–2228. liu rc, hahn c and hurtt me (1996) The direct effect of hepatic peroxisome proliferators on rat Leydig cell function in vitro. Fundam. Appl. Toxicol., 30, 102–108. luijten m, tomenko v, popov v and piersma ah (2005) A scenariospecific prioritization list of chemicals bearing a high potential of affecting human fecundity. D1.1. Contract No. FOOD-CT-2004-513953. mantero f and lucarelli g (2000) Aldosterone antagonists in hypertension and heart failure. Ann. Endocrinol (Paris), 61, 52–60. marik j and hulka j (1978) Luteinized unruptured follicle syndrome: a subtle cause of infertility. Fertil. Steril., 29, 270–274. marti b (1991) Health effects of recreational running in women. Some epidemiological and preventive aspects. Sports Med., 11, 20–51. masala a, delitala g, devilla l, alagna s and rovasio pp (1979) Enhancement of insulin-induced prolactin secretion by fluoxetine in man. J. Clin. Endocrinol. Metab., 49, 350–352. matuszkiewicz-rowinska j, skorzewska k, radowicki s, sokalski a, przedlacki j, niemczyk s, wlodarczyk d, puka j and switalski m (1999) The benefits of hormone replacement therapy in pre-menopausal women with oestrogen deficiency on haemodialysis. Nephrol. Dial. Transplant, 14, 1238–1243. melman a, fersel j and weinstein p (1984) Further studies on the effect of chronic alpha-methyldopa administration upon the central nervous system and sexual function in male rats. J. Urol., 132, 804–808. meltzer hy, fang vs, tricou bj and robertson a (1982) Effect of antidepressants on neuroendocrine axis in humans. Adv. Biochem Psychopharmacol., 32, 303–316. meston cm (2004) A randomized, placebo-controlled, crossover study of ephedrine for SSRI-induced female sexual dysfunction. J. Sex Marital Ther., 30, 57–68. meston cm, gorzalka bb and wright jm (1997) Inhibition of subjective and physiological sexual arousal in women by clonidine. Psychosom. Med., 59, 399–407. metcalfe cd, metcalfe tl, kiparissis y, koenig bg, khan c, hughes rj, croley tr, raymond em and potter t (2001) Estrogenic potency of chemicals detected in sewage treatment plant effluents as determined by in vivo assays with Japanese medaka (Oryzias latipes). Environ. Toxicol. Chem., 20, 297–308. metcalfe cd, miao xs, koenig bg and struger j (2003) Distribution of acidic and neutral drugs in surface waters near sewage treatment plants in the lower Great Lakes, Canada. Environ. Toxicol. Chem., 22, 2881–2889. miao xs and metcalfe cd (2003) Determination of carbamazepine and its metabolites in aqueous samples using liquid chromatography-electrospray tandem mass spectrometry. Anal. Chem., 75, 3731–3738. miao xs, bishay f, chen m and metcalfe cd (2004) Occurrence of antimicrobials in the final effluents of wastewater treatment plants in Canada. Environ. Sci. Technol., 38, 3533–3541. mikkonen k, tapanainen p, pakarinen aj, paivansalo m, isojarvi ji and vainionpaa lk (2004) Serum androgen levels and testicular structure during pubertal maturation in male subjects with epilepsy. Epilepsia, 45, 769–776. mimeault c, woodhouse aj, miao xs, metcalfe cd, moon tw and trudeau vl (2005) The human lipid regulator, gemfibrozil bioconcentrates and reduces testosterone in the goldfish, Carassius auratus. Aquat. Toxicol., 73, 44–54. mitch ra (1989) Carbamazepine toxicity precipitated by intravenous erythromycin. DICP, 23(11), 878–879.

Pharmaceuticals

513

molitch me (1992) Pathologic hyperprolactinemia. Endocrinol. Metab. Clin. North Am., 21, 877–901. montejo-gonzalez al, llorca g, izquierdo ja, ledesma a, bousono m, calcedo a, carrasco jl, ciudad j, daniel e, de la gj, derecho j, franco m, gomez mj, macias ja, martin t, perez v, sanchez jm, sanchez s and vicens e (1997) SSRI-induced sexual dysfunction: fluoxetine, paroxetine, sertraline, and fluvoxamine in a prospective, multicenter, and descriptive clinical study of 344 patients. J. Sex Marital Ther., 23, 176–194. montgomery sa, baldwin ds and riley a (2002) Antidepressant medications: a review of the evidence for drug-induced sexual dysfunction. J. Affect. Disord., 69, 119–140. morrissette dl, skinner mh, hoffman bb, levine re and davidson jm (1993) Effects of antihypertensive drugs atenolol and nifedipine on sexual function in older men: a placebo-controlled, crossover study. Arch. Sex Behav., 22, 99–109. morrison jl, riggs kw and rurak dw (2005) Fluoxetine during pregnancy: impact on fetal development. Reprod. Fertil. Dev., 17, 641–650. moss hb (1983) More cases of anorgasmia after MAOI treatment. Am. J. Psychiatry, 140, 266. moss hb and procci wr (1982) Sexual dysfunction associated with oral antihypertensive medication: a critical survey of the literature. Gen. Hosp. Psychiatry, 4, 121–129. muller sc, el damanhoury h, ruth j and lue tf (1991) Hypertension and impotence. Eur. Urol., 19, 29–34. murdia a, mathur v, kothari lk and singh kp (1978) Sulpha-trimethoprim combinations and male fertility. Lancet, 2, 375–376. nafziger an, bertino js, jr, goss-bley ai and kashuba ad (1999) Incidence of sexual dysfunction in healthy volunteers on fluvoxamine therapy. J. Clin. Psychiatry, 60, 187–190. nash jp, kime de, van der ven lt, wester pw, brion f, maack g, stahlschmidtallner p and tyler cr (2004) Long-term exposure to environmental concentrations of the pharmaceutical ethynylestradiol causes reproductive failure in fish. Environ. Health Perspect., 112(17), 1725–1733. newman rj and salerno hr (1974) Letter: Sexual dysfunction due to methyldopa. Br. Med. J., 4, 106. nulman i and koren g (1996) The safety of fluoxetine during pregnancy and lactation. Teratology, 53, 304–308. oehninger s and alexander nj (1991) Male infertility: the focus shifts to sperm manipulation. Curr. Opin. Obstet. Gynecol., 3, 182–190. o’grady jp, caldwell bv, auletta fj and speroff l (1972) The effects of an inhibitor of prostaglandin synthesis (indomethacin) on ovulation, pregnancy, and pseudopregnancy in the rabbit. Prostaglandins, 1, 97–106. olive dl, lindheim sr and pritts ea (2003) Endometriosis and infertility: what do we do for each stage? Curr. Womens Health Rep., 3, 389–394. orme mle, back dj and breckenbridge am (1983) Clinical pharmacokinetics of oral contraceptives. Clin. Pharmacokinetics, 8, 95–136. orme m, back dj, ward s, et al. (1991) The pharmacokinetics of ethynylestradiol in the presence and absence of gestodene and desogestrel. Contraception, 43, 305–316. orn s, holbech h, madsen th, norrgren l and petersen gi (2003) Gonad development and vitellogenin production in zebrafish (Danio rerio) exposed to ethinylestradiol and methyltestosterone. Aquat. Toxicol., 65(4), 397–411. ou yc, hwang ti, yang cr, chang cl, chang ch, wu hc and hwang yf (1991) Hormonal screening in impotent patients. J. Formos. Med. Assoc., 90, 560–564.

514

Endocrine-disrupting chemicals in food

pall m, friden be and brannstrom m (2001) Induction of delayed follicular rupture in the human by the selective COX-2 inhibitor rofecoxib: a randomized doubleblind study. Hum. Reprod., 16, 1323–1328. papoulias dm, noltie db and tillitt de (1999) An in vivo model fish system to test chemical effects on sexual differentiation and development: exposure to ethinyl estradiol. Aquatic Toxicol., 48, 37–50. patel r and rosengren rj (2001) Acetaminophen elicits anti-estrogenic but not estrogenic responses in the immature mouse. Toxicol. Lett., 122, 89–96. pentland b, anderson da and critchley ja (1981) Failure of ejaculation with indoramin. Br. Med. J. (Clin. Res. Ed.), 282, 1433–1434. peterson ew, davis rk and orndorff ha (2001) 17β-Estradiol as an indicator of animal waste contamination in mantled karst aquifers. J. Environ. Qual., 29, 826–834. pierce jr, jr (1983) Cimetidine-associated depression and loss of libido in a woman. Am. J. Med. Sci., 286, 31–34. pizarro s, bargay j and d’agosto p (1990) Gemfibrozil-induced impotence. Lancet, 336, 1135. pohl r (1983) Anorgasmia caused by MAOIs. Am. J. Psychiatry, 140, 510. pohland rc, byrd tk, hamilton m and koons jr (1989) Placental transfer and fetal distribution of fluoxetine in the rat. Toxicol. Appl. Pharmacol., 98, 198–205. prescott lf (2000) Paracetamol: past, present, and future. Am. J. Ther., 7, 143–147. purdom ce, hardiman pa, bye vj, eno nc, tyler cr and sumpter j (1994) Estrogenic effects of effluents from sewage treatment works. Chem. Ecol., 8, 275–285. ranney re (1977) Comparative metabolism of 17α-ethynyl steroids used in oral contraceptives. J. Toxicol. Environ. Health, 3, 139–166. rathner m von and sonneborn m (1979) Biologishe wirksame Oestrogene in Trinkund Abwasser. Forum Staedte. Hygiene, 30, 45–49. reel jr, lawton ad and lamb jc (1992) Reproductive toxicity evaluation of acetaminophen in Swiss CD-1 mice using a continuous breeding protocol. Fundam. Appl. Toxicol., 18, 233–239. richardson ml and bowron jm (1985) The fate of pharmaceutical chemicals in the aquatic environment. J. Pharm. Pharmacol., 37, 1–12. roberts ph and thomas kv (2006) The occurrence of selected pharmaceuticals in wastewater effluent and surface waters of the lower Tyne catchment. Sci. Total Environ., 356(1–3), 143–153. robinson cd, brown e, craft ja, davies im, moffat cf, pirie d, robertson f, stagg rm and struthers s (2003) Effects of sewage effluent and ethynyl oestradiol upon molecular markers of oestrogenic exposure, maturation and reproductive success in the sand goby (Pomatoschistus minutus, Pallas). Aquatic Toxicol., 62, 119–134. rosen rc and marin h (2003) Prevalence of antidepressant-associated erectile dysfunction. J. Clin. Psychiatry, 64 Suppl 10, 5–10. rosen rc, kostis jb and jekelis aw (1988) Beta-blocker effects on sexual function in normal males. Arch. Sex Behav., 17, 241–255. rosen rc, lane rm and menza m (1999) Effects of SSRIs on sexual function: a critical review. J. Clin. Psychopharmacol, 19, 67–85. rossi ag, soares jm, jr, motta el, simoes mj, oliveira-filho rm, haidar ma, rodrigues dl and baracat ec (2002) Metoclopramide-induced hyperprolactinemia affects mouse endometrial morphology. Gynecol. Obstet. Invest., 54, 185–190. roste ls, tauboll e, haugen tb, bjornenak t, saetre er and gjerstad l (2003) Alterations in semen parameters in men with epilepsy treated with valproate or carbamazepine monotherapy. Eur. J. Neurol., 10, 501–506.

Pharmaceuticals

515

rudkin l, taylor mj and hawton k (2004) Strategies for managing sexual dysfunction induced by antidepressant medication. Cochrane Database Syst. Rev., CD003382. schardein jl (1980) Studies on the components of an oral contraceptive agent in albino rats. II. Progestogenic component and comparison of effects of the components and the combined agent. J. Toxicol. Environ. Health, 6, 895–906. schlegel pn, chang ts and marshall ff (1991) Antibiotics: potential hazards to male fertility. Fertil. Steril., 55(2), 235–242. scholz s and gutzeit ho (2000) 17-a-ethinylestradiol affects reproduction, sexual differentiation and aromatase gene expression of the medaka (Oryzias latipes). Aquatic Toxicol., 50, 363–373. segraves rt (1988) Sexual side-effects of psychiatric drugs. Int. J. Psychiatry Med., 18, 243–252. shen ww and mallya ar (1983) MAOI-induced inhibited female orgasm. Am. J. Psychiatry, 140, 1275. shen ww and park s (1982) Thioridazine-induced inhibition of female orgasm. Psychiatr. J. Univ. Ott., 7, 249–251. shen ww and sata ls (1983) Inhibited female orgasm resulting from psychotropic drugs. A clinical review. J. Reprod. Med., 28, 497–499. shen ww and sata ls (1990) Inhibited female orgasm resulting from psychotropic drugs. A five-year, updated, clinical review. J. Reprod. Med., 35, 11–14. sidi aa, cameron js, duffy lm and lange ph (1986) Intracavernous drug-induced erections in the management of male erectile dysfunction: experience with 100 patients. J. Urol., 135, 704–706. silvestri a, galetta p, cerquetani e, marazzi g, patrizi r, fini m and rosano gm (2003) Report of erectile dysfunction after therapy with beta-blockers is related to patient knowledge of side effects and is reversed by placebo. Eur. Heart J., 24, 1928–1932. slob w and pieters mn (1998) A probabilistic approach for deriving acceptable human intake limits and human health risks from toxicological studies: general framework. Risk Analysis, 18, 787–798. smaldone m, sukkarieh t, reda a and khan a (2004) Epilepsy and erectile dysfunction: a review. Seizure, 13, 453–459. smith pj and talbert rl (1986) Sexual dysfunction with antihypertensive and antipsychotic agents. Clin. Pharm., 5, 373–384. soliman ga and abla am (1999) Effects of antiepileptic drugs carbamazepine and sodium valproate on fertility of male rats. Dtsch. Tierarztl. Wochenschr., 106, 110–113. sparwasser c, schmelz hu, drescher p, eckert r and madsen po (1998) Role of intracellular Ca2+ stores in smooth muscle of human penile erectile tissue. Urol. Res., 26, 189–193. stanczyk fz (1994) Structure–function relationships, potency, and pharmacokinetics of progestogens. In: Lobo RA, ed. Treatment of the Postmenopausal Woman: Basic and Clinical Aspects. New York: Raven Press; 69–89. stanczyk fz (1997) Pharmacokinetics of the new progestins and influence of gestodene and desogestrel on ethinylestradiol metabolism. Contraception, 55, 273–282. stanczyk fz and roy s (1990) Metabolism of levonorgestrel, norethindrone, and structurally related contraceptive steroids. Contraception, 42, 67–96. stanczyk fz, leyden j, kempers s, et al. (2000) The effect of an oral contraceptive containing 20 μg ethinyl estradiol and 100 μg levonorgestrel (EE 20/LNG 100) on ovarian, adrenal, and peripheral androgens. Gynecol Endocrinol., 14(suppl), 137. Abstract R.T59.

516

Endocrine-disrupting chemicals in food

stanford ms and patton jh (1993) In utero exposure to fluoxetine HCl increases hematoma frequency at birth. Pharmacol. Biochem Behav., 45, 959–962. stoker te, goldman jm and cooper rl (2001) Delayed ovulation and pregnancy outcome: effect of environmental toxicants on the neuroendocrine control of the ovary (1). Environ. Toxicol. Pharmacol., 9, 117–129. stolker aa, niesing w, hogendoorn ea, versteegh jf, fuchs r and brinkman ua (2004) Liquid chromatography with triple-quadrupole or quadrupole-time of flight mass spectrometry for screening and confirmation of residues of pharmaceuticals in water. Anal. Bioanal. Chem., 378, 955–963. stone s, khamashta ma and nelson-piercy c (2002) Nonsteroidal anti-inflammatory drugs and reversible female infertility: is there a link? Drug Safety, 25, 545–551. stumpf m, ternes ta, haberer k and baumann w (1996) Nachweis von naturlichen und synthetischen Ostrogenen in Klaranlagen und Fliessgewassern. Vom Wasser (in German), 87, 251–261. tabak h, bloomhuff rn and bunch rl (1981) Steroid hormones as water pullutants II, Dev. Ind. Microbiol., 22, 497–519. tadokoro n, vollenhoven b, clark s, baker g, kovacs g, burger h and healy d (1997) Cumulative pregnancy rates in couples with anovulatory infertility compared with unexplained infertility in an ovulation induction programme. Hum. Reprod., 12, 1939–1944. tamura k, abe y and kogo h (2000) Phenytoin inhibits both the first ovulation and uterine development in gonadotropin-primed immature rats. Eur. J. Pharmacol., 398, 317–322. tanaka h, yakou y, takahashi a, higashitami k and komori m (2001) Comparison between estrogenicities estimated from DNA recombinant yeast assay and from chemical analyses of endocrine disruptors during sewage treatment. Water Sci. Technol., 43/2, 125–132. tanyildizi s and bozkurt t (2003) The effects of lincomycin–spectinomycin and sulfamethoxazole–trimethoprim on hyaluronidase activities and sperm characteristics of rams. J. Vet. Med. Sci., 65(7), 775–880. tauxe-wuersch a, de alencastro lf, grandjean d and tarradellas j (2005) Occurrence of several acidic drugs in sewage treatment plants in Switzerland and risk assessment. Water Res., 39, 1761–1772. teichmann at (1996) Empfängnisverhütung: eine vergleichende Übersicht aller Methoden, Risiken und Indikationen. Stuttgart, Germany: Georg Thieme Verlag; 23–31. ternes ta, stumpf m, mueller j, haberer k, wilken rd and servos m (1999) Behavior and occurrence of estrogens in municipal sewage treatment plants – I. Investigations in Germany, Canada and Brazil. Sci. Total Environ., 225, 81–90. thorneycroft i, stanczyk f, bradshaw k, et al. (1999) Effect of low-dose oral contraceptives on androgenic markers and acne. Contraception, 60, 255–262. timmermans l (1974) Influence of antibiotics on spermatogenesis. J. Urol., 112, 348–349. tinajero jc, fabbri a and dufau ml (1993) Serotonergic inhibition of rat Leydig cell function by propranolol. Endocrinology, 133, 257–264. turkington rw and macindoe jh (1972) Hyperprolactinemia in sarcoidosis. Ann. Intern. Med., 76, 545–549. uhler ml, hsu jw, fisher sg and zinaman mj (2001) The effect of nonsteroidal antiinflammatory drugs on ovulation: a prospective, randomized clinical trial. Fertil. Steril., 76, 957–961. vange n van der, blankenstein ma, kloosterboer hj, et al. (1990) Effects of several low-dose combined oral contraceptives on sex hormone binding globulin, corticosteroid binding globulin, total and free testosterone. Contraception, 41, 345–352.

Pharmaceuticals

517

vermeire t, pieters m, rennen m and bos p (2001) Probabilistic assessment factors for human health risk assessment. RIVM report 601516005, TNO report V3489. vogel d, gehring m, tennhardt l, weltin d and bilitewski b (2003) Mobility and fate of endocrine disrupting compounds (EDCs) in soil after application of sewage sludge to agricultural land. In: Pullammanappallil, Pr., McComb, A., Diaz, L. F., Bidlingmaier, W. (eds.): Proc. 4th Int. Conference ORBIT Ass. on Biol. Process. Organics: Adv. Sustain. Soc., April 30–May 2, 2003, Perth, Australia, vol. 1, 241–250. vose d (2000) Risk Analysis: A Quantitative Guide. John Wiley & Sons, Chichester. waghmarae d (1972) Infertility. East Afr. Med. J., 49, 698–702. wardle pg, mitchell jd, mclaughlin ea, ray bd, mcdermott a and hull mg (1985) Endometriosis and ovulatory disorder: reduced fertilisation in vitro compared with tubal and unexplained infertility. Lancet, 2, 236–239. warneke l (1994) Phenelzine and sexuality. J. Clin. Psychiatry, 55, 216–217. webb s (2000) Risk assessment for pharmaceuticals. In: Proceedings of International Seminar on Pharmaceuticals in the Environment. 9 March 2000, Brussels, Technological Institute. webb s, ternes t, gibert m and olejniczak k (2003) Indirect human exposure to pharmaceuticals via drinking water. Toxicol. Lett., 142, 157–167. webster j (1979) Male sexual dysfunction and cimetidine. Br. Med. J., 1, 889. weiner ri (1996) Cellular basis of the GnRH pulse generator. Nippon Sanka Fujinka Gakkai Zasshi, 48, 573–577. weiss rj (1991) Effects of antihypertensive agents on sexual function. Am. Fam. Physician., 44, 2075–2082. wenzel a, schäfers c, vollmer g, michna h and diel p (2001) Research Efforts Towards the Development and Validation of a Test Method for the Identification of Endocrine Disrupting Chemicals. Final Report Contract B6-7920/98/000015. white wb, andreoli jw and cohn rd (1985) Alpha-methyldopa disposition in mothers with hypertension and in their breast-fed infants. Clin. Pharmacol. Ther., 37, 387–390. wiegel s, aulinger a, brockmeyer r, harms h, loffler j, reincke h, schmidt r, stachel b, von tumpling w and wanke a (2004) Pharmaceuticals in the river Elbe and its tributaries. Chemosphere, 57, 107–126. wilens te, cohen l, biederman j, abrams a, neft d, faird n and sinha v (2002) Fluoxetine pharmacokinetics in pediatric patients. J. Clin. Psychopharmacol., 22, 568–575. wirshing da, pierre jm, marder sr, saunders cs and wirshing wc (2002) Sexual side effects of novel antipsychotic medications. Schizophr Res., 56, 25–30. xu s, zhu bt, cai mx and conney ah (2001) Stimulatory effect of clofibrate on the action of estradiol in the mammary gland but not in the uterus of rats. J. Pharmacol. Exp. Ther., 297, 50–56. yarkoni s, spitz im, ben david m and polishuk wz (1978) Failure of bromocriptine to suppress prolactin in majeptil-induced hyperprolactinemia. Acta. Eur. Fertil., 9, 185–188. young ah and ashton ch (1996) Pharmacokinetics of fluoxetine. Trends Pharmacol. Sci., 17, 400. ying g-g, kookana rs and ru x-j (2002) Occurrence and fate of hormone steroids in the environment. Environment Int., 28, 545–551. zanagnolo v, dharmarajan am, endo k and wallach ee (1996) Effects of acetylsalicylic acid (aspirin) and naproxen sodium (naproxen) on ovulation, prostaglandin, and progesterone production in the rabbit. Fertil. Steril., 65, 1036–1043. zawar mp, deshpande nm, gadgil pa and mahanta aa (2003) Histopathological study of endometrium in infertility. Indian J. Pathol. Microbiol., 46, 630–633.

518

Endocrine-disrupting chemicals in food

zillioux ej, johnson ic, kiparissis y, metcalfe cd, wheat jv, ward sg and liu h (2001) The sheepshead minnow as an in vivo model for endocrine disruption in marine teleosts: a partial life-cycle test with 17-alpha-ethynylestradiol. Environ. Toxicol. Chem., 20, 1968–1978. zipper j, bruzzone me, angelo s, munoz v and wheeler rg (1982) Effect of topically applied adrenergic blockers on fertility. Int. J. Fertil., 27, 242–245. zuccato e, castiglioni s and fanelli r (2005) Identification of the pharmaceuticals for human use contaminating the Italian aquatic environment. J. Hazard. Mater., 122, 205–209. zuehlke s, duennbier u and heberer t (2004) Determination of polar drug residues in sewage and surface water applying liquid chromatography–tandem mass spectrometry. Anal. Chem., 76, 6548–6554.

19 Endocrine-active ultraviolet filters and cosmetics M. Schlumpf and W. Lichtensteiger, GREEN Tox, Switzerland

Abstract: Several UV filters used in sunscreens and other cosmetics exhibit endocrine activity. 4-Methylbenzylidene camphor and 3-benzylidene camphor were studied in rats for actions on the developing organism, a sensitive life stage. Effects on reproductive organs and brain, including gene expression, were seen in neonatal and adult offspring. Puberty was delayed in males, sexual behavior impaired in females. Evidence for transdermal passage and presence in fish suggests potential human exposure. For information at population level, we conducted a chemical-analytical study on human milk. Some 78.8% of women reported use of products containing UV filters in a questionnaire, and 76.5% of milk samples contained UV filters. Key words: UV filters, 4-methylbenzylidene camphor (4-MBC), 3-benzylidene camphor (3-BC), developmental toxicity, human milk, food chain.

19.1

Chemicals used as ultraviolet filters

Ultraviolet (UV) filters are organic molecules containing light-absorbing chromophores that absorb light quanta in defined wavelength ranges; UVA (400–320 nm), UVB (320–280 nm) and UVC (below 280 nm). The energy resulting from the absorption process is converted into infrared light. Physical filters such as titanium dioxide and zinc oxide mainly scatter and reflect UV rays. Organic UV filters with different individual absorption maxima can be combined to offer protection over the entire spectrum of relevant wavelengths (Salvador and Chisvert, 2005). Legislation for UV filters used in cosmetics differs among countries (cf. Salvador and Chisvert, 2007), but initiatives for global harmonization of sunscreen testing are under way (agreement of cosmetics industry bodies, Colipa Annual General Meeting, 2008). In the US and in Australia, sunscreen products are classified as OTC (over the counter) drugs. In the US,

520

Endocrine-disrupting chemicals in food

the regulation is described in Title 21 of the Code of Federal Regulations (21 CFR), Part 352, published by the American Food and Drug Administration (FDA). In Australia, the corresponding information can be found in the Australian Regulatory Guidelines for OTC Medicines (ARGOM, 2003) published by the Therapeutic Goods Administration (TGA). In the European Union (EU), Japan and Switzerland, sunscreen products are classified as cosmetics. Regulations are laid down for the EU in Annex VII of the EU Cosmetics Directive (Council Directive 76/786/EEC, with periodic review of allowed products by the Scientific Committee on Consumer Products (SCCP)), for Japan in the Pharmaceutical Affairs Law (PAL) and Appendix 4 of the Standards for Cosmetics (Ministry of Health and Welfare, 2000), for Switzerland in the ‘Verordnung des EDI vom 23.11.2005 über kosmetische Mittel’ (Vkos, last update 1 April 2008). In the various legislations, UV filters allowed for use in sunscreen products are listed in positive lists. Other UV-absorbing chemicals used exclusively for protection of the product against UV rays (UV absorbers) are not considered in these documents and need not be declared, in spite of considerable overlap in chemical structure. The compounds used in sunscreen products for skin protection can also be used for product protection and are then designated as UV absorbers. The toxicological evaluation of a new UV filter includes tests for acute toxicity, irritation, sensitization, phototoxicity, photosensitization, subchronic and chronic toxicity, reproductive toxicity, genotoxicity, photogenotoxicity and carcinogenicity (Nohynek and Schaefer, 2001). The endocrine activity of certain UV filters has not been detected by the existing regulatory process. This problem is not a special feature of UV filters, but is observed with a broad range of chemicals used in very different products. Problems for assessment of safety of cosmetics are to be expected in the future because the EU ‘Directive 2003/15 EC’ prohibits marketing of cosmetic products or ingredients that have been subject of animal testing. With respects to tests concerning repeated-dose toxicity, reproductive toxicity and toxicokinetics, for which there presently is no alternative, the period for implementation has been limited to a maximum of 10 years after the entry into force of the Directive.

19.2

Endocrine activity and developmental toxicity of ultraviolet filters

19.2.1 Endocrine activity of ultraviolet filters Studies in MCF-7 cells and in uterotrophic assays on immature rats have revealed that several UV filters exhibit estrogenic activity (Schlumpf et al., 2001a, 2004a). This finding was confirmed in cell cultures as well as in mammalian and fish test systems (Holbech et al., 2002; Schreurs et al., 2002;

Endocrine-active ultraviolet filters and cosmetics

521

Tinwell et al., 2002; Inui et al., 2003; Mueller et al., 2003; Jarry et al., 2004, Seidlova-Wuttke et al., 2004; Klammer et al., 2005; Kunz and Fent, 2006). Some UV filters also exhibit anti-androgenic activity in vitro (Ma et al., 2003; Schreurs et al., 2005). In chronic (3 months) toxicity studies in adult ovariectomized rats, 4methylbenzylidene camphor (4-MBC), octylmethoxycinnamate (OMC, now called ethylhexyl-methoxycinnamate (EHMC)) and benzophenone-2 (Bp-2) have been shown to have estradiol (E2)-like effects on fat deposits, lipid metabolism and serum leptin. 4-MBC and Bp-2 also have antiosteoporotic effects (Seidlova-Wuttke et al., 2004, 2005, 2006a,b), and 4-MBC, OMC/EHMC and Bp-2 interact with the thyroid axis (Schmutzler et al., 2004; Schlumpf et al., 2004b; Seidlova-Wuttke et al., 2005; Maerkel et al., 2007).

19.2.2

Developmental toxicity of 4-methylbenzylidene camphor and 3benzylidene camphor During pre- and early postnatal life, the developing organism is highly sensitive to estrogenic and anti-androgenic chemicals because sex hormones control developmental processes. In mammals, the male sexual phenotype is established by testosterone synthesized by the fetal testis. In many target tissues, the developmental action of testosterone is mediated by androgen receptors (AR), but in some targets, such as the brain, testosterone is converted to 17β-estradiol by intraneuronal aromatase (MacLusky and Naftolin, 1981; Lauber and Lichtensteiger, 1994), and the developmental effect, in this case male sexual brain differentiation, is achieved by 17β-estradiol acting on estrogen receptors (ER) (MacLusky and Naftolin, 1981). Some peripheral tissues in the developing male, such as the urogenital sinus that gives rise to the prostate, express 5α-reductase as well as aromatase activities (George, 1993; Pezzi et al., 2003) and hence can convert testosterone into the AR agonist dihydrotestosterone or the ER agonist 17β-estradiol. Since these tissues express ER and AR, they provide sites of action for estrogenic as well as anti-androgenic chemicals. With the exception of tissues that convert testosterone to 17β-estradiol, endogenous estrogen levels are very low in fetuses of both sexes in rodents (Habert and Picon, 1984) and also comparatively low in humans (SilerKhodr, 1998), thus facilitating the competition of weak xenoestrogens for the ERs. The potential relevance of human exposure to estrogenic UV filters was investigated in a mammalian model, the rat, using 4-MBC and 3benzylidene camphor (3-BC) (Schlumpf et al., 2001b, 2004b, 2008a,b; Durrer et al., 2005, 2007; Maerkel et al., 2005, 2007). Both compounds exhibit ERβ binding preference (Schlumpf et al., 2004a), but are also active in ERαtypical tests. The activity of 4-MBC is possibly due to the presence of an

522

Endocrine-disrupting chemicals in food

hydroxylated metabolite (Völkel et al., 2006). In order to mimic real exposure conditions, 4-MBC and 3-BC were administered to the parent generation, to the females during pregnancy and lactation, and to the offspring until adulthood. The presence of UV filters such as 4-MBC in human milk (see below) indicates exposure of the infant, but it is also a measure of internal exposure of the mother and thus indicates potential exposure of the unborn. Given its lipophilicity and a molecular weight below 500 (254.4), it is possible that 4-MBC crosses the placenta. The chemicals were given in the feed, which is the least stressful mode of application. Pregnancy and early postnatal period Female rats exposed to 4-MBC underwent normal pregnancies, while those exposed to 3-BC had lower weight gains during pregnancy. Higher doses of both UV filters reduced survival rates (Table 19.1; Schlumpf et al., 2001b, 2004b), and lowered thymus weights, indicating impairment of the immune system. Sexual development: puberty One of the most sensitive parameters, the onset of puberty, is delayed in males exposed to 4-MBC or 3-BC (Table 19.1; Durrer et al., 2007; Schlumpf et al., 2008a,b). 4-MBC and 3-BC thus mimic the typical effect of 17βestradiol in males (Biegel et al., 1998). Normal body weights at puberty indicate that the delay of puberty in males is not due to nutritional effects. In contrast, puberty is not significantly affected in females. In this sex, increased levels of 17β-estradiol advance the onset of puberty (Biegel et al., 1998). Reproductive organs of adult offspring Exposure to UV filters during pre- and postnatal life affects the reproductive organs of male rats. The testes of rats exposed to 4-MBC show decreased weight before puberty, at postnatal day 14 (Schlumpf et al., 2001b) and a small increase in relative weight at the highest dose (47 mg/kg) when the rats reach adulthood (postnatal day 84). The adult finding is reminiscent of reports on the effects of neonatal administration of weak environmental estrogens (Atanassova et al., 2000; Putz et al., 2001). A marked decrease in prostate weight at higher doses of 4-MBC resembles the effect of pre- or early postnatal administration of the potent ER agonist diethylstilbestrol (DES, vom Saal et al., 1997, Atanassova et al., 2000). This suggests a differential sensitivity among different male target organs. Exposure to 3-BC also decreases prostate weight, but only at the lowest dose tested (Table 19.1), and does not affect testis weight. The two camphor derivatives also have different effects on the weights of reproductive organs in female rats (Table 19.1). Adult body weights are unaffected by 4-MBC but slightly reduced by the highest dose of 3-BC, possibly as a result of the estrogenic action of the compound (Biegel et al., 1998).

Gene expression in adult F1 prostate, mRNA/protein Androgen receptor (AR) dorsolateral prostate (DP) Androgen receptor (AR) ventral prostate (VP) AR mRNA down-regulation by estradiol in VP N-CoR protein, DP N-CoR protein, VP ↓ Ø

Ø/Ø Ø/Ø

↓/Ø Ø/(↓) ↓ ↓ (↓)

Ø ↓ Ø

Reproductive organs of adult offspring Testis, relative weight Prostate, ventral lobe relative weight2 Uterus, relative weight Ø Ø Ø

Delayed Ø

Ø Ø Ø /↓

↓ ↓ Ø /↓

Puberty Males (preputial separation) Females (vaginal opening)

Ø ↑ ↓

7

Ø Ø

0.7

↓/↓ ↓/↓ ↓ (↓) ↓

Ø ↓ ↑

↓/(↓) Ø/–

↑ ↓ Ø

Delayed Ø

↓

↓

Delayed Ø

↓

47

↓

24

4-Methylbenzylidene camphor (mg/kg peroral)

Effects of 4-MBC and 3-BC on selected end points in rat offspring

Early postnatal period Survival rate, PN 14 Prostate, PN 11, duct number and volume Testis, relative weight, PN 14 Uterus, PN 6, gene expression, mRNA Estrogen receptor-alpha Vascular endothelial growth factor (VEGF) mRNA ecNOS/iNOS mRNA

Table 19.1

↓ ↓ ↓ /↓

Ø Ø

0.07

↓ Ø

↑/↓ ↑/↑

Ø ↓ Ø

Ø Ø

↓ ↓ (↓) /↓

Ø Ø

0.24

↓ Ø

Ø/↓ ↑/Ø

Ø Ø Ø

Ø Ø

Ø

0.7

Ø/– ↑/–

Ø Ø Ø

Delayed Ø

Ø

2.4

↑/– Ø/–

Ø Ø Ø

Delayed Ø

↓

7

3-Benzylidene camphor (mg/kg peroral)

Continued

Ø

↓

↓ ↓ Ø ↓

↓/Ø ↓ (↓)

24

Ø

↓ ↓ Ø ↓

↑ Ø Ø Ø

(↓)

Ø/Ø

0.7

↓

Ø ↑ ↑ ↓

↓

Ø/Ø

2.4

irregul. irregul. irregul.

↑ Ø Ø ↑

Ø

↓

0.24

Ø/↑

0.07

↓/Ø

47

irregul.

↓

Ø ↑ ↑ ↓

Ø

↑/Ø

7

3-Benzylidene camphor (mg/kg peroral)

2

1

PN 1 = day of birth. Adult offspring 12 weeks old. Gene expression analysis in females in diestrus. Weight at 12 weeks only for ventral lobe for methodological reasons. 13 months’ data show decreased weight of ventral and dorsal + lateral lobes (Durrer et al., 2007). Ø: not significant. Blank or –: not analyzed. Data from Schlumpf et al. (2001b, 2008b) and work in preparation, Durrer et al. (2005, 2007), Maerkel et al. (2005, 2007), Hofkamp et al. (2008), O. Faass et al., in preparation, M. Fuetsch, C. Ehnes, C. Gaille, in preparation.

Ø

Estrous cycle

Ø ↓ Ø ↓

Ø/Ø ↓ Ø

7

↓

Ø/↓ Ø ↓

0.7

4-Methylbenzylidene camphor (mg/kg peroral)

Female sexual behavior

Adult functions

Ventromedial hypothalamic nucleus, adult F1, mRNA Estrogen receptor alpha, Males Females Progesterone receptor, Males Females

Gene expression in adult F1 uterus, mRNA/protein Progesterone receptor (PR-A protein) PR mRNA up-regulation by estradiol SRC-1 protein

Table 19.1

Endocrine-active ultraviolet filters and cosmetics

525

Estrogen target gene expression in prostate and uterus of adult offspring Exposure to 4-MBC or 3-BC during development significantly affects the expression of ER and estrogen target genes in specific tissues of the ventral and dorsolateral (dorsal + lateral) prostate and the uterus at mRNA and protein levels. The decrease in prostate weight induced by 4-MBC is accompanied by a decrease in AR and ER alpha at mRNA and protein levels, and by a reduction of insulin-like growth factor-I (IGF-I) at the mRNA level (Table 19.1, Fig. 19.1; Durrer et al., 2007). In the uterus of female rates exposed to 4-MBC, the target genes involved include ER alpha and progesterone receptors (PR) (Table 19.1; Durrer et al., 2005). As illustrated in Table 19.1, the two camphor derivatives also show different effects at the molecular level (Schlumpf et al., 2008a,b), in spite of their close structural relationship, while they have similar actions in acute in vitro and in vivo assays, and similar ER binding characteristics. Estrogen sensitivity and steroid receptor coregulators 4-MBC reduces the acute up-regulation of PR and IGF-I and the downregulation of ER alpha and AR by 17β-estradiol in the uterus, and decreases the acute down-regulation of AR and IGF-I in the ventral prostate (Table 19.1; Durrer et al., 2005, 2007). The reduced up-regulation of estrogen target genes in the uterus is accompanied by a decrease in steroid receptor coactivator-1 (SRC-1) protein levels, while the reduced repression of estrogen target genes in the prostate is accompanied by reduced levels of nuclear receptor corepressor (N-CoR) protein (Durrer et al., 2005, 2007). This suggests that changes in coregulator levels may be involved in the changes in estrogen sensitivity. Coregulators were among the most sensitive genes studied, with a lowest effective level of 0.7 mg/kg 4-MBC for both SRC-1 and N-CoR protein. Female sexual behavior and sexually dimorphic gene expression in brain As outlined above, the testis releases testosterone during early development of the embryo, which leads to elevated levels of intracellular 17β-estradiol in tissues of the male that convert testosterone to 17β-estradiol, such as the brain. In contrast, 17β-estradiol levels in the female remain low throughout pre- and early postnatal life, suggesting that the female brain should be particularly sensitive to estrogenic chemicals during early development. The original model of sexual brain differentiation assumed that the female brain developed without control by sex hormones, but recent studies have shown that female brain development depends on hormone (estrogen) actions (Bakker et al., 2002). While male sexual brain differentiation in primater may be mediated by androgen rather than estrogen receptors, female sexual behavior is estrogen – dependent in rodents as well as primates (Wallen and Zehr, 2004). Both 4-MBC and 3-BC strongly impair sexual behavior at low doses (Table 19.1, Fig. 19.1; Schlumpf et al., 2008a,b; O. Faass et al., 2009), affecting proceptive behavior (jumping and ear wiggling) displayed to attract

Endocrine-disrupting chemicals in food

100

*

75

**

**

50 25 0 C

(c) 2.0 1.5

*

0

+++ 8

8 8

Female behavior

(d)

8 *** ** *** 7 7 8

**

50

C 0.7 C 7 C 24 4-MBC (mg/kg)

7

0.5 0.0

100

C 7 24 47 C 7 24 47 Female Male 4-MBC (mg/kg)

7 6 5 4 3 2 1 0

Proceptive Lordosis Rejective behavior quotient behavior 100 **

** 80

** **

40 20

** C 7 24

60

%

1.0

AR protein DP 150

0.7 7 24 47 4-MBC (mg/kg/day) PR mRNA VMH

2.5 Normalized to Cyclophilin

(b)

AR mRNA DP 125

Percent of control

Percent of control

(a)

Mean events/15 min

526

C 7 24 C 7 24 4-MBC (mg/kg)

0

Fig. 19.1 Gene expression in reproductive organs and brain of adult rat offspring and female sexual behavior following pre- and postnatal exposure to 4methylbenzylidene camphor (4-MBC, 0.7, 7, 24, 47 mg/kg body weight per day in chow) or control chow (c). (a) Androgen receptor (AR) mRNA. (b) AR protein in dorsal + lateral prostate (DP). (c) Progesterone receptor (PR) in male and female ventromedial hypothalamic nucleus (VMH). mRNA measured by real-time reverse transcription polymerase chain reaction (RT PCR), values normalized to cyclophilin and expressed as percentage of the mean of untreated controls (VMH: untreated female controls). Mean ± SEM, n = 7–8 per group. Proteins analyzed by western blot in the same homogenates as used for mRNA determination, quantitated by densitometry relative to actin, and expressed as percentage of the mean of the corresponding control (mean ± SEM, n = 7–9). (d) Female sexual behavior: Proceptive behavior (jumping and ear wiggling, number of events/15 min observation period, left Y axis), receptive (lordosis) behavior (lordosis quotient, LQ = number of lordosis responses/number of mounts × 100, right Y axis). Rejective behavior towards the male (percentage of females exhibiting rejection of the male, right Y axis). Different from control, * p < 0.05, ** p < 0.01, *** p < 0.001. +++ Male controls different from female controls, p < 0.001. Data from Durrer et al. (2007), Maerkel et al. (2007), Faass et al. 2009.

the male, as well as receptive (lordosis) behavior. Female rats exposed to 4MBC during pre- and postnatal life have normal estrous cycles despite their disturbed sexual behavior, whereas pre- and postnatal exposure to 3-BC causes irregular cycles. Control of gonadal functions and control of sexual behavior are thus differentially affected. Gene expression studies of the two brain regions involved in the control of gonadal function and sexual behavior, the medial preoptic region (MPO) and ventromedial hypothalamic nucleus (VMH) (Maerkel et al., 2005, 2007;

Endocrine-active ultraviolet filters and cosmetics

527

Schlumpf et al., 2008a,b), have shown that developmental exposure to 4MBC and 3-BC causes sex- and region-specific changes in estrogen target genes (Maerkel et al., 2007; O. Faass et al., 2009, Schlumpf et al., 2008b). While there are again differences between the two UV filters, both result in a drop of PR mRNA in female VMH down to male levels at behaviorally active doses. Females exposed to these doses exhibited impaired female sexual behavior (Table 19.1; O. Faass et al., 2009). A similar correlation between PR mRNA level and behavior has been observed with a brominated flame retardant, the pentabrominated dipenylether, PBDE 99 (Lichtensteiger et al., 2004). A commercial mixture of polychlorinated biphenyls (PCB), Aroclor 1254, which has also been found to interfere with female sexual behavior (Chung et al., 2001) shows an analogous effect on PR mRNA in the VMH (Lichtensteiger et al., 2004). Lordosis behavior is directly correlated with the expression of PR mRNA in the female rat VMH (Pollio et al., 1993; Ogawa et al., 1994). The loss of sexual dimorphism of PR mRNA in VMH of the diestrus female thus appears to represent altered regulation of PR linked with behavioral impairment across different endocrine disrupters. Low-dose effects in neonatal rat offspring We are currently investigating the effects of low doses of UV filters on neonatal offspring, in collaboration with Jesus Tresguerres (University Complutense, Madrid), and effects on the developing prostate with Luke Hofkamp and Barry Timms (University of South Dakota, Vermillion, USA). In early postnatal uterus (postnatal day 6), significant effects on estrogen target gene mRNA levels were observed at doses as low as 0.07 mg/kg 3-BC and 0.7 mg/kg 4-MBC, the lowest doses studied (Table 19.1). The same doses also affected gene expression in postnatal brain (M. Fuetsch, C. Gaille, in preparation). The reduced expression of mRNAs encoding for vascular epithelial growth factor (VEGF), inducible nitric oxide synthase (iNOS) and, in part, endothelial cell nitric oxide synthase (ecNOS) in postnatal uterus seems particularly interesting. It suggests that angiogenesis and blood flow may be influenced. Effects on the developing prostate could be seen in the male neonate. 4-MBC increased the size of prostate, seminal vesicles and coagulating gland on the day of birth. In line with other findings on estrogenic chemicals, marked differences in growth responses of specific regions of the prostate were observed (Table. 19.1; Hofkamp et al., 2008).

19.3 19.3.1

Exposure to ultraviolet filters and other cosmetic ingredients

Presence of ultraviolet filters and fragrances in sewage sludge and surface waters The increasing use of UV filters in sunscreens with high sun protection factors (SPFs) justifies questions concerning the environmental impact of sunscreen ingredients. Publications on UV filters have increased from a

528

Endocrine-disrupting chemicals in food

single article in 1970 to more than 80 papers in 2004 (Salvador and Chisvert, 2005). UV filters applied to the skin may be directly introduced into surface waters during swimming and bathing. Another relevant source for UV filters is wastewater from private households and industries. Sewage sludge contains important xenobiotics and represents a matrix from which chemicals can be released from the anthroposphere into the environment (Plagellat et al., 2006). Sewage sludge samples in Switzerland were tested for synthetic perfume substances, including polycyclic musks, such as HHCB (Galaxolide), AHTN (Tonalide), ADBI (Celestolide), AHDI (Phantolide), ATII (Traseolide) and DPMI (Cashmeran), and a metabolite of HHCB (HHCB-lactone). Mean values in stabilized sludge from 16 wastewater treatment plants were 20.3 mg/kg dry matter (d.m.) HHCB, 7.3 mg/kg d.m. AHTN, and 1.8 mg/kg d.m. HHCB-lactone, with the remaining polycyclic musks in the range of 0.1–1.8 mg/kg d.m (Fig. 19.2; Kupper et al., 2004). Another important group of compounds found in pharmaceutical and personal care products (PPCPs) are organic UV filters. The concentrations of some UV filters found in sewage sludge are in the range of or higher than those of polychlorinated biphenyls (PCB) or polybrominated diphe-

10 000

mg / kg d.m.

1000 100 10 1 0.1

0.001

Carb. Diuron TPT PCB Perm. OMC HBCD TBT PBDE* DBP 4-MBC OC OT PAK AHTN HHCB DEHP NP LAS

0.01

Fig. 19.2 Concentrations of UV filters and synthetic perfumes in Swiss sewage sludge in relation to other environmental contaminants (mean values and range). UV filters: OMC: octylmethoxycinnamate (=EHMC), 4-MBC: 4-methylbenzylidene camphor, OC: octocrylene, OT: octyl-triazone. Polycyclic musks: AHTN: tonalide, HHCB: galaxolide. Other contaminants: Carb.: Carbendazim, TPT: triphenyl tin, PCB: polychlorinated biphenyls, Perm.: permethrin, HBCD: hexabromecyclododecane, TBT: tributyl tin, PBDE: polybrominated diphenylethers, DBP: dibutyl-phthalate, PAK: polycyclic aromatic hydrocarbons, DEHP: diethylhexylphthalate, LAS: lysine acetylsalicylate. Data compiled by Thomas Kupper (University of Applied Sciences for Agriculture, Zollikofen, Switzerland) from Plagellat et al., 2006 (UV filters), Kupper et al. (2004) (polycyclic musks), Berset and Holzer (1999, 2001) (PCB, phthalates), Kupper et al. (in preparation) (PBDE), with permission.

Endocrine-active ultraviolet filters and cosmetics

529

nylethers (PBDE) used as flame retardants (Berset and Holzer, 1999; Kupper et al., 2006; Plagellat et al., 2006). Octocrylene (OC) and octyltriazone (OT) have been found at concentrations of 320–18 740 μg/kg d.m. and 700–27 700 μg/kg d.m., respectively and 4-MBC and EHMC/OMC at concentrations of 150–4980 μg/kg d.m. and 10–90 μg/kg d.m. (Fig. 19.2, Plagellat et al., 2006). Phthalates were also present, mostly at concentrations below 1 mg/kg d.m. (Berset and Etter-Holzer, 2001), with the most common being di-(ethylhexyl) phthalate (DEHP, 21–114 mg/kg d.m.). UV filters present in sewage sludge originate mainly from private households, with surface runoff and industry being additional sources (Plagellat et al., 2006). The authors argue that UV filters might also be released to the environment by volatilization or leaching from plastics and other materials. Wastewater treatment removes a significant amount of polycyclic musks and UV filters (Kupper et al., 2006), but 1–8% of UV filters are released into the environment, which may represent considerable quantities in the case of highvolume chemicals. In Swiss lake water, several UV filters, including 4-MBC, OMC/EHMC, OC, OT and benzophenone-3 (Bp-3), have been detected at concentrations up to 125 ng/L, with the highest levels found in summer. River water analysis and comparisons with chemicals released exclusively from wastewater treatment plants (WWTPs) indicate that UV filters in surface waters originate from direct inputs as well as from WWTPs (Table 19.2; Poiger et al., 2004; Balmer et al., 2005).

19.3.2 Ultraviolet filters and fragrances in the food chain: fish The release of PPCPs into wastewater and their presence in surface waters, soil and sediments may lead to bioaccumulation in the tissues of aquatic organisms. This has raised the question of secondary effects in these animals

Table 19.2 UV filters in Swiss lake water and in fish from lakes and rivers. Concentration range across different seasons, river fish caught in September. Data from Poiger et al. (2004), Balmer et al. (2005), Buser et al. (2006) 4-MBC

Bp-3

OC

EHMB/OMC

Lake Zurich (large) Water (ng/L) Fish (roach) (ng/g lipid)

<2–22 73–80

<2–20 92–112

<2 n.d.

<2–26 64

Lake Hütten (small) Water (ng/L) Fish (perch) (ng/g lipid)

<2–82 44–166

5–125 66–123

<2–27 <2–25

<2–19 n.d.

7 Rivers Fish (perch) (ng/g lipid)

50–1800

105–2400

530

Endocrine-disrupting chemicals in food

(Ramirez et al., 2007). Two polycyclic musks, HHCB and AHTN, have been detected in a broad range of trophic levels, including lugworm, clam, crustacean, fish, marine mammals and birds (Nakata et al., 2007). Fish are a primary organism for monitoring the occurrence of persistent lipophilic contaminants. The presence of UV filters in fish was originally reported by Nagtegaal and coworkers (1997) and has recently been analyzed in more detail in Swiss lakes at different times of the year. The UV filters 4-MBC, OMC/EHMC, OC and Bp-3 have been detected in white fish, roach and perch (Table 19.2; Balmer et al., 2005). Brown trout (Salmo trutta fario) caught in seven different rivers between 0.1 and 0.7 km downstream of WWTPs have been found to have lipid weight-based concentrations of up to 1800 ng/g 4-MBC and 2400 ng/g OC (Table 19.2, Buser et al., 2006). These values were considerably higher than those found in white fish (Coreganus sp.) and roach (Rutilus rutilus) from Swiss lakes with inputs from WWTPs (Balmer et al., 2005). The data suggest a higher availability of these contaminants for fish in rivers than in lakes, and at the same time confirm that WWTPs are a major source of UV filters in the aquatic environment. There is a much greater variation in UV filter (4-MBC) concentrations in river fish of similar age collected at the same site than in lake fish, suggesting larger exposure differences in river fish (Buser et al., 2006). While synthetic musks were also detected in fish from remote alpine lakes, indicating airborne transport of these chemicals (Schmid et al., 2007), a search for UV filters in one Alpine lake was negative (Balmer et al., 2005). Bioaccumulative substances are usually defined as hydrophobic, fatsoluble chemicals with high octanol/water partition coefficients (Kow). However, Kelly and coworkers (2007) have recently drawn attention to the fact that bioaccumulation in air-breathing animals (including humans) presents different conditions. Certain poorly metabolized, moderately hydrophobic organic chemicals with low Kow that have a low rate of respiratory elimination (high Koa = octanol/air coefficient) may also show bioaccumulation.

19.3.3

Human exposure to cosmetics: ultraviolet filters and fragrances in human milk Assessing chemical risks requires information on the quality and quantity of chemicals present during critical and sensitive life stages. UV filters may reach the human body through direct application of cosmetics (Hayden et al., 1997; Janjua et al., 2004; Schauer et al., 2006) or through the food chain. Analysis of human milk provides information on the exposure of mother and fetus and on the quality and degree of contamination of the food given to the young infant. Most of the data on human milk pollution relates to organochlorine compounds, many of which are now banned. Epidemiological studies provide suggestive evidence that prenatal and/or postnatal exposure to organochlorines can produce health deficits, including effects on the

Endocrine-active ultraviolet filters and cosmetics

531

developing nervous and immune systems, thyroid hormone status, liver function, physical size and activity (Hoover, 1999). Unfortunately, the fact that the levels of organochlorine compounds in human milk exhibit a decreasing trend is often mistaken as evidence that the problems of chemical exposure for babies are being solved, without considering possible additional exposure to new contaminants appearing in food chains, cosmetics, pharmaceutical preparations and household detergents. It is important to note that a German study of breast-feeding found that it was associated with about 30% higher median concentrations of DDE, HCB and PCBs and a 30% increase in mean PCDD/PCDF levels (Link et al., 2005). Aside from organochlorines, several groups of chemicals have recently been identified in human milk, including phthalates and constituents of body care products, synthetic musk perfumes and UV filters (Main et al., 2006; Schlumpf et al., 2008b and in preparation). The Basel human milk cohort Information on internal human exposure to environmental chemicals apart from classical persistent organic pollutants (POPs; Wong et al., 2005; Sun et al., 2006) is sparse and only few studies include more recent pollutants (Inoue et al., 2006; So et al., 2005). Enormous information gaps exist with regard to exposure to xenobiotics such as cosmetic ingredients, food additives and indoor air pollutants. To begin to address this, a human milk monitoring study was carried out at the Department of Obstetrics of the University Hospital of Basel (Switzerland) to investigate different groups of xenobiotics, with a focus on endocrine-active substances including UV filters (4-MBC, 3-BC, Bp-2, Bp-3, EHMC/OMC, OC, octyldimethyl-PABA, homosalate), synthetic perfumes (musk xylene, musk ketone, six polycyclic musks including HHCB (Galaxolide) and AHTN (Tonalide)), brominated flame retardants (six polybrominated diphenylethers), organochlor pesticides (including DDT/DDE, methoxychlor, α-, β-, γ-cyclohexane, hexachlorobenzene), seven PCB congeners, cylcodiene insecticides (aldrin, dieldrin, chordane, endrine, bromocyclene), and others. Participating mothers completed questionnaires about lifestyle and their personal use of cosmetics, including sunscreens, lipsticks, skin care creams, perfumes, body lotions, shower gels and deodorants, and frequency of application. Studies were conducted in 2004 (pilot study), 2005 and 2006 (Table 19.3). The studies comprised 13 (fall 2004), 21 (2005) and 20 (2006) mother–child pairs. Milk sampling started in the first week of life because this could yield information regarding exposure of the mother during pregnancy and of the neonate following birth, and because the neonatal period was considered to be particularly sensitive. Presence of xenobiotics in human milk Sampling, extraction and chemical analyses of UV filters in human milk were performed as described by Schlumpf et al. (2008b). UV filters were

Table 19.3 Selected UV filters, synthetic fragrances, phthalates and organochlorines in human milk1,2

Compounds

Women using product with compound n = 34

Milk samples with compound n = 34

Concentration range in human milk fat of positive samples

% Total

%Total

ng/g fat

Concentration range in human whole milk as a function of fat content3 ng/kg milk 1.5% fat

UV filters UV filters in sunscreens UV filters in different cosmetics All products with UV filters EHMC/OMC OC Bp-3 4-MBC

5.4% fat

45.5 60.6 78.8

76.5

58.8 38.2 14.7 17.7

64.7 47.1 18.2 11.8

Synthetic fragrances Total synthetic 96.97 fragrances Nitro musks Not declared in Musk xylene product Musk ketone Not declared in Polycyclic product musks Galaxolide, HHCB Tonalide, AHTN Polychlorinated biphenyls (PCB) PCB 118 General PCB 138 environmental PCB 153 contaminants PCB 180 Organochlorines Total DDT Source and DDT not known metabolites pp′-DDE

2.1–78.1 4.7–77.5 7.3–121.4 6.7–19.0

32–1172 71–1163 110–1821 101–285

113–4217 254–4185 394–6556 362–1026

97.1 82.0

0.3–31.6 0.3–12.0

4.5–474 4.5–180

16.2–1706 16.2–648

91.2

8.7–309.7

131–4646

470–16 724

17.7

2.7–28.8

40.5–432

145.8–1556

100.0 100.0 100.0 100.0

4.1–35.4 18.2–101.0 20.8–133.0 8.3–83.8

61.5–531 273–1515 312–1995 124.5–1257

221.4–1912 982.8–5454 1123–7182 448.2–4525

100.0

59.5–522.8

893–7842

3213–28 231

100.0

47.3–431.9

710–6479

2556–23 323

100.0

Pilot study (2004) n = 13, Study 1 (2005) n = 21, Schlumpf et al. (2008b). Combined analysis of all three cohorts including 2006 (n = 54) in preparation. Most frequently detected UV filters: EHMC/OMC (ethylhexylmethoxy cinnamate/octylmethoxy cinnamate), OC (octocrylene), Bp-3 (benzophenone-3), 4-methylbenzylidene camphor (4-MBC). Data for 3-benzylidene camphor, benzophenone-2, homosalate, and octyldimethylamino-benzoic acid, see Schlumpf et al. (2008b). 3 1.5% = lowest and 5.4% highest fat content determined in the study. n.d.: not detected. 1

2

Endocrine-active ultraviolet filters and cosmetics

533

determined by gas chromatography coupled to low-resolution mass spectrometry (GC-LRMS) with mass spectrometry detection with electrospray ionization (MSD-EI) mode. An overview of the preliminary data is given in Table 19.3. Synthetic perfumes were found in all milk samples, even though a few women declared they did not use synthetic perfumes so it must be assumed that the perfumes originated from other cosmetics. Some 78.8% of the women used products containing UV filters, and UV filters were detected in 76.5% of the milk samples. Concentration ranges were large. Table 19.3 illustrates the role of fat content. Chemicals that are measured per gram of milk fat will have higher concentrations at increased milk fat levels. This aspect is relevant because of the large variation of fat concentration in human milk, but is sometimes overlooked since data are usually expressed in relation to milk fat. For the two most frequently occurring compounds, EHMC/OMC and octocrylene, presence in milk was significantly correlated with use on an individual basis (p < 0.05). Interestingly, about half the women used UV filter-containing products other than sunscreens and hence may not have been aware that they used UV filters.

19.4

Considerations of human risk

19.4.1 Neonatal sensitivity Protecting humans and animals from the adverse effects of chemical toxicants has become a major challenge to society today. During the last decade, it has become increasingly evident that health can be affected by toxic chemicals at much lower levels than previously believed. The developing organism is particularly susceptible to chemical insults (Lichtensteiger and Schlumpf, 1993; Pohl and Hibbs, 1996). Epidemiological studies have revealed associations between decreased anogenital distance in male infants (indicating insufficient male sexual differentiation) and low-level phthalate exposure (Swan et al., 2005), and between the occurrence of congenital cryptorchidism and levels of PBDEs or pesticides in human milk (Damgard et al., 2006; Main et al., 2007). Reproductive hormone profiles were also found to be altered in newborn boys in relation to phthalate exposure (Main et al., 2006). Exposure to chemicals at an early age may not cause effects that are immediately detectable in the baby. A chemical-induced functional failure might only become manifest at the age when functional challenge occurs, e.g. at school age for mental development or at adult age when reproductive functions start (Vandenberg et al., 2007). According to our preliminary data (Table 19.3; Schlumpf et al., 2008b), xenobiotic patterns in human milk are impressive both with regard to the variety of xenobiotics present and the quantities of certain xenobiotics. The milk samples of individual mothers represent pooled samples of between two and ten feeds, which reduces the impact of daily variations. Since sampling started during the first week after parturition, the data provide

534

Endocrine-disrupting chemicals in food

indirect information on in utero exposure and direct information on levels administered to the suckling infant. As xenobiotic loads in human milk diminish with continuous nursing, exposure is highest during the early postnatal period.

19.4.2 Persistent organic pollutants: unexpected high concentrations An unexpected finding was the comparatively high levels of some organochlorines (DDT/DDE, hexachlorobenzene, PCB congeners, Table 19.3). Some of these compounds belong to the so-called dirty dozen POPs that are due to be phased out by 2012. Calculated as tolerable daily intake for the baby on the basis of milk fat concentrations of individual samples, some values found for DDT exceeded the concentration tolerances given in the Swiss Ordinance for Xenobiotics in Food (FIV, 2005).

19.4.3

Ultraviolet filter levels in relation to developmental toxicity in the rat model The results of our monitoring study demonstrate that UV filters are present in human milk at significant levels, in the range of the three PCB congeners that are most abundant in biological specimens and interact with ER (PCB138, PCB153, PCB180; Bonefeld-Jorgensen et al., 2001). Some 76.5% of the women studied had UV filters in their milk. This indicates that internal human exposure could be more widespread and relevant than may have been expected from transdermal absorption models and acute experiments in humans (Janjua et al., 2004; Schauer et al., 2006). According to our studies, pre- and postnatal exposure to the UV filters 4-MBC and 3-BC interferes with the development of reproductive organs and brain and behavior. Conventional developmental and reproductive end points (puberty, organ weights, behavior) exhibited LOAEL (lowest observed adverse effect level) values for 4-MBC of 7 mg/kg body weight, and NOAEL (no observed adverse effect level) values of 0.7 mg/kg. The corresponding values for 3-BC were 0.24 mg and 0.07 mg/kg, respectively. Molecular end points in the brain and periphery were found to be affected at the lowest doses studied, i.e. 0.7 mg/kg 4-MBC and 0.07 mg/kg 3-BC. At the LOAEL dose of 7 mg/kg 4MBC for conventional end points, rat milk was found to contain 208.6 ng 4-MBC/g milk lipid. This is only 11 times the maximal 4-MBC concentration detected in human milk of the first two cohorts (2004, 2005) (Schlumpf et al., 2008b). This suggests that an acceptable margin of safety (MOS), defined as the quotient MOS = exposure level/NOAEL ≥ 100, may not be reached. The finding that one week’s percutaneous administration of 4-MBC and two other UV filters (EHMC/OMC and Bp-3) to adult human volunteers did not significantly affect sex hormone levels (Janjua et al., 2004), is not surprising, considering the efficiency of homeostatic mechanisms in the adult. We think that the main potential target is the developing organism.

Endocrine-active ultraviolet filters and cosmetics

535

For risk assessment, it is important to realize that individual endocrineactive UV filters act in concert with other chemicals having an analogous mechanism of action. They contribute to the effect of the entire mixture of exogenous endocrine-active chemicals and endogenous hormones present in living organisms. As a consequence of the non-linearity of dose–response relationships and concentration addition, such mixture effects may be considerably greater than the sum of the individual effects (Silva et al., 2002). Unfortunately, there is no published data on developmental toxicity of other UV filters exhibiting endocrine activity. With respect to risk management, two observations from our monitoring study are of particular interest: 1. UV filters in cosmetics other than sunscreens were the source of internal exposure in about half of the participating women (Table 19.3). These women were mostly not aware of the fact that they were using products containing UV filters. Better information seems necessary. 2. Internal exposure patterns of cosmetics exhibited marked variability between individuals, and were found to be significantly correlated with usage patterns in the case of the more frequently used UV filters. The situation thus differs from that encountered with POPs such as PCBs, since the consumer could reduce internal exposure levels of UV filters and other cosmetic constituents during critical life stages such as pregnancy and lactation by limiting the use of the relevant products.

19.5

Acknowledgments

The investigations of the authors were supported by Swiss National Science Foundation (National Research Program 50), EU 5th Framework Programme (EURISKED/CREDO Cluster), Swiss Federal Office for the Environment (BAFU), Hartmann-Müller Stiftung, Olga Mayenfisch Stiftung. We thank Thomas Kupper (Zollikofen, Switzerland) for Fig. 19.2.

19.6

References

argom. (2003). Australian regulatory guidelines for OTC medicines (ARGOM), Therapeutic Goods Administration (TGA), 1 July 2003, Chapter 10, Sunscreens. atanassova, n., mckinnell, c., turner, k.j., walker, m., fisher, j.s., morley, m., millar, m.r., groome, n.p., sharpe, r.m. (2000). Comparative effects of neonatal exposure of male rats to potent and weak (environmental) estrogens on spermatogenesis and puberty and the relationship to adult testis size and fertility: evidence for stimulatory effects of low estrogen levels. Endocrinology 141: 3898–3907. bakker, j., honda, s., harada, n., balthazart, j. (2002). The aromatase knockout mouse provides new evidence that estradiol is required during development in the female for the expression of socio-sexual behaviors in adulthood. J. Neurosci. 22: 9104–9112.

536

Endocrine-disrupting chemicals in food

balmer, m.e., buser, h.r., müller, m.d., poiger, t. (2005). Occurrence of some organic UV filters in wastewater, in surface waters and in fish from Swiss lakes. Environ. Sci. Technol. 39 (4): 953–962. berset, j.d., etter-holzer, r. (1999). Quantitative determination of polycyclic aromatic hydrocarbons, polychlorinated biphenyls and organochlorine pesticides in sewage sludges using supercritical fluid extraction and mass spectrometric detection. J. Chromatogr. A 852: 545–558. berset, j.d., etter-holzer, r. (2001). Determination of phthalates in crude extracts of seven sludges by high resolution capillary gas chromatography with mass spectrometric detection. J. AOAC Intern. 84: 383–391. biegel, l.b., flaws, j.a., hirshfield, a.n., o’connor, j.c., elliott, g.s., ladics, g.s., silbergeld, e.k., van pelt, c.s., hurtt, m.e., cook, j.c., frame, s.r. (1998). 90-Day feeding and one-generation reproduction study in Crl:CD BR rats with 17βestradiol. Toxicol. Sci. 44: 116–142. bonefeld-jorgensen, e.c., andersen, h.r., rasmussen, t.h., vinggaard, a.m. (2001). Effect of highly bioaccumulated polychlorinated biphenyl congeners on estrogen and androgen receptor activity. Toxicology 158: 141–153. buser, h.r., balmer, m., schmid, p., kohler, m. (2006). Occurrence of UV filters 4methylbenzylidene camphor and octocrylene in fish from various Swiss rivers with inputs from wastewater treatment plants. Environ. Sci. Technol. 40: 1427–1431. chung, y.-w., nunez, a.a., clemens, l.g. (2001). Effects of neonatal polychlorinated biphenyl exposure on female sexual behavior. Physiol. Behav. 74: 363–370. damgaard, i.n., skakkebaek, n.e., toppari, j., virtanen, h.e., shen, h., schramm, k.-w., petersen, j.h., jensen, t.k., main, k.m., and nordic cryptorchidism study group (2006). Persistent pesticides in human breast milk and cryptorchidism. Environ. Health Perspect. 114: 1133–1138. durrer, s., maerkel, k., schlumpf, m., lichtensteiger, w. (2005). Estrogen target gene regulation and coactivator expression in rat uterus after developmental exposure to the UV filter 4-methylbenzylidene camphor. Endocrinology 146: 2130–2139. durrer, s., ehnes, c., fuetsch, m., maerkel, k., schlumpf, m., lichtensteiger, w. (2007). Estrogen sensitivity of target genes and expression of nuclear receptor coregulators in rat prostate after pre- and postnatal exposure to the ultraviolet filter 4-methylbenzylidene camphor. Environ. Health Perspect. 115, Suppl. 1: 42–50. faass, o., schlumpf, m., reolon, s., henseler, m., maerkel, k., durrer, s., lichtensteiger, w. (2009) Female sexual behaviour, estrous cycle and gene expression in sexually dimorphic brain regions after pre- and post natal exposure to endocrine active UV filters. Neuro Toxicology, in press (online). fiv (2005). Verordnung des EDI über Fremd- und Inhaltsstoffe in Lebensmitteln (Fremd- und Inhaltsstoffverordnung, FIV), Eidgenössisches Departement des Innern, Bern, 23 November 2005. george, f.w. (1993). Postnatal expression of high rates of 5α-reductase in the female rat urogenital tract. J. Devl. Physiol. 19: 187–191. habert, r., picon, r. (1984). Testosterone, dihydrotestosterone and estradiol-17 beta levels in maternal and fetal plasma and in fetal testes in the rat. J. Steroid. Biochem. 21: 193–198. hayden, c.g.j., roberts, m.s., benson, h.a.e. (1997). Systemic absorption of sunscreen after topical application. Lancet 350: 863–864. hofkamp, l., bradley, s., tresguerres, j.a., lichtensteiger, w., schlumpf, m., timms, b. (2008). Region-specific growth effects in the developing rat prostate following fetal exposure to estrogenic UV filters. Environ. Health Perspect. 116: 867– 872.

Endocrine-active ultraviolet filters and cosmetics

537

holbech, h., norum, u., korsgaard, b., bjerregard, p. (2002). The chemical UV-filter 3-benzylidene camphor causes an oestrogenic effect in in vivo fish assay. Pharmacol. Toxicol. 91: 204–208. hoover, s.m. (1999). Exposure to persistent organochlorines in canadian breast milk: a probabilistic assessment. Risk Analysis 19(4): 527–545. inoue, k., harada, k., takenaka, k., uehara, s., kono, m., shimizu, t., takasuga, t., senthilkumar, k., yamashita, f., koizumi, a. (2006). Levels and concentration ratios of polychlorinated biphenyls and polybrominated diphenyl ethers in serum and breast milk in Japanese mothers. Environ. Health Perspect. 114 (8): 1179–1185. inui, m., adachi, t., takenaka, s., inui, h., nakazawa, m., ueda, m., watanabe, h., mori, c., iguchi, t., miyatake, k. (2003). Effect of UV screens and preservatives on vitellogenin and choriogenin production in male medaka (Oryzias latipes). Toxicology 194: 43–50. janjua, n.r., mogensen, b., andersson, a.-m., petersen, j.h., henriksen, m., skakkebaek, n.e., wulf, h.c. (2004). Systemic absorption of the sunscreens benzophenone-3, octyl-methoxycinnamate, and 3-(4-methyl-benzylidene) camphor after whole-body topical application and reproductive hormone levels in humans. J. Invest. Dermatol. 123: 57–61. jarry, h., christoffel, j., rimoldi, g., koch, l., wuttke, w. (2004). Multi-organic endocrine disrupting activity of the UV screen benzophenone 2 (BP2) in ovariectomized adult rats after 5 days treatment. Toxicology 205: 87–93. kelly, b.c., ikonomou, m.h., blair, j.d., morin, a.e., gobas, f.a.p.c. (2007). Food webspecific biomagnifications of persistent organic pollutants. Science 317: 236–239. klammer, h., schlecht, c., wuttke, w., jarry, h. (2005). Multi-organic risk assessment of estrogenic properties of octyl-methoxycinnamate in vivo. A 5-day subacute pharmacodynamic study with ovariectomized rats. Toxicology 215: 90–96. kunz, p.y., fent, k. (2006). Multiple hormonal activities of UV filters and comparison of in vivo and in vitro estrogen activity of ethyl-4-aminobenzoate in fish. Aquatic Toxicology 79: 305–324. kupper, t., berset, j.d., etter-holzer, r., furrer, r., tarradellas, j. (2004). Concentrations and specific loads of polycyclic musks in sewage sludge originating from a monitoring network in Switzerland. Chemosphere 54: 1111–1120. kupper, t., plagellat, c., brändli, r.c., de alencastro, l.f., grandjean, d., tarradellas, j. (2006). Fate and removal of polycyclic musks, UV filters and biocides during wastewater treatment. Water Res. 40: 2603–2612. lauber, m.e., lichtensteiger, w. (1994). Pre- and postnatal ontogeny of aromatase cytochrome P450 messenger ribonucleic acid expression in the male rat brain studied by in situ hybridization. Endocrinology 135: 1661–1668. lichtensteiger, w., schlumpf, m. (1993). Prenatal nicotine exposure: biochemical and neuroendocrine bases of behavioral dysfunction. Dev. Brain Dysfunct. 6: 279–304. lichtensteiger, w., faass, o., ceccatelli, r., schlumpf, m. (2004). Developmental exposure to PBDE 99 and PCB affects estrogen sensitivity of target genes in rat brain regions and female sexual behavior. Organohalogen Compounds 66: 3965–3970. link, b., gabrio, t., zoellner, i., piechotowski, i., paepke, o., herrmann, t., felderkennel, a., maisner, v., schick, k.h., schrimpf, m., schwenk, m., wuthe, j. (2005). Biomonitoring of persistent organochlorine pesticides, PDD/PCDF and dioxinlike PCBs in blood of children from South West Germany (Baden-Wuerttemberg) from 1993–2003. Chemosphere 58: 1185–2000. ma, r., cotton, b., lichtensteiger, w., schlumpf, m. (2003). UV filters with antagonistic action at androgen receptors in the MDA-kb2 cell transcriptional-activation assay. Toxicol. Sci. 74, 43–50.

538

Endocrine-disrupting chemicals in food

maclusky, n.j., naftolin, f. (1981). Sexual differentiation of the central nervous system. Science 211: 1294–1303. maerkel, k., lichtensteiger, w., durrer, s., conscience, m., schlumpf, m. (2005). Sex- and region-specific alterations of progesterone receptor mRNA levels and estrogen sensitivity in rat brain following developmental exposure to the estrogenic UV filter 4-methylbenzylidene camphor. Environ. Toxicol. Pharmacol. 19: 761–765. maerkel, k., durrer, s., henseler, m., schlumpf, m., lichtensteiger, w. (2007). Sexually dimorphic gene regulation in brain as a target for endocrine disrupters: Developmental exposure of rats to 4-methylbenzylidene camphor. Toxicol. Appl. Pharmacol. 218: 152–165. main, k.m., mortensen, g.k., kaleve, m.m., boisen, k.a., damgaard, i.m., chellakooty, m., schmidt, i.m., suomi, a.m., virtanen, h.e., petersen, j.h., andersson, a.m., toppari, j., skakkebaek, n.e. (2006). Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age. Environ. Health Perspect. 114: 270–276. main, k.m., kiviranta, h., virtanen, h.e., sundqvist, e., tuomisto, j.t., tuomisto, j., vartiainen, t., skakkebaek, n.e., toppari, j. (2007). Flame retardants in placenta and breast milk and cryptorchidism in newborn boys. Environ. Health Perspect. 115: 1519–1526. mueller, s.o., kling, m., firzani, p.a., mecky, a., duranti, e., shields-botella, j., delansorne, r., broschard, t., kramer, p.j. (2003). Activation of estrogen receptor α and ERβ by 4-methylbenzylidene-camphor in human and rat cells: comparison with phyto- and xeno-estrogens. Toxicol. Lett. 142: 89–101. nagtegaal, m., ternes, t.a., baumann, w., nagel, r. (1997). UV-Filtersubstanzen in Wasser und Fischen. UWSF-Z. Umweltchem. Ökotox. 9: 79–86. nakata, h., sasaki, h., takemura, a., yoshioka, m., tanabe, s., kannan, k. (2007). Bioaccumulation, temporal trend and geographical distribution of synthetic musks in the marine environment. Environ. Sci. Technol. 41: 2216–2222. nohynek, g.j., schaefer, h. (2001). Benefit and risk of organic ultraviolet filters. Regul. Toxicol. Pharmacol. 33: 285–299. ogawa, s., olazabal, u.e., parhar, i.s., pfaff, d.w. (1994). Effects of intrahypothalamic administration of antisense DNA for progesterone receptor mRNA on reproductive behavior and progesterone receptor immunoreactivity in female rat. J. Neurosci. 14: 1766–1774. pezzi, v., mathis, j.m., rainey, w.e., carr, b.r. (2003). Profiling transcript levels for steroidogenic enzymes in fetal tissues. J. Steroid Biochem. Mol. Biol. 87: 181–189. plagellat, c., kupper, t., furrer, r., de alencastro, l.f., grandjean, d., tarradellas, j. (2006). Concentrations and specific loads of UV filters in sewage sludge originating from a monitoring network in Switzerland. Chemosphere 62: 915–925. pohl, h.r., hibbs, b.f. (1996). Breastfeeding exposure of infants to environmental contaminants – a public health risk assessment viewpoint: chlorinated dibenzodioxins and dibenzofurans. Toxicol. Ind. Health 12 (5): 593–609. poiger, t., buser, h.-r., balmer, m.e., bergquist, p.-a., müller, m.-d. (2004). Occurrence of UV filter compounds from sunscreens in surface waters: regional mass balance in two Swiss lakes. Chemosphere 55: 951–963. pollio, g., xue, p., zanisi, m., nicolin, a., maggi, a. (1993). Antisense oligonucleotide blocks progesterone-induced lordosis behavior in ovariectomized rats. Mol. Brain Res. 19: 135–139. putz, o., schwartz, c.b., leblanc, g.a., cooper, r.l., prins, g.s. (2001). Neonatal lowand high-dose exposure to estradiol benzoate in the male rat: II. Effects on male puberty and the reproductive tract. Biol. Reprod. 65: 1506–1517.

Endocrine-active ultraviolet filters and cosmetics

539

ramirez, a.j., mottaleb, m.a., brooks, b.w., chambliss, c.k. (2007). Analysis of pharmaceuticals in fish using liquid chromatography-tandem mass spectrometry. Anal. Chem. 79: 3155–3163. salvador, a., chisvert, a. (2005). Sunscreen analysis: a critical survey on UV filters determination. Analy. Chim. Acta 537: 1–14. salvador, a., chisvert, a., eds. (2007). Analysis of Cosmetic Products. Elsevier, Amsterdam. schauer, u.m.d., völkel, w., heusener, a., colnot, t., broschard, t.h., van landenberg, f., dekant, w. (2006). Kinetics of 3-(4-methylbenzylidene)camphor in rats and humans after dermal application. Toxicol. Appl. Pharmacol. 216: 339–346. schlumpf, m., cotton, b., conscience, m., haller, v., steinmann, b., lichtensteiger, w. (2001a). In vitro and in vivo estrogenicity of UV screens. Environ. Health Perspect. 109: 239–244. schlumpf, m., berger, l., cotton, b., conscience-egli, m., durrer, s., fleischmann, i., haller, v., maerkel, k., lichtensteiger, w. (2001b). Estrogen active UV screens. SFÖW-Journal 127: 10–15. schlumpf, m., jarry, h., wuttke, w., ma, r., and lichtensteiger, w. (2004a). Estrogenic activity and estrogen receptor β binding of the UV filter 3-benzylidene camphor. Comparison with 4-methylbenzylidene camphor. Toxicology 199: 109–120. schlumpf, m., schmid, p., durrer, s., conscience, m., maerkel, k., henseler, m., gruetter, m., herzog, i., reolon, s., ceccatelli, r., faass, o., stutz, e., lichtensteiger, w. (2004b). Endocrine activity and developmental toxicity of cosmetic UV filters – an update. Toxicology 205: 113–122. schlumpf, m., durrer, s., faass, o., ehnes, c., fuetsch, m., gaille, c., henseler, m., hofkamp, l., maerkel, k., reolon, s., timms, b., tresguerres, j.a.f., lichtensteiger, w. (2008a). UV filters: developmental toxicity of UV filters and environmental exposure. A review. Int. J. Androl. 31: 144–151. schlumpf, m., kypke, k., vökt, c.c., birchler, m., durrer, s., faass, o, ehnes, c., fuetsch, m., gaille, c., henseler, m., hofkamp, l., maerkel, k., reolon, s., zenker, a., timms, b., tresguerres, j.a.f., lichtensteiger, w. (2008b). Endocrine active UV filters: Developmental toxicity and exposure through breast milk. Chimia. 62: 345–351. schmid, p., kohler, m., gujer, e., zennegg, m., lanfranchi, m. (2007). Persistent organic pollutants, brominated flame retardants and synthetic musks in fish from remote alpine lakes. Chemosphere 67: S16–S21. schmutzler, c., hamann, i., hoffmann, p.j., kovacs, g., stemmler, l., mentrup, b., schomburg, l., ambrugger, p., grüters, a., seidlowa-wuttke, d., jarry, h., wuttke, w., köhrle, j. (2004). Endocrine active compounds affect thyrotropin and thyroid hormone levels in serum as well as endpoints of thyroid hormone action in liver, heart and kidney. Toxicology 205: 95–112. schreurs, r., lauser, p., seinen, w., van den burg, b. (2002). Estrogenic activity of UV filters determined by an in vitro reporter gene assay and an in vivo transgenic zebrafish assay. Arch. Toxicol. 76: 257–261. schreurs, r.h., sonneveld, e., jansen, j.h., seinen, w., van der burg, b. (2005). Interaction of polycyclic musks and UV filters with the estrogen receptor (ER), androgen receptor (AR), and progesterone receptor (PR) in reporter gene bioassays. Toxicol. Sci. 83: 264–272. seidlova-wuttke, d., jarry, h., wuttke, w. (2004). Pure estrogenic effect of benzophenone-2 (BP2) but not of bisphenol A (BPA) and dibutylphthalate (DBP) in uterus, vagina and bone. Toxicology 205: 103–112. seidlova-wuttke, d., jarry, h., christoffel, j., rimoldi, g., wuttke, w. (2005). Effects of bisphenol A, dibutylphthalate (DBP), benzophenone-2 (BP2), procymidone (Proc), and linurone (Lin) on fat tissue, a variety of hormones and metabolic

540

Endocrine-disrupting chemicals in food

parameters: a 3 months comparison with effects of estradiol (E2) in ovariectomized (ovx) rats. Toxicology 213: 13–24. seidlova-wuttke, d., christoffel, j., rimoldi, g., jarry, h., wuttke, w. (2006a). Comparison of effects of estradiol (E2) with those of octylmethoxycinnamate (OMC) and 4-methylbenzylidene camphor (4MBC) – 2 filters of UV light – on several uterine, vaginal and bone parameters. Toxicol. Appl. Pharmacol. 210: 246–254. seidlova-wuttke, d., christoffel, j., rimoldi, g., jarry, h., wuttke, w. (2006b). Comparison of effects of estradiol with those of octylmethoxycinnamate and 4methylbenzylidene camphor on fat tissue, lipids and pituitary hormones. Toxicol. Appl. Pharmacol. 214: 1–7. siler-khodr, t.m. (1998). Endocrine and paracrine function of the human placenta. In: Fetal and Neonatal Physiology, Vol. 1, R.A. Polin and W.W. Fox, Eds., W.B. Saunders Co., Philadelphia, pp. 89–102. silva, e., rajapakse, n., kortenkamp, a. (2002). Something from ‘nothing’ – eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ. Sci. Technol. 36: 1751–1756. so, m.k., yamashita, n., taniyasu, s., jiang, q., giesy, j.p., chen, k., lam, p.k. (2006). Health risks in infants associated with exposure to perfluorinated compounds in human breast milk from Zhoushan China. Environ. Sci. Technol. 40 (9): 2924–2929. sun, s.-j., zhao, j.-h., liu, h.-j., lu, d.-w., ma, y.-x., li, l., horiguchi, h., uno, h., iida, t., koga, m., kiyonari, y., nakamura, m., sasaki, w.m.h., leung, a.o.w., chan, j.k.y., choi, m.p.k. (2005). A review on the usage of POP pesticides in China, with emphasis on DDT loadings in human milk. Chemosphere 60: 740–752. swan, s.h., main, k.m., liu, f., stewart, s.l., kruse, r.l., calafat, a.m., mao, c.s., redmon, j.b., terrand, c.l., sullivan, s.,teague, j.l., and the study for future families research team (2005). Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. 113: 1056–1061. tinwell, h., lefevre, p.a., moffat, g.j., burns, a., odum, j., spurway, t.d., orphanides, g., ashby, j. (2002). Confirmation of uterotrophic activity of 3 (4-methylbenzylidene) camphor in the immature rat. Environ Health Perspect. 110: 533–536. vandenberg, l.n., maffini, m.v., wadia, p.r., sonnenschein, c., rubin, b.s., soto, a.m. (2007). Exposure to environmentally relevant doses of the xenoestrogen bisphenol-A alters development of the fetal mouse mammary gland. Endocrinology 148: 116–127. völkel, w., colnot, t., schauer, u.m.t., broschard, t.h., dekant, w. (2006). Toxicokinetics and biotransformation of 3-(4-methylbenzylidene)camphor in rats after oral administration. Toxicol. Appl. Pharmacol. 216: 331–338. vom saal, f.s., timms, b.g., montano, m.m., palanza, p., thayer, k.a., nagel, s.c., dhar, m.d., ganjam, v.k., parmigiani, s., whelshons, w.v. (1997). Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses. Proc. Natl. Acad. Sci. 94: 2056–2061. wallen, k., zehr, j.l. (2004) Hormones and history: the evolution and development of primate female sexuality. J. Sex Res. 41: 101–112. wong, m.h., leung, a.o.w., chan, j.k.y., choi, m.p.k. (2005). A review on the usage of POP pesticides in China, with emphasis on DDT loadings in human milk. Chemosphere 60: 740–775.

20 Mechanisms of action of particular endocrine-disrupting chemicals F. Pakdel, O. Kah and B. Jégou, Université de Rennes 1, France

Abstract: In vertebrates, nuclear receptors are important for the growth, differentiation and development of many organs. A large number of chemicals including pesticides, plastic components, heavy metals, hydrocarbons or pharmaceutical, present in food and in the environment are able to interfere with these receptors producing adverse effects during vertebrate development or reproductive cycle. In this review, we summarise the molecular mechanisms of action of nuclear receptors. We discuss how these receptors may be activated in vivo by different exogenous chemicals and discuss reporter assays and cell experimental models that provided useful mechanistic information about several classes of environmental contaminants. Key words: nuclear receptors, environmental contaminants, estrogens, androgens, hydrocarbons.

20.1

Introduction

In all vertebrate species, endogenous steroid hormones, particularly estrogens and androgens, play a crucial role in the development, growth, maintenance and function of both the female and male reproductive tracts as well as reproductive behaviour. These sex steroids also exert positive and negative feedback effects on the hypothalamo-pituitary axis to regulate the secretion of gonadotropic and other pituitary hormones. Steroid hormones, notably 17β-estradiol (E2) and testosterone (T), are also involved in reproductive disorders such as breast, endometrial and testicular cancers. Many natural and synthetic chemicals present in the environment and in food are suspected to interfere with steroid hormones and could exert adverse effects on the organism, leading to endocrine disruption. These compounds are generally termed endocrine-disrupting chemicals (EDCs).

542

Endocrine-disrupting chemicals in food

Many studies have reported a variety of reproductive alterations and appearance of adverse health effects in aquatic and wildlife species living within or near contaminated areas (Colborn et al., 1993; Fent, 2004; Guillette et al., 1994; Sumpter and Jobling, 1993). Male fish exposed to wastewater treatment work effluents, in UK rivers, developed intersex (female ovarian tissue within the testes) and also produced vitellogenin, a protein required for egg yolk production in females under the control of estrogens (Sumpter, 1995; Sumpter and Jobling, 1995). Moreover, several in vitro and cell-based assays demonstrated that many substances generated from widespread environmental chemicals such as pesticides, herbicides, plastic components, heavy metals or pharmaceuticals, have hormonal activity and particularly showed estrogenic potency (Balaguer et al., 1996; Flouriot et al., 1995; Lassurguere et al., 2003; Le Guevel and Pakdel, 2001; Petit et al., 1997; Soto et al., 1998). In parallel, a series of epidemiological studies have suggested a deterioration of the male reproductive functions over the past decades, in particular a reduction in the sperm number and spermatic quality, an increase in malformations of reproductive tissues as well as an increased frequency of testicular cancers (Jegou et al., 1999; Toppari et al., 2006). These deteriorations in reproductive parameters have also been detected in different animal species such as birds, alligators, frogs and fish, suggesting an origin possibly dependent on environmental factors. Although the precise reasons for the increased occurrence of reproductive disorders in humans remain unclear, it is well known that estrogens and androgens play crucial roles in testicular functions (Carreau et al., 2002; Toppari et al., 2006). Notably, the regulation of testicular descent is tightly controlled by androgens and, on the other hand, estrogens participate in the differentiation process of testicular cells (Hess et al., 1997; Staub et al., 2005). Moreover, environmental estrogens such as bisphenol A or ethinyl-estradiol have been reported to cause cryptorchidism or to significantly increase prostate size in experimental animals. Together, these observations led to the conclusion that environmental contaminants could interfere with normal hormonal processes, and act as estrogenic or anti-androgenic chemicals (Sharpe and Skakkebaek, 1993; Sohoni and Sumpter, 1998; Sonnenschein and Soto, 1998). Among EDCs, many are formed as a consequence of human activities. These include polycyclic aromatic hydrocarbons (PAH) and halogenated aromatic hydrocarbons (HAH) such as polychlorinated dibenzo-p-dioxins and dibenzofurans, or polychlorinated biphenyls (PCBs) that are highly persistent and widespread environmental contaminants. These chemicals were reported to have adverse effects on different species, including humans (Barouki et al., 2007; Safe, 2001a). Dioxin-like compounds could be involved in the development of endometriosis, because of the presence of higher amounts of these pollutants in women suffering from this disease, and because of the appearance of endometriosis in monkeys exposed to a dioxin

Mechanisms of action of particular endocrine-disrupting chemicals

543

treatment for several years. The actions of dioxin and dioxin-like chemicals are mediated by the aryl hydrocarbon receptor (AhR), a basic helix–loop– helix transcription factor that functions as nuclear receptors (NRs), although there is no homology of structure between the AhR and NR families. In addition to their classical direct actions through the specific AhR, both estrogenic and anti-estrogenic effects of dioxin-like compounds have been reported (Abdelrahim et al., 2006; Chen et al., 2001; Khan et al., 2006; Kuiper et al., 1998; Navas et al., 2004; Ohtake et al., 2003). In summary, the effects of environmental contaminations as EDCs have raised many questions concerning their impact on human and animal health and caused great public concern. Although there are several studies showing good correlations between reproductive disorders in wildlife animals and exposure to environmental contaminations, there is uncertainty about the impact of EDCs on human health, except the clear relationship between diethylstilbestrol and anomalies of the reproductive tracts (Yamashita, 2006). Because of the paucity of available data, the possible involvement of EDCs in the increase of endometrial or breast cancer incidence or even in male reproductive health remains hypothetical. The suggested mechanism of action for these chemicals is their capacity to interact with nuclear receptors, in particular estrogen receptors (ERs) and androgen receptors (ARs), and also with the specific receptor for dioxin (AhR). In this review, we summarise the molecular mechanisms of action of nuclear receptors, particularly ERs. We discuss how these receptors may be activated in vivo by different exogenous chemicals and discuss reporter-assays and cell experimental models that provided useful mechanistic information about several classes of environmental contaminants.

20.2

Nuclear receptor family: estrogen receptors

The effects of estrogens are mainly mediated by two different estrogen receptors, ERα and ERβ, which are members of the NR family. This superfamily of receptors that function as ligand-activated transcription factors regulate various physiological functions, from development to homeostasis, in metazoans to human. The NRs are characterised by distinct structural and functional domains including a conserved zinc finger DNA-binding domain (DBD) and a ligand-binding domain (LBD). The NR family includes 48 members including the glucocorticoid receptor (GR), progesterone receptor (PR), androgen receptor (AR), vitamin D receptor (VDR), thyroid hormone receptor (TR), retinoic acid (RAR), and also an important number of orphan receptors for which ligands have not yet been identified (Fig. 20.1) (Germain et al., 2006; Robinson-Rechavi et al., 2003). It is believed that the NR family appeared very early during evolution and that the ancestral NR is likely to be an orphan receptor that acquired ligand-binding

544

Endocrine-disrupting chemicals in food A/B

(a)

C

E/F

D

185 250 317

1

595 ER

1

421 486 526

777 GR

1

933

567 633 680

PR 559 624 676

1

919 AR

1 24

89

427

192

VDR 1

102 169 241

456 TR

1

88

462

153 198

RAR

(b)

185 250 317

1

595 ERa

30% 1

96% 30%

53%

149 214 248

530 ERb

DNA binding Ligand binding Dimerisation

Transactivation

AF-1

AF-2

Fig. 20.1 The evolutionary conserved domains of several membres of nuclear receptors. (a) The structural domains of these receptors (A/B, C, D, and E/F), as well as the N-terminal and C-terminal transactivation functional domains (AF-1, AF-2) are shown. (b) Schematic structure–activity of human ER alpha and human ER beta with the percentage amino acid identity of each domain between the two receptors. Domains invoved in the DNA/ligand binding, dimerisation and in the ligandindependent transactivation function AF-1 and in the ligand-dependent transactivation function AF-2 are also shown.

Mechanisms of action of particular endocrine-disrupting chemicals

545

ability during metazoan evolution. NRs classically regulate the expression of target genes by direct interaction with specific DNA sequences in recruiting cofactors necessary for the transcription (Klinge, 2000) (see also below). Analysis of the three-dimensional (3D) structure of several NRs has demonstrated that after specific ligand binding, the LBD induces a conformational change, allowing the recruitment of coactivator proteins. These factors that are recruited in an ordered series of sequential receptor–coactivator interactions possess diverse functional domains with notably acetyltransferase, methyltransferase or ubiquitin ligase activity. These enzymatic activities are important for nucleosome disruption and chromatin remodelling and thus play critical roles in modifying transcription of target genes. Several coactivators have been described that interact with NRs, the most common being the p160 family members, SRC-1 (also called NCOA-1), SRC-2 (also called GRIP1/TIF-2) and SRC-3 (also called AIB1/ACTR/pCIP) which possess histone acetytransferase activity (HAT). The two ERs are generated by two distinct genes, have partially distinct expression patterns and their activities are modulated differently by some ligands called selective ER modulators (SERMs) (Gustafsson, 1998; Katzenellenbogen et al., 2000; Kuiper et al., 1997). Like other members of the nuclear receptor superfamily, ERα and ERβ exhibit a modular structure. The N-terminal A/B domains are poorly conserved between members of the nuclear receptor family and also between the two ERs. This region contains a ligand-independent activation function, AF1 (Fig. 20.1). Although little is known about this function, it has been reported that a highly conserved helical structure and phosphorylation sites in the B domain are crucial elements for the AF1 activity. Consequently growth factors, such as IGF-1, EGF or TNF-α, can be synergistic with E2 through activation of intracellular protein kinases pathway which in turn induce the phosphorylation of the serine residues of the N-terminal AF1 domain. The C domain including DBD, is highly conserved between both ERs (more than 95% of identity), and also between the members of the NR superfamily (40–50% of identity), and contains two zinc fingers, permitting the recognition of a specific DNA hormone-responsive element. In the case of ERs, this element, known as estrogen-responsive element (ERE), is a palindromic sequence composed of two core 5′-AGGTCA-3′ motifs with a 3 base pair (bp) spacer (AGGTCAnnnTGACCT). Located in the C-terminal end of the C domain, the poorly conserved D domain is involved in the three-dimensional structure of the ERs and permits the stabilisation of the DNA binding. Moreover, several lysine residues in this domain could be acetylation sites, suggesting that this region is involved in regulation of the ER nuclear translocation and interaction with other factors. The multifunctional E/F-domains, also including the LBD, presents a globular organisation which consists of a conserved arrangement of 12 helices (H1–H12) forming the hydrophobic ligand pocket closed by an

546

Endocrine-disrupting chemicals in food

antiparallel β-sheet and by helix H12. This region shows 50–55% identity between ERα and ERβ, but is not conserved between the members of nuclear receptors. The LBD of ERs recognises a variety of compounds, exhibiting a great diversity in their size and in their chemical properties, and susceptibility to induce different conformations of the ER protein. Consequently, agonist binding induces an intramolecular shift, causing a repositioning of H12 known to be involved in the transactivation function AF2 and able to interact with cofactors. In contrast, the binding of antagonist is accompanied by major structural reorganisation disrupting the topography of the AF2 and preventing the orientation of helix H12 to an agonist position. Although the unliganded ERs are found mainly in the nucleus, they are in an inactive state and not tightly bound to nuclear components. Ligand binding triggers conformational changes in the receptor inducing receptor dimerisation, receptor-DNA interaction, and recruitment of coactivators and the general transcription factors (Fig. 20.2). In the nucleus, ERs modulate gene transcription, and the resulting protein products determine the biological actions of estrogens. Several studies have indicated that regulatory sequences other than ERE could also be targeted by ER, suggesting that other mechanisms of transcriptional activation may also exist. Indeed, analysis of the E2-dependent regulation of several genes such as c-fos, cathepsin D, c-myc and HSP27 gene promoters showed that even if these

Cytosol

Nucleus

Estrogens

Pol II ERE

Target gene Pol II

Growth factors

MAPK, PI3K, …

Sp1

Target gene

Cellular effects: proliferation differentiation survival

Pol II AP1 ‘Non-genomic action’

Target gene

‘Genomic action’

Fig. 20.2 Transcriptional regulation of target genes by estrogen receptor. The hormone enters the cell, binds to the inactive receptor, either in the cytoplasm, or directly in the nucleus. In the nucleus, the activated complex hormone-receptor forms a dimer to fix tightly DNA directly at the ERE sites or indirectly at Sp1 or AP1 sites. The activated ER is then able to recruit cofactors and the RNA polymerase II (pol. II), which allow the transcription of target genes.

Mechanisms of action of particular endocrine-disrupting chemicals

547

genes are devoid of consensus ERE, ERα was able to modulate their transcription by establishing protein/protein interactions with other transcription factors such as AP-1 or Sp1 complexes (Nilsson et al., 2001; Safe, 2001a,b; Safe and Kim, 2004) (Fig. 20.2). As an alternative to the classic ‘dogma’ of molecular mechanisms of ER interactions with the transcriptional machinery, co-activators/co-repressors and other transcription factors, multiple levels of cross-talks between ERs and other intracellular pathways have now been reported. Indeed, although ERs are mainly in the nucleus, a small pool of ERs are present in the cytosol or close to the plasma membrane that could activate several signal transduction cascades such as mitogen-activated protein kinase (MAPK), protein kinase C and phosphatidylinositol 3-kinase (PI3K). This type of interactions may be responsible for the so-called ‘non-genomic’ effects of estrogens (Fig. 20.2). For example, rapid estrogen-induced activation of signal transduction cascades (i.e. PKC, MAP kinase, PI3 kinase) can occur even in cells expressing only a mutant ERα lacking the DBD. Activation of the MAP kinase pathway seems to be necessary to activate non-genomic actions of estrogens on cell survival (Levin, 2003; Levin and Pietras, 2008). Moreover, ligand-independent activation of the ER AF1 through phosphorylation by growth factors such as insulin-like growth factor-I (IGF-I), transforming growth factor-α (TGFα) and epidermal growth factor (EGF) have also been reported. Through tyrosine kinase receptor systems, growth factors activate a series of downstream cellular kinases (MAPKs, PI3K, PKC) and signalling pathaways involved in multiple cell processes such as protection against cell death, proliferation or migration (Levin, 2003). The cross-talk between growth factors and ER can also contribute to the termination of the ligand response. It is also important to mention that mechanisms and pathways for ER activation by growth factors are likely to be dependent on the gene and/or gene promoter as well as on the cell context. These different modes of regulation illustrate the complexity of estrogen signalling pathways their deregulation could lead to cancers or other diseases.

20.3

Estrogenic/anti-estrogenic potency of endocrine-disrupting chemicals

Differences between ER subtypes in relative ligand binding affinity for a range of natural and synthetic agonists or antagonists have been described. Such chemicals can induce distinct conformational changes in the tertiary structure of the ERs, therefore affecting the sequential recruitment of cofactors. These ligand-dependent interactions of ERs and coactivators/ corepressors are critical steps in ER-mediated transcriptional regulation and consequently, modulation of the expression of E2-target genes. In

548

Endocrine-disrupting chemicals in food

addition, although both ERs share the same modular organisation, they may have different target genes. In the zebrafish (Brachydanio rerio), we have reported the cloning and characterisation of three functional ER forms, zfERα, zfERβ1 and zfERβ2 (Menuet et al., 2002, 2004). Zebrafish represent an attractive in vivo model organism for developmental toxicology of EDCs, because of the external and rapid development of the transparent embryos, relatively short generation time, together with the fact that many techniques in molecular biology and genetics can be applied to zebrafish. The existence of three distinct ER genes coding for three distinct receptors in zebrafish provides an interesting in vivo model for understanding the mechanisms of action and for the evaluation of the impact of environmental estrogens during organogenesis. Recently, we showed that during development, ERs regulate the expression of zebrafish aromatase B (cyp19b; Aro-B) gene. Aro-B is a crucial enzyme that aromatises androgens into estrogens in the brain of fish. The local production of estrogens by Aro-B early during development is likely to be very important for the growth and sex differentiation of the brain (Menuet et al., 2005; Pellegrini et al., 2005). The expression of Aro-B gene was used as a sensitive marker of xenoestrogens effects on the central nervous system during embryogenesis and in zebrafish juveniles (Kazeto et al., 2004; Kishida et al., 2001; Le Page et al., 2006). To test the impact of environmental estrogenic chemicals, on distinct ER activity, we developed a glial cell-based assay using an Aro-B promoter-luciferase reporter. Dose–response analyses with several structurally different xenoestrogens showed that estrogenic potency of the xenoestrogens markedly differed depending on ER subtype in this assay (Le Page et al., 2006). Moreover, the combination of these agents showed an additive effect according to the concept of concentration addition (CA). This confirmed that the combined additive effect of the xenoestrogens leads to an enhancement of the estrogenic potency, even when each single agent is present at low effect concentrations (Le Page et al., 2006; Rajapakse et al., 2002; Silva et al., 2002). Studies in ERα and/or ERβ knock-out in mice showed that these two receptors mediate distinct biological effects of estrogens in many tissues such as mammary gland, bone, brain and the vascular system, in both males and females. Therefore, because ERα and ERβ show partially different tissue distribution and distinct physiological functions, SERM are developed to display either agonist or antagonist activity in a tissue-selective manner. These molecules are likely to provide novel and improved hormonal therapy of ER-dependent diseases at different tissue levels. Moreover, the use of SERMs after menopause may decrease the risk of breast and endometrial cancers. Such molecules acting in an estrogen-like way on the bone or the cardiovascular system and to the contrary as anti-estrogens on the mammary and endometrial tissues may be used to treat osteoporosis without promoting mammary and endometrial pathologies. In addition, recent studies demonstrated that many genes regulated by ERα are distinct

Mechanisms of action of particular endocrine-disrupting chemicals

549

from those regulated by ERβ (Chang et al., 2006a; Frasor et al., 2003; Kian Tee et al., 2004; Kim et al., 2005). These results indicate that SERMs could exert tissue-specific effects by regulating unique sets of targets genes through ERα and ERβ. The natural phytoestrogens with estrogenic or anti-estrogenic properties are SERMs. Some of them have significant pharmacological interest, in particular as hormonal substitutes at menopause or for hormonal treatments of breast cancer. It is well known that proliferation of ERα-positive breast cancer cells is enhanced by estrogens that modulate multiple growth factors and genes involved in cell cycle progression. Although ERα has a proliferative effect, recent data from functional gene analysis and cell proliferation studies suggested that ERβ could act as a negative regulator of ERα in breast cancer cells and may act to counteract the mitogenic effect of estrogens (Chang et al., 2006a; Frasor et al., 2003). Interestingly, in all ER-selective bioassays reported, such as proliferation of breast cancer cell lines, gene reporter assays in mammalian or nonmammalian cells and also in the recombinant yeast transcription assays, it was found that genistein (a phytoestrogen from soy) displayed high specificity toward ERβ transactivation. Genistein showed antiproliferative effects on a variety of cancer cells, including breast and prostate cancer. However, although the antiproliferative effect of genistein can be mediated, at least in part, by ERβ, other molecular mechanisms are envisioned. For instance, on certain prostate or breast cancer cells expressing only ERβ, genistein has been shown to exhibit antiproliferative effects and pro-apoptotic activities through caspase-3 activation. However, small interfering RNAmediated down-regulation of ERβ did not entirely reverse the antiproliferative effect of genistein in these cell lines. Indeed, genistein may inhibit cancer cell growth directly by inhibiting tyrosine kinase activity from growth factor receptors and cytoplasmic tyrosine kinases. In addition, a recent study showed that inhibition of prostate cancer cell proliferation by genistein is associated with a reduction in telomerase activity which is crucial for cells to gain immortality and proliferation ability (Jagadeesh et al., 2006). On the other hand, the antimitotic activity of genistein was reported to be biphasic in ER-positive MCF-7 cells, stimulation at low concentrations and inhibition at high concentrations. These observations demonstrate the complexity of genistein and in general phyto-estrogen actions in the context of their anti-cancer properties. For the last 50 years, a variety of synthetic compounds including the synthetic agonist diethylstilbestrol (DES) and SERMs, such as tamoxifen, have been synthesised in order to develop hormonal replacement therapy and prevention or treatment of ER-dependent diseases. Tamoxifen (OHT), which is a non-steroid and clinically important antiestrogen, is widely used to inhibit the growth of mammary carcinoma. However, OHT acts as partial estrogen agonist in the endometrial tissues and has been reported to be associated with a small, but significant, increased incidence (4%) of

550

Endocrine-disrupting chemicals in food

endometrial cancer in OHT-treated patients (Jordan and Brodie, 2007). A number of studies were reported to explain the molecular mechanisms of agonist/antagonist of OHT and so the increased incidence of endometrial carcinomas. Notably, OHT shows strong stimulatory activity at AP-1 sites with ERβ, but not with ERα (Paech et al., 1997). Attempts to improve the pharmacological profile of OHT have resulted in a series of 7α-substituted estradiols such as ICI-164,384 that have been identified as pure estrogen antagonists, but with little cytotoxicity. Raloxifene is another SERM of clinical use: it was developed to avoid some of the undesirable estrogen agonist actions of other SERMs such as OHT. Raloxifene is clinically successful mostly for prevention and treatment of postmenopausal osteoporosis. In the breast, raloxifene acts as a classical antiestrogen to inhibit the growth of mammary carcinoma and to reduce the risk of ER-positive invasive breast cancers. However, raloxifene treatment is associated with lower incidence of uterine cancers than those observed with OHT (Jordan and Brodie, 2007). Interestingly, in several gene-reporter assays and cell-based systems, raloxifene showed ERα-selective anti-estrogen potency while RU486 showed ERβ-selective anti-estrogen potency (Escande et al., 2006; Kuiper et al., 1997). On the other hand, 16α-LE2 (3,17-dihydroxy-19-nor-17αpregna-1,3,5(10)-triene-21,16α-lactone) and PPT (4-propyl-pyrazole-1,3,5triyl-triphenol) were reported to have high ERα-selective agonist potency whereas DNP (2,3-bis(4-hydroxyphenyl)-propionitrile) and 8β-VE2 (8vinylestra-1,3,5(10)-triene-3,17β-diol) showed ERβ-selectivity with 8β-VE2 being the most potent and selective ERβ agonist (Escande et al., 2006; Kuiper et al., 1997; Stauffer et al., 2000). Crystallography and mutagenesis studies of the LBD, in particular the binding cavity of ERα and ERβ showed two amino acid changes between ERα and ERβ (ERαLeu384→ ERβMet336 and ERαMet421→ERβIle373) are extremely important for ER-ligand binding selectivity and transactivation. These changes could subtly affect the size and shape of the binding cavity of ERs (Brzozowski et al., 1997; Pike et al., 2000).

20.4

Androgenic/anti-androgenic potency of endocrinedisrupting chemicals

For more than three decades it has been known that estrogens alter the development of the mammalian reproductive system. The fact that EDCs are able to interact with ERs has attracted a great deal of attention (McLachlan, 2001; McLachlan et al., 2001). However, it has now become evident that besides the EDCs with estrogenic agonist and antagonist activities, some EDCs can exert anti-androgenic or androgenic actions via their interactions with the AR in vivo as well as in vitro (Gray et al., 2001, 2006). Moreover, besides this receptor-mediated mode of action, a number of anti-

Mechanisms of action of particular endocrine-disrupting chemicals

551

androgenic EDCs are known instead to exert their effect via a disruption of testosterone synthesis (Table 20.1). Among the latter are the phthalates which are industrial chemicals found abundantly in many consumer products such as flexible plastics. Di-(2-ethylhexyl) phthalate (DEHP) is the most commonly used additive in the fabrication of plastics and is distributed in very large amounts in the environment (ATSDR, 2002). Thus, humans and animals are highly exposed to phthalates (Koch et al., 2006; Wormuth et al., 2006). Once absorbed, DEHP is rapidly metabolised in the liver, gut and blood into mono-(2-ethylhexyl) phthalate (MEHP) known to be the active molecule (Gray and Gangolli, 1986). Numerous studies have shown that administration of DEHP disrupts male reproductive development. However, most of these studies were performed in utero (Barlow et al., 2003; Borch et al., 2006; Gray et al., 2001) which, though very useful, does not allow the discrimination between direct and indirect effects. Owing to the rapid metabolism of these chemicals, such approaches either cannot identify the phthalate(s) (actually active) at the level of the gonads, or cannot determine the dose at which effects occur. To overcome these limitations, we use an organotypic culture model coupled to morphological and functional methods to analyse the development of the fetal testis in the presence or absence of these compounds. We named this model FEGA for FEtal Gonad Assay and it was developed originally to investigate the mechanisms of action of gonadotropin hormones and regulatory factors on fetal rat testes (Habert et al., 1991; Livera et al., 2001). It has also been used to study the direct effect of estradiol and diethylstilbestrol (DES) on fetal testis in culture (Lassurguere et al., 2003). In this assay, the testes of rat fetuses at 14.5 days post-coitum (dpc) are dissected out and immediately explanted in vitro. They are cultured with the adjacent mesonephros on Millipore filters (pore size: 0.45 μm) as previously described (Habert et al., 1991). In our study, the immunopositive cells were counted using the Computer Assisted Stereology Toolbox (CAST) Grid System (Olympus, Copenhagen, Denmark) connected to a light microscope. All counts and measurements were done blind (Lassurguere et al., 2003). Our results show that DEHP did not alter the general testis morphology, as shown by WT1 immunostaining (a marker of Sertoli cells). By contrast, MEHP exposure produced dysgenic tubules, with Sertoli cells that appeared misplaced and, notably, a number of Leydig cells which were mislocalised within the tubules. Furthermore, a portion of the gonocytes had disappeared within the exposed testis as revealed by the occurrence of vacuoles in the seminiferous epithelium. However, the total number of Leydig and Sertoli cells did not vary with either treatment. While the gonocyte number did not vary in the DEHP-treated samples, the MEHP reduction of gonocytes resulted both from a decrease in gonocyte mitosis and an increase of gonocyte apoptosis. In conclusion, FEGA is the first in vitro system allowing the study of the direct effects of phthalates on fetal testis development and function. The

Linuron

pp′-DDT and its metabolite pp-DDE

Polybrominated diphenyl ethers (PBDES)

Phthalates

Insecticide

Flame retardant

Plasticisers

Inhibit testosterone synthesis

AR antagonist

AR antagonists

AR antagonist; also inhibits fetal testosterone synthesis of steroidogenic enzymes

AR antagonists

Vinchlozolin and procymidone

Herbicide

AR antagonist; also inhibits 17, 20 lyase and aromatase

Prochloraz

Fungicides

Mode of action

Toxicant

Induce damage to the reproductive organs (e.g. testis) in rats and rabbits; induce agenesis of the epididymis and gubernaculum; induce cryptorchidism and decreased anogenital distance; reduce accessory sex gland weight

Delays puberty; reduces accessory sex gland weight

Delay puberty in the male rat; reduces androgen-dependent tissues/organs; induce hypospadia and nipple retention

Induces testicular damage; inhibits fetal testosterone; reduces the weight of androgen-dependent tissues/organs

Reduces testosterone in the male rat fetus; disturbs normal development of the male, induces hypospasdias Perturb the development of puberty and of external genitalia; decrease ano-genital distance in male rats; reduce androgen-dependent organs; induce nipples and cryptorchidism in male rats; alter central nervous system sexual differentiation

Effects in the male

Diverse mechanisms and effects of anti-androgenic chemicals

Usage

Table 20.1

Carruthers and Foster (2005), Gray et al. (1999c), Higuchi et al. (2003), Stroheker et al. (2005)

Stoker et al. (2004, 2005)

Gray et al. (1999b, c), Kelce et al. (1997)

Gray et al. (1999c), Lambright et al. (2000), McIntyre et al. (2000), Wilson et al. (2007)

Gray et al. (1999a, b), Hotchkiss et al. (2003), Kelce et al. (1997), Monosson et al. (1999), Ostby et al. (1999)

Noriega et al. (2005), Vinggaard et al. (2002), Wilson et al. (2004, 2007)

References

Mechanisms of action of particular endocrine-disrupting chemicals

553

setting of this new tool paves the way for the study of the anti-androgenic action of the phthalates and of the screening of phthalate metabolites, as well as other hormones and xenohormones, for effects on the male fetal gonad.

20.5

Dioxin-like potency of endocrine-disrupting chemicals

The highly persistent and widespread environmental contaminants hydrocarbons (HAHs, PAH), PCB, benzo(α)pyrene (BaP), 2,3,7,8tetrachlorodibenzofuran (TCDF) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD, dioxin) that are the most potent members of this class of chemicals, produce a wide variety of species- and tissue-specific toxic and carcinogenic effects in laboratory animals and humans. The effects of these chemicals are mediated by the specific cytosolic receptor for dioxin, the aryl hydrocarbon receptor (AhR; Fig. 20.3). This receptor is a basic helix–loop–helix (bHLH)and Per–Arnt–Sim (PAS)-containing transcription factor that dimerises with its DNA binding partner, Arnt (AhR nuclear translocator). Arnt exhibits a structure similar to that of AhR and other members of the bHLH/ PAS transcription factor family (Fig. 20.3). The AhR/Arnt heterodimer complex is then able to bind, with high affinity, to the dioxin-responsive element (DRE) which is generally located upstream of target genes. DRE is also called xenobiotic-responsive element (XRE) or aryl hydrocarbon-

bHLH

PAS domain A

NLS NES

Transactivation domain

B

Q rich

HSP 90 binding Ligand binding AhR/Arnt/DRE complex formation

Fig. 20.3 Domain structure of the AhR, a member of the bHLH-PAS (basic helix– loop–helix/Per-Arnt-Sim) superfamily of transcription factors. The AhR (848aa, 96 kDa in human and 805aa, 90 kDa in mouse) contains a bHLH region that contains sequences important for both AhR nuclear localisation (NLS) and nuclear export (NES). The bHLH region is also involved in dimerisation with Arnt and DNA binding. The AhR PAS domain contains two structural repeats (PAS A and PAS B) which are involved in AhR/Arnt dimerisation (PAS A) and AhR ligand and hsp90 binding (PAS B). Hsp90, a molecular chaperone protein, is thought to be important for correct folding of the AhR LBD and for nuclear localisation. The C-terminal region of the AhR contains the transactivation domain including a rich region in glutamine residues (Q-rich).

554

Endocrine-disrupting chemicals in food

responsive element (AhRE) and its consensus core sequence is 5′TNGCGTGAGA-3′. The interaction of AhR/Arnt heterodimer with DRE leads to chromatin and nucleosome disruption, causing an increase in promoter accessibility and transcription rate of the specific target genes (Fig. 20.4). A number of cytochrome P450 (CYP) genes, including CYP 1A1, CYP1A2, CYP1B1 known to play major roles in xenobiotic and drug metabolism in the liver and also in extrahepatic tissues, contain DRE in their promoter. These genes, and also some involved in steroid metabolism (e.g. CYP19) that are regulated by AhR in a ligand-dependent manner, have been used as model systems (in particular CYP1A1) to study the mechanisms of action of dioxin toxicity in different species. In fish, dioxin showed high toxicity, most likely because of the existence of two distinct AhR in these species. Thus, fish CYPs genes have been used as interesting targets to study AhR functions and the mechanisms of dioxin toxicity, especially during development (Incardona et al., 2006; Whyte et al., 2000).

Ligand (TCDD)

Hsp 90 Cytoplasm AhR

Arnt

Nucleus Cofactors +

DRE

Xenobiotic response genes

Fig. 20.4 Transcriptional regulation of target genes by the AhR. Ligand binding to the AhR in the cytosol induces the release of the protein chaperone complexes such as Hsp90. This makes nuclear localisation signals accessible for importing binding and nuclear translocation. In the nucleus, AhR dimerises with Arnt to form a transcriptionally active heterodimer that can bind to dioxin response sequences (DRE) located upstream in the promoter of target genes. The active AhR/Arnt formed is then able to recruit cofactors and the basic transcriptional machinery, which allow the transcription of target genes.

Mechanisms of action of particular endocrine-disrupting chemicals

555

Knock-out of AhR expression in mice revealed that, in addition to the detoxification function of xenobiotic, AhR is involved in the control of cell proliferation and differentiation in diverse physiological processes. In fact, AhR is involved in development and homeostasis of liver, ovary, immune and cardiovascular systems, in neurogenesis, and it can promote migration of breast cancer cells and tumour development (Barouki et al., 2007). Several natural and endogenous ligands were also found to activate the AhR signalling pathway, although the majority of these chemicals appear to be relatively weak AhR ligands. Among dietary constituents, flavonoids, carotinoids or curcumine are able to induce endogenous CYP1A1 in vivo or bind to AhR. Resveratrol, a red wine component, acts as an antagonist of AhR. Although resveratrol induces AhR translocation in the nucleus, AhR/Arnt dimerisation and DNA binding, it inhibits the endogenous CYP1A1 gene in vivo. On the other hand, tryptophan (an aromatic amino acid) and lipid, lipoxin A4 produced from arachidonic acid metabolism have been reported as endogenous ligands of AhR. Although these chemicals appear to be relatively weak agonists, 7-ketocholesterol was reported to be a weak antagonist of AhR. Many studies reported that AhR ligands can also affect estrogenic signalling, as can be seen by the proliferation of mammary or uterine cells exposed to estrogens or the expression of estrogen-target genes in several vertebrate species. However, mechanistic studies on the interaction between dioxin and estrogen produced conflicting results probably because of the cellspecific actions and the complexity of dioxin’s hormonal effects. Recently, Ohtake and collaborators (Ohtake et al., 2003) reported that an AhR agonist, methylcholanthrene (3MC), is able to activate a reporter gene containing an ERE only in presence of both receptors, AhR and ER, without binding directly to ERs or affecting expression levels of ERs. When this promoter is activated by estradiol, 3MC has an antagonistic effect. These functional interactions are correlated with physical interaction between AhR and ER demonstrated by several techniques such as co-immunoprecipitation. This interaction depends on the presence of the AhR ligand, but is independent of the presence of estradiol. The proposed model (Fig. 20.5; Brosens and Parker, 2003) suggested that 3MC activates AhR which dimerises with Arnt. The AhR/Arnt heterodimer can directly associate with unliganded ER to activate estrogen-sensitive gene transcription by recruiting the coactivator p300 (Ohtake et al., 2003). This model is consolidated by in vivo experiments performed in the mouse. Indeed, trophic effect of 3MC on the uterus has been observed in the ovariectomised mouse, but not in AhR knock-out mice (AhR−/−) or ERα knock-out mice (ERα−/−). These studies suggest an original mechanism of activation of the ERα in the absence of estradiol since ligand-activated AhR is able to cooperate with the unliganded ER to activate the transcription (Fig. 20.5). However, in the absence or in the presence of estrogen, the AhR ligands, including 3MC, could behave as partial agonist or antagonist on ER signalling pathways.

556

Endocrine-disrupting chemicals in food (a) ER

E2 or xeno-E2

Cofactor

Estrogen target gene

ERE (b) AhR

Arnt

Dioxin

Cofactor

ERE

Estrogen target gene

Fig. 20.5 Interaction between AhR and ER at the estrogen-responsive-element. (a) Estrogens (E2) or xenoestrogens (xeno-E2) binding to ER induces DNAbinding at the ERE and recruitment of transcription factors that allow modulation of E2 target genes. (b) Dioxin binding to arylhydrocarbon receptor (AhR) induces formation of a transcriptionally active heterodimer AhR with arylhydrocarbon nuclear translocator (Arnt). AhR/Arnt heterodimer then interacts at the ERE site and modulates the transcription of E2 target gene by direct interaction with ER.

Several mechanisms have also been proposed to explain the antiestrogenicity of AhR ligands (Safe and Wormke, 2003). Liganded AhR may interfere with transcriptionally active ER/SP1 or ER/AP-1 complexes (Khan et al., 2006; Matthews et al., 2005; Pocar et al., 2005). Liganded AhR can also suppress the binding of liganded ER with ERE sites by direct association with this receptor (Ohtake et al., 2003). An AhR-mediated reduction in ERα protein may explain anti-estrogenic effects of dioxins as well (Wormke et al., 2003), but the AhR-mediated degradation rates may vary according to the specific cellular context (Safe and Wormke, 2003). Altogether, these studies highlight the importance of estrogen concentrations that might determine the estrogenic or anti-estrogenic effects elicited by AhR ligands. These findings should be taken into account when interpreting the results of studies investigating the estrogenic effects of AhR ligands, particularly in mixtures. Further studies on interplay between AhR and ER; and their ligands, target genes and cofactors are necessary to understand the estrogenic/anti-estrogenic actions of dioxins. Moreover, the understanding of the interplay between these receptors, crucial for develop-

Mechanisms of action of particular endocrine-disrupting chemicals

557

ment and reproduction, may allow preventing adverse effects that might be caused by the two major groups of environmental chemicals, xenoestrogens and dioxins. The expression of CYP19 aromatase genes has been generally linked with the process of sexual differentiation and was used as a biomarker of exposure to dioxin-like compounds. In zebrafish, E2 strongly upregulated the expression of CYP19B aromatase (Aro-B) gene in vivo (Kishida et al., 2001; Le Page et al., 2006; Menuet et al., 2005). Aro-B thus represents a good model to study the molecular mechanisms of action and possible cross-talk of estrogen- and dioxin-like compounds. Recently, we and others showed in zebrafish that co-exposure of E2 with AhR ligands (TCDD, B[a]P) led to significant down-regulation of E2-stimulation of Aro-B gene, both in vivo and in vitro (Cheshenko et al., 2007; Kazeto et al., 2004). This effect could either partially (in vivo) or fully (in vitro) be rescued by the addition of the AhR antagonist, ANF. The inability to observe full rescue in vivo could be explained by the complexity of in vivo models, including the rate of uptake, metabolism and biodegradation of the chemicals (Cheshenko et al., 2007). Moreover, in transfection experiments, we observed slight up-regulation of Aro-B promoter activity in the presence of zebrafish ERα, AhR2 and ARNT2b, but in the absence of E2. This up-regulation was, however, abolished by the addition of either an ER or AhR antagonist, suggesting the involvement of both receptors in the mechanism. Detailed studies confirmed the independence of this AhR action from the putative DRE elements predicted in the zebrafish CYP19B promoter. These effects observed in zebrafish (Cheshenko et al., 2007) correlated in many ways with previous studies on human ER (Ohtake et al., 2003).

20.6

Conclusions and future trends

The origin of EDCs in ecosystems is diverse and the sources of exposure to these chemicals are multiple. EDCs can be present in the various compartments of the environment (air, waters, soils) or can come from food, products of combustion, agricultural spraying, detergents and chemical industry in general. In addition to a widespread origin, EDCs show varying structural complexity and may act differently depending on cell context. For instance, although natural and synthetic xenoestrogens have generally phenolic or carbon ring structures, some heavy metals such as cadmium (Cd) have been reported to act as EDCs capable of interfering with estrogenic signalling (Le Guevel et al., 2000; Safe, 2003; Stoica et al., 2000). Both potent estrogenic and anti-estrogenic effects of cadmium have been demonstrated in vivo in the rodent uterus and in vitro in mammary cell lines, in recombinant yeast assays or in fish hepatocyte culture (Johnson et al., 2003; Le Guevel et al., 2000; Silva et al., 2006). Moreover, it has been shown that Cd has androgen-like activities in vivo and in vitro, by directly binding to

558

Endocrine-disrupting chemicals in food

androgen receptors. However, the precise mechanisms underlying the effects of Cd as an endocrine disruptor remain unclear (Takiguchi and Yoshihara, 2006). Mechanistic studies suggested that Cd could directly interact with the LBD of ERα resulting in the repositioning of helix 12 of the receptor which is critical for ER activation (Martin et al., 2003). Conversely, a study from our laboratory (Le Guevel et al., 2000) suggested that the interaction of Cd with the LBD of ERα reduces the interaction of activated ERα with DNA and consequently reduces ERα transcriptional activity. A recent paper also supports our hypothesis that the interaction of Cd with the C-terminal domain of ER does not result in activation of the receptor, but induces conformational changes of the DBD which could reduce the DNA binding activity of the receptor (Le Guevel et al., 2000; Silva et al., 2006). This further exemplifies the diversity and complexity of EDC effects and the need for further understanding of the diversity of their molecular mechanisms of actions. The assessment of environmental xenobiotics is rapidly increasing on the international level and many cell-based or in vitro models have been reported (Andersen et al., 1999; Le Guevel and Pakdel, 2001; Le Page et al., 2006; Legler et al., 1999; Petit et al., 1997; Routledge and Sumpter, 1997; Soto et al., 1995). Although such in vitro assays do not entirely reflect the in vivo situation, some of them constitute generally powerful assay because (i) of the high degree of specificity of the response, (ii) of their possible utilisation for high-throughput screening of large numbers of chemicals and (iii) these assays often allow determination of the mechanisms of molecular and cellular action of EDCs. In addition to ‘classical’ action of EDCs (i.e. modification of specific gene regulation) and phenotype alterations in the organism, recent studies suggested that exposure during development to environmental contaminants could induce epigenetic modifications and alterations in gene expression that can pass from one generation to the next (Jirtle and Skinner, 2007) (see Chapter 1). Indeed, exposure of developing rats to the fungicide vinclozolin affects the fertility of treated male animals that is transmitted through four generations without further exposure to vinclozolin (Anway et al., 2005, 2006; Chang et al., 2006b). Although, the fungicide vinclozolin shows anti-androgenic activity and its metabolites can interact with several steroid receptors (Kavlock and Cummings, 2005; Molina-Molina et al., 2006), it is unclear whether the effects of vinclozolin are mediated by its interference with the hormonal signalling during development. Further characterisation of how EDCs induce epigenetic modifications of DNA sequences and associated chromatin is of primary importance for specific gene regulations and phenotype alterations. It is important to stress that EDC effects are not only mediated through estrogen and androgen receptors. Indeed, EDCs can also interact with other nuclear receptors such as the thyroid receptor (TR), peroxisome proliferator-activated receptor (PPAR), retinoid X receptor (RXR), liver

Mechanisms of action of particular endocrine-disrupting chemicals

559

X receptor (LXR), and steroid and xenobiotic receptors (SXR; also called PXR) that are potential targets for these chemical compounds (Grun et al., 2006; Mitro et al., 2007; Zhou et al., 2007; Zoeller, 2005). In addition, EDCs do not necessarily mediate their effects by binding to specific NRs. Indirect effects are also possible. For instance, modifications of the level or activity of some coactivators, protein kinases, enzymes or transcription factors necessary for the activity of specific nuclear receptors (DNA binding, phosphorylation, transactivation, degradation, subcellular translocation). On the other hand, recent identification of the G-protein-coupled receptor homologue GPR30 as the plasma membrane receptor for estrogens provides a previously unevaluated mechanism of action of these hormones (Hewitt et al., 2005; Revankar et al., 2005). GPR30 is able to bind E2 and allows fast non-genomic responses of estrogens, such as stimulation of MAPK pathways, adenylyl cyclase or c-fos expression, in the breast cancer cell line SKBR3 which does not express the classical ERs (Filardo et al., 2002; Maggiolini et al., 2004; Vivacqua et al., 2006a,b). Interestingly, GPR30 is also sensitive to several phytoestrogens such as genistein and quercitin, or other xenoestrogens such as bisphenol A, zearalonone, nonylphenol that have also been shown to bind to the membrane estrogen receptor (Thomas and Dong, 2006; Watson et al., 2007). Since GPR30 is expressed in a wide number of cell types, it could thus potentially mimic estrogen actions of environmental chemicals in a broad range of tissues. Further characterisation of cellular and tissue distribution, as well as the mode of action of GPR30 and other plasma membrane receptors for steroid hormones, will likely contribute to a better understanding of the complexity of EDC actions in relation to the wide range of physiological roles played by steroid hormones.

20.7

Acknowledgements

This research was supported by grants from European Union Commission (EDEN Project QLK4-CT-2002-00603) and the French Ministry of Ecology and Sustainable Development (Programme National de Recherche sur les Perturbateurs Endocriniens du Ministère de l’Ecologie et du Développement Durable, CV no.05000194).

20.8

References

abdelrahim, m., ariazi, e., kim, k., khan, s., barhoumi, r., burghardt, r., liu, s., hill, d., finnell, r., wlodarczyk, b., et al. (2006). 3-Methylcholanthrene and other aryl hydrocarbon receptor agonists directly activate estrogen receptor α. Cancer Res 66, 2459–2467. andersen, h. r., andersson, a. m., arnold, s. f., autrup, h., barfoed, m., beresford, n. a., bjerregaard, p., christiansen, l. b., gissel, b., hummel, r., et al. (1999).

560

Endocrine-disrupting chemicals in food

Comparison of short-term estrogenicity tests for identification of hormonedisrupting chemicals. Environ Health Perspect 107 Suppl 1, 89–108. anway, m. d., cupp, a. s., uzumcu, m., and skinner, m. k. (2005). Epigenetic transgenerational actions of endocrine disruptors and male fertility. Science 308, 1466–1469. anway, m. d., memon, m. a., uzumcu, m., and skinner, m. k. (2006). Transgenerational effect of the endocrine disruptor vinclozolin on male spermatogenesis. J Androl 27, 868–879. atsdr (2002). Toxicological profile of Di(2-ethylhexyl)phthalate. Agency for Toxic Substances and Disease Registry, Public Health Service, US Department of Health and Human Services, Atlanta. balaguer, p., joyeux, a., denison, m. s., vincent, r., gillesby, b. e., and zacharewski, t. (1996). Assessing the estrogenic and dioxin-like activities of chemicals and complex mixtures using in vitro recombinant receptor–reporter gene assays. Can J Physiol Pharmacol 74, 216–222. barlow, n. j., phillips, s. l., wallace, d. g., sar, m., gaido, k. w., and foster, p. m. (2003). Quantitative changes in gene expression in fetal rat testes following exposure to di(n-butyl) phthalate. Toxicol Sci 73, 431–441. barouki, r., coumoul, x., and fernandez-salguero, p. m. (2007). The aryl hydrocarbon receptor, more than a xenobiotic-interacting protein. FEBS Lett 581, 3608–3615. borch, j., metzdorff, s. b., vinggaard, a. m., brokken, l., and dalgaard, m. (2006). Mechanisms underlying the anti-androgenic effects of diethylhexyl phthalate in fetal rat testis. Toxicology 223, 144–155. brosens, j. j., and parker, m. g. (2003). Gene expression: oestrogen receptor hijacked. Nature 423, 487–488. brzozowski, a. m., pike, a. c., dauter, z., hubbard, r. e., bonn, t., engstrom, o., ohman, l., greene, g. l., gustafsson, j. a., and carlquist, m. (1997). Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389, 753–758. carreau, s., bourguiba, s., lambard, s., galeraud-denis, i., genissel, c., and levallet, j. (2002). Reproductive system: aromatase and estrogens. Mol Cell Endocrinol 193, 137–143. carruthers, c. m., and foster, p. m. (2005). Critical window of male reproductive tract development in rats following gestational exposure to di-n-butyl phthalate. Birth Defects Res B Dev Reprod Toxicol 74, 277–285. chang, e. c., frasor, j., komm, b., and katzenellenbogen, b. s. (2006a). Impact of estrogen receptor β on gene networks regulated by estrogen receptor α in breast cancer cells. Endocrinology 147, 4831–4842. chang, h. s., anway, m. d., rekow, s. s., and skinner, m. k. (2006b). Transgenerational epigenetic imprinting of the male germline by endocrine disruptor exposure during gonadal sex determination. Endocrinology 147, 5524–5541. chen, i., hsieh, t., thomas, t., and safe, s. (2001). Identification of estrogen-induced genes downregulated by AhR agonists in MCF-7 breast cancer cells using suppression subtractive hybridization. Gene 262, 207–214. cheshenko, k., brion, f., le page, y., hinfray, n., pakdel, f., kah, o., segner, h., and eggen, r. i. (2007). Expression of zebra fish aromatase cyp19a and cyp19b genes in response to the ligands of estrogen receptor and aryl hydrocarbon receptor. Toxicol Sci 96, 255–267. colborn, t., vom saal, f. s., and soto, a. m. (1993). Developmental effects of endocrine-disrupting chemicals in wildlife and humans. Environ Health Perspect 101, 378–384. escande, a., pillon, a., servant, n., cravedi, j. p., larrea, f., muhn, p., nicolas, j. c., cavailles, v., and balaguer, p. (2006). Evaluation of ligand selectivity using

Mechanisms of action of particular endocrine-disrupting chemicals

561

reporter cell lines stably expressing estrogen receptor α or β. Biochem Pharmacol 71, 1459–1469. fent, k. (2004). Ecotoxicological effects at contaminated sites. Toxicology 205, 223–240. filardo, e. j., quinn, j. a., frackelton, a. r., jr, and bland, k. i. (2002). Estrogen action via the G protein-coupled receptor, GPR30: stimulation of adenylyl cyclase and cAMP-mediated attenuation of the epidermal growth factor receptor-toMAPK signaling axis. Mol Endocrinol 16, 70–84. flouriot, g., pakdel, f., ducouret, b., and valotaire, y. (1995). Influence of xenobiotics on rainbow trout liver estrogen receptor and vitellogenin gene expression. J Mol Endocrinol 15, 143–151. frasor, j., danes, j. m., komm, b., chang, k. c., lyttle, c. r., and katzenellenbogen, b. s. (2003). Profiling of estrogen up- and down-regulated gene expression in human breast cancer cells: insights into gene networks and pathways underlying estrogenic control of proliferation and cell phenotype. Endocrinology 144, 4562–4574. germain, p., staels, b., dacquet, c., spedding, m., and laudet, v. (2006). Overview of nomenclature of nuclear receptors. Pharmacol Rev 58, 685–704. gray, l. e., jr, ostby, j., cooper, r. l., and kelce, w. r. (1999a). The estrogenic and antiandrogenic pesticide methoxychlor alters the reproductive tract and behavior without affecting pituitary size or LH and prolactin secretion in male rats. Toxicol Ind Health 15, 37–47. gray, l. e., jr, ostby, j., monosson, e., and kelce, w. r. (1999b). Environmental antiandrogens: low doses of the fungicide vinclozolin alter sexual differentiation of the male rat. Toxicol Ind Health 15, 48–64. gray, l. e., jr, wolf, c., lambright, c., mann, p., price, m., cooper, r. l., and ostby, j. (1999c). Administration of potentially antiandrogenic pesticides (procymidone, linuron, iprodione, chlozolinate, p,p′-DDE, and ketoconazole) and toxic substances (dibutyl- and diethylhexyl phthalate, PCB 169, and ethane dimethane sulphonate) during sexual differentiation produces diverse profiles of reproductive malformations in the male rat. Toxicol Ind Health 15, 94–118. gray, l. e., ostby, j., furr, j., wolf, c. j., lambright, c., parks, l., veeramachaneni, d. n., wilson, v., price, m., hotchkiss, a., et al. (2001). Effects of environmental antiandrogens on reproductive development in experimental animals. Hum Reprod Update 7, 248–264. gray, l. e., jr, wilson, v. s., stoker, t., lambright, c., furr, j., noriega, n., howdeshell, k., ankley, g. t., and guillette, l. (2006). Adverse effects of environmental antiandrogens and androgens on reproductive development in mammals. Int J Androl 29, 96–104; discussion 105–108. gray, t. j., and gangolli, s. d. (1986). Aspects of the testicular toxicity of phthalate esters. Environ Health Perspect 65, 229–235. grun, f., watanabe, h., zamanian, z., maeda, l., arima, k., cubacha, r., gardiner, d. m., kanno, j., iguchi, t., and blumberg, b. (2006). Endocrine-disrupting organotin compounds are potent inducers of adipogenesis in vertebrates. Mol Endocrinol 20, 2141–2155. guillette, l. j., jr, gross, t. s., masson, g. r., matter, j. m., percival, h. f., and woodward, a. r. (1994). Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environ Health Perspect 102, 680–688. gustafsson, j. a. (1998). Therapeutic potential of selective estrogen receptor modulators. Curr Opin Chem Biol 2, 508–511. habert, r., devif, i., gangnerau, m. n., and lecerf, l. (1991). Ontogenesis of the in vitro response of rat testis to gonadotropin-releasing hormone. Mol Cell Endocrinol 82, 199–206.

562

Endocrine-disrupting chemicals in food

hess, r. a., bunick, d., lee, k. h., bahr, j., taylor, j. a., korach, k. s., and lubahn, d. b. (1997). A role for oestrogens in the male reproductive system. Nature 390, 509–512. hewitt, s. c., deroo, b. j., and korach, k. s. (2005). Signal transduction. A new mediator for an old hormone? Science 307, 1572–1573. incardona, j. p., day, h. l., collier, t. k., and scholz, n. l. (2006). Developmental toxicity of 4-ring polycyclic aromatic hydrocarbons in zebrafish is differentially dependent on AH receptor isoforms and hepatic cytochrome P4501A metabolism. Toxicol Appl Pharmacol 217, 308–321. jagadeesh, s., kyo, s., and banerjee, p. p. (2006). Genistein represses telomerase activity via both transcriptional and posttranslational mechanisms in human prostate cancer cells. Cancer Res 66, 2107–2115. jegou, b., auger, j., multigner, l., pineau, c., thonneau, p., spira, a., and jouannet, p. (1999). The saga of the sperm count decrease in humans wild and farm animals. In: Ganon, C. (ed), The Male Gamete: From basic science to clinical applications. Cach River Press, Vienna, IL, USA, 1999, Chap 41, pp 445–454. higuchi, t. t., palmer, j. s., gray, l. e., jr., and veeramachaneni, d. n. (2003). Effects of dibutyl phthalate in male rabbits following in utero, adolescent, or postpubertal exposure. Toxicol Sci 72, 301–313. hotchkiss, a. k., ostby, j. s., vandenbergh, j. g., and gray, l. e., jr (2003). An environmental antiandrogen, vinclozolin, alters the organization of play behavior. Physiol Behav 79, 151–156. jirtle, r. l., and skinner, m. k. (2007). Environmental epigenomics and disease susceptibility. Nat Rev Genet 8, 253–262. johnson, m. d., kenney, n., stoica, a., hilakivi-clarke, l., singh, b., chepko, g., clarke, r., sholler, p. f., lirio, a. a., foss, c., et al. (2003). Cadmium mimics the in vivo effects of estrogen in the uterus and mammary gland. Nat Med 9, 1081–1084. jordan, v. c., and brodie, a. m. (2007). Development and evolution of therapies targeted to the estrogen receptor for the treatment and prevention of breast cancer. Steroids 72, 7–25. katzenellenbogen, b. s., choi, i., delage-mourroux, r., ediger, t. r., martini, p. g., montano, m., sun, j., weis, k., and katzenellenbogen, j. a. (2000). Molecular mechanisms of estrogen action: selective ligands and receptor pharmacology. J Steroid Biochem Mol Biol 74, 279–285. kavlock, r., and cummings, a. (2005). Mode of action: reduction of testosterone availability – molinate-induced inhibition of spermatogenesis. Crit Rev Toxicol 35, 685–690. kazeto, y., place, a. r., and trant, j. m. (2004). Effects of endocrine disrupting chemicals on the expression of CYP19 genes in zebrafish (Danio rerio) juveniles. Aquat Toxicol 69, 25–34. kelce, w. r., lambright, c. r., gray, l. e., jr., and roberts, k. p. (1997). Vinclozolin and p,p′-DDE alter androgen-dependent gene expression: in vivo confirmation of an androgen receptor-mediated mechanism. Toxicol Appl Pharmacol 142, 192–200. khan, s., barhoumi, r., burghardt, r., liu, s., kim, k., and safe, s. (2006). Molecular mechanism of inhibitory aryl hydrocarbon receptor–estrogen receptor/Sp1 cross talk in breast cancer cells. Mol Endocrinol 20, 2199–2214. kian tee, m., rogatsky, i., tzagarakis-foster, c., cvoro, a., an, j., christy, r. j., yamamoto, k. r., and leitman, d. c. (2004). Estradiol and selective estrogen receptor modulators differentially regulate target genes with estrogen receptors alpha and beta. Mol Biol Cell 15, 1262–1272. kim, s. h., tamrazi, a., carlson, k. e., and katzenellenbogen, j. a. (2005). A proteomic microarray approach for exploring ligand-initiated nuclear hormone

Mechanisms of action of particular endocrine-disrupting chemicals

563

receptor pharmacology, receptor selectivity, and heterodimer functionality. Mol Cell Proteomics 4, 267–277. kishida, m., mclellan, m., miranda, j. a., and callard, g. v. (2001). Estrogen and xenoestrogens upregulate the brain aromatase isoform (P450aromB) and perturb markers of early development in zebrafish (Danio rerio). Comp Biochem Physiol B Biochem Mol Biol 129, 261–268. klinge, c. m. (2000). Estrogen receptor interaction with co-activators and corepressors. Steroids 65, 227–251. koch, h. m., preuss, r., and angerer, j. (2006). Di(2-ethylhexyl)phthalate (DEHP): human metabolism and internal exposure – an update and latest results. Int J Androl 29, 155–165; discussion 181–155. kuiper, g. g., carlsson, b., grandien, k., enmark, e., haggblad, j., nilsson, s., and gustafsson, j. a. (1997). Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors α and β. Endocrinology 138, 863–870. kuiper, g. g., lemmen, j. g., carlsson, b., corton, j. c., safe, s. h., van der saag, p. t., van der burg, b., and gustafsson, j. a. (1998). Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor β. Endocrinology 139, 4252–4263. lambright, c., ostby, j., bobseine, k., wilson, v., hotchkiss, a. k., mann, p. c., and gray, l. e., jr (2000). Cellular and molecular mechanisms of action of linuron: an antiandrogenic herbicide that produces reproductive malformations in male rats. Toxicol Sci 56, 389–399. lassurguere, j., livera, g., habert, r., and jegou, b. (2003). Time- and dose-related effects of estradiol and diethylstilbestrol on the morphology and function of the fetal rat testis in culture. Toxicol Sci 73, 160–169. le guevel, r., and pakdel, f. (2001). Assessment of oestrogenic potency of chemicals used as growth promoter by in-vitro methods. Hum Reprod 16, 1030–1036. le guevel, r. l., petit, f. g., goff, p. l., metivier, r., valotaire, y., and pakdel, f. (2000). Inhibition of rainbow trout (Oncorhynchus mykiss) estrogen receptor activity by cadmium. Biol Reprod 63, 259–266. le page, y., scholze, m., kah, o., and pakdel, f. (2006). Assessment of xenoestrogens using three distinct estrogen receptors and the zebrafish brain aromatase gene in a highly responsive glial cell system. Environ Health Perspect 114, 752–758. legler, j., van den brink, c. e., brouwer, a., murk, a. j., van der saag, p. t., vethaak, a. d., and van der burg, b. (1999). Development of a stably transfected estrogen receptor-mediated luciferase reporter gene assay in the human T47D breast cancer cell line. Toxicol Sci 48, 55–66. levin, e. r. (2003). Bidirectional signaling between the estrogen receptor and the epidermal growth factor receptor. Mol Endocrinol 17, 309–317. levin, e. r., and pietras, r. j. (2007). Estrogen receptors outside the nucleus in breast cancer. Breast Cancer Res Treat 708, 351–361. livera, g., rouiller-fabre, v., and habert, r. (2001). Retinoid receptors involved in the effects of retinoic acid on rat testis development. Biol Reprod 64, 1307–1314. maggiolini, m., vivacqua, a., fasanella, g., recchia, a. g., sisci, d., pezzi, v., montanaro, d., musti, a. m., picard, d., and ando, s. (2004). The G protein-coupled receptor GPR30 mediates c-fos up-regulation by 17β-estradiol and phytoestrogens in breast cancer cells. J Biol Chem 279, 27008–27016. martin, m. b., reiter, r., pham, t., avellanet, y. r., camara, j., lahm, m., pentecost, e., pratap, k., gilmore, b. a., divekar, s., et al. (2003). Estrogen-like activity of metals in MCF-7 breast cancer cells. Endocrinology 144, 2425–2436. matthews, j., wihlen, b., thomsen, j., and gustafsson, j. a. (2005). Aryl hydrocarbon receptor-mediated transcription: ligand-dependent recruitment of estrogen

564

Endocrine-disrupting chemicals in food

receptor α to 2,3,7,8-tetrachlorodibenzo-p-dioxin-responsive promoters. Mol Cell Biol 25, 5317–5328. mcintyre, b. s., barlow, n. j., wallace, d. g., maness, s. c., gaido, k. w., and foster, p. m. (2000). Effects of in utero exposure to linuron on androgen-dependent reproductive development in the male Crl:CD(SD)BR rat. Toxicol Appl Pharmacol 167, 87–99. mclachlan, j. a. (2001). Environmental signaling: what embryos and evolution teach us about endocrine disrupting chemicals. Endocr Rev 22, 319–341. mclachlan, j. a., burow, m., chiang, t. c., and li, s. f. (2001). Gene imprinting in developmental toxicology: a possible interface between physiology and pathology. Toxicol Lett 120, 161–164. menuet, a., pellegrini, e., anglade, i., blaise, o., laudet, v., kah, o., and pakdel, f. (2002). Molecular characterization of three estrogen receptor forms in zebrafish: binding characteristics, transactivation properties, and tissue distributions. Biol Reprod 66, 1881–1892. menuet, a., le page, y., torres, o., kern, l., kah, o., and pakdel, f. (2004). Analysis of the estrogen regulation of the zebrafish estrogen receptor (ER) reveals distinct effects of ERα, ERβ1 and ERβ2. J Mol Endocrinol 32, 975–986. menuet, a., pellegrini, e., brion, f., gueguen, m., anglade, i., pakdel, f., and kah, o. (2005). Expression and the estrogen-dependent regulation of the zebrafish brain aromatase gene. J Comp Neurol 485, 304–320. mitro, n., mak, p. a., vargas, l., godio, c., hampton, e., molteni, v., kreusch, a., and saez, e. (2007). The nuclear receptor LXR is a glucose sensor. Nature 445, 219–223. molina-molina, j. m., hillenweck, a., jouanin, i., zalko, d., cravedi, j. p., fernandez, m. f., pillon, a., nicolas, j. c., olea, n., and balaguer, p. (2006). Steroid receptor profiling of vinclozolin and its primary metabolites. Toxicol Appl Pharmacol 216, 44–54. monosson, e., kelce, w. r., lambright, c., ostby, j., and gray, l. e., jr (1999). Peripubertal exposure to the antiandrogenic fungicide, vinclozolin, delays puberty, inhibits the development of androgen-dependent tissues, and alters androgen receptor function in the male rat. Toxicol Ind Health 15, 65–79. navas, j. m., chana, a., herradon, b., and segner, h. (2004). Induction of cytochrome P4501A (CYP1A) by clotrimazole, a non-planar aromatic compound. Computational studies on structural features of clotrimazole and related imidazole derivatives. Life Sci 76, 699–714. nilsson, s., makela, s., treuter, e., tujague, m., thomsen, j., andersson, g., enmark, e., pettersson, k., warner, m., and gustafsson, j. a. (2001). Mechanisms of estrogen action. Physiol Rev 81, 1535–1565. noriega, n. c., ostby, j., lambright, c., wilson, v. s., and gray, l. e., jr (2005). Late gestational exposure to the fungicide prochloraz delays the onset of parturition and causes reproductive malformations in male but not female rat offspring. Biol Reprod 72, 1324–1335. ohtake, f., takeyama, k., matsumoto, t., kitagawa, h., yamamoto, y., nohara, k., tohyama, c., krust, a., mimura, j., chambon, p., et al. (2003). Modulation of oestrogen receptor signalling by association with the activated dioxin receptor. Nature 423, 545–550. ostby, j., kelce, w. r., lambright, c., wolf, c. j., mann, p., and gray, l. e., jr. (1999). The fungicide procymidone alters sexual differentiation in the male rat by acting as an androgen-receptor antagonist in vivo and in vitro. Toxicol Ind Health 15, 80–93. paech, k., webb, p., kuiper, g. g., nilsson, s., gustafsson, j., kushner, p. j., and scanlan, t. s. (1997). Differential ligand activation of estrogen receptors ERα and ERβ at AP1 sites. Science 277, 1508–1510.

Mechanisms of action of particular endocrine-disrupting chemicals

565

pellegrini, e., menuet, a., lethimonier, c., adrio, f., mm., g., tasscon, c., anglade, i., pakdel, f., and kah, o. (2005). Relationships between aromatase and estrogen receptors in the brain of teleost fish. General Comparative Endocrinol 142, 60–66. petit, f., le goff, p., cravedi, j. p., valotaire, y., and pakdel, f. (1997). Two complementary bioassays for screening the estrogenic potency of xenobiotics: recombinant yeast for trout estrogen receptor and trout hepatocyte cultures. J Mol Endocrinol 19, 321–335. pike, a. c., brzozowski, a. m., and hubbard, r. e. (2000). A structural biologist’s view of the oestrogen receptor. J Steroid Biochem Mol Biol 74, 261–268. pocar, p., fischer, b., klonisch, t., and hombach-klonisch, s. (2005). Molecular interactions of the aryl hydrocarbon receptor and its biological and toxicological relevance for reproduction. Reproduction 129, 379–389. rajapakse, n., silva, e., and kortenkamp, a. (2002). Combining xenoestrogens at levels below individual no-observed-effect concentrations dramatically enhances steroid hormone action. Environ Health Perspect 110, 917–921. revankar, c. m., cimino, d. f., sklar, l. a., arterburn, j. b., and prossnitz, e. r. (2005). A transmembrane intracellular estrogen receptor mediates rapid cell signaling. Science 307, 1625–1630. robinson-rechavi, m., escriva garcia, h., and laudet, v. (2003). The nuclear receptor superfamily. J Cell Sci 116, 585–586. routledge, e. j., and sumpter, j. p. (1997). Structural features of alkylphenolic chemicals associated with estrogenic activity. J Biol Chem 272, 3280–3288. safe, s. (2001a). Molecular biology of the Ah receptor and its role in carcinogenesis. Toxicol Lett 120, 1–7. safe, s. (2001b). Transcriptional activation of genes by 17-β-estradiol through estrogen receptor–Sp1 interactions. Vitam Horm 62, 231–252. safe, s. (2003). Cadmium’s disguise dupes the estrogen receptor. Nat Med 9, 1000–1001. safe, s., and kim, k. (2004). Nuclear receptor-mediated transactivation through interaction with Sp proteins. Prog Nucleic Acid Res Mol Biol 77, 1–36. safe, s., and wormke, m. (2003). Inhibitory aryl hydrocarbon receptor–estrogen receptor alpha cross-talk and mechanisms of action. Chem Res Toxicol 16, 807–816. sharpe, r. m., and skakkebaek, n. e. (1993). Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet 341, 1392–1395. silva, e., rajapakse, n., and kortenkamp, a. (2002). Something from ‘nothing’ – eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ Sci Technol 36, 1751–1756. silva, e., lopez-espinosa, m. j., molina-molina, j. m., fernandez, m., olea, n., and kortenkamp, a. (2006). Lack of activity of cadmium in in vitro estrogenicity assays. Toxicol Appl Pharmacol 216, 20–28. sohoni, p., and sumpter, j. p. (1998). Several environmental oestrogens are also antiandrogens. J Endocrinol 158, 327–339. sonnenschein, c., and soto, a. m. (1998). An updated review of environmental estrogen and androgen mimics and antagonists. J Steroid Biochem Mol Biol 65, 143–150. soto, a. m., sonnenschein, c., chung, k. l., fernandez, m. f., olea, n., and serrano, f. o. (1995). The E-SCREEN assay as a tool to identify estrogens: an update on estrogenic environmental pollutants. Environ Health Perspect 103 Suppl 7, 113–122. soto, a. m., michaelson, c. l., prechtl, n. v., weill, b. c., sonnenschein, c., oleaserrano, f., and olea, n. (1998). Assays to measure estrogen and androgen agonists and antagonists. Adv Exp Med Biol 444, 9–23; discussion 23–28.

566

Endocrine-disrupting chemicals in food

staub, c., rauch, m., ferriere, f., trepos, m., dorval-coiffec, i., saunders, p. t., cobellis, g., flouriot, g., saligaut, c., and jegou, b. (2005). Expression of estrogen receptor ESR1 and its 46-kDa variant in the gubernaculum testis. Biol Reprod 73, 703–712. stauffer, s. r., coletta, c. j., tedesco, r., nishiguchi, g., carlson, k., sun, j., katzenellenbogen, b. s., and katzenellenbogen, j. a. (2000). Pyrazole ligands: structure– affinity/activity relationships and estrogen receptor-α-selective agonists. J Med Chem 43, 4934–4947. stoica, a., katzenellenbogen, b. s., and martin, m. b. (2000). Activation of estrogen receptor-alpha by the heavy metal cadmium. Mol Endocrinol 14, 545–553. stoker, t. e., laws, s. c., crofton, k. m., hedge, j. m., ferrell, j. m., and cooper, r. l. (2004). Assessment of DE-71, a commercial polybrominated diphenyl ether (PBDE) mixture, in the EDSP male and female pubertal protocols. Toxicol Sci 78, 144–155. stoker, t. e., cooper, r. l., lambright, c. s., wilson, v. s., furr, j., and gray, l. e. (2005). In vivo and in vitro anti-androgenic effects of DE-71, a commercial polybrominated diphenyl ether (PBDE) mixture. Toxicol Appl Pharmacol 207, 78–88. stroheker, t., cabaton, n., nourdin, g., regnier, j. f., lhuguenot, j. c., and chagnon, m. c. (2005). Evaluation of anti-androgenic activity of di-(2-ethylhexyl) phthatate. Toxicology 208, 115–121. sumpter, j. p. (1995). Feminized responses in fish to environmental estrogens. Toxicol Lett 82–83, 737–742. sumpter, j. p., and jobling, s. (1993). Male sexual development in ‘a sea of oestrogen’. Lancet 342, 124–125. sumpter, j. p., and jobling, s. (1995). Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ Health Perspect 103 Suppl 7, 173–178. takiguchi, m., and yoshihara, s. (2006). New aspects of cadmium as endocrine disruptor. Environ Sci 13, 107–116. thomas, p., and dong, j. (2006). Binding and activation of the seven-transmembrane estrogen receptor GPR30 by environmental estrogens: a potential novel mechanism of endocrine disruption. J Steroid Biochem Mol Biol 102, 175–179. toppari, j., virtanen, h., skakkebaek, n. e., and main, k. m. (2006). Environmental effects on hormonal regulation of testicular descent. J Steroid Biochem Mol Biol 102, 184–186. vinggaard, a. m., nellemann, c., dalgaard, m., jorgensen, e. b., and andersen, h. r. (2002). Antiandrogenic effects in vitro and in vivo of the fungicide prochloraz. Toxicol Sci 69, 344–353. vivacqua, a., bonofiglio, d., albanito, l., madeo, a., rago, v., carpino, a., musti, a. m., picard, d., ando, s., and maggiolini, m. (2006a). 17β-Estradiol, genistein, and 4-hydroxytamoxifen induce the proliferation of thyroid cancer cells through the g protein-coupled receptor GPR30. Mol Pharmacol 70, 1414–1423. vivacqua, a., bonofiglio, d., recchia, a. g., musti, a. m., picard, d., ando, s., and maggiolini, m. (2006b). The G protein-coupled receptor GPR30 mediates the proliferative effects induced by 17β-estradiol and hydroxytamoxifen in endometrial cancer cells. Mol Endocrinol 20, 631–646. watson, c. s., alyea, r. a., jeng, y. j., and kochukov, m. y. (2007). Nongenomic actions of low concentration estrogens and xenoestrogens on multiple tissues. Mol Cell Endocrinol 274, 1–7. whyte, j. j., jung, r. e., schmitt, c. j., and tillitt, d. e. (2000). Ethoxyresorufin-Odeethylase (EROD) activity in fish as a biomarker of chemical exposure. Crit Rev Toxicol 30, 347–570.

Mechanisms of action of particular endocrine-disrupting chemicals

567

wilson, v. s., lambright, c., furr, j., ostby, j., wood, c., held, g., and gray, l. e., jr. (2004). Phthalate ester-induced gubernacular lesions are associated with reduced insl3 gene expression in the fetal rat testis. Toxicol Lett 146, 207–215. wilson, v. s., howdeshell, k. l., lambright, c. s., furr, j., and earl gray, l., jr. (2007). Differential expression of the phthalate syndrome in male Sprague-Dawley and Wistar rats after in utero DEHP exposure. Toxicol Lett 170, 177–184. wormke, m., stoner, m., saville, b., walker, k., abdelrahim, m., burghardt, r., and safe, s. (2003). The aryl hydrocarbon receptor mediates degradation of estrogen receptor α through activation of proteasomes. Mol Cell Biol 23, 1843–1855. wormuth, m., scheringer, m., vollenweider, m., and hungerbuhler, k. (2006). What are the sources of exposure to eight frequently used phthalic acid esters in Europeans? Risk Anal 26, 803–824. yamashita, s. (2006). Expression of estrogen-regulated genes during development in the mouse uterus exposed to diethylstilbestrol neonatally. Curr Pharm Des 12, 1505–1520. zhou, c., zhang, s., nanamori, m., zhang, y., liu, q., li, n., sun, m., tian, j., ye, p. p., cheng, n., et al. (2007). Pharmacological characterization of a novel nonpeptide antagonist for formyl peptide receptor-like 1. Mol Pharmacol 72, 976–983. zoeller, r. t. (2005). Environmental chemicals as thyroid hormone analogues: new studies indicate that thyroid hormone receptors are targets of industrial chemicals? Mol Cell Endocrinol 242, 10–15.

21 Epilogue I. Shaw, University of Canterbury, New Zealand

So are endocrine-disrupting chemicals (EDCs) in food a risk to consumers’ health? This text has brought together a great deal of thinking from some of the top scientists in the field and it is clear that on one side of the risk– benefit equation are the negative health impacts such as precocious puberty, reduced sperm count and effects on gene expression in utero and on the other the positive hormonal effects that consuming phytoestrogen-rich foods might have in menopausal osteoporosis and redressing hormone imbalance immediately post menopause. I am convinced by the arguments and examples given amidst the millions of words in this text that the risk– benefit balance lies tilted quite steeply towards risk. It is obvious to me that we must take the risks presented by EDCs in food seriously. Whether we should regulate EDCs or not is a moot point. However, I think we should regulate endocrine-disrupting components of plastics used in food packaging – this is where our highest doses of synthetic EDCs come from. However the highest food-related intake of any EDC is from soy – clearly it would be difficult to set safe limits for soy consumption, but we could through educational processes and good food labelling reduce the ‘esoteric’ use of soy in places where it is used as a protein or carbohydrate source to replace more conventional protein sources – for example, in soy ‘milk’. Clearly soy ‘milk’ has an important part to play in preventing illness in people with, for example, lactose intolerance, but in my opinion it should not be used to replace real milk for those without medical conditions. If consumers were aware of the issues and food was labelled as soy-containing, some people might choose not to buy that food; the market would then regulate soy intake for us. A good example is the use of soy flour in breadmaking – soy flour is used to increase the protein content of low-gluten

Epilogue

569

wheat flours. Educate people to question the use of soy and label the bread properly and we might reduce phytoestrogen consumption in the breadeating parts of the world. We must balance the risk with the benefit and perhaps even introduce a precautionary approach in cases where we do not fully understand the implications to consumers’ health. In bringing this text together we have inevitably missed important issues related to EDCs in food; for example their role in non-genotoxic carcinogenesis. Chemicals such as xenoestrogens might result in carcinogenesis by a non-DNA altering mechanism. For example xenoestrogens occupy breast cell endocrine receptors (ERs) and stimulate cell proliferation. Such proliferation increases the risk of a genetic material transcription errors that might in turn initiate carcinogenesis. In addition xenoestrogens might act as promoters in the initiator–promotor carcinogenesis hypothesis. For example a breast cell with a genetic abnormality might lie dormant until its ER is occupied by estrogen or an estrogen mimic. Such occupancy would promote the proliferation of the cell and might ‘wake up’ the cancer potential. These are important considerations when assessing the risks of human exposure to EDCs. The discovery of EDCs began with observations that exposure to pesticides and other environmental chemicals might be causing hormone-related effects. Guillette’s work (Guillette et al, 1994) with alligators is seminal in this respect. He attributed the effects to dichlorodiphenyltrichloroethane (DDT) contamination of the alligators’ environment. Since Guillette’s work many studies on pesticides have shown the potential for a broad array of these agriculturally important, but environmentally risky chemicals to mimic sex hormones. For example the molecular structure of DDT has been shown to nicely fit the electronically recreated ER with its two electronwithdrawing ring-chlorines mimicking the two hydroxyl groups of 17βestradiol and the long hydrophobic phenyl groups and trichloromethyl group of the DDT molecule mimicking the equally long hydrophobic region of the steroid ring structure of 17β-estradiol (see Shaw & Chadwick, 1998). At the other end of the spectrum there have been concerns about farm workers exposed to Vinclozalin – this is a fungicide used in horticulture that appears to occupy the androgen receptor and blocks access of its natural ligands so reducing maleness following high-level prolonged exposure. Farm workers in the UK illegally spraying greenhouse lettuces with Vinclozalin (WPPR, 1998) were one such highly exposed group and spraying the lettuces exposed consumers to residues. When the farmers realised that there was risk to their maleness the misuse stopped! Very much more recent work has shown that some apparently very lowrisk (from the human effects point of view) pesticides might be converted in the environment (e.g. by microbial metabolism) or by human metabolism to estrogenic compounds. A good example is the pyrethroid insecticide cypermethrin which degrades in the environment or is metabolised in mammals to 3-phenoxybenzyl alcohol, 3-(4-hydroxy-3-phenoxy)benzyl

570

Endocrine-disrupting chemicals in food

alcohol and 3-phenoxybenzaldehyde, all of which were positive in the yeast estrogenicity assay (YES) (McCarthy et al., 2006). What the environmental impact of this might be is very difficult to speculate on, but since EDC effects are receptor-mediated they are at least additive, so examples such as this add further to the estrogenicity of the ‘sea of estrogens’ in which we all live. I have included this discussion on pesticides here because of their importance as EDC residues in food and because there is not a chapter in this book on the subject. Barbara Thomson, a recent PhD student of mine, determined the hierarchy of food-related EDC intake in the New Zealand diet (Thomson et al, 2003) and found that the top two EDCs are bisphenol A and genistein and that pesticides came a long way down the intake hierarchy. Clearly, if we are to tackle concerns about EDC effects on human consumers we should start with the highest intake substances if we are to achieve maximum effect. There is no doubt in my mind that we should reduce exposure to EDCs and that bisphenol A is the place to start.

References guillette lj, jr, gross ts, masson gr, matter jm, percival hf, woodward ar (1994) Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environ. Health Perspect., 102, 680–688. mccarthy ar, thomson bm, shaw ic, abell ad (2006) Estrogenicity of pyrethroid insecticide metabolites. J. Environ. Monit. 8, 197–202. shaw ic, chadwick j (1998) Principles of Environmental Toxicology, Taylor & Francis Ltd, London, p 53. thomson bm, cressey pj, shaw ic (2003) Dietary exposure to xenoestrogens in New Zealand. J. Environ. Monit. 5, 229–235. wppr (1998) Annual Report of the Working Party on Pesticide Residues, Ministry of Agriculture, Fisheries & Food, UK, p iii.

Index

17α-ethynylestradiol see ethynylestradiol 16α-LE2, 550 A-screen, 275 acetaminophen, 470 acetylcholinesterases-based enzymatic biosensors, 187 active substances, in pharmaceuticals availability and scope of risk assessment guidelines, 368 regulatory frameworks, 362 toxicity data requirements, 366 AHH assay, 279 AHRR polymorphism, 294–6 Aldomet, 472–3 alkyl phenols, 233 Allen–Doisy assay, 263 amenorrhoea, secondary, 462 androgen see reproductive hormones antibiotics, 494–7 defined daily dose consumption in the Netherlands, 495 doxycycline, 496 erythromycin, 496–7 minocycline, 496 sulfamethoxazole–trimethoprim combination, 494–6 tetracycline, 496 anticonvulsants, 473–5 carbamazepine, 473–4

defined daily dose consumption in the Netherlands, 473 phenobarbital, 474 phenytoin, 475 valproic acid, 474 antihypertensive drugs, 472–3 antipyretic drugs, 470 APRIN, 275 AR CALUX, 272 AR-EcoScreen, 298–9 Aroclor, 387 Aroclor 1254, 527 aromatase, 25 arylhydrocarbon-responsive element, 553–4 atenolol, 478–9 Atlantic croaker fish see Micropogonias undulates 17β-estradiol and bisphenol A, 407 β-galactosidase, 266, 274, 280 8β-VE2, 550 BADGE, 408 Basel human milk cohort, 531 3-BC see 3-benzylidene camphor Belgian dioxin crisis in 1999, 393–4 benzophenone-2, 521 benzophenone-3, 529 3-benzylidene camphor, 521

572

Index

beta-blockers, 477–9 atenolol, 478–9 defined daily dose consumption in the Netherlands, 478 metoprolol, 478 propranolol, 478 biomonitoring, 343 biorecognition, 186–9 biosensors see also specific biosensors for endocrine-disrupting chemicals, 183–200 schematic representation, 185 vs traditional methods, 186 general structure biocomponents and transducers for biosensor construction, 192 biorecognition, 186–9 interface techniques, 189–91 transducing elements, 191–3 whole-cell-based concept, 188 2,3-bis(4-hydroxyphenyl)-propionitrile) see DNP bisphenol A, 9–10, 197, 233–4, 406–27, 570 applications, 406 and 17β-estradiol, 407 effects assessments, 369–70 exposure selection, 60 future trends, 427 in humans, 414–15 exposure estimates based on food intake, 415 exposure estimates based on measurements, 414–15 metabolism and pharmacokinetics, 414 levels in canned food and beverages, 410–12 levels in foods and beverages in unknown/no containers, 413 levels in milk stored in polycarbonate plastic containers, 413 mechanism of action, 296–7, 416 migration into food and beverages, 407–13 products with food-contact uses, 408 resulting concentrations in food and beverages, 409, 413 as model compound, 357–8 positions of government bodies on human health risks, 423–7 Canada, 425–6

Europe, 424–5 Japan, 425 United States, 426–7 risks to human health, 416–23 carcinogenic and mutagenic effects, 421–2 health effects, 422–3 immunological effects, 420–1 neurological effects, 419–20 reproductive and developmental effects, 418–19 structure, 358 bisphenol A diglycidyl ether, 408 bone health effects of endocrine-disrupting chemicals, 85 methods to study the effects of endocrine-disrupting chemicals, 84–5 and nutritional phytoestrogens, 83–96 bacterial degradation resulting to biological transformation, 86 estradiol-17β, genistein, daidzein and equol effects on bone mineral density, 90 estradiol-17β, genistein, daidzein and equol effects on uterine weights, 88 estradiol-17β and isoflavones structure, 85 future trends, 96 isoflavones, 86–91 PCNA positivity effects in mamma structure from estradiol-17β, genistein, daidzein and equol, 89 selective estrogen receptor modulators, 84 postmenopause, 91, 96 soy/red clover/isoflavone doubleblind placebo controlled clinical studies, 92–5 trends, 83–4 bone mineral density estradiol-17β, genistein, daidzein and equol, 90 Bp-2 see benzophenone-2 Brachidanio rerio, 437, 439, 448 Brachydanio rerio, 548, 557 aromatase B gene, 548 breastfeeding current recommendations, 345–6 health-related issues, and EDCs in human milk, 322–47

Index brominated flame retardants, 395–9 analysis, 398 exposure, 398–9 Michigan incident, 395–6 structures, 396 toxicity, 396–8 lowest observable adverse effect level, 397 no observable adverse effect level, 397 brown trout see Salmo trutta fario cadmium, 46–7, 557–8 CAFLUX, 280 CALUX assay, 279–80, 398 cancer studies, 75–6 endometrial cancer, 76 prostate cancer, 76 testis cancer, 75–6 vaginal and cervical adenocarcinoma, 75 carbamazepine, 473–4 carcinogenicity, 366 Cashmeran, 528 Celestolide, 528 cell proliferation assays, 274–5 chemical body burden, 343 ChIP-chip array, 291, 302 chromatography coupled to mass spectrometry for endocrine-disrupting chemicals, 149–74 future trends, 171–4 coupling bioassays with mass spectrometry, 171–2 mass spectrometric fingerprinting, 172–4 clawed frogs see Xenopus laevis Clethrionomys glareolus, 448 clofibrate, 471 comparative molecular field analysis, 313 Computer Assisted Stereology Toolbox Grid System, 551 congenital malformations, 18 consumer exposure assessments and dietary intake calculations, 139–42 ‘at risk’ groups, 140–1 mixtures, 141–2 precautionary principle, 140 survey design, 129–31 sampling, 131–3, 135 EDCs in food sample collection and preparation, 134

573

recording sample details, 132 sample handling, 133 sample preparation, 133, 135 sample purchase, 131–2 survey design choice of sampling locations, 131 geographical coverage, 131 objectives, 130 origin of samples, 131 previous or existing knowledge, 131 sample size, 130–1 sample units, 130 timing, 131 contemporary use pesticides see nonpersistent pesticides Coreganus sp., 530 coumestans, 218 coumestrol, 443 cryptorchidism, 18–19, 295 CYP1A1, 554–5 cypermethrin, 569–70 3D-QSAR see 3D-quantitative structure-activity relationship 3D-quantitative structure-activity relationship, 312–14 calculation of field-based descriptors, 313 vs QSAR, 313–14 daidzein see also puerarin bacterial degradation resulting to biological transformation, 86 estradiol-17β, genistein, and equol effects on bone mineral density, 90 estradiol-17β, genistein, and equol effects on uterine weights, 88 structure, 85 DDT see dichlorodiphenyltrichloroethane decaBDE, 235 DEHP, 551 dehydroepiandrosterone, 274 di-(2-ethylhexyl) phthalate see DEHP dichlorodiphenyltrichloroethane, 296, 569 and other organochlorine pesticides, 338–9 diclofenac, 469 dietary intake calculations and consumer exposure assessments, 139–42

574

Index

‘at risk’ groups, 140–1 mixtures, 141–2 precautionary principle, 140 diethylstilbestrol, 297, 549 3,17-dihydroxy-19-nor-17α-pregna1,3,5(10)-triene-21,16α-lactone see 16α-LE2 dimethylbenzanthracene, 389 dioxin-responsive element, 553 dioxins, 60, 237–9, 383–95, 557 analytical methods, 392–4 assessing toxic effects, 388–92 effects in humans and animals, 388–90 toxic equivalents principle, 390–2 congener patterns, 386 current exposure, 394–5 limits in food and feed, 395 and dioxin-like polychlorobiphenyls, 384–8 incidents, 384–6 sources and production, 386–7 effects assessments, 371 as model compound, 358–9 structure, 358 structures of PCDD, PCDF, and PCB, 385 TEF values assigned by WHO, 391 DNP, 550 doxycycline, 496 DR CALUX, 261–2 E-screen, 266, 274, 275, 444, 449 EHMC see ethylhexyl-methycinnamate embryo/fetal toxicity, 366 endocrine-disrupting chemicals, 568–70, xix–xxi androgenic/anti-androgenic potency, 550–1, 553 diverse mechanisms and effects, 552 aryl hydrocarbon receptor domain structure, 553 and ER interaction at estrogenresponsive-element, 556 transcriptional regulation of target genes, 554 biosensors for, 183–200 chromatography coupled to mass spectrometry-related techniques, 149–74 advances in gas chromatography, 152–3

advances in liquid chromatography, 162–4 case studies, 153, 155–6, 158–60, 162, 164–5, 168–71 future trends, 171–4 contribution to total estrogenicity for average Western and Asian diets, 248 development of testing and assessment methods, 375–7 dietary exposure, 211–50 dietary exposure estimates bisphenol A, 233 coumestrol, 218 dioxins and dioxin-like PCBs, 240–1 flavonoids, 220–1 isoflavones genistein and daizein, 223–4 lignans enterolactone and enterdiol, 226 nonylphenol, 233 organochlorine pesticides, 229–30 PAH and benzo(a)pyrene, 245 phthalates, 236 polybrominated diphenyl ethers, 238 synthetic pyrethroid pesticides, 232 total polychlorinated biphenyls, 243 zearalenone, 227 dioxin-like potency, 553–7 discovery, 569 effect of exposure during critical periods of development, 15–16 effect on developing fetus, 3–27 effect on human development, 5–7 effect on the placenta, 23–5 placental hormones, 24–5 placental transporters, 24 xenobiotic and steroid metabolizing enzymes, 25 effects assessments, 369–72 bisphenol A, 369–70 comparison, 372 dioxins, 371 ethinyl estradiol, 371–2 results of hazard assessments, 373 toxicological principles and assumptions, 376 vinclozolin, 372 effects on bone health, 85 endometriosis and fibroids, 74

Index epidemiological evidence on impaired reproductive function and cancer, 58–77 estimate of total estrogenicity from food, 246, 247 and estrogen and androgen structures, 293 estrogenic/anti-estrogenic potency, 547–50 exposure assessment environmental contaminants, 237–9, 242 estrogenic EDC, 244–9 future trends, 249–50 industrial chemicals, 233–5, 237 methodologies, 214–15 mycotoxins, 225–6 pesticides, 227–8, 231 phytoestrogens, 218, 220, 222, 225 exposure to total estrogenicity, 215–18 blood levels of estrogenic EDC, 217–18 classical mechanism of estrogenresponsive effect, 216 cumulative, 217 estrogenic EDC, 215 relative estrogenic potency, 215–17 female reproductive dysfunction, 23 food surveillance, 126–43 dietary intake calculations and consumer exposure estimates, 139–42 environmental risk assessment vs dietary exposure estimates, 127–8 future trends, 142–3 monitoring time trends, 142 sampling, 131–3, 135 surveillance programmes, 135–6, 139 survey design, 129–31 four model compound, 357–60 bisphenol A, 357–8 dioxins, 358–9 ethinyl estradiol, 359 mechanism of action and main sources of consumer exposure, 358 vinclozolin, 359–60 future trends, 249–50 gene expression modulation through epigenetic mechanisms, 300–1

575

genetic variability in susceptibility, 294–6 AHRR polymorphism, 294–6 SNP, 294–6 human epidemiologic studies and altered hormone levels, 36–51 future trends, 50–1 human exposure, 11–16 hierarchy of xenoestrogen exposure, 11 pharmacologically relevant doses of xenoestrogens, 12–16 structures of 17β-estradiol and xenoestrogens, 12 in human milk and health-related issues of breastfeeding, 322–47 implication on food industry, 249 mechanisms of action, 541–59 future trends, 553–7 menstrual cycle disturbances, 73–4 metabolism, 7–10 2,3,7,8-tetrachloro-dibenz-dioxin, 9 bisphenol A, 9–10 detoxification, 7 fetal protection from maternal dietary contaminants, 8–9 placental metabolism, 9 metals, 46–9 cadmium, 46–7 lead, 47–8 manganese, chromium and other metals, 48–9 methods of studying the effects on bone health, 84–5 microarrays and related techniques for detection of effects, 296–7 monitoring bisphenol A, 197 detection techniques, 194 estrogen receptor biosensors, 197 pesticides, 193, 195–6 phenols, 199 polychlorinated compounds, 198 polycyclic aromatic hydrocarbons, 198 steroids, 196 surfactants, 198–9 tributylin, 199 natural sources mycotoxins, 108 natural steroid hormones, 104–6 phytoestrogens, 106–7 in non-genotoxic carcinogenesis, 569

576

Index

non-persistent pesticides, 42–4 association of testosterone level with increasing TCPY quintiles, 44 nuclear receptor family, 543, 544–7 DNA-binding domain, 543 evolutionary conserved domains, 544 ligand-binding domain, 543 transcriptional regulation of target genes, 546 occurrence in food or produced from food constituents, 213 organochlorines, 37–42 association with reproductive hormones, 38–9 thyroid hormones association with, 40–2 origins, fates and transmission into the food chain, 103–20 natural sources found in the food chain, 105 other emerging compounds, 49–50 bisphenol A, 49–50 brominated flame retardants, 49 perchlorate, 49 polybrominated diphenyl ethers, 49 PDE4D4 gene epigenetic regulation in prostrate tissue, 300 pharmacological effect, xx phthalates, 44–6 association of free T4 level with increasing MEHP quintiles, 45–6 association with infant hormone level, 45 di(2-ethylhexyl) phthalate [DEHP] exposure and endogenous hormone levels, 44–5 placental transfer of xenoestrogens, 25–6 regulatory frameworks, 360–3 active substances in pharmaceuticals, 362 comparison, 363 environmental pollutants in food, 361–2 existing active substances in plant protection products, 362–3 main components for the four model compounds, 361 priority existing industrial substances, 360–1

reproductive abnormalities at birth, 62, 71 risk assessment, 356–78 availability and scope of guidelines, 367–9 issued guidelines to the risk assessor, 370 risk-benefit equation, 568 role of genetics, epigenetics and genomic technologies, 291–302 future trends, 301–2 screening methods for detection in food and environment AR-EcoScreen, 298–9 human androgen receptor assay, 298 receptor binding and reporter gene assays, 298 in vivo chromatin immunoprecipitation assay, 298 selection, 213–14 semen quality, 71–2 signaling through AhR, 299 in silico method of potency prediction, 306–18 3D-QSAR, 312–14 future trends, 317–18 QSAR, 307–12 results and implications, 315–17 virtual docking, 314–15 from soy, 568–9 structural diversity, 212 structures 16 EPA priority pollutant PAH, 244 different flavonoid groups, 219 dioxins and dibenzofurans, 239 organochlorine pesticides, 228 phthalate esters, 234–5 polybrominated diphenyl ethers, 237 polychlorinated biphenyls, 241 principal lignans found in food, 225 synthetic pyrethroids, 231 synthetic sources food additive chemicals, 116 food packaging/contact chemicals, 116 industrial chemicals, 114–15 pesticides, 112, 114 pharmaceuticals, 108, 111 suspected endocrine-disrupting synthetic chemicals, 109–10 veterinary medicines, 111–12

Index time to pregnancy, 75 toxicity data requirements, 363–7 active substances in pharmaceuticals, 366 comparison, 366–7 environmental pollutants in food, 363 existing active substances in plant protection products, 366 from the manufacturer for risk assessment, 364–5 priority existing industrial substances, 363 toxicological assumptions and principles, 372, 374–5 comparison, 375 dose–response relationship, 374–5 mode of action, 374 transmission into the food chain, 116–20 endocrine system fetal development, 16 endometrial hyperplasia, 87 endometrial prostaglandinendoperoxide synthase, 297 endometriosis, and fibroids, 74 environmental contaminants, 237–9 dioxins, 237–9 polychlorinated biphenyls, 239, 242 polycyclic aromatic hydrocarbons, 242 ‘environmental pollutants in food’ availability and scope of risk assessment guidelines, 368 regulatory frameworks, 361–2 toxicity data requirements, 363 environmental risk assessment vs dietary exposure estimates, 127–8 food surveillance, 128 risk assessment process, 128 equol, 441 estradiol-17β, genistein, daidzein and effects on bone mineral density, 90 estradiol-17β, genistein, daidzein and effects on uterine weights, 88 ER CALUX, 267 ERα CALUX, 270–1 EROD assay, 279 erythromycin, 496–7 estradiol-17β in bone health, 83–4 genistein, daidzein and equol effects on bone mineral density, 90

577

genistein, daidzein and equol effects on uterine weights, 88 and isoflavones structure, 85 estrogen see reproductive hormones estrogen equivalency, 440 estrogen mimics, 4 estrogen receptor biosensors, 197 estrogen receptors, 543, 545–7 estrogen-responsive element, 545 estrogenicity, 297, 440, 486–94 ethinyl estradiol effects assessments, 371–2 as model compound, 359 structure, 359 ethylhexyl-methycinnamate, 521 EU-FIRE project, 398 EU-FP6 project ‘Food and Fecundity’ pharmaceutical products affecting human fecundity and their mechanism of action, 467 antibiotics, 494–7 anticonvulsants, 473–5 antihypertensive drugs, 472–3 antipyretic drugs, 470 beta-blockers, 477–9 nonsteroidal anti-inflammatory drugs, 467–70 peroxisome proliferators, 471–2 risk assessment, 497–503 serotonin reuptake inhibitors, 475–7 steroid contraceptives, 479, 481–94 European risk assessment, 369, 370 exposures bisphenol A, 60 persistent organochlorine, 59–60 dichlorodiphenyltrichloroethane, 59 dioxins, 60 hexachlorobenzene, 59 polychlorinated biphenyls, 59 phthalates, 60 eyed grayling see Thymallus thymallus female sexual dysfunction, 463 fetal development effect of endocrine-disrupting chemicals, 3–27 endocrine system, 16 FEtal Gonad Assay, 550, 553 fetal loss, 20 field voles see Clethrionomys glareolus FireMaster FF-1, 396

578

Index

flavonoids, 218, 220, 222 see also isoflavones fluoxetine hydrochloride, 475–6 flutamide, 374 fluvoxamine maleate, 476–7 food endocrine-disrupting chemicals from, 211–50 food additive chemicals, 116 food packaging/contact chemicals, 116 Fusarium spp., 437 Galaxolide, 528 gas chromatography-mass spectrometry advances, 152–3 benefit for EDCs analysis, 152–3 main application, 152 case studies, 153, 155–6, 158–60, 162 diagnostic ion chromatograms, 157 GC x GC-TOFMS chromatogram, 161 halogenated persistent organic pollutants, 158–60, 162 steroid hormones, 153, 155–6, 158 synthesis and mass spectrometric behaviour of pentafluorobenzylbromide, trimethylsilylated estradiol derivative, 154 gemfibrozil, 471–2 genistein, 549, 570 estradiol-17β, daidzein and equol effects on bone mineral density, 90 estradiol-17β, daidzein and equol effects on uterine weights, 88 gibberellic acid, 4 GPR30, 559 halogenated persistent organic pollutants, 158–60, 162 Hansch’s paradigm Hershberger assay, 264, 377 hexabromocyclododecane, 395 HHCB-lactone, 528 hormones see specific hormones H295R steroidogenesis assay, 278 human milk, 333 contaminants with suspected endocrine activity, 327–32 DDT total mean concentrations, 334 dl-polychlorinated biphenyls, 336 EDCs in, and health-related issues of breastfeeding, 322–47

nutritional phytoestrogens, 326, 333 PBDE total mean concentrations, 338 polychlorinated biphenyls, 337 polychlorodibenzdioxins/furans, 335 recommendations on breastfeeding, 345–6 time trend of persistent organochlorine pollutants, 324 xenobiotics assessment of exposure, 343–4 DDT and other organochlorine pesticides, 338–9 factors influencing transfer of xenobiotics, 325 factors that influence passive diffusion rate and transfer, 324–5 non-persistent pesticides, 340–1 persistent perfluorinated chemicals, 342 pharmaceuticals and personal care products, 341–2 phenol compounds, 342–3 phthalates, 341 polybrominated diphenyl ethers, 340 polychlorinated biphenyls, 339–40 polychlorinated dioxins and furans, 340 range and distribution, 333–4, 338–43 risk assessment, 344–5 transmission into milk, 324–6 20-hydroxyecdysone, 294 hypospadias, 19–20, 295 ibuprofen, 468–9 ICI-164,384, 550 indomethacin, 469–70 industrial chemicals, 114–15, 233–7 alkyl phenols, 233 biomagnification of DDT, 114 bisphenol A, 233–4 phthalates, 234–5 polybrominated diphenyl ethers, 235, 237 industrial substances availability and scope of risk assessment guidelines, 367–8 regulatory frameworks, 360–1 toxicity data requirements, 363 infertility, 461–2 initiator–promoter carcinogenesis hypothesis, 569

Index interface techniques, 189–91 immobilization, 189–91 absorption, 189–90 encapsulation, 190 entrapment, 190–1 Langmuir-Blodgett, 191 isoflavones, 86–91, 222, 326, 333, 440–3 endometrial hyperplasia from, 87 Kaneclor, 387 Langmuir-Blodgett, 191 lead, 47–8 libido, altered, 464 lignans, 222, 225, 443–4 liquid chromatography-mass spectrometry advances, 162–4 benefit for EDCs analysis, 163–4 main application, 162–3 case studies, 164–5, 168–71 diagnostic ion chromatograms for phytoestrogen compounds, 167 entropic contaminants, 169–71 naturally occurring contaminants, 164–5, 168–9 phytoestrogen fragmentation pathway, 166 UPLC-MS/MS chromatogram of apple crude extract, 170 liver X receptor, 558–9 low-dose hypothesis, 417, 418, 427 luciferase, 266, 274, 280 Luvox, 476–7 male sexual dysfunction, 462–3 manganese, chromium and other metals, 48–9 4-MBC see 4-methylbenzylidene camphor medaka fish see Oryzias latipes medial preoptic region, 526 developmental exposure effects of 4MBC and 3-BC, 526 medroxyprogesterone 17-acetate, 262 MEHP see mono-(2-ethylhexyl) phthalate menstrual cycle disturbances, 73–4 metabolomics, 172–4 metabolonomics, 296 metals, 46–9 exposure and hormone levels cadmium, 46–7 lead, 47–8

579

manganese, chromium and other metals, 48–9 4-methylbenzylidene camphor, 521 methylcholanthrene, 555 methyldopa, 472–3 metoprolol, 478 micropenis, 294–5 Micropogonias undulates, 444 Microtus agrestis, 439 minocycline, 496 MMV-Luc assay, 267 molecular interaction fields, 313 mono-(2-ethylhexyl) phthalate, 550 association with free T4 level, 46 morning-after pills, 371 musk fragrances, 341–2 nitro musk musk ketone, 341 musk xylene, 341 polycyclic musk Celestolide, 342 Galaxolide, 342 Phantolide, 342 Tonalide, 342 Traesolide, 342 MVLN assay, 267 mycotoxins, 108, 225–6 zearalenone, parabens and hydroxylated butanol molecular structure, 107 Myzus persicae, 294 naproxen, 469 natural steroid hormones, 104–6 non-persistent pesticides, 42–4 association of testosterone level with increasing TCPY quintiles, 44 epidemiological studies, 42–3 organophosphates, 42 urinary and serum biomarkers, 43 non-steroidal anti-inflammatory drugs, 467–70 consumption in Netherlands in defined daily dose, 468 diclofenac, 469 ibuprofen, 468–9 indomethacin, 469–70 naproxen, 469 norfluoxetine, 476 nuclear receptor corepressor protein, 525 nuclear receptor family, 543, 544–7 coactivators p160 family, 544

580

Index

functional domains DNA-binding domain, 543 ligand-binding domain, 543 members androgen receptor, 543 glucocorticoid receptor, 543 progesterone receptor, 543 retinoid acid, 543 thyroid hormone receptor, 543 vitamin D receptor, 543 Nutrimaster, 395–6 nutritional phytoestrogens bacterial degradation resulting to biological transformation, 86 and bone health, 83–96 estradiol-17β, genistein, daidzein and equol effects on bone mineral density, 90 estradiol-17β and isoflavones structure, 85 PCNA positivity effects in mamma structure from estradiol-17β, genistein, daidzein and equol, 89 octaBDE, 235 octocrylene, 529 octylmethoxycinnamate, 521 octyltriazone, 529 OHT see tamoxifen OMC see octylmethoxycinnamate organochlorine, 227–8 Oryzias latipes, 442 osteoporosis, 87 PAH see polycyclic aromatic hydrocarbons paracetamol see acetaminophen partial least squares, 313 PBDE see polybrominated diphenylethers PCB see polychlorobiphenyls PCDD see polychlorinated dibenzo-p-dioxins PCDF see polychlorinated dibenzofurans PDE4D4 gene, 300 peach-potato aphid see Myzus persicae pentaBDE, 235 perfluorinated chemicals, persistent, 342 perfluorononanoic acid, 342 perfluorooctanesulfonamide, 342 perfluorooctanesulfonate, 342 perfluorooctanoic acid, 342

peroxisome proliferator-activated receptor, 558 peroxisome proliferators, 471–2 clofibrate, 471 defined daily dose consumption in the Netherlands, 471 gemfibrozil, 471–2 persistent organic compound, 118 persistent organic pollutants, 323 persistent organochlorine, 37–42 exposure selection, 59–60 organochlorine pesticides, 37–8 dichlorodiphenyltrichloroethane, 38 hexachlorobenzene, 38 polychlorinated biphenyls, 37–8 persistent organohalogen pollutants, 295 pesticides, 112, 114, 193, 227–32, 570 food chain contamination by pesticides and other EDCs, 113 organochlorine, 227–8 synthetic pyrethroids, 228, 231 pesticides, non-persistent, 340–1 Phantolide, 528 pharmaceutical products, 459–504, 467, 503 affecting human fecundity and their mechanism of action, 467 antibiotics, 494–7 anticonvulsants, 473–5 antihypertensive drugs, 472–3 antipyretic drugs, 470 beta-blockers, 477–9 classification of mechanisms that affect fecundity, 461–4 altered libido, 464 female sexual dysfunction, 463 infertility, 461–2 male sexual dysfunction, 462–3 sexual dysfunction, 462 defined daily dose consumption in Netherlands antibiotics, 495 anticonvulsants, 473 beta blockers, 478 estrogens with progestogens, 486 lipid regulators, 471 NSAIDS, 468 exposure pathways in food, 464–7 main pathways ending in human food chain, 465 non-steroidal inflammatory drugs, 467–70

Index peroxisome proliferators, 471–2 risk assessment, 497–503 dose-response assessment, 498–500 exposure assessment, 500–1 quantitative structure–activity relationships, 502–3 risk characterization, 501 serotonin reuptake inhibitors, 475–7 steroid contraceptives, 479, 481–94 available combination oral contraceptives, 480 contraceptive progestins, 479 contraceptive progestins family tree, 481 physicochemical properties and environmental fate data, 490 pharmaceuticals and personal care products, 341–2 phenobarbital, 474 phenols, 199, 342–3 phenytoin, 475 phthalates, 234, 341 exposure selection di(2-ethylhexyl) phthalate [DEHP], 60 phytoestrogens, 106–7, 218–25 adverse effects determination, 440–6, 448–9 comparative genotoxicity, 447 coumestans, 218 coumestrol, 106 dietary intake assessment, 449–51 estimated human dietary intakes, 450 flavonoids, 218, 220, 222 future trends, 452–3 human maximum tolerable daily intake, 451 isoflavones, 106, 222 lignans, 106, 222, 225 major dietary and key metabolites, 438 and phytosterols, 437–53 in food and endocrine disruption, 437, 439 risk management, 452 risks and benefits assessment, 451–2 secoisolariciresinol diglucoside, 439 uterine growth effects, 447 phytosterols, 445–6, 448 adverse effects determination, 440–6, 448–9 dietary intake assessment, 449–51 future trends, 452–3 major dietary, 446

581

and phytoestrogens, 437–53 in food and endocrine disruption, 437, 439 risk and benefits assessment, 451–2 risk management, 452 placental fetal barrier, 8–9 placental metabolism, 9 plant protection product availability and scope of risk assessment guidelines, 368 regulatory frameworks, 362–3 toxicity data requirements, 366 polybrominated biphenyls, 395 polybrominated diphenylethers, 235, 237, 340, 395 polychlorinated biphenyls, 239, 242, 339–40, 383–95 assessing toxic effects, 388–92 effects in humans and animals, 388–90 TEQ principle, 390–2 current exposure, 394–5 limits in food and feed, 395 dioxin-like, and dioxins, 384–8 incidents, 384–6 sources and production, 387–8 structures of PCDD, PCDF, and PCB, 385 TEF values assigned by WHO, 391 polychlorinated compounds, 198 polychlorinated dibenzo-p-dioxins and polychlorobiphenyls and polychlorinated dibenzofurans, 385 polychlorinated dibenzodioxins see dioxins polychlorinated dibenzofurans and polychlorobiphenyls and polychlorinated dibenzo-pdioxins, 385 polychlorinated dioxins and furans, 340 polychlorobiphenyls and polychlorinated dibenzofurans and polychlorinated dibenzo-pdioxins, 385 polycyclic aromatic hydrocarbons, 198, 242 polymerase chain reaction, 297 postmenopause, 91, 96 soy/red clover/isoflavone doubleblind placebo controlled clinical studies, 92–5 PPT, 550 precautionary principle, 140

582

Index

pregnancy, 75 preterm birth, 20–1 principal component analysis, 310 progesterone, 24–5 progestins, 479–84 prohormones, 263, 274, 281 propranolol, 478 4-propyl-pyrazole-1,3,5-triyl-triphenol see PPT Prozac, 475–6 puerarin, 90–1 pyrethroid insecticide, 569–70 QSAR see quantitative structureactivity relationship quantitative structure-activity relationship, 307–12 major classes of descriptor types, 309 modeling step, 308–11 molecular descriptors, 308 schematic view of the process, 310 validation procedures, 311–12 cross- validation, 312 external validation, 312 graphical representation of overfitting problem, 312 internal validation, 311, 312 vs 3D-QSAR, 313–14 vs virtual docking, 315 raloxifene, 550 receptor binding assays, 265–6 Registration, Evaluation, Authorization of Chemicals, 357 reproductive abnormalities at birth cryptorchidism, 62 feminisation, 71 hypospadias, 62 reproductive hormones Ah-receptor assays, 275–7 association with organochlorines, 38–9 estrogen, 39 follicle stimulating hormone, 39 luteinizing hormone, 39 sex hormone binding globulin, 38 testosterone, 38–9 bioassays for detection of activities, 259–82 classical mechanism of estrogen responsive effect, 265

effect of anti-androgen flutamide, and some brominated flame retardants on 5α-DHT response, 274 estrogenic compounds response in yeast cells, 268 future trends, 281–2 other hormonal bioassays, 280–1 potential influence of metabolism of any EDC, 276 practical application, 277 pure anti-estrogen ICI 182780 and RU58668 response, 270 spectrum of bioassays, 261 yeast androgen bioassay based on yEGFP expression, 273 compounds with hormonal activity, 262–3 in vitro bioassays, 264–77 for androgenic activities, 272 cell proliferation assays, 274–5 ER CALUX, 267 for estrogenic activities, 267 indirect effects on endogenous hormone levels, 278 MMV-Luc assay, 267 MVLN assay, 267 receptor binding assays, 265–6 test chemicals metabolism, 275–7 transcription activation assays, 266–74 in vivo bioassays, 263–4 Allen–Doisy assay, 263 Hershberger assay, 264 retinoid X receptor, 558 roach see Rutilus rutilus RU486, 550 Rutilus rutilus, 530 Saccharomyces cerevisiae, 277 ‘safety assessment,’ 362 Salmo trutta fario, 560 selective ER modulators, 545, 548, 549 selective estrogen receptor modulator, 84, 440 semen quality, 71–2 serotonin reuptake inhibitors, 475–7 fluoxetine hydrochloride, 475–6 fluvoxamine maleate, 476–7 sertraline, 477 sertraline, 477 sex hormone binding globulin, 440 sexual dysfunction, 462

Index Silastic capsule implants, 422 Simplified Molecular Input Line Entry System, 308 single nucleotide polymorphism, 294, 295 SMILES, 308 SNP see single nucleotide polymorphism steroid and xenobiotic receptors, 559 steroid contraceptives, 479–94 available combination oral contraceptives, 480 defined daily doses consumption of estrogens with progestogens, 486 ethynylestradiol, 486–94 effects to fecundity, 487–9 evidence for presence in the environment, 492–4 exposure routes, 491 persistence in the environment, 489–91 production volumes and use, 489 physicochemical properties and environmental fate data, 490 progestins, 479–84 adverse effects, 484 bioavailability, 482 biologically active forms, 481–2 changes in biochemical markers of androgenicity, 484 dose/ovulation inhibition dependence, 483 effect on sex hormone binding globulin/testosterone, 483 effects on androgens, 483 ethynodiol diacetate, 484 family tree, 481 levonorgestrel, 485 norethindrone, norethindrone acetate, 484–5 and oral contraceptives, 483 pharmacological effects in oral contraceptives, 479, 481 plasma levels, 482 relative binding affinities, 482–3 serum half-lives, 482 steroid hormones, 153, 155–6, 158 steroid receptor coactivator-1, 525 steroids, 196 sulfamethoxazole–trimethoprim combination, 494–6 sun protection factors, 527 surfactants, 198–9

583

surveillance programmes, 135–6 current surveillance efforts and findings, 135–6, 139 exposure source and level of EDCs with regulation levels, 137–8 result interpretation and follow-up action, 139 synthetic pyrethroids, 228, 231 tamoxifen, 549–50 TCPY see 3,5,6-trichloro-2-pyridinol Technical Guidance Document, 367 TEF see toxic equivalency factor TEQ see toxic equivalents principle tetrabromobisphenol A, 395 2,3,7,8-tetrachloro-dibenz-dioxin, 9 tetracycline, 496 tetrahydrogestrinone, 274 Thymallus thymallus, 446, 448 thyroid binding globulin, 389–90 thyroid hormones in relation to organochlorine exposure, 40–2 thyroid-stimulating hormone, 40–1 thyroid receptor, 558 Tonalide, 528 total diet study, 129 toxic equivalency factor values assigned by WHO, 391 toxic equivalents principle, 390–2 transcription activation assays, 266–74 androgens, 271–4 estrogens, 266–8 yeast and mammalian-cell-based, specific properties for estrogens, 268–71 transducing elements, 191–3 Traseolide, 528 tributylin, 199 3,5,6-trichloro-2-pyridinol, 43–4 association with testosterone level, 44 tryptophan, 555 tundra voles see Microtus agrestis TUNEL analysis, 445 ultraviolet filters chemicals used, 519–20 considerations of human risk, 533–5 levels and developmental toxicity in rat model, 534–5

584

Index

neonatal sensitivity, 533–4 persistent organic pollutants, 534 developmental toxicity and endocrine activity, 525–7 developmental toxicity of 4-MBC and 3-BC, 521–2, 525–7 effects on selected end points in rat offspring, 523–4 estrogen sensitivity and steroid receptor coregulators, 525 female sexual behavior and sexually dimorphic gene expression in brain, 525–7 gene expression in reproductive organs and brain, 526 low-dose effects in neonatal rat offspring, 527 pregnancy and early postnatal period, 522 prostate and uterus estrogen target gene expression, 525 reproductive organs of adult offspring, 522 sexual development, 522 endocrine-active, and cosmetics, 519–35 endocrine activity, 520–1 and fragrances in food chain: fish, 529–30 and fragrances in human milk, 530–1, 533 Basel human milk cohort, 531 presence of xenobiotics, 531, 533 selected UV filters, synthetic fragrances, phthalates and organochlorines, 532 presence in sewage sludge and surface waters, 527–9 risk assessment, 535 risk management, 535 in Swiss lake water and in fish from lakes and rivers, 529 and synthetic perfumes concentration in Swiss sewage sludge, 528 urine biomonitoring, 414, 416, 418, 423 uterotrophic assays, 377

uterus estradiol-17β, genistein, daidzein and equol, 88 valproic acid, 474 vasotocin neurons, 442 ventromedial hypothalamic nucleus developmental exposure effects of 4MBC and 3-BC, 526–7 veterinary medicines, 111–12 bovine growth hormone, 112 diethylstilbestrol, 111–12 vinclozin, 301 vinclozolin, 558, 569 effects assessments, 372 as model compound, 359–40 structure, 360 8-vinylestra-1,3,5(10)-triene-1,17β-diol see 8β-VE2 virtual docking, 314–15 protein and ligand mutual recognition process, 314 vs QSAR, 315 white fish see Coreganus sp. xenobiotic-responsive element, 553 xenoestrogens, 548, 557 bimodal dose-response effect, 13 endocrine disruptors, 11 pharmacologically relevant doses high-dose effect, 14–16 low-dose effect, 12–14 placental transfer, 25–6 Xenopus laevis, 448 Yasmin, 369, 371 yeast-enhanced green fluorescent protein, 266 yeast estrogen bioassay, 266 yeast estrogen screen, 266, 441, 570 zearalenone, 444–5 zebrafish see Brachidanio rerio; Brachydanio rerio Zoloft, 467