Microbial toxins in dairy products 9781118823095

Table of contents :
Content: List of Contributors xi Preface to the Technical Series xv Preface xvi 1 Microbial Toxins An Overview 1 R. Early and A.Y. Tamime 1.1 Introduction 1 1.2 Microbial toxins: modes of action 2 1.3 Bacterial toxins 4 1.3.1 Staphylococcal enterotoxins (SEs) 4 1.3.2 Bacillus cereus group enterotoxins 6 1.3.3 Clostridium botulinum nerotoxin 6 1.4 Mycotoxins 7 1.4.1 Background 7 1.4.2 General aspects of mycotoxins 7 1.4.3 Postscript on mycotoxins 13 1.5 Biogenic amines (BAs) 14 1.6 Conclusions 14 References 16 2 Incidences of Mould and Bacterial Toxins in Dairy Products 19 M.L.Y. Wan, N.P. Shah and H.I. El -Nezami 2.1 Background 19 2.2 Bacterial toxins 19 2.2.1 Emetic toxin produced by Bacillus cereus 20 2.2.2 Enterotoxins produced by Staphylococcus aureus 24 2.2.3 Botulinum neurotoxins produced by Clostridium botulinum 26 2.3 Mould toxins (mycotoxins) 33 2.3.1 Aflatoxins B1 and M1 34 2.3.2 Sterigmatocystin 42 2.3.3 Ochratoxin A 43 2.4 Other mycotoxins 47 2.5 Conclusion 49 References 50 3 Bacterial Toxins Structure, Properties and Mode of Action 71 J.W. Austin 3.1 Background 71 3.2 Bacillus cereus toxins 72 3.2.1 Bacillus cereus emetic toxin 73 3.2.2 Bacillus cereus enterotoxins 74 3.2.3 Bacillus cereus haemolysin BL (Hbl) 76 3.2.4 Bacillus cereus non -haemolytic enterotoxin (Nhe) 76 3.2.5 Cytotoxin K (Cyt K) 77 3.3 Botulinum neurotoxin 77 3.3.1 Outbreaks of botulism caused by dairy products 78 3.3.2 Structure of botulinum neurotoxin 79 3.3.3 Mode of action of BoNTs 80 3.4 Staphylococcus aureus enterotoxin 80 3.5 Conclusions 83 References 83 4 Biogenic Amines in Dairy Products 94 V. Ladero, D.M. Linares, M. Perez, B. del Rio, M. Fernandez and M.A. Alvarez 4.1 Introduction 94 4.2 Biochemistry: biosynthesis pathways, enzymes and transporters 98 4.2.1 Tyramine 98 4.2.2 Histamine 99 4.2.3 Putrescine 100 4.2.4 Cadaverine, -phenylethylamine, tryptamine 101 4.3 Biogenic amine -producing micro -organisms 101 4.3.1 Genes involved in the biosynthesis of biogenic amines 103 4.3.2 Is the production of BAs a strain- or species-dependent characteristic? 105 4.3.3 Physiological functions of BAs biosynthesis 106 4.4 Toxicological effects 107 4.4.1 Tyramine 108 4.4.2 Histamine 109 4.4.3 Putrescine and polyamines 110 4.4.4 Cadaverine, tryptamine and -phenylethylamine 111 4.4.5 Recommended limits of BAs 111 4.5 Factors affecting BAs accumulation in dairy products 113 4.5.1 Presence of BAs-producing bacteria 113 4.5.2 Physiochemical factors 114 4.5.3 Technological factors 117 4.6 Other preventive methods 119 4.7 Conclusions 119 Acknowledgements 120 References 120 5 Contamination of Raw Milk: Sources and Routes Up to the Farm Gate 132 R. Early 5.1 Introduction 132 5.2 The concept of contamination 132 5.2.1 What does contamination mean to the concept of food? 134 5.2.2 Contamination and cow health 135 5.3 Sources of contamination 136 5.3.1 Biological contamination 136 5.3.2 Chemical contamination 142 5.3.3 Mycotoxins 147 5.3.4 Physical contamination 148 5.4 Conclusion 150 References 150 6 Milk Product Contamination After the Farm Gate 154 R. Early 6.1 Introduction 154 6.2 The significance of microbial contamination 154 6.2.1 Product spoilage 154 6.2.2 Food-borne illness 155 6.2.3 Microbial toxins 155 6.3 Factories, processes and people 157 6.4 Raw milk handling 158 6.5 Milk -processing and dairy products manufacture 160 6.5.1 Liquid milk and cream processing 161 6.5.2 Packing, storage, distribution and the retail environment 162 6.6 Buttermaking 164 6.7 Cheesemaking 165 6.8 Yoghurt 168 6.9 Milk powders 170 6.10 Evaporated milk and sweetened condensed milk 172 6.10.1 Evaporated milk 172 6.10.2 Sweetened condensed milk 175 6.11 Ice-cream 175 6.12 Hygiene, food safety management and cleaning -in -place (CIP) 177 6.13 Packaging, storage, distribution and the retail environment 178 6.14 Conclusions 180 References 180 7 Techniques for Detection, Quantification and Control of Bacterial Toxins 183 L. Ramchandran, A. Warnakulasuriya, O. Donkor and T. Vasiljevic 7.1 Introduction 183 7.2 Bacterial toxins 184 7.3 Control of toxins 186 7.4 Methods for identification and detection of microbial toxins 187 7.4.1 Traditional biological assays 189 7.4.2 Antibody and immunoassay 191 7.5 Conclusion 196 References 196 8 Techniques for Detection, Quantification and Control of Mycotoxins in Dairy Products 201 O. Donkor, L. Ramchandran and T. Vasiljevic 8.1 Introduction 201 8.2 Methods for detection and quantification of mycotoxins 203 8.2.1 Sample pre-treatment method 203 8.2.2 Liquid-liquid extraction 204 8.2.3 Supercritical fluid extraction 204 8.2.4 Solid phase extraction 204 8.3 Separation methods 206 8.3.1 Thin layer chromatography 206 8.3.2 High pressure liquid chromatography (HPLC) 208 8.3.3 Gas chromatography (GC) 210 8.3.4 Capillary electrophoresis (CE) 210 8.3.5 Biosensors 210 8.3.6 Enzyme-linked immmunosorbent assay (ELISA) method 212 8.3.7 Electrochemical immunoassay 214 8.3.8 Polymerase chain reaction (PCR)-based detection and quantification 215 8.4 Mathematical model (exposure assessment of mycotoxins in dairy milk) 216 8.5 Control of mycotoxin 217 8.5.1 Physical methods 218 8.5.2 Chemical methods 218 8.5.3 Biological methods 219 8.5.4 Activated carbon (AC) 219 8.6 Conclusion 220 References 220 9 Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production 229 N. Murru, R. Mercogliano, M. -L. Cortesi, F. Leroy, R. Condoleo and M.F. Peruzy 9.1 Background on risk analysis 229 9.2 Focus on cheese risk assessment 231 9.2.1 Source of milk 231 9.2.2 Raw and/or heat-treated milk cheeses 231 9.2.3 Level of moisture in cheese 232 9.2.4 Methods of manufacture 232 9.2.5 Fat content 232 9.2.6 Maturation indices 232 9.2.7 Washed or mould cheeses 233 9.3 Staphylococcus aureus 233 9.3.1 Background 233 9.3.2 Staphylococcus aureus and production of staphylococcal eneterotoxins 234 9.3.3 Cheese production and hazard characterisation of Staphylococcus aureus 237 9.3.4 Cheesemaking conditions and exposure assessment of Staphylococcus aureus 238 9.3.5 Predictive modelling and risk assessment of Staphylococcus aureus and enterotoxin production in cheese 242 9.4 Escherichia coli 243 9.4.1 Hazard identification 243 9.4.2 Growth and inactivation 244 9.4.3 Hazard characterisation 244 9.4.4 Exposure assessment 246 9.4.5 Risk characterisation 248 9.5 Listeria monocytogenes 251 9.5.1 Hazard characterisation 251 9.5.2 Exposure assessment 253 9.5.3 Hazard characterisation 256 9.5.4 Risk characterisation 258 9.6 Cheese - chemical risk assessment 259 9.6.1 Background 259 9.6.2 Biogenic amines in cheese 260 9.6.3 Occurrence of biogenic amines in cheese: hazard and exposure assessment 260 9.6.4 Method for controlling biogenic amines in food 263 9.7 Modelling of growth and inactivation: kinetic approaches 264 9.7.1 Model categories: an overview 264 9.7.2 Growth - no growth interface: a probabilistic approach 267 9.7.3 Modelling of toxin production 268 9.8 Conclusions 268 References 269 10 Regulatory Measures for Microbial Toxins 287 M. Hickey 10.1 Introduction and background 287 10.2 The evolution and economic significance of heat -treated milk and milk products 288 10.2.1 Fluid milk 289 10.2.2 Evaporated and sweetened condensed milks 289 10.2.3 Milk and dairy powders 290 10.2.4 Cheeses (natural and processed) and fermented milks 291 10.3 Bacterial toxins 292 10.3.1 Staphylococcal enterotoxins 292 10.3.2 Bacillus cereus toxins 293 10.3.3 Clostridium botulinum toxins 294 10.4 Regulatory provisions on bacterial toxins in milk and milk products 297 10.4.1 European regulations on food hygiene and food safety 297 10.4.2 US milk hygiene and food safety standards 303 10.4.3 International perspective on food hygiene and safety Codex Alimentarius 307 10.5 Mycotoxins 310 10.5.1 Aflatoxins 311 10.5.2 Other mycotoxins 311 10.5.3 Aflatoxins M1 and B1 and their regulatory provisions 312 10.5.4 EU legislations on aflatoxins in milk, milk products and animal feed 313 10.5.5 Regulations of aflatoxins of importance in milk and milk products in the USA 314 10.5.6 Regulations of aflatoxins of importance in milk and milk products in Canada 314 10.5.7 Regulation of aflatoxins in Australia and New Zealand 315 10.5.8 MERCOSUR standard on aflatoxins 315 10.6 Conclusions 316 References 316 Index 323

Citation preview

Microbial Toxins in Dairy Products

Society of Dairy Technology Series Series Editor(s): A.Y. Tamime. The Society of Dairy Technology has joined with Wiley‐Blackwell to produce a series of technical dairy‐related handbooks providing an invaluable resource for all those ­ involved in the dairy industry; from practitioners to technologists working in both traditional and modern large‐scale dairy operations. Biofilms in the Dairy Industry by Koon Hoong Teh, Steve Flint, John Brooks, and Geoff Knight (Editors) Milk and Dairy Products as Functional Foods by Ara Kanekanian (Editor) Membrane Processing: Dairy and Beverage Applications by Adnan Y. Tamime (Editor) Processed Cheese and Analogues by Adnan Y. Tamime (Editor) Technology of Cheesemaking, 2nd Edition by Barry A. Law and Adnan Y. Tamime (Editors) Dairy Fats and Related Products by Adnan Y. Tamime (Editor) Dairy Powders and Concentrated Products by Adnan Y. Tamime (Editor) Milk Processing and Quality Management by Adnan Y. Tamime (Editor) Cleaning‐in‐Place: Dairy, Food and Beverage Operations, 3rd Edition by Adnan Y. Tamime (Editor) Structure of Dairy Products by Adnan Y. Tamime (Editor) Brined Cheeses by Adnan Y. Tamime (Editor) Fermented Milks by Adnan Y. Tamime (Editor) Probiotic Dairy Products by Adnan Y. Tamime (Editor)

Microbial Toxins in Dairy Products Edited by

Adnan Y. Tamime Consultant in Dairy Science and Technology, Ayr, UK

This edition first published 2017 © 2017 by John Wiley & Sons Ltd Registered Office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Offices 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 111 River Street, Hoboken, NJ 07030‐5774, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley‐blackwell. The right of Adnan Y. Tamime to be identified as the author of the editorial material in this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty: While the publisher and author(s) have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging‐in‐Publication Data Names: Tamime, A. Y., editor. Title: Microbial toxins in dairy products / edited by Adnan Y. Tamime. Other titles: Society of Dairy Technology series. Description: Chichester, West Sussex, UK ; Hoboken, NJ : John Wiley & Sons, Inc., 2017. |   Series: Society of Dairy Technology series | Includes bibliographical references and index. Identifiers: LCCN 2016032565 (print) | LCCN 2016033156 (ebook) | ISBN 9781118756430 (cloth) |   ISBN 9781118823651 (pdf) | ISBN 9781118823149 (epub) Subjects: | MESH: Dairy Products–microbiology | Bacterial Toxins–toxicity | Dairy Products–toxicity Classification: LCC QR121 (print) | LCC QR121 (ebook) | NLM QW 85 | DDC 579.3/7–dc23 LC record available at https://lccn.loc.gov/2016032565 A catalogue record for this book is available from the British Library. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Cover image: Dr O. Donkor Set in 10/12.5pt Times by SPi Global, Pondicherry, India

10 9 8 7 6 5 4 3 2 1

Contents

List of Contributors Preface to the Technical Series Preface 1 Microbial Toxins – An Overview R. Early and A.Y. Tamime

xi xv xvi 1

1.1 Introduction 1 1.2 Microbial toxins: modes of action 2 1.3 Bacterial toxins 4 1.3.1 Staphylococcal enterotoxins (SEs) 4 1.3.2 Bacillus cereus group enterotoxins 6 1.3.3 Clostridium botulinum nerotoxin6 1.4 Mycotoxins 7 1.4.1 Background 7 1.4.2 General aspects of mycotoxins 7 1.4.3 Postscript on mycotoxins 13 1.5 Biogenic amines (BAs) 14 1.6 Conclusions 14 References16 2 Incidences of Mould and Bacterial Toxins in Dairy Products M.L.Y. Wan, N.P. Shah and H.I. El‐Nezami

19

19 2.1 Background 2.2 Bacterial toxins 19 2.2.1 Emetic toxin produced by Bacillus cereus20 2.2.2 Enterotoxins produced by Staphylococcus aureus24 2.2.3 Botulinum neurotoxins produced by Clostridium botulinum26 33 2.3 Mould toxins (mycotoxins) 2.3.1 Aflatoxins B1 and M1 34 2.3.2 Sterigmatocystin 42 2.3.3 Ochratoxin A 43 2.4 Other mycotoxins 47 2.5 Conclusion 49 References50

vi  Contents

3 Bacterial Toxins – Structure, Properties and Mode of Action J.W. Austin

71

3.1 Background 71 3.2 Bacillus cereus toxins 72 3.2.1 Bacillus cereus emetic toxin 73 3.2.2 Bacillus cereus enterotoxins 74 3.2.3 Bacillus cereus haemolysin BL (Hbl) 76 3.2.4 Bacillus cereus non‐haemolytic enterotoxin (Nhe) 76 3.2.5 Cytotoxin K (Cyt K) 77 3.3 Botulinum neurotoxin 77 3.3.1 Outbreaks of botulism caused by dairy products 78 3.3.2 Structure of botulinum neurotoxin 79 3.3.3 Mode of action of BoNTs 80 3.4 Staphylococcus aureus enterotoxin 80 3.5 Conclusions 83 References83 4 Biogenic Amines in Dairy Products V. Ladero, D.M. Linares, M. Pérez, B. del Rio, M. Fernández and M.A. Alvarez

94

4.1 Introduction 94 4.2 Biochemistry: biosynthesis pathways, enzymes and transporters 98 4.2.1 Tyramine 98 4.2.2 Histamine 99 4.2.3 Putrescine 100 4.2.4 Cadaverine, β‐phenylethylamine, tryptamine 101 4.3 Biogenic amine‐producing micro‐organisms 101 4.3.1 Genes involved in the biosynthesis of biogenic amines 103 4.3.2 Is the production of BAs a strain- or species-dependent characteristic?105 106 4.3.3 Physiological functions of BAs biosynthesis 4.4 Toxicological effects 107 4.4.1 Tyramine 108 4.4.2 Histamine 109 4.4.3 Putrescine and polyamines 110 4.4.4 Cadaverine, tryptamine and β‐phenylethylamine111 111 4.4.5 Recommended limits of BAs 4.5 Factors affecting BAs accumulation in dairy products 113 4.5.1 Presence of BAs-producing bacteria 113 4.5.2 Physiochemical factors 114 4.5.3 Technological factors 117 4.6 Other preventive methods 119 4.7 Conclusions 119 Acknowledgements120 References120

Contents  vii

5 Contamination of Raw Milk: Sources and Routes Up to the Farm Gate R. Early

132

5.1 Introduction 132 5.2 The concept of contamination 132 5.2.1 What does contamination mean to the concept of food?  134 5.2.2 Contamination and cow health 135 5.3 Sources of contamination 136 5.3.1 Biological contamination 136 5.3.2 Chemical contamination 142 5.3.3 Mycotoxins 147 5.3.4 Physical contamination 148 5.4 Conclusion 150 References150 6 Milk Product Contamination After the Farm Gate R. Early

154

6.1 Introduction 154 6.2 The significance of microbial contamination 154 6.2.1 Product spoilage 154 6.2.2 Food-borne illness 155 6.2.3 Microbial toxins 155 6.3 Factories, processes and people 157 6.4 Raw milk handling 158 6.5 Milk‐processing and dairy products manufacture 160 6.5.1 Liquid milk and cream processing 161 6.5.2 Packing, storage, distribution and the retail environment 162 6.6 Buttermaking 164 165 6.7 Cheesemaking 6.8 Yoghurt 168 6.9 Milk powders 170 6.10 Evaporated milk and sweetened condensed milk 172 6.10.1 Evaporated milk 172 6.10.2 Sweetened condensed milk 175 6.11 Ice-cream 175 6.12 Hygiene, food safety management and cleaning‐in‐place (CIP) 177 6.13 Packaging, storage, distribution and the retail environment 178 6.14 Conclusions 180 References180 7 Techniques for Detection, Quantification and Control of Bacterial Toxins L. Ramchandran, A. Warnakulasuriya, O. Donkor and T. Vasiljevic 7.1 Introduction 7.2 Bacterial toxins

183 183 184

viii  Contents

7.3 Control of toxins 186 7.4 Methods for identification and detection of microbial toxins 187 7.4.1 Traditional biological assays 189 7.4.2 Antibody and immunoassay 191 7.5 Conclusion 196 References196 8 Techniques for Detection, Quantification and Control of Mycotoxins in Dairy Products O. Donkor, L. Ramchandran and T. Vasiljevic

201

8.1 Introduction 201 8.2 Methods for detection and quantification of mycotoxins 203 8.2.1 Sample pre-treatment method 203 8.2.2 Liquid-liquid extraction 204 8.2.3 Supercritical fluid extraction 204 8.2.4 Solid phase extraction 204 8.3 Separation methods 206 8.3.1 Thin layer chromatography 206 8.3.2 High pressure liquid chromatography (HPLC) 208 8.3.3 Gas chromatography (GC) 210 8.3.4 Capillary electrophoresis (CE) 210 8.3.5 Biosensors 210 8.3.6 Enzyme-linked immmunosorbent assay (ELISA) method 212 8.3.7 Electrochemical immunoassay 214 8.3.8 Polymerase chain reaction (PCR)-based detection and quantification 215 8.4 Mathematical model (exposure assessment of mycotoxins in dairy milk) 216 217 8.5 Control of mycotoxin 8.5.1 Physical methods 218 8.5.2 Chemical methods 218 8.5.3 Biological methods 219 8.5.4 Activated carbon (AC) 219 8.6 Conclusion 220 References220 9 Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production N. Murru, R. Mercogliano, M.‐L. Cortesi, F. Leroy, R. Condoleo and M.F. Peruzy 9.1 Background on risk analysis 9.2 Focus on cheese risk assessment 9.2.1 Source of milk 9.2.2 Raw and/or heat-treated milk cheeses 9.2.3 Level of moisture in cheese

229

229 231 231 231 232

Contents  ix

9.2.4 Methods of manufacture 232 9.2.5 Fat content 232 9.2.6 Maturation indices 232 9.2.7 Washed or mould cheeses 233 9.3 Staphylococcus aureus233 9.3.1 Background 233 9.3.2 Staphylococcus aureus and production of staphylococcal eneterotoxins234 9.3.3 Cheese production and hazard characterisation of Staphylococcus aureus237 9.3.4 Cheesemaking conditions and exposure assessment of Staphylococcus aureus238 9.3.5 Predictive modelling and risk assessment of Staphylococcus aureus and enterotoxin production in cheese 242 9.4 Escherichia coli243 9.4.1 Hazard identification 243 9.4.2 Growth and inactivation 244 9.4.3 Hazard characterisation 244 9.4.4 Exposure assessment 246 9.4.5 Risk characterisation 248 9.5 Listeria monocytogenes251 9.5.1 Hazard characterisation 251 9.5.2 Exposure assessment 253 9.5.3 Hazard characterisation 256 9.5.4 Risk characterisation 258 9.6 Cheese ‐ chemical risk assessment 259 9.6.1 Background 259 9.6.2 Biogenic amines in cheese 260 9.6.3 Occurrence of biogenic amines in cheese: hazard 260 and exposure assessment  9.6.4 Method for controlling biogenic amines in food 263 9.7 Modelling of growth and inactivation: kinetic approaches 264 9.7.1 Model categories: an overview 264 9.7.2 Growth ‐ no growth interface: a probabilistic approach 267 9.7.3 Modelling of toxin production 268 9.8 Conclusions 268 References269 10 Regulatory Measures for Microbial Toxins M. Hickey 10.1 Introduction and background 10.2 The evolution and economic significance of heat‐treated milk and milk products 10.2.1 Fluid milk 10.2.2 Evaporated and sweetened condensed milks

287 287 288 289 289

x  Contents

10.2.3 Milk and dairy powders 290 10.2.4 Cheeses (natural and processed) and fermented milks 291 10.3 Bacterial toxins 292 10.3.1 Staphylococcal enterotoxins 292 10.3.2 Bacillus cereus toxins 293 10.3.3 Clostridium botulinum toxins 294 10.4 Regulatory provisions on bacterial toxins in milk and milk products 297 10.4.1 European regulations on food hygiene and food safety 297 10.4.2 US milk hygiene and food safety standards 303 10.4.3 International perspective on food hygiene and safety – Codex Alimentarius307 10.5 Mycotoxins 310 10.5.1 Aflatoxins 311 10.5.2 Other mycotoxins 311 10.5.3 Aflatoxins M1 and B1 and their regulatory provisions 312 10.5.4 EU legislations on aflatoxins in milk, milk products and animal feed313 10.5.5 Regulations of aflatoxins of importance in milk and milk  products in the USA 314 10.5.6 Regulations of aflatoxins of importance in milk and milk  products in Canada 314 10.5.7 Regulation of aflatoxins in Australia and New Zealand 315 10.5.8 MERCOSUR standard on aflatoxins 315 10.6 Conclusions 316 References316 Index

323

List of Contributors

Editor Dr A.Y. Tamime Dairy Science & Technology Consultant Ayr Scotland ‐ United Kingdom Tel. +44 (0)1292 265498 Fax +44 (0)1292 265498 Mobile +44 (0)7980 278950 E-mail: [email protected]

Contributors Dr M.A. Alvarez Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Head of the Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain Tel. +34 985893418 Fax +34 985892233 E-mail: [email protected]. Dr J.W. Austin Health Canada Food Directorate Health Products and Food Branch Ottawa Ontario Canada Tel. + 1 613 957-0902 Fax +1 613 941-0280 E-mail: [email protected]

Dr R. Condoleo Istituto Zooprofilattico Sperimentale Lazio e Toscana M. Aleandri Roma Italy Tel. +39 06 79099360 Fax +39 06 79099312 E-mail: [email protected] Dr M.‐L. Cortesi Research Group of Food Inspection of Department of Veterinary Medicine and Animal Production Napoli Italy Tel. +39 081 2536469 Fax +39 081 458683 E-mail: [email protected] Dr O. Donkor Victoria University Advanced Food Systems Research Unit College of Health and Biomedicine Werribee Campus Melbourne Victoria Australia Tel. + 61 (0)3 9919 8062 Fax + 61 (0)3 9919 8284 E-mail: [email protected]

xii  List of Contributors

Dr R. Early Harper Adams University Department of Food Science and Agri‐ Food Supply Chain Management Newport Shropshire England – United Kingdom Tel. +44 (0) 1952 815365 Mobile +44 (0) 7792 453319 Fax +44 (0) 1952 814783 E-mail: [email protected] & [email protected] Dr M. Fernández Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain Tel. +34 985893352 Fax +34 985892233 E-mail: [email protected] M. Hickey Michael Hickey Associates Food Consultancy, Derryreigh Creggane Charleville Co. Cork Ireland Tel. +353 (0)63 89392 Mobile +353 (0)87 2385653 E-mail: [email protected] Dr V. Ladero Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain

Tel. +34 985892131 Fax +34 985892233 E-mail: [email protected] Dr F. Leroy Vrije Universiteit Research Group of Industrial Microbiology and Food Biotechnology (IMDO) Faculty of Sciences and Bio‐engineering Sciences Brussels Belgium Tel. +32 (0)2 6293612 Fax +32 (0)2 6292720 E-mail: [email protected] Dr D.M. Linares Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain Tel. +34 985892131 Fax +34 985892233 E-mail: [email protected] Dr R. Mercogliano Research Group of Food Inspection of Department of Veterinary Medicine and Animal Production Napoli Italy Tel. +39 081 2536062 Fax +39 081 458683 E-mail: [email protected] Dr N. Murru Research Group of Food Inspection of Department of Veterinary Medicine and Animal Production Napoli Italy

List of Contributors   xiii

Tel. +39 081 2536062 Fax +39 081 458683 E-mail: [email protected] Dr H.I. El‐Nezami University of Hong Kong 5S‐13 Kadoorie Biological Sciences Building Hong Kong Tel. +852 2299 0835 Fax +852 22990364 E-mail: [email protected] Dr M.F. Peruzy Research Group of Food Inspection of Department of Veterinary Medicine and Animal Production Naples Italy Tel. + 0039 03 391884980 E-mail: mariafrancesca.peruzy@gmail. com Dr M. Pérez Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain Tel. +34 985892131 Fax +34 985892233 E-mail: [email protected] Dr L. Ramchandran Victoria University Advanced Food Systems Research Unit College of Health and Biomedicine Werribee Campus Melbourne Victoria Australia Tel. + 61 (0)3 9919 8062 Fax + 61 (0)3 9919 8284 E-mail: [email protected]

Dr B. del Rio Spanish National Research Council (CSIC) Dairy Research Institute (IPLA‐CSIC) Department of Biotechnology of Dairy Products Villaviciosa Asturias Spain Tel. +34 985892131 Fax +34 985892233 E-mail: [email protected] Dr N.P. Shah University of Hong Kong 6N‐08, Kadoorie Biological Sciences Building Dairy and Probiotic Unit Food and Nutritional Science Programme Hong Kong Tel. +852 2299 0836 Fax +852 2559 9114 E-mail: [email protected] Dr T. Vasiljevic Victoria University Advanced Food Systems Research Unit College of Health and Biomedicine Werribee Campus Melbourne Victoria Australia Tel. + 61 (0)3 9919 8062 Fax + 61 (0)3 9919 8284 E-mail: [email protected] Dr M.L.Y. Wan University of Hong Kong 5S‐12 Kadoorie Biological Sciences Building Hong Kong Tel. +852 2299 0835 Fax +852 2299 0364 E-mail: [email protected]

xiv  List of Contributors

Dr A. Warnakulasuriya Victoria University Advanced Food Systems Research Unit College of Health and Biomedicine Werribee Campus Melbourne

Victoria Australia Tel. + 61 (0)3 9919 8062 Fax + 61 (0)3 9919 8284 E-mail: asalangika.warnakulasuriya@ vu.edu.au

Preface to the Technical Series

For more than 60 years, the Society of Dairy Technology (SDT) has sought to provide education and training in the dairy field, disseminating knowledge and fostering personal development through symposia, conferences, residential courses, publications and its journal, the International Journal of Dairy Technology (previously published as the Journal of the Society of Dairy Technology). Over the last 150 years there has been an enormous gain in our understanding of the role of the microbial flora in food preservation, spoilage and the threats to our health. At the same time, improvements in process technology have been accompanied by massive changes in the scale of many milk processing operations, and the manufacture a wide range of dairy and other related products. The Society has embarked on a project with Wiley-Blackwell to produce a technical series of dairy‐related books to provide an invaluable source of information for practicing dairy scientists and technologists, covering the range from small enterprises to modern large‐scale operation. This, the fourteenth volume in the series, on Microbial Toxins in Dairy Products, provides a timely and comprehensive update of the potential and possible routes for contamination, techniques for detection and how this knowledge can be used to reduce the risks to the consumer. Andrew Wilbey Chairman of the Publications Committee ‐ SDT October 2015

Preface

Food‐borne diseases, including dairy products, have been recognised as major threats to human health and can affect the national economies of both industrialised and developing countries worldwide. It can be argued that some of the causes associated with dairy food‐borne diseases are the use of raw milk in the manufacture of dairy products, faulty processing conditions during the heat treatment of milk, contamination of products after post‐processing, failure in due‐diligence (i.e. adding processed food to dairy products, for example, the case of botulism in hazel nut yoghurt in the United Kingdom), in‐adequate clean water supply, and so on. Primarily, dairy food‐borne diseases affecting human health are associated with certain strains of bacteria (e.g. belonging to the genera of Clostridium, Bacillus, Escherichia, Staphylococcus and Listeria) that are capable of producing toxins, plus moulds that are similarly capable of producing mycotoxins, such as aflatoxins, sterigmatocytin and ochratoxin. In addition, biogenic amines can accumulate in dairy products through microbial activity (e.g. starter cultures in cheeses), where the ingestion of high concentrations can be dangerous to human health. Incidences of human illnesses associated with dairy products are relatively low compared with foodborne illnesses in general. The purpose of this book, which is written by a team of well‐known international scientists, is to review the latest scientific knowledge/developments in these fields, such as surveying the incidences of human illnesses caused by the consumption of dairy products, updating the analytical techniques required to examine bacterial and mould toxins, and the potential for contamination of milk as it passes along the food chain (i.e. from ‘farm‐to‐fork’). This is complemented by a review of current approaches used to model microbial growth and toxin production plus the associated risks, and the regulatory measures available in different countries for control of microbial toxins in dairy products. It is anticipated that this up‐dated information will help to further minimise the incidences of dairy food‐borne illnesses in humans. I would like to acknowledge the time and effort that the expert contributors have given to make this edition possible. Although some overlap of scientific data have occurred in some chapters, I have felt justified in allowing this overlap because it has resulted in making the chapter(s) easier to read and understand. Adnan Y. Tamime October 2015

1 Microbial Toxins – An Overview R. Early and A.Y. Tamime

1.1 Introduction Microbial toxins can be and often are troublesome to human health and well‐being. History records numerous occurrences of death and suffering caused by disease organ­ isms, such as Yersinia pestis, of bubonic plague or Black Death fame, Corynebacterium diptheriae and Vibrio cholerae, the causative organism of diptheria and cholera respec­ tively, as their names suggest, Bordetella pertussis, which causes whooping cough, and Salmonella enterica subsp. enterica serotype Typhi, which causes typhoid. Although these bacterial pathogens exhibit very different aetiologies in terms of vectors and modes of infection, they are all similar in that they cause disease by means of toxins. The word ‘toxin’ is derived from the ancient Greek language, and refers to poison pro­ duced by living cells or organisms; although today it has a wider application, and can apply to synthetic compounds. Modern medicine, linked with improvements in sanitation and other public health measures, has reduced the incidence of many bacterially mediated diseases, ­particularly in technologically developed societies. Since the nineteenth century, our scientific understanding of the mechanisms by which these organisms proliferated and caused disease has increased greatly. Our ability to treat the diseases by means of vaccines and antibiotics has meant that for many people today, the spectre of the once common ill­ nesses and death that the organisms represented no longer hovers close to society. This is not to say, however, that illness and disease caused by micro‐organisms and their toxins are no longer problematic. While diseases, such as diptheria, cholera and typhoid are relatively uncommon today, modern consumers all too often encounter the actions and consequences of microbial toxins, particularly bacterial toxins. When they con­ sume food products that have been contaminated and/or mismanaged, often during ­production, a consequence can be affliction with food‐borne disease, commonly referred to as food poisoning. All food businesses engaged in the production, processing, manufacture, distribution and sale of food products to consumers have to be concerned about food safety and the problem of food‐borne disease. The dairy industry is no exception. Although dairy products are amongst the safest of food products, because of the exceptionally high

Microbial Toxins in Dairy Products, First Edition. Edited by Adnan Y. Tamime. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

2   Microbial Toxins in Dairy Products

standards of general management, and specifically hygiene and food safety m ­ anagement, employed throughout the dairy chain, from farm to supermarket, the potential exists for dairy products to be involved in the occurrence of food‐borne disease. The food safety management strategies of dairy farmers, milk processors and milk products manufactur­ ers are designed and implemented precisely to safeguard consumers. An important part of the strategies concerns understanding the nature of the food‐borne disease organisms that may be associated with dairy products, and how to minimise the risk of their occur­ rence in products. The purpose of this chapter is to review the issue of toxins associated with dairy products and specifically those of microbial origin, thereby setting the scene for the rest of the book.

1.2  Microbial toxins: modes of action Toxigenesis is the ability to produce toxins and the sources of microbial toxins that may be associated with dairy products are two‐fold: (a) those produced by bacteria, and (b) those produced by fungi (or moulds). Bacterial toxigenesis is very significantly of greater concern to the dairy industry and consumers of milk products than are toxins produced by fungi, although the latter are not unimportant. Bacterial toxins are commonly designated as endotoxins and exotoxins. Most gram‐ negative bacteria and all Enterobacteriacea produce endotoxins (Lüderitz et al., 1966). These toxins are commonly components of bacterial cell walls in the form of lipopolysac­ charides, and are compounds composed of three units: (a) the lipid A component, which is hydrophobic in nature, (b) the core oligosaccharide (consisting of an inner and outer core), the structure of which varies diversely according to bacterial species and subspecies, and (c) the O polysaccharide, or O antigen as it is also known. The somatic antigen is located in the cell wall of both gram‐positive and gram‐negative bacteria. However, the somatic O antigen is exhibited by many organisms, including salmonellas and the well‐known Escherichia coli O157:H7, of which the letter O is at times erroneously reproduced as the numeral zero. Like the core oligosaccharide, the structure of the O antigen also varies widely according to bacterial species and subspecies. As biologically active, heat stable endotoxins, lipopolysaccharides are commonly associated with the infections and variety of symptoms caused by gram‐negative bacte­ ria. The lipopolysaccharide structures are released from the cell walls of pathogenic bacteria as they break down, or autolyse, following the normal path of death and decay, or experience externally mediated lysis such as when attacked by the immune system of the host organism. Lipid A is the toxic component of the lipopolysaccharide, considered by some to be the most potent element, although Bishop (2005) noted that other inflam­ matory compounds may be derived from bacteria, such as the diacylgly­cerylcysteine component of bacterial lipoproteins and nucleic acids amongst others. Because of its hydrophobic nature, lipid A attaches to and becomes embedded in the cell wall of the host cell, interfering with the normal transport and cell regulation ­processes undertaken by the cell wall. This has the consequence of causing inflammatory and other responses, including in some instances toxic shock, to the host organism.

Microbial Toxins – An Overview   3

Exotoxins, in contrast to endotoxins, are toxic compounds released by bacterial cells. They are commonly enzymes, although some are polypeptides. In some few cases, lipopolysaccharides may be secreted as exotoxins. Exotoxins are proteinaceous com­ pounds secreted by bacterial cells locally to the point of action or at some ­distance from the tissue sites that they affect. For example, amongst the range of toxins ­produced by the gram‐positive organism Staphylococcus aureus, one is a protein enterotoxin that may be secreted onto food. Although the Staph. aureus organism may be killed by heat treatment, the toxin, which is quite heat stable, may remain unaffected and cause subsequent food‐ borne illness and the commonly exhibited emetic food poisoning symptoms. In contrast, Clostridium botulinum, the most heat resistant food‐related pathogen of concern, pro­ duces an exotoxin, which is readily denatured by heating. Terms, such as enterotoxin and neurotoxin associated with, in this instance, Staph. aureus and the also gram‐positive C. botulinum, describe either the mode of action of the toxin, or the target tissues. Some bacterial toxins have specific cytotoxic activity, such as C. botulinum neurotoxin (BoNT), of which there are seven antigenic types labelled BoNT A to G, affecting only nerve tis­ sues and preventing the proper functioning of neurotransmitters. Dembek et al. (2007) reported that human botulism was caused by BoNT A, B, E and F, and that the different neurotoxins exhibit different toxicities and persistence in cells. Other bacteria, such as Bacillus cereus, have a broad, non‐specific cytotoxic activity, affecting various cells and tissue types causing non‐­ specific cellular necrosis. Bacterial protein toxins exhibit strongly antigenic properties, although in some instances antitoxins can be used to neu­ tralise their toxicity. In the fight against bacterial disease, toxoids can be produced from bacterial protein toxins by exposing the compounds to a combination of reagents, such as formalin and organic acids, and moderate heat. The toxoids can then be used to provide artificial ­immunisation against diseases, such as diphtheria. Such immunisation is desir­ able where infectious bacterial diseases are concerned, but is not normally practiced in relation to food‐borne disease organisms of the kind associated with food poisoning. Mycotoxins are a class of toxic compound produced by fungi. In contrast to bacterial toxins, they are of lesser concern to the dairy industry as mycotoxin producing fungi do not commonly grow on dairy products. Fungi normally associated with dairy products tend to be limited to organisms, such as Penicillium roqueforti used to produce mould ripened blue cheeses, for example, Stilton and Roquefort, and Geotrichum candidum used in the manu­ facture of white mould ripened soft cheeses, for example, Neufchatel, Camembert and Brie. The mycotoxins of concern to food safety are secondary metabolites, released from fungal cells during growth. They include aflatoxins (AFs), produced by Aspergillus species, such as Aspergillus flavus and Aspergillus parasiticus, Fusarium spp. mould toxins, ochratoxin pro­ duced by various Aspergillus spp. and Penicillium spp., patulin, produced by Penicillium expansum, amongst other species, and ergot, the cause of ergotism and produced by the fungus Claviceps purpurea. Mycotoxins are generally associated with the spoilage of com­ modity crops, such as cereals spoiled in the field by, for example, Fusarium spp. or Clav. purpurea, and harvested seeds and nuts kept in store, such as cereal grains, peanuts, and so on, contaminated with aflatoxins from the growth of Aspergillus spp. From the perspective of the dairy industry, perhaps the main cause of concern with mycotoxins is the possibility of the transmission of these toxic compounds from con­ taminated animal feed through the cow (or other milk producing animal) into milk. The

4   Microbial Toxins in Dairy Products

risk then exists that mycotoxin residues in milk may be carried into dairy products, affecting consumers. As stated by the International dairy Federation (IDF, 2012), the ability of a mycotoxin or its metabolite to be excreted in milk will depend on the ability of the compound to pass the blood‐milk barrier. The IDF records that the only myco­ toxin which has been shown to possess this ability to any significance is AF B1, which is excreted in milk as AF M1. Aflatoxin M1 is presumed to be carcinogenic, affecting the liver, but is considerably less so than AF B1, by a factor of 10.

1.3  Bacterial toxins It is a well established fact that bacterial pathogens can cause food‐borne diseases in humans, and the possible routes of infection with relevance to dairy products are: (a) ingestion of already produced toxin(s) (i.e. sensu stricto), and (b) ingestion of a patho­ genic bacterium that is capable of producing toxins in the gastrointestinal (GI) tract (in situ). The micro‐organisms that have been identified to cause food‐borne i­llness via the consumption of dairy products belong to the genera of Staphylococcus (i.e. production of staphylococcal enterotoxin  –  SE), Clostridium (production of BoNT) and Bacillus (emetic type). These micro‐organisms including their toxin p­ roduction are reviewed extensively in Chapter 3; however, readers are referred to the following selected refer­ ences for complete discussion of food‐borne diseases including dairy products (Cary et al., 2000; Hui et al., 2001; Labbe & Garcia, 2001; De Buyser et al., 2001; de Leon et al., 2003; Jay et al., 2005; Granum, 2006; Heidinger et al., 2009; Argudín et al., 2010; Hale, 2012; Claeys et al., 2013, 2014; Hadrya et al., 2013).

1.3.1  Staphylococcal enterotoxins (SEs) According to Paulin et al. (2012), some characteristics of SEs in milk and dairy ­products are summarised as follows:

• • • • • • •

The SEs pass through the stomach into the intestinal tract where they stimulate eme­ sis and diarrhoea. Common symptoms are nausea, vomiting, retching, abdominal cramping and diarrhoea. Symptoms start 1–6 h after consuming food containing SEs and resolve within 1–3 d without the need for treatment. Food poisoning containing SEs is not usually fatal, but some fatalities can occur in very young or old people. Dairy‐borne outbreaks in many countries are associated with consumption of dairy products made from raw milk that cause SEs intoxications. Pasteurisation of milk inactivates Staph. aureus, whilst cheeses made from raw milk do not have such an elimination step; thus, safety is not guaranteed. In general, staphylococci are inactivated by D60°C of 6 min and the presence of ­lactoperoxidase in milk enhances the inactivation of Staph. aureus, decreasing its D value 15‐fold; however, SEs are heat‐stable at 121°C for 15 min.

Microbial Toxins – An Overview   5

•

At present, 23 SE serotypes have been identified, and the potential for enterotoxin to occur in cheese can be defined as: (a) the initial concentration of Staph. aureus in milk prior to cheesemaking must be sufficient, (b) the genes in Staph. aureus must be able to encode SE production in cheese milk, (c) the environmental conditions of pH, temperature and other factors must be suitable to permit bacterial growth and enterotoxin production, and (d) subsequent treatments, such as scalding the curd and brining may inactivate Staph. aureus, but any enterotoxin which has been formed is unlikely to be destroyed in the cheese.

It is of interest to note that some strains of staphylococci are used as secondary starter cultures for the manufacture of certain cheese varieties, and according to Bockelmann (2010), “Staphylococcus xylosus and Staphylococcus carnosus are used in certain varie­ ties of cheese to optimise the texture and aroma development. They are used as cheese adjuncts in starter cultures, or can be brushed or sprayed onto the cheese surface. These strains exhibit medium proteolytic and low lipolytic and aminopeptidase activities.” “Staphylococcus equorum is ubiquitous in cheese brines. It became available as starter culture only recently, and it has similar technological properties as Staph. xylosus, which is used to optimise the texture and aroma development in the cheese. In combination with Debaromyces hansenii, Staph. equorum supports the growth of other smear type bacteria when the ripening of the cheese starts, and it has a mould‐inhibiting effect. In addition, it can contribute to colour development when pigmented strains are used. Although the common consensus is that coagulase‐negative staphylococci (CNS), such as Staph. equorum do not represent a concern with respect to food‐borne disease, Irlinger et al. (2012) suggested that this may be changing and that CNS may produce enterotoxins harmful to humans.” In dairy‐mediated staphylococcal food poisoning, cheese has been the most fre­ quently incriminated. In France, it accounted for about 90% of the staphylococcal‐medi­ ated outbreaks with raw‐milk cheese representing 96.2% of the recorded cheese‐borne staphylococcal intoxications. Also, the high incidence of SE‐producing Staph. aureus in cheese compared to other dairy products appears to be a general tendency, probably because this product provides an optimal medium for the growth of enterotoxigenic Staph. aureus, which thrives in media rich in protein and with a high salt content (Singh et  al., 2012), as is the case for many types of cheeses. Also, the commonly applied mesophilic temperature (25–37°C) during fermentation allows the pathogen to grow rapidly and produce enterotoxins before conditions are no longer favourable (active development of lactic acid by lactic acid bacteria (LAB) and consequent pH drop). “Staphylococcus spp. are salt‐ and acid‐tolerant micro‐organisms, which can grow at the early stages of cheese ripening when the pH is still below 6, and they are found in all kinds of surface ripened cheeses. Like the yeasts, Staph. equorum is found in the cheese brine, sometimes at high cell counts (max. 105 colony forming units (cfu) mL−1). When the cheese brine is pasteurised, frequently to reduce the yeast counts (a practice adopted by many soft cheese producers), no or very low concentrations of staphylo­ cocci are present. Species most frequently observed on smear cheeses are Staph. equorum (natural flora), Staph. xylosus (cultural flora), and the nonfood‐grade Staphylococcus saprophyticus (natural contamination).”

6   Microbial Toxins in Dairy Products

“Staphylococcus equorum and Staph. xylosus seem to be the typical, naturally o­ ccurring species in cheese brines and on most smear cheeses. In a different study, all 150 cocci of a smeared Gouda cheese and a Bergkaese isolated from organic farmhouse cheese producer in Northern Germany were classified as Staph. equorum by amplified ribosomal deoxyribonucleic acid (DNA) restriction analysis (ARDRA) method. This was ­confirmed when staphylococci, which were isolated from French smeared soft cheeses of three ­different producers, were identified by species and strain level. The Staph. equorum flora consisted of a variety of strains, typical of a house flora, whereas all Staph. xylosus isolates showed identical DNA restriction patterns in pulsed‐field gel‐electrophoresis, which matched the pattern of a commercial Staph. xylosus strain, indicating that this organism was added as a starter culture.” “Staphylococcus saprophyticus is a nonfood‐grade species, and it is repeatedly ­isolated in low numbers from smear‐ripened cheeses and brine. Acid curd cheese (Harzer) seems to be an exception, where Staph. saprophyticus can be predominant in the staphylococcal surface flora, and can grow to high counts (e.g. 109 cfu cm−2)” – (Source: Technology of Cheesemaking, reproduced with permission of Wiley‐Blackwell).

1.3.2  Bacillus cereus group enterotoxins A wide range of enterotoxins are produced by B. cereus (see Chapter 3), and some char­ acteristics of these toxins include: (a) the identified toxins are: emetic toxins (heat‐stable at 126°C for 90 min), haemolysin BL (Bbl) (heat‐labile and inactivated at 56°C for 30 min, for details refer to Chapter 3), non‐haemolytic enterotoxin (Nhe) isoforms A, B and C (heat‐labile and inactivated at 56°C for 30 min), and cytotoxic toxin (Cyt) ­isoforms K1 and K2 (heat‐labile and inactivated at 56°C for 30 min), (b) the symptoms of food‐ or dairy‐ borne illnesses include nausea, abdominal cramps, watery diarrhoea and/or vomiting, and (c) the food‐poisoning symptoms are caused by intoxication with the peptide cereulide, and the diarrheal form of B. cereus food poisoning is caused by enterotoxins produced by growth of B. cereus in the small intestine after ingestion of viable cells or spores.

1.3.3  Clostridium botulinum nerotoxin The symptoms of botulism are caused by the ingestion of highly soluble exotoxin pro­ duced by C. botulinum while growing in foods or dairy products compared to the food poisoning of Clostridium perfringens where large numbers of viable cells must be ingested (for more details, refer to Chapter 3). The toxin produced is known as BoNT, and seven serotypes have been identified, for example, BoNT A to G; only types A, B, E, F and G cause diseases in humans (Jay et al., 2005). The thermal D values of endospores of C. botulinum (i.e. BoNT A to G) are: D110°C of 2.7‐2.9; D110°C of 1.3‐1.7‐2.9; not reported; D80°C of 0.8; D110°C of 1.6‐1.8; D80°C of 0.3‐0.8; and D110°C of 0.5, respectively. Also, all BoNTs are produced as single polypeptides (Jay et al., 2005). The chemical formula of the toxin is C6760H10447N1743O2010S32 (http://en. wikipedia.org/wiki/Botulinum_toxin) – accessed on 22nd April 2015, and the s­ tructure of the BoNT is reported by Silvaggi et al. (2007).

Microbial Toxins – An Overview   7

1.4 Mycotoxins 1.4.1  Background Fungal metabolites, which are toxic to humans and animals, are known as mycotoxins and consist of aflatoxins (AF – also known as A. flavus toxin – A‐fla‐toxin), ochratoxins (OTs), trichothecenes, zearalenone (ZEN), fumonisins (F), t­remorgenic toxins, tri­ chothecenes and ergot alkaloids (Zain, 2011). The International Agency for Research on Cancer (IARC, 2002a, 2002b) of the World Health Organisation (WHO) classified the carcinogenicity of mycotoxins to humans as follows: (a) AF is carcinogenic (Group 1), (b) OT and F are possibly carcinogenic (Group 2B), and (c) trichothecenes and ZEN are not carcinogenic to humans (Group 3) (IARC, 1993a, 2002a, 2002b; see also www.afro. who.int/des). The most likely predominant genera of fungi to produce mycotoxins in dairy products are Penicillium and Aspergillus (Frazier & Westhoff, 1988; Yousef & Juneja, 2003; Jay et al., 2005). The former organism could originate in milk due to unhygienic milk ­production (i.e. cheesemaking using raw milk), or the use of secondary starter cultures (e.g. Penicillium roqueforti) for the manufacture of Blue Veined cheeses (Roquefort, Stilton, Gorgonzola, Blue d’Auvergne, Cabrales, Blauschimmelkase, Tulum and Danablue) and white mould cheeses (Penicillium camemberti), such as Camembert, Brie and Gammelost. Although some penicilia spp. have been reported in old dairy books (Penicillium caseiocolum, Penicillium caseiocola, Penicillium ­candidum and Penicillium album), these are now considered biotypes of, or synonyms for, P. camemberti (Tamime, 2002). However, Aspergillus spp. can contaminate animal feed (e.g. aflatoxins ‐ AF), which can be excreted into the milks after being consumed by lactating cows. Another mould specie, Byssochlamys fulva, has been found in milk and can pro­ duce toxins (e.g. byssotoxin A, byssochlamic acid, ­patulin, fumitremorgin A and C, verruculogen, fischerin and eupenifeldin) (Tournas, 1994), but none have been impli­ cated in dairy‐borne products outbreaks, and they will not be reviewed in this chapter; however, more detailed infromation of incidences mycotoxins in dairy products that can be implicated in human health risk are reviewed in Chapter 2.

1.4.2  General aspects of mycotoxins Aflatoxin According to Frazier & Westhoff (1988), IARC (2002a, 2002b), Yousef & Juneja (2003) and Jay et al. (2005), AF B1 and B2 are produced by A. flavus, whilst A. parasiticus produces AF B1, B2, G1 and G2. However, some aspergilla strains (Aspergillus nominus, Aspergillus bombycis, Aspergillus pseudotamari and Asper­gillus ochraceoroseus) and Emericella venezuelensis are also AF‐producers, but they are encountered less fre­ quently especially in dairy products, and will not be reviewed in this book. Aflatoxin B1 is produced by all aflatoxin mould‐producers, which is the most potent form of all. Some AF serotypes (AF L, LH1, Q1 and P1) are derived from AF B1, and some of the main AF characteristics are shown in Table (AF 1.1). The potent toxicity of the main six AF in descending order is as follows: AF B1 (blue) > M1 (blue/violet) > G1 (green) > B2

8   Microbial Toxins in Dairy Products

(blue) > M2 (violet) and G2 (green/blue) – data in parenthesis illustrate the fluorescence noted when viewed under ultraviolet (UV) light, where the colours designate the AF serotype (IARC, 2002a, 2002b; Jay et al., 2005). Aflatoxins cross the human placenta, and exposure has been associated with growth impairment in young children. In general, AF production by moulds occurs in a growth environment of water activity (aw) of 0.85 and at a temperature of 25–40°C. Ochratoxin Also, the genera of Aspergillus (e.g. A. ochraceus, A. westerdijkiae, A. niger, A. carbonarius, A. lacticoffeatus and A. sclerotioniger) and Penicillium (P. verrucosum and P. nordicum), which consist of many filamentous fungi, produce mycotoxin known as ochratoxin (OT) (el Khoury & Atoui, 2010; Sorrenti et al., 2013). The metabolite, which has first identified, is known as OT A, and its related metabolites, such as OT B (i.e. dechloro analogue of OT A) and OT C (i.e. the isocoumaric derivative of OT A) (see Table 1.1). The International Union of Pure and Applied Chemistry (IUPAC) names of OT A to C are as follows: N‐​{[(3R)‐​5‐​chloro‐​8‐​hydroxy‐​3‐​methyl‐​1‐​oxo‐​3,​4‐​dihydro‐​ 1H‐​isochromen‐​7‐​yl]​carbonyl}‐​L‐​phenylalanine, (2S)‐2‐[[(3R)‐8‐hydroxy‐3‐methyl‐ 1‐oxo‐3,4‐dihydroisochromene‐7‐carbonyl]amino]‐3‐phenylpropanoic acid and ethyl (2S)‐2‐[[(3R)‐5‐chloro‐8‐hydroxy‐3‐methyl‐1‐oxo‐3,4‐dihydroisochromene‐7‐carbonyl] amino]‐3‐phenylpropanoate, respectively. However, at present another 16 related metabolites of OT A have been identified, and for detailed information refer to the review by el Khoury and Atoui (2010). In addition to OT A to C, another sixteen OT A related derived metabolites have been also identified (el Khoury & Atoui, 2010), and some examples are OT α, OT β, OT A methyl‐ester, OT B methyl‐ester, OT B ethyl‐ester, 4‐R‐hydroxy‐OT A, 4‐s‐hydroxy‐OT A, and 10‐hydroxy‐OT A. The florescence colour noted viewed under UV light for OT A is greenish, whilst OT B emits blueish (Jay et al., 2005). According to Frazier & Westhoff (1988), although the effects of OTs to human beings are unknown or possibly slightly toxic, the significance of OTs in food is of interest for the following aspects:

• • • •

OTs are toxic to certain animals; Some OTs are heat resistant, and are not destroyed after prolonged autoclaving; Many OTs‐producing moulds are able to grow and produce mycotoxin at tempera­ tures below 10°C; and Ochratoxins have been isolated from many foods.

Citrinin  Mould organisms belonging to the following genera: Aspergillus (A. niveus, A. ochraceus, A. oryzae and A. terreus), Monascus (M. ruber and M. ­purpureus) and Penicillium (P. citrinum and P. viridicatum and P. camemberti) have been reported to produce myco­ toxin known as citrinin (Jay et al., 2005; http://en.wikipedia.org/wiki/Citrinin). Citrinin is also known as antimycin and, according to  the IUPAC, it is known as (3R,4S)‐8‐ hydroxy‐3,4,5‐trimethyl‐6‐oxo‐4,6‐dihydro‐3H‐isochromene‐7‐carboxylic acid.

C17H12O7

AF G1

Carcinogenic, hepatocarcinogens

It is the 2,3‐dihydroform of AF B1, which reduces the mutagenicity by 200‐ to 500‐fold Carcinogenic, hepatocarcinogens, mutagenic, teratogenic, and causes immunosuppression

C17H12O6

AF B2

Comments

Catalysing 3‐hydroxylation of AF B1 to yield the AF Q1 metabolite Carcinogenic, immunosuppressive, hepatocarcinogens, genotoxic Binds covalently to live mitochondrial deoxynucleic acid

Chemical formula

Aflatoxins (AF) B1 C17H12O6

Name of mycotoxin

Table 1.1  Some general characteristics of mycotoxins detected in dairy products.

O

O

O

H

H

H

H

H

H

O

O

O

O

O

O

O

O

O

O

OCH3

O

OCH3

O

OCH3

Structure

(Continued)

It is hepatotoxic, nephrotoxic, neurotoxic, teratogenic and immunotoxic Carcinogenic to humans (Group 2B), and weakly mutagenic It is neurotoxic and cause immunosuppression and immunotoxicity in animals Can cause Balkan endemic nephropathy

Ochratoxin (OT) A C20H18ClNO6

Hydroxylated product of AF B1 Hepatotoxic, mutagenic (the C2─ C3 double bond in the dihydrofurofurane moiety), carcinogenic, immunotoxic, teratogenic, Less toxic than the parent compound AF B1

It is a 4‐hydroxy AF B2 Less toxic than the parent compound AF B2

C17H12O7

AF M1

It is the 2,3‐dihydroform of AF G1 Carcinogenic, hepatocarcinogens

Comments

AF M2

C17H14O7

Chemical formula

AF G2

Name of mycotoxin

Table 1.1  (Continued)

O

O

O

OH

H

O

O

O

N H

COOH O

O

O

O

OH

H

H

O

O

O

O

Cl

HO

OCH3

O

OCH3

O

OCH3

Structure

O

O

O

H

CH3

C20H19NO8

C22H22CINO6

C13H14O5

C7H6O4

C22H23N5O2

OT C

Citrinin (also known as antimycin)

Patulin

Roquefortine C

Chemical formula

OT B

Name of mycotoxin

Neurotoxin Its chemical nature is indole alkaloid

Toxicity ‐ neurotoxic, hepathotoxic, nephrotoxic, genotoxic/teratogenotoxic, pulmonary congestion and edema Demonstrated as carcinogen It is toxic primarily through affinity to sulfhydryl groups (SH) Its chemical nature is polyketide lactone

It is a nephrotoxin, and can permeate through the human skin. Although no significant health risk is expected after dermal contact in agricultural or residential environments but, nevertheless, it should be limited. Under long‐wave UV light → fluoresces lemon yellow Compound produces nephritis in mice Possess antimicrobial activity Its chemical nature is quinonemethine

Most likely it has similar properties as above

Most likely it has similar properties as above, but not toxic

Comments

N

O

O

O

O

CH3

OH

O

O

O Cl

H

N N H

O

O

CH3

CH2

CH3

O

OH

H H3C

CH3

O

O H O

OH

H N

NH HN

O

HOOC

O

O

HO

H N

Structure

O

(Continued)

C8H10O4

Penicillic acid

HO

H

O

O

O

HN O

O

N

CH3

OH

HO

OCH3 CH2

H

O

O

HH

H

Structure

O

O

O

O

OH

CH3

OCH3 CH2

Note: the reported incidences of mycotoxins in dairy products are detailed in Chapter 2. Data compiled from ICMSF (1978), Frazier & Westhoff (1988), IARC (2002a, 2002b), Yousef & Juneja (2003) and Jay et al. (2005); see also www.bing.com, http://wwww.fermentek.co.il/ aflatoxin_M2.htm, http://en.wikipedia.org, http://pubchem.ncbi.nlm.nih.gov/compound/20997#section=2D‐Structure

Demonstrated some toxic effects on laboratory animals

Toxic and it is related to dermatoxin Closely related to AF B1, which is a potent liver carcinogenic, mutagenic and teratogenic

C18H12O6

Sterigmatocystin

Comments

Appears to be toxic in high concentrations

Chemical formula

Cyclopiazonic acid C20H20N2O3 (CPA)

Name of mycotoxin

Table 1.1  (Continued)

Microbial Toxins – An Overview   13

Patulin Penicilla species (Penicillium claviform, Penicillim expansum, Penicillium patulum, P. roqueforti, Penicillium clavigerum, Penicillium griseofulvum, Penicillium crustosum and P. enicilliumpaneum), aspergilla species (e.g. Aspergillus clavatus, Aspergillus terreus and others), and Byss. fulva and Byssochlamys nivea ­produce mycotoxin known as patulin. When first discovered, it was described as an anti­ biotic and its chemical structure is synonymous with penicidin, clavatin, clavi­ formin, calvicin, mycoin C, expansin, and gigantic acid (Frazier and Westhoff, 1988), and the IUPC name is 4‐hydroxy‐4H‐furo[3,2‐c]pyran‐2(6H)‐one. Miscellaneous mycotoxins A wide range of mycotoxins have been found in dairy products, mainly cheeses (see Chapter 2) and, in brief, their characteristics are as follows:

•

•

•

•

•

Roquefortine ‐ Excessive growth of P. roqueforti in Blue Vein cheese (Roquefort, Stilton, Danablue, etc.) during the maturation period results in the production of a toxin known as Roquefortine C. It acts as a neurotoxin to animals (e.g. mice) when injected into the body leading to convulsive seizers (Frazier & Westhoff, 1988). Cyclopiazonic acid (CPA) – Chemically, it is known as an indole tetramic acid, and is isolated from the Penicillium cyclopium, Penicillium griseofulvum, P. camemberti, Penicillium commune, A. flavus, and Aspergillus versicolor. The IUPAC name is (6aR,11aS,11bR)‐10‐Acetyl‐11‐hydroxy‐7,7‐dimethyl‐2,6,6a,7,11a,11b‐hexahy­ dro‐9H‐pyrrolo[1’,2’:2,3]isoindolo[4,5,6‐cd]indol‐9‐one, (Holzapfel, 1968 ; http:// en.wikipedia.org/wiki/Cyclopiazonic_acid), accessed on 21st April 2015. Sterigmatocystin  –  This mycotoxin is isolated from the crust of hard cheeses. The  IUPAC name is (3aR,12cS)‐8‐hydroxy‐6‐methoxy‐3a,12c‐dihydro‐7H‐ furo[3’,2’:4,5]furo[2,3‐c]xanthen‐7‐one, and is produced by moulds of the g­ enera Aspergillus, and the classification by the IARC of sterigmatocystin is Group 2B (http://en.wikipedia.org/wiki/Sterigmatocystin), accessed on 21st April 2015. Penicillic acid – This mycotoxin is produced by Penicillium roqueforti, Penicillium spp. and Aspergillus spp. This toxin was found in hard cheeses and Roquefort. The  IUPAC name is 5‐hydroxy‐5‐isopropenyl‐4‐methoxy‐furan‐2‐one (http:// en.wikipedia.org/wiki/Penicillic_acid), accessed on 21st April 2015. Andrastin A‐D and isofumigaclavines A and B – They are secondary metabolites of P. roqueforti, and the latter mycotoxin is considered neurotoxic. As can be expected, their occurrences have been in Blue Vein cheeses.

1.4.3  Postscript on mycotoxins It is evident that different mycotoxins have been found in dairy products which may pose a slight or more severe health risks to humans depending on the ingested amounts of these toxin. However, some emerging mycotoxins (i.e. the masked, bound and or conjugated types) may pose potential health risk to humans if they appear in dairy ­products, but further studies are required to categorise their toxicity. Some of

14   Microbial Toxins in Dairy Products

these masked mycotoxins are produced by Fusarium spp., such as deoxynivalenol, zearalenone, fumonisins, nivalenol, fusarenon‐X, T‐2 toxin, HT‐2 toxin, and fusaric acid) (Berthiller et al., 2013; see also Scott, 1989; Murphy et al., 2006; Zinedine et al., 2007; Tsakalidou, 2011; Singh et al., 2012; Todd et al., 2014), including some bound and conjugated mycotoxins (enniatins, beauvericin and fusaproliferin) should not be overlooked. Last, but least, Cladosporium spp. have been isolated from Cheddar cheese (Hocking & Faedo, 1992; Basilico et al., 2003; Panelli et al., 2014), and the descriptive term used of the contaminant as ‘Thread Mould’. The source of the Cladosporium spp. contamination in Cheddar cheese made using the Block‐ Forming unit was the factory environment including the compressed air system (A.Y. Tamime, unpublished data); however, the problem was contained by installation of microbiological filters in the compressed air‐line entering the cheesemaking area. Nevertheless, more work is required to establish the toxicity of such mould metabo­ lite in humans and animals.

1.5  Biogenic amines (BAs) Biogenic amines are nitrogenous substances with one or more amine groups, which are formed mainly by decarboxylation of amino acids or by amination and transamination of aldehydes and ketones. In the present context, BAs are only found in cheeses, and they are formed by the metabolic activity/catabolism of the starter culture and their enzymes including any other bacteria that could be present in the milk prior to cheese­ making, and coagulating enzymes. The catabolism by bacteria and possibly moulds of raw material can produce BAs in milk, or they are generated by microbial decarboxyla­ tion of amino acids. It is well established that BAs play an important role as source of nitrogen and precursor for the synthesis of hormones, alkaloids, nucleic acids, proteins, amines and food aroma compounds. However, dairy products (e.g. cheeses) containing high amounts of BAs may have toxicological effects, and some characteristics of BAs are summarised in Table 1.2.

1.6 Conclusions This chapter has proposed through its argument and illustration that toxin‐producing micro‐organisms represent an important source of food‐borne hazard to consumers, which must therefore be managed. The control of endotoxin and exotoxin producing organisms must be an aim of all food processors and manufacturers, not just those involved in the manufacture of dairy products. However, as with the manufacture of any food product, the targets set for levels of toxin producing micro‐organisms in dairy products must be tempered with knowledge and common sense. For instance, the infec­ tive dose level of E. coli O157:H7 is low at 10–100 cfu g─1, whereas the requirement for the formation of Staph. aureus exotoxin sufficient to cause illness is much higher at some 106 cfu g−1. While every measure must be taken to  prevent or eliminate E. coli O157:H7 contamination of dairy products, it must also be recognised that in the case of  some dairy products, hand‐made, specialist cheeses for instance, saleable product

C5H9N3

C4H12N2

C4H12N2

C8H11N

C10H12N2

Histamine

Putrescine

Cadavarine

β‐phenylethylamine (β‐PEA)

Tryptamine

2‐(1H‐Indol‐3‐yl) ethanamine

2‐phenylethylamine

pentane‐1,5‐diamine

butane‐1,4‐diamine

2‐(1H‐imidazol‐4‐yl) ethanamine

4‐(2‐aminoethyl) phenol

IUPAC1

Heterocyclic amine, Monoamine It contains an indole ring structure, and is structurally similar to the amino acid tryptophan

Cyclic aromatic amine, monoamine

Aliphatic amine, diamine It is a toxic diamine with the formula NH2(CH2)5NH2, which is similar to putrescine

Aliphatic amine, diamine It is a tetramethylenediamine organic chemical compound NH2(CH2)4NH2 (1,4‐ diaminobutane or butanediamine) that is related to cadaverine

Heterocyclic amine, diamine

Cyclic aromatic amine, monoamine and trace of amine derived from the amino acid tyrosine It is also known as 4‐ hydroxyphenethylamine, para‐tyramine, mydrial or uteramin

Classification and other names applied

1

 IUPAC = International Union of Pure and Applied Chemistry. Data compiled from EFSA (2011), data from Chapter 4, and (http://en.wikipedia.org/wiki/) [Accessed 6th April 2015].

C8H11NO

Chemical formula

Tyramine

Name of BAs

7

6

5

HN 1

H2 N

H2 N

HN

N

HO

4

2

3

β α

NH2

NH2

Structure

Table 1.2  Some physical characteristics of biogenic amine (BAs) that have been identified in dairy products, mainly cheeses that are potential health risks to humans.

NH2

NH2

NH2

NH2

16   Microbial Toxins in Dairy Products

cannot be made if the specification for Staph. aureus is set at impractically and ­unnecessarily low levels. The management of food safety in dairy products manufacture must be based on good and informed judgement. Reliable and informative sources able to impart the knowledge and understanding required to inform judgement are, therefore, important to milk processors and dairy products manufacturers. It is hoped that this chapter, and the book of which it is part, will serve those who work in the dairy industry, as well as those who have academic or scholarly interests in the topic, enabling them to make appropriate decisions about matters of microbial food safety and dairy products, and particularly those concerning microbial toxins.

References Argudín, M.A., Mendoza, M.C. & Rodicio, M.R. (2010) Food poisoning and Staphylococcus aureus enterotoxins. Toxin, 2, 1751–1773. Berthiller, F., Crews, C., Dall’Asta, C., De Saeger, S., Haesaert, G., Karlovsky, P., Oswald, I.P., Seefelder, W., Speijer, G. & Stroka, J. (2013) Masked mycotoxins: A review. Molecular Nutrition & Food Research, 57, 165–186. Bishop, R.E. (2005) Fundamentals of exotoxin structure and function. Concepts in Bacterial Virulence (eds. W. Russell & H. Herwald), pp. 1–27, Karger, Basel. Bockelmann, W. (2010) Secondary cheese starter cultures. Technology of Cheesemaking (eds. B.A. Law & A.Y. Tamime) 2nd Edition, pp. 193–230, Wiley‐Blackwell, Oxford. Cary, J.W., Linz, J.E. & Bhatnager, D. (eds.) (2000) Microbial Foodborne Diseases: Mechanisms of Pathogenesis and Toxin Synthesis, Technico Publishing, Lancaster. Claeys, W.L., Cardoen, S., Daube, G., De Block, J., Dewettinck, K., Dierick, K., De Zutter, L., Huyghebaert, A., Imberechts, H., Thiange, P., Vandenplas, Y. & Herman, L. (2013) Raw or heated cow milk consumption: Review of risks and benefits. Food Control, 31, 251–262. Claeys, W.L., Verraes, C., Cardoen, S., De Block, J., Huyghebaert, A., Raes, K., Dewettinck, K. & Herman, L. (2014) Consumption of raw or heated milk from different species: An evaluation of the nutritional and potential health benefits. Food Control, 42, 188–201. De Buyser, M.L., Dufour, B., Maire, M. & Lafarge, V. (2001) Implication of milk and milk prod­ ucts in food‐borne diseases in France and in different industrialised countries. International Journal of Food Microbiology, 67, 1–17. Dembek Z.F., Smith L.A. & Rusnak J.M. (2007) Botulinum toxin. Textbooks of Military Medicine: Medical Aspects of Biological Warfare (ed. Z.F. Dembek), Vol. 16, pp. 337–353, The Borden Institute, Washington, D.C. EFSA (2011) Scientific Opinion on risk based control of biogenic amine formation in fermented foods ‐ European Food Safety Authority Panel on Biological Hazards (BIOHAZ). EFSA Journal, 9, 2393–2486. Frazier, W.C. & Westhoff, D.C. (1988) Food‐borne poisonings, infections, and intoxication: ­nonbacterial. Food Microbiology, 4th Edition, pp. 440–460, McGraw‐Hill Book Company, New York. Granum, P.E. (2006) Bacterial toxins as food poisons. The Comprehensive Sourcebook of Bacterial Protein Toxins (eds. J.E. Alouf & J.H. Feer), 3rd Edition, pp. 949–958, Elsevier, Boston. Hadrya, F., Benlarabi, S., Ben Alia, D., Hami, H., Soulaymani, A. & Soulaymani‐Bencheikh, R. (2013) Epidémiologie des intoxications liées aux produits laitiers au Maroc. Nature & Technologie, B‐ Sciences Agronomiques et Biologiques, 8, 17–22. Hale, M.L. (2012) Staphylococcal enterotoxins, staphylococcal enterotoxin B and bioterrorism. Bioterrorism (ed. S. Morse), pp. 41–64, InTech, Available from: http://www.intechopen.com/books/ bioterrorism/staphylococcal‐enterotoxins‐staphylococcal‐enterotoxin‐b‐andbioterrorism [Accessed 25th March 2015].

Microbial Toxins – An Overview   17

Heidinger, J.C., Winter, C.K. & Cullor, J.S. (2009) Quantitative microbial risk assessment for Staphylococcus aureus and Staphylococcus enterotoxin A in raw milk. Journal of Food Protection, 72, 1641–1653. Hocking, A.D. & Faedo, M. (1992) Fungi causing thread mould spoilage of vacuum packaged Cheddar cheese during maturation. International Journal of Food Microbiology, 16, 123–130. Holzapfel, C.W. (1968) The isolation and structure of cyclopiazonic acid, a toxic metabolite of Penicillium cyclopium Westling. Tetrahedron, 24, 2101–2119. Hui, Y.H., Pierson, M.D. & Gorham, J.R. (eds.) (2001) Diseases Caused by Bacteria, Vol. 1, 2nd Edition, Marcel Dekker, New York. IARC (1993) Mycotoxins. Some Naturally Occurring Substances: Food Items and Constituents, Heterocyclic Aromatic Amines and Mycotoxins, International Agency for Research on Cancer (IARC) Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 56, 245–524, WHO, Lyon. IARC (2002a) Some Mycotoxins. Some Traditional herbal Medicines, Some Mycotoxins, Naphthalene and Styrene, International Agency for Research on Cancer (IARC) Monographs on the Evaluation of Carcinogenic risks to humans, Vol. 82, 169–366, WHO, Lyon. IDF (2012) Feed‐Associated Mycotoxins in the Dairy Chain: Occurrence and Control, Bulletin No. 444, International Dairy Federation, Brussells. IARC (2002b) International Agency for Research on Cancer, Aflatoxins http://monographs.iarc.fr/ ENG/Monographs/vol100F/mono100F‐23.pdf [Accessed 16th April 2015]. ICMSF (1978) Food‐borne microbial Toxins. Micro‐organisms in Foods ‐1Their significance and Methods of Enumeration, 2nd Edition, 68–89, International Commission on Microbiological Specifications for Foods, University of Toronto, Toronto. Irlinger, F., Loux, V. Bento, P., Gibrat, J‐F., Straub, C., Bonnarme, P., Landaud, S. & Monnet, C. (2012) Genome sequence of Staphylococcus equorum subsp. equorum Mu2, isolated from a French smear‐ripened cheese. Journal of Bacteriology, 194, 5141–5142. Jay, J.M., Loessner, M.J. & Golden, D.A. (2005) Modern Food Microbiology, 7th Edition, pp. 567–590 and 709–726, Springer Science + Business Media, New York. el Khoury, A. & Ali Atoui, A. (2010) Ochratoxin A: General overview and actual molecular status. Toxin, 2, 461–493. Labbe, R.G. & Garcia, S. (2001) Guide to Foodborne Pathogens, Wiley and Sons Inc., New York. de Leon, S.Y., Meacham, S.L. & Claudio, V.S. (2003) Global Handbook on Food and Water Safety, Chas. C. Thomas, Springfield. Lüderitz, O., Staub, A.M. & Westphal, O. (1966) Immunochemistry of O and R antigens of ­salmonella and related Enterobacteriaceae. Bacteriology Reviews, 30, 192–255. Murphy, P.A., Hendrich, S., Landgren, C. & Bryant, C.M. (2006) Food mycotoxins: An update. Journal of Food Science, 71, 51–65. Panelli, S., Buffoni, J.N., Bonacina, C. & Feligini, M. (2014) Identification of moulds from the Taleggio cheese environment by the use of DNA barcodes. Food Control, 28, 385–391. Paulin, S., Horn, B. & Hudson, J.A. (2012) Factors Influencing Staphylococcal Enterotoxin Production in Dairy Products, MPI Technical, Paper No. 2012/07, Publications Logistics Officer, Ministry for Primary Industries, Wellington. Scott, P.M. (1989) Mycotoxigenic fungal contaminants of cheese and other dairy products. Mycotoxins in Dairy Products (ed. H.P. Van Egmond), pp. 193–259, Elsevier, Amsterdam. Silvaggi, N.R., Boldt, G.E., Hixon, M.S., Kennedy, J.P., Tzipori, S., Janda, K.D. & Allen, K.N. (2007) Structures of Clostridium botulinum neurotoxin serotype A light Chain complexed with small‐molecule inhibitors highlight active‐site flexibility. Chemistry & Biology, 14, 533–542. Singh, S.K., Pandey, V.D. & Verma, V.C. (2012) Bacterial food intoxication. Microbial Toxins and Toxigenic Microbes (eds. S.K. Singh & V.D. Pandey), pp. 215–232, Studium Press LLC, New Delhi. Sorrenti, V., Di Giacomo, C., Acquaviva, R., Barbagallo, I., Bognanno, M. & Galvano, F. (2013) Toxicity of Ochratoxin A and its modulation by antioxidants: A review. Toxins, 5, 1742–1766.

18   Microbial Toxins in Dairy Products

Tamime, A.Y. (2002) Microbiology of starter cultures. Dairy Microbiology Handbook  –  The Microbiology of Milk and Milk Products (ed. R.K. Robinson), 3rd Edition, pp. 261–366, John Wiley and Sons Inc., New York. Todd, E.C.D. (2014) Foodborne diseases: Overview of biological hazards and foodborne diseases. Encyclopedia of Food Safety (eds. Y. Motarjemi, G. Moy & E.C.D. Todd), pp. 221–286, Elsevier, London. Tournas, V. (1994) Heat‐resistant fungi of importance to the food and beverage industry. Critical Reviews in Microbiology, 20, 243–263. Tsakalidou, E. (2011) Microbial flora. Safety Analysis of Foods of Animal Origin. Part III: Milk and Dairy Foods (eds. L.M.L. Nollet & F. Toldra), pp. 781–798, CRC Press, Boca Raton. Yousef, A.E. & Juneja, V.K. (eds.) (2003) Microbial Stress Adaptation of Food Safety, CRC Press (Taylor & Francis Group), Boca Raton. Zain, M.E. (2011) Impact of mycotoxins on humans and animals. Journal of Saudi Chemical Society, 15, 129–144. Zinedine, A., Soriano, J.M., Molto, J.C. & Mañes, J. (2007) Review on the toxicity, occurrence, metabolism, detoxification, regulations and intake of zearalenone: an oestrogenic mycotoxin. Food and Chemical Toxicology, 45, 1–8.

2 Incidences of Mould and Bacterial Toxins in Dairy Products M.L.Y. Wan, N.P. Shah and H.I. El‐Nezami

2.1 Background Milk and dairy products form an essential part of the human diets, as they are rich in nutrients. However, due to their high nutrient values, they favour the rapid growth of micro‐organisms. Some of these micro‐organisms are beneficial, while some are undesirable. In many cases, milk and dairy products are easily contaminated by pathogens or preformed microbial toxins. Microbial toxins are toxins produced by micro‐organisms, including bacteria and fungi. These toxins, namely bacterial toxins and mould toxins, when introducing into host animals, play important roles in causing various food intoxication (Williams & Clarke, 1998). The Centres for Disease Control and Prevention (CDC) has published data demonstrating that dairy products accounted for more food‐ borne‐illness (13.8%), hospitalisations (16.2%) and deaths (9.7%) than 16 other food commodities in all foodborne disease outbreaks in the United States for 1998–2008 (Painter et al., 2013). Mould growth on cheese and other fermented dairy products is a common and recurring problem. Potential mycotoxin contamination is serious since some moulds can grow and produce mycotoxins at temperatures as low as −2 to 10°C. This chapter will review the common bacterial toxins and mycotoxins involved in food poisoning in milk and dairy products.

2.2  Bacterial toxins According to the Food and Drug Administration (FDA, 2010), there are numerous significant outbreaks of illnesses associated with milk and dairy products with ­ Listeria  monocytogenes, Salmonella spp., Yersinia enterocolitica, enterohemorrhagic Escherichia coli (EHEC) 0157:H7 and Clostridium botulinum. Food‐borne bacterial infection occurs when food contaminated with pathogenic ­bacteria is consumed. The bacteria produce toxins that link to and/or invade intestinal epithelial cells, and cause damage to specific cells. Food intoxication is caused by the consumption of toxins produced by bacterial growth, rather than bacterium itself. There are three bacterial genera that are considered to be important causes of the intoxication

Microbial Toxins in Dairy Products, First Edition. Edited by Adnan Y. Tamime. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

20   Microbial Toxins in Dairy Products

type of food poisoning. Bacillus cereus (emetic), C. botulinum and Staphylococcus aureus are the most common species that are capable of causing i­llnesses by producing pre‐formed toxins in food. Although the incidence of these bacteria has frequently been reported in different food‐stuffs including dairy ­products, there is no information about the occurrence and analysis of their toxins produced in food.

2.2.1  Emetic toxin produced by Bacillus cereus Bacillus cereus is rod‐shaped, gram‐positive, spore‐forming bacterium that is widely distributed in the environment, and are frequently found in a variety of foods including milk and dairy products, cereals (especially rice), and food additives (Reyes et al., 2007). The bacterium plays an important role as the causative agent of two different types of food poisonings: (a) the emetic syndrome, caused by ingestion of a preformed toxin in the food (Granum & Lund, 1997), and (b) the diarrhoeal syndrome, caused by complex enterotoxins (Beecher & Wong, 1997; Lund & Granum, 1997) that can either be formed in the food or produced during vegetative growth of B. cereus in the small intestine (Drobniewski, 1993; Granum, 1994). The diarrheal syndrome is characterised by abdominal pain, a profuse watery diarrhoea and rectal tenesmus, which appears 8 to 16 h after ingestion of contaminated food usually associated with proteinaceous foods, whereas the emetic syndrome is characterised by nausea, vomiting and malaise, starting 1 to 5 h after consuming food mostly associated with farinaceous foods, such as cooked rice (Kramer & Gibert, 1989). The food involved in both types of food has usually been heat treated, and the surviving spores are the source of the food poisoning (Granum et al., 1996). Diarrheal and emetic types of food poisonings are caused by very different types of toxins. The diarrheal disease is caused by at least three enterotoxins (Agata et al., 1995a, 1995b; Beecher & Wong, 1997; Lund & Granum, 1997). All the diarrheal causing toxins are unstable and are inactivated by low pH and digestive enzymes. As a consequence, any preformed toxins in food should be destroyed as they pass through the stomach. Therefore, it is generally accepted that the diarrheal syndromes are not caused by the preformed toxins, but by the production of toxins by B. cereus in the small intestine of the host (Notermans & Batt, 1998). Compared with the diarrheal toxins, the emetic toxin is thought to be more dangerous as it has been associated with life‐threatening acute conditions, such as fulminant liver failure and rhabdomyolysis (Mahler et al., 1997). The emetic toxin is extremely stable consisting of small dodecadepsipeptide named cereulide (Melling & Capel, 1978; Agata et al., 1994, 1995a,b), which acts as a potassium ionophore (Agata et al., 1994). This toxin is unique among enterotoxins since it is resistant to proteolytic ­degradation, extreme pH and high temperature, surviving 121°C for 90 min (Granum & Lund, 1997). The production of toxin is maximal during the late exponential to stationary growth phases (and may be associated with sporulation) at optimal ­temperatures of 25 to 30°C, but minimal below 8°C or above 40°C (Drobniewski, 1993; Häggblom et al., 2002). The major type of diseases caused by B. cereus varies from countries to countries (Granum & Lund, 1997). For example, ten times more emetic type is reported in Japan than diarrheal type, whereas in Europe and North America, more diarrheal type is



Incidences of Mould and Bacterial Toxins in Dairy Products   21

reported (Granum, 2005). The majority of B. cereus emetic food poisoning cases have been attributed to rice and rice dishes (e.g. risotto, curries, rice salad), when the rice is usually left at room temperature, spores germinate and vegetative cells produce toxins at room temperature (Kramer & Gibert, 1989; Drobniewski, 1993). However, other types of food, including vanilla slices, pasteurised cream, milk pudding, chicken supreme, reconstituted infant formula and cooked pasta, have also been implicated (Kramer & Gibert, 1989). This may be due to the addition of protein in the form of meat or egg, which enhances the production of toxins in food, and subsequent cooking is usually inadequate to inactivate the toxins (Drobniewski, 1993). Bacillus cereus is of some importance for dairy industry. Since it can produce endospores, which survive pasteurisation, affecting the quality of pasteurised food, such as milk and dairy products (Dzieciol et al., 2013). Psychotropic strains are an important cause of spoilage in cold‐stored dairy products because they are able to  grow and produce toxins during storage at refrigerated temperature (Svensson et al., 2006). Despite the fact that B. cereus is a common contaminant in milk and dairy products, reports on the illness are rare since the symptoms of B. cereus infection are usually mild (Dierick et al., 2005), and only large scales of food poisoning attributed to B. cereus are reported (Granum & Lund, 1997). A large outbreak of food‐borne d­ iseases caused by milk containing B. cereus in Japan was reported, in which both syndromes of emesis and diarrhoea were noted among the patients (Shinagawa, 1993). Some strains of B. cereus may be able to produce diarrheal toxins at reduced temperature (Christiansson et  al., 1989; Fermanian et al., 1996), and some other strains, which produce emetic toxin, have been recovered from dairy products (Beattie & Williams, 1999). Some B. cereus strains haven shown to grow and produce emetic toxins in skimmed milk at 30°C (Szabo et al., 1991; Finlay et al., 1999), whereas some psychotropic strains have also been shown to grow at low temperatures. For example, of the 40% of skimmed milk samples in the Netherlands that were shown to be contaminated with B. cereus, 53% of the strains tested were able to grow at 7°C, suggesting that strains can adapt to environmental conditions in milk and produce enterotoxins that may be a health hazard (Te Giffel et al., 1997). In another study of 125 samples of skimmed milk powder (SMP), white soft cheese, processed cheese, Kareish cheese and rice with milk, 39.2% of the samples were contaminated with B. cereus. Both local isolated and reference strains of B. cereus were examined for toxin production in sterilised milk at 10°C. Detectable amount of toxins was produced in sterilised milk at the end of the storage period when the two tested strains were inoculated at 105 colony forming units (cfu) mL−1 milk (Sadek et al., 2006). This was in agreement with another study which has shown that psychrotrophic strains of B. cereus could produce enterotoxin during growth in milk at refrigeration temperature, but high bacterial numbers (>1 × 107 cfu mL−1) were required before detectable levels of toxin were synthesised (Baker & Griffiths, 1995). Also, in an  outbreak of nausea and vomiting in the Netherlands associated with B. cereus in pasteurised milk, strains isolated were able to grow at 4°C; however, the emetic toxin production were not tested in this study (van Netten et al., 1990). Yet, in another study by Finlay et al. (2000), they have reported that significantly higher emetic toxins were produced at 12 or 15°C than 30°C, albeit over longer incubation periods. No detectable

22   Microbial Toxins in Dairy Products

Table 2.1  Growth and toxin production by Bacillus cereus nc7401 in various food samples at 30°C for 24 h. Food Egg and its products Milk Milk shake Soy milk Soy milk shake Soybean curd

Viable count (colony forming units (cfu) g−1)a

Toxin titer (ng g−1)b

3.8 × 107 ± 4.4 x 102 9.1 × 107 ± 4.4 x 10 4.3 × 108 ± 1.3 x 10 7.7 × 107 ± 3.9 x 10 2.9 × 108 ± 1.9 x 10 3.8 × 106 ± 1.5 x 102

NDc –10 ND – 10 640 ND – 10 320 ND – 20

 Each value is a mean ± standard deviation (SD).  Each value is an average of five samples. c  ND: not detected (75% of outbreaks), followed by SE D, C and B. Outbreaks with SE E are very rarely reported (Vernozy‐Rozand et al., 2004). Regarding the toxic dose, most of the studies are referred to SE A. It had been suggested that the amount of SEs necessary to cause symptoms in humans is about 100 to 200 ng (Bergdoll, 1991). Other authors estimated that the amount of SE A needed to cause vomiting and diarrhoea was 144 ng, and such amount was recovered from 0.28 L carton of a 2 g 100 g−1 chocolate milk (Evenson et al., 1988). In a SFP in Japan, the total intake of SE A in low‐fat milk per capita was e­ stimated mostly at ~20–100 ng (Asao et al., 2003). Such level was slightly lower than the chocolate milk outbreak reported by Evenson et al. (1988), which was mainly attributed to uneven ­distribution of SE A in the low‐fat milk. Numerous biological tests and immunoassays are available to detect SEs (Dolman & Wilson, 1940; Ewald, 1988; Park et al., 1996; Normanno et al., 2001; Vernozy‐Rozand et al., 2004; Bennett, 2005). Molecular biology methods, such as polymerase chain reaction (PCR), are able to detect the genes encoding for the SEs (Johnson et al., 1991; Mehrotra et al., 2000; Fischer et al., 2007; Zouharova & Rysanek, 2008; Ertas et al., 2010; see also Chapter 7). The presence of enterotoxigenic staphylococci is normally associated with meat, poultry or their products (Wieneke et al., 1993). Staphylococcal enterotoxins in dairy products, such as raw milk and cheese, have frequently been associated with food poisoning (Simeão do Carmo et al., 2002; LeLoir et al., 2003; Jorgensen et al., 2005b; Akineden et al., 2008). Raw cow’s milk has generally shown to have a low incidence (2.6 to 9.4%) of enterotoxin producers (Casman & Bennett 1965; Casman et al., 1967; Harvey & Gilmour, 1985). Regardless, outbreaks associated with raw cow’s milk have been occasionally reported. In a large‐scale study collecting 440 bulk tank milk samples from 298 dairy herds, 70 were proven to be positive for Staph. aureus (15.9%). Staphylococcal enterotoxins genes were detected in 39 isolates (55.7%), and the most commonly detected ones were sei (38.6%), seg (31.4%) and sea (27.1%). Enterotoxin genes of seb, seh, sed, sej and sec were observed in 10%, 4.3%, 2.9%, 2.9% and 1.4% of strains, respectively. Toxin production, however, was observed in 9 (12.9%) Staph. aureus isolates, where 7 strains were detected producing SE B (10%) and 2 as SE D (2.9%) (Zouharova & Rysanek, 2008). In 2003, another outbreak in Norway was associated with a mashed potato product made with raw cow’s milk c­ ontaining Staph. aureus that subsequently produced sufficient SE H to cause food poisoning (Jørgensen et al., 2005a). Raw cow’s milk cheeses have shown high frequency (2.8 to 5.4%) of contamination (Thatcher et al., 1959, Zehren & Zehren, 1968). Normally, if the levels of Staph. aureus in the cheese batch exceed 105 cfu g−1, the batch is required to be tested for SE as required by food safety criteria and withdrawn or recalled from the market if p­ resent (EU, 2005). In most cheeses, highest levels of Staph. aureus will be reached 2–3 days after production, and may be reduced significantly during storage. If the levels exceed 105 cfu g−1 at any point, there will be a significant high risk of enterotoxin p­ roduction

26   Microbial Toxins in Dairy Products

that will remain in the cheese regardless of the remaining recoverable level of this organism (Little et al., 2008). Sheep’s milk is usually used for cheesemaking, and in many cases without a prior heat treatment (Bautista et al., 1988). In a study by Bautista et al. (1988), 78 out of 124 staphylococcal strains isolated from sheep’s milk were found to produce SE A to D, with SE A (44 out of 78) and SE D (43 out of 78) showed the highest incidence (Bautista et al., 1988). Staphylococcal intoxication after consuming sheep’s cheeses have also been reported in France (Buyser et al., 1985) and the United Kingdom (UK) (Wieneke & Gilbert, 1987; Bone et al., 1989). In a study analysing 150 samples (consisting of 50 dairy dessert samples and 100 sheep’s milk cheeses), the enterotoxin genes (sea, seb, sec and/or sed) were found in 13 (3.02%) out of 80 Staph. aureus isolates. Also, enterotoxin genes sea, seb and sed were detected in 5 (1.6%), 2 (0.6%), 1 (0.3%), respectively, from cheese isolates, whereas sea, sec and sed were detected in 3 (2.3%), 1 (0.76%), 1 (0.76%), respectively, from dairy dessert isolates. The presence of SEs was identified in 12 (2.8%) out of 80 isolates by using enzyme‐linked immunosorbent assay (ELISA) technique. It was determined that these SEs had a ­distribution of 7 (1.6%) SE A, 2 (0.46%) SE B, 1 (0.23%) SE C, and 2 (0.46%) SE D. Staphylococcal enterotoxins were found in 7 (2.3%) cheese and 5 (3.8%) dairy dessert isolates (Ertas et al., 2010). The incidence of Staph. aureus and their SEs reported in the literature varies widely, which might be due to the differences in the reservoir in the various countries or ecological origin of strains, the sensitivity of detection ­methods, detected genes, as well as number and types of samples included in these studies (Ertas et al., 2010).

2.2.3  Botulinum neurotoxins produced by Clostridium botulinum Food‐borne botulism is a severe type of intoxication resulting from the consumption of pre‐formed botulinum neurotoxins (BoNTs) in food (Peck et al., 2008). They are among the most lethal toxins known to exist because as little as 30 ng neurotoxins, which could be in just a few micrograms in food, are sufficient to cause serious illnesses or even death (Lund & Peck, 2000). For example, in 2002, a 32‐year‐old man ate a mouthful of foil‐ wrapped baked potatoes, found it to be foul tasting and spat it out, but he had consumed sufficient neurotoxin to require extensive medical treatment that included more than 6 months in hospital (Bhutani et al., 2005). Symptoms from food‐borne ­botulism appear several hours to within a few days (range from 2 h to 8 days) after consumption of contaminated food (Arnon et al., 2001). The rapidity of onset of symptoms and severity of illness may depend on both the toxin serotype and the amount of toxin ingested (Woodruff et al., 1992; Arnon et al., 2001). Intoxication with BoNT can result in a neuroparalytic disease, and symptoms of botulism often commence with blurred vision. Flaccid paralysis of the respiratory muscles can result in death (Hatheway, 1988). In the absence of medical intervention, mortality rates of BoNT intoxication could be as high as 50% to 60% (Shapiro et al., 1998; Nishiura, 2007). Clostridium botulinum is a gram‐positive spore‐forming anaerobic rod bacterium, which produces a potent neurotoxin (Brown, 2000). The spores are thermal resistant,



Incidences of Mould and Bacterial Toxins in Dairy Products   27

and can survive in foods that are incorrectly or minimally processed (LeLoir et al., 2003). Studies have shown that C. botulinum are able to grow and produce toxins at pH values lower than 4.6 (Raatjes & Smelt, 1979). Therefore, the acidic, anaerobic conditions, as well as certain combinations of storage temperature and preservatives used in food preservation and canning, may encourage the growth of the bacteria, germination of spores, and with subsequent BoNT production in products (Dembek et al., 2007; WHO, 2014). This happens most often in lightly preserved foods and in inadequately processed, home‐canned or home‐bottled foods (WHO, 2014). There are seven types of botulinum neurotoxins, which are designated by the letters A through G (BoNT A to G), with the neurotoxin formed being dependent on the producing organism (Peck, 2006). They are structurally related, but antigenically distinct large protein neurotoxins (Sharma et al., 2005, 2006). Three of these types (BoNT A, B, and E) are the ones most frequently associated with botulism in humans, whilst BoNT F has only been associated with several outbreaks among humans, and BoNT G has never been linked to human botulism (Sharma et al., 2006). In addition, BoNT C and D can cause animal botulism, and some animals, including cattle, are also ­susceptible to BoNT A, B, or E (Lindström et al., 2010). The BoNT complex is formed by the synthesis of a botulinum neurotoxin and, depending on the serotype, 2–6 nontoxic neurotoxin associated proteins (NAPs), which are known to protect the BoNTs from the low pH environment and proteases of the gastrointestinal tract (Sakaguchi, 1982; Singh et al., 1995; Sharma & Singh, 1998, 2000). The exact toxicity of BoNTs in humans is unclear. However, animal studies using nonhuman primates and from cases of human botulism, it is estimated that the lethal oral dose of BoNT is between 10 ng and 1 μg kg−1 body weight (Herrero et al., 1967; Lund, 1990). In general, the lethal toxicity is dependent on the serotype of BoNT and the route of exposure (Arnon et al., 2001; Cheng et al., 2008), and it may vary among individuals (Weingart et al., 2010). Foodborne botulism is a severe and deadly disease, and outbreaks in most countries have been associated with homemade foods, where control measures have not been properly implemented (Peck, 2006). Outbreaks involving commercial foods and ­restaurants more rarely occur (Peck et al., 2008). Two groups of C. botulinum are responsible for most cases of food‐borne botulism: (a) C. botulinum group I (proteolytic strain), and (b) C. botulinum group II (non‐proteolytic strain) (Peck, 2006). Proteolytic C. botulinum is a mesophile, which has a minimum growth temperature of 10–12°C, and forms BoNT A, B or F (or in some cases two toxin types). Non‐proteolytic C. botulinum is a psychrotroph that grows and produces toxin at 3°C, and forms BoNT B, E or F (Lund & Peck, 2000), and it is of a greater concern in foods that are intended to be stored chilled. Owing to the fact that only an exceedingly small quantity of BoNTs is needed for intoxication (60

Time to toxigenesis (days)

(Continued)

(Billon et al. 1980) Franciosa et al. (1999) Glass & Johnson (2004)

Kaufmann & Brillaud (1964) Wagenaar & Dack (1958)

Read et al. (1970) Glass et al. (1999)

References

5.7 5.9 6.0 5.7 5.7

1.5g

1.9h 2.0i 2.7g 2.7g

5.6

4.0e

5.7–6.2

5.6

4.0e

0–2f

5.8 5.7

pH

2.7e 4.1e

Sodium chloride (NaCl – g 100 g−1)

0.96 0.96 NR NR

0.97

NR

0.96

0.97

0.98 0.96

Water activity (aw)

58 58 50–69 50–69

56

52–56

62

62

60 62

Moisture (g 100 g−1)

I/A, B (103) I/A, B (103) I/A, B (103) I/A, B (101–102)

I/A (102–104)

I/A, B (103)

I/A, B (103)

I/A, B (103)

I/A, B (103) I/A, B (103)

Spore inoculum (group/ type, number of spores g−1)

b

a

30 30 25 25

30

30

30

30

30 30

Storage temperature (°C)

 NR: not reported.  Physiological group has not been reported. c  Cheese matured with yeast and bacteria. d  The spores were inoculated onto the straw bedding in contact with un‐inoculated cheese. e  Total salt of NaCl plus disodium phosphate. f  Added of sodium chloride (NaCl) (g 100 g−1); the cheese mass also contained disodium phosphate (2.0–2.5 g 100 g−1). g  Added NaCl (g 100 g−1); the use of emulsifiers is not reported. h  Total salt of NaCl plus sodium citrate; cheese mass contained also 0.2% lactic acid. i  Total salt of NaCl plus di‐ and tri‐sodium phosphate or sodium citrate; cheese mass contained also 0.2% lactic acid. Adapted from Lindström et al. (2010).

Processed cheese spread

Processed cheese spread (fat‐free, 0.8

60 50 39

52 31 28.21

2006 2006 2007–2008

72 193 132

60 82.4 82.6

0.16–7.26 0.10–5.20 0.25

2001–2002 2002–2003 2004

 NR = not reported.

a

100 82.5 76.6 81.9 60 100

Range (μg kg−1)

0.111–0.413 0.15–2.41 0.052–0.785 0.030–1.200 0.041–0.374 0.032–0.506

References

Prado et al. (2000) Brezina et al. (1983) El‐Sayed Abd Alla et al. (2000) Amer & Ekbal Ibrahim (2010) Piva et al. (1988) Zerfiridis (1985) Piva et al. (1988) Barbieri et al. (1994) Pietri et al. (1997) Minervini et al. (2001) Manetta et al. (2009) Kamkar (2006) Fallah et al. (2009) Fallah (2010a) Tavakoli et al. (2012) Nilchian & Rahimi (2012) Dashti et al. (2009) Elkak et al. (2012) Elgerbi et al. (2004) Iqbal & Asi (2013) Barrios et al. (1996) Yaroglu et al. (2005) Sarımehmetoglu et al. (2004) Tekinşen & Tekinşen (2005) Gürbay et al. (2006) Aygun et al. (2009) Ardic et al. (2009) Tekinşen & Eken (2008) Kav et al. (2011) Ertas et al. (2011) Oruc & Sonal (2001)

Incidences of Mould and Bacterial Toxins in Dairy Products   41



Table 2.7  Incidence and level of aflatoxin M1 in different dairy products. Country

Iran

Italy

Year

Number of incidences per surveyed samples (%)

Mean (μg L−1)

NRa

Yoghurt Butter Ice‐cream Yoghurt

45/68 (66.1) 8/31 (25.8) 25/36 (69.4) 14/40 (35)

0.032 0.005 0.041 0.1305

1996

Yoghurt

73/120 (61)

0.0091

2003

25/25 (100) 25/25 (100) 16/48

0.034 0.18 0.05112

33/74 (45) 59/96 (61) 25/27 (92.6)

0.156 0.147 NR

2009

Product

Pakistan

2010–2011

Turkey

2002–2003

Whey Curd Yoghurt flavoured with strawberries Butter Yoghurt Butter

2005

Butter

92/92 (100)

0.236

2010

Yoghurt Dairy desserts Butter

28/50 (56) 26/50 (52) 3/10 (30)

0.0303 0.0262 0.057

NR

References

Fallah (2010a)

Nilchian & Rahimi (2012) Galvano et al. (2001) Manetta et al. (2009)

Iqbal & Asi (2013) Aycicek et al. (2005) Tekinşen & Uçar (2008) Ertas et al. (2011) Var & Kabak (2009)

 NR: not reported.

a

of AF M1 found in dairy drinks might be less than that of the milk being used for the process, which may be due to the addition of many other food ingredients to the milk during preparation of these products (Iha et al., 2011), and/or other factors, such as low pH, formation of organic acids or other fermentation by‐products, or even to the presence of lactic acid bacteria (Govaris et al., 2002). The low pH during fermentation may alter the structure of milk proteins (mainly the caseins), which may affect the association of AF M1 with these proteins causing adsorption or occlusion of the toxin (Brackett & Marth, 1982c). Low levels and frequency of AF M1 contamination in butter samples have been detected in some studies (Var & Kabak, 2009; Fallah, 2010a), but high levels of AF M1 in butter have been reported by Tekinşen & Uçar (2008) and Iqbal & Asi (2013). During buttermaking, the protein component around the fat globule membrane is broken and, as a consequence, the serum phase is separated. Due to the insolubility of AF M1 in butter and its affinity to casein, it binds to this fraction of protein. Hence, ­butter may contain less amount of AF M1 comparing to the other dairy products (Bakirci, 2001). Albeit, the variations in the reported amounts of AF M1 in butter could be attributable to the differences in the levels of AF M1 in the milk used for processing according to the seasonal variation (Galvano et al., 1996; Bakirci, 2001; Sarımehmetoglu et al., 2004; Tekinşen & Tekinşen, 2005), the different processing techniques and analytical methods used (Tekinşen & Uçar, 2008). Moreover,

42   Microbial Toxins in Dairy Products

d­ ifferences in the hygiene and storage conditions in dairies and retail outlets are other key factors for the variations of the AM M1 content in butter (Wiseman & Marth, 1983a; Galvano et al., 1996).

2.3.2  Sterigmatocystin Other than aflatoxins, Aspergillus spp. are able to produce mycotoxins known as ­sterigmatocystin (STC), which has been found to be produced naturally in dairy products, such as cheese (Northolt et al., 1980). Sterigmatocystin is a fungal secondary metabolite produced by many aspergilli species, such as Aspergillus versicolor, Aspergillus chevalieri, Aspergillus ruber, Aspergillus amstelodami, Aspergillus aureolatus, Aspergillus quadrilineatus and Aspergillus sydowi (Rabie et al., 1977; Abdell‐ Mallek et al., 1993; Lund et al., 1995; Reijula and Tuomi, 2003). In addition, there are also other fungal genera that are able to produce this mycotoxins, such as Penicillium, Bipolaris, Chaetomium, Emiricella (Holzapfel et al., 1966; Schroeder & Kelton, 1975; Davis, 1981; Terao, 1983; Frisvad & Samson, 2004; Frisvad et al., 2005; Shanawany et al., 2005). Among these, A. versicolor is the main producer for STC. It is generally xerophilic, which means it can grow at low aw (180/120 mmHg). This can cause damage to the heart or central nervous system (Blackwell, 1963). Tyramine toxicity has also been related to an increased risk of carcinogenesis since it can react with dietary nitrites to form 3‐diazotyramine, which induces oral cancer in rats (Fujita et al., 1987; Shalaby, 1996). Tyramine also affects the adherence of some bacteria to the intestinal epithelial cells (Fernandez de Palencia et al., 2011) and, certainly, it enhances the adherence of enteropathogenic E. coli O157:H7 to these cells (Lyte, 2004). In addition, tyramine has been shown to be cytotoxic in an in vitro model of human intestinal epithelium (Linares et al., 2016).

4.4.2  Histamine Histamine intolerance, enteral histaminosis and scombroid poisoning are terms used to describe a group of specific adverse reactions that occur after the ingestion of histamine‐ rich food. Scombroid poisoning refers to the illness that can follow the consumption of fish belonging to the suborder Scombroidea, such as tuna and mackerel (Ladero et al., 2010a). After ingesting a food with a high histamine content, the toxin comes into contact with detoxifying enzymes (e.g. DAO) produced by the intestinal epithelium, However, this may not be able to inactivate all the histamine present, and some may enter the systemic circulation and trigger adverse neurological (headache, migraine), gastrointestinal (abdominal cramps, nausea, flatulence, vomiting and diarrhoea), circulatory (hypotension and palpitations) and respiratory (bronchospasms and respiratory distress) symptoms, as well as flushing, rash, and urticaria (Shalaby, 1996; Ladero et al., 2010a; Stratta & Badino, 2012). Symptoms can start any time between a few minutes and a few hours following the consumption of histamine‐rich food (Lehane & Olley, 2000). It is important to highlight that histamine poisoning is not an allergy, although the symptoms resemble those seen in allergic reactions (Stratta & Badino, 2012) since histamine is the main inducer of allergic responses. This confusion can lead to an incorrect diagnosis of histamine intolerance and, thus, the under‐estimation of histamine intoxication outbreaks (Schwelberger, 2009; Ladero et al., 2010a).

110   Microbial Toxins in Dairy Products

When a healthy person ingests food with a low concentration of histamine (400 mg kg–1 is considered dangerous to health (Taylor, 1985), and certainly a dose of 1000 mg can cause severe intoxication (Rauscher‐Gabernig, 2009). Impairment of histamine metabolism (genetic or otherwise) seems to play an important role in the severity of histamine poisoning. The reduction of histamine metabolism is mainly a problem of reduced DAO and/or N‐methyltransferase activity; this can lead to histamine intolerance and the clinical symptoms detailed above (Maintz & Novak, 2007). Some drugs used for the treatment of depression, Alzheimer’s and Parkinson’s disease, and even antibiotics and agents used to reduce intestinal motility, can behave as DAO inhibitors (Maintz & Novak, 2007; EFSA, 2011). By reducing the catabolism of extracellular histamine the concentration in plasma increases, bringing on intoxication symptoms (Sattler et al., 1988). A reduction in DAO activity has been associated with inflammatory diseases, such as Crohn’s disease (Schmidt et al., 1990) and ulcerative colitis (Mennigen et al., 1990), and with colorectal neoplasms (Raithel et al., 1998; Maintz & Novak, 2007). In some patients with reduced DAO activity, reduced N‐methyltranferase activity has also been reported; this can render a person almost completely unable to degrade histamine (Kuefner et al., 2004). As mentioned earlier, putrescine and cadaverine can potentiate the toxicity of histamine (Lehane & Olley, 2000) since they impair its catabolism by competitive ­inhibition of DAO (Mongar, 1957; Shalaby, 1996; Lehane & Olley, 2000). These BA are usually found in high concentrations in histamine‐rich cheeses (Fernandez et al., 2007a; Ladero et al., 2010a; Linares et al., 2011). Tyramine, tryptamine and β‐phenylethylamine are also strong inhibitors of DAO and histamine N‐methyltransferase, and can, therefore, also enhance histamine toxicity (Taylor & Lieber, 1979; Lehane & Olley, 2000).

4.4.3  Putrescine and polyamines Putrescine is usually referred to as a polyamine, as are spermidine and spermine. Their physiological and toxicological effects are usually studied and described together since these compounds are interconvertable (Pegg & Casero, 2011). Polyamines are polycationic compounds essential for optimal cell growth rates in all organisms, but dietary polyamines have been described to have non‐direct toxic effects; although they can increase cardiac output, which could lead to tachycardia or hypotension (Ladero et al.,



Biogenic Amines in Dairy Products   111

2010a; Wunderlichova et al., 2014). Certainly, the extracellular content of polyamines and their metabolising enzymes are strongly associated with the proliferation of neoplasms in the gastrointestinal tract and in the promotion of malignancy (Seiler et al., 1998; Gerner & Meyskens, 2004). In fact, patients with familial adenomatous polyposis are recommended to reduce their dietary putrescine intake, and studies performed in mice have shown that dietary putrescine increases the malignancy grade of adenomas (Ignatenko et  al., 2006). In addition, alterations in the intracellular polyamine concentration has been observed in colorectal cancer cells (Wallace & Caslake, 2001), indicating the potential importance of exogenous putrescine in the development of neoplasms (Gerner & Meyskens, 2004). Polyamines are key effectors of the carcinoma caused by micro‐organisms, such as Helicobacter pylori (Alam et al., 1994). A further carcinogenic effect of polyamines is seen in their capacity to react with nitrites and produce nitrosamines, compounds known to be carcinogenic (Shalaby, 1996; Coneski & Schoenfisch, 2009). This is of particular importance with respect to some BAs‐rich fermented products in which nitrites and nitrates are added as preservatives (Ruiz‐ Capillas & Jimenez‐Colmenero, 2004). The technological process to which foods are subjected (e.g. frying or toasting) is also important in this respect since, even though the raw food may be free of nitrosamines, heat promotes the formation of nitrosamines from the polyamines present in it (Shalaby, 1996). Polyamines can also enhance the virulence of pathogens (Shah & Swiatlo, 2008; Pegg & Casero, 2011), such as Yersinia pestis (Patel et al., 2006) and Vibrio cholerae (Karatan et al., 2005), in which they are involved in biofilm formation and pathogenicity. In Proteus mirabilis, exogenous putrescine can activate the swarming phenotype needed for pathogenesis (Sturgill & Rather, 2004).

4.4.4  Cadaverine, tryptamine and β‐phenylethylamine Non‐direct toxic effects have been described for cadaverine, tryptamine and β‐­ phenylethylamine ‐ certainly they can increase the toxic effects of the other BAs. Moreover, these BAs are usually found in high concentrations in histamine‐rich ­fermented ­products (Spano et al., 2010; Linares et al., 2011), thus potentiating the risk of ­histamine intoxication. They can also react with nitrites added to food to form ­nitrosamines, compounds with known carcinogenic activity (Shalaby, 1996; Coneski & Schoenfisch, 2009).

4.4.5  Recommended limits of BAs Although the toxicity of BAs is beyond doubt, no clear maximum limits for them have been established in foodstuffs that would help to ensure the safety of consumers. The establishment of such limits is very difficult since the toxic effect of each BA is different and depends on the combination of other factors, such as the presence of other toxicity enhancers (inhibitors of MAO and DAO, alcohol, tobacco), individual susceptibility, health status, and individual efficiency in terms of BA catabolism (EFSA, 2011). In addition, the absorption, metabolism and, hence, the toxicity of one BA might be

112   Microbial Toxins in Dairy Products

modified by the presence of another, for example, in cheeses with high concentrations of several BAs (Table 4.1). This might explain the increased toxicity of much matured cheeses compared to the toxicity of an equivalent dose of histamine in aqueous solution (Taylor, 1986). Most food safety agencies have only set maximum histamine limits, and then only for fish and fish products. However, the European Food Safety Agency (EFSA) has established a maximum histamine concentration of 100–200 mg kg–1 in fresh fish, and 200–400 mg kg–1 in cured fish products (Commission Regulation 2073/2005 ‐ EU, 2005). The EFSA reported a “no‐observed adverse‐effect‐level” (NOAEL) for histamine in healthy volunteers of 25–50 mg kg–1 food, but for people with histamine intolerance even very small amounts caused health issues. Thus, only histamine levels below detectable limits can be considered completely safe (EFSA, 2011). However, the EFSA recommendations indicate that fish products containing 1000 mg kg–1 as toxic and unsafe (EFSA, 2011). The European Commission has suggested that a maximum of 300 mg kg–1 for total BA in fish and fish products may be an appropriate limit. Recently, the Food and Drug Administration (FDA) of the United States of America (USA) established a guidance of histamine concentration of 50 mg kg–1 for the consumption of scombroid or scombroid‐like fish (FDA, 2011), and >500 mg kg–1 is considered to be a threat to human health. This agrees with reports indicating that the ingestion of food with histamine contents of >500 mg kg–1 can cause histamine intoxication (Ladero et  al., 2010a). A legal vacuum exists, however, with respect to the regulation of threshold limits for histamine in other histamine‐rich foods, such as cheese, wine and other fermented products (Konakovsky et al., 2011). Switzerland temporally introduced a legal histamine threshold of 10 mg L–1 for wine, although it was abandoned in 2008 when regulations were adjusted to current European Union (EU) standards (Konakovsky et al., 2011). No legal threshold limit exists for tyramine. Some studies show it to be toxic at >25 mg kg–1, although the concentration needed to cause a toxic reaction in persons undergoing monoamine oxidase inhibitor (MAOI) treatment is much lower (McCabe, 1986). Patients receiving such treatment should reduce their ingestion of foods ­potentially rich in tyramine (cheese and other fermented foods). According to EFSA there is currently insufficient scientific evidence to set a NOAEL for tyramine in humans (EFSA, 2011). The EFSA considers a tyramine uptake of 600 mg per meal to have no adverse effect on healthy individuals not taking MAOI. A ­maximum of 50 mg of tyramine is advised for people taking third generation MAOI, and only 6 mg per meal in people taking traditional MAOI (EFSA, 2011). Although the toxicity of putrescine, cadaverine, tryptamine and β‐phenylethylamine is generally assumed to be less than that associated with other BAs, their capacity to enhance the toxicity of histamine and tyramine needs to be remembered (Hui & Taylor, 1985), as does their involvement in the formation of carcinogenic nitrosamines (Coneski & Schoenfisch, 2009). Since no human dose‐response data are available for these BAs, the EFSA has established a recommended limit of 180 mg of putrescine and



Biogenic Amines in Dairy Products   113

cadaverine per kg of body weight based on a limited number of animal studies (EFSA, 2011). Few studies have been performed to establish a NOAEL for tryptamine and cadaverine. β‐phenylethylamine toxicity has been detected after the administration of 5 mg to healthy volunteers, whereas as little as 3 mg appear to be sufficient to induce adverse effects in individuals who have reduced MAO‐B activity and suffer migraines (EFSA, 2011). The capacity of these BAs to enhance the toxicity of histamine should also be remembered when histamine‐rich foods are ingested.

4.5  Factors affecting BAs accumulation in dairy products The biosynthesis and accumulation of BAs in dairy foods requires the presence of bacteria with decarboxylase and/or deiminase activities, plus the presence of substrate amino acids (released from milk proteins during fermentation). The final concentration of BAs in a product depends on the conditions for bacterial growth and the expression of the necessary bacterial genes. Knowledge of the conditions that favour the accumulation of BAs in dairy products could help in the implementation of measures designed to prevent it.

4.5.1  Presence of BAs-producing bacteria There is no doubt that BAs‐producing micro‐organisms are required for BAs synthesis and accumulation in dairy products. Thus, measures taken to reduce their presence and growth (e.g. heat treatment of the milk prior to processing, refrigeration) would reduce the BAs concentration in the product. In dairy products, the main BAs producers are LAB that play an essential role in the manufacture of fermented milk products and cheese; however, LAB starter cultures are responsible for the initial rapid conversion of lactose into lactic acid with the concomitant acidification of the milk that facilitates its clotting, and non‐starter lactic acid bacteria (NSLAB) participate in the conversion of proteins, sugars and fats into compounds that provide the final flavour and texture of the product. Some of these LAB can produce BAs, but this is not a general trait of all LAB strains and species (see section 4.3.2). In dairy products made by the addition of a starter culture, the selection of strains unable to produce BAs is vital in any attempt to reduce the BA concentration of the final product. There are several methods available for assessing the capacity of a strain to produce BAs, with some based on differential microbiological media, and some on analytical methods designed to detect the presence of BAs after the addition of the required corresponding precursor amino acid to the growth medium (Bover‐Cid &  Holzapfel, 1999; Garcia‐Moruno et al., 2005). These methods call for specific media and appropriate culture conditions, require the checking of the expression of the c­orresponding genes, and entail the monitoring of the activity of enzymes and ­transporters ‐ such analyses are not always possible. Culture independent methods based on PCR have, therefore, come to the fore as means of determining the potential capacity of bacteria to produce BAs and, thus, for deciding upon the suitability of starter cultures (Landete et al., 2007; Ladero et al., 2015).

114   Microbial Toxins in Dairy Products

Several authors have postulated that a minimum number of BAs‐producing micro‐ organisms need to be present in the food matrix to pose a risk of harmful BAs accumulation (Joosten & Northolt, 1987; Ladero et al., 2008). The development of methods that would allow the quantification of these micro‐organisms ­during manufacture would be useful when examining measures designed to reduce BAs accumulation. The qPCR is a main contender (Martinez et al., 2011) and, indeed, several qPCR methods have been developed for identifying and  ­quantifying BAs‐producers in dairy products (Ladero et al., 2010b, 2012a; La Gioia et al., 2011). Other BAs‐producing micro‐organisms can enter a product as contaminants from the environment during manufacture or storage. For example histamine‐producing LAB were shown to contaminate cheese products during processing and after maturation as grating or slicing (Ladero et al., 2009). In addition, Gram‐negative BAs‐producers, such as Enterobacteriaceae can contaminate products if hygiene conditions are poor, but in general their impact on the accumulation of BAs in dairy products is very low and limited to such scenarios.

4.5.2  Physiochemical factors The pH, temperature and salt concentration are important in the manufacture of dairy products, but they also have an impact on BAs accumulation since they can influence the growth of BAs‐producing bacteria, their gene expression, and the activity of their enzymes and transporters. Although some authors have noted that good bacterial growth conditions frequently go hand in hand with tyramine production (Gardini et al., 2001), others have reported the factors most influential in microbial growth not to greatly affect tyramine production (Connil et al., 2002b; Marcobal et al., 2006b). This might be related to the fact that stress conditions induced by low pHs, high salt concentrations, or a low temperature that can reduce the growth capacity of BAs‐producers, but might increase the production of some BAs via the induction of BA gene clusters as a means of defence (Wolken et al., 2006; Linares et al., 2009; Fernandez de Palencia et al., 2011). pH Dairy fermentations require a low pH (4.6) for milk to clot. Such acidity also helps prevent microbial spoilage. However, since the BAs‐producers in dairy products are mainly LAB, they are able to grow under these conditions. Moreover, BAs synthesis may act as a mechanism for survival under acid stress conditions, although this is still poorly understood. Trip et al. (2012) showed that the heterologous expression in Lac. lactis subsp. cremoris of the histidine decarboxylase cluster of Str. thermophilus enabled cells to survive at pH values of as low as 3 for at least 2 h ‐ conditions under which the wild type Lac. lactis subsp. cremoris host cells were rapidly killed (see section  4.3.3). In LAB, histamine production is more intense at low pHs, a process that seems to be mediated by greater enzymatic activity of HdcA (Coton et al., 1998; Landete et al., 2008; Schelp et al., 2001) rather than any induction of the HDC genes (Landete et al., 2008). The largest studies on pH and BAs have examined its effect on tyramine production. Moreno‐Arribas & Lonvaud‐Funel (2001) showed that the optimum pH of TdcA from



Biogenic Amines in Dairy Products   115

Lb. brevis was 5.0, and that it had an activity range over pH 2.0 to 9.0. The most c­ omplete studies on the influence of pH in TDC gene cluster expression have been performed on Ent. faecalis V583 as a model strains and on Ent. Durans 655, that is, a dairy strain (Linares et al., 2009; Perez et al., 2015) in which a significant induction of the tyrosine decarboxylase and antiporter genes were demonstrated at low pH. The greater expression of the encoding genes and the greater activity of TdcA may explain the greater tyramine production of Ent. durans at pH 5.0 compared to pH 6.8 (Fernandez et  al., 2007b). Similarly, La Gioia et al. (2011) reported a slight increase in expression at low pH in Str. thermophilus, although in this case pH was not the only impacting condition ‐ depletion of the main carbohydrate source was also related to the increase in gene expression. These authors suggest that a combination of stress conditions, more than single factors, such as pH, participate in the induction of the Str. thermophilus TDC gene cluster. In the production of putrescine via the AGDI pathway, ammonium ions are another main product (Fig. 4.4); a low pH and a low pH resistance mechanism might, therefore, play important roles in BAs production. However, the AGDI of L. monocytogenes shows its greatest activity at pH 7.5 (Cheng et al., 2013). Moreover, conflicting data have been reported regarding the influence of pH on AGDI gene induction. For example, Suarez et al. (2013) showed that the AGDI cluster in Ent. faecalis not to be induced by low pH, while the same conditions have been described to induce it in Streptococcus mutans (Griswold et al., 2006) and L. monocytogenes (Chen et al., 2013). It should be remembered that great diversity in the nucleotide sequence of genes exist between the different clusters, and even within clusters of the same genera and species. This might be related to the origin of the BAs‐producing pathway ‐ some traits are species‐specific, while others have only been recently acquired by horizontal gene transfer and may not be fully integrated into the general metabolism and gene regulation network of the acquiring strain. In conclusion, the acidic environment typical of dairy products offers perfect conditions for the accumulation of BAs. Unfortunately, this is very difficult to modify since these conditions are intrinsic to the fermentation process, which help to ensure microbiological safety by inhibiting the growth of most pathogens and spoilage micro‐organisms, facilitate milk clotting, and influence the flavour of the product. Salt concentration The main role of salt in cheese production is to act as a preservative by reducing water activity values. Undesirable pathogens and spoilage micro‐organisms are either inactivated or their growth is limited, and the chloride ions inhibit the germination of microbial spores. It also helps control the growth of LAB during the maturation stage of the product. Salt is also a component of the expected cheese taste and acts as a flavour enhancer. With few exceptions, the salt content of cheese is 0.5–2 g 100 g–1. Cheeses, such Blue Vein varieties and Feta, can have salt ­contents of 3–7 g 100 g–1. Salt concentrations of 0.99).  Aerobic (anaerobic 0.92–>0.99). Data compiled from Tatini (1973), Crowther & Holbrook (1980), Baird‐Parker (1990), ICMSF (1996) and SCVMPH (2003). 1 2

guaranteed. In addition, when the total counts in milk are low, Staph. aureus can reach high levels during the early stages of cheesemaking (Meyrand et al., 1998). Despite the possibility of staphylococcal counts decreasing during the maturation period and storage of the cheese, the enterotoxins may persist and be consumed (EFSA, 2003). Some physical conditions that may interact with SE formation include (see Table 9.3):

• •

• •

Natural antimicrobial substances ‐ Even if milk contains some natural antimicrobial substances (lactoperoxidase system, lysozyme, lactoferrin), they are not very effective against Staph. aureus (SCVMPH, 2003). Growth temperature – The optimal growth temperature of Staph. aureus is ~37°C but, in skimmed milk, the maximum temperature is 48.9°C (Stiles & Witter, 1965; SCVMPH, 2003); however, the bacterium can grow at 46°C when protected by 1 M NaCl (El‐Banna & Hurst, 1983; Medeved’ova et al., 2009; Valero et al., 2009). Pasteurisation of the milk (i.e. at 57.2°C for 80 min, 60°C for 24 min, 62.8°C for 6.8 min, 65.6°C for 1.9 min and/or 71.7°C for 0.14 min) inactivates staphylococci species. The D‐value ranges for Staph. aureus in milk are: D62°C of 12–27 s, or D72°C of 1.5 s (Firstenberg‐Eden et al., 1977; Walker & Harmon, 1996) (Table 9.4). Heat treatment of milk at 121°C for 3–8 min is required to inactivate 90% of the other SEs (Stewart, 2003). However, in dairy products, the bacterium may become more heat resistant as the aw is lowered and, at a aw level of 0.70–0.80, the resistance begins to decline (Troller, 1986; Bergdoll, 1989). Thermisation  –  Milk can be thermised at different temperatures (e.g. 57–68°C for  10–20 s), and such heat treatment is not sufficient to reduce significantly the population of vegetative cells of Staph. aureus. Product ‐ In hard and semi‐hard cheeses, the minimum temperatures for SE p­roduction ranges between 14 and 38°C, and the maximum temperatures between 35 and 45°C (Schmitt et al., 1990). Generally, crude SE A and B seems to be considerably more heat resistant than purified SE A (Minor & Marth, 1976), and the thermal inactivation of SEs is often accompanied by loss of the serological activity

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   237

Table 9.4  Effect of the minimum thermisation (62°C for 15 s) and minimum storage/maturation of cheese (2°C for 90 days). Organism Staphylococcus aureus

Effect of thermisation

Effect of storage/maturation of cheese

~1 log10 cycle reduction D62°C of 12–27 s D65°C of 1.8–10 s D72°C of 1.5 s

Largely inactivated in Swiss‐type cheese Staph. aureus (initial concentration of 5 × 105 colony forming units (cfu) mL−1 survived for more than 60 days, but not 90 days in Swiss‐type cheese)

Data compiled from Bachmann & Spahr (1995) and FS AU & NZ (2009).

•

• •

• •

(Bergdoll, 1989). Also, SE A is comparatively more heat stable at pH 6.0, or higher than at pH 4.5‐5.5, and reactivation may occur during the cooking stage and s­torage period (Tatini, 1976). Presence of oxygen ‐ Staphylococcus aureus grows best in the presence of oxygen but, under strict anaerobic conditions, growth is slow. Thus, with slower anaerobic growth, relatively less SE A is produced than under aerobic conditions (Belay & Rasooly, 2002). pH  –  Under both anaerobic and aerobic conditions, most Staph. aureus strains produce detectable amounts of SEs below pH 5.7 and 5.1, respectively. Water activity (aw) – Staphylococcal enterotoxin A and D production occurs slightly under the aw optimal growth conditions of the organism, but the production of SE B and C are very sensitive to reduced aw condition, and hardly any is produced at aw 0.93 (Ewald & Notermans, 1988). Redox potential – The range for bacterial growth lies between  +200 mV redox potential and, at > +200 mV, SE is produced (Tatini, 1973; Crowther & Holbrook, 1980; Baird‐Parker, 1990; ICSMF, 1996). Competitive flora ‐ Staphylococcus aureus does not grow well in milk in the presence of starter cultures because they lower the pH, produce organic acids, or hydrogen peroxide (H2O2), or inhibitory substances like antibiotics, and volatile compounds (Smith et al., 1983; Genigeorgis, 1989) that help to prevent SE formation.

9.3.3  Cheese production and hazard characterisation of Staphylococcus aureus Cheese made from raw or pasteurised milk, the presence of Staph. aureus is a well‐ known health hazard because the product is a good substrate for the organism to grow (Bone et al., 1989; de Buyser et al., 200 l; Lindquist et al., 2002). In raw milk, counts of Staph. aureus vary from 106 cfu mL−1. Similarly in cheese, the prevalence of staphylococci vary from below the limit of detection to a maximum in excess of 106 cfu mL−1 (Ramsey & Funck, 2009; Rosengren et al., 2010). The enterotoxinogenic Staph. intermedius produces SE C, while Staph. hyicus produces SEs other than A to E, producing an emetic response in monkey feeding tests (Adesiyun et al., 1984); however, Staph. intermedius strains p­ roduce toxin liable to cause a possible health hazard (Hirooka et al., 1988).

238   Microbial Toxins in Dairy Products

Staphylococcal enterotoxins A and D are most commonly involved in food poisoning, and are formed during the exponential growth of the bacteria, whereas the other types of toxins are predominantly formed when the bacteria enter the stationary phase (ICSMF, 1996). There are no simple relationships available between the number of bacteria, concentration of the toxins and dose‐response relationships. The severity of the illness depends on level of toxins, the amount of the food ingested, and the general health of the victim (Lindquist et al., 2002; Smith et al., 2004). The amount of enterotoxin, which can cause illness in humans, is generally believed to be in the range 0.1–1.0 µg kg−1, when Staph. aureus counts exceed 105 cfu g−1 (ICSMF, 1996; Ash, 1997). Acute effects follow ingestion of pre‐formed SEs after a short incubation period (1 to 6 h). Normally, the symptoms wear off within 24 h (Murray et al., 2002; Havelaar et al., 2003), and recovery is usually between 1‐3 days, requiring no medical treatment (i.e. self‐limiting). Symptoms appear around 3 h after ingestion, but can occur in as little as 1 h (i.e. self‐limiting), and include copious vomiting, abdominal pain, diarrhoea (ICSMF, 1996). In severe cases, blood and mucus may be observed in the stools and vomit. In Staph. aureus intoxication, a dose of 0.5 µg of SE A causes vomiting; however, an adult has to ingest 100–200 ng of SE A to get symptoms (Bergdoll, 1970; Mossel et al., 1995). Furthermore, the amount of SE A needed to cause vomiting and diarrhoea in contaminated chocolate milk is 0.144 µg in average (Evenson et al., 1988), and in low‐fat milk 20–224 ng as total intake of SE A per capita (Asao et al., 2003). Milk products, which are involved in staphylococcal foodborne disease, could be attributed to the following aspects:

• • •

Occurrence of high counts of coagulase‐positive staphylococci in raw milk; Cross‐contamination during the process; and/or Possible post processing contamination (SCVMPH, 2003).

Staphylococcus aureus is one of main causes of mastitis in a dairy herd. If a farm has a sub‐clinical mastitis problem, the bacteria may multiply rapidly during cheesemaking (SCVMPH, 2003; Paulin et al., 2012). The amount of SE A in milk increased linearly with time at temperatures between 14 and 32°C once the cell population reached 3.2 × 106 cfu mL−1 (Fujikawa & Morozumi, 2006).

9.3.4  Cheesemaking conditions and exposure assessment of Staphylococcus aureus Within existing cheesemaking methods, the exposure assessment is regulated by the processing conditions favouring SE production. According to SCVMPH (2003) in exposure assessment, the effect of processing conditions on Staph. aureus growth and enterotoxin production is related to the different categories of cheeses, that is, fresh, soft, semi‐hard and/or hard.

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   239

Fresh and whey cheeses In fresh cheeses (e.g. Cream cheese, Quesco Blanco, Mozzarella, Crescenza) and also in whey cheeses (Ricottone, Mytzithra, Mysot) prepared from raw or pasteurised milk, the curd is fermented by the ‘wild’ lactic acid bacteria (LAB) present in milk, or the added starter culture. The pH value usually drops rapidly to levels below 5.0, and the moisture content remains >55 g 100 g−1 (aw 0.95–0.97). Thus in these cheese varieties, SEs are produced when the starter culture activity is reduced. Staphylococcal enterotoxin A, and more rarely SE D, has been detected in these cheeses prepared from raw or pasteurised milk, when the count of Staph. aureus reaches 106 cfu g−1 of cheese. In Fresh cheese with a high LAB count, staphylococcal levels decline rapidly during processing but, in whey cheese, as the count reaches >107 cfu g−1, SEs are produced in the absence of lactic fermentation, high aw (0.94–0.96) and high pH (6.0–6.2) of the product (Karaioannoglou et al., 1983). Soft and white brined cheeses Cheese varieties, such as Petit Swiss, Mont d’Or, Kingston, Minas Gerias, are produced from raw milk, thermised milk (65–68°C for 15 min) or pasteurised milk (72°C for 15 s) with the use of selected starter cultures (Scott, 1986; Meyrand & Vernozy‐Rozand, 1999). In soft cheeses made with added starter culture, only a large inoculant of Staph. aureus (105 cfu mL−1 of milk) could result in enterotoxin production. Consequently, pasteurisation of milk and the use of starter culture is the best method to prevent the growth of Staph. aureus and enterotoxin formation. Similarly in white brined cheeses, such as Feta, Brinza, Domiati, staphylococci can only multiply during the first 2–3 h of processing when a large initial inoculant (>103 cfu mL−1) is used (Mantis, 1973; Ahmed et al., 1983). It is evident, however, that soft cheeses made with starter culture, including white brined cheese, represent an unfavourable environment for growth and enterotoxin production by Staph. aureus. Surface white mould and smeared cheeses By contrast, cheese varieties, such as Brie, Camembert and Brick, represent a more favourable environment for Staph. aureus to grow and produce enterotoxin, if the initial number is higher than 103 cfu mL−1 (Mueller et al., 1996, Meyrand et al., 1998). In surface smeared type cheese, Brevibacterium linens and/or Brevibacterium casei grow on the surface of the product during the maturation period and, as a consequence, these organisms produce a slimy smear and alter the pH (i.e. from low to high) of the surface of the product giving the opportunity for pathogenic contaminants to multiply. Semi-hard and hard cheeses Edam, Gouda, Credos and Providence cheese are classified as semi‐had varieties, and are produced from pasteurised milk with added starter cultures. The pressed curd is normally matured and stored for a period up to 3 months. However, hard cheeses

240   Microbial Toxins in Dairy Products

(e.g. Grana, Asiago, Cheddar, Pecorino, Emmenthal, Gruyere, Romano, Manchego) are also produced from pasteurised milk with added starter cultures, and the curd is scalded/ heated/cooked from 32 to 55°C. These cheeses have a maturation period up to 2 years. In Semi‐hard or hard cheeses, the growth of Staph. aureus is always possible and enterotoxins can be produced if the initial population of the pathogen in the milk is higher than 103 cfu mL−1. In addition, bad manufacturing practice allows Staph. aureus to multiply from 3 to 5 log10 cfu g−1, until the pH drops to inhibitory levels (Tatini et al., 1971; Koening & Marth, 1982). Blue Vein cheeses Roquefort, Gorgonzola, Blue Vein, and Adelost cheeses are classified as semi‐soft varieties, and they provide a very hostile environment for the growth Staph. aureus due to the combined inhibitory effects of Penicillium spp. and starter cultures (Tatini et al., 1973; Meyrand & Vernozy‐Roland, 1999). Thus, these cheeses do not represent a potential hazard for staphylococci foodborne intoxication. Pasta Filata cheeses Mozzarella, Provolone, Keseri and Kaskaval (Kashkaval) are typical examples of Pasta Filata‐type cheeses. The curd is initially heated to 45–48°C in the whey, plasticised at 78–80°C, formed into blocks, cooled and brined (at 13–17°C and the pH drops), and matured for different durations depending on the cheese variety. However, Mozzarella cheese is not matured, and is normally consumed fresh (Tatini et al., 1973). The acidity of the curd (pH 5.0–5.2) and the h­eating of cheese mass up to 80°C reduce Staph. aureus populations and enterotoxin production in these products. Processed cheese The manufacturing technology of this type of cheese involves the addition of polyphosphates followed by heating blends of ground cheeses to 80–85°C which, as a consequence, does not give rise to staphylococcal food poisoning (Johnson et al., 1990; Glass et al., 1998). Referring to the exposure assessment evaluation, official data of Sanitary Surveillance Services of the State of Rio Grande do Sul (Southern Brazil), which was collected during the years of 2000 to 2002, have identified Staph. aureus as the only agent responsible for 57 foodborne outbreaks, where 42 (74%) were confirmed by microbiological analyses and 15 (26%) by clinical symptoms and/or epidemiological data (Heeyoung et al., 2014). Staphylococcal outbreaks were responsible for the infection of 5 991 persons, with the most affected age group in people between 20 to 49 years old, that is, men (48%) and women (52%). The main food vehicles involved were meats (35%), followed by pastries (25%), cheese (23%), pasta (11%) and potato salad with homemade mayonnaise (11%). The majority of outbreaks occurred in private homes (33%) followed by commercial food establishments (28%). Inadequate control of temperature and failures

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   241

Table 9.5  Consequences of exposure determinations after pathogen infection. organism

Staphylococcus aureus

Severity of illness and general population susceptible

Consequences of exposure ‐ general population susceptible

Moderate: not usually life‐threatening; normally short duration; symptoms are self‐limiting symptoms; >20% of cases require medical attention

Mild: sometimes requires medical intervention

Data compiled from Ross & Sumner (2002), ICMSF (2002, 2009) and FS AU & NZ (2009). Table 9.6  Hazard characterisation and infective dose. Organism Staphylococcus aureus

Infective dose

Consequences of exposure ‐ general susceptible population

0.92 µg kg−1 cheese; three out of four people became ill (75% attack rate). Staphylococcus aureus in ewe’s milk is likely to increase in number by more than 3 log10 cfu g−1, and exceed the counts from 105 to 106 cfu g−1 in a short period of time, particularly under farm conditions. It can compete with starter cultures even when present at higher numbers. While exposure of Staph. aureus through consumption of ewe’s cheese made from raw milk is high, its consequences are mild, and severity may be assessed as negligible. The overall risk is considered as low (Tables 9.5 and 9.6 ‐ Qualitative framework developed by Food Science Australia ‐ Risk Rangers; FS AU & NZ, 2009), and some type of confirmation indicates an official incidence of s­taphylococcal poisoning at 0.02 per 100 000 population (i.e. lower than the EU a­verage of 0.06 per 100 000 population).

242   Microbial Toxins in Dairy Products

9.3.5  Predictive modelling and risk assessment of Staphylococcus aureus and enterotoxin production in cheese Different predictive models are used to study pathogens in food products. In cheese, primary predictive models are very useful, based on data describing changes in numbers of Staph. aureus, or levels of SE in the product. However, secondary predictive models are based on the effects of environmental factors (temperature and pH), or on the parameters in primary models (i.e. the maximum specific rate of growth of staphylococci). Tertiary predictive models are implemented with user‐friendly software, for example, Staph. aureus models, and include the growth and survival m­odels as used in the US Federal and Drug Administration pathogen modelling p­rogramme (PMP), the growth model in the ComBase modelling toolbox, and the ComBase predictor. To evaluate the safety of the process, Valero et al. (2009) developed a model where optimal levels of pH and aw were required for the probability of Staph. aureus growth, and found an abrupt transition of the interface between growth and no growth at low temperatures. Obeso et al. (2010) described the effect of initial lytic phage titres and initial Staph. aureus contamination of pasteurised milk on the probability of survival of this organism at different temperatures. Vora et al. (2003) used a probability simulation approach to evaluate the effect of the contamination level of Staph. aureus on the survival/gradual decline in intermediate‐ moisture foods. They reported no effect on initial contamination levels, but both simulations and observations indicated a wide variation in rates of decline, including occasional increases in population. Interactions with other micro‐organisms, naturally present in food or added, may have profound effects on Staph. aureus growth (as exploited by the use of starter cultures in fermented foods), and SEs production. For example, Le Marc et al. (2009) developed a kinetic model that described the inhibitory effect of a starter culture on Staph. aureus growth in milk, when the LAB had exceeded a critical density. Very few predictive models of SE production are available. Fujikawa and Morozumi (2004) developed a model based on observations that SE A was detectable at levels greater than the count of the organism of 6.5 log10 cfu mL−1, and increased linearly during the whole growth phase in a sterile milk medium. The rate of SE A production increased linearly with temperature from about 15 to 32°C, and was described by the following equation:

p

0.0376 t 0.559

Equation (1)

where p is the rate of SE A production (ng mL−1 h−1), and t is the temperature (°C); however, SE A was still produced at temperatures above 32°C, but the rate of increase with temperature levelled off. Thus, there is a lack of predictive SE models. In risk characterisation based on the number of Staph. aureus (cfu mL−1) or the concentration of enterotoxin (ng g−1) or per serving, the initial contamination levels, temperatures, storage/holding times and pH have the greatest impact on the assessment endpoints. Buchanan et al. (2000) developed a growth rate estimate model for Staph. aureus, and compared the accuracy of his model with the growth rate predictions of the

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   243

PMP, version 7. The former model predicted a faster growth rate than the PMP. The equation for generation time is in a general formula, which is derived from fitted parameters of a Gompertz function (Equation 2). A reduced version of the growth model is used in risk assessment under the assumption that nitrite is neither used nor present during the manufacture of cheese (Equation 3). However, equations 3 and 4 are derived equations for parameters B and C based on the temperature (T, °C), pH (P), and salt concentration (S, g 100 g−1), (FS AU & NZ, 2009). Generation Time GT, min



log10 2 e BC

Equation (2)

where “e” refers to Euler’s Number (~2.71828) ln B 

ln MPN 

10.8812 0.2551 T 1.0648P 0.2653S 0.00133TP 0.00516 TS (3) 0.00723PS 0.00273 T 2 0.0563P 2 0.00308S2 Equation

ln A C 1.4074 0.00765 T 0.1588P 0.0330 S 0.00241TP 0.0000980 TS 0.00355PS 0.000413 T 2 0.0129P 2 0.00122S2 Equation (4)

where MPN is the most probable number, and C can be solved ignoring the contribution of salt and assuming that A = 3 (see FS AU & NZ (2009) for further explanations and mathematical calculations).

9.4  Escherichia coli 9.4.1  Hazard identification Escherichia coli is a gram‐negative, facultative anaerobe, within the family Enterobacteriaceae, and is generally a commensal bacterium in the gastrointestinal tract of mammals and birds in a mutually beneficial relationship (Tchaptchet & Hansen, 2011). Most strains of E. coli strains are not pathogenic; however, some can acquire virulence factors from other bacteria and, in some cases, cause serious diseases. Pathogenic strains of this bacterium can be classified into six types:

• • • • • •

Enterotoxigenic E. coli (ETEC), Enteropathogenic E. coli (EPEC), Enteroaggregative E. coli (EAEC), Enteroinvasive E. coli (EIEC), Diffusely adherent E. coli (DAEC), and Verocytotoxin‐producing/Shiga‐toxin‐producing E. coli (VTEC/STEC).

244   Microbial Toxins in Dairy Products

The VTEC strains produce potent cytotoxins, which was originally designated as verocytotoxins (Vtx) because of their cytotoxic activity against kidney Vero cells of the African green monkey Chlorocebus sabaeus. Later, and due to their high structural and functional similarities with Shiga‐toxin produced by Shigella dysenteriae, they were also called (Shiga/Shiga‐like toxins, STx) and, consequently, the producer strains were also named Shiga‐toxin‐producing E. coli (STEC). The verocytotoxins can be divided into two groups, Vtx1/Stx1 and Vtx2/Stx2, and strains that produce sub‐type c of Vtx2 (Vtx2c) are clinically the most important because they induce the haemolytic uremic syndrome (HUS) more frequently than those producing Vtx1 (Friedrich et al., 2002; Bosilevac & Koohmaraie, 2011). Karmali et al. (2003) proposed a classification scheme of VTEC strains into five seropathotypes. This classification range from high‐risk seropathotype A to seropathotypes D and E with minimal risk, and relies on the frequency and severity in human disease and on association with outbreaks. Several serogroups exist, but only some are associated to pathogenic strains. The most important serogroups pathogenic to humans are: O26, O91, O103, O111, 0145 and O157 (EFSA, 2014). A subset of serotypes of VTEC is represented by Enterohaemorrhagic E. coli (EHEC) that can cause bloody diarrhoea and HUS in industrialised countries (Nataro & Kaper, 1998; Caprioli et al., 2005). In addition, this category holds the gene eae that encodes another toxin intimin Aea) responsible of attaching and effacing (A/E) lesions in infected cells. However, it is also known that some atypical EHEC, which do not carry the locus of enterocyte effacement (LEE) island where eae is located, can still cause the HUS and occasional outbreaks (Bonnet et al., 1998; Nataro & Kaper, 1998; Paton et al., 1999; Feng et al., 2001; Karch et al., 2005).

9.4.2  Growth and inactivation Growth of E. coli ranges from 7–8°C to 46°C, with an optimum temperature of 35–40°C (ICSMF, 1996). Although VTEC strains are not heat resistant, Rasooly & Do (2009) demonstrated that Stx2/Vtx2 are heat‐stable, and is not inactivated by p­asteurisation, but the vegetative cells are killed. In order to inactivate heat‐stable Stx2, a temperature of 100°C for 5 min is needed. The optimum pH for E. coli growth is 6‐7, but it can survive in pH ranging from 4.4 to 10.0 (Arnold et al., 1995; Leyer et al., 1995; Lin et al., 1996; Desmarchelier & Fegan, 2003). Nevertheless, a study by Molina et al. (2003) confirmed that several VTEC strains can survive at pH 2.5–3.0 for over 4 h. The minimum aw necessary for E. coli to grow is 0.95 but, if the temperature or pH conditions are not optimal, a higher aw is required for growth (Desmarchelier & Fegan, 2003).

9.4.3  Hazard characterisation Acquirement of the disease Verocytotoxin‐producing E. coli have been isolated from different species of animals that can spread the bacteria in their faeces. In fact, the most important reservoir is the gastrointestinal tract of healthy ruminants. Humans could become infected through the

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   245

consumption of contaminated food of cow and sheep origin, of faecal contaminated vegetables and drinking water, by direct transmission among people, or from infected animals or faecal contaminated environments to humans (EFSA, 2014). Disease mechanisms These bacteria can affect the food safety and, therefore, human health with their ability to cause gastrointestinal diseases and even death (Meng et al., 2007; Solomakos et al., 2009). These bacteria produce Vtx/Stx in the colon; the toxin is absorbed through the bowel and carried in the blood to reach and damage endothelial cells, above all in the kidney tissue and in the central nervous system, generating thrombin and fibrin deposits in the microvasculature. This progresses to cause leakage and tissue oedema. The small blood vessels, in part, occluded by thrombi can cause damage to the erythrocytes and, subsequently, haemolysis (Lynn et al., 2005). Illness Verocytotoxin‐producing E. coli requires low‐infectious dose (1 to 100 cells) to cause severe illness (Teunis et al., 2004). Clinically, the manifestations associated with VTEC infection may differ as they can include non‐specific diarrhoea and haemorrhagic colitis. The incubation period is usually 3‐4 days after exposure to infection, but may be as short as 1 day, or as long as 10 days (CDC, 2011). The illness lasts on average 8 days, and the patients excrete EHEC for about one week (ICSMF, 1996). However, the victim may develop HUS (Karmali, 1989; Griffin & Tauxe, 1991), which is characterised by acute renal failure, anaemia and thrombocytopenia (Farrokh et al., 2012). In addition, HUS can be “classical D+ (diarrhoea‐associated)” or “atypical”. Classical is associated with VTEC infection, and is the most common form of HUS. The prodrome of atypical HUS is usually a respiratory illness, and can affect children if inherited; it can also occur in adults (females who are pregnant or taking oral contraceptives), suffering from malignant hypertension or various chronic illnesses. Whereas the prodrome of VTEC‐associated HUS is an acute diarrhoeal illness, and has its highest incidence in children (Karmali et al., 2010). According to EFSA (2012), 5671 confirmed human VTEC infections were reported in the EU with 1.15 cases per 100 000 population (EFSA, 2014). The prevalence of confirmed cases from 2007 to 2012 is shown in Table 9.7. Around 6.3% of patients Table 9.7  Prevalence of confirmed food poisoning cases in European countries from 2007 to 2012. Year

Confirmed cases

Confirmed cases per 100 000

2012 2011 2010 2009 2008 2007

5671 9487 3656 3583 3162 2905

1.15 1.93 0.83 0.75 0.70 0.60

Data compiled from EFSA (2009, 2010, 2011, 2012, 2013, 2014).

246   Microbial Toxins in Dairy Products

Table 9.8  Prevalence of confirmed Escherichia Coli food poisoning in European countries from 2007 to 2012 of the most important serogroups. Year

Serogroup (%) O157

2012 2011 2010 2009 2008 2007

41.1 41.2 41.1 51.7 53.0 54.1

O104 20.1

O26

O91

O103

O145

12.0 5.4 7.1 5.4 5.3 4.7

3.6 2.2 1.6 1.3 1.6 1.5

3.3 2.7 2.5 2.3 2.8 2.7

2.9 1.4 1.7 1.3 1.6 1.1

Data compiled from EFSA (2009, 2010, 2011, 2012, 2013, 2014).

developed HUS and 4.6% died (Gould et al., 2009). Moreover, HUS is most common in children up to five years old and the elderly (Fitzpatrick, 1999), and majority of the cases are reported in the summer months (EFSA, 2014). Table 9.8 shows the prevalence of food poisoning outbreaks in the EU from 2007 to 2012 with the most important serogroups reported. During this period, several VTEC disease outbreaks were caused by E. coli O157 (EFSA, 2014). Dose response It is unethical to evaluate the dose response from studies in humans; however, data can be obtained by conducting experimental studies on animals. Moreover, another problem for EHEC is the large number of serotypes that make the dose‐response relationship very difficult. Pai et al. (1986) inoculated 3 infant rabbits intra‐gastrically with EHEC O157:H7, and examined them for clinical symptoms, bacterial colonisation, presence of detectable free Vtx in the intestines, and histological changes. Using the response of diarrhoea, the data were fitted to the beta‐Poisson model by Haas et al. (2000). These researchers developed a dose response of diarrhoea associated with E. coli O157:H7 infection. Comparing this model with two human outbreaks, one foodborne and the other waterborne, it was estimated that the dose necessary to cause illness 50% of the exposed population was 5 × 105 cfu g−1, and the probability of illness from ingesting 100 organisms was 2.6 × 10−4 cfu g−1 (Haas et al., 2000).

9.4.4  Exposure assessment Cheeses are ready‐to‐eat products, and do not require any treatment before eating. Since the consumption of raw milk and dairy food made from raw milk is a possible cause of concern, it is important to prevent contaminations with foodborne pathogens, such as VTEC. The sources of milk contamination by VTEC can be either faecal or intra‐mammary (i.e. from mastitis in the udder of the cow) (Stephan & Kuhn, 1999). Accordingly, if these bacterium are present in the soil, soiled teats will be contaminated and subsequently the milk could be infected during milking. Farrokh et al. (2012) reported that

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   247

the prevalence of VTEC in raw milk varies between 0% and 4%, but published data have employed different approaches for the detection of VTEC that show a huge d­ifference between the prevalence rate in different samples. The EFSA (2007a) reported that human outbreaks diseases were more frequently associated with raw milk cheeses, particularly with soft and semi‐soft varieties even if these cheeses are considered to have better aroma and flavour than cheeses made from pasteurised milk. While the EU (2004) defines raw milk as: “Milk produced by the secretion of the mammary gland of farmed animals that has not been heated to more than 40°C or undergone any treatment that has an equivalent effect”. The same regulation also establishes that: “During transport the cold chain must be maintained and, on arrival at the establishment of destination, the temperature of the milk must not be more than 10°C”, and “Food business operators must ensure that, upon acceptance at a p­rocessing establishment, milk is quickly cooled to not more than 6°C and kept at that temperature until processed”. However, food business operators may keep milk at a higher temperature if:

• •

Processing begins immediately after milking, or within four hours of acceptance at the processing establishment; and The competent authority authorises a higher temperature for technological r­easons concerning the manufacture of certain dairy products.

Storage of the milk at temperatures 53°C resulted in the inactivation of bacteria and destruction of toxins, except for Staph. aureus. Furthermore, a recent study confirmed that a temperature >53°C caused a rapid decrease in VTEC population in cooked cheeses (Peng et al., 2011). Conversely, Usajewicz & Nalepa (2006) observed that, if a temperature of 55°C was used, E. coli was inactivated after 180 min but, if the cheese was stored at a temperature between 7°C and 37°C (i.e. thermal abuse), the bacteria were able to repair their damage. Salting is usually done to assist supplementary drainage of whey, and is achieved by brining (i.e. immersion of moulded curd in salt solution) or dry salting (i.e. rubbing salt on the surface of the milled curd or moulded curd). The salt content in different cheeses ranges between 0.7–7.0 g 100 g−1 (Fox et al., 2000) and, it has been observed by Glass et al. (1992) that in Tryptone soya broth (TSB), E. coli O157:H7 was inhibited by ≥8.5 g 100 −1 salt. However, in the presence of 6.5 g 100 g−1 salt, the pathogen could grow, and dry salting of the milled curd (i.e. on the surface only) does not reach the whole curd mass so that the micro‐organisms inside the curd could initially continue to grow. Ingham et al. (2000) showed that E. coli O157:H7 was able to survive in cheese brine for several weeks under typical brining conditions. The minimum aw for the growth of E. coli is 0.95, and a short duration of maturation does not lead to a significant reduction of aw. However, storing the cheese at a low temperature in conjunction with low pH and aw may reduce the count of VTEC (Moretro et al., 2010).

9.4.5  Risk characterisation Although the number of confirmed human VTEC infections in Europe is not high (Table 9.7), the low infectious dose and the severity of the disease increases the risk to children. Verocytotoxin‐producing E. coli can survive during the manufacture of cheese; Farrokh et al. (2012) reviewed several studies that show the prevalence of VTEC in the product (Table  9.9). They concluded that different approaches have been used by researchers for the detection of this bacterium, and it is difficult to c­ompare the results obtained. Miszczycha et al. (2013) studied the behaviour of different VTEC serotypes (O157:H7; O26:H11, O103:H2 and O145:H28) during cheesemaking made from raw milk. They observed the following: (a) VTEC did not grow in cooked curd, (b) the organism did not grow in the presence of starter culture, but was still detectable in the matured cheese, and (c) the organism grew in freshly made Blue Vein cheese, but decreased after maturation. In smear cheeses, Maher et al. (2001) demonstrated that E. coli O157: H7 grew in milk that was spiked at a level of 1.52 log10 cfu mL−1, and the count increased by 1.3 log10

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   249

Table 9.9  Some selected studies reporting on the occurrence of verotoxigenic Escherichia coli (VTEC) and prevalence of Stx genes in dairy products. Region/country

Dairy product

European Union (EU)

Cheese made from raw milk1

Belgium

Cheese made from raw cow’s milk Cheese made from raw goat’s and sheep’s milk Butter and cream made from raw cow’s milk Butter, yoghurt, cheese, ice‐cream and fresh cheese made from raw cow’s milk Cheese made from raw milk1 Cheese made from pasteurised milk1 Mozzarella cheese made from raw buffalo’s milk Dairy products made from pasteurised cow’s milk Dairy products made from raw cow’s milk Dairy products made from pasteurised sheep’s milk Dairy products made from raw sheep’s milk Mozzarella cheese made from buffalo’s milk1 Cheese made from raw cow’s, goat’s an sheep’s milk Market milk, cheese curd and cheese made from raw cow’s, goat’s an sheep’s milk Fresh cheese made from goat’s and sheep’s milk1 Cheese1

Italy

Portugal

Spain

Scotland

Castello2 cheese made from raw sheep’s milk Cheese made from raw cow’s milk

STEC isolate occurrence (%) and/or stx gene prevalence (%) (N = number of samples tested)

References

0.2% (N = 2876) STEC in 2005 2.4% (N = 1064) STEC in 2006 0.5% (N = 1961) STEC in 2007 1.8% (N = 700) STEC in 2008 5.6% (N = 71) Escherichia coli O157:H7 0% (N = 222) E. coli O157

EFSA (2010)

De Reu et al. (2002) Imberechts et al. (2007)

0% (N = 181) E. coli O157 0% (N = 64, 9, 16, 7 and 4, respectively) E. coli O157

De Reu et al. (2004)

0% (N = 143) E. coli O157

Civera et al. (2007)

0% (N = 60) E. coli O157 0% (N = 93) E. coli O157 0% (N = 657) E. coli O157

Martucciello et al. (2008) Conedera et al. (2004)

0% (N = 811) E. coli O157 0% (N = 477) E. coli O157 0% (N = 502) E. coli O157 0% (N = 501) E. coli O157 0% (N = 70) E. coli O157

Almeida et al. (2007)

1.8% (N = 502) STEC (45% stx + samples)

Rey et al. (2006)

0% (N = 103) STEC (4% stx + samples) 0% (N = 39) STEC (2% stx + samples) 2.4% (N = 84) STEC 0% (N = 739) E. coli O157

Caro & Garcia‐ Armesto (2007) Coia et al. (2001) (Continued)

250   Microbial Toxins in Dairy Products

Table 9.9  (Continued) Region/country

Switzerland

France

Dairy product

Cheese3 made from raw cow’s, goat’s and sheep’s milk Cheese3 made from raw cow’s, goat’s and sheep’s milk Cheese made from raw milk1 Cheese4 made from raw goat’s milk Cheese1 Cheese made from raw milk1 Cheese made from raw cow’s and goat’s milk

Germany

Cheese made from raw cow’s milk

STEC isolate occurrence (%) and/or stx gene prevalence (%) (N = number of samples tested)

References

2% (N = 786) STEC (4.9% stx + samples)

Stephan et al. (2008)

1.9% (N = 1502) STEC (5.7% stx + samples)

Zweifel et al. (2010)

11.7% (N = 180) STEC (30.5% stx + samples) 13.1% (N = 1039) STEC

Fach et al. (2001) Vernozy‐Rozard et al. (2005a) Pradel et al. (2000) Auvray et al. (2009) Madic et al. (2011)

1% (N = 603) STEC (10% stx + samples) 27.7% (N = 112) stx + samples 5.5% (N = 400) STEC, including 1.8% STEC O26:H11 (29.8% stx + samples) 0.48% (N = 209) STEC

Messelhäusser et al. (2008)

 Type of milk not reported and/or whether raw or pasteurised.  Matured hard cheese variety. 3  Semi‐hard, hard and soft varieties. 4  Soft, hard, fresh, blue mould varieties. Adapted from Farrokh et al. (2012). 1 2

cycle during the different stages of cheesemaking, surviving for 70 days maturation or 90 days after enrichment of the test media. In soft cheese, Leuschner & Boughtflower (2002) inoculated the milk with low levels of E. coli O157, and d­emonstrated that the bacteria could survive the cheesemaking process. The presence of pathogenic organisms in matured hard cheese made from raw milk has been demonstrated by Peng et al. (2013). They have investigated 3 d­ ifferent strains of VTEC, having inoculated the raw milk at a level of 101 and 103 cfu mL−1 and cooked the curd at 40 and 46°C. All the strains survived cheesemaking (i.e. at both spiking levels and cooking temperatures); however, VTEC were recovered at >10 cfu g−1 after 16 weeks maturation only in cheeses made with the lower ­inoculation level. Other studies on the survival of pathogens in different cheeses were as follows: Pecorino Romano Escherichia coli survived a storage temperature of 4°C in sheep’s milk, but growth was inhibited. According to statutory requirement of the Pecorino Romano, “curd made from thermised milk is cooked at 45–48°C for 10 minutes and is held at this temperature for up to 30 minutes while the curd is pressed under whey”. This processing step inhibited growth of E. coli, resulting in a reduction of the cell count by 2–3 log10 cycles. However, Hudson et al. (1997) reported an increase of O157:H7 during different stages of manufacture, for example: (a) an increase of 1.7 log10 cycle during curd formation and after

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   251

65 h in the brine (22 g 100 g−1 NaCl), and (b) a 3 log10 cycles reduction in counts after maturating the cheese for 30 days at 11–13°C. Feta and Telemes In Feta and Telemes, Govaris et al. (2002) reported an increase in cell count of E. coli O157:H7 during the first 10 h after inoculating pasteurised ewe’s and cow’s milk, but no detectable increase in matured cheese (i.e. after 44  and 36 days, respectively). Feta cheese made with pasteurised cow’s milk had an increase of 1 cfu g−1 of E. coli O157:H7 after curd formation (Hudson et al., 1997), but a reduction of 3 log10 cycles in the counts was observed after 27 days of maturation. However, Ramsaran et al. (1998) showed that E. coli O157 can survive during Feta cheesemaking, and can persist after storage for 75 days (counts were >106 cfu mL−1). Manolopoulou et al. (2003) also observed an increase in E. coli count between 2.2–3.8 log10 cycles in the first 10 days followed by a decrease in numbers and, at 120 days, no cells were detected. Camembert Montet et al. (2009) reported an increase of VTEC count of 2.2 log10 cfu g−1 during the first stages of cheesemaking after inoculating the raw milk at a rate of 3 log10 cfu mL−1; E. coli O157 survived during the manufacture of Camembert, and persisted for 65 days after storage (Reitsma & Henning, 1996). Cheddar Escherichia coli O157 can survive the Cheddar cheesemaking and persist after maturation (Reitsma & Henning, 1996), when the milk was inoculated with the pathogen (103 cfu mL−1). However, the count was reduced by 2 log10 cycles after maturing the product for 60 days, but the pathogen could still be detected after 158 days. After spiking the milk with a low inoculum, cells were not detectable after 130 days. D’amico et al. (2010) demonstrated that on inoculating raw milk with E. coli O157:H7 (1.3 log10 cfu mL−1), the strain remained detectable (>270 days) after enrichment of the test medium.

9.5  Listeria monocytogenes 9.5.1  Hazard characterisation The genus Listeria includes six species: L. monocytogenes, L. ivanovii, L. innocua, L.  seeligeri, L. grayi and L. welshimeri. Among these species, L. monocytogenes is by  far the most frequently implicated in animal and human listeriosis (McLauchlin et  al., 2004). There is evidence that not all L. monocytogenes strains are pathogenic or present the same level of virulence. Only few (1/2a, 1/2b, and 4b) of the thirteen L. monocytogenes serotypes appear to be most frequently involved in foodborne outbreaks. However, other serotypes have sporadically shown to be able to cause human listeriosis after consumption of contaminated dairy products (McLauchlin, 1997;

252   Microbial Toxins in Dairy Products

Table 9.10  Human outbreaks of food‐borne listeriosis associated with dairy products. Year

Country/region

Number of reported cases

1985

United States of America (USA) Switzerland France

142

Mexican‐style soft cheese

122 36 14 25 2

1983–87 1995 1997 1998–99 1999

Source of infection

2000

USA

13

Soft cheese Brie de Meaux cheese Livarot, Pont‐Lʹévêque cheeses Butter Cheese (variety was not reported) Mexican‐style soft cheese

2001

Sweden

50

Fresh cheese

2001

Japan

38

2002

Canada

17

2005 2006

Switzerland Czech Republic

10 78

Cheeses (varieties were not reported) Raw milk cheese (variety was not reported) Tomme cheese Soft cheese

2006 2008 2009 2012

Germany Canada Austria USA

6 38 14 22

Hard cheese Soft cheese Quarg Ricotta salata cheese

Finland England

References McLauchlin et al. (2004) and EFSA (2007b) De Buyser et al. (2001) McLauchlin et al. (2004) MacDonald et al. (2005) Carrique‐Mas et al. (2003) Makino et al. (2005) Gaulin (2003) Bille et al. (2006) EFSA (2007a) and Vit et al. (2007) EFSA (2007a) Bille et al. (2006) Fretz et al. (2010) CDC (2012)

McLauchlin et al., 2004). For this reason, L. monocytogenes strains should be considered pathogenic regardless of their serotype (SCVMPH, 1999; EFSA, 2007b) (Table 9.10). Pathogenic L. monocytogenes can carry several virulence factors (e.g. listeriolysin O (LLO), or protein regulatory factors – PrfA). Even though some virulence factors are regularly detected in strains isolated from human patients affected by listeriosis, it is still not possible to discriminate with certainty non‐pathogenic strains from pathogenic ones by analysis of virulence genes. Hence, any serotype of L. monocytogenes detected in dairy products should be considered as a potential hazard for human health in risk assessment (EFSA, 2007b; FDA & Health Canada, 2012). Listeria monocytogenes is able to survive or grow in a wide range of adverse environmental conditions (Gandhi & Chikindas, 2007), including growth from 0 to 47°C (ICSMF, 1996). This is an important characteristic for food safety since the bacterium is able to multiply in refrigerated dairy products in addition to survive and eventually grow at low pH (down to pH 4.4) and high salt concentrations (up to 10 g 100 g−1 NaCl) (Fox et al., 2000; Koutsoumanis & Sofos, 2005; Cataldo et al., 2007). In food, the conditions, which permit listerial growth, are limited. EU (2005) categorises the microbiological criteria of those food products with pH ≤4.4 and aw ≤0.92, or with pH ≤5.0 and aw ≤0.94 that are unable to support the growth of L. monocytogenes. Consequently, most dairy products have physical characteristics that potentially allow the growth of this bacterium. Despite being a non‐spore‐forming bacterium, L. monocytogenes shows remarkable resistance to heat treatments (Sofos, 1993). Although the pathogen is sensitive to pasteurisation, viable cells can still be recovered from milk heated for 15 min at 62.8°C (Farber et al., 1988; Knabel et al., 1990).

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   253

The persistence of Listeria spp. in the environment can be enhanced by its capacity to form biofilms, a group of microbes embedded in an extracellular organic matrix adhering to a surface (Carpentier & Cerf, 1993). In processing plants, biofilm formation is likely to occur in sites where water and organic soil may stagnate and/or if it is more difficult to carry out accurate cleaning operations (Carpentier & Cerf, 2011). Disinfectants, heat, ultraviolet (UV) application, and other treatments generally used for cleaning and disinfection/sanitation of equipment cannot effectively penetrate the biofilm matrix. As a result, the bacteria are protected and can survive for long periods (Carpentier & Cerf, 1993).

9.5.2  Exposure assessment Dairy products represent a very heterogeneous food category where each product is characterised by a distinctive manufacturing process, and each stage can potentially be associated with contamination by L. monocytogenes. The major exposure means to hazards are discussed below. At farm level Primary contamination of raw milk can occur during its collection at the farm. Many studies reported prevalence of L. monocytogenes in bulk tanks, and the percentage of positive analysis is related to several factors, such as the country where the survey and the experimental design were conducted. FDA and Health Canada (2012) estimated that the median cow herd prevalence is 2.2% and 3.7% in the USA and Canada, respectively, whilst FS AU & NZ (2009) listed more than fifty international surveys, which had investigated the presence of L. monocytogenes in raw cow’s milk where the highest reported prevalence was 60%; however, in most studies it was under 10%. Less information is available regarding sheep and goat farms. A study conducted on 283 sheep farms in Europe showed a prevalence of 2.2%, and a similar percentage (2.4%) was found in bulk tank milk samples in Asia (Rodriguez et al., 1994; Jamali et al., 2013). However, the occurrences of listerial contamination of goat’s milk samples collected by two international surveys were 3.8% and 2.5%, respectively (Gaya et al., 1996; Abou‐Eleinin et al., 2000). Contamination of the bulk tank during milking can be ascribed mainly to two different sources for the survival and/or growth of L. monocytogenes, that is, the dairy farm environment (e.g. floors, work surfaces, hair, b­eddings, lagoons, manure and equipment) and the health condition of the udder of milking animals (e.g. infected with mastitis) (Fthenakis et al., 1998; Wagner et al., 2000; Winter et al., 2004; Oliver et al., 2005; Gameiro et al., 2007; Pintado et al., 2009; Schoder et al., 2011; Osman et al., 2014). As a consequence, the animals can develop clinical symptoms of listeriosis or become a healthy carrier of the pathogen (Nightingale et al., 2004). Although the environment appears to be the most frequent cause of contamination, the resulting concentration of L. monocytogenes in the milk bulk tank is usually very low. Nevertheless, in a quantitative risk assessment concerning soft cheeses from raw cow’s milk, FDA & Health Canada (2012) used data from several studies to estimate a concentration substantially below 10 cfu mL−1 in contaminated bulk tanks where animals were mastitis‐free.

254   Microbial Toxins in Dairy Products

Infected animals are able to secret the pathogen and contaminate the entire production chain, which can be responsible for significant concentrations of L. monocytogenes in the final product (Bourry et al., 1995; Schoder et al., 2003; Winter et al., 2004; Gameiro et al., 2007; Pintado et al., 2009; Osman et al., 2014). The counts of Listeria spp. excreted by an udder can vary during the milking period (maximum of ~4 × 104 cfu mL−1), and similar counts have been reported in sheep (Wagner et al., 2000; Schoder et al., 2003). Its occurrence on‐farm is sporadic and there is only little information available about its frequency but, in milking herds, positive percentages (0.01–0.1%) of cows have been observed (Jensen et al., 1996). Regarding the occurrence of Listeria spp. in sheep and goats, the published data is insufficient and only descriptions of sporadic cases can be found, often investigated only after an accidental isolation of the pathogen in cheese (Schoder et al., 2003; Gameiro et al., 2007; Pintado et al., 2009). At the farm, during transport and before processing, the milk is kept under refrigeration. Different temperature distributions have been used to model the bulk tank temperature on farm and in dairy product risk assessments. FS AU & NZ (2009) estimated the temperature range using a triangular (2, 4, 10°C) distribution function. Bemrah et al. (1998) used the same type of distribution, but with a narrow range of values (triangular ‐ 4, 5, 6°C), while Latorre et al. (2011) and Bemrah et al. (1998) adopted a uniform (7, 10°C) distribution function. The results of these studies suggest that L. monocytogenes is able to grow in milk at the temperatures above those reported (Walker et al., 1990; Augustin et al., 2005). Walker et al. (1990) reported a minimum growth temperature in milk was between −0.1 and −0.4°C. Even though L. monocytogenes multiplies slowly and the lag time can be prolonged at refrigerator temperatures (4–5°C), an increase of some degrees can significantly change the behaviour of the micro‐organism (Walker et al., 1990). For these reasons, risk assessment should be particularly attentive to the modelling of raw milk storage temperatures and temperature abuse scenarios should be hypothesised. The processing plant Several cheesemaking parameters affect the growth and survival of L. monocytogenes in the product and the milk used for manufacturing. These factors include heat treatments, acidification, whey separation, and maturation, and will be discussed below. Also, contamination from the plant environment may be relevant. Heat treatment of the milk is the most efficient control measure to ensure the absence of L. monocytogenes in the raw milk or cheese (FS AU & NZ, 2009). Milk pasteurisation inactivates high concentrations of the pathogen, and is considered a reliable method to control milk’s primary contamination (Farber et al., 1988; MacDonald & Sutherland, 1993), but not in artisanal cheeses made from raw or thermised milk. The higher the processing temperature and/or duration of the thermising treatment, the greater the decrease of the listerial count in milk. Consequently, if the thermal treatment is low, the pathogen will survive, for example, the thermisation process (Doyle et al., 2001). For certain cheese types, additional heat treatments can take place, such as curds can undergo cooking after the coagulation stage or heated for the stretching of Pasta Filata cheeses

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   255

(e.g. Mozzarella). As for the thermisation processes, the temperature and duration of these treatments vary according to the type of cheese, so that sufficiently lethal effects on L. monocytogenes will only occur at certain t­emperature/time combinations (Kim et al., 1998; FS AU & NZ, 2009). Milk acidification is a second major anti‐listerial safety factor of importance. During the coagulation stage, the growth of LAB reduces the milk pH from 6.6 to typically 5.0–4.6. When the curd reaches these pH values, Listeria spp. can survive, but is not able to grow especially if other intrinsic and extrinsic factors of importance (e.g. aw and t­emperature) are unfavourable. Furthermore, the presence of LAB generally limits listerial growth in the curd and during the subsequent stage of maturation (Fox et al., 2000; Schvartzman et al., 2011a). Nevertheless, problems may arise in the case of a delayed or inhibition of growth of the starter culture due to poor milk quality (e.g. presence of antibiotics, bacteriophages and/or a poor hygienic quality) or poor starter culture viability (FS AU & NZ, 2009; Schoder et al., 2008). During acid‐induced coagulation, some cells of Listeria spp. in the milk are lost in the whey, but the large majority remains trapped in the curd. After six contamination trials, Papageorgiu & Marth (1989) reported that an average of 3.6% of the cells of the inoculum were lost in the whey during the manufacturing of a cow’s milk cheese. Several studies have reported ~1 log10 cycle increase between listerial counts in the milk and in the curd at the end of the coagulation stage, but it is not clear if this increase is only due to a pathogen concentration or if growth occurs (Papageorgiou & Marth, 1989; Margolles et al., 1997; Morgan et al., 2001). The subsequent maturation stage plays a fundamental role in the growth or survival of L. monocytogenes in cheeses. This is influenced by many factors, such as pH, temperature, moisture content, aw, competing microbiota, and the presence or absence of additives. Hence, modelling the fate of the pathogen can be very complex (Fox et al., 2000), and the situation is further obscured due to the many varieties of cheeses with idiosyncratic process characteristics. Nevertheless, it is well known that high values of aw considerably affect the growth of Listeria spp. in cheeses. Several studies have shown that L. monocytogenes is able to grow in soft cheese under favourable conditions, and surveillance data reported were frequently associated to human cases of listeriosis (Bolton & Frank, 1999; Fox et al., 2000; MacDonald et al., 2005; EFSA, 2007b; Fretz et al., 2010). This is particularly evident for mould and smeared cheeses because the development of moulds on the surface and/or an increase in pH within the product during maturation significantly creates optimal growth conditions (Pearson & Marth, 1990; D’Amico et al., 2008a; FDA & Health Canada, 2012; Bernini et al., 2013). In contrast, semi‐hard and hard cheeses can be considered safer, because the decrease of aw during the maturation period, especially at low pH and the presence of competitive microbiota, does not permit the growth of L. monocytogenes, which dies off progressively (Pearson & Marth, 1990; Fox et al., 2000; Wemmenhove et al., 2013; Dalmasso & Jordan, 2014). Nevertheless, attention should be always paid to this type of cheese as some strains of Listeria spp. can survive for weeks under such adverse conditions. As a result, human listeriosis outbreaks caused by consumption of hard cheeses have been reported by EFSA (2007b) and Yde et al. (2012).

256   Microbial Toxins in Dairy Products

In addition, listerial contamination in the milk or cheese may be introduced from the processing plant (D’Amico & Donnelly, 2008), or the processing environment, such as walls, ceilings, drains, floor puddles, sinks, refrigeration units and condensation in the compressed air line (FS AU & NZ, 2009; Hoelzer et al., 2012). If cleaning and disinfection procedures are inadequate, the pathogen can be found on food contact surfaces (tanks, vats, moulds, baskets, shelves, utensils) and, consequently, will contaminate the milk or the product during the manufacturing stages (D’Amico & Donnelly, 2008; Hoelzer et al., 2011). Although it is difficult to estimate the rate of bacterial transfer  from the dairy plant environment into the product (Hoelzer et al., 2012), cross‐­ contamination is generally considered low. FDA & Health Canada (2012) has estimated a median transfer of 30 cells in 250 g of cheese after contamination. Cross‐contamination during processing is particularly relevant for those types of dairy products that support growth of L. monocytogenes (e.g. soft or unmatured cheeses). Distribution and consumption Since Listeria spp. are psychrotrophic micro‐organisms, the maintenance of the cold chain during distribution and home storage is crucial for dairy products. FDA (2003) estimated, in its assessment of ready‐to‐eat products, the temperature has to be at 1–5°C for various dairy products sold in the retail chain (FDA & FSIS, 2003). However, refrigerator temperatures are not constant and thermal abuses can occur. Several surveys of European retail refrigerators have shown that the mean temperature ranges from 3.7 to 5.6°C, with a maximum of 12.2°C (EFSA, 2007b). The situation is even more variable for temperature distributions in domestic refrigerators. Several factors, such as the external temperature and the positioning of the product in the refrigerator (e.g. door shelf or upper shelf), can significantly affect the temperature of the milk products (EFSA, 2007b). Temperature distributions in home refrigerators were reported by EFSA (2007b) using data from two American and fourteen European surveys. In the USA, the temperature was under 4°C in 45% of the observations (Audits International, 2000; Kosa et al., 2007), while in Europe it was under 6–8°C for the same percentile (EFSA, 2007b).

9.5.3  Hazard characterisation Acquirement of the disease Although the mode of transmission of L. monocytogenes may include vertical (mother to child) and zoonotic (animal to man) transmission, most cases of human listeriosis, both sporadic and epidemic, are caused by ingestion of contaminated food (Low & Donachie, 1997; Vazquez‐Boland et al., 2001). Ready‐to‐eat products are most frequently involved in cases of human listeriosis. According to surveillance and published data, listeriosis outbreaks are often associated with the intake of dairy products (especially soft cheeses) (McLauchlin et  al., 2004; EFSA, 2007b). Data from the Centers for Disease Control and Prevention (CDC) showed that 17% of the 71 foodborne disease outbreaks associated with cheese during the period 1998–2011 in the USA were caused by L. monocytogenes (Gould et al., 2014).

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   257

Illness Human listeriosis can occur in different forms distinguished on the basis of the clinical symptoms, developing into invasive systemic disease or non‐invasive or gastroenteritic disease (Ray & Bhunia, 2007). In general, invasive listeriosis has an incubation period of 2‐3 weeks, but can reach up to 3 months. The symptoms include fever, headache, vomiting, visual disorders, sepsis and meningitis. With lower frequency of infection, the symptoms including endocarditis, myocarditis, arteritis, pneumonia, pleuritis, hepatitis, colecystitis, peritonitis, localised abscesses, arthritis, osteomyelitis, and sinusitis may occur (Fox et al., 2000; Vazquez‐Boland et al., 2001). The mortality rate is 20 to 30%, both for outbreaks and sporadic cases. Some categories of the population, such as pregnant women, the elderly, immunocompromised and children, are particularly susceptible to Listeria spp. infection because their immune systems are not able to control the infection with the same efficiency. Invasive listeriosis associated to pregnancy is frequent (up to 44% of all cases), and can be responsible for abortion, stillborn foetus, septicaemia, and neonatal meningitis (Siegman‐Igra et al., 2002; Vazquez‐Boland et al., 2001). Listeriosis, in its non‐invasive form, is not well characterised yet, and seems to be associated mainly to healthy adults exposed to highly contaminated foods. The symptoms are less severe compared to invasive listeriosis, and consist essentially of diarrhoea (Aureli et al., 2000; Hof, 2001). Dose‐response The minimum infectious dose of L. monocytogenes required to cause disease in humans is unclear, but epidemiological data from foodborne outbreaks show that human listeriosis is usually associated with the consumption of foods that are highly contaminated (>106 cfu g−1) (McLauchlin et al., 2004; Mateus et al., 2013). However, cases of the disease caused by products with lower bacterium counts of the pathogen can occur, although less likely (FAO, 2004). The infectious dose depends on several factors, such as the virulence characteristics of the strain, the susceptibility of the host and any attributes of the food that alter microbial or host status. The majority of the L. monocytogenes strains are able to provoke listeriosis, but can be more or less virulent based on the virulence genes carried (FAO, 2004). As previously mentioned, some categories of people are more susceptible to i­nfection so that a lower infectious dose is sufficient to cause the disease. Other risk factors may increase the risk of acquiring listeriosis, such as chronic diseases (AIDS, cirrhosis, diabetes) or immunosuppressive and antacid therapies (FAO, 2004; Goulet et al., 2012). Several mathematical models have been developed to correlate the number of bacterial cells ingested with the likelihood of developing invasive listeriosis (Farber et al., 1996; Buchanan et al., 1997; FAO, 2004; Lindqvist & Westoo, 2000; FDA & Health Canada, 2012). However, the biological end‐points of these models are different, affecting the prediction of the likelihood of infection (probability of colonisation of the host which is not necessarily associated to symptoms), the morbidity or mortality. No model has been developed with data from human volunteers because of the severity of the symptoms and trial tests are not possible. As a result, all dose‐response studies are

258   Microbial Toxins in Dairy Products

derived from epidemiological data, foodborne outbreak investigations, trials on animals, and/or combinations of these studies (FAO, 2004). First attempts to estimate a dose‐ response relationship were done using the limited data cited in the literature (Farber et al., 1996; FAO, 2004). Buchanan et al. (1997) used a completely different approach based on epidemiological data to correlate the annual incidence of listeriosis caused by food in Germany with the level of exposure of the population. The resultant exponential model estimated that the dose that would be expected to produce severe illnesses in half of a population of immunocompromised individuals was equal to 5.9 × 109 cfu (Buchanan et al., 1997; FAO, 2004). The FDA & FSIS (2003) developed a dose‐response model to estimate the mortality in humans that takes into account the variability of virulence of L. monocytogenes strains, population exposure to the pathogen, human susceptibility, and dose‐response data from animal studies. Therefore, the median probability of mortality for the population (intermediate‐age, neonatal, and elderly) after ingesting a contaminated portion of food (i.e. 109 cfu per serving) would be 1.2 × 10–7, 1.4 × 10–4, and 3.4 × 10–6, respectively. A conversion factor based on surveillance data that reflects the overall relationship between illness and mortality across the entire dose range was used to predict the risk of serious listeriosis (FDA & FSIS, 2003; FAO, 2004). Finally, the FAO (2004) developed an exponential model using an approach similar to that adopted by Buchanan et al. (1997), but based on detailed exposure assessment from FDA & FSIS (2003) risk assessments. This model is represented by the following formula:

P ill | D, r

1 exp

rD

r 0, D 0

Equation (5)

where P{ill|D,r} is the probability of illness (occurrence of invasive listeriosis, i.e. response); D is the ingestion dose of a certain number of L. monocytogenes cells; r is a constant parameter that can be intended as the probability of illness in a random person after the ingestion of a single cell (Haas & Eisenberg, 2001). FAO (2004) and FDA & Health Canada (2012) used a median r‐value of 1.06 × 10−12 for a population with increased susceptibility (pregnant, elderly, infants, or immunocompromised), and a median r‐value of 2.37 × 10−14 for a healthy population, for their risk assessments on ready‐to‐eat products and raw milk soft‐cheeses, respectively.

9.5.4  Risk characterisation Few risk assessments have been published to estimate the risk of human listeriosis associated with dairy products or to study what factors can mainly affect it. The comparison of the outputs of these studies is difficult because they used different models and very different data sources. Several risk assessments studies have been conducted on soft cheeses because these products are widely marketed, and are considered to put consumers at a risk. Bemrah et al. (1998) calculated the probability of human listeriosis and death associated with the consumption of soft cheese made from raw milk on a small number of farms. The median probability of illness associated with the consumption of one cheese serving was 1.86 × 10−8 for a high‐risk sub‐population and 9.74 × 10−13 for a low‐risk

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   259

sub‐population. Similarly Sanaa et al. (2004) calculated the risk associated with the consumption of two traditional French soft cheeses made from ​​ raw milk (Camembert from Normandy and Brie from Meaux). The study was based on the level of Listeria spp. in milk, which was estimated by a survey conducted specifically for the purpose of risk assessment. The source of the contamination was not specified, and the contamination and cross‐contamination during the cheesemaking, transport and distribution was not taken into account (Bemrah et al., 1998). However, the risk assessment estimated that the number of cases of severe listeriosis for 100 million serving follows a Poisson distribution with parameters 3.46 × 10−3 for Brie and 5.11 × 10−4 for Camembert cheeses (Sanaa et al., 2004). FS AU & NZ (2009) evaluated the microbial risk of three bacterial strains (including L. monocytogenes) associated with consumption of different varieties of cheeses made from different mammalian raw milk (cow, sheep and goat). A qualitative approach was used to estimate the risk of listeriosis associated with extra‐hard, Swiss‐type, Cheddar, Blue Vein, Feta and Camembert cheeses. The risk assessment estimated a negligible risk (for general population) or a very low/low risk (susceptible populations) for the former three cheeses, while the risk was considered higher for Blue Vein, Feta and Camembert cheeses (low for general population and high for susceptible populations). FDA & Health Canada (2012) assessed the risk of listeriosis associated with the consumption of a soft‐type cheese (Camembert‐like) in the USA and Canada by a quantitative approach. The model included all steps of the food chain (“from farm to fork”), and was based on published data that had been subjected to previous risk assessments (Bemrah et al., 1998; FDA & FSIS, 2003; FAO, 2004; Sanaa et al., 2004) and expert sources. The baseline model, which considered the manufacture of cheese made from pasteurised milk, predicted the mean risk of invasive listeriosis per serving was 7.3 × 10−9, 1.8 × 10−8, 5.2 × 10−9 for the susceptible population, for example, the elderly, pregnant women and immunocompromised, respectively, and 1.2 10−10 for the general population in the USA and Canada (FDA & Health Canada, 2012). Interestingly, the simulation considered the contamination of cheeses (frequency and bacterial concentration) from the plant environment during cheesemaking using data surveyed by Gombas et al. (2003) and FDA & Health Canada (2012). In the USA, the predicted mean risk for invasive listeriosis from consuming a single serving of raw milk cheese was estimated to be 112, 96, 157 and 157 times higher (for the elderly, pregnant women, immunocompromised and general populations, respectively) than the mean risk for pasteurised milk cheese (FDA & Health Canada, 2012).

9.6  Cheese ‐ chemical risk assessment 9.6.1  Background The main biogenic amines (BAs), which can be found in food, are the aromatic monoamines (tyramine, histamine, β‐phenylethylamine and tryptamine), the aliphatic diamines (putrescine, cadaverine) and polyamine (agmatine); however, aliphatic amines (spermidine and spermine) are not produced by direct decarboxylation of amino acids

260   Microbial Toxins in Dairy Products

by micro‐organisms. Cheese and other fermented foods are frequently associated with BAs poisoning (EFSA, 2011a,b), and their formation can occur during cheesemaking and storage (Novella‐Rodriguez et al., 2003). The occurrence of BAs in cheese is considered undesirable since they may be toxic. Tyramine can cause hypertensive reactions in patients treated with MAOI drugs, known as a ‘cheese reaction’ (Takeba et al., 1990), and provoke migraine headaches in sensitive individuals. Histamine can cause ‘histamine poisoning’ (Lehane & Olley, 2000), which is characterised by an incubation period ranging from few minutes to hours; however, the symptoms developed are respiratory distress, nasal secretion, bronchospam, tachycardia, extrasystoles, headache, hypotension, edema and urticaria (Forth et al., 2001; Jarisch, 2004). Other amines, such as putrescine and cadaverine, may potentiate the toxic effects of histamine and tyramine by inhibiting monoamine oxidase, diamine oxidase and hydroxymethyl‐transferase (Bardócz, 1995; Straub et al., 1995). However, BAs are reviewed in detail in Chapter 4.

9.6.2  Biogenic amines in cheese Biogenic amines in cheese originate from product degradation by bacteria. If their concentration in food is high, or the detoxification process of amine oxidase is disturbed, BAs become toxic metabolites responsible of serious human health problems (Ladero et  al., 2010). Biogenic amines accumulation in cheese have been reported by many researchers (Chander et al., 1989; Vale & Glòria, 1997a,b; Ordóñez et al., 1997; Gardini et al., 2001; Roig‐Sagués et al., 2002; Novella‐Rodríguez et al., 2003; Marcobal et al., 2006; Fernández et al., 2006, 2007; Ladero et al., 2009, 2010), and different microbiota, which are used in cheesemaking and can produce BAs in the product and, in brief, they are as follows:

•

•

•

Traditional starter cultures belonging to the following genera Streptococcus, Lactobacillus, Leuconostoc, and Lactococcus (Calles‐Enrìquez et al., 1989; Fernàndez et al., 2004; Nieto‐Arribas et al., 2009; La Gioia et al., 2011; Ladero et al., 2009, 2011a, 2011b; Linares et al., 2011, 2012). Secondary starter cultures, for example, yeasts (genera of Debaryomyces, Yarrowia, Kluyveromyces, Candida, Saccharomyces and Rhodosporidium) when added to pasteurised cheese milk to mimic the natural yeast flora or raw milk and improve cheese flavour (Bockelman, 2010), or blue and white moulds (e.g. Penicillium spp.). Contaminants of milk, such as E. coli, Hafnia alvei, Klebsiella pneumonia, Pseudomonas spp. and Enterococcus spp.

9.6.3  Occurrence of biogenic amines in cheese: hazard and exposure assessment Tyramine levels from 600 mg up to 2 000 mg have been administrated in a meal to cause a minimal systolic blood pressure increase. A dose‐response curve was generated by Patat et al. (1995), where 1100 mg of tyramine corresponds to the effective dose (ED50

Approaches to Assess the Risks/Modelling of Microbial Growth and Toxin Production   261

Table 9.11  Dose‐response (mg kg−1) for humans and alimentary histamine in different cheeses. Products Cheese (variety was not reported) Cheese (variety was not reported) Swiss Gouda

Level and classification

References

1000 – intoxication 1870 – clinical case intoxication 850 – clinical case intoxication

Durlu‐Özkaya (2002) Roig‐Sauges et al. (1998) Taylor et al. (1982) Doeglas et al. (1967)

value, i.e. the dose at which 50% of the individuals responded) and there was no correlation with sex, age or weight (Bieck & Antonin, 1988). The situation is different in those individuals on medication with MAOI drugs, and the impaired f­unction of the enzyme does not allow the complete metabolism of tyramine, leading to elevated plasma tyramine levels after food ingestion. In these cases, much less d­ietary tyramine is needed to produce similar effects (dose‐response curve shifts c­onsiderably to the left so lowering the ED50 value). Doses of 6 mg of tyramine can provoke mild headaches in humans, and 10–25 mg severe headache with intracranial haemorrhage in patients treated with ‘classical’ MAOI (McCabe, 1986). Moreover, 50–150 mg tyramin would be well tolerated by patients under the ‘new generation’ of MAOI treatment (Korn et al., 1988; Dingemanse et al., 1998; Patat et al., 1995, Audebert et al., 1992). For example, the reversible inhibitors of monoamine oxidase, a class of drugs, which selectively and reversibly inhibit the isoform A of the monoamine oxidase‐A (MAO‐A), are particularly effective in treating depression, and dysthymia. Therefore, the bioavailability of tyramine as a dietary constituent seems to be drastically reduced and systematic concentrations are reduced by approximately 2‐ to 3‐fold (Patat et al., 1995; Van den Berg et al., 2003; Azzaro et al., 2006). The syndrome of ‘histamine poisoning’ occurs after an incubation period ranging from few minutes to hours (Lehane & Olley, 2000), and the symptoms observed include respiratory distress, nasal secretion, bronchospasm, tachycardia, extra systoles, headache, hypotension, oedema and urticaria (Jarisch, 2004). Reported outbreaks related to BAs in food include ingestion of cheeses made from raw and pasteurised milk (Lund et al., 2000; Stratton et al., 1991). In cheese, a histamine content of 105 cfu g−1 are detected Butter and cream made from raw milk or milk that has undergone a lower heat treatment than pasteurisation Milk powder and whey powder

Pasteurised milk and other pasteurised liquid dairy products Cheeses made from heat treated milk or whey that have undergone heat treatment Cheeses made from raw milk

Product

Enterobacteriaceae Coagulase‐positive Staphylococci Enterobacteriaceae Enterobacteriaceae Presumptive Bacillus cereus Enterobacteriaceae L. monocytogenes

Coagulase‐positive staphylococci Staphylococcal enterotoxins E. coli

5

Coagulase‐positive staphylococci Coagulase‐positive staphylococci

2

5

5 5 5

10 10 5

0 0 0

0 0 1

0 2

0

5

5 5

2

2

2

0 2

c

5

5

5 5

n

Sampling plan

Enterobacteriaceae Escherichia coli

Micro‐organism

Table 10.3  Microbiological criteria for milk and milk products in European Union legislation.

M

Absence in 10 g

10 cfu g−1

500 cfu g−1

100 cfu g−1

100 cfu g−1

100 cfu g−1

Not detected in 25 g

100 cfu g−1

1000 cfu g−1

105 cfu g−1

(Continued)

Absence in 10 g * 100 cfu g−1 Absence in 25g (applies before the food has left the immediate control of the FBO who has produced it

50 cfu g−1

10 cfu g−1

10 cfu g−1

10 cfu g−1

10 cfu g−1

100 cfu g−1

104 cfu g−1

10 colony forming units (cfu) mL−1 100 cfu g−1 1000 cfu g−1

m

Limits

0

10 5 5 5 30 30 30

L. monocytogenes Salmonella spp. Salmonella spp. Salmonella spp. Salmonella spp. Salmonella spp. Chronobacter spp., Enterobacter sakazakii

0 0

0

0 0

0

0

c

5

n

Sampling plan

L. monocytogenes

Micro‐organism

* Absence in 10 g

* Absence in 25 g

* 100 cfu g−1

m

Limits M

Abbreviations:‐ n = number of sample units comprising the sample which should not exceed ‘M’ in one sample; c = Number of sample units where the bacterial count may be between ‘m’ and ‘M’, the sample is acceptable if the count of the other sample units is ‘m’ or less; m = threshold value for the number of bacteria which should not exceed ‘M’; M = maximum value for the number of bacteria; FBO = Food Business Operator. * Applies for products placed on the market during their shelf‐life. Data compiled from Regulation 2073/2005, as amended (EU, 2005a).

Ready‐to‐eat foods unable to support the growth of L. monocytogenes, other than those intended for infants and for special medical purposes Ready‐to‐eat foods intended for infants and ready‐to‐eat foods for special medical purposes Cheeses, butter and cream made from raw milk or milk that has undergone a lower heat treatment than pasteurisation Milk powder and whey powder Ice-cream, excluding products where the manufacturing process or the composition of the product will eliminate the salmonella risk Dried infant formulae and dietary foods for special medical purposes intended for infants below 6 months of age Dried follow‐on formulae Dried infant formulae and dietary foods for special medical purposes intended for infants below 6 months of age

Product

Table 10.3  (Continued)

Regulatory Measures for Microbial Toxins   303

Raw milk for direct consumption and for the manufacture of products made with raw milk, is addressed in Article 10 of Regulation 853/2004 (EU, 2004b), which permits Member States to adopt national measures adapting the specific requirements which do not compromise the objectives of the regulation being achieved. Furthermore, a Member State may, on its own initiative, maintain or establish national rules prohibiting or restricting the placing on the market of raw milk for direct consumption. This results in different provisions across the EU. Some countries permit sale, usually with strict controls, while other Member States prohibit the sale. Raw milk that is intended for direct consumption must be labelled with the words “raw milk” and, in the case of products made with raw milk for which the manufacturing process does not include any heat treatment or any physical or chemical treatment, the words “made with raw milk” must be used (Annex III, Section XI, Chapter IV, 1 (a) and (b) of Regulation 853/2004 – EU, 2004a).

10.4.2  US milk hygiene and food safety standards The Public Health Service (PHS) in the USA, through its agency the Food and Drug Administration (FDA), does not have direct legal jurisdiction in the enforcement of milk hygiene standards throughout the entire country; its jurisdiction applies only where milk moves across state boundaries. Therefore, the PHS serves only in an advisory capacity and its functions are intended primarily to assist the State and Local Regulatory Agencies in their functions, and its aims are: (a) to promote the establishment of effective and well‐balanced milk hygiene programmes in each state, (b) to stimulate the adoption of adequate and uniform State and local milk control legislation, and (c) to encourage the application of uniform enforcement procedures through appropriate legal and educational measures. The broader interest in the milk hygiene programme of the PHS derives from two important public health considerations. Firstly, the importance of milk as a major nutritional source for the maintenance of good health, especially for young children and the elderly; the consequence of this is the PHS promotes increased milk consumption. Secondly, the recognition that, historically, milk has had potential to serve as a significant vehicle for the transmission of disease and indeed in the past been associated with major disease outbreaks. The PHS and FDA activities in the area of milk hygiene and food safety began at the start of the twentieth century with studies on the role of milk in the spread of disease. These led to the recognition that effective control required the application of hygiene procedures and practices throughout the full product chain from production, through handling, processing and product distribution. There followed research to identify and evaluate the hygiene requirements that should be adopted and enforced to control disease. These included studies that led to improvement of the pasteurisation process. In 1950, it was proposed that raw milk containing the causative organism of Q‐Fever (Coxiella burnettii) might be the cause of a significant number of cases of this disease in California (Bell et al., 1950). The effect of pasteurisation on this organism was investigated (Lennette et al., 1952) and, based on a research project by PHS and University of California (Enright et al., 1957) that showed that there could be some survivors if

304   Microbial Toxins in Dairy Products

high numbers were present in raw milk, led to the recommendation that the low heat treatment (LHT) pasteurisation temperature be raised from 61.7°C for 30 min to 62.8°C for 30 min to ensure adequate destruction. No change in the high temperature and short time (HTST) pasteurisation was deemed necessary as a result of this research. Over the years, the incidence of milk borne illness in the United States, which constituted 25% of all disease outbreaks due to infected foods and contaminated water in 1938, has been reduced significantly. In the Pasteurised Milk Ordinance 2013 Revision, it was stated that milk and fluid milk products are now associated with less than 1% of such reported outbreaks (PHS, 2014). Nowadays, the FDA is recognised as having made a major contribution to the improvement of the national milk supply through its technical assistance, training, research, standards development, evaluation and certification activities. Despite the progress that has been made, occasional milk borne outbreaks still occur, and this requires continued vigilance throughout the full product chain from farm to fork (or “from stable to table”, as this principle is expressed commonly in the USA). The situation has been complicated by to the introduction of new products and processes, the use of new packaging materials and new marketing patterns. Thus, considerable efforts continue to be expended in the development and use of the HACCP based systems throughout the entire dairy industry in the United States. In the USA, the federal legislation on milk and milk products and their hygienic production and food safety is laid down in the Grade “A” Pasteurised Milk Ordinance (PMO) and in the Code of Federal Regulations (National Archives and Records Administration – NA and RA, 2014). The FDA is now responsible for about 80% of the food supply in the United States. The exceptions are as regards the safety, wholesomeness, labelling and packaging of meat, poultry and certain egg products, which areas are the responsibility of the United States Department of Agriculture (USDA). In 1924, the PHS developed the model regulation that became known as the Standard Milk Ordinance for voluntary adoption by State and Local Milk Control Agencies. To provide for the uniform interpretation of this Ordinance, an accompanying Code was published in 1927, which provided administrative and technical details on compliance. On‐going revisions of the PMO incorporate new knowledge and technology into effective and practicable legislation. The Ordinance has been revised and updated many times in the intervening period, the title changing to the present one in 1965. This model of milk regulation, now titled the Grade “A” Pasteurised Milk Ordinance (Grade “A” PMO), 2013 Revision, incorporates the provisions governing the processing, packaging, and sale of Grade “A” milk and milk products, including buttermilk and buttermilk products, whey and whey products, and condensed and dried milk products (PHS, 2014). The PMO is used as the hygiene regulation for milk, and milk products, for interstate carriers; it is recognised by the Public Health Agencies, the milk industry, and many others as the national standard for milk hygiene. It represents a consensus of current knowledge and experiences and, as such, is said to represent a practical and equitable milk hygiene standard in the USA. It has been adopted by 46 of the 50 States for their own standards  –  California, Pennsylvania, New York and Maryland being the exceptions, and these have adopted similar standards. However, where it is adopted locally, its enforcement becomes a function of the Local or State authorities.

Regulatory Measures for Microbial Toxins   305

The PHS and the FDA do not produce the PMO on its own. It is developed with input and assistance from Milk Regulatory and Rating Agencies at every level of Federal, State, and Local Government, including both Health and Agriculture Departments, all segments of the dairy industry, including producers, milk processors, equipment manufacturers, and representative dairy associations, inputs from educational and research institutions and with comments from many individual hygienic experts and others. The Grade “A” PMO is the basic standard used in the voluntary State‐PHS and FDA Programme for the Certification of Interstate Milk Shippers. This is a programme in which all 50 States participate, together with the District of Columbia and  Trust Territories in the USA. The National Conference on Interstate Milk Shipments (NCIMS) recommends changes and modifications to the FDA at its biennial conferences. The standards for Grade “A” raw milk for processing in the USA are outlined in Table 10.4. Pasteurisation is defined as the process of heating every particle of milk in properly designed and operated equipment, and held at or above the temperatures for at least the times specified as outlined in Table 10.5. In addition, the standards for Grade “A” Pasteurised Milk and microbiological standards for some concentrated or dried milk and whey products are shown in Tables 10.6 and 10.7, respectively. Ultra‐pasteurisation means that milk or milk product shall be heat treated as outlined in Table 10.5 either before or after packaging so as to produce a product that has an ESL under refrigerated conditions. The definition of Aseptic Processing is a milk product that has been subjected to sufficient heat processing and packaged in a hermetically sealed container so that the product meets the definition of commercial sterility as outlined in §113.3(e)1 of the Code of Federal Regulations (NA & RA, 2014). This requires the milk or milk product to be free of micro‐ organisms capable of growing under normal non‐refrigerated storage and distribution and is free of viable micro‐organisms (including spores) of public health significance. The PMO also incorporates administrative and technical requirements for the manufacture of condensed and dried milk products, and condensed and dried whey products included in the Grade “A” Condensed and Dry Milk Ordinance – Supplement I to the Grade “A” Pasteurised Milk Ordinance (PHS, 2014). The PMO runs to 430 pages, and Table 10.4  Standards for grade “A” raw milk and milk products for pasteurisation, u­ltra‐pasteurisation and aseptic processing supply in the US legislation (PMO). Criterion Temperature

Bacterial limits – individual producer Bacterial limits – mixed milk prior to pasteurisation Drugs/antibiotics Somatic cell count (SCC) – individual producer

Requirement Cooled to ≤10°C within 4 h or less of commencement of first milking and to ≤7°C within 2 h of completion of milking, provided that the blend temperatures after the first milking and subsequent milkings do not exceed 10°C ≤ 100 000 colony forming units (cfu) mL−1 prior to mixing with other producers milk ≤ 300 000 cfu mL−1 No positive results on drug residue detection methods; β‐lactam detection methods acceptable ≤750 000 cells mL−1

Abbreviation:‐ PMO = Pasteurised Milk Ordinance. Data compiled from (PHS, 2014).

306   Microbial Toxins in Dairy Products

Table 10.5  Heat treatment requirements specified in the US legislation (PMO and CFR). Type

Temperature (minimum)

Pasteurisation

Time (minimum) 30 min 15 s 1.0 s 0.5 s 0.1 s 0.05 s 0.01 s 2 s

63°C 72°C1 89°C 90°C 94°C 96°C 100°C 130°C 1

Ultra‐pasteurisation

Abbreviations:‐ PMO = Pasteurised Milk Ordinance; CFR = Code of Federal Regulations. 1  If the fat content is ≥10 g 100 g−1 then these minimum temperature should be increased by 3°C. Data compiled from CFR (NA & RA, 2014) and PMO (PHS, 2014).

Table 10.6  Standards for grade “A” pasteurised milk and milk products in the US legislation (PMO). Criterion Temperature Bacterial limits – individual producer Coliforms Phosphatase

Drugs/antibiotics

Requirement Cooled to ≤7°C and maintained thereat ≤20 000 colony forming units (cfu) mL−1 ≤10 cfu mL−1