Modifying Flavour in Food

Modifying flavour in food

Related titles: Flavour in food (ISBN 978-1-85573-960-4) The flavour of a food is one of its most important qualities. Edited by two leading authorities in the field, and with a distinguished international team of contributors, this important collection summarises the wealth of recent research on how flavour develops in food and is then perceived by the consumer. The first part of the book reviews ways of measuring flavour. Part II looks at the ways flavour is retained and released in food. It considers the way flavour is retained in particular food matrices, how flavour is released during the process of eating, and the range of influences governing how flavour is perceived by the consumer. Flavour in food guides the reader through a complex subject and provides the essential foundation in both understanding and controlling food flavour. Taints and off-flavours in foods (ISBN 978-1-85573-449-4) Taints and off-flavours are a major problem for the food industry. Part I of this important collection reviews the major causes of taints and off-flavours, from oxidative rancidity and microbiologically derived offflavours, to packaging materials as a source of taints. Part II discusses the range of techniques for detecting taints and off-flavours, from sensory analysis to instrumental techniques, including the development of new rapid, on-line sensors. Colour in food – Improving quality (ISBN 978-1-85573-590-3) The colour of a food is central to consumer perceptions of quality. This important collection reviews key issues in controlling colour quality in food, from the chemistry of colour in food to measurement issues, improving natural colour and the use of colourings to improve colour quality. Details of these books and a complete list of Woodhead’s titles can be obtained by: • •

visiting our web site at www.woodheadpublishing.com contacting Customer Services (e-mail: sales@woodhead-publishing. com; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 130; address: Woodhead Publishing Ltd, Abington Hall, Abington, Cambridge CB21 6AH, England)

Modifying flavour in food Edited by Andrew Taylor and Joanne Hort

Cambridge, England

Published by Woodhead Publishing Limited, Abington Hall, Abington Cambridge CB21 6AH, England www.woodheadpublishing.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA © 2007, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-074-8 (book) Woodhead Publishing ISBN 978-1-84569-336-7 (e-book) CRC Press ISBN 978-1-4200-4389-1 CRC Press order number WP4389 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elementary chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by SNP Best-set Typesetter Ltd., Hong Kong Printed by TJ International Limited, Padstow, Cornwall, England

Contents

Contributor contact details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2

3

Modifying flavour: an introduction. . . . . . . . . . . . . . . . . . . . . . . . . . A. J. Taylor and J. Hort, University of Nottingham, UK 1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flavouring substances: from chemistry and carriers to legislation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Baines, Baines Food Consultancy Ltd, UK 2.1 The importance of olfaction in the appreciation of flavour. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Flavouring substances in foods. . . . . . . . . . . . . . . . . . . . . . . 2.3 Flavouring substances legislation . . . . . . . . . . . . . . . . . . . . . 2.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extraction of flavourings from natural sources. . . . . . . . . . . . . . . . G. Cravotto, University of Turin, Italy and P. Cintas, University of Extremadura, Spain 3.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 General overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Supercritical fluid extraction: SC-CO2 . . . . . . . . . . . . . . . . . 3.4 Continuous subcritical water extraction (CSWE) . . . . . . . 3.5 Ultrasound-assisted extraction (UAE) . . . . . . . . . . . . . . . .

x 1 1 9

10

10 12 25 36 37 37 41

41 42 50 52 53

vi

Contents 3.6 3.7 3.8 3.9 3.10 3.11

4

Microwave-assisted extraction (MAE) . . . . . . . . . . . . . . . . Extraction in the analysis of flavours. . . . . . . . . . . . . . . . . . Drying methods and solvent distillation . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

From fermentation to white biotechnology: how microbial catalysts generate flavours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. G. Berger, Leibniz University of Hannover, Germany 4.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Flavour formation along known pathways . . . . . . . . . . . . . 4.3 Flavours from complex substrates . . . . . . . . . . . . . . . . . . . . 4.4 White biotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Sources of further information . . . . . . . . . . . . . . . . . . . . . . . 4.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54 56 58 59 59 59

64 64 67 78 81 83 85 86

5

New developments in yeast extracts for use as flavour enhancers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 B. Noordam and F. R. Meijer, DSM Food Specialties, The Netherlands 5.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 5.2 Developments in yeast extracts . . . . . . . . . . . . . . . . . . . . . . 96 5.3 Materials and methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 5.4 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 5.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 5.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

6

Chiral chemistry and food flavourings . . . . . . . . . . . . . . . . . . . . . . . S. Serra, C.N.R. Milano, Italy 6.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Chiral flavour – synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Chiral flavour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Key flavour compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Sources of further information and advice . . . . . . . . . . . . . 6.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

Formulating low-fat food: the challenge of retaining flavour quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Hort and D. Cook, University of Nottingham, UK 7.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Lowering the fat content of food: what happens to the flavour? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Strategies for replacing fat in foods and the implications for flavour . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

107 107 108 115 121 127 128

131 131 132 134

Contents

vii

Why is fat so hard to replace? . . . . . . . . . . . . . . . . . . . . . . . Representation of fat in the brain . . . . . . . . . . . . . . . . . . . . Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

138 140 142 143

New pungent and cooling compounds for use in foods. . . . . . . . . C. C. Galopin, Givaudan Flavors Corporation, USA 8.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3 Fundamental differences between chemesthetics and flavours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Physiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 Common pungent chemicals and their activity . . . . . . . . . 8.6 Common cooling compounds and their activity. . . . . . . . . 8.7 Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

146

Controlled release of flavour in food products . . . . . . . . . . . . . . . . G. Reineccius, University of Minnesota, USA 9.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Industrial approaches for protecting flavourings from deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Industrial approaches to achieve controlled release of flavourings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4 Needs in flavour encapsulation/controlled release . . . . . . 9.5 Advice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

169

7.4 7.5 7.6 7.7 8

9

10

11

Developments in sweeteners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Kemp, Cadbury Schweppes, UK and M. Lindley, Lindley Consulting, UK 10.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Mechanism of sweetness perception . . . . . . . . . . . . . . . . . . 10.3 Novel sweeteners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4 Sweetness potentiators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5 Sweetness inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6 Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7 Sources of further information and advice . . . . . . . . . . . . . 10.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

146 146 147 151 154 157 162 163 163

169 170 174 181 182 182 185

185 187 189 195 197 197 198 199

Enhancing umami taste in foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 J. B. Marcus, Kendall College, USA 11.1 Umami: what it is, what it does and how it works. . . . . . . 202 11.2 Culinary history of umami in flavour enhancement . . . . . 203

viii

Contents 11.3 11.4 11.5 11.6 11.7 11.8 11.9 11.10 11.11 11.12 11.13 11.14 11.15 11.16 11.17 11.18 11.19 11.20

12

13

Scientific background of umami in flavour enhancement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How the culinary aspects and science of umami interact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Asian condiments that impart umami taste and tasteactive components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Western foods that impart umami taste and tasteactive components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other umami taste-activators . . . . . . . . . . . . . . . . . . . . . . . . Taste-active components and umami synergy . . . . . . . . . . Umami formed in ripening, drying, curing, ageing and fermenting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Practical aspects of umami in consumer acceptance . . . . . Consumer applications of umami. . . . . . . . . . . . . . . . . . . . . Food technology applications of umami . . . . . . . . . . . . . . . Umami applications that maximise flavour, food acceptance and food preference. . . . . . . . . . . . . . . . . . . . . . Umami applications for the development and enhancement of recipes and products . . . . . . . . . . . . . . . . . US umami initiatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . European umami initiatives . . . . . . . . . . . . . . . . . . . . . . . . . Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of further information and advice . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

203 204 205 205 206 207 207 208 209 210 210 215 216 217 217 218 218 220

Bitter blockers in foods and pharmaceuticals . . . . . . . . . . . . . . . . . R. McGregor, Linguagen Corporation, USA 12.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.2 Why reduce bitterness in foods and pharmaceuticals? . . . 12.3 Current approaches to reducing bitterness. . . . . . . . . . . . . 12.4 Advantages to bitter blocked pharmaceutical formulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.5 The science of taste perception . . . . . . . . . . . . . . . . . . . . . . 12.6 Identifying compounds that decrease the perception of bitterness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.7 Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.8 Sources of further information and advice . . . . . . . . . . . . . 12.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

221

Masking agents for use in foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Gascon, Wixon Inc., USA 13.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2 How masking agents work . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3 Ingredients used to formulate masking agents. . . . . . . . . .

232

221 221 223 224 225 228 230 230 230

232 233 235

Contents 13.4 13.5 13.6 13.7 14

ix

What to consider when working with masking agents . . . Masking agents: when to use or not to use. . . . . . . . . . . . . Outlook and perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

239 240 240 241

Selecting the right flavourings for a food product . . . . . . . . . . . . . K. B. de Roos, Givaudan Nederland B.V., The Netherlands 14.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 Creation of the desired flavour profile . . . . . . . . . . . . . . . . 14.3 Stability of the flavouring . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.4 Solving flavour release problems . . . . . . . . . . . . . . . . . . . . . 14.5 Solving flavour stability problems in products . . . . . . . . . . 14.6 Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.7 Sources of further information and advice . . . . . . . . . . . . . 14.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

243

Index

243 244 247 249 253 265 267 267

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

Contributor contact details

Chapter 1

Chapter 2

A. J. Taylor Professor of Flavour Technology Division of Food Sciences University of Nottingham Sutton Bonington Campus Loughborough LE12 5RD UK

D. Baines Baines Food Consultancy Ltd 22 Elizabeth Close Thornbury Bristol BS35 2YN UK e-mail: [email protected]

e-mail: Andy.Taylor@nottingham. ac.uk and J. Hort Lecturer in Sensory Science Division of Food Sciences University of Nottingham Sutton Bonington Campus Loughborough LE12 5RD UK e-mail: joanne.hort@nottingham. ac.uk

Chapter 3 Giancarlo Cravotto* Dipartimento di Scienza e Tecnologia del Farmaco University of Turin Via P. Giuria 9 I-10125 Torino Italy e-mail: [email protected] and

Contributors Pedro Cintas Departamento de Química Orgánica Facultad de Ciencias University of Extremadura Avenida de Elvas s/n e-06071 Badajoz Spain e-mail: [email protected]

Chapter 4 R. G. Berger Leibniz University of Hannover Centre of Applied Chemistry Institute of Food Chemistry Wunstorfer Strasse 14 D-30453 Hannover Germany e-mail: [email protected]. de

Chapter 5 B. Noordam and F. R. Meijer DSM Food Specialties R&D/FTD/Application PO-box 1 2600MA Delft The Netherlands e-mail: [email protected] [email protected]

xi

Chapter 6 S. Serra C.N.R. Istituto di Chimica del Riconoscimento Molecolare, presso Dipartimento di Chimica Materiali ed Ingegneria Chimica del Politecnico Via Mancinelli 7 20131 Milano Italy e-mail: [email protected]

Chapter 7 J. Hort* and D. Cook Division of Food Sciences University of Nottingham Sutton Bonington Campus Loughborough LE12 5RD UK e-mail: joanne.hort@nottingham. ac.uk

Chapter 8 C. C. Galopin Givaudan Flavors Corporation Research and Development 1199 Edison Drive Cincinnati OH 45216 USA e-mail: christophe.galopin@pmusa. com

xii

Contributors

Chapter 9

Chapter 11

G. Reineccius Dept. Food Science and Nutrition University of Minnesota 1334 Eckles Ave St. Paul MN 55108 USA

J. B. Marcus Kendall College School of Culinary Arts 900 N. North Branch Street Chicago Illinois 60622 USA

e-mail: [email protected]

e-mail: [email protected] [email protected] [email protected]

Chapter 10 S. Kemp Head of Global Sensory and Consumer Guidance Global Science Centre Cadbury Schweppes The Lord Zuckerman Research Centre Pepper Lane Reading RG6 6LA UK e-mail: [email protected] and M. G. Lindley Lindley Consulting 17 Highway Crowthorne Berkshire RG45 6HE UK e-mail: [email protected]

Chapter 12 R. McGregor Linguagen Corp. 2005 Eastpark Boulevard Cranbury NJ 08512-3515 USA e-mail: richard.mcgregor@ linguagen.com

Chapter 13 M. Gascon Wixon Inc. 1390 East Bolivar Avenue St Francis WI 53235 USA e-mail: mariano_gascon@wixon. com

Chapter 14 K. B. de Roos De Eelinkes 37 7101 PZ Winterswijk The Netherlands e-mail: [email protected]

1 Modifying flavour: an introduction A. J. Taylor and J. Hort, University of Nottingham, UK

1.1

Introduction

Modifying the flavour of foods by adding other materials is an ancient practice. The use of salt as a preservative, and as a flavour enhancer, has been practised for hundreds, if not thousands, of years. Similarly, the use of heat to cook food has its roots in antiquity and presumably made the food easier to eat (compare the texture of cooked meat versus raw) and more digestible (native starch is hard to digest compared with gelatinised). It also reduced the load of food poisoning organisms on the food and increased food safety while adding more volatile compounds to the flavour profile through chemical reactions such as caramelisation and the Maillard reaction. Today, we cook foods for all these reasons but we are more interested in the flavour aspect, as the other issues are no longer of primary importance, at least for that part of the world that has an adequate supply of food. Chefs routinely use herb and spice flavourings to complement the natural flavours of staple foods such as meat, fish and cereal products, and many popular dishes owe their characteristic flavour, not to the major ingredients, but to the minor added flavours. Thus modifying food flavour is a longstanding practice that has developed in response to various human factors, starting with our initial attempts to ensure survival of the human race but now focused more on the hedonics of food. Food production and consumption patterns have changed markedly in Western society over the past few decades, with a move away from home cooking using basic ingredients, to mass-produced foods that can be ‘cooked’ at home simply by reheating in an oven. Convenience foods and preprepared meals form much of the average person’s diet and the scale and scope of food manufacture create different preparation techniques

2

Modifying flavour in food

compared with traditional cooking in the home, or restaurant cooking, both of which serve only a relatively small number of people. For example, scaling up the dishes prepared by development chefs is notoriously difficult as the whole time and temperature conditions of cooking can change. A good example is sauce manufacture where the come-up time (the time to reach the desired temperature) of a 2 tonne batch cooker is of the order of 60 min. This is followed by a cooking time and then further holding times during packing, meaning that products can be held at elevated temperatures for several hours. This change in thermal processing has a significant effect on the chemistries that produce or destroy flavour compounds, with the result that the flavour of a 2 tonne batch will not necessarily resemble the flavour of the 200 g sample from the development kitchen. One way of addressing this situation is to use the flavourings produced by flavour houses. They can be added to the batch late in the cook and the level of dosage can be varied to produce the desired flavour intensity. Another example of a mass manufacture challenge, is the necessity to maintain the flavour of foods during shelf-life and/or during the reheating cycle. The flavour industry has developed special forms of flavour to address the shelflife issue (e.g. encapsulated flavours) and has the knowledge to select flavour compounds that will perform better in reheating scenarios than the naturally occurring flavours. Added flavourings are also essential ingredients in the production of goods such as confectionery, beverages, cakes, biscuits and snack foods. In all these examples, the flavour of the base composition is modified in some way to deliver the desired properties to the foods. Although mass-produced foods are very popular, in some parts of the world, the food and flavour industry is facing a new challenge as consumer organisations raise questions about the quality of the mass-produced products. In an open society, this is a valid topic that should be debated in an open way. However, the tone taken by many of the consumer groups (and the media) is quite aggressive and is based on the premise that the food and flavour industries are providing unhealthy food and that they are only out to make money. In the United Kingdom, at least, rational discussion is largely missing from the newspaper and television coverage and, in these authors’ opinion, the interviewers rarely challenge the representatives of the consumer groups as to their credentials and their motives since, after all, many of these people are also making money and creating careers out of the situation. The message that consumers receive from these sources is that any food component considered synthetic must be bad and that anything natural must be good. Although irrational, this is a deeply rooted belief and is illustrated by a recent UK government pronouncement that ‘natural’ remedies can make health claims without going through the extensive testing procedures required for synthetic drugs (MHRA 2004). Although the issues described above are non-technical and, although the assertion about unhealthy food and profit making is largely untrue, these issues have had a major effect on the attitude of consumers and also the

Modifying flavour: an introduction

3

approach taken by food and flavour manufacturers. Some companies are changing product formulation so that all ingredients can be declared as natural. Other companies are exploring an approach that decrees that their products will be made using only ingredients that could be found in ‘the kitchen store cupboard’. Additives are thus excluded and only ingredients recognised by the consumer (e.g. salt, wheat flour, butter) can be used. These consumer-driven changes impact on the flavour industry in two ways. First, they are required to produce greater quantities of natural flavourings and, second, the debate about whether mass-produced natural flavourings will be considered as ‘kitchen store cupboard’ ingredients (and therefore accepted by consumers) remains to be settled. In this latter case, maybe the common domestic usage of flavourings such as vanilla extract will lead to a certain level of acceptance. Another current, associated concern is the problem of obesity in many Western countries. There is considerable pressure on food manufacturers to produce foods with reduced amounts of fat, salt and sugar to address the nutritional guidelines issued by governments. However, the new products have to taste just as good as the originals or the population will not consume them and therefore will gain no nutritional benefit. The paragraphs above largely describe the market factors that influence flavour production and marketing, commonly called ‘market pull’. There is also a complementary, research-led ‘push’ which influences flavour production and it is interesting to examine how research findings have influenced the way that flavours are developed commercially. Such an examination is subjective and the following topics are the editors’ interpretation of research progress since 1995. The most notable development is our increased understanding of the flavour receptors in the nose and mouth. Buck and Axel received the 2004 Nobel prize in Physiology and Medicine for their work on olfactory receptors, an award that raised the profile of flavour research worldwide. Parallel work on the taste receptors, and the associated transduction mechanisms, have increased our knowledge of the way tastants act on the system. These advances have resulted in some direct applications, for example, the Senomyx and Linguagen companies in the United States are using knowledge of the taste receptors, coupled to molecular biology techniques, to screen compounds for various taste activities, such as the ability to block bitterness or to enhance flavour. Other potential applications will need more knowledge on the systems biology of the taste and smell systems. Examples of these applications are an understanding of aroma interactions at the receptor level (i.e. how does one odour molecule affect the binding of a different odour), the dynamics of binding, the potential interactions between neural signals as they are processed at the different levels in the brain and, the ultimate goal, an understanding of why a particular mixture of aroma and taste compounds produces such strong emotions when sensed.

4

Modifying flavour in food

Knowledge of the mechanisms and kinetics of the Maillard reaction have also advanced, thanks to the application of new analytical and data processing tools. The pathways of reactants can be followed using labelling techniques such as CAMOLA (Schieberle, 2005) while on-line monitoring of the process can be achieved (Channell and Taylor, 2005; Pollien et al., 2003). Models of the Maillard reaction, which predict the products from different reactant levels and different processing conditions, are also being developed (Martins and Van Boekel, 2005). There are obvious applications for these techniques in optimising thermal flavour production or in controlling the amounts of undesirable products such as acrylamide and 3-chloro-propan1,2-diol (MCPD). The way in which flavour components modify the perceived flavour of foods has also advanced from a stage where we were simply aware of the situation, to an understanding of the interactions involved at a qualitative and quantitative level. Many examples exist in the literature but the observation that the perceived flavour of a solution decreases as viscosity increases, illustrates the point well (Pangborn and Szczesniak, 1974). One of the original explanations for this behaviour was that viscosity hindered the release of aroma from the sample when eaten and caused a decrease in aroma signal. This hypothesis was disproved (Hollowood et al., 2002) and, coupled with work from other laboratories, reinforced the idea that flavour was a multi-modal construct in which cross-modal interactions were significant. The seminal work of Dalton et al. (2000) who showed that subthreshold levels of benzaldehyde and sweetener interacted cognitively to produce a perception, has been complemented by data from our laboratory on taste – aroma and taste – viscosity – aroma interactions in sweet and savoury flavour (see for example Davidson et al., 1999; Pfeiffer et al., 2005). We now know that tastant release is altered in viscous solutions and the change in sensory properties is caused by a reduced taste–aroma interaction. Commercial examples of this knowledge are now evident, such as the recently launched chewing gum in the United States with the ‘ridiculously long-lasting flavour’ (Cadbury Adams, Stride gum). The prolonged flavour sensation seems to work by delivering sugar over a longer time period to enhance the sensation of mintiness as suggested previously (Davidson et al., 1999). Other, not so obvious, applications are in confectionery and reduced fat foods. The result is that the principles of multi-modal flavour perception have been established through fundamental research and are now being used by industry to address real needs; this is the ultimate goal of applied scientific research and is gratifying to see the loop being closed. With the ready availability of analytical methods to monitor flavour release in vivo (Linforth and Taylor, 2003) and for brain imaging to investigate cognitive interactions between taste and aromas (Verhagen and Engelen, 2006), further progress in our understanding is envisaged. For the reasons outlined above, the manufacture, development and uses of flavourings are driven by a very wide range of economic, scientific, legal

Modifying flavour: an introduction

5

and social factors. The current needs for modifying flavour are therefore different from those experienced by the industry in the mid-1990s and the content of this book attempts to address some of the issues that are relevant at the current time. The flavouring industry has used technology to develop a range of pleasant flavours with high flavour potency that can be used to modify the overall flavour of a product. The development of those technologies has formed the basis for R&D in flavour houses across the world and has led to novel and improved ways of modifying the flavour of foods to meet the needs of industry and consumers. The book starts with an account of how the flavour industry has developed its technology over the past century or so (Chapter 2; Baines). The initial focus on chemical synthesis of the flavour compounds found in nature created an innovative industry which improved the flavour, and the general quality, of foods from the turn of the 20th century. This was followed by the adoption of the newly discovered separation and identification techniques in the 1960s to extract new flavour compounds from natural materials and then synthesise analogues to overcome the problems associated with natural sources, such as seasonality and variations in flavour content. The current trend is for ‘natural’ flavours and ingredients and, in some countries, this is being pursued to a very high level. Baines finishes his chapter by pointing out that the current EU legislation is likely to favour the development of new flavours in markets outside the EU, owing to the costs and complications of clearing new flavouring compounds for use in the EU. With the consumer pressure for natural food ingredients, the next two chapters tackle the production of natural flavours (Chapters 3 and 4). This could be considered as modifying the form in which the flavour is delivered, i.e. using natural means to manufacture the flavours rather than organic chemical syntheses. Chapter 3 is concerned with the extraction of natural flavourings from plant materials and how the new extraction techniques can provide high-quality materials while also meeting the additional needs of protecting the environment and ensuring sustainability. The use of solvents such as supercritical carbon dioxide and pressurised water has many environmental attractions as they are natural and non-toxic agents. Like many potential solutions to the problems of energy and environment, the full impact of these techniques has not been fully considered, e.g. although CO2 is environmentally friendly, how much extra energy do we spend extracting, compressing, transporting and applying CO2 in the extraction of flavours? It will need a bigger book than this to fully discuss these issues but the adverse energy requirements of bioethanol production have already shown the potential problems of these ‘green’ technologies. Chapter 4 describes new advances in biotechnology to produce natural flavours in a vast array of microorganisms, using our advanced knowledge of metabolic pathways which can now be expressed in most any organism through genetic modification. It is ironic that while genetically modified food is not acceptable to

6

Modifying flavour in food

some sections of Europe, drugs and flavours produced by these means are totally acceptable. The extraction of tastants from yeast is described in Chapter 5. Yeast extracts are well known for their meaty and savoury (bouillon) flavour properties but new methodology delivers yeast extracts with little or no savoury flavour and with very interesting enhancing properties. This is an exciting development as the components of these yeast extracts seem to exert a cross-modal effect on some unexpected sensory attributes, e.g. the mouthfeel of a low-fat dairy product. The next set of chapters represents some of the more fundamental research work that will provide the future scientific knowledge for further innovations by the flavour industry. Chapter 6 describes fundamental work on the chirality of flavours and it has long been recognised that stereoisomers of certain compounds have different odours, with menthone being the classic example (Fisher and Scott, 1997). However, with the consumer demand for natural flavours, there is also a drive to mimic as closely as possible the form in which flavours are found in nature. The authors of Chapter 6 remark that although consumers apparently want their flavours to be natural, the costs of delivering flavour enantiomers that match the natural state will have to be accepted by the consumer first; then the industry will need to establish the feasibility of the current laboratory-scale methods for large-scale production. This chapter neatly illustrates how the development of technologies depends very much on the potential markets and is an example of an area where technology is available at a fundamental level but where the ‘market pull’ is not yet established. In Chapter 7, the role of fat in food flavour is discussed from the fundamental point of view. The basic tenet is that fat has many roles in our appreciation of flavour quality. We know of its role in the partition (and therefore the release) of aromas, in the viscosity of the food as well as the all-important, but difficult to define, aspect of mouthfeel. Thus, if fat content in a food is changed, we can expect a host of effects on sensory quality and we will need not just a single fat replacer but a range of adjustments to recreate the many properties contributed by fat to a food’s overall flavour and acceptance. Indeed, the authors of this chapter question whether manufacturers will ever be able to replace fat successfully considering how adept the sensory mechanisms of the human body are at detecting fat. Therefore, applying this knowledge to the commercial situation, it seems that if we want to make a low-fat analogue of an existing product, we need to add a range of materials to fully replace fat in foods. While this may create a technical solution, it is unlikely to meet the consumer demands for ‘additive-free’ food products. The alternative solution is to create new food products with lower fat levels and a different flavour from conventional products. There are signs of both types of activity in the food area. New microstructures patented by Unilever (see, for example, Appelquist et al. 2000) were designed to decrease fat content while maintaining effective

Modifying flavour: an introduction

7

flavour delivery and demonstrate the development of new food structures. The development of new low-fat products can be seen in many supermarkets where clearly identified product ranges have been developed and marketed using words such as ‘lite’ or ‘low-fat’ to differentiate them from conventional products. The next chapter in this section (Chapter 8), addresses the subject in a fundamental way by considering how pungent and cooling sensations are perceived by humans. With this knowledge, methods for manipulating the sensation by using appropriate structures and delivery mechanisms can be postulated and trialled. Linked to this is the concept of controlling flavour delivery to elicit optimum and novel flavours. Although flavour delivery can be easily controlled in model systems to study fundamental aspects of flavour perception (Hort and Hollowood, 2004), controlling delivery in real foods is much more difficult. Factors such as cooking and oral processing (e.g. mastication and hydration) can remove the effects of some delivery systems and cancel out any potential advantage. Reineccius (Chapter 9) reviews the methods currently available to modify the delivery of flavours through encapsulation techniques and gives examples of applications where encapsulation offers advantages. The next group of chapters (Chapters 10–13) describes methods and materials to modify flavours, by masking or enhancement. One of the main applications for these techniques is the result of food reformulation to reduce fat, salt and sugar in our diet or to add components considered as nutritionally beneficial such as fibre or antioxidants which have inherent poor taste properties. Despite the simplistic view of some government agencies that meeting nutritional requirements is simply a question of adding or removing components from food, manufacturers and flavour houses who want to satisfy consumers’ flavour expectations know differently. Changing food formulations in this way creates flavour imbalance, and sensory analysis often identifies flavour defects such as bitterness or poor mouthfeel, which adversely affect the overall flavour of foods. The first topic to be addressed is the use of sweeteners in food. Although considerable advances have been made in identifying alternatives to sucrose, the economic manufacture of foods with the same clean taste quality and temporal sweetness as sucrose eludes researchers. Chapter 10 reviews a range of compounds that have been investigated for their sweet taste. Some mention is also given to those compounds that may suppress sweetness, enabling the use of sucrose for its other functional properties without adding sweetness. Although only recently recognised as the fifth taste, umami with its flavour-enhancing capability has been recognised in culinary circles for centuries. In Chapter 11, Marcus summarises what is known about the compounds and sensory mechanisms involved in umami sensation. She catalogues the presence of umami compounds in a wide range of foods and illustrates particular ingredients that are used in cooking that make use of

8

Modifying flavour in food

umami to optimise flavour. The information in this chapter complements some of the material presented in Chapter 5 on the flavour-enhancing properties of yeast extracts. The next two chapters investigate the use of blocking or masking agents to modify the flavour of foods. In Chapter 12, McGregor reviews the current approaches taken to reduce bitterness and highlights how an increased understanding of taste receptor mechanisms can, and will continue to, advance the identification of suitable bitter blockers. In Chapter 13 Gascon outlines some of the masking agents available in the industry’s ‘toolbox’. However, more importantly, he highlights the fact that every product is likely to need its own unique ‘masking’ solution and that the use of blends of masking agents is more likely to be the solution. Without doubt, masking bitterness is a major challenge to both the food and pharmaceutical industry. Many of the previous chapters of the book have addressed modifications to the production of flavours to meet the needs of consumers or to modify flavours as a result of reformulation. However, combining these flavour ‘raw materials’ into successful commercial flavours that will function in a wide variety of foods is a major task. In Chapter 14, the science behind the design of flavours that will function in the different food matrices is explained from the commercial angle by de Roos. Our understanding of how flavour is released from foods, before eating and during eating has advanced considerably over the last 10 years and it is satisfying to see these basic principles being applied and further developed to improve food flavours. This is an area in which one of the editors has had a significant interest and it is interesting to see how much the field has changed since an early review (Taylor, 1996). Besides introducing the current work on modifying flavour to meet market and consumer needs, the chapters collectively demonstrate the breadth of scientific disciplines used in flavour production and application. Extraction involves principles of chemical engineering, physical and organic chemistries, bioprocessing brings together biochemistry and molecular biology, while chirality needs sophisticated analyses coupled to organic syntheses. The development of the various agents used to enhance or mask flavours requires knowledge of flavour perception at a receptor and cognitive level, coupled with the ability to produce compounds and materials that can deliver the necessary properties at an economic price. Moving on to the later chapters, we see that the role of human physiology during the eating of food also needs to be considered. For many researchers in the food area, it is this diversity that makes the field so interesting and, at times, so difficult. The diversity of sciences involved also reinforces the widely held view (see summary in Section 9.5) that significant advances in flavour will only be made through multi-disciplinary teams where all members have some awareness of other scientific disciplines.

Modifying flavour: an introduction

1.2

9

References

appelquist, i. a. m., brown, c. r. t., homan, j. e., jones, m. g., malone, m. e., norton, i. t., appleqvist, i. a. and brown, c. r. (2000), ‘Low-fat food emulsions for e.g. spreads, dressing or mayonnaise, have gel particles for delaying the release of the flavor molecules’, Patents WO200007462-A, EP1102548-A, WO200007462-A1, EP1102548-A1, EP1102548-B1. channell, g. a. and taylor, a. j. (2005), ‘On line monitoring of the Maillard reaction using a film reactor coupled to ion trap mass spectrometry’. Process and reaction flavors. Weerasinghe, D. K. and Sucan, M. K. Washington, DC, American Chemical Society. 905, 181–191. dalton, p., doolittle, n., nagata, h. and breslin, p. a. s. (2000), ‘The merging of the senses: Integration of subthreshold taste and smell.’ Nature Neuroscience, 3 (5), 431–432. davidson, j. m., hollowood, t. a., linforth, r. s. t. and taylor, a. j. (1999), ‘The effect of sucrose on the perceived flavour intensity of chewing gum.’ Journal of Agricultural and Food Chemistry, 47, 4336–4340. fisher, c. and scott, t. r. (1997), Food Flavours: Biology and Chemistry. Cambridge, Royal Society of Chemistry. hollowood, t. a., linforth, r. s. t. and taylor, a. j. (2002), ‘The effect of viscosity on the perception of flavour.’ Chemical Senses, 27, 583–591. hort, j. and hollowood, t. a. (2004), ‘Controlled continuous flow delivery system for investigating taste–aroma interactions.’ Journal of Agricultural and Food Chemistry, 52, 4834–4843. linforth, r. s. t. and taylor, a. j. (2003), ‘Direct mass spectrometry of complex volatile and non-volatile flavour mixtures.’ International Journal of Mass Spectrometry, 223–224, 179–191. martins, s. i. f. s. and van boekel, m. a. j. s. (2005), ‘Kinetics of the glucose/glycine Maillard reaction pathways: influences of pH and reactant initial concentrations.’ Food Chemistry, 92 (3), 437–448. mhra (2004), Medicines and Healthcare products Regulatory Agency, http://www. mhra.gov.uk/home. pangborn, r. m. and szczesniak, a. s. (1974), ‘Effect of hydrocolloids and viscosity on aromatic flavour compounds.’ Journal of Texture Studies, 4, 467–482. pfeiffer, j., hort, j., hollowood, t. a. and taylor, a. j. (2005), ‘Temporal synchrony and integration of sub-threshold taste and smell signals.’ Chemical Senses, 30, 1–7. pollien, p., lindinger, c., yeretzian, c. and blank, i. (2003), ‘Proton transfer reaction mass spectrometry, a tool for on-line monitoring of acrylamide formation in the headspace of Maillard reaction systems and processed food.’ Analytical Chemistry, 75 (20), 5488–5494. schieberle, p. (2005), ‘The carbon module labeling (camola) technique: a useful tool for identifying transient intermediates in the formation of Maillard-type target molecules.’ Annals of the New York Academy of Sciences, 1043, 236–248. taylor, a. j. (1996), ‘Volatile flavor release from foods during eating.’ Critical Reviews in Food Science and Nutrition, 36 (8), 765–784. verhagen, j. v. and engelen, l. (2006), ‘The neurocognitive bases of human multimodal food perception: sensory integration.’ Neuroscience and Biobehavioral Reviews, 30 (5), 613–650.

2 Flavouring substances: from chemistry and carriers to legislation D. Baines, Baines Food Consultancy Ltd, UK

2.1

The importance of olfaction in the appreciation of flavour

The appreciation of flavour is realised through our chemical senses of olfaction and gustation and, to a lesser extent, through certain molecules that interact with trigeminal nerves located in the mouth, throat and nasal cavity. The trigeminal effect, or chemesthesis, is a minor but integral part of the flavour sensation involving the pain receptors that sense, for example, the heat from chillies, the cooling from peppermint and the burning from horseradish.

2.1.1 Gustation Gustation, or the sense of taste, is experienced through five types of interaction in the mouth; sweetness, sourness, bitterness, saltiness and ‘umami’, the latter being the Japanese word for succulence or deliciousness and elicited by the salts of glutamic acid, ribonucleotides and a few other taste active chemicals (Bell and Watson, 1999; Dewis, 2005). Taste responds primarily to non-volatile, water-soluble or saliva-soluble flavouring substances and serves a number of purposes. It monitors the quality of food through the sweet and umami taste receptors which respond to the presence of sugars and glutamate respectively. Detecting sugars determines carbohydrate quality and the supply of energy, and the detection of glutamate determines protein quality and the supply of essential amino acids for healthy metabolic functioning. The bitter and sour receptors serve a protective role and defend the body from noxious and poisonous substances and the regulation

Flavouring substances: from chemistry and carriers to legislation

11

of sodium chloride in the bloodstream is facilitated by the salty taste receptor.

2.1.2 Olfaction Our real ability to appreciate flavour, and to discriminate and understand the subtleties and nuances of foods and drinks, lies mainly with olfaction and is due to a quite amazing organ, the olfactory epithelium, which detects volatile organic chemicals in the environment and flavouring substances in food. This organ, located high in the nasal cavity, is a forward projection of the brain that interfaces with the outside world. It would have been one of the first organs to develop, even before the brain, and its original function would have been to detect chemical nutrients in an aqueous environment necessary for the survival of the organism. The olfactory epithelium has therefore had a very long time to evolve (Stoddart, 1999) and its ability to detect volatile organic chemicals is highly developed and tuned into the flavouring substances that are important in the foods we eat. It is estimated that the olfactory epithelium can detect and discriminate around 10 000 different smells through around 6 million receptor cells which are replaced by the body every 4–10 weeks (Mackay-Sim and Kittel, 1991; Key, 1999). These olfactory receptor cells each possess several hair-like structures called cilia which protrude into a mucous layer covering the olfactory epithelium. The cilia act as ‘docking ports’ for the volatile flavour substances that evaporate from the tongue and throat into the nasal cavity and absorb into the olfactory mucus during the process of eating and drinking. The flavour chemicals absorbing in the mucous layer couple with specialised olfactory binding proteins located in the ciliary membrane (Buck and Axel, 1991). A signal then passes to the olfactory bulb where it is amplified and the ‘signal to noise’ ratio minimised before being relayed to the olfactory cortex in the limbic brain where it is interpreted. The limbic system is the most ancient part of the brain and is also involved with emotion, mood, sexual behaviour, reproductive control and fear. Smells play a number of important roles through the limbic system such as neonatal bonding, flight response from prey, the selection of a mate and the monitoring of food. Smells can evoke memory and emotions, trigger fear, influence mood and are important for our enjoyment when eating. The ability of the olfactory epithelium to detect and interact with flavouring substances varies by many orders of magnitude and compared with gustation it is highly developed. If we compare the sensitivities of aroma and taste, our ability to detect sweetness through sugar is of the order of 3400 parts per million, the threshold level at which ordinary people start to detect sugar in water. If we relate this to time, which is easier to conceptualise, it is equivalent to approximately 1 second in an hour. By comparison, our ability to detect volatile flavouring substances is many orders of magnitude more sensitive and the flavouring substance with the lowest recorded

12

Modifying flavour in food

threshold, maple furanone, has a threshold in water of 0.00001 parts per billion, which equates to a time scale of 1 second in 3.2 million years. Humans have an extraordinary ability to perceive and discriminate between thousands of volatile organic compounds and an astonishing sensitivity to certain molecules that in evolutionary terms have been an important aspect in the survival of the species. Such is the amazing chemistry, biochemistry and neurobiology of the olfactory epithelium; nanotechnology in action initiated by flavouring substances in food and other volatile organic chemicals in the environment. Our flavour senses are more complex than the simplistic overview described above and recent research has revealed that flavour perception is multisensory in nature, where one sensory input can modify the perception of another, showing that olfaction, gustation and chemesthesis are strongly interactive. A number of excellent texts are available describing olfaction, gustation and the multimodal relationship that exists between the senses in more detail, and the reader is directed to these for further information (Bell and Watson 1999; Taylor and Roberts 2004).

2.2

Flavouring substances in foods

Enormous progress has been made since the 1960s with the identification of volatile organic compounds in foods. In 1965 approximately 700 had been found but today the number of discrete volatile organic compounds detected in foods stands at over 7670; these are listed in the Nutrition and Food Research Institute of the Netherlands (TNO) publication website Volatile Compounds in Foods (VCF) (http://www.vcf-online.nl). The VCF register is the recognised world authority on the occurrence of volatile compounds in food and the latest version, 9.1, lists over 680 food sources in which flavouring substances have been identified. The compounds listed in the VCF have only been included if the validity of the identification has been authenticated by two methods of analysis, usually a retention time and a mass spectrum, and more than 105 700 individual occurrences are now registered in the database. Over 1000 discrete volatiles have been found in coffee and nearly the same number in beef. Many organic volatiles are ubiquitous across foods, while others are specific to a particular food group or even an individual food. It is this that makes the subject so fascinating and challenging for flavour chemists to identify which of the volatile organic compounds in foods are actually creating the flavour signals in the human brain that allow us to recognise the flavour of the food we eat. Not all volatile compounds found in food can be classified as flavouring substances because some do not possess distinctive aromas, while others have very high odour thresholds and are not detected among other lower threshold flavouring substances present in foods. Others may have undesirable aroma profiles and some volatile compounds found in food cannot be

Flavouring substances: from chemistry and carriers to legislation CHO

13

O OH

OR OH (I): R = Me (II): R = Et

Fig. 2.1

O

R

(III): R = Me (IV): R = Et

Vanillin and maltol and their artificial homologues.

used as flavouring substances because they are toxic. For example, benzene, which is a carcinogen, is listed in 90 food products in the TNO register, including major foods such as apples, oranges, beef, chicken, pork and cheese. Conversely, not all flavouring substances are found in food. These are the artificial flavouring substances of which there are now relatively few but some are very useful such as ethyl vanillin (II) which has a sweet, warm, creamy, vanilla-like flavour (Arctander, 2000) and is four times the flavour strength of vanillin (I) and ethyl maltol (IV) which has a sweet, fruity, jamlike, slightly bready flavour (Arctander, 2000) and is four to six times the tenacity of its methyl analogue, maltol (III), which is widely distributed in nature (Fig. 2.1). By definition, therefore, flavouring substances are volatile organic compounds, not necessarily present in foods that are useful in the formulation of food flavourings.

2.2.1 Flavouring substances – a historical perspective The first flavouring substance of commercial use was vanillin (I), which was extracted from vanilla beans (Gobey, 1858) and synthesised by Tiemann and Haarmann (1874). The structure was subsequently proved by Riemer (1876) who synthesised vanillin from guaiacol. This heralded the beginning of modern flavour chemistry and the discovery of new and interesting flavouring substances through organic synthesis. It also gave birth to the world’s first synthetic flavour and fragrance factory, Haarmann & Reimer, which was founded in 1874 and in which perfumery and flavouring substances were produced. In 2003, Haarmann & Reimer was merged with Dragoco to form the world’s fourth largest flavour company, Symrise.

2.2.2 The synthetic era Quite a number of flavouring substances were discovered first by synthesis and then later found in the food product which represented their aroma character. Raspberry ketone (V) is a notable example, which was first synthesised by Japanese workers (Nomura and Nozawa, 1919). It has a sweet,

14

Modifying flavour in food

fruity and warm odour resembling raspberry preserve (Arctander, 2000) and was used extensively as a raspberry flavouring substance well before its discovery in the berry (Schinz and Seidel, 1957). Raspberry flavouring is an example of the forces driving flavour companies in the early decades of the 20th century, competing to secure market share and satisfy the needs of food companies to provide ever more appealing food products for consumers. The synthesis and patenting of a novel flavour chemical gave a flavour company the exclusive rights to that chemical during the lifetime of the patent and the sale of improved flavourings containing the novel compound gave the company a significant commercial benefit over its competitors. The major flavour houses were investigating new synthetic routes to flavour compounds that could help them create improved raspberry flavours and a large number of closely related derivatives of raspberry ketone (V) were synthesised and their olfactory properties evaluated (Winter, 1961) in the laboratories of Firmenich. Winter concluded that the naturally occurring ketone, raspberry ketone, had the strongest and finest organoleptic properties although the methoxy derivative was also reported to be useful as a raspberry flavouring substance. Considerable interest was shown in the ability of synthetic derivatives of the ionones to impart fresh raspberry notes to various flavour formulations. Both α-ionone (VI) and β-ionone (VII) had been found in raspberry and although they make an important contribution to the flavour they do not give the character of raspberry required to create authentic raspberry flavours. The flavour companies Givaudan and Firmenich filed patents (Helmlinger et al., 1972; Ohloff and Sundt, 1973) on compounds derived from the ionone nucleus indicated by the general formula (VIII) and a subsequent Givaudan patent (Helmlinger et al., 1975) named 4-(1,1,3trimethyl-3-cyclohexen-2-yl)-4-acetylmercapto-butan-2-one (IX) as useful for generating raspberry aromas in foodstuffs. International Flavours and Fragrances (IFF) investigated methods of modifying the ionone structure and patented the compounds 4-(1,1,3-trimethyl-2,4-cyclohexadien-2-yl)-2butanol (X) and 4-(1,1-dimethyl-3-methylene-4-cyclohexen-2-yl)-2-butanol (XI) (Mookherjee et al., 1975) and their corresponding acetates and βcylohomocitral (XII), a fragment of the ionone structure, which they claimed ‘rounded out’ and contributed to a very natural fresh aroma and taste as found in full ripe raspberries (Pittet et al., 1976). Another compound, 2-(4hydroxy-4-methylpentyl)norbornadiene (XIII), was patented for use in raspberry flavours by IFF (Sanders and Vock, 1976) and was reported to give the full aroma of ripened raspberries with the taste of ripe raspberry and its seedy kernel note. A compound with this structure is unlikely to be found in nature but it is interesting to note the structural similarities of both artificial and naturally occurring flavouring substances exhibiting raspberry character, a factor that would not have gone unnoticed to synthetic chemists attempting to create the flavour of raspberry. Subsequently damascenone

Flavouring substances: from chemistry and carriers to legislation O

15

O

HO (V)

(VI) SR

O

O

R' R = H or Ac R′ = H or Me (VII) SAc

(VIII) O

OH

(X)

(IX) OH

CHO

(XI)

(XII) OH

(XIII)

Fig. 2.2

O

(XIV)

Flavouring substances used in raspberry flavourings.

(XIV) was found in raspberries and this contributes a natural sweet fruity rose plum raspberry sugar type of character and makes a valuable contribution to raspberry flavourings (Fig. 2.2). Strawberry flavour also presented another major challenge for the flavour industry being one of the most popular consumer flavours, especially in ice cream and, for many decades, two synthetic compounds, ethyl-3-phenylglycidate (EPG) (XV) and ethyl-3-methyl-3-phenyl glycidate (EMPG) (XVI) also known as ‘strawberry aldehyde’, formed the basis of strawberry flavourings (Fig. 2.3). EMG has a warm caramellic-fruity odour reminiscent of cooked strawberries with a strawberry jam taste and EMPG

16

Modifying flavour in food COOC2H5

O

HO

O

O

R

R (XV): R = H (XVI): R = Me

Fig. 2.3

(XVII): R = Me (XVIII): R = H

Flavouring substances used in strawberry flavouring.

is described as having a sweet, fruity, candy-like warm odour with a similarity to strawberry juice and a sweet distinctly fruity taste with good resemblance to strawberry and other berry fruits (Arctander, 2000). The discovery of 4-hydroxy-2,5-dimethyl-3-(2H)furanone, known as furaneol (XVII), in strawberries (Willhalm et al., 1965) in the laboratories of Firmenich with its fruity, strawberry, pineapple character gradually replaced the use of the glycidates which have never been identified in nature. Furaneol (XVII) turned out to be one of the most versatile flavouring substances ever discovered and has a distinct flavour-enhancing effect in addition to contributing its own flavour notes. Shortly after its discovery in strawberry, it was found in pineapple and it is now known to play a key role in raspberry, guava, mango, passionfruit, tomato, arctic bramble and others. Not only does it contribute an important sweet fruity flavour character to many fruits and other food products, it also has a very important role in cooked products such as coffee, malt and roasted almonds. It has been identified as one of the flavouring substances of importance in beef (Tonsbeek et al., 1968) and other meat products along with its homologue 4-hydroxy-5-methyl-(2H)furan-3-one, known as norfuraneol (XVIII) and for two decades furaneol was protected by patents from Firmenich for its use in fruit flavours and from Unilever for its use in meat and savoury flavours. The latter was filed because furaneol furnishes an important flavour character to beef and chicken but also because both furaneol and norfuraneol were also found to be the precursors of a number of highly potent sulphur compounds that are formed during the cooking of beef and chicken and that occupy a pivotal role in meat flavour. Furaneol is truly the chameleon of flavouring substances.

2.2.3 The analytical era The synthetic era gradually gave way to the analytical era, as analytical techniques developed and became increasingly capable of detecting and identifying the flavouring substances in foods that actually cause the flavour. The fact that these flavouring substances could be called nature identical in Europe gave them a considerable commercial advantage over synthetic artificial compounds as consumer trends shifted towards a demand for

Flavouring substances: from chemistry and carriers to legislation

17

natural additives and ingredients in food products. The discovery of βdamascenone (XIV) in rose by workers in Firmenich in 1970, then its subsequent discovery in raspberry, grape and wine, and its use in combination with α- and β-ionone, raspberry ketone and other nature identical flavouring substances, gradually obviated the need for artificial synthetic chemicals in the creation of raspberry flavourings. The analytical era gathered momentum in the mid-1970s as analytical methods became more sophisticated and nature could be ‘peeled open’ to uncover new, naturally occurring flavouring substances. This became an important strategic goal for the researchorientated flavour companies that had invested heavily in the new analytical methods. The discovery of a new flavouring substance clearly had significant commercial benefits for flavour companies but it was the consumer who was the eventual beneficiary through improvements in flavours which contributed significantly to the increasing quality of foodstuffs. The advent of gas chromatography coupled to mass spectrometry and odour port assessment of gas chromatograms was a leap forward (Karasek and Clement, 1988) but the real advance was the link-up of the computer to the mass spectrometer which made an enormous difference to the quality and speed of information available to the flavour chemist. From analysing UV-sensitive chart paper impregnated by the ion pattern from the source of the mass spectrometer, followed by reconstruction of chemical structures from fragmentation patterns, the flavour chemist found that, with the aid of the computer, the information was readily available and much easier to interpret. As mass spectrometry libraries developed, known compounds could be instantly identified and valuable data associated with unknown flavouring substances became more accessible. Important new data previously buried in the enormous amount of information generated by the mass spectrometer could be collected and analysed. Specific detectors for elements such as sulphur, the advent of chemical ionisation mass spectrometry, quadrupole mass spectrometry and more recently tandem mass spectrometry and time-of-flight mass spectrometry have made further inroads into the identification of volatile organic chemicals in food. Likewise extraction techniques became increasingly capable of removing flavour chemicals from the matrix of a food product without damaging or rearranging the critical flavouring substances. Relatively crude solvent extraction techniques gave way to headspace analysis as analytical methods became more sensitive. In the 1970s very large quantities of a food product had to be extracted to concentrate enough of the flavouring substances to allow their detection by mass spectrometry. More sophisticated solvent extraction methods such as solvent-assisted flavour evaporation (SAFE) (Engel et al., 1999) have been developed to protect flavouring substances during the extraction and concentration process but probably the greatest advance came from the water industry in Canada with the development of the headspace technique, solid phase micro-extraction (SPME) (Berlardi

18

Modifying flavour in food

and Pawliszyn, 1989), which has had widespread application in flavour research. As more and more flavouring substances were found in food products, it became a challenge to identify which flavouring substances were making the greatest contribution to the characteristic flavour of an individual food. Two elegant methods were developed that enabled research chemists to single out the key flavouring substances; CHARM analysis (Acree, 1993) and aroma extraction dilution analysis (Ullrich and Grosch, 1987). Both methods evaluate the contribution that each individual flavouring substance is making by sniffing the effluent of the gas chromatograph in a series of dilutions of the original flavour extract. Several injections are needed to reach a dilution of the flavour extract where the odour detection threshold of most flavouring substances present has been exceeded and only those making the greatest contribution to the flavour remain.

2.2.4 The importance of model reactions As extraction and analytical techniques developed, many new compounds were identified in foods such as fruits, vegetable products, herbs and spices where the flavouring substances contributing to flavour are formed by biogenetic processes occurring in the plant. Conversely, the flavouring substances generated in foods during cooking processes and formed by the Maillard reaction presented a new challenge to analytical chemists. This was due to the incredibly low flavour thresholds exhibited by some of the important character impact compounds in foods such as meat and bread and beverages such as coffee – the techniques available then were not sufficiently sensitive to detect then. It was also because many of the flavouring substances involved contained sulphur and early methods of extraction and analysis were notorious for degrading these compounds. The discovery of compounds of importance in cooked foods was facilitated through the study of model reaction systems where flavour precursors such as reducing sugars and amino acids were reacted to form concentrated mixtures containing higher levels of the flavouring substances than in the cooked foods. The compounds now recognised as being of importance to meat flavour (Fig. 2.4) were reported in a model system (Evers et al., 1976) in which thiamine hydrochloride, cysteine hydrochloride and a hydrolysed vegetable protein were reacted together. Among the compounds formed was a thiol, 2-methyl-3-furanthiol (XIX) and its related disulphide (XX), which had been patented by the author prior to the publication (Evers, 1970). The compounds were subsequently found in other model systems such as the reaction between 4-hydroxy-5-methyl-3(2H)-furanone and hydrogen sulphide (van den Ouweland and Peer, 1975), the reaction of cysteine and ribose (Farmer and Mottram, 1990), the heat degradation of thiamine (van der Linde et al., 1979) and in a model system containing cysteine, thiamine, glutamate and ascorbic acid (Werkhoff et al., 1990). It was

Flavouring substances: from chemistry and carriers to legislation SH

O (XIX)

S

SMe

S

O

O (XX)

19

O (XXI)

Fig. 2.4 Character impact compounds of meat flavour.

not until 1986 that 2-methyl-3-(methylthio)furan (XXI) was found in cooked beef (MacLeod and Ames, 1986) and shortly after that when 2-methyl-3furan thiol (XIX) was identified along with its corresponding disulphide, bis-(2-methyl-3-furyl) disulphide (XX), as the major contributors to the aroma of cooked beef, chicken and pork (Gasser and Grosch, 1988). This latter compound is reported to possess a rich aged-beef, prime rib aroma (Rowe, 2000) and this, and the other furan thiols, play an important role as the character impact compounds of meat flavour. Since then, many related sulphur-substituted furans and thiophenes have been discovered in the volatiles of cooked meat (Mottram, 1998). Without the knowledge gained through the study of model systems, it would have taken considerably longer to identify the character impact compounds of many cooked and baked foods.

2.2.5 High-impact flavouring substances – threshold values The compound bis-(2-methyl-3-furyl) disulphide found in beef has one of the lowest threshold levels ever detected. It has been measured at 0.00002 ppb in water or, translating this to a time scale, is approximately equivalent to 1 second in 1.6 million years. Grapefruit mercaptan discovered in the fruit in 1982 also has an incredibly low threshold. Its two enantiomers, (S)-(−)-1-p-menthen-8-thiol (XXII) and (R)-(+)-1-p-menthen8-thiol (XXIII), have thresholds of 0.00008 ppb and 0.00002 ppb respectively (Lehmann et al., 1995) and the (S)-(−) enantiomer possesses a more fruity and less sulphury flavour profile, demonstrating how the human olfactory epithelium can distinguish between chiral compounds even at ppb levels (Fig. 2.5). One can postulate that this interaction between a flavouring substance in a fruit and the human olfactory epithelium may have had a purpose – that natural selection has resulted in a plant in which a flavouring substance, 1-p-menthen-8-thiol, had been selected because it is attractive to fruit-eating primates. Is it at all possible, as it is with other species, that a symbiotic relationship existed between early hominids and the plants they consumed; the plant deriving the benefit of successful propagation and the primate a source of nutrients and an enjoyable and an attractive flavour? Have some of the flavour chemicals that we most enjoy been designed by plants for us and is this one of the reasons why we smell what we smell?

20

Modifying flavour in food

SH

SH

(XXII)

Fig. 2.5

(XXIII)

Enantiomers of 1-p-menthen-8-thiol.

OH

O (XXIV)

O

Fig. 2.6 Maple furanone.

That would infer that some flavouring substances, possibly low threshold compounds and character impact compounds, have had a special role to play in human evolution. The lowest odour threshold ever recorded is associated with the flavouring substance 5-ethyl-3-hydroxy-4-methyl-2(5H)-furanone (XXIV), also called ‘maple furanone’ (Fig. 2.6), which has a maple and caramel aroma with a sweet, burnt, caramel taste and a threshold in water of 0.00001 ppb. This compound was first identified in soy sauce where it makes an important flavour contribution and interestingly has recently been found in raspberry (Klesk et al., 2004).

2.2.6 The discovery of new flavouring substances There are many flavouring substances in food products that have not yet been discovered. Flavour scientists have concentrated their attention on identifying the flavour chemicals in the most commercially important food products such as meat, coffee, fruits, etc., and many foods have not been subjected to the rigors of analysis. Cheese, for example, has been extensively studied but roasted cheese used as a topping on products such as pizza and lasagne or cheese used in sauces has received very little attention. Another example is duck, which is consumed worldwide and is a major source of protein in China but is not one of the 680 food sources listed in the VCF and hence no flavour chemicals for this important meat have been recorded. Food products such as meat, fish, vegetables and spices, are cooked as blends to create dishes such as stews, soups, burgers and casseroles, and the

Flavouring substances: from chemistry and carriers to legislation

N

N

(XXV)

21

N

(XXVI)

(XXVII)

N (XXVIII)

Fig. 2.7

New flavouring substances discovered in roasted spotted shrimp.

many flavour precursors generated by each raw material can interact to produce new flavouring substances that create some of the unique and characteristic flavours of many dishes. A recent study of the flavour volatiles present in a cooked mixture of beef and onion, a relatively common food combination, is a case in point (Dewis and Kendrick, 2002). Two compounds, furfurylpropyldisulphide and 3-(allylthio)propanal, were identified in the cooked mixture that could not be found in either cooked beef or sautéed onion. Another example of a recent discovery of some interesting new flavouring substances is from a study of the volatile compounds of roasted spotted shrimp by workers in Japan (Tachihara et al., 2005). A number of related pyrrolidine compounds were found (Fig. 2.7, XXV–XXVII), which possessed both roasted seafood and seafood flavour character. When the two stereoisomers of N-(2-methylbutyl)pyrrolidine (XXVI) were synthesised and evaluated, they demonstrated a chiral flavour effect where one had a roasted seafood aroma and the other a seafood aroma. The same study also found a range of new butylidine compounds such as 2-methyl-N-(2methylbutylidine)butan-1-amine (XXVIII) possessing dried seafood flavour characters. A recent paper from Firmenich reported the discovery of two new monoterpenes in honey (Fig. 2.8): 4-isopropenylcyclohexa-1,3-diene-1-carboxylic acid (XXIX) and 4-(hydroxyl-1-methylethyl)cyclohexa-1,3-diene-1-carboxylic acid (XXX) (Naef et al., 2005). Roasted shrimp and honey have received relatively little attention from flavour chemists and these discoveries demonstrate that there are many interesting and potentially commercial flavouring substances yet to be found. It is also quite extraordinary that, after over a hundred years of detailed and extensive research into the monoterpenes, that two new compounds from these genera have been recently discovered.

22

Modifying flavour in food

Fig. 2.8

COOH

COOH

(XXIX)

OH (XXX)

New flavouring substances discovered in linden honey.

2.2.7 The classification of flavouring substances Flavouring substances can be classified through their functional groups such as acids, esters, alcohols and aldehydes, but this is relatively meaningless for the flavourist. Attempts have been made to group flavouring substances by their organoleptic properties and a number of flavour companies have developed quite complex systems. Recently, Oxford Chemicals has developed a flavour wheel (Rowe, 2002) for high-impact flavouring substances (Fig. 2.9). The criteria set for qualification of a flavouring substance for inclusion on the wheel are: • Low odour threshold: <10 ppb. • Character impact: the flavouring substance has a recognisable character. The character also needs to be a desirable one at the level of use – many chemicals have powerful aromas but would not be used as flavouring substances. • Available to the flavourist: the flavouring substance is permitted, is stable and is economically viable. The compounds on the wheel are potent flavouring substances used by the industry in compounding commercial flavourings. Examples of high-impact flavouring substances for each organoleptic sector of the wheel are depicted as their chemical formulae in Fig. 2.9. Clockwise around the wheel from ‘mushroom’ to ‘vegetable’, the materials are formed by biogenesis in plants and the remaining nine sectors, from ‘nutty’ to ‘rancid and cheesy’, are formed in the Maillard reaction during the cooking of foods or in fermentation processes. Combinations of different sectors can be used to create a flavour. For example, coffee flavour can be formulated by combinations of burnt/roasted, smoky, caramel, nutty, blackcurrant/cassis, earthy/mushroom and beefy.

2.2.8 Flavour delivery systems Liquid blends of flavouring substances are dispersed in solvent carriers but the number available is quite limited and includes ethanol, propylene glycol,

Flavouring substances: from chemistry and carriers to legislation Green, grassy

Onion N

Garlic

Herbaceous

S

Spicy

S

S

Truffle

S

S

S

O

S

Mushroom, earthy

CH3

HO

S

S S

23

O

O

CH3

Fruity, ester-like

O

SH

Rancid, cheesy

O S

S

CH3 H

Tropical

SH O

Fatty

O

O

SH

Blackcurrant, cassis

CH3 S

Savoury, bouillon

O

CH3 S

OH N

S

Pork, OH lamb, chicken

O

SH S

O

Beefy

O

S

Burnt, roasted

Fig. 2.9

Vegetable

N

OH OCH3

Nutty

O

Caramel Smoky

A flavour wheel for high impact chemicals (reproduced by kind permission of Oxford Chemicals, Hartlepool, UK).

glycerol, isopropyl alcohol (IPA), diacetin, triacetin, triethyl citrate and medium chain triglycerides (MCT). Internationally, the number is reduced for technical and cultural reasons; ethanol is not permitted in Muslim countries, glycerol is a poor solvent, IPA is banned in the EU and the United States and diacetin and triacetin are easily hydrolysed. This leaves propylene glycol and MCT as the universally accepted flavouring solvents. Powder flavours are mainly produced by spray drying and the carriers in greatest use are the maltodextrins and gum arabic or combinations of both. In combination with a maltodextrin, gum arabic performs the function of an encapsulating agent and the flavour is locked into the dried particle until it is added to water, when flavour release is instant. The flavour industry has invested considerably in the development of encapsulating agents, which not only lock-in and protect the flavour but can

24

Modifying flavour in food

also be used to control the delivery of the flavour into a food matrix. The main encapsulating technologies used by the industry are: •

•

•

• •

Spray cooling: a variation of spray drying in which ambient or chilled air is used to solidify a molten drop (usually a fat) containing a flavour. Coacervation: encapsulation in a protein, usually gelatin, by precipitating the protein in the presence of a flavour at the isoelectric point of the protein. Melt extrusion: converts a liquid flavouring into a durable, free-flowing amorphous glassy particulate through extrusion of the flavouring with a carrier – usually a sugar. Molecular encapsulation: formation of inclusion complexes between a flavouring substance and β-cyclodextrin. Yeast encapsulation: the technology involves trapping a flavouring inside the walls of a yeast cell.

For more information on encapsulation and delivery systems, the reader is referred to Baines and Knights (2005).

2.2.9 Creating flavours The creative skills of the flavourist take many years to hone and the creation of new flavours is one that is surrounded in mystery. ‘Putting nature back together’ is a skill that is more complex now than it was 50 years ago because of the increased number of raw materials available. Flavourists have to familiarise themselves with an enormous number of flavouring substances, understand their flavour profiles at different concentrations and their chemical and sensory interactions with other flavour substances, flavour carriers and other ingredients used in flavourings. It not only requires experience, judgement, flair and intuition but an extra spark of creativity to produce commercially successful flavours. Very little has been written about the training of flavourists but the reader is directed to Wright (2004), a book that gives a comprehensive coverage of this subject. As well as creating flavours, the flavourist is involved with matching flavours. The analytical technology used to discover the flavouring substances in foods can also be used to analyse the flavouring substances in a competitor’s flavours and the mass spectrometers of the flavour industry are being used more and more for this purpose. The cost of the technology is now within the grasp of the smallest of flavour companies; matching has become standard practice in the industry and is eroding flavour margins and profitability. The industry has moved from the heady days of high-quality, almost ivory tower, flavour research to an industry driven by matching and this is reducing the potential for flavour companies to indulge in the type of flavour research that was so productive in the latter part of the twentieth century.

Flavouring substances: from chemistry and carriers to legislation

2.3

25

Flavouring substances legislation

2.3.1 European legislation Flavour substances have been regulated in the European Union since the late 1980s (EC, 1988) by a negative list system in which restrictions have been placed upon certain flavouring substances that have been found to be deleterious to health. The list is small, containing 12 substances, and the industry has been free to use all other flavouring substances in the food supply. In addition to flavouring substances, the Directive also covers flavouring preparations, process flavourings, smoke flavourings and mixtures of these flavour groups. Flavours Directive 88/388/EEC (EC, 1988) defines a flavouring substance as ‘a defined chemical substance with flavouring properties’, which is obtained: (i)

by appropriate physical processes (including distillation and solvent extraction) or enzymatic or microbiological processes from material of vegetable or animal origin either in the raw state or after processing for human consumption by traditional food-preparation processes (including drying, torrefaction and fermentation). (ii) by chemical synthesis or isolated by chemical processes and which is chemically identical to a substance naturally present in material of vegetable or animal origin described in (i), (iii) by chemical synthesis but which is not chemically identical to a substance naturally present in material of vegetable or animal origin as described in (i).

The legislation has set the tone for the supply of flavourings in Europe for almost two decades and the industry has been providing natural, nature identical and artificial flavours for use in the food supply based around the definitions above. Hence flavourings containing flavouring substances exclusively from category (i) are referred to as ‘natural’. Flavourings containing one or more flavouring substances from category (ii) and no substances from category (iii) are classed as ‘nature identical’ and flavourings containing one or more flavouring substances from category (iii) are artificial. Natural flavourings in Europe have been produced from category (i) flavouring substances and from flavouring preparations or combinations of both. Flavouring preparations are produced from natural food sources, examples are: essential oils, oleoresins, solvent extracts and even concentrated fruit juices used for their flavouring properties. This legislative system has been very open and has given impetus to research with the result that flavour companies have been actively searching for new flavouring substances in foods, which they have been free to use in confidence without disclosure or the need for regulatory authorisation. This has had considerable commercial benefits for the companies investing in research and making the discoveries and, for a period of time, especially between the 1970s and 1990s, companies were competing to find and use

26

Modifying flavour in food

new nature identical compounds to establish market position. Consequently, Europe led the way in the search for new flavouring substances and this resulted in significant improvements in the quality and authenticity of flavourings being supplied to the food industry and, thus, of benefit to the consumer.

2.3.2 The Flavouring Substances Register Incorporated into Council Directive 88/388/EC is the provision to lay down a list of flavouring substances which may be used in or on foodstuffs. The process was initiated in 1996 (EC, 1996) and subsequent Commission Decisions (EC, 1999, 2000, 2002a, b, 2005, 2006a) with the objective of establishing a list of flavouring substances ‘the use of which is authorised to the exclusion of all others’. This, in effect, changes the regulation of flavouring substances in the EU from a negative list system to a positive list system. The flavour industry in Europe, through the European Flavours and Fragrances Association, was mandated to provide a list of flavouring substances in use in the EU to the European Commission and this list has been compiled into the Flavouring Substances Register and contains approximately 2650 discrete flavour chemicals. The list is currently being evaluated by the European Food Safety Authority (EFSA). Its remit is to consider the safety in use of each flavouring substance and to establish that they present no risk to the health of the consumer. To undertake this enormous task, a group-based approach has been adopted and the register has been divided into 34 groups of substances of structurally related compounds which are expected to show some metabolic and biological behaviour in common. The groups adopted are listed in Table 2.1 and their complexity underlines the difficulties in attempting to classify flavouring substances. The information considered essential for the evaluation of each flavouring substance includes its purity, chemical specification, natural occurrence in food, total amount added to foods and results of toxicological and metabolic studies. The evaluation process should be completed by the end of 2008 when the register will be published and, at some stage after that, will become a binding positive list on the industry. As the list is evaluated, substances that do not meet the safety criteria set by the EFSA are deleted from the register or restrictions are placed on their usage. To date, the substances that have been deleted on safety grounds are methyleugenol, estragole, capsaicin and pentan-2,4-dione.

2.3.3 New flavouring substances Special provisions were requested by industry for the evaluation of new ‘confidential’ flavouring substances. It had been recognised that a number of flavour companies invest considerable resources in research to identify

Flavouring substances: from chemistry and carriers to legislation

27

Table 2.1 Chemical groups for flavouring substances in the EU Group 1

Straight chain primary aliphatic alcohols/aldehydes/acids, acetals and esters with esters containing saturated alcohols and acetals containing saturated aldehydes. No aromatic or heteroaromatic moiety as a component of an ester or acetal.

Group 2

Branched-chain primary aliphatic alcohols/aldehydes/acids, acetals and esters with esters containing branched-chain alcohols and acetals containing branched-chain aldehydes. No aromatic or heteroaromatic moiety as a component of an ester or acetal.

Group 3

α,β-Unsaturated (alkene or alkyne) straight-chain and branchedchain aliphatic primary alcohols/aldehydes/acids, acetals and esters with esters containing α,β-unsaturated alcohols and acetals containing α,β-unsaturated alcohols or aldehydes. No aromatic or heteroaromatic moiety as a component of an ester or acetal.

Group 4

Non-conjugated and accumulated unsaturated straight-chain and branched-chain aliphatic primary alcohols/aldehydes/acids, acetals and esters with esters containing unsaturated alcohols and acetals containing unsaturated alcohols or aldehydes. No aromatic or heteroaromatic moiety as a component of an ester or acetal.

Group 5

Saturated and unsaturated aliphatic secondary alcohols/aldehydes/ acids, acetals and esters with esters containing secondary alcohols. No aromatic or heteroaromatic moiety as a component of an ester or acetal.

Group 6

Aliphatic, alicyclic and aromatic saturated and unsaturated tertiary alcohols and esters with esters containing tertiary alcohols. Esters may contain any acid component.

Group 7

Primary alicyclic saturated and unsaturated alcohols/aldehydes/acids, acetals/esters with esters containing alicyclic alcohols. Esters/acetals may contain aliphatic acyclic or alicyclic acids or alcohol component.

Group 8

Secondary alicyclic saturated and unsaturated alcohols/ketones/ ketals/esters with ketals containing alicylic alcohols or ketones and esters containing secondaryalicylic alcohols. Ester may contain aliphatic acyclic or alicyclic acid components.

Group 9

Primary aliphatic saturated and unsaturated alcohols/aldehydes/acids, acetals/esters with a second primary, secondary or tertiary oxygenated functional group including aliphatic lactones.

Group 10

Secondary aliphatic saturated and unsaturated alcohols/ketones/ ketals/esters with a second secondary or tertiary oxygenated functional group.

Group 11

Alicyclic and aromatic lactones.

Group 12

Maltol derivatives and ketodioxane derivatives.

Group 13

Furanones and tetrahydrofurfuryl derivatives.

Group 14

Furfuryl and furan derivatives with and without additional side-chain substituents and heteroatoms.

Group 15

Phenylethyl alcohols, phenylacetic acids, related esters, phenoxyacetic acids and related esters. (Continued)

28

Modifying flavour in food

Table 2.1

Continued

Group 16

Aliphatic and alicyclic ethers.

Group 17

Propenylhydroxybenzenes.

Group 18

Allylhydroxybenzenes.

Group 19

Capsaicin-related substances and related amides.

Group 20

Aliphatic and aromatic mono- and di-thiols and mono- and di-, triand polysulphides with or without additional oxygenated functional groups.

Group 21

Aromatic ketones, secondary alcohols and related esters.

Group 22

Aryl-substituted primary alcohol/aldehyde/acid/ester/acetal derivatives, including saturated ones.

Group 23

Benzyl alcohols/aldehydes/acids/esters/acetals. Benzyl and benzoate esters included. May also contain aliphatic acyclic or alicyclic ester or acetal component.

Group 24

Pyrazine derivatives.

Group 25

Phenol derivatives containing ring-alkyl, ring alkoxy and side-chains with an oxygenated functional group.

Group 26

Aromatic ethers including anisole derivatives.

Group 27

Anthranilate derivatives.

Group 28

Pyridine, pyrrole and quinoline derivatives.

Group 29

Thiazoles, thiophene, thiazoline and thienyl derivatives.

Group 30

Miscellaneous substances.

Group 31

Aliphatic and aromatic hydrocarbons.

Group 32

Epoxides.

Group 33

Aliphatic and aromatic amines.

Group 34

Amino acids.

new flavour chemicals in food and disclosure could significantly harm their competitive position. Coupled to this, there is the new requirement to submit a great deal of information about new flavouring substances to the Commission, including toxicology and metabolic studies, which adds further to the cost, and the discovering company does need time to recoup its investment through the sale of flavours containing the new flavouring substance. It was recognised that, without some form of protection, the future of flavour research in Europe could be seriously impaired and an arrangement was made to protect the identities of confidential flavour chemicals for a period of 5 years from the date of submission. Around 100 flavouring substances were placed on this list. However, this decision has been revoked

Flavouring substances: from chemistry and carriers to legislation

29

and it is proposed that flavour companies will now be required to divulge the identity of any new flavouring substances on submission to the register. Representations are being made to develop a new legislative system to protect intellectual property. 2.3.4 The new European Directive It was recognised by legislators in Brussels, that Council Directive 88/388/ EC (EC, 1988) needed to be substantially amended to take into account technological and scientific developments in flavourings and it will be replaced by a new European Directive (EC, 2006b). The European Commission was aware that large numbers of flavouring substances were being used by the industry without regulatory control. Consequently the Flavouring Substances Register, and their conditions of use, will be annexed to the new Directive which will be called the ‘Food Improvement Agents Directive’ and will contain legislation on flavourings, food additives, enzymes and authorisation procedures. Six categories of flavourings will be included: flavouring substances, flavouring preparations, thermal process flavourings, smoke flavourings, flavour precursors and ‘other flavourings’. The latter is a ‘catch all’ to include all flavourings not covered by the previous categories and new developments in flavours. The definition of flavouring substances will change and the nature identical and artificial categories will disappear. The proposed new wording is: • ‘Flavouring substance’ shall mean a chemically defined substance with flavouring properties. • ‘Natural flavouring substance’ shall mean a flavouring substance obtained by appropriate physical, enzymatic or microbiological processes from material of vegetable, animal or microbiological origin either in the raw state or after processing for human consumption by one or more of the traditional food preparation processes listed in Annex II.

The ‘traditional processes’ defined for the production of natural flavouring substances are outlined in Table 2.2. Three new categories for natural flavourings are proposed in the draft Directive: 1.

The term ‘natural’ may only be used in combination with a reference to a food, food category or a vegetable or animal flavouring source, if at least 90% [by w/w] of the flavouring component has been obtained from the source material referred to.

For example, a raspberry flavour containing 90% of a natural raspberry extract along with flavouring preparations and natural flavouring substances from other sources may be called ‘natural raspberry flavour’.

30

Modifying flavour in food

Table 2.2 List of traditional food preparation processes by which natural flavouring substances and natural flavouring preparations are obtained (Annox II) Chopping Cooking, baking, frying (up to 240 ºC) Cutting Drying Evaporation Fermentation Grinding Infusion Microbiological processes Peeling Pressing Roasting/grilling Steeping

Coating Cooling Distillation/rectification Emulsification Extraction, including solvent extraction Filtration Heating Maceration Mixing Percolation Refrigeration/freezing Squeezing

2. ‘Natural << food(s) or food category or source(s) >> flavouring with other natural flavourings’ may only be used if the flavouring component is partially derived from the source material referred to and can be clearly recognised.

For example, a raspberry flavour containing 60% of a natural raspberry extract along with flavouring preparations and natural flavouring substances from other sources may be called ‘natural raspberry flavour with other natural flavourings’ if it is clearly recognisable as raspberry. 3. The term ‘natural flavouring’ may only be used if the flavouring component is derived from different source materials and where a reference to the source materials would not reflect their flavour or taste.

This category applies to flavours that are well recognised but where their names do not reflect their flavouring components such as barbecue, cola and tutti frutti flavours. The labelling of flavourings in the EU will be limited to ‘flavouring’, ‘smoke flavouring’ and natural flavours as defined above. Smoke flavourings are singled out to control the levels of the carcinogen, 3,4-benzopyrene, and other polycyclic aromatic hydrocarbons produced during the pyrolysis of wood in the production of smoke flavourings.

2.3.5 Implications for the flavour industry in Europe The removal of the nature identical and artificial categories of flavouring substances has a number of implications for industry. From a risk assessment perspective, it is irrelevant whether flavouring substances are natural, nature identical or artificial but whether they are safe and this, on the face of it, is an eminently sensible approach to take. However, the consumer sees

Flavouring substances: from chemistry and carriers to legislation

31

it very differently and regards natural as desirable, familiar, safe, wholesome, etc., a myth that is continuously reinforced by the media, by food advertising, multiple retailers and pressure groups. The consumer does not understand the concept of ‘nature identical’ and, when it is explained, considers it to be artificial. Nature identical has held a pseudo-natural status and operated as a bridge between the undesirable ‘artificial’ and the desirable ‘natural’ and its removal is likely to increase the demand for natural flavourings. Nature identical flavouring substances produced by synthetic techniques will, more and more, be regarded as artificial. This will increase the demand for natural flavouring substances but, considering the current make-up of the flavouring substances register, this will be very difficult to achieve in the short term as the vast majority of flavour chemicals on the register are synthetic nature identicals. Considerable research is being expended to provide natural equivalents of nature identical flavouring substances using extraction processes, distillation techniques, enzyme technology, microbial transformations, fermentation technology, plant cell culture and other biotechnological developments. However, it will be very difficult in the short term, to reproduce, by natural processes, the many hundreds of nature identical flavouring substances that are formed in cooked foods by the Maillard reaction. It is likely therefore that there will be a considerable gap between what the consumer and the food industry want in terms of natural flavours and what the flavour industry can provide. In the short term, it is anticipated that this gap will be filled by natural flavourings produced as flavouring preparations but these are considerably weaker in flavour strength than flavourings produced by blends of flavouring substances and this will have a distinct bearing on quality and cost. There will also be implications on the costs of flavours because flavouring substances produced by processes deemed to be natural, are considerably more expensive than their synthetic counterparts and these costs will be eventually borne by the consumer. It will also be impossible, in some cases, to cost-effectively produce natural equivalents of synthetic flavouring substances. Another implication of the removal of the nature identical and artificial status is that the consumer deception of the phrase ‘contains no artificial flavours’, used by some food companies in their advertising, will no longer be able to be used to mislead the consumer into imagining that the flavour used is natural. One very contentious aspect of the new legislation is a requirement on industry to report to the Commission the annual amounts of flavouring substances added to foods in the EU and the usage level for each food category in the EU. Member States are additionally required to establish systems to monitor the consumption and use of flavourings and flavouring substances and report them to the Commission and EFSA each year. While it is appreciated that the motivation for these controls is to monitor food safety and the Commission and EFSA should have knowledge about flavour

32

Modifying flavour in food

usage in Europe, the measures they are proposing place an enormous burden on the industry and are considered neither practical nor appropriate. The flavour manufacturer will have to divulge details about his or her formulation to the customer and the food manufacturer will have to divulge usage levels to the supplier. This dissemination of proprietary commercial information is likely to have severe economic consequences and is a prime example of how over-zealous legislation can interfere with normal business activities. The major impact of the Directive, however, on the flavour industry is the move to a positive list system and the rejection of an industry request for a 5 year moratorium on the disclosure of new flavouring substances. This is likely to have a detrimental effect on research and the discovery of new flavouring substances by companies in Europe, and European industry will most probably lose its leading position in the global flavourings market.

2.3.6 Flavouring substances legislation in the United States The United States has been using a positive list system for flavouring substances since 1965 when the first FEMA GRAS list was published in response to the Food Additives Amendment (US, 1958). FEMA is the US Flavor and Extract Manufacturers Association and GRAS substances are ‘generally recognised as safe’ under conditions of intended use following evaluation by an independent panel of experts and, as such, are exempt from the definition of food additives. The FEMA expert panel, known as FEXPAN, consists of ‘experts qualified by scientific training and experience to evaluate the safety of flavouring ingredients’ and includes toxicologists, pharmacologists and biochemists. For most of the lifetime of the GRAS programme, they have concentrated on the evaluation and re-evaluation of 1900 chemically identified flavouring substances but their brief has been widened to cover natural flavouring complexes such as essential oils and oleoresins and chemical substances with non-flavour functions used in the compounding of flavourings such as antioxidants, emulsifiers and flavour modifiers. A series of GRAS reports have been produced since 1965 and the most recent, number 22 (Smith et al., 2005), added a further 185 flavouring substances bringing the FEMA GRAS list to just below 2100. Many of the additions to the FEMA GRAS list were as a direct result of European food analytical research which, with the new EU regime, will not be economically viable. Nature identical is not a term that is recognised in the United States, where all synthetic flavouring substances, whether they have been found in a natural product or not, are regarded as artificial and labelled as such. Flavours containing a mixture of natural and artificial flavouring substances are labelled as natural and artificial depending on the relative quantities and natural flavourings fall into three categories (Knights, 2002):

Flavouring substances: from chemistry and carriers to legislation

33

• FTNF: ‘From the named fruit’, which includes extracts, distillates and condensates derived exclusively from the named fruit and can include flavouring substances providing that they are also derived from the named fruit. Flavours from this category tend to be weak and expensive but are in increasing demand. • WONF: ‘With other natural flavourings’. Flavours in this category must contain 51% of the flavour from the named source but can also contain other natural materials and natural flavouring substances, • Natural Flavour: Flavours that contain all natural ingredients and/or flavouring substances but the source is not defined.

Unlike Europe, the United States is well advanced in the evaluation and approval of flavouring substances and the FEMA GRAS list provides quite comprehensive information on the use of flavouring substances in various food products. However, the regulation of flavouring substances has restricted research into the identification of new flavouring substances in the United States, which has been moving forward at pace whereas in Europe positive list regulation is a recent phenomenon. The approval of new flavouring substances onto the FEMA GRAS list has been streamlined and is efficient compared to the cumbersome system being evolved in the EU. However, it is anticipated that the positive list systems adopted in the United States and the EU will stimulate research into the identification of new flavouring substances in India and China where the results of research are not so hamstrung by bureaucracy. This will be driven by the enormous emerging markets for flavourings in South East Asia where the regulatory control of flavourings has not evolved to any extent.

2.3.7 Flavour substances legislation in Japan Japan operates a positive list system for flavouring substances under the auspices of the Food Sanitation Law No. 233 (Japan, 2004a). Flavouring substances are listed in the Specifications and Standards section (Japan, 2004b) and it is surprisingly large, containing over 2600 discrete flavour chemicals and new compounds are being frequently added. The list is subdivided into 18 structural groupings: aliphatic higher alcohols, aliphatic higher aldehydes, aliphatic higher hydrocarbons, aromatic alcohols, aromatic aldehydes, esters, ethers, fatty acids, furfural and its derivatives, indole and its derivatives, isothiocyanates, ketones, lactones, phenols, phenol ethers, terpene hydrocarbons, thioethers and thiols. The list can be quite confusing because some of the flavouring substances are listed under their chemical names and others under their trivial names, such as raspberry ketone, and a number of flavouring substances that are in international use, such as the alkyl pyrazines, were not included until quite recently because they did not fit into any of the structural groups. Limiting aliphatic alcohols and aldehydes to ‘higher’ also excluded common flavour chemicals such as acetaldehyde and isoamyl alcohol but these omissions have now been rectified as

34

Modifying flavour in food

a result of an effort by the International Organisation of the Flavour Industry (IOFI) Science Board to achieve acceptance in Japan of substances that are used internationally. Chemicals ‘generally regarded as highly toxic’ are excluded from the Japanese structural groups and there is no distinction between nature identical and artificial but there is a definition of natural (Japan, 2000b): ‘Natural flavouring agent means additives intended for use for flavouring food which are substances obtained from animals or plants or mixtures thereof’.

2.3.8 Flavour substances legislation in Australasia In 2002, Australia and New Zealand developed the Food Standards Code (FSANZ, 2002a) which is a joint set of food labelling and composition rules for the two countries to allow free trade without technical obstruction. A user guide (FSANZ, 2002b) has also been published, providing information on the regulation of flavourings and flavour enhancers. The two countries now operate a ‘mixed list’ system where both natural and nature identical flavouring substances are recognised and are controlled by a negative list system, i.e. they can be used unless specifically regulated, and artificial flavouring substances are controlled by positive list system, i.e. they can only be used if they are on the positive list. The Flavour and Fragrance Association of Australia and New Zealand has compiled a list of 381 artificial flavouring substances and this is appended to the user guidelines. The definitions of the three groups of flavouring substances used are (FSANZ, 2002a): • Natural flavouring substances means flavouring substances obtained from plant or animal raw materials by physical, microbiological or enzymatic processes. They can be either used in their natural state or processed for human consumption but cannot contain any nature-identical or artificial flavouring substances. • Nature identical flavouring substances means flavouring substances that are obtained by synthesis or isolated through chemical processes which are chemically identical to flavouring substances naturally present in products intended for human consumption. They cannot contain any artificial flavouring substances. • Artificial flavouring substances means flavouring substances not identified in a natural product intended for human consumption, whether or not the product is processed.

Food products containing flavourings are labelled with the words ‘flavouring’ or ‘flavour’ or with a specific name or description of the flavouring. The artificial flavouring substance, ethyl maltol can be labelled either as a ‘flavour enhancer’ or as a ‘flavouring’ depending on the function it is judged to be performing in the food. This regulatory system bears some resemblance to the existing European legislation and the ‘mixed list’ system which operated in Germany prior to 1988.

Flavouring substances: from chemistry and carriers to legislation

35

2.3.9 Flavour substances legislation in South America A new technical regulation (MERCOSUR, 2006) on flavouring additives was adopted in June 2006 by Argentina, Brazil, Paraguay and Uruguay. The harmonisation of the technical regulations on flavours is aimed at eliminating barriers to trade created between prevailing national regulations. Under the new regulations flavourings are defined as natural and synthetic, and flavouring substances fall into the following categories with synthetic flavouring substances being further subdivided into nature identical and artificial (MERCOSUR, 2006): • Natural flavouring substance are chemically defined substances obtained by physical, microbiological or enzymatic processes from natural aromatic raw materials or from natural flavourings. • Synthetic flavouring substances are chemically defined compounds obtained by chemical processes and include: 1. Nature identical flavouring substances; chemically defined substances obtained by synthesis and/or isolated by chemical processes from raw materials of animal, vegetable or microbial origin, which present a chemical structure identical to the substances present in the reference natural materials (either processed or not). 2. Artificial flavouring substances; chemical substances obtained by synthesis, as yet not identified in products of animal, vegetable or microbial origin, when used in their primary state or after preparation for human consumption.

This new regulation and the new regulations in Australasia recognise the category of ‘nature identical’ whereas in the northern hemisphere the US regulations, the new EU Directive and the Japanese regulations no longer recognise this term. This new regulation in South America also introduces the concept of ‘synthetic’ flavourings into legislation, a term that is not likely to be very consumer friendly. Elsewhere in the world, extensive use is made of the FEMA GRAS list as a means to control the supply of flavours, to ensure food safety and protect commercial interests. Countries that have just entered, or are seeking entry, into the EU are adopting the EU system.

2.3.10 Global harmonisation of flavouring legislation Flavour legislation is failing to keep pace with the acceleration of globalisation in the food supply. The above sections demonstrate that where flavour legislation has developed there are significant differences between the regulatory systems in the control of flavouring substances which could impede global trade. Some inroads are being made by IOFI to counter this and it is important that where positive lists exist there is coordination to ensure that they correlate with each other. Better still would be the adoption of a global system to harmonise the control of flavourings and the systems are there to achieve this through

36

Modifying flavour in food

Table 2.3 Summary of evaluations performed by the Joint FAO/WHO Expert Committee on Food Additives on 4-(p-hydroxyphenyl)-2-butanone COE No.: FEMA No.: JECFA No.: Chemical names: Synonyms: Functional class: Latest evaluation: ADI: Comments: Report: Specifications: Tox monograph:

755 2588 728 4-(4-Hydroxyphenyl)butan-2-one p-Hydroxybenzylacetone; 1-p-hydroxyphenyl-3-butanone; raspberry ketone; rastone; oxyphenylon; oxanone Flavouring agent 2000 Acceptable No safety concern at current levels of intake when used as a flavouring agent TRS 901-JECFA 55/44 Compendium addendum 8/FNP 52 Add.8/172 FAS 46-JECFA 55/165

JECFA, the Joint Food and Agriculture Organization of the United Nations/ World Health Organization Expert Committee on Food Additives, the Codex Alimentarius Commission (CAC) and the World Trade Organization (WTO). JECFA serves as the scientific advisory body to the CAC on all matters relating to food additives and Codex standards carry legal significance with all countries signed up to the WTO Agreement on the Application of Sanitary and Phytosanitary Measures (WTO/GATT, 1994). The SPS agreement encourages nations to develop regulatory controls based around the international standards developed by CAC and those that differ from the Codex standards may be challenged as trade barriers. JECFA was formed in 1956 and since then has evaluated over 1750 flavouring substances. It has made recommendations on their safe level of use by elaborating specifications for their identity and purity and by evaluating toxicological data and estimating acceptable intakes for humans. The evaluation status of each flavouring substance is reported in a format as shown for raspberry ketone in Table 2.3. Codex has not yet considered the regulation of flavourings in the draft General Standard for Food Additives but this may not be appropriate as many countries regard flavourings as exempt from food additive status. A separate Codex standard for flavourings utilising the evaluation work of JECFA, and drawing from the experience of FEXPAN and EFSA, could form the basis of a globally harmonised regulatory system.

2.4

Conclusions

Approximately one-third of all volatile organic chemicals found in food are useful as flavouring substances and these have made significant improvements to the quality of the global food supply. Most of these flavouring substances are synthetic ‘nature identical’ flavour chemicals and there will

Flavouring substances: from chemistry and carriers to legislation

37

be a growing need to develop natural equivalents through biotechnology and by isolation from natural source materials as the consumer demand for natural flavourings increases. Consumers will require more information about what is added to food, including flavourings, and the industry should be ready to supply this in a balanced and informed manner. The consumer desire for new flavour sensations and new eating experiences will continue unabated and the sales of flavourings will increase accordingly, as they have done since the flavour industry began. New opportunities will be created for flavour companies as food manufacturers expand worldwide, adopting core supplier policies and trading with flavour companies with global capabilities. To meet the challenge, flavour companies will have to be highly innovative, offering a wide portfolio of added value flavouring solutions and excellence in customer service. Legislation will need to catch up with the process of globalisation, by the adoption of an international system to harmonise the regulation of flavourings and facilitate free trade and this can be achieved through JECFA, Codex and the WTO, supported by IOFI. However, positive list legislation will inhibit research into the discovery of new flavouring substances in the West, especially the regulatory system being developed in the EU. More efficient and less costly procedures, which still take into account public safety, should be adopted for the approval of new flavouring substances. It is also envisaged that research into the discovery of new flavouring substances will find a home in South East Asia where the regulatory control of flavourings has not developed to any extent and where large emerging markets and new centres of technical excellence will stimulate research into culturally important foods. The flavouring substances in use, numbering around 2700, represent the chemicals that are important in the food we eat and some have played a key role in the survival of our species. The knowledge, understanding and use of these flavouring substances is therefore of great importance to mankind for the supply of palatable foods in a world of increasing population growth, climate change and the consequential pressures that this will place on the food supply.

2.5

Acknowledgement

The author would like to thank Jack Knights, Northampton, UK, for his valuable advice in reviewing this chapter.

2.6

References

acree, t. e. (1993), ‘Gas chromatography – olfactometry’, in Ho C T and Manley C H, Flavour Measurement, New York, Marcel Dekher, 77–94.

38

Modifying flavour in food

arctander, s. (2000), Perfume and Flavour Chemicals, Caral Stream, IL, Auured Publishing Corp. baines, d. and knights, j. (2005), ‘Applications I: Flavors’, in Rowe J D Chemistry and Technology of Flavors and Fragrances, Oxford, Blackwell, 274–304. bell, g. a. and watson, a. j. (1999), Tastes and Aromas: The Chemical Senses in Science and Industry, Sydney, UNSW. berlardi, r. and pawliszyn, j. (1989), ‘The application of chemically modified fused silica fibers in the extraction of organics from water matrix samples and their rapid transfer to capillary columns’, Water Pollution Res J Canada, 24 (1), 179–191. buck, l. and axel, r. (1991), ‘A novel multigene family may encode odorant receptors: a molecular basis for odor recognition’, Cell, 65 (1), 175–187. dewis, m. l. (2005), ‘Molecules of taste and sensation’, in Rowe J D Chemistry and Technology of Flavors and Fragrances, Oxford, Blackwell, 199–243. dewis, m. l. and kendrick, l. (2002), ‘Creation of flavors and the synthesis of raw materials inspired by nature’, in Swift K A D, Advances in Flavours and Fragrances, Cambridge, Royal Society of Chemistry, 147–160. ec (1988), Council Directive 88/388/EEC on the approximation of the laws of the Member States relating to flavourings for use in foodstuffs and to source materials for their production, Off J European Communities, L184, 61. ec (1996), Council Regulation (EC) No. 2232/96 laying down a Community procedure for flavouring substances used or intended for use in or on foodstuffs, Off J European Communities, L299, 1. ec (1999), Commission Decision 1999/217/EC of 23 February 1999 adopting a register of flavouring substances used in or on foodstuffs drawn up in application of Regulation (EC) No. 2232/96 of the European Parliament and of the Council of 28 October 1996, Off J European Communities, L84, 1. ec (2000), Commission Decision 2000/489/EC of 18 July 2000, Off J European Communities, L197, 53. ec (2002a), Commission Decision 2002/113/EC of 23 January 2002, Off J European Communities, L49, 1. ec (2002b), Commission Decision 2004/357/EC of 7 April 2004, Off J European Communities, L113, 28. ec (2005), Commission Decision 2005/389/EC of 18 May 2005, Off J European Communities, L128, 73. ec (2006a), Commission Decision 2006/252/EC of 27 March 2006, Off J European Communities, L91, 48. ec (2006b), Regulation No. 2006/0147 (COD), COM(2006)0427 on flavourings and certain food ingredients with flavouring properties for use in and on foods and amending Council Regulation (EEC) No. 1576/89, Council Regulation (EEC) No. 1601/91, Regulation (EC) No. 2232/96 and Directive 2000/13/EC. engel, w., bahr, w. and schieberle, p. (1999), ‘Solvent assisted flavour evaporation – a new and versatile technique for the careful and direct isolation of aroma compounds from complex food matrices’, Eur Food Res Tech, 209 (3/4), 237–241. evers, w. j. (1970), Sulphur-containing compounds, a process for their preparation and compositions containing them, GB1256462. evers, w. j., heinsohn, h. h., mayers, b. j. and sanderson, a. (1976), ‘Furans substituted in the three position with sulphur’, in Charalambous G and Katz I, Phenolic, Sulphur and Nitrogen Compounds in Food Flavors, Washington DC, American Chemical Society, 184–193. farmer, l. j. and mottram, d. s. (1990), ‘Recent studies on the formation of meat aroma compounds’, in Bessiere Y and Thomas A F, Flavour Science and Technology, Chichester, Wiley, 113–116.

Flavouring substances: from chemistry and carriers to legislation

39

fsanz (2002a), The Australia New Zealand Food Standards Code, Food Standards Australia New Zealand. fsanz (2002b), The Australia New Zealand Food Standards Code User Guide, Flavourings and Flavour Enhancers, Food Standards Australia New Zealand. gasser, u. and grosch, w. (1988), ‘Identification of volatile flavour compounds with high aroma values from cooked beef’, Z Lebensm Unters Forsch, 186, 489–494. gobley, m. (1858), ‘Recherches sur le principe odorant de la vanilla’, J Pharm Chim, 34 (1), 401–405. helmlinger, d., lamparsky, d., schudel, p. and wild, j. (1972), Bezeichnung: Neue Jonon- und Ironderivate, German Patent 2159924. helmlinger, d., lamparsky, d., schudel, p., sigg-grutter, t. and wild, j. (1975), Novel mercaptoderivatives of ionones and irones, US3883572. japan (2004a), Law no. 233, Chapter 1; Food, Food Additives, Apparatus, and Containers/Packages, Article 12, Food Sanitation Law in Japan (2004). japan (2004b), Specifications and Standards for Foods, Food Additives, etc, Part III, Food Additives, Section 4, Food Additives with Standards of Use, Flavorings, Food Sanitation Law in Japan, 158. karasek, f. w. and clement, r. e. (1988), Basic Gas Chromatography–Mass Spectrometry, London, Elsevier. key, b. (1999), ‘Anatomy of the peripheral chemosensory systems: how they grow and age in humans’, in Bell G A and Watson A J, Tastes and Aromas, Sydney, UNSW, 138–148. klesk, k., qian, m. and martin, r. r. (2004), ‘Aroma extract dilution analysis of cv. Meeker red raspberries from Oregon Washington’, J Agric Food Chem, 52 (16), 5155–5161. knights, j. (2002), ‘Flavour Legislation’, in Taylor, A J, Food Flavour Technology, Sheffield, CRC Press, 276–295. lehmann, d., dietrich, a., hener, u. and mosandl, a. (1995), ‘Stereoisomeric flavour compounds LXX, 1-p-menthene-8-thiol: separation and sensory evaluation of the enantiomers by entioselective gas chromatography/olfactometry’, Phytochem Anal, 6, 255–257. mackay-sim, a. and kittel, p. w. (1991), ‘On the life-span of olfactory receptor neurons’, Eur J Neurosci, 3 (3), 209–215. macleod, g. and ames, j. m. (1986), ‘2-methyl-3-(methylthio)furan: a meaty character impact aroma compound identified from cooked beef’, Chem & Ind, 3 March, 175–177. mercosur (2006), MERCOSUR/GMC/RES No. 10/06, Mercado Común del Sur Technical Regulation Concerning Flavouring Additives. mookherjee, b. d., vock, m. h., benaim, c. and shuster, e. j. (1975), Altering raspberry flavored foodstuffs with 4-(2,6,6-trimethyl-1,3-cyclohexadien-1-yl)-2-butanol and/or 4-(6,6-dimethyl-2-methylene-3-cyclohexen-1-yl)-2-butanol, and/or acetates thereof, US3899597. mottram, d. s. (1998), ‘The chemistry of meat flavour’, in Shahidi F, Flavor of Meat, Meat Products and Seafoods, 2nd Edition, London, Blackie Academic & Professional, 5–26. naef, r., jaquier, a., velluz, a. and bachofen, b. (2005), ‘From the linden flower to the linden honey – volatile constituents of linden nectar, the extract of bee stomach and ripe honey’, in Kraft P and Swift K A D, Perspectives in Flavor and Fragrance Research, Zürich, Wiley-VCH, 31–40. nomura, h. and nozawa, f. (1919), Chem Abs, 13, 118. ohloff, g. and sundt, e. (1973), Utilisation de composes carbonyls soufrés comme agents aromatisants, CH531313.

40

Modifying flavour in food

pittet, a. o., klaiber, e. m., vock, m. h., shuster, e. j. and vinals, j. (1976), Food or flavor containing 2,6,6-trimethyl-1-cyclohexen-1-ylacetaldehyde US3940499. reimer, k. (1876), ‘Ueber eine neue Bildungsweise aromatischer Aldehyde’, Berichte der deutschen chemischen Gesellschaft, 9 (1), 423–424. rowe, d. j. (2000), ‘More fizz for your buck: high-impact aroma chemicals’, Perfumer & Flavourist, 25 (5), 1–19. rowe, d. j. (2002), ‘High impact aroma chemicals’, in Swift K A D, Advances in Flavours and Fragrances, Cambridge, Royal Society of Chemistry, 202–226. sanders, j. m. and vock, m. h. (1975), Foodstuffs containing 2-(4-hydroxy-4methylpentyl) norbornadiene, US3886289. schinz, h. and seidel, c. (1957), ‘Untersuchungen über Aromastoffe 1. Mitteilung. über das Himbeeraroma’, Helv Chim Acta, 40 (6), 1839–1859. smith, r. l., cohen, s. m., doull, j., feron, v. j., goodman, j. i., marnett, l. j., portoghese, p. s., waddell, w. j., wagner, b. m. and adams, t. b. (2005), ‘GRAS flavoring substances 22’, Food Tech, 59 (8), 24–62. stoddart, d. m. (1999), ‘The senses: meeting biological needs’, in Bell G A and Watson A J, Tastes and Aromas, Sydney, UNSW, 1–10. tachihara, t., ishizaki, s., ishikawa, m. and kitahara, t. (2005), ‘Studies on the volatile compounds of roasted spotted shrimp’ in Kraft P and Swift K A D, Perspectives in Flavor and Fragrance Research, Zürich, Wiley-VCH, 197–206. taylor, a. j. and roberts, d. d. (2004), Flavor Perception, Oxford, Blackwell Publishing. tiemann, f. and haarmann, w. (1874), ‘Ueber das Coniferin und seine Umwandlung in das aromatische Princip der Vanille’, Ber, 7 (1), 608–623. tonsbeek, c. h. t., plancken, a. j. and v. d. weerdhof, t. (1968), ‘Components contributing to beef flavour: isolation of 4-hydroxy – 5-methyl-3(2H)furanone and its 2,5-dimethyl homolog from beef broth’, J Agric Food Chem, 16 (6), 1016–1021. ullrich, f. and grosch, w. (1987), ‘Identification of the most intense volatile flavour compounds formed during autoxidation of linoleic acid’, Z Lebensm Unters Forsch, 184 (4), 277–282. van den ouweland, g. a. m. and peer, h. g. (1975), ‘Components contributing to beef flavour. Volatile compounds produced by the reaction of 4-hydroxy-5-methyl3(2H)-furanone and its thio analog with hydrogen sulphide’, J Agric Food Chem, 23 (3), 501–505. van der linde, l. m., van dort, j. m., de valois, p., boelens, h. and de rijke, d. (1979), ‘Volatile compounds from thermally degraded thiamine’, in Land D G and Nursten H E, Progress in Flavor Research, London, Applied Science, 219–224. us (1958), Food Additives Amendment to the Federal Food, Drug and Cosmetics Act, Public Law 85–929, 72 Stat 1784. werkhoff, p., bruening, j., emberger, r., guentert, m., koepsel, m., kuhn, w. and surburg, h. (1990), ‘Isolation and characterisation of volatile sulphur-containing meat flavour components in model systems’, J Agric Food Chem, 38 (3), 777–791. willhalm, b., stoll, m. and thomas, a. f. (1965), ‘2,5-Dimethyl-4-hydroxy-2,3 dihydrofuran-3-one’, Chem & Ind, 18 September, 1629–1630. winter, m. (1961), ‘Odeur et constitution XIX. Sur des homologues et analogues de la p-hydroxyphényl-1-butanone-3 (“cétone de framboise”)’, Helv Chim Acta, 44 (7), 2110–2121. wright, j. (2004), Flavor Creation, Carol Stream, IL, Allured Publishing Corporation. wto/gatt (1994), Agreement on the Application of Sanitary and Phytosanitary Measures, GATT Doc MTN/FA li-A1A-4, in Final Act Embodying the Results of the Uruguay Round of Multilateral Trade Negotiations, GATT Doc MTN/FA 33 I.L. M. 9.

3 Extraction of flavourings from natural sources G. Cravotto, University of Turin, Italy and P. Cintas, University of Extremadura, Spain

3.1

Introduction

Because food flavourings strongly influence food evaluation and consumer satisfaction, they have always received great attention from the food industry. As far as records can be traced to ancient civilisations, people tried to obtain flavours and perfumes from natural sources by empirical extraction methods. For a long time, plant tissues were the only sources of flavour compounds. Three main difficulties were encountered in isolating these volatile molecules: the compound under consideration may be present in low quantities or in a bound form; the chemical complexity of the source may require elaborate procedures of extraction/purification; and, finally, the natural plant material may be seasonal or not easily available. Natural flavours have simple chemical structures and low toxicities, and are effective at very low concentrations, resulting in very low levels of human exposure or consumption. Naturalness of flavours became an increasingly important requirement for the food market from the mid1980s: nowadays more than 65% of all flavouring ingredients used in Europe and the United States are labelled as natural. This development is putting enormous pressure on the industrial production of natural flavours and stressing its capacity to an extent that can no longer be met by traditional recovery processes. Hence the need to develop new, more efficient extraction methods; biotechnology may also provide new techniques for the microbial and enzymatic production of food flavourings (Dufosse et al., 2002). Commercial enzyme preparations are already being used to improve quality and yield in the processing of fruit juices (Speiser, 1993).

42

Modifying flavour in food Regulation valve

CO2 pressure 90–500 bar

Evaporator

Plant material

40–70 bar Separator -

Extractor

Condenser -

-

CO2 reservoir

Extract (

(

Heater

CO2 pump

Fig. 3.1

3.2

Chiller

Scheme of supercritical CO2 extraction.

General overview

Current techniques of solid – liquid extraction (SLE) are essentially based on diffusion and osmosis. The usual ways to shorten extraction times and improve yields are to increase the temperature of the treatment and/or to repeat it several times with fresh solvent. The simplest and cheapest technique is maceration, where the surface area of a solid matrix is increased and then the material is soaked in a liquid to which it releases its soluble components. As the extraction process is chiefly diffusion controlled, it may take so long that putrefaction sets in before it is completed. At any rate, if local saturation is to be avoided at the matrix surface, stirring must also be used so that diffusion may work more efficiently. In the production of essential oils and, generally, of volatile compounds, steam distillation often proves very useful, although it involves an exposure to heat that may cause degradation of thermally labile molecules. The same drawback affects Soxhlet extraction, the usual preliminary step in the processing of solid samples according to many official methods of analysis. Industrial plants, where large amounts of extract must be produced in a short time, resort to percolation, in which the solvent drips through large cylinders packed with up to several cubic metres of solid matrix. As a single pass would yield but a limited amount of extract, the effluent is enriched by returning it several times to the top of the column. Because efficiency limits are ultimately set by diffusion rates that increase with temperature, the above-mentioned principles can be applied to speed up diffusion, for example by working with hot solvents. A more recent technique (SFE, supercritical fluid extraction) replaces common organic solvents with a gas (e.g. CO2) that, under supercritical conditions, behaves like a non-polar solvent (Fig. 3.1).

Extraction of flavourings from natural sources

Fig. 3.2

43

Supercritical CO2 extractors (Burgundy – ARKOPHARMA, France).

At the end of the process the gas is easily removed at low temperatures and extracted materials are recovered in a dry form. Despite the large outlay for apparatus, this method is used in the food and beverage industries (Raventos et al., 2002; Brunner, 2005). Many industrial plants (Fig. 3.2) use it for decaffeinating coffee and tea (Peker et al., 1992), extracting beer flavouring agents from hops or nicotine from tobacco and separating oils or oleoresins from spices (Kerrola, 1995). The use of ultrasound to extract active principles and flavours from plants carries great advantages because it both saves time and improves the quality of extracts (Povey and Mason, 1998; Vinatoru et al., 1999; Toma et al., 2001; Vinatoru, 2001). Since extraction time depends mainly on diffusion rates, ultrasound greatly speeds up the process by breaking cell walls and disrupting tissues. If the matrix has been previously dried, ultrasound accelerates its rehydration and swelling. In the laboratory, one can employ either simple ultrasound baths or the more efficient probe systems. On the industrial scale one can use either large tubs resembling a giant ultrasound bath, with transducers applied underneath, or mechanically stirred

44

Modifying flavour in food

Fig. 3.3

A 400 litre ultrasound-extractor (Martin Bauer Spa, Nichelino, Italy).

cylindrical reactors with transducer plates mounted on the lower part of the wall (Fig. 3.3). Microwave heating of fresh plant material is a simple way to achieve direct distillation of essential oils (Lucchesi et al., 2004) (Fig. 3.4). It also can strongly enhance extractions with organic solvents (Pare et al., 1994). Liquid-phase extraction (often used to isolate essential oils from plants) is based on the different absorbance of microwaves by materials with different dielectric constants. The matrix to be extracted (usually water-rich) is mixed with a solvent having a low dielectric constant, so that most of the heating effect will be concentrated in the plant material. Ganzler et al. (1986) were the first to report that this method had many advantages over conventional techniques for the extraction of natural compounds. Microwave-assisted extraction (MAE) requires less solvent and less time, yielding better products at lower costs. Many flavours have been obtained in this way, e.g. glycosylated flavour precursors from grape juice and grapes (Bureau et al., 1996), rosemary, peppermint and other essential oils (Chen and Spiro, 1994; Collin et al., 1991). All the above-mentioned extraction techniques have also been exploited in various combinations, e.g. mechanical stirring combined with ultrasound or SFE with ultrasound. We studied ultrasound-assisted extraction (UAE) and MAE of soybean germ using several sorts of ultrasound apparatus (cup horn, immersion horn, cavitation tube) working at different frequencies (19, 25, 40 and 300 kHz); we also combined two transducers by inserting an immersion horn in the cavitation tube. In a newly developed apparatus, we achieved simultaneous ultrasound and microwave irradiation by inserting a Pyrex horn in the multimode oven (Fig. 3.5). Optimum extraction times were determined and yields were compared with those obtained with MAE (open and closed vessel) and with conventional methods. The best

Extraction of flavourings from natural sources

Fig. 3.4

45

Microwave promoted dry-distillation of fresh plants (Milestone Spa, Sorisole – BG, Italy).

oil yield was obtained with the cavitation tube (19 kHz) and with a double sonication (Fig. 3.6) employing an additional immersion horn (25 kHz) (Table 3.1). Compared with conventional extraction methods, much higher yields were achieved with closed-vessel microwave at 120 ºC and with simultaneous ultrasound/microwave irradiation (Fig. 3.5). Extraction times were reduced by up to 10-fold and yields increased by 50–120%. Gas chromatography (GC) analyses showed only slight or negligible differences in the methyl ester profiles of oils extracted under high-intensity ultrasound and those in Soxhlet.

46

Modifying flavour in food

Fig. 3.5

(a)

Reactor for simultaneous ultrasound/microwave irradiation (Prof. Cravotto’s laboratories, University of Turin).

(b)

Fig. 3.6 Double ultrasound irradiation: (a) the titanium horn is inserted in the cavitation tube; (b) top view of the cavitation tube (a hollow titanium cylinder) (Prof. Cravotto’s laboratories, University of Turin).

Extraction of flavourings from natural sources

47

Table 3.1 Extraction of soybean germ: optimal extraction times and oil yields as percentage of matrix weight Extraction method

Separation funnel (static) Soxhlet Immersion horn 25.0 kHz, 80 W Immersion horn 40.0 kHz, 80 W Cup horn 300 kHz, 70 W Cavitation tube 19.0 kHz, 80 W Cavitation tube 19.0 kHz, 65 W and immersion horn 25.0 kHz, 60 W Combined ultrasound/ microwave (Pyrex horn) 50 W Microwave open-vessel† Microwave† under pressure

Soybean germ (g)

Solvent* (ml)

Temperature (ºC)

15 15 8

50 100 50

Ambient Reflux 40

8 4 1

3.5 8.6 9.5

8

50

40

1

8.3

8 8

40 50

40 40

1 1

6.4 17.7

8

50

40

0.5

17.9

8

40

40

1

14.1

3 2

40 35

Reflux 120

1 0.5

10.0 16.5

Time (h)

Yield (%)

* Solvent = petroleum ether. † Milestone magnetic stirrer (P/N 86116) and weflon button for apolar solvents (P/N WO1703).

There are several ways to extract essential oils from plants and the molecular composition of the product will depend to some extent on the method used, although many oils can be extracted by only one method: •

Solvent extraction – this method is not suitable for flavouring oils because solvent residues will taint flavours. Besides maceration and percolation, we can also list boiling with water (infusion), extraction with cold fat (enfleurage) and extraction with hot fat. • Cold pressing (expression) – probably the most desirable method for flavouring oils, it is normally used only for citrus oils. Terpenes are extracted from the plant material by this method and a subsequent process, such as distillation or counter-current extraction, is required to make a terpene-free oil. • Distillation – direct essential oil distillation, water and steam distillation and steam distillation. This is probably the most common extraction method. Steam is used to break down the plant material and the essential oils, released as vapour, are then condensed. This method is not recommended if an alternative is at hand, because insoluble terpenes are extracted and more terpenes may be produced when the plant material is heated. Moreover, some of the more volatile flavour molecules may

48

Modifying flavour in food be lost. Terpenes may be removed from the oil by a subsequent distillation process, which will, however, entail a further loss of volatile flavour components.

Non-conventional extraction techniques (Luque de Castro et al., 1999) include: • • • • • •

vortical- or turbo-extraction (high-speed mixing) (List and Schmidt, 1990); UAE; pressurised solvent extraction (PSE); MAE; extraction by electrical energy; SFE – liquid CO2 at 10 ºC and a very high pressure is used as a solvent to remove essential oils from plant material. Once atmospheric pressure is restored, CO2 rapidly escapes, leaving no residue. As this process does not remove terpenes, it results in a terpene-free oil.

Volatile aroma compounds can be recovered by pervaporation, a vapourphase separation process based on selective permeation through a dense organophilic membrane (Baudot and Marin, 1997). Essential oils are widely used in the perfume and food industries. They are complex mixtures that basically consist of two fractions, the more abundant one (90–95%) being made up of volatiles (monoterpenes, sesquiterpenes, hydrocarbons and their oxygenated derivatives, along with aliphatic aldehydes, alcohols and esters). The non-volatile residue, amounting to 5– 10% of the whole oil, contains hydrocarbons, fatty acids, sterols, carotenoids, waxes, coumarins, psoralens and flavonoids (Luque de Castro et al., 1999). Terpenes, ranging from approximately 99% of the volatile fraction in grapefruit oil to 60% in bergamot oil, make little contribution to the flavour or fragrance of the oil. Moreover, as they are mostly unsaturated compounds, they are decomposed by heat, light and oxygen to produce undesirable compounds that can be detrimental to flavour and aroma. The oxygenated fraction is highly odoriferous and is mainly responsible for the characteristic flavour. The isolation, concentration and purification of essential oils have been important processes for centuries, as a consequence of the widespread use of these compounds. Commonly used methods to this day are mainly based on solvent extraction and steam distillation. The drawbacks to these techniques have led to a search for new alternative extraction processes. Another important goal in the extraction industry is to improve the quality of the oil. As essential oils used in the food and perfume industries are commonly isolated by cold pressing, they contain more than 95% of monoterpene hydrocarbons, mainly limonene. The industrial practice of deterpenation or ‘folding’ removes some of the limonene along with other unstable terpenes, and thus concentrates the oxygenated fraction. This process aims to improve product stability, to increase the

Extraction of flavourings from natural sources

49

solubility of the oil in low-proof alcohol, food solvents and water, besides reducing storage and transportation costs. The commercial methods currently used for ‘folding’ are fractional vacuum distillation, selective solvent extraction and chromatographic separation. All these methods have serious drawbacks, such as low yields, formation of by-products (owing to exposure to heat) and the presence in the extracts of potentially toxic organic residues. In this context the adoption of new techniques is an urgent issue. Extraction with supercritical CO2 (SC-CO2) and especially with subcritical water are effective methods to isolate high-quality essential oils. Many authors have compared conventional techniques (including discontinuous, continuous and hybrid approaches) with SC-CO2, subcritical water extraction, UAE and MAE. Pressurised liquid extraction (PLE) is another competitive technique for flavour extraction that has grown in recent years. Superheated water extraction (SWE) used in essential oil extractions (Ollanketo et al., 2002; Gogus et al., 2006) also belongs to this group of techniques. The crucial features of SWE are temperature, pressure, flow rate and extraction time. In many SWE studies, a flow rate of 2 ml/min, moderate pressure (enough to keep water in the liquid state) and various temperatures have been used. By appropriate temperature programming it is possible to modify, in a certain range, the polarity of water. In other words, under SWE, the solvent may behave like water or like methanol. The advantages and drawbacks of each method have to be considered before they can be applied commercially. The use of non-toxic extracting solvents such as carbon dioxide and water is highly attractive as they carry great economic and environmental benefits and deliver products that are likely to be better accepted by consumers. The enormous demand for flavourings has been met mostly by chemical synthesis that produces the desired flavour compounds from readily available sources. Today this approach is markedly declining because of new regulations concerning food additives (see Chapter 2) and because of consumer aversion to chemicals. For these reasons in the past ten years the bio-production of natural flavours has risen from 5–10% to 75–80% of the total output. Microorganisms have been identified as potential sources of natural flavours, and interest in new fermentation processes, as well as in enzymatic conversions, is now growing. Organic solvent extraction must be used with caution on fermentations because they may exert toxic effects on microorganisms. One way to overcome this problem might be extraction with SC-CO2, one of the new techniques for recovering molecules such as aromas from dilute fermentation broths (Fabre et al., 1999). Owing to its good solvent properties and good transport properties (low viscosity and high diffusivity), this approach appears very promising for the extraction of natural compounds. Moreover, CO2 is non-toxic, non-flammable and a physiological molecule, a favourable aspect in the food industry. Numerous studies highlighted that SC-CO2 efficiently removed terpenes, limonene, eugenol and thymol from plant tissues (Kerrola, 1995); it also made it

50

Modifying flavour in food

possible to obtain aromatic extracts from lilac and rose, from vanilla, ginger, rosemary, sage, wild rose and clove. These studies highlighted the good conservation of the organoleptic properties of the extracts (Quirin, 2004). One of the most common flavours is natural vanillin (4-hydroxy-3methoxybenzaldehyde), which is used in a broad range of food products. In the vanilla bean, it makes up 0.2% of the dry weight and is associated with many other compounds. Approximately 12 000 tonnes of vanillin are consumed annually, of which only 20 are extracted from vanilla beans (Ramachandra Rao and Ravishankar, 2000). Beside the relatively high cost of conventional vanilla cultivation, the common extraction methods used to obtain the flavour from the cured bean lack efficiency and speed. Percolation (ethanol/water) takes from 48 to 72 h, while the oleoresin method consists of circulating ethanol over the pulverised whole beans under vacuum (45 ºC) and takes about 8–9 days. However, by the latter process, an approximately 10-fold stronger vanilla extract may be obtained. Commercially, natural vanillin is sold as a dilute ethanolic extract. High-pressure techniques have found application in food industry. This technology, referred to as ultra-high pressure (UHP) and high hydrostatic pressure (HHP), may replace the conventional pasteurisation treatment of food. Thermally pasteurised fruit juices often suffer from a loss of fresh flavour notes. Organoleptic alterations in a juice mixture composed of orange, lemon and carrot (OLC) were analysed as a function of high pressure treatment and storage time (up to 21 days at 4 ºC). In standard pasteurised juices, odour and flavour quality decreased significantly during storage. In OLC juice treated at 500 MPa for 5 min, the changes in odour and flavour were but slight (Butz and Tauscher, 2002). HHP treatment was advantageous for strawberry flavour because β-glucosidase was significantly inactivated at 600–800 MPa (Zabetakis et al., 2000). The following sections present the innovative extraction techniques in greater detail, with a brief survey of published studies in which a comparison was made of several extraction methods of flavours from natural sources (Luque de Castro et al., 1999; Szentmihalyi et al., 2002).

3.3

Supercritical fluid extraction: SC-CO2

Supercritical CO2, the sole supercritical fluid used in the food industry, exists when carbon dioxide is heated under pressure (>31 ºC, >74 bar) (Fig. 3.7). The supercritical phase exhibits a very low viscosity and excellent solvent properties, and its use to extract natural matrices generally yields products of better quality than those obtained using organic solvents, whose residues, even if present as traces, are objectionable. SFE offers some important advantages, as it can be performed at low temperatures, avoiding

Pressure, P

Extraction of flavourings from natural sources

51

Supercritical phase Liquid phase Solid phase

Tc = 31.1 °C Pc = 73.8 bar Critical point

Triple point (–56.6 °C, 5.19 bar)

Gaseous phase

Temperature, T

Fig. 3.7

CO2 phase diagram.

the degradation of thermolabile compounds and limiting hydrolytic phenomena that lead to loss of water-soluble compounds. One of the greatest hurdles besetting supercritical fluid extraction of plant raw materials is the very slow kinetics of the process (Brunner, 1994). The classical way to accelerate the process is via the application of a mechanical agitation system. As the high pressures required by SFE do not allow it, an alternative must be considered. The use of ultrasound (both high and low intensity) has been applied in food technology, especially to modify the texture of food products. In particular, the use of high-intensity ultrasound (18–100 kHz) has been suggested to enhance heat and mass transfer, dehydration processes, inactivation of enzymes and de-gasification of liquids. The use of acoustic waves in SFE can effectively accelerate the process, owing to the high density of the solvent (the velocity of propagation increases with density) (Riera et al., 2004). The literature reports a plethora of applications of supercritical CO2 to extraction processes in food technology, such as decaffeination of green coffee beans, production of hop extracts, recovery of aromas and flavours from spices and herbs, extraction and fractionation of edible oils and removal of contaminants. Yonei et al. (1995) extracted ginger flavour in a two-step separation procedure working in a certain range of pressure and temperature (7.9–29.5 MPa, 288–353 K); in this way they concentrated 6-gingerol to produce a highly pungent ginger flavour. Volatile extracts of marjoram leaves (Majorana hortensis Moench) obtained by hydrodistillation (HD) and SFE were compared (El-Ghorab

52

Modifying flavour in food

et al., 2004). The antioxidant power of the latter was markedly greater than that of the former, possibly because it had a higher content of γ-terpinene, terpinolene and thymol. Since 1995, several reviews have comprehensively covered extraction methods for flavours and essential oils, comparing SFE with other techniques (Quirin, 2004). Beside affording high extraction yields, supercritical CO2 can achieve a better degree of selectivity (Senorans et al., 2003; Pourmortazavi et al., 2005; Bhattacharjee et al., 2005).

3.4

Continuous subcritical water extraction (CSWE)

The use of water in a dynamic mode, at temperatures between 100 and 374 ºC (the critical point of water lies at 221 bar and 374 ºC) and pressures high enough to maintain the liquid state, is emerging as a powerful alternative method for the extraction of solids (Luque de Castro and JimenezCarmona, 1998). Its outstanding feature is the easy manipulation of the dielectric constant (er), that can be made to vary over a wide range just by changing the temperature. Thus, at room temperature and pressure, water has a dielectric constant of ca. 80, which makes it an extremely polar solvent. This value can be drastically lowered by raising the temperature under moderate pressure. For instance, subcritical water at 250 ºC and a pressure just over 40 bar has an er value of 27, which is close to that of ethanol, and is suitable for the leaching of low-polarity compounds. The recent application of this technique to the isolation of essential oils is very promising. The extraction rate is determined by compound partition between the matrix and water, rather than by the diffusion rate of a compound out of the matrix. Superheated water under pressure between 125 and 175 ºC has been shown to rapidly extract the oxygenated fragrance and flavour compounds from Rosmarinus officinalis, while the monoterpenes are extracted slowly, and only very small amounts of the sesquiterpenes, waxes and lipids are removed. The yields of oxygenated compounds are higher than those obtained by steam distillation and the organoleptic properties are better preserved. Although larger amounts of water are needed for extraction by superheated water, the energy costs are competitive because it is not necessary to vaporise the water and most of the heat required can be recycled (Basile et al., 1998). Similar results are reported for the isolation of marjoram essential oil with water at 50 bar and 150 ºC. An in-depth study of the variables affecting the process showed that temperature was a key variable. In the range between 100 and 175 ºC the optimum temperature was 150 ºC, providing essential oil of the best quality (Jiménez-Carmona et al., 1999). Subcritical water extraction has also been shown to be a promising technique for the deterpenation of lemon oil, removing approximately 96% of the monoterpenes.

Extraction of flavourings from natural sources

3.5

53

Ultrasound-assisted extraction (UAE)

Techniques have been developed to obtain valuable products from plant sources on an industrial scale. Recent trends in extraction techniques are largely focused on finding solutions that minimise, or even dispense with, the use of solvents (Chemat et al., 2004b). The application of ultrasound to enhance extraction yields began in the 1950s at the laboratory level (Mason, 1997). It has been recently shown to improve the extraction from plant materials containing polysaccharides and essential oils, mainly through the phenomenon of cavitation (Hromàdkovà et al., 1999).The mechanical effect of ultrasound accelerates the release of organic compounds contained within the plant body by disrupting cell walls, by enhancing mass transfer and facilitating solvent access to the cell content. This effect is much stronger at low frequencies (18–40 kHz), and practically negligible at 400–800 kHz, assuming of course the same wave intensity. UAE has proved to be a versatile technique that can be scaled up to the great benefit of industrial production. Studies on the UAE of the main constituents of sage (Salvia officinalis) showed that the recovery of cineole, thujone and borneol was better when sonication was carried out with a probe system rather than in an ultrasonic cleaning bath or with conventional techniques. The use of ultrasound afforded purer extracts in better yields, in shorter times and working at lower temperatures (Povey and Mason, 1998; Vinatoru et al., 1999; Toma et al., 2001; Vinatoru, 2001). A comparative study of aqueous ethanol extracts obtained from Salvia officinalis (see Table 3.2) and valerian roots by conventional and UAE also showed the latter to be more efficient. Vinatoru et al. (1997) showed that the main advantage of UAE is the enhanced hydration process that takes place simultaneously with matrix fragmentation without any appreciable chemical degradation. In the last few years low-intensity ultrasound, a non-destructive technique, has been put to more extensive use in the food industry, whether for analysis and production control, or for processing and modifying foods. High-intensity ultrasound has been combined with other extraction Table 3.2 Extraction of Salvia officinalis components (mg/kg) under various conditions Method Stirring Ultrasound bath Stirring Ultrasound bath Ultrasound + stirring Horn (20 kHz) * Average temperature.

t (h)

T (ºC)

Cineole

Thujone

Borneol

3 3 3 3 3 2

20 20 30 30 20 20*

10.3 13.0 10.0 16.5 22.7 24.3

63.6 81.6 71.7 118.1 141.9 167.2

3.8 4.3 3.0 3.6 6.3 5.8

54

Modifying flavour in food

techniques. One of the drawbacks of SFE, one that sets limits for some industrial applications, is that it is not compatible with mechanical agitation systems, such as are commonly used to increase mass transfer (Riera et al., 2004). On the other hand, high-intensity ultrasound causes heating of the medium, precipitation of some dissolved materials, degasification of liquids by bubble collapse and formation of acoustic micro-currents, phenomena that all concur to an interface shrinking, that is, to enhance mass transfer coefficients during the extraction. Rice-bran oil (Cravotto et al., 2004) and soybean-germ oil (Li et al., 2004) (SG) (see Table 3.1) were obtained in higher yield under UAE with only minimal degradation of thermosensitive components (Cravotto et al., 2005). Sonication of sunflower oil did not cause significant changes; however, some off-flavour compounds such as n-hexanal and limonene resulting from its ultrasonic degradation were identified by gas chromatography – mass spectrometry (GC-MS) (Chemat et al., 2004a). Acoustic cavitation induced by ultrasonic cutting devices which are used for food packaging, may also induce changes on edible oil composition (Schneider et al., 2006). UAE and MAE (Section 3.6) were compared in sample preparation for the determination in a traditional Chinese medicine composed of three water-soluble bioactive constituents (danshensu, puerarin and ferulic acid) (Jiantao et al., 2004). MAE worked very fast: using 30% EtOH as solvent (pH 2.5) the extraction was complete in 2 min at 750 W, while UAE took 30 min for the same result working however at a lower temperature. A good chemical tally was found between the two extracts. The literature reports several examples of UAE of flavours, e.g. the extraction of volatile compounds from citrus flowers and citrus honey (Alissandrakis et al., 2003) and that of filbertone from hazelnut oil (Ruiz del Castillo and Herraiz, 2003).

3.6

Microwave-assisted extraction (MAE)

The use of dielectric heating in analytical laboratories started in the late 1970s (Yeo and Shibamoto, 1991; Parliment et al., 1994). Ten years later, MAE was used for the first time to prepare samples for chromatographic determination of natural products and pesticides in foods and soil matrices (Sanghi and Kannamkumarath, 2004). In continental Europe microwave generators usually work at 2450 MHz but, in the United Kingdom and the United States, 915 MHz is the operating frequency. The lower frequency confers greater penetration but requires bigger generators and apparatus. Microwave dielectric heating depends on the ability of the solvent or the plant material to absorb microwave energy and convert it to heat. Unlike conductive heating, microwaves heat the whole sample volume simultaneously. Microwaves also disrupt weak hydrogen bounds by promoting the rotation of molecular dipoles, an effect that is opposed by the viscosity of the medium. Furthermore, the movement of dissolved ions increases solvent

Extraction of flavourings from natural sources

55

penetration into the matrix and thus facilitates analyte solvation. The resulting friction with the medium releases heat by Joule effect. This phenomenon depends on the size and charge of the ions present in the solution. The effect is strongly dependent on the nature of both the solvent and the solid matrix. Commonly used solvents cover a wide range of polarities, from heptane to water. In most cases, the chosen solvent has a high dielectric constant and strongly absorbs microwaves; however, the ability of the medium to interact with microwaves and the selectivity of extraction can be modulated by using mixtures of solvents. Sometimes the matrix itself interacts with microwaves while the surrounding solvent, having a low dielectric constant, remains relatively cold. This set-up, which presents obvious advantages when dealing with thermosensitive compounds, has been successfully used for the extraction of essential oils (Kaufmann and Christen, 2002). In fact, microwaves interact selectively with the polar molecules present in glands, trichomes or vascular tissues. Localised heating causes the expansion of cells and rupture of their walls, followed by the release to the solvent of essential oils (Craveiro et al., 1989). This situation also occurs when a dry sample has been rehydrated before extraction (Pare et al., 1994). Moisture content is essential in MAE because water locally superheats and promotes the release of analytes into the surrounding medium; control of the matrix water content also leads to more reproducible results. The recent literature shows that usually MAE works better than conventional methods, as it saves time and solvent, accomplishes a more thorough extraction and yields better products (Cesare et al., 1995; Sides et al., 2000). While the Soxhlet method usually takes hours, up to 20 or more, MAE only takes a few minutes, using at least 10-fold less solvent. Although for the time being MAE has not proceeded beyond the small pilot scale, the apparatus is simpler and cheaper than is required by SFE and can be built to higher capacities. The method can be used for a greater variety of analytes and with less limit of the polarity of the solvent. Two types of instruments are commercially available, using different approaches. The most common involves working in a closed vessel under controlled pressure and temperature; the other uses an open vessel under atmospheric pressure. There is a series of commercial microwave reactors, operating at either single-mode or multimode conditions, enabling accurate and reproducible results. Large-scale industrial equipment is also available (Singh, 1993; Stuerga and Delmotte, 2002). In the MAE of essential oils, if samples are suspended in an apolar solvent such as hexane, MW will heat almost exclusively the inner glandular and vascular systems of the fresh plant material owing to their high moisture content. The resulting disruption of cell membranes releases the analytes to the solvent. These applications, together with the direct distillation of volatile oils from plants and other biological material have been patented by Paré at the beginning of the 1990s. The latter procedure was applied by

56

Modifying flavour in food

Fig. 3.8

Microwave-promoted dry-distillation of fresh plants under nitrogen stream (Prof. Cravotto’s laboratories, University of Turin).

many authors to obtain essential oils (e.g. Cuminum cyminum L. and Zanthoxylum bungeanum M. (Wang et al., 2006) or Rosmarinus officinalis (Lo Presti et al., 2005). After microwave extraction, scanning electron micrographs of the leaves showed structural changes in the oil-containing glands: most of them were collapsed and some completely disintegrated. Two explanations of these facts are plausible, one involving diffusion of the essential oil across the gland wall and the other a rupture of the wall and liberation of the constituents into the solvent. Either may apply, depending on the maturation stage of the gland. Using a domestic microwave oven, MAE has been compared with classical hydrodistillation for the extraction of essential oils from 10 different plant species (Lucchesi et al., 2004). The yields were generally similar, but the chromatographic profiles varied dramatically, especially in the ratios between different peaks. Microwave irradiation of fresh leaves promotes the distillation of volatiles and an air or nitrogen stream passing through it facilitates the stripping (Fig. 3.8). The yields obtained after a few minutes were comparable to those obtained after 60–90 min of steam distillation, and both qualitative and quantitative results were superior to conventional method.

3.7

Extraction in the analysis of flavours

This field is widely covered by many reviews dealing with modern techniques for the extraction and quantitative analysis of flavours (Modey et al., 1996; Wilkes et al., 2000; Chaintreau, 2001; Werkhoff et al., 2002). Traditional methods

Extraction of flavourings from natural sources

57

for isolating volatile components have been complemented by solid-phase microextraction (SPME), GC combined with head-space volatile-compound sampling, high-pressure extraction with supercritical CO2, solvent-assisted flavour evaporation, stir-bar sorptive extraction and GC-MS. Simultaneous distillation–extraction (SDE) stretches over 50 years of history in organic synthesis, but its original use was in analytical applications when Likens and Nickerson (1964) designed the original device. This onestep isolation–concentration of flavour constituents allowed a dramatic time saving and a great reduction of treated liquid volumes because of continuous recycling. SDE remains one of the most frequently used isolation methods for recovering a wide range of compounds from a non-volatile matrix without further clean-up. This does not mean that it is a universal technique; yet, when properly used, i.e. taking into account its possibilities and limitations, it often achieves the highest recoveries. Modelling of the distillation–extraction process and computer-simulated operation enables the analyst to optimise conditions. New improvements are at hand, such as preparative and analytical isolations carried out at room temperature that minimise the risk of artefact formation. For heat-sensitive products, vacuum simultaneous distillation– solvent extraction (V-SDE) and headspace sampling play complementary roles and are conceivably the most selective techniques for the separation of volatiles from non-volatiles. Used in combination, and without submitting the sample to heat, they can produce extracts of almost all the volatile flavours and fragrances. An interesting application of SDE is the recovery of water-soluble constituents from aqueous solutions obtained during rose oil production (Eikani et al., 2005). The most common perfumed extracts from the flowers are rose oil and rose water, derived from steam distillation. The water which forms the major portion of the distillate may hold in solution an appreciable amount of the valuable part of the oil; by SDE the rose oil components may be identified and quantified from the aqueous phases obtained in both small- and large-scale rose oil production. Conventional Soxhlet extraction of vegetal solid matrices for analytical sample preparation was compared with extraction under ultrasound and under microwave (Rashmi and Kannamkumarath, 2004). Although recovery of standard pesticides was more efficient with MAE, this method has to be re-optimised whenever a new matrix is analysed. The efficiency of ultrasound-assisted extraction can often be related to the power, frequency and type of instrument used, e.g. frequencies higher than 45 kHz are less efficient than 18–25 kHz. Pressurised hot water extraction, also called subcritical water extraction, is a novel approach for the extraction of a wide variety of substances, as pressurised water at higher temperatures becomes progressively less polar. Fundamental and practical aspects of this technique have been reviewed (Eskilsson et al., 2004).

58

Modifying flavour in food

All reported applications have shown that pressurised hot water extraction is a viable alternative to conventional techniques (Soxhlet and/or standard solvent extraction procedures). There is also evidence suggesting that it may compete favourably with other recent extraction techniques such as SFE and PLE with organic solvents. Early work with subcritical water demonstrated its ability to selectively extract different classes of compounds, the more polar organics being extracted at lower temperatures and the less polar ones (phenols, hydrocarbons) at higher temperatures. Thus, the selectivity of subcritical water extraction (a preference for more polar organics at milder conditions) runs contrary to that of supercritical CO2 (preferring non-polar over polar organics). Recently, solubilities of oxygenated flavour and fragrance components were measured in subcritical water and found to be substantially higher than non-oxygenates. For example at 100 ºC (ca. 65 bar), the oxygenates, carvone and eugenol, have solubilities two orders of magnitude higher than the non-oxygenated compound d-limonene. These solubility trends correspond to studies demonstrating high recoveries of oxygenated flavour in solvents. The analysis of essential oils may combine adsorption on inert material and desorption by a supercritical fluid. To increase the ratio of oxygenated flavour compounds to terpene hydrocarbons in orange oil, this is fractionated on porous silica gel, then further purified by desorption into SC-CO2 (Shen et al., 2002).

3.8

Drying methods and solvent distillation

Most drying methods are known to affect volatiles (Coumans et al., 1994). For instance, oven-drying at 30 ºC did not deplete volatiles in sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) significantly but drying at 60 ºC caused a marked loss (Venskutonis, 1997). Air-drying preserved more volatiles of parsley (Petroselinum crispum L.) compared with oven-drying at 45 ºC and freeze-drying. The volatiles lost through the latter methods were the monoterpenes that make up the most of parsley aroma (p-mentha1,3,8-triene and apiole) (Diaz-Maroto et al., 2002, 2003). Freeze-drying generally has the most pronounced effect and consistently fails to preserve the volatile profile of the fresh plant, usually to the detriment of the compounds that give the plant its unique aroma (Abascal et al., 2005). Several flavour components of blueberries (Vaccinium spp.) are lost, including 1,8-cineole which gives this fruit its characteristic aroma. On the other hand, freeze-drying preserved more of the aroma volatiles of dill (Anethum graveolens) than hot air drying but both methods failed to capture over 80% of benzofuranoids, the most important fraction (Huopalahti et al., 1985). Over 50% of the total volatiles in dried and freeze-dried samples consisted of secondary aroma compounds (mostly phytadienes) formed

Extraction of flavourings from natural sources

59

during processing. In the final analysis, the freeze-dried sample preserved only 25% of the total primary aroma compounds found in fresh dill. Freezedrying preserved fairly well total volatiles in sage but increased the absolute thymol content by 33%, indicating a change in the volatile profile of the plant. Both oven- and freeze-drying increased the percentage in the headspace of unknown volatiles, presumably degradation products. Sometimes, drying procedures have a favourable effect on aroma, e.g. all drying methods increase the characteristic minty odour of spearmint which may be a plus for some uses, such as in herbal teas. The behaviour of volatile compounds of banana during dehydration was determined during air-drying (AD), during vacuum microwave-drying (VMD) and during freeze-drying. The process that yielded the most crisp banana chips with significantly higher volatile levels and sensory ratings was the combination of AD and VMD (Mui et al., 2002). A comparison between microwave heating and conventional roasting of cumin seeds (Cuminum cyminus L.) showed that microwave-heated samples better retained the characteristic flavour compounds, such as typical aldehydes (Behera et al., 2004).

3.9

Conclusion

Conventional extraction processes for flavours and essential oils are timeconsuming, laborious, involve large amounts of solvents and ultimately cause some thermal decomposition of the target molecules and partial loss of volatiles. Great improvements may be in the offing with the new techniques and their combinations; each matrix, however, will need careful optimisation of operating conditions.

3.10

Acknowledgements

This work has been carried out within the European Union COST Action D32 (Working Group D32/006/04: Microwave and high-intensity ultrasound in the synthesis of fine chemicals). Financial support from MIUR-COFIN (Italy) and the Spanish Ministry of Education and Science (grant CTQ200507676/BQU) are also acknowledged.

3.11

References

abascal, k., canora, l. and yarnell, e. (2005), ‘The effect of freeze-drying and its implications for botanical medicine: a review’, Phytother Res, 19, 655–660. alissandrakis, e., daferera, d., tarantilis, p. a., polissiou, m. and harizanis, p. c. (2003), ‘Ultrasound-assisted extraction of volatile compounds from citrus flowers and citrus honey’, Food Chem, 82 (4), 575–582.

60

Modifying flavour in food

basile, a., jimenez-carmona, m. m. and clifford, a. a. (1998), ‘Extraction of rosemary by superheated water’, J Agric Food Chem, 46, 5205–5209. baudot, a. and marin, m. (1997), ‘Pervaporation of aroma compounds: comparison of membrane performances with vapor–liquid equilibria and engineering aspects of process improvement’, Food Bioprod Processing, 75, 117–142. behera, s., nagarajan, s. and mohan rao, l. j. (2004), ‘Microwave heating and conventional roasting of cumin seeds (Cuminum cyminus L.) and effect on chemical composition of volatiles’, Food Chem, 87, 25–29. bhattacharjee, p., kshirsagar, a. and singhal, r. s. (2005), ‘Supercritical carbon dioxide extraction of 2-acetyl-1-pyrroline from Pandanus amaryllifolius Roxb’, Food Chem, 91 (2), 255–259. brunner, g. (1994), Gas Extraction: An Introduction to Fundamentals of Supercritical Fluids and the Application to Separation Processes, New York, Springer, 1994. brunner, g. (2005), ‘Supercritical fluids: technology and application to food processing’, J Food Eng, 67, 21–33. bureau, s., razungles, a., baumes, r. and bayonove, c. (1996), ‘Glycosylated flavor precursor extraction by microwaves from grape juice and grapes’, J Food Sci, 61 (3), 557–559. butz, p. and tauscher, b. (2002), ‘Emerging technologies: chemical aspects’, Food Res Int, 35 279–284. cesare, l. f., diriva, m. and schiraldi, a. (1995), ‘Microwave extraction of basil aroma compounds’, Develop Food Sci, 37, 857–868. chaintreau, a. (2001), ‘Simultaneous distillation-extraction. From birth to maturity. Review’, Flavour Fragr J, 16 (2), 136–148. chemat, f., grondin, i., costes, p., moutoussamy, l., sing, a. s. c. and smadja, j. (2004a), ‘High power ultrasound effects on lipid oxidation of refined sunflower oil’, Ultrason Sonochem, 11, 281–285. chemat, s., lagha, a., ait-amar, h., bartels, p. v. and chemat, f. (2004b), ‘Comparison of conventional and ultrasound-assisted extraction of carvone and limonene from caraway seeds’, Flavour Fragr J, 19, 188–195. chen, s. s. and spiro, m. (1994), ‘Study of microwave extraction of essential oil constituents from plant materials’, J Microwave Electromagn Energy, 29, 231–241. collin, g. j., lord, g., allaire, j. and gagnon, d. (1991), ‘Huiles essentielles et extraits microondes’, Parfums Cosmèt Arômes, 97, 105–112. coumans, w. j., kerkhof, p. and bruin, s. (1994), ‘Theoretical and practical aspects of aroma retention in spray drying and freeze drying’, Drying Technol, 12 (1–2), 99–149. craveiro, a. a., matos, f. j. a., alencar, j. w. and plumel, m. m. (1989), ‘Microwave oven extraction of an essential oil’, Flavour Fragr J, 4 (1), 43–44. cravotto, g., binello, a., merizzi, g. and avogadro, m. (2004), ‘Improving solventfree extraction of policosanol from rice bran by high-intensity ultrasound treatment’, Eur J Lipid Sci Techn, 106, 147–151. cravotto, g., robaldo, b., viola, f., oliaro-bosso, s. and avogadro, m. (2005), ‘Soybean-germ oil inhibits oxidosqualene cyclase in 3T3 fibroblasts’, Eur J Lipid Sci Technol, 107 (10), 701–705. diaz-maroto, m. c., perez-coello, m. s. and cabezudo, m. d. (2002), ‘Effect of drying method on the volatiles in bay leaf (Laurus nobilis L.)’, J Agric Food Chem, 50 (16), 4520–4524. diaz-maroto, m. c., perez-coello, m. s., gonzalez viñas, m. a. and cabezudo, m. d. (2003), ‘Influence of drying on the flavor quality of spearmint (Mentha spicata L.)’, J Agric Food Chem, 51 (5), 1265–1269. dufosse, l., blin-perrin, c., souchon, i. and feron, g. (2002), ‘Microbial production of flavors for the food industry. A case study on the production of

Extraction of flavourings from natural sources

61

gamma-decalactone, the key compound of peach flavor, by the yeasts Sporidiobolus sp.’, Food Sci Biotech, 11 (2), 192–202. eikani, m. h., golmohammad, f., rowshanzamir, s. and mirza, m. (2005), ‘Recovery of water-soluble constituents of rose oil using simultaneous distillation–extraction’, Flavour Fragr J, 20, 555–558. el-ghorab, a. h., mansour, a. and el-massry, k. f. (2004), ‘Effect of extraction methods on the chemical composition and antioxidant activity of Egyptian marjoram (Majorana hortensis M.)’, Flavour Fragr J, 19, 54–61. eskilsson, c. s., hartonen, k., mathiasson, l. and riekkola, m.-l. (2004), ‘Pressurized hot water extraction of insecticides from process dust – comparison with supercritical fluid extraction’, J Sep Sci, 27, 59–64. fabre, c. e., condoret, j.-s. and marty, a. (1999), ‘Extractive fermentation of aroma with supercritical CO2’, Biotech Bioeng, 64 (4), 392–400. ganzler, k., salgo, a. and valko, k. (1986), ‘A novel sample preparation method for chromatography’, J Chromatogr, 371, 299–305. gogus, f., oze, m. z. and lewis, a. c. (2006), ‘Extraction of essential oils of leaves and flowers of Achillea monocephala by superheated water’, Flavour Fragr J, 21 (1), 122–128. hromàdkovà z., ebringerovà a. and valachovic, p. (1999), ‘Comparison of classical and ultrasound-assisted extraction of polysaccharides from Salvia officinalis L.’, Ultrason Sonochem, 5, 163–168. huopalahti, r., kesaelahti, e. and linko, r. (1985), ‘Effect of hot air and freeze drying on the volatile compounds of dill (Anethum graveolens L.) herb’, J Agric Sci Finl, 57 (2), 133–138. jiantao, h., zhihong, s. and wenbao, c. (2004), ‘Comparison of microwave-assisted and ultrasound-assisted extraction for determination of main water-soluble bioactive constituents in traditional Chinese medicinal preparation tongmaichongji by HPLC-DAD’, J Liq Chrom Rel Technol, 27 (11), 1769–1784. jimenez-carmona, m. m., ubera, j. l. and luque de castro, m. d. (1999), ‘Comparison of continuous subcritical water extraction and hydrodistillation of marjoram essential oil’, J Chromatogr A, 855 (2), 625–632. kaufmann, b. and christen, p. (2002), ‘Recent extraction techniques for natural products: microwave-assisted extraction and pressurised solvent extraction’, Phytochem Anal, 13, 105–113. kerrola, k. (1995), ‘Literature review: isolation of essential oils and flavor compounds by dense carbon dioxide’, Food Rev Int, 11 (4), 547–573. li, h., pordesimo, l. and weiss, j. (2004), ‘High intensity ultrasound-assisted extraction of oil from soybeans’, Food Res Int, 37, 731–738. likens, s. t. and nickerson, g. b. (1964), ‘Detection of certain hop oil constituents in brewing products’, Am Soc Brewing Chemists Proc, 5, 5–13. list, p. h. and schmidt, c. c. (1990), Phytopharmaceutical Technology, Cleveland, CRC Press. lo presti, m., raqusa, s., trozzi, a., dugo, p., visinoni, f., fazio, a., dugo, g. and mondello, l. (2005), ‘A comparison between different techniques for the isolation of rosemary essential oil’, J Sep Sci, 28 (3), 273–280. lucchesi, m. e., chemat, f. and smadja, j. (2004), ‘Solvent-free microwave extraction of essential oil from aromatic herbs: comparison with conventional hydrodistillation’, J Chromatogr A, 1043, 323–327. luque de castro, m. d. and jimenez-carmona, m. m. (1998), ‘Potential of water for continuous automated sample-leaching’, Trends Anal Chem, 17 (7), 441–447. luque de castro, m. d., jimnez-carmona, m. m. and fernandez-pèrez, v. (1999), ‘Towards more rational techniques for the isolation of valuable essential oils from plants’, Trends Anal Chem, 18 (11), 708–716.

62

Modifying flavour in food

mason, t. j. (1997), ‘Ultrasound in synthetic organic chemistry’, Chem Soc Rev, 26, 443–451. modey, a., mulholland, w. k., dulcie, a. and raynor, m. w. (1996), ‘Analytical supercritical fluid extraction of natural products’, Phytochem Anal, 7 (1), 1–15. mui, w. w. y., durance, t. d. and scaman, c. h. (2002), ‘Flavor and texture of banana chips dried by combinations of hot air, vacuum, and microwave processing’, J Agric Food Chem, 50, 1883–1889. ollanketo, m., peltoketo, a., hartonen, k., hiltunen, r. and riekkola, m. l. (2002), ‘Extraction of sage (Salvia officinalis L.) by pressurized hot water and conventional methods: antioxidant activity of the extracts’, Eur Food Res Technol, 215, 158–163. pare, j. r. j., belanger, j. m. r. and stafford, s. s. (1994), ‘Microwave-assisted process (MAP(TM)) – a new tool for the analytical laboratory’, Trends Anal Chem, 13, 176–183. parliment, t. h., morello, m. j. and mcgorrin, r. j. (1994), Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes, Washington, DC, ACS Symposium Series 543. peker, h., srinivasan, m. p., smith, j. m. and mccoy, b. j. (1992), ‘Caffeine extraction rates from coffee beans with supercritical carbon dioxide’, AIChE J, 38 (5), 761– 770. pourmortazavi, s. m., ghadiri, m. and hajimirsadeghi, s. s. (2005), ‘Supercritical fluid extraction of volatile components from Bunium persicum Boiss. (black cumin) and Mespilus germanica L. (medlar) seeds’, J Food Comp Anal, 18 (5), 439–446. povey, m. and mason, t. j. (1998), Power Ultrasound in Food Processing – The Way Forward, London, Blackie Academic & Professional. quirin, k. w. (2004), ‘SFE: The best technology for high grade essential oils’, Innov Food Technol, 22, 40–43. ramachandra rao, s. and ravishankar, g. a. (2000), ‘Vanilla flavour: production by conventional and biotechnological routes’, J Sci Food Agric, 80, 289–304. rashmi, s. and kannamkumarath, s. s. (2004), ‘Comparison of extraction methods by Soxhlet, sonicator, and microwave in the screening of pesticide residues from solid matrices’, J Anal Chem, 59 (11), 1032–1036. raventos, m., duarte, s. and alarcon, r. (2002), ‘Application and possibilities of supercritical CO2 extraction in food processing industry: an overview’, Food Sci Technol Int, 8 (5), 269–284. riera, e., golas, y., blanco, a., gallego, j. a., blasco, m. and mulet, a. (2004), ‘Mass transfer enhancement in supercritical fluids extraction by means of power ultrasound’, Ultrason Sonochem, 11 (3–4), 241–244. ruiz del castillo, m. l. and herraiz, m. (2003), ‘Ultrasonically assisted solid-phase extraction and GC analysis of filbertone in hazelnut oil’, J Am Oil Chem Soc, 80 (4), 307–310. sanghi, r. and kannamkumarath, s. s. (2004), ‘Comparison of extraction methods by soxhlet, sonicator, and microwave in the screening of pesticide residues from solid matrices’, J Anal Chem, 59 (11), 1032–1036. schneider, y., zahn, s., hofmann, j., wecks, m. and rohm, h. (2006), ‘Acoustic cavitation induced by ultrasonic cutting devices: a preliminary study’, Ultrason Sonochem, 13, 117–120. senorans, f. j., ruiz-rodriguez, a., ibanez, e., tabera, j. and reglero, g. (2003), ‘Isolation of brandy aroma by countercurrent supercritical fluid extraction’, J Supercrit Fluids, 26 (2), 129–135. shen, z., mishra, v., imison, b., palmer, m. and fairclough, r. (2002), ‘Use of adsorbent and supercritical carbon dioxide to concentrate flavor components from orange oil’, J Agric Food Chem, 50, 154–160.

Extraction of flavourings from natural sources

63

sides, a., robards, k. and helliwell, s. (2000), ‘Developments in extraction techniques and their application to analysis of volatiles in foods’, Trends Anal Chem, 19 (5), 322–329. singh, k. p. (1993), ‘Challenges and opportunities in essential oil processing industries’, Res Ind, 38 (2), 83–89. speiser, w. (1993), ‘Enzymes and the fruit juice industry – a love affair?’, Fruit Process, 2, 8–39. stuerga, d. and delmotte, m. (2002), ‘Wave-material interactions, microwave technology and equipment’, in Loupy A, Microwaves in Organic Synthesis, Weinheim, Wiley-VCH, 22–31. szentmihalyi, k., vinkler, p., lakatos, b., illes, v. and then, m. (2002), ‘Rose hip (Rosa canina L.) oil obtained from waste hip seeds by different extraction methods’, Biores Technol, 82 (2), 195–201. toma, m., vinatoru, m., paniwnyk, l. and mason, t. j. (2001), ‘Investigation of effects of ultrasound on vegetal tissues during solvent extraction’, Ultrason Sonochem, 8 (2), 137–142. venskutonis, p. r. (1997), Effect of drying on the volatile constituents of thyme (Thymus vulgaris L.) and sage (Salvia officinalis L.), Food Chem, 59 (2), 219– 227. vinatoru, m. (2001), ‘An overview of the ultrasonically assisted extraction of bioactive principles from herbs’, Ultrason Sonochem, 8 (3), 303–313. vinatoru, m., toma, m., radu, o., filip, p. i., lazurca, d. and mason, t. j. (1997), ‘The use of ultrasound for the extraction of bioactive principles from plant materials’, Ultrason Sonochem, 4 (2), 135–139. vinatoru, m., toma, m. and mason, t. j. (1999), ‘Ultrasonically assisted extraction of bioactive principles from plants and their constituents’, in Mason T J, Advances in Sonochemistry, Mason Stamford, CT, JAI Press Inc., 209–247. wang, z., ding, l., li, t., zhou, x., wang, l., zhang, h., liu, l., li, y., liu, z., wang, h., zeng, h. and he, h. (2006), ‘Improved solvent-free microwave extraction of essential oil from dried Cuminum cyminum L. and Zanthoxylum bungeanum Maxim, J Chromatogr A, 1102 (1–2), 11–17. werkhoff, p., brennecke, s., bretschneider, w. and bertram, h. j. (2002), ‘Modern methods for isolating and quantifying volatile flavor and fragrance compounds’, Food Sci Technol, 115, 139–204. wilkes, j. g., conte, e. d., kim, y., holcomb, m., sutherland, j. b., miller, d. w. (2000), ‘Sample preparation for the analysis of flavors and off-flavors in foods’, J Chromatog A, 880, 3–33. yeo, h. c. h. and shibamoto, t. (1991), ‘Chemical comparison of flavours in microwaved and conventionally heated foods’, Trends Food Sci Technol, 2, 329–332. yonei, y., ohinata, h., yoshida, r., shimizu, y. and yokoyama, c. (1995), ‘Extraction of ginger flavor with liquid or supercritical carbon dioxide’, J Supercrit Fluids, 8, 156–161. zabetakis, i., koulentianos, a., orruño, e. and boyes, i. (2000), ‘The effect of high hydrostatic pressure on strawberry flavour compounds’, Food Chem, 71, 51–55.

4 From fermentation to white biotechnology: how microbial catalysts generate flavours R. G. Berger, Leibniz University of Hannover, Germany

4.1

Introduction

4.1.1 Microbial generation of flavour Volatile flavours are long-range chemical signals. Fruits, vegetables and other plants produce them to attract pollinators, birds or other animals, or to repel microbial and animal invaders trying to feed on them. The recent interest in functional properties of food constituents has shown that many flavour compounds possess additional bioactivities (Berger, 1995): bipolar terpenes of essential oil plants and some carbonyls are antimicrobial, antifungal, anti-inflammatory or even antiviral; some volatile phenols are strong antioxidants; some volatile aldehydes inhibit the production of nitric oxide, a factor contributing to the regulation of blood pressure. Both the signalling properties and other functional properties are usually associated with one distinct form of two or more stereo-forms of the flavour compound (Berger, 1995). The recent example of the (±)-massoilactones, the (R)-isomer exhibiting a butter-coconut odour and the (S)-isomer being weaker and less specific (Watanabe, 2005), adds to a long list of structure–activity relationships of flavours. The biochemical precursors of flavours, mainly fatty and amino acids, isoprene units and phenylpropanoids, are ubiquitous in nature. Microorganisms modify these substrates along efficient and selective enzymatically accelerated pathways, similar to those operating in plant cells which, in the form of intact tissues, are called ‘food’. Still, the perception that microorganisms may deliver ‘edible’ food constituents is causing scepticism among consumers. Reasons may be that microorganisms are more associated with dangerous diseases and food spoilage than with fermented food and vital events during human digestion. Today, the microbial production

How microbial catalysts generate flavours

65

of flavours has developed into one of the most rapidly emerging areas of industrial biotechnology. More than 100 compounds are on the market, or are used for in-house flavour compositions, among them volatile aliphatic and aromatic alcohols (such as fusel alcohols and cinnamyl alcohol), acids and esters (such as 2-phenylethyl acetate and methyl jasmonate), aldehydes (such as hexenals and vanillin), ketones and methyl ketones, a series of 4and 5-alkanolides, and a few terpenoids (such as perillyl alcohol and nootkatone) (Berger, 1995).

4.1.2 Substrates, biocatalysts and naturalness of flavours Although many flavours and fragrances are easily produced through chemical synthesis, natural plant sources continue to provide most of the materials used in the food industry in the EU and the United States. Often natural flavours are more efficient and useful as building blocks. Microbial production of flavours benefits from the high selectivity and reaction rate of many microbial enzymes, as well as from the possibility of conducting multi-step syntheses in intact cells. The mild reaction conditions and the possibility of using renewable substrates are other, important advantages. In contrast to conventional agricultural production of plant material, none of these bioprocesses requires ecologically problematic fertilisers or pesticides, and they are not adversely affected by climate, weather, soil, trade barriers, sociopolitical instability and the like. The label ‘natural’, which is assigned if the precursor substrates used are of natural origins, carries positive connotations. Most consumers equate ‘natural’ with ‘unadulterated’ and ‘fresh’, and prefer such a food product over a competing product. As there is simply no other labelling option for food manufacturers, apart from the flavour ingredient used, they have to fall back on ‘natural’ flavours, if they want the magic word to appear on the tag. The high demand for ‘natural’ flavours is reflected by market prices up to 11000 per kg and higher, which is asked for popular flavours such as vanillin, raspberry ketone (II), 2,6-(E,Z)-nonadienal (cucumber; V) or nootkatone (grapefruit; I) (Fig. 4.1). This explains the plethora of research efforts in both the flavour industry and in academia. The world’s largest flavour producer, International Flavors and Fragrances, has recently published some fine examples of bioprocesses from its own labs (Kim & Eckert, 2003), and the CEO of Symrise, Germany’s leading flavour house, has stated that they ‘plan to grow mainly in the area of biotechnology’ (Hannoversche Allgemeine Zeitung, 2006). All kinds of microorganisms from small prokaryotic bacilli to complex eukaryotic fungi have been used for flavour modification and generation. Organisms generally recognised as safe (GRAS) (Schallmey et al., 2004) as classified by the US Food and Drug Administration are preferred, but numerous other species and even pathogens, such as Enterobacter or Serratia, have been patented for food flavour production. This sounds

66

Modifying flavour in food O

O HO O (II): Raspberry ketone (raspberry) (I): (+)-Nootkatone (grapefruit) O

(III): Perillene (rose)

O (V): (2E,6Z)-Nonadienal (cucumber)

O

HO

O (IV): Methyl jasmonate (lemon)

Fig. 4.1

O

O (VI): α-Sinensal (orange)

O

(VII): Furaneol (strawberry jam)

Some of the most sought-after natural flavours.

peculiar at first glance, but volatile flavours generally are small and hydrophobic molecules, hence, a large number of extraction, distillation and adsorption methods can be applied to separate a perfectly safe product from the by-products of fermentation. Because of the universal biochemical pathways, several microbial species can produce a certain flavour chemical. Roughly more than 20 different species were shown, for example, to produce 4-decanolide. It does matter, however, which strain is chosen, whether in concerted biotransformations or in traditional fermentations: large differences in the flavour profiles were recurrently observed when different strains were used (Estevez et al., 2004; Leclercq-Perlat et al., 2004; Paraggio & Fiore, 2004). As a result, the development of a successful bioprocess is inseparably linked to an extended screening of strains, substrates and nutrient media. This first, tedious phase of manual work often deters industrial researchers, but opportunities missed at this stage cannot be compensated by any means at a later stage of process development. Microorganisms can be used to carry out single-step reactions (biotransformation), sequential reactions (bioconversion), or de novo syntheses; they can likewise produce (Coghe et al., 2004) or eliminate (Blagden & Gilliland, 2005; Hambraeus & Nyberg, 2005) off-flavours. Indirect applications of microbial catalysts, not detailed in this chapter, are the hydrolytic disruption of polysaccharide walls in waste fruit or in barks, seeds and rhizomes (Galkina et al., 2004) for increasing the extraction yields and the treatment of plant tissues or extracts with glycosidases to liberate flavours from acetal precursors (Sarry & Gunata, 2004). The microbial formation of flavours will be presented, according to increasing levels of biochemical complexity and biotechnological effort, ranging from a single precursor to complex food substrates. Molecular engineering will be discussed as a tool for making genes related to flavour formation function on a larger technical scale. Finally, the biotechnological achievement of flavour production, the concept of ‘white biotechnology’

How microbial catalysts generate flavours

67

and future prospects in the field will be outlined using some examples from the most recent literature.

4.2

Flavour formation along known pathways

4.2.1 Isoprenoids Essential oils of higher plants contain numerous mono- and sesquiterpene hydrocarbons and an even larger number of oxyfunctionalised derivatives. The hydrocarbons dominate quantitatively, but are prone to autoxidation, resulting in the formation of turbidity and off-flavours and are therefore distilled off. Many of the oxyfunctionalised minor and trace constituents are valuable flavours, but the low concentrations often make isolation uneconomic. As early as the 1950s, this situation stimulated attempts to microbially transform or convert the waste hydrocarbons to more valuable flavour compounds. The common procedure starts with one of the readily available limonenes or pinenes as the precursor substrate, a local microbial isolate from a terpene-rich plant material, such as citrus peel, and a shake flask incubation in a standard medium (Kaspera et al., 2005b; Divyashree et al., 2006). Pseudomonas are the preferred bacterial strains, while fungi of the genus Aspergillus appear to be the eukaryotic strains of choice. Most yeasts, and particularly the common wine yeasts, are sensitive to higher concentrations of these monoterpenes and are thought to be useful only for esterifications to yield compounds such as geranyl and citronellyl acetate (King & Dickinson, 2003). Recently it was shown that microaerobic conditions combined with higher levels of nitrogen in the medium favoured the de novo synthesis of monoterpeneols by a strain of Saccharomyces cerevisiae (Carrau et al., 2005). It was speculated that the actual concentrations of linalool, a monoterpenol with a low odour threshold in water, could be sensorially relevant. Oxygen insertion at the menthadiene skeleton regularly occurs at allylic positions resulting in the predominant formation of verbenol, verbenones, carveols, carvones and related compounds (Nazhat-ul-Ainn et al., 2002; de Carvalho & da Fonseca, 2005). In rare cases, an attack on nonactivated carbons was observed (Fig. 4.2) (Park et al., 2003; Krings et al., 2005). Typical problems encountered with this approach are the formation of mixtures of volatiles (and non-volatiles), cytotoxicity of both precursor and products (some flavours are antimicrobials!), and loss of both through the gas phase. Several options exist to overcome these drawbacks: screening for better strains, fed-batch addition of substrate, intermittent or continuous removal of product, use of tolerant or resting cells, immobilisation of the biocatalyst and genetic manipulation. An intriguing idea to overcome the poor water solubility of the substrate is to use a terpene alcohol as a precursor (Hrdlicka et al., 2004), but the primary (di-oxo) product is usually too polar to still possess flavour

68

Modifying flavour in food 1

(VIII)

O

OH

9

(X)

(IX)

O

OH (XI)

(XII)

Fig. 4.2 Unusual non-allylic transformation of γ-terpineol (VIII) by Stemphylium botryosum yielded p-mentha-1,4-dien-9-ol (IX), p-cymene-9-ol (XI), and the corresponding aldehydes (X,XII) (Krings et al., 2005).

properties. If, however, one of the oxo-functions is eliminated in a second step, a new flavour compound (here α-sinensal (VI) from nerolidol) can result. Another way to circumvent cytotoxicity, the use of isolated enzymes, was patented for the laccase catalysed conversion of valencene to nootkatone (Hitchman et al., 2005). This reaction is perhaps the ‘holy grail’ of flavour biotechnology, as the target compound is one of the most soughtafter character impact compounds (Fig. 4.1) (Kaspera et al., 2005a). Many papers and patents describing the reaction have been published, but attempts to reproduce the results have often failed. Even green algae (Chlorella fusca and Ch. pyrenoidosa) conduct this transformation with excellent yields (Furusawa et al., 2005), but the enzyme(s) behind its (their) regulation remain a mystery. The source of the relatively cheap ‘natural’ nootkatone that entered the market recently is unknown. The detection of the existence of two independent pathways to isopentenyl diphosphase was a great surprise for the biochemical community. In plants it appears that monoterpenes are generated through the novel plastidial DOXP/MEP (1-deoxy-d-xylulose/2-C-methyl-d-erythritol-4phosphate) pathway, while sesquiterpenes are formed along the classical cytosolic mevalonate and the novel pathway (Hampel et al., 2004). Both pathways show different discrimination of 13C. Thorough analyses of linalool enantiomers in strawberry have shown that the (R)-enantiomer was formed through the DOXP/MEP pathway, whereas the opposite enantioform was only found labelled after administration of [5,5-2H2]-mevalonic acid lactone (Hampel et al., 2006). These data will help in mapping quantitative trait loci (QTL) responsible for flavour properties of edible plants

How microbial catalysts generate flavours

69

(Sevini et al., 2004). A number of geranyl and farnesyl diphosphase cyclases have been identified on the cDNA level and have been cloned (Guterman et al., 2002; El Tamer et al., 2003; Sharon-Asa et al., 2003; Pechous & Whitaker, 2004, 2005), and first patents in the field have appeared (Aharoni et al., 2004, 2002; Green et al., 2004). Recombinant over-expression in homologous and heterologous plant hosts, for example in tobacco or in Arabidopsis, has not always been successful, but transcript analysis or antibody testing has sometimes proven the principle. The expression of valencene cyclase in Citrus was shown, among other factors, to be regulated by ethene, although citrus fruits are non-climacteric (Sharon-Asa et al., 2003). This result calls for a closer look at the various stages of fruit ripening and the role of ethene. While still little is known on the genetic basis of the introduction of oxygen in essential oil plants, some genes coding for redox enzymes are under investigation (Davis et al., 2005). It is believed that sets of genes can be transferred to microbial hosts such as Escherichia coli (Keasling et al., 2004). A metabolically engineered E. coli strain could combine the rapid growth and straightforward bioprocessing of a prokaryont with the complex flavour-producing capabilities of a higher plant. First examples of a functional over-expression of sesquiterpene cyclases (Keasling et al., 2004) and polyketide complexes (Wenzel & Mueller, 2005) have demonstrated the feasibility of this idea. Some members of the C-13 norisoprenoid family (ionones, damasc(en)ones) belong to the most potent flavours on earth. Several lines of evidence (structural similarities, co-occurrence in non-heated fruit tissues, correlation of pigmentation and pattern of volatiles; Zhou et al., 2004; Lewinsohn et al., 2005a, b) suggested their formation through an oxidative cleavage of the 9,10 double bond of tetraterpenoid precursor molecules, the carotenoids (Fig. 4.3). However, the existence of competing chemical degradation pathways (autoxidation, photo-oxidation, thermal degradation and combinations of the pathways; Nonier et al., 2004; Baldermann et al., 2005) complicated the assessment of an enzymatic contribution. The final proof was missing until the identification of the maize enzyme VP14 (Schwartz et al., 1997). VP14 cleaves 9-(Z)-epoxy carotenoids specifically at the 11,12 double bond. Today, a set of regiospecific carotenoid degrading enzymes from plant and animal sources has been well characterised, for example from tomato (Simkin et al., 2004a). A circadian regulation of a carotenoid dioxygenase was identified in petunia flowers, and the lightinduced expression of the coding gene was followed on the transcription level (Simkin et al., 2004b). Biotic carotenoid degradation by microorganisms, however, is less explored, although some 100 million tonnes of carotenoids are synthesised annually in nature and obviously disposed subsequently. Mixed cultures of Trichosporon asahii and Paenibacillus amylolyticus degraded lutein extracted from marigold flowers (RodriguezBustamante et al., 2005). The presence of both species was required for the formation of the norisoprenoid flavours, with 7,8-dihydro-β-ionol as the

70

Modifying flavour in food

1

1

1

O

O O

O (XIV): β-Cyclocitral

(XIII): Isophorone

9

1

(XV): Dihydroactinidiolid

(XVI)

(HO)

O

O 1

9

(XVII): β-Ionone

1

(XVIII): β-Damascenone

1

(XIX): (E,E)-Megastigma-4,6,8-triene

Fig. 4.3 β,β-carotene and other carotenoids are precursors of norisoprenoid flavours.

main compound. The same researchers have found another community composed of a Geotrichum species capable of cleaving lutein, while a Bacillus species further converted the primary product, β-ionone, to the above ionol and the corresponding ketone (Maldonado-Robledo et al., 2003). An extracellular, so-called ‘versatile’ peroxidase of the edible fungus Pleurotus eryngii (previously erroneously classified as Lepista irina) degraded β, β-carotene to β-ionone, β-cyclocitral, dihydroactinidiolide and 2-hydroxy-2,6,6-trimethylcyclohexanone (Zorn et al., 2003a). β-Apo-10′carotenal was identified as a non-volatile degradation product using APcI+LC-MS (atmospheric pressure chemical ionisation (positive mode)–liquid chromatography–mass spectrometry). A similar enzyme, also capable of asymmetric carotenoid cleavage, was isolated from Marasmius scorodonius (Zorn et al., 2003b). Versatile peroxidases are a novel class of peroxidases which combine the catalytic properties of lignin peroxidase and manganese peroxidase. Basidiomycetes are obviously forming them in a response to their versatile chemical environment. As saprophytic organisms they depend on an exo-proteome with a broad substrate specificity to survive in their natural habitat: they live and feed on wood (Zorn et al., 2005). The first specific carotenoid cleaving enzymes, however, probably coevolved with the first phototrophic organisms, the cyanobacteria (blue algae), because carotenoids are accessory pigments of the photosynthetic

How microbial catalysts generate flavours O

O

Carboxylation

O

O

HO

O

Hydrogenation

O

(XX)

O

HO

O

HO SCoA

(XXV)

OH

Dehydratisation

O

CoA-transfer

O (XXII)

(XXI)

O

OH

HO

OH

OH

71

(XXIV)

Hydrogenation

O O

HO

OH

(XXIII)

OH

Isomerisation (B12)

O HO

O

Decarboxylation SCoA

(XXVI)

Fig. 4.4

O

O

CoA-re-transfer SCoA

OH

(XXVII)

(XXVIII)

The methyl malonyl-CoA pathway of Propionibacterium starts with lactate oxidation to pyruvate.

apparatus. The occurrence of several norisoprenoids in freshwater contaminated by cyanobacteria suggests that similar cleavage enzymes must be present (Hockelmann and Jüttner, 2005). As in lipoxygenase-initiated oxidation reactions, some of the volatiles are generated after disintegration of the cells, perhaps as allelopathic signals. Aquatic nematodes which feed on cyanobacteria perceive these volatiles and respond to single or multicomponent stimuli. As ultra-traces of some norisoprenoids are also perceived by humans, the abundant occurrence of cyanobacteria in eutrophic waters (‘water bloom’) regularly evokes consumers complaints.

4.2.2 Lipid metabolites, esters and alcohol acyltransferases There are several pathways leading to short-chain carboxylic acids: ester hydrolysis (E.C. 3.1.1.1 and 3.1.1.11), lipolysis (E.C. 3.1.1.3 and phospholipases) followed by β-oxidation, oxidation of fermentation alcohols or of Strecker aldehydes, or carboxylation of pyruvate through methylmalonylCoA to eventually yield propanoate. Lipases and the β-oxidation multienzyme complex are ubiquitous, but the oxidation of methanol, ethanol, other alcohols or short-chain aliphatic aldehydes is a pronounced trait of strains of Acetobacter, Gluconobacter, lactococci and lactobacilli (Ganesan et al., 2004; Tungjaroenchai et al., 2004). The methyl malonyl-CoA pathway operates in coryneform bacteria of the genus Propionibacterium and contributes much to the typical sensory character of Swiss cheese (Fig. 4.4)

72

Modifying flavour in food

(Thierry et al., 2005). The same pathway operates in the reverse direction to degrade branched chain skeletons as in l-valine, l-isoleucine and oddchain fatty acids. Lipoxygenases (E.C. 1.13.11.12) catalyse the regio- and stereoselective introduction of molecular oxygen into unsaturated fatty acids. In the microbial world, the occurrence of active lipoxygenases is typical of many fungi. Strains of the genera Aspergillus (Hall et al., 2004) and Penicillium (Husson et al., 2004; Hall et al., 2005) were the focus of interest. The main products were 10-hydroperoxides with (S)-oriented chiral carbon. Cleavage of these hydroperoxides yields 1-octen-3-(R)-ol, (Z)-1,5-octadien-3-(R)-ol, and corresponding ketones (plus a non-volatile oxo-acid), compounds that determine the typical ‘mushroom’ odour of edible fungi and the surfaces of Camembert cheese and mould-ripened sausages. Supplementation of submerged cultures of Penicillium camemberti with soybean oil enhanced the concentration of 1-octen-3-ol (Husson et al., 2004). Much better investigated are the isoforms of lipoxygenases in plants. The antisense technique was used to show the contribution of TomloxC to the flavour formation of tomato fruit. The chloroplastidal association of the isoform was confirmed using a TomloxC-green fluorescent protein fusion product and imaging of recombinant tobacco leaves (Chen et al., 2004). Esters of short-chain carboxylic acids and alcohols impart a large variety of delicious fruity flavour notes and are not only important in fruit and all sweet flavours, but also in alcoholic beverages and dairy products (Liu et al., 2004; Stewart, 2005). Direct ester condensation is unlikely to occur in most foods, but may play a limited role in low water activity systems stored over extended periods of time, such as Parmesan cheese (Holland et al., 2005). In the field of industrial flavour biotechnology, however, the same reaction (reverse hydrolysis) using lipases isolated from yeasts, such as Candida rugosa (Bezbradica et al., 2006) or Candida antarctica (Larios et al., 2004) in organic solvents, has become one of the biggest success stories. Numerous esters including those possessing aryl or isoprenoid moieties are amenable to reverse hydrolysis through a simple, one-pot reaction. Sometimes, the alkyl moiety may serve as the solvent; removal of the reaction water from the equilibrium is mandatory and can be achieved by condensation, molecular sieve, distillation or a membrane technique. If both acyl and alkyl moieties were derived from natural sources, the ester thus obtained is classified as ‘natural’. In line with this mechanism, a sake yeast was constructed by transferring the gene of an isoamyl acetate hydrolysing enzyme from Saccharomyces cerevisiae (Fukuda, 2004). Most esters in food, however, owe their formation to thermodynamically preferred alcohol acetyltransfer reactions (AAT, E.C.2.3.1.84) (Fig. 4.5). Saccharomyces cerevisiae contains at least three alcohol acetyltransferases (Verstrepen et al., 2003b). Deletion and over-expression experiments have established functional correlations (Van Laere et al., 2005). Factors found to affect the enzyme activity in the living cell are oxygen partial pressure,

How microbial catalysts generate flavours O

R'

O

OH

+ R

SCoA (XXIX)

AAT

73

R

OR'

HSCoA

(XXX)

Fig. 4.5 Alcohol acetyltransferases catalyse the microbial formation of flavour esters.

presence of unsaturated fatty acids, and glucose concentration (Verstrepen et al., 2003a; Fujiwara, 2004). Low substrate affinity of a purified AAT indicated that the formation of volatile esters may be just a side-aspect of an unknown physiological role (Van Laere et al., 2005). As the enzyme is located in lipid particles, a specific function in the cell’s lipid or sterol metabolism was suggested (Verstrepen et al., 2004b). To counteract the public concern about genetically engineered organisms, a ‘self-cloning’ sake yeast was constructed (Hirosawa et al., 2004). The yeast contained a constitutive promoter from the glyceraldehyde-3-phosphate dehydrogenase gene fused to the transferase gene, and produced increased concentrations of isoamyl acetate. The authors claimed that this was the first example of an expression-controlled industrial yeast. Heterologous expression of AAT genes from yeast and strawberry in E. coli was also successful, as proven again by increased levels of isoamyl acetate (Vadali et al., 2004; Horton and Bennett, 2006). Such strains will neither replace reverse lipolysis nor brewer’s yeast in traditional fermentations, because they depend on supplementation of exogenous substrate. They present, however, valuable genetic models to estimate the specificity and efficiency of the various isoenzymes, and to study means of enhancing CoA or acetyl-CoA levels in the cell. AATs of melon (Yahyaoui et al., 2002), strawberry and banana (Beekwilder et al., 2004), apple (Souleyre et al., 2005) and grape (Wang & De Luca, 2005) have now been characterised on the protein or nucleic acid level. These are all candidate genes for studies in prokaryotic hosts. Availability of the alkyl precursor moiety was limiting in the E. coli system (Beekwilder et al., 2004), as well as in ripe apple fruit (Echeverria et al., 2004). This agrees with earlier experiments which showed a strong increase of ethyl ester concentrations in apples stored in an atmosphere containing ethanol vapours (Berger & Drawert, 1984). The common β-oxidation of fatty acids proceeds, as the name suggests, via 3-hydroxy acyl-CoA intermediates. If hydroxylation of the acyl chain occurs in the γ- or δ-position instead, CoA-esters are formed which easily cyclise upon hydrolysis of the thioester bond to form five or six membered lactone rings. The resulting alkanolides are potent flavours imparting fruity, nutty, buttery or coconut-like notes. As with the open-chain esters, decreasing water activity, as occurs for example during ripening of Gouda cheese, favours lactone formation (Alewijn et al., 2005). It has not been unambiguously determined if the hydroxy function is derived from a microbial

74

Modifying flavour in food

activity or from an autoxidative side reaction. The primrose pathway to high yields of lactones has turned out to be due to the presence of large amounts of exogenous γ- or δ-hydroxy fatty acid. Castor oil and Massoia oil contain the required precursors. Geotrichum fragrans and other Geotrichum species accumulated γ-decalactone after enzymatically hydrolysed Castor oil was supplemented (Neto et al., 2004). The alkane-assimilating yeast Yarrowia lipolytica is particularly suited for the uptake, transport and oxidation of hydrophobic moieties and produced significant amounts of γ-decalactone from methyl ricinoleate (Fickers et al., 2005). To increase the actual precursor concentration in the aqueous medium, surfactants were added (Aguedo et al., 2004). It turned out, however, that some of the surfactants tested were toxic themselves, and the availability of methyl ricinoleate was not limiting the conversion rate, as was shown by zeta potential measurements (Aguedo et al., 2005). A Yarrowia lipolytica that lacked one of the five acyl-CoA oxidase isoforms was constructed and showed 10 times increased lactone levels (Groguenin et al., 2004). This finding not only provided evidence for the involvement of this enzyme in lactone degradation, but also revealed that microbial flavour compounds, which have been target compounds in the past, have often been erroneously considered as end products of the metabolism. Labelling studies, performed using Sporobolomyces odorus, another lactone-producing yeast, demonstrated an enzymatic Bayer–Villiger oxidation and provided a sound experimental basis on which to develop a bioprocess (Garbe & Tressl, 2006). If the rate of degradation is in the same order of magnitude as the rate of formation, accumulation of product will never reach anything like an attractive level. Blocking of the degrading enzymes by metabolic engineering is one option; selective in situ recovery of the target product is another one (Dufosse et al., 2002).

4.2.3 Phenolics After the structural elucidation of benzaldehyde, coumarin, vanillin and other aroma compounds was accomplished towards the end of the nineteenth century, the early natural product chemists believed that flavour properties were linked to the presence of an ‘aromatic’ ring. Today we know that volatile flavours may carry all kinds of functional groups, and the above benzene derivatives and many others, such as phenylethyl and cinnamyl derivatives, have been described from many microbial cultures. The most popular phenolic flavour of the world, vanillin, has attracted a lot of research. Genuine vanillin is obtained from Vanilla planifolia Andrews by solvent extraction (Stockfleth & Fiebig, 2004). The epiphytic orchid is difficult to grow and to pollinate (Havkin-Frenkel et al., 2005), and the pods need a multi-step treatment called curing (hot water scalding, sweating, drying, conditioning; Havkin-Frenkel et al., 2004) so that endogenous glycoside precursors can make contact with glycosidases located in the outer fruit wall region to achieve the full flavour (Setyaningsih et al., 2005). Isolation

How microbial catalysts generate flavours HO

O

HO

O

HO

OCH 3 OH (XXXI)

Fig. 4.6

HO O

OCH3 OH (XXXII)

O – Acetate

O

OCH3 OH (XXXIII)

75

OCH3 OH (XXXIV)

Ferulic acid is the most common precursor of biotech vanillin.

and extraction may be improved by treating the pods with cellulases (Ponzone et al., 2004). All together, these obstacles limit the supply of vanillin ex Vanilla, resulting in small market and a high price (up to $4000 per kg in years of poor harvest). Biotechnology offers alternative solutions. More than 20 bioprocesses have been described to produce vanillin, from the use of isolated enzymes, to complex plant cell cultures as catalysts. Recent work has concentrated on the bioconversion of phenylpropanoid precursors, such as ferulic, pcoumaric or cinnamic acid or coniferyl aldehyde which contain a pre-formed vanillin structure and are readily available from agricultural wastes, such as rice or oat bran (Fig. 4.6). Haematococcus pluvialis (Tripathi et al., 2002), E. coli (Torre et al., 2004) or Streptomyces halstedii and many other Actinomycetes (Brunati et al., 2004) formed vanillin and related volatiles in submerged culture and also when immobilised and operated in a fixed-bed column (Torre et al., 2004). Transient accumulation and low peak concentrations indicated a further conversion of the target compound, an observation also made with suspension cultures of Capsicum annuum L. fed with ferulic acid (Kang et al., 2005). Some of the phenolic intermediates are prone to one or two electron oxidation reactions, yielding stable radicals and phenolic acids. They are funnelled to lignin biosynthesis and reappear in plant cell walls in bound and free forms (Suresh et al., 2003). One may recall that alkaline hydrolysis of corn hulls started chemical vanillin production and industrial flavour production in Germany. The construction of a Pseudomonas strain, which has vanillin oxidase activity blocked, results in a yield of several grams of vanillin per litre and has solved the problem of catabolism (Priefert et al., 2001). Chemical losses due to reactions of the aldehyde group with constituents of the nutrient medium or with itself may be reduced by application of an adsorbent or another appropriate in situ technique.

4.2.4 Amino acids and derivatives Some amino acids are important precursors in microbial flavour syntheses, especially in lactic acid bacteria (Ardoe, 2006). Acetoin and diacetyl in

76

Modifying flavour in food L-Ile

O

C

or L-Leu COOH

L-Glu

O

C

COOH

O

C

COOH

or COOH

(XXXVI)

(XXXVII)

(XXXV) – CO 2 O

O

C

C

or (XXXVIII)

Fig. 4.7

(XXXIX)

Typical fermentation flavours from α-ketoglutarate assisted transaminations.

lactobacilli are derived from aspartate (Kieronczyk et al., 2004), while the synthesis of the branched-chain amino acids (Val, Leu, Ileu) yields 2methylpropanol, n-butanal, 2- and 3-methylbutanal and related redox derivatives as side products. The analogous volatiles derived from the aromatic (Phe, Tyr, Trp) and the sulphur-containing amino acids (Cys, Met) are also contributors to many plant and fermentation flavours. The addition of blends of amino acids to alcoholic or lactic fermentations must therefore lead to intensified or modified flavours (Capelle, 2004; Herranz et al., 2005). Enzymatic hydrolysis of endogenous proteins by added endo- or exopeptidases is an obvious alternative (Blinkovsky et al., 2004; Jin et al., 2004; Kania & Stasinska, 2004). More recently, peptidases and transpeptidases with improved substrate specificity were discovered for treating dough (Edens & Hille, 2005), dry sausage (Benito et al., 2005a), cocoa nibs (Sousa de Brito et al., 2004), or to elevate the levels of the umami compounds glutamic and aspartic acid (Sakamoto et al., 2004) and theanine (Suzuki & Kumagai, 2004). Many of the novel peptidases are from filamentous fungi, typically from Aspergillus strains such as A. oryzae (‘Flavorzyme®’ from Novo Nordisk Bioindustrials). Solid-state fermentation appears to be the most efficient way for their production (Nampoothiri et al., 2005). The metabolic driving force in lactic fermentations appears to be regeneration of α-ketoglutarate (Fig. 4.7) (Helinck et al., 2004; Tjener et al., 2004a). As a general amino group acceptor, the keto acid is key to the formation of other keto acids from amino acids. These are then either converted to acyl-CoA intermediates or decarboxylated and the resulting aldehydes are reduced or oxidised to typical fermentation volatiles. A diverse microflora of the genera Lactococcus, Lactobacillus, Streptococcus, Leuconostoc, Bifidobacterium, Propionibacterium, Brevibacterium, Corynebacterium and Arthrobacter is found in fermented dairy products,

How microbial catalysts generate flavours

77

but a broader screening has indicated that only a few of these strains possessed high decarboxylase activity (B. A. Smit et al., 2004a). Pronounced metabolic differences were found between various strains of lactobacilli (Thage et al., 2005), between non-starter strains (Williams et al., 2004a) and between strains of the smear population (Williams et al., 2004b). Genetic work in this area has been initiated, for example, the cloning of a decarboxylase gene of Lactococcus lactis (B. A. Smit et al., 2005). Volatile breakdown products of amino acids also control the overall sensory quality of dry sausages (Tjener et al., 2004a). Pediococcus pentosaceus and Staphylococcus carnosus or S. xylosus dominate the microflora. High inoculation levels of S. carnosus stimulated the formation of the methyl-branched acids and aldehydes, phenylacetaldehyde, dimethylsulphide, and dimethyltrisulphide (Tjener et al., 2004b). Studies using isotopelabelled leucine showed that α-ketoglutarate was a less efficient amino group acceptor than α-keto-β-methylvaleric acid and α-ketoisovaleric acid (Beck et al., 2004), and that the concentration of 3-methylbutanoic acid was lowered and α-hydroxyisocaproic acid was increased by adding nitrate or nitrite (Olesen et al., 2004). It was concluded that the sensorially important compounds are intermediates on a route leading to branched-chain fatty acids, as they constitute the major fraction in cell wall lipids.

4.2.5 Thio-compounds The large sulphur atom is easily polarised and interacts rapidly with electrophilic compounds and Lewis acids. This explains why trace concentrations of many volatile thio-compounds are sufficient to change receptor protein conformation and to trigger the cascade of sensory perception. The only immediate metabolic sources of sulphur are l-cysteine and l-methionine. Brevibacteria are often cited for their pronounced ability to form volatile sulphur compounds, as they possess a lyase converting l-methionine to methanethiol, α-ketobutanoate and ammonia (Amarita et al., 2004). The lyase gene of B. linens was recently cloned (Amarita et al., 2004), and transformation of the same species using common plasmids from lactic acid bacteria was demonstrated (Nardi et al., 2005). The degradation of methionine by yeasts usually starts with transamination yielding 4-methylthio2-oxo-butanoic acid (Bondar et al., 2005). Yeasts can also transform cysteine–carbonyl conjugates to release volatile thiols (Fig. 4.8). This reaction was applied for the concerted production of furfurylthiol, a coffee flavour compound (Huynh-Ba et al., 2003), but also plays a role in the formation of 4-mercapto-4-methylpentan-2-one and 2-methyltetrahydrothiophen-3-one in wine (Howell et al., 2005). Similar thio-ether bond cleaving activities of Eubacterium limosum and E. coli were used for the enantioselective liberation of 3-mercaptohexanol (Wakabayashi et al., 2002). Wild yeasts isolated from the surface of truffles (Buzzini et al., 2005a) and basidiomycetous yeasts (Buzzini et al., 2005b), when grown on

78

Modifying flavour in food

O

+

COOH

HS

O

RT

H N O

NH2

(XL)

(XLI) SH

+

b-Lyase

COOH

S (XLII)

NH3 + Pyruvate

O (XLIII)

Fig. 4.8

Thiazolidine ‘conjugates’ can be enzymatically converted to potent thiol flavours.

methionine as the sole nitrogen source, produced numerous volatile thiocompounds. As soon as the respective genes have been identified, cloning and overexpression into more convenient hosts are expected.

4.3

Flavours from complex substrates

4.3.1 Fermentations dominated by lactic acid bacteria The starting materials of traditional food fermentations, such as milk, butter, cereals, soybean, meat, fruit musts and vegetables (cabbage, olives), contain all of the above substrates for microbial flavour modification pathways. In the presence of complex isoprenoids, lipids, phospholipids and peptides, a broad spectrum of volatiles is formed, even by simple prokaryotics. Some recent highlights in this field of flavour modification are presented. The secondary or malolactic fermentation of (red) wine by Oenococcus or Lactobacillus strains not only converts malate to the less acidic lactate (plus CO2), but is also known to alter the composition of amino acids and volatile flavours (Pozo-Bayon et al., 2005). The known sensory changes have not been analytically substantiated for a long time. Recently, it was found that the concentration of ethyl esters from butanoate to octanoate and 3-methylbutyl acetate increased (Ugliano & Moio, 2005). Glycosidases of lactobacilli were able to hydrolyse flavour precursors of wine, a trait that was formerly supposed to be more associated with fungi (Grimaldi et al., 2005). Analogous findings were reported for malt whisky fermentations. Lactic acid bacteria formed esters and led to increased concentrations of flavours, such as damascenone (Priest & Cachat, 2004), a potent norisoprenoid compound known to occur in the glycoside fraction of many fruits. Microbial decarboxylation of coumaric and ferulic acid to 4-vinylphenol and 4-vinylguaiacol was beneficial to the flavour of the distilled beverage. The contribution of Lactobacillus frucivorans and Lactobacillus sanfranciscensis to the flavour modification of sourdough has been established in a thorough study (Czerny et al., 2005). Among the key volatiles were 2,4-(E,E)-decadienal, (2E)-nonenal, 2- and 3-methylbutanal, ethyl butanoate and acetic acid.

How microbial catalysts generate flavours

79

These data will support future screenings for other lactobacilli with a stronger flavour formation potential. The important transformation of arginine to ornithine, the precursor of the key volatile acetylpyrroline in bread, has also been investigated (Ganzle et al., 2005).

4.3.2 Fermentations dominated by yeasts Hundreds of yeast strains have been selected and deposited in culture collections over the last decades for the optimum fermentation of beer wort and grape must. Still, a need for strains with novel properties appears to exist, for example, to release glycosidically bound flavours (Alexandre et al., 2004) or to produce icewine (Erasmus et al., 2004). If a satisfactory result cannot be achieved with a Saccharomyces strain, another yeast may be applied. Pichia membranaefaciens, long regarded as one of the undesirable ‘wild’ yeasts, converted Chardonnay and Pinot noir musts into products with a sparkling wine flavour (Mamede et al., 2004), and other wild strains gave Cachaças (a sugarcane spirit) with acceptable flavour (Oliveira et al., 2005). Even a member of the subdivision Basidiomycotina, the edible fungus Pleurotus ostreatus, was successfully used in rice fermentations to produce a ‘high flavour’ alcoholic beverage (Takizawa et al., 2005). As in alcoholic fermentations, wild yeast are regarded as undesirable in cheeses, but artisanal products in particular are rich in Debaromyces, Kluyveromyces, Geotrichum and Candida strains (Fadda et al., 2004). The flavours generated by Debaromyces have been evaluated in more detail (Flores et al., 2004; Andrade et al., 2005). Ethyl esters and the above discussed volatiles from branched-chain amino acids were most prominent. Enhanced proteolysis, as was achieved in ham with purified fungal peptidases (Benito et al., 2004, 2005a), was also the main motive for suggesting the addition of a nontoxigenic strain of Penicillium chrysogenum to dry-cured sausage (Benito et al., 2005b), while the role of Penicillia occurring as the surface flora on Salami-type sausages is obviously limited to the protection of meat lipids from autoxidation (Hierro et al., 2005).

4.3.3 Sequential and combined fermentations Advanced proteolysis of (milk) proteins bears the risk of generating bitter peptides. Microbial communities may remedy the situation, as they provide an integrative set of complementing peptidolytic activities. Lactic fermentations were improved by Leuconostoc strains isolated from artisanal cheese (Sanchez et al., 2005), by wild heterofermentative lactobacilli (Ortigosa et al., 2005), or by a combination of three lactic acid bacteria (Garde et al., 2006). Increased levels of free amino acids and decreased bitterness due to lower concentrations of hydrophobic peptides were found. The close association of some yeasts and lactic acid bacteria in nature was imitated in dairy fermentations using wild yeasts combined with lactic acid bacteria and

80

Modifying flavour in food

Brevibacterium linens (Arfi et al., 2004), or combinations of yeasts with starter cultures (De Wit et al., 2005). The effect of the yeast may be indirect, by raising the pH value of the fermentation, or direct, for example, by transforming methanethiol accumulated by the bacterium to S-methylthioacetate (Arfi et al., 2005). When curd was sequentially treated with proteolytic and lipolytic enzymes, an improved enzyme-modified cheese flavour was obtained (Kilcawley et al., 2006). A synthetic combination of five species comprising yeasts, lactic acid and acetic acid bacteria mimicked the regular fermentation of cocoa and resulted in a good quality chocolate (Schwan & Wheals, 2004). Amylomyces rouxii, a mould and S. cerevisiae in the form of starter granules, worked together in rice wine brewing (Dung et al., 2005). The European counterpart to this idea was the use of Pichia fermentans in a first must fermentation to produce flavours, and of S. cerevisiae in a second step to increase ethanol levels to the required 12–13%vol (ClementeJimenez et al., 2005).

4.3.4 Progress through molecular engineering Saccharomyces cerevisiae has developed into the eukaryotic workhorse of the molecular biologists. Now that the entire genome is known, concerted improvements of bioprocesses are possible. Common malo-lactic fermentation, for example, was replaced by a transfer of the malo-ethanolic genes from Schizosaccharomyces pombe to S. cerevisiae. To this end, malate permease and malic enzyme genes were put under regulation elements from the 3-phosphoglycerate kinase and heterologously expressed in S. cerevisiae. The marker gene was removed so that the final recombinant construct contained yeast genes only. An efficient malo-ethanolic pathway without negative side effects was hence achieved in a GRAS organism (Volschenk et al., 2004). The genes of amino acid degradation are at the centre of interest. It was observed that lowering the standard fermentation pH of 5 to 3 was followed by maximum formation of 3-methylbutanol (Schoondermark-Stolk et al., 2006). A genome-wide expression analysis using a cDNA microarray showed that more than 700 genes were induced or repressed by the pH shift, among them 7 genes related to the 3-methylbutanol pathway. However, the constitutive expression of the only decarboxylase gene which was transcriptionally up-regulated when leucine, methionine or phenylalanine was the only source of nitrogen, did not lead to any 2-oxoacid decarboxylase activity in a wild strain grown on ammonia (Vuralhan et al., 2005). It was concluded that an unknown second protein or transcriptional regulation accounted for this unexpected result. A single base exchange in a structural gene can completely alter metabolic equilibria. Yeast mutants with enhanced leucine uptake produced significantly more isoamyl alcohol and isoamyl acetate (Abe & Horikoshi, 2005). For one of these mutants, a defect of the LEU4 gene (coding for α-isopropylmalate synthase) caused by the exchange of

How microbial catalysts generate flavours

81

just one amino acid was found (Matsuta et al., 2004). These experiments have also confirmed the causal correlation of amino acid synthesis and fusel alcohol formation. The substitution of Met for Thr-62 in a farnesyl diphosphase synthase produced a mutant with an altered accumulation of higher terpenoids (Watanabe et al., 2002). The accumulating volatile terpenols imparted a Muscat-like aroma to sake produced with this strain. The transfer of glucosidase genes from Saccharomycopsis fibuligera increased terpenol levels by a more efficient hydrolysis of the glucoside precursors present in wine musts (van Rensburg et al., 2005). A more exploratory study transferred a methyl transferase of Antirrhinum majus (an ornamental plant) into S. cerevisiae (Farhi et al., 2006). Using benzoic acid as the substrate, small amounts of (antifungal) methyl benzoate were formed. Until recently, this kind of biotransformation relied on enzymes present in the selected strain. The novel tools of molecular biology now open access to any metabolite, not only flavours, in a host microorganism, although experimental hurdles have often been observed.

4.4

White biotechnology

4.4.1 Flavours from renewable substrates and agro-industrial wastes The steadily increasing price of crude oil is a clear signal of the limited resources of the planet. Alarming reports, such as the one issued by the Club of Rome, have eventually pricked the political consciousness. The European Commission has decided to support white biotechnologies; these are biocatalyst-based, sustainable processes of high ecological compatibility, i.e. using renewable resources, little energy and water, no toxic solvents or processing aids, and producing less waste streams than conventional production plants. A recent survey stated that the chemical industry in Germany produces more than 2 million tonnes of raw materials (equal to 12% of the total demand) from renewable sources (www.chemische-industrie.de/). The most familiar biotransformations in the flavour field start from renewables; simple hydrolyses of oils, proteins, starch, or of whole cells yield peptides, amino acids, nucleotides, fatty acids and small saccharides. Formation of volatile esters by reverse hydrolysis and some redox reactions on carbonyls starting from renewable compounds, such as fatty and amino acids or terpenes, were developed in the 1980s and found industrial application. The ‘green chemistry’ approach in combination with the ancient dream of the alchemists – production of gold from base metals – has resulted in a new look at agro-industrial waste streams. Side-streams of food processing, such as flavour condensates, and waste streams of traditional fermentations, such as starter culture distillates and wine lees oil are still used as flavouring materials. Modern biotechnology offers new options: fish waste was hydrolysed to a fish sauce containing more than 100 volatiles (Shih et al., 2003), and similarly squid skin (Huang et al., 2004) and a by-product of snow crab

82

Modifying flavour in food

processing (Baek et al., 2003) were treated enzymatically to yield flavour. Refined processes for the improvement of yeast (Komorowska et al., 2003) and gluten (Romero & Ho, 2005) hydrolysates are under development. A novel aspect is heating the hydrolysates to generate thermal flavours (Huang et al., 2004; Baek et al., 2003; Romero & Ho, 2005), and also the reverse approach, to first react a solution of Maillard precursors by heating and then to ferment them using baker’s yeast has been reported (Ota, 2005). Lactic acid bacteria growing in a semi-solid corn-based medium produced a flavour rich in diacetyl (Escamilla-Hurtado et al., 2005), and cassava bagasse, a large-scale agro-waste in Brazil, was fermented in packed-bed reactors using Kluyveromyces marxianus to generate volatile flavours (Medeiros et al., 2001). Fruity flavours, mainly smaller esters, were produced by Ceratocystis fimbriata grown on coffee husks in a solid-state fermentation (Medeiros et al., 2006). Today, attention is directed towards elucidating unusual substrates (Santos et al., 2003) and novel species, such as Basidiomycetes (Jang et al., 2004). The cold synthesis of thermal flavours (Rungsardthong & Noomhoom, 2005) and the highly enantioselective synthesis of both enantiomers of a given compound (Cappaert & Larroche, 2004) have presented challenging research goals. Not all of the suggested processes will find their way into industry, but the ubiquitous occurrence of flavour precursors and the universal biochemical pathways will facilitate future developments.

4.4.2 Advances in bioprocessing After a stable and productive microorganism or enzyme has been identified, it is the role of the bioengineer to select the most appropriate physical working conditions, to install peripheral devices for process monitoring and product recovery, and to scale-up the process. In the past, attempts to produce flavours through biotechnology have often neglected these aspects. Flavour precursors and products are special: often unstable, cytotoxic, usually hydrophobic, and easily lost through the off-gas stream of an aerobic cultivation. The status of a bioprocess is monitored on-line by measuring pH value, dissolved oxygen and cell density. These data are thought to correlate with metabolic activity and events, but are obviously not very meaningful. Adverse effects of cultivation parameters on flavour accumulation and the phase of maximum product concentration may well be missed. Fibre optic and fluorescence probes are available to measure intra- and extracellular metabolite concentrations (Hagedorn et al., 2004), and electronic noses to monitor, for example, wine fermentation, even in the presence of excess ethanol (Pinheiro et al., 2002). Genotypes of lactobacilli were distinguished using a mass spectrometer as a particularly efficient ‘electronic nose’ (Marilley et al., 2004). Our insufficient knowledge of the microbial pathways leading to volatile flavours is often reflected by trial-and-error approaches. Labelling studies

How microbial catalysts generate flavours COOH

COOH

NH2

OH

OAc

O

TA

(XLIV)

O

83

DC

(XLV)

Red

(XLVI)

AAT

(XLVII)

(XLVIII)

Fig. 4.9 Conversion of l-phenylalanine through transamination, decarboxylation, hydrogenation and esterification (TA, transaminase; DC, decarboxylase; Red, reductase; AAT, alcohol acyltransferase).

are still rare (Beck et al., 2004; Hampel et al., 2004; Zorn et al., 2004). After fundamental biochemical data on the metabolism of phenylalanine were discovered (Wittmann et al., 2002), a yeast-based process for the generation of gram per litre concentrations of 2-phenylethanol and 2-phenylethylacetate was successfully developed (Etschmann et al., 2004, 2005) (Fig 4.9). Both a hydrophobic precursor (edible oil, terpene hydrocarbon) and a flavour product may exert toxic effects on the biocatalyst. As the established separation technologies of the flavour industry (distillation, extraction, squeezing) inevitably harm the biocatalyst, other processing strategies have been developed. Fed-batch application of precursor and variants of in situ product recovery, such as adsorption and membrane techniques, were introduced. Organophilic pervaporation is the method of choice if the precursor, as in the above case of phenylalanine (Etschmann et al., 2005), is a compound highly soluble in water. On the other hand, reaction water was withdrawn from the esterification reactions leading to 3-methylbutylacetate or bornyl acetate, using hydrophilic membranes to drive the equilibrium towards the product side (Ehrenstein et al., 2005; Izak et al., 2005). A white biotechnology process is well complemented by such cold separation techniques. A tremendous number of experiments are required to determine the most appropriate biological, chemical and physical parameters of a bioprocess, with many of the variables being interdependent. Statistical methods, such as response surface methodology, accelerate bioprocess development by selecting a set of most relevant experimental parameters. The syntheses of butyl butanoate (Kumar & Rao, 2004), l-menthyl butanoate (Shih et al., 2007) and citronellyl esters (Melo et al., 2005) were optimised in that way.

4.5

Future trends

Large-scale application of a microorganism in an empirical food fermentation is an excellent motive for a genome sequencing project. The complete genomes of Lactobacillus sakei (1.88 Mbp), Lactobacillus plantarum

84

Modifying flavour in food

(3.7 Mbp) and Staphylococcus carnosus (2.56 Mpb) have been sequenced. One of the most recent examples is Oenococcus oeni (1.8 Mbp) which contained 1701 open reading frames, of which 75% were assigned to a metabolic function (Mills et al., 2005). On our way to the individual human genome new micro-methods are expected to increase the speed of genome sequencing (the so-called ‘$1000 genome’). Various promising methods are currently under development. Automated reading of base-specific fluorescence markers, combinations of specific primers and labelled stencils, or moving a single DNA strand through a nanopore driven by an electric field have been proposed. The sequence data will have to be linked to physiological properties and key metabolites to become really relevant (‘metabolomics’), but they still represent a big step forwards in understanding the metabolism of a GRAS organism and potential flavour producer. Nowadays it appears that large amounts of genetic information are rapidly acquired, whereas knowledge on physiological networks is growing more slowly. Only a few recent studies have dealt with this complicated matter (Passoth et al., 2006), and more efforts including labelling studies will be needed in the future. Progress in molecular biology is usually discussed in terms of improved biocatalysts: new substrates, new traits, new conversion capabilities and the like. Enzymes coupled to redox pairs with co-factor regeneration have been published, but even more elegant would be to couple desired enzymes in a recombinant microorganism, assuring regeneration not only of the cofactor, but also of the flavour-forming enzymes. Some other, already existing techniques, such as microsatellite polymerase chain reaction (PCR) (Howell et al., 2004) or 16S rRNA analyses (Zhang et al., 2005) will facilitate monitoring population dynamics. Multidimensional fast protein liquid chromatography (FPLC) methods for protein separation and particularly DNA microarray techniques (Kondo et al., 2003) will support expression analysis. Commercial microarray chips of S. cerevisiae will detect the expression of thousands of genes during fermentation. A human two-chip genotyping set with 500 000 DNA sequences (single nucleotide polymorphisms, SNPs) for medical diagnostics is now available for less than 1200. The microarray (or PCR)-based search for genes in environmental samples without cultivating the source microorganism (‘metagenome’ technique; Streit et al., 2004) is supposed to lead to novel catalysts with exciting properties. Databases of millions of amino acid and nucleobase sequences are electronically searchable, and everyone can identify functional properties of a gene or enzyme by homology studies, compare gene families and develop novel enzymes in silico (Martin & Bohlmann, 2004). Some established procedures and well-accepted beliefs will have to be critically re-examined. Is it physiologically reasonable to feed a flavour producer with excess glucose or sucrose, just because these carbon sources are readily catabolised (Verstrepen et al., 2004a)? Is it reasonable cost-wise to work in dilute aqueous solutions, or could unusual reaction media such as supercritical gases provide an alternative (Kumar et al., 2005)? Will mag-

How microbial catalysts generate flavours

85

netic beads replace the present techniques of enzyme immobilisation and facilitate enzyme recovery (Gomes et al., 2005), or will foam fractionation of enzymes provide an alternative (Gerken et al., 2005; Linke et al., 2005)? Will the surprising observation of an enhanced reaction rate in the presence of an electric current through a bioreactor be confirmed and explained (Mustacchi et al., 2005)? Will high-throughput measurement of volatile metabolites accelerate the selection and improvement of flavour producing strains (B. A. Smit et al., 2004b)?

4.6

Sources of further information

The chemical properties and industrial applications of flavours can be taken from recent editions of classical books (Burdock, 2004; Surburg and Panten, 2006). A collection of odour thresholds and other books on flavours are offered at www.thresholdcompilation.com and at dfa.leb.chemie.tu muenchen.de/DBuecher.html. More specifically, an overview on the biocatalytic generation and modification of flavours up to 1994 can be found in a monograph (Berger, 1995). Progress achieved during the last decade is covered by some comprehensive reviews (Ehrenstein, 2005; Etschmann et al., 2004; Frey, 2005); more specific topics, as addressed above, have also been reviewed (Table 4.1).

Table 4.1 Recent reviews covering a single topic as mentioned in this chapter Main subject

Reference

Sources of natural flavours Glycoside hydrolases Esters in beer and dairy products Fungal cultures, low-cost substrates, downstream processing Fungi with emphasis on Ceratocystis, volatile terpenes Biotransformation, vanillin, monoterpenes, non-volatiles Enzymes and dairy flavour

Frey (2005) Sarry and Gunata (2004) Liu et al. (2004), Stewart (2005) Divyashree et al. (2006)

Flavour of sourdough Grape and wine flavour Flavours from plant cell culture Scale-up of natural product formation Lipase from Candida antarctica in oxidation reactions Lipoxygenase, cytochrome P450, recombinant enzymes

Hanssen (2002) Suresh et al. (2006) G. Smit et al. (2004, 2005), Wilkinson and Kilcawley (2005) Gul et al. (2005), Hansen and Schieberle (2005), Katina (2005) Swiegers et al. (2005), Pretorius and Hoj (2005) Hrazdina (2006) Berger et al. (2005) Gatfield et al. (2005, 2006a,b) Casey and Hughes (2004)

86

Modifying flavour in food

Most flavour producers are organised in national associations, and these fused in 1969 to form the International Organization of the Flavor Industry (IOFI)(//iofi.org/). IOFI is focusing on activities revolving around consumer safety and provides links of general interest on its web page. How the director of research of a flavour company is looking at the term ‘natural’ can be found at www.chemsoc.org/chembytes/ezine/1998/sine.htm. The site of Leffingwell & Associates provides numerous links to other flavour sites, for example, to updates of the GRAS lists and properties of chiral flavours (www.leffingwell.com/links.htm). An interactive tool for the calculation of logP and logW data is provided by www.logp.com/. A wealth of information on enzymes is presented at //brenda.bc.uni-koeln.de/, www.expasy.ch/ enzyme/, and www.rcsb.org/pdb/. Nucleotide sequences and much more can be searched at the web site of the European Bioinformatics Institute at srs6. ebi.ac.uk/. One of the bioinformatics tools for metabolome analysis can be downloaded at mzmine.sourceforge.net/.

4.7

References

abe, f. and horikoshi, k. (2005), Cellular & Molecular Biology Letters, 10, 383–388. aguedo, m., wache, y., coste, f., husson, f. and belin, j.-m. (2004), Journal of Molecular Catalysis B: Enzymatic, 29, 31–36. aguedo, m., wache, y., belin, j.-m. and teixeira, j. a. (2005), Biotechnology Letters, 27, 417–422. aharoni, a., jongsma, m. a., verhoeven, h. a. and bouwmeester, h. j. (2002, 2004) US Pat. Appl. Publ. 2004, Cont.-in-part of WO 2002 64,764 US 2004161819 A1 20040819. alewijn, m., sliwinski, e. l. and wouters, j. t. m. (2005), International Dairy Journal, 15, 733–740. alexandre, h., charleux, p., lavesvre, a., nedjma, m. and chassagne, d. (2004), Revue Francaise d’Oenologie, 205, 9–11. amarita, f., yvon, m., nardi, m., chambellon, e., delettre, j. and bonnarme, p. (2004), Applied and Environmental Microbiology, 70, 7348–7354. andrade, m. j., casado, e., rodriguez, m. m., cordoba, j. j., martin, a., perez, f. and cordoba, m. g. (2005), Alimentacion Equipos y Tecnologia, 198, 42–47. ardoe, y. (2006), Biotechnology Advances, 24, 238–242. arfi, k., leclercq-perlat, m.-n., baucher, a., tache, r., delettre, j. and bonnarme, p. (2004), Lait, 84, 435–447. arfi, k., leclercq-perlat, m. n., spinnler, h. e. and bonnarme, p. (2005), International Dairy Journal, 15, 883–891. baek, h. h., jang, j. t. and park, g. h. (2003), Repub. Korean Kongkae Taeho Kongbo. KR 2003028089 A 20030408. baldermann, s., naim, m. and fleischmann, p. (2005), Food Research International, 38, 833–836. beck, h. c., hansen, a. m. and lauritsen, f. r. (2004), Journal of Applied Microbiology, 96, 1185–1193. beekwilder, j., alvarez-huerta, m., neef, e., verstappen, f. w. a., bouwmeester, h. j. and aharoni, a. (2004), Plant Physiology, 135, 1865–1878. benito, m. j., rodriguez, m., martin, a., aranda, e. and cordoba, j. j. (2004), Meat Science, 67, 497–505.

How microbial catalysts generate flavours

87

benito, m. j., rodriguez, m., cordoba, m. g., andrade, m. j. and cordoba, j. j. (2005a), Journal of the Science of Food and Agriculture, 85, 273–280. benito, m. j., nunez, f., cordoba, m. g., martin, a. and cordoba, j. j. (2005b), Food Microbiology, 22, 513–519. berger, r. g. (1995), Aroma Biotechnology, Springer, Berlin, Heidelberg, New York. berger, r. g. and drawert, f. (1984), Journal of the Science of Food and Agriculture, 35, 1318–1325. berger, r. g., scheper, t. and schuegerl, k. (2005), Molecular Nutrition & Food Research, 49, 732–743. bezbradica, d., karalazic, i., ognjanovic, n., mijin, d., siler-marinkovic, s. and knezevic, z. (2006), Journal of the Serbian Chemical Society, 71, 31–41. blagden, t. a. d. and gilliland, s. e. (2005), Journal of Food Science, 70, M186–M189. blinkovsky, a., brown, k., golightly, e. and byun, t. (2004), US 6800467 B1 20041005. bondar, d. c. t., beckerich, j.-m. and bonnarme, p. (2005), Applied and Environmental Microbiology, 71, 4585–4591. brunati, m., marinelli, f., bertolini, c., gandolfi, r., daffonchio, d. and molinari, f. (2004), Enzyme and Microbial Technology, 34, 3–9. burdock, g. a. (2004), Fenaroli’s Handbook of Flavor Ingredients, 5th edn. CRC Press, Washington, DC. buzzini, p., gasparetti, c., turchetti, b., cramarossa, m. r., vaughan-martini, a., martini, a., pagnoni, u. m. and forti, l. (2005a), Archives of Microbiology, 184, 187–193. buzzini, p., romano, s., turchetti, b., vaughan, a., pagnoni, u. m. and davoli, p. (2005b), FEMS Yeast Research, 5, 379–385. capelle, s. (2004), PCT Int. Appl. WO 2004043153 A1 20040527 WO 2003-BE197 20031114. cappaert, l. and larroche, c. (2004), Biocatalysis and Biotransformation, 22, 291–296. carrau, f. m., medina, k., boido, e., farina, l., gaggero, c., dellacassa, e., versini, g. and henschke, p. a. (2005), Seccion Enologia, FEMS Microbiology Letters, 243, 107–115. casey, r. and hughes, r. k. (2004), Food Biotechnology, 18, 135–170. chen, g., hackett, r., walker, d., taylor, a., lin, z. and grierson, d. (2004), Plant Physiology, 136, 2641–2651. clemente-jimenez, j. m., mingorance-cazorla, l., martinez-rodriguez, s., las heras-vazquez, f. j. and rodriguez-vico, f. (2005), International Journal of Food Microbiology, 98, 301–308. coghe, s., benoot, k., delvaux, f., vanderhaegen, b. and delvaux, f. r. (2004), Journal of Agricultural and Food Chemistry, 52, 602–608. czerny, m., schieberle, p., brandt, m. and hammes, w. (2005), Getreidetechnologie, 59, 15–19. davis, e. m., ringer, k. l., mcconkey, m. e. and croteau, r. (2005), Plant Physiology, 137, 873–881. de carvalho, c. c. c. r. and da fonseca, m. m. r. (2005), Food Chemistry, 95, 413– 422. de wit, m., osthoff, g., viljoen, b. c. and hugo, a. (2005), Enzyme and Microbial Technology, 37, 606–616. divyashree, m. s., george, j. and agrawal, r. (2006), Journal of Food Science and Technology, 43, 73–76. dufosse, l., blin-perrin, c., souchon, i. and feron, g. (2002), Food Science and Biotechnology, 11, 192–202.

88

Modifying flavour in food

dung, n. t. p., rombouts, f. m. and nout, m. j. r. (2005), Innovative Food Science & Emerging Technologies, 6, 429–441. echeverria, g., graell, j., lopez, m. l. and lara, i. (2004), Postharvest Biology and Technology, 31, 217–227. edens, l. and hille, j. d. r. (2005), PCT Int. Appl., WO 2005117595 A1 20051215. ehrenstein, u., hennig, t., kabasci, s., kreis, p. and gorak, a. (2005), Chemie Ingenieur Technik, 77, 1551–1556. el tamer, m. k., smeets, m., holthuysen, n., lucker, j., tang, a., roozen, j., bouwmeester, h. j. and voragen, a. g. j. (2003), Journal of Biotechnology, 106, 15–21. erasmus, d. j., cliff, m. and van vuuren, h. j. j. (2004), American Journal of Enology and Viticulture, 55, 371–378. escamilla-hurtado, m. l., valdes-martinez, s. e., soriano-santos, j., gomez-pliego, r., verde-calvo, j. r., reyes-dorantes, a. and tomasini-campocosio, a. (2005), International Journal of Food Microbiology, 105, 305–316. estevez, p., gil, m. l. and falque, e. (2004), International Journal of Food Science and Technology, 39, 61–69. etschmann, m. m. w., sell, d. and schrader, j. (2004), State-of-the-Art in Flavour Chemistry and Biology, Proceedings of the 7th Wartburg Symposium on Flavour Chemistry and Biology, Eisenach, Germany, 21–23, April Editors: Hofmann, T., Rothe, M., Schieberle P 248–253. etschmann, m. m. w., sell, d. and schrader, j. (2005), Biotechnology and Bioengineering, 92, 624–634. fadda, m. e., mossa, v., pisano, m. b., deplano, m. and cosentino, s. (2004), International Journal of Food Microbiology, 95, 51–59. farhi, m., dudareva, n., masci, t., weiss, d., vainstein, a. and abeliovich, h. (2006), Journal of Biotechnology, 122, 307–315. fickers, p., benetti, p.-h., wache, y., marty, a., mauersberger, s., smit, m. s. and nicaud, j.-m. (2005), FEMS Yeast Research, 5, 527–543. flores, m., dura, m.-a., marco, a. and toldra, f. (2004), Meat Science, 68, 439–446. frey, c. (2005), Natural Flavors and Fragrances ACS Symposium Series no. 908, 3–19. fujiwara, d. (2004), Abstracts of Papers, 228th ACS National Meeting, Philadelphia, PA, United States, 22–26 August, AGFD-066. ACS, Washington, DC. fukuda, k. (2004), Seibutsu Kogaku Kaishi, 82, 50–55. furusawa, m., hashimoto, t., noma, y. and asakawa, y. (2005), Chemical & Pharmaceutical Bulletin, 53, 1513–1514. galkina, g. v., illarionova, v. i. and gorbatova, e. v. (2004), Proizvodstvo Spirta, i Likerovodochnyh Izdelii, 2, 21–23. ganesan, b., seefeldt, k., koka, r. c., dias, b. and weimer, b. c. (2004), International Dairy Journal, 14, 237–246. ganzle, m. g., thiele, c., vermeulen, n. and vogel, f. (2005), Getreidetechnologie, 59, 148–152. garbe, l.-a. and tressl, r. (2006), Chemistry & Biodiversity, 1, 900–915. garde, s., avila, m., gaya, p., medina, m. and nunez, m. (2006), Journal of Dairy Science, 89, 840–849. gatfield, i., hilmer, j.-m. and bertram, h.-j. (2005), Perfumer & Flavorist, 30, 46–49. gatfield, i. l., hilmer, j.-m. and bertram, h.-j. (2006a), Advances in Food Science, 28, 139–141. gatfield, i. l., hilmer, j.-m., lorenz, s. and bertram, h.-j. (2006b), Advances in Food Science, 28, 142–145. gerken, b. m., wattenbach, c., linke, d., zorn, h., berger, r. g. and parlar, h. (2005), Analytical Chemistry, 77, 6113–6117.

How microbial catalysts generate flavours

89

gomes, c. s. g., petersen, t. l., hobley, t. j. and thomas, o. r. t. (2005), 7th World Congress of Chemical Engineering, Glasgow, United Kingdom, 10–14 July, 86907/1–86907/10. green, s. a., friel, e. n., beuning, l. l. and macrae, a. e. (2004), PCT Int. Appl. 2004, WO 2004035791 A1 20040429. grimaldi, a., bartowsky, e. and jiranek, v. (2005), Journal of Applied Microbiology, 99, 1061–1069. groguenin, a., wache, y., garcia, e. e., aguedo, m., husson, f., ledall, m.-t., nicaud, j.-m. and belin, j.-m. (2004), Journal of Molecular Catalysis B: Enzymatic, 28, 75–79. gul, h., ozcelik, s., sagdic, o. and certel, m. (2005), Process Biochemistry, 40, 691– 697. guterman, i., shalit, m., menda, n., piestun, d., dafny-yelin, m., shalev, g., bar, e., davydov, o., ovadis, m., emanuel, m., wang, j., adam, z., pichersky, e., lewinsohn, e., zamir, d., vainstein, a. and weiss, d. (2002), Plant Cell, 14, 2325–2338. hagedorn, a., levadoux, w., groleau, d. and tartakovsky, b. (2004), Biotechnology Progress, 20, 361–367. hall, c. e., husson, f. and kermasha, s. (2004), Journal of Molecular Catalysis B: Enzymatic, 29, 201–209. hall, c. e., husson, f. and kermasha, s. (2005), Flavour and Fragrance Journal, 20, 690–702. hambraeus, g. and nyberg, n. (2005), Journal of Agricultural and Food Chemistry, 53, 8714–8721. hampel, d., mosandl, a. and wuest, m. (2004), State-of-the-Art in Flavour Chemistry and Biology, Proceedings of the 7th Wartburg Symposium on Flavour Chemistry and Biology, Eisenach, Germany, 21–23 April, Editors: Hofmann, T., Rothe, M., Schieberle P 267–273. hampel, d., mosandl, a. and wuest, m. (2006), Journal of Agricultural and Food Chemistry, 54, 1473–1478. Hannoversche Allgemeine Zeitung. (2006), 26 May, 2006 (in German) (www.haz.de). hansen, a. and schieberle, p. (2005), Trends in Food Science & Technology, 16, 85–94. hanssen, h.-p. (2002), Pharmazeutische Zeitung, 147, 38, 40. havkin-frenkel, d., french, j. c., graft, n. m., joel, d. m., pak, f. e. and frenkel, c. (2004), Acta Horticulturae, 629, 93–102. havkin-frenkel, d., french, j., pak, f. e. and frenkel, c. (2005), Perfumer & Flavorist, 30, 36–43, 46–55. helinck, s., le bars, d., moreau, d. and yvon, m. (2004), Applied and Environmental Microbiology, 70, 3855–3861. herranz, b., de la hoz, l., hierro, e., fernandez, m. and ordonez, j. a. (2005), Food Chemistry, 91, 673–682. hierro, e., ordonez, j. a., bruna, j. m., pin, c., fernandez, m. and hoz, l. (2005), European Food Research and Technology, 220, 494–501. hirosawa, i., aritomi, k., hoshida, h., kashiwagi, s., nishizawa, y. and akada, r. (2004), Applied Microbiology and Biotechnology, 65, 68–73. hitchman, t., robertson, d. e., hiraiwa, m., phillips, y. and gray, k. a. (2005), PCT Int. Appl., WO 2005021714 A2 20050310. hockelmann, c. and jüttner, f. (2005), Flavour and Fragrance Journal, 20, 387– 394. holland, r., liu, s.-q., crow, v. l., delabre, m.-l., lubbers, m., bennett, m. and norris, g. (2005), International Dairy Journal, 15, 711–718. horton, c. e. and bennett, g. n. (2006), Enzyme and Microbial Technology, 38, 937–943.

90

Modifying flavour in food

howell, k. s., bartowsky, e. j., fleet, g. h. and henschke, p. a. (2004), Letters in Applied Microbiology, 38, 315–320. howell, k. s., klein, m., swiegers, j. h., hayasaka, y., elsey, g. m., fleet, g. h., hoj, p. b., pretorius, i. s., de barros, l. and miguel, a. (2005), Applied and Environmental Microbiology, 71, 5420–5426. hrazdina, g. (2006), Journal of Agricultural and Food Chemistry, 54, 1116–1123. hrdlicka, p. j., sorensen, a. b., poulsen, b. r., ruijter, g. j. g., visser, j. and iversen, j. j. l. (2004), Biotechnology Progress, 20, 368–376. huang, g., jiang, j. and ying, y. (2004), Zhongguo Tiaoweipin, 1, 27–29, 37. husson, f., krumov, k. n., cases, e., cayot, p., bisakowski, b., kermasha, s. and belin, j.-m. (2004), Process Biochemistry, 40, 1395–1400. huynh-ba. t., matthey-doret, w., fay, l. b. and rhlid, r. b. (2003), Journal of Agricultural and Food Chemistry, 51, 3629–3635. izak, p., mateus, n. m. m., afonso, c. a. m. and crespo, j. g. (2005), Separation & Purification Technology, 41, 141–145. jang, i.-h., kang, m.-s. and chae, h. j. (2004), Han’guk Eungyong Sangmyong Hwahakhoeji, 47, 49–54. jin, z., xu. j., niu, s., ma. c., cheng, w. and wang, q. (2004), Shipin Kexue, 25, 45– 48. kang, s.-m., jung, h.-y., kang, y.-m., min, j.-y., karigar, c. s., yang, j., kim, s.-w., ha, y.-r., lee, s.-h. and choi, m.-s. (2005), Journal of Agricultural and Food Chemistry, 53, 3449–3453. kania, m. and stasinska, b. (2004), Biotechnologia, 3, 260–270. kaspera, r., krings, u., nanzad, t. and berger, r. g. (2005a), Applied Microbiology and Biotechnology, 67, 477–483. kaspera, r., krings, u., pescheck, m., sell, d., schrader, j. and berger, r. g. (2005b), Z Naturforsch.C, 60c, 459–466. katina, k. (2005), Sourgough:. A Tool for the Improved Flavour, Texture and Shelf-life of Wheat Bread, VTT Publications no. 569, 1–176. keasling, j. y. d., martin, v., pitera, d., withers, s., yoshikuni, y., newman, j., reiling, k., renninger, n. s. and marshall, j. (2004), Abstracts of Papers, 228th ACS National Meeting, Philadelphia, PA, United States, 22–26 August, 2004. kieronczyk, a., skeie, s., langsrud, t., le bars, d. and yvon, m. (2004), International Dairy Journal, 14, 227–235. kilcawley, k. n., wilkinson, m. g. and fox, p. f. (2006), Food Research International, 39, 619–627. kim, a. y. and eckert, m. (2003), Abstracts of Papers, 226th ACS National Meeting, NY, United States. king, a. j. and dickinson, j. r. (2003), FEMS Yeast Research, 3, 53–62. komorowska, a., sieliwanowicz, b., mrowka, e., stecka, k. and halasinska, a. (2003), Electronic Journal of Polish Agricultural Universities, 6. kondo, h., yomo, h., takatani, t., izumi, t., hatanaka, h., fujita, a. and fukui, n. (2003), Proceedings of the Congress – 29th European Brewery Convention, 49/1–49/13. krings, u., brauer, b., kaspera, r. and berger, r. g. (2005), Biocatalysis and Biotransformation, 23, 457–463. kumar, g. v. and rao, m. n. (2004), Journal of Food Science and Technology, 41, 560– 562. kumar, r., modak, j. and madras, g. (2005), Biochemical Engineering Journal, 23, 199–202. larios, a., garcia, h. s., oliart, r. m. and valerio-alfaro, g. (2004), Applied Microbiology and Biotechnology, 65, 373–376.

How microbial catalysts generate flavours

91

leclercq-perlat, m.-n., latrille, e., corrieu, g. and spinnler, h.-e. (2004), Journal of Dairy Research, 71, 355–366. lewinsohn, e., sitrit, y., bar, e., azulay, y., ibdah, m., meir, a., yosef, e., zamir, d. and tadmor, y. (2005a), Trends in Food Science & Technology, 16, 407–415. lewinsohn, e., sitrit, y., bar, e., azulay, y., meir, a., zamir, d. and tadmor, y. (2005b), Journal of Agricultural and Food Chemistry, 53, 3142–3148. linke, d., zorn, h., gerken, b., parlar, h. and berger, r. g. (2005), Lipids, 40, 323–327. liu, s.-q, holland, r. and crow, v. l. (2004), International Dairy Journal, 14, 923– 945. maldonado-robledo, g., rodriguez-bustamante, e., sanchez-contreras, a., rodriguez-sanoja, r. and sanchez, s. (2003), Applied Microbiology and Biotechnology, 62, 484–488. mamede, m. e. o., cardello, h. m. a. b. and pastore, g. m. (2004), Food Chemistry, 89, 63–68. marilley, l., ampuero, s., zesiger, t. and casey, m. g. (2004), International Dairy Journal, 14, 849–856. martin, d. m. and bohlmann, j. (2004), Phytochemistry, 65, 1223–1229. matsuta, y., ishigaki, h. and koseki, t. (2004), Nippon Jozo Kyokaishi, 99, 127–133. medeiros, a. b. p., pandey, a., christen, p., fontoura, p. s. g., de freitas, r. j. s. and soccol, c. r. (2001), World Journal of Microbiology & Biotechnology, 17, 767– 771. medeiros, a. b. p., pandey, a., vandenberghe, l. p. s., pastore, g. m. and soccol, c. r. (2006), Food Technology and Biotechnology, 44, 47–51. melo, l. l. m. m., pastore, g. m. and macedo, g. a. (2005), Process Biochemistry, 40, 3181–3185. mills, d. a., rawsthorne, h., parker, c., tamir, d. and makarova, k. (2005), FEMS Microbiology Reviews, 29, 465–475. mustacchi, r., knowles, c. j., li. h., dalrymple, i., sunderland, g., skibar, w. and jackman, s. a. (2005), Biotechnology and Bioengineering, 89, 18–23. nampoothiri, k. m., nagy, v., kovacs, k., szakacs, g. and pandey, a. (2005), Letters in Applied Microbiology, 41, 498–504. nardi, m., sextius, p., bonnarme, p., spinnler, h. e., monnet, v. and irlinger, f. (2005), Journal of Dairy Research, 72, 179–187. nazhat-ul-ainn, r., premanath and agrawal, r. (2002), Asian Journal of Microbiology, Biotechnology & Environmental Sciences, 4, 265–269. neto, r. s., pastore, g. m. and macedo, g. a. (2004), Journal of Food Science, 69, C677–C680. nonier, m.-f., de gaulejac, n. v., vivas, n. and vitry, c. (2004), Comptes Rendus Chimie, 7, 689–698. olesen, p. t., stahnke, l. h. and talon, r. (2004), Meat Science, 68, 193–200. oliveira, e. s., cardello, h. m. a. b., jeronimo, e. m., souza, e. l. r. and serra, g. e. (2005), World Journal of Microbiology & Biotechnology, 21, 707–715. ortigosa, m., arizcun, c., torre, p. and izco, j. m. (2005), Journal of Dairy Research, 72, 168–178. ota, m. (2005), Jpn. Kokai Tokkyo Koho JP 2005304377 A2 20051104. paraggio, m. and fiore, c. (2004), World Journal of Microbiology & Biotechnology, 20, 743–747. park, y. j., kim, i. c. and chang, h. c. (2003), Journal of Microbiology and Biotechnology, 13, 636–639. passoth, v., fredlund, e., druvefors, u. a. and schnurer, j. (2006), FEMS Yeast Research, 6, 3–13. pechous, s. w. and whitaker, b. d. (2004), Planta, 219, 84–94.

92

Modifying flavour in food

pechous, s. w. and whitaker, b. d. (2005), Acta Horticulturae, 682 (Proceedings of the 5th International Postharvest Symposium, 2004, Volume 1), 395–402. pinheiro, c., rodrigues, c. m., schäfer, t. and crespo, j. g. (2002), Biotechnology and Bioengineering, 77, 632–640. ponzone, c., d’adago, v. and bertani, m. (2004), PCT Int. Appl., WO 2004091316 A1 20041028 WO 2003-EP14702 20031222. pozo-bayon, m. a., g-alegria, e., polo, m. c., tenorio, c., martin-alvarez, p. j., calvo de la banda, m. t., ruiz-larrea, f. and moreno-arribas, m. v. (2005), Journal of Agricultural and Food Chemistry, 53, 8729–8735. pretorius, i. s. and hoj, b. (2005), Australian Journal of Grape and Wine Research, 11, 83–108. priefert, h., rabenhorst, j. and steinbüchel, a. (2001), Applied Microbiology and Biotechnology, 56, 296–314. priest, f. g. and cachat, e. (2004), Proceedings of the 28th Convention of the Institute & Guild of Brewing, Asia Pacific Section, Hanoi, 21–26 March, 102–105. rodriguez-bustamante, e., maldonado-robledo, g., ortiz, m. a., diaz-avalos, c. and sanchez, s. (2005), Applied Microbiology and Biotechnology, 68, 174–182. romero, m., villafuerte and ho, c.-t. (2005), Abstracts of Papers, 230thACS National Meeting, Washington, DC, 28 Aug.–1 Sept. 2005. rungsardthong, v. and noomhoom, a. (2005), Flavour and Fragrance Journal, 20, 710–714. sakamoto, t., nakamura, n., kodera, t., nio, n., yamada, n. and wakabayashi, h. (2004), PCT Int. Appl. WO 2004105503 A1 20041209. sanchez, j., ignacio, m. b. and rodriguez, a. (2005), International Journal of Food Microbiology, 105, 377–387. santos, a. s., pereira, n. jr, da silva, i. i., sarquis, m. i. and antunes, o. a. c. (2003), Applied Biochemistry and Biotechnology, 105–108, 649–657. sarry, j.-e. and gunata, z. (2004), Food Chemistry, 87, 509–521. schallmey, m., singh, a. and ward, o. p. (2004), Canadian Journal of Microbiology, 50, 1–17. schoondermark-stolk, s. a., jansen, m., veurink, j. h., verkleij, a. j., verrips, c. t., euverink, g.-j. w., boonstra, j. and dijkhuizen, l. (2006), Applied Microbiology and Biotechnology, 70, 237–246. schwan, r. and wheals, a. (2004), Critical Reviews in Food Science and Nutrition, 44, 205–221. schwartz, s. h., cai tan, b., gage, d. a., zeevaart, j. a. d. and mccarty, d. r. (1997), Science, 276, 1872. setyaningsih, d., apriyantono, a. and suhartono, m. t. (2005), Food Flavor and Chemistry Special Publication – Royal Society of Chemistry, 300, 145–155. sevini, f., marino, r., grando, m. s., moser, s. and versini, g. (2004), Acta Horticulturae, 652 (Proceedings of the 1st International Symposium on Grapevine Growing, Commerce and Research, 2003), 439–446. sharon-asa, l., shalit, m., frydman, a., bar, e., holland, d., or, e., lavi, u., lewinsohn, e. and eyal, y. (2003), Plant Journal, 36, 664–674. shih, i.-l., chen, l.-g., yu, t.-s., chang, w.-t. and wang, s.-l. (2003), Enzyme and Microbial Technology, 33, 154–162. shih, i.-l., hung, s.-h., chen, f.-y. and shieh, c.-j. (2007), Food Chemistry, 100, 1223– 1228. simkin, a. j., schwartz, s. h., auldridge, m., taylor, m. g. and klee, h. j. (2004a), Plant Journal, 40, 882–892. simkin, a. j., underwood, b. a., auldridge, m., loucas, h. m., shibuya, k., schmelz, e., clark, d. g. and klee, h. j. (2004b), Plant Physiology, 136, 3504–3514. smit, b. a., engels, w. j. m., wouters, j. t. m. and smit, g. (2004a), Applied Microbiology and Biotechnology, 64, 396–402.

How microbial catalysts generate flavours

93

smit, b. a., burgering, m., engels, w., de jong, c., smit, g. editors: hofmann, t., rothe, m. and schieberle, p. (2004b), State-of-the-Art in Flavour Chemistry and Biology Proceedings of the 7th Wartburg Symposium on Flavour Chemistry and Biology, Eisenach, Germany, 21–23 April, 457–463. smit, b. a., van hylckama vlieg, j. e. t., engels, w. j. m., meijer, l., wouters, j. t. m. and smit, g. (2005), Applied and Environmental Microbiology, 71, 303– 311. smit, g., wouters, j. t. m. and meijer, w. c. (2004), in Handbook of Food and Beverage Fermentation Technology, editors: Hui, Y. H., Goddich, L. M., Hansen, A. S., Josephsen, J., Nip, W-K, Stanfield, P. S., Toldra, P., Food Science and Technology no. 134, CRC Press Boca Raton, FC, 89–111. smit, g., smit, b. a. and engels, w. j. m. (2005), FEMS Microbiology Reviews, 29, 591–610. souleyre, e. j. f., greenwood, d. r., friel, e. n., karunairetnam, s. and newcomb, r. d. (2005), FEBS Journal, 272, 3132–3144. sousa de brito, e., garcia, n. h. p. and amancio, a. c. (2004), Brazilian Archives of Biology and Technology, 47, 553–558. stewart, g. g. (2005), Proceedings of the Congress – 30th European Brewery Convention, Prague, 14–19 May, 100/1–100/10. stockfleth, r. and fiebig, b. (2004), Deutsche Lebensmittel-Rundschau, 100, 343– 347. streit, w. r., daniel, r. and jaeger, k.-e. (2004), Current Opinion in Biotechnology, 15, 285–290. surburg, h. and panten, j. (2006), Common Fragrance and Flavor Materials – Preparation, Properties and Uses, 5th edn. Wiley-VCH, Weinheim. suresh, b., ritu, t. and ravishankar, g. a. (2003), Biocatalysis and Biotransformation, 21, 333–340. suresh, b., ritu, t. and ravishankar, g. a. (2006), in Food Biotechnology, editors Shetty, K., Paliyath, G., Pometto, A., Levin, R. E., Food Science and Technology no. 148 CRC Press, Boca Raton, FC, 1655–1690. suzuki, h. and kumagai, h. (2004), Challenges in Taste Chemistry and Biology, ACS Symposium Series, 867, 223–237. swiegers, j. h., bartowsky, e. j., henschke, p. a. and pretorius, i. s. (2005), Australian Journal of Grape and Wine Research, 11, 139–173. takizawa, r., watari, j. and kawamura, a. (2005), Jpn. Kokai Tokkyo Koho, JP 2005210939 A2 20050811. thage, b. v., broe, m. l., petersen, m. h., petersen, m. a., bennedsen, m. and ardoe, y. (2005), International Dairy Journal, 15, 795–805. thierry, a., maillard, m.-b., richoux, r., kerjean, j.-r. and lortal, s. (2005), Lait, 85, 57–74. tjener, k., stahnke, l. h., andersen, l. and martinussen, j. (2004a), Meat Science, 67, 711–719. tjener, k., stahnke, l. h., andersen, l. and martinussen, j. (2004b), Meat Science, 67, 447–452. torre, p., de faveri, d., perego, p., ruzzi, m., barghini, p., gandolfi, r. and converti, a. (2004), Annals of Microbiology, 54, 517–527. tripathi, u., ramachandra, r. s. and ravishankar, g. a. (2002), Process Biochemistry, 38, 419–426. tungjaroenchai, w., white, c. h., holmes, w. e. and drake, m. a. (2004), Journal of Dairy Science, 87, 3224–3234. ugliano, m. and moio, l. (2005), Journal of Agricultural and Food Chemistry, 53, 10134–10139. vadali, r. v., bennett, g. n. and san, k.-y. (2004), Metabolic Engineering, 6, 294– 299.

94

Modifying flavour in food

van laere, s. d. m., saerens, s. m. g., dufour, j.-p., van dijck, p., derdelinckx, g., thevelein, j. m., verstrepen, k. j. and delvaux, f. r. (2005), Proceedings of the Congress – 30th European Brewery Convention, Prague, 14–19 May, 36/1–36/4. van rensburg, p., stidwell, t., lambrechts, m. g., cordero, o., ricardo and pretorius, i. s. (2005), Annals of Microbiology, 55, 33–42. verstrepen, k. j., moonjai, n., bauer, f. f., derdelinckx, g., dufour, j.-p., winderickx, j., thevelein, j. m., pretorius, i. s., and delvaux, f. r. in Editor: smart, k. a. (2003a), Brewing Yeast Fermentation Performance. (2nd Edition), Blackwell Publishing, Ames, 1A, 234–248. verstrepen, k. j., van laere, s. d. m., vanderhaegen, b. m. p., derdelinckx, g., dufour, j.-p., pretorius, i. s., winderickx, j., thevelein, j. m. and delvaux, f. r. (2003b), Applied and Environmental Microbiology, 69, 5228–5237. verstrepen, k. j., iserentant, d., malcorps, p., derdelinckx, g., van dijck, p., winderickx, j., pretorius, i. s., thevelein, j. m. and delvaux, f. r. (2004a), Trends in Biotechnology, 22, 531–537. verstrepen, k. j., van laere, s. d. m., vercammen, j., derdelinckx, g., dufour, j.-p., pretorius, i. s., winderickx, j., thevelein, j. m. and delvaux, f. r. (2004b), Yeast, 21, 367–377. volschenk, h., viljoen-bloom, m., van staden, j., husnik, j. and van vuuren, h. j. j. (2004), South African Journal of Enology and Viticulture, 25, 63–73. vuralhan, z., luttik, m. a. h., tai, s. l., boer, v. m., morais, m. a., schipper, d., almering, m. j. h., koetter, p., dickinson, j. r., daran, j.-m. and pronk, j. t. (2005), Applied and Environmental Microbiology, 71, 3276–3284. wakabayashi, h., wakabayashi, m., engel, k.-h. editors: le quere, j.-l. and etievant, p. x. (2002), Flavour Research at the Dawn of the Twenty-First Century Proceedings of the 10th Weurman Flavor Research Symposium, Beaune, France, 25–28 June, 350–355. wang, j. and de luca, v. (2005), Plant Journal, 44, 606–619. watanabe, h. (2005), Koryo, 226, 113–118. watanabe, m., ueda, m., asai, t. and nishimura, a. (2002), Jpn. Kokai Tokkyo Koho, JP 2002238572 A2 20020827. wenzel, s. c. and mueller, r. (2005), BIOspektrum, 11, 628–631. wilkinson, m. g. and kilcawley, k. n. (2005), International Dairy Journal, 15, 817–830. williams, a. g., beattie, s. h. and banks, j. m. (2004b), International Journal of Dairy Technology, 57, 7–13. williams, a. g., noble, j. and banks, j. m. (2004a), Letters in Applied Microbiology, 38, 289–295. wittmann, c., hans, m. and bluemke, w. (2002), Yeast, 19, 1351–1363. yahyaoui, f. e. l., wongs-aree, c., latche, a., hackett, r., grierson, d. and pech, j.-c. (2002), European Journal of Biochemistry, 269, 2359–2366. zhang, w., qiao, z., shigematsu, t., tang, y., hu, c., morimura, s. and kida, k. (2005), Journal of the Institute of Brewing, 111, 215–222. zhou, j., yang, h., lin, g. and yang, s. (2004), Hunan Nongye Daxue Xuebao, 30, 20–23. zorn, h., langhoff, s., scheibner, m., nimtz, m. and berger, r. g. A peroxidase from Lepista irina cleaves β,β-Carotene to flavor compounds. Biol. Chem. (2003a), 384, 1049–1056. zorn, h., langhoff, s., scheibner, m. and berger, r. g. (2003b), Applied Microbiology and Biotechnology, 62, 331–336. zorn, h., neuser, f. and berger, r. g. (2004), Journal of Biotechnology, 107, 255–263. zorn, h., peters, t., nimtz, m. and berger, r. g. (2005), Proteomics, 5, 4832–4838.

5 New developments in yeast extracts for use as flavour enhancers B. Noordam and F. R. Meijer, DSM Food Specialties, The Netherlands

5.1

Introduction

Yeast extracts are widely used in savoury-type food applications because of their very interesting composition. For instance, autolytic yeast extracts are used because they are rich in protein, peptides, amino acids and B vitamins. On the other hand, hydrolytic yeast extracts are mainly used for the presence of the taste-enhancing 5′nucleotides, 5′-guanine monophosphate (5′GMP) and 5′-inosine monophosphate (5′IMP). Both types of yeast extract provide, in addition, a savoury or bouillon-like taste. This basic or bouillon-like taste of yeast extracts makes them extremely suited for application in savoury-type foods, but hampers application outside the savoury food area, such as dairy-type foods where the introduction of bouillon notes is not desired. New developments in yeast extract production have resulted in yeast extracts with a very low bouillon-like taste, which makes application in special savoury-type applications, such as ketchup, possible. In addition, extracts with extremely low bouillon-like taste can be made and this feature makes application possible outside the savoury area, in products such as low-fat fruit drink yoghurts and light beverages. An overview of developments in yeast extracts is presented as well as an insight into the newly developed production route for high-nucleotide yeast extracts. In addition, the idea behind the composition of highnucleotide yeast extracts in relation to their application is discussed and results in special savoury applications and in a few applications outside the savoury area are discussed.

96

5.2

Modifying flavour in food

Developments in yeast extracts

Yeast is one of the most applied microbial organisms in food. Its original applications were found in ethanol production (beer and wine) and in baking bread. Besides live yeast, there has been significant interest in inactive yeast and, about 50 years ago, the first commercially produced yeast extracts entered the market. Initially, these yeast extracts were produced to take care of over-production of baker’s yeast. Later, these yeast extracts were recognised for their interesting components such as peptides and, amino acids, especially glutamic acid and B vitamins. The yeast extracts also provided a pleasant, savoury-base or bouillon-like flavour. The flavour of these yeast extracts is, determined by the peptide and amino acid composition, as well as the presence of typical Maillard reaction products (Weenan et al., 1997; Mlotkiewicz, 1998), such as 2-furanmethanethiol (roasty, coffeelike), 2-methyl-3-furanthiol (meat-like), 3-methylbutanal (malty) and methional (cooked potato) (Münch et al., 1997; Münch and Schieberle, 1998). All these attributes made yeast extracts useful for application in savoury foods, such as soups, sauces and snacks. The first available type of yeast extract was made by autolysis, i.e. a process where the yeast’s own enzymes play an important role in the liberation of the yeast cell content and in the degradation of proteins into peptides and amino acids. The yeast extract composition is determined by the yeast cell composition and the extraction process conditions. Yeast extracts, rich in proteins, peptides and amino acids, can best be made from yeast having a high protein content. A high-nucleotide extract for taste enhancement can best be made from yeast, rich in RNA. The extraction process can also be used to influence the composition of the yeast extract. If, for instance, a yeast extract with a high glutamic acid level (>6% glutamic acid) is desired, an elevated level of free glutamic acid in the yeast is essential and a high degree of protein hydrolysis, affected during the extraction process, is necessary. The yeast cell composition (see Table 5.1) can be influenced through strain selection, but also by changing the fermentation conditions and choice of raw materials (e.g. carbon source used for yeast fermentation). About 30 years ago, the first 5′nucleotide-containing yeast extracts entered the market. This kind of extract is usually made through a hydrolytic process, i.e. a process where the yeast’s own enzymes are inactivated prior to the extraction process and where the cells are opened up (lysed) by adding agents, such as enzymes. An example of this kind of yeast extracts is DSM’s Maxarome® Plus which contains the nucleotides 5′GMP and 5′IMP. This kind of yeast extract is particularly beneficial for taste enhancement owing to the presence of 5′nucleotides which create a taste-enhancing synergy with glutamic acid, forming a sensory attribute known as ‘umami’. This umami taste was discovered long ago in Japan by Ikeda in 1908 (Ikeda, 1908) and has more recently been recognised as the

New developments in yeast extracts for use as flavour enhancers

97

Table 5.1 The composition of baker’s yeast (Nagodawithana, 1995) Cell component

Cellular part

Amount (%)

β-Glucans Mannoproteins Chitin Lipids Proteins RNA (source of 5′ nucleotides) Minerals (ash) Vitamins (water soluble) Carbohydrates (trehalose, glycogen) Sterols Glutathione

Cell wall Cell wall Cell wall Cell wall Intracellular Intracellular Intracellular Intracellular Intracellular Intracellular Intracellular

6–12 6–12 1–3 6–7 45–65 5–10 4–6 <1 3–15 <1 <1

fifth taste sensation next to the four basic taste sensations, namely sweet, salt, bitter and acid. The synergy between 5′IMP and MSG (the monosodium salt of glutamic acid) was first determined by Yamaguchi (1967) and was later extended to 5′GMP by Komata (1990). This flavour synergy makes nucleotide-containing yeast extracts, which also contain glutamic acid, very powerful in taste-enhancement applications. Nucleotides originate from the RNA, present in the yeast cell, which is specifically broken down by the enzyme 5′-phosphodiesterase (5′FDase). This breakdown results in the nucleotides 5′GMP, 5′AMP, 5′CMP and 5′UMP. Of these four, only 5′GMP has taste-enhancement properties. In order to further increase the taste enhancement of this kind of yeast extracts, the 5′AMP is converted by the enzyme deaminase into 5′IMP, which also has taste-enhancing features. The nature of RNA and its hydrolysis into 5′nucleotides was studied for optimisation of its breakdown in nucleotidecontaining yeast extract production processes. Three different types of RNA exist in living cells, i.e. messenger-RNA (m-RNA), transfer-RNA (t-RNA) and ribosomal-RNA (r-RNA). Some properties of the different RNA types are relevant for RNA hydrolysis: •

•

•

m-RNA may be associated with DNA during the transcription process and may therefore be partially inaccessible to 5′FDase. However, the amount of m-RNA is rather small compared with the other types. In addition, m-RNA is rather unstable. t-RNA transfers amino acids to the ribosomes for protein synthesis. t-RNA may contain up to 10% ‘odd’ bases in addition to the normal bases A, G, C and U. r-RNA constitutes up to 65% of the weight of ribosomes in the yeast cell. r-RNA is associated with protein in the ribosomes. Ribosomes will dissociate when the Mg2+ concentration is reduced.

Some properties of the different types of RNA present in yeast (Woolford and Warner, 1991) are presented in Table 5.2. The properties of the different

98

Modifying flavour in food

Table 5.2

RNA types in Saccharomyces

RNA type

Molecular weight (kDa)

No. of nucleotides

25–1000 23–30 41–1200

75–3000 75–90 120–3400

m-RNA t-RNA r-RNA

Total cell RNA (%) 2 10 88

RNA types in yeast may differ somewhat from each other, but it is expected that their behaviour in RNA hydrolysis by 5′Fdase will be quite similar. In addition, the ratio and amount of the different RNA types in the yeast cell may vary with the growth rate of the yeast. Most conventional nucleotide yeast extracts have levels of 5′GMP + 5′IMP of 4–10% (sum of 5′GMP + 5′IMP). Owing to the inevitably high level of the bouillon-like taste of yeast extracts, these conventional nucleotide-containing yeast extracts could be used only in savoury-type applications. However, knowledge of the nature of RNA has revealed other possible production routes, which may lead towards higher nucleotide purities, lower levels of the inevitable basic or bouillon-like taste and consequently towards new application areas. We tested whether it was possible to separate high molecular weight RNA from other soluble components, released from the yeast cell after a hydrolytic yeast extraction process, by using ultrafiltration. This lab-scale test resulted in a nucleotide fraction having a fairly high purity. The resulting high-nucleotide fraction was used to make new yeast extract prototypes, which have been tested in some novel applications. In other words, this new development in yeast extract production focuses on providing taste enhancement and taste modification, without contributing the characteristic bouillon-like taste found in conventional yeast extracts. These newly developed high-nucleotide yeast extracts (14–20% 5′GMP + 5′IMP) have a clean to very clean bouillon-like taste profile. The taste-enhancing or flavour-modifying properties of these yeast extracts are very well expressed in products on the borderline of savoury foods.

5.3

Materials and methods

5.3.1 Yeast extract preparation A cream of baker’s yeast (Saccharomyces cerevisiae) was heat shocked to inactivate the endogenous yeast enzymes. Next, the yeast cells were opened up by a treatment with protease. The reaction mixture was ultra-filtered over a hollow-fibre membrane to remove the solubles. The retentate fraction, containing the yeast cell walls, also contained most of the RNA originally present in the yeast cells. The cell wall fraction was further incubated with a 5′phosphodiesterase preparation in order to hydrolyse the RNA into

New developments in yeast extracts for use as flavour enhancers

99

Table 5.3 Prototypes of high-nucleotide yeast extracts Prototype 1 2

5′GMP + 5′IMP (% on salt-free dry matter) 14 20

Salt (%) 20 <1

Bouillon-like taste Low Very low

nucleotides. The generated 5′AMP was further converted into 5′IMP by an incubation with deaminase. After this, the incubation mixture was ultrafiltered again over a hollow-fibre membrane. The resulting filtrate, having high-nucleotide purity, was concentrated and standardised. Finally, the concentrated yeast extract was spray dried. Using the described process, two different prototype yeast extracts (see Table 5.3) were made, which were tested in borderline savoury applications, a dairy application and an artificially sweetened beverage application.

5.3.2 Yeast extract application examples A high-nucleotide yeast extract containing 14% 5′GMP + 5′IMP (prototype 1) was used for testing in a range of products: (a) ketchup, (b) tomato pasta sauce, (c) reduced fat mayonnaise and (d) cheese spreads. This highnucleotide yeast extract (prototype 1) was tested by addition to the standard recipes of the products mentioned. Several dosage levels (between 0.01 and 0.1%) of the prototype 1 yeast extract were applied and after an equilibration time of at least 12 hours, the samples were assessed by an expert sensory panel. This equilibration time was necessary to obtain the optimal contribution of the yeast extract. The panellists defined the key sensory attributes necessary to describe the sensory profile of the products and also selected the dosage that gave the optimum flavour. The product with optimal dosage was then further assessed sensorially, by scoring the intensity of the key sensory attributes of these products on a categorised scale from 1 to 6 points (1 is low intensity, 6 is high intensity). The same procedure was carried out for the reference product, i.e. the product without high-nucleotide yeast extract. The panellists were asked specifically if they could detect a bouillon-like note in the different applications. A similar procedure as described above was followed for a yeast extract containing 20% 5′GMP + 5′IMP (prototype 2). As explained earlier, this prototype has almost no bouillon-like taste, is low in salt and is expected to have an impact on the sensory properties of products at only low dosage levels. This prompted us to test the flavour enhancement properties of this innovative yeast-derived high-nucleotide product at very low dosages (0.005–0.05%) in a low-fat fruit (strawberry) yoghurt drink and in a cola light beverage.

100

Modifying flavour in food

5.4

Results and discussion

5.4.1 Yeast extract preparation The process as described in Section 5.3 resulted in two different prototype yeast extracts having high nucleotide purity. The basic composition and properties of the two different prototypes are listed in Table 5.3. Prototype 1 is especially targeted towards less obvious savoury applications, such as tomato ketchup and tomato sauce. Prototype 2 is targeted towards applications outside the savoury area, i.e. applications with low salt levels and applications where introduction of the normally present basic or bouillon-like taste of yeast extracts is not desirable. Examples of this kind of applications are low-fat fruit drink yoghurts and light beverages.

Basic taste in extract

5.4.2 The effect of nucleotide purity on taste in food application The characteristic bouillon taste of yeast extracts is due to the various components in yeast extracts but the 5′nucleotides do not contribute. To demonstrate that this hypothesis was true, the bouillon-taste of different yeast extracts containing different amounts of the 5′nucleotides was determined using a sensory panel. The results in Fig. 5.1 show that, as the nucleotide content increases from 5 to 20%, the basic, bouillon taste decreases, thus confirming the hypothesis. Figure 5.1 clearly demonstrates the effect of nucleotide content on bouillon-like taste in yeast extracts, but, in food applications, the effect of the bouillon-like taste is rather different as less of the high-nucleotide material is required to deliver the taste enhancement effect and, consequently, the bouillon taste is minimal. Figure 5.2 shows the results of sensory testing to determine the basic, bouillon taste as a function of nucleotide content. It is clear that there is a dramatic decrease in the bouillon-like taste of yeasts extracts as the nucleotide content increases and shows that, at GMP + IMP contents above 20%, the effect is diminishing.

100 90 80 70 60 50 40 30 20 10 0

0

5

10

15

20

25

GMP + IMP content (%)

Fig. 5.1

The effect of the nucleotide concentration on the basic or bouillon-like taste of yeast extracts (taste intensity score from 1 to 100).

Basic taste in application

New developments in yeast extracts for use as flavour enhancers 100 90 80 70 60 50 40 30 20 10 0

0

5

10

15

20

101

25

GMP + IMP content (%)

Fig. 5.2 The effect of the nucleotide content on the basic or bouillon-like taste in application.

5.4.3

Application of high-nucleotide yeast extracts in less obvious savoury applications High-nucleotide yeast extract (prototype 1) was added to ketchup, tomato pasta sauce, reduced fat mayonnaise and cheese spreads and the flavour assessed using a sensory panel. The optimal level of addition was first determined, then the sensory properties of the optimised were determined and compared with the reference (standard) product using spider plots (Fig. 5.3–5.6). Ketchup The expert panel tested a range of dosages (0.01–0.1%) and concluded that the optimal dosage for ketchup was 0.02% of the high-nucleotide yeast extract. As is demonstrated in Fig. 5.3, even this extremely low dosage of

Length of aftertaste

Fullness mouthfeel

Sour flavour 6 5 4 3 2 1 0

Liquorice

Sweet flavour

Reference Salt

Enhanced

Tomato

Fresh

Fig. 5.3

The effect of 0.02% high-nucleotide yeast extract on the sensory flavour profile of ketchup compared with the standard reference product.

102

Modifying flavour in food

high-nucleotide yeast extract had an enormous impact on the taste profile of this typical product. As a result of the rather acidic conditions, the tomato-based products do not specifically show the savoury (umami) character. Where, traditionally, yeast extracts have functioned by delivering this savoury (umami) taste, we have now demonstrated other effects on taste. The panel results demonstrate that, in this low pH (3.7) product, sourness and freshness are more important attributes than umami or savoury. However, the addition of the high-nucleotide yeast extract changed sensory attributes not associated with savoury or bouillon flavour as shown by the increase in fullness (mouthfeel), sourness and length of aftertaste. In addition the overall flavour profile of the product was significantly influenced by the addition of this yeast extract prototype. It should be noted that during the testing of various dosages it was found that the high-nucleotide yeast extract can easily be over-dosed (>0.1%), decreasing the acid taste further, leaving a less fresh-tasting product. Tomato pasta sauce Addition of high-nucleotide yeast extracts had similar effects to those found in the ketchup. The fullness and length of aftertaste attributes increased and the sourness attribute was again modified. In contrast to the findings in the ketchup, the enhanced pasta sauce product was perceived to be more fresh, with a slightly higher dosage of 0.05% of prototype 1 in the final product (Fig. 5.4). Reduced fat mayonnaise The optimal high-nucleotide yeast extract dosage for this reduced fat mayonnaise was found to be 0.03%. A lower dosage (0.01%) made the product less creamy, where a higher dosage (0.1%) clearly enhanced the oily taste. Sour flavour 6 5 4

Length of aftertaste

3 2 1 0

Fullness mouthfeel

Sweet flavour

Salt

Reference Enhanced

Herbs flavour

Tomato

Bitter

Fresh

Fig. 5.4 The effect of 0.05% high-nucleotide yeast extract on sensory flavour profile of tomato pasta sauce compared with the standard reference product.

New developments in yeast extracts for use as flavour enhancers

103

Sour odour 6 Fullness mouthfeel

5

Egg odour

4 3 Creamy flavour

2

Oil odour

1

Reference

0

Enhanced

Fresh flavour

Sour taste

Egg flavour

Sweet taste Oil flavour

Fig. 5.5 The effect of 0.03% high-nucleotide yeast extract on flavour profile of reduced fat mayonnaise compared with the standard reference product.

This combined effect had a negative effect on the freshness of the mayonnaise. Still the optimised dosage (0.03%) clearly shows an additional benefit of high-nucleotide yeast extract besides increased fullness, i.e. creaminess was significantly improved (Fig. 5.5). These benefits are very well suited to the improvement of flavour in reduced fat products. Reduced fat cheese spread The reduced fat cheese spread is the most savoury of the tested products. Figure 5.6 shows that this savoury character was clearly intensified and expressed through elevated scores on the attributes fullness, cheese and mature taste. During the screening for the optimal dosage for cheese spreads, it was observed that the whole range, even up to 0.1% dosage, performed well. Higher dosages result in a stronger effect. Above 0.05%, a saturation effect on taste perception was observed, which prompted the panel to indicate 0.05% as the optimal dosage. The taste in the four food applications described above was not negatively influenced by the introduction of a noticeable amount of basic or bouillonlike taste, normally present in yeast extracts, owing to the low dosage levels and strongly reduced bouillon-like taste of this prototype. 5.4.4

Application of high-nucleotide yeast extracts in applications outside the savoury area The second part of the application experiment was focused on non-savoury applications, such as a commercial low-fat strawberry yoghurt drink and a cola light beverage. The high-nucleotide yeast extract, prototype 2, was used

104

Modifying flavour in food

Mature taste

Length of aftertaste

Salty flavour 6 5 4 3 2 1 0

Fullness mouthfeel

Sour flavour

Sweet flavour Reference Enhanced Fatty flavour

Bitter flavour

Creamy mouthfeel Cheese flavour

Fig. 5.6 The effect of 0.05% high-nucleotide yeast extract on flavour profile of reduced fat cheese spread compared with the standard reference product.

for these experiments. Sensory panel results from the assessment of the products containing an optimised dosage of high-nucleotide yeast extract added to commercial products are compared with the reference in the spider plots in Figs. 5.7 and 5.8. Low-fat strawberry yoghurt drink The expert panel tested a range of dosages (0.005–0.05%) and concluded that the optimal dosage for this high-nucleotide yeast extract prototype 2 was 0.01%. As is demonstrated in Fig. 5.7, even this extremely low dosage

Length of aftertaste

Sweet flavour 6 Sour flavour 5 4

Fullness mouthfeel

3 2

Fruity flavour Reference

1 0 Fatty mouthfeel

Chalky mouthfeel Creamy mouthfeel

Enhanced Strawberry flavour

Fresh flavour Yoghurt flavour

Fig. 5.7 The effect of 0.01% high-nucleotide yeast extract prototype 2 on the flavour profile of low-fat strawberry yoghurt drink compared with the standard reference product.

New developments in yeast extracts for use as flavour enhancers

Length of aftertaste

Fullness mouthfeel

Sweet flavour 6 5 4 3 2 1 0

Artificial flavour

105

Sour flavour

Reference Cola flavour

Enhanced

Refreshing flavour

Bitter taste

Fig. 5.8

The effect of 0.008% high-nucleotide yeast extract on the flavour profile of Cola light compared with the standard reference product.

of prototype 2 had an enormous impact on the taste profile of this product, particularly on mouthfeel attributes such as fullness, fattyness and creaminess. Applying the optimal, low dosage is very important. When too high dosages are used, they tend to have a negative effect on freshness and sourness. When choosing the optimal dosage, the idea was to cause little change in freshness and sourness, but to improve mouthfeel attributes. Fruity and strawberry types of attributes were also enhanced by this special yeast extract. Cola light A light beverage such as cola light was considered to be the most challenging of the applications tested. Here absolutely no bouillon-like taste introduction is desirable, so as to leave the cola taste unspoiled. However, in light beverages, an artificial taste, or bitter aftertaste, is often found owing to low sugar levels and the use of artificial sweeteners. The objective of this test was to improve these characteristics, without interfering with the other sensory properties of this commercial product. The optimal dosage level was determined as 0.008%, added to the commercial cola light. Figure 5.8 shows that, although flavour profile of the cola (e.g. the cola, refreshing or sourness attributes) did not change significantly, the artificial flavour character and bitter after taste were suppressed significantly.

5.5

Conclusions

A completely new approach to the production of yeast extracts has led to a new production process for the generation of high-nucleotide yeast

106

Modifying flavour in food

extracts with up to 20% 5′GMP + 5′IMP. As a consequence, these highnucleotide yeast extracts possess less basic or bouillon-like taste, normally present in yeast extracts (nucleotides have almost no taste) and can be used at lower dosage levels, further reducing any bouillon-like savoury flavour introduced with these yeast extracts in food products. The novel aspect of the applications is that the high-nucleotide yeast extracts show powerful interactions with other non-savoury, flavour modalities. Flavour improvements in products through increased ‘fullness’, ‘creaminess’ or ‘freshness’ or a decrease of ‘bitter aftertaste’ when high-nucleotide yeast extracts are added are unexpected, yet very welcome, to flavourists trying to reformulate products that have had their fat content reduced or that contain artificial sweeteners. These new yeast extracts therefore offer the possibility of improving flavours in a wide range of savoury and non-savoury applications and further research will no doubt discover other uses for these natural food ingredients.

5.6

References

ikeda, k. (1908), Japanese patent 4805. komata, y. (1990), ‘Umami taste of seafoods’, Food Revi Int, 6 (4), 457–487. mlotkiewicz, j. a. (1998), ‘The role of Maillard reaction in food’, in O’Brien, J., Nursten, H. E., Crabbe, M. J. C. and Ames, J. M., The Maillard Reaction in Foods and Medicine, Cambridge, The Royal Society of Chemistry, 19–28. münch, p. and schieberle, p. (1998), ‘Quantitative studies on the formation of key odorants in thermally treated yeast extracts using stable isotope dilution assays’, J Agric Food Chem, 46, 4695–4701. münch, p., hofmann, t. and schieberle, p. (1997), ‘Comparison of key odorants generated by thermal treatment of commercial and self-prepared yeast extracts: influence of the amino acid composition on odorant formation’, J Agric Food Chem, 45, 1338–1344. nagodawithana, t. w. (1995), Savoury flavours, Wisconsin, Esteekay Associates, Inc. weenen, h., kerler, j. and van der ven j. g. m. (1997), ‘The Maillard reaction in flavour formation’, in Swift, K. A. D., Flavours & Fragrances, Cambridge, The Royal Society of Chemistry, 153–171. woolford, j. l. and warner, j. r. (1991), ‘The ribosome and its synthesis’, in Broach, J. R., Pringle, J. R. and Jones, E. W., The Molecular and Cellular Biology of the Yeast Saccharomyces: Volume 1: Genome Dynamics, Protein Synthesis, and Energetics, New York, Cold Spring Harbor Laboratory Press, 587–626. yamaguchi, s. (1967), ‘The synergistic taste effect of monosodium glutamate and disodium 5′-inosinate’, J Food Sci, 32, 473–478.

6 Chiral chemistry and food flavourings S. Serra, C.N.R. Milano, Italy

6.1

Introduction

The flavour of a food or beverage is the overall sensation that arises by interaction of taste, odour and structural feeling during consumption. Some substances are responsible only for taste sensation and are generally nonvolatile at room temperature. This is the case for different sugars, flavours enhancers, bitter and salty compounds. Volatile flavours contribute mainly to the odour sensation, but can sometimes contribute to the taste sensation (e.g. menthol). Attention is focused here on the aroma substances since they are the most relevant in flavouring processes. Since the 1960s, the improvement of analytical and synthetic knowledge has allowed the isolation, characterisation and preparation of an impressive number of aroma compounds. At an industrial scale, many of these compounds have found widespread application in food, beverages, cosmetic, and pharmaceutical products. The worldwide turnover of fragrances and flavours was estimated at around US$17 billion in 2004 (http://www.leffingwell.com/top_10.htm): half of this amount was covered by the flavours market. If we consider the latter data, the relevance of flavour in food manufacture is unquestionable. Moreover, modern production of aroma is oriented to the use of highquality ingredients. Consumers are demanding ‘natural’-like formulations and this goal is obtainable by accurate reconstruction of the required flavour. Therefore, the chemical study of the so-called ‘character impact compounds’ has become a fundamental part of flavour research. Since flavours are often present in nature as a complex mixture of isomers, identification and sensory evaluation of each component are necessary. In this context, chiral flavours play a primary role. It is known (Brenna et al., 2003)

108

Modifying flavour in food

that the enantiomeric composition of an odorous compound greatly affects its aroma properties in terms of either its perceived aroma or its odour threshold. Consequently, only one enantiomer, or a balanced combination of the enantiomers, might be used in aroma formulation. Otherwise, preparation of enantio-enriched compounds needs asymmetric synthesis or the use of optical active staring materials, greatly increasing the cost of production. Therefore a question arises: when should chiral flavours be used? Economic, chemical and marketing considerations have affected the industrial production. The cost and availability of the enantio-enriched flavours (Section 6.2) and the feasibility of the chemical processes (Section 6.3) should be economically supported by the needs of consumers. Very few relevant industrial processes have actually produced enantio-enriched flavours. Although a great number of asymmetric syntheses of relevant key aroma ingredients have been reported (Section 6.4), their production is still restricted to the racemic form. In spite of this fact, the major manufacturers of flavours and fragrances (Section 6.5) have extensively studied the enantioselective preparation of aroma compounds, clearly demonstrating that this approach will be important and successful in the future.

6.2

Chiral flavour – synthesis

6.2.1 General The availability of chiral flavours is strictly dependent upon their origin. The compounds could be accessible in the following ways: isolation from natural resources, synthesis from natural chiral precursors, resolution of racemic material, asymmetric synthesis and biotransformation. The most cost-effective path is often used in industrial production, though, for some flavours, different processes are equally convenient. Moreover, a great number of key aroma compounds were found in nature in minute amounts or as trace components. Therefore, extraction is not a suitable or realistic approach and only the synthetic or microbiological pathways allow their production. It is noteworthy to realise that the origin of flavours also affects their commercial values. Recent US (Food and Drug Administration, 1985) and European (Council of the European Communities, 1988) legislations have officially stated that ‘natural’ flavour substances can only be prepared either by physical processes (extraction from natural sources) or by enzymatic or microbial processes, which involve precursors isolated from nature. This classification has created a dichotomy in the market because compounds labelled as ‘natural’ become profitable products, whereas other flavours that occur in nature but are obtained through chemical methods must be indicated as ‘nature identical’ and are less appreciated by the consumers. From a chemical point of view, there is no difference between a compound synthesised in nature and the identical molecule produced in the laboratory, but still the price of a flavour sold as natural is often significantly higher than those prepared by chemical synthesis.

Chiral chemistry and food flavourings

109

6.2.2 Flavours extracted from natural sources A great number of chiral flavours are extracted from plants. Mono- and sesquiterpenes are the most relevant ones. The terpene hydrocarbons, which occur as constituents of several essential oils, are of marginal odour value but their abundance and low cost make them ideal starting products for the preparation of nature-identical odorants (Section 6.3). The most relevant source for the latter hydrocarbons is crude pine resin (Table 6.1). Steam distillation of this raw material yields 20–25% of an essential oil known as turpentine. Chemical composition of turpentine depends upon the Pinus species extracted. The main constituents are (+)-(I) and (−)-α-pinene (II), (−)-β-pinene (III) and (+)-3-carene (IV) (Fig. 6.1). Further relevant chiral hydrocarbons are the p-menthane terpenes, limonene and phellandrene that occur abundantly in many essential oils. (+)-Limonene (V) is present in citrus peel oils at a concentration of over 90%. Otherwise, the (−)-isomer is found in conifers and Mentha species. (−)-α-Phellandrene (VII) is isolated from eucalyptus oil whereas the β isomers (VIII) and (IX) are available from water fennel oil and Pinus species, respectively. All the latter hydrocarbons are used as starting materials for flavours and fragrances preparation or for reconstruction of essential oils. Oxygenated terpenes (X)–(XXIV) are nevertheless important chiral building blocks but they are also used directly in flavouring processes. (−)-Menthol (XI), (+)pulegone (XII) and (−)-carvone (XIV) are isolated from different Mentha species and are used in a number of foods, beverage and oral hygiene products as well as (+)-carvone (XIII) and (−)-piperitone (XV). Otherwise, owing to their characteristic penetrating odour, (+)-camphor (XVI) and the borneol isomers (XVII) and (XVIII) are used mainly for fragrance formulations or in the reconstruction of the essential oils in which they occur naturally. Both linalool isomers (XIX) and (XX) are used to impart fruity and flowery notes to flavours and fragrances whereas citronellol isomers (XXI) and (XXII) are used in citrus compositions. Lastly, citronellal isomers (XXIII) and (XXIV) are used in a limited extent for flavouring since their are mainly employed as a starting material for the production of isopulegol and citronellol. Many chiral flavours are available in both enantiomerich forms. In some cases, for example β-phellandrene and citronellal, the odour features of both enantiomerse are identical. However, for example limonene and carvone, the enantiomeric composition has a major effect on the overall flavour sensation, thus allowing a number of formulations with different sensory properties.

6.2.3 Flavours found in nature and prepared by different methods The majority of the chiral flavours available on the market are not obtained directly from natural sources. They are usually prepared by a variety of pathways depending upon the specific characteristic of each flavour. The use of racemic or enantio-enriched starting materials of extractive or synthetic origin may be combined with different kinds of chemical

110

Table 6.1 Flavours extracted from natural sources Source

Odour

(+)-α-Pinene (I) (−)-α-Pinene (II) (−)-β-Pinene (III) (+)-3-Carene (IV) (+)-Limonene (V) (−)-Limonene (VI) (−)-α-Phellandrene (VII) (+)-β-Phellandrene (VIII) (−)-β-Phellandrene (IX) (−)-α-Terpineol (X) (−)-Menthol (XI) (+)-Pulegone (XII) (+)-Carvone (XIII) (−)-Carvone (XIV) (−)-Piperitone (XV) (+)-Camphor (XVI) (+)-Borneol (XVII) (−)-Borneol (XVIII) (−)-Linalool (XIX) (+)-Linalool (XX)

Turpentine oil. Pinus halepensis Turpentine oil. Pinus pinaster Turpentine oil. Pinus ponderosa needle oil Turpentine oil of Pinus sylvestris Orange oil. Citrus sinensis White spuce oil. Abies alba Eucaliptus oil. Eucalyptus sp. Water fennel oil. Phellandrium aquaticum Needle oil of Pinus contorta Turpentine oil. Eucalyptus oil Cornmint oil. Mentha arvensis Pennyroyal oil. Mentha pulegium Caraway oil. Carum carvi Spearmint oil. Mentha spicata Eucalyptus dives Camphor oil. Cinnamomum camphora Camphor oil. Siberian fir needle oil. Abies sibirica Ho oil. Cinnamomum camphora Oil of coriander. Coriandrum sativum

Harsh, terpene-like, minty Harsh, terpene-like, coniferous

(−)-Citronellol (XXI) (+)-Citronellol (XXII) (−)-Citronellal (XXIII) (+)-Citronellal (XXIV)

Geranium and rose oils. Different species Oil of Boronia citriodora and Eucalyptus citriodora Backhousia citriodora Citronella oil. Cymbopogon winterianus

Fresh citrus, orange-like Harsh, turpentine-like, lemon note Citrus, slight peppery note Mint, terpentine Coniferous odour, character tarry Very cooling, fresh, sweet, minty Peppermint, camphor, fresh herbal Caraway, fresh herbal Sweet spearmint, fresh herbal Herbal, minty, camphoraceous Camphoraceous Camphoraceous, earthy-peppery note Camphoraceous and woody odour Floral, woody lavender note Sweet, floral; odour reminiscent of petitgrain and lavender Floral, rose-like odour, reminiscent of geranium oil Citronella oil-like Powerful, fresh, herbaceous-citrus

Modifying flavour in food

Flavour

Chiral chemistry and food flavourings

(+)-δ-3-Carene (IV)

(+)-α-Pinene (–)-α-Pinene (–)-β-Pinene (I) (II) (III)

(+)-Limonene (–)-Limonene (V) (VI)

(–)-α-Phellandrene (+)-β-Phellandrene (–)-β-Phellandrene (VII) (VIII) (IX)

O OH OH (–)-α-Terpineol (X)

(–)-Menthol (XI)

O O

O

(+)-Pulegone (XII)

O

HO

HO

(+)-Camphor (XVI)

(+)-Borneol (XVII)

(–)-Borneol (XVIII)

HO

111

(+)-Carvone (–)-Carvone (–)-Piperitone (XIII) (XIV) (XV)

OH

OH (–)-Linalool (+)-Linalool (XIX) (XX)

Fig. 6.1

(–)-Citronellol (XXI)

OH (+)-Citronellol (XXII)

O

O (–)-Citronellal (XXIII)

(+)-Citronellal (XXIV)

Some relevant flavours extracted from natural sources.

transformation. Racemic, diastereoselective or enantioselective synthesis are used as well as biotransformations. A few, selected examples on the latter topics may be illustrative of the multiplicity of the possible problems. Flavours available from natural sources, prepared by asymmetric synthesis (−)-Isopulegol (XXV) (Fig. 6.2) occurs in a large number of essential oils, often in optically active form. Since citronellal readily cyclises to isopulegol, the latter is frequently found in citronellal-containing essential oils. Although its extraction is possible, the isopulegol flavour is usually prepared

112

Modifying flavour in food

HO O

OH

O

O

OH (XXVII)

(–)-Isopulegol (XXV)

(–)-cis Rose oxide (XXVI) O

(XXVIII)

Linalool oxide

O

O α-Ionone (XXIX)

S

O

α-Damascone (XXX)

S

O R

(XXXII)

Fig. 6.2

(S)-Filbertone (XXXI)

(XXXIII)

O (XXXIV)

O

R

O

R = aliphatic O alkyl chain

(XXXV)

Some relevant flavours found in nature and prepared by different methods.

by asymmetric synthesis. In effect (XXV) is one of the key intermediates in the preparation of (−)-menthol (Section 6.3) and is obtained in large amounts and very high enantiomerical purity by the Takasago process (Akugatagawa, 1992). Moreover, recent investigations (Yamamoto, 1998) have shown that (−)-isopulegol with e.e. (enantiomeric excess) up to 99.7 is odourless and can be used as a cooling agent. As a consequence, the main part of optically active (XXV) on the market is prepared by enantioselective synthesis. Flavours obtained from natural chiral precursors by a non-regioselective pathway e.g. (−)-cis rose oxide (XXVI) and the isomeric linalool oxides (XXVII) and (XXVIII), occur in small quantities in rose oil and fruit aromas, respectively (Wüst and Mosandl, 1999). They cannot be conveniently obtained by extraction and are prepared by a non-regioselective synthesis from optically active starting materials (Fig. 6.3). Since the (−)isomers of rose oxide are naturally present in rose oil and cis-(XXVI) is the most powerful stereoisomer in aroma terms, the easily available (−)citronellol is used as starting material (Bauer et al., 2001) (Fig. 6.1). Photooxidation of (XXI) with oxygen and visible light give a mixture of hydroperoxides that were reduced to diols (XXXVI) and (XXXVII). Treatment with diluted sulphuric acid resulted in allylic rearrangement and spontaneous cyclisation of isomer (XXXVII) to afford the diastereoisomeric mixture of enantiopure rose oxide (XXVI) and (XXXVIII). Similarly,

Chiral chemistry and food flavourings

O2, hν HO

reduction

OH

HO

113

H2SO4

HO

O

O

OH (–)-Citronellol (XXI)

(XXXVI)

(XXXVII)

HO HO O OH (R)-(XXVII)

(R)-(XXVIII)

OH

O OH (S)-(XXVII)

(–)-Linalool (+)-Linalool (XIX) (XX)

LiAlH4 OH

O

(S)-(XXVIII)

O

(XLI)

NH2CSNH2, red Al OH

(+)-Tartrate

O

MnO2 OH

(XL)

(S)-(XXXIX)

HO

Peracid

K2Cr2O2 LiC⬅CMe

(–)-trans rose oxide (XXXVIII)

OH

Peracid O

(–)-cis rose oxide (XXVI)

(XLIIIa)

(S)-Filbertone (XXXI)

SH OH (XLIVa)

S

O

(+)-(XXXII)

OH (XLII) (–)-Tartrate

O

NH2CSNH2, red Al

SH

OH

OH (XLIVb)

(XLIIIb) SH OH

S

O

(–)-(XXXII)

CH2CHO, H+ rac-(XXXII)

(XLIV)

Fig. 6.3

The synthesis of rose oxide, linalool oxide, filbertone and 2-methyl4-propyl-1,3-oxathiane.

both regioisomers (XXVII) and (XXVIII) occur in nature as a diastereoisomeric mixture with an absolute configuration that depends on the plant origin (Wüst and Mosandl, 1999). Accordingly, treatment of (−)-(R)-linalool and (+)-(S)-linalool with peracids gave a mixture of (R)-(XXVII), (R)(XXVIII) and (S)-(XXVII), (S)-(XXVIII), respectively (Bauer et al., 2001). The preparation of (XXVII)/(XXVIII) mixtures in racemic or (R), (S) enantio-enriched forms could be evaluated on the basis of the specific aroma preparation.

114

Modifying flavour in food

Flavours used in racemic form α-Ionone (XXIX) and α-damascone (XXX) are natural norterpenoids with pronounced violet and rose scents, respectively. They are present in nature in minute amounts and their extraction is very expensive. The industrial preparation of ionone and damascone isomers is strictly related to carotenoid synthesis in which the same starting materials should be used. Therefore, the commercial racemic compounds (XXIX) and (XXX) are very cheap and are used mainly in perfume compositions. Although resolution of synthetic materials and some enantioselective synthesis have been reported (Brenna et al., 2002), their production is still restricted to racemic compounds. Flavours used both in racemic and enantio-enriched form Filbertone (XXXI) is the principal flavour component of filberts (hazelnuts) where it occurs in low concentration. The sensory evaluation (Jauch et al., 1989) of the enantiomeric forms, established a marked difference between the isomers showing that (S)-(XXXI) is about ten times more powerful. Natural filbertone exhibits only rather low enantiomeric excess, varying with the origin of the hazelnuts, with the prevalence of (S)-isomers. Its enantiospecific synthesis is possible (Jauch et al., 1989) by the use of enantiopure (S)-(−)-2-methylbutan-1-ol (XXXIX) available from the chiral pool. Oxidation of alcohol (XXXIX) to the corresponding aldehyde, followed by the addition of the lithium salt of propyne, gave propargylic alcohol (XL). The trans regioselective reduction of the triple bond was achieved by meana of LiAlH4 whereas the MnO2 oxidation of the obtained allylic alcohol gave the desired (S)-filbertone. Since natural filbertone is not enantiopure and alcohol (XXXIX) is rather expensive, the latter procedure is applied both to the preparation of racemic and enantio-enriched flavours. Flavour prepared in racemic and diastereoisomerically enriched form 2-Methyl-4-propyl-1,3-oxathiane (XXXII) and (XXXIII) was isolated (Winter et al., 1976) as a 10:1 isomeric mixture from the juice of the yellow passion fruit (Passiflora edulis). The latter compounds are responsible of the typical odour of passion fruit and are also suitable for a wide range of tropical fruits, grapefruit, citrus and fruity flavour formulations. The (+) and (−)- cis isomers (XXXII) are the most powerful odorants, with threshold values in water of 2 and 4 ppb, respectively. In order to investigate the olfactory features of each enantiomer, their enantioselective synthesis was reported (Pickenhagen and Brönner-Schindler, 1984). The starting material was (E)-2-hexen-1-ol (XLII) that was converted into epoxides (XLIIIa) and (XLIIIb) by Sharpless epoxidation using (+) and (−)-tartrate, respectively. Transformation of oxirane to tia-derivative and the following reductive ring opening, afforded (R)- and (S)-3-mercapto-hexanol (XLIVa) and (XLIVb), that reacted with acetaldehyde to give enantiopure (+)- and (−)-

Chiral chemistry and food flavourings

115

(XXXII), respectively. The olfactory evaluation demonstrated that only (+)-(XXXII) possessed the typical sulphur notes of tropical fruits, whereas (−)-(XXXII) was slightly camphoraceus, more woody, and not very typical. In spite of the latter facts, the above-mentioned oxatianes are not prepared in optical active forms. This is due to the high complexity of enantioselective synthesis and to a recent study that underlines the presence of (−)-(XXXII) in passionfruits (Singer et al., 1986). In addition the compound (XXXII) was prepared by reaction of racemic 3-mercapto-hexanol with acetaldehyde that gave mainly the suitable cis isomer (XXXII). Flavour prepared by biotransformation in racemic or enantio-enriched form Lactones of type (XXXIV) and (XXXV) (Fig. 6.2) are well-known naturally occurring chiral flavour compounds. Whereas γ lactones (XXXIV) are important aroma components of many fruits, δ lactones (XXXV) are widespread in fermented food, milk products and in a variety of dairy products (Dufossé et al., 1994). Both γ- and δ-lactones occur in minute amounts and their extraction is not suitable for commercial production. Specific and characteristic enantiomeric ratios have been reported for different natural lactones (Brenna et al., 2003). Although the sensory properties of the enantiomers of the latter compounds are somewhat different, the majority of nature-identical lactones are prepared in racemic form by a number of lowcost starting materials. Otherwise, natural lactones are very expensive and different biotechnological methods have been developed for their synthesis (Serra et al., 2005). Some of these materials are manufactured by degradation of natural hydroxy-fatty acids, via β-oxidation. Specifically, the relevant γ-decalactone (peach flavour) is obtained by chain shortening of C-18 ricinoleic acid (from castor oil) by different microorganisms. Improvement of the processes caused the selling price of the latter compound to decrease from US$12 000/kg in 1986 to US$500/kg in 1998. Similarly, some precious γ-lactones containing an odd number of carbon atoms are accessible by degradation of natural hydroxy acids. Interestingly, δ-decalactone can be obtained by natural modification, either by oxidation of hydroxy-fatty acids or by enzymic reduction of the α,β-unsaturated compound (massoia lactone) the main component of massoi bark oil. For all the processes mentioned above, the lactones can be obtained in racemic or enantio-enriched form, depending on the starting materials and the microrganism/enzyme used in the transformation.

6.3

Chiral flavour

6.3.1 General As mentioned in the previous paragraphs, the preparation of chiral flavours is based on the following synthetic pathways: chemical transformation of

116

Modifying flavour in food

HO Oxidation OH Pyrolysis

Pyrolysis

(LI) (XX)

(XLV) Reduction

Hydration

HO

HO Pyrolysis

(XIX)

(–)-α-Pinene (II)

Oxidation

(XXII) Ziegler procedure

OH

Pyrolysis

Pyrolysis

Pyrolysis

(XLIX)

(XLVI)

(LIII)

(XLVII) Reduction

Pyrolysis

Oxidation

(XX) Hydration

(+)-α-Pinene HO (I) (XLVIII)

(L) Ziegler procedure

(LII)

OH

OH

HO

Pyrolysis

(LIV) Oxidation

(XIX)

HO (XXI)

Fig. 6.4

The pinene pathway to enantio-enriched flavours.

extractive chiral compounds, asymmetric synthesis, resolution of a racemic flavour and biotechnological transformation. Although a great number of syntheses of chiral flavour have been reported, the majority of them were of academic interest or had a limited number of applications. Therefore, in this chapter, we focus our attention on a few, very relevant, processes that are used for the industrial production of flavour materials. The synthesis of flavour from chiral, extracted materials is well illustrated by the description of pinene and limonene transformations. The account of menthol synthesis in Section 6.3.3 is descriptive of all the synthetic approaches cited above.

6.3.2 Synthesis of key flavours from natural pinene and limonene Turpentine oil is the most abundant essential oil produced worldwide. Similarly, terpene hydrocarbons from citrus species are side products from orange juice production. Therefore, their main components α-pinene and limonene are very cheap and are the ideal chiral building block for flavour synthesis. α-Pinene is used in the industrial production of linalool, citronellol, borneol and camphor (Ohloff, 1994). The diastereoselective hydrogenation (Fig. 6.4) of (−)- and (+)-α-pinene affords pinane (XLV)/(XLVI) and (XLVII)/(XLVIII), respectively. The pyrolysis of the latter pinane, above 385 ºC, leads to the formation of 3,7-dimethylocta-1,6-diene. In line with Fig. 6.4, pyrolysis of diastereoisomeric hydrocarbons (XLV) and (XLVII) gave (−)-(3R)-(XLIX), whereas under similar conditions, compounds (XLVI) and (XLVIII) gave (+)-(3S)-(L). According to the Ziegler alcohol synthesis, enantiomeric 3,7-dimethylocta-1,6-diene (XLIX) and (L) are treated with

Chiral chemistry and food flavourings HCl

Cl

(–)-α-Pinene (LV) (II) CI NOCl

O

Base

O

Oxidation

(XVII)

NO

(+)-Limonene (LVIII) (V)

Fig. 6.5

HO

–

OH

OH–

Cl

(+)-Camphor (–)-Camphor (XVIII) (XVI) (LVII)

NOH+

H /H2O

(LX)

HO

Oxidation

O

O H+/H O

(–)-Carvone (+)-Carvone (XIV) (XIII)

2

(LXI)

117

HCl

(LVI) (+)-α-Pinene (I) NOH

Base

Cl

NO

NOCl

(LIX) (–)-Limonene (VI)

The preparation of camphor and carvone from pinene and limonene.

aluminium diisobutyl hydride. The latter reagent gives regioselective addition, in an anti-Markovnikov fashion, to the least substituted double bonds of the enantiomeric dienes. The resulting organo-aluminium derivatives are oxidised with air and then saponified with water. Accordingly, (−)-(3R)(XLIX) and (+)-(3S)-(L) gave (+)-citronellol (XXII) and (−)-citronellol (XXI), respectively. The above-mentioned pinanes are also intermediates in linalool synthesis. The oxidation by air under radical conditions of the four enantio-enriched pinanes (XLV)–(XLVIII) gave the corresponding 2-hydroperoxides, which were then reduced by hydrogenation to pinan2-ols (LI)–(LIV), respectively. The latter alcohols are also obtainable directly from (−)- and (+)-α-pinene by diastereoselective hydration of the double bond. Pyrolysis of alcohols (LI) and (LIII) gave (+)-linalool (XX), whereas, under similar conditions, compounds (LII) and (LIV) gave (−)linalool (XIX). The acid-catalysed Wagner–Meerwein rearrangement of pinene is a useful reaction for the synthesis of camphor derivatives. According to Fig. 6.5, anhydrous HCl converts (−)- and (+)-α-pinene into (−)- and (+)-bornyl chlorides (LV) and (LVI), respectively. Complete saponification of the latter compounds provided the borneol enantiomers (+)-(XVII) and (−)-(XVIII) that were oxidised to camphor enantiomers (+)-(XVI) and (−)-(LVII). Limonene is a further important chiral starting material that has been used for a number of industrial applications. One very relevant process leading to chiral flavours, starting from limonene is the case of carvone production (Pybus and Sell, 1999). (−)-Carvone is the principal component of spearmint oil and both natural and synthetic flavours are used as ingredients in mint formulation. The enantiomeric composition is fundamental since (+)-enantiomer has an odour reminiscent of caraway and dill rather than spearmint. The synthesis is based on the regioselective addition of the nitrosyl cation to the more electron rich endocyclic double bond of limonene. Consequently, (+) or (−)-limonene are treated with nitrosyl chloride to afford the derivatives (LVIII) and (LIX), respectively. The latter compounds are not isolated but treated with base to give elimination of hydrogen chloride and rearrangement of the nitrosyl group. The obtained oximes

118

Modifying flavour in food

m-Cresol propene

Six + isomers Fractionation

Conversion to benzoate OH

Crystallisation

Mother OCOPh liquor (LXIV)

OCOPh (LXIII)

Rac. menthol Lipase hydrolysis

Hydrolysis

Reduction OH Thymol (LXII)

OH

Epimerisation

Eight isomers Lipasemediated Seven + acetylation isomers

(+)-Menthol (LXV)

Freezing Mentha arvensis oil

OH (–)-Menthol (XI)

O O (LXVII)

Hydrolysis Reduction

Cyclisation

Asymmetric synthesis

OH

O

(–)-Isopulegol (XXV)

(+)-Citronellal (XXIV)

Myrcene (LXVI)

Fig. 6.6 Industrial production of (−)-menthol. Dashed arrows indicate Haarmann & Reimer, extractive and Takasago processes. Dotted arrows give the new biocatalytic processes of Haarmann & Reimer and AECI.

(LX) and (LXI) are then hydrolysed with acid to give (−)- and (+)-carvone, respectively.

6.3.3 Synthesis of (-)-menthol Menthol is one of the most important flavours and it is used extensively as a food additive, in pharmaceuticals, cosmetics, toothpastes, oral-care products, tobacco and chewing gum. This compound is broadly used for its mint taste and odour and especially for its cooling effect. The desired organoleptic properties of this monoterpene are related to its absolute configuration and, from the eight possible isomers, only the natural (−)-(1R,3R,4S) isomer is suitable as a flavouring. In 1998, the estimated world production of menthol was 11 800 tonnes (Clark, 1998). The majority of (−)-menthol is still obtained by freezing the oil of Mentha arvensis to crystallise the menthol present (Fig. 6.6). Since the main part of the extractive menthol is produced in the Peoples’ Republic of China, its price and availability fluctuate as a result of the season variability and of the market supply and demands. Therefore, many efforts have been devoted to the production of (−)-menthol from other readily available raw materials. The Haarmann & Reimer and the Takasago processes are the most competitive commercial synthetic sources of this flavour (Pybus and Sell, 1999). The first is based on a classical resolution procedure and starts from inexpensive m-cresol and propylene to produce thymol (LXII). This compound is hydrogenated to give a mixture of the eight isomers of menthol. The physical properties of the diastereoisomeric forms (menthol, neomenthol, isomenthol, neoisomenthol) are different, allowing their separation. Fractional distillation gives racemic menthol (two isomers) which is converted to the corresponding benzoate (LXIII). The latter ester is resolved

Chiral chemistry and food flavourings

119

by fractional crystallisation, seeding the racemic material with enantiopure benzoate of (−)-menthol. The obtained (LXIV) is saponified to produce (−)menthol (XI) whereas the mother liquor gives (+)-menthol (LXV). The (+)-menthol and the other six isomers are recycled in a separate racemisation step. Takasago uses asymmetric synthesis (Akugatagawa, 1992) in the key step of its process. Myrcene (LXVI) is converted into diethylgeranylamine, which is isomerised to (+)-citronellal (XXIV) in the presence of a chiral rhodiun phosphine catalyst (RhI-(S)-BINAP). The transformation of citronellal into (−)-menthol is performed through cyclisation to (−)-isopulegol (XXV) followed by hydrogenation. Two new processes, similar to the first described route, but based on lipase resolution, have been recently proposed. Haarmann and Reimer accomplished the resolution of racemic menthol benzoate (LXIII) by lipase-mediated (e.g. Candida rugosa) enantioselective hydrolysis to provide (−)-menthol with essentially complete enantioselectivity (Serra et al., 2005). Furthermore, the AECI Ltd process starts directly from the mixture of the eight isomers of menthol. The enantioand diastereoselective acylation of this mixture using lipases yields menthyl acetate (LXVII) in at least 96% enantiomeric excess. The ester is separated from the unreacted isomers by distillation and then hydrolysed to yield (−)menthol. In both processes, the undesired isomers are recycled by isomerisation. Although neither the biocatalytic H&R or AECI process have yet been commercialised, these means are based on the well-established route of racemic menthol preparation and have the potential for future production. The amount of (−)-menthol prepared by the processes described above accounts for the majority of the worldwide production. In addition, some different, minor processes afford (−)-menthol starting from terpenes of extractive origin. Although they are not really competitive from a quantitative point of view, the particular availability of the raw materials or the local economic conditions can make them effective. For instance, (−)-βphellandrene (IX) and (+)-δ-carene (IV) (Fig. 6.7) occur abundantly in some turpentine oils of specific plant species. The conjugated dienes (IX) are treated (http://www.leffingwell.com/menthol-1/menthingo.htm) with hydrogen chloride to give phellandrene hydrochloride (predominantly, 1chloro-2-menthene) (LXVIII) as a mixture of diastereoisomers which can be converted to a mixture of optically active piperityl acetates by allylic displacement with sodium acetate in acetic acid. Hydrolysis affords a separable diastereoisomeric mixture of enantio-enriched trans and cis piperitol (LXIX) and (LXX). (+)-trans-Piperitol (LXIX) could be hydrogenated with high selectivity to obtain (+)-isomenthol (LXXII) (99%) containing a small amount of (+)-menthol (LXXI)(1%) whereas (−)-cis-piperitol (LXX) gave a mixture of (+)-neoisomenthol (LXXIII) and (−)-neomenthol (LXXIV). After separation, (+) (LXXII) and (+)-(LXXIII) can be isomerised into (−)-menthol (XI) using aluminium isopropoxide as a catalyst. The

120

Modifying flavour in food Cl Reduction

HCl +

HO (–)-β-Phellandrene (IX)

HO

(LXIX)

O

(LXX)

(–)-Piperitone (XV)

(LXVIII) Reduction

(IV)

HO

(+)-δ-3-Carene

HO (LXXI)

(LXXII)

HO

HO (LXXIII)

O

(LXXIV)

(+)-Pulegone (XII)

Isomerisation

Isomerisation

Reduction HO

O Oxidation (LXXV)

(LXXVI)

H2

Reduction

cat. HO (LXXVIII)

HO (LXXVII)

(–)-Menthol (XI)

Heat

O

+ O (LXXXII) Epimerisation

(LXXXIII)

Fractionation Isomerisation (LXXIX)

Fig. 6.7

H2 cat. (LXXX)

Oxidation (LXXXI) O

(LXXXII)/(LXXXIII)

Industrial production of (−)-menthol from natural resources.

same key steps are used in the production of (XI) from (−)-piperitone (XV) (Pybus and Sell, 1999). The latter compound is a major constituent of the oil of Eucalyptus dives produced in Australia at low cost. The reduction of (XV) using LiAlH4 gave again a mixture of (+)-trans-piperitol (LXIX) (64%) and (−)-cis-piperitol (LXX) (36%) that was processed as described above. (+)-δ-3-Carene (IV) provides another monoterpene raw material for the potential production of (−)-menthol (Pybus and Sell, 1999). Catalytic isomerisation provides the needed (+)-δ-2-carene (LXXV) required for either of two synthetic routes. In one approach, reaction of (+)-(LXXV) with peracid yields (+)-cis-2,8-p-menthadien-l-ol (LXXVII) via the intermediate cis-2-carene epoxide (LXXVI). Allylic displacements with acetic or formic acid in a buffered solution affords a mixture of trans- and cis-isopiperitenyl acetates (or formates) which on hydrolysis gave mainly (−)-transisopiperitenol (LXXVIII) contaminated by the corresponding (+)-cisisomer. These alcohols are separated by fractional distillation and the cis-isomer is recycled by the same buffered carboxylic acid system used above to produced isopiperitenyl esters. In this manner pure (−)-trans(LXXVIII) is obtained with minimum losses due to (+)-cis isomer formation. Compound (LXXVIII) is desired here so as to be able to selectively obtain (−)-menthol directly on hydrogenation. The latter reaction is performed using Pd/C as catalyst to produce mainly (−)-menthol along with some undesired (−)-isomenthol. Efficient fractional distillation gives (−)menthol of good purity. A second route involves pyrolysis of (+)-δ-2-carene (LXXV) to (+)trans-2,8-p-menthadiene (LXXIX) with generation of an asymmetric centre

Chiral chemistry and food flavourings

121

at C-1. Isomerisation of (LXXIX) to (+)-2,4(8)-p-menthadiene (LXXX) can be accomplished either catalytically or via hydrochlorinationdehydrochlorination. Hydrogenation over a poisoned catalyst gives (+)-3p-menthene (LXXXI), which is epoxidised and the epoxide rearranged to afford a mixture of (−)-menthone (LXXXII) and (+)-isomenthone (LXXXIII). The latter compounds are separable and the base-mediated epimerisation efficiently increases the percentage of menthone isomer in the mixture. The following reduction (usually hydrogenation) gives predominantly (−)-menthol, which can be separated from the unwanted isomers by distillation or crystallisation. Similarly, the ketones (LXXXII)/(LXXXIII) are intermediates in the menthol production from pennyroyal oil (Pybus and Sell, 1999). The latter essential oil is produced principally in Southern Europe and North Africa. Its major component is (+)-pulegone (XII) that is hydrogenated to afford the above-mentioned menthone/isomenthone mixture.

6.4

Key flavour compounds

6.4.1 General The characteristic odour and taste of a flavour formulation are due to the combination of the organoleptic properties of each chemical substance present in the mixture. Furthermore, the relevant importance of each ingredient on the final flavour quality are often very different. The odour features of a chemical compound are somewhat substitutable with one another, without great changes in the final flavour character. However, some particular chemical compounds are so relevant that their exclusion from the final formulation does not allow the preparation of the same flavour formulation. The latter compounds are therefore called ‘key flavour compounds’ and their characterisation and preparation are of great interest in flavour production. Moreover, some of these compounds may occur in nature in very low amounts and their analytical detection is often a challenging objective. A previously presented key flavour compound with these features is 2methyl-4-propyl-1,3-oxathiane (XXXII) that was detected in 1976, patented in 1980 by Firmenich with the trade name of Oxane and until now widely used as key ingredient in tropical type formulations. The characterisation of new key flavour compounds is a topical research field especially in the search of new chiral flavours where their stereoselective synthesis is necessary for their complete olfactory evaluation. In this context the case of wine lactone is presented below. This compound is very relevant for its coumarinlike odour. The latter features are especially demanded since coumarin itself is not allowed in flavouring for toxicological reasons. 6.4.2 p-Menthane lactones: the relevant case of wine lactone Among the monoterpenes of the p-menthane family, many natural products show oxygen functionalisation both in positions 3 and 9 of the p-menthane

122

Modifying flavour in food

O

O

O

O (LXXXIV)

(LXXXV)

O

O

O

O

(LXXXVI) (LXXXVII)

OH O

O O (LXXXVIII)

Fig. 6.8

O O

(LXXXIX)

(XC)

Some relevant p-menthane flavours.

framework. Several compounds of different chemical structures belong to this class and most of them are well known as flavouring ingredients (Fig. 6.8). The 3-hydroxy-p-menthan-10-oic acid lactone (LXXXIV), (−)-mintlactone (LXXXV), (+)-isomintlactone (LXXXVI)-, (1R)-menthofurolactone (LXXXVII), (1R)-3-hydroxy-mintlactone (LXXXVIII) as well as (+)menthofuran (LXXXIX) were found in peppermint oil (Takahashi et al., 1980; Näf and Velluz, 1998), while the isomeric (+)-mintlactone and (−)isomintlactone are key aroma components of the wood of Bursera graveolens (Iwabuchi, 1997). In this context, p-menthane lactones are very interesting compounds since the most prominent members of this class, lactones (LXXXIV)– (LXXXVIII), are used as commercial flavouring ingredients (Gaudin, 1995; Köpsel and Emberger, 1983). Moreover, recent studies have highlighted the relevance of the isomeric wine lactone (XC) (Guth, 1996) and of the saturated analogues p-menthane lactones (Gaudin, 2000) due to their low odour thresholds and the powerful coumarin-like odour properties. In addition, these studies show that either the isomeric and/or the enantiomeric composition of these lactones determined their olfactory features, and different researchers have demonstrated that both the character and intensity of their odour are related to their relative and absolute stereochemistry (Guth, 1996). This is the case of wine lactone (XC) which was first identified as an animal metabolite (Southwell, 1975), then recognised as a key flavour compound of different white wines (Guth, 1997) and subsequently it was found in orange juice and black pepper (Hinterholzer and Schieberle, 1998; Jagella and Grosch, 1999). The eight possible isomers of wine lactone (Fig. 6.9) were prepared by Guth (1996) both in racemic and enantio-enriched form. The racemic synthesis (Fig. 6.10) started from 3-methylcyclohexan-2-one (XCI). The kinetic

O

5 4

O (3S,3aS,7aR) (XCa)

6

7 7a O1 3a 2 3 O

O (3R,3aR,7aS) (XCb)

O (3S,3aS,7aS) (XCe)

Fig. 6.9

O

b

c, d OH

O COOMe COOMe (XCII) (XCIII)

(XC) O

g

+

OH

O

COOMe (XCIV)

f

+

(XCVa)

+

COOH COOH (XCVIa) (XCVIb)

(XCVb)

h g

+ O

O

O (XCa)

O (XCc)

+ COOMe COOMe (XCVIIa) (XCVIIb)

c, d

+

+

OH

COOMe COOMe (XCVa) (XCIVb)

O

O

O (XCe)

O (XCg)

+

(VI) (–)-Limonene

O

O

COOMe COOMe (XCIIIb) (XCIIIa) b

OH

O (3S,3aR,7aR) (XCh)

The stereoisomers of wine lactone.

OH (V) (+)-Limonene

O

O (3R,3aS,7aS) (XCg)

I

e

O (3S,3aR,7aS) (XCd)

O

O (3R,3aR,7aR) (XCf)

a (XCI)

O

O (3R,3aS,7aR) (XCc)

O

O

Wine lactone (XC)

O

O

+

and

O

O

O

O

O (XCh)

O (XCf)

O (XCb)

O (XCd)

Fig. 6.10 The Guth synthesis of racemic and enantio-enriched wine lactone (a, lithium diisopropylamide (LDA), tetrahydrofuran (THF); b, NaBH4, CaCl2, iPrOH; c, NaOH, MeOH, H2O; d, dicyclohexyl carbodiimide (DCC), benzene; e, 9-9-borabicyclo[3.3.1]nonane (BBN), THF; f, pyridinium dichromate (PDC), N,Ndimethylformamide (DMF); g, PDC, benzene, tBuOOH, h, MeOH, H2SO4).

124

Modifying flavour in food Table 6.2 Odour properties of wine lactone stereoisomers Configuration

Compound

Odour threshold (ng/l air)

(3S,3aS,7aR) (3R,3aR,7aS) (3R,3aS,7aR) (3S,3aR,7aS) (3S,3aS,7aS) (3R,3aR,7aR) (3R,3aS,7aS) (3S,3aR,7aR)

(XCa) (XCb) (XCc) (XCd) (XCe) (XCf) (XCg) (XCh)

0.00001–0.00004 >1000 >1000 80–160 0.007–0.014 14–28 8–16 0.05–0.2

formation of lithium enolate of (XCI) and the following alkylation with 2iodo-propanoate (XCII) afforded the keto-ester (XCIII). The regioselective reduction of the ketonic functionality by means of NaBH4/CaCl2 reagent gave the hydroxy-ester (XCIV) that, after saponification and DCCmediated lactonisation, afforded wine lactone (XC) as a mixture of eight isomers. Moreover, (+)- and (−)-limonene (V) and (VI) were the chiral building blocks for the enantioselective path to (XC). Regioselective hydroboration of (+)-(V) with 9-BBN gave a mixture of the diastereoisomeric alcohols (XCVa) and (XCVb) that were oxidised by means of PDC (pyridinium dichromate) to the corresponding acids (XCVIa) and (XCVIb), respectively. Concurrent oxidation at position 7a and lactonisation were achieved by the use of PDC/tBuOOH reagent. The wine lactone (XCa) and epi-wine lactone (XCc) obtained were then separated by chromatographic methods. In a different way, acids (XCVIa) and (XCVIb) were converted into the methyl esters (XCVIIa) and (XCVIIb) and then oxidised to the keto esters (XCIIIa) and (XCIIIb), respectively. The latter compounds were submitted to diastereoselective reduction to produce hydroxy esters (XCIVa) and (XCIVb) showing the desired trans configuration at positions 7a and 3a. The following saponification and DCC (N,N′dicyclohexylcarbodiimide)-mediated lactonisation gave a separable mixture of wine lactone isomers (XCe) and (XCg). The lacking four stereoisomers with 3aR configuration were therefore obtained starting from (−)-limonene (VI), using the same synthetic path. The availability of all eight stereoisomers allowed their olfactory evaluation. This study revealed (Table 6.2) that (XCa) is the more powerful isomer with an odour threshold <0.04 pg/l, whereas the weakest isomers show a threshold >106 pg/l. It is worth noting that compound (XCa) is the natural enantiomer. These characteristics clearly demonstrate that odour features can be dramatically influenced by the chemical configuration. The latter outstanding results prompted the studies of a more direct synthesis of enantiopure (XCa). Again, starting from an enantiopure build-

Chiral chemistry and food flavourings

a

125

b +

+

OH (LXXIX)

(XCVIIIa)

OH (XCVIIIb)

COOH (XCIXa)

COOH (XCIXb) c

I +

O

O (XCc)

O

O

O

O (XCa)

I

d

+

O

O (Ca)

(Cb)

Fig. 6.11 The Chavan et al. synthesis of enantio-enriched wine lactone (a, 9-BBN, THF; b, Jones’s oxidation; c, NaI, FeCl3, CH3CN, reflux; d, diazabicyclo[5.4.0]undec7-ene (DBU), THF).

ing block, Chavan et al. (2001) described (Fig. 6.11) the conversion of (+)isolimonene (LXXIX) in (−)-wine lactone and its C(8)-epimer by means of iodolactonisation as the key step. Accordingly, (+)-(LXXVIII) was regioselectively converted into the alcohols (XCVIIIa) and (XCVIIIb) that were oxidised to acids (XCIXa) and (XCIXb), respectively. The following lactonisation step was achieved by means of the NaI/FeCl3 reagent that produced iodo-lactones (Ca) and (Cb), showing cis configuration at the ring junction. The latter compounds were then submitted to an elimination reaction using DBU (1,8-diazobicyclo[5.4.0]undec-7-ene) as a base to produce wine lactone (XCa) and its epimer (XCc). The first enantioselective asymmetric synthesis of the same two compounds was reported by Bergner and Helmchen (2000) (Fig. 6.12). The latter approach is based on a palladiumcatalysed enantioselective reaction and involved the use of asymmetric ligands and a number of chemical manipulations. The starting material was racemic acetate (CI) that was submitted to a palladium-catalysed enantioselective allylic substitution using the lithium salt of dimethylmalonate as a nucleophile and chiral compound (CII) as a palladium ligand. The obtained enantio-enriched (95% ee) diester (CIII) was converted into the monoester (CIV) by hydrolysis and elimination of carbon dioxide. The following saponification and iodolactonisation gave cis lactone (CV). The stereoselective introduction of the missing two methyl groups was performed by a multi-step sequence. Elimination of iodine atom by DBU gave unsaturated lactone (CVI) that reacted with MeMgCl in the presence

126

Modifying flavour in food PPh2•BH3 (CII)

b

COOH

I

c, d

a

O

OAc

MeOOC

(CI)

f

COOMe (CIII)

d

+

O (CVI) O

e

COOH (CVIIa)

e

COOH (CVIIb)

COOMe (CIV)

I

(CV)

O

I +

O

O

(CVIIIa) O

(CVIIIb) O

g O (CIX) O

O (XCa)

O

Fig. 6.12 The Bergner and Helmchen synthesis of enantio-enriched wine lactone (a, (C3H5PdCl)2 cat.; (CII) cat., LiCH(COOMe)2, THF; b, NaCl, H2O, dimethyl sulphoxide (DMSO), heat; c, NaOH, heat; d, KI, I2, NaHCO3, H2O; e, DBU, THF, reflux; f, MeMgCl, CuBr. Me2S cat., THF, Me2S; g, LDA, MeI, THF).

of a copper catalyst (SN2′ substitution) to give acids (CVIIa) and (CVIIb). Repeating the iodolactonisation step afforded lactones (CVIIIa) and (CVIIIb) that, by elimination, gave the same unsaturated lactone (CIX). Diastereoselective alkylation of (CIX) with LDA and MeI afforded the suitable wine lactone (XCa). Moreover, a furter enantioselective synthesis of compounds (XCa) and (XCc) was reported by Serra and Fuganti (2002, 2004) (Fig. 6.13). The latter approach is based on the lipase-mediated resolution of diastereoisomerically pure p-mentha-1,8(10)-diene-3,9-diol (CXVI). This compound was prepared from the easily available diene (CX) and dienophile (CXI) in turn obtained from butynol (CXII) and butyndiol (CXIII), respectively. Accordingly, the Diels–Alder reaction of (CX) and (CXI) afforded cis derivative (CXIV) with high diastereoselectivity (cis/trans: 95/5). The introduction of the missing carbon atom was performed by methylenation (Wittig reaction) to give dienic acetate (CXV). The following saponification and purification by crystallisation produced diastereoisomerically pure (up to 99% de) (CXVI). The latter compound was submitted to lipase-mediated acetylation to give enantio-enriched monoacetate (CXVII) and diacetate (CXVIII). The separation of these products and the following saponifica-

Chiral chemistry and food flavourings OH

(CXII)

(CX) OAc a O

OAc (CXI)

HO

127

49% OH e OAc c, d OH OAc 46% (CXV) (CXVI)

OAc b OAc O (CXIV)

OH (CXIII) f OH OH (+)-(CXIX)

g O O (–)-(CXX)

c OH OH 93% OH OAc (+)-(CXIX) (CXVII) 94% ee c OH OAc 96% OAc OH (–)-(CXIX) (CXVIII) 99% ee

h, i + (XCa) + (+)-(XCc) O O (XCc)/(XCa)>7:1 (XCa)/(XCc)>4:1 O (XCa) O (XCc)

Fig. 6.13 The Serra and Fuganti synthesis of enantio-enriched wine lactone (a, (CX) + (CXI), benzene, reflux; b, Ph3PCH2, THF, reflux; c, KOH, MeOH, reflux; d, crystallisation from hexane/AcOEt 3 : 1; e, vinyl acetate, t-BuOMe, lipase (PS); f, bis-acetoxyiodobenzene (BAIB), CH2Cl2, TEMPO (cat.); g, Mg, MeOH; h, KOtBu, tBuOH; i, saturated aqueous NH4Cl solution).

tion gave (+)- and (−)-(CXIX), respectively, both in high enantiomeric purity. Diol (+)-(CXIX) was then used in the synthesis of (XCa). Its regioselective oxidation by means of BAIB (bis-acetoxyiodobenzene) and TEMPO (2,2,6,6-tetramethyl-1-piperidinyloxy, free radical) as catalysis afforded the unsaturated lactone (CXX) that was reduced by magnesium in methanol to a mixture of epi-wine lactone (XCc) and wine lactone (XCa) where the former compound was predominant. The treatment of the latter mixture with strong bases (KOtBu) followed by quenching, produced a mixture of (XCa) and (XCc) where wine lactone was the main component and from which it can be easily separated.

6.5

Sources of further information and advice

Flavour production is strongly connected to fragrance production. Although the compounds are used for different application (flavours mainly in foods and beverages while fragrances in perfumes), their preparation is often related. For example, different flavours such as pinene, limonene, terpineol, camphor, linalool, citronellol and citronellal are used also as fragrances or as a suitable chiral starting materials in their synthesis. Therefore, it should be noted that many sources of information about fragrances (availability, preparation, use) are also relevant to flavours.

128

Modifying flavour in food

Some pertinent books are: bauer, k., garbe, d. and surburg, h. (2001), Common Fragrance and Flavor Materials, fourth ed., Weinheim (Germany), Wiley-VCH Verlag GmbH ohloff, g. (1994), Scent and Fragrances, Berlin, Heidelberg (Germany), Springer-Verlag pybus, d. h. and sell, c. s. (1999), The Chemistry of Fragrances, Cambridge (UK), The Royal Society of Chemistry belitz, h.-d. and grosch, w. (1987), Food Chemistry, Berlin, Heidelberg (Germany), Springer-Verlag Some pertinent reviews are: brenna, e., fuganti, c. and serra, s. (2003), ‘Enantioselective perception of chiral odorants’, Tetrahedron: Asymm., 14, 1–42 serra. s., fuganti, c. and brenna, e. (2005), ‘Biocatalytic preparation of natural flavours and fragrances’, Trends Biotecnol., 23, 193–198 Some information is also available on the following website: http://www.leffingwell.com Further information are available on the websites of the major flavour and fragrances companies, for instances: www.givaudan.com www.iff.com www.firmenich.com www.symrise.com www.questintl.com www.takasago.com

6.6

References

akugatagawa, s. (1992), ‘A practical synthesis of (−)-menthol with the Rh-BINAP catalyst’, in Collins A N, Sheldrake G N, Crosby J, Chirality in Industry, Chichester, John Wiley & Sons Ltd, 313–339. bauer, k., garbe, d. and surburg, h. (2001), Common Fragrance and Flavor Materials, fourth ed., Weinheim (Germany), Wiley-VCH Verlag GmbH. bergner, e. j. and helmchen, g. (2000), ‘Synthesis of enantiomerically pure (−)-wine lactone based on a palladium-catalyzed enantioselective allylic substitution’, Eur. J. Org. Chem., 419–423. brenna, e., fuganti, c., serra, s. and kraft, p. (2002), ‘Optically active ionones and derivatives: preparation and olfactory properties’, Eur. J. Org. Chem., 967–978. brenna, e., fuganti, c. and serra, s. (2003), ‘Enantioselective perception of chiral odorants’, Tetrahedron: Asymm., 14, 1–42. chavan, s. p., kharul, r. k., sharma, a. k. and chavan, s. p. (2001), ‘An efficient and simple synthesis of (−)-wine lactone’, Tetrahedron: Asymm., 12, 2985–2988. clark, g. s. (1998), ‘Menthol’, Perfumer & Flavorist, 23 (5), 33–46.

Chiral chemistry and food flavourings

129

council of the european communities (1988), Council Directive 88/388/EEC of 22 June 1988. dufossé, l., latrasse, a. and spinnler, h. e. (1994), ‘Importance des lactones dans les arômes alimentaires: structures, distribution, propriétés sensorielles et biosynthèse’, Sci. Aliment., 14, 17–21. food and drug administration (1985), US Code of Federal Regulations, 21, 101.22a.3.Washington, DC. gaudin, j.-m. (1995), ‘Use of furanones as perfuming ingredients’, United States Patent, US 5464824. gaudin, j.-m. (2000), ‘Synthesis and organoleptic properties of p-menthane lactones’, Tetrahedron, 56, 4769–4776. guth, h. (1996), ‘Determination of the configuration of wine lactone’, Helv. Chim. Acta, 79, 1559–1571. guth, h. (1997), ‘Identification of character impact odorants of different white wine varieties’, J. Agric. Food Chem., 45, 3022–3026. hinterholzer, a. and schieberle, p. (1998), ‘Identification of the most odour-active volatiles in fresh, hand-extracted juice of Valencia late oranges by odour dilution techniques’, Flavour Fragrance J., 13, 49–55. iwabuchi, h. (1997), Abstract Paper pp. R1–R4, 41st Symposium on the Chemistry of Terpenes, Essential Oil and Aromatics, Morioka, Japan. jagella, t. and grosch, w. (1999), ‘Flavour and off-flavour compounds of black and white pepper (Piper nigrum L.)’, Eur. Food Res. Technol., 209, 16–21. jauch, j., schmalzing, d., schurig, v., emberger, r., hopp, r., köpsel, m., silberzahn, w. and werkhoff, p. (1989), ‘Isolation, synthesis, and absolute configuration of filbertone – the principal flavor component of the hazelnut’, Angew. Chem., Int. Ed. Engl., 28, 1022–1023. köpsel, m. and emberger, r. (1983), ‘Di- and tetra-hydrobenzofuranones as scents and aroma substances’, United States Patent, US 4407740. näf, r. and velluz, a. (1998), ‘ Phenols and lactones in italo-mitcham peppermint oil Mentha × Piperita L.’, Flavour Fragr. J., 13, 203–208. ohloff, g. (1994), Scent and Fragrances, Berlin, Heidelberg (Germany), SpringerVerlag; and references cited therein. pickenhagen, w. and brönner-schindler, h. (1984), ‘Enantioselective synthesis of (+)- and (−)-cis-2-methyl-4-propyl-1,3-oxathiane and their olfactive properties’, Helv. Chim. Acta, 67, 947–952. pybus, d. h. and sell, c. s. (1999), The Chemistry of Fragrances, Cambridge (UK), The Royal Society of Chemistry; and references cited therein. serra, s. and fuganti, c. (2002), ‘Enzyme-mediated preparation of enantiomerically pure p-menthan-3,9-diols and their use for the synthesis of natural p-menthane lactones and ethers’, Helv. Chim. Acta, 85, 2489–2502. serra, s. and fuganti, c. (2004), ‘Natural p-menthene monoterpenes: synthesis of the enantiomeric forms of wine lactone, epi-wine lactone, dill ether, and epi-dill ether starting from a common intermediate’, Helv. Chim. Acta, 87, 2100–2109. serra, s., fuganti, c. and brenna, e. (2005), ‘Biocatalytic preparation of natural flavours and fragrances’, Trends in Biotecnol., 23, 193–198; and references cited therein. singer, g., heusinger, g., fröhlich, o., schreier, p. and mosandl, a. (1986), ‘Chirality evaluation of 2-methyl-4-propyl-1,3-oxathiane from the yellow passion fruit’, J. Agric. Food Chem., 34, 1029–1033. southwell, i. a. (1975), ‘Essential oil metabolism in the koala. III Novel urinary monoterpenoid lactones’, Tetrahedron Lett., 24, 1885–1888. takahashi, k., someya, t., muraki, s. and yoshida, t. (1980), ‘A new keto-alcohol, (−)-mintlactone, (+)-isomintlactone and minor components in peppermint oil’, Agric. Biol. Chem., 44, 1535–1543.

130

Modifying flavour in food

winter, m., furrer, a., willhalm, b. and thommen, w. (1976), ‘Identification and synthesis of two new organic sulfur compounds from the yellow passion fruit (Passiflora edulis f. flavicarpa)’, Helv. Chim. Acta, 59, 1613–1620. wüst, m. and mosandl, a. (1999), ‘Important chiral monoterpenoid ethers in flavours and essential oils – enantioselective analysis and biogenesis’, Eur. Food Res. Technol., 209, 3–11. yamamoto, t. (1998), ‘Method for purifying (−)-N-isopulegol and citrus perfume composition containing (−)-N-isopulegol obtained by the method’, United States Patent, US 5773410.

7 Formulating low-fat food: the challenge of retaining flavour quality J. Hort and D. Cook, University of Nottingham, UK

7.1

Introduction

The term ‘fat’ can encompass a range of different materials but throughout this chapter we use the term to refer to the triglyceride molecules present in food that are composed of a glycerol backbone esterified to three fatty acids. Fat contributes to the characteristic texture and flavour of many food products (Drewnowski, 1997a) which, more often than not, elicit a positive hedonic response by the consumer (Tuorila, 1992). Fat has an important role in the diet and, in addition to initiating a rewarding sensory response, it provides a concentrated source of energy, insulation for the body and protection for vital organs. Fats form integral elements of cell membranes, and contribute to cholesterol transport, blood clotting, immune function and the manufacture of hormones for reproductive function. Certain fatty acids, linoleic acid and alpha-linolenic acid (essential fatty acids) must also be obtained directly from the diet as they cannot be made in the body. However, there is now little doubt that a diet high in fat is linked to the chronic health problems of Western society: obesity, heart disease, diabetes and colon cancer (Department of Health 1984, 1991, 1994, 1998). According to the National Diet and Nutrition survey, over 50% of the UK population are overweight or obese, indicating that total energy intake is still too high (Hoare et al., 2004). Fat has a higher energy value than protein or carbohydrate but does not appear to reduce hunger or induce satiety to the same extent as these other macronutrients (Johnson and Vickers, 1993; Stubbs et al., 1995; Bell et al., 1998; Bray and Popkin, 1998). As a consequence, fat is easily over-consumed and so contributes to increased levels of obesity and its related health problems.

132

Modifying flavour in food

Health campaigns are not just concerned with the amount of fat in the diet but also fat composition, i.e. the saturated to unsaturated fatty acid ratio. Diets high in saturated fatty acids and low in unsaturated fatty acids are linked to higher blood cholesterol levels and subsequent complications such as coronary heart disease (Department of Health, 1984). Not surprisingly then, the UK population is currently advised to reduce the total fat content in its diet to around 35% of energy intake with no more than 11% from saturated fatty acids (Department of Health, 1991). Although we appear to be meeting the 35% target, energy derived from saturated fatty acids in UK diets is still too high at around 13% (Hoare et al., 2004). The food industry has made considerable steps towards facilitating changes in dietary patterns through the manufacture of low-fat alternatives but such products fail to meet the sensory expectations of today’s discerning consumer. In this chapter, we look at the changes in the sensory properties of food when fat is removed and review the strategies currently used by industry to replace fat. We then consider why it might be such a challenge for the food industry to succeed in replacing fat by looking at the biological mechanisms involved in its perception. We conclude by considering the implications of current research for future product development policy and by directing the reader to further sources of information.

7.2

Lowering the fat content of food: what happens to the flavour?

It has long been appreciated that modifying the fat content of foods has a considerable impact upon the processes of flavour release and perception (e.g. Kinsella, 1969). Poor flavour quality is one of the key difficulties to surmount when developing reduced fat alternatives to established products and the complexities of doing so reflect the multifunctional nature of fat as a food ingredient. Fat in foods can be considered to influence food flavour via four principal mechanisms: 1. 2.

3.

Since many odorous chemicals are hydrophobic in nature, fat acts as a carrier and reservoir for significant aroma chemicals in food systems. The fat content of foods stimulates the senses during eating, both via a range of oral textural parameters which are influenced by fat content (e.g. oiliness, lubrication, creaminess) but also, as has recently been established, due to the existence of specific fat receptors on the palate (Mattes, 2003). Fat can act as a precursor for certain flavours which would not be formed in its absence, e.g. it is known that the kind of lipid present during cooking influences the volatile aroma profile formed in the Maillard reaction (Mottram et al., 1998; Elmore et al., 2002).

Formulating low-fat food: the challenge of retaining flavour quality 4.

133

The nature, physical state (oil or crystalline) and amount of fat present can have a considerable bearing on the physical structure of food systems. This directly influences the partitioning and release of flavour compounds and determines kinetic parameters related to flavour generation, stability or the generation of rancidity.

Naturally, the relative importance of each of the above factors varies with product type and what is an acceptable reduction in fat for one product in percentage terms may turn out to be disastrous for another. 7.2.1 Influence of fat content on aroma partition and release Aroma release from foods during eating is a dynamic process which is governed by two main factors (de Roos, 1997): partition phenomena (which determine the potential extent of flavour release) and mass transport phenomena related to the texture, surface area and surface renewal of a product during eating. Fat exerts a major influence on each of these factors. In fat-containing foods, lipophilic aroma compounds are bound to fat molecules by weak, reversible van der Waals and hydrophobic interactions. However, in the absence of fat, lipophilic flavours are poorly bound to the food matrix and the resulting headspace concentrations are far too high (Plug and Haring, 1993). The retention of flavour in products over time (hence shelf-life) is likewise adversely affected (de Roos, 1997). Studies utilising a range of foodstuffs have shown that the fat content of foods influences both the balance of an aroma and the temporal profile of its release. An illustrative study is that reported by Brauss et al. (1999) who investigated the release of selected aroma compounds (with differing physicochemical properties) from yoghurts containing 0.2, 3.5 and 10% fat. Flavour release from the systems was measured during eating using a realtime mass spectrometry technique and panellists’ perception of flavour was followed using a time–intensity (TI) method. The influence of fat content depended upon the hydrophobicity of a particular aroma compound, which was quantified in terms of its oil–water partition coefficient (log P). Compounds with a high log P (e.g. terpinolene, anethole) are lipophilic in nature and were influenced most profoundly by changes in fat content. Flavour release measurements showed that lipophilic aroma compounds were released more rapidly (lower Tmax) and with greater intensity (higher Imax) in low-fat formulations. These measurements were in agreement with sensory TI ratings of the products. The differences were principally between the lowest fat yoghurt (0.2%) and the 3.5% and 10% fat yoghurts which behaved similarly to one another. Other studies have likewise noted the influence of aroma compound polarity on the effects of fat reduction. Chung (2004) worked with ice creams varying in fat content and fat replacers and flavoured with an artificial cherry aroma. Faster release rates were reported for hexanal and benzaldehyde with decreasing matrix fat content, while the opposite trend was observed for the more hydrophillic vanillin.

134

Modifying flavour in food

In general, it is true to say that most studies of this type have demonstrated the effects of fat content on temporal elements of aroma release more clearly than has been true of the corresponding differences in sensory perception (Miettinen et al., 2004). This may reflect the sensory complexities of identifying rapidly changing flavour qualities over short timescales, the level of complexity of the flavouring used in some studies or, indeed, the ability to detect changes in the composition of aroma mixtures. It follows from the above that the effect of fat content on aroma perception depends largely on whether or not the key character impact compounds for the flavour system are lipophilic in nature. Such findings highlight one of the key difficulties in reformulating low-fat foods for better flavour: removing fat selectively modifies the partitioning and release of the most lipophilic compounds, relative to hydrophilic compounds, leading to flavour imbalance. In particular, the more rapid and intense release of hydrophobic aroma compounds can make the flavour appear quite imbalanced and less smooth. Flavour manufacturers are well aware of this fact because complex reformulation of flavours may be required to tailor them for new applications (de Roos, 1997; Plug and Haring, 1993). The fat content of foods is the most significant parameter that must be taken into account in such a process.

7.3

Strategies for replacing fat in foods and the implications for flavour

The most simple, yet limited, fat replacement strategy involves removing fat from a food system (e.g. as in the development of semi-skimmed and skimmed milks) either with or without changes in formulation using established ingredients. It is highly likely that the flavour quality and other significant product quality parameters will be affected, although these modifications may either be deemed acceptable (e.g. are out-weighed by the advantages offered by a low-fat product) or may be overcome using a combination of technical reformulation, process modification or novel storage and/or marketing approaches. However, for the majority of food systems, obtaining a significant reduction in fat content requires a more sophisticated approach involving the use of functional ingredients developed as fat replacers. The range of ingredients marketed as fat replacers is vast, which reflects both the importance of this developing market area and the complexities of replacing all of the functionalities of fat. The emphasis in this chapter is on understanding how fat replacement strategies influence food flavour, hence there is not room here for a detailed description of individual products. However, consideration of the different categories of fat replacers that have been developed gives an insight into the main strategies that have been adopted (Table 7.1).

Formulating low-fat food: the challenge of retaining flavour quality

135

Table 7.1 Subcategorisation of types of fat replacers Category

Description

Example

Fat substitute

Substance that replaces fat on a weight by weight basis. Generally resembles triglycerides in structure, but with significantly lower digestibility. Synthetic triglycerides featuring unusual (less metabolised) fatty acids.

‘Olestra’ (sucrose polyester). ‘Salatrim’ (short and long acyltriglyceride molecule.

Fat mimetic

Substance that modifies the aqueous phase of foods, thereby simulating some of the physical properties of fat in foods.

Maltodextrins (e.g. ‘Paselli’) and starchderived products. ‘Simplesse’ (microparticulated protein).

Fat extender

Emulsions containing 25 or 33% standard fats or oils – make the fat go further!

‘Veri-Lo’.

Fat substitutes may be used as 1 : 1 replacements for fats and oils and exhibit similar (but distinct) functionalities. Olestra can, for example, be used as a frying oil. A general consequence of removing fat from foods is that the proportion of aqueous phase is increased. One of the key functionalities of materials that function as fat replacers is that they replace some of this structure by ordering the aqueous phase, thus mimicking the presence of fat. Fat mimetics are a large and diverse group of materials that primarily act in this fashion, e.g. maltodextrins, microparticulated proteins, microcrystalline cellulose, fibre-based fat replacers and inulin (to name but a few).

7.3.1 Impact of fat replacement strategies on food flavour Bearing in mind the vast diversity of chemical structures represented by fat replacers, it is clear that generic advice regarding their impact upon food flavour carries certain caveats. For example, individual binding interactions between fat replacers and key aroma compounds can modify flavour release and perception, irrespective of other structural/sensory considerations. In certain circumstances this may be advantageous, e.g. if the interactions help to maintain the ‘reservoir’ effect of fat in binding and distributing flavour, but are also reversible within the time-scale of food consumption, so that flavour can be released and perceived. In other circumstances, increased interactions of an irreversible nature may make certain flavour components unavailable for release, e.g. the binding of aldehydes to proteins high in arginine/lysine (Overbosch et al., 1991). Fat substitutes are the least likely fat replacers to have an impact upon the structure, distribution and partitioning of flavour compounds within and

136

Modifying flavour in food

from the food matrix. For example, Brauss et al. (2000) reported that Benefat (trade name for salatrim, Table 7.1) caused the least disruption of aroma release in vivo relative to the full-fat control when a range of commercial fat replacers were used to completely substitute the fat in a biscuit formulation. Biscuits were flavoured by the addition of a series of ethyl esters (ethyl butyrate, hexanoate and octanoate). Each of the carbohydrate fat replacers investigated (Litesse, Paselli, Raftiline) resulted in more rapid and intense (by a factor of around 2.5) release of the aroma system, relative to the fullfat control, in parallel with the effects of reduced fat content discussed in the foregoing section. Although the effect of Benefat was closer to that of a full-fat formulation, the maximum intensity of release of ethyl butyrate (the least lipophilic of the aromas investigated) almost doubled, whereas the more lipophilic components were slightly retained by Benefat, relative to the control. Hence, even the best candidate investigated was likely to ‘skew’ the aroma profile of a product, owing to subtle changes in the physicochemical nature of the fat component.

7.3.2 Fat replacement strategies and food texture Much work has concentrated on the textural properties that fat confers to food systems and how these can best be mimicked. Once again, the precise effects of fat on perceived textural attributes is hugely product dependent – varying, for example, with the fat content, composition, physical state (e.g. liquid, crystalline or amorphous solid) and system structure (e.g. emulsion, dispersion, particle size and nature of the continuous phase(s)). A widely adopted approach has been to mimic the effect of dispersed oil/fat droplets in food systems by generating fat replacers that have been structured to form microparticles. This approach was based upon the significance of fat droplet size in determining the perceived creaminess of foodstuffs. Drewnowski (1992) stated that food particles less than 3 µm in size are not detected individually by the tongue, resulting in an overall creamy mouthfeel. Thus, a typical target particle size range for fat replacers has been 0.1–3.0 µm. A widely quoted example of this approach was the development of Simplesse®, a microparticulated whey protein product formed by heating protein under controlled conditions of shear. Under such conditions the protein denatures but associates to form insoluble microparticles. Other products have been developed by coating microparticulated proteins with an anionic polysaccharide (e.g. gum arabic, carboxy methyl cellulose, CMC) to produce dispersed ‘fat droplet replacements’ which do not tend to coalesce. A review of the development of such fat ‘droplet’ replacers (Jones, 1996) suggested that, although some structuring effect on the aqueous phase was clearly required in order to mimic fat, the need to gener-

Formulating low-fat food: the challenge of retaining flavour quality

137

ate particles of a particular size or shape was less clearly defined, i.e. successful products were developed which had irregular particle shapes and with sizes beyond the original target range stated above. A potential reason for this discrepancy is the over-simplification of assuming that the fat-associated textural properties of foods can be mimicked by food structure and/ or rheology alone. Later work has shown that sensory attributes such as creaminess are clearly complex sensory constructs which are both product-specific and incorporate interactions with other sensory information such as aroma and taste (Kilcast and Clegg, 2002). This is entirely consistent with contemporary views on the multisensory nature of flavour perception and the body of published work relating to texture–flavour interactions. Thus, taking a holistic approach, it now appears entirely logical that fat replacement strategies must address texture and flavour, and their potential interactions, simultaneously. Furthermore, recent developments (described more fully below) have highlighted the limitations of treating fat purely as a textural stimulus (de Araujo and Rolls, 2004).

7.3.3 Sensory properties of fat replacement products Typically, the flavour of reduced-fat formulations, containing fat replacers, suffers by comparison with the original product. This reflects the difficulties of finding substitutes to replace all the effects that fat has on the sensory properties of a food system. Thus, some of the more successful products that have been developed are those such as low-fat spreads, which open up a new market by offering clear advantages to the consumer (e.g. healthy, spreads direct from fridge, less greasy) and are thus judged as acceptable alternatives to the original product, but with added value. The literature is full of examples of just how difficult it is to match the textural and flavour quality of food systems using fat replacement strategies. For example, Roland et al. (1999) reported that ice creams (<0.5% fat) containing fat replacers (maltodextrin, polydextrose or milk protein concentrate) melted faster and were judged to have a less creamy flavour than 10% fat ice cream. Both of these aspects are clearly key quality criteria for ice creams. Devereux et al. (2003) formulated low-fat versions of cookies, blueberry muffins, carrot cake, chocolate cake, lemon cheesecake, ice cream and beef sausages, containing inulin and oligofructose as fat replacers. In sensory tests each of the low-fat formulations were scored as less acceptable than their full-fat counterparts and it was noted that texture was more important than flavour in determining the acceptability of the low-fat products. It would be easy, although not particularly informative, to list many more examples illustrating the sensory deficiencies of low-fat products when judged relative to a full-fat control. However, in an attempt to formulate

138

Modifying flavour in food

ideas regarding a more holistic approach to fat replacement strategies, it is time now to consider why fat is hard to replace.

7.4

Why is fat so hard to replace?

From an evolutionary perspective, it is not difficult to conceive that an ability to detect the presence of fat in food would be advantageous considering its important role in the body, and that the risk of over-consumption would not be an issue when food was not so readily available. Furthermore, mechanisms that initiate a sense of enjoyment would also facilitate adequate consumption of such a nutrient. For some time the sole mechanism for fat detection was believed to be through textural stimulation and there is considerable evidence that fat contributes to the textural quality of food. Indeed Mela and Marshall (1992) describe fat’s contribution to textural quality as unique. However, describing the general textural properties elicited by fat is difficult as they are very dependent upon the nature of the fat and how it is incorporated into the physical structure of the specific food stuff. Virtually any textural attribute can be attributed to fat and these have been summarised by Drewnowski (1997b). Where fat is dispersed through a liquid in droplet form as in dairy products, salad dressings, soups and sauces, it imparts a smooth, thick and creamy texture. Its water-holding capability in animal tissue and cake batter is responsible for the juiciness or moistness of these products respectively and its shortening capability accounts for the short texture of pastry. Finally, Drewnoski credits the heat transfer capability of frying oil for the crisp and crunchy textures associated with fried products such as French fries and other snack-type products. Fat content or type also contributes to the hardness of some foods such as cheese and butter or other fat spreads. Recently de Wijk and Prinz (2005) identified that the lack of lubrication by fat in the mouth when eating low-fat foods accounted for the poor mouthfeel, including increased astringency, attributed to such products. The majority of these properties are regarded as positive attributes in the particular foods mentioned but it is also worth mentioning that some attributes associated with fat have negative connotations. Too much fat can be perceived as greasy and, if the fat has a high melting temperature, this leads to textures that are too hard and chewy. Although there tends to be a general positive association between fat content and hedonic liking, this can be reversed at higher fat levels (Mela and Marshall, 1992). Not surprisingly, with such a multitude of sensory characteristics associated with fat, various sensory mechanisms can be credited with the detection of these textural attributes in the oral cavity. The determination of viscosity, which may be related to smooth, thick and creamy textures, is complex with inputs from receptors in various parts of the oral cavity such as filiform papillae on the tongue and nerve fibres in muscles and tendons,

Formulating low-fat food: the challenge of retaining flavour quality

139

sensing the amount of pressure being applied to the food product (Meilgaard et al., 1999). A range of touch receptors and nerve endings in the surface of glabrous skin respond to somesthetic sensations such as particle size, smoothness and temperature. It is clear that perceived fat content is strongly associated with textural cues (Mela and Marshall 1992). However, evidence has now been published that suggests that there may be a gustatory element to fat detection. The sense of taste occurs when ion channels on taste receptor cells, located mainly on the surface of the tongue but also on other tissue in the oral cavity, are stimulated. Gilbertson et al. (1997) discovered that, although triglyceride molecules were not able to stimulate taste receptor cells, free fatty acids were able to inhibit the response of ‘delayed rectifying K+’ (DRK) channels in rat taste cells. Not all fatty acids demonstrated this effect. In fact, the effect was limited to compounds containing unsaturated fatty acids. Fatty acid methyl esters also had no effect but linoleic acids were moderately effective. The authors thus concluded that, to initiate a gustatory response, the fatty acid had to be unsaturated, contain a charged carboxyl carbon and have at least one double bond in the cis configuration (Gilbertson and Kim, 2002). In addition, Fukuwatari et al. (1997) have identified a fatty acid transporter mechanism in circumvallate papillae, particularly in the apical area of the taste cells. Although 98% of fat consumed is in the form of triglyceride molecules (Gunstone and Norris, 1983), there is now evidence that lingual lipase enzymes secreted by the von Ebers glands located at the base of circumvallate papillae are effective at releasing fatty acids from triglyceride molecule within 5 seconds in rats, which may enable the stimulation of taste receptors to take place (Kawai and Fushiki, 2003). It does seem reasonable, from a nutritional perspective, to postulate that there should be detection mechanisms for the essential fatty acids. However, rats do not necessarily mimic human responses and it is not clear whether the amount of lipase enzyme secreted in human saliva would be sufficient to elicit a gustatory response (Spielman et al., 1993). There is now additional evidence that fat elicits a chemosensory response in the oral cavity. Through monitoring postprandial levels of triacylglycerol (TAG in the blood), a marker that is very responsive to diet and also implicated as a risk factor in coronary heart disease, Mattes (2001a) has demonstrated that oral stimulation (mastication followed by expectoration), and in particular ‘taste’, had more impact on TAG levels than olfactory stimulation. When subjects’ oral cavities were stimulated with a cream cheese and cracker snack (with and without nostrils pinched) following a fat load (safflower oil capsules), significant rises in TAG levels were observed compared with simply sniffing the odour of the snack (olfactory stimulation). In similar experiments Mattes (2001b) demonstrated that oral stimulation with butter also elevated TAG levels but that the fat replacers also tested did not.

140

Modifying flavour in food

Oral fat exposure is also known to initiate cephalic responses that include gastric lipase secretion, modulated gastrointestinal (GI) transit and mobilisation of stored lipid (Mattes, 2005).

7.5

Representation of fat in the brain

Ultimately, it is the brain that processes the signals from sensory receptors interacting with fat and that initiates a sensory, emotional and/or physiological response. Recent advances in neuroscience and brain imaging techniques are beginning to elucidate how the brain represents fat when it interacts with the mouth. Particular areas of the brain (the taste cortex, olfactory cortex, somatosenory cortex, visual cortex) have been identified as processing initial inputs from the sensory organs, and more recently secondary areas such as the orbitofrontal cortex (OFC) and amygdala have also been implicated as areas where representations of food flavour are formed (Rolls, 2004). Neuronal recordings in these and other areas of the brain have now expanded our understanding of how the brain might represent the presence of fat in the mouth. It is important to note that much of this work has been carried out on primates as, in comparison with rats, the neural pathways of the taste system are much more comparable to those of humans (Verhagen et al., 2004). When recording neuronal activity in the primate OFC, Rolls et al. (1999) discovered that the population of neurons investigated was responsive to a range of different sensory stimuli, including fat. Specific neurons that responded to triglycerides (e.g. cream, triolein) were identified and, as these neurons were also shown to respond to other stimuli, which had a similar texture to, but were chemically different from, fat (e.g. silicone and mineral oils), the authors concluded that the textural characteristic of fat, rather than a chemical (or gustatory) stimulus, activated these neurons. In addition, bimodal neurons, which responded to both the texture of fat and specific tastants, were identified. Rolls and co-workers (1999) also discovered that the fat-sensitive neurons responded less when the fatty stimulus was eaten to satiety, indicating that these neurons also played a role in the satiety, and hence pleasure, response to fat. Verhagen et al. (2003) investigated the response of fat-sensitive neurons in the primate OFC to try to further elucidate the characteristics of fat represented in this area of the brain. The presence of fat-sensitive neurons that also responded to substances of a similar texture to fat, but that were chemically different, was confirmed in agreement with the study of Rolls et al. (1999). In Verhagen’s study, neuronal recordings were also made when primates consumed CMC solutions, prepared as a range of different viscosities, and neurons that responded to both the fat and the CMC viscosity stimuli were identified. However, what was particularly interesting was that

Formulating low-fat food: the challenge of retaining flavour quality

141

some neurons were found to respond to fat but not to the viscous CMC solutions and furthermore some neurons responded to the CMC solutions but not to the fat or fat-like stimuli. Consequently, these workers concluded that there are characteristics of fat, independent of its viscosity, that the brain utilises to recognise a fat stimulus. Furthermore, to test Gilbertson’s (1998) theory that free fatty acids may initiate a gustatory response, the response of neurons in the OFC to a range of saturated and unsaturated fatty acids was recorded. The results showed that there was no difference in response between the fatty acids and water and thus they concluded that any response to fat in the OFC was not chemically based but that some other textural attribute, in addition to viscosity, must provide the cue for fat detection. It is also interesting to note that many of the neurons were bi- or trimodal in nature, which indicates a capacity for taste and textural stimuli to interact when initiating a response. Previous work has also identified bimodal neurons responsive to taste and aroma, leading to the conclusion that flavour percepts are formed in this area of the brain. The OFC is known to receive projections from the primary taste cortex and so it is not surprising that Verhagen and co-workers (2004) went on to investigate the response to oral stimuli in the area of the brain initially receiving information from the taste receptors. Neuronal recordings from primates demonstrated that, in addition to neurons sensitive to the different taste modalities, there were neurons in the taste cortex sensitive to fat texture and viscosity but none that responded to visual or olfactory stimuli. Furthermore, as in the OFC, there were neurons that responded to fat stimuli independent of viscosity. Bimodal neurons, which responded to taste and fat or viscosity, were also identified, indicating that some convergence between taste and somatosensory stimuli occurs even at this lower level of the brain. Most recently Kadohisa et al. (2005) confirmed that taste, in-mouth texture and also temperature and trigeminal inputs were all represented in the primary taste cortex, OFC and also the amygdala. Indeed, they also confirmed that there was no relationship between fat and fatty acid activation in any of these regions. What was apparent was that the OFC was much more able to give a distinct representation of oral stimuli than the primary taste areas, probably because of more convergence, i.e. the presence of multimodal neurons, which suggests that this may be the area of the brain where recognition of the stimuli as fat (and other stimuli) is more likely to occur. Few studies have investigated representation of fat in the human brain as direct measurement methods (such as direct recording of neural signals using electrodes) are not ethical. Nevertheless, although less specific, techniques such as functional magnetic resonance imaging (fMRI) can help to identify which areas of the brain are activated by particular stimuli. fMRI detects changes in blood flow indirectly linked to increased neuronal activity in the brain. de Araujo and Rolls (2004) used this technique to compare

142

Modifying flavour in food

the brain response in humans with their neuronal observations in primates. fMRI confirmed that the primary taste cortex is activated by both taste and viscosity and also by fat. Brain scans also identified an area in the midinsular region, behind the taste cortex that was activated by both fat and viscosity but not taste. Finally, an area in the anterior cingulate was identified, which responded to fat independent of viscosity. This area of the brain is linked with hedonic response and so is likely to correspond to the pleasantness associated with fat. More research concerning the brain’s response to fat is required but, even at this early stage, it has emerged that the brain responds to the textural properties of fat independent of viscosity and, that through the identification of multimodal fat responsive neurons, responses to fat may be modulated by other sensory modalities. Certainly the brain response to fat is complex and further work in this area will provide more opportunity to understand the mechanisms involved in fat detection.

7.6

Future trends

It is hard to predict where the ‘quantum leaps’ in the science and technology of fat replacement will occur, but it seems most likely that they will result from an improved understanding of the gustatory and physiological influences of fat on consumer pleasure and satiety. By developing a more holistic picture of the effects of fat intake, it should be possible to develop more sophisticated approaches to fat replacement. Alongside such developments we foresee the continued introduction of low-fat products that offer perceived health advantages to the consumer in addition to the benefits associated with fat reduction, e.g. the development of ‘synbiotic’ foods where polysaccharides adopt the role of both fat replacer and prebiotic, or the addition of compounds such as phytosterol esters in order to decrease the absorption of cholesterol from foods. The human body appears to be very well equipped to distinguish between fat and potential impostors. To succeed, it appears that a fat replacer would need to elicit the range of sensory properties (not simply textural) offered by fat and induce the same brain and physiological responses. Unfortunately, the only ingredient that appears to be capable of doing this at present is fat. Formulating low-fat products with fat replacers does not seem to be a sustainable solution to reducing fat in the diet. A different strategy is needed, which no doubt will encompass educating the public. The success stories of low-fat spreads and semi-skimmed milk provide an insight into what these may be. By offering alternative rewards, consumers may be willing to change their lifestyle – with low-fat spreads, ‘spreadability’ offers a reward for reduced sensory performance. With semi-skimmed milk, regular use appears to have educated the consumer’s palate to the point that it is now preferred to whole milk.

Formulating low-fat food: the challenge of retaining flavour quality

143

Although fat replacers may provide a solution in reducing the fat content of some products, the real solution may lie in designing new foods that are naturally low in fat and have attractive sensory properties, or, which consumers ‘learn’ to like through repeated exposure.

7.7

References

bell, e. a., castellanos, v. h., pelkman, c. l., thorwart, m. l. and rolls, b. j. (1998), ‘Energy density of foods affects energy intake in normal-weight women’, American Journal of Clinical Nutrition, 67, 412–420. brauss, m. s., linforth, r. s. t., cayeux, i., harvey, b. and taylor, a. j. (1999), ‘Altering the fat content affects flavor release in a model yogurt system’, Journal of Agricultural and Food Chemistry, 47, 2055–2059. brauss, m. s., linforth, r. s. t. and taylor, a. j. (2000), ‘Effect on flavour release of reducing the fat content of biscuits and substituting with fat replacers’. In Frontiers of Flavour Science, Schieberle, P. & Engel, K-H. (Eds). Garching, Deutsch Forsch Sanstalt Lebensmittelchim, 279–282. bray, g. a. and popkin, b. m. (1998), ‘Dietary fat intake does affect obesity’, American Journal of Clinical Nutrition, 68, 1157–1173. chung, s. j. (2004), ‘Release of artificial cherry flavor from ice creams varying in fat and fat replacers’, Journal of Sensory Studies, 19, 211–236. de araujo, i. e. and rolls, e. t. (2004), ‘Representation in the human brain of food texture and oral fat’, Journal of Neuroscience, 24, 3086–3093. de roos, k. b. (1997), ‘How lipids influence food flavor’, Food Technology, 51, 60–62. de wijk, r. a. and prinz, j. f. (2005), ‘The role of friction in perceived oral texture’, Food Quality and Preference, 16, 121–129. department of health (1984), Diet and Cardiovascular Disease – Report of the Panel on Diet in Relation to Cardiovascular Disease of the Committee on Medical Aspects of Food Policy. Report on Health and Social Subjects 28. HMSO, London. department of health (1991), Dietary Reference Values, A Guide. HMSO, London. department of health (1994), Report on Health and Social Subjects 46. Nutritional Aspects of Cardiovascular Disease. Report of the Cardiovascular Review Group Committee on Medical Aspects of Food and Nutrition Policy. HMSO, London. department of health (1998), Report on Health and Social Subjects 48. Nutritional Aspects of the Development of Cancer. Report of the Working Group on Diet and Cancer of the Committee on Medical Aspects of Food and Nutrition Policy. Stationary Office, London. devereux, h. m., jones, g. p., mccormack, l. and hunter, w. c. (2003), ‘Consumer acceptability of low fat foods containing inulin and oligofructose’, Journal of Food Science, 68, 1850–1854. drewnowski, a. (1992), ‘Sensory properties of fats and fat replacements’, Nutrition Reviews, 50, 17–20. drewnowski, a. (1997a), ‘Taste preferences and food intake’, Annual Reviews of Nutrition, 17, 237–253. drewnowski, a. (1997b), ‘Why do we like fat?’, Journal of the American Dietetic Association, 97 (7), S58–S62. elmore, j. s., campo, m. m., enser, m. and mottram, d. s. (2002), ‘Effect of lipid composition on meat-like model systems containing cysteine, ribose, and polyunsaturated fatty acids’, Journal of Agricultural and Food Chemistry, 50, 1126–1132.

144

Modifying flavour in food

fukuwatari, t., kawada, t., tsuruta, m., hiraoka, t., sugimoto, e. and fushiki, t. (1997), ‘Expression of the putative membrane fatty acid transporter (FAT) in taste buds of the circumvallate papillae in rats’, FEBS Letters, 44, 461–464. gilbertson, t. a. (1998), ‘Gustatory mechanisms for the detection of fat’, Current Opinion in Neurobiology, 8, 447–452. gilbertson, t. a. and kim, i. (2002), ‘Chemoreception of fat’. In Chemistry of Taste, Given, P. & Paredes, D. (Eds.). ACS, Washington, DC, 180–191. gilbertson, t. a., fontenot, d. t., liu, l., zhang, h. and monroe, w. t. (1997), ‘Fatty acid modulation of K1 channels in taste receptor cells: gustatory cues for dietary fat’, American Journal of Physiology, 272, C1203–C1210. gunstone, f. d. and norris, f. a. (1983), Lipids in Foods: Chemistry, Biochemistry and Technology. Pergamon Press, Oxford. hoare, j., henderson, l., bates, c. j., prentice, a., birch, m., swan, g. and farron, m. (2004), The National Diet & Nutrition Survey: adults aged 19 to 64 years summary report Volume 5. HMSO, London. johnson, j. and vickers, z. (1993), ‘Effects of flavor and macronutrient composition of food servings on liking, hunger and subsequent intake’, Appetite, 21, 25–39. jones, s. a. (1996), ‘Physical, chemical and sensory aspects of fat replacement’. In Handbook of fat replacers, Roller, S., Jones, S.A. (Eds.). CRC Press, Boca Raton, FL, 59–86. kadohisa, m., rolls, e. t. and verhagen, j. v. (2005), ‘Neuronal representation of stimuli in the mouth: the primate insular taste cortex, orbitofrontal cortex and amygdala’, Chemical Senses, 30, 401–419. kawai, t. and fushiki, t. (2003), ‘Importance of lipolysis in oral cavity for orosensory detection of fat’, American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 285, R447–R454. kilcast, d. and clegg, s. (2002), ‘Sensory perception of creaminess and its relationship with food structure’, Food Quality and Preference, 13, 609–623. kinsella, j. (1969), ‘What makes fat important in flavours’, American Dairy Review, 31, 36–40. mattes, r. d. (2001a), ‘The taste of fat elevates postprandial triacylglycerol’, Physiology and Behavior, 74, 343–348. mattes, r. d. (2001b), ‘Oral exposure to butter, but not fat replacers elevates postprandial triacylglycerol in humans’, Journal of Nutrition, 131, 1491–1496. mattes, r. d. (2003), ‘Fat: the sixth taste – implications for health’, Food Australia, 55, 510–514. mattes, r. d. (2005), ‘Fat taste and lipid metabolism in humans’, Physiology and Behavior, 86, 691–697. meilgaard, m., civille, g. v. and carr, b. t. (1999), Sensory Evaluation Techniques. CRC Press, Boca Raton, FL. mela, d. j. and marshall, r. j. (1992), ‘Sensory properties and perceptions of fats’, In Dietary Fats: Determinants of Preference, Acceptance, and Consumption, Mela, D.J. (Ed.). Elsevier Applied Science, London, 43–57. miettinen, s. m., hyvonen, l., linforth, r. s. t., taylor, a. j. and tuorila, h. (2004), ‘Temporal aroma delivery from milk systems containing 0–5% added fat, observed by free choice profiling, time intensity, and atmospheric pressure chemical ionization-mass spectrometry techniques’, Journal of Agricultural and Food Chemistry, 52, 8111–8118. mottram, d. s., elmore, j. s., enser, m. and wood, j. d. (1998), ‘The effect of lipid composition on the flavour of beef’, Abstracts of Papers of the American Chemical Society, 215, 074-AGFD. overbosch, p., afterof, w. g. m. and haring, p. g. m. (1991), ‘Flavor release in the mouth’, Food Reviews International, 7, 137–184.

Formulating low-fat food: the challenge of retaining flavour quality

145

plug, h. and haring, p. (1993), ‘The role of ingredient–flavour interactions in the development of fat-free foods’, Trends in Food Science & Technology, 4, 150–152. roland, a. m., phillips, l. g. and boor, k. j. (1999), ‘Effects of fat replacers on the sensory properties, color, melting, and hardness of ice cream’, Journal of Dairy Science, 82, 2094–2100. rolls, e. t. (2004), ‘Multisensory neuronal convergence of taste, somatosensory, visual, olfactory and auditory inputs’. In The Handbook of Multisensory Processes, Calvert, G., Spence, C. & Stein, B (Eds.). MIT Press, London. rolls, e. t., critchley, h. d., browning, a. s., hernadi, i. and lenard, l. (1999), ‘Responses to the sensory properties of fat of neurons in the primate orbitofrontal cortex’, Journal of Neuroscience, 19 (4), 1532–1540. spielman, a. l., d’abundo, s., field, r. b. and schmale, h. (1993), ‘Protein analysis of human von Ebner saliva and a method for its collection from foliate papilla’, Journal of Dental Research, 72, 1331–1335. stubbs, r. j., harbron, c. g., murgatroyd, p. r. and prentice, a. m. (1995), ‘Covert manipulation of dietary fat and energy density: effect on substrate flux and food intake in men eating ad libitum’, American Journal of Clinical Nutrition, 62, 316–329. tuorila, h. (1992), ‘Preferences and attitudes related to fat-containing foods’. In Dietary Fats: Determinants of Preference, Acceptance, and Consumption, Mela, D.J. (Ed.). Elsevier Applied Science, London, 43–57. verhagen, j. v., rolls, e. t. and kadohisa, m. (2003), ‘Neurons in the primate orbitofrontal cortex respond to fat texture independently of viscosity’, Journal of Neurophysiology, 90 (3), 1514–1525. verhagen, j. v., kadohisa, m. and rolls, e. t. (2004), ‘Primate insular/opercular taste cortex: neuronal representations of the viscosity, fat texture, grittiness, temperature, and taste of foods’, Journal of Neurophysiology, 92 (3), 1685–1699.

8 New pungent and cooling compounds for use in foods C. C. Galopin, Givaudan Flavors Corporation, USA

8.1

Introduction

Spices and herbs have been used for a very long time to augment the pleasure of food consumption. In ancient times, they were primarily used to palliate the blandness of cereals and to hide the smell and taste of salted or spoiled meat. As food supply and storage conditions improved, they have been used to add aroma and taste sensation to cooked dishes. Two spices, mint and peppers, are of particular interest because they not only have a unique aroma but they also add mouth sensations, cold and burning, respectively, that are usually associated with touch rather than taste or smell. As a result, cold and burning sensations not only have value in foods, but also in pain- and congestion-relieving ointments, such as Capzasin® or Vicks VapoRub® (Eccles 1994; Robbins 2000). Hot peppers were also used in Mayan medicine and are believed to have antimicrobial properties (Cichewicz and Thorpe 1996). After a brief history, a discussion is provided as to why cold and hot sensations are neither aroma nor taste but belong to a different class, often called chemesthetics. Knowledge of physiology, relevant currently known chemicals and their structure–activity relationships are reviewed. The chapter concludes with future trends and with important references.

8.2

History

Peppers and mint have been used since prehistoric times. Use of red pepper from the capsicum family seems to date back to 7000 bc in South America, black pepper may have been used since 5000 bc in India and mint is described

New pungent and cooling compounds for use in foods

147

in Greek and Roman mythology.1 While varieties of mint have always grown on all continents, the origins of black pepper seem to trace back to India, while those of red pepper are clearly traced back to South America. Just like potatoes, red peppers were brought back by Christopher Columbus from the New World and were completely unknown to the Old World (Govindarajan 1982a). Unlike potatoes, however, red peppers were not readily adapted into European cuisine because they were thought too strong for European palates used to the milder black pepper and ginger. The only exception was in Hungary, where red peppers were used in a powder form, called paprika, which used to be much stronger than its modern version, as can be learned at the Paprika Museum of Szeged, Hungary. Asia and Africa, on the other hand, embraced the use of the strong red peppers in cuisines from the Sahara to the Far East. Nowadays, red peppers are considered typical of Mexican, Maghrebian, Indian, Chinese (Szechuan), Vietnamese, Thai and Indonesian cuisines. More information about the history, cultivation, production and consumption of red peppers and ginger can be found in the thorough, even though slightly outdated, reviews of Govindarajan and coworkers (Govindarajan 1982a,b, 1985, 1986a,b; Govindarajan et al. 1987; Govindarajan and Sathyanarayara 1991). For centuries, black and red peppers were used whole or ground. Their active pungent components, piperine (Pelletier 1821, 1832) and capsaicin (Thresh 1876; Nelson 1919), respectively, were isolated only in the nineteenth century. Because of their low concentration in peppers, piperine and capsaicin were only made commercially available through modern chemistry in the twentieth century (Crombie et al. 1955). Even today, capsaicin still remains an expensive chemical to make (Kaga et al. 1989, 1996) and it is often substituted by a simpler molecule, N-vanillyl nonamide, also called synthetic capsaicin, or used as capsicum oleoresin, the capsaicin-rich residue from the distillation of red pepper oil. Mint, on the other hand, has long been used as a steam-distilled oil. As a result, menthol, the cooling component of mint oil, has also been known for centuries because mint oils may contain up to 80% of menthol, which crystallises easily at ambient or cool temperatures. Menthol was originally called mint crystal and it is difficult to determine when it was first isolated pure and identified.

8.3

Fundamental differences between chemesthetics and flavours

8.3.1 Different concepts Chemesthetic (also called somatosensorial) sensations, such as cold and hot are not flavours. Indeed they are neither taste nor aroma. An easy way to 1

The Earth nymph Minthe is transformed into a plant by Hades (Pluto in Latin) to save her from the rage of the goddess Persephone (Proserpina in Latin). During the transformation, the plant would have captured the perfume of the goddess, which gives mint its subtle fragrance.

148

Modifying flavour in food

be convinced by this simple fact is to consider that chemesthesis occurs all over the body; it is not limited to the nose (like aroma) or to the mouth (like taste). The signalling mechanism is also different since cold and hot sensations involve transient receptor potential (TRP) ion channels (Koltzenburg 2004) while aroma and taste involve G-protein coupled receptors (GPCR), as explained in Section 8.4. Moreover, chemesthesis is related to the pain pathway (Caterina 2003), and signalling is conveyed by neither olfactory nor gustatory nerves but by the trigeminal nerve. Hot sensations can be sensed in every part of the body, including the nose, and some volatile hot/pungent chemicals, such as the isocyanates of mustard oils, can also be experienced in the nose. However, there is confusion around nasal pungency. Most chemicals will exhibit nasal pungency at some level (Cometto-Muñiz et al. 2005). That type of nasal pungency, which should be called nasal irritation, is a different trigeminal sensation that indicates a dangerous cell-damaging concentration of chemicals in the air. It follows a non-specific mechanism (Cometto-Muñiz and Cain 1994) that is unlikely to activate the same receptors as hot sensations. In a similar way, while cold sensations can be sensed by the nose, it is important to recognise that the anti-congestive effect of menthol is perceptual, not actual. It is possible that the cooling sensation is interpreted by the brain as increased airflow and is therefore perceived as decongestion, but it has been well established that inhalation of menthol has no effect on nose air flow (Eccles and Jones 1983; Eccles 1994).

8.3.2 Different evaluation requirements Now that we have established that chemesthesis is different from aroma, it should be understandable that it cannot be measured with the same sensory tools that have been developed for aroma evaluation. While this concept was considered shocking a few years ago, most researchers in the chemesthesis and taste area agree that specific mouth methods, such as half-mouth and half-tongue methods, are key to unequivocal measurements of chemical intensity (O’Mahony 1995). The following paragraphs will show why standard evaluation methods, such as threshold determination or Scoville point scale, are inaccurate and inappropriate for chemesthetics. First of all, the maximal activation of a receptor is compound-dependent (Galopin et al. 2005a). As a result, chemical A may activate a receptor at a low concentration and reach an early plateau, while chemical C may require a higher activation concentration but may also deliver a higher plateau (see Fig. 8.1). The use of sensory threshold would conclude that A is more potent than C, when in fact, at the appropriate dosage, C is far superior to A. Figure 8.1 also shows that two compounds of similar thresholds may have different intensities. Thomas Hofmann (Hofmann and Stark 2005) has recently shown that the activity of some chemicals with low threshold rapidly reached a plateau and that these chemicals do not play a major role in the taste of

New pungent and cooling compounds for use in foods

149

Dose–response curves

EC50 of A

Intensity

Threshold of A

Max of A

Compound A Compound B Compound C Concentration

Fig. 8.1

Comparisons of activities of fictitious compounds A, B and C.

botanicals despite their low threshold and high concentration. The fact that the correlation between detection threshold and intensity is satisfactory for aroma compounds should be considered as an exception rather than the rule. The intensity of chemesthetic and taste compounds must therefore be established with a full dose–response curve. To take the shape of the curve into account when comparing the intensity of several substances, we have recommended (Galopin et al. 2005a) the use of the maximal effect (Max) and the EC50 (efficient concentration at which 50% of the maximal effect is perceived) of the curve. To make comparison easier, both values can be combined into one ‘Power’ function either through a linear or logarithmic relationship (Equation 8.1). This function can also be normalised so that the power of a reference compound is equal to 100 (Equation 8.2): Power =

Max EC50

Power = 100

or Power = −Max × log ( EC50 )

Max  EC50   1 − log   Max ref   EC50 ref  

[8.1] [8.2]

Finally, hot and cold sensations have a unique feature that makes their evaluation more complex than that of aroma or taste. This feature is the time-dependency of their response. Because TRP receptors are not present

150

Modifying flavour in food

in taste buds but buried within the skin and mucosa tissues, chemicals need to penetrate through the tissue to be effective and then must ooze out for the signal to recess. Depending on the structure of the chemical and its affinity for the mucosa and for the saliva, some chemicals will penetrate quickly while others will be slower, therefore creating different sensation onsets. Most people have experienced the ‘surprise’ effect of some spicy foods that are mildly pungent at first and become intolerable after a few seconds. The delayed onset is a unique feature of chemesthetic sensations such as hot and cold. Appropriate evaluation methods must take into account that a taste solution must stay in the mouth long enough for the chemesthetic effect to take place and that the panellist should wait a set amount of time before giving their assessment of the sensation. Hot and cold sensations also have lingering effects, meaning that the sensation stays in the mouth even after swallowing. Lingering, however, is not unique to chemesthesis and is also known for taste compounds, such as artificial sweeteners and bitter compounds (Deis 2005). Therefore, accurate evaluation methods must include a mouth washing stage if several samples are to be tasted in the same session. Since the lingering effect is observed for lipophilic compounds that have a higher affinity to the mucosa than to saliva, a washing with water is not enough; a fatty solution, such as milk or yoghurt, is much more efficient. Even when great care is taken in the design of the sensory experiment, fatigue of the panellists can be observed to negatively impact the accuracy of the results when more than four tastings are scheduled in one session. Finally, chemesthesis also seems to have a ‘memory effect’ (Rousseau et al. 2002), which makes it difficult for panellists to assess the relative strength of samples that are tasted at more than 90 second intervals. Unfortunately, because of delayed onsets and lingering effects, sensory experiment designs must often include 5 to 10 minute lags between samples. Therefore, the best method is to taste both samples at the same time by using the half-mouth or half-tongue methods (O’Mahony 1995). Both methods involve the exposure of only one half of the oral cavity to each sample to be evaluated. Partial exposure is accomplished either by placing sample-soaked filter papers on one part of the tongue or by spreading tasting solutions on the inside of one cheek (Hofmann et al. 2004). When both samples have been applied to their respective location (usually left and right sides), the panellist is simply asked to rate which side is stronger. Statistical treatment of the panellists’ answers determines the d′ or R-index values, which show how different two samples are from each other. This method has numerous advantages. First of all, it eliminates the error caused by the great variation in sensitivity that panellists have for hot and cold sensations, since the method is only asking which compound is stronger, not how much stronger it is. Also, because both samples are in the oral cavity at the same time, the aroma that the samples may have has no influence on the results. For example, if sample A has a minty aroma while B is

New pungent and cooling compounds for use in foods

151

odourless, results of classical sensory methods may favour A as being more intense because the smell of mint enhances the perception of cooling. However, when A and B are tested in a half-mouth experiment, the minty smell of A enhances the perception of A and B in a similar way. In that instance, B may be found more strongly cooling, because the smell of A, although it is still interfering with the experiment, is no longer giving advantage to A. Finally, as with any sensory evaluation, it is important that results be recorded along with the matrix used for the experiment. Strong variations can be observed when chemicals are tested in water, milk, chewing gums, toothpastes or sauces. For example, the well-known cooling compounds WS-3 and WS-23 are known to show drastically different intensities depending on the matrix they are in. When tasted in water or in toothpaste, WS-3 is found to be about four times more potent than WS-23; but in chewing gums, WS-23 is often preferred to WS-3. This result can be explained only by major differences between the interactions of the compounds with the matrices they are dissolved in.

8.4

Physiology

As mentioned in the previous section, hot and cold sensations activate TRP (Nilius and Voets 2005). The family of TRPs involved in thermal sensations (Voets and Nilius 2003; Patapoutian 2005) has been dubbed thermoTRP and a list is given in Table 8.1. Several recent reviews and articles describe their modes of action (Calixto et al. 2005; Reid 2005). Some of these TRPs (TRPV1 and TRPA1 in particular) have been demonstrated to be nociceptors (Numazaki and Tominaga 2004; McKemy 2005; Tominaga 2005), which are receptors involved in the pain pathway (Dray 1992; Tominaga and Julius 2000; Tominaga and Caterina 2004; Scholtz 2002). The connection between Table 8.1 Known thermoTRPs and their temperature ranges Preferred name

Other name

Temperature range

Major references

TRPV1 TRPV2 TRPV3

VR1 VRL1

>43 ºC >52 ºC 34–38 ºC

TRPV4 TRPA1

ANKTM1

27–35 ºC <17 ºC

TRPM8

CMR1

25–28 ºC

Caterina et al. 1997 Caterina et al. 1999 Smith et al. 2002, Xu et al. 2002 Güler et al. 2002 Story et al. 2003, Jordt et al. 2004 McKemy et al. 2002, Peier et al. 2002 Talavera et al. 2005

TRPM5

Involved in GPCR cascade signalling

152

Modifying flavour in food

temperature sensation and pain is not surprising since the primal role of these sensors is to warn the body of potentially damaging hot or cold temperatures. TRPA1 is particularly intriguing since it is activated by low temperatures (Story et al. 2003) but it is also activated by pungent compounds from mustard oil (Bandell et al. 2004; Jordt et al. 2004; Bautista et al. 2005; McPherson et al. 2005). TRPA1 seems to be the receptor for painful hot and cold sensations (Namer et al. 2005; Obata et al. 2005; Reid 2005). TRPV1 is the nociceptor activated by capsaicin from chili peppers (Caterina et al. 1997). This link between hot and painful sensations sheds light on the longknown anaesthetic effect of capsaicin (Robbins 2000; Scholtz 2002) and it also shows that the confusion between hot and pungent sensations is natural. Nevertheless, trained panellists can differentiate between hot, pungent and painful sensations because the signal quality is different. Most pungent compounds will only activate one or two receptors (TRPV1, TRPV2 or TRPA1, for example), while a thermally hot sensation will activate the whole battery of TRPVs. A painful sensation will also certainly involve cell damage, which will activate another pain pathway. Therefore, what we call pungency is an incomplete hot/pain signal, which we have nicknamed a ‘pleasurable pain in the mouth’ (Galopin et al. 2004). The signalling of cold sensations seems to be simpler as it involves one main receptor, TRPM8 – also known as CMR1 for cold menthol receptor (McKemy et al. 2002; Peier et al. 2002). It is unlikely to be a nociceptor even though it has been found to be coupled with TRPV1 in some neurons (Okazawa et al. 2004). As a result, it is almost impossible to distinguish a cold sensation that is chemically activated from one that is thermally activated. Finally, other thermoTRPs, such as TRPM5, do not seem to be activated by temperature but are regulated by temperature (Talavera et al. 2005). The presence of TRPM5 in the GPCR cascade (see below) could explain the temperature dependency of taste sensations. TRPM5 may be the reason why over-sweetened beverages taste just right when served ice-cold. TRPs are sodium/calcium ion channels that are cell-membrane proteins with six trans-membrane α-helices (Fig. 8.2). It is believed that, when activated, the channel opens to let in a flow of cations that depolarises the cell and creates a neuronal signal. While this is the most widely accepted mechanism, some scientists have proposed alternatives (Reid 2005). Both N and C-termini of TRP are within the cell, which means that potential ligands must bind to the trans-membrane domain of the receptor since it is the only part that is accessible from the extracellular domain. The absence of an extracellular terminus suggests that the primary agonist of these receptors is not a chemical but rather a physical effect such as temperature or pressure. Taste or olfactory GPCRs, on the other hand, follow a completely different mechanism (Xu et al. 2004; Chen 2005). GPCRs are cell-membrane proteins with seven trans-membrane α-helices. As a result, they have one

New pungent and cooling compounds for use in foods Na+

Na+

Ca2+ Ca2+

Na+

Ca2+

Ca2+

Ca2+ Na+

Na+ Ca2+

Ca2+

Extracellular domain

153

Ca2+

Na+

Transmembrane domain

alciu

Fig. 8.2

Ca2+

e C

Calcium-sensitive dye

-sens

dy

m

itive

Intracellular domain

Transient receptor potential.

terminus in the extracellular domain (likely to be the binding site of their native ligands) and one terminus in the intracellular domain, which is coupled to a G-protein (such as gustducin). When the GPCR is activated, either by binding to the extracellular domain or to the trans-membrane domain, the G-protein breaks apart to start what is commonly referred to as the ‘GCPR cascade’. The mechanisms of sweet and umami receptors, which are known to be dimers, may even be more complex. While this cascade is still not fully understood, it eventually opens (a) cation channel(s) – like TRPM5 – that cause(s) depolarisation of the neuron signalling the brain. Even though GPCRs and TRPs are two completely different families of receptors, they share a similar final event, and therefore the activation of both types of receptor can be studied by cell clamping or calcium imaging (Gunthorpe et al. 2004). Calcium imaging is usually accomplished by incorporation of a calcium-sensitive dye into the neuron or host cell. When the concentration of calcium in the cell increases, the dye binds to the calcium and the dye–calcium complex emits a fluorescent light. The intensity of this light can be measured and is related to the amount of calcium the ion channel lets in and therefore to the degree of activation of the ion channel. Since TRPs are ion-channels themselves, the intensity of the fluorescent emission is directly correlated to the binding affinity of the tested ligand. This method can be fully automated (high-throughput screening or HTS) and allows the rapid screening of a large amount of chemicals without toxic impact on humans and panellist’s fatigue or unreliability. Nevertheless, assays must be validated by tasting. Indeed, an HTS that is used to screen irrelevant or badly designed libraries is bound to fail even if the screening

154

Modifying flavour in food

method is flawless (Lipinski 2000; Rishton 2003). Therefore, the sensory methods described in Section 8.3 may still currently be the best way to evaluate hot and cooling compounds; especially if sensory scientists apply knowledge of physiological mechanisms to the design of their experiments and the interpretation of their results.

8.5

Common pungent chemicals and their activity

The two following sections are only meant to be an introduction to some important compounds. Detailed information can be obtained from Leffingwell’s FlavorBase database by searching on the descriptive words ‘hot’, ‘warm’, ‘cool’ or ‘pungent’. Several reviews also exist on the subject (Galopin 2003; Eccles 1994).

8.5.1 Common compounds The number of pungent compounds on the market is fairly limited (Galopin 2003). Piperine and capsaicin are still the most widely used pungent chemicals. They are both available as natural ingredients (natural capsaicin is usually used as capsicum oleoresin). They can be used to create mild (piperine) to strong (capsaicin) pungent sensations. A recent compound from Takasago (Fig. 8.3), called TK1million, TK1MM, or HotAct® 1MM, is claimed to create an extreme sense of heat in the oral cavity (Nakatsu et al. 1996). Although it is too early to tell whether the chemical will be of value or not, its menthane ring was shown to overlap closely to the carbon chain of capsaicin (Green et al. 2004). Takasago has also recently revived an old compound, vanillyl butyl ether, as a cost-effective pungent chemical and cooling enhancer (Shiroyama et al. 2001). Some volatile hot/ pungent chemicals such as cinnamic aldehyde, methyl salicylate, gingerone or mustard oil components (isothiocyanates) can also be used to augment and round off the pungency of capsaicin and piperine. In recent years there has been a lot of interest in new pungent sensations (Katz 2005) such as tingling (Galopin et al. 2004) and warming without pain (Yang and Eilerman 1999). Some tingling compounds have been made available as natural extracts of Hua Jiao (sanhool) or Jambu (spilanthol). Several pungent dialdehydes, such as polygodial or aframodial (Szallasi and Blumberg 1996; Abe et al. 2002; Galopin 2003) have also been described but their complex structure may prevent their commercial use (Fig. 8.4).

8.5.2 Insight on structure–activity relationship Unusual to the field of food flavour, a wealth of information is available on structure–activity of hot/pungent compounds because this sensation is linked to the pain pathway, which has attracted the interest of pharmaceuti-

New pungent and cooling compounds for use in foods OH

OH

OMe

CO2Me

155

H N O (I) Capsaicin (Fema # 3404)

(II) Methyl salicylate (Fema # 2745) O O

S

N

C N

O

(III) 2-Phenylethylisothiocyanate (Fema # 4014) OH

(IV) Piperine (Fema # 2909) OH

O

OMe

OMe (V) Vanillyl butyl ether (Fema # 3796) OMe

O (VI) Zingerone (Fema # 3124) OMe H N

OH

O

OH

O O

O

(VII) HotAct 1MM (Fema # 3904) (VIII) Capsaicine in a menthane-like conformation

Fig. 8.3

Commercial hot/pungent flavour chemicals.

O

O O (IX) Galangal

N H

(X) α-Sanshool

OAc

OH

CHO CHO

O O (XI) Spilanthol

N H (XII) Aframodial

Fig. 8.4

Other published pungent chemicals.

156

Modifying flavour in food A-region HO

B-region

C-region

H N O O S

H N

A

A

O

Ph

A

N H

O

(XIII) A–NHCS–styrene

Fig. 8.5

(XIV) A–CONH–adamantane

Ph N N H H (XV) A–NHCONH– diphenylmethane

The ABC model and some basic BC combinations.

cal research, and has produced many publications on the activation of TRPV1. A model of particular interest is the ABC model developed at Sandoz (Walpole et al. 1993a,b,c), which decomposes capsaicinoid TRPV1 ligands in three regions (Fig. 8.5): 1. 2. 3.

A-region, or aromatic ring, is a benzyl or phenyl ring with a required substitution in para and an optional substitution in meta. B-region, or amide region, must be a polar group preferably with an H-bonding or enolisable carbonyl group. C-region, or chain region, is a lipophilic group which can be a saturated fatty chain of 4 to 12 carbons or a longer unsaturated chain. The C-group can also be replaced by a styrene – preferably in E configuration with para substitution, an adamantane or a diphenylmethane.

The same group at Sandoz has shown that this basic model could be applied to more complex compounds than capsaicinoids (Walpole et al. 1996) and the model can be used to simplify and optimise TRPV1 ligands (Janusz et al. 1993). This model also successfully predicts that capsaicinoates (where B is an ester) found in sweet peppers are much less pungent than capsaicinoids (Macho et al. 2003). While the research had been done without knowledge of Walpole’s work, we also had shown that the acid moiety (Cregion) of sanshool chemicals can be replaced by a cinnamic acid backbone (Galopin et al. 2004), which is consistent with the ABC model. Nevertheless, Fig. 8.6 shows that not all pungent chemicals fit the ABC model. As explained in Section 8.4, hot/pungent sensations are modulated by an array of receptors which, despite a strong sequence homology (Huang 2004), may each have individual binding requirements.

New pungent and cooling compounds for use in foods A-region B-region

C-region

A-region O

O

B

157 C

OAc

N H

HO

O

A-region

C-region

A-region B-region

C-region

O OHC

O

N OHC

O

Fig. 8.6

8.6

B O

Proposed applications of the ABC model to non-capsinoids.

Common cooling compounds and their activity

8.6.1 Common compounds The creation of cool flavours is more subtle than that of pungent flavour. It requires use of various matrices (toothpaste, mouth wash, candies, chewing gums, beverage, ointments) and the management of both intensity and lastingness (very desirable in oral care products but not in beverages, for example). As a result, there is a much wider variety of cooling compounds on the market and the research field is much more active. In fact, during the preparation of this chapter, new cooling entities have been patented (Sun 2006), a known compound (WS-5) has been commercialised (Erman 2005), an old compound (menthyl lactate) has been found in nature (Gassenmeier 2006) and several patent applications by the author are still to be published. For up-to-date information about cooling compounds, the very informative and well-referenced webpage http://www.leffingwell.com/cooler_than_ menthol.htm maintained by Leffingwell and Associates is recommended. The main natural cooling compound is l-menthol. It is the most potent and the most pleasant of all menthol isomers and of all natural cooling compounds (Enberger and Hopp 1988). According to rough estimates (Clark 1998) 9400 tons/year of natural l-menthol are produced in high purity (>99.5%) by crystallization from arvensis oil. Another 2500 tons/year are made synthetically by Symrise, through an impressive engineering process starting with m-cresol, and by Takasago, through an amazingly efficient chiral synthesis designed from myrcene by Nobel-laureate Ryoji Noyori (Akutagawa 1997). Invaluable information about menthol, its properties and syntheses can be found on the Leffingwell webpage http://www. leffingwell.com/menthol1/menthol1.htm and in several reviews (Leffingwell and Shackelford 1974; Hopp 1993; Croteau et al. 2005). Eucalyptol, menthone, camphor and highly purified isopulegol (sold by Takasago as CoolAct® P) are other commercially available natural cooling

158

Modifying flavour in food

O OH

H N

O

O OH

(XVI) l-Menthol (XVII) l-Menthyl lactate Fema # 2665 Fema # 3748 O

OH

HN

(XX) WS-23 Fema # 3804

O

O

OH (XXVI) Frescolat MPC Fema # 3806

Fig. 8.7

OH

O

(XXI) CoolAct 10 Fema # 3849

O

O (XIX) WS-5 Fema # TBA

O (XVIII) WS-3 Fema # 3455

O OH

O

O

H N

OH

OH (XXII) (XXIII) (XXIV) (XXV) Eucalyptol D-Menthone Isopulegol PMD-38 Fema # 2465 Fema Fema Fema # 2465 # 2962 # 4053 OH O O O NH N n OH O NO2

(XXVII) n = 2: Monomenthyl succinate, Fema # 3946 (XXVIII) n = 3: Monomenthyl glutarate, Fema # 4006

(XXIX) Icilin

Common cooling compounds.

compounds. However, they are much weaker and are usually used as cooling enhancers rather than the main cooling component of a flavouring composition (Fig. 8.7). Although it is not yet offered as natural by any reliable supplier, lmenthyl lactate has recently joined the list of nature-identical cooling compounds (Gassenmeier 2006). Menthyl lactate is slightly more potent than menthol; however, its value is not in its strength but in its taste profile, which is less bitter and less odorant than that of menthol. Its cooling profile is slightly delayed but lasts longer than that of menthol. Because menthyl lactate is an ester, however, it may hydrolyse in applications that are basic (carbonate toothpaste) or acidic (sodas). The application may then develop an off-flavour because of the release of lactic acid. Because many cooling compounds are in applications (oral care, cosmetics and confectionery) where artificial ingredients are common, there is a wide range of chemically designed cooling compounds to cater to many flavour creation needs.

New pungent and cooling compounds for use in foods

159

The most popular ones are clearly the Wilkinson-Sword chemicals WS-3, WS-23 and the recently commercialised WS-5. WS-5 is the most potent of the series but it is too early to comment on its profile and applications. One apparent drawback, though, is the presence of an ester group that may cause stability concerns. WS-3 is well established. It is about 1.5 times as strong as menthol. Its cooling profile is even more delayed than that of menthyl lactate and lasts for about 30 minutes. WS-3 is odourless and tasteless although it may display some bitterness and throat burning at high concentration. It can be used in any flavour (minty or not) either to bring cooling or simply to enhance freshness (of fruit flavours, for example). WS23 is also well established, odourless, tasteless and displays no throat burning. Its cooling profile is closer to that of menthyl lactate (slight delay and moderate lastingness). WS-23 is weaker than WS-3, however it is less lipophilic (log P = 2.7) than WS-3 (log P = 3). As a result, it seems to release faster from complex matrices (such as chewing gums) and may appear to be as potent. CoolAct 10 (log P = 2.1) and Frescolat MGA (log P = 2.7) are other cooling chemicals with reduced lipophilicity. Their cooling profile is similar to that of WS-23. Icilin is an old compound. Its cooling potency was discovered in academia by accident while assessing its toxicity (Wei and Seid 1983). The chemical is much more potent than any cooling chemical on the market. However, its tricky synthesis, its poor solubility and the presence of a nitro group make it an undesirable candidate for use in food or cosmetics. Icilin also seemed very different from all the known cooling compounds. However, no chemical research seems to have followed and icilin was archived as an oddity. Interest in icilin was revived recently when it was found to activate TRPM8 as well as TRPA1 (McKemy et al. 2002, McPherson et al. 2005). A nitro-free derivative was synthesised (Foster et al. 2004) but it is still unlikely that the icilin backbone would become commercially available. In recent years, the industry has realised that the profile of cooling compounds depended on the application where they were used. To develop more versatile products, many flavour houses are marketing mixtures of cooling chemicals in order to take advantage of their various qualities while exploiting the apparent additivity of their cooling effects (Erman et al. 2005, Galopin et al. 2006). For example, a mixture of menthol, menthyl lactate, WS-23 and WS-3 gives a cooling which starts immediately (thanks to menthol), reaches a plateau (as menthyl lactate and WS-23 kick in) and still lasts 30 minutes or more (long-lastingness of WS-3). The throat burning of WS-3 and the bitterness of menthol and menthyl lactate are avoided (because these negative effects are diluted) and the cost-in-use can be controlled (since menthol and menthyl lactate are much cheaper than WS-3 or WS-23). Enhanced cooling can also be obtained by a combination of cooling compounds with sub-threshold concentrations of hot pungent chemicals and vice versa (Kumamoto and Ohta 2002).

160

Modifying flavour in food

H N n

O

H N

OR

O O O

n OR

O

O

O (XXX) Hasegawa

(XXXI) Millennium (WS-5)

(XXXII) Firmenich

X H N n O (XXXIII) Qaroma

O

H N

OR

Y n

O (XXXIV) Givaudan

Fig. 8.8

N O

Z

O (XXXV) IFF (Fema # 4230)

Recently released cooling chemicals.

Finally several new chemicals have been introduced these past four years. Mono-menthyl succinate and glutarate, p-menthane-3,8-diol (PMD38), menthol propylene glycol carbonate (Frescolat MPC) seem to exhibit weak cooling. PMD-38 and mono-menthyl succinate have been approved as natural identical by IOFI but the status of mono-menthyl glutarate is still pending. Frescolat MPC seems to be a menthol precursor more than a cooling agent. Recently commercialised or recently published cooling compounds (Fig. 8.8) all seem to have been designed by expansion of the sidechain of the substituent at the 3-position.

8.6.2 Insight on structure–activity relationship In order to find odourless, stronger and longer-lasting compounds, researchers at Wilkinson-Sword started to create structure–activity relationship data as early as 1970 (Watson et al. 1978). That knowledge was augmented by the work of Haarman & Reimer, Takasago and Givaudan. Moreover, the recent discovery (Tsavaler et al. 2001; Zhang and Barritt 2004) of the role of the cold receptor TRPM8 in the apoptosis of cancer cells has accelerated the enrichment of this field (Reynolds and Polakis 2005). The basic Wilkinson-Sword model lists four requirements for cooling compounds: 1. 2. 3. 4.

A hydrogen bonding group. A compact hydrocarbon skeleton. A log P between 1.0 and 5.0. A molecular weight between 150 and 350 g/mol.

Because cooling sensations involve mostly one receptor, and in contrast with hot/pungent sensations, all common cooling compounds fit the basic

New pungent and cooling compounds for use in foods

161

Hydrogen bonding group ? OH O

O

O OH O O Log P: 2.65 Molecular weight: 256.34

N

O OH Log P: 2.77 Molecular weight: 228.33

NH

? Log P: 2.13 Molecular weight: 311.29

Compact hydrocarbon skeleton

H N

H N CH3 O Log P: 2.7 Molecular weight: 171.28

Log P: 3.0 O Molecular weight: 211.34 Hydrogen bonding group

Fig. 8.9

O2N

O

OH

OH Log P: 2.06 Molecular weight: 230.34

Fit of various cooling compounds to the Wilkinson-Sword model.

Wilkinson-Sword description. Even icilin may fit to the model somewhat (Fig. 8.9). Researchers at Wilkinson-Sword tried to improve cooling intensity by modifying the carbon skeleton. They explored more than ten different skeletons but none gave intensities as strong as the menthane skeleton (Watson et al. 1978). Recently, researchers at Givaudan (Galopin et al. 2005b), IFF (e.g. Fema # 4230), Hasegawa (Takazawa et al. 2004) and Qarôma (Sun 2006) have shown independently that enhanced intensities could be obtained by functionalising the side-chain of the molecule (Fig. 8.10). The IFF, Qaroma and Hasegawa side-chains are based on a three-carbon chain with an omegasubstituent while the Givaudan side-chain is based on a phenyl ring with a para-substituent, Fig. 8.10. While completely independent, these two approaches are in fact very similar since a three-carbon chain is a known bioisostere of a phenyl (Burger 1970, Lipinski 1986). However, the rotational flexibility of the three-carbon chain makes them much weaker than the Givaudan chemicals.

162

Modifying flavour in food

H N

H N

O X

O (XXXVII) Qaroma

(XXXVI) Givaudan

O

H N

O

O

O O

O (XXXVIII) Hasegawa

Fig. 8.10

8.7

N

(XXXIX) IFF (Fema # 4230)

Improved cooling chemicals have an expanded side-chain.

Future trends

It is common to hear that, in the area of hot/pungent and cold, trends are always towards stronger, more intense flavours (Finz 2006). While the author agrees to some extent, some nuances must be introduced, as well as a reality check about opportunity for new ingredients. For the past two decades, salsas have indeed become spicier and toothpastes have become more cooling. It seems that consumers never get enough sensation in their mouths. However, the impact on development of new hot or cold chemicals has been very different. In the area of cold sensation, this trend has created vibrant research from the food and flavour industry as well as from academia (see Section 8.6). The oral care and confectionery applications, which are the primary markets, have a high tolerance for artificial ingredients. Therefore, ingredients for these applications are judged on their profile not on their natural/artificial status. While there is a focus on highly potent cooling compounds, there also seems to be room for weaker cooling compounds in markets such as beverages (Landi 2005). Since products in this category are likely to be an extension of existing flavours, the profile of these coolants must be completely tasteless and odourless so they do not interfere with the established flavours. In the area of hot pungency, on the other hand, the need for increased intensity is not as obvious. The strength of capsaicin is so intense that it is rarely used at its full potential. More than intensity, research in pungent sensations should be focused on novelty and subtlety. For example, few

New pungent and cooling compounds for use in foods

163

good solutions exist for warmth without pain, tingling or bubbly sensations. The research field is in need of good natural lead compounds to work with. As a result, many opportunities are left to be explored and progress in the understanding of the physiology of touch may also help design new mouth sensations.

8.8

Conclusion

Cold and hot/pungent sensations are unique and ancestral mouth stimulations. Neither aroma nor taste, they require their own sensory evaluation methods, such as half-mouth methods. Because each chemical has a unique profile, comparison of two chemesthetic chemicals should be done by taking their whole dose–response curve into account; for example, by considering a Power function combining maximal and mid-point (EC50) values. They share similar physiology (thermoTRPs) and are both involved in the pain pathway (nociception). Many generally recognised as safe (GRAS) chemicals are commercially available, especially in the area of cold sensations. Even though a large amount of research already exists, there is still a high potential for discovery of new structures with improved profiles. Many scientific challenges still lie ahead; from the defying design of evermore potent coolants to the exciting discovery of new refined pungent sensations.

8.9

References

abe, m., ozawa, y., uda, y., yamada, y., morimitsu, y., nakamura, y. and osawa, t. (2002), ‘Labdane-type diterpene dialdehyde, pungent principle of Myoga, Zingiber mioga Roscoe’: Biosci. Biotechnol. Biochem., 66 (12), 2698–2700. akutagawa, s. (1997), ‘Enantioselective isomerization of allylamine to enamine: practical asymmetric synthesis of (−)-menthol by Rh-BINAP catalysts’ Topics in Catalysis, 4 (3,4), 271–274. bandell, m., story, g. m., hwang, s. w., viswanath, v., eid, s. r., petrus, m. j., earley, t. j. and patapoutian, a. (2004), ‘Noxious cold ion channel TRPA1 is activated by pungent compounds and bradykin’ Neuron, 41, 849–857. bautista, d. m., mohaved, p., hinman, a., axelsson, h. e., sterner, o., högestätt, e. d., julius, d., jordt, s. e. and zygmunt, p. m. (2005), ‘Pungent products from garlic activate the sensory ion channel TRPA1’ PNAS, 102 (34), 12248–12252. burger, a. (1970), in Burger A, Medicinal Chemistry, Wiley-Interscience, New York, Part 1, Chapter 6. calixto, j. b., kassuya, c. a. l., andré, e. and ferreira, j. (2005), ‘Contribution of natural products to the discovery of the transient receptor potential (TRP) channels family and their functions’ Pharmacol. Therapeut., 106, 179–208. caterina, m. j. (2003), ‘Vanilloid receptors take a TRP beyond the sensory afferent’ Pain, 105, 5–9. caterina, m. j., schumacher, m. a., tominaga, n., rosen, t. a., levine, j. d. and julius, d. (1997), ‘The capsaicin receptor: a heat activated ion channel in the pain pathway’ Nature, 389, 816–824.

164

Modifying flavour in food

caterina, m. j., rosen, t. a., tominaga, m., brake, a. j. and julius, d. (1999), ‘A capsaicin-receptor homologue with a high threshold for noxious heat’ Nature, 398, 436–441. chen, c. k. (2005), ‘The vertebrate phototransduction cascade: amplification and termination mechanisms’ Rev. Physiol. Biochem. Pharmacol., 154, 101–121. cichewicz, r. h. and thorpe, p. a. (1996), ‘The antimicrobial properties of chile peppers (Capsicum species) and their uses in Mayan medicine’ J. Enthopharmacol., 52, 61–70. clark, g. s. (1998), ‘Menthol’, Perfumer Flavorist, 23 (5), 33–46. cometto-muñiz, j. a. and cain, w. s. (1994), ‘Sensory reactions of nasal pungency and odor to volatile organic compounds: the alkylbenzenes’ Am. Ind. Hyg. Assoc. J., 55 (9), 811–817. cometto-muñiz, j. a., cain, w. s. and abraham, m. h. (2005), ‘Determinants for nasal trigeminal detection of volatile organic compounds’ Chem. Senses, 30, 627–642. crombie, l., dandegaonker, s. h. and simpson, k. b. (1955), ‘Amides of vegetable origin, Part V, synthesis of capsaicin’ J. Chem. Soc., 1025–1027. croteau, r. b., davis, e. m., ringer, k. l. and wildung, m. r. (2005), ‘(−)-Menthol biosynthesis and molecular genetics’ Naturwissenschaften, 92 (12), 567–577. deis, r. (2005), ‘How sweet it is: using polyols and high-potency sweeteners’ Food Product Design, October issue. dray, a. (1992), ‘Mechanism of action of capsaicin-like molecules on sensory neurons’ Life Sci., 51 (23), 1759–1765. eccles, r. (1994), ‘Menthol and related cooling compounds’ J. Pharm. Pharmacol., 46, 618–630. eccles, r. and jones, a. s. (1983), ‘The effect of menthol on nasal resistance to airflow’ J. Laryng. Otol., 97 (8), 705–709. enberger, r. and hopp, r. (1988), ‘Synthesis and sensory characteristics on menthol enantiomers and their derivatives for the use of nature-identical peppermint oils’ in Berger RG, Nitz S, Schrier P and Eichorn H Topics in Flavor Research, H. Eichorn, Marzling, Germany, 201–218. erman, m. (2005), ‘Physiological cooling compositions containing highly purified ethyl ester of N-[[5-methyl-2-(1-methylethyl)cyclohexyl] carbonyl]glycine’ US Patent Application 20050222256. erman, m., whelan, p. j. and snow, j. w. (2005), ‘Physiological cooling compositions’ US Patent Application 20050265930. finz, s. (2006), ‘America’s mean cuisine: more like it hot, from junk food to ethnic dishes, spicy flavors are the rage’ San Francisco Chronicle, 16 April. foster, a. j., van der logt, c. p. e. and tareilus, e. w. (2004), ‘Novel compounds and their uses’, US Patent Application 20040067970. galopin, c. (2003), ‘The world of non-volatiles: heating compounds for today and tomorrow’ in Le Quere JL and Etievant P Flavour Research at the Dawn of the 21st Century; proceedings of the 10th Weurman symposium, Beaunes, France, 610–613. galopin, c., furrer, s. and goeke, a. (2004), ‘Pungent and tingling compounds in Asian cuisine’ ACS Symposium Series, 867, 159–162. galopin, c., tigani, l. and furrer, s. (2005a), ‘Structure property relationship of chemosensates’, in Hofmann T, Rothe M and Schieberle P State-of the-art in Flavour Chemistry and Biology; proceedings of the 7th Wartburg Symposium, Eisenach, Germany, 61–67. galopin, c., krawec, p. v., slack, j. p. and tigani, l. (2005b), ‘N-substituted p-menthane caboxamides’ PCT Patent Application WO 2005049553 A1. galopin, c., moraes, e. and tigani, l. (2006), ‘Coolant solutions and compositions comprising the same’ US Patent Application 20060051301.

New pungent and cooling compounds for use in foods

165

gassenmeier, k. (2006), ‘Identification and quantification of l-menthyl lactate in essential oils from Mentha arvensis L. from India and model studies on the formation of l-menthyl lactate during essential oil production’ Flavor Fragrance J., 21 (4), 725–730. govindarajan, v. s. (1982a), ‘Ginger – chemistry, technology and quality evaluation, Part 1’ Crit. Rev. Food Sci. Nutr., 17 (1), 1–96. govindarajan, v. s. (1982b), ‘Ginger – chemistry, technology and quality evaluation, Part 2’ Crit. Rev. Food Sci. Nutr., 17 (3), 189–258. govindarajan, v. s. (1985), ‘Capsicum production, technology chemistry, and quality, Part I: History, botany, cultivation and primary processing’ Crit. Rev. Food Sci. Nutr., 22 (2), 109–176. govindarajan, v. s. (1986a), ‘Capsicum production, technology chemistry, and quality, Part II: Processed products, standards, world production and trade’ Crit. Rev. Food Sci. Nutr., 23 (3), 207–288. govindarajan, v. s. (1986b), ‘Capsicum production, technology chemistry, and quality, Part III: Chemistry of the color, aroma and pungency stimuli’ Crit. Rev. Food Sci. Nutr., 24 (3), 245–355. govindarajan, v. s. and sathyanarayana, m. n. (1991), ‘Capsicum production, technology chemistry, and quality, Part V: Impact on physiology, pharmacology, nutrition and metabolism; structure, pungency, pain and desensitization sequences’ Crit. Rev. Food Sci. Nutr., 29 (6), 435–474. govindarajan, v. s., rajalakshmi, d. and chand, n. (1987), ‘Capsicum production, technology chemistry, and quality, Part IV: Evaluation of quality’ Crit. Rev. Food Sci. Nutr., 25 (3), 185–282. green, c., pucarelli, f., mankoo, a. and manley, c. (2004), ‘Recreating flavors from nature’ Food Technol., 58 (11), 28–34. güler, a. d., lee, h., iida, t., shimizu, i., tominaga, m. and caterina, m. (2002), ‘Heatevoked activation of the ion channel TRPV4’ J. Neurosci., 22 (15), 6408– 6414. gunthorpe, m. j., rami, h. k., jerman, j. c., smart, d., gill, c. h., soffin, e. m., hannan, s. l., lappin, s. c., egerton, j., smith, g. d., worby, a., howett, l., owen, d., nasir, s., davies, c. h., thompson, m., wyman, p. a., randull, a. d. and davis, j. b. (2004), ‘Identification and characterization of SB-366791, a potent and selective vanilloid receptor (VR1/TRPV1) antagonist’ Neuropharmacology, 46, 133–149. hofmann, t. and stark, t. (2005), ‘Structures, sensory activity, and dose/response functions of 2,5-diketopiperazines in roasted cocoa nibs (Theobroma cacao)’, J. Agric. Food Chem., 53 (18), 7222–7231. hofmann, t., scharbert, s. and holzmann, n. (2004), ‘Identification of the astringent taste compounds in black tea infusions by combining instrumental analysis and human biorcsponse’, J. Agric. Food Chem., 52 (11), 3498–3508. hopp, r. (1993), ‘Menthol: its origins, chemistry, physiology and toxicological properties’ Rec. Adv. Tobacco Sci., 19, 3–46. huang, c. l. (2004), ‘The transient receptor potential superfamily of ion channels’ J. Am. Soc. Nephrol., 15 (7), 1690–1699. janusz, j. m., buckwalter, b. l., young, p., lahann, t. r., farmer, r. w., kasting, g. b., loomaus, m. e., kerckaert, g. a., maddin, c. s., berman, e. k., bohn, r. l., cupps, t. l. and milstein, j. r. (1993), ‘Vanilloids 1. Analogs of capsaicin with antinociceptive and anti-inflammatory activity’ J. Med. Chem., 36, 2595–2604. jordt, s. e., bautista, d. m., chuang, h. h., mckemy, d. d., zygmunt, p. m., högestätt, e. d., meng, i. d. and julius, d. (2004), ‘Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1’, Nature, 427, 260–265. kaga, h., miura, m. and orito, k. a. (1989), ‘A facile procedure for synthesis of capsaicin’ J. Org. Chem., 54, 3476–3477.

166

Modifying flavour in food

kaga, h., goto, k., takahashi, t., hino, m., tokuhashi, t. and orito, k. (1996), ‘A general stereoselective synthesis of capsaicinoids via orthoester Claisen rearrangement’, Tetrahedron, 52 (25), 8451–8470. katz, f. (2005), ‘Hot, cool and bubbly’ FoodProcessing.com, 21 February. koltzenburg, m. (2004), ‘The role of TRP channels in sensory neurons’ Novartis Found. Symp., 260, 206–220. kumamoto, h. and ohta, h. (2002), ‘Warming composition’ US Patent Application 20020142015. landi, h. (2005), ‘Beverage ingredient innovations’ Beverage World Nov/Dec. leffingwell, j. c. and shackelford, r. e. (1974), ‘Laevo-menthol – syntheses and organoleptic properties, cosmetics and perfumery’, Cosmetics Perfumery, 89 (6), 69–89. lipinski, c. (1986), ‘Bioisosterism in drug-design’ Annu. Rep. Medicinal Chem., 21 (VI-27), 283–291. lipinski, c. (2000), ‘Drug-like properties and the causes of poor solubility and poor permeability’ J. Pharmacolog. Toxicolog. Methods, 44, 235–249. macho, a., lucena, c., sancho, r., daddario, n., minassi, a., muñoz, e. and appendino, g. (2003), ‘Non-pungent capinoid from sweet peppers’ Eur. J. Nutr., 42, 2–9. mckemy, d. d. (2005), ‘How cold is it? TRPM8 and TRPA1 in the molecular logic of cold sensation’ Molecular Pain, 1, 16; www.molecularpain.com/content/1/ 1/16. mckemy, d, d., neuhausser, w. m. and julius, d. (2002), ‘Identification of a cold receptor reveals a general role for TRP channels in thermosensations’ Nature, 416, 52–58. mcpherson, l. j., geierstanger, b. h., viswanath, v., bandell, m., eid, s. r., hwang, s. w. and patapoutian, a. (2005), ‘The pungency of garlic: activation of TRPA1 and TRPV1 in response to allicin’ Current Biol., 15, 929–934. nakatsu, t., green, c. b., reitz, g. a. and kang, r. k. (1996), ‘4(1-Menthoxymethyl)2-phenyl-1,3-dioxolane or its derivatives and flavor composition containing the same’ US Patent 5 545 424. namer, b., seifert, f., handwerker, h. o. and maihöfner, c. (2005), ‘TRPA1 and TRPM8 activation in humans: effects of cinnamaldehyde and menthol’ NeuroReport, 16 (9), 955–959. nelson, e. k. (1919), ‘The constitution of capsaicin, the pungent principle of capsicum’ J. Am. Chem. Soc., 41, 1115–1123. nilius, b. and voets, t. (2005), ‘TRP chanels: a TR(i)P through a world of multifunctional cation channels’ Pflugers Arch. – Eur. J. Physiol., 451, 1–10. numazaki, m. and tominaga, m. (2004), ‘Nociception and TRP channels’ Current Drug Targets CNS Neurol. Disord., 3, 479–485. obata, k., katsura, h., mizushima, t., yamanaka, m., kobayashi, k., dai, y., fukuoka, t., tokunagu, a., tominaga, m. and noguchi, k. (2005), ‘TRPA1 induced in sensory neurons contributes to cold hyperalgesia after inflammation and nerve injury’ J. Clin. Invest., 115 (9), 2393–2401. okazawa, m., inoue, w., hori, a., hosokawa, h., matsumura, k. and kobayashi, s. (2004), ‘Noxious heat receptors present in cold-sensory cells in rats’ Neurosci. Lett., 359, 33–36. o’mahony, m. (1995), ‘Who told you the triangle test was simple?’ Food Qual. Prefe., 6 (4), 227–238. patapoutian, a. (2005), ‘TRP channels and thermosensation’ Chem. Senses, 30 (suppl. 1), i193–i194. peier, a. m., moqrich, a., hergarden, a. c., reeve, a. j., andersson, d. a., story, g. m., earley, t. j., dragoni, i., mcintyre, p., bevan, s. and patapoutian, a. (2002), ‘A TRP channel that sense cold stimuli and menthol’ Cell, 108, 705–715.

New pungent and cooling compounds for use in foods

167

pelletier, p. j. (1821), Ann. Chim. (Paris), 16 (2), 344. pelletier, p. j. (1832), Ann. Chim. (Paris), 51 (2), 199. reid, g. (2005), ‘ThermoTRP channels and cold sensing: what are they really up to?’ Pflugers Arch. – Eur. J. Physiol., 451, 250–263. reynolds, m. and polakis, p. (2005), ‘Trp-p8 active compounds and therapeutic treatment methods’ US Patent Application 20050090514. rishton, g. m. (2003), ‘Nonleadlikeness and leadlikeness in biochemical screening’ Drug Discovery Today, 8 (2), 86–95. robbins, w. (2000), ‘Clinical applications of capsaicinoids’ Clinical J. Pain, 16 (2), S86–S89. rousseau, b., stroh, s. and o’mahony, m. (2002), ‘Investigating more powerful discrimination tests with consumers: effect of memory and response bias’ Food Qual. Pref., 13 (1), 39–45. scholtz, a. (2002), ‘Mechanisms of (local) anaesthetics on voltage-gated sodium and other ion channels’ Br. J. Anaesth., 89, 52–61. shiroyama, k., sawano, k. and ohta, h. (2001), ‘Cool feeling composition’ US Patent 6 328 982. smith, g. d., gunthorpe, m. j. and kelsell, r. e. (2002), ‘TRPV3 is a temperaturesensitive vanilloid receptor-like protein’ Nature, 418, 186–190. story, g. m., peier, a. m., reeve, a. j., eid, s. r., mosbacher, j. n., hlricik, t. r., earley, t. j., heigarden, a. c., andersson, d. a., hwany, s. w., mclntyre, r., jegla, t., bevan, s. and papoutian, a. (2003), ‘ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures’ Cell, 112, 819–829. sun, h. (2006), ‘Novel compounds with physiological cooling effect’ US Patent 7 030 273. szallasi, a. and blumberg, p. m. (1996), ‘Vanilloid receptors: new insights enhance potential as therapeutic agent’, Pain, 68, 195–208. takazawa, o., watanabe. h. and iso, m. (2004), ‘p-Menthane derivative and cold sensing agent containing the same’ Japanese Patent Application 2004059474. talavera, k., yasumatsu, k., voets, t., droogmans, g., shigemura, n., ninomiya, y., margolskee, r. f. and nilius, b. (2005), ‘Heat activation of TRPM5 underlies thermal sensitivity of sweet taste’ Nature, 438 (7070), 1022–1025. thresh, l. t. (1876), ‘Capsaicin, the active principle of capsicum fruits’ Pharm. J. Trans., 7, 21. tominaga, m. (2005), ‘Molecular mechanisms of trigeminal nociception and sensation of pungency’ Chem. Senses, 30 (suppl. 1), i191–i192. tominaga, m. and caterina, m. j. (2004), ‘Thermosensation and pain’ J. Neurobiol., 61, 3–12. tominaga, m. and julius, d. (2000), ‘Capsaicin receptor in the pain pathway’ Jpn. J. Pharmacol., 83, 20–24. tsavaler, l., shapero, m. h., morkowski, s. and laus, r. (2001), ‘Trp-p8, a novel prostate-specific gene, is up-regulated in prostate cancer and other malignancies and shares high homology with transient receptor potential calcium channel proteins’, Cancer Res., 61, 3760–3769. voets, t. and nilius, b. (2003), ‘TRPs make sense’ J. Membrane Biol., 192, 1–8. walpole, c. s. j., wrigglesworth, r., bevan, s., campbell, e. a., dray, a., james, i. f., masdin, k. j., perkins, m. n. and winter, j. (1993a), ‘Analogues of capsaicin with agonist activity as novel analgesic agents; structure–activity studies. 1. The aromatic A-region’ J. Med. Chem. 36, 2362–2372. walpole, c. s. j., wrigglesworth, r., bevan, s., campbell, e. a., dray, a., james, i. f., masdin, k. j., perkins, m. n. and winter, j. (1993b), ‘Analogues of capsaicin with agonist activity as novel analgesic agents; structure–activity studies. 2. The amide bond B-region’ J. Med. Chem., 36, 2373–2380.

168

Modifying flavour in food

walpole, c. s. j., wrigglesworth, r., bevan, s., campbell, e. a., dray, a., james, i. f., masdin, k. j., perkins, m. n. and winter, j. (1993c), ‘Analogues of capsaicin with agonist activity as novel analgesic agents; structure–activity studies. 3. The hydrophobic side-chain C-region’ J. Med. Chem., 36, 2381–2389. walpole, c. s. j., wrigglesworth, r., bevan, s., campbell, e. a., dray, a., james, i. f., masdin, k. j., perkins, m. n. and winter, j. (1996), ‘Similarities and differences in the structure–activity relationship of capsaicin and resiniferatoxin analogues’ J. Med. Chem., 39, 2939–2952. watson, h. r., hems, r., rowsell, d. g. and spring, d, j. (1978), ‘New compounds with the menthol cooling effect’ J. Soc. Cosmet. Chem., 29, 185–200. wei, e. t. and seid, d. a. (1983), ‘AG-3-5: a chemical producing sensations of cold’ J. Pharm. Pharmacol., 35, 110–112. xu, h., ramsey, i. s. and kotecha, s. a. (2002), ‘TRPV3 is a calcium-permeable temperature-sensitive cation channel’ Nature, 418, 181–186. xu, h., staszewski, l., tang, h., adler, e., zoller, m. and li, x. (2004), ‘Different functional roles of T1R subunits in the heteromeric taste receptors’ PNAS, 101 (39), 14258–14263. yang, x. and eilerman, r. g. (1999), ‘Pungent principle of Alpinia galangal (L.) Swartz and its applications’ J. Agric. Food Chem., 47, 1657–1662. zhang, l. and barritt, g. j. (2004), ‘Evidence that TRPM8 is an androgen-dependent Ca2+ channel required for the survival of prostate cancer cells’ Cancer Res., 64, 8365–8373.

9 Controlled release of flavour in food products G. Reineccius, University of Minnesota, USA

9.1

Introduction

Protecting and ultimately delivering the aroma component of a food during eating are essential to creating the desired flavour perception. Unfortunately, aroma compounds are readily lost during storage of a food. They are inherently subject to loss through evaporation since by definition they must be volatile to be sensed by the olfactory system. They are also subject to loss through chemical reaction (e.g. with other aroma compounds, the food base, or oxygen) as also discussed in Chapter 14. Thus, in many food applications, it is desirable to use some method of encapsulation for the protection of food flavourings from these losses. Encapsulation is also used in some applications to afford a controlled release property to the flavouring. Controlled release is a means of further offering protection to a flavouring in that the flavour is ‘locked’ up in an encapsulation matrix until some point later in the processing, storage, home preparation or even consumption stage, when it is released for sensing. This delay in release can minimise flavour volatilisation, oxidation or chemical reaction, thereby improving the performance of the flavouring. This chapter provides an overview of the processes used to encapsulate flavourings and discusses controlled release in this context. Owing to the wide range in techniques used for this purpose, only major processes will be discussed. Broader discussions of these topics are available in more detailed reviews (e.g. Druri and Pawlik, 2001; Reineccius and Buffo, 2001; Vilstrup, 2001; Uhlemann et al., 2002; Porzio, 2004; Sébastien, 2004).

170

Modifying flavour in food

9.2

Industrial approaches for protecting flavourings from deterioration

The primary industrial approaches used to encapsulate flavourings include spray drying, extrusion, coacervation and cyclodextrins. Spray drying is by far the largest volume process of these methods.

9.2.1 Spray drying Spray drying is a reasonably simple and well-established unit operation in the flavour industry (about 70 years old). It begins with dissolving the encapsulation matrix (most commonly gum acacia, modified starch or starch hydrolysate alone or as blends with glucose, sucrose or other mono-/ disaccharides) in water at the highest solids level that is still readily atomised (e.g. gum acacia at 30–35%, modified starch (40–45%) or starch hydrolysate (40–55%)). The highest practical solids level is used since minimising the amount of water that needs to be evaporated during spray drying increases the retention of volatile flavourings and increases production capacity. The hydrocolloid is often heated to facilitate hydration. Flavouring is added to the solubilised encapsulation matrix and, if the flavouring is water insoluble (as most are), the ‘solution’ is homogenised. Homogenisation is a critical step in that it breaks the flavour emulsion into smaller sized droplets that result in better flavour retention during the drying process (Soottitantawat et al., 2003; Risch and Reineccius, 1988). The emulsion is fed into a spray dryer using dryer inlet air temperatures ranging from about 160 to 260 ºC and exit air temperatures ranging from 80 to 100 ºC. The flavour emulsion is atomised using either a single fluid spray nozzle or a centrifugal wheel atomiser. The dryer using a single fluid spray nozzle is taller and of narrower diameter while the dryer using a centrifugal wheel atomiser is shorter and wider. These drying chamber configurations accommodate the spray direction of the respective atomisers. The emulsion dries very quickly (in seconds) owing to the very large surface to volume ratio of the atomised emulsion. The cooled air (from evaporating water) is separated from the powder in a cyclone (or a bag house). The powder is sieved and packaged. If one prepares the infeed emulsion properly and operates the spray dryer under optimal conditions, the retention of volatiles is quite good for all but the most volatile compounds (Reineccius, 2004). Retention generally is close to 95% for aroma compounds with six or more carbons but may drop to 20% for small molecular weight volatiles (e.g. acetaldehyde). This range in volatile retention results in the powdered flavouring not being an identical match for the liquid flavouring going into the dryer which can be problematic when a customer likes a liquid flavour and this flavour cannot be reproduced in the dry form. While the retention of volatiles during spray drying is an important consideration in the manufacture of flavourings, one is also concerned

Controlled release of flavour in food products

171

about the protection afforded to a flavouring by the spray dried capsules: any flavouring containing a citrus oil is very prone to oxidation. The primary determinant of oxidative stability of spray dried flavourings is the encapsulation matrix used in the infeed emulsion. The inclusion of some proportion of low molecular weight sugars (e.g. 10–20% by weight of glucose, sucrose or corn syrup solids) into a gum acacia, modified starch or maltodextrin matrix greatly enhances protection against oxidation (Anandaraman, 1984; Anandaraman and Reineccius, 1986; Westing et al., 1988; Baisier and Reineccius, 1989).

9.2.2 Extrusion The use of extrusion for the manufacture of food flavourings is also a wellestablished process in the industry (used for about 50 years). The process initially involved dispersing a flavouring in a hard candy melt, allowing it to harden and then milling the product to the desired particle size. The process quickly evolved into extruding the hard candy melt through a multiple hole die into a cold isopropanol bath (Westing et al., 1988). The extruded molten carbohydrate mass (maltodextrin, sucrose, emulsifier and flavouring) solidified rapidly in the bath and the hardened strands were broken into pieces about equal in length to their diameter through vigorous mixing. This process has ultimately been adapted to a twin screw extruder which has offered a significant advantage in that much lower moisture contents can be used (Benczedi and Bouquerand, 2003; Porzio and Zasypkin, 2004). This minimises the need to remove water from the extruded product, thereby eliminating the isopropanol bath. Innovations in the process have resulted in much lower manufacturing costs, making it competitive with spray drying. This type of flavouring offers the advantages of large particle size and excellent protection against oxidation (Westing et al., 1988). As will be mentioned later, the matrix can be designed to offer some controlled release properties. The primary disadvantages of the process are poor emulsion stability and large particle size (if this is not desirable).

9.2.3 Cyclodextrins Cyclodextrins are cyclic polymers of glucose available as 6 (α), 7 (β) or 8 (γ) membered rings. When we speak of flavour encapsulation by this technique, we refer to the formation of ‘inclusion complexes’ or ‘molecular encapsulation’. In effect, the centres of the cyclodextrin rings are hydrophobic and, thus, the hydrophobic portion of a flavour molecule is free to move into the centre of the cyclodextrin. The strength of this complex varies with the chemistry and geometry of the flavour molecule in relation to the cyclodextrin. This complex generally comprises a single molecule of flavouring per single cyclodextrin molecule. This limits the flavour loading to the ratio

172

Modifying flavour in food

of flavour molecular weight to the molecular weight of the cyclodextrin : flavour complex: this is often 10–15%. Cyclodextrin complexes are typically manufactured by dispersing the desired cyclodextrin in water at 25– 40% solids and then adding the flavouring. The cyclodextrin : flavouring suspension is stirred for some time period to allow the flavouring components to move into the cyclodextrin cavity and then is spray dried to remove the water (Reineccius et al., 2002). Cyclodextrins are selective in which molecules they include. In order for a flavour compound to be incorporated into a cyclodextrin, it must have a suitable molecular geometry: the molecule must fit into the cyclodextrin cavity. Interestingly, eugenol is included very effectively into β-cyclodextrin but isoeugenol is not – it does not fit. Since the different cyclodextrin forms have a different number of glucose units, the cavities of these cyclodextrins vary and thus one type of cyclodextrin may bind one flavour molecule better than a different cyclodextrin (Goubet et al., 2001; Reineccius et al., 2002, 2003). The primary advantages of cyclodextrins are the oxidative stability they impart to any molecule that is included therein, and the uniqueness of a molecular inclusion. In terms of oxidative stability, molecules that are included in a cyclodextrin generally are not susceptible to oxidation (Westing et al., 1988; Seck et al., 2000). The molecular inclusion property results in one molecule being effectively isolated from another molecule. This fact results in the stabilisation of molecules that might react with themselves to form dimers. No other mechanism of encapsulation offers this advantage since they pool flavour compounds in droplets (when insoluble) or dispersed in the carrier (when soluble). The primary weaknesses of cyclodextrins include cost, limited release in an application and their selective inclusion.

9.2.4 Coacervation Coacervation has a long history of use in the carbon-less paper industry. In this application, parts of the ink system are encapsulated separately by coacervation and coated onto a sheet of paper. As one writes on the paper, the ink capsules are crushed by the pen point, allowing the ink parts to mix, which results in colour. This process has had some early application in the food flavouring field but it was not particularly successful. The need for controlled release flavourings has renewed development efforts in this technique and it is now finding use in the flavour industry. Coacervates may be produced by different processes: ‘complex’ coacervation is used for flavourings (Soper, 1995; Wampler et al., 1998; Soper et al., 2000). Complex coacervation is a specific coacervation process that involves the use of two different hydrocolloids that come together at the

Controlled release of flavour in food products

173

droplet (flavouring droplet) interface to form a complex, i.e. the droplet wall. The hydrocolloids most commonly used in flavouring applications are gelatin and gum acacia. The process may be described by the following steps: 1.

2.

3.

4.

5.

6.

Disperse both hydrocolloids (2–10% solids) in water separately. (The solutions must be at or above the the solidification temperature of the gelatin solution.) To one of the hydrocolloid dispersions, the flavouring is added with gentle stirring. (The shear from the stirring breaks the flavouring materials into droplets, the diameter of which determines the final encapsulate size since a wall is established around these droplets.) The second hydrocolloid dispersion is added to the flavouring : hydrocolloid dispersion and the pH is lowered. (This results in the gelatin being positively charged and the gum acacia negatively charged. The hydrocolloids come together to form a wall around the flavouring droplets.) The dispersion is diluted with water and allowed to cool. (This dilutes the gelatin so, when the dispersion cools, the gelatin concentration in the bulk phase is too low to gel the overall mass but a fragile solid gelatin : gum acacia capsule is formed around the flavouring droplets. At this stage the flavouring may be sold as a suspension or be crosslinked and dried.) While the soft capsules may be dried without cross-linking, the most durable (and insoluble) capsules are produced by chemically crosslinking the capsules. Traditionally, this is done by the addition of glutaraldehyde and pH adjustment (to allow reaction). The capsules can then be dried by various techniques to result in dry capsules.

Coacervation produces a water-insoluble capsule that has controlled release properties. While one can envision rupture as a release mechanism, diffusion and dissolution at elevated temperatures are also important. While the coacervate particles that have been cross-linked do not dissolve, as do the capsules formed in most other encapsulation processes, the capsule walls become permeable as they pick up moisture in a food application. Thus, the flavouring slowly diffuses from the capsule core to the food product. This delays release of the flavouring. In addition to offering controlled release properties, coacervation produces the core/shell structure. This structure allows for very high flavour loadings (claimed up to 90% loading). It also permits relatively good control of particle size and size range. Since the size of the original emulsion determines the final particle size, one can create particles with well-controlled size and a tight distribution around the mean.

174

Modifying flavour in food

Table 9.1 Influence of particle size on coating required to apply a 10 µm thick layer if coating is uniformly applied to particles of the specified diameter (Jones, 1995) Particle diameter (mm) 4 1 0.5 0.074 0.044

9.3

Coating amount in final product (%) 1.18 4.49 8.75 45.1 63

Industrial approaches to achieve controlled release of flavourings

In the sections above, there has been some discussion of controlled release. Unfortunately, the major flavour encapsulation process (spray drying) offers no controlled release properties other than through moisture pick-up by the particle. Increasing moisture without particle dissolution results in a diffusion-controlled release dependent upon the moisture content and carrier properties, while increasing moisture content to the point of dissolution results in immediate release. The spray drying process requires that the carrier material be water soluble (the manufacturing vehicle) and thus water is the only possible mechanism of release. Imparting controlled release properties to a spray dried flavouring requires secondary processing to be applied to the dry flavouring, e.g. applying a coating. Coacervation and extrusion both impart controlled release in the process itself through cross-linking (coacervation) the particle matrix or through matrix choice (extrusion and coacervation). The means of imparting controlled release to dry flavourings will be discussed in the following sections.

9.3.1

Secondary coatings (centrifugal disk, fluidised bed and granulation) Secondary coatings affording different release mechanisms (e.g. thermal or diffusion) can theoretically be applied to any dry flavouring. However, coating particles that are small and/or low in density is problematic. While the ideal product specifications for optimal coating are dependent upon the process, particles >100 µm in diameter and as dense as possible are generally more efficiently coated. The importance of particle size on the amount of coating needed to provide controlled release is illustrated in Table 9.1. From these data it is clear that a lot of material is required to coat spray dried flavourings (average particle size is about 40 µm). Encapsulation processes which yield larger particle sizes (e.g. extrusion or coacervation) require much less coating material. With this introduction, the three most common

Controlled release of flavour in food products Concentrated suspension

Thin fluid film

175

Coated core particles

Small droplets of excess coating

Fig. 9.1 Apparatus used to accomplish centrifugal-suspension coating. (Source: Sparks and Mason, 1987.)

means of coating particles will be discussed, i.e. centrifugal-suspension coating, fluidised bed and granulation. Centrifugal-suspension coating In this process, the dry flavouring is mixed with molten fat (or another matrix that solidifies on cooling) and is metered onto a spinning disc (Sparks and Mason, 1987; Sparks et al., 1995; Fig. 9.1). The fat and flavouring are thrown from the disc and travel through ambient temperature air. As the particles and fat are going through the air, the fat solidifies. The fat without a flavour particle forms a smaller particle and does not have the same momentum as fat containing a flavour particle; thus, the ‘loaded’ fat droplets fall further from the disc and are thereby separated from the unused coating material. While this process has been available for several years, this author has not had the opportunity to see any products made with it or evaluate its coating efficiency. The process is attractive in that it is a continuous process as opposed to the alternative coating methods which are batch methods. Fluidised bed coating Fluidised bed coating is very commonly used in the pharmaceutical industry to apply controlled release coatings to pharmaceuticals. Fluidised beds are batch operations which are costly but afford excellent control of the process (Jones, 1995). A fluidised bed works by initially fluidising a dry flavouring and then spraying the coating material onto it. The coatings may be a fat or other edible materials (e.g. insoluble protein or shellac; Fig. 9.2). If fat is used, the air temperature must be chosen to permit the solidification of the fat in the coating chamber. If other water-insoluble coatings are used, then issues of solvent flammability (e.g. ethanol) become an issue. These systems may be designed to be explosion proof, permitting the use of non-aqueous solvents but costs become limiting for food applications.

176

Modifying flavour in food

Controlled particle flow

Coating partition

Nozzle (hydraulic or pneumatic)

Coating spray Air flow

Air distribution plate

Fig. 9.2

Diagram of a bottom coater fluidised bed (Source: Courtesy of Glatt Air Technologies.)

As mentioned earlier, coating spray dried particles is problematic because of their small size and low density. This is particularly true for fluidised bed coating. The very small spray dried flavouring particles are readily lost from the coating system owing to their small size. They penetrate the air filtering system used to separate the fluidising air from the product, which results in product loss. Also, the material being sprayed onto the spray dried flavouring (e.g. atomised fat) is approximately of the same dimensions as the spray dried particles. The fat will adhere as a droplet to the spray dried flavouring but not spread over it to coat it. Thus, one does not get a coating but droplets of fat stuck on the surface of the dry flavouring, resulting in powder agglomeration. Once the agglomerated particles reach a larger size (e.g. 150 µm), then one begins to effect a coating operation. The product thereby obtained is an agglomerate of fat and flavouring that is not effectively coated until the particles reach 200–300 µm (Fig. 9.3). This may be too large for a specific application or dilute the flavouring too much to offer the intensity desired. Despite these limitations, coating a spray dried flavouring with fat adds controlled release properties to the flavouring. A paper on the effect of encapsulation and fat coating on the retention of a model flavouring during cookie preparation and baking was published

Controlled release of flavour in food products

177

Fig. 9.3 Electron micrograph of fat-coated flavour (50% coating, each line in figure is 10 µm).

Retention (in % of [c] before baking

100

Ethanol

80 N-Lok

60 40

N-Lok/fat (29%)

20 Gum acacia

e hy

Gum acacia/fat (23%)

Et

ap lc Et hy

lc

ry

la

ap ra t

te

ro at e ap lc

Et hy

lb Et hy

Et

hy

la

ce

ut yr

ta t

at e

e

0

Fig. 9.4 Retention of a series of ethyl esters when added to cookie dough in different flavour forms and then baked. (Source: Heiderich and Reineccius, 2001.)

by Heiderich and Reineccius (2001). While the encapsulated flavourings (no fat coating) are soluble in water and one would expect their release in the cookie dough, they were retained substantially better during subsequent baking than simply adding the model flavour compounds in liquid form (Fig. 9.4). It appears that there was not enough moisture in the cookie dough to release the flavouring until later in baking and thus they were retained better. The carrier used in encapsulation influences the retention of esters as well. Gum acacia appears to provide better retention of all of

178

Modifying flavour in food

the esters compared with using N-Lok (chemically modified starch blend from National Starch, Bridgewater, NJ) as the encapsulating matrix. There are no data to aid in speculation as to why this difference in performance is observed. The addition of controlled release properties, i.e. a fat coating, to the spray dried flavourings again increased their retention during baking (Fig. 9.4). The effect was most pronounced for the most volatile esters (ethyl acetate and ethyl butyrate) and was insignificant for the least volatile esters. The implications of this is that often the most volatile aroma compounds carry the freshness of a flavouring and these notes are very readily lost during baking. This leaves the baked product strawberry flavoured, for example, but not ‘fresh’ strawberry flavoured. Thus, the differences we observe for encapsulated flavouring (or those with controlled release properties) may have significant implications in how one achieves ‘fresh’ flavours in baked products. Granulation Granulation offers a means of producing larger particles that may form a product suitable for secondary coating using additional processing, or be a means of applying a coating to an existing powdered flavouring (Uhlemann et al., 2002). In a spray granulator (not pictured), the entire process is carried out in a specially designed spray dryer (incorporates a fluidised bed). Particles remain in the drying chamber, colliding with atomised infeed and thereby continuing to increase in size, initially by agglomeration and finally layering, until the desired particle size is obtained. Since this is a continuous, one-step process with the dryer infeed serving as the agglomerating liquid, there is no opportunity to apply a coating different from the basic particle wall matrix. Thus, one cannot apply a unique coating to the particles to impart controlled release properties. However, granulation in a rotor granulator (Fig. 9.5) may apply a secondary coating to an existing dry flavouring. This is done in an apparatus similar to a fluidised bed coater (Wurster coater) but the rotation of the particles in the apparatus yields a smoother, more spherical particle structure. Like fluidised bed coaters, this process suffers when one wishes to coat very small or low-density particles.

9.3.2 Insoluble capsule matrices Some processes used in flavour encapsulation do not require that the wall material be water soluble (e.g. extrusion, centrifugal extrusion, spray chilling or matrix solidification), or the wall matrix may be made insoluble through its chemical cross-linking (coacervation). These processes permit the manufacture of controlled release encapsulated flavourings (not released by moisture) without secondary processing. This approach has an inherent cost benefit over the processes that require secondary processing. These

Controlled release of flavour in food products

179

Exhaust air

Spray solution

Drying air

Granulated product

Fig. 9.5

Schematic of a rotor granulator. (Source: Uhlemann et al., 2002.)

processes will be briefly discussed in terms of their controlled release properties. Insoluble wall matrices If an encapsulation process uses water as the manufacturing vehicle (e.g. spray drying), then one must use a water-soluble wall matrix and flavour release is due to water absorption. However, if the encapsulation process does not require water as the manufacturing vehicle, then one can make flavourings where the release mechanism is not water absorption. Encapsulation processes that can use water-insoluble matrices include extrusion (partially soluble carbohydrates or proteins; McIver et al., 2005), centrifugal-extrusion or centrifugal-suspension coating (congealable matrix such as gelatin, meltable resin or fat; Schlameus, 1995), spray chilling (fat or gelatin) or matrix solidification (e.g. gelatin or meltable resin). The extrusion process has been presented earlier in this chapter and mention made of its ability to use less water-soluble matrices. The process of matrix solidification can be applied in several different ways. A common practice is to dissolve (or disperse) the flavouring in a molten matrix (fat, gelatin or resin) and then spray it into a cold spray dryer (may be termed ‘prilling’ or ‘spray chilling’). The infeed material is atomised to produce small particles which then solidify as they cool, falling through the dryer to yield a free-flowing particulate. One can appreciate that each of the potential matrices offers unique release properties. If one uses a solid

180

Modifying flavour in food

fat, one usually considers melting as the release mechanism (Uhlemann et al., 2002). However, there is reasonable question as to the protection during storage and release in application one can achieve in a fat encapsulated matrix. Since most flavouring substances are fat soluble, they would readily migrate from the fat capsules to the bulk food system and undergo chemical reaction, evaporative losses or oxidation. Also, there are no data to confirm that flavour release is significantly more rapid when the fat is a solid or on melting. Flavour release is dictated by diffusion from the fat irrespective of the physical form. Thus, this author questions the value of fat encapsulation for controlled release of flavourings. In the case of non-fat-based matrix solidification systems, the primary release mechanism is diffusion controlled. Since the matrices are water insoluble and meltable (depending upon the application), release is controlled by the permeability of the flavouring to the particle matrix. The presence of moisture in an application (e.g. a baked good or a chewing gum), typically increases the permeability of the matrix and thus flavourings are slowly released in application. This slow rate of release may offer substantial protection to the flavouring from harsh processing environments. Rendering wall matrices insoluble In the case of coacervated flavourings, the wall materials (typically gum acacia and gelatin) are water soluble but appear to remain around the flavouring droplets even in aqueous systems at ambient (or lower) temperatures without cross-linking. As is noted later, uncross-linked capsules afford some controlled release properties. Chemically cross-linking the wall gelatin late in the process renders the capsule insoluble and affords greater capsule integrity and, therefore, controlled release. The release mechanism in these products, whether cross-linked or not, is typically diffusion (as opposed to rupture as is used in the carbonless paper industry). The uncross-linked coacervates may also release flavourings through dissolution of the wall in thermal applications. As noted earlier, glutaraldehyde has been traditionally used for crosslinking. The toxicity of free glutaraldehyde has resulted in a search for alternative cross-linking agents, i.e. those that are less toxic and food ‘friendly’. A patent has been issued where enzymes have been used for this purpose (Soper and Thomas, 1998). Controlled release properties of coacervated flavourings have been demonstrated for the release of garlic oil in baked goods (Graf and Soper, 1996, Miller et al., 1997) and a general model application for thermally processed foods (Yoon et al., 2005). The objective in the patents on baked goods was to delay the release of garlic oil into a yeast-leavened baked product to permit proper proofing. The direct addition of garlic oil to a flour-based yeast-leavened product results in a loss of loaf volume (inadequate rising due to an effect on gluten formation). Incorporating a coacervated garlic

Controlled release of flavour in food products

181

oil into the dough sufficiently delayed the release of the garlic oil to allow normal proofing. Yoon et al. (2005) dealt more generally with the issue of flavour release in thermally processed foods. They used the standard gelatin : gum acacia wall system and no cross-linking. They noted that significant temperature initiated release of their model system (essential oils with a dye marker). They hypothesised that this release was due to a melting of the wall matrix (gelatin). Salt was also found to initiate release through disruption of the gelatin : gum acacia complex.

9.4

Needs in flavour encapsulation/controlled release

9.4.1 Needs in flavour encapsulation A recent survey of food industry needs listed heat stable flavours among the most important. Flavour stability and controlled delivery were also listed as important. Current encapsulation methods address these issues only to a limited extent. In a review on flavour encapsulation methods published by Uhlemann et al. (2002), only 3 of 11 processes discussed offered any mechanism of release other than moisture. The only thermal release process mentioned was spray chilling (fat) and this author questions the effectiveness of this method for this purpose. They did not list coacervates as being a thermal release product. Thus, we need encapsulation methods to isolate flavourings from the harsh environment during thermal processing. A prime example is the production of extended shelf-life flavoured milks. These products are often processed using direct steam injection heating. The steam and subsequent vacuum treatment (to remove added water) are extremely harsh on flavourings. Ideally, we need technologies that isolate the flavouring from harsh environments, and even the food base itself. Many flavour components are very reactive and are readily lost through chemical reactions with food components. Coacervation offers some thermal protection to foods but is acceptable only where the presence of capsules is unnoticed (e.g. baked goods as opposed to beverages, or capsules are sufficiently small). We do reasonably well in protecting flavourings in the dry form. For example, cyclodextrins (when suitable), extrusion and spray drying have all proven to be effective in protecting flavourings from oxidation. Unfortunately, some flavouring materials (e.g. diacetyl) migrate readily through most of our standard wall materials even when stored at low relative humidities. It is extremely difficult to retain diacetyl in any product during storage. Some other compounds such as thiols or 2-acetyl-1-pyrroline react with each other (or dimerise) to result in undesirable flavour changes. The only encapsulation process that isolates flavour compounds as individual molecules are the cyclodextrins. But some molecules do not fit into the cyclodextrin cavities (α, β or γ forms) and thus are not protected. Also

182

Modifying flavour in food

cyclodextrins do not readily give up molecules that are complexed well. It would be desirable to have a more generally inclusive encapsulation method that individually protected flavouring molecules. In terms of materials for encapsulation (specifically coacervation), we have few choices for the cationic hydrocolloid in coacervation. Gelatin has largely been the only choice for this purpose (we have many anionic hydrocolloids). Gelatin is problematic in that it is not kosher (most gelatins are made from pork, although gelatin from beef and fish has been used), is unacceptable to vegetarian consumers and is a protein. Proteins lack solubility at low pHs and thus, may become cloudy or precipitate in beverage applications. They also readily participate in numerous chemical reactions with food flavourings. 9.4.2 Needs in controlled release of flavourings The panacea for the industry would be to develop techniques for efficiently and effectively coating small particles with either hydrophilic or lipophilic materials: materials that are water insoluble or only slowly soluble are in great need. One can envision a vapour phase coating process that uniformly applies food grade coatings on micron-sized flavouring particles. Unfortunately, current, inefficient coating techniques require as much coating as base flavouring to apply any effective barrier. We need encapsulated flavourings that protect a flavouring during retorting (for example) but readily release the flavouring when warmed on a stove or are placed in the mouth. We need capsules/matrices that are very protective when initially manufactured but slowly erode or degrade to free the flavouring when desired.

9.5

Advice

The innovations needed in flavour encapsulation/controlled release will probably not come from the food or flavour industry. The technologies we are using today, even the ‘new’ ones, are based on decades-old methodologies that are slowly being refined to serve our needs. We need to cross disciplines to study approaches being used by vastly different fields, and be innovative enough to take these methods and apply them to our industry. This will require a change in mentality, e.g. we should not attend our usual scientific meetings but go to, for example, pharmaceutical, cosmetic, agrochemical, oil drilling, electronic or polymer science meetings.

9.6

References

anandaraman, s. (1984), Encapsulation, analysis and stability of orange peel oil. Ph.D., University of Minnesota. Minneapolis, 278.

Controlled release of flavour in food products

183

anandaraman, s. and reineccius, g. a. (1986), Stability of encapsulated orange peel oil. Food Technol., 40, 88–93. baisier, w. and reineccius, g. a. (1989), Spray drying of food flavours. V. Factors influencing shelf-life of encapsulated orange peel oil. Perfum. Flavourist, 14, 48– 50, 52–53. benczedi, d. and bouquerand, p. e. (2003), Process for the preparation of granules for the controlled release of volatile compounds. US Patent, 6 607 771, Assignee: Firmenich SA (Geneva, Swiss). druri, m. and pawlik, a. (2001), Microencapsulation techniques of food flavours. Przemysl Spozywczy, 55, 31–33. goubet, i., dahout, c., semon, e., guichard, e., quere, j. and voilley, a. (2001), Competitive binding of aroma compounds by beta-cyclodextrin. J. Agric. Food Chem., 49, 5916–5922. graf, e. and soper, j. c. (1996), Flavoured flour containing allium oil capsules and method of making flavoured flour dough product. US Patent 5 536 513, Assignee: Tastemaker, Cincinnati, OH. heiderich, s. and reineccius, g. a. (2001), The influence of fat content, baking method, and flavour form on the loss of volatile esters from cookies, Perfum. Flavourist, 16, 14–21. jones, d. m. (1995), Controlling particle size and release properties. In Flavour Encapsulation (Eds, Risch, S.J. and Reineccius, G.A.) ACS Books, Washington, DC, pp. 158–176. mciver, r. c., vlad, f., golding, r. a., travis, j. and benczedi, d. (2005), Encapsulated flavour and/or fragrance composition. US Patent 6 932 982, Assignee: Firmenich SA (Geneva, Swiss). miller, r. a., hoseney, r. c., graf, e. and soper, j. c. (1997), Garlic effects on dough properties. J. Food Sci., 62, 1198–1201. porzio, m. (2004), Flavour encapsulation: a convergence of science and art. Food Technol., 58, 40–42, 44, 46–47. porzio, m. a. and zasypkin, d. (2004), Encapsulation compositions and process for preparing the same. US Patent, 6 790 453, Assignee: McCormick & Company, Inc (Sparks, MD). reineccius, g. a. (2004), The spray drying of food flavours. Drying Technol., 22, 1289–1324. reineccius, g. a. and buffo, r. a. (2001), Developments in flavour encapsulation, Proceedings – 28th International Symposium on Controlled Release of Bioactive Materials and 4th Consumer & Diversified Products Conference, San Diego, CA, United States, 23–27 June 1, 91–92. reineccius, t. a., reineccius, g. a. and peppard, t. l. (2002), Encapsulation of flavours using cyclodextrins: comparison of flavour retention in alpha, beta and gamma types. J. Food Sci., 67, 3271–3279. reineccius, t. a., reineccius, g. a. and peppard, t. l. (2003), Comparison of flavour release from alpha-, beta- and gamma-cyclodextrins. J. Food Sci., 68, 1234–1239. risch, s. j. and reineccius, g. a. (1988), Effect of emulsion size on flavour retention and shelf-stability of spray dried orange oil. In Flavour Encapsulation (Eds, Risch, S.J. and Reineccius, G.A.) American Chemical Society, Washington, DC, pp. 67–77. schlameus, w. (1995), Centrifugal extrusion encapsulation. In Encapsulation and Controlled Release of Food Ingredients (Eds, Risch, S.J. and Reineccius, G.A.) ACS Books, Washington, DC, pp. 96–103. sébastien, g. (2004), Microencapsulation industrial appraisal of existing technologies and trends. Trends Food Sci. Technol., 15, 330–347. seck, j. k., gu, b. p., chung, b. k., sang, d. p., mun, y. j., jeong, o. k. and yeong, l. h. (2000), Improvement of oxidative stability of conjugated linoleic acid (CLA) by microencapsulation in cyclodextrins. J. Agric. Food Chem., 48, 3922–3929.

184

Modifying flavour in food

soottitantawat, a., yoshii, h., furuta, t., ohkawara, m. and linko, p. (2003), Microencapsulation by spray drying: influence of emulsion size on the retention of volatile compounds, J. Food Sci., 68, 2256–2262. soper, j. c. (1995), Utilization of coacervated flavours. In Encapsulation and Controlled Release of Food Ingredients (Eds, Risch, S.J. and Reineccius, G.A.) ACS Books, Washington, DC, pp. 104–112. soper, j. c. and thomas, m. t. (1998), Preparation of protein-encapsulated oil particles using enzyme-catalyzed crosslinking, EP 856 355 A2, 7. soper, j. c., yang, x. and josephson, d. b. (2000), Encapsulation of flavours and fragrances by aqueous diffusion into microcapsules. Assignee: Givaudan Roure (International) Sa, Switz., pp. 7. sparks, r. a. and mason, n. s. (1987), Method for coating particles or liquid droplets, US Patent, 4 675 140, Washington University Technology Associates (St. Louis, MO). sparks, r. e., jacobs, i. c. and mason, n. s. (1995), Centrifugal suspension-separation for coating food ingredients. In Encapsulation and Controlled Release of Food Ingredients (Eds, Risch, S.J. and Reineccius, G.A.) ACS Books, Washington, DC, pp. 87–95. uhlemann, j., schleifenbaum, b. and bertram, h.-j. (2002), Flavour encapsulation technologies: an overview including recent developments. Perfum. Flavourist, 27, 52, 54–61. vilstrup, p. (Ed.). (2001), Microencapsulation of Food Ingredients, Leatherhead Publ., Leatherhead, UK. wampler, d. j., soper, j. c. and pearl, t. t. (1998), Method of flavouring and mechanically processing foods with polymer encapsulated flavour oils. US Patent 5 759 599, Assignee: Tastemaker (Cincinatti, OH). westing, l. l., reineccius, g. a. and caporaso, f. (1988), Shelf life of orange oil. Effects of encapsulation by spray-drying, extrusion, and molecular inclusion. In Encapsulation and Controlled Release of Food Ingredients (Eds, Risch, S.J. and Reineccius, G.A.) ACS Books, Washington, DC, pp. 110–121. yoon, y., bellas, e., firestone, w., langer, r. and kohane, d. s. (2005), Complex coacervates for thermally sensitive controlled release of flavour compounds, J. Agric. Food Chem., 53, 7518–7525.

10 Developments in sweeteners S. Kemp, Cadbury Schweppes, UK and M. Lindley, Lindley Consulting, UK

10.1

Introduction

Sucrose is the most important sweet-tasting food ingredient with annual world production in excess of 100 million tonnes. However, its importance to the food industry is a consequence of many factors in addition to an ability to deliver a clean sweetness that is devoid of aftertastes or side-tastes. For example, the physical properties delivered to foods by sucrose influence the appearance, colour, texture and shelf-life of a range of important foods such as baked goods, confectionery products and ice cream. In fact, many of these products could not have developed into major industries without the functional properties of this most versatile food ingredient. In spite of (or possibly because of) this central position in the developed food industry of today, for many years there has been intense food and drink product development activity that has focused on removing sucrose from products and using replacement ingredients to deliver the functionalities it provides. The rationale for this effort is based largely on the fact that sucrose is caloric and can contribute to the aetiology of dental caries (Alvarez and Navia, 1989). Successful development of good-tasting products that do not contain sucrose ensures that they can then be marketed on the basis that they may be ‘low calorie’, ‘sugar-free’ or ‘diet’, all of which are descriptors that resonate well with many consumers. Naturally, the successful development of high-quality foods and beverages that do not contain sucrose requires the availability of alternative ingredients that deliver the myriad of functionalities normally provided by sucrose. Therefore, discovery and development of bulk alternatives to sucrose and high-potency sweeteners has been an important activity of suppliers of speciality food ingredients, leading to the successful

186

Modifying flavour in food

commercialisation of many novel sugar-replacement products. Of these, it is undoubtedly the high-potency sweeteners that have attracted most attention. Historically, most high-potency sweeteners were discovered by accident. Some were products of studies designed to develop an understanding of the basic chemistry of particular chemical classes (e.g. cyclamate and sucralose), others of unrelated chemical synthesis programmes (e.g. aspartame, discovered out of a programme seeking anti-ulcer drugs) and still others as part of structure–activity studies into other taste modalities (e.g. neohesperidin dihydrochalcone, in this case into bitter taste). There are a number of reasons why most high-potency sweeteners have been discovered accidentally. Firstly, the relationships between molecular structure and taste are incompletely understood and, where there are clear relationships, they can rarely be used to predict sweet structures from chemical classes not included in the development of those relationships. In addition, the physiological processes responsible for sweet taste perception and the detailed three-dimensional structure of the relevant taste receptors are not known in their entirety. However, recent developments have greatly increased the level of understanding of receptor structure and the mechanisms whereby sweeteners are perceived as being sweet. This increased understanding may be expected to lead to the rational design of new sweet structures, as well as helping to identify compounds that may potentiate or inhibit sweetness. Sweet taste potentiators are expected to be valuable tools for the food industry for a variety of reasons: •

•

•

An effective sweetness potentiator might reduce costs by enabling the same level of sweetness to be achieved using less sweetener (assuming the potentiator added was less expensive than the sweeteners eliminated). Effective potentiation of the sweetness of sucrose by a compound with no intrinsic taste would permit the development of lower calorie products, particularly beverages, having the taste quality normally associated with sucrose. Effective potentiation of the sweetness of sucrose would allow development of reduced calorie products, particularly beverages, while not requiring the use of artificial sweeteners.

Sweet taste inhibitors are already known and commercially available in some markets. Their value to the food industry is that they permit the use of sucrose or other sweet carbohydrates for their valuable functional properties in products that do not need to taste sweet. In addition, control of sweetness levels permits use of sucrose at concentrations where its functional properties can be used to maximum advantage without concerns that the resulting products would be perceived as being too sweet. In summary, therefore, there are solid justifications for seeking to understand the mechanisms whereby compounds elicit sweetness and for attempt-

Developments in sweeteners

187

ing to unravel the relationships between molecular structure and sweet taste; the commercial and nutritional drivers are both strong. Thus, it is appropriate to review recent developments in the sweetness and sweeteners arena and to discuss the potential consequences of success.

10.2

Mechanism of sweetness perception

In humans and other animals, perception of taste occurs in specialised structures called taste buds. Taste buds are groups of elongated taste receptor cells whose purpose is to bind tastants present in the mouth and transduce this physical binding into an electrochemical signal relayed to the brain. Taste science made significant progress in the early 1990s with the discovery of the protein gustducin in taste-responsive cells (McLaughlin et al., 1992). Gustducin was shown to interact with receptors in the taste cell membrane upon binding of certain tastants, its activation resulting in taste cell activation. Gustducin is a member of a family of proteins called G proteins and hence the receptor proteins it interacts with are known as G protein coupled receptors (GPCRs). A number of G proteins have subsequently been identified in the taste system (McLaughlin et al., 1994). Upon tastant binding, a change in conformation of the receptor leads to activation of G protein. G protein activation in turn results in second messenger activation, the switching on of effector enzymes that modulate the concentration of molecules within the cell. Changes in second messenger levels activate various ion channels in the cell membrane and in intracellular membranes, resulting in the depolarisation of the taste cell as ions flow down their respective concentration gradients. An electrical signal is transmitted through the nervous system to the brain where it is interpreted as taste (Macgregor, 2006). Recent advances in the field of genetics have led to significant advances in identification and characterisation of sweet taste receptors. Sweet taste is transduced through GPCR heterodimers consisting of two receptor subunits, T1R2 and T1R3, that form the functional sweet taste receptor (e.g. Bachmanov et al., 2001; Kitagawa et al., 2001; Max et al., 2001; Li et al., 2002). The T1Rs are members of the class C family of GPCRs, which all have a large extracellular domain containing the binding site for taste ligands. Four distinct receptor-binding sites for sweeteners have so far been postulated. Xu et al. (2004) have identified the C-terminal transmembrane domain of T1R3 as being essential for the binding of the sweetener cyclamate and lactisole (an inhibitor of sweet taste). Recently, progress has been made in determining the structure–activity relationships of the human T1R2 and T1R3. A model has been published identifying key binding sites for a variety of sweet-tasting compounds ranging from carbohydrates, amino acids and sweet molecules such as saccharin, to peptides and proteins (Morini et al., 2005). More details have been provided by the research of

188

Modifying flavour in food Aspartame neotame

Cyclamate

Brazzein Lactisole

VFTM

CRD

HD

C- tail T1R2

T1R3

Fig. 10.1 Interactions of various sweeteners and lactisole with specific areas of the sweet taste receptor. VFTM, venus flytrap module; CRD, cysteine-rich domain; HD, heptahelical transmembrane domain; C-tail, intracellular carboxyl-terminal tail. (Source: from Naim, M., Chapter 1 ‘Stimulation of taste cells by sweet taste compounds: receptors, downstream signal transduction components and the implications to sweet taste quality’ in Optimising Sweet Taste (Naim et al., 2006), Figure 1.1.).

Nie and coworkers (2005), working with the expressed and purified N-terminal domain (NTD) of T1R2 and T1R3, who demonstrated that sucrose, glucose and the high-potency sweetener sucralose all bind (or induce liganddependent conformational changes) to the T1R3NTD (see Fig. 10.1). The sweet taste of sugars, especially that of sucrose, is regarded as ‘pure’ in humans, whereas many non-sugar sweeteners possess inferior taste quality. Sugars and sugar alcohols show linear concentration–response relationships while high-potency sweeteners yield hyperbolic dependence. It is apparent that there are multiple reasons that may be related for the dissimilarity in sweet taste between sugars and non-sugar sweeteners. Many (though not all) non-sugar sweeteners possess additional taste sensations such as bitter, metallic and liquorice. Although the molecular basis for such sensations is not clear, the sweeteners saccharin and acesulfame-K were recently found to stimulate both bitter and sweet taste receptors (Kuhn et al., 2004). A further distinguishing factor that separates sugars and non-sugar sweeteners on the basis of taste quality is their temporal properties. Time– intensity relationship studies have indicated that, compared with sucrose, it takes a longer time for the sensation of a non-sugar sweetener to reach a maximal sweet taste intensity, and more time (sometimes minutes) for the

Developments in sweeteners

189

sweetness to be extinguished (lingering aftertaste) (Naim et al., 1986). This lingering phenomenon is known to occur also for a variety of bitter stimuli. To date, the molecular basis for the ‘slow onset’ and ‘lingering aftertaste’ phenomena are not known, even though these have significant implications regarding the acceptance of a variety of food products (Naim et al., 2006). There are also genetic variations in taste perception. Individuals carry their own unique set of taste receptors that gives unique taste perception (Bufe et al., 2005) so that individuals vary in their perception of and preference for sweetness and also for the bitterness that may appear as a side taste in alternative sweeteners (Bartoshuk, 2000). Other factors such as age, environment and culture also play a part in taste preference (Beauchamp, 1999) and this means that consumers with different demographics and/or locations are likely to have different preferences for sweetness. As has been described, over the last decade, progress in our understanding of the mechanism of sweet taste perception has been rapid. With full characterisation of sweet receptors, by first modelling and subsequently crystallisation, the full structure–activity relationship for sweeteners will be determined. This will enable full predictive modelling of the properties of molecules by computer and enable ultra-high-throughput screening. Computer-generated hits will only be synthesised once their sweetener properties have been determined (Walters, 2006).

10.3

Novel sweeteners

Developments in sweeteners have been covered extensively in the recent book ‘Optimising Sweet Taste in Foods’ (Spillane, 2006). The book includes chapters giving overviews of well-known sweeteners such as sucrose (Cooper, 2006), low-calorie potent sweeteners (Kemp, 2006), reducedcalorie sweeteners (von Rymon Lipiski, 2006) and polyols (Embuscado, 2006). Therefore, this chapter will review only those sweeteners that have been described most recently in the literature and that are not yet widely used commercially. Some or all of these novel sweeteners will, in time, become important additional tools in the armoury of sweet ingredients for use by the food and beverage industries.

10.3.1 Novel high-potency sweeteners Introduction In the latter part of the 1990s, it was reported that there was little incentive to identify and manufacture new potent sweeteners as the range of sweeteners already existing was adequate for most purposes and the cost and time involved in developing and approving new potent sweeteners were prohibitive (Grenby, 1996; Lindley, 1999). The priority shifted to identifying

190

Modifying flavour in food

and developing technological advances to improve sweetener functionality, either by developing other ingredients to replace the functionality of sugar, such as bulking agents or taste modifiers, or by processing sweeteners in novel ways, such as agglomeration to improve handling and encapsulation to improve stability of, for example, aspartame. The development of the aspartame–acesulfame salt (Twinsweet®) was a step-wise development that enabled new benefits to be gained from existing potent sweeteners. As has been discussed, one ultimate outcome of research aimed at the molecular characterisation of sweet receptors will be a complete understanding of structure–activity relationships for sweeteners. It is expected that this will permit the design of novel sweeteners, more or less at will. Although receptor-based assays allow high-throughput screening of libraries of compounds for tastants and taste modifiers and will, almost inevitably, identify novel sweet structures, no novel sweeteners identified in this way have yet been described in the literature. The novel sweeteners described recently have been identified by one of the following routes: • discovery and identification of sweeteners of natural origin; • classical structure–activity relationship studies. Novel natural sweeteners Natural high-potency sweeteners represent a ‘holy grail’ for the food industry. Consumers increasingly express a desire for natural products that are free from ‘artificial’ additives. Natural potent and non-nutritive sweeteners provide a route to delivering low-calorie, tooth-friendly products free from the negative aspects associated with artificial additives. Natural sweeteners are not intrinsically novel, of course, although they may only recently have become scientifically and commercially of interest. Examples of natural sweeteners currently of developmental interest include brazzein, lo han guo and monatin. Brazzein is a small molecular weight protein detected in the ripe fruit of Pentadiplanadria brazzeana (van der Wel et al., 1989). Owing to its small molecular size, it has been an interesting molecule for study with the effects of mutations on structure and function being examined (Assadi-Porter et al., 2005) and models of multi-site brazzein– receptor interactions developed. These models have been used in two studies to design smaller molecular weight peptides that contain the putatively critical structural features necessary for receptor interaction, but neither has yielded sweet peptides, thus suggesting a more extensive structure is required for sweetness (Temussi, 2002; Assadi-Porter et al., 2005). Although an efficient bacterial production system has been developed for brazzein (Assadi-Porter et al., 2000), and it is also possible to express brazzein protein in corn seed embryo, thus opening the interesting possibility of producing pre-sweetened cereals with ‘no added sugar’ (Lamphear, 2005). The taste characteristics of the

Developments in sweeteners

191

sweetener are substantially different from those of sucrose and so its commercial utility is projected to be limited. Lo han guo is a Chinese plant that is a member of the Cucurbitacae and is named Siraitia grosvenorii (Swingle) C. Jeffrey. The skin and flesh of the fruit of the plant contain a variety of sweet-tasting triterpene glycosides known as mogrosides. The major sweet principle is known as mogroside V. Lo han guo is considered to be a ‘food’ for regulatory purposes in some markets. Recent development activity on this sweetener has focused on plant selection and tissue culture techniques to produce plant seedlings that contain high levels of mogrosides. Monatin is a derivative of the amino acid tryptophan. It is found in the roots of a South African shrub, Schlerochiton ilicifolius, but since the shrub is not amenable to cultivation, successful commercialisation of monatin will depend on either identification of a synthetic route suitable for scale-up and/or a biochemical route that may preserve this sweetener’s natural status, at least in some markets. Recent studies indicate that monatin can be synthesised according to synthetic schemes that might be commercially viable (Amino and Kawahara, 2003) and that biochemical routes to its synthesis may also be a viable alternative (Hicks and McFarlan, 2005). The sweetener and its derivatives have also been used as model compounds in attempts to model the sweet taste chemoreception mechanism (Bassoli et al., 2005). Work to identify new natural potent sweeteners continues, as they are perceived to be able to pass more quickly through the regulatory process and enjoy a more consumer-friendly image. There are many inaccessible areas of the world, particularly rainforests, containing undiscovered plant species that may yield new compounds with potent sweet taste. Some companies have entered into commercial bio-divining agreements with local bodies that may yield new flavour compounds while providing a source of revenue for ecological preservation programmes. Potential new natural sweet-tasting compounds can be screened rapidly using genomic techniques (Kemp, 2006). Genetic modification may provide a route to production of natural potent sweeteners. Easily grown crop plants may be implanted with the genes necessary to express naturally occurring sweeteners (Fry, 2005). Attempts have been made to modify yeast cells to produce thaumatin (Weickmann et al., 1989) and more stable forms of monellin (Kim et al., 1991). The ideal is to find a natural, stable, safe potent sweetener with the taste properties of sucrose, as this would be more acceptable to consumers and potentially easier to gain regulatory approval (Lindley, 1999; Fry, 2005). Novel sweeteners from structure–activity relationship studies Tinti and Nofre (1991a) developed a model of the sweet taste pharmacophore that incorporates Shallenberger and Acree’s classical AH and B sites

192

Modifying flavour in food

(Shallenberger and Acree, 1967) plus six other sites arranged at specific points in three-dimensional space. The model led to the discovery of two classes of extremely potent sweeteners, the N-carbamoyl dipeptides and the guanidine-acetic acids (Tinti and Nofre, 1991b). A novel series of arylurea sweeteners was designed following the use of conformational energy calculations leading to construction of a van der Waals surface around selected superimposed sweeteners. An electrostatic potential was subsequently mapped onto this surface and the model used to identify promising targets for synthesis. As noted, it was used in the successful design of a new series of arylureas, and it correctly predicted the active stereoisomer of compounds in this series (Madigan et al., 1989; Muller et al., 1991). One difficulty with the use of structure–activity relationships has been that such relationships are naturally developed to encompass known compounds, frequently from within a single, specific chemical class. Thus, it has often been observed that models so developed have limited predictive ability within classes of compounds not used to help develop the model. However, with characterisation of sweet receptors at the atomic level, future research will be concentrated on receptor structure. Since the extracellular binding domains of the T1R2 and T1R3 have been expressed and purified successfully, their structures are likely to be determined by X-ray crystallography in the near future. This should lead to an even more accurate picture of the receptor binding sites and will be expected to lead to identification of novel sweet structures, almost at will.

10.3.2 Novel bulk sweeteners Introduction As discussed in the introduction to this chapter, successful replacement of sucrose in foods requires the availability of alternative ingredients that deliver sweetness with low/no energy and others that provide the functional properties lost when sucrose is removed. Alternative bulk sweeteners that enjoy significant markets include high-fructose corn syrups (similar sweetness at reduced cost), a broad range of polyols (similar functionalities from non-cariogenic ingredients) and a small number of novel mono- and disaccharide sugars that provide some functional properties similar to sucrose, but with different nutritional profiles. Although none of these established or novel bulk sweeteners was identified through application of the mechanistic work discussed earlier, for completeness, a brief description of novel bulk sweeteners is provided. Isomaltulose Isomaltulose (I) is a reducing glucose–fructose disaccharide in which glucose and fructose are linked through their 1 and 6 carbon atoms, respectively (Fig. 10.2). It occurs naturally in honey and sugar cane extract and is pre-

Developments in sweeteners

193

OH O

O O

OH

OH

HO

OH HO

OH

OH

(I)

Fig. 10.2

Isomaltulose.

pared by the enzymatic rearrangement of sucrose using the enzyme sucroseglucosylmutase, itself obtained from the microorganism Protaminobacter rubrum. The sweetness of isomaltulose relative to sucrose is generally considered to be 0.4–0.45 times that of sucrose. The taste delivered, like that from most small molecular weight carbohydrates, is purely sweet with no side-tastes or aftertastes. The glycosidic linkage between glucose and fructose is known to be much more stable to hydrolysis than the linkage in sucrose, thus ensuring that, when formulated into products such as fruit beverages at acidic pH values, isomaltulose will not hydrolyse to invert sugar. However, isomaltulose readily undergoes browning reactions due to the presence of a free reducing group. The sugar is said to be non-hygroscopic. Ingested isomaltulose is metabolised by the sucrase–isomaltase complex in the intestinal mucosa. Equal parts of glucose and fructose are produced and absorbed, contributing 4 kcal/g, as does sucrose. The key difference between the metabolism of sucrose and isomaltulose is that the hydrolysis of isomaltulose proceeds at a much slower rate than that of sucrose. The main impact of the slower hydrolysis rate of isomaltulose is that the resulting glycaemic and insulin responses of healthy subjects and those with type II (non-insulin dependent) diabetes are attenuated. Isomaltulose has also been proposed as being an interesting ingredient for the development of ‘tooth-friendly’ confectionery since it has been shown to be hypoacidogenic in dental plaque. Since the tolerance to isomaltulose has been shown to be excellent – single doses of 50 g induce no gastrointestinal discomfort – ‘tooth-friendly’ confectionery products that induce no laxation effects can be prepared. a,a-Trehalose α,α-Trehalose (trehalose, II) is a naturally occurring disaccharide consisting of two glucose moieties linked through their respective anomeric carbon atoms (1,1) by an α-glycosidic bond (Fig. 10.3). Trehalose can be found in bacteria, yeast, fungi and algae, and some higher plants. It occurs in a number of foods consumed as part of a reglar diet, including mushrooms,

194

Modifying flavour in food OH O HO HO OH O HO OH OH

O OH (II)

Fig. 10.3 α,α-Trehalose.

bread, fermented beverages and honey. Commercially, food grade starch is treated in a multi-step enzyme process involving hydrolysis to glucose followed by enzymatic synthesis of trehalose by maltoligosyl trehalose synthase and maltooligosyl trehalose trehalohydrolase. Both enzymes are obtained from a strain of Pseudomonas amyloderamosa. Trehalose delivers a clean sweetness that is approximately one-half as sweet as sucrose. There are indications in the literature that at iso-sweet concentrations, the sweetness from trehalose lingers a little longer than that from sucrose, but no quantitative data have been published. Some flavourmodifying properties have also been mooted, with trehalose claimed to be able to ameliorate the bitterness associated with the salt substitute potassium chloride and some high-potency sweeteners. The metabolism of ingested trehalose resembles that of maltose (or starch) in that both products are absobed in the form of glucose, and so it delivers 4 kcal/g. Mammals contain intestinal trehalase enzymes and the tolerance to trehalose is therefore high. There have been indications from the market that trehalose has valuable functionalities in sports rehydration beverages and that it also may have a role to play in enhancing the rates of fat oxidation. This latter possibility might be of important interest to the producers of products designed to assist in calorie and weight control. D-Tagatose

d-Tagatose is a keto-hexose monosaccharide differing from fructose only in the orientation of a single hydroxyl group. It is prepared from lactose, which is hydrolysed to galactose and glucose and the galactose moiety is converted directly to tagatose by alkaline isomerisation. This small difference in structure from fructose results in major differences in physical

Developments in sweeteners

195

functionality as well as in its nutritional characteristics (Bertelsen et al., 1999). Tagatose is almost as sweet as sucrose on a weight basis and delivers a clean, sucrose-like taste. At low concentrations it is able to modify the taste of high-potency sweeteners, conferring a flavour closer to that of sucrose, particularly with respect to the perceived mouthfeel or syrupy character. It also ameliorates bitter/metallic aftertastes frequently associated with highpotency sweeteners. Nutritionally, tagatose delivers reduced energy (c. 1.5 kcal/g), is noncariogenic and is an effective prebiotic. Thus, on technical and nutritional grounds, it is an alternative to sucrose that is of substantial interest.

10.4

Sweetness potentiators

10.4.1 Introduction As discussed in the introduction to this chapter, there is a sound rationale for identifying and developing compounds that potentiate sweetness. An effective sweetness potentiator might generate sweetness economically, taste quality improvements might be anticipated and there are perceived marketing benefits consequent on reducing or eliminating the use of ‘artificial’ sweetener ingredients. There is currently significant research effort seeking to identify sweetness potentiators. Belief in their existence is based, at least in part, on the observation that many potent sweeteners contain hydrophilic structural features that are responsible for their sweet taste and lipophilic structural features believed to govern their potency. Thus, identification of a suitable structure that interacts at the binding site ‘for potency’ and tasting it in combination with a sweetener may enhance the sweetness of that sweetener. While this hypothesis is yet to be proven, it provides enough justification to conduct an appropriate search for such compounds. As with the design of new sweeteners, however, sweet taste potentiator structures are expected to be identified more readily following detailed determination of receptor structure and receptor binding sites. Until then, high-throughput screening and standard structure–activity relationship studies must be employed. That said, some sweet taste potentiator compounds have been described.

10.4.2 Alapyridaine Alapyridaine is a Maillard reaction product formed when a mixture of glucose and l-alanine is heated. Chemically, alapyridaine is the inner salt of N-(1-carboxyethyl)-6-(hydroxymethyl)pyridinium-3-ol. Naturally, it is one of many compounds formed when glucose and alanine are heated, but its isolation and confirmation that it functions as a sweetness enhancer were

196

Modifying flavour in food

achieved by applying the comparative taste dilution analysis technique (Ottinger et al., 2003). Ottinger and coworkers demonstrate that alapyridaine reduces the sweetness threshold concentrations of quite diverse structures, including glucose, sucrose, l-alanine and aspartame. They also show that, in common with the structure–activity relationships of sweet compounds, the stereochemistry of alapyridaine is critical: (+)-(S)-alapyridaine is effective as a sweet taste enhancer, (−)-(R)-alapyridaine is not. Interestingly, there is a strong pH-dependency to the effect, with alapyridaine proving to be substantially more effective at pH 7 and pH 9 than at pH 5 or below. This observation suggests it is the de-protonated pyridinium-3-olate that is the physiologically active form of alapyridaine and also has important commercial implications. Since the most likely commercial applications for a sweetness potentiator are soft drinks, the great majority of which are prepared at acidic pH (around pH 3 or pH 4), alapyridaine would therefore not be a functional sweetness potentiator in these applications. Nonetheless, the discovery of alapyridaine has some important implications. Firstly, it appears to confirm that compounds capable of potentiating sweetness do exist. Secondly, although unlikely to be of commercial utility itself, alapyridaine could be an important lead compound in the development of structure–activity relationships in the search for further novel sweetness potentiators.

10.4.3 Substituted benzoic acids There is a wide range of benzoic acid derivatives that display tastemodifying characteristics. Some are known to enhance sweetness whereas others have been described that inhibit sweet taste perception. 2,4-Dihydroxybenzoic acid is a flavour ingredient generally recognised as safe (GRAS) by the Flavor and Extract Manufacturers’ Association (FEMA) in the United States for use in a broad range of food and drink applications. Although slightly sweet in its own right, it also has the capacity to enhance the sweetness delivery of other sweeteners and, at appropriate use-levels, to improve the quality of taste delivered by high-potency sweeteners (Merkel and Lindley, 2006). The taste effects of this hydroxybenzoic acid are similar to those described (Barnett and Yarger, 1989) for other dihydroxybenzoic acids as well as benzoic acid derivatives substituted with hydroxy and amino groups, e.g. 3-hydroxy-4-aminobenzoic acid. Barnett and Yarger (1986) also describe the sweetness-enhancing effects of 3-aminobenzoic acid and, interestingly, demonstrate a pH-dependency of the enhancing effect opposite to that of alapyridaine. 3-Aminobenzoic acid enhances sweetness when present in solution at acid pH, but not at neutral pH. All of these modified benzoic acids deliver some sweetness when tasted alone, thus raising the possibility that they are functioning more as efficient sweetness synergists, rather than as true sweetness potentiators.

Developments in sweeteners OH

197

OCH3

OH O

N H3C

COOH H3C

COOH

(III) N-(1-Carboxyethyl)-6-(hydroxymethyl) (IV) 2-(4-Methoxyphenoxy) propanoic acid pyridinium-3-ol inner salt (Lactisole) (Alapyridaine)

Fig. 10.4

10.5

Structural analogy between the sweetness inhibitor lactisole and the sweetness enhancer alapyridaine.

Sweetness inhibitors

The commercial justification for seeking to identify and develop an ingredient capable of inhibiting sweetness is less obvious than that for a sweetness potentiator. However, there are some products that are perceived as being over-sweet, but whose formulation demands the presence of sucrose or other small molecular weight sweet-tasting carbohydrate for textural and/or stability reasons. In such circumstances, an ability to moderate high sweetness levels can lead to significant product advantages. A number of sweetness inhibitors have been described (Barnett, 1985; Lindley, 1986, 1991), including the commercial inhibitor known as ‘lactisole’. All these inhibitors have been shown to be functional against all sweeteners and, in the case of lactisole, exhibit similar stereo-specificity to other tastants; the S-(−) enantiomer of lactisole is active as a sweetness inhibitor, whereas the R-(+) enantiomer is inactive. In addition, there is an intriguing structural relationship between lactisole (IV), a sweetness inhibitor, and alapyridaine (III), a putative sweetness potentiator (Fig. 10.4). While it is clear that the structural parallels are perhaps tenuous, nonetheless, it would be interesting to examine structural analogues of lactisole, perhaps those with hydroxyl substituents on the phenoxy moiety, and evaluate their sweetness potentiation capacity. Lactisole is, as noted, a commercial product, but its primary value is as an antagonist ligand in mechanistic studies. As described, Xu et al. (2004) have identified the C-terminal transmembrane domain of T1R3 as being essential for the binding of this sweetness inhibitor. However, although the binding site of lactisole has been identified, the mechanism whereby its inhibitory effect is induced remains to be elucidated.

10.6

Future trends

Recent advances in the field of genomics hold great promise for the future. New routes for design and production of sweeteners include design

198

Modifying flavour in food

based on receptor structure, modification of perception based on molecular mechanisms, individually tailored perception based on genomics and production of natural potent sweeteners using genetic modification (Kemp, 2006). With the characterisation of sweet receptors at the atomic level, by first modelling and subsequently crystallisation, the full structure–activity relationship for sweeteners will be determined. This will enable predictive modelling of the properties of molecules by computer and facilitate ultrahigh-throughput screening. These activities should lead to the rational design of sweeteners and sweet taste potentiator compounds and this can only lead to consumer benefits. Of course, the taste profile of a sweetener consists of a number of factors, not just overall sweetness, but also time for onset, duration of effect and off-tastes. Full knowledge of structure–activity relationships will ultimately enable sweeteners to be designed for specific purposes, leading to a situation where the food industry will be able to draw on numerous sweeteners and sweetness modifiers (potentiators and inhibitors) to satisfy the requirements of specific applications. It has long been known that individuals vary in their taste abilities. Now it will be possible to combine precise molecular and perceptual information from individuals, so that all stages of the individual taste perception process from taste receptor genes to taste receptors and their mechanisms to taste response can be characterised, linked and understood (Kim et al., 2004). This approach, known as tastomics, has shown that individuals carry their own unique set of taste receptors that gives unique taste perception (Bufe et al., 2005). Similar investigations in the field of olfaction, when combined with taste research, will provide insights on individual variation in flavour perception (Kemp, 2006). The ultimate goal of all this work is to provide consumer choices that do not require compromises to be made on taste, and hopefully, successful completion of these studies will help the food industry to achieve that goal.

10.7

Sources of further information and advice

The proceedings of an American Chemical Society symposium on sweetener chemistry have been published (Walters et al., 1991) and that volume provides an excellent grounding on these topics. Many excellent and relevant chapters can be found in the book Optimising Sweet Taste in Foods edited by Spillane and published by Woodhead Publishing in 2006. The book series Alternative Sweeteners provides detailed information on individual sweeteners (O’Brien Nabors and Gelardi, 1986, 1991; O’Brien Nabors, 2001).

Developments in sweeteners

10.8

199

References

alvarez, j. o. and navia, j. m. (1989), ‘Nutritional status, tooth eruption, and dental caries: a review’. Am. J. Clin. Nutr., 49, 417–426. amino, y. and kawahara, s. (2003), ‘Process for producing gamma-hydroxyamino acid derivatives and monatins’. US Patent Application 20030228403 A1. assadi-porter, f. m., aceti, d. j., cheng, h. and markley, j. l. (2000), ‘Efficient production of recombinant brazzein, a small, heat-stable, sweet-tasting protein of plant origin’. Arch. Biochem. Biophys., 376, 252–258. assadi-porter, f. m., abildgaard, f., blad, h., cornilescu, c. c. and markley, j. l. (2005), ‘Brazzein, a small, sweet protein: effects of mutations on its structure, dynamics and functional properties’. Chem. Senses, 30 (Supplement 1), i90–i91. bachmanov, a. a., li, x., reed, d. r., ohmen, j. d., li, s., chen, z., tordoff, m. g., de jong, p. j., wu, c., west, d. b., chatterjee, a., ross, d. a. and beauchamp, g. k. (2001), ‘Positional cloning of the mouse saccharin preference (Sac) locus’. Chem. Senses, 26, 925–933. barnett, r. e. (1985), ‘Foodstuffs containing sweetness inhibiting agents’. US Patent 4 544 565. barnett, r. e. and yarger, r. g. (1986), ‘Edible material containing m-aminobenzoic acid or salt’. US Patent 4 613 512. barnett, r. e. and yarger, r. g. (1989), ‘Foodstuffs containing hydrobenzene organic acids as sweetness modifying agents’. US Patent 4 871 570. bartoshuk, l. m. (2000), ‘Comparing sensory experiences across individuals: recent psychophysical advances illuminate genetic variation in taste perception’. Chem. Senses, 25, 447–460. bassoli, a., borgonovo, g., busnelli, g., morini, g. and merlini, l. (2005), ‘Monatin, its stereoisomers and derivatives: modelling the sweet taste chemoreception mechanism’. Eur. J. Org. Chem., 2518–2525. beauchamp, g. k. (1999), ‘Factors affecting sweetness’ in Corti, A., Low Calorie Sweeteners: Present and Future: World Conference on Low-calorie Sweeteners, 25–28 April 1999, Barcelona, Spain. World review of nutrition and dietetics, 85, S. Karger, New York and London, pp 10–17. bertelsen, h., jensen, b. b. and buemann, b. (1999), ‘d-Tagatose – a novel low-calorie bulk sweetener with prebiotic properties’ in Corti, A, Low Calorie Sweeteners: Science and Technology, Karger, Basel, pp 98–109. bufe, b., breslin, p. a. s., kuhn, c., reed, d. r., tharp, c. d., slack, j. p., kim, u.-k., drayna, d. and meyerhof, w. (2005), ‘The molecular basis of individual differences in phenylthiocarbamide and propylthiouracil bitterness perception’, Curr. Biol., 15, 322–327. cooper, j. m. (2006), ‘Sucrose’, in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge, pp 135–152. embuscado, m. e. (2006), ‘Polyols’ in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge, pp 153–174. fry, j. (2005), ‘The future for low-calorie sweeteners’. Food Sci. Technol., 19 (2), 26–30. grenby, t. h. (1996), ‘Preface’ in Grenby, TH, Advances in Sweeteners, Blackie Academic and Professional, Glasgow, pp v–vii. hicks, p. m. and mcfarlan, s. c. (2005), ‘Polypeptides and biosynthetic pathways for the production of monatin and its precursors’. US Patent Application 2005 0282260 A1. kemp, s. e. (2006), ‘Low-calorie sweeteners’, in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge, pp. 175–251. kim, s.-h., kang, c.-h. and cho, j.-m. (1991), ‘Sweet proteins. Biochemical studies and genetic engineering’, in Walters, DE, Orthoefer, FT and Dubois, GE (Eds.)

200

Modifying flavour in food

Sweeteners: Discovery, Molecular Design and Chemoreception. ACS Symposium Series 450, American Chemical Society, Washington, DC, pp 28–40. kim, u.-k., breslin, p. a. s., reed, d. and drayna, d. (2004), ‘Genetics of human taste perception’. Dent. Res., 83, 448–453. kitagawa, m., kusakabe, y., miura, h., ninomiya, y. and hino, a. (2001), ‘Molecular genetic identification of a candidate receptor gene for sweet taste’. Biochem. Biophys. Res. Commun., 283, 236–242. kuhn, c., bufe, b., winnig, m., hofmann, t., frank, o., behrens, m., lewtschenko, t., slack, j. p., ward, c. d. and meyerhof, w. (2004), ‘Bitter taste receptors for saccharin and acesulfame-K’. J. Neurosci., 24, 10260–10265. lamphear, b. j. (2005), ‘Expression of the sweet protein brazzein in maize for production of a new commercial sweetener’. Plant Biotechnol., 3, 103–114. li, x., staszewski, l., xu, h., durick, k., zoller, m. and adler, e. (2002), ‘Human receptors for sweet and umami taste’. Proc Natl Acad Sci USA, 99 (7), 4692–4696. lindley, m. g. (1986), ‘Method for inhibiting sweetness’. US Patent 4 567 053. lindley, m. g. (1991), ‘Phenoxyalkanoic acid sweetness inhibitors’ in Walters, DE, Orthoefer, FT and Dubois, GE (Eds.) Sweeteners: Discovery, Molecular Design and Chemoreception, American Chemical Society, Washington, DC, pp 251–260. lindley, m. g. (1999), ‘New developments in low-calorie sweeteners’, in Corti, A., Low Calorie Sweeteners: Present and Future: World Conference on Low-calorie Sweeteners, 25–28 April, Barcelona, Spain. World review of nutrition and dietetics, 85, S. Karger AG, New York and London, pp 44–51. madigan, d. l., muller, g. w., walters, d. e., culberson, j. c., dubois, g. e., carter, j. s., nagarajan, s., klix, r. c., ager, d. j. and klade, c. a. (1989), ‘Substituted aryl ureas as high potency sweeteners’. Eur. Patent Appl. 89 115587.1, 23 August 1989; Eur. Patent Specification 0 355 819 A1, 1994. Europe. max, m., shanker, y. g., huang, l., rong, m., liu, z., campagne, f., weinstein, h., damak, s. and margolskee, r. f. (2001), ‘Tas1r3, encoding a new candidate taste receptor, is allelic to the sweet responsiveness locus Sac’. Nat. Genet., 28, 58–63. mcgregor, r. (2006), ‘Future directions: using biotechnology to discover new sweeteners, bitter blockers and sweetness potentiators’, in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge, pp 404–414. mclaughlin, s. k., mckinnon, p. j. and margolskee, r. f. (1992), ‘Gustducin is a tastecell-specific G protein closely related to the transducins’. Nature, 357, 563–569. mclaughlin, s. k., mckinnon, p. j., spickofsky, n., danho, w. and margolskee, r. f. (1994), ‘Molecular cloning of G proteins and phosphodiesterases from rat taste cells’. Physiol. Behav., 56 (6), 1157–1164. merkel, c. m. and lindley, m. g. (2006), ‘Method of improving sweetness delivery of sucralose’. US Patent 6 998 144. morini, g., bassoli, a. and temussi, p. a. (2005), ‘From small sweeteners to sweet proteins: Anatomy of the binding sites of the human T1R2_T1R3 receptor’. J. Med. Chem., 48 (17), 5520–5529. muller, g. w., madigan, d. l., culberson, j. c., walters, d. e., carter, j. s., klade, c. a., dubois, g. e. and kellogg, m. s. (1991), ‘High potency sweeteners derived from beta-amino acids’ in Walters, DE, Orthoefer, FT and Dubois, GE (Eds.), Sweeteners: Discovery, Molecular Design and Chemoreception, American Chemical Society, Washington, DC. naim, m., dukan, e., zehavi, u. and yaron, l. (1986), ‘The water-sweet aftertaste of neohesperidin dihydrochalcone and thaumatin as a method of determining their sweet persistence’, Chem. Senses, 11, 361–370 naim, m., huang, l., spielman, a. i., shaul, m. e. and aliluiko, a. (2006), ‘Stimulation of taste cells by sweet taste compounds: receptors, downstream signal transduction components, and the implications to sweet taste quality’, in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge.

Developments in sweeteners

201

nie, y., vigues, s., hobbs, j. r., conn, g. l. and munger, s. d. (2005), ‘Distinct contributions of T1R2 and T1R3 taste receptor subunits to the detection of sweet stimuli’. Curr. Biol., 15, 1948–1952. o’brien nabors, l. (2001), Alternative Sweeteners, third edition, revised and expanded, Marcel Dekker, Inc, New York. o’brien nabors, l. and gelardi, r. c. (1986), Alternative Sweeteners, Marcel Dekker, Inc, New York. o’brien nabors, l. and gelardi, r. c. (1991), Alternative Sweeteners, second edition, revised and expanded. Marcel Dekker, Inc, New York. ottinger, h., soldo, t. and hofmann, t. (2003), ‘Discovery and structure determination of a novel Maillard-derived sweetness enhancer by application of the comparative taste dilution analysis (cTDA)’. J. Agric. Fd Chem., 51, 1035–1041. shallenberger, r. s. and acree, t. e. (1967), ‘Molecular theory of sweet taste’. Nature, 216, 480–482. spillane, w. j. (Ed.) (2006), Optimising Sweet Taste in Foods. Woodhead Publishing Ltd, Cambridge. temussi, p. a. (2002), ‘Why are sweet proteins sweet? Interaction of brazzein, monellin and thaumatin with the T1R2-T1R3 receptor’. FEBS Lett., 526, 1–4. tinti, j.-m. and nofre, c. (1991a), ‘Why does a sweetener taste sweet? A new model’, in Walters, DE, Orthoefer, FT and DuBois, GE (Eds.) Sweeteners: Discovery, Molecular Design, and Chemoreception, American Chemical Society, Washington, DC. tinti, j.-m. and nofre, c. (1991b), ‘Design of sweeteners: a rational approach’, in Walters, DE, Orthoefer, FT and DuBois, GE (Eds.) Sweeteners: Discovery, Molecular Design, and Chemoreception, American Chemical Society, Washington, DC. van der wel, h., larson, g., hladik, a., hellekant, g. and glaser, d. (1989), ‘Isolation and characterisation of pentadin, the sweet principle of Pentadiplandra brazzeana Baillon’. Chem. Senses, 14, 75–79. von rymon lipinski, g.-w. (2006), ‘Developments in reduced-calorie sweeteners and low-caloric alternatives’, in Spillane, W (Ed.), Optimising Sweet Taste in Foods, Woodhead Publishing Ltd, Cambridge, pp 252–280. walters, d. e. (2006), ‘Analysing and predicting properties of sweet tasting compounds’, in Spillane, W, Optimising Sweet Taste in Food, Woodhead Publishing Ltd, Cambridge. pp 283–291. walters, d. e., orthoefer, f. t. and dubois, g. e. (Eds.) (1991), Sweeteners: Discovery, Molecular Design, and Chemoreception. American Chemical Society, Washington, DC. weikmann, j. l., lee, j.-h., blair, l. c., ghosh-dastidar, p. and koduir, r. k. (1989), ‘Exploitation of genetic engineering to produce novel protein sweeteners’, in Grenby, TH, Progress in Sweeteners, Elsevier Applied Science Ltd, London and New York, pp 47–69. xu, h., staszewski, l., tang, h., adler, e., zoller, m. and li, x. (2004), ‘Different functional roles of T1R subunits in the heteromeric taste receptors’. Proc. Natl Acad. Sci. USA, 101, 13972–13073.

11 Enhancing umami taste in foods J. B. Marcus, Kendall College, USA

11.1

Umami: what it is, what it does and how it works

The fifth basic taste, umami, has been used in culinary preparations for over 2000 years, Chinese gastronomes have written about umami for 1200 years, and Japanese researchers identified dietary components that produce the umami taste a century ago. Science is now catching up to what great cooks around the world have intuitively known: foods and flavour enhancers with umami can be useful in creating a savoury, delicious taste, and rounding out and heightening food flavours. This chapter focuses on the practical aspects of the use of umami in recipe, menu and product development, and explains how umami, alone or synergistically, enhances food selection for people of all ages. Components that naturally impart an umami taste are featured, as well as other substances and processes that help create an umami taste, including curing, ageing and fermenting. Examples of how to match the inherent flavour attributes in foods with the appropriate use of ingredients are provided. Methods are described to help develop flavourful and satiating food products, recipes and menus featuring umami in a variety of food settings. Attention is given to the use of flavour-enhanced foods in ageing and weight management, growing global issues. Flavour enhancement is important because highly flavourful foods are satisfying, even in small quantities. As the global population grows exponentially, the quantity and quality of food are important considerations for economical food selection, nutritional status, health and well-being. Good tasting, highly flavourful food with umami can be an important tool in meeting expanding global food and nutrition needs.

Enhancing umami taste in foods

11.2

203

Culinary history of umami in flavour enhancement

Although umami taste receptors have only recently been confirmed (Chaudhari et al., 2000), from a culinary perspective, the umami taste is not new. Fermented fish sauces and intense meat and vegetable extracts have been valued in world cuisines for over 2000 years (Ninomiya, 2002). Consider Roman garum or liquamen, Thai num pla, Vietnamese nuoc mum tom cha, Indonesian terasi, Burmese ngapi, Philippine pagoon and British beef tea and their concentrated flavours. In 1825, the French gastronome, Brillat-Savarin, in The Physiology of Taste, described a meaty taste as ‘toothsome,’ and predicted that the future of gastronomy ‘belongs to chemistry’ (Brillat-Savarin and Fisher, 1978). His description of a meaty taste is similar to the Japanese interpretation of umami as ‘deliciousness.’ It is the chemistry within foods with glutamate that helps create the umami perception.

11.3

Scientific background of umami in flavour enhancement

The culinary roots of umami have paralleled the scientific research of umami in flavour enhancement. Ritthausen identified glutamic acid, the amino acid that elicits a unique umami taste in human sensations in 1866 (Takechi, 2004). Ikeda proposed umami as a separate, distinct taste in 1908. Ikeda was struck by the distinct flavour that dashi, a fish and seaweed broth, added to a bowl of tofu cooked with it. The umami taste was due to the glutamate in the kelp and kombu (seaweed), and dried bonito flakes that flavoured the dashi (Takechi, 2004). The seasoning monosodium glutamate (MSG) was formulated in Japan in 1909, introduced into the United States in 1917, and opened a wealth of flavour enhancement possibilities worldwide. The nucleotide inosine 5′-monophosphate (IMP) was isolated from dried bonito tuna by Kodama in 1914. In 1960, the nucleotide guanosine 5′-monophosphate (GMP) was isolated from shiitake broth by Kuninaka. Kuninaka also discovered a synergy between glutamate and inosinate and guanylate (Takechi, 2004). The l-glutamate taste receptor, taste-mGluR4, was discovered by Chaudhari et al. (2000). This taste receptor regulates the so-called ‘firing’ of taste receptor cells. The researchers likened this discovery to ‘turning a key in a lock that then starts an engine’. Nelson et al. (2002) subsequently identified a broadly-tuned amino acid receptor, T1R1+3, highly stimulated by l-orientated amino acids (of which free glutamate is a member). These receptors provide evidence for the evolutionary importance of detecting amino acids for nutritional survival. These scientific findings corresponded with the development of international organisations and symposia on umami. The Society for Research on Umami Taste originated in 1982. Symposia have since included the

204

Modifying flavour in food

International Symposia on Umami in 1985 and 1990, and the International Symposia on Olfaction and Taste (ISOT) in 1993, 1997, 2000 and 2004 in Kyoto, Japan, among others.

11.4

How the culinary aspects and science of umami interact

Science is now catching up with what great cooks around the world have detected for years: that there is an inexplicable, delicious taste sensation from the interaction of certain foods and ingredients with umami. While umami has ancient roots, culinarians worldwide have since incorporated its flavour-enhancing abilities, and are now experimenting with its versatility (Gugino, 2003). However, culinary differences exist between Eastern and Western approaches, both in tradition and palate. This may be due, in part, to nomenclature. Asian cuisine is based on traditional umami-rich ingredients such as dashi for fullness, depth of flavour, meatiness, inexplicable mouthfeel, complexity, versatility and deliciousness (Fig. 11.1). In comparison, European and

Fig. 11.1

Typical dashi broth with mushrooms illustrates the delicious umami taste.

Enhancing umami taste in foods

205

American cuisines have few traditional umami seasonings and umamirich ingredients. These cuisines tend to favour fat for richness and fullness or sodium for saltiness to help make food taste delicious. If Americans recognise umami at all, they tend to think of it as savoury (Hegenbart, 1992). According to Mark Miller, chef/owner of the Coyote Café, in Sante Fe, New Mexico, USA, ‘Westerners have a lineal palate, set up on sweet and salty tastes with few counterpoints and harmonies. In Asian cuisine, however, you use all tastes at the same time. You’re eating circularly. You must train your mind to go after flavour characteristics and look for flavours in different parts of your mouth’ (Labensky and Hause, 1999). Globalisation has provided opportunities to fuse Eastern and Western cuisines and philosophies. Culinarians around the world are applying scientific research about umami in their culinary applications. For example, US wine specialist Tim Hanni has promoted the pairing of umami-rich food and wine that complement the deliciousness of meals. Chef Jonathan Pratt, of the Umami Café in Croton-on-Hudson in New York, USA, Chef Heston Blumenthal, of the Fat Duck in Berkshire, UK, and Chef Yoshihiro Murata, of Kikunoi in Kyoto, Japan, have integrated umami principles in their contemporary local cuisines.

11.5

Asian condiments that impart umami taste and taste-active components

The basis of these culinary applications are the Asian condiments in dashi broth that impart an umami taste with their taste-active ingredients: dried bonito flakes and bonito, with disodium inosinate (IMP); black mushrooms, with disodium gyanylate (GMP), and kelp (kombu), soy sauce and fish sauce with l-glutamate and disodium adenylate (AMP) (Fig. 11.2).

11.6

Western foods that impart umami taste and taste-active components

Bouillon has been used in Western countries as an inexpensive way of making nutritious soup since 1892 when the Swiss flour manufacturer Julius Maggi commercially produced it. Bouillon (from the French word ‘bouilli’ or boiled) is savoury, meat-like and concentrated. Similar to dashi broth, bouillon imparts a meaty flavour from umami due to glutamate or nucleotides. Glutamate is most abundant in protein hydrolysates produced by enzymatic hydrolysis. Westerners appreciate the flavour enhancement ability of umami when glutamate is present in this ‘free’ form, rather than bound with other amino acids, such as in mushrooms (especially dried), aged cheese, cured ham, sun-dried tomatoes, peas, sardines and anchovies.

206

Modifying flavour in food

Fig. 11.2

Kelp (kombu), a taste-active ingredient and foundation of savoury dashi broth.

Despite the frequent description of umami as meaty, foods such as peas and aged cheese actually have a higher level of free glutamate than an equivalent amount of beef or pork. This is why foods cooked with tomatoes or blended with pungent cheese such as Roquefort or blue cheese appear to have a rounder, fuller flavour than when consumed alone. Mackerel, sea bream, tuna and aged beef are high in taste-active nucleotides, as are shiitake, matsutake and enokitake mushrooms – common in Asian cuisine.

11.7

Other umami taste-activators

Other umami taste-activators are hydrolysed proteins and MSG, the natural occurring sodium salt of glutamic acid, extracted from seaweed or fermented from molasses or sugar beets. Monosodium glutamate comprises

Enhancing umami taste in foods

207

78% glutamic acid and 12% sodium. It has been speculated that the sodium in MSG may activate the glutamate to produce the umami effect (Hegenbart, 1992).

11.8

Taste-active components and umami synergy

When food containing taste-active components such as l-glutamate, nucleotides, such as disodium inosinate, disodium gyanylate and/or disodium adenylate, and MSG are combined in a recipe or menu, there is a synergistic effect, and the umami character is magnified. Exactly how this synergy is achieved has been debated, but depending on the system, up to eight times the enhancing effect can be observed by using, for example, a 50 : 50 blend of MSG and IMP (Hegenbart, 1992).

11.9

Umami formed in ripening, drying, curing, ageing and fermenting

Concentrated flavours, from the processes of ripening, roasting, drying, curing, ageing and fermenting, liberate glutamic acid and increase umami levels, as detected in cured pork products (Fig. 11.3). These are popular in the United States as well as in many global cuisines, such as in the local or regional chorizo in Spain, pepperoni in Italy, kielbasa in Poland and frankfurters in Germany. A ripe tomato has ten times more glutamate than an unripe tomato. Dried shiitake mushrooms contain 1060 mg/100 g of glutamate versus 71 mg/100 g in fresh shiitake mushrooms. During the salt-cure of fermented protein products, such as anchovies, the protein breaks down into a wide variety of free amino acids and nucleotides. Aged beef has more glutamate than fresh beef. Fermentation is used, in part, to imbue soy sauce, Asian fish sauce, hot sauce and Worcestershire sauce with more umami. The fermentation in beer and wine has a similar effect. Milk from cows, sheep and goats provides a rich source of glutamic acid. When milk is cultured with enzymes, bacilli or moulds, the glutamic acid is liberated, as is the flavour. The harder and more aged the cheese, such as Parmesan or Gruyère, the higher the glutamate content, and the better it is for flavouring a dish. While yogurt and sour cream have less glutamate than aged cheese, they are both ancient and traditional sources of umami, as seen in Mediterranean and Middle Eastern cuisine. Food processing methods that break down food into smaller units of flavour help make tastes easier to detect. Simply stated, cooking increases umami by deconstructing it. A long, slow stew usually has a higher level of umami than a quick sauté. Oven-roasting or oven-drying unripe tomatoes

208

Modifying flavour in food

Fig. 11.3

Roasted peppers add concentrated flavour to this basic white bean dish.

concentrates their flavour and boosts umami. If a pinch of sugar is added to tomato sauce, this mimics the ripening process and also boosts the umami (Gugino, 2003).

11.10

Practical aspects of umami in consumer acceptance

While consumers may want great-tasting food, they may not understand why. This is where umami may be effective. By utilising the flavour principles of umami, consumers can create homemade, well-rounded flavours with savoury taste. These full-bodied flavours that convey hours of cooking time may help today’s consumer who has little time or interest to cook, but still seeks great taste.

Enhancing umami taste in foods

209

Umami may also be effective in food selection. Glutamate appears to drive the appetite for protein-rich foods, just as the sweet taste probably drives the appetite for carbohydrates, and the salty taste for minerals. This may be similar to what people long ago experienced, but this evolutionary function is difficult to prove (Yuan, 2003). Some people are very fond of glutamate-rich foods and will select them before others. It may be that one’s liking for high-glutamate food is directly related to individual perception and appreciation of glutamate itself (Yuan, 2003). Umami is useful in sodium reduction. This is because umami highlights the sweetness and saltiness in a salty food, such as soup, without actually increasing the salt. The sweetness partially counterbalances the saltiness. Umami also lessens the bitterness in foods that contain some sweetness. This function provides one of the more important consumer applications of umami: its effectiveness in sodium-reduced diets. Umami can contribute up to 50% salt reduction in foods while retaining desirability. Soup without glutamate does not become palatable until the salt concentration reaches 0.75%. With glutamate, however, the same soup is palatable with a salt concentration of just 0.4% salt concentration. In an average serving of soup (200 g or 1 cup), this represents a 0.7 g reduction of sodium. This sodium reduction attribute of umami may have great implications to both consumers and food scientists because hypertension worldwide is a growing global concern (Devika, 2005).

11.11

Consumer applications of umami

Desirability is key in food acceptance and enjoyment. The umami taste is hard to recognise, but easy to enjoy. In the United States, the popular Caesar salad and salad dressing are good examples. The individual ingredients are not as tasty as their sum. In a Caesar salad, umami in aged Parmesan cheese helps make relatively tasteless romaine lettuce tastier. Anchovy, used in very small quantities, provides functional amounts of nucleotides. The resultant taste is amplified and synergistic. Umami may also be the reason why Americans love pizza and pasta with Parmesan cheese, why they find French fries and hamburgers so much more savoury with ketchup, why tomatoes are present in most ordinary lettuce salads, and why glutamate helps consumers select foods that are ‘just right,’ because it is a measure of produce readiness. Sweet peas and corn have two of the highest levels of free glutamate when very fresh and ripe, as opposed to vegetables that have lingered long in the fields, allowed to sit for extended time after picking, or undergone cooking. This importance to consumers cannot be understated: the enhanced smell and taste of umami-rich ingredients stimulates the palate, implies strong flavours will follow, and increases the desire for tasty, and often nutritious, food.

210

11.12

Modifying flavour in food

Food technology applications of umami

Umami is useful for food technologists because it helps improve flavour in product development. It is beneficial in processed foods when home preparation to develop rich, meaty flavours is not feasible. Foods cooked in steam kettles may not achieve the full, meaty and roasted flavours that are available to the home cook. Flavour enhancement can provide the missing link in institutional cooking to help recreate a savoury taste. It is practical since its amino acid content helps fill out flavour profiles, such as when a little soy sauce is added to a formula. This applies to chicken or beef entrées, soups and/or cured meats. Umami is also economical since its natural flavour may allow for a reduction of more expensive umami-rich ingredients. For example, by using flavour enhancers with umami, one might be able to reduce the amount of expensive dried mushrooms that are called for in a soup or entrée (Frank, 1999). MSG, glutamate and nucleotides help provide food technologists with the ‘missing links’ in recipes that pull flavour profiles together. They work in synergy to round out food flavours when a food is missing something that is not clearly identifiable. They can also help make a finished product taste ‘more than a sum of the ingredients’ (Frank, 1999). An unbalanced product tends to create taste fatigue. It may taste good initially, but lose its appeal after a few bites. This can give the eater the impression that the food does not taste good, unless there is a delicious, lingering aftereffect, which umami activators can provide. The imbalances may be so great that reformulation is required. This situation may occur with low-sodium and/or low-fat food. An umami-rich taste-activator such as MSG can impart a fuller flavour with less total sodium and fewer calories – important concerns in the prevention of hypertension and obesity.

11.13

Umami applications that maximise flavour, food acceptance and food preference

The inherent flavour attributes of foods can be matched with umami-rich ingredients for maximum flavour, acceptance and preference. Some of the ingredients that impart/affect the umami taste are soy sauce, Parmesan and other pungent cheese, seaweed, anchovies, smoked fish, tomatoes and tomato sauce, bacon and other smoked meats, meat and fowl stock, mushrooms and wine.

11.13.1 Soy sauce Cooks have long appreciated the umami taste of soy sauce with its 300 flavour compounds, including glutamic acid and volatile aromatic substances

Enhancing umami taste in foods

211

that impart a meaty taste and light colour to meat, soups and sauces. Soy sauce has an affinity to sweet food such as fish, and works well to balance an acidic taste, such as when it is served at the same meal with sake. Soy sauce also enhances the sweetness in bitter foods, such as stir-fried broccoli or Chinese greens. When soy sauce is added to tomato sauce, the umami increases.

11.13.2 Parmesan cheese Parmesan cheese weighing 100 g contains approximately 12 g of natural free glutamate. Its salty, umami nature adds appeal to bland, sweet foods such as pasta and rice, and depth to sour dishes, such as bolognese sauce. A sprinkle of pungent cheese will lift the blandness of ordinary macaroni and cheese. Parmesan cheese reduces bitterness by accentuating any sweetness in bitter foods, such as when it is mixed with basil in pesto sauce.

11.13.3 Salt plus umami While the basic taste of salt does not impart an umami taste, umami appears to work in tandem with salt. For instance, umami and bitterness increase a salty perception (Yamaguchi and Takahashi, 1984). A touch of preserved food, such as artichoke hearts, capers, olives or gherkins, adds an instant salty note. For a more sophisticated salty/umami taste, ingredients such as the ones that follow can be added to a recipe, but they may also add bitter notes: anchovies (fresh, dried), bacon, bresaola, capers, caper berries, caviar, chorizo, clams or clam juice, dashi, dulse, fermented black beans, fish sauce (nam pla or nuoc nam), glasswort (also samphire), ham, lox, monosodium glutamate, niboshi (Japanese dried sardines), nori, olives, oysters or oyster sauce, prosciutto, roe (lobster, shad or salmon), salt cod (baccala), salted nuts or seeds (pumpkin or sunflower), salt pork, sambal (Southeast Asian chili paste), sardines, sea urchins (also uni), seaweed, shrimp paste, soy sauce, tamari and/or Worcestershire sauce.

11.13.4 Seaweed Kelp, wakame and nori are rich in iodine and glutamic acid, and add a complex umami taste to recipes. The Japanese today still use kelp (kombu) in savoury dashi stock with dried bonito flakes (Fig. 11.4) (Takechi, 2004). Seaweed’s salty taste enhances sweet ingredients such as prawns and tofu, and bittersweet foods such as shiitake mushrooms (with nucleotides) or aubergine eggplant. Seaweed tastes sweeter if combined with other salty foods, such as soy sauce.

212

Modifying flavour in food

Fig. 11.4

Dried bonito flakes with iodine and glutamic acid impart distinct taste to dashi broth.

11.13.5 Anchovies/smoked fish The salty-fishiness of chopped or mashed anchovies into olive oil or butter brings out the sweetness of pasta and bittersweet foods, such as cauliflower, courgette (zucchini) or fennel by offsetting the bitterness. When anchovy butter is topped on glutamate-rich steak, poultry or fish, the umami is boosted. Minced anchovies or anchovy paste added to tomato sauce will also boost umami. The sauce will benefit from both the salt-cured and oily characteristics of the anchovies. Umami-rich smoked fish, such as smoked haddock or salmon, with a delicate, salty taste, imbues a recipe with flavour. These fish are often paired with relatively sweet, bland foods, such as rice, potatoes, butter, blinis or bread to enhance their flavour.

Enhancing umami taste in foods

213

Fig. 11.5 Classic tomato sauce with Parmesan cheese adds savoury ingredients to plain pasta.

11.13.6 Tomatoes Tomatoes contain 0.0007 oz of free glutamate/4 oz of tomato (0.02 g/110 g), as well as sweet, sour, salty and bitter tastes. Cooked or raw tomatoes will enhance the flavour of most savoury foods, and bring out sweet, sour and salty tastes while they counterbalance any bitterness. For this reason, some chefs even use tomatoes in desserts. A classic tomato sauce is an easy way to add a savoury, salty element to a variety of dishes (Fig. 11.5). A small piece of ham adds a salty/umami depth. If a tasteless alcohol, such as vodka, is added to the tomato sauce, the flavour is increased even more. Vodka acts as a solvent and releases the umami in the tomatoes. Because it has more alcohol than wine, vodka tends to be a more effective vehicle than wine for this purpose.

214

Modifying flavour in food

11.13.7 Bacon/smoked meat Bacon gains its salty, umami taste from curing with dry salt (dry cure) or brine (wet cure). Bacon’s salty/umami nature enhances sweetness in eggs or peas (also with umami). It intensifies the perception of sourness in a wine sauce, and counteracts bitterness; for example, when bacon is cooked with game or cabbage.

11.13.8 Meat or fowl stock A homemade meat or fowl stock has natural umami and saltiness from its protein that adds a luscious depth. The bones and vegetables should be roasted before turning into stock to intensify the meaty taste.

11.13.9 Mushrooms Dried shiitake, matsutake and enokitake mushrooms, and fresh shiitake mushrooms are rich sources of nucleotides (Fig. 11.6). The umami in mushrooms will intensify if roasted. Some mushrooms, such as portabella, have a pronounced meaty taste. That may be why portabella mushrooms are popular in the United States as hamburger substitutes, enhanced by tomato products.

11.13.10 Wine plus umami According to Tim Hanni, US wine specialist, ‘If you understand what a particular taste will do to the taste of wine you want to drink, you can adjust the way food is seasoned or sauced in preparation so that a particular dish matches the wine. The wine doesn’t change, but the perception about the wine has’ (Hanni, 1996). Sweet foods may increase the perception of acidity in wine. Sweetness comes from fruit and fruit juice, but also from hoisin, teriyaki, cocktail and tomato sauces with umami. Acidic foods from vinegar, citrus or dry wine reductions may also affect the sour taste in wine. Bitter foods, such as endive, broccoli and foods charred in cooking increase the perception of bitterness in wine. Umami also increases the perception of bitterness and astringency in wine. Astringency is a term used to describe the tactile sensation of wine. The judicious addition of salt and acid to food, especially sauces and other foods high in umami, can be useful to tone down the bitterness and astringency of some wines, and help the compatibility of some food and wine. If a dish is balanced, the perception of the wine is that it ‘matches’, such as long-cooking beef bourguignon and aged Burgundy wine. If a recipe is high in umami, a young wine may taste bitter due to increased tannins, such as matching umami-rich shrimp cocktail with a young Zinfandel or an aged umami-filled steak with Chianti. But if salt and a squirt of lemon are added

Enhancing umami taste in foods

Fig. 11.6

215

Shiitake mushrooms provide a rich source of nucleotides and umami taste to recipes and product development.

to the shrimp cocktail or steak, the tannins decrease and the perception of the wine paired with these dishes improves.

11.14

Umami applications for the development and enhancement of recipes and products

According to Michael Roberts, former chef of Trumps, New York, ‘Primary flavours are those that are obvious. Secondary flavours are secret ingredients that make a dish more quintessential and complex’ (Roberts, 1988). Consider umami as a ‘secret ingredient’ that partners, layers, balances and acts as a flavour catalyst to help make a dish or a product more interesting and desirable. It can partner with other flavours to create a new flavour, balance flavours or act as a flavour catalyst.

216

Modifying flavour in food

11.14.1 Umami as a flavour partner As a flavour partner, umami helps to create something new or superior to an original dish or product. For example, a standard seafood bisque is a combination of seafood, potatoes and cream. An umami-enhanced seafood bisque may partner seafood, with its natural umami, fortified wine and umami-rich mushrooms to maximise flavour. The synergy is accomplished with natural glutamate, nucleotides and wine (with umami).

11.14.2 Umami as a flavour layerer As a flavour layerer, umami helps different flavours peak at different times, then combine. For example, ordinary cocktail sauce is a combination of tomato products and heat, such as ketchup and chilli sauce. An umamienhanced cocktail sauce, with natural umami from tomatoes, may first reveal a tomato taste; then it displays sharpness from wasabi, boosted by soy sauce with more umami for a hot/savoury unified finish.

11.14.3 Umami as a flavour balancer As a flavour balancer, umami highlights flavours by contrasting, cancelling or balancing. Chinese five-spice powder, curry powder, garam masala, garlic and ginger are pungent. If any of these ingredients are combined with an Asian sauce or soy sauce with umami and a touch of sugar, they mellow, and their hot/savoury taste, rather than pungent taste, is magnified, losing their unique characteristics for the greater good of the dish.

11.14.4 Umami as a flavour catalyst As a flavour catalyst, umami provides the backbone flavour in a recipe and keeps primary flavours from disappearing. For example, a standard grilled steak already tastes delicious from the umami in the beef. To uplift the umami even more, select aged beef, enhance it with truffle butter (with umami), a squirt of lemon and a touch of salt. The rich umami flavour of the beef is boosted by this combination of ingredients; likewise, umami increases a salty perception. Italian chefs intuitively achieve this blending of flavours in the recipe ‘Bistecca alla Fiorentina’.

11.15

US umami initiatives

At the Umami Café in Croton-on-Hudson, New York, Chef Jonathan Pratt has used the common theme of umami in every dish, such as: ‘Truffled Mac and Cheese’ in a cream sauce of Gruyère and Fontina cheese with black truffle butter and white truffle oil; ‘Pulled Pork and Foie Gras Panini’ with smoked port, duck foie gras and quince paste; and ‘Grilled Marinated

Enhancing umami taste in foods

217

Shrimp’ with mushrooms and sweet soy glaze, served with coconut rice, umami-rich vegetables and sweet soy-sriacha glaze.

11.16

European umami initiatives

Similarly, at the Fat Duck in Bray, Berkshire, United Kingdom, Chef Heston Blumenthal has combined seafood, rich in nucleotides, with Parmesan cheese, rich in glutamate such as: ‘Marinated Squid with Parmesan Cheese’ with just the right amount of Parmesan cheese to balance the umami in the squid; ‘Poached Sea Bream with Kombu Broth’ for a dense, meaty taste; and ‘Green Bean and Tomato Salad with Soy and Fish Sauce Dressing’ that transforms ordinary Western ingredients into a umami-rich melody.

11.17

Future trends

While it is challenging to predict new applications of umami, trends indicate more complex tasting, individualised food are on the rise. In the United States, people are experimenting with bolder, more aggressive flavours and seeking new taste experiences. ‘People are looking for flavours that are louder, brighter, deeper and more complex’, Chef Chris Schlesinger of East Coast Grill in New York City reports (‘Experience Flavour’ brochure; see Section 11.19.6). There is a sense that, ‘Personalized and healthy food that connects specific metabolic needs with individual taste preferences is certainly a goal for innovations to come’ (le Coutre, 2003).

11.17.1 Formulation initiatives The exact mechanism for how MSG, nucleotides and glutamate work synergistically is not known, but nutritional scientists suspect that when these ingredients are mixed with yeast-based enhancers, the synergy increases even more (Yuan, 2003). The flavour profile is enhanced and lengthened, as when lower sodium-containing IMP and GMP are added to vegetables. This may have implications for creating foods with nutritional enhancement for the ageing whose sharpness of taste has decreased owing to longevity or medications. Other formulation initiatives include the use of soy sauce and yeast extracts used to decreased sodium without losing the perception of salt in meat products, soups, sauces, gravies, rubs and spice mixes. Wine powder and tomato products can be used in combination to increase the perception of sodium in sauces. Foods high in nutritional value, such as bitter green leafy vegetables, can be paired with umami-rich ingredients, and modified to taste better to consumers who find them distasteful. Alapyridaine, a tasteless compound isolated from beef stock that relies on GMP synergism, is

218

Modifying flavour in food

used to augment salt, sweet, and umami flavours in food, and to help make bitter chocolate taste sweeter. 11.17.2 Weight management and nutritional enhancement In 2004, the Institute of Food Technologists (IFT) presented the top ten functional food trends. According to their findings, Americans were focusing on weight management and eating better, and health-conscious choices rivalled convenience as the major factor determining what foods are being purchased. Among other trends were eating better by avoiding or reducing the intake of some types of foods and/or seeking foods with nutritional enhancement (Sloan, 2004). Flavour-enhanced foods with umami may help decrease consumption of foods with excess sodium and calories and increase consumption of highly nutritious foods.

11.18

Conclusion

A Chinese proverb of unknown origin states, ‘The more you eat, the less flavour; the less you eat, the more flavour’. A little great tasting, wellprepared or well-formulated flavour-enhanced food may be very satisfying. This may be an important factor from both economic and nutritional standpoints in meeting global food and nutritional needs.

11.19

Sources of further information and advice

Books ackerman, d. (1990), A Natural History of the Senses. New York: Vintage Books. dispirito, r. (2003), Flavour. New York: Hyperion. doty, r. l. (1995), Handbook of Olfaction and Gustation. New York: Marcel Dekker. filer, l. j., jr, garattini, s., kare, m., reynolds, w. a. and wurtman, r. j. (1979), Glutamic Acid: Advances in Biochemistry and Physiology. New York: Raven Press. getshell, t. v., doty, r. l, bartoshuk, l. m. and snow, j. b. (1991), Smell and Taste in Health and Disease. New York: Raven Press. kapoor, s. (2003), Taste A New Way to Cook. North Vancouver, BC: Whitecap Books. kawamura, y. and kare, m. r. (1987), Umami: A Basic Taste. New York: Marcel Dekker, Inc. kunz, p. and kaminsky, p. (2001), The Elements of Taste. New York: Little Brown and Co. roberts, m. (1988), Secret Ingredients: The Magical Process of Combining Flavours. New York: Bantam Books.

Journals Food Reviews International. Special Issue on Umami, 4:2, 3, 1998. The Journal of Nutrition. International Symposium on Glutamate, 130:4S, Oct, 1998.

Enhancing umami taste in foods

219

Book chapters mattes, r. d. (1999), ‘Nutrition and the Chemical Senses.’ In: Shills, Olson, Shike, Ross. Modern Nutrition in Health and Disease. New York: Williams & Wilkins. steiner, j. (1987), In: Kuwarura Y, Kare M. What the Neonate Can Tell Us About Umami: A Basic Taste. New York: Marcel Dekker, Inc.

Articles chaudhari, n., landin, a. m. and roper, s. d. (2000), A metabotropic glutamate receptor variant functions as a taste receptor. Nature Neuroscience, 3, 2. cox, m., menagh, m., kunes, e., kaplan, j. and goldberg, j. (1989), A question of taste. OMNI 43–46; 78, 80, 81. drewnowski, a. and gomez-carmeros, c. (2000), ‘Bitter taste, phytonutrients and the consumer: a review.’ American Journal of Clinical Nutrition, 72, 1424–1435. frank, p. (1999), Design elements – flavour tricks. Food Product Design. Available at: www.foodproductdesign.com/archive/1999/1299de.html gugino, s. (2003), Umami, the fifth taste. Wine Spectator. Available at: www.wbmeersburg.de/infos/umami.htm hegenbart, s. (1992), Flavour enhancement: making the most of what’s there. Food Product Design. Available at: www.foodproductdesign.com/archive/1993/0293CS. html hess, m. a. (1992), Taste: the neglected nutritional factor. Topics in Clinical Nutrition, 7 (4), 1–6. international food information council (1999), Taste matters. Food Insight, July/August. kafka, b. (2000), Taste matters. The Culinarian, February. li, x. (2002), Human receptors for sweet and umami taste. PNAS, 99 (7), 4692–4696. maeder, t. (2001), The fifth avenue of taste. The Red Herring, February. nelson, g. (2002), An amino acid taste receptor. Nature, 416 (6877), 199–202. pelchat, m. l. (1991), Aging, olfaction and food preferences. Chemical Senses, 16, 567. schiffman, s. s. and warwick, z. s. (1993), Effect of flavour enhancement of foods for the elderly on nutritional status: food intake, biochemical indices and anthropometric measures. Physiology and Behavior, 53, 395–402. smith, d. l. and margolskee, r. f. (2001), Making sense of taste. Scientific American, March. steiman, h. (1997), The fifth element. Wine Spectator, 31 July. steingarten, j. (1999), Take a powder. Vogue, March. willoughby, j. (1998), A chemical mystery that excites the taste buds. The New York Times, 14 January. yuan, k. (2003), Can’t get enough of umami: revealing the fifth element of taste. The Journal of Young Investigators, Inc. Available at: www.jvi.org/volumes/volume9 zhao, g. q., zhang, y., hoon, m. a., chandrashekar, j., erlenback, i., ryba, n. j. p. and zuker, c. s. (2003), The receptors for mammalian sweet and umami taste. Cell, 115, 255–266.

Interviews hurd, j. Director of Nutrition and Food Service, Baptist Medical Center, Arkadelphia, AR, USA. Interview, Jan. 2000. dollarhide, p. Director of Nutrition Services, St. Joseph Medical Center, Wichita, KS, USA. Interview, Jan. 2000. saile, d. Einstein Health Systems, Philadelphia, PA, USA. Interview, Jan. 2000.

220

Modifying flavour in food

Other Better Eating for Better Ageing. International Food Information Council, Washington, D.C. Cooking Healthy with Great Taste. Ajinomoto, USA, Inc., Paramus, NJ. Experience Flavour. Lawry’s Foods, Inc. Monrovia, CA. Flavour Enhancement – A Practical Guide. Ajinomoto, USA, Inc., Paramus, NJ. Flavour Essentials. McCormick & Company, Inc. Hunt Valley, MD. Flavour Trends 2000. McCormick & Company, Inc. Hunt Valley, MD. Forcasting Flavour 2004. McCormick & Company, Inc. Hunt Valley, MD. Sweet, Sour, Salty, Bitter and Umami. Umami Information Center, Tokyo, Japan.

11.20

References

brillat-savarin, j. a. and fisher, m. f. k. (trans) (1978), The Physiology of Taste. New York: Harcourt Brace Jovanovich. chaudhari, n., landin, a. m. and roper, s. d. (2000), A metabotropic glutamate receptor variant functions as a taste receptor. Nature Neuroscience, 3 (2), 113–119. devika, s. (2005), Global burden of hypertension: analysis of worldwide data. Lancet, 365, 217–223. frank, p. (1999), Design elements – flavour tricks. Food Product Design [Online]. Available at: www.foodproductdesign.com/archive/1999/1299de.html gugino, s. (2003), Umami, the fifth taste. Wine Spectator, 31 May, 32–34 [Online]. Available at: www.wb-meersburg.de/infos/umami.htm hanni, t. (1996), The cause and effect of food and wine. Wines of the Pacific Rim Fair Seminar, Hong Kong. hegenbart, s., ed. (1992), Flavour enhancement: making the most of what’s there. Food Product Design [Online]. Available at: www.foodproductdesign.com/archive/ 1993/0293CS.html labensky, s. r. l. and hause, a. m. (1999), On Cooking: A Textbook of Culinary Fundamentals, 2nd edn. Englwood Cliffs, NJ: Prentice-Hall. le coutre, j. (2003), Taste: the metabolic sense. Food Technology, 57 (4), 34–37. nelson, g., chandrashekar, j., hoon, m. a., feng, l., zhao, g., ryba, n. j. and zuker, c. s. (2002), An amino acid taste receptor. Nature, 416 (6877), 199–202. ninomiya, k. (2002), Umami: a universal taste. Food Reviews International, 18 (1), 23–38. roberts, m. (1988), Secret Ingredients: The Magical Process of Combining Flavours. New York: Bantam Books. sloan, a. (2004), The top 10 functional food trends of 2004. Food Techonology, 58 (4), 28–51. takechi, y., ed. (2004), The Fifth Taste of Human Being: Umami the World. London: Cross Media Limited. yamaguchi, s. and takahashi, c. (1984), Interactions of monosodium glutamate and sodium chloride on saltiness and palatability of a clear soup. Journal of Food Science, 49, 82–85. yuan, k. (2003), Can’t get enough of umami: revealing the fifth element of taste. Journal of Young Investigators, 9 (2) [Online]. Available at: www.jyi.org/volumes/ volume9

12 Bitter blockers in foods and pharmaceuticals R. McGregor, Linguagen Corporation, USA

12.1

Introduction

The ability to sense potentially harmful molecules entering the digestive system has evolved as a warning system in animals known as bitter taste perception. However, many ingredients in foods, and most pharmaceuticals, are sensed as bitter, which can lead to decreased acceptability. The chapter begins with an examination of why reducing bitterness is important for food and beverage applications, and the current approaches to reducing bitterness. The chapter then turns to recent advances in the understanding of the mechanisms of taste, and how this knowledge is enabling the application of pharmaceutical research methods in the discovery of novel taste-modifying compounds, such as bitter blockers. The chapter concludes with an appraisal of future trends in the development of bitter blockers.

12.2

Why reduce bitterness in foods and pharmaceuticals?

There is significant pressure on producers of foods and beverages to improve the nutritional value of their products. These pressures come in the form of government guidelines and regulations, along with increased consumer awareness of the importance of a healthy diet. The industry is also seeing considerable growth potential in the marketing of functional foods and beverages. However, consumer acceptance of the taste of a product has, and always will be, of paramount importance in the success of foods and beverages. It is an accepted fact that a purportedly healthy product will fail in the marketplace if it does not taste good. Many of the compounds in foods

222

Modifying flavour in food

that provide health benefits, such as polyphenols in soy and chocolate, and phytonutrients in nutritional products and functional foods, are bitter. The same is true for functional ingredients such as hydrolysed proteins used as stabilisers and texturants in nutritional products. One area in which pressure on producers has been mounting is in the marketing of products with reduced sodium content. It is now widely accepted that, for a significant proportion of the population, those known as the salt-sensitive subpopulation, increasing intake of salt (sodium chloride) leads to increased blood pressure (Jones, 2004). While it is a common perception with the public that a significant quantity of salt in the diet is inherently present, added during cooking or shaken on at the table, in fact, 75% of the salt in the typical Western diet is added during processing. This has led to increasing government pressure, both in Europe and North America, to reduce the levels of sodium in processed foods in order to decrease the growing burden hypertension, and its related diseases, are placing on national healthcare systems. This problem is increasing as the populations of these countries continue to age into the at-risk population. In addition to making foods more salty, the sodium in foods fulfils a number of other roles, including as a texturant, preservative, antibacterial, for colour, curing, and water retention, and as an antioxidant. It has therefore proved difficult to find ways of reducing sodium without negatively impacting a food’s overall product profile. This is evident when comparing the number of low-sodium products on supermarket shelves with those of low-sugar and low-fat products, where acceptable alternatives to nutritive sweeteners and fats have been developed. With the characterisation of the taste system at the molecular level, there are a number of approaches available that involve taste modification to reduce the levels of sodium in processed foods. The three most likely approaches, in descending order of technical difficulty, are: replacement of sodium with a novel salty tasting molecule, potentiation of sodium taste, and substitution of potassium salts in place of sodium salts. Potentiating salty taste, or replacing sodium chloride with a salty tasting alternative that does not negatively impact blood pressure, can go some way to reducing sodium levels in foods; however, the real benefit will probably come with the development of products in which sodium, in all its forms and functions, is replaced. This points to the development of blockers of the bitter taste of potassium as a key advance in the fight against hypertension, as potassium is interchangeable with sodium as the counterion in many sodium-containing ingredients added to processed foods. In addition, the understanding of, and technology surrounding, bitter blocker development is currently more advanced than for salty taste perception. Still further, most people, particularly in Western countries, do not consume enough potassium in their diet. The latest dietary guidelines in the United States, released in January 2005 by the Departments of Health and Human

Bitter blockers in foods and pharmaceuticals

223

Services and Agriculture, recommend a potassium intake of 4700 mg per day. However, the average American consumes only 2500 mg of potassium per day. Therefore, in addition to allowing for a reduction in sodium content in foods, the use of potassium as a replacement would help counter the deficit in potassium intake in the typical diet. The ability to modify the taste of pharmaceutical formulations is opening up new opportunities to increase patient compliance, improve treatment efficacy, expand formulation options, and extend product life. Non-compliance due to poor taste is still a major health problem in many applications, such as antibiotics, and in important patient populations, including children. For the approach most commonly used to mask unpalatable taste, tableting, compliance can be reduced in certain patient populations, particularly children, the elderly and patients with dysphagia, who have difficulty swallowing tablets. Improving taste allows more palatable liquid and orally disintegrating formulations, and enables drug formulations for oral absorption, which can increase the speed of onset for treatments of conditions such as migraine and breakthrough pain. Beyond the benefits to patients, the pharmaceutical industry has a number of compelling reasons for improving taste, such as the desire to extend a drug’s life cycle with new patented formulations, and to distinguish products from competitors with better tasting, more convenient dosage forms.

12.3

Current approaches to reducing bitterness

The food industry uses a number of approaches to improve taste acceptability of products, which can be summarised as masking, enhancing, complementing or removing tastants. Salt and sugar are routinely used to decrease the perception of bitterness, and microencapsulation can be used to physically prevent the bitter tastant from interacting with the tastereceptive cells of the tongue. The nucleotides inosine monophosphate and guanosine monophosphate are used to enhance savoury taste by potentiating the umami taste of monosodium glutamate. One of the best examples of complementing taste is found in the oral healthcare field, in the use of mint in toothpastes and mouthwashes, where the astringency of the active ingredients is far more acceptable to the consumer as it is positively associated with the taste of mint. The final approach to improving taste is to remove the offending compound from the final product. However, with many important nutrients possessing less than acceptable taste, this approach can impact negatively on the nutritional value of the product. There are currently two broadly distinct approaches used to improve the taste of pharmaceutical formulations. In the first, technically simpler and more cost-effective approach, compounds are added to the formulation to improve the palatability. These compounds include sweeteners, amino acids and flavours. However, as anyone with young children can attest to, this

224

Modifying flavour in food

approach is not 100% effective, particularly with active pharmaceutical ingredients (APIs) that are either highly bitter and/or highly water soluble. In addition, the use of caloric sweeteners is counter-indicated for those with diabetes and those prone to dental caries. The second approach, which can prove technically challenging, involves physically sequestering the unpalatable pharmaceutical and thus preventing it from contacting the taste-receptive cells of the mouth. This can be achieved using a number of approaches depending upon the specifics of the formulation and drug. Increasing the viscosity of the formulation with lipids, gums or carbohydrates effectively coats the taste buds, preventing interaction with the drug. Alternatively, the drug can be coated with hydrophilic molecules such as carbohydrates (e.g. cellulose), proteins (e.g. Zein), surrounded with a complexing agent such as cyclodextrin, or bound to ionexchange resins that make the drug insoluble until it reaches the acidic environment of the stomach. However, physical sequestration of the API limits formulation options, can decrease speed of onset, and can add significantly to the cost of goods. In addition, this approach can have its own compliance issues as water is typically required to take the medication, and certain populations have difficulty swallowing pills. Outside these two main strategies, other approaches to taste masking that may be possible with certain drugs include effervescence and the use of non-bitter salts or prodrugs of the API.

12.4

Advantages to bitter blocked pharmaceutical formulations

A key advantage with the use of a taste modifier in a pharmaceutical formulation is that it enables the development of liquids, fast-dissolve tablets and orally absorbed formulations for unpalatable actives that were not previously possible due to taste: •

Liquid formulations and orally disintegrating tablets can significantly improve ease of administration in patient populations that have difficulty swallowing tablets, such as children and the elderly. Clearly, if these formulations have acceptable taste, then compliance will increase. • Orally absorbed formulations can improve the efficacy of treatments where speed of onset is important, such as in anaesthesia, asthma, analgesia and migraine. • Orally absorbed formulations are possible for drugs that undergo significant first pass metabolism, and are today administered either by injection, or orally at high concentrations. The reduced quantity of API required in the new formulation can significantly reduce both toxicity to the patient of hepatic by-products, and the cost of API production for the manufacturer.

Bitter blockers in foods and pharmaceuticals

225

•

Taste modification enables formulations for rapid onset controlled selfmedication, such as lozenges and sticks containing drug that can be administered by the patient until symptoms are alleviated in conditions such as breakthrough pain and asthma. • As the effect of a taste modifier is in the mouth, it can be designed to have low bioavailability, and hence little or no systemic effect, either alone or through interaction with the API. • The use level of the taste modifier will be minimal, in the low micromolar range, compared with the large quantities of excipient necessary for (micro)encapsulation. This can significantly decrease the cost of goods of excipients, particularly for drugs given at high dose. For pharmaceutical companies, developments in taste science have a number of advantages throughout the entire life cycle of a drug: •

During the discovery process, palatability of lead actives can be determined and can factor in to the decision process as to which lead to take forward into development. • Knowing the taste profile of a lead compound allows formulators to address taste problems at an earlier stage of development than has been previously possible, thus reducing formulation-related delays in marketing the drug. • Improved taste can lead to improved clinical trial performance as compliance is increased, and the most appropriate delivery method is used regardless of the taste of the API. • Improved taste and formulation options provide a powerful means of differentiating a product from its competitors, both while it is covered by patent protection, and upon patent expiration and market entry of generic competitors. • For drugs reaching the end of their patent life, the new formulation options available due to taste modification allow a pharmaceutical company to maintain revenue either through marketing exclusivity or patent protection on a new formulation.

12.5

The science of taste perception

The perception of taste occurs predominantly in specialised structures called taste buds, clustered into taste papillae on the surface of the tongue. Taste buds contain groups of elongated taste receptor cells linked to the peripheral nervous system. The apical surface of these cells contains protein receptors and channels, which bind to, or allow entry of, tastants that come into contact with the taste bud in the mouth cavity. The purpose of taste receptor cells is to transduce the physical interaction of tastants at the apical surface into release of neurotransmitter from the cell’s basolateral surface. The basolateral surface is exposed to the peripheral nervous system and

226

Modifying flavour in food

the resulting increase in neurotransmitter concentration triggers activation of nerve cells resulting in a nerve impulse to the brain. For a review of the basics of taste, read Smith and Margolskee (2001). Five taste modalities are currently recognised: sweet, salty, sour, bitter and umami (savoury) (Smith and Margolskee, 2001; Kim et al., 2004). Aetiologically speaking, recognition of these various tastes has evolved to enable humans and animals to discern important information about the quality of food. Sweet, salty and umami tastes are associated with foods that contain nutrients important for well-being. Sweet-tasting foods are typically high in carbohydrates, salty food contains important minerals, and umami taste is coupled to the presence of amino acids. Sour and bitter taste perception is characteristically a protective mechanism against ingesting substances that may be deleterious to the body, such as spoiled food or poisons, respectively. Other tastes, such as the taste of polysaccharides, may be accepted as distinct taste modalities in the future (Sclafani, 2004). Since 1990 there have been rapid advances in the elucidation of the mechanisms by which the physical interaction of tastants at the exposed surface of taste receptor cells is transduced into neurotransmitter release from the basolateral surface of these cells (Fig. 12.1). In the early 1990s, the

Tastant

G protein coupled receptor G protein Effector enzyme Second messenger

Intracellular stores Ion channel

Ions

Neurotransmitter Nerve cell to the central nervous system

Ion channel

Fig. 12.1 Overview of taste transduction. Tastants bind to receptors in the apical surface of the taste cell (there are at least 25 G protein coupled receptors in the case of bitter taste). Receptor activation in turn triggers G proteins within the cell that switch on effector enzymes. Effector enzyme activity modulates the levels of second messengers within the cell, which leads to opening of ion channels. Ion influx results in depolarisation of the cell and subsequent release of neurotransmitter.

Bitter blockers in foods and pharmaceuticals

227

laboratory of Robert F. Margolskee reported the identification of a protein involved in taste (McLaughlin et al., 1992). This protein, called gustducin, interacts with receptor proteins in the apical membrane of taste cells and was the first taste protein to be characterised at the molecular level. Gustducin is a member of a family of proteins called G proteins, and hence the receptor proteins it interacts with are known as G protein coupled receptors (GPCRs). A number of G proteins have subsequently been identified in the taste system (McLaughlin et al., 1994). Upon tastant binding in the bitter, sweet and umami systems, a change in conformation of the receptor leads to activation of G protein. G protein activation in turn results in the switching on of intracellular proteins known as effector enzymes. These effector enzymes modulate the concentration of molecules within the cell called second messengers. In the resting state, taste cells are polarised owing to the active maintenance of concentration gradients of ions across the cell membrane. As second messenger levels change, various ion channels are activated both in the cell membrane and in intracellular membranes, resulting in the depolarisation of the taste cell as ions flow down their respective concentration gradients. Depolarisation results in release of neurotransmitter from the taste cell into the synaptic cleft and subsequent activation and depolarisation of the adjacent neuron. This electrical signal is transmitted through the nervous system to the brain, where it is interpreted as taste. There are at least 25 GPCRs proposed to be involved in bitter taste (Adler et al., 2000; Matsunami et al., 2000). These GPCRs, known as the T2Rs, enable the detection of the wide variety of structural classes of compounds that are bitter. This large family of receptors also speaks to the aetiological importance of bitter taste detection. If only one receptor was involved in bitter taste, then a non-functioning mutation of this receptor would probably result in death as there would be no first-line detection apparatus for many poisons. It appears that not all bitter tastants exert their bitter taste by activating T2R receptors. It has been reported that certain amphiphilic substances, including H1-receptor antagonists, can directly activate G proteins (Naim et al., 1994; Burde et al., 1996). It is possible that the GPCR-independent mechanism of G protein activation represents a bitter sensing mechanism for many bitter tastants present at high concentrations. Other bitter compounds, such as the methyl xanthines (e.g. caffeine in coffee, theophylline in tea and theobromine in cocoa) are known to interact with phosphodiesterase (PDE), which is one of the effector enzymes present in taste cells. Studies have shown that in response to caffeine and theophylline, taste cell levels of the substrate for PDE rises (Rosenweig et al., 1999). Indeed, it can be predicted that the taste of any compound that modulates the level of activation of a protein involved in the taste transduction cascade could be altered if the compound comes into contact with that protein. The taste system also contains a number of ion channels that sense other characteristics such as astringency, burning and cooling. These channels

228

Modifying flavour in food

belong to the transient receptor potential (TRP) family of gated ion channels, and can have positive or negative impacts on taste acceptability.

12.6

Identifying compounds that decrease the perception of bitterness

Knowledge of how unpalatable tastants are perceived at the molecular level enables the application of biopharmaceutical methods to identify taste modifiers. By using recombinant DNA technology, proteins involved in bitter taste transduction can be overexpressed and regulated in cellular systems and their modulation by aversive tastants and test compounds quantified. These engineered cell lines are formatted to enable the rapid screening of thousands of molecules to identify taste-modifying compounds. This throughput is many times greater than is possible with traditional approaches that rely on laborious human sensory testing at an early stage of the discovery process. Using these methods, taste modifiers with suitable characteristics of taste profile, stability and safety are being identified. Recombinant DNA techniques are also being used to determine the structure–activity relationship (SAR) between taste modifiers and the proteins with which they interact. Mutagenesis of receptor proteins enables determination of essential amino acid residues and regions of the receptors involved in binding of tastants and other taste cell components. An early success with using this biopharmaceutical approach to improve taste was achieved with an assay that monitored the activation of the taste G protein by bitter tastants. Adenosine 5′-monophosphate (AMP) was identified as a compound that reduces taste cell activation by bitter compounds (Ming et al., 1999). Mouse preference studies confirmed that AMP was improving the palatability of bitter solutions, and electrophysiological recordings showed a decreased activation of nerve responses by bitter compounds in the presence of AMP. Human sensory studies have subsequently demonstrated that one of AMP’s most robust activities is reducing the bitter taste of potassium. Figure 12.2 shows the effect of AMP on the taste of a low-sodium soup in which potassium chloride (KCl) was added to replace the absent sodium. It can be seen that the soup containing KCl/AMP is perceived as significantly less bitter than the soup with KCl alone. Larger-scale consumer testing has confirmed these findings. An interesting observation coming out of these studies is that the addition of AMP leads to increased saltiness and umami in the soup. This is probably due to two factors, the inherent umami taste of AMP at higher concentrations and the reduction of bitterness, leading to the saltiness and umami taste of the soup being more pronounced. Further research has identified formulations containing AMP and other ingredients that work better than AMP alone at improving the taste of low

Bitter blockers in foods and pharmaceuticals 5

Control AMP (800 ppm)

4 Intensity scores

229

3 2 1 0 Bitter*

Salt*

Chicken flavour*

Flavour attributes

Fig 12.2 AMP reduces the bitterness of low-sodium chicken broth supplemented with 1.5% potassium chloride, and increases the saltiness and overall flavour. *Statistically significant, 95% CI; 19 trained panellists; 0–8 intensity scale, 8 = highest intensity.

8 7 6 5 4 3 2 1 0 Salt

Bitter

100% NaCl

50% NaCl + 0.9% KCl

Chicken flavour 50% NaCl + 0.9% KCl + BetraTM

Fig 12.3 BetraTM-supplemented low-sodium chicken soup with potassium chloride gives a product with taste attributes similar to the full sodium product. Attribute ratings: 9-point structured scale where 0 = no taste and 8 = very strong.

sodium, KCl-containing products. This bitter blocker formulation, known as BetraSaltTM, has been shown to effectively function as a NaCl substitute in a number of applications. Figure 12.3 demonstrates the effects of AMP and BetraSaltTM on the taste of a low-sodium gravy containing KCl. Decreasing the amount of sodium in the gravy has an adverse effect on saltiness and overall flavour quality. The addition of KCl improves the

230

Modifying flavour in food

saltiness of the gravy, but introduces a significant bitterness that prevents the overall flavour profile from coming close to the full-sodium product. The addition of BetraSaltTM to a 50% reduced sodium, KCl-containing formulation significantly reduces this bitterness and improves the saltiness, beef flavour and overall flavour of the gravy to the point where, it is equivalent to the full sodium gravy.

12.7

Future trends

As high-throughput screening provides more data on the taste properties of specific compound classes, and the proteins involved in taste transduction are characterised at the atomic level by modelling and crystallisation, the full SAR between tastants, receptors and transduction components will be determined. This will enable ever more focused libraries of compounds to be screened to discover novel taste modifiers, and allow predictive modelling of the taste properties of molecules by computer, and ultrahigh throughput screening in silico. Computer-generated hits will only be synthesised once their taste-modifying properties have been determined. Full knowledge of SAR will allow scientists to design specific properties into taste modifiers, such as time of onset, duration of action and target compounds to modify. As there are over two dozen bitter taste receptors, it can be envisioned that blocking a subset of these receptors would target a particular population of bitter compounds, whereas interceding further downstream in the transduction process would give bitter blocking to a broader range of compounds. All these advances will broaden the approaches food, beverage and pharmaceutical companies can use to improve the taste of their products, while at the same time making them more nutritionally beneficial and efficacious.

12.8

Sources of further information and advice

For an introduction to taste, Smith and Margolskee (2001) is a useful primer. A more detailed review of the genetics of taste perception is given by Kim et al. (2004).

12.9

References

adler, e., hoon, m. a., mueller, k. l., chandrashekar, j., ryba, n. j. p. and zuker, c. s. (2000), A novel family of mammalian taste receptors. Cell, 100, 693–702. burde, r., dippel, e. and seifert, r. (1996), Receptor-independent G protein activation may account for the stimulatory effects of first generation H1-receptor

Bitter blockers in foods and pharmaceuticals

231

antagonists in HL-60 cells, basophils and mast cells. Biochem. Pharmacology, 51, 125–131. jones, d. w. (2004), Dietary sodium and blood pressure. Hypertension, 43, 932–935. kim, u.-k., breslin, p. a. s., reed, d. and drayna, d. (2004), J. Dent. Res, 83 (6), 448– 453. matsunami, h., montmayeur, j.-p. and buck, l. b. (2000), A family of candidate taste receptors in human and mouse. Nature, 404, 601–604. mclaughlin, s. k., mckinnon, p. j. and margolskee, r. f. (1992), Gustducin is a tastecell-specific G protein closely related to the transducins. Nature, 357, 563–569. mclaughlin, s. k., mckinnon, p. j., spickofsky, n., danho, w. and margolskee, r. f. (1994), Molecular cloning of G proteins and phosphodiesterases from rat taste cells. Physiol. Behav, 56 (6), 1157–1164. ming, d., ninomiya, y. and margolskee, r. f. (1999), Blocking taste receptor activation of gustducin inhibits gustatory responses to bitter compounds. Proc. Natl Acad. Sci. USA, 96, 9903–9908. naim, m., seifert, r., nurnberg, b., grunbaum, l. and schultz, g. (1994), Some taste sunstances are direct activators of G proteins. Biochem. J, 297, 451–454. rosenweig, s., yan, w., dasso, m. and spielman, a. i. (1999), Possible novel mechanism for bitter taste mediated through cGMP. J. Neurophysiol, 81 (4), 1661–1665. sclafani, a. (2004), The sixth taste? Appetite, 43 (1), 1–3. smith, d. v. and margolskee, r. f. (2001), Making sense of taste. Scientific American, 284, 32–39.

13 Masking agents for use in foods M. Gascon, Wixon Inc., USA

13.1

Introduction

During the past few years, the leading forces in the development of new food and beverage products have been convenience, health and wellness. Not so long ago, the idea of getting your entire meal within a pill was only in science fiction movies. Today, the concept of food in a pill is not so farfetched. Meal-replacement products, energy bars and sports drinks are quite familiar to most consumers. Many are turning to quick nutritional fixes to support on-the-go lifestyles. Consumers are becoming aware of the potential health values of specific food compounds. Numerous food staples are being reformulated to provide better nutritional value with healthy ingredients, fewer calories, more fibre, less fat, lower levels of carbohydrates and more protein, but above all, it is imperative that they maintain the same traditional taste to which consumers have been accustomed. Food companies continually cope with the emerging challenges of creating products that meet consumer demands and still taste good. Supermarkets are now filled with new food products that are fortified, enriched or enhanced with dietary supplements. Often, though, the addition of healthful components adds off-flavours. Unpleasant notes – described as excessive salt, sour, bitter, astringent, metallic, etc. – require delicate adjustments from one application to the next. To keep up with consumer demands for health, convenience and good flavour, masking agents have become an essential tool for food technologists. Masking agents are flavour ingredients that have little or no taste or smell of their own but complement, enhance or otherwise modify the flavour of a food product. They may improve overall flavour, modify the character

Masking agents for use in foods

233

of off-flavours, enhance desirable ones or inhibit the impact of undesirable ones. There are many flavour ingredients and combinations for masking agents which will be further described in this chapter. However, it is important to understand that there is no single ingredient capable of fixing all flavour issues. Rather, the solution to masking undesirable flavour notes in food and beverage systems relies on the right combination of different agents.

13.2

How masking agents work

Masking agents work in different ways depending upon what they are masking and the food application. Most commercial masking agents are complex mixes of modifiers, inhibitors and enhancers that cover an undesirable flavour characteristic by providing other sensations (taste modifying) or by competing with specific receptor sites (taste suppressing), or accentuating other flavours (taste enhancers). Taste is both psychological and physiological. It is important to remember that taste receptors are in part a self-defence mechanism against potential hazards. They are nature’s way of telling you not to eat poisonous food since rotten food, poisonous plants and toxic chemicals often taste bitter or sour. The function of masking agents is then to trick your taste into letting you swallow otherwise unpalatable foods or those with flavour attributes that the receptors recognise as potentially harmful. Masking tastes is embedded in our culture: for instance, sugar has been used for centuries to mask any bitterness or sourness. When drinking a cup of coffee, it is common practice to add some sugar to reduce the bitter taste. Even Mary Poppins tells us how ‘a spoonful of sugar helps the medicine go down’. A further example can be found from the low-fat food craze at the end of the last century. Most fat-free salad dressings were particularly sweet because the oil in traditional salad dressings mellows the acidity of vinegar; however, when the oil was removed, the formulators often disguised the acidity with sugar. Because masking agents are often a blend of ingredients, how they work can be explained by understanding some of the taste phenomena in which they may be involved. These include, but are not limited to, adaptation, cross-adaptation, taste blocking and taste modifying. The sense of taste exhibits almost complete adaptation with continuous exposure to a taste stimulus and the perception of a substance fades to almost nothing in seconds. This experience is called adaptation and it is a form of fatigue. When applied to a single substance it is known as self-adaptation. For example, if adaptation to the sweetness of sucrose diminishes the sweetness intensity of a second sweetener, it is called cross-adaptation (Bartoshuk, 1987). Moreover, adaptation between other tastes can also

234

Modifying flavour in food

occur (e.g. it has been reported that adaptation to bitter taste lowers the perception threshold for other tastes; Dallenback and Dallenback, 1943). In other words, adaptation to one taste may lower or even increase the perception threshold of the other taste perceptions. For the present writer, if the tongue is treated with a dilute solution of citric acid until the solution is no longer perceived as being overwhelmingly sour, a subsequent taste of a dilute solution of quinine may not be perceived as bitter. This personal experience agrees with reports that, while bitter adaptation does not reduce saltiness, sour adaptation will reduce bitterness (McBurney, 1974). Alternatively, adaptation to citric acid and bitter compounds did not reduce sweetness, but enhanced it (Shallenberger, 1992). Another taste phenomenon useful in the design of masking agents is taste blocking. There are known substances that have the capacity to suppress all taste, like local anesthaesia used at the dentist. There are also known substances that suppress a particular taste perception such as the triterpene saponin glycosides, including gymnemic acid, present in Gymnema sylvestre, a native plant of India; that suppress the perception of sweetness (Kolodny and Kennedy, 1988). Low concentration solutions of gymnemic acid are reported to suppress only sweetness and have little effect on sourness, saltiness and bitterness; however, higher concentrations suppress all tastes (Kurihara, 1971). Another phenomenon to be considered is taste modification. The typical example for this phenomenon is miraculin. This protein is present in the so-called miracle fruit and it has the potential to turn sour tastes into sweet perceptions (Faus and Sisniega, 2003). Finally, in assessing the right combination of masking agents, the concepts of taste suppression and taste synergism may also be explored. For example, mixtures of sugars have shown synergism at low concentrations while, at the same time, showing suppression effects at high concentrations (Curtis et al., 1984). When substances with different taste properties are mixed, suppression or enhancement of some of the tastes is likely to occur; however, the total intensity of the taste perceived is not easily determined. There are extensive discussions on the mode of action of taste phenomena (see Shallenberger, 1992). Ultimately, evaluating combinations of taste mixtures may be difficult, and requires trained flavour chemists who understand how each flavour component acts independently and can combine all needed flavour modifiers into one customised agent for a specific product. The importance of working with a flavour house is emphasised by another consideration: to ensure that the flavour does not react with any other ingredient in the food product, rendering it inactive. Creating a masking agent may be best left to a flavourist because the development often requires creativity in formulating unique combinations of masking agents. Individual flavour ingredients have limited masking capability, but in combination, may have synergy and, consequently, enhanced masking capability.

Masking agents for use in foods

13.3

235

Ingredients used to formulate masking agents

Masking is the most difficult trick of the flavour trade. Each challenge is likely to be application-specific, often requiring innovative and unique approaches for the given problem. There is currently no specific structure– activity theory dictating which ingredients should be used to design masking agents. Rather, they are developed through an in-depth knowledge of ingredients that have specific functional taste properties. For example, some high-intensity sweeteners may elicit flavour-enhancing or even flavourmodification properties when used below their taste threshold concentrations. This, however, is not applicable to all high-intensity sweeteners. Only in-depth knowledge of the ingredients allows them to be utilised to full benefit. For the purposes of this chapter, the ingredients used to formulate masking agents will be broadly classified into two categories: those that possess intensive sweetness and those that exhibit little or no sweetness response. A classification of the various ingredients that can be used to make masking agents is provided in Table 13.1. 13.3.1 Ingredients that exhibit high-potency sweetness Glycyrrhizin and its salts Glycyrrhizin, also known as glycyrrhizic acid, is primarily known as a flavouring agent and flavour enhancer, though exhibiting sweetness. It has been rated as approximately 50–100 times sweeter than sucrose with a characteristic slow onset of sweet taste and a long aftertaste. It has a synergistic sweetening effect in the presence of sucrose. This compound, an oleanane-type triterpene glycoside, is extracted from the roots of liquorice (Glycyrrhiza glabra L., Fabaceae) and other species in the same genus (Kinghorn and Compadre, 2001). The US Food and Drug Administration (FDA) classifies the ammonium salt as GRAS (generally recognised as safe) under parts 184 of the code of Federal regulations. There are many applications using this compound as a flavour modifier and flavour enhancer. Neohesperidin dihydrochalcone Neohesperidin dihydrochalcone is an intense non-nutritive sweetener derived from neohesperidin, a naturally occurring bitter-tasting flavanone from citrus fruit. Although its relative sweetness diminishes rapidly with increasing concentration, at a sweetness level equivalent to 5% sucrose, it is about 250 times as sweet as sucrose (Shallenberger, 1992). Like other highly sweet glycosides, such as glycyrrhizin, neohesperidin DC exhibits a long-lasting sweetness at high concentrations, associated with a licorice-like aftertaste and it has an apparent synergism with citric acid (Leffingwell, 2004). It has not been approved as a sweetener in the United States, although it is considered GRAS as a flavour enhancer.

236

Compound class

Compound name

Triterpene glycosides Dihydrochalcones Proteins

Glycyrrhizin and its salts (MagnifiqueTM a) Neohesperidin dihydrochalcone Thaumatin (TalinTM b) Miraculin 3-Hydroxy-2-methyl-pyran-4-one (maltol) 3-Hydroxy-2-ethyl-pyran-4-one (ethyl maltol) N-Neohexyl-alpha-aspartyl-1-phenylalanine methyl ester (NeotameTM c) 2-(4-Methoxyphenoxy)propanoic acid and its salts (LactisoleTM b) Alanine Glycine Lysine Adenosine monophosphate (AMP)

Pyrones Aspartic acid peptides Phenoxyacetic acid derivatives Amino acids Nucleotides a b c

Trademarked by Wixon Industries. Trademarked by Tate & Lyle Industries. Trademarked by NutraSweet Property Holdings.

High-potency sweetness response

Commercially available in the US

Yes Yes Yes No No No Yes

Yes Yes Yes No Yes Yes Yes

No

Yes

No No No No

Yes Yes Yes Yes

Modifying flavour in food

Table 13.1 High-potency sweetness response and structural type of ingredients that may be used to make masking agents

Masking agents for use in foods

237

Thaumatin Thaumatin is primarily known as a high-intensity sweetener. However, the FDA classifies it as a flavouring agent. It has been rated as approximately 3000 times sweeter than sucrose on a weight basis. The thaumatins are a family of very sweet proteins present in the fruits of the tropical plant Thaumatococcus daniellii Benth. At high concentrations, it exhibits a longlasting sweet effect that might not be considered acceptable to some palates (Faus and Sisniega, 2003). Thaumatin is included in the GRAS list of approved natural flavouring agents by the FDA. There are many applications for this compound as a flavour modifier and flavour enhancer. NeotameTM NeotameTM, an aspartyl-derived dipeptide, is a high-intensity sweetener. It has been reported to have a sweetness intensity 8000 times sweeter than sucrose. Like any other high-intensity sweetener, neotame often produces a sweet taste that has a different temporal effect from sugar and a lingering aftertaste (Gerlat et al., 2002; Prakash et al., 2003; Shoichi, 2003, 2004). Approved by the FDA for general use in July 2002, NeotameTM has not been widely applied, but there are several potential applications using this compound as a flavour modifier and flavour enhancer. 13.3.2 Ingredients that have little or no sweetness response Miraculin Miraculin is a well-known taste modifier. This protein was first isolated from the so-called ‘miracle fruit’. These are the red berries of Synsepalum dulcificum Daniell or Richardella dulcifica, a small evergreen shrub native of West Africa (Theerasilp and Kurihara, 1988). The fruit is a small bright red berry containing a single seed. Miraculin by itself does not elicit a sweet response but it can modify a sour taste into a sweet taste (Faus and Sisniega, 2003). In other words, when a single fruit is eaten and the fleshy pulp has coated the taste buds of the tongue and inside of the mouth, an interesting effect occurs. The fruit allows one to eat a slice of lemon without detecting the intense sourness. One can taste the flavour and inherent sweetness of the lemon but the sourness is almost completely diminished. This effect is persistent and it may remain for about 20 minutes or more depending on the concentration of protein in the mouth (Shallenberger, 1992). Although the isolated protein is no longer commercially available in the United States as a food additive (O’Brien Nabors and Inglett, 1986), it is known that some miracle fruit extracts are being used in food products, particularly in Asia. Maltol Maltol (3-hydroxy-2-methyl-pyran-4-one), is the common choice when it comes to flavour enhancers. This material has the characteristic burnt sugar aroma of ‘cotton candy’ prepared by the melting of sugar. On dilution, it

238

Modifying flavour in food

takes on a fruity (strawberry–pineapple) note. It is widely used in flavours but it is also used as a ‘sweetener’ and flavour enhancer (Leffingwell, 2004). It has an enhanced effect in the presence of some other flavour ingredients such as vanillin, some of which act as enhancers with maltol (Arctander, 1969). Maltol is included in the GRAS list of approved flavouring agents by the FDA. Ethyl maltol Ethyl maltol, 3-hydroxy-2-ethyl-pyran-4-one, is an analogue of maltol. It is the obvious choice when it comes to flavour enhancers. It has a sweet caramel-like odour and fruity taste, and it is a lot more powerful than maltol (Leffingwell, 2004). This material is almost the same in its odour and flavour profile as Maltol except it is about five times stronger in effect (Arctander, 1969), and is included on the GRAS list of approved flavouring agents by the FDA (Leffingwell, 2004). LactisoleTM LactisoleTM or PMP is the sodium salt of 2-(4-methoxyphenoxy)-propionic acid. It is a selective competitive inhibitor of sweet taste and it is reported to reduce the sweetness of sucrose and other sweeteners by as much as 80%, or more, when present in relatively low concentrations (Lindley, 1986; Schiffman et al., 1999). The effect is reversible and first order with respect to inhibitor concentration. It has a slightly bitter taste by itself and it does not impact on the salt, sour or bitter taste (Shallenberger, 1992). LactisoleTM can actually enhance sweetness in the presence of monoammonium glycyrrhizinate, neohesperidin dihydrochalcone and thaumatin (Schiffman et al., 1999). 2-(4-Methoxyphenoxy)-propionic acid was found to be a naturally occurring constituent of roasted Colombian coffee beans (Rathbone et al., 1989) and its sodium salt, LactisoleTM, is designated as GRAS by the FDA. Alanine Alanine is an α-amino acid. It exists as two distinct enantiomers – l-alanine and d-alanine. Both forms possess a pleasant sweet taste when tasted in solution (Shallenberger, 1992). Anecdotally, the author has found that the racemic mixture of alanine has a moderate enhancing effect on the sweetness of sucrose and a mild suppressing effect on the acidity of citric acid solutions. Alanine is included on the GRAS list of approved flavouring agents by the FDA. It has been reported that as a food additive, dl-alanine may be safely used as a flavour enhancer for sweeteners in pickling mixtures (Leffingwell, 2004). Glycine Glycine is a neutral amino acid with no chiral centre and therefore no enantiomeric forms. It possesses a pleasant sweet taste when tasted by itself

Masking agents for use in foods

239

(Shallenberger, 1992). It is a well-known flavour enhancer. Additionally, it has been reported to reduce the bitter aftertaste of saccharin and improve the mouthfeel in carbonated dietetic beverages as well as improving the taste in aspartame/saccharin blends (Leffingwell, 2004). Glycine is included in the GRAS list of approved flavouring agents by the FDA. Lysine Lysine is a basic amino acid. It is an essential amino acid that exists as two enantiomers – l-lysine and d-lysine. Only the l-form possesses a pleasant bitter/sweet taste (Shallenberger, 1992). Lysine has GRAS status by the FDA. It has been reported that lysine masks the bitter aftertaste of potassium chloride when used as a salt replacer and in dietary supplements (Berglund and Alizadeh, 1999). Adenosine monophosphate Linguagen Corporation obtained FDA approval in September 2004 to employ adenosine monophosphate (AMP), a nucleotide, as a bitter blocker to be used as an additive in foodstuffs. This ingredient, along with other nucleotide compounds, is capable of inhibiting the activation of Gproteins involved in the perception of bitterness (McGregor and Gravina, 2005). This subject is discussed more fully in Chapter 12.

13.4

What to consider when working with masking agents

It is important to remember that there is no single ingredient capable of fixing all flavour problems. Rather, the solution to masking undesirable flavour notes in food and beverage systems relies on applying the right combination of individual flavour masking agents. There are a few other considerations to take into account when working with masking agents. First, they are very application-specific. Food developers should consider working closely with their flavour supplier to have a masking agent specifically designed for their application. Second, keep in mind that masking agents always function as additive ingredients because we add them on top of the undesirables. There is always the possibility that some of the components of a masking agent may react with other ingredients in the food system. It is important that the entire food or beverage system be evaluated, including processing and packaging. All of these variables may affect how the masking agent performs. Third, it should be noted that the function of masking agents is to mask undesirables: excessive use may suppress desirable flavours in a food system. Lastly, when using masking agents, remember that, if the product is reformulated or any part of the manufacturing process changes, the effectiveness of the masking agent may need to be re-evaluated.

240

13.5

Modifying flavour in food

Masking agents: when to use or not to use

The appearance of unexpected off-flavours has been a common challenge to many food developers who attempt to modify standard formulations. For example, fat can either accentuate or cover up flavours. If the fat content of a product is reduced, one may start seeing other flavours come through disproportionately to the situation before the modification. The addition of healthful ingredients may also upset the mix by introducing unwanted notes. For instance, soy protein may introduce a ‘beany’ character, certain extracts can impart grassiness, and minerals can add bitterness, a metallic taste or even astringency. Before adding masking agents, it is essential to look at the whole formula first. One should consider whether adjusting the sweeteners, acids, starches or fats may help suppress off-notes. For minor problems, there are commercially available masking flavors formulated to suppress certain specific off-flavours, such as green notes, bitterness and astringency. More complex flavour problems generally require using a combination of masking agents. Taste is a delicate balance and once one taste perception is modified, the other tastes may become unbalanced. For example, when masking soy, some excessive sweetness or sourness may appear. This may require adjustments in the bulk formulation or may require the addition of a second masking agent to balance. Frequently, off-flavours involve a combination of different factors, not one specific one. For off-flavours that are intense and persistent, a lot of fine-tuning work is required. This often means the food technologist must work closely with the technical staff of masking agent suppliers.

13.6

Outlook and perspectives

Since the mid-1990s, there has clearly been a need for new and improved masking agents; particularly with today’s marketplace emphasising the need for nutrition and wellness. Several companies have ongoing research and development programmes to meet these needs that have led to a better understanding of the chemistry of taste. Consequently, the flavour industry continually evolves with the introduction of new ingredients and technologies; for example, the novel use of biochemical and molecular biological approaches that has identified substances such as nucleotides, capable of blocking the bitter taste of bitter compounds. Additionally, an extraordinary number of sweet-tasting proteins have been discovered – monellin, mabilin, pentadin and brazzein among others – opening the possibility of exploring new combinations of masking agents (Faus and Sisniega, 2003). Furthermore, additional work has also been made on substances capable of taste modification, for example curculin, a protein isolated from the plant Curculigo latifolia can modify taste perception, like miraculin, and convert a sour taste into a sweet taste (Faus and Sisniega, 2003).

Masking agents for use in foods

241

In conclusion, masking is the trickiest part of the flavour trade. Each challenge is unique and will require a unique approach. There will be times when the problem is to mask a very specific bitter compound, driving the need to utilise bitter blockers technology. In cases where the problems come from complex ingredients or multiple sources, for example, when the problem in question is to mask a good-for-you ingredient that elicits sourness, bitterness and astringency all at the same time, the use of blends of ingredients that target different taste pathways remains the only current solution. Ultimately, the function of masking agents is to hide the negatives and accentuate the positives.

13.7

References

arctander, s. (1969), Perfume and Flavor Chemicals, Montclair, Steffen Arctander. bartoshuk, l. m. (1987), ‘Is sweetness unitary? An evaluation of the evidence for multiple sweets’ in Sweetness, Dobbing J, London, Springer-Verlag, 33–47. berglund, k. a. and alizadeh, h. (1999), Composition and method for producing a salty taste, US Patent 5897908. curtis, d. w., stevens, d. a. and lawless, h. t. (1984), Perceived intensity of the taste of sugar mixtures and acid mixtures. Chem Senses, 9, 107–120. dallenback, j. w. and dallenback, k. m. (1943), The effects of bitter-adaptation on sensitivity to other taste-qualities. Am. J. Psychol., 56, 21–31. faus, i. and sisniega, h. (2003), ‘Sweet-tasting proteins’ in Fahnestock SR and Steinbüchel A, Biopolymers, Weinheim, Wiley-VCH, 203–220. gerlat, p. a., walters, g. c., bishay, i. e., prakash, i., jarrett, t. c., desai, n., sawyer, h. a. and bechert, c. t. (2002), Use of additives to modify the taste characteristics of N-neohexyl-α-aspartyl-l-phenylalanine methyl ester, US Patent 6368651. kinghorn, a. d. and compadre, c. m. (2001), ‘Less common high-potency sweeterners’ in Food Science and Technology, New York, Marcel Dekker, 209–234. kolodny, d. e. and kennedy, l. m. (1988), A model system for receptor cell studies with the taste modifier, hodulcin. Chem Senses, 13, 545–557. kurihara, k. (1971), ‘Taste modifiers’ in Handbook of Sensory Physiology. IV. Chemical Senses, Berlin, Springer-Verlag, 363–378. leffingwell, j. c. (2004), Flavor-Base 2004-database, Georgia, Leffingwell & Associates. lindley, m. g. (1986), Method of inhibiting sweetness, US Patent 4567053. mcburney, d. h. (1974), Are these primary taste for man? Chem. Senses Flavour, 1, 17–28. mcgregor, r. a. and gravina, s. a. (2005), Nucleotide compounds that block the bitter taste of oral compositions, US Patent 6942874. o’brien nabors, l. and inglett, g. e. (1986), ‘A review of various other alternative sweeteners’ in Alternative Sweeteners, O’Brien Nabors L and RC Gelardi, New York, Marcel Dekker, 309–323. prakash, i., guo, z., schroeder, s. and wachholder, k. l. (2003), Synthesis of N[N-(3,3-dimethylbutyl)-l-α-aspartyl]-l-phenylalanine 1-methyl ester using l-αaspartyl-l-phenylalanine 1-methyl ester precursors, US Patent 6642406. rathbone, e. b., patel, g. d., butters, r. w., cookson, d. and robinson, j. l. (1989), Occurrence of 2-(4-methoxyphenoxy)propanoic acid in roasted coffee beans: analysis by gas-liquid chromatography and by high-performance liquid chromatography. J Agric Food Chem, 37, 54–58.

242

Modifying flavour in food

schiffman, s. s., booth, b. j., sattely-miller, e. a., graham, b. g. and gibes, k. m. (1999), Selective inhibition of sweetness by the sodium salt of ±2-(4methoxyphenoxy)propanoic acid. Chem. Senses, 24, 439–447. shallenberger, r. s. (1992), Taste Chemistry, New York, Chapman & Hall. shoichi, i. (2003), Sweetener compositions and uses thereof, US Patent 6652901. shoichi, i. (2004), Sweetener compositions containing aspartyl dipeptide ester compounds, US Patent 6761922. theerasilp, s. and kurihara, y. (1988), Complete purification and characterization of the taste-modifying protein, miraculin, from miracle fruit. J. Biol. Chem., 263, 11536–11539.

14 Selecting the right flavourings for a food product K. B. de Roos, Givaudan Nederland B.V., The Netherlands

14.1

Introduction

The perceived flavour of a food or beverage is the result of the combined perception of aroma, taste and mouthfeel resulting from the stimulation of receptors in the oral and nasal cavities (Ney, 1988; Taylor et al., 2003). From a flavour performance point of view, it is most convenient to define aroma, taste and mouthfeel as follows: •

Aroma or odour is the sensation resulting from the interaction of volatile chemicals with receptors in the nose. To reach the receptors in the nose, aroma compounds must be volatile. • Taste is the sensation resulting from the interaction of chemicals with receptors in the mouth. To be perceptible, taste compounds should be soluble in the saliva so that they can be transferred from the product to the receptors in the mouth. According to this definition, all chemicals that elicit a sensation in the mouth are referred to as taste compounds, including the trigeminal (pain) sensations such as astringency and pungency. • Mouthfeel covers the tactile sensations, which are the sensations that can be perceived by touch such as texture and temperature. In contrast with taste, which is the result of interactions with chemicals, mouthfeel is the result of physical interactions. Examples of aroma, taste and mouthfeel sensations as defined above are given in Table 14.1. The average consumer does not distinguish between these flavour attributes; he or she perceives flavour as a single sensation. Most flavourings consist only of aroma compounds and do not provide taste and mouthfeel. A few flavourings (e.g. savoury flavourings) also

244

Modifying flavour in food

Table 14.1 Examples of aroma, taste and mouthfeel sensations that contribute to the flavour Aroma: caramel, vanilla, cinnamon, buttery, minty Taste: sweet, acid, salty, bitter, umani, astringent, pungent, soapy, cooling, metallic Mouthfeel: liquid, solid, gas, fibrous, warm, viscous, elastic, creamy, juicy, cold

contain taste compounds but, in general, it is the food company that takes care of the taste. Since most flavourings are just aroma compositions, the focus in this chapter will be on volatile flavourings. To create a flavouring that generates the desired flavour experience in a product, the flavour creation process has to go through the following steps: 1. 2. 3. 4.

Creation of the desired flavour profile. Reformulation of the flavouring for higher stability, before use, if necessary. Rebalancing of the flavouring to compensate for the differences in flavour release between products. Reformulation of the flavouring for higher stability in the product, if necessary.

Each of these steps will be discussed in a separate section (14.2 to 14.5). The final sections 14.6 and 14.7 deal with future trends and sources of further information.

14.2

Creation of the desired flavour profile

14.2.1 Gathering information about the flavour composition The creation of a new flavour starts with gathering information about its composition. This information is usually obtained by analysis of the volatile compounds of a natural product that has the desired flavour profile. The first critical step of the analysis is the preparation of a representative aroma concentrate. To separate the aroma compounds from the food, use is made of one of the following two categories of physical separation methods (Weurman, 1969; Wilkes et al., 2000): •

•

Methods making use of the volatility of aroma compounds (steam distillation, high vacuum distillation or sublimation, headspace analysis). These methods have the advantage of providing clean aroma concentrates suitable for direct analysis by gas chromatography-mass spectroscopy (GC-MS) analysis. A disadvantage is the often low recovery of poorly volatile compounds (Krings et al., 2003). Methods making use of the differences in hydrophobicity between aroma compounds and other food components. Most frequently used are

Selecting the right flavourings for a food product

245

solvent extraction and adsorption (Weurman, 1969; Krings et al., 1993). The disadvantage of these methods is that non-volatile food ingredients are also extracted. Further clean-up is then often required such as (high-) vacuum distillation or sublimation (Sen et al., 1991) or gel permeation chromatography (de Roos and Graf, 1995). To prevent artefact formation during flavour isolation, high temperatures should be avoided. Therefore, distillations should be carried out in vacuo (Budin et al., 2001; Sen et al., 1991). However, at low temperatures there is a risk of enzymatic reactions taking place, in particular with damaged fruits and vegetables (Tressl et al., 1981). To avoid enzymatic oxidation reactions, vacuum steam distillations should be carried out under nitrogen, starting with the intact fruits or vegetables. When the fruit or vegetable disintegrates during distillation and releases its enzymes, no oxidation can then take place. Another method to reduce the chance of enzymatic reaction is by inhibition with methanol or calcium chloride (Schreier et al., 1977). The next step in the analysis of the flavour is the identification and quantification of the constituents of the flavour extract. In most cases identification is done by GC-MS. For the identification of new compounds, infrared spectroscopy and nuclear magnetic resonance (NMR) are used as well, provided that sufficient quantities can be made available. The result of the analysis is a long list of volatile compounds. Unfortunately, these compounds normally have all kinds of odours, except for the odour of the analysed product itself. Therefore, one of the first questions that has to be answered is, which combination of compounds is needed to create the desired aroma profile? To answer that question, it is important to know which compounds make the highest contribution to the total flavour impact. The following methods are often used to select these compounds: •

•

Gas chromatography-olfactometry (GC-sniff). The selection of the most important aroma compounds is often based on the results of the semiquantitative GC-olfactometric methods known as aroma extract dilution analysis (AEDA; Grosch, 1994, 2001) and CHARM analysis (Acree, 1997). However, these methods ignore that the aroma concentrations in the extracts are different from those released during consumption, which gives a distorted picture of their relative importance (de Roos and Nelissen, 2000). AEDA of headspace (the air above the sample) gives a better impression of the relative importance of the aroma compounds because it analyses the released aroma concentrations, although not exactly in the proportions that are released during consumption (see also Table 14.2 below). Calculation of the odour-activity values. A better, although not yet perfect, criterion for the selection of the most important aroma compounds is their odour activity during consumption (Frijters, 1978; Grosch, 1994). Odour-activity values (OAVs) can easily be calculated if the

246

Modifying flavour in food

odour thresholds C at in the air and the released aroma concentrations Ca during consumption are known: AOV =

Ca Cat

[14.1]

The resulting list of the most odour-active aroma compounds is still likely to contain artefacts and unreliable quantitative data owing to incomplete recovery during the aroma isolation. So, the creation of a flavouring requires more than just mixing the listed odour-active compounds. A further complication is that these compounds are not always available in the required purity. Therefore, it can easily happen that impurities in the synthetic and natural aroma ingredients disturb the profile of the flavouring. 14.2.2 Creation of the desired flavour profile With 20–30 of the most important compounds, a good reconstitution of a food flavour can in general be made (Grosch, 2001). The traditional method of flavour creation consists of mixing the selected aroma compounds in the proportions found during the analysis and tasting the resulting compositions in a simple test medium. This process of mixing and tasting has to be repeated several times since the first creations are often far from satisfactory owing to inaccuracies in the supplied quantitative data and impurities in the available flavour ingredients. Since this process is very timeconsuming, flavour companies have looked for methods that can speed up the flavour creation process. An interesting new tool for flavour creation is the so-called virtual aroma synthesiser (VAS) as is being used by Givaudan. The VAS allows flavour creation in real time by mixing the vapours of the 20 most important flavour ingredients. Flavourists are able to control the flow rate of each ingredient, thus allowing them to instantly make and evaluate the changes in the aroma profile. After the desired aroma profile has been obtained, the composition of the aroma in the vapour phase is calculated from the volatility of the aroma compounds in their solutions (air–solution partition coefficients) and the flow rates passing through these solutions (Espinosa Díaz et al., 2003). The software of the VAS finally calculates the composition that is required for a flavouring to generate the same aroma profile during the consumption of a product. If a food company is not satisfied with the existing flavourings or is looking for something new, a product developer from the food company can also be invited to create the desired aroma profile on the VAS. So, the VAS allows a flavour company to communicate with a food company in odours rather than in words and this can save a lot of time in achieving the desired result. If the desired aroma profile has finally been created, it might turn out that the resulting flavouring is not stable. Further work is then necessary.

Selecting the right flavourings for a food product

14.3

247

Stability of the flavouring

A flavouring should have a shelf-life of at least a few months. This is not always easy to achieve. Aroma extracts that deteriorate rapidly after their isolation from the natural source are difficult to duplicate in a stable form. The chance of solving flavour stability problems varies with the physical state of the flavouring (liquid or solid).

14.3.1 Liquid flavourings In liquid flavourings, aroma compounds can react not only with other aroma compounds but also with the solvent. Use of a flavour solvent is often necessary to obtain the desired dilution for easy and accurate dosage to a product. Selection criteria for a flavour solvent are its ability to dissolve the flavour compounds and its solubility in the application medium. It is the solvent that determines whether a flavouring is water or oil soluble. A further demand is that the flavour solvent does not adversely affect the product properties. For example, flavourings for hard candies should not contain propylene glycol because it works as a plasticiser and makes the candies sticky. Flavour solvents can affect the stability of flavourings in different ways. Well known are the reactions of aldehydes and acids with ethanol and propylene glycol to form acetals and esters. This is not always a problem because these reactions are reversible. In particular, acetals are rapidly hydrolysed in aqueous acidic media to regenerate the original aldehydes. In this case, acetal formation can be an advantage because it protects the aldehydes against oxidation and polymerisation in the flavouring and sometimes also in the application medium (Alka Sharma et al., 1998). In low-moisture and non-acidic products the formed acetals and esters remain intact. The impact of the flavouring in the product is then a function of its age because the reactions proceed with time. The chance of acetal and ester formation can be reduced by adding water to the flavouring, just enough to avoid phase separation or precipitation. However, if these reactions continue to be a problem, one has to switch to a different solvent or to the dry flavour form.

14.3.2 Dry flavourings Most dry flavourings are produced by spray drying or extrusion and consist of droplets of flavour oil dispersed in a glassy carbohydrate or other hydrophilic matrix (Porzio, 2004). In contrast with hydrophobic matrices, hydrophilic matrices are excellent barriers for aroma compounds, oxygen and other hydrophobic compounds (Benczédi, 1999; Blake et al., 2003; Soottitantawat et al., 2004). The barrier properties of hydrophilic materials decrease with the hydrophilicity of the aroma compounds.

248

Modifying flavour in food

Therefore, it is difficult to effectively immobilise such hydrophilic aroma compounds as acetaldehyde and ethanol (Dronen and Reineccius, 2003). For effective immobilisation of the aroma molecules in a glassy carbohydrate matrix, it is important to keep the concentrations of small polar molecules as low as possible. Polar compounds such as water, propylene glycol, acetaldehyde and ethanol work as plasticisers of the glassy hydrophilic matrix, thus increasing the mobility of the entrapped aroma compounds (Soottitantawat et al., 2004; Bohn et al., 2005). In certain cases, unstable liquid flavourings can be stabilised by encapsulating the reactive molecules in separate particles. Entrapment in separate compartments seems also to be the mechanism by which unstable aroma compounds are stabilised in natural products. For example, 2-furfurylthiol in coffee beans remains stable over extended periods of time but disappears rapidly after its release during brewing by reaction with the simultaneously released melanoidins (Hofmann and Schieberle, 2002). So, to prepare stable coffee flavourings containing 2-furfurylthiol and coffee extract, these ingredients need to be encapsulated in separate particles (Kindel et al., 2003). Encapsulation in separate particles does not prevent polymerisation reactions between identical molecules (vinylpyrazines, vinylphenols, aldehydes and polyunsaturated hydrocarbons such as myrcene and ocimene). From a technical point of view, molecular inclusion in cyclodextrin or amylose would be the best solution here because it effectively keeps the molecules separated. Unfortunately, the use of cyclodextrins is restricted because of concerns about their safety. Moreover, flavour loads are low and encapsulation costs are high. Dry flavourings also have disadvantages. During spray drying or extrusion, high losses of very volatile compounds (acetaldehyde, dimethyl sulphide) may occur. Production costs are also higher and there is a higher chance of oxidation. To minimise the risk of oxidation, it is important that there is no flavour oil on the surface of the particles. In general, flavour companies have sufficient knowledge in-house to create flavourings with a shelf-life of at least 6 months. This knowledge can be easily made available to the flavourists by incorporation in the flavour creation programmes run on their computers. When a flavourist has succeeded in creating a stable flavouring that delivers the desired flavour profile in a test medium, the next problem that she or he will often have to solve is how to generate the same aroma profile in a wider variety of application media. This means that the flavourist has to solve the following flavour application problems: • •

Flavour distortion due to the effect of the product matrix on flavour release during consumption. Flavour instability in the product environment.

Selecting the right flavourings for a food product

249

Table 14.2 Factors by which the aroma concentrations in a creamy yoghurt containing 4.1% (w/w) fat have to be increased for the same aroma impact as in a soft drink Correction for same equilibrium headspace (25 ºC) Cis-3-hexenol Dimethyl sulphide Ethyl butanoate Linalool Gamma-decalactone Hexyl acetate Methyl cinnamate Linalyl acetate

Correction factors for same aroma impact Orthonasal (20 ºC)*

Retronasal

1.3 2.9 3.1 6.9 11 6.6 20 30

1.2 1.6 2.1 4.4 5.2 5.3 6.6 7.5

1.1 1.4 2.8 8.2 12 28 28 190

* Correction for same intensity after opening of the package and stirring of the contents.

14.4

Solving flavour release problems

14.4.1 Effect of the product matrix on the flavour release A flavouring applied to different products will usually generate totally different sensations because of the effect of the product matrix on the flavour release. Not only the flavour strength is affected, but also the flavour profile and the longevity of the flavour (de Roos and Wolswinkel, 1994). To deliver the same flavour profile in different products, the flavouring needs to be reformulated. Table 14.2 shows an example of the corrections required to obtain the same aroma impact by sniff (orthonasal aroma) and by mouth (retronasal aroma) from two products of different composition. The differences in flavour release during sniffing and consumption often make it impossible to adjust a flavour composition to achieve both the same orthonasal and retronasal impact. Because the retronasal aroma is more important than the orthonasal aroma, the standard procedure is to reformulate for same aroma perception by mouth and accept the unavoidable differences in the aroma perception by sniff. Since food companies do not wish to wait a long time for a new flavouring, the flavour companies must be able to do the flavour reformulation and optimisation in a short period of time. Owing to the complexity of the flavour–matrix interactions, this is only possible if the flavour release is predictable. The predictive models used for this purpose will be briefly discussed in the next section.

14.4.2 Prediction of flavour release The factors that have to be taken into account when predicting the flavour release are the volatility of the aroma compounds in the product (the

250

Modifying flavour in food

thermodynamic factor) and the rate of mass transport (the kinetic factor). The volatility of an aroma compound in a product is most conveniently expressed by its air–product partition coefficient Pap, which is defined as follows (de Roos, 2000): Pap =

Ca Cp

[14.2]

where Ca and Cp are the aroma compound concentrations at equilibrium in the air and product, respectively. The air–product partition coefficients vary with the nature of the aroma compound and the composition of the product. They can be predicted with sufficient accuracy if the following information is available (de Roos, 2006a): •

Flavour volatility in the aqueous phase, as determined by: • the air–water partition coefficient Paw; • the effect of dissolved molecules on the phase partitioning (Conner et al., 1998; Nahon et al., 2000; Tsachaki et al., 2005); • the effect of flavour binding/complex formation (King and Solms, 1982; Andriot et al., 2000; Kant et al., 2003); • acid–base equilibria (de Roos, 2006a). • Flavour volatility in the lipid phase as determined by the air–triglyceride and air–phospholipid partition coefficients (de Roos, 2006b). • Volumes of the crystalline fractions (de Roos, 2006b). • Temperature (de Roos, 2006a). As soon as the phase equilibria are disturbed, the rate of mass transport starts to play a role in determining the headspace concentrations. When air flowing over the surface of a product dilutes the headspace, mass transport takes place from product to air in an attempt to restore the disturbed phase equilibria. The resulting concentration gradient developed at the surface of the product provides here the driving force for mass transport from the bulk phase to the surface of the product (see Fig. 14.1). The mass flux Jp to the surface of the product is then given by (de Roos, 2000, 2006a): Jp = kp(C pi − Cp)

[14.3] i

where kp is the mass transport coefficient and C p is the aroma compound concentration in the product at the air–product interface. The higher the headspace dilution, the lower the aroma compound concentration C pi at the product surface (C pi = Ca/Pap). At very high headspace dilution (Ca → 0), the flavour depletion at the surface of the product will be nearing completion (C pi → 0) and the mass flux will then achieve its maximum given by: Jp = −kpCp

[14.4]

Under these conditions, the release of the volatile compound has become completely mass transfer (kp) controlled and is then independent of com-

Selecting the right flavourings for a food product Volatile compound – equilibrium Air

o o

o

o

o o

o

o o

Product o o

o

o

o o

o

o o

o

o

o o

o

o o

o

o

o

o

o

o

o o

o o

o

o

Cpee

o

o

o o

o

Poorly volatile compound – equilibrium

o

Product

o

o

o

o o

o o

o o

o

o

Low

o

o o

o

o

o

o o o

o

o o

o o

o

o

o o o

o o

o o

Concentration

o

o o o

o

o o

o o

Cpe e

o

Ca

o

o

o

o

o

i

Cppi

o

o o

o

o o o

o o

o

o

o

o

o o

Cp

o

o

o

o

o o

Air

o

ii

Caa

o o

o

Product

o

High

o o o

o

o o

Low

o

o o

o o

o o

o

o

o o

o o

o

o o

Ppa wa

o

o o

o o

o

Ppa wa o o

o o

Product o

o

o o o o o o o o

o

Poorly volatile compound – non-equilibrium

o

Cae o

o

o

o

o

Air

o

o

o

o

Ppa

i Caai

o

o

Air o

o

o

o o

o

o

o

o

o

o

Ppa

o

o

Ca

o

o

Cae

o

Volatile compound – non-equilibrium o

o

251

o

o o

o

o o

o o o

i

Cppi

o

o

o o

Concentration

o o o

o

o o

o

o o o

Cp

o o o o o

High

Fig. 14.1 Aroma compound concentrations released into headspace under equilibrium and non-equilibrium conditions. Under non-equilibrium conditions, the sharpest decrease of the surface concentrations C pi is observed for the most volatile compounds. Ppa = Cp/Ca.

pound volatility. The rate at which an aroma molecule can diffuse from the bulk phase to the air–product interface then determines the release rate. Higher air flow rates over the product surface under these circumstances do not result in higher release rates. The headspace concentrations just become more diluted and the aroma intensity decreases. Higher aroma intensity can then be achieved only if the mass transport in the product phase is increased. The value of the mass transport coefficient k is a function of the diffusion mechanism. In stagnant systems, mass transport can take place only by molecular diffusion (also known as static diffusion), which is caused by the random movement of the molecules. In a monophasic system, the rate of molecular diffusion varies only slightly with flavour type and size (de Roos, 2006a). The other diffusion mechanism that may apply is the eddy or convective diffusion, which transports elements, or eddies, of the fluid from one location to another, carrying with them the dissolved solutes. The rate of eddy diffusion is completely independent of the nature of the flavour compounds. The overall result is that, under strongly diffusion-controlled conditions, the release rates from monophasic systems tend to become very similar for all aroma compounds (Fig. 14.1). The chance of having a uniform diffusion-controlled release is highest in systems in which the aroma volatility is high and that is, in general, in aqueous systems at elevated temperatures. For example, in hot beverages it can easily happen that the release of the most volatile compounds becomes already limited at relatively low temperatures. The consequence is that their

252

Modifying flavour in food

intensity hardly increases with temperature in contrast to the intensity of the less volatile compounds. Therefore, the contribution of very volatile compounds to the orthonasal aroma of beverages decreases with the increase in temperature (de Roos, 2006c). The release can also become strongly diffusion-controlled during consumption. This happens, for example, with terpenes in aqueous solutions (Linforth et al., 2002; Espinosa Díaz, 2004). It is the high flow air flow rate in the mouth/throat that is responsible for the high depletion of terpenes at the surface of the liquid (de Roos, 2006c). With beverages, the released concentrations during consumption are therefore usually lower than before consumption. In solid products, both the eddy and static diffusion are strongly hindered, which has a negative effect on the orthonasal impact (de Roos, 2003; Pozo-Bayon et al., 2004). The strongest odour intensity is perceived immediately after opening of the packaging, when the headspace concentrations are still close to those at equilibrium. Since the smell of a product also affects consumer acceptance, it is important that a flavouring does provide not only a nice ‘taste’ but also a nice smell. Flavourings for thermally processed products should even meet three demands: a nice aroma during heating, after heating and during consumption. When predicting flavour release, it is often assumed that the resistance to mass transport in the product controls the flavour release and that the resistance in the air can be ignored. This, however, is not always true (Marin et al., 1999, 2000). When the mass transfer coefficients in the air and product are comparable, i.e. differ by a factor of less than 10, the mass transport in both phases have to be taken into account by using the overall mass transfer coefficient ko given by: 1 1 Pap = + ko ka kp

[14.5]

For the prediction of the flavour release, it is a favourable circumstance that the temperature of a product has only a marginal effect on the release during consumption. This holds even for beverages, which are quickly swallowed (Boelrijk et al., 2003). The effect of thickening agents on the retronasal aroma release is often also negligible (Darling et al., 1986; Malone et al., 2000; Bylaite et al., 2003; Paçi Kora et al., 2004). So, the effects of temperature and thickening agents may often be ignored when reformulating flavourings to deliver the desired flavour experience in a different product. 14.4.3 Flavour reformulation for optimum release Computer simulation of the flavour release has allowed flavour companies to respond quickly when a flavouring has to be reformulated to deliver the same aroma profile in a different product. Computer programs help to solve the following kinds of flavour release problems:

Selecting the right flavourings for a food product •

•

• • • •

253

Rebalancing of flavour compositions to deliver the desired orthonasal or retronasal impact. Since ‘taste’ is more important than smell, the standard procedure is to reformulate for same retronasal impact (de Roos and Wolswinkel, 1994). Creation of flavourings with relatively high odour impact making use of the difference between the orthonasal and retronasal release (Espinosa Díaz, 2004). Calculation of the odour activity values for ortho- and retronasal aroma perception (odour thresholds should be known). Reduction of the flavouring costs by replacing compounds with a low release by more effective alternatives. Calculation of the flavour longevity during consumption. This is of particular importance for chewing gum (de Roos, 2003, 2006d). Calculation of the persistence of aroma compounds after consumption, which increases with decreasing volatility in the product medium (Doyen et al., 2001; Malone et al., 2003). Differences in persistence affect the aftertaste (Hodgson et al., 2004), which is one of the factors determining the consumer preference for toothpaste and full-fat products.

14.5

Solving flavour stability problems in products

The stability of flavourings in products does not only depend on the nature of the flavour compounds but also on the product composition, the packaging and the process and storage conditions. For flavour houses, this means that they should be able to predict flavour stability as a function of these variables in order to be able to respond quickly to customer requests for stable flavourings in new products. When a flavour stability problem has to be solved, it is important to realise that there are two fundamentally different causes of flavour instability: •

Chemical instability. In this case, the aroma compound concentrations decrease through decomposition in the product. • Physical instability. In this case, the aroma compounds remain intact but disappear from the product or product ingredient by volatilisation, migration, permeation or other physical processes. The solutions to these different categories of stability problems are often totally different and will therefore be discussed in separate sections. 14.5.1 Solving problems of chemical instability It is important to be aware that a high chemical stability is not always a guarantee for high flavour stability. If the decomposition products have a very strong smell, a flavour compound will already be perceived as unstable if only a very small percentage has been converted.

254

Modifying flavour in food a

Flavouring remaining (%)

100

b, mild conditions

80

60 b, less harsh conditions 40

20

b, harsh conditions

0 0

1

2

3

4

5

6

Time (weeks)

Fig. 14.2 Effect of mild and harsh conditions on the perceived stability of the unstable flavouring b. Less harsh conditions have a negative effect on the perceived stability after 1 week (b, less harsh conditions), unless the conditions are really mild (b, mild conditions). (Curve a represents stable flavour.)

On the other hand, an unstable flavouring may be perceived as stable if it is undergoes rapid conversion into a more stable composition before the product is sold (Fig. 14.2, curve b, harsh conditions). In that case, improvement of the chemical stability, for example, by storage at lower temperature, might have an adverse effect on the perceived stability (Fig. 14.2, curve b, less harsh conditions). Depending on the kind of the flavour instability, different flavouring types or concepts can be used to solve the problem. In this section, the following flavour systems will be discussed in more detail: • liquid flavourings; • encapsulated flavourings; • flavour precursor systems. Liquid flavourings (a) What flavour companies can do to improve flavour stability To create flavourings that are stable in the target application, flavour companies should be able to predict the stability of the aroma compounds as a function of the product composition, the process conditions and the storage conditions. To allow prediction of the chemical stability, the reaction mechanisms should be known. The most frequently occurring reactions are the first-order and second-order reactions of which the reaction rates are given by:

Selecting the right flavourings for a food product

255

−d [ A ] = k1 [ A ] dt

[14.6]

−d [ A ] = k2 [ A ][ B] dt

[14.7]

where k1 and k2 are the first-order and second-order rate constants and [A] and [B] are the concentrations of the flavour chemicals A and B in the hydrophilic or lipid phase of the product. In foods, it often happens that the second-order reactions take place with an excess of the food ingredient. In that case, the concentration of ingredient B remains constant during the reaction and equation 14.7 is reduced to: −d [ A ] = k2′ [ A ] dt

[14.8]

where k2′ = k2[B]. The reaction is then said to be pseudo first-order. The progress of (pseudo) first-order reactions is easy to predict because integration of equations 14.6 and 14.8 results in a simple linear relationship between log(fraction retained) and time: Log

[A] = −kt [ A0 ]

[14.9]

Since most reactions of aroma compounds in foods are (pseudo) first-order, we will restrict the discussion to this type of reaction. The (pseudo) firstorder rate constants are dependent only on the temperature and the catalyst concentrations. The most reactive food ingredients and strongest catalysts are found in the hydrophilic (aqueous) phase. Among them, the following are most relevant in connection with flavour stability: • Water: involved in hydrolysis and hydration reactions. • Proteins: disulphide bridges react with thiols and disulphides (Parker et al., 2003). • Proteins: amino groups react with carbonyl compounds (Le Guen and Vreeker, 2003). • Sulphite and cysteine: form complexes with aldehydes (Kallen, 1971; Esterbauer et al., 1976; Dufour et al., 2000). • Microorganisms: reduce aldehydes to alcohols. • Acids: catalyse several reactions including reactions with water (reacts with terpenes, esters, lactones). • Enzymes: catalyse oxidation reactions (Anklam et al., 1997; Gassenmeier, 2003), hydrolysis reactions (Drawert et al., 1965), etc. • Metal ions: catalyse oxidation reactions. Aroma compounds in the lipid phase are protected against the threats in the hydrophilic phase. Therefore, the flavour stability normally increases

256

Modifying flavour in food

with the fat content of the product and the lipophilicity of the aroma compounds (de Roos, 1997). The concentrations in the hydrophilic phase can be calculated if the lipid–water partition coefficients of the aroma compounds and their binding constants with the food ingredients are known (de Roos 2006a). For hydrophilic aroma compounds, a reverse relationship exists between stability and fat content because the aqueous concentrations of these compounds increase with fat content. Absorption by plastic coatings can also provide protection for aroma compounds against decomposition in aqueous solution (Reynier et al., 2004). This explains why the hydrophobic aroma compounds in a beverage are more stable in laminated aluminium packaging than in glass. When the aqueous aroma concentrations in the beverage decrease by decomposition, the absorbed aroma compounds are released again into the beverage, which keeps the aqueous concentrations at an acceptable level over a longer period of time. In view of what has been said above, it is not surprising that the highest chance of flavour instability is in no- and low-fat products. In particular, soft drinks and juices have a bad reputation. Owing to the low pH of these beverages, the conditions are very harsh for the aroma compounds. Among the most extensively studied flavour stability problems in beverages are those with citrus flavourings. Therefore, stability problems with citrus flavourings will serve here as examples to illustrate the kind of flavour stability problems that flavour and food companies have to deal with. Among the major reactions of citrus flavourings in beverages are the acid-catalysed hydrolysis and hydration reactions (Clark and Chamblee, 1992). Hydration reactions result in the addition of water to double bonds of terpenes to yield terpene alcohols. These reactions can be very fast; the bicyclic terpenes (pinene, thujene, sabinene) are converted almost completely to alcohols within a few weeks (Fig. 14.3). The aldehyde concentrations undergo also rapid changes, showing either a decrease to negligible levels (citronellal, neral and geranial) or an increase to a constant higher level (decanal). The rapid increase of the decanal concentration indicates that it was initially present in acetal form (probably an acetal with terpene alcohols). Because of these rapid changes, it makes sense to postpone transport of soft drinks to supermarkets until the flavour composition has stabilised. The rates of the (pseudo) first-order hydrolysis and hydration reactions in citrus beverages are linearly related to the hydrogen ion concentrations (Clark and Chamblee, 1992) and therefore, are easy to predict. The effect of temperature can also easily be accounted for. Since hydrolysis and hydration reactions require the presence of water, these reactions do not take place in an anhydrous environment, even if these products contain high acid concentrations. This is illustrated in Table 14.3 where decomposition of flavour compounds in pastilles is either negligible or much slower than in the aqueous, soft drink system. Whereas all

Selecting the right flavourings for a food product pH 3.5

Concentration (g/l)

100

10

1

1

0.1

0.1

0.01 0

1

2

Concentration (g/l)

100

3

4

pH 3.5

100

10

5

6

7

0.01

0

1

2

100

pH 2.7

10

10

1

1

0.1

0.1

0.01

257

3

4

5

6

7

3 4 5 Time (days)

6

7

pH 2.7

0.01 0

1

Limonene α-Thujene Fenchol

2

3 4 5 Time (days) β-Pinene Sabinene Borneol

6

7

α-Pinene α-Terpineol

0

1

2

Geranial Decanal

Neral Citronellal

Fig. 14.3 Effect of pH on the stability of lime oil fractions in a soft drink base at 30 ºC. The fractions were obtained by preparative liquid chromatography on silica gel. The terpene fraction was free from alcohols and other oxygen compounds. (Source: Stijnman and de Roos, unpublished results.)

Table 14.3 Effects of citric acid on lemon flavour stability in pastilles and soft drinks (Stijnman and de Roos, unpublished results) Change during storage at room temperature (%/day)*

Neral Geranial α-Pinene β-Pinene Fenchol Borneol α-Terpineol

Pastilles

Soft drink

−1.9 −1.3 0.0 0.0 0.0 0.0 0.0

−16 −14 −7 −24 ++ ++ ++

* A change of 1.9% per day means that the fraction retained after n days is (1 − 0.019)n.

258

Modifying flavour in food

acid-catalysed reactions of terpenes to alcohols require water, the aldehydes neral and geranial can also be concerted along pathways that do not require the involvement of water (Clark and Chamblee, 1992). (b) What food companies can do to improve flavour stability It is not only the flavour company that can do something to improve the flavour stability, the food or beverage company can contribute as well. The standard approach to solve a flavour stability problem caused by low pH is to ask the flavour company for a more acid-stable flavouring. With citrus flavourings, this is possible by removing the terpenes from the essential oils (to yield terpeneless citrus oils) or by converting the unstable terpenes to the more stable alcohols (as with distilled lime oil). The beverage company, on the other hand, can improve the flavour stability by creating less harsh conditions in the product, for example, by increasing the pH of the beverage. Since sourness perception is a function of both the free and bound hydrogen ion concentrations (Ganzevles and Kroeze, 1987), it is possible to improve the flavour stability without adversely affecting the perceived sourness by increasing the pH by use of buffers. Also in the next example, where it is the non-volatile constituents of a flavouring that are causing the flavour stability problem, it is the beverage company that can offer the better solution. In this case, the problem was that a citrus beverage in one-way (non-reusable) poly(ethylene terephthalate) (PET) bottle developed a green-metallic off-flavour, when exposed to light. The origin of the off-flavour could be traced back to the oxidation of the unsaturated waxes in the cold-pressed citrus oil that was used as a flavour ingredient. This was concluded from the identification of tr-2-cis-6undecadienal and tr-2-cis-6-dodecadienal as the compounds responsible for these off-notes and from the observation that these off-notes were only developed in containers that were permeable to oxygen. To solve this problem, two approaches can be followed: either the flavour company can eliminate the waxes from the flavouring or the beverage company can use containers that are less oxygen permeable. The first solution solves only part of the problem since there are other oxidation reactions that can take place with citrus oils (Ueno et al., 2004). So, a change of packaging by the beverage company is then the better solution. Although citral is the citrus flavour compound that is most sensitive to light (Iwanami et al., 1997), its role in the above example is negligible because its acid-catalysed degradation is much faster than its light-induced conversion to photocitral. Moreover, at the low concentrations formed, photocitral does not significantly affect the citrus flavour profile. There are more examples where beverage companies can do more than flavour companies to improve the flavour stability. This certainly holds if the product base or packaging is the real cause of the stability problem rather than the added flavouring. Since in such cases, the added flavouring is usually blamed for the stability problem, a flavour house should be able

Selecting the right flavourings for a food product

259

to recognise the problem and to help the food producer to prevent the problem (Anklam et al., 1997; Gassenmeier, 2003). The next example of stability problems with citrus beverages shows how unsuitable packaging can induce a sequence of reactions in citrus juices, resulting in discoloration and off-flavour formation. To prevent oxidation of aroma compounds in citrus juices, ascorbic acid (vitamin C) is usually added as an oxygen scavenger. However, the reaction of ascorbic acid with oxygen yields dehydroascorbic acid and this is an effective catalyst for Maillard reactions that are taking place between the amino acids and sugars in the juices. The Maillard reactions result not only in the formation of strong smelling volatile compounds such as methional and 2-methylfuran-3-thiol (Bezman et al., 2001), but also in brown pigments. The higher the permeability of the containers to oxygen, the stronger is the off-flavour development. The permeability of containers to oxygen is also the cause of the oxidation of other indigenous citrus juice components such as p-vinylguiacol, which is oxidised to vanillin (Wolswinkel and de Roos, unpublished results). Since the stability of the added citrus flavourings is not significantly affected by the permeability of the container, a flavour company cannot do much more than advise the beverage company to switch to less permeable containers for the juices. The examples given above show that the creation of stable flavourings requires a thorough knowledge of the effects of the product matrix, packaging and the process and storage conditions on the flavour stability. Close cooperation of the flavour company with food companies is often required to obtain all the necessary information required for the creation of a stable flavouring. An additional advantage for the food companies is that they can profit from knowledge of flavour companies about flavour stability problems caused by the product matrix and improper packaging. Encapsulated flavourings In dry products, flavourings can be stabilised by immobilisation of the reactive compounds in separate particles. Therefore, instant product often provide the better environment for unstable flavours such as those of fresh fruits or freshly brewed coffee (Kindel et al., 2003). For effective immobilisation, encapsulation in glassy, water-soluble carbohydrate matrices is the method of choice. In contrast to hydrophobic matrices, hydrophilic matrices are excellent barriers for the hydrophobic aroma compounds. The stability problems in dry products differ widely from those in aqueous solutions. Hydration and hydrolysis reactions cannot take place (cf. Table 14.3), whereas the reverse reactions (dehydration and condensation) occur more easily. In dry products there is also a higher risk of oxidation owing to the relatively high volume of the gas phase within the package. The high stability of dry flavourings in instant products is sometimes also a disadvantage. In citrus beverage powders, for example, the conversion of terpenes to terpene alcohols does not take place. As a consequence, the

260

Modifying flavour in food

resulting beverage has the typical aroma profile of citrus peel oil. Moreover, an earthy off-flavour may rapidly develop in the beverage through conversion of the bicyclic terpenes (pinene, sabinene, thujene) to borneol and fenchol. The latter problem can easily be avoided by using so-called folded citrus oils from which the volatile bicyclic terpenes have been removed by distillation. In the presence of moisture, the barrier properties of glassy hydrophilic matrices decrease sharply and effective separation of reactive compounds is no longer possible (Blake et al., 2003; Soottitantawat et al., 2004; Bohn et al., 2005). For that reason, it is not possible to improve the flavour stability in humid environments over a longer period of time by encapsulation in food-grade materials. The solubility of hydrophilic capsules in water is an advantage during consumption because it allows a quick and complete release of the entrapped aroma molecules. It is also an advantage in those instant products, such as beverage powders, where a immediate release of the encapsulated compounds is required. Flavour precursor systems Use of flavour precursors is worth considering if an unstable aroma has to be imparted to a product. To fully enjoy an unstable aroma, the aroma should be generated just prior to consumption. The trigger for aroma release can be water, heat or any other stimulus that can induce a reaction. For example, water has been used as a trigger for the release of acetaldehyde from its precursor upon dissolution of a beverage powder (DeSimone and Byrne, 1986; Gassenmeier et al., 2004). For heat-induced aromas, the following two approaches can be used: •

Flavour generation in situ. For this purpose, the key intermediates in the chain of reactions from primary precursors to aroma compounds are preferred if costs and stability permit. Examples of such flavour precursor systems are the pre-ferments used in bread (Gassenmeier and Schieberle, 1995) and the Amadori compounds claimed for use in bakery (Doornbos and van den Ouweland, 1977) and meat products (de Roos et al., 2005). For products heated to dryness, such as biscuits and crackers, in situ flavour generation has the additional advantage of stabilising the generated compounds through in situ immobilisation in the dry hydrophilic starch phase (Grab and Gfeller 2000; de Roos, 2003). The applicability of these precursors is limited because of the fixed set of conditions that is required for optimum flavour development. • Flavour generation off-line. This has become the most popular approach for generating the delicate flavours of freshly baked bread, boiled and roasted meats and other thermally processed foods. This can be concluded from the hundreds of patents that are issued per year on process

Selecting the right flavourings for a food product

261

and reaction flavours (Sucan and Weerasinghe, 2005). The process flavour approach has the advantage of allowing optimisation for high flavour yield. The aroma compounds are stabilised by immobilisation in the flavouring during drying (Blank, 2004). Moreover, unstable flavour compounds lost during storage and heating of the product can often easily be replenished from the residual precursor activity still present in these flavourings. The complexity of the reactions taking place during thermal processing makes it difficult and time-consuming to generate the right flavour profile. To overcome this problem, Patel et al. (2001) have developed a method of simulating the complex chemical reactions of the Maillard process in sufficient detail to allow prediction of the composition of the generated volatile flavour. 14.5.2 Solving problems of physical stability Although loss by volatilisation is the major cause of the poor physical stability of flavourings in products, sometimes the migration of aroma compounds can also cause problems. The risk of migration is highest in an aqueous environment because most aroma compounds are hydrophobic and have a strong tendency to be ad- or absorbed by the hydrophobic food ingredients. For example, fruit aromas in fruit yoghurts are not present in the pieces of fruit but in the fat phase of the yoghurt. In this case, the consumer does not seem to consider this as a serious problem, but in other cases it is often not acceptable. If migration is indeed a serious problem, physical separation of the different foods or food ingredients is the only possible solution. Flavour–packaging interactions can also be an important cause of the physical instability of added flavourings. In plastic containers, the potential for flavour loss by absorption or permeation is highest with hydrophobic aroma compounds in aqueous solutions (Gremli, 1996; Risch, 2000). The problems can be avoided by using glass or reusable PET bottles and/or by using less hydrophobic (‘PET-stable’) aroma compounds. Migration of odour compounds from package to product may also occur. In that case, the chance of migration increases with the oil/fat content of the product. Only use of a different container can solve this problem. By far the major cause of the physical instability of flavourings is loss of volatile compounds by volatilisation. The highest risk of aroma loss exists during thermal processing in open systems such as baking, frying and boiling (de Roos and Graf, 1995; Heiderich and Reineccius, 2001; de Roos, 2006c). Improper storage is another cause of flavour loss. The following approaches have been used to solve these stability problems: • Use of poorly volatile liquid flavourings. • Flavour encapsulation. • In situ flavour generation from precursor systems.

262

Modifying flavour in food

Poorly volatile liquid flavourings The risk of flavour loss can be reduced by using aroma compounds with a low volatility in the product base. However, a problem is that a low volatility also results in a low release during consumption, which may neutralise the positive effect of high retention. For example, in spite of the higher retention of aroma compounds in high-fat than in low-fat biscuits (due to lower volatility in high-fat medium), the aroma concentration, released from the high-fat biscuits during consumption were found to be lower than those from the low-fat biscuits (Brauss et al., 1999). The following factors have been found to control the aroma retention in thermally processed foods (de Roos and Graf, 1995; de Roos and Mansencal, 2003; Dimelow et al., 2005): •

The volatility of the aroma compounds in the product environment as determined by their air–product partition coefficients of the aroma compounds. • The mass transfer as influenced by size, texture and composition of the product. • The percentage of water loss. • The heating method. Relatively simple relationships exist between the flavour retention during baking, on the one hand, and the water loss and air–product partition coefficients of the aroma compounds, on the other (de Roos and Graf, 1995; de Roos and Mansencal, 2003). Only at low moisture contents can deviations occur, owing to immobilisation of the hydrophilic aroma compounds in the drying starch phase, which prevents them from being released according to their volatility. When the product is completely dry, flavour loss can take place only from the lipid phase, which is a poor barrier for aroma compounds (de Roos, 2003). The heating method can also influence the flavour retention. For example, the differences in aroma loss during convection baking of cakes are less pronounced than those during microwave baking (Fig. 14.4). In general, water loss is higher during microwave heating and this affects also the flavour loss. When creating ‘microwave-stable flavourings’, these factors have to be taken into account. Flavour loss is not always the only reason why flavourings have a relatively low impact in thermally processed foods. Poor release during consumption can be another reason. For example, to obtain the same retronasal impact of linalool and ethyl octanoate from biscuits as from soft drinks, a dose one hundred times higher is necessary, the major correction being required to compensate for the differences in the flavour release (de Roos and Mansencal, 2003). The poor release from biscuits is due to the low volatility of hydrophobic compounds in low-moisture, fat-containing products. The volatility of hydrophilic flavour compounds is much less affected by fat and that is the reason why compounds such as vanillin have a relatively high impact in bakery products (de Roos and Mansencal, 2003). In products

Selecting the right flavourings for a food product

263

100 1. 2,3-Dimethyl-pyrazine 2. Naphthalene 3. p-Cresol 4. α-Ionone 5. Indole 6. δ-2-Deceno-lactone 7. Raspberry ketone

Flavour loss (%)

80

60

Microwave 3 2

1 Convection

5 4

40 6 20 7 0 1

10

100 Volatility (Pap x

106

1000

10 000

at 100 °C)

Fig. 14.4 Effect of heating method on flavour retention during baking of cakes. Water losses during microwave and convection heating were 85% and 73%, respectively. (Source: de Roos and Graf, 1995.)

that are boiled the use of hydrophilic aroma compounds should be avoided because losses here not only increase with the volatility of the aroma compounds in water but also with their solubility in it. Flavour losses from boiling liquids, such as soups, are much higher than from solid products. The reason is that the high turbulence in the boiling liquid results in a very effective mass transport to the vapour phase. However, if the hot liquid is stagnant, the release of the most volatile aroma compounds can become completely diffusion controlled and their release will no longer increase with temperature as is the case for the less volatile compounds. In particular, in hot beverages there is a high chance of a completely diffusion-controlled release due to the high volatility of aroma compounds in aqueous solution (de Roos, 2006c). The positive effect is that, at high temperatures, the retention of these volatile compounds is relatively high. However, the negative effect is that their contribution to the total flavour impact is low. Encapsulated flavourings For the successful application of encapsulated flavourings in heated products, the following conditions should be fulfilled: • •

The encapsulation should increase the aroma retention during heating without adversely affecting the aroma release during eating. There is a major flavour retention or release problem. If not, no major improvement may be achieved that justifies the additional costs of the encapsulation.

264

Modifying flavour in food

The first condition can most easily be fulfilled in anhydrous application media such as chocolates, chewing gum and hard candies. In these products, the risk of aroma loss by volatilisation can be effectively reduced by encapsulating the aroma compounds in glassy hydrophilic particles (Blake et al., 2003). During consumption, these particles dissolve in the saliva and completely release the entrapped aroma compounds. This is an advantage in comparison with liquid flavourings, which are often incompletely released owing to their absorption by fat, the chewing gum base or other lipophilic food ingredients (de Roos, 2000). Flavourings encapsulated in water-soluble particles can also be used to flavour dry products afterwards. Successful examples of this approach are the dustings on savoury snacks (chips/crisps) and the coatings on cereals and bakery products. A possible disadvantage of this approach is that the release is too fast and completed long before chewing has finished. Therefore, in many cases it is preferred to incorporate the flavour before thermal processing. Since most thermally processed products require the use of water at some stage of the production process, the encapsulation systems for these products should be relatively insensitive to moisture. In that case, however, flavour release by capsule dissolution during eating is no longer possible, with the result that the advantage of higher retention can be completely nullified by lower release during consumption (Reineccius et al., 2004). Fortunately, it appears that this problem can be solved by using capsules that delay flavour loss instead of completely preventing it. Alginate beads (Bouwmeesters, and de Roos 1998) and microcapsules obtained by complex coacervation (coacervation capsules: Soper, 1995) have been found to be useful for this purpose (Fig. 14.5). With these capsules it is possible to achieve higher retention without compromising on the ortho- and retrona-

Flavour retention (%)

80 70 60 50 40 30 20 10 0 Ethyl Butyl Ethyl butanoate butanoate octanoate Liquid flavour

Fig. 14.5

Menthyl acetate

Alginate gel beads, 1 mm

Hexanol

Linalool

Borneol

Coacervation capsules, 0.25 mm

Effect of encapsulation on flavour retention in crackers.

Selecting the right flavourings for a food product

265

sal aroma impact. This demonstrates that the encapsulated aroma compounds must have already been completely released during the heating process and that it is not necessary to rupture the capsules during eating in order to release the flavour (de Roos 2006c). The benefits from encapsulation vary with the properties of the aroma compounds and the conditions during thermal processing. When the flavour losses are low, only a marginal improvement of the aroma retention is possible and the benefits from encapsulation do not always justify the cost of the encapsulation. Therefore, it is important that a flavour company can estimate the benefits from encapsulation. This requires that the retention of the encapsulated and unencapsulated flavourings can be easily compared. Computer simulation of the aroma loss from heated products has proven its value here. Flavour precursors For imparting a physically unstable flavour to a product, in situ flavour generation is worthy of consideration. Since volatile aroma compounds are easily lost during and after heating, their flavour can be fully enjoyed only if they are generated in situ and the product is consumed soon after the thermal processing (Schieberle and Grosch, 1992). The method is of particular interest for imparting the authentic flavour profiles of traditional products to their cheaper or healthier alternatives (low-fat, low-carbohydrate, vegetarian, etc.). In the latter products, the flavour precursor composition is out of balance and needs to be adjusted to reproduce the delicate baked, fried or roasted aroma profiles of the traditional products. An additional advantage of in situ flavour generation is that relatively high concentrations of the generated volatile compounds can be built up by immobilisation in the drying starch phase of biscuits, crackers or other products that are heated to dryness (de Roos, 2006c). A further advantage is that these volatile compounds are well retained during storage and easily released from the hydrated product matrix during consumption. Rebalancing of the flavour precursor mix in modified products can easily be done if the precursor concentrations in the original and modified products are known.

14.6

Future trends

Today, the important aroma compounds of the common foods are known. Therefore, it has become more difficult for flavour companies to distinguish themselves from the competition by having unique aroma chemicals at their disposal. Most nature-identical and artificial flavourings are just commodities and the competition is here on price. With natural flavourings it is slightly easier to distinguish from the competition because these flavourings are more difficult to match.

266

Modifying flavour in food

An area where it is now easier to distinguish from the competition is the flavour performance in products. Here we see the continuation and emergence of the following trends. 14.6.1

Computerised optimisation of the aroma profile during consumption Here the following problems will (continue to) receive attention: •

Flavour profile distortion due to the differences in flavour release from products. To meet the need for quick response to customer requests, computerised compensation for the differences in the flavour release is gaining popularity (de Roos and Wolswinkel, 1994; Normand et al., 2004). Improvement of the predictive models will probably continue to receive attention. More focus on cost-effectiveness through elimination or replacement of aroma chemicals with poor release. • Flavour profile distortion due to the flavour instability in products. Encapsulation and flavour precursors will probably continue to receive attention for delivering the unstable flavours of fresh foods in products that are processed or stored under harsh conditions. Whether these methods will be able to compete with the cheaper liquid flavourings that are optimised for high stability in the target application remains questionable. The creation of cost-effective stable flavourings requires that the stability of aroma compounds can be predicted as a function of their microenvironment and the process and storage conditions. 14.6.2 Delivering the desired total flavour impact In the past, flavour and food companies have relied mainly on aroma compositions for improving flavours and ignored the effects of interactions between aroma, taste and mouthfeel. This has sometimes resulted in disappointing results because the perceived aroma is not only determined by the concentrations of volatile compounds in the nose, but also by the simultaneous taste and mouthfeel sensations (Valdés et al., 1956; Frank and Byram, 1988; Noble et al., 1993; Noble, 1996; Baek et al., 1999; Davidson, et al., 1999; Weel et al., 2002; King et al., 2003; Taylor et al., 2003; Juteau et al., 2004; Lethuaut et al., 2005; Pfeiffer et al., 2005). This is clearly demonstrated by the fact that a long-lasting aroma in chewing gum does not generate a longlasting flavour sensation if the taste compounds are not long-lasting (Duizer et al., 1996). To deliver the desired total flavour sensation, flavour houses need to understand the relationship between aroma, taste and texture. To achieve this total flavour sensation in a product, close cooperation between flavour and food companies is necessary because none of them has full control over all flavour attributes. To deliver the desired total flavour impact, it is important that flavour companies can also deliver some special taste sensations such as cooling,

Selecting the right flavourings for a food product

267

tingling, pungent and other trigeminal effects. The ability of a flavour company to help food companies in improving the sensory properties of their products will become an important competitive advantage. Flavour companies will have to shift their focus from the sensory properties of the flavouring to those of the food. This means that flavour companies should be able to solve all kinds of off-flavour problems, independent of whether they are caused by the flavouring or by the product base and the process and storage conditions. 14.6.3 Relating total flavour to consumer preference For flavour companies, it is important to understand the relationship between total flavour impact and consumer preference, in particular, the role of aroma and taste in this relationship. Because of the high practical significance of the answers to these questions, this subject may be expected to receive more attention in the future than it receives today.

14.7

Sources of further information and advice

baigrie, b., ed (2003), Taints and Off-flavours in Foods, Woodhead Publishing Ltd, Cambridge, UK. charalambous, g., ed (1992), Off-flavours in Foods and Beverages, Elsevier Science B.V., Amsterdam. marsili, r., ed (2001), Flavor, Fragrance and Odor Analysis, Marcel Dekker Inc, New York. nursten, h. e. (2005), The Maillard Reaction – Chemistry, Biology and Implications, Springer, Berlin, Germany. roberts, d. d. and taylor, a. j., eds (2000), Flavour Release, ACS Symposium Series 763, American Chemical Society, Washington, DC. rowe, d., ed (2004), Chemistry and Technology of Flavours and Fragrances, Blackwell Publishing, Oxford, UK. taylor, a. j., ed (2002), Food Flavour Technology, Sheffield Academic Press, Sheffield, UK. taylor, a. j. and roberts, d., eds (2004), Flavour Perception, Blackwell Publishing, Oxford, UK. voilley, a. and etievant, p., eds (2006), Flavour in Food, Woodhead Publishing Ltd, Cambridge, UK. weerasinghe, d. k. and sucan, m. k., eds (2005), Process and Reaction Flavours, ACS Symposium Series 905, American Chemical Society, Washington DC, USA.

14.8

References

acree, t. e. (1997), ‘GC/Olfactometry’, Anal Chem News Features, 69, 170A–175A. alka sharma nagarajan, s. and gurudutt, k. n. (1998), ‘Stabilization of aldehydes as propylene glycol acetals’, J Agric Food Chem, 46, 654–656. andriot, i., harrison, m., fournier, n. and guichard, e. (2000), ‘Interactions between methylketones and β-lactoglobulin: sensory analysis, headspace analysis and mathematical modelling’, J Agric Food Chem, 48, 4246–4251.

268

Modifying flavour in food

anklam, e., gaglione, s. and muller, a. (1997), ‘Oxidation behaviour of vanillin in dairy products’, Food Chem, 60 (1), 43–51. baek, i., linforth, r. s. t., blake, a. and taylor, a. j. (1999), ‘Sensory perception is related to the rate of change of volatile concentration in-nose during eating of model gels’, Chem Senses, 24, 155–160. benczédi, d. (1999), ‘Estimation of the free volume of starch-water barriers’, Trends Food Sci Technol, 10, 21–24. bezman, y., rouseff, r. l. and naim, m. (2001), ‘2-Methyl-3-furanthiol and methional are possible off-flavours in stored orange juice: aroma-similarity, NIF/SNIFF GCO, and GC analyses’, J Agric Food Chem, 49, 5425–5432. blake, a., cantergiani, e. and harvey, b. (2003), ‘Direct monitoring of flavour release from encapsulated flavours’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 154–158. blank, i. (2004), ‘Self-assembly structures in food flavour generation by thermal processes’, in Hofmann T, Rothe M and Schieberle P, State-of-the Art in Flavour Chemistry and Biology Garching, Germany, Deutsche Forschungsanstalt für Lebensmittelchemie, 171–181. boelrijk, a., basten, w., burgering, m., gruppen, h., voragen, f. and smit, g. (2003), ‘The effect of co-solvent on the release of key flavours in alcoholic beverages; comparing in vivo with artificial mouth-MS nose measurements’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 204–207. bohn, d. m., cadwallader, k. r. and schmidt, s. (2005), ‘Development and validation of a dynamic vapour sorption-fat gas chromatography-flame ionization detection method for rapid analysis of volatile release from glassy matrices’, J Agric Food Chem, 53, 3149–3155. bouwmeesters, j. f. g. and de roos, k. b. (1998), ‘Process for preparing beads as food additive’, Patent WO 98/15192. brauss, m. s., balders, b., linforth, r. s. t., avison, a. and taylor, a. j. (1999), ‘Fat content, baking time, hydration and temperature affect flavour release from biscuits in model-mouth and real systems’, Flavour Fragr J, 14, 351–357. budin, j. t., milo, c. and reineccius, g. a. (2001), ‘Perceivable odorants in fresh and heated sweet cream butters’, in Special Publication, Royal Society of Chemistry 274 (Food Flavours and Chemistry), Cambridge, Royal Society of Chemistry, 85–96. bylaite, e., meyer, a. s. and adler-nilsen, j. (2003), ‘Changes in macroscopic viscosity do not affect the release of aroma aldehydes from pectinaceous food model system of low sucrose content’, J Agric Food Chem, 51, 8020–8026. clark, b. c. and chamblee, t. s. (1992), ‘Acid-catalyzed reactions of citrus oils and other terpene-containing flavours’, in Charalambous G, Off-flavours in Foods and Beverages, Amsterdam, Elsevier Science B.V., 229–285. conner, j. m., birkmyre, l., paterson, a. and piggott, j. r. (1998), ‘Headspace concentration of ethyl esters at different alcoholic strengths’, J Sci Food Agric, 77, 121–126. darling, d. f., williams, d. and yendle, p. (1986), ‘Physico-chemical interactions involved in aroma transport processes from solution’, in Birch G G and Lindley M G, Interactions of Food Components, London, Elsevier Applied Science, 165–188. davidson, j. m., linforth, r. s. t., hollowood, t. a. and taylor, a. j. (1999), ‘Effect of sucrose on the perceived flavour intensity of chewing gum’, J Agric Food Chem, 47, 4336–4340. de roos, k. b. and wolswinkel, c. (1994), ‘Non-equilibrium partition model for predicting flavour release in the mouth’, in Maarse H and Van der Heij H G, Trends in Flavour Research, Amsterdam, Elsevier Science B.V., 15–32.

Selecting the right flavourings for a food product

269

de roos, k. b. (1997), ‘How lipids influence food flavour’, Food Technol, 51 (1), 60–62. de roos, k. b. (2000), ‘Physicochemical models of flavour release from foods’, in Roberts D D and Taylor A J, Flavour Release, Washington, DC, American Chemical Society, 126–141. de roos, k. b. (2003), ‘Effect of texture and microstructure on flavour retention and release’, Intern Dairy J, 13, 593–605. de roos, k. b. (2006a), ‘Modelling aroma interactions in food matrices’, in Voilley A and Etiévant P, in Flavour in Food, Cambridge, UK, Woodhead Publishing Ltd, 229–259. de roos, k. b. (2006b), ‘How lipids influence flavour perception’, in Shahidi F and Weenen H, Food Lipids. Chemistry, Flavour, and Texture, Washington, DC, American Chemical Society, ACS Symposium Series 920, 145–158. de roos, k. b. (2006c), ‘Understanding and controlling the behaviour of aroma compounds in thermally processed foods’, Trends Food Sci Technol, 17, 236–243. de roos, k. b. (2006d), ‘Flavourings for chewing gum’, in Fritz D, Formulation and Production of Chewing and Bubble Gum, London, Kennedy’s Publications Ltd, 205–238. de roos, k. b. and graf, e. (1995), ‘Non-equilibrium partition model for predicting flavour retention in microwave and convection heated foods’, J Agric Food Chem, 43, 2204–2211. de roos, k. b. and mansencal, r. (2003), ‘Poor performance of flavourings in heat processed products: A problem of poor flavour retention or poor flavour release?’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentiethfirst Century, Paris, Editions Tec & Doc, 27–32. de roos, k. b. and nelissen, j.-p. (2000), ‘The relationship between GC-olfactometric data and the perceived flavour during consumption’, in Schieberle P and Engel K-H, Frontiers of Flavour Science, Garching, Germany, Deutsche Forschungsanstalt für Lebensmittelchemie, 248–252. de roos, k. b., wolswinkel, c. and sipma, g. (2005), ‘Amadori compounds of cysteine and their role in the development of meat flavour’, in Weerasinghe D K and Sucan M K, Process and Reaction Flavours, ACS Symposium Series 905, Washington DC, American Chemical Society, 117–129. desimone, r. s. and byrne, b. (1986), ‘Aldehyde generators for flavours’, Perfumer & Flavorist, 11 (6), 15–16, 18, 20, 22, 24–26. dimelow, c. p., linforth, r. s. t. and taylor, a. j. (2005), ‘Model studies on retention of added volatiles during breadcrumb production’, J Agric Food Chem, 53, 3572–3576. doornbos, t. and van den ouweland, g. a. m. (1977), ‘Flavouring with Amadori compounds’, US patent 4022920. doyen, k., carey, m., linforth, r. s. t., marin, m. and taylor, a. j. (2001), ‘Volatile release from an emulsion: headspace and in mouth studies’, J Agric Food Chem, 49, 804–810. drawert, f., heimann, w., emberger, r. and tressel, r. (1965), ‘Enzymic modification of the natural apple flavour during the processing’, Naturwissenschaften, 52 (11), 304–305. dronen, d. m. and reineccius, g. a. (2003), ‘Mass transfer parameters of volatile compounds in dry carbohydrate matrices’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 75–78. dufour, j.-p., leus, m. and hayman, a. r. (2000), ‘Chemical equilibria of beer staling: reaction of bisulphite with unsaturated aldehydes’, in Schieberle and Engel K-H, Frontiers of Flavour Science, Garching, Germany, Deutsche Forschungsanstalt für Lebensmittelchemie, 579–583.

270

Modifying flavour in food

duizer, l. m., bloom, k. and findlay, c. j. (1996), ‘Dual-attribute time–intensity measurement of sweetness and peppermint perception of chewing gum’, J Food Sci, 61, 636–638. espinosa-díaz, m. (2004), ‘Comparison of orthonasal and retronasal flavour perception at different concentrations’, Flavour Fragr J, 19, 499–504. espinosa díaz, m., de roos, k. b. and antenucci, r. n. (2003), ‘Prediction of flavour release under dynamic conditions’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 75–78. esterbauer, h., ertl, a. and scholz, n. (1976), ‘The reaction of cysteine with α, βunsaturated aldehydes’, Tetrahedron, 32 (2), 285–289. frank, r. a. and byram, j. (1988), ‘Taste-smell interactions are taste and odorant dependent, Chem Senses, 13, 445–455. frijters, j. e. r. (1978), ‘A critical analysis of the odour unit number and its use’, Chem Senses, 3, 227–233. ganzevles, g. j. and kroeze, j. h. a. (1987), ‘The sour taste of acids. The hydrogen ion and the undissociated acid as sour agents’, Physiol Behaviour, 40 (5), 641– 646. gassenmeier, k. (2003), ‘Formation of cardboard off-notes in vanilla ice-cream’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentiethfirst Century, Paris, Editions Tec & Doc, 91–94. gassenmeier, k. and schieberle, p. (1995), ‘Potent aromatic compounds in the crumb of wheat bread (French type) – influence of pre-ferments and studies on the formation of key odorants during dough processing’, Z Lebensmittel Unters u Forschung, 201, 241–248. gassenmeier, k., nelissen, j. p., daniher, a. and furrer, s. m. (2004), ‘Preparation of acetaldehyde-precursor flavour and fragrance compounds’, Patent No. WO 2004048305. grab, w. and gfeller, h. (2000), ‘Flavorspace: a technology for the measurement of fast dynamic changes of flavour release during eating’, in Roberts D D and Taylor A J, Flavour Release, ACS Symposium Series 763, Washington, DC, American Chemical Society, 33–43. gremli, h. (1996), ‘Flavour changes in plastic containers: a literature review’, Perfumer & Flavorist, 21 (Nov/Dec), 1–8. grosch, w. (1994), ‘Determination of potent odorants in foods by aroma extract dilution analysis (AEDA) and calculation of odour activity values (OAVs)’, Flavour Fragr J, 9, 147–158. grosch, w. (2001), ‘Review. Evaluation of the key odorants of foods by dilution experiments, aroma models and omission’, Chem Senses, 26, 533–545. heiderich, s. and reineccius, g. (2001), ‘The loss of volatile esters from cookies’, Perfumer & Flavorist, 26 (6), 14–21. hofmann, t. and schieberle, p. (2002), ‘Chemical interactions between odor-active thiols and melanoidins involved in the aroma staling of coffee beverages’, J Agric Food Chem, 50, 319–326. hodgson, m., parker, a., linforth, r. s. t. and taylor, a. j. (2004), ‘In vivo studies on the long-term persistence of volatiles in the breath’, Flavour Fragr J, 19, 470– 475. iwanami, y., tateba, h., kodama, n. and kishino, k. (1997), Changes of lemon flavour components in an aqueous solution during UV irradiation, J Agric Food Chem, 45, 463–466. juteau, a., tournier, c. and guichard, e. (2004), ‘Influence of type and amount of gelling agent on flavour perception: physicochemical effect or interaction between senses?’, Flavour Fragr J, 19, 483–490.

Selecting the right flavourings for a food product

271

kallen, r. g. (1971), Equilibria for the reaction of cysteine and derivatives with formaldehyde and protons’, J Amer Chem Soc, 93, 6227–6235. kant, a., linforth, r. s. t. and taylor, a. j. (2003), ‘Flavour retention and release from starch solutions’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 103–106. kindel, g., krammer, g. and trautzsch, s. (2003), Coffee aroma with improved stability, Patent WO 03/101216. king, b. m. and solms, j. (1982), ‘Interactions of volatile flavour compounds with propyl gallate and other phenols as compared with caffeine’, J Agric Food Chem, 30, 838–840. king, b. m., arents, p., derks, e. p. p. a., duineveld, c. a. a., boelrijk, a. e. m. and burgering, m. j. m. (2003), ‘Flavour perception and flavour release in fruit flavoured yoghurt’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 27–32. krings, u., kelch, m. and berger, r. g. (1993), ‘Adsorbents for the recovery of aroma compounds in fermentation processes’, J Chem Technol Biotechnol, 58 (3), 293–299. krings, u., banavara, d. s. and berger, r. g. (2003), ‘Thin layer high vacuum distillation to isolate the flavour of high-fat food’, Eur Food Res Technol, 217 (1), 70–73. le guen, s. and vreeker, r. (2003), ‘Interactions between flavour compounds and milk proteins under static and dynamic conditions’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 182–187. lethuaut, l., brossard, c., meynier, a., rousseau, f., llamas, g., bousseau, b. and genot, c. (2005), ‘Sweetness and aroma perceptions in dairy desserts varying in sucrose and aroma levels and in textural agent’, Intern Dairy J, 15, 485–493. linforth, r., martin, f., carey, m., davidson, j. and taylor, a. (2002), ‘Retronasal transport of aroma compounds’, J Agric Food Chem, 50, 1111–1117. malone, m. e., appelqvist, i. a. m. and norton, i. t. (2003), ‘Oral behaviour of food hydrocolloids and emulsions. Part 2. Taste and aroma release’, Food Hydrocolloids, 17, 775–784. marin, m., baek, i. and taylor, a. j. (1999), ‘Volatile release from aqueous solutions under dynamic headspace dilution conditions’, J Agric Food Chem, 47, 4750– 4755. marin, m., baek, i. and taylor, a. j. (2000), ‘Flavour release as a unit operation: a mass transfer approach based on a dynamic headspace dilution method’, in Roberts D D and Taylor A J, Flavour Release, Washington, DC, American Chemical Society, 153–165. nahon, d. f., harrison, m. and roozen (2000), ‘Modelling flavour release from aqueous sucrose solutions, using mass transfer and partition coefficients’, J Agric Food Chem, 48, 1278–1284. ney, k. h. (1988), ‘Sensogramme, eine methodische Erweiterung der Aromagramme’, Gordian, 88 (1), 19. noble, a. c. (1996), ‘Taste–aroma interactions’, Trends Food Sci Technol, 7, 439–444. noble, a. c., kuo, y. and pangborn, r. m. (1993), ‘Temporal patterns of nasal, oral, and retronasal perception of citral and vanillin and interaction of these odourants with selected tastants’, Int J Food Sci Technol, 28, 127–137. normand, v., avison, s. and parker, a. (2004), ‘Modelling the kinetics of flavour release during drinking’, Chem Senses, 29, 235–245. paçi kora, e., souchon, i., latrille, e., martin, n. and marin, m. (2004), ‘Composition rather than viscosity modifies the aroma compound retention of flavoured stirred yogurt’, J Agric Food Chem, 52, 3048–3056.

272

Modifying flavour in food

parker, j. k., mottram, d. s. and adams, r. l. (2003), ‘Interactions of sulphurcontaining aroma compounds with proteins in both model systems and real food systems’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 45–50. patel, s., rabone, j., russell, s., tissen, j. and klaffke, w. (2001), ‘Iterated reaction graphs: simulating complex Maillard reaction pathways, J Chem Inf Comput Sci, 41 (4), 926–933. pfeiffer, j. c., hollowood, t. a., hort, j. and taylor, a. j. (2005), ‘Temporal synchrony and integration of sub-threshold taste and smell signals’, Chem Senses, 30, 539–545. porzio, m. (2004), ‘Flavour encapsulation: a convergence of science and art’, Food Technol, 58 (7), 40–42, 44, 46–47. pozo-bayon, m.-a., demaizieres, d. and guichard, e. (2004), ‘Flavouring of sponge cake: Retention of aroma compounds in function of composition’, in Hofmann T, Rothe M and Schieberle P, State-of-the-Art in Flavour Chemistry and Biology, Garching, Germany, Deutsche Forschungsanstalt für Lebensmittelchemie, 408–411. reineccius, t. a., reineccius, g. a. and peppard, t. l. (2004), ‘Utilization of βcyclodextrin for improved flavour retention in thermally processed foods’, J Food Sci, 69, 58–62. reynier, a., dole, p., fricoteaux, f., saillard, p. and feigenbaum, e. (2004), ‘Stabilization of aroma compounds through sorption-release by packaging polymers’, J Agric Food Chem, 52, 5653–5662. risch, s. j. (2000), ‘Flavour and package interactions’, in Risch S J and Ho C-T, Flavour Chemistry, ACS Symposium Series 756, Washington, DC, American Chemical Society, 94–100. schieberle, p. and grosch, w. (1992), ‘Changes in the concentrations of potent crust odorants during storage of white bread’, Flavour Fragr J, 7, 213–218. schreier, p., drawert, f., junker, a. and mick, w. (1977), ‘The quantitative composition of natural and technologically changed aromas of plants. II. Aroma compounds in oranges and their changes during juice processing’ (translation, written in German). Z Lebensm Unters-Forschung, 164 (3), 188–193. sen, a., laskawy, g., schieberle, p. and grosch, w. (1991), Quantitative determination of β-damascenone in foods using a stable isotope dilution assay. J Agric Food Chem, 39, 757–759. soottitantawat, a., yoshii, h., furuta, t., ohgawara, m., forssell, p., partanen, r., poutanen, k. and linko, p. (2004), ‘Effect of water activity on the release characteristics and oxidative stability of d-limonene encapsulated by spray drying’, J Agric Food Chem, 52, 1269–1276. soper, j. c. (1995), ‘Utilization of coacervated flavours’, in Risch S J and Reineccius G A, Encapsulation and Controlled Release of Food Ingredients, ACS Symposium Series 590 Washington, DC, American Chemical Society, 104–112. sucan, m. k. and weerasinghe, d. k. (2005), ‘Process and reaction flavours: an overview’, in Weerasinghe D K and Sucan M K, Process and Reaction Flavours, ACS Symposium Series 905, Washington, DC, American Chemical Society, 117–129. taylor, a., hollowood, t., davidson, j., cook, d. and linforth, r. (2003), ‘Relating taste and mouthfeel stimuli to overall flavour perception’, in Le Quéré J L and Etiévant P X, Flavour Research at the Dawn of the Twentieth-first Century, Paris, Editions Tec & Doc, 194–199. tressl, r., bahri, d. and engel, k. h. (1981), ‘Lipid oxidation in fruits and vegetables’, in Teranishi R and Barerabenitez H, Quality of Selected Fruits and Vegetables of North America, ACS Symposium Series 170, Washington, DC, American Chemical Society, 213–232.

Selecting the right flavourings for a food product

273

tsachaki, m., linforth, r. s. t. and taylor, a. j. (2005), ‘Dynamic headspace analysis of the release of volatile organic compounds from ethanolic systems by direct APCI-MS’, J Agric Food Chem, 53, 8328–8333 ueno, t., masuda, h. and ho, c.-t. (2004), ‘Formation mechanism of p-methylacetophenone from citral via tert-alkoxy radical intermediate’, J Agric Food Chem, 52, 5677–5684. valdés, r. m., hinreiner, e. h. and simone, m. j. (1956), ‘Effect of sucrose and organic acids in apparent flavour intensity’, Food Technol, 10 (6), 282–285. weel, k. g. c., boelrijk, a. e. m., alting, a. c., van mil, p. j. j. m., burger, j. j., gruppen, h., voragen, a. g. j. and smit, g. (2002), ‘Flavour release and perception of flavoured protein gels: perception is determined by texture rather than by release’, J Agric Food Chem, 50, 5149–5155. weurman, c. (1969), ‘Isolation and concentration of volatiles in food odour research’, J Agr Food Chem, 17, 370–384. wilkes, j. g., conte, e. d., kim, y., holcomb, m., sutherland, j. b. and miller, d. w. (2000), ‘Sample preparation for the analysis of flavours and off-flavours in foods’, J Chromatogr A, 880 (1+2), 3–33.

Index

ABC model 156–7 acids 255 active pharmaceutical ingredients (APIs) 224–5 adaptation 233–4 adenosine 5′-monophosphate (AMP) 205, 228–30, 239 AECI Ltd 119 ageing 207–8 agro-industrial wastes 81–2 air-drying (AD) 59 alanine 238 alapyridaine 195–6 alcohol 213 sugar 188 alcohol acyltransferases 71–4 acetyltransferases reactions (AAT) 72–3 Amadori compounds 260 amino acids 75–7 analysis extraction in 56–8 techniques of 16–18 anchovies 212 Argentina 35 ARKOPHARMA (France) 43 aroma 243–4 ‘aromatic’ ring 74 computerised optimisation 266

extraction dilution analysis (AEDA) 18, 245 artificial flavouring substances FSANZ definition 34 MERCOSUR definition 35 Asian cuisine 204, 205, 206 aspartame-acesulfame salt 190 astringency 214 asymmetric synthesis 111–13 Australasian legislation 34, 35 Australia 34, 120 bacon 214 baker’s yeast composition 97 Benefat 136 benzaldehyde 4 benzoic acids 196 BetraSaltTM 229–30 biocatalysts 65–7 bioconversion 66 bioprocessing advances 82–3 biotransformation 66, 115 bitter blockers 221–31 advantages to pharmaceuticals 224–5 current approaches 223–4 future trends 230 information sources 230 overview 8

276

Index

reasons for 221–3 suitable compounds 228–30 taste perception 225–8 bonito flakes 203, 205, 211–12 brain, representation of fat 140–2 Brazil 35, 82 brazzein 190, 240 bulk sweeteners 192–5 calcium imaging 153 CAMOLA 4 Canada 17 Capzasin® 146 carbon dioxide (CO2) extraction 42 liquid 48 supercritical extraction (SC-CO2) 49–50, 50–2, 57, 58 carboxyl methyl cellulose (CMC) 136, 140–1 carotenoids 69–71 centrifugal-suspension coating 175 CHARM analysis 18, 245 cheese Parmesan 211, 213 reduced fat spread 103, 104 chemesthetics and flavour 147–51 concepts 147–8 evaluation 148–51 chemical cross-linking 173, 178, 180–1 chemical groups 27–8 chemical instability 253–61 China 20, 33, 54, 118, 147 chiral flavour 107–30 information sources 127–8 (−)-menthol 118–21 natural identical products 108, 109–15 natural products 108, 109, 110, 111 overview 6 pinene/limonene 116–18 p-menthane lactones 121–7 classification of flavour substances 22, 23 Club of Rome 81 coacervation 24, 172–3, 178, 181–2 steps in process 173 Codex Alimentarius Commission (CAC) 36, 37

cola light 105 cold menthol receptor (CMR1) 152 cold pressing 47 combined fermentations 79–80 common compounds 154 computer simulation of flavour release 252 consumers 267 groups 2 umami and 208–9 continuous subcritical water extraction (CSWE) 52 controlled release see flavour release CoolAct® 10, 159 CoolAct P 157 cooling compounds see pungent/cooling sensations Council of European Communities 108 creating flavour 24, 244–9 information 244–6 process steps 244 cross-adaptation 233 cross-linking 173, 178, 180–1 curing 207–8 customer preference 267 cyclodextrins 171–2, 181–2, 248 cysteine 255 cytosolic mevalonate pathway 68 dashi 203, 204, 205, 206, 211–12, 215 de novo syntheses 66 definitions, flavouring substances 29–30 delivery systems 22–4 deterioration, flavour 170–3 deterpenation 48–9 diastereoisomerically enriched form 114–15 disodium adenylate (AMP) 205, 228–30, 239 disodium gyanylate (GMP) 203, 205, 217 disodium inosinate (IMP) 203, 205, 207, 217 distillation 47–8 DNA technology 228 microarray techniques 84 dry hydrophilic phase 260 dry substances 247–8

Index drying methods 58–9, 207–8 see also spray drying DSM 96 ‘electronic nose’ 82 enantio-enriched form 114–15 encapsulation 170–3 coacervation 172–3 cyclodextrins 171–2 extrusion 171 needs 181–2 spray drying 170–1 stability 259–60, 263–5 Enterobacter 65 enzymes 255 Escherichia coli 69, 73 essential oils 42, 44, 47, 48, 56, 58 esters 71–4 ethyl maltol 238 Europe 54, 222 cuisine 147 umami initiatives 217 European Food Safety Authority (EFSA) 26, 31, 36 European Union (EU) 32–3, 37, 65, 81 legislation 16, 25–32, 34, 108 Expert Committee on Food Additives (FAO/WHO) 36 extraction, natural flavour 41–63 in analysis 56–8 by electrical energy 48 continuous subcritical water (CSWE) 52 drying methods 58–9 microwave assisted (MAE) 54–6 non-conventional 48–50 overview 5, 42–50 solvent distillation 58–9 supercritical2 (SC-CO2) 49–50, 50–2, 57, 58 techniques 17 ultrasound assisted (UAE) 53–4 extrusion 171, 181, 248 fast protein liquid chromatography (FPLC) 84 fat -coated flavour 177 see also low-fat food

277

fat replacement difficulties 138–40 flavour impact 135–6 food texture and 136–7 sensory properties 137–8 strategies 134–8 types of 135 FEMA GRAS list 32–3, 35 fermentations 78–80, 207–8 ferulic acid 75 Firmenich 14, 16, 17, 21, 121 fish, smoked 212 Flavor and Extract Manufacturers Association (FEMA) 32, 196 expert panel (FEXPAN) 32, 36 Flavour and Fragrance Association of Australia and New Zealand 34 Flavour Industry Science Board, International Organisation of (IOFI) 34, 35, 37, 160 flavour release 169–84 computer simulation 252 controlled release 174–81, 182 insoluble capsule matrices 178–81 secondary coatings 174–8 encapsulation 170–3, 181–2 overview 7 problems 249–53, 252–3 prediction 249–52 product matrix effect 249 reformulation 252–3 flavour wheel 22–3 flavour-packaging interactions 261 flavouring substances 12–24 analytical techniques 16–18 Australasian legislation 34, 35 chemical groups 27–8 classification 22, 23 creating 24 definitions 29–30 delivery systems 22–4 European legislation 16, 25–32, 34, 108 harmonising legislation 35–6 historical perspective 13 Japanese legislation 33–4 model reactions 18–19 new 20–2, 26–9 overview 5

278

Index

South American legislation 35 synthetic 13–16 threshold values 19–20 United States legislation 32–3, 35, 108 Flavouring Substances Register 26, 29 fluidised bed coating 175–8 ‘folding’ 48–9 food acceptance and umami 210–15 texture and fat replacement 136–7 Food and Drug Administration, US (FDA) 65, 108, 235, 237, 238–9 ‘Food Improvement Agents Directive’ 29, 35 Food Sanitation Law (Japan) 33 Food Standards Code (Australasia) 34 formulation initiatives 217–18 fowl stock 214 FRANZ (Food Standards Australia, New Zealand) 34 freeze-drying 58–9 Frescolat MGA 159 Frescolat MPC 160 FTNF (‘from the named fruit’) 33 functional magnetic resonance imaging (fMRI) 141–2 Furaneol (XVII) 16 G protein coupled receptors (GPCRs) 187, 227 cascade 152–3 gas chromatography (GC) 17, 45, 57 -mass spectrometry (GC-MS) 54, 57, 244–5 -olfactometry (GC-sniff) 245 gelatin 173, 180, 181, 182 General Standard for Food Additives 36 ‘generally recognised as safe’ (GRAS) 32, 65, 80, 84, 163, 196 masking agents 235, 237, 239, 240 generation, flavour off-line 260 in situ 260 Germany 34, 65, 75, 207 Givaudan 14, 160–1, 246 global legislation 35–6 glutamate 207, 209, 210, 213, 217

glutamic acid 211–12 glutaraldehyde 180 glycine 238–9 glycyrrhizin 235 government departments Agriculture (USA) 223 Department of Health (UK) 131–2 Health and Human Services (USA) 222 GPCR see G protein coupled receptors (GPCRs) granulation 178 GRAS see ‘generally recognised as safe’ (GRAS) guanosine 5′-monophosphate (GMP) 203, 205, 217 gum acacia 173, 177, 181 gustation 10–11 gustducin 187, 227 gymnemic acid 234 Haarmann & Reimer 13, 118–19, 160 half-mouth method 150–1 half-tongue method 150–1 Hasegawa 161 headspace analysis 57, 244 heating methods 263 herbs 146 high hydrostatic pressure (HHP) 50 high vacuum distillation 244 high-nucleotide prototypes 99 applications 100–1, 101–3, 103–5 high-potency sweeteners 186, 189–92 high-pressure techniques 50, 57 high-speed mixing 48 high-throughput screening (HTS) 153– 4, 230 historical perspective 13 honey 21–2 hot/cold sensations see pungent/cooling sensations HotAct®154 Hua Jiao (sanhool) 154 Hungary 147 hydration 256, 259 hydrodistillation (HD) 51 hydrolysis 256, 259 hydrophilic phase 255–6 hydrophobicity 244

Index Icilin 159 impact, desired 266–7 India 33, 147 Indonesia 147 industrial approaches to controlled release 174–81 to encapsulation 170–3 inosine 5′-monophosphate (IMP) 203, 205, 207, 217 insoluble capsule matrices 178–81 wall matrices 179–81 Institute of Food Technologists (IFT) 218 International Flavours and Fragrances (IFF) 14, 65 International Symposia on Olfaction and Taste (ISOT) 204 on Umami 204 isomaltulose 192–3 isoprenoids 67–71 isopropyl alcohol (IPA) 23 Italy 207 Jambu (spilanthol) 154 Japan 21, 35, 96, 203, 205, 211 legislation 33–4 Joint Food and Agriculture Organization of UN/WHO Expert Committee on Food Additives (JECFA) 36, 37 kelp 203, 205, 206, 211–12 ketchup 101–2 α-ketoglutarate 76–7 kombu (seaweed) 203, 205, 206, 211–12 lactic acid bacteria 78–9 lactisole 197 Lactisole™ 238 Leffingwell and Associates 157 FlavorBase database 154 legislation Australasian 34, 35 European 16, 25–32, 34, 108 harmonising 35–6 Japanese 33–4 South American 35 United States 32–3, 35, 108 Lewis acids 77

279

limonene 116–18 Linguagen 3 lipid metabolites 71–4 liquids 247, 254–9, 262–3 cabon dioxide (CO2) 48 liquid-phase extraction 44 lo han guo 191 low-fat food 104–5, 131–45 aroma influence 133–4 brain representation 140–2 fat replacement difficulties 138–40 strategies 134–8 flavour influence 132–3, 134–8 future trends 142–3 overview 6–7 lysine 239 maceration 42 Maghreb 147 Maillard reaction 4, 96, 132, 195 flavouring substances and 18, 22, 31 precursors 82 stability and 259, 261 maltol 237–8 maple furanone 20 ‘market pull’ 3, 6 masking agents 232–42 appropriate use 240 considerations 239 future trends 240–1 ingredients 235–9 high-potency sweetness 235–7 no sweetness response 237–9 overview 8 workings of 233–4 mass spectrometry 17 Maxarome® Plus 96 Mayan medicine 146 mayonnaise, reduced fat 102–3 meat 18–19 smoked 214 stock 214 medium chain triglycerides (MCT) 23 melt extrusion 24 ‘memory effect’ 150 p-menthane -3,8-diol (PMD-38) 160

280

Index

flavours 122 lactones 121–7 menthol 147 (−)-menthol 118–21 l-menthol 157 menthyl lactate 158 MERCOSUR (Mercado Común del Sur) 35 ‘metagenome’ technique 84 metal ions 255 methyl xanthines 227 Mexico 147 microbial catalysts 64–94 amino acids 75–7 flavour generation 64–5 future trends 83–5 information sources 85–6 isoprenoids 67–71 lactic acid bacteria 78–9 lipid metabolites/esters/alcohol acyltransferases 71–4 molecular engineering 80–1 overview 5–6 phenolics 74–5 sequential/combined fermentations 79–80 substrates/biocatalysts 65–7 thio-compounds 77–8 white biotechnology 81–3 yeasts 79 microorganisms 255 microwave -assisted extraction (MAE) 44, 48, 49, 54–6, 57 baking 262 irradiation 45–6 mint 146–7 miraculin 234, 237, 240 ‘mixed list’ system 34 model reactions 18–19 molecular encapsulation 24 molecular engineering 80–1 monatin 191 monellin 191 monosodium glutamate (MSG) 203, 206–7, 210, 217 mouthfeel 243–4 mushrooms 214, 215

National Diet and Nutrition survey (UK) 131 natural flavouring substances 29, 34 natural products 108, 109, 110, 111 sweeteners 190–1 naturalness, flavour 33, 65–7 see also extraction, natural flavour nature identical products 108, 109–15 flavouring substances 16, 34, 35 neohesperidin dihydrochalcone 235 NeotameTM 237 Netherlands 12 New Zealand 34 non-sugar sweeteners 188–9 norisoprenoids 70–1 North America 222 nuclear magnetic resonance (NMR) 245 nucleotides 210, 217 purity 100–1 Nutrition and Food Research Institute (TNO) 12–13 nutritional enhancement 218 odour -activity values (OAVs) 245–6 port assessment of gas chromatograms 17 properties 124 off-flavours 66 oleoresin method 50 olfaction 11–12, 198 Optimising Sweet Taste in Foods (Spillane) 189 orbitofrontal cortex (OFC) 140–1 oven-drying 58–9, 207 oven-roasting 207–8 Oxane 121 Oxford Chemicals 22 Paraguay 35 Parmesan cheese 211, 213 pentadin 240 peppers 146–7 perception sweetness 187–9 taste 225–8 percolation 42, 50

Index pharmaceuticals, bitter blockers and 224–5 phenolics 74–5 phosphodiesterase (PDE) 227 physical instability 261–5 physiology and hot/cold sensations 151–4 Physiology of Taste (Brillat-Savarin) 203 pinene/limonene 116–18 plastic coatings 256 plastidial DOXP/MEP pathway 68 Poland 207 polyethylene terephthalate (PET) bottle 258, 261 potassium chloride (KCl) 228–30 potentiators, sweetness 195–6 precursor systems 260–1, 265 prediction, flavour release 249–52 pressurised hot water extraction 57–8 pressurised liquid extraction (PLE) 49, 58 pressurised solvent extraction (PSE) 48 ‘prilling’ 179 product development, umami and 215–16 product matrix effect 249 proteins 255 pungent/cooling sensations 146–68 chemesthetics and flavour 147–51 cooling compounds 157–62 structure-activity 160–2 future trends 162–3 overview 7 physiology and 151–4 pungent chemicals 154–7 structure-activity 154–7 Qarôma 161 quantitative trait loci (QTL) 68 racemic form 114–15 raspberry flavouring 13–15, 17 recipe development, umami and 215–16 reformulation, flavour 252–3 release of flavour see flavour release renewable substrates 81–2

281

‘reservoir’ effect 135 ripening 207–8 RNA 96–8 rotor granulator 178, 179 sage (Salvia officinalis) 53, 58 salt 211 Sandoz 156 Sanitary and Phytosanitary Measures (SPS) 36 Scoville point scale 148 seaweed (kombu) 203, 205, 206, 211–12 secondary coatings 174–8 selection, flavour 243–73 creation 244–9 flavour release 249–53 future trends 265–7 information sources 267 stability 247–9, 253–65 self-adaptation 233 self-cloning 73 Senomyx 3 sensory properties, fat replacement products 137–8 sequential fermentations 79–80 sequential reactions (bioconversion) 66 Serra and Fuganti synthesis 127 Serratia 65 shrimps 21 Simplesse® 136 simultaneous distillation-extraction (SDE) 57 single step reactions (biotransformation) 66, 115 smoked food 212, 214 Society for Research on Umami Taste 203 solid phase micro-extraction (SPME) 17, 57 solid-liquid extraction (SLE) 42 solid-state fermentation 76 solvents 22–4 -assisted flavour evaporation (SAFE) 17, 57 distillation 58–9 extraction 47 somatosensorial sensations see pungent/cooling sensations

282

Index

South America 146–7 legislation 35 South East Asia 33, 37 Soxhlet extraction 42, 45, 55, 57–8 soy sauce 210–11 soybean germ 47 Spain 207 spices 146 spray cooling 24, 179 spray drying 170–1, 174, 181, 248 flavouring particles 176 stability, flavour 247–9, 253–65 chemical 253–61 dry substances 247–8 liquids 247 physical 261–5 steam distillation 42, 244 stir-bar sorptive extraction 57 stock, meat/fowl 214 storage, improper 261 strawberry flavouring 15–16 low fat yoghurt drink 104–5 structure-activity relationship (SAR) 228, 230 cooling compounds 160–2 pungent chemicals 154–7 sweeteners 191–2 subcritical water extraction 57–8 sublimation 244 substrates 65–7 complex 78–81 renewable 81–2 sugars 188 sulphite 255 supercritical CO2 (SC-CO2) extraction 49–50, 50–2, 57, 58 supercritical fluid extraction (SFE) 42– 3, 44, 48, 51–2, 54, 55, 58 superheated water extraction (SWE) 49 ‘surprise’ effect (spices) 150 sweet taste high-potency 235–7 inhibitors 186 potentiators 186 receptors 192 sweeteners 4, 185–201 future trends 197–8 information sources 198

inhibitors and 197 novel 189–95 bulk 192–5 high-potency 189–92 overview 7 perception and 187–9 potentiators and 195–6 Symrise 13, 65, 157 synbiotic foods 142 synergy, umami 207 synthetic flavour 13–16, 35 d-Tagatose 194–5 Takasago process 118–19, 154, 157, 160 taste 243–4 -activators 206–7 blocking 233–4 modalities 226 modification 233, 234 perception 225–8 receptor cells 189, 203 suppression 234 synergism 234 transduction 226 see also sweet taste tastomics 198 terpenes 48 Thailand 147 thaumatin 191, 237 thio-compounds 77–8 threshold values 19–20 time-intensity (TI) method 133 TNO (Nutrition and Food Research Institute) 12–13 tomatoes 213 pasta sauce 102 transient receptor potential (TRP) 148, 149, 151–3, 156, 159, 228 α,α-trehalose 193–4 triacylglycerol (TAG) 139 turbo-extraction 48 Twinsweet® 190 ultra-high pressure (UHP) 50 ultra-high-throughput screening 189 ultrasound 43–4 -assisted extraction (UAE) 44, 48, 49, 53–4 irradiation 46

Index umami 202–20 consumer and 208–9 culinary aspects 203, 204–5 defined 202 European initiatives 217 food acceptance/preference and 210–15 food technology and 210 formation processes 207–8 future trends 217–18 information sources 218 overview 7–8 recipes/products development 215–16 science of 203–4, 204–5 synergy 207 taste-activators 206–7 United States initiatives 216–17 Western foods and 205–6 Unilever 16 United Kingdom (UK) 2, 54, 131–2, 205 United States of America (USA) 3, 4, 54, 65, 196, 222–3, 237 legislation 32–3, 35, 108 umami and 203, 205, 207, 209, 214, 216–17 Uruguay 35 vacuum microwave-drying (VMD) 59 vacuum simultaneous distillationsolvent extraction (V-SDE) 57 van der Waals interaction 133 vanillin 13, 50, 74–5, 133 Vicks VapoRub® 146 Vietnam 147 virtual aroma synthesiser (VAS) 246

283

viscosity 140 vodka 213 Volatile Compounds in Foods (VCF) (TNO) 12, 20 volatilisation 261 vortical-extraction 48 wasabi 216 water 255 weight management 218 Western foods, umami and 205–6 white biotechnology 66, 81–3 bioprocessing advances 82–3 renewable substrates/agro-industrial wastes 81–2 Wilkinson-Sword 159, 160–1 wine lactone 121–7 umami and 214–15 WONF (‘with other natural flavourings’) 33 World Trade Organization (WTO) 36, 37 yeast extracts 95–106 applications 99 baker’s yeast composition 97 developments 96–8 high-nucleotide prototypes 99 applications 100–1, 101–3, 103–5 nucleotide purity 100–1 overview 6 preparation 98–9 yeasts 79 encapsulation 24 yoghurt drink, low-fat strawberry 104–5