Milk Proteins: From Expression to Food [3 ed.] 9780128152522, 0128152524, 9780128152515

Table of contents :
Content: 1. The World Supply of Food and the Role of Dairy Protein 2. Milk: An Overview 3. The Comparative Genomics of Monotremes, Marsupials, and Pinnipeds: Models to Examine the Functions of Milk Proteins 4. Significance, Origin, and Function of Bovine Milk Proteins: The Biological Implications of Manipulation or Modification 5. Post-translational Modifications of Caseins 6. Casein Micelle Structure and Stability 7. Structure and Stability of Whey Proteins 8. Effect of Non-thermal Processing of Milk Protein Interactions and Functionality 9. The Whey Proteins in Milk: Thermal Denaturation, Physical Interactions, and Effects on the Functional Properties of Milk 10. Effect of UHT Processing and Storage on Milk Proteins 11. Effects of Drying and Storage on Milk Proteins 12. Interactions and Functionality of Milk Proteins in Food Emulsions 13. Milk Protein-Polysaccharide Interactions 14. Interactions between Milk Proteins and Micronutrients 15. Application of Milk and Whey Protein Ingredients in Foods 16. Milk Protein Gels 17. Milk Proteins: A Cornucopia for Developing Functional Foods 18. Milk Proteins and Human Health 19. Structural Changes to Milk Protein Products during Gastrointestinal Digestion 20. Milk Proteins: Digestion and Absorption in the Gastrointestinal Tract 21. Milk Proteins: The Future

Citation preview

MILK PROTEINS THIRD EDITION

This page intentionally left blank

MILK PROTEINS FROM EXPRESSION TO FOOD THIRD EDITION Edited by

MIKE BOLAND HARJINDER SINGH

Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1650, San Diego, CA 92101, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom © 2020 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN 978-0-12-815251-5 For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Charlotte Cockle Acquisition Editor: Megan Ball Editorial Project Manager: Ruby Smith Production Project Manager: Surya Narayanan Jayachandran Cover Designer: Mark Rogers Typeset by SPi Global, India

Dedication

We dedicate this book to the memory of W.J. (Jim) Harper—a pioneer in dairy science and a mentor and friend to many of us.

This page intentionally left blank

Contents 3. The comparative genomics of monotremes, marsupials, and pinnipeds: Models to examine functions of milk proteins

Contributors xi Preface to the Third Edition xiii 1. World supply of food and the role of dairy protein

Julie Sharp, Christophe Lefe`vre, Kevin R. Nicholas

Mike Boland, Jeremy Hill

Introduction 99 The monotremes 101 The tammar wallaby (Macropus eugenii) 107 A role for milk in the control of mammary function 121 The fur seal 122 A new player in milk bioactives: miRNA 130 Conclusions 131 References 132

Introduction and outline of chapter 1 Hunger and need for food 2 The importance of protein in world nutrition 5 The dietary essential amino acids in proteins 10 Demographic changes, aging populations, and the need for quality protein and dietary essential amino acids 11 Global trade in proteins and the long term prospects, with a focus on dairy foods 14 Conclusions 17 References 17 Further reading 19

4. Defining the origin and function of bovine milk proteins through genomics: The biological implications of manipulation and modification

2. Milk proteins: An overview

Sarah Berry, Paul Sheehy, Peter Williamson, Julie Sharp, Karensa Menzies, Christophe Lefe`vre, Matthew Digby, Chad Harland, Stephen Davis, Russell Snell

D.A. Goulding, P.F. Fox, J.A. O’Mahony

Introduction 21 Bovine milk composition 22 Milk protein system 27 Casein 33 Whey proteins 39 Differences between casein and whey proteins 43 Minor milk proteins 47 Analytical considerations for milk proteins 51 Milk protein ingredients 69 References 81

Introduction 143 Milk genomics: A contemporary approach to milk composition 144 Comparative milk genomics 149 Origins of milk proteins 150 Constraints and opportunities for evolution or manipulation of bovine milk proteins 152 Conclusions 163 References 164

vii

viii

Contents

5. Posttranslational modifications of caseins Etske Bijl, John W. Holland, Mike Boland

Introduction 173 The caseins 175 Sources and significance of casein heterogeneity Caseins from other species 195 Conclusions 202 Acknowledgments 202 References 202

183

6. Casein micelle structure and stability David S. Horne

Introduction 213 Primary structure and interactions of caseins 215 Casein micelle properties 221 Models of casein micelle structure 225 Dual-binding model for micelle assembly and structure 225 Calcium phosphate nanoclusters 227 Application of the dual-binding model 231 Concluding remarks 243 References 244

7. Structure and stability of whey proteins Patrick J.B. Edwards, Geoffrey B. Jameson

Introduction 251 Bovine β-lactoglobulin 252 α-Lactalbumin 269 Serum albumin 272 Immunoglobulins 276 Lactoferrin 278 Concluding remarks 280 Acknowledgments 280 References 280 Further reading 291

8. Effect of nonthermal processing on milk protein interactions and functionality Pranav K. Singh, Thom Huppertz

Introduction 293 High-pressure processing US processing 302

295

PEF processing 306 UV irradiation processing 312 Concluding remarks 316 References 316 Further reading 324

9. The whey proteins in milk: Thermal denaturation, physical interactions, and effects on the functional properties of milk Skelte G. Anema

Introduction 325 The casein micelle 326 The heat treatment of milk 330 Relationships between denaturation/interactions of the whey proteins in heated milk and the functional properties of milk products 353 Conclusions 376 References 376

10. The effect of UHT processing and storage on milk proteins Hilton C. Deeth

Introduction 385 The UHT process 386 Protein changes during processing and storage 388 Conclusions 414 References 414

11. Effects of drying and storage on milk proteins Alan Baldwin, Kerianne Higgs, Mike Boland, Pierre Schuck

Introduction 423 World dairy powder situation 425 Properties of spray-dried milk products 428 Principles of spray drying 428 Drying of proteins 433 Characterization of insolubility 440 Changes in milk proteins during storage of dry powders 444 Rehydration of protein powders 456 Conclusions 461 References 461 Further reading 466

ix

Contents

12. Interactions and functionality of milk proteins in food emulsions

15. Model food systems and protein functionality

Harjinder Singh, Aiqian Ye

W. James Harper, Sheelagh A. Hewitt, Lee M. Huffman

Introduction 467 Adsorption of milk proteins during the formation of emulsions 469 Stability of milk protein–based emulsions 476 Process-induced changes in milk protein–based emulsions 481 Behavior of milk protein–stabilized emulsions under physiological conditions 486 Conclusions 491 References 491

13. Milk protein-polysaccharide interactions Kelvin K.T. Goh, Anges Teo, Anwesha Sarkar, Harjinder Singh

Introduction 499 Mixing behavior of biopolymers 500 Phase diagram 502 Nature of interactions in protein-polysaccharide systems 504 Milk protein-polysaccharide interactions in the aqueous phase and at the interface 506 Rheological properties and microstructures of protein-polysaccharide systems 516 Concluding remarks 525 References 527 Further reading 535

14. Interaction between milk proteins and micronutrients There`se Considine, John Flanagan, Simon M. Loveday, Ashling Ellis

Introduction 537 Interaction between milk proteins and micronutrients 538 Effect of processing on milk protein structure 555 Conclusions 560 References 561

Introduction 573 Protein functionality in foods 574 Role of interactions in determining food characteristics 575 Processing effects 580 Uses of model food systems 581 Applications of model food systems 584 Use of model food systems for other food components 590 Limitations 591 Conclusions 591 References 592

16. Milk protein gels John A. Lucey

Introduction 599 Rennet-induced gels 600 Acid-induced milk gels 609 Mixed gels made with rennet and acid 616 Whey protein gels 617 Conclusions 624 References 625 Further reading 632

17. Milk proteins: A rich source of bioactives for developing functional foods Paul J. Moughan

Introduction 633 Functional foods 634 Milk proteins as a source of amino acids: Specialized nutritionals 636 Milk proteins as a source of amino acids: Specific physiological roles 639 Milk proteins as a source of amino acids: Role in providing calories and in promoting satiety 642 Milk protein as a source of bioactive peptides 643 Holistic properties of foods 646 Conclusions 646 References 646 Further reading 649

x

Contents

18. Milk proteins and human health Sally D. Poppitt

Introduction 651 Milk proteins, metabolic health, and type 2 diabetes 652 Milk proteins, obesity, and weight control 653 Milk proteins, muscle wasting, and sarcopenia 656 Milk proteins and heart health 658 Milk proteins and bone health 660 Milk proteins and infant health 661 Conclusions 663 References 663

19. Structural changes to milk protein products during gastrointestinal digestion Aiqian Ye, Debashree Roy, Harjinder Singh

Introduction 671 Stress conditions of the GI tract 672 Coagulation of milk protein under gastric conditions 674 Coagulation of milk during gastric digestion 675 Effect of processing treatments 678 Impact of milk coagulation on the release of fat globules during digestion 681 Behavior of milk fat globule membrane proteins during digestion 682 Milk protein ingredients 685 Digestion of nonbovine milks 688

Concluding remarks 693 References 694

20. Milk proteins: Digestion and absorption in the gastrointestinal tract Didier Dupont, Daniel Tome

Introduction 701 Digestion of milk proteins 702 Milk protein hydrolysis in the intestinal lumen 703 Peptides released during digestion 705 Impact of processing on milk protein digestion and absorption 707 Conclusions 711 References 712 Further reading 714

21. Milk proteins: The future Mike Boland

Introduction 715 Global issues for food 715 Consumer demands and trends for food and ingredients 720 New technologies and their possible effect on milk protein ingredients and products 724 Conclusions 727 References 728

Index 731

Contributors Skelte G. Anema Fonterra Research and Development Centre, Palmerston North, New Zealand

Kelvin K.T. Goh School of Food and Advanced Technology, Massey University, Palmerston North, New Zealand

Alan Baldwin Formerly Fonterra Research Centre, Palmerston North, New Zealand

D.A. Goulding School of Food and Nutritional Sciences, University College Cork, Cork, Ireland

Sarah Berry Auckland War Memorial Museum, Auckland, New Zealand

Chad Harland Livestock Improvement Corporation, Hamilton, New Zealand

Etske Bijl Dairy Science and Technology, Food Quality and Design Group, Wageningen University and Research, Wageningen, The Netherlands

W. James Harper†

Mike Boland The Riddet Institute, Massey University, Palmerston North, New Zealand

Sheelagh A. Hewitt Fonterra Research and Development Centre, Fonterra Co-operative Group Ltd, Palmerston North, New Zealand

There`se Considine Fonterra Co-operative Group Ltd, Palmerston North, New Zealand

Kerianne Higgs Formerly Fonterra Research Centre, Palmerston North, New Zealand

Stephen Davis Livestock Improvement Corporation, Hamilton, New Zealand

Jeremy Hill The Riddet Institute, Massey University; Fonterra Co-operative Group, Palmerston North, New Zealand

Hilton C. Deeth School of Agriculture and Food Sciences, University of Queensland, Brisbane, QLD, Australia

John W. Holland Drug Toxicology Unit, Forensic and Analytical Science Service, NSW Health Pathology, North Ryde, NSW, Australia

Matthew Digby Department of Zoology, University of Melbourne, Melbourne, VIC, Australia Didier Dupont STLO, INRA, Agrocampus Ouest, Rennes, France

David S. Horne Wisconsin Center for Dairy Research, University of Wisconsin—Madison, Madison, WI, United States

Patrick J.B. Edwards School of Fundamental Sciences, Massey University, Palmerston North, New Zealand

Lee M. Huffman New Zealand Institute for Plant & Food Research Ltd, Palmerston North, New Zealand

Ashling Ellis The Riddet Institute, Massey University, Palmerston North, New Zealand

Thom Huppertz Dairy Science and Technology Group, Wageningen University, Wageningen; FrieslandCampina, Amersfoort, The Netherlands

John Flanagan

Naturex S.A., Avignon, France

P.F. Fox School of Food and Nutritional Sciences, University College Cork, Cork, Ireland

†

Geoffrey B. Jameson School of Fundamental Sciences; The Riddet Institute, Massey University, Palmerston North, New Zealand

Deceased.

xi

xii

Contributors

Christophe Lefe`vre Peter MacCallum Cancer Research Institute, East Melbourne; Division of Bioinformatics, Walter and Eliza Hall Medical Research Institute; Department of Medical Biology (WEHI), The University of Melbourne, Melbourne, VIC, Australia Simon M. Loveday AgResearch; The Riddet Institute, Massey University, Palmerston North, New Zealand John A. Lucey Wisconsin Center for Dairy Research, University of Wisconsin-Madison, Madison, WI, United States Karensa Menzies Agri-Food Development Specialist, Melbourne, VIC, Australia Paul J. Moughan Riddet Institute, Massey University, Palmerston North, New Zealand Kevin R. Nicholas Department of Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, VIC, Australia J.A. O’Mahony School of Food and Nutritional Sciences, University College Cork, Cork, Ireland Sally D. Poppitt School of Biological Sciences; Department of Medicine; Human Nutrition Unit, University of Auckland, Auckland; The Riddet Institute, Massey University, Palmerston North, New Zealand

Debashree Roy Riddet Institute, Massey University, Palmerston North, New Zealand Anwesha Sarkar School of Food Science and Nutrition, University of Leeds, Leeds, United Kingdom Pierre Schuck

INRA, Rennes, France

Julie Sharp Agri-Food Development Specialist, Melbourne, VIC, Australia Paul Sheehy Faculty of Science, Sydney School of Veterinary Science, University of Sydney, Sydney, NSW, Australia Pranav K. Singh College of Dairy Science & Technology, GADVASU, Ludhiana, India; Riddet Institute, Massey University, Palmerston North, New Zealand Harjinder Singh Riddet Institute, Massey University, Palmerston North, New Zealand Russell Snell School of Biological Sciences, University of Auckland, Auckland, New Zealand Anges Teo Abbott Nutrition Research and Development, Singapore, Singapore Daniel Tom e PNCA, INRA, AgroParisTech, Paris, France Peter Williamson Faculty of Science, School of Life and Environmental Sciences, University of Sydney, Sydney, NSW, Australia Aiqian Ye Riddet Institute, Massey University, Palmerston North, New Zealand

Preface to the Third Edition It is now another 5 years since the second edition of Milk Proteins: From Expression to Food was published. In that time, there have, again, been considerable advances in the science covered in the first and second editions, which means that all chapters needed to be updated. In the past 5 years, the way we think about milk proteins and more broadly about animalderived protein has changed. The environmental burden of animal proteins has become socially important, leading to “flexitarian” and “vegan” diets and lower levels or zero levels of consumption of animal proteins—making it even more important to understand how we can best use them. Increasing consumption of animal protein products in developing countries has led to a demand that will become unsustainable, and there is a need to balance overconsumption of protein in many Western economies with underconsumption in developing regions. New science has led to advances in all aspects of milk proteins but specifically advances in our understandings

of UHT processing and its effects on proteins; the use of new types of nonthermal processing and new understanding of the structural basis of protein digestion have resulted in three new chapters: many other chapters have been extensively rewritten. All chapters have been revised and updated. Not all of the original authors were available to update their chapter, and so, others have come aboard to rewrite and update. We thank these new authors joining the team. We also thank the original authors for their chapters and particularly those who have gone through them again to bring them completely up to date. Finally, we would like to thank the excellent staff at Elsevier, particularly Ruby Smith, and also Ansley Te Hiwi of the Riddet Institute and Claire Woodhall, our technical editor, for their efforts in making this volume a reality.

xiii

Mike Boland Harjinder Singh

This page intentionally left blank

C H A P T E R

1 World supply of food and the role of dairy protein Mike Bolanda, Jeremy Hilla,b a

The Riddet Institute, Massey University, Palmerston North, New Zealand b Fonterra Co-operative Group, Palmerston North, New Zealand

Introduction and outline of chapter As one of the basic necessities of life, the availability, quality, and affordability of food are of concern to individuals and nations alike. The “Green Revolution” started in the late 1960s and involved the introduction of high-yielding new seed varieties and better use of irrigation and fertilizers. Following this, the world enjoyed several decades of relative stability in the price of basic items of food, including food grains. The period from the early 1970s to 1990 saw the world output of food grains, and oilseeds rise steadily by an average of 2.2% a year, with periodic fluctuations. With the exception of parts of the African continent, the rate of growth of food crops exceeded that of the world population, leading to an increase in their per capita availability, and to relatively stable prices. Indeed, world food prices in real terms were at their lowest in 100 years in 2000 (Trostle, 2008). However, from the early 1990s, the global rate of the growth of grain and oilseed production declined, reaching a net 0% for the decade to 2016. Over the same decade, milk production grew by, on average, 2%. The inflation-adjusted Food and Agriculture Organization (FAO) of the United Nations Food Price Index stood at 141.1 points in October 2018, having risen from its 2002 to 2004 base of 100.0 points (FAO, 2018) and reached a high of 229.9 points in 2011. The October 2018 decline in the Food Price Index was the result of falling dairy, meat, and oil prices, which more than offset a surge in sugar prices and a more moderate increase in the prices of cereals. The FAO Dairy Price Index averaged 181.8 points in October 2018, continuing a downward trend for a fifth consecutive month. The price decrease reflects the growing evidence of increased export supplies across all major dairy products. In contrast, the FAO Cereal Price Index averaged 166.3 points in October 2018, representing a 13.6 point (8.9%) year-on-year increase. Among the major cereals, maize quotations from the United States firmed the most,

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00001-3

1

# 2020 Elsevier Inc. All rights reserved.

2

1. World supply of food and the role of dairy protein

supported by strong export sales, and wheat prices also averaged higher, driven by a tighter supply outlook, especially in view of deteriorating crop prospects in Australia. Side by side with this worsening situation with respect to the availabilities and prices of the major cereals, there has been another recent development involving food consumption patterns in a number of countries. With growing affluence, tastes change, and consumers shift to more varied diets, which usually include larger proportions of noncereal items. Over the last few decades, several poorer countries, among them, the two most populous, namely, China and India, have experienced rapid growth and transformation in their economies. This has helped to lift several million people out of extreme poverty and to make many others more affluent, bringing in its trail significant changes in food consumption. One of the more noticeable changes has been an increase in the consumption of animal protein. These developments have important policy implications for the global food economy. At the same time, hunger still afflicts a large number of people globally and is in need of policies to resolve or mitigate it. This chapter examines several interrelated issues relating to the evolving world food situation. In particular, it investigates the issue of nutrition and the role of proteins and their constituent amino acids. It starts by looking at the issue of hunger, its measurement, its global incidence, and mitigation targets. It then addresses the issue of nutrition, its global and regional perspectives, and the role of animal and vegetable proteins. The evolving global demographic trends, with a rapidly increasing elderly population that has special nutritional needs, call for policies to deal with the issue of nutrition for the aged, and the role of proteins in it. As the production and the consumption of protein foods in different countries do not always match, significant international trade in protein products exists. This is briefly discussed to identify the major players in the global protein markets. The chapter concludes with some observations on the policy implications deriving from its discussions. In the examination of proteins in their various dimensions, the particular focus of this chapter is the evolving role of dairy proteins and its implications for future policy.

Hunger and need for food Every day, millions of people around the globe do not get enough food to eat and remain hungry. Hunger has been referred to as “the uneasy or painful sensation caused by a lack of food” and “the recurrent and involuntary lack of access to food” (Anderson, 1990). There is no assurance that these hungry people will get the minimum required quantity of food on a daily basis. This unpredictability about where the next meal will come from is called “food insecurity.” The FAO (2000) defines food insecurity as: “A situation that exists when people lack secure access to sufficient amounts of safe and nutritious food for normal growth and development and an active and healthy life.” Following this definition, people are hungry if they do not get enough energy supply from food (fewer than about 1800 kilocalories a day), or if the food they consume is not of sufficiently high quality (containing essential nutrients). Hunger is usually understood to refer to the discomfort associated with a lack of food (von Grebmer et al., 2012).

Hunger and need for food

3

Hunger-reduction targets The United Nations 2030 Agenda for Sustainable Development (United Nations, 2015a) is a framework for advancing the well-being of humankind. It aims to ensure peace and prosperity for all on a healthy planet. The 17 sustainable development goals address global challenges related to poverty, inequality, climate, environmental degradation, prosperity, and peace and justice and are targeted to be reached by 2030. Goal 2 specifically addresses hunger and is simply stated as: Goal 2: Zero Hunger. The goal has five subgoals. 2.1 By 2030, end hunger and ensure access by all people, in particular the poor and people in vulnerable situations, including infants, to safe, nutritious, and sufficient food all year round. 2.2 By 2030, end all forms of malnutrition, including achieving, by 2025, the internationally agreed targets on stunting and wasting in children under 5 years of age, and address the nutritional needs of adolescent girls, pregnant and lactating women, and older persons. 2.3 By 2030, double the agricultural productivity and incomes of small-scale food producers, in particular women, indigenous peoples, family farmers, pastoralists, and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition, and nonfarm employment. 2.4 By 2030, ensure sustainable food production systems and implement resilient agricultural practices that increase productivity and production; that help to maintain ecosystems; that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding, and other disasters; and that progressively improve land and soil quality. 2.5 By 2020, maintain the genetic diversity of seeds, cultivated plants, and farmed and domesticated animals and their related wild species, including through soundly managed and diversified seed and plant banks at the national, regional, and international levels, and promote access to and fair and equitable sharing of benefits arising from the utilization of genetic resources and associated traditional knowledge, as internationally agreed. Goal 2.2 is particularly important with respect to protein supply, as insufficient protein is one of the major contributors to childhood stunting. According to the United Nations, childhood stunting is one of the most significant impediments to human development, globally affecting approximately 162 million children under the age of 5 years. Stunting, or being too short for one’s age, is defined as a height that is more than two standard deviations below the World Health Organization (WHO) Child Growth Standards median. It is a largely irreversible outcome of inadequate nutrition and repeated bouts of infection during the first 1000 days of a child’s life (WHO, 2014). In 2012, the World Health Assembly Resolution 65.6 endorsed a comprehensive implementation plan on maternal, infant, and young child nutrition. It specified six global nutrition targets for 2025, with the first target being a 40% reduction in the number of children under 5 years who are stunted (WHO, 2014). World hunger and undernutrition status According to the most recent FAO report, the total number of undernourished people in the world was estimated to be 815 million in 2016, up from 777 million in 2015 although still

4

1. World supply of food and the role of dairy protein

13 38 58 Sub-Saharan Africa Central and South Asia 70

307 East and South-East Asia West Asia and North Africa Latin America North America & Europe 212

FIG. 1.1 Undernourishment (severe food insecurity) in 2016, by region (millions). Data from FAO, IFAD, UNICEF, WFP, WHO, 2017. The State of Food Security and Nutrition in the World 2017. Building Resilience for Peace and Food Security. Food and Agriculture Organization of the United Nations, Rome, Italy.

down from about 900 million in 2000 (FAO et al., 2017). The largest numbers of undernourished people live in the developing countries. More than half live in just seven countries (Bangladesh, China, the Democratic Republic of the Congo, Ethiopia, India, Indonesia, and Pakistan) and over 40% live in China and India alone (data from FAO et al., 2017). The territory with the highest number of undernourished people is sub-Saharan Africa, with an estimated 307 million (Fig. 1.1). The food security situation has worsened in recent times, in particular in parts of sub-Saharan Africa, South-Eastern Asia, and Western Asia. Deteriorations have been observed most notably in situations of conflict and conflict combined with droughts or floods (FAO et al., 2017). Global hunger index The International Food Policy Research Institute has introduced the “Global Hunger Index” (GHI) tool to measure and track global hunger. The index combines three hunger indicators as follows: (1) the number of the undernourished as a proportion of the total population, (2) the proportion of underweight children under the age of 5 years, and (3) the mortality rate of children under the age of 5 years. The three indicators are assigned equal weights. On a 100-point scale, the higher is the value of the index, the worse is the incidence of hunger, implying that a score of 0 indicates no hunger and a score of 100 indicates the worst possible hunger—both of these extremes are, obviously, just notional and not observed in practice. Different hunger scenarios are defined with the help of the GHI. An index value of less than 4.9 indicates “low hunger”; values between 5 and 9.9 indicate “moderate hunger”; values between 10 and 19.9 indicate “serious hunger”; values between 20 and 29.9 indicate “alarming hunger”; values in excess of 30 indicate “extremely alarming hunger.”

5

The importance of protein in world nutrition

7.8

14.4 9.3

9.7

10

8.4

17.1

'17

16.6

'92 '00 '08

12.6

28.9 20.5 14.4

12.8

19.3

16.7

34.8

20

13.6

Under-five mortality rate Prevalence of wasting in children Prevalence of stunting in children Proportion of undernourished

43.5

29.4

30.9

34.9

46.3 35.2

38.2 25.7

21.8

30

29.9

GHI score

50 40

48.3

60

'17

'92 '00 '08

'17

0 '92 '00 '08 '17

World

'92 '00 '08

'17

South Asia

'92 '00 '08

'17

Africa south of the Sahara

Near East & North Africa

'92 '00 '08

'17

East & Southeast Asia

'92 '00 '08

Latin America Eastern Europe & Commonwealth & Caribbean of Independent States

FIG. 1.2 Global and regional trends of the Global Hunger Index: contribution of components in 1992, 2000, 2008, and 2017. From von Grebmer K., Bernstein J., Hossain N., Brown T., Prasai N., Yohannes Y., Patterson F., Sonntag A., Zimmermann S.-M., Towey O., Foley C., 2017. 2017 Global Hunger Index: The Inequalities of Hunger. International Food Policy Research Institute, Washington, DC. Reproduced with permission from the International Food Policy Research Institute, www. ifpri.org. The publication from which this figure originates can be found online at https://doi.org/10.2499/9780896292710.

The 2017 world GHI showed improvement over the 2000 value, decreasing by 27%. The countries of greatest concern (rated alarming or extremely alarming) are mostly in subSaharan Africa. At the regional level, South Asia and Africa south of the Sahara have the highest 2017 GHI scores, indicating serious levels of hunger. Fig. 1.2 graphs these values for the different regions.

The importance of protein in world nutrition Most reports on hunger and undernutrition are primarily focused on calories and rightly so. If a person does not get enough calories, their well-being will be compromised. However, calories are a necessary, but not sufficient, condition for good nutrition. Many nutrients and micronutrients, such as essential fatty acids, omega-3 lipids, and complex lipids, as well as vitamins and minerals, are also an important part of the diet, but are not the focus of this discussion. A further aspect of undernutrition occurs when people are not getting enough protein and particularly not enough of the dietary essential amino acids. This aspect of undernutrition and the role of dairy protein in meeting these needs are the focus of the rest of this chapter. Table 1.1 lists the recommended daily intake of protein and of the dietary essential amino acids. It is necessary to have an adequate intake of total protein and of all the essential amino acids to maintain health.

Protein and its composition and bioavailability The assessment of protein nutrition is more complex than that for calories, because proteins vary widely in terms of their composition and bioavailability.

6

1. World supply of food and the role of dairy protein

TABLE 1.1 Recommended daily intakes for adults of protein and dietary essential amino acids Dietary item

Recommended daily intake (mg/kg body weight)

Protein

800

Histidine

10

Isoleucine

20

Leucine

39

Lysine

30

Methionine + cystine

15

Phenylalanine + tyrosine

25

Threonine

15

Tryptophan

4

Valine

26

Data from WHO, 2007. Protein and Amino Acid Requirements in Human Nutrition: Report of a Joint FAO/WHO/UNU Expert Consultation. World Health Organization, Geneva, Switzerland.

Protein composition All proteins are composed of linear chains of amino acids, and each species of protein has its own defined amino acid sequence, which is determined by the genetics of the producing organism. Thus, the amino acid composition and, by implication, the amount of each essential amino acid in a given protein are defined. In practice, most food protein sources contain a complex mixture of proteins; nevertheless, the overall composition can be determined empirically and is generally quite constant over time and geographic origin. This has enabled the compositions of almost all the major food protein sources in terms of essential amino acids to be determined, and thus, intakes of dietary essential amino acids can be estimated from the knowledge of the types and amounts of food protein in a diet. In practice, it turns out that the intake of most essential amino acids in most diets is adequate, provided the total protein intake is adequate (although the timing of intake is also important—see later in this chapter). The exception to this is the dietary essential amino acid lysine. This is discussed in detail later in this chapter. Bioavailability Adequate protein supply is one aspect of protein nutrition. A further important aspect is bioavailability: getting the amino acids from the food structures in the gastrointestinal tract to the cells that need them throughout the body. In the adult gastrointestinal tract, proteins must be broken down to very small oligopeptides (at most di- or tripeptides) to be taken up and to single amino acids to enter most metabolic pathways. Thus, it is necessary for the protein in foods to be both accessible to digestive enzymes and broken down by the digestive enzymes in the stomach and small intestine in particular. Furthermore, it is necessary that the brokendown protein is able to be taken up into the bloodstream, where it can be redistributed to the

7

The importance of protein in world nutrition

tissues that need it. The efficiency of the digestion of most of the common food protein sources has been determined, using a range of different methods. Past methodology has largely been based on so-called fecal digestibility. This method is now known to be flawed, particularly with respect to foods with poor digestibility; however, it has been widely used and is the only method for which literature values for most common foods are available. For a full discussion on protein quality and nutritional requirements, the reader is referred to an FAO report on “Protein quality evaluation in human nutrition” (FAO, 2013). Using digestibility values, dietary intakes can be converted to an estimated uptake into the body for these proteins, noting that, as these values are derived from a flawed methodology, particularly in the case of plant-derived proteins, they represent an upper limit of their true bioavailability. The digestibilities of a range of dietary proteins are given in Fig. 1.3. Animal-derived proteins generally have good bioavailability and content of dietary essential amino acids, but many plant proteins are deficient in one or more dietary essential amino acids, and many are not efficiently digested, and therefore, the constituent amino acids are not highly bioavailable. As most western countries are characterized by a high-protein diet with a strong emphasis on animal-derived protein, protein nutrition is not generally a problem (although there may be some issues with the protein nutrition of the elderly, as discussed later in this chapter). Most developing countries are very dependent on plant protein as the main dietary source; that protein may be inadequate, because of poor digestibility and poor amino acid balance, particularly in the case of lysine.

World protein supply and its regional distribution Information on the amount of protein available per capita and by protein source in each country is available from the FAOSTAT database (http://faostat.fao.org/). In Fig. 1.4, we present the average protein availability for individual countries, along with their total 1.1 1 0.9 0.8 0.7 0.6 Oats

Beans

Rye

Sorghum

Millet

Nuts

Barley

Peas

Maize

Potatoes

Wheat

Soyabeans

Mutton & goat meat

Poultry meat

Demersal fish

Milk, whole

Rice

Eggs

Cheese

Bovine meat

0.5

FIG. 1.3 “True protein digestibility” of common food proteins. Data from FAO, 1970. Amino-Acid Content of Foods and Biological Data on Proteins. Food and Agriculture Organization of the United Nations, Rome, Italy, http://www.fao.org/ docrep/005/ac854t/ac854t00.htm (Accessed 10 February 2011).; WHO/FAO, 1991. Protein Quality Evaluation. Report of Joint FAO/WHO Expert Consultation. Food and Agriculture Organization, Rome, Italy.

8

1. World supply of food and the role of dairy protein

Protein supply (g/day/capita)

Population (Millions)

2500

2000

1500

1000

500

0

FIG. 1.4 Protein availability by population on a country average basis. Each bar represents the total population of countries with average per capita daily protein availability in bands of 5 g per capita per day. Protein figures for 2013 from FAOSTAT September 2018; 2013 population data from United Nations, 2018. World Population Prospects 2017. Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, New York, NY.

populations. This is presented as the total number of people who reside in countries with corresponding average protein availability for each incremental 5-g band. The figure is striking in having two clear peaks, one in the range 60–65 g of protein per capita per day and another in the range 95–100 g of protein per capita per day, with a considerable tail to the right. The first of these peaks represents nearly a third of the world’s population, including the populations of India and Indonesia. It is simple to calculate that, with a standard body weight of 70 kg and based on the dietary recommendations for protein requirements, a person will need 56-g protein/day to stay healthy. This band is 5–10 g of protein per capita per day more than that but will include a significant proportion of the population below this average. The band at 55–60 g of protein per capita per day is of greater concern as it is on the margin. It notably includes the countries of Bangladesh and the Philippines and a range of sub-Saharan African countries. This band is of great concern because, although the average availability figure is just above the minimum requirement, disparities of income and situation in these countries will mean that a large proportion of these people are not getting enough protein. Furthermore, these figures are for total protein supply only, with no correction for bioavailability. When the dietary pattern is corrected for the digestibility of the main protein components of the diet (from FAOSTAT), the situation is more serious, with more than 150 million people getting, on average, less than 55-g protein/day based on FAOSTAT figures for 2013. Although this is an improvement on the situation previously, there is still a large requirement to be met.

Plant and animal protein sources with a focus on dairy foods In the consideration of global sustainability, there is considerable debate over the merits of vegetarianism and eating only vegetable-origin foods. It is often estimated that the

The importance of protein in world nutrition

9

production of 1 kg of animal-origin food requires 10 kg of plant-origin food, leading on to the simplistic assumption that 10 times as many people could be fed from the same resources if everyone was vegetarian. For a full discussion of the subject, the reader is referred to Fairlie (2010). In the case of dairy produce and eggs, the situation is somewhat better than it is for meat, because the animal can continue producing throughout its adult lifetime. This leads to conversion ratios of about 4:1. In fact, the argument is much more complex, partly because of the role of animals in subsistence agriculture, largely eating food waste or processing residue, or grazing and browsing plant species that are not suitable for human consumption and partly because of the niche many animals occupy in developed agricultural economies, either grazing pastures intensively or being farmed on country that is not suitable for arable cultivation (see, e.g., Elferink et al., 2008). In “Livestock: On our plates or eating at our table? A new analysis of the feed/food debate” by the FAO, it is estimated that, although livestock consume 6 billion tonnes of feed each year—including one-third of global cereal production— 86% is composed of materials currently not eaten by humans (Mottet et al., 2017). In a recent analysis of the global food chain (Ritchie et al., 2018), it is estimated that 72 g of humandigestible protein per person per day is fed to livestock (all livestock, including monogastrics), but that meat and dairy (cows and buffaloes) on their own contribute more than half of this amount, that is, 37 g per person per day, to the human diet. In a simple analysis, we have calculated that, in the Canterbury Plains in New Zealand (the main wheat growing area), the yield per hectare per annum of protein from the wheat crop, processed to the form of white flour and its consequent baked products, is somewhat less that the yield of protein from milk that is produced over the same period in the same area. Thus, the efficiencies of production need to be considered in the context of what is the target of that production (there is no doubt that wheat produces the greater number of calories). Nevertheless, it must be recognized that the changing protein consumption patterns, involving more animal-based products, have significant implications for global land-use patterns, agriculture, agrifood industries, cereal prices, and the environment.

Growing global demand for animal proteins and implications The demand for animal protein foods is expected to increase to about double the present consumption by 2050, driven off population growth and by emerging middle classes in developing countries (FAO, 2006). As people get more money, one of the first priorities is better food, and this usually means animal protein foods. This was first described by Bennett (1941), who related comparative studies of the consumption of staple foods, leading to what has come to be known as “Bennett’s law,” the empirical generalization that there is an inverse relationship between the percentage of total calories derived from cereals and other staple foods and per capita income. This has since become generalized to mean a move away from carbohydrate-based foods to protein-based foods. A simple extrapolation from past increases in animal production indicates that we should be able to meet this demand if past rates of increase can be sustained (Boland et al., 2013). However, past increases have been based on bringing in new land for farm production, increases in efficiency through breeding gains, better livestock management and nutrition, and other factors of the Green Revolution covered earlier. Most of these options are reaching their limits or entering a phase of diminishing returns in developed economies, but there is

10

1. World supply of food and the role of dairy protein

considerable scope for increased efficiencies of production in developing economies (Mottet et al., 2017). The carbon footprint of livestock production is a further constraint, although the good news is that, as animal production has intensified, the carbon footprint has massively decreased. For example, Capper et al. (2009) have calculated that the carbon footprint for milk in the United States in 2007 was just 37% of that for the same milk in 1944. Nonetheless, past increases will not continue, and infinitum and new ways of sustainably meeting the increasing demand are needed.

The dietary essential amino acids in proteins Although there are nine dietary essential amino acids, it is rare for a diet with adequate overall protein intake to be deficient in any of them. The exception is lysine. Lysine may be an issue for two reasons. The first is that many staple protein sources, particularly the cereals, are deficient in lysine. The second is that lysine is chemically unstable under heating and undergoes a range of reactions when food is heated. The most important of these is the Maillard reaction, in which the side chains of the lysine residues in the protein cross-react with sugar molecules to produce glycosyl lysine side chains that are indigestible and thus no longer bioavailable. This reaction can occur under mild heating conditions and, under more extreme conditions, is responsible for much of the browning of food that occurs on cooking. Another reaction of importance for dairy products is the reaction with phosphoserine, leading to the formation of lysinoalanine, which is not bioavailable. This problem is mostly restricted to casein-containing products (mainly milk powders and caseinate), because of their high phosphoserine content (see Chapter 11 for a detailed discussion of this reaction); however, it is seldom a nutritional problem because dairy proteins are rich in lysine.

Identifying the countries deficient in dietary essential amino acids In an attempt to get an understanding of the dietary availability of the essential amino acids, countries with low intakes of protein were analyzed to determine the dietary essential amino acid content of the mix of protein sources for that country (from FAOSTAT), corrected for digestibility for each protein source. Because literature values were unavailable for some minor protein sources, a sensitivity test was performed, changing the digestibility figure from 1.0 to 0.8 for plant proteins and from 1.0 to 0.9 for animal proteins in these cases. As this change did not make a noticeable difference to the overall lysine bioavailability for the countries in question, the method was considered to be robust. The countries found to be lysine deficient are given in Table 1.2, together with information about the main dietary protein sources. It is noted that the number of countries and the degree of deficiency are considerably less than we have previously reported (Chatterjee et al., 2014). The countries that are lysine deficient show a clear pattern of low levels of consumption of animal protein and strong dependence on cereals for their protein.

Protein and dietary essential amino acid contents of food items The amino acid composition and particularly the lysine content of proteins are of particular concern for countries that tend to be protein deficient. Thus, the maintenance of an adequate

Demographic changes, aging populations, and the need for quality protein and dietary essential amino acids

11

TABLE 1.2 Countries with low levels of lysine in the average diet; data are for 2013 and from FAOSTAT Country

Lysine (%RDIa)

Animal protein in diet (%)

Main protein source(s)

Main protein source (%)

Liberia

82

22

Rice

50

Guinea-Bissau

90

20

Rice

43

Mozambique

93

14

Maize

25

Zimbabwe

93

24

Maize

40

Madagascar

99

21

Rice

48

a

RDI, recommended daily intake.

12.00 Animal protein

Lysine (g/100 g)

10.00

Plant protein

8.00 6.00 4.00 2.00 0.00

FIG. 1.5 Bioavailable lysine content in a range of food proteins (in g/100 g protein, corrected for digestibility).

intake of lysine, particularly in populations with a high dependence on cereals, requires attention. Fig. 1.5 indicates the levels of bioavailable lysine in a range of common dietary protein sources. Meat is clearly the best source of lysine but may not be a suitable dietary component for many because of cost, of utility—as meat can be used in only a number of food applications and of cultural restrictions. Dairy protein is also an excellent source of lysine and can be used in a wide range of food applications. The inclusion of supplementary dairy protein in the diet may offer an effective solution that is acceptable to vegetarians, price notwithstanding.

Demographic changes, aging populations, and the need for quality protein and dietary essential amino acids Recent global demographic trends indicate a steady increase in the number of people aged 60 years and over. The projection is for this population to more than triple from 600 million in

12

1. World supply of food and the role of dairy protein

2000 to over 2.1 billion in 2050 (United Nations, 2015b). As a consequence, in the more developed world, the fastest growing section of population is adults aged 80 years or over. This clearly presents unique challenges for healthcare, diets, and nutrition, as well as certain age-specific clinical conditions. Both the number and the proportion of older persons are growing in virtually all countries, and these trends are likely to continue worldwide (United Nations, 2015b). In 2015, one in eight people worldwide was aged 60 years or over. By 2030, older persons are projected to account for one in six people globally. By the middle of the 21st century, one in every five people will be aged 60 years or over. Fig. 1.6 shows the predicted percentage of elderly population in 2050 in major continents. By 2050, it is projected that there will be more than 1.3 billion elderly people (> 60 years) in Asia alone. Aging is a continuous, ongoing, and progressive process of damage accumulation. It is associated with a reduction in muscle mass and function and reduced physical activity. The loss of muscle mass with aging is known as sarcopenia. With the aging of the population globally, the prevalence of sarcopenia is likely to increase. Sarcopenia is accelerated by inadequate diet, mainly because of a lack of quality protein in optimal quantity and a lack of essential amino acids. The issue of the nutritional needs of the growing aging population, in terms of the role of dietary protein and essential amino acids with particular reference to sarcopenia, is described in more detail in the following sections. Overall, a strong case can be made that an aging population will require a substantially increased intake of protein, and of essential amino acids (particularly leucine), a demand that milk proteins are particularly well suited to meet.

Protein nutritional needs of the elderly The aging process is characterized by changes in body composition, with a progressive loss of muscle and bone mass, strength, and metabolic function. The loss of muscle with aging is the result of a chronic imbalance between muscle protein synthesis and breakdown. There are many causes of sarcopenia, and an understanding of the complex mechanism is evolving. FIG. 1.6 Percentage of elderly population in 2050

Chart title 13.2 122.7 220.3 200

Africa Asia

242

Europe Latin America Oceania 1293.7

North America

(numbers in millions). Data from United Nations, 2015b. World Population Ageing 2015 (ST/ESA/SER.A/ 390). Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, New York, NY.

Demographic changes, aging populations, and the need for quality protein and dietary essential amino acids

13

This degenerative loss of skeletal muscle occurs at a rate of 3%–8% per decade after the age of 30 years and accelerates with advancing age; chronic muscle loss is estimated to affect 30% of people older than 60 years and 50% of those older than 80 years (Katsanos et al., 2006; PaddonJones et al., 2008). With the aging of the population, the prevalence of sarcopenia and the resulting burden of disability are likely to increase. Strategies to prevent sarcopenia are, therefore, of considerable importance, and there is a need for public awareness, as simple health strategies can be effective. Researchers have identified two measures that can play a role in the fight against sarcopenia: diet and exercise. However, in the case of many elderly individuals, the ability to perform exercise is compromised because of disease and disability. In this case, a daily high-quality protein intake can be helpful to slow or prevent muscle protein loss. Different protein sources have been found to stimulate muscle protein synthesis to varying degrees. The most important factor is the amount of essential amino acids in the protein, in particular, leucine. Differences in the digestibility and the bioavailability of certain protein-rich foods may also influence muscle protein synthesis (Paddon-Jones et al., 2008). Currently, there is no agreement on whether dietary protein needs change with advancing age. For adults, the recommended dietary allowance (RDA) for protein is 0.8 g of protein per kilogram of body weight per day (WHO, 2007). It is recommended in the report of the FAO/ WHO/UNU expert consultation, published in 2007, that the essential amino acid requirement for elderly people should be the same as that for adults, as the current acceptable methodologies are not apposite for making a separate set of essential amino acid values for elderly people (WHO, 2007). A more recent FAO-sponsored expert consultation has failed to resolve this. One group maintains that: “The data based on the currently acceptable methodologies … are inadequate to make a separate recommendation for dietary IAA requirements in elderly people” (Pillai and Kurpad, 2012); another group advises that: “Dietary protein intake, and the resulting increased availability of plasma amino acids, stimulates muscle protein synthesis. If all other variables are controlled, increased muscle protein synthesis leads to improved muscle mass, strength and function over time. Increased muscle mass, strength and function are related to improved health outcomes in older individuals. Since adverse effects of reasonable increases in protein intake above the RDA of 0
8 g protein/kg/day have not been reported, it is reasonable to conclude that the optimal protein intake for an older individual is greater than the RDA” (Wolfe, 2012). It has been suggested in some studies that, instead of the RDA value, an intake of 1.0–1.5 g of protein per kilogram of body weight per day or about 15%–20% of the total caloric intake is essential to preserve a proper nitrogen balance in the healthy elderly (Morais et al., 2006; Wolfe et al., 2008). Many older people struggle to consume even the current recommended intake.

Role of essential amino acids in nutrition of the elderly It is recognized that essential amino acids are mainly responsible for the stimulation of muscle protein anabolism in the aged (Volpi et al., 2003). It is considered that 15 g of essential amino acids taken as a bolus is required for maximum stimulation of muscle protein synthesis (Wolfe, 2002). This indicates that the quality of protein is very important in the diet of the elderly.

14

1. World supply of food and the role of dairy protein

Preliminary data from a recent randomized controlled trial indicate that it is more important to ingest a sufficient amount (25–30 g) of high-quality protein with each meal rather than one large bolus, because more than 30 g in a single meal may not further stimulate muscle protein synthesis (Symons et al., 2009). It is also recognized in recent studies that the intake of whey protein brings beneficial effects to muscle protein anabolism in the elderly. Further, the ingestion of intact whey protein has been found to provide a greater anabolic benefit than the ingestion of the equivalent essential amino acids alone. Thus, whey protein may be more than just a simple source of essential amino acids with respect to providing a stimulus for enhancing muscle protein anabolism in the elderly (Katsanos et al., 2008). For a fuller discussion of the function of whey proteins and other milk proteins in human health, the reader is referred to Chapter 18 of this volume. There is general agreement that the essential amino acid leucine increases protein anabolism and decreases protein breakdown (Paddon-Jones and Rasmussen, 2009). Leucine-rich food sources include legumes such as soybeans and cowpea and animal products such as beef, fish, and particularly dairy proteins (whey protein). It is reported that amino acid supplements without adequate leucine do not stimulate protein synthesis (Rieu et al., 2007; Hayes and Cribb, 2008). Leucine has recently been acknowledged to be especially important as a signaling molecule and a building block for muscle. Rat studies show that leucine can directly stimulate muscle protein synthesis through increasing mRNA translation (Anthony et al., 2000). Insulin and leucine are anabolic stimuli for muscle, and both share a common pathway of action via activation of a kinase known as mTOR. mTOR is the main regulator of cell growth and acts by phosphorylating target proteins involved in mRNA translation. Because insulin sensitivity decreases with age, one possible mechanism by which amino acids (mainly leucine) might improve muscle mass is by providing another anabolic stimulus to activate the mTOR-controlled pathway (Gaffney-Stomberg et al., 2009; Casperson et al., 2012). No differences exist in the protein balance in the elderly relative to that in the young following the administration of either 30 g of beef protein or 15 g of essential amino acids as a bolus (Paddon-Jones et al., 2004). However, when 6.7 g of a mix of the dietary essential amino acids is given, the overall protein synthetic response is reduced in the elderly relative to the young (Katsanos et al., 2005). This anabolic resistance has been attributed to a decrease in leucine sensitivity and may be overcome by increasing the proportion of this amino acid in the diet. For example, when a 6.7-g bolus of dietary essential amino acids enriched with leucine (46% leucine compared with the 26% normally found in whey protein) was given to elderly individuals, protein synthesis was fully restored (Katsanos et al., 2006).

Global trade in proteins and the long term prospects, with a focus on dairy foods Global food consumption patterns have been changing in recent decades in several significant ways. Among them is the noticeable and continuing shift in favor of proteins, especially animal proteins. The global consumption of protein is forecast to grow significantly by 2050, although the extent of this growth will depend upon various assumptions (Henchion et al., 2017). Henchion et al. (2017) determined that consumption would increase 32%–78% by 2050. The growth is largely due to the rising incomes in the developing world, more particularly in

Global trade in proteins and the long term prospects, with a focus on dairy foods

15

some of the more populous countries such as China, Brazil, and, to a lesser extent, India. As only a few countries currently have surplus protein to export, the projected increase in its demand is likely to pose serious challenges to these countries and to the world in general. In overall world protein nutrition, milk products, representing about 10% of all protein consumption, are the third most important source of protein after cereals (40%) and meat (18%) (data for 2016 from FAOSTAT). When the low levels of lysine in cereals are taken into account (about one-third of that in dairy products), it is clear that milk protein plays a very important nutritional role in the world today.

The global dairy food scene: An overview Milk and other dairy products have always been among the major everyday food items in human consumption in many cultures. It is a particularly useful food for the large, and possibly growing, number of vegetarians and flexitarians around the world. Its value for both infants and the elderly is easily recognized. Apart from its consumption in liquid form, there are many other ways in which milk is transformed and consumed. Innovations to make new milk-based products available in the market keep occurring. With the growing world population and changing food habits, the production and the consumption of milk and other milk-based products have also been increasing over time. Over the five decades since 1961, world milk production more than doubled from 3.4 million tonnes to 7 million tonnes in 2009 (FAOSTAT) and 9.7 million tonnes in 2017 (IDF, 2018). Of this, 7 million tonnes or 82% is cows’ milk, with the rest from buffalo, goat, sheep, and camel. The pattern of the regional production of cows’ milk (IDF, 2018) reveals that, as of 2017, Asia’s share is the highest—with 30% of global production; Europe comes next, with 32%, followed by North America with 18%, South America with 9%, Africa with 5%, and Oceania with 4.5%. It is important to note that much of the milk produced is consumed in the country (or economic bloc, in the case of the EU) in which it is produced and that just over 11% of dairy production is involved in international trade (2017 figures; IDF, 2018). In this context, it is noted that about 50% of whole milk powder, 49% of skim milk powder, and only about 12% of cheese are traded internationally (IDF, 2018). In 2017, the major exporters of cheese were the EU-27, the United States, New Zealand, Belarus, and Australia, and the main importers were Russia, Japan, the United States, Saudi Arabia, Korea, Mexico, Australia, and China (IDF, 2018). There is, evidently, some intraindustry trade in cheese. Given its variety and established regional specialties, this is not difficult to understand. Five major exporters account for nearly 80% of world dairy trade in cows’ milk (IDF, 2018). In 2017, they were New Zealand (26%), the EU-28 (24%), the United States (14%), and Australia and Belarus (5% each). All of the exported “milk” consists of processed products, of which the main protein-containing products (in order of importance) are whole milk powder, skimmed milk powder, and cheese. Different countries dominate the export markets of these products (Table 1.3). One notable feature of the discussion earlier is perhaps the absence of the poorer developing countries among the major exporters and importers of such processed high value milk

16

1. World supply of food and the role of dairy protein

TABLE 1.3 Major dairy protein exporters, 2017 Country

Whole milk powder (000 tonnes)

Skim milk powder (000 tonnes)

Cheese (000 tonnes)

Total protein (calculated)a (000 tonnes)

EU-28 (A)

394

781

830

579

New Zealand

1351

404

344

565

The United States

27

606

341

304

Australia

55

158

172

112

Belarus

29

114

189

94

a

Excludes other products such as casein, whey products, and liquid and condensed milks. Values are based on protein of whole milk powder 25%, skim milk powder 35%, and cheese 25%. The cheese value is based on the bulk of traded cheese being cheddar. Data from IDF, 2018. World Dairy Situation 2018. Bulletin of the International Dairy Federation 494/2018. International Dairy Federation, Brussels, Belgium.

products as butter and cheese. One explanation for this may be that the consumption of these products is income sensitive. In some of these countries (e.g., Algeria and Mexico), the major purchase of dairy imports is by the government as part of a cost-effective strategy to support good nutrition. This presumption is further confirmed when the trade patterns with respect to milk powders are examined. Milk powders are usually reconstituted for consumption as liquid milk and the demand for them is likely to be less income sensitive. Among the major importers of whole milk powder are China, Algeria, Saudi Arabia, Sri Lanka, and Nigeria; the major exporters are New Zealand, the EU-28, and Uruguay. For skimmed milk powder, the major importing countries are Mexico, China, Indonesia, Algeria, and the Philippines, and the major exporters are the EU-28, the United States, New Zealand, and Australia. Whereas China features as a milk importer, India, perhaps surprisingly, is neither a major importer nor a major exporter of dairy products, although it has the largest bovine herd in the world. With the introduction around the mid-1960s of a system of dairy cooperatives under the umbrella of the National Dairy Development Board (NDDB), India’s dairy industry has achieved a remarkable transition. Set up in 1965, the NDDB oversaw the dairy cooperatives collecting the often-small marketable surplus milk from the small herds scattered around the villages and supplying the growing market for milk in the urban areas. This linking of the milk producers with the markets—both groups being scattered in locations and large in numbers—generated a fivefold growth in India’s milk production in three decades from the late 1960s, as domestic consumption of milk also rose steadily (Chatterjee, 1990; Brown, 2009). This transformation is all the more remarkable in that India’s dairy industry is built almost entirely on crop residues—wheat or rice straw, corn stalks, vegetable residues, and grass gathered from the roadside—a rather different protein production model. Although the consumption of dairy products is projected to grow as the standards of living improve in the developing world, some new developments, so far mainly in the more affluent countries, have also been creating additional demand for certain types of foods that are referred to as “specialty foods.” These include functional foods, defined as “food and drink

References

17

products making a specific health claim,” organic foods, and genetically modified foods. Although international trade in specialty foods is still relatively small and confined to a few countries, evidence suggests that it is growing rapidly (Chatterjee, 2012). Dairy products feature prominently among both functional foods and organic foods that are currently traded internationally; other animal protein products feature less. The market for these products is likely to grow over time as the rising affluence spreads globally. Resources, including land devoted to mostly export-oriented organic farming, for example, have also been growing, particularly in the developing countries of Asia, Africa, and Latin America.

Conclusions World hunger continues to be a major problem, although some of the drivers have changed, particularly the effects of climatic events and of conflict. Hunger has several dimensions, notably the need first and foremost for an adequate intake of calories; a close second is the need for adequate intakes of protein and of dietary essential amino acids. Protein nutrition is more complex than calories because all proteins are not equal: nutritional value depends on the type of protein and how it has been treated prior to consumption, as much as the amount of protein itself. Milk protein is a very high-quality protein, with a good supply of the dietary essential amino acids and high bioavailability. It can therefore be used to supplement poorer plant-derived proteins, such as cereal protein, to greatly improve the nutritional value of the combination. Milk production is growing globally, and the amounts of dairy products (and implicitly milk protein) traded internationally are also growing. Milk protein already accounts for 10% of the global food protein supply and makes a disproportionate contribution to global protein nutrition, based on its bioavailability and desirable composition. The future role of milk proteins in the global food protein economy continues to deserve special attention.

References Anderson, S.A., 1990. Core indicators of nutritional state for difficult-to-sample populations. J. Nutr. 120, 1557–1600. Anthony, J.C., Yoshizawa, F., Anthony, T.G., Vary, T.C., Jefferson, L.S., Kimball, S.R., 2000. Leucine stimulates translation initiation in skeletal muscle of postabsorptive rats via a rapamycin-sensitive pathway. J. Nutr. 130, 2413–2419. Bennett, M.K., 1941. Wheat in National Diets. vol. 18. Wheat Studies of the Food Research Institute, Stanford University, California, pp. 37–76. Boland, M.J., Rae, A.N., Vereijken, J.M., Meuwissen, M.P.M., Fischer, A.R.H., van Boekel, M.A.J.S., Rutherfurd, S.M., Gruppen, H., Moughan, P.J., Hendriks, W.H., 2013. The future supply of animal-derived protein for human consumption. Trends Food Sci. Technol. 29, 62–73. Brown, L.S., 2009. Plan B 4.0: Mobilising to Save Civilization. Earth Policy Institute, Washington, DC. http://www. earth-policy.org/books/pb4/pb4ch9_ss4. (Accessed 14 February 2013). Capper, J.L., Cady, R.A., Bauman, D.E., 2009. The environmental impact of dairy production: 1944 compared with 2007. J. Anim. Sci. 87, 2170–2177. Casperson, S.L., Sheffield-Moore, M., Hewlings, S.J., Paddon-Jones, D., 2012. Leucine supplementation chronically improves muscle protein synthesis in older adults consuming the RDA for protein. Clin. Nutr. 31, 512–519. Chatterjee, S., 1990. Aid, trade and rural development: a review of New Zealand’s assistance to Indian dairying. In: Doornbos, M., Nair, K.N. (Eds.), Resources, Institutions and Strategies: Operation Flood and Indian Dairying. Indo-Dutch Studies on Development Alternatives (Chapter 15). Sage Publications, New Delhi, India, pp. 319–338.

18

1. World supply of food and the role of dairy protein

Chatterjee, S., 2012. Changing global food consumption patterns: an economic perspective. In: Ghosh, D., Das, S., Bagchi, D., Smarta, R.B. (Eds.), Innovation in Healthy and Functional Foods (Chapter 9). CRC Press, Taylor and Francis Group, New York, NY, pp. 125–140. Chatterjee, S., Sarkar, A., Boland, M.J., 2014. The world supply of food and the role of dairy protein. In: Singh, H., Boland, M., Thompson, A.K. (Eds.), Milk Proteins: From Expression to Food, second ed. Academic Press, Waltham, MA, pp. 1–18. Elferink, E.V., Nonhebel, S., Moll, H.C., 2008. Feeding livestock food residue and the consequences for the environmental impact of meat. J. Clean. Prod. 16, 1227–1233. Fairlie, S., 2010. Meat: A Benign Extravagance. Chelsea Green Publishing, White River Junction, VT. FAO, 2000. The State of Food Insecurity in the World. Food and Agriculture Organization of the United Nations, Rome, Italy. FAO, 2006. World Agriculture: Towards 2030/2050, Interim Report. Food and Agriculture Organization of the United Nations, Rome, Italy. FAO, 2013. Protein Quality Evaluation in Human Nutrition. Report of an Expert Consultation. FAO Food and Nutrition Paper 92. Food and Agriculture Organization of the United Nations, Rome, Italy. FAO, 2018. World Food Situation: FAO Food Price Index. Food and Agriculture Organization of the United Nations, Rome, Italy. http://www.fao.org/worldfoodsituation/foodpricesindex/en/. (Accessed November 2018). FAO, IFAD, UNICEF, WFP, WHO, 2017. The State of Food Security and Nutrition in the World 2017. Building Resilience for Peace and Food Security. Food and Agriculture Organization of the United Nations, Rome, Italy. Gaffney-Stomberg, E., Insogna, K.L., Rodriguez, N.R., Kerstetter, J.E., 2009. Increasing dietary protein requirements in elderly people for optimal muscle and bone health. J. Am. Geriatr. Soc. 57, 1073–1079. Hayes, A., Cribb, P.J., 2008. Effect of whey protein isolate on strength, body composition and muscle hypertrophy during resistance training. Curr. Opin. Clin. Nutr. Metab. Care 11, 40–44. Henchion, M., Hayes, M., Mullen, A.M., Fenelon, M., Tiwari, B., 2017. Future protein supply and demand: strategies and factors influencing a sustainable equilibrium. Foods 6, 53. https://doi.org/10.3390/foods6070053. IDF, 2018. World Dairy Situation 2018. Bulletin of the International Dairy Federation 494/2018. International Dairy Federation, Brussels, Belgium. Katsanos, C.S., Kobayashi, H., Sheffield-Moore, M., Aarsland, A., Wolfe, R.R., 2005. Aging is associated with diminished accretion of muscle proteins after the ingestion of a small bolus of essential amino acids. Am. J. Clin. Nutr. 82, 1065–1073. Katsanos, C.S., Kobayashi, H., Sheffield-Moore, M., Aarsland, A., Wolfe, R.R., 2006. A high proportion of leucine is required for optimal stimulation of the rate of muscle protein synthesis by essential amino acids in the elderly. Am. J. Physiol.-Endocrinol. Metab. 291, E381–E387. Katsanos, C.S., Chinkes, D.L., Paddon-Jones, D., Zhang, X.J., Aarsland, A., Wolfe, R.R., 2008. Whey protein ingestion in elderly persons results in greater muscle protein accrual than ingestion of its constituent essential amino acid content. Nutr. Res. 28, 651–658. Morais, J.A., Chevalier, S., Gougeon, R., 2006. Protein turnover and requirements in the healthy and frail elderly. J. Nutr. Health Aging 10, 272–283. Mottet, A., de Haan, C., Falcucci, A., Tempio, G., Opip, C., Gerber, P., 2017. Livestock: on our plates or eating at our table? A new analysis of the feed/food debate. Global Food Secur. 14(1–8). Paddon-Jones, D., Rasmussen, B.B., 2009. Dietary protein recommendations and the prevention of sarcopenia. Curr. Opin. Clin. Nutr. Metab. Care 12, 86–90. Paddon-Jones, D., Sheffield-Moore, M., Zhang, X.J., Volpi, E., Wolf, S.E., Aarsland, A., Wolfe, R.R., 2004. Amino acid ingestion improves muscle protein synthesis in the young and elderly. Am. J. Physiol.-Endocrinol. Metab. 286, E321–E328. Paddon-Jones, D., Short, K.R., Campbell, W.W., Volpi, E., Wolfe, R.R., 2008. Role of dietary protein in the sarcopenia of aging. Am. J. Clin. Nutr. 87, 1562S–1566S. Pillai, R.R., Kurpad, A.V., 2012. Amino acid requirements in children and the elderly population. Br. J. Nutr. 108, S44–S49. Rieu, I., Balage, M., Sornet, C., Debras, E., Ripes, S., Rochon-Bonhomme, C., Pouyet, C., Grizard, J., Dardevet, D., 2007. Increased availability of leucine with leucine-rich whey proteins improves postprandial muscle protein synthesis in aging rats. Nutrition 23, 323–331. Ritchie, H., Reay, D.S., Higgins, P., 2018. Beyond calories: a holistic assessment of the global food system. Front. Sustain. Food Syst. 2, 57. https://doi.org/10.3389/fsufs.2018.00057.

Further reading

19

Symons, T.B., Sheffield-Moore, M., Wolfe, R.R., Paddon-Jones, D., 2009. A moderate serving of high-quality protein maximally stimulates skeletal muscle protein synthesis in young and elderly subjects. J. Am. Diet. Assoc. 109, 1582–1586. Trostle, R., 2008. Global Agricultural Supply and Demand: Factors Contributing to the Recent Increase in Food Commodity Prices. Economic Research Service, United States Department of Agriculture, Washington, DC. United Nations, 2015a. Transforming Our World: The 2030 Agenda for Sustainable Development. Report A/RES/70/1, United Nations, New York, NY. https://sustainabledevelopment.un.org/content/documents/21252030% 20Agenda%20for%20Sustainable%20Development%20web.pdf. (Accessed 12 December 2018). United Nations, 2015b. World Population Ageing 2015 (ST/ESA/SER.A/390). Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, New York, NY. Volpi, E., Kobayashi, H., Sheffield-Moore, M., Mittendorfer, B., Wolfe, R.R., 2003. Essential amino acids are primarily responsible for the amino acid stimulation of muscle protein anabolism in healthy elderly adults. Am. J. Clin. Nutr. 78, 250–258. von Grebmer, K., Ringler, C., Rosegrant, M.W., Olofinbiyi, T., Wiesmann, D., Fritschel, H., Badiane, O., Torero, M., Yohannes, Y., Thompson, J., von Oppein, C., Rahall, J., 2012. 2012 Global Hunger Index: The Challenge of Hunger: Ensuring Sustainable Food Security Under Land, Water, and Energy Stresses. International Food Policy Research Institute, Washington, DC. WHO, 2007. Protein and Amino Acid Requirements in Human Nutrition: Report of a Joint FAO/WHO/UNU Expert Consultation. World Health Organization, Geneva, Switzerland. WHO, 2014. WHO Global Nutrition Targets 2025: Stunting Policy Brief. http://www.who.int/nutrition/topics/ globaltargets_stunting_policybrief.pdf. (Accessed November 2018). Wolfe, R.R., 2002. Regulation of muscle protein by amino acids. J. Nutr. 132, 3219S–3224S. Wolfe, R.R., 2012. The role of dietary protein in optimizing muscle mass, function and health outcomes in older individuals. Br. J. Nutr. 108, S88–S93. Wolfe, R.R., Miller, S.L., Miller, K.B., 2008. Optimal protein intake in the elderly. Clin. Nutr. 27, 675–684.

Further reading FAO, 1970. Amino-Acid Content of Foods and Biological Data on Proteins. Food and Agriculture Organization of the United Nations, Rome, Italy. http://www.fao.org/docrep/005/ac854t/ac854t00.htm. (Accessed 10 February 2011). United Nations, 2018. World Population Prospects 2017. Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, New York, NY. von Grebmer, K., Bernstein, J., Hossain, N., Brown, T., Prasai, N., Yohannes, Y., Patterson, F., Sonntag, A., Zimmermann, S.-M., Towey, O., Foley, C., 2017. 2017 Global Hunger Index: The Inequalities of Hunger. International Food Policy Research Institute, Washington, DC. WHO/FAO, 1991. Protein Quality Evaluation. Report of Joint FAO/WHO Expert Consultation. Food and Agriculture Organization, Rome, Italy.

This page intentionally left blank

C H A P T E R

2 Milk proteins: An overview D.A. Goulding, P.F. Fox, J.A. O’Mahony School of Food and Nutritional Sciences, University College Cork, Cork, Ireland

Introduction Milk is a fluid that is secreted by the female of all mammals, of which there are about 4500 extant species (about 80% of mammalian species are extinct), primarily to meet the complete nutritional requirements of the neonate. The proteins of milk are essential in meeting many of the neonate’s nutritional requirements as they provide essential amino acids and amino groups for the biosynthesis of nonessential amino acids. Milk proteins and peptides, including immunoglobulins, enzymes, enzyme inhibitors, growth factors, hormones, and antibacterial agents, also perform most of milk’s physiological functions. The other requirements are for energy (supplied by lipids and lactose and, when in excess, by proteins), essential fatty acids, vitamins, inorganic elements, and water. Because the nutritional requirements of the neonate depend on its maturity at birth, its growth rate, and its energy requirements (which depend mainly on environmental temperature), the gross composition of milk shows large interspecies differences, which reflect these requirements (see Fox and McSweeney, 1998; Fuquay et al., 2011; McSweeney and Fox, 2013; Fox et al., 2015). Although milk is designed to be species specific, humans have consumed the milk of other mammalian species for at least 8000 years (Kindstedt, 2012). World milk production has increased rapidly in recent decades, rising from 500 million tons in 1983 to 811 million tons in 2017 (FAO, 2018). Assuming an average protein content of
3.5%, this equates to
28 million tons of milk protein. The world’s largest milk producers are India and the EU with
20% of global production each, followed by the United States, China, Pakistan, and Brazil. Dairy is an integral part of the human diet in many parts of the world, with more than 6 billion people worldwide consuming milk and milk products; the majority of these people live in developing countries (FAO, 2018). Of the milk constituents, the protein fraction represents the most “value-added” component. This is primarily due to an increasing acceptance that bovine milk proteins have superior nutritional and functional attributes compared with many other protein sources.

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00002-5

21

# 2020 Elsevier Inc. All rights reserved.

22

2. Milk proteins: An overview

Consequently, many types of milk protein ingredients are produced globally to satisfy market demands for a variety of industrial applications. The various milk protein ingredient types differ in nutrient composition, protein profile, and functional properties, facilitating a diverse range of industrial applications. The demand for milk protein ingredients will probably continue into the future because of constant developments in global socioeconomic conditions and large increases in world population (Lagrange et al., 2015). By 2050, the FAO predicts that the world’s population will increase to 9.8 billion (United Nations, 2017). This increase in population will be coupled with the need for a 70% increase in food production to meet the added nutritional demands of the global population (Lagrange et al., 2015). The latest OECD/FAO report projects a 22% increase in global milk production by 2027, compared with the 2015–17 period, with most of that growth being in developing countries (OECD/FAO, 2018). The chemistry and the physicochemical properties of milk have been studied for about 200 years and are now understood in considerable detail, as described in a voluminous literature. This chapter aims to provide an overview of milk proteins. The main chemical and physicochemical properties of the individual proteins are documented, followed by a summary of the analytical methods used to study milk proteins and an overview of the various milk protein ingredients available commercially throughout the world. Numerous textbooks and review articles are cited to assist the reader in identifying further reading material.

Bovine milk composition Milk is a very complex fluid containing several hundred molecular species (several thousand, if all triglycerides are counted individually). The principal constituents are water, lipids, sugar (lactose), and proteins. In addition, there are numerous minor constituents, mostly at trace levels (e.g., minerals, vitamins, hormones, enzymes, and miscellaneous compounds). Reflecting mainly the nutritional and physiological requirements of the neonate, the composition of milk and even the profile of the constituents therein change markedly during lactation. The changes are most marked during the first few days postpartum, especially in the immunoglobulin fraction of proteins. The composition of milk remains relatively constant during midlactation but changes considerably in late lactation, reflecting the involution of the mammary gland tissue and the greater influx of blood constituents. The main constituents of bovine milk and their relative concentrations are listed in Table 2.1. The nonprotein constituents of milk are summarized here, and the protein system of milk is described in later sections.

Water Water is the principal constituent in the milk of most species. In addition to meeting the requirement of the neonate for water, the water in milk serves as a solvent for milk salts, lactose, and proteins and affects their properties and stability. It controls the rate of many reactions (e.g., Maillard browning, lipid oxidation, enzyme activity, and microbial growth), thus affecting the stability of milk and milk products. For further detail on the chemistry of water, its measurement, and its significance in foods, including dairy products, the reader is referred to Duckworth (1975), Rockland and Beuchat (1987), Fennema (1996), Roos (1997, 2011), Franks (2000), Chieh (2006), McSweeney and Fox (2009), and Simatos et al. (2009, 2011).

23

Bovine milk composition

TABLE 2.1 Gross composition of bovine milk (Fox et al., 2015) Component

%

Total solids

12.7

Fat

3.7

Protein

3.5

Casein

2.8

Whey proteins

0.7

Lactose

4.8

Ash

0.7

Lipids Historically, the fat of milk was regarded as its most valuable constituent, and until recently, milk was valued largely, or totally, on its fat content. Milk lipids are very complex chemically and exist as a rather unique emulsion. Milk lipids are commonly divided into three classes: • Neutral lipids: these are esters of glycerol and one, two, or three fatty acids for mono-, di-, and triglycerides, respectively. Neutral lipids are by far the dominant class of lipids in all foods and tissues, representing 98.5% of the total milk lipids. • Polar lipids (a complex mixture of fatty acid esters of glycerol or sphingosine): many contain phosphoric acid, a nitrogen-containing compound (choline, ethanolamine, or serine), or a sugar/oligosaccharide. Although present at low levels (
1% of the total milk lipids), the polar lipids play critical roles in milk and dairy products. They are very good natural emulsifiers and are concentrated in the milk fat globule membrane, which maintains the milk lipids as discrete globules and ensures their physical and biochemical stability. • Miscellaneous lipids: a heterogeneous group of compounds, which are unrelated chemically to each other or to neutral or polar lipids. This group includes cholesterol; carotenoids; and the fat-soluble vitamins, A, D, E, and K. The carotenoids are important for two reasons. They are natural pigments (yellow, orange, and red) and are responsible for the color of butter and cheese; some consumers prefer highly colored cheese, a color that is achieved by adding a carotenoid-containing extract from annatto beans. In addition, some carotenoids are converted to vitamin A in the liver. Milk lipids are chemically similar to all other lipids but contain a very wide range of fatty acids (up to 400 fatty acids have been reported in milk lipids, although most of these are present at trace levels). The milk lipids of ruminants are unique in that they are the only natural lipids that contain butyric (butanoic) acid (C4:0); they also contain substantial amounts of medium-chain fatty acids [hexanoic (C6:0), octanoic (C8:0), and decanoic (C10:0)], the only other sources of which are coconut oil and palm kernel oil. The short- and medium-chain fatty acids are water soluble and volatile and have a strong aroma and taste. Ruminant milk fats contain low levels of

24

2. Milk proteins: An overview

polyunsaturated fatty acids (PUFAs) because PUFAs in the diet are hydrogenated by bacteria in the rumen. Biohydrogenation can be prevented by encapsulating dietary PUFAs or PUFA-rich sources in cross-linked protein or cross-linked crushed oilseeds. Incomplete biohydrogenation by the rumen bacterium, Butyrivibrio fibrisolvens, results in the formation of conjugated linoleic acid (also called rumenic acid), which has potent anticarcinogenic properties. Eight isomers of conjugated linoleic acid are possible but cis-9, trans-11 is the most biologically active. In milk, the lipids are dispersed in the milk serum (specific gravity, 1.036) as globules with a diameter in the range from 6%, the lactose crystallizes as the α hydrate, the crystals of which form interlocking masses and clumps, which may render the powder unusable if very extensive; that is, inadequately crystallized lactose-rich powders are hygroscopic. • Among sugars, lactose has a low level of sweetness; it is only about 16% as sweet as sucrose at 1% in solution and hence has limited value as a sweetening agent, the principal application of sugars in foods. However, it is a useful bulking agent when excessive sweetness is undesirable. • Lactose is important in the manufacture of fermented dairy products, where it serves as a carbon source for lactic acid bacteria, which produce lactic acid. Mammals cannot absorb disaccharides from the small intestine, where they are hydrolyzed to monosaccharides, which are absorbed. Lactose is hydrolyzed by β-galactosidase (lactase), which is secreted by cells in the brush border of the small intestine. The young of most mammalian species secrete an adequate level of lactase, but as the animal ages, lactase secretion declines and eventually becomes inadequate to hydrolyze undigested lactose, which enters the large intestine, causing an influx of water and resulting in diarrhea, and is metabolized by bacteria with the production of gas, which causes cramps and flatulence. In humans, this may occur at 8–10 years of age and causes many individuals to exclude milk from the diet. The problems may be avoided by prehydrolyzing the lactose using exogenous β-galactosidase (see Mahoney, 1997; Rehman, 2009). The frequency and intensity of lactose intolerance/malabsorption varies widely among populations from
100% in southeast Asia to
5% in northwest Europe (Mustapha et al., 1997; Ingram and Swallow, 2009; Mattar et al., 2012; Storhaug et al., 2017; Ugidos-Rodrı´guez et al., 2018). Lactose has been reviewed by Whittier (1925, 1944), Weisberg (1954), Zadow (1984, 1992), Fox (1985, 1997), Fox and McSweeney (1998), McSweeney and Fox (2009), Fox (2011), Paterson (2011), Schuck (2011b), Huppertz and Gazi (2016), and Chen and Ganzle (2017).

26

2. Milk proteins: An overview

Oligosaccharides In addition to lactose, bovine milk contains other free saccharides, mainly oligosaccharides. The concentration of oligosaccharides is higher in colostrum than in milk. General reviews on the oligosaccharides in milk include Newburg and Newbauer (1995), Mehra and Kelly (2006), Urashima et al. (2001, 2009, 2011), and Oliveira et al. (2015). Almost all of the oligosaccharides have lactose at the reducing end, contain three to eight monosaccharides, may be linear or branched, and contain either or both of two unusual monosaccharides, fucose (a 6-deoxyhexose) and N-acetylneuraminic acid. The oligosaccharides are synthesized in the mammary gland, catalyzed by special transferases that transfer galactosyl, sialyl, N-acetylglucosaminyl, or fucosyl residues from nucleotide sugars to the core structures. Bovine milk contains relatively low levels of oligosaccharides, which have been characterized (see Urashima et al., 2001, 2009, 2011; Liu et al., 2017). The significance of oligosaccharides is not entirely clear, but the following aspects are generally deemed to be important: for any particular level of energy, they have a smaller impact on osmotic pressure than smaller saccharides, they are not hydrolyzed by β-galactosidase, and fucosidase or neuraminidase is not secreted in the intestine. Hence, oligosaccharides are not hydrolyzed and absorbed in the gastrointestinal tract and function as soluble fiber and prebiotics, which affect the microflora of the large intestine. It is claimed that they prevent the adhesion of pathogenic bacteria in the intestine. Galactose and, especially, N-acetylneuraminic acid are important for the synthesis of glycolipids and glycoproteins, which play a role in brain development; hence, it has been suggested that oligosaccharides are important for brain development (see Kunz and Rudloff, 2006). In addition to lactose and free oligosaccharides, the milk of all species examined contains small amounts of monosaccharides; some milk proteins, especially κ-casein, are glycosylated, and there are low levels of highly glycosylated glycoproteins, especially mucins, and glycolipids in the MFGM. There is considerable interest in the development of oligosaccharide-enriched ingredients from bovine milk (O’Mahony and Tuohy, 2013), primarily for infant formula applications, because of the demonstrated bioactive functionality of these compounds in humans (Kunz et al., 2000; Barile et al., 2011; Zivkovic and Barile, 2011; Dallas et al., 2014).

Milk salts In milk, there are salts of sodium, potassium, chloride, sulfate, phosphate, citrate, calcium, and magnesium. Milk also contains 20–25 elements at very low or trace levels. These microelements are very important from a nutritional viewpoint; many, for example, Zn, Fe, Mo, Cu, Ca, Se, and Mg, are present in enzymes, many of which are concentrated in the MFGM; some microelements, for example, Fe and Cu, are very potent lipid prooxidants. Although the salts are relatively minor constituents of milk, they are critically important for many technological and nutritional properties of milk. Some of the salts in milk are fully soluble, but others, especially calcium phosphate, exceed their solubility under the conditions in milk and occur partly in the colloidal state, associated with the casein micelles; these salts are referred to as colloidal calcium phosphate, although some magnesium, citrate, and traces of other elements are also present in the micelles.

Milk protein system

27

The distribution of species between the soluble and colloidal phases is markedly affected by pH and temperature. The solubility and the ionization status of many of the principal ionic species are interrelated, especially H+, Ca2+, PO4 3 , and citrate3. These relationships have major effects on the stability of the casein system and consequently on the processing properties of milk. The status of various species in milk can be modified by adding certain salts to milk; for example, [Ca2+] is reduced by adding PO4 3 or citrate3; the addition of CaCl2 affects the distribution and ionization status of calcium and phosphate and the pH of milk. There are substantial changes in the concentrations of the macroelements in milk during lactation, especially at the beginning and end of lactation and during mastitic infection. Changes in the concentration of some of the salts in milk, especially calcium phosphate and citrate, have major effects on the physicochemical properties of the casein system and on the processability of milk, especially rennet coagulability and related properties and heat stability. Reviews on the chemistry of milk salts include Pyne (1962), Holt (1985, 1997, 2003), Lucey and Fox (1993), Holt et al. (1996, 1998), Lucey and Horne (2009), Choi et al. (2011), Holt and Carver (2012), and Fox et al. (2015). Reviews on the nutritional significance of milk salts include Flynn and Power (1985), Flynn and Cashman (1997), Cashman (2003a,b, 2006), and Hunt and Nielsen (2009).

Milk vitamins Milk contains all the vitamins in sufficient quantities to enable normal maintenance and growth of the neonate. Bovine milk is a very significant source of some vitamins, especially biotin (B7), riboflavin (B2), and cobalamin (B12), in the adult human diet. For general information on the vitamins and for specific aspects in relation to milk and dairy products, including stability during processing and storage, the reader is referred to Roginski et al. (2003), Morrissey and Hill (2009), Nohr and Biesalski (2009), Fuquay et al. (2011), Amador-Espejo et al. (2015), and Sharabi et al. (2018). In addition to their nutritional significance, four vitamins are significant for other reasons: vitamin A (retinol) and carotenoids are responsible for the yellow-orange color of fat-containing products made from cows’ milk; vitamin E (tocopherols) is a potent antioxidant; vitamin C (ascorbic acid) is an antioxidant or prooxidant, depending on its concentration; and vitamin B2 (riboflavin), which is greenish yellow, is responsible for the color of whey or ultrafiltration permeate, cocrystallizes with lactose and is responsible for its yellowish color, which may be removed by recrystallization or bleached by oxidation, and acts as a photocatalyst in the development of light-oxidized flavor in milk, which is due to the oxidation of methionine.

Milk protein system Milk proteins The properties of milk and most dairy products are affected more by the proteins they contain than by any other constituent. The milk proteins have many unique properties; because of this and their technological importance, they have been studied extensively and are probably the best characterized food protein system. Research on milk proteins dates from the early 19th century. Pioneering work was reported by J.J. Berzelius in 1814; by H. Schubler in

28

2. Milk proteins: An overview

1818, on the physicochemical status of the milk proteins; and by H. Braconnot in 1830, who published the first paper in which the word casein was used. A method for the preparation of protein from milk by acid precipitation was described in 1938 by J.G. Mulder, who coined the term “protein” (“primary” or “of first rank”). The acid-precipitated protein was referred to as casein (some early authors called acid-precipitated milk protein caseinogen, which was converted by rennet to casein, which coagulated in the presence of Ca2+) (van Slyke and Bosworth, 1913); this situation is analogous to the conversion of fibrinogen in blood by thrombin to fibrin, which coagulates in the presence of Ca2+; later, the term “casein” was universally adopted as the English word for the pH 4.6 insoluble protein in milk. The method for acid (isoelectric) precipitation of casein was refined by O. Hammarsten in 1883–85; consequently, isoelectric casein is frequently referred to as “casein nach Hammarsten.” An improved method for the isolation of casein was published by L.L. van Slyke and J.C. Baker in 1918. As early as 1846, J.E. Schlossberger reported the separation of casein into two fractions (see Woodward, 1976), and in 1880, A. Danilewsky and P. Radenhausen suggested that isoelectric casein is heterogeneous (see Hammarsten, 1883; Lindqvist, 1963), but Hammarsten (1883) maintained that properly produced casein is homogeneous. Based on differential solubility in ethanol-HCl solutions, evidence began to emerge from the work of T.B. Osborne and A.J. Wakeman in 1918 and of K. Linderstrøm-Lang and collaborators during the period 1925–29 that isoelectric casein is heterogeneous. Heterogeneity of casein was confirmed by K.O. Pederson in 1936, using analytical ultracentrifugation, and by O. Mellander in 1939, using free-boundary electrophoresis (see McMeekin, 1970, for references to the early literature). The liquid whey remaining after isoelectric precipitation of casein from skim or whole milk is a dilute solution of proteins (whey or serum proteins;
0.7% in bovine milk), lactose, organic and inorganic salts, vitamins, and several constituents at trace levels. By salting-out with MgSO4, the whey proteins were fractionated by J. Sebelein, in 1885, into soluble (albumin) and insoluble (globulin) fractions. According to McMeekin (1970), A. Wichmann, in 1899, crystallized a protein from the albumin fraction of whey by the addition of (NH4)2SO4 and acidification, a technique that is used to crystallize blood serum albumin and ovalbumin. Using the techniques available at that time, the whey proteins were found to be generally similar to the corresponding fractions of blood proteins and were considered to have passed directly from blood to milk; consequently, the whey proteins attracted little research effort until the 1930s. In addition to the caseins and whey proteins, milk contains two other groups of proteinaceous materials, proteose peptones and nonprotein nitrogen (NPN), which were recognized in 1938 by S.J. Rowland, who observed that, after heating milk at 95°C for 10 min, the whey proteins coprecipitated with the caseins on acidification to pH 4.6. When the pH 4.6 soluble fraction of heated milk was made to a 12% trichloroacetic acid (TCA) solution, some nitrogenous compounds, which were designated “proteose peptone,” precipitated; nitrogenous compounds that remained soluble in 12% TCA were designated NPN (see Fig. 2.1). A modified version of Rowland’s scheme is commonly used to quantify the principal nitrogenous groups in milk (Aschaffenburg and Drewry, 1959). Thus, by 1938, the complexity of the milk protein system had been described, that is, caseins, lactalbumin, lactoglobulin, proteose peptones, and NPN, which represent approximately 78%, 12%, 5%, 2%, and 3%, respectively, of the nitrogen in bovine milk. However,

29

Milk protein system

Total nitrogen/crude protein

(–) Nonprotein nitrogen Material soluble in 12% TCA

True protein

Casein

Whey protein

Minor proteins

FIG. 2.1 Nitrogen fractions in milk.

knowledge of the milk protein system was rudimentary and vague at this stage. Advancement of the knowledge on the chemistry of milk proteins during the 20th and 21st centuries can be followed through the progression of textbooks and reviews on dairy chemistry (see Fox, 1982, 1992, 2003; Fox and McSweeney, 2003; O’Mahony and Fox, 2013; Fox et al., 2015; McSweeney and O’Mahony, 2015).

Nomenclature of milk proteins During the period of greatest activity on the fractionation of casein (1950–70), several casein (and whey protein) fractions that either were similar to proteins already isolated and named or were artifacts of the isolation procedure were prepared. To standardize the nomenclature of the milk proteins, the American Dairy Science Association established a Nomenclature Committee in 1955, which has published seven reports, the most recent of which is Farrell et al. (2004). In addition to standardizing the nomenclature of the milk proteins, the characteristics of the principal milk proteins are summarized in these reports.

Interspecies comparison of milk proteins Of the 4500 species of mammal, the milk of only about 180 species has been analyzed, and of these, the data for only about 50 species are to be considered reliable (a sufficient number of properly taken samples, representative sampling, and adequate coverage of the lactation period). Milks from the commercially important species, cattle, goat, sheep, buffalo, yak, horse, camel, and pig, are quite well characterized; human milk is also reasonably well characterized, as is that of experimental laboratory animals, especially rats and mice. Reviews on nonbovine milks include general (Evans, 1959; Jenness and Sloan, 1970; Park et al., 2006a;

30

2. Milk proteins: An overview

Fuquay et al., 2011; Medhammar et al., 2011; Tsakalidou and Papadimitriou, 2016); buffalo (Laxminarayana and Dastur, 1968), goat (Parkash and Jenness, 1968; Haenlein, 1980; Jenness, 1980; Zenebe et al., 2014), sheep (Bencini and Pulina, 1997; Balthazar et al., 2017), sheep and goats (IDF, 1996; Jandal, 1996; Park et al., 2007), camel (Rao et al., 1970; Farah, 1993; Yadav et al., 2015; Hailu et al., 2016; Ipsen, 2017), horse (Doreau and Boulot, 1989; Solaroli et al., 1993; Doreau, 1994; Park et al., 2006b; Uniacke-Lowe and Fox, 2012; Pieszka et al., 2016), human (Atkinson and Lonnerdal, 1989; Jensen, 1989, 1995), sow (Verstegen et al., 1998), donkey (Madhusudan et al., 2017), and yak (Ma et al., 2017b). The books edited by Park et al. (2006a), Park and Haenlein (2013), and Fuquay et al. (2011) are particularly useful sources of information on the milk of nonbovine mammals; they include chapters on goat, sheep, buffalo, mare, camel, yak, reindeer, sow, llama, minor species (moose, musk ox, caribou, alpaca, ass, elk, seals, sea lion, and polar bear), and human. The concentrations of the principal constituents vary widely among species—lipids, 2%– 55%; proteins, 1%–20%; and lactose, 0%–10%—reflecting mainly the energy requirements (lipids and lactose) and growth rate (mainly proteins) of the neonate. The concentrations of the minor constituents also vary very widely (e.g., the concentrations of lysozyme and lactoferrin in equine, human, and bovine milks). Also, within any species, the composition of the milk varies among individual animals, between breeds, with the stage of lactation, feed, and health of the animal, and many other factors. The protein content of milk varies widely depending on the species, ranging from
1% to
20%. The protein content reflects the growth rate of the neonate of the species, that is, its requirements for essential amino acids. The milks of all species for which data are available contain two groups of protein, caseins and whey proteins, but the ratio of these varies widely. Both groups show genus- and even species-specific characteristics, which presumably reflect some unique nutritional or physiological requirements of the neonate of the species. Interestingly and perhaps significantly, of the milks that have been characterized, human and bovine milks are more or less at opposite ends of the spectrum. Publications on the interspecies comparisons of milk proteins include Woodward (1976), Jenness (1973, 1979, 1982), Ginger and Grigor (1999), Martin et al. (2003, 2013a), and El-Hatmi et al. (2015). Publications on milk proteins of individual species include buffalo (Addeo et al., 1977; Buffoni et al., 2011), goat (Trujillo et al., 1997, 2000; Olumee-Shabon and Boehmer, 2018), sheep (Amigo et al., 2000), camel (Ochirkhuyag et al., 1997; Kappeler et al., 1998), yak (Ochirkhuyag et al., 1997), horse (Ochirkhuyag et al., 2000; Park et al., 2006b; Uniacke-Lowe et al., 2010; Uniacke-Lowe and Fox, 2011), and sow (Gallagher et al., 1997). Table 2.2 provides a summary of the interspecies differences in milk protein profiles and concentrations. There is considerably more and better information on the interspecies comparison of individual milk proteins than on the overall milk composition, probably because only one sample of milk from one animal is sufficient to yield a particular protein for characterization. The two principal milk-specific whey proteins, α-lactalbumin and β-lactoglobulin, from quite a wide range of species have been characterized and, in general, show a high degree of homology (see Brew, 2003, 2013; Sawyer, 2003, 2013). The caseins show much greater interspecies diversity, especially in the α-casein fraction; most of the species that have been studied contain a protein with an electrophoretic mobility similar to that of bovine β-casein (see Fig. 2.2), but the β-caseins that have been sequenced show a low level of homology (Martin et al., 2003, 2013a). Sheep’s milk is used mainly for cheese production, with small amounts used for the

TABLE 2.2 Protein profile (g L1) of the milk of different species (Crowley, 2016) Human

Bovine

Caprine

Ovine

Buffalo

Equine

Camel

Yak

Reindeer

Total casein

4.8

2.4–4.2

24.6–28

23.3–46.3

41.8–46

32–40

9.4–13.6

22.1–26.0

34.3–45.8

70–80

αs1-Casein

1.74

0.77

8.0–10.7

0.0–13.0

15.4–22.1

8.9

2.4

n.d.

9.3–13.1

n.d.

αs2-Casein

0.48

–

2.8–3.4

2.3–11.6

n.d.

5.1

0.2

n.d.

3.6–6.5

n.d.

β-Casein

1.62

3.87

8.6–9.3

0.0–29.6

15.6–17.6

12.6–20.9

10.66

n.d.

15.0–20.6

n.d.

κ-Casein

0.6

0.14

2.3–3.3

2.8–13.4

3.2–4.3

4.1–5.4

0.24

–

4.9–8.5

n.d.

γ-Casein

0.36

–

0.8

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

Total whey protein

7.2

6.2–8.3

5.5–7.0

3.7–7.0

10.2–11

6

7.4–9.1

5.9–8.1

n.d.

13.4

β-Lactoglobulin

–

–

3.2–3.3

1.5–5.0

6.5–8.5

3.9

2.55

–

3.4–10.1

n.d.

α-Lactalbumin

1.4–2.3

1.9–3.4

1.2–1.3

0.7–2.3

1.0–1.9

1.4

2.37

0.8–3.5

0.2–1.7

n.d.

Serum albumin

n.d.

0.4–0.5

0.3–0.4

n.d.

0.4–0.6

0.29

0.37

7–11.9

0.2–3.1

n.d.

Proteose peptone

n.d.

0.8–1.2

n.d.

n.d.

3.31

n.d.

n.d.

n.d.

n.d.

Lactoferrin

0.05

0.02–0.5

0.02–0.2

0.03–3.4

0.1–2.0

0.02–7.28

n.d.

n.d.

(120–152)  106

0.5–1.33

(60–1350)  106

n.d.

n.d.

1.5–2.0

0.8 6

100  10

6

Lysozyme

n.d.

0.1–0.9

(70–600)  106

250  10

Osteopontin

0.01

0.138

0.018

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

Immunoglobulins

n.d.

1.0–1.3

0.5–1.0

n.d.

0.7

10.66

1.63

n.d.

n.d.

n.d.

IgG

n.d.

0.03

0.15–0.8

0.1–0.4

n.d.

0.37–1.34

0.38

n.d.

n.d.

n.d.

IgA

n.d.

0.96

0.05–0.14

0.03–0.08

n.d.

0.01–0.04

0.47

n.d.

n.d.

n.d.

IgM

n.d.

0.02

0.04–0.1

0.01–0.04

n.d.

0.04–1.91

0.03

n.d.

n.d.

n.d.

CN:WP ratio

40:60

29:71–33:67

82:18

78:22

76:24

82:18

52:48

73:27–76:24

82:18

80:20–83:17

Milk protein system

Infant formula

n.d., no data.

31

32

2. Milk proteins: An overview

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

FIG. 2.2 Urea-polyacrylamide gel electrophoretograms of milk from 15 species. Lanes: 1, bovine; 2, camel; 3, equine; 4, asinine; 5, human; 6, rhinoceros; 7, caprine; 8, ovine; 9, Asian elephant; 10, African elephant; 11, vervet monkey; 12, macaque monkey; 13, rat; 14, canine; 15, porcine. From O’Mahony, J.A., Fox, P.F., 2014. Milk: an overview. In: Milk Proteins. Academic Press, pp. 19–73.

production of fermented milks; hence, the coagulation and gel-forming properties of ovine milk are particularly important. The αs1-casein of caprine milk is very heterogeneous; not only do the properties of the variants differ, but also the concentration of αs1-casein varies from 0% to 26% of the total casein; consequently, the total protein content varies considerably, and this, in turn, has major effects on the rennet-induced coagulation properties of ovine milk and caprine milk and on the yield and quality of cheeses produced therefrom (Amigo et al., 2000; Clark and Sherbon, 2000a,b; Martin et al., 2013a,b). Human β-casein occurs in multiphosphorylated forms (0–5 mol P per mol protein; see Atkinson and Lonnerdal, 1989), as does equine β-casein (Ochirkhuyag et al., 2000; Girardet et al., 2006; Uniacke-Lowe et al., 2013). Considering the critical role played by κ-casein, it would be expected that all casein systems contain this protein, but Ochirkhuyag et al. (2000) failed to identify κ-casein in mare’s milk and suggested that the micelle-stabilizing role was played by β-casein with a zero or low level of phosphorylation; more recent work has shown that equine milk does contain a low level of κ-casein (Iametti et al., 2001; Egito et al., 2002; Uniacke-Lowe et al., 2013). Human κ-casein is very highly glycosylated, containing 40%–60% carbohydrate (compared with
10% in bovine κ-casein). The αs-casein fraction differs markedly between species; human milk probably lacks an αs-casein (Darragh and Lonnerdal, 2011), and the α-casein fractions in horse milk and donkey milk are very

Casein

33

heterogeneous. Urea-polyacrylamide gel electrophoretic analysis of the protein profile of milk from 15 species is shown in Fig. 2.2. There are very considerable interspecies differences in the minor proteins of milk. The milk of those species that have been studied in sufficient depth contains approximately the same profile of minor proteins, but there are very marked quantitative differences. Most of the minor proteins in milk have some biological or physiological function, and the quantitative interspecies differences presumably reflect the requirements of the neonate of the species. Greatest interspecies differences, in some cases 4000-fold, seem to occur in the indigenous enzymes (Fox and Kelly, 2006a,b). In the milk of all species, the caseins exist as micelles (at least the milks appear white), the properties of which have been studied in many nonbovine species: caprine (Ono and Creamer, 1986; Ono et al., 1989), ovine (Ono et al., 1989), buffalo (Patel and Mistry, 1997), camel (Attia et al., 2000), mare (Welsch et al., 1988; Ono et al., 1989; Uniacke-Lowe, 2011), human (Sood et al., 1997, 2002), and donkey (Malacarne et al., 2017). The appearance and the size of casein micelles in the milks of 19 species (guinea pig, rat, coypu (nutria), dog, cat, grey seal, rabbit, donkey, horse, alpaca, dromedary camel, cow, red deer, sheep, pig, water buffalo, goat, porpoise, and human) were studied by Buchheim et al. (1989); the structure of all micelles appeared to be similar using electron microscopy, but there were large interspecies differences in size—human micelles were smallest (64 nm), whereas those of the alpaca, goat, camel, and donkey were very large (300–350 nm); the micelles in equine milk are as large as 700 nm (Uniacke-Lowe, 2011). In addition to this work, the casein and casein micelle structures of 20 different species were recently studied by Holt (2016), showing further interspecies variations.

Casein Heterogeneity and fractionation of the caseins Hammarsten (1883) believed that isoelectric casein was a homogeneous protein, but during the early years of the 20th century, evidence was presented by T.B. Osborne and A.J. Wakeman and especially by K. Linderstrøm-Lang and collaborators that it was heterogeneous (see McMeekin, 1970). By extraction with ethanol-HCl mixtures, K. Linderstrøm-Lang and S. Kodoma obtained three major casein fractions, which contained about 1.0%, 0.6%, or 0.1% P, and several minor fractions. The heterogeneity of casein was confirmed by analytical ultracentrifugation and free-boundary electrophoresis by K.O. Pedersen and O. Mellander, respectively (see McMeekin, 1970). Electrophoresis resolved isoelectric casein into three proteins, which were named α, β, and γ, in order of decreasing electrophoretic mobility and represented about 75%, 22%, and 3% of whole casein, respectively. Following the demonstration of casein’s heterogeneity, several attempts were made to isolate the individual caseins. The first reasonably successful method was developed in 1944 by R.C. Warner, who exploited differences in the solubility of the α- and β-caseins at pH 4.4 and 2°C. A much more satisfactory fractionation method was developed in 1952 by N.J. Hipp and coworkers, based on the differential solubility of α-, β-, and γ-caseins in urea solutions at pH 4.9. This method was widely used for many years until the widespread application of

34

2. Milk proteins: An overview

ion-exchange chromatography (Visser et al., 1986) and reverse-phase chromatography (Visser et al., 1991; Bobe et al., 1998) for such applications. Reviews describing the application of high-performance liquid chromatography and fast protein liquid chromatography for the analysis of milk and dairy products include Gonzalez de Llano et al. (1990), Strange et al. (1992), and Dupont et al. (2013). α-Casein was resolved by Waugh and von Hipple (1956) into calcium-sensitive and calcium-insensitive proteins, which were called αs- and κ-caseins, respectively. κ-Casein, which represents
12% of the total casein, is responsible for the formation and the stabilization of casein micelles and affects many technologically important properties of the milk protein system. Numerous chemical methods were soon developed for the isolation of κ-casein (see Fox, 2003; O’Mahony and Fox, 2013). αs-Casein, prepared by the method of Waugh and von Hipple, contains two proteins, now called αs1- and αs2-caseins (Annan and Manson, 1969). Chemical methods for fractionation of the caseins have largely been superseded by ion-exchange chromatography, which gives superior results when urea and a reducing agent are used (see Strange et al., 1992; Imafidon et al., 1997). Microheterogeneity of the caseins αs1-, αs2-, β-, and κ-Caseins represent approximately 38%, 10%, 35%, and 12%, respectively, of whole bovine casein. However, starch gel electrophoresis (SGE) or polyacrylamide gel electrophoresis (PAGE) indicates much greater heterogeneity because of small differences in one or more of the caseins, referred to as microheterogeneity, which arises from five factors: 1. Variability in the degree of phosphorylation All the caseins are phosphorylated but to a variable degree (αs1-, 8 or 9P; αs2-, 10, 11, 12, or 13P; β-, 4 or 5P; κ-, 1 or 2P per molecule). The number of phosphate residues is indicated: thus, αs1-CN 8P, β-CN 5P, etc. 2. Genetic polymorphism In 1955, R. Aschaffenburg and J. Drewry discovered that β-lactoglobulin exists in two forms (variants and polymorphs), now called A and B, that differ by only two amino acids. The variant in the milk is genetically controlled, and the phenomenon is called genetic polymorphism. It was soon shown that all milk proteins exhibit genetic polymorphism and at least 45 polymorphs have been detected electrophoretically, which differentiates on the basis of charge, and therefore only polymorphs that differ in charge have been detected (NgKwai-Hang, 2011). It is very likely that only a small proportion of the genetic polymorphs of the milk proteins have been detected. The potential of peptide mapping of enzymatic hydrolyzates by high-performance liquid chromatography (HPLC)-mass spectrometry (MS) has been assessed (Leonil et al., 1995). The genetic polymorphs present are indicated by a Latin letter as follows: β-CN A 5P, αs1-CN B 9P, κ-CN A 1P, etc. Genetic polymorphism also occurs in the milks of sheep, goat, buffalo, pig, horse, and probably of all species. Technologically important properties of milk, for example, rennetability, heat stability, yield, and proportions of milk proteins, are affected by the genetic polymorphs of the milk proteins present, and work in this area is being expanded and refined ( Jakob and Puhan, 1992). The extensive literature on the genetic polymorphism of milk proteins has been the subject of several reviews, including Ng-Kwai-Hang and Grosclaude (2003), Ng-Kwai-Hang (2011), and Martin et al. (2013b).

Casein

35

3. Disulfide bonding αs1- and β-Caseins lack cysteine and cystine, but both αs2- and κ-caseins contain two 1/2 cystine residues, which occur as intermolecular disulfide bonds. αs2-Casein exists as a disulfide-linked dimer, and up to 10 κ-casein molecules may be linked by disulfide bonds. Inclusion of a reducing agent (usually 2-mercaptoethanol) in SGE or PAGE gels is required for good resolution of κ-casein; in its absence, αs2-casein appears as a dimer (originally called αs5-casein). 4. Variations in the degree of glycosylation κ-Casein is the only glycosylated casein; it contains galactose, N-acetylgalactosamine, and N-acetylneuraminic (sialic) acid, which occur as tri- or tetrasaccharides, the number of which varies from 0 to 4 per molecule of protein (i.e., a total of nine variants) attached to threonine residues. 5. Hydrolysis of the caseins by plasmin Milk contains several indigenous proteinases, the principal of which is plasmin, a trypsin-like serine-type proteinase from blood; it is highly specific for peptide bonds with a lysine or arginine at the P1 position. The preferred casein substrates are β- and αs2-; αs1casein is also hydrolyzed, but κ-casein is very resistant, as are the whey proteins. All the caseins contain several lysine and arginine residues, but only a few bonds are hydrolyzed rapidly. β-Casein is hydrolyzed rapidly at the bonds Lys28dLys29, Lys105dHis106, and Lys107dGlu108. The resulting C-terminal peptides are the γ-caseins (γ1: β-CNf29–209; γ2: β-CNf106–209; γ3: β-CNf108–209), and the respective N-terminal peptides are proteose peptones 5, 8slow, and 8fast. The γ-caseins, which represent
3% of the total casein, are evident in PAGE gels (O’Flaherty, 1997). Other plasmin-produced peptides are probably present, but either they are too small to be readily detectable by PAGE, or their concentrations are very low relative to those of the principal caseins. Although αs2-casein in solution is also quite susceptible to plasmin, αs2- derived peptides have not been identified in milk.

Casein micelles It has been known since the work of H. Schuler in 1818 that the casein in milk exists as large particles, now called casein micelles. The stability of the micelles is critically important for many of the technologically important properties of milk and consequently has been the focus of much research, especially since the discovery of κ-casein by Waugh and von Hipple (1956). Early views and research on casein micelles was reviewed by Fox and Brodkorb (2008), and a recent review is that of McMahon and Oommen (2013). Although views on the detailed structure of the casein micelle are divided, there is widespread, or unanimous, agreement on their general structure and properties. Electron microscopy shows that casein micelles are spherical with a diameter in the range 50–500 nm (average
120 nm) and a mass ranging from 106 to 3  109 Da (average
108 Da) for bovine micelles. There are numerous small micelles, but these represent only a small proportion of the mass. There are 1014–1016 micelles/mL milk, and they are roughly two micelle diameters (
250 nm) apart. The dry matter of the micelles is
94% protein and 6% low-molecular mass species, referred to collectively as colloidal calcium phosphate (CCP), and consisting mainly of calcium phosphate with some magnesium and citrate and trace amounts of other species. The micelles bind
2.0 g H2O/g protein. They

36

2. Milk proteins: An overview

scatter light, and the white color of milk is due largely to light scattering by the casein micelles; the white color is lost if the micelles are disrupted by dissolving CCP with citrate, EDTA, or oxalate, by increasing the pH, or by urea (>5 M) or ethanol (
35% at 70°C). Stability of casein micelles The micelles are quite stable to the principal processes to which milk is normally subjected. They are very stable at high temperatures and withstand heating at 140°C for 15–20 min at pH 6.7. Coagulation is caused by heat-induced changes, for example, a decrease in pH because of the pyrolysis of lactose to acids, dephosphorylation of casein, cleavage of the carbohydraterich moiety of κ-casein, denaturation of the whey proteins and their precipitation on the casein micelles, and precipitation of soluble calcium phosphate on the micelles (see O’Connell and Fox, 2003). The micelles are stable to compaction (e.g., they can be sedimented by ultracentrifugation and redispersed by mild agitation), to commercial homogenization, and to ionic calcium concentrations ([Ca2+]) up to at least 200 mM at temperatures up to 50°C. The effects of high pressure (up to 800 MPa) on the casein micelles in bovine, ovine, caprine, and buffalo milks have been studied; the size of the micelles increases up to 200–300 MPa but decreases at higher pressure (see Huppertz et al., 2006; Huppertz and de Kruif, 2007). As the pH of milk is reduced, the CCP dissolves and is fully soluble at pH 4.9. Acidification of cold (4°C) milk to pH 4.6, followed by dialysis against bulk milk, is a convenient technique for altering the CCP content of milk. If acidified cold milk is readjusted to pH 6.7, the micelles reform provided the pH had not been reduced below 5.5 (Lucey et al., 1996). This result suggests that most of the CCP can be dissolved without destroying the structure of the micelles. Some proteinases, especially chymosin, catalyze a very specific hydrolysis of κ-casein, as a result of which the casein coagulates in the presence of Ca2+ or other divalent ions. This is the key step in the manufacture of most cheese varieties. The proteinase preparations used for cheesemaking are called rennets. At room temperature, the casein micelles are destabilized by
40% ethanol at pH 6.7 or by a lower concentration if the pH is reduced (see Horne, 2003b). However, if the system is heated to 70°C, the precipitate redissolves, and the system becomes translucent. When the system is recooled, the white appearance of the milk is restored, and a gel is formed if the ethanol-milk mixture is held at 4°C, especially if concentrated milk is used. If the ethanol is removed by evaporation, very large aggregates (average diameter
3000 nm) are formed. The dissociating effect of ethanol is promoted by increasing the pH (35% ethanol causes dissociation at 20°C and pH 7.3) or by adding NaCl (Horne, 2003b). Methanol and acetone have an effect similar to that of ethanol, but propanol causes dissociation at
25°C. The mechanism by which ethanol dissociates casein micelles has not been established, but it is not due to the dissolution of CCP (O’Connell et al., 2003). The micelles are also reversibly dissociated by urea (5 M) (McGann and Fox, 1974; Holt, 1998), sodium dodecyl sulfate (SDS; Lefebvre-Cases et al., 1998), or raising the pH >9; under these conditions, the CCP is not dissolved. The micelles are destabilized by freezing (cryodestabilization) because of a decrease in pH and an increase in the [Ca2+] in the unfrozen phase of milk; concentrated milk is very susceptible to cryodestabilization (Moon et al., 1988, 1989). Cryodestabilized casein can be dispersed by warming the thawed milk to 55°C to give particles with micelle-like properties.

Casein

37

Micelle structure There has been speculation since the beginning of the 20th century on how the casein particles (micelles) are stabilized (see Fox and Brodkorb, 2008), but no significant progress was possible until the isolation and characterization of κ-casein by Waugh and von Hipple (1956). The first attempt to describe the structure of the casein micelle was made by Waugh (1958); since then, numerous models have been made and refined. Progress has been reviewed regularly; recent reviews include de Kruif and Holt (2003), Horne (2002, 2003a, 2006), Farrell et al. (2006b), and McMahon and Oommen (2013). The principal features that must be met by any micelle model are the following: κ-Casein, which represents
12%–15% of the total casein, must be so located as to be able to stabilize the calcium-sensitive αs1-, αs2-, and β-caseins, which represent approximately 85% of the total casein; chymosin and other rennets, which are relatively large molecules [molecular weight (MW)
35 kDa], very rapidly and specifically hydrolyze most of the κ-casein; when heated in the presence of whey proteins, κ-casein and β-lactoglobulin (MW
36 kDa) interact to form a disulfide-linked complex, which modifies the rennet and heat coagulation properties of the micelles in milk. The arrangement that would best explain these features is a surface layer of κ-casein surrounding the calciumsensitive caseins, somewhat analogous to a lipid emulsion in which the triglycerides are surrounded by a thin layer of emulsifier. Most models of the casein micelle propose a surface location for κ-casein, but some early models envisaged κ-casein serving as nodes in the interior of the micelle. Removal of CCP causes the disintegration of the micelles into particles of MW
106 Da, suggesting that the casein molecules are held together in the micelles by CCP. The properties of the CCP-free system are very different from those of normal milk (e.g., it is precipitated by relatively low levels of Ca2+, it is more stable to heat-induced coagulation, and it is not coagulable by rennets). Many of these properties can be restored, at least partially, by an increased concentration of calcium. However, CCP is not the only integrating factor, as indicated by the dissociating effects of urea, SDS, ethanol, or alkaline pH. At low temperatures, casein, especially β-casein, dissociates from the micelles. There has been strong support for the view, first proposed by C.V. Morr in 1967, that the micelles are composed of submicelles (
106 Da and 10–15 nm in diameter) linked together by CCP, giving a micelle with an open, porous structure. On removing the CCP (by acidification/dialysis, EDTA, citrate, or oxalate), the micelles disintegrate. Disintegration may also be achieved by treatment with urea, SDS, 35% ethanol at 70°C, or pH >9. These reagents do not solubilize CCP, suggesting that hydrophobic interactions and hydrogen bonds contribute to the micelle structure. Much of the evidence for a submicellar structure relies on electron microscopy studies that appear to show variations in electron density, a raspberry-like structure, which was interpreted as indicating submicelles. Views on the proposed structure of the submicelles have evolved over the years (see McMahon and McManus, 1998; Dalgleish et al., 2004; Dalgleish, 2011; McMahon and Oommen, 2008, 2013). Proposals have included the following: A rosette-type structure similar to that of a classical soap micelle, in which the polar regions of the αs1-, αs2-, and β-caseins are orientated toward the outside of the submicelles to reduce electrostatic repulsion between neighboring charged groups; each submicelle was considered to be surrounded by a layer (coat) of κ-casein, thus providing a κ-casein coat for the entire micelle. Several authors have suggested that the submicelles are not covered completely by κ-casein and that there are

38

2. Milk proteins: An overview

κ-casein-rich hydrophilic and κ-casein-deficient hydrophobic regions on the surface of each submicelle. The latter aggregate via the hydrophobic patches such that the entire micelle has a κ-casein-rich surface layer but with some of the other caseins also on the surface. In a popular version of this model, it was proposed that the hydrophilic C-terminal region of κ-casein protrudes from the surface, forming a layer 5–10 nm thick and giving the micelles a hairy appearance. This hairy layer, functioning as an ionic brush, is considered to be responsible for the stability of the micelle, through major contributions to the zeta potential (20 mV) and steric stabilization. If the hairy layer is removed through the specific hydrolysis of κ-casein or collapsed (e.g., by ethanol), the colloidal stability of the micelles is destroyed, and they aggregate. A further variant of the subunit model envisages two main types of subunit: one, consisting of αs1-, αs2-, and β-caseins, is present in the core of the micelle; the other, consisting of αs1-, αs2-, and κ-caseins, forms a surface layer. It has also been proposed that β-casein associates to form thread-like structures with which αs1- and αs2-caseins associate hydrophobically to form the core of the micelle or submicelles, which are surrounded by a layer of κ-casein; CCP cements neighboring submicelles within the micelle. A recent study by Bouchoux et al. (2010), using small-angle X-ray scattering analysis, supports a submicelle structure. Although the submicelle model of the casein micelle explains many of the principal features of and the physicochemical reactions undergone by the micelles and has been supported widely, it has never enjoyed unanimous support. Indeed, new electron microscopy techniques have cast doubts on the authenticity of submicelles. Using cryopreparation electron microscopy with stereo imaging, McMahon and McManus (1998) found no evidence to support the submicelle model and concluded that, if the micelles do consist of submicelles, these must be smaller than 2 nm or less densely packed than previously presumed. Like other forms of electron microscopy, field-emission scanning electron microscopy showed that casein micelles have an irregular surface, but Dalgleish et al. (2004) concluded that the caseins form tubular structures rather than spherical submicelles; in principle, this model seems basically similar to earlier subunit models. McMahon and Oommen (2008) also found no evidence for a submicellar structure using high-resolution transmission electron microscopy. Alternative models have been proposed; Visser (1992) proposed that the micelles are spherical conglomerates of randomly aggregated casein molecules held together by amorphous calcium phosphate and hydrophobic bonds, with a surface layer of κ-casein. Holt (1992) considered the casein micelle to be a tangled web of flexible casein molecules forming a gel-like structure in which microgranules of CCP are an integral feature and, from the surface of which, the C-terminal region of κ-casein extends, forming a hairy layer. In what he referred to as the dual-binding model, Horne (1998, 2002, 2003a, 2006) described how casein molecules interact hydrophobically and through calcium phosphate nanoclusters to form micelles. These three models retain the key features of the submicelle model, that is, the cementing role of CCP and the predominantly surface location and micelle-stabilizing role of κ-casein, and differ from it mainly with respect to the internal structure of the micelle. Recent reviews on aspects of casein micelle structure and stability include McMahon and Oommen (2013), de Kruif (2014), Day et al. (2015), Holt (2016), Hindmarsh and Watkinson (2017), Horne (2017), Huppertz et al. (2017), Bijl et al. (2018), Huppertz et al. (2018), and Lucey and Horne (2018).

Whey proteins

39

Whey proteins About 20% of the total protein of bovine milk is whey (serum) protein. The total whey protein fraction is prepared by any of the methods used in the preparation of casein, that is, the proteins that are soluble at pH 4.6, in saturated NaCl, or after rennet-induced coagulation of the caseins are permeable on microfiltration or are not sedimented by ultracentrifugation. The proteins prepared by these methods differ somewhat: acid whey contains proteose peptones; immunoglobulins are coprecipitated with the caseins by saturated NaCl; rennet whey contains the macropeptides, produced from κ-casein by rennet, plus small amounts of casein; small casein micelles remain in the ultracentrifugal serum. On a commercial scale, whey protein-rich products are prepared by the following: • Ultrafiltration/diafiltration of acid casein or rennet whey, to remove various amounts of lactose, and spray-dried to produce whey protein concentrates (30%–85% protein) • Ion-exchange chromatography and spray-dried to yield whey protein isolate, containing
95% protein • Demineralization by electrodialysis or ion exchange, thermal evaporation of water, and crystallization of lactose • Thermal denaturation, the removal of precipitated protein by filtration/centrifugation and spray-drying, to yield lactalbumin, which has very low solubility and poor functionality

Fractionation of whey proteins It was recognized early that acid whey contains two well-defined groups of proteins: lactalbumins, which are soluble in 50% saturated (NH4)2SO4 or saturated MgSO4, and lactoglobulins, which are salted out under these conditions. The lactoglobulin fraction contains mainly immunoglobulins. The lactalbumin fraction contains two principal proteins, β-lactoglobulin and α-lactalbumin, and several minor proteins, including blood serum albumin and lactoferrin, which have been isolated by various procedures and crystallized (see Imafidon et al., 1997; Fox, 2003; O’Mahony and Fox, 2013). There is considerable interest in the production of the major and many minor whey proteins on a commercial scale for nutritional, nutraceutical, or functional applications. Several methods have been developed for the industrial-scale production of several whey proteins (see Mulvihill and Ennis, 2003).

β-Lactoglobulin β-Lactoglobulin (β-Lg) represents
50% of the whey proteins, that is,
12% of the total protein, in bovine milk. It is a typical globular protein and has been very well characterized. The extensive literature has been reviewed, among others, by Sawyer (2003, 2013) and Creamer et al. (2011). β-Lg is the principal whey protein in the milks of cattle, buffalo, sheep, and goat, although there are slight interspecies differences. Initially, it was considered that β-Lg occurs only in the milk of ruminants, but it is now known that a similar protein occurs in the milk of many other species, including the pig, horse, kangaroo, dolphin, and manatee. However, β-Lg does not occur in the milks of human, rat, mouse, guinea pig, camel, llama, or alpaca, in which α-lactalbumin is the principal whey protein.

40

2. Milk proteins: An overview

Bovine β-Lg consists of 162 amino acid residues per monomer, with a MW of
18 kDa; its amino acid sequence and that of several other species have been established. Its isoelectric point is
pH 5.2. It contains two intramolecular disulfide bonds and 1 mol of cysteine per monomer. The cysteine is especially important because it reacts, following thermal denaturation, with the intermolecular disulfide of κ-casein and significantly affects the rennet coagulation and heat stability of milk. It is also responsible for the cooked flavor of heated milk. Some β-Lgs (e.g., porcine) lack a sulfhydryl group. Twelve genetic variants of bovine β-Lg have been identified, the most common being A and B. Genetic polymorphism also occurs in the β-Lgs of other species. β-Lg is a highly structured protein: in the pH range 2–6, 10%–15% of the molecule exists as α-helices, 43% as β-sheets, and 47% as unordered structures, including β-turns; the β-sheets occur in a β-barrel-type calyx. The molecule has a very compact globular structure; each monomer exists almost as a sphere, about 3.6 nm in diameter. β-Lg exists as a dimer, MW
36 kDa, in the pH range 5.5–7.5, as a monomer at pH 7.5, and as an octamer (MW
144 kDa) in the pH range 3.5–5.5. Porcine and other β-Lgs, which lack a free thiol, do not form dimers, a property that is probably not due to the absence of a thiol group. β-Lg is very resistant to proteolysis in its native state; this feature suggests that its primary function is not nutritional. It may have either or both of two biological roles. • It binds retinol (vitamin A) in a hydrophobic pocket, protects it from oxidation, and transports it through the stomach to the small intestine, where the retinol is transferred to a retinol-binding protein, which has a similar structure to β-Lg. It is not clear how retinol is transferred from the core of the fat globules, where it occurs in milk, to β-Lg and why some species lack this protein. β-Lg can bind many hydrophobic molecules, and hence, its ability to bind retinol may be incidental. β-Lg is a member of the lipocalin family, all of which have binding properties (Akerstrom et al., 2000). • Through its ability to bind fatty acids, β-Lg stimulates lipase activity, which may be its most important physiological function. β-Lg is the most allergenic protein in bovine milk for human infants, and there is interest in producing whey protein products with reduced β-Lg levels for use in infant formulas. β-Lg has very good thermogelling properties and determines the gelation of whey protein concentrates.

α-Lactalbumin About 20% of the protein of bovine whey (3.5% of the total milk protein) is α-lactalbumin (α-La), which is the principal protein in human milk. It is a small protein containing 123 amino acid residues, with a mass of
14 kDa, and has been well characterized; the literature has been reviewed, among others, by McKenzie and White (1991) and Brew (2003, 2011, 2013). α-La contains four tryptophan residues per mole, giving it a specific absorbance at 280 nm of
20. It contains four intramolecular disulfide bonds per mole but no cysteine, phosphate, or carbohydrate. Its isoionic point is
pH 4.8. The milk of Bos taurus breeds contains only one genetic variant (B) of α-La, but Zebu cattle produce two variants (A and B). α-La has been isolated from the milks of cattle, sheep, goat, sow, human, buffalo, rat, guinea pig, horse, and many other species; there are minor interspecies differences in the composition and properties.

Whey proteins

41

The primary structure of α-La is homologous with lysozyme; out of a total of 123 amino acid residues in α-La, 54 are identical to corresponding residues in chicken egg white lysozyme, and 23 others are structurally similar. α-La is a compact, highly structured globular protein. The tertiary structure of α-La is similar to that of lysozyme (McKenzie and White, 1991), and X-ray crystallography-based analyses of apo- and holo-bovine α-La have been published by Chrysina et al. (2000). In evolutionary terms, lysozyme is a very ancient protein; it is believed that α-La evolved from it through gene duplication (see Nitta and Sugai, 1989). The biological function of α-La is its contribution to lactose synthetase, the enzyme that catalyzes the final step in the biosynthesis of lactose. There is a direct correlation between the concentrations of α-La and lactose in milk. Interestingly, the milk of the California sea lion and the hooded seal contain no α-La. α-La is a metalloprotein containing two Ca2+ per molecule in a pocket containing four Asp residues. The calcium-containing protein is the most heat stable of the principal whey proteins, or more correctly, the protein renatures following heat denaturation, which occurs at a relatively low temperature, as indicated by differential scanning calorimetry. When the pH is reduced to 20% protein, thereby increasing the cost of drying and resulting in powders of low bulk density. The lack of stable tertiary structures means that the caseins are not denaturable stricto sensu and, consequently, are extremely heat stable; sodium caseinate, at pH 7, can withstand heating at 140°C for several hours without visible change. This very high heat stability makes it possible to produce heat-sterilized dairy products with very little change in physical appearance, whereas other major food systems undergo major physical and sensorial changes on such severe heating. The caseins have a very strong tendency to associate mainly as a result of hydrophobic bonding. Even in sodium caseinate, the most soluble form of casein, the molecules form aggregates of 250–500 kDa, that is, containing 10–20 casein protein molecules. This strong tendency to associate makes it difficult to fractionate the caseins, for which a dissociating agent (e.g., urea or SDS) is required. However, a tendency to associate is important for some functional applications and in the formation and stabilization of casein micelles. In contrast, the whey proteins are molecularly dispersed in solution.

Differences between casein and whey proteins

45

Because of their high content of phosphate groups, which occur in clusters, αs1-, αs2-, and β-caseins have a strong tendency to bind metal ions, which, in the case of milk, are mainly Ca2+. This property has many major consequences, the most important from a technological viewpoint being that these three proteins, which represent approximately 85% of the total casein, are insoluble at calcium concentrations >
6 mM at temperatures >20°C. As bovine milk contains
30 mM calcium, it would be expected that the caseins would precipitate under the conditions prevailing in milk. However, κ-casein, which contains only one organic phosphate group, binds calcium weakly and is soluble at all calcium concentrations found in dairy products. Furthermore, when mixed with the calcium-sensitive caseins, κ-casein can stabilize and protect up to
10 times its mass of the former by forming large colloidal particles called casein micelles. The micelles act as carriers of inorganic elements, not only Ca and P but also Mg and Zn, and are, therefore, very important from a nutritional viewpoint. Through the formation of micelles, it is possible to solubilize much higher levels of Ca and PO4 than would otherwise be possible.

Comparison of key properties of casein and whey proteins • Solubility at pH 4.6. The caseins are, by definition, insoluble at pH 4.6, whereas the whey proteins are soluble under the ionic conditions of milk. The isoelectric precipitation of casein is exploited in the production of caseins and caseinates, fermented milk products, and acid-coagulated cheeses. • Coagulability following limited proteolysis. The caseins are coagulable following specific, limited proteolysis, whereas the whey proteins are not. This property of the caseins is exploited in the production of rennet-coagulated cheese (
75% of all cheeses) and rennet casein. • Heat stability. The caseins are very heat stable. Milk at pH 6.7 may be heated at 100°C for 24 h without coagulation and withstands heating at 140°C for up to 20–25 min; aqueous solutions of sodium caseinate may be heated at 140°C for several hours without apparent changes. The heat stability of the whey proteins is typical of globular proteins; they are denatured completely on heating at 90°C for 10 min. The remarkably high heat stability of the caseins, which is probably due to their lack of typical stable secondary and tertiary structures, permits the production of heat-sterilized dairy products with relatively small physical changes. • Amino acid composition. The caseins contain high levels of proline (17% of all residues in β-casein), which explains their lack of α- and β-structures. The caseins are phosphorylated, whereas the principal whey proteins are not. Whole isoelectric casein contains approximately 0.8% phosphorus, but the degree of phosphorylation varies among the individual caseins. The phosphate is attached to the polypeptides as phosphomonoesters of serine; the presence of phosphate groups has major significance for the properties of the caseins, for example, molecular charge and related properties, such as hydration, solubility and heat stability, and metal binding, which affects their physicochemical, functional, and nutritional properties. Metal binding by casein is regarded as a biological function because it enables a high concentration of calcium phosphate to be carried in milk in a soluble form (to supply the requirements of the neonate); otherwise, calcium phosphate would

46

2. Milk proteins: An overview

precipitate in and block the ducts of the mammary gland, leading to the death of the gland and perhaps of the animal. • Sulfur content. The caseins are low in sulfur (0.8%), whereas the whey proteins are relatively rich (1.7%). The sulfur in casein is mainly in methionine, with little cystine or cysteine; the principal caseins are devoid of the latter two amino acids. The whey proteins are relatively rich in cystine and/or cysteine, which have major effects on the physicochemical properties of these proteins and of milk. • Site of biosynthesis. The caseins are synthesized in the mammary gland and are unique to this organ. Presumably, they are synthesized to meet the amino acid requirements of the neonate and as carriers of important metals required by the neonate. The principal whey proteins are also synthesized in and are unique to the mammary gland, but several minor proteins in milk are derived from blood, either by selective transport or because of leakage. Most of the whey proteins have a biological function. • Physical state in milk. The whey proteins exist in milk as monomers or as small quaternary structures, whereas the caseins exist as large aggregates, known as micelles, with a mass of
108 Da and containing about 5000 molecules. The white color of milk is due largely to the scattering of light by the casein micelles. The structure, properties, and stability of the casein micelles are of major significance for the technological properties of milk and have been the subject of intensive research.

Preparation of casein and whey proteins The protein fractions may be prepared from whole or skimmed milk, but the latter is usually used because the fat is occluded in isoelectric casein and interferes with further characterization of the proteins. The fat is easily removed from milk by centrifugation (e.g., 3000  g for 30 min), and any remaining fat may be removed by washing the precipitated protein with ether. Isoelectric precipitation is the most widely used method for separating the casein and noncasein fractions of milk protein, but several other techniques are used in certain situations (see Fox, 2003; O’Mahony and Fox, 2013). • Isoelectric precipitation at
pH 4.6 at 20°C: the precipitate is recovered by filtration or lowspeed centrifugation. Essentially similar methods are used to prepare casein on a laboratory or industrial scale. • Ultracentrifugation: in milk, the casein exists as large micelles, which may be sedimented by centrifugation at 100,000  g for 1 h; the whey proteins are not sedimentable. The casein pellet can be redispersed in a suitable buffer as micelles with properties similar to those of natural micelles. • Salting-out methods: casein can be precipitated by any of several salts, usually by (NH4)2SO4 at 260 g/L or saturated NaCl. Under these conditions, the immunoglobulins coprecipitate with the caseins. • Ultrafiltration and microfiltration: all the milk proteins are retained by small-pore, semipermeable membranes and are separated from lactose and soluble salts. This process, ultrafiltration, is used widely for the industrial-scale production of whey protein concentrates and to a lesser extent for the production of milk protein concentrates. Intermediate-pore membranes are used to separate casein micelles from whey proteins.

Minor milk proteins

• •

•

•

•

47

In microfiltration, using large-pore membranes (1.4 μm), both the caseins and the whey proteins are permeable, but >99.9% of bacteria and other large particles are retained; microfiltration is used for the production of extended-shelf-life beverage milk or cheese milk or to remove lipoprotein particles from whey to improve the functionality of whey protein concentrate. Gel filtration: it is possible to separate the caseins from the whey proteins by permeation chromatography, but this method is not used industrially and rarely on a laboratory scale. Precipitation by ethanol: the caseins are precipitated from milk by
40% ethanol, whereas the whey proteins remain soluble; however, precipitation by ethanol is rarely used, on either a laboratory scale or an industrial scale, for the precipitation of casein. Cryoprecipitation: caseins, in a micellar form, may be destabilized and precipitated by freezing milk or, preferably, concentrated milk, at about 10°C. Precipitation is caused by a decrease in pH and an increase in [Ca2+]; the precipitated micelles may be redispersed as micelles by heating to about 55°C. Alternatively, the cryoprecipitated casein may be recovered, washed, and dried; it has many interesting properties for food applications but is not produced commercially. Rennet coagulation: The casein micelles are destabilized by specific, limited proteolysis and coagulate in the presence of Ca2+. The properties of rennet-coagulated casein are very different from those of isoelectric casein, and it is very suitable for certain food applications (e.g., cheese analogs). Caseinates: isoelectric casein is insoluble in water, but may be converted to water-soluble caseinates by dispersion in water and adjusting the pH to
6.7 with alkali, usually NaOH to yield sodium caseinate. KOH, NH4OH, and Ca(OH)2 give the corresponding caseinates, which may be freeze-dried or spray-dried.

Minor milk proteins Milk contains several proteins at very low or trace levels, many of which are biologically active (see Schrezenmeir et al., 2000; Korhonen, 2009; Wynn and Sheehy, 2013; Fox et al., 2015; Nongonierma and FitzGerald, 2015); some are regarded as highly significant and have attracted considerable attention as nutraceuticals. When ways of increasing the value of milk proteins are discussed, the focus is usually on these minor proteins, but they are, in fact, of little economic value to the overall dairy industry. They are found mainly in whey, but some are also located in the MFGM. Reviews on the minor proteins of milk include Fox and Flynn (1992), Fox and Kelly (2003), Haggarty (2003), Wynn et al. (2011), O’Mahony and Fox (2013), Wynn and Sheehy (2013), Fox et al. (2015), Morgan et al. (2016), and Crowley et al. (2017b).

Metal-binding proteins Milk contains several metal-binding proteins: the caseins (Ca, Mg, Zn), α-La (Ca), xanthine oxidase (Mo, Fe), alkaline phosphatase (Zn, Mg), lactoperoxidase (Fe), catalase (Fe), ceruloplasmin (Cu), glutathione peroxidase (Se), lactoferrin (Fe), and transferrin (Fe). Lactoferrin, a

48

2. Milk proteins: An overview

nonheme iron-binding glycoprotein, is a member of a family of iron-binding proteins, which includes transferrin and ovotransferrin (conalbumin) (see Lonnerdal, 2003; Lonnerdal and Suzuki, 2013). It is present in several body fluids, including saliva, tears, sweat, and semen. Lactoferrin has several potential biological functions: it improves the bioavailability of Fe; is bacteriostatic (by sequestering Fe and making it unavailable to intestinal bacteria); and has antioxidant, antiviral, antiinflammatory, immunomodulatory, and anticarcinogenic activity. Human milk contains a very high level of lactoferrin (
20% of the total nitrogen), and therefore, there is interest in fortifying bovine milk-based infant formulas with lactoferrin. The pI of lactoferrin is
9.0, that is, it is cationic at the pH of milk, whereas most milk proteins are anionic, and it can be isolated on an industrial scale by adsorption on a cation-exchange resin. Hydrolysis of lactoferrin by pepsin yields peptides called lactoferricins, which are more bacteriostatic than lactoferrin, and their activity is independent of the iron status. Milk also contains a low level of serum transferrin. Milk contains a copper-binding glycoprotein, ceruloplasmin, also known as ferroxidase (EC 1.16.3.1). Ceruloplasmin is an α2-globulin with a MW of
126 kDa; it binds six atoms of copper per molecule and may play a role in delivering essential copper to the neonate. Glutathione peroxidase (GTPase) is a selenium-containing protein. It has been reported that milk contains GTPase and that it binds 30% of the total selenium in milk (see Fox and Kelly, 2006b; O’Mahony et al., 2013). GTPase has no known enzymatic function in milk, and the activity attributed to GTPase in milk may be due to sulfhydryl oxidase (Stagsted, 2006).

β2-Microglobulin β2-Microglobulin, initially called lactollin, was first isolated from bovine acid-precipitated casein by M.L. Groves in 1963. Lactollin, reported to have a MW of 43 kDa, is a tetramer of β2-microglobulin, which consists of 98 amino acids, with a calculated MW of 11,636 Da. β2-Microglobulin, a component of the immune system, is probably produced by proteolysis of a larger protein, mainly within the mammary gland, and it has no known significance in milk.

Osteopontin Osteopontin is a highly phosphorylated acidic glycoprotein, consisting of 261 amino acid residues with a calculated MW of 29,283 Da (total MW of the glycoprotein,
60,000 Da). Osteopontin has 50 potential calcium-binding sites, about half of which are saturated under normal physiological concentrations of calcium and magnesium. Osteopontin occurs in bone (it is one of the major noncollagenous proteins in bone), in many other normal and malignant tissues, and in milk and urine and can bind to many cell types. It is believed to have a diverse range of functions (Bayless et al., 1997), but its role in milk is not clear. A rapid method for the isolation of osteopontin from milk was reported by Azuma et al. (2006), who showed that it binds lactoferrin, lactoperoxidase, and immunoglobulins and may serve as a carrier for these proteins. Reviews on the biological functions of osteopontin include Christensen and Sørensen (2016), Jiang and L€ onnerdal (2016), and Demmelmair et al. (2017).

Minor milk proteins

49

Vitamin-binding proteins Milk contains binding proteins for at least the following vitamins: retinol (vitamin A, i.e., bound by β-Lg), biotin, folic acid, and cobalamin (vitamin B12). The precise role of these proteins is not clear, but they probably improve the absorption of vitamins from the intestine or act as antibacterial agents by rendering vitamins unavailable to bacteria. The concentration of these proteins varies during lactation, but the influence of other factors such as individuality, breed, and nutritional status is not known. The activity of these proteins is reduced or destroyed on heating at temperatures somewhat higher than regular pasteurization (see Wynn et al., 2011; Wynn and Sheehy, 2013).

Angiogenins Angiogenins induce the growth of new blood vessels, that is, angiogenesis. They have high sequence homology with members of the RNase A superfamily of proteins and have RNase activity. Two angiogenins (ANG-1 and ANG-2) have been identified in bovine milk and blood serum; both strongly promote the growth of new blood vessels in a chicken membrane assay. The function(s) of the angiogenins in milk is unknown. They may be part of a repair system to protect either the mammary gland or the intestine of the neonate and/or part of the host defense system (see Wynn et al., 2011; Wynn and Sheehy, 2013).

Kininogen Two kininogens have been identified in bovine milk, a high-MW form (88–129 kDa, depending on the level of glycosylation) and a low-MW form (16–17 kDa). Bradykinin, a biologically active peptide containing nine amino acids, which is released from the high-MW kininogen by the action of the enzyme, kallikrein, has been detected in the mammary gland and is secreted into milk, from which it has been isolated. Plasma kininogen is an inhibitor of thiol proteinases and has an important role in blood coagulation. Bradykinin affects smooth muscle contraction and reduces hypertension. The biological significance of bradykinin and kininogen in milk is unknown (see Wynn et al., 2011; Wynn and Sheehy, 2013).

Glycoproteins Many of the minor proteins discussed earlier are glycoproteins; in addition, several other minor glycoproteins have been found in milk and especially in colostrum, the functions of which have not been elucidated. One of the high-MW glycoproteins in bovine milk is prosaposin, a neurotrophic factor that plays an important role in the development, repair, and maintenance of nerve tissue. It is a precursor of saposins A, B, C, and D, which have not been detected in milk. The physiological role of prosaposin in milk is not known, although saposin C, released by digestion, could be important for the growth and development of the young (Patton et al., 1997; Campana et al., 1999; Patton, 1999).

50

2. Milk proteins: An overview

Proteins in the MFGM About 1% of the total protein in milk is in the MFGM. Most of the proteins are present at trace levels, including many of the indigenous enzymes in milk. The principal proteins in the MFGM include mucins, adipophilin, butyrophilin, and xanthine oxidoreductase (XOR). Publications on MFGM proteomics include Fong et al. (2007), Cavaletto et al. (2008), Affolter et al. (2010), Yang et al. (2016), and Demmelmair et al. (2017).

Growth factors Milk contains many peptide hormones, including epidermal growth factor, insulin, insulin-like growth factors 1 and 2, three human growth factors (α1, α2, and β), two mammary-derived growth factors (I and II), colony-stimulating factor, nerve growth factor, platelet-derived growth factor, and bombasin. It is not clear whether these factors play a role in the development of the neonate, or in the development and functioning of the mammary gland, or both (see Fox and Flynn, 1992; Gauthier et al., 2006; Wynn and Sheehy, 2013).

Indigenous milk enzymes Milk contains about 70 indigenous enzymes, which are minor but very important members of the milk protein system (see Fox and Kelly, 2006a,b; O’Mahony et al., 2013; Fox et al., 2015). The enzymes originate from the secretory cells or the blood; many are concentrated in the MFGM and originate in the Golgi membranes of the cell or the cell cytoplasm, some of which becomes entrapped as crescents inside the encircling membrane during exocytosis. Plasmin and lipoprotein lipase are associated with the casein micelles, and several enzymes are present in the milk serum; many of the latter are derived from the MFGM, which is shed as the milk ages. The indigenous enzymes are significant for several reasons: • Deterioration of product quality: plasmin, lipoprotein lipase, acid phosphatase, and XOR • Bactericidal agents: lactoperoxidase and lysozyme • Indices of the thermal history of milk: alkaline phosphatase, γ-glutamyltransferase, and lactoperoxidase • Indices of mastitic infection: catalase, acid phosphatase, and especially N-acetylglucosaminidase The concentration/activity of indigenous enzymes in milk shows greater interspecies variability than that of any other constituent, for example, 3000-fold greater concentration of lysozyme in equine and human milks than in bovine milk. Lactoperoxidase is a major enzyme in bovine milk but is absent from human milk. Human milk and the milks of a few other species contain bile salt-stimulated lipase, but the milks of most species lack this enzyme; the principal lipase in milk is lipoprotein lipase, of which there is
500 times as much in guinea pig milk as in rat milk. Bovine milk has a high level of XOR activity, but all other milks that have been studied have low XOR activity because the protein lacks molybdenum. (XOR plays a major role in the excretion of fat globules from the mammocyte, but it does not act as an enzyme in this function.) The reasons for these interspecies differences are not known, but some of them may be significant.

Analytical considerations for milk proteins

51

Nonprotein nitrogen The nonprotein nitrogen (NPN) fraction of milk contains those nitrogenous compounds that are soluble in 12% TCA; it represents
5% of the total nitrogen (
300 mg/L). The principal components are urea, creatine, uric acid, and amino acids. Human milk contains a high level of taurine, which can be converted to cysteine and may be nutritionally important for infants. Urea, the concentration of which varies considerably, has a significant effect on the heat stability of milk (Walstra and Jenness, 1984). The amino acids present in the NPN fraction also support the growth of lactic acid bacteria.

Biologically active cryptic peptides One of the most exciting recent developments in milk proteins is the discovery that all milk proteins contain sequences that have biological/physiological activities when released by proteolysis. The best studied are phosphopeptides, angiotensin-converting enzyme inhibitory peptides, platelet-modifying peptides, opiate peptides, immunomodulating peptides, and the caseinomacropeptides, which have many biological properties (see FitzGerald and Meisel, 2003; Korhonen, 2006; Korhonen and Pihlanto, 2006; Mills et al., 2011; Nongonierma and FitzGerald, 2015; Park and Nam, 2015; Mohanty et al., 2016).

Analytical considerations for milk proteins The most common, but not all, analytical techniques used in the analysis of milk proteins are summarized in this section.

Total protein determination The principal methods for total protein determination are summarized here: Kjeldahl, Dumas, and infrared spectroscopy. There are many other methods (mainly colorimetric) that are not used routinely for the analysis of milk and dairy products but may be used in research; these include Lowry, Bradford, Biuret, ultraviolet (UV)-visible spectroscopy, and bicinchoninic acid assays (Lowry et al., 1951; Bradford, 1976; Smith et al., 1985; Olson and Markwell, 2007). For both the Kjeldahl method and the Dumas method for protein determination, the nitrogen content of the product is measured indirectly. In milk, the majority (
95%) of the nitrogen is present as “true protein” with the rest being NPN (
5%) (DePeters and Ferguson, 1992; Walstra, 1999). Collectively, true protein and NPN may be referred to as crude proteins. NPN comprises urea, creatine, uric acid, orotic acid, peptides, ammonia, amino acids, and all other nitrogenous materials that are soluble in 12% TCA (Dupont et al., 2013; O’Mahony and Fox, 2014). Kjeldahl method The Kjeldahl method is recognized by Codex Alimentarius as the standard for quantifying milk protein (FAO, 2017). Although initially developed in 1883 by Johan Kjeldahl, the most

52

2. Milk proteins: An overview

recent version of the method is based on interlaboratory studies by Barbano and Clark (1990) and Barbano et al. (1991). The Kjeldahl method involves a three-step approach to the quantification of protein: digestion, distillation, and titration. Digestion of organic material is achieved using concentrated H2SO4, heat, K2SO4 (to raise the boiling point), and a catalyst (e.g., selenium) to speed up the reaction. This process converts any nitrogen in the sample to ammonium sulfate. The digestate is neutralized by the addition of NaOH, which converts the ammonium sulfate to ammonia, which is distilled off and collected in a receiving flask of excess boric acid, forming ammonium borate. The residual boric acid is then titrated with a standard acid with the use of a suitable end-point indicator to estimate the total nitrogen content of the sample. Following determination of the total nitrogen, the use of a specific conversion factor is needed to convert the measured nitrogen content to the crude protein content. A nitrogen-to-protein conversion factor of 6.38 was proposed by Hammarsten and Sebelien in 1892 based on purified acid casein typically containing 15.67% nitrogen (Dupont et al., 2013). The Kjeldahl conversion factor of 6.38 is not accurate for all proteins in milk as it depends on the nitrogen content of a protein. Karman et al. (1987) determined the specific Kjeldahl conversion factors for many of the milk proteins (see Table 2.4). Although nitrogen conversion factors can be used to convert nitrogen content to crude protein content, it is important to acknowledge that the actual content of NPN in milk products is TABLE 2.4 Kjeldahl nitrogen-to-protein conversion factors for milk proteins (Dupont et al., 2013) Without carbohydrate Protein

N%

Kjeldahl factor

αs1-Casein

15.77

6.34

αs2-Casein

15.83

6.30

β-Casein

15.76

6.34

κ-Casein

16.26

6.15

γ-Casein

15.87

6.30

β-Lactoglobulin

15.68

6.38

α-Lactalbumin

16.29

6.14

BSA

16.46

6.07

Ig

16.66

6.00

PP, 8f, 8s

15.30

6.54

PP3

16.97

Lactoferrin

With carbohydrate N%

Kjeldahl factor

15.67

6.38

16.14

6.20

5.89

15.27

6.55

17.48

5.72

16.29

6.14

Transferrin

17.00

5.88

16.10

6.21

Total MFGM

15.15

6.60

14.13

7.08

Milk

15.87

6.30

15.76

6.34

Analytical considerations for milk proteins

53

product specific. Adulteration, whereby nitrogenous compounds are added to milk products to increase the apparent (crude) protein content, is also possible (Finete et al., 2013). Importantly, phospholipids, amino sugars, and nucleic acids can contribute to the measured nitrogen content of milk and dairy products using the Kjeldahl method. Dumas method The Dumas method for nitrogen determination uses combustion (700–1020°C) to convert the organic and inorganic nitrogen in a sample to nitrogen gas, which is then quantified (Tremblay et al., 2003). The Dumas method has many advantages over the Kjeldahl method, such as a shorter analysis time, the possibility for automation, and the absence of toxic reagents. Limitations include the instrumental cost and the inaccuracies associated with using nitrogen conversion factors for estimating the crude protein content of samples. Infrared spectroscopy Infrared (IR) spectroscopy can be used for the quantification of protein. When IR light is applied to a protein sample, IR energy is absorbed at frequencies that correspond to the vibrational energy of selected molecular chemical groups. The IR spectrum can be divided into three regions based on wavelength (or wave number): near-IR (NIR) refers to wavelengths from 0.7 to 2.5 μm and wave numbers from 14,285 to 4000 cm1; mid-IR (MIR) refers to wavelengths from 2.5 to 25 μm and wave numbers from 4000 to 400 cm1; far-IR (FIR) refers to wavelengths from 25 to 100 μm and wave numbers from 400 to 100 cm1 (Dupont et al., 2013). MIR is used mainly to analyze liquid products by transmission, whereas NIR is used to analyze either solid or liquid products by reflectance or transmittance (Dupont et al., 2013). For proteins, MIR analysis by Fourier transform infrared spectrometers is normally used (Wang et al., 2018). Fourier transformation is a mathematical tool that is used to analyze the interference patterns produced by subjecting a sample to IR light (Sheehan, 2009). In milk, the IR absorption band at wave number
1550 cm1 (i.e.,
6.46 μm) is primarily used for protein quantification (Etzion et al., 2004; Dupont et al., 2013). This absorption band is known as Amide II and is one of the resulting bands caused by peptide bond IR absorption. The Amide II absorption band is attributed mainly to NdH bending (40%–60%) and CdN stretching (18%–40%) (Krimm and Bandekar, 1986). Throughout the IR transmission spectrum, other macromolecules (e.g., lipids and carbohydrates) and water also absorb IR energy strongly that may cause interference. Such interference is typically overcome by applying correction factors or spectral subtraction. This can also be an advantage as lipid and carbohydrate analysis can be performed concomitantly (see Fig. 2.3).

Protein separation, purification, identification, and quantification To separate, enrich, purify, identify, and quantify particular milk proteins, certain molecular properties (e.g., solubility, molecular size, molecular charge, polarity, and affinity characteristics) of proteins are exploited. Knowledge of a particular protein’s molecular properties is essential when choosing a suitable analytical method for the particular protein(s) of interest. An overview of some common analytical techniques used to isolate, identify, and quantify the proteins in milk/milk products is given in this section.

54

2. Milk proteins: An overview

100

Fat

90

Water

Protein

80 Transmittance

70

Fat

60 50 Lactose

40 30 20 10 0 3.5

5.7

6.5

9.6

FIG. 2.3 Infrared transmission spectrum of milk versus water. From Dupont, D., Croguennec, T., Brodkorb, A., Kouaouci, R., 2013. Quantitation of proteins in milk and milk products. In: McSweeney, P.L.H., Fox, P.F. (Eds.), Advanced Dairy Chemistry: Volume 1A: Proteins: Basic Aspects, fourth ed. Springer, New York, pp. 87–134.

Electrophoresis methods 1. Native gel electrophoresis Electrophoresis involves the application of a uniform electric field to a liquid sample, causing the separation of the components based (partly) on their electrophoretic mobility. Native gel electrophoresis can be performed using starch, polyacrylamide, or polyacrylamide-agarose gels without the use of a denaturing agent (e.g., SDS). Starch gels give good separation of milk proteins, but such gels are often brittle and opaque after staining (Smithies, 1955; Tremblay et al., 2003); for this reason, polyacrylamide gels are now more widely used. Polymerizing acrylamide to form a gel network of controlled porosity is desirable as the gel is mechanically strong and chemically inert and provides high resolution, and gel gradients are easily achieved to allow the analysis of a wide range of molecular masses (Sheehan, 2009). As native electrophoresis avoids the use of denaturing agents, separation is almost solely based on charge and friction to movement through the gel. The net charge carried by a protein is determined by the pH of the sample and the pH of the gel and the running buffer. The velocity of a protein through the gel matrix is determined by the molecular size/shape of the protein, the viscosity of the buffer, and the pore size of the gel network (Sheehan, 2009). Small proteins will migrate faster through the gel than larger proteins, which experience more resistance to migration. Native electrophoresis retains proteins in their native conformation and can be useful for detecting any changes to the native conformation at set conditions (Arakawa et al., 2006). Once separated, the native proteins can also be excised for further analysis such as mass spectrometry. 2. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) SDS is an anionic surfactant that binds strongly to proteins, mainly via hydrophobic interactions. SDS is a denaturing agent, which, at concentrations greater than 0.1 mM, causes

Analytical considerations for milk proteins

55

conformational changes to proteins (Reynolds and Tanford, 1970). The amount of SDS bound by the protein is proportional to the weight of the protein,
1.4 g SDS/g protein (Reynolds and Tanford, 1970). Each SDS molecule has a net negative charge because of a sulfate group in its polar head. Masking of protein intrinsic charges by SDS gives all proteins a very similar surface charge, and the net negative charge means that all proteins should migrate toward the anode with a constant electrophoretic mobility when a constant electric field is applied (Dupont et al., 2013). When a polyacrylamide gel is used, the sieving effect of the gel network causes proteins of higher MW to travel more slowly, hence separating the proteins based almost entirely on their MW (Shapiro et al., 1967). The procedure involves casting the polyacrylamide gel at a specific gradient suitable for the application of interest. Gels usually have a stacking section and a separating section with different degrees of cross-linking. The most widely used buffer system is the Tris-glycine version described by Laemmli (1970), which was used to study the proteins in the head of bacteriophage T4. Gels can be prepared by polymerizing acrylamide to different levels [e.g., 4% (v/v) in the stacking gel and
12% in the separating gel] with a cross-linker such as methylene bisacrylamide. Water, buffers, and SDS are added to the mixture. The use of a catalyst and a radical initiator begins the polymerization process. The solution is then poured into the cast and allowed to set at optimum temperature and catalyst concentration. A comb is inserted at the top of the gel before it sets to create sample application wells. Alternatively, precast gels are commercially available; they eliminate this step of the process. Precast gels are usually composed of one continuous buffer (Bis-Tris methane) at a pH range 6.4–7.2 (Hachmann and Amshey, 2005). Prior to loading samples into the gel, samples are mixed with sample buffer. The mixture is then heated to 95°C for a few minutes to disrupt secondary and tertiary structures. A reducing agent such as β-mercaptoethanol, dithiothreitol, or dithioerythritol may also be added to disrupt disulfide linkages. Prepared samples are loaded into the gel slots. A MW standard ladder marker is commonly applied to the gel, from which sizes of proteins in the added samples can be estimated. Application of a voltage (typically 150–200 V) causes the proteins to migrate to the anode. Following separation of the proteins, the gels are stained and subsequently destained. Proteins can also be excised from the gels for further analysis (e.g., by mass spectrometry). For staining gels, Coomassie blue, silver, amido black, fast green, fluorescent stains, and immunological staining such as Western blotting are some of the options available. Destaining uses a methanol-water-acetic acid mixture to give a clear background for easy identification of the stained protein bands. Densitometry can be conducted using appropriate software to analyze an image of the gel (usually placed on acetate film and scanned or photographed). As quantitative analysis by gel electrophoresis can be difficult, it remains primarily a qualitative or preparative technique. Mass estimates from SDS-PAGE are termed “apparent molecular mass,” as they depend on comparison with known standards as opposed to direct measurements (Sheehan, 2009). The primary disadvantage of SDS-PAGE is that the proteins are denatured in the course of analysis. Alternative techniques such as native-PAGE do not cause such effects and retain proteins in the naturally folded conformation. β-Casein and αs1-casein behave anomalously in SDS-PAGE, causing inaccuracies in MW estimation (Creamer and Richardson, 1984). This is because β-casein, which has very high surface hydrophobicity, binds a

56

2. Milk proteins: An overview

disproportionately high amount of SDS and, consequently, displays a higher electrophoretic mobility than αs1-casein, although it is a larger molecule. 3. Urea-PAGE Urea is a chaotropic agent that, at high concentrations, denatures proteins by disrupting the hydrogen bonds that stabilize a protein’s secondary, tertiary, and quaternary structures (Sheehan, 2009). Urea has a neutral net charge, meaning that it does not migrate in an electric field. It is also a good dissociating agent, which is very useful for the analysis of casein as it prevents aggregation and precipitation of the proteins (Wake and Baldwin, 1961; Strange et al., 1992; Sheehan, 2009). Similar to SDS-PAGE, a reducing agent may be included in the sample preparation to break disulfide linkages. Urea-PAGE may be a good alternative denaturant to SDS-PAGE when further analysis of excised proteins is required, as the removal of urea is easier than that of SDS (Sheehan, 2009); urea-PAGE is also better for resolution of the caseins, which have similar masses. 4. Capillary electrophoresis In capillary electrophoresis, submillimeter capillary tubes separate proteins in the presence of an electric field through electrolyte solutions or gels. Proteins or peptides are separated based on ionic mobility and mass. The use of thin capillaries with small internal volumes and large surface areas allows the use of a higher voltage for electric field generation as the heat produced is easily removed (Sheehan, 2009). The sample size required is far smaller than for other routine gel electrophoresis methods, and the resolution and the sensitivity are superior. Capillary tubing links a source vial to a destination vial, passing a buffer solution through the system. In the sample vial, the sample is pulled into the capillary tubing by capillary action, pressure, and electroosmotic flow (pH dependent). Within the capillary tubing, the proteins in the sample will migrate because of electroosmotic flow and electrophoretic flow (depending on their attraction to the cathode or the anode). At the outlet of the capillary tube, a detector (e.g., UV or diode array) records the migration of each protein, which provides the basis for the relevant electropherogram produced. The high-resolution separation provided by capillary electrophoresis makes it suitable for subsequent chromatography or mass spectrometry analysis, if desired (Sheehan, 2009). To further understand the use of capillary electrophoresis for milk protein analysis, see de Jong et al. (1993), Recio et al. (1995), Lindeberg (1996), Recio et al. (1997), and Dong (1999). 5. Isoelectric focusing Isoelectric focusing separates proteins based on their isoelectric point. The protein samples are loaded into a polyacrylamide gel that has a pH gradient, before applying an electric field. The gel used can be adapted to the protein of interest by selecting gels that include the desired pH range. When the electric field is applied, the protein will migrate to the pH at which its net charge is zero (its isoelectric point). At pH values above its isoelectric point, a protein will carry a net negative charge and will migrate toward the anode, whereas, at pH values below its isoelectric point, a protein will carry a net positive charge and will migrate toward the cathode (Sheehan, 2009). The resolution of protein separation by isoelectric focusing is controlled by the slope of the pH gradient and the electric field strength (Andrei, 2008). Further reading on the topic of isoelectric focusing for milk protein analysis includes Trieu-Cuot and Gripon (1981), Seibert et al. (1985), Bovenhuis and Verstege (1989), Kim and Jimenez-Flores (1994), and Khaldi and Shields (2011).

Analytical considerations for milk proteins

57

6. Two-dimensional gel electrophoresis Two-dimensional gel electrophoresis (2D GE) was introduced by O’Farrell (1975). This technique combines two molecular properties of proteins to facilitate resolution of proteins that have similar MWs or isoelectric points (O’Donnell et al., 2004). For the first dimension, isoelectric focusing is performed to separate and fix proteins based on their isoelectric point. For the second dimension, the isoelectrically separated protein strip is subjected to standard SDS-PAGE to separate proteins based on their MW. 2D GE of the major milk proteins has also been coupled to mass spectrometry (Galvani et al., 2000, 2001). Reviews on the topic of 2D GE include G€ org et al. (2000) and Westbrook et al. (2001). A disadvantage of 2D GE is that the upper and lower limits of the MW analysis are
150 and
8 kDa, respectively (O’Donnell et al., 2004). In addition, proteins at low concentration, very acidic proteins, or very basic proteins can be difficult to detect, requiring prefractionation of samples in many cases. 7. Lab on a chip Lab on a chip refers to a microfluidic technique, whereby sample preparation, separation, and detection are all conducted on a single-chip surface. The attractions of using this technique over other electrophoresis methods are its miniature form, the extremely short time needed to perform the analysis (1–3 min), the low cost of the analysis, the high sensitivity, and the small sample size required (Nazzaro et al., 2012). Usually, SDS and heat are used to denature the proteins (similar to SDS-PAGE) and then capillary gel electrophoresis, staining, and destaining all occur on the chip surface. Yao et al. (1999) were among the first groups to use microfabricated channels for SDS-PAGE. Since then, the use of this method has increased greatly, especially in the area of DNA analysis. Regarding milk proteins specifically, Anema (2009) compared the method with SDS-PAGE; using the chip method, α-La, β-Lg, and αs-, β-, and κ-caseins were resolved in a milk sample, but immunoglobulins, lactoferrin, and blood serum albumin could not be resolved. Although the sample run time was as low as 3 min per sample, the author claimed that the standard error was too high for reliable analysis (Anema, 2009). The “lab on a chip” technique has been discussed in greater detail by Bousse et al. (2001), Butikofer et al. (2006), and Tran et al. (2009). For further information on the use of electrophoresis to analyze milk proteins, see Swaisgood (1975), Shalabi and Fox (1987), Creamer (1991), Strange et al. (1992), McSweeney and Fox (1997), Tremblay et al. (2003), Chevalier (2011a,b), and Dupont et al. (2013). Chromatography methods 1. Reversed phase (RP)-high-performance liquid chromatography In the simplest form, a HPLC system consists of a pump, a mobile phase, a sample injection port (autosampler), a column packed with a stationary phase, an oven, a detector, and a recorder. The mobile phase refers to a liquid solvent that flows through a specific column at a controlled rate with the aid of a pump. The column is filled with immobilized, solid, adsorbent material known as the stationary phase (Aguilar, 2004). The sample of interest is injected into the mobile phase and passes through the column under controlled pressure. Each individual component of the injected sample interacts differently with the material in the stationary phase and, therefore, travels at a different speed through the column (retention time postinjection). The composition of the mobile phase used can remain constant throughout a run (isocratic) or can be adjusted (gradient) to accelerate sample elution from the system. The mobile phase often includes an ion-pairing agent

58

2. Milk proteins: An overview

(such as 0.1% trifluoroacetic acid), which behaves as a counterion to the ionic species in the sample being analyzed and creates an ion pair of sufficient hydrophobicity to be retained by the stationary phase (Sheehan, 2013). Upon elution of the sample components from the column, they pass through light of a specific wavelength, and their absorbance is measured using a detector. The absorbance is recorded throughout a HPLC run, allowing the data to be translated into a chromatogram output of absorbance as a function of time. RP-HPLC separates proteins using a nonpolar stationary phase. This solid material usually consists of surface-modified silica particles. The more polar sample components will be eluted first with more nonpolar components having longer retention times. RP-HPLC commonly uses gradient elution, whereby the polarity of the mobile phase is increased over the duration of the run. Usually, a UV, fluorescence, or diode array detector is employed for the identification of the proteins by HPLC. The specific wavelength used for detection depends on the sample being analyzed; for example, peptide bonds absorb light at 214 nm and aromatic side chains absorb light at 280 nm. RP-HPLC is widely used for protein separation, identification, and quantification in research groups analyzing milk proteins. Fig. 2.4 provides an example of a typical RP-HPLC chromatogram showing the separation of some major milk proteins. Some studies using RP-HPLC for milk protein analysis include Andrews et al. (1985), Carles (1986), le Maire et al. (1986), Gonzalez de Llano et al. (1990), Parris and Baginski (1991), Parris et al. (1991), Visser et al. (1991), Strange et al. (1992), Leonil et al. (1995), Bobe et al. (1998), Manso et al. (2005), Bonfatti et al. (2008), Wang et al. (2009), Le et al. (2017), and Ma et al. (2017a). 2. Size exclusion chromatography Size exclusion chromatography, also known as gel filtration or gel permeation chromatography, separates proteins based on their hydrodynamic volume using a porous column αs1-Casein

(A)

α-Lactalbumin β-Casein

Area units (–)

κ-Casein

β-Lactoglobulin

(B)

(C)

Retention time (min)

FIG. 2.4 Typical reversed-phase high-performance liquid chromatography chromatograms showing major milk proteins in skim milk (A), permeate from microfiltration of skim milk (B), and retentate from microfiltration of skim milk (C) (Crowley, 2016).

Analytical considerations for milk proteins

59

that is composed of cross-linked polymers such as dextran and agarose. Pores in the column trap and retard the migration of smaller molecules, whereas larger molecules pass through more quickly. Therefore, larger molecules elute first (short retention times), and smaller molecules elute later (longer retention times). If a calibration curve (retention time versus MW) is obtained, using molecules of known MW, size exclusion chromatography may be used to determine the MW of a protein. It is important to use standards that are conformationally similar to the samples of interest for MW calibration, as structural differences can lead to differences in elution behavior and hence inaccuracies in MW calculations (le Maire et al., 1986; Goetz et al., 2004). With the availability of size exclusion chromatography columns of various pore sizes, the range of MW that can be analyzed by this technique is vast (from 0.1 to
10,000 kDa) (Goetz et al., 2004). A major advantage of liquid chromatography methods is their noninvasive impact on the protein sample being analyzed; this can be useful when further analysis (e.g., mass spectrometry) is desired following analyte elution. 3. Ion-exchange chromatography Ion-exchange chromatography (IEX) separates proteins based on their electrostatic attraction to charged sites on the ion-exchange resin (stationary phase). The net charge carried by milk proteins depends on the pH of the environment in which they are present. If the pH is below the isoelectric point of a protein, it will carry a net positive charge, whereas, at pH values above the isoelectric point of a protein, it will carry a net negative charge. The net charge carried by a protein will determine its electrostatic affinity to the charged stationary phase in ion-exchange columns. If the protein is oppositely charged to the resin, it will be attracted to it and adsorb, whereas unadsorbed material will be eluted from the ion exchanger. At the end of an IEX run, the adsorbed material is washed and desorbed from the resin for recovery. The desorption of resin-bound proteins can be achieved using techniques such as pH adjustment or salt solution addition to reduce the protein affinity to the resin. IEX is primarily used for protein fractionation. For further information on the use of IEX in the analysis of milk proteins, refer to the following: Davies and Law (1977), Manji et al. (1985), Guillou et al. (1987), Hollar et al. (1991), Gerberding and Byers (1998), Fee and Chand (2006), Holland et al. (2010), and Faraji et al. (2017). 4. Mass spectrometry methods Mass spectrometry (MS) allows proteins to be resolved, identified, and quantified. The molecular mass and the amino acid sequence of a protein can be determined as well as information about any posttranslational modifications, such as the type and positioning of the modifications. In MS, proteins are first separated by some of the aforementioned separation techniques (electrophoresis and chromatography) before being analyzed by the mass spectrometer based on differences in mass-to-charge ratio. Depending on the type of experimentation, MS analysis may require the protein sample to be adequately digested prior to analysis. This can be achieved before, during, or after separation depending on the preparative technique selected, and the peptides produced on proteolysis can be identified by comparison with peptide databases (O’Donnell et al., 2004). A common partnership for protein analysis using MS is liquid chromatography coupled with mass spectrometry (LC-MS). The primary advantage of LC-MS is that molecules are analyzed by MS upon elution from the chromatography column, thereby combining sample separation with the MS analysis. 2D GE coupled with MS is also a useful combination and has been used for the analysis of milk proteins (Appella et al., 2000).

60

2. Milk proteins: An overview

In the simplest description, a mass spectrometer system may be considered to be an ionization device that is coupled to a mass analyzer and a detector (O’Donnell et al., 2004). MS analyzers filter different analytes based on their behavior in an electric or electromagnetic field, therefore ionization of molecules in the mass spectrometer is required before they are sublimed into the gaseous phase for mass analysis. The mass analyzer, such as quadrupole (Q), time-of-flight (TOF), ion trap, ion cyclotron resonance analyzer, or orbitrap mass analyzer, selects analytes based on the mass-to-charge ratio (m/z) of each ion produced. For a given set of conditions within the electric or electromagnetic field, only ions with a particular m/z will have the stable trajectory required to reach the detector. By varying these conditions, each m/z can be sequentially selected for and the detector is then used to record the number of ions present at each m/z by converting the ion current into an electrical current (Fenn et al., 1989; Mamone et al., 2009; Dupont et al., 2013; Le et al., 2017). MS may be conducted in one single stage or in tandem, whereby two mass analyzers are used; these can be of the same type or different (hybrid) and usually feature a collision cell in between to fragment analytes. The use of a single-stage mass spectrometer measures molecular mass, whereas a multistage system allows more detailed analysis to be achieved by working on the fragmented ion from the initial mass analysis (O’Donnell et al., 2004). Analytes fragment in predictable and reproducible patterns based on their composition and structure. Therefore, if two unrelated analytes share the same m/z value, these precursor ions will not fragment in the same way and can subsequently be distinguished based on their product ions. Tandem MS may be used for applications including amino acid sequencing and the analysis of protein modifications. Some examples of multistage mass analyzer combinations used in sequence include triple quadrupole, TOF-TOF and Q-TOF (Morris et al., 1996; Dupont et al., 2013). Quadrupole refers to a mass filter, which uses metal rods to generate an electric field, and a mass spectrum is created by detecting ion trajectories through the system (Morris et al., 1996). Ionization of protein molecules for MS predominantly uses either matrix-assisted laser desorption ionization (MALDI) or electrospray ionization (ESI). The development of these “soft” techniques in the 1980s has allowed macromolecules, such as proteins, to be ionized and transformed to the gaseous phase mostly intact for MS, whereas the solvent is evaporated (Alomirah et al., 2000). MALDI, first developed by Karas and Hillenkamp (1988), generates gaseous, charged molecules by mixing an excess of dissolved matrix material with the liquid sample to be analyzed. Both the matrix material and the sample need to be soluble under the same conditions to prevent precipitation. The mixture of matrix and sample is applied to the tip of a mass spectrometer and allowed to dry, whereby crystallization of the matrix occurs with the sample molecules scattered throughout. Once dried, the mixture is irradiated by nanosecond nitrogen laser pulses (usually 337 nm) causing the ionization of molecules. Common matrices used for protein include ferulic acid, sinapic acid, α-cyano-4-hydroxycinnamic acid, and 2,5-dihydroxybenzoic acid (Vorm et al., 1994). The choice of matrix can be influenced by the type of MS being conducted, as ions need to be stable for milliseconds in ion trap experiments, whereas the requirement is microseconds in TOF experiments (Mann et al., 2001). Depending on the matrix used, molecule fragmentation levels can differ and hence affect the output and stability of the ionic species produced. As MALDI can fragment proteins and impact the analysis, it is used mostly for the analysis of peptides or proteins up to
300 kDa (Mann et al., 2001; Goetz et al., 2004). MALDI is mostly used in combination with TOF mass analyzers

Analytical considerations for milk proteins

61

(Goetz et al., 2004). In MALDI-TOF, the short bursts of ions produced by the laser are accelerated along a flight path in a vacuum (Dupont et al., 2013). Smaller ions travel faster than larger ions, leading to the formation of a TOF spectrum (Mann et al., 2001). ESI, developed by Fenn et al. (1989), works differently from MALDI and involves pumping the sample/solvent mixture slowly through a needle at high voltage, producing a fine spray of electrostatically charged droplets in the atmosphere (versus in vacuum with MALDI) (Mann et al., 2001). These droplets rapidly desolvate, resulting in the ionization of the sample molecules (Dupont et al., 2013), and the spray produced is typically stabilized by a nebulizer gas (Mann et al., 2001). Unlike MALDI which typically produces ions with a charge state of +1, ESI often produces multiple charge states, particularly for relatively large molecules that have several sites available for charging. This results in the molecule being detected at multiple m/z, sometimes needing a mathematical algorithm to deconvolute the data for molecular mass determination (Labowsky et al., 1993). The ESI device is often combined with quadrupole mass analyzers. Another type of mass analyzer used in protein analysis is the ion trap mass spectrometer. This instrument traps the beam of ions in a three-dimensional electric field before selectively ejecting ions one by one for detection, producing a mass spectrum (Louris et al., 1987). The analytical capability of mass spectrometers is superior to that of most other instruments in protein research laboratories. The high sensitivity (0.1% mass accuracy with sensitivity to pico- and femtomole) and the fast run times make its contribution invaluable in proteomics, including the detection of both natural and process-induced protein modifications (Alomirah et al., 2000; Le et al., 2017). Natural modifications refer to genetic polymorphisms and posttranslational modifications such as phosphorylation, glycosylation, and disulfide bond formation (Ng-Kwai-Hang and Grosclaude, 2003; Martin et al., 2013b; Le et al., 2017). Process-induced protein modifications are a result of oxidation, glycation, deamidation, cross-linking, and proteolysis, for example. Both types of modification can affect the quality, functionality, health effects, and applications of milk protein ingredients. Le et al. (2017) provide further information on the different types of protein modification encountered as well as some published studies that have exploited the suitability of MS for detecting such modifications. A more in-depth understanding of MS for milk protein analysis can be obtained from Leonil et al. (1995), Catinella et al. (1996), Alomirah et al. (2000), Mann et al. (2001), O’Donnell et al. (2004), Cunsolo et al. (2011), Dupont et al. (2013), and Le et al. (2017).

Protein structural analysis The structure of protein is categorized into four levels: primary, secondary, tertiary, and quaternary. Primary sequence refers to the chain of genetically determined amino acid residues that make up a polypeptide backbone. Individual amino acids are connected to each other by peptide bonds, which join the amino group of one amino acid with the carboxylic group on another, forming an amide bond. Secondary structure can be in the form of α-helices, β-sheets, or β-turns, and any structures that are not encompassed in these categories are termed “‘random coil” or “unordered structure.” Tertiary structure arises when the protein’s secondary structures fold and pack tightly because of interactions between the side chains of individual amino acids. The associative interactions that occur between side chains are usually weak and can be disrupted easily, causing unfolding, when exposed to thermal or

62

2. Milk proteins: An overview

chemical denaturants. Interactions include hydrophobic, electrostatic, van der Waals, hydrogen bonds, and disulfide bonds. The final level of protein structure is the quaternary structure, which occurs when individual protein molecules associate because of interactions similar to those contributing to tertiary structure (e.g., micelle formation of casein molecules). Fourier transform infrared spectroscopy Fourier transform infrared (FTIR) spectroscopy is a rapid, nondestructive technique that can be used to analyze the secondary structure of proteins. This technique can be conducted rapidly and represents a more accessible alternative to techniques such as X-ray crystallography or nuclear magnetic resonance, which provide detail on tertiary protein structure. The principle of FTIR spectroscopy is based on the ability of samples to absorb certain wavelengths of IR light, each sample producing its own characteristic spectrum (Gallagher, 1997). A typical IR spectrum shows the intensity of IR absorption by a sample across a specific range of wavelengths (Kong and Yu, 2007). For protein analysis, these data are interpreted as vibrations to some structural regions of the polypeptide chain, of which there are nine characteristic IR absorption bands (amide A, B, and I–VII) (Kong and Yu, 2007). Of the nine IR absorption bands, it is primarily the amide I and amide II bands that are used for structural interpretation. The amide I absorption band (1600–1700 cm1) arises mainly (
80%) from vibrations caused by C]O stretching (Byler and Susi, 1986). This band is a broad overlap of many narrower bands, each representative of individual contributory structural units, which can be resolved by mathematical deconvolution methods (e.g., Fourier selfdeconvolution and second derivative analysis) (Kong and Yu, 2007). Table 2.5 outlines some TABLE 2.5 Deconvoluted amide 1 band assignments of protein secondary structure in water (Yang et al., 2015) Wave number (cm21)

Band assignment

1624  1.0

β-Sheet

1627  2.0

β-Sheet

1633  2.0

β-Sheet

1638  2.0

β-Sheet

1642  1.0

β-Sheet

1648  2.0

Random

1656  2.0

α-Helix

1663  3.0

310-Helix

1667  1.0

β-Turn

1675  1.0

β-Turn

1680  2.0

β-Turn

1685  2.0

β-Turn

1691  2.0

β-Sheet

1696  2.0

β-Sheet

Analytical considerations for milk proteins

63

of the amide I band assignments following deconvolution. These frequencies can differ depending on the medium. The reason for the individual narrow bands is that the C]O stretching of each type of secondary structure occurs at slightly different frequencies because of its molecular geometry and hydrogen bonding (Kong and Yu, 2007). Upon resolution of the individual secondary structure features, quantitative estimations or changes developed because of treatments can be analyzed (Dong et al., 2000). The amide II band is formed mainly from in-plane NdH bending (40%–60%) and CdN stretching vibrations (18%–40%) (Krimm and Bandekar, 1986). However, as the amide II band is not as useful as the amide I band for the quantification of protein secondary structure, the focus lies primarily with the amide I absorption band (Gallagher, 1997). Fig. 2.5 illustrates the vibrations within peptide bonds responsible for the amide I and amide II infrared absorption bands. The use of FTIR spectroscopy for secondary structural analysis of proteins has been greatly enhanced by the development of biomolecule-specific accessories. Such accessories include insulated temperature control cells, attenuated total reflectance cells, transmission cells, and software packages that allow deconvolution and spectral subtraction to remove overlapping IR absorption bands of interfering compounds such as water (Dupont et al., 2013). This means that protein analysis can be performed in both liquid form and solid form. Further information on FTIR spectroscopy for analyzing protein structure is available from Byler and Susi (1986), Kumosinski and Farrell (1993), Jackson and Mantsch (1995), Gallagher (1997), Haris and Severcan (1999), Singh (1999), Barth (2007), Kong and Yu (2007), Carbonaro and Nucara (2010), and Smith (2011). Circular dichroism Circular dichroism (CD) is a form of chiroptical spectroscopy that uses a spectropolarimeter to measure differences in the absorption of left and right circularly polarized light rays by an asymmetric molecule (Woody, 1996). Spectropolarimeters are usually run in modulation mode, whereby the incident light source is switched between left and right circularly polarized light. This continuous switching is achieved by applying an alternating electric field to a modulator, which is usually a quartz crystal accompanied by a thin plate of isotropic material (Kelly et al., 2005a; Sheehan, 2009). Light passing through an optically active sample can be refracted or absorbed, absorption being measured as the molar extinction coefficient. After circularly polarized light passes through an optically active chiral sample, there is a shift from circularity (because of absorbance differences between left and right circularly polarized light), and instead, an elliptical path is traced by the electric vector around the axis of the transmitted beam (Sheehan, 2009). As proteins are chiral molecules, they absorb circularly polarized light unequally; hence, the light leaving the sample becomes elliptical as opposed FIG. 2.5 The vibrations in peptide bonds responsible for

O C

the amide I and amide II infrared absorption bands of proteins.

O

=

=

Amide I vibration

N H

C

N H

Amide II vibration

64

2. Milk proteins: An overview

to circular (Woody, 1996). The CD signal is 103–105 times the normal absorbance and, when detected by the photomultiplier of the spectropolarimeter, converts absorbance into ellipticity (θ, degree of deviation from circularity) with units of millidegrees (Kelly et al., 2005a; Dupont et al., 2013). This unit can be positive or negative, depending on the differential absorption of left circularly polarized light and right circularly polarized light (Dupont et al., 2013). CD spectra (a plot of ellipticity over wavelength) of proteins arise mainly from the signals of peptide bonds (absorption TR=>TR[BP = 5564.3, 1441]

22---------C----37

Voyager Spec #1=>BC=>RSM500=>MC=>TR=>TR[BP = 4013.9, 1583]

4014 1583 22---------C------------45 100 5568 80 60 40 5182 4223 4655 4967 20 0 0 3900 4300 4700 5100 5500 5900

22---------C----37

22---------C----37

22---------C---------42

22---------C------------45 5563 2980

Voyager Spec #1=>BC=>RSM500=>MC=>TR[BP = 5562.2, 2981]

100 80 4015 60 40 20 0 3900 4300

4650 4963 4700

5100

5507 5500

0 5900

22---------C----37 5565 1441 100 90------------------C--------118 80 60 4014 40 5183 4514 20 0 0 3900 4300 4700 5100 5500 5900

22---------C----37

Voyager Spec #1=>BC=>RSM500=>MC=>TR[BP = 5568.6, 844]

5567 843.6 90------------------C--------118 100 4014 5184 80 60 40 4517 20 0 0 3900 4300 4700 5100 5500 5900

FIG. 5.7 Distribution of disulfide-linked isoforms of κ-casein on a nonreducing 2-D gel. Dimers, trimers, and tetramers of κ-casein are labeled to show the participating monomeric forms. The dimers and trimers run as doublets depending on the disulfide linkages. Matrix-assisted laser desorption/ionization time of flight mass spectra show the disulfide-linked peptides obtained from tryptic digests of the homodimer and homotrimer of κ-casein B-1P.

Sources and significance of casein heterogeneity

185

genetically different traits. The unique Ser-Xxx-Asp motif of Ser56 in αs1-casein-9P might be phosphorylated by a different enzyme than that in αs1-casein-8P, because protein kinase enzymes differ in their specificity for phosphorylation sites (Ubersax and Ferrell, 2007). The Golgi-enriched fraction casein kinase (GEF-CK) has been recognized to phosphorylate caseins (Moore et al., 1985). This fraction has a consensus sequence for the Ser-Xxx-Glu/pSer motif and fails to recognize Asp in a Ser-Xxx-Asp motif (Lasa-Benito et al., 1996). In addition, it was shown that the kinase FAM20C, which is one of the candidates for the GEK-CK fraction, specifically phosphorylates the Ser-Xxx-Glu motif (Tagliabracci et al., 2012). It is not known which enzyme could be responsible for the phosphorylation of the Ser-Xxx-Asp motif in αs1-casein-9P. Although αs1-casein and αs2-casein are two different proteins that exhibit different phosphorylation patterns, they seem to be connected by the phosphorylation regulatory system in the mammary gland. The degree of phosphorylation of αs1-casein was found to be correlated to that of αs2-casein (Heck et al., 2008; Fang et al., 2018). Further studies on phenotypic correlations and hierarchical clustering (Fig. 5.8) between αs1-casein with 8 and 9 phosphate groups and αs2-casein with from 9 to 14 phosphate groups suggest that there are at least two regulatory systems for αs-casein phosphorylation (Fang et al., 2016). One system would be responsible for αs-casein with low levels of phosphorylation (αs1-casein-8P, αs2-casein-10P,

FIG. 5.8 Dendrogram of phenotypic correlations among αs-caseins with different levels of phosphorylation. Reprinted from Fang, Z.H., Visker, M.H.P.W., Miranda, G., Delacroix-Buchet, A., Bovenhuis, H., Martin, P., 2016. The relationships among bovine alpha(S)-casein phosphorylation isoforms suggest different phosphorylation pathways. J. Dairy Sci. 99, 8168–8177. https://doi.org/10.3168/jds.2016-11250.

186

5. Posttranslational modifications of caseins

and αs2-casein-11P), and one would be responsible for higher levels of phosphorylation (αs1casein-9P, αs2-casein-12P, αs2-casein-13P, and αs2-casein-14P). Fang et al. (2016) suggested that the αs-caseins with higher levels of phosphorylation may be phosphorylated by another mammary casein kinase on its four Thr-Xxx-Glu motifs. Parity, stage of lactation, genetic variation of cows, and breeds have been indicated in changes in the degree of phosphorylation of the αs-caseins (Buitenhuis et al., 2016; Fang et al., 2017, 2018). Functional significance A few decades ago, it was not possible to determine the variation in the degree of phosphorylation of the αs-caseins; however, with the use of capillary zone electrophoresis or improved separation profiles by liquid chromatography, this is feasible today (Heck et al., 2008; Bonfatti et al., 2009). The use of these techniques has opened up the possibility of studying the functional significance of the phosphorylation of αs-casein. The coagulation properties of milk are important for cheesemaking. Interesting correlations between the phosphorylation of αscasein and the coagulation properties of milk have been found (Frederiksen et al., 2011; Jensen et al., 2012b; Poulsen et al., 2016). Improved milk coagulation was associated with higher fractions of αs1-8P and αs2-11P. Variation in the phosphorylation of αs1-casein also seems to affect the hydrolysis of milk by plasmin (Zhang et al., 2018), which can play a role in the destabilization of UHT milk during storage. UHT milk incubated with plasmin showed a faster decrease in αs1-9P than αs1-8P during the onset of gelation. Bijl et al. (2014c) determined proteolysis during ripening in a model cheese system. Chymosin-induced hydrolysis of αs1-casein-8P and αs1-casein-9P was monitored over time in milk gels. It was found that, at 50% hydrolysis, 15% more αs1-8P than αs1-9P was hydrolyzed. The authors suggested that changes in the physical conformation of the caseins were responsible for the observed differences. When the secondary structure of αs1-8P is compared with that of αs1-9P, the ninth phosphate group is positioned just in front of an alpha-helix region (Creamer et al., 1981; Kumosinski et al., 1991). Therefore, it is likely that the additional phosphate group on this specific position will impact secondary structure formation. Additionally, the ninth phosphate group might affect calcium phosphate nanocluster formation because it can form a strong binding site together with two other phosphorylated residues in close proximity (Holt, 2004; Bijl et al., 2018).

κ-Casein heterogeneity Although the heterogeneity of κ-casein has been recognized for many decades and the structural elements are now fairly well defined, less is known about the source of the heterogeneity, particularly its glycosylation, and its functional role(s). One of the reasons that this subject has not been well studied has been the absence of high-throughput characterization methods. Rapid developments in the field of “glycoproteomics” are leading to new insights on the topic. The available techniques are beyond the scope of this chapter but have been summarized elsewhere (Wuhrer et al., 2007; Le et al., 2017; Mulagapati et al., 2017; Darula and Medzihradszky, 2018).

Sources and significance of casein heterogeneity

187

Early studies on the influence of the glycosylation of κ-casein on its functional and biological properties have been reviewed previously (Dziuba and Minikiewicz, 1996). They highlighted a number of studies addressing factors that could influence the degree of glycosylation and the influence that glycosylation might have on micelle stability. The following sections cover some of those studies again and highlight more recent work related to the sources, functional significance, and biological significance of κ-casein heterogeneity. Sources of heterogeneity A large number of studies have examined the influence of milk protein polymorphism on milk composition and yield, and these have been extensively reviewed (e.g., Ng-Kwai-Hang, 1997; Martin et al., 2002; Heck et al., 2009). However, in many cases, the results have been inconsistent, which is probably a reflection of the multifactorial nature of milk production. It is difficult to isolate the effects of the polymorphism of a single protein from those of the other major milk proteins, especially as there appears to be a substantial degree of coordination of their expression. There are also a number of environmental or cow-related factors such as feed type and lactation stage that are frequently interrelated, as they all vary with the seasonal changes in dairy farming. Studies on specific effects of κ-casein variants have largely focused on the common A and B variants, and there appears to be a consensus that milk from B variant cows contains more fat, protein, casein, and κ-casein than milk from A variant cows (Bovenhuis et al., 1992; Ng-Kwai-Hang, 1997; Bobe et al., 1999). Progress has been made in studies relating to the glycosylation status of κ-casein. Robitaille et al. (1991a) identified a number of factors that appeared to contribute to variation in the neuraminic acid (NeuAc) content of bovine κ-casein. The NeuAc content was higher in cows with the κ-casein AB phenotype than in cows with the AA phenotype; it decreased with increasing parity and varied over the course of lactation, dropping to a minimum at 2–3 months after calving before increasing over the next 9–10 months. They also examined the association between the glycosylation of κ-casein and milk production/composition (Robitaille et al., 1991b). Although there appeared to be a statistically significant association between the NeuAc content of κ-casein and the milk yield, the most striking result of these investigations was the variability of the NeuAc/κ-casein measurements (mean, 64
21 μg/mg; range, 23–166 μg/mg), which suggests that other factors could have had a large impact on glycosylation or that the inherent variability in the assay masked any true associations. Limited 2-D gel analyses suggest that the pattern of glycosylation is far more consistent than these measurements indicate (Holland et al., 2004, 2005). Significant differences in the content of nonglycosylated κ-casein in milk have been reported for cows of different κ-casein genotypes (Lodes et al., 1996). Nonglycosylated κ-casein (as a percentage of total protein) was determined by electrophoretic analysis. The levels were higher in milk from cows with the B variant than in milk from cows with the A variant. The rarer variants, C and E, were generally associated with lower levels. However, as no measurements of glycosylated κ-casein levels were reported, no effect of genetic variant on glycosylation can be inferred from this report. Electrophoretic and chromatographic techniques have been used to profile the caseinomacropeptide from cows of the AA and BB phenotypes. Coolbear et al. (1996) found that the B variant macropeptide was more highly glycosylated than the A variant macropeptide, with an increased content of both hexosamine (i.e., GalNAc) and sialic acid (i.e., NeuAc). After

188

5. Posttranslational modifications of caseins

anion-exchange HPLC on a MonoQ column, the elution profile of the B variant contained more peaks, suggesting that an increased number of oligosaccharide chains were attached. These results were consistent with other studies suggesting more extensive glycosylation of the B variant than the A variant (Vreeman et al., 1986; Molle and Leonil, 1995; Bijl et al., 2014a), despite the fact that the A variant contains an extra (potential) glycosylation site (Pisano et al., 1994). From these and other results, Coolbear et al. (1996) suggested that there were generally consistent patterns of glycosylation for the genetic variants but that the overall extent of glycosylation could vary. It was also discovered that the effect of the κ-casein variant on the relative contents of glycosylated and nonglycosylated κ-casein was consistent across breeds (Poulsen et al., 2016). Variations in κ-casein glycosylation during lactation have already been touched on. Early studies indicated a higher degree of glycosylation of κ-casein in colostrum than in mature milk as well as the presence of two additional sugar moieties, N-acetylglucosamine (GlcNAc) (Guerin et al., 1974; Fournet et al., 1975) and fucose (Fiat et al., 1988). Subsequently, a number of studies addressed the structure of the oligosaccharides attached to colostral κ-casein and how they varied with time after parturition (Saito et al., 1981a, b, 1982; van Halbeek et al., 1981; Fiat et al., 1988). As well as the structures in normal milk that we have already identified, the following structures have been reported: the acidic hexasaccharide, NeuAcα(2–3)Galβ(1–3) [NeuAcα(2–3)Galβ(1–4)GlcNAcβ(1–6)]GalNAc; the acidic pentasaccharide, NeuAcα(2–3)Galβ (1–3)[Galβ(1–4)GlcNAcβ(1–6)]GalNAc; the acidic tetrasaccharide, GlcNAcβ(1–3)Galβ(1–3) [NeuAcα(2–6)]GalNAc; the neutral pentasaccharide Galβ(1–3)[Galβ(1–4){Fucα(1–3)}GlcNAcβ (1–6)]GalNAc; the neutral tetrasaccharide, Galβ(1–3)[Galβ(1–4)GlcNAcβ(1–6)]GalNAc; and the neutral trisaccharide, Galβ(1–3)[GlcNAcβ(1–6)]GalNAc. This extra complexity is already observable 15 min after parturition but decreases to normal over about 66 h (Fiat et al., 1988). These results suggest changes in the expression profiles of the glycosyltransferases that are responsible for assembling the O-linked glycans on κ-casein. The initial step of attachment of GalNAc to a threonine residue, also known as a mucin-type linkage, is catalyzed by UDPGalNAc:polypeptide N-acetylgalactosaminyltransferase (ppGaNTases) (Ten Hagen et al., 2003). In building the tetrasaccharide, the next steps involve transfer of Gal from UDP-Gal and of Neu5Ac from CMP-Neu5Ac donor sugars by galactosyltransferase and sialyltransferases, respectively (Chokhawala and Chen, 2007). The formation of linear and branched glycans follows the same biosynthetic pathways as other mucin-type O-glycans (Corfield and Berry, 2015). Hydrolysis of the glycosidic bonds is catalyzed by glycosidase enzymes. Elevated levels of glycosidase activities have been detected in bovine colostrum (O’Riordan et al., 2014a). Of seven different glucosidases, N-acetyl-β-D-glucosaminidase, α-Lfucosidase, α-galactosidase, and N-acetyl-neuraminidase appeared to be the most biologically relevant. In mature milk, the levels decreased to a minimal but constant level. Also, between animals, only a low level of variation was observed. It is therefore important to consider both the expression profiles of glycosyltransferases and the activity of glycosidases in studying glycosylation during lactation. Another source of variation in glycosylation status that is related to lactation is the length of the cow’s dry period. In cows with a zero-day dry period, the relative amount of glycosylated κ-casein in the protein fraction, as determined by HPLC, increased from 8% to 12% between 6 and 2 weeks prepartum (de Vries et al., 2015). Furthermore, after calving, glycosylation was higher for cows with a zero-day dry period (6.7%) than for cows with a normal dry period

Sources and significance of casein heterogeneity

189

length of 60 days (5.2%). There was a negative correlation between glycosylation and milk yield 2-week postpartum, which was suggested to be related to a reduced productivity of the mammary epithelial cells. In line with these studies, Maciel et al. (2017) found a significant effect of days before calving (DBC), determined at 180, 90, and 60 days before expected calving, on the level of glycosylation relative to total κ-casein. Between 180 and 90 DBC, the level increased from 42.4% to 47.5%, while the calving interval of either 15 or 18 months did not show significant differences in level of glycosylation. Other seasonal factors related to climate, such as heat stress, drought, and nutrition (e.g., pasture vs. fodder), can have an impact on milk production and composition. However, we are not aware of any specific studies on their effect on κ-casein glycosylation. The genetic background of κ-casein glycosylation (expressed as percentage on total protein) was studied by Bonfatti et al. (2014) and Buitenhuis et al. (2016). Simmental breed had a high heritability of 0.46, also Danish Holsteins showed high heritability of 0.64, while Danish Jerseys had a lower heritability of 0.14. Buitenhuis et al. (2016) detected several SNPs that were specific for κ-casein glycosylation. Although no obvious candidate genes were found for genetic regulation of κ-casein glycosylation, two genes were related to the PTM of caseins. The first gene was casein kinase 1, gamma 3 (CSNK1G3) on BTA7, which is a serine-/threonine-specific protein kinase that phosphorylates caseins (Rowles et al., 1991). The second gene was protein kinase c. theta (PRKCQ) on BTA13, which belongs to the family of protein kinase C (PKC) that are also involved in phosphorylation of serine and threonine. If these genes have a specific role in glycosylation remains to be elucidated. Functional significance Casein micelle size

κ-Casein plays a key role in micelle stability by acting as a hairy layer that provides both steric repulsion and electrostatic repulsion between micelles, preventing aggregation. Glycosylation of κ-casein increases both the size of the hydrophilic C-terminal “hairs” and their charge, because of the bulk of the hydrophilic sugar residues with their hydration shells and the negative charge of the neuraminic acid groups, respectively. Theoretically, the higher is the degree of glycosylation of κ-casein, the greater should be its stabilizing ability. As such, it might be expected that the degree of glycosylation of κ-casein would have a marked effect on both the size and the stability of the casein micelles. Several studies have made use of bulk milk to determine correlations between size and milk composition (McGann et al., 1980; Davies and Law, 1983; Donnelly et al., 1984; Dalgleish et al., 1989). Casein micelles from bulk milk were fractionated into different size classes, and it was shown that their size was inversely related to their κ-casein content (relative to total casein). In studies on bulk milk, there was no clear correlation between micelle size and degree of glycosylation. Slattery (1978) found an apparent inverse relationship between the proportion of glycosylated κ-casein and the micelle size, but it did not apply to all of the size fractions isolated. In contrast, Dalgleish (1985, 1986) found that the proportions of glycosylated and nonglycosylated κ-casein did not vary with the micelle size. The studies of Yoshikawa et al. (1982) and O’Connell and Fox (2000) showed an apparent increase in κ-casein glycosylation with increasing micelle size. Some of these discrepancies are probably the result of differences in the method of fractionation and differences in

190

5. Posttranslational modifications of caseins

the analysis methods of glycosylated κ-casein. Also, variation in the bulk milk itself might have caused the contradictory findings. More recently, the milk of individual cows was studied to determine the link between casein micelle size and milk composition (Bijl et al., 2014a). Two sets of nine cows that produced milk with high or low average casein micelle size were selected in this study, and the detailed milk composition, including the degree of glycosylation of κ-casein, was determined. It was found that the average casein micelle size in milk of individual cows correlated significantly with genetic variants A and B of κ-casein and the degree of glycosylation of κ-casein. Small casein micelles were associated with the B variant of κ-casein, as was also shown in bulk milk studies. Interestingly, small casein micelle size was also associated with a high percentage of glycosylated κ-casein in the total protein. It was suggested that glycosylation influences micelle stabilization during or after the formation of the casein micelle in the mammary gland and thereby its size. The underlying mechanism is unknown and needs further study. One way to approach this is to study the effect of purified fractions on casein micelle stability. Takeuchi et al. (1985) used ion-exchange chromatography to prepare nine fractions of κ-casein A-1P that varied in their level of glycosylation. The ability of these subfractions to stabilize αs1-casein was shown to increase with increasing carbohydrate content. Further study on purified casein fractions may help to increase our understanding of casein micelle stability and the role of PTMs. Functionality in cheese

As already stated, micelle stability, or controlled destabilization in the case of cheese and yogurt manufacture, is of key importance in dairy manufacture. A number of authors have investigated the effects of κ-casein heterogeneity on micellar aggregation and the processing properties of milk. In cheese manufacture, the initial step is the chymosin-(rennet)catalyzed cleavage of the Phe126-Met127 bond in κ-casein, resulting in the release of the hydrophilic caseinomacropeptide from the micelle surface, which leads to micellar aggregation or clotting. Doi et al. (1979) examined the susceptibility to chymosin action of κ-casein preparations with different degrees of glycosylation. They found that more highly glycosylated forms were less susceptible to hydrolysis not only by chymosin but also by other proteases. Others have also found an inverse relationship between glycosylation and chymosin susceptibility for purified κ-casein fractions (Addeo et al., 1984; Vreeman et al., 1986) and in model systems (Addeo et al., 1984; Leaver and Horne, 1996). Compared with these model systems, in milk, the relationship is not always consistent. Using bulk milk, Chaplin and Green (1980) claimed that all κ-casein molecules were hydrolyzed with equal efficiency, whereas van Hooydonk et al. (1984) found that the rate of chymosin-catalyzed hydrolysis decreased with increasing glycosylation. The latter was also found in a study of individual milk samples ( Jensen et al., 2015): the highest rate of hydrolysis occurred in nonglycosylated κ-casein, and all glycosylated κ-casein isoforms had lower reaction rates. Within the nonglycosylated κ-casein group, variant A with one or two phosphate groups had the highest reaction rate, followed by variant B 1-2P and variant E 1-2P. Considering the consistent results between the studies of van Hooydonk et al. (1984) and Jensen et al. (2015), it seems convincing that glycan modifications negatively influence the reaction rate of chymosin hydrolysis. The rennet coagulation time (RCT), the rate of curd firming, and the curd firmness have been measured to assess the effect of κ-casein glycosylation on the coagulation properties

Sources and significance of casein heterogeneity

191

of milk (Robitaille et al., 1993). Whereas no effect on the RCT was observed, the rate of curd firming decreased, and the curd firmness increased at higher glycosylation levels. A more recent study on more than 2000 samples of milk from Simmental cows (Bonfatti et al., 2014) showed that the RCT decreased when the total κ-casein (as a proportion of total casein) and the glycosylated κ-casein increased, whereas nonglycosylated κ-casein exhibited a slightly unfavorable effect on the onset of the coagulation process. A decrease of 2 minutes in the RCT was also observed for milks with a high percentage of glycosylation of κ-casein on the total protein compared with milk with the lowest percentage of glycosylation. A favorable effect of κ-casein, glycosylated κ-casein, and degree of glycosylation on curd firmness was also detected. In another recent study, a detailed characterization of κ-casein isoforms was conducted by 2-D gel electrophoresis coupled with MS on milks with extremes of good or poor coagulation properties, selected from 892 samples of milk from Holstein Friesian and Jersey cows ( Jensen et al., 2012a). Six κ-casein isoforms, varying in phosphorylation and glycosylation levels, from each of the genetic variants of κ-casein were separated and identified, along with an unmodified κ-casein form at low abundance. Relative quantification showed that around 95% of the total κ-casein was phosphorylated with one or two phosphates attached, whereas approximately 35% of the identified κ-casein was glycosylated with from one to three tetrasaccharides. When isoforms from individual samples were compared, a very consistent ĸ-casein isoform pattern was found, with only minor differences in relation to breed, ĸ-casein genetic variant, and milk coagulation ability. The effect of PTMs on the coagulation properties of milk was studied by liquid chromatography/electrospray ionizationmass spectrometry using the same sample set of 892 cows (Poulsen et al., 2016). In Holstein milk, a higher relative content of κ-casein to total protein and a higher content of glycosylated κ-casein were associated with improved milk coagulation. In milk from Danish Jersey cows, the same tendency was observed, although it was not significant. Similar to glycosylated κ-casein as discussed by Poulsen et al. (2016), smaller casein micelles were associated with improved milk coagulation (Glantz et al., 2010). Therefore, the negative correlation between casein micelle size and glycosylated κ-casein (Bijl et al., 2014a) might explain the positive association between the degree of glycosylation and the coagulation properties. However, to date, there have been no studies that have determined the casein micelle size of fresh milk in relation to the coagulation properties to confirm this hypothesis. Differences in rennet coagulation properties have also been observed for genetic variants of κ-casein. Shorter RCTs, higher rates of curd firming, and higher curd firmness have been reported for milk from cows with the BB variant than for milk from cows with the AA variant (Walsh et al., 1998; and references therein). These differences between A and B variant milks were maintained after heat treatments of up to 80°C for 2 min, despite an overall deterioration in the coagulation properties at elevated temperatures (Choi and Ng-Kwai-Hang, 2003). Milks containing the rarer κ-casein C variant form rennet gels even more slowly than the A or B variant milks, possibly because of the substitution of histidine for arginine at residue 118, which may affect chymosin binding (Smith et al., 1997). Similar results have been observed for the κ-casein G variant, which has cysteine at residue 118 (Erhardt et al., 1997), and a similar explanation has been proposed (Smith et al., 1997). The influence of genetic variation in the major milk proteins on individual coagulation properties in 1299 Danish Holstein, Danish Jersey, and Swedish Red cows was determined by Poulsen et al. (2013). The authors showed that κ-casein B, αs1-casein C, and β-casein B were associated with good milk coagulation properties, whereas β-casein A2 had a negative effect.

192

5. Posttranslational modifications of caseins

Also in this case, it would be interesting to follow up whether a link between genetic variants and casein micelle size can explain the shorter RCTs of milk with different genetic variants. Acid coagulation

The coagulation of milk can also be induced by acid, as is the case in yogurt manufacture. Because this step does not involve the cleavage of κ-casein, a lesser or different effect might be expected. There are fewer studies on the effect of the glycosylation of κ-casein on acid coagulation. Cases et al. (2003) found that partial deglycosylation with neuraminidase had little effect on micellar surface charge and solvation but caused a decrease in the acid gelation time, a higher rate of gel firming, and a higher final firmness. Heat treatment

Heat treatment of milk can also destabilize the casein micelle structure. The heat-induced coagulation of milk is a very complex process that is affected by many parameters (O’Connell and Fox, 2003). A number of studies have examined the influence of the genetic variants of κ-casein on heat stability parameters, and it is generally accepted that B variant milks are more stable than A variant milks (FitzGerald and Hill, 1997). The reason may be more related to the effects on κ-casein concentration and micelle size, as already mentioned, than to the structural differences between the variants (Smith et al., 2002). Again, there are fewer studies related to the influence of the glycosylation of κ-casein on heat stability. Using a model system composed of casein micelles in simulated milk ultrafiltrate, Minkiewicz et al. (1993) showed that enzymatic removal of neuraminic acid using neuraminidase caused a decrease in heat stability. However, Robitaille and Ayers (1995), using whole milk, could not find a significant effect of neuraminidase treatment on heat stability. When milk is heated above 65°C, β-lactoglobulin denatures, exposing a previously buried sulfhydryl group that can participate in disulfide exchange reactions with other cysteine-containing proteins including κ-casein. This interaction has been recognized for many decades (Sawyer, 1969) and has been the subject of numerous investigations and reviews over the years; a detailed analysis is beyond the scope of this chapter (for an extensive review, see Chapter 9 of this volume). Recent studies have addressed both the mechanism of formation (Guyomarc’h et al., 2003) and the impact on product quality (Vasbinder et al., 2003) of disulfide-linked complexes. Despite the vast amount of literature on this topic, there do not appear to be any studies that have addressed the impact of the variable glycosylation of κ-casein on its ability to form disulfide-linked complexes either with itself or with β-lactoglobulin; however, this is, perhaps, not surprising because the sulfhydryl amino acids are in the para-κ-casein domain of the molecule, and the glycosylation sites are in the caseinomacropeptide. UHT milk

Heat-induced changes in micelle structure are particularly relevant for the production and storage of UHT milk. The extremes of heat treatment (of the order of 140–145°C for 4–10 s) produce a number of changes in the milk, not least of which is the formation of κ-casein-βlactoglobulin complexes. On storage, UHT-treated milks show a variable tendency to form gels, and this phenomenon, known as age gelation, affects product shelf life (for a review, see Datta and Deeth, 2001). From a theoretical perspective, higher initial levels of

Sources and significance of casein heterogeneity

193

glycosylation may act to temper the deleterious effects of heat treatment through effects on micelle size, micellar stability, and the formation of disulfide-linked complexes. The heat treatment itself may affect the glycosylation level at the surface of the micelle either indirectly, through loss of κ-casein in complex formation with β-lactoglobulin, or directly, through degradation of glycosidic residues (van Hooydonk et al., 1987; as quoted in Dziuba and Minikiewicz, 1996). Subsequent changes in the glycosylation level during storage could be mediated by the action of heat-stable glycosidases originating from psychrotrophic bacteria present in the raw milk (Marin et al., 1984). Release of monosaccharides during the storage of UHT milk has been observed (Recio et al., 1998; Belloque et al., 2001). Thus, both the initial glycosylation level of the κ-casein and the residual amount after UHT treatment may affect the storage properties of UHT-treated milk. As the actions of heat-resistant proteases can contribute to the age gelation of UHT milk, the inhibitory effects of glycosylation on the activity of proteases such as plasmin (Doi et al., 1979) may be important for prolonging shelf life. Unraveling specific effects will require the application of modern proteomic technologies for κ-casein analysis (Claverol et al., 2003; Holland et al., 2004, 2005, 2006; O’Donnell et al., 2004). Using these technologies, it will be possible to elaborate the heterogeneous glycoforms of κ-casein in raw milk, after pretreatment(s), after UHT processing, and during storage leading up to gelation. This will allow a definitive assessment of the functional significance of κ-casein glycosylation. Emulsification

The functional performance of the caseinomacropeptide is of importance both for use of the caseinomacropeptide as a food ingredient in its own right and for the properties it can impart to cheese whey products. Emulsifying performance has been found to vary with glycosylation (Kreuss et al., 2009a). Nonglycosylated caseinomacropeptide showed significantly better emulsifying properties than glycosylated caseinomacropeptide. Whereas nonglycosylated caseinomacropeptide showed an emulsifying activity index of 150.7 g/m2, glycosylated caseinomacropeptide achieved a value of only 98.5 g/m2. The stability of the emulsions was 1.4 times higher for nonglycosylated caseinomacropeptide than for glycosylated caseinomacropeptide. Droplet size measurements and creaming studies showed a marked influence of pH on both fractions, with minimal emulsion stabilities at pH 4.1 (glycosylated caseinomacropeptide) and pH 4.9 (nonglycosylated caseinomacropeptide). Investigation of the flocculation behavior and variations in the ionic strength indicated that the glycan side chains induced a combination of electrostatic, steric, and hydrophilic effects, preventing an ordered adsorption of glycosylated caseinomacropeptide molecules at the oil/water interface. In contrast, nonglycosylated caseinomacropeptide built a stable network at the oil/water interface. Foaming

Caseinomacropeptide fractions, both glycosylated and nonglycosylated, were also studied in detail for their foaming properties (Kreuss et al., 2009b). The nonglycosylated caseinomacropeptide-stabilized foams showed significantly higher foam rigidity and stability than the glycosylated caseinomacropeptide-stabilized foams, whereas both fractions yielded a high foaming ability with overruns of around 600%. The glycosylated caseinomacropeptide-stabilized foams, in particular, were considerably influenced by pH

194

5. Posttranslational modifications of caseins

and showed reduced foaming properties above the pI but superior properties at strong acidic pH, below the pI. This influence was less significant for nonglycosylated caseinomacropeptide. An increase in ionic strength did not appear to influence either fraction. Similar to the effects observed in oil-water emulsions, the combination of electrical, steric, and hydrophilic barriers caused by the glycosylation of glycosylated caseinomacropeptide appeared not to allow an ordered adsorption at the air/water interface. Also in this foam system, the nonglycosylated caseinomacropeptide could build a stable network at the air/water surface. Biological significance One aspect of κ-casein heterogeneity that has not been considered earlier is its influence on the biological properties of milk. This area has been reviewed extensively for cows’ milk (Dziuba and Minikiewicz, 1996). Some other reviews provide a broader perspective on the biological properties of all milk glycoproteins and glycolipids (Peterson et al., 2013; O’Riordan et al., 2014b). Several areas, all of which directly or indirectly link to health-related aspects, need to be considered: digestibility and bioavailability, the nutritional value of glycans, and the structure-function relationship of glycosylated groups and the bioactivity of peptides. Firstly, there is the effect of PTM on digestibility and bioavailability. This area has received attention only relatively recently. The effect of glycosylation on the hydrolysis by chymosin has been discussed earlier and has been studied intensively because of its importance for cheesemaking. The digestion of the resultant glycosylated or nonglycosylated caseinomacropeptide by brush border membrane peptidases has been described by Boutrou et al. (2008). Their key finding was that the digestions of nonglycosylated and glycosylated caseinomacropeptide through the action of exopeptidases were similar but that the activity of endopeptidases on glycosylated caseinomacropeptide was limited, certainly because of the attached O-glycosylations. Consequently, many more peptides were identified from the nonglycosylated caseinomacropeptide than from the glycosylated caseinomacropeptide. In addition, the glycosylation core, as well as the number of the attached glycosylated chains, modified the kinetics of digestion, the most heavily glycosylated forms being digested most slowly. Adding to the complexity, it was found by Petrat-Melin et al. (2016) that different genetic variants of κ-casein A, B, and E that varied in degrees of glycosylation were differentially affected by gastrointestinal digestion. Secondly, there is the nutritional contribution of the carbohydrate residues in κ-casein, particularly neuraminic acid (NeuAc). The importance of NeuAc and its roles in numerous biological functions have been reviewed (Schauer, 2000). NeuAc is commonly found as the terminal sugar residue on mammalian glycoproteins. Although mammals can synthesize NeuAc, the high levels in milk and especially colostrum may be related to a high demand for neonatal growth and development. The normal glycans on κ-casein are part of a class known as the Thomsen-/Friedenreich-related antigens (Dall’Olio and Chiricolo, 2001). The terminal NeuAc residues may play a key role in preventing colonization of the gut by pathogenic organisms by providing alternative binding sites that minimize binding to the normal gut epithelium. In this case, not only is the type of glycan important, but also the structure of the glycan will determine its functionality. The final aspect relates to the enormous interest in bioactive peptides derived from milk proteins (Clare and Swaisgood, 2000; Kilara and Panyam, 2003). Numerous in vitro activities

Caseins from other species

195

have been ascribed to κ-casein, its caseinomacropeptide, or peptides derived from them (Dziuba and Minikiewicz, 1996; Brody, 2000). These activities include prebiotic, antimicrobiological, and immunomodulatory effects (O’Riordan et al., 2014b). Some of these activities appear to be associated with particular forms of κ-casein (Malkoski et al., 2001) and can be glycosylation dependent (Li and Mine, 2004; O’Riordan et al., 2018). Whether or not the same activities occur in vivo is not always clear because it requires both generation and absorption of the active component during digestion, and this is not easy to detect. In vivo production of caseinomacropeptide is known to occur after the ingestion of milk (Ledoux et al., 1999; and references therein) and has been detected in the plasma of infants after milk ingestion (Chabance et al., 1995). Any naturally occurring bioactivity of caseinomacropeptide-derived peptides could be strongly influenced by the glycosylation status of κ-casein either directly, by modifying the activity of the peptide, or indirectly, by affecting proteolysis of κ-casein and hence release of the peptide. For future research, it will be interesting to see what we can learn from the biological significance of other mucin-type O-glycans.

Caseins from other species The biodiversity in the composition of milk between species is remarkable. Large differences can be identified by considering only the casein fraction: Fast growing animals such as rabbits can have milk with 90 g/L of caseins, whereas cows’ milk contains 27 g/L, and human milk contains only 5.8 g/L (Holt et al., 2013). Also, the composition of the casein fraction varies. Whereas the bovine milk protein fraction consists of the four caseins discussed in this chapter, the milk of humans and mares contains no or little αs2-casein, and elephant milk contains only β-casein and κ-casein. Differences between species are highly relevant. Firstly, the milks of some of these species provide an alternative source of nutrition. For the milks of all species, an understanding of between-species differences contributes to our knowledge of casein micelle stability and the significance of the heterogeneity of caseins. The differences between species on a genetic level have been reviewed previously (Ginger and Grigor, 1999), but the information on variation in the PTMs is more scattered. We therefore attempt to give an overview of the between-species variation with emphasis on β-casein phosphorylation and κ-casein glycosylation.

Phosphorylation It is apparent that considerable variations in phosphorylation of the caseins occur between species. Whereas the β-casein in bovine milk is mainly present with five phosphate groups, β-casein phosphorylation in human milk and horse milk is more complex. The biological significance of these variations is currently not known but is probably related to the fact that β-casein is the major casein fraction in the milks of these species. Also, the other species, caprine, ovine, and donkey, show variation in the phosphorylation of β-casein and α-casein compared with that in bovine milk. The only exception here is the water buffalo, which shows high consistency with its bovine counterpart.

196

5. Posttranslational modifications of caseins

Human The β-casein of human milk exists as six different forms with from zero to five phosphates (Greenberg et al., 1984) and is therefore more complex than that of bovine milk. In a recent study, the phosphorylation status of β-casein was investigated in human milk samples over time (Molinari et al., 2013). It varied significantly between term and preterm milk. A longitudinal trend among the term population was observed, with the phosphorylation state of β-casein decreasing during lactation in seven of the eight mothers analyzed. In the preterm population, no change in the distribution of phosphorylated isoforms was observed over time in 10 of the 16 mothers, whereas in the other six mothers, the level of phosphorylation increased during lactation. Horse Equine β-casein also shows variation in phosphorylation, with typically from three to seven phosphates on full length β-casein (Girardet et al., 2006) and from one to seven phosphates on a low-molecular-weight form that arises from an internal deletion (Miclo et al., 2007). The phosphorylation sites for equine milk have recently been reported (Mateos et al., 2010). In the mature protein, the isoform 4P was found to be phosphorylated on residues Ser9, Ser23, Ser24, and Ser25. Addition of phosphate groups on Ser18, Thr12, and Ser10 led to the formation of the isoforms 5P–7P, respectively. The results indicate that the in vivo phosphorylation of equine β-casein follows a sequential path and is not random. Ovine and caprine Ovine β-casein has also been reported to be variably phosphorylated, with from zero to seven phosphates (Ferranti et al., 2001), although the position of the seventh phosphorylation site is not clear. Caprine β-casein appears to be more like bovine β-casein, with the same five phosphorylation sites and an additional site on Thr27 (Neveu et al., 2002) in the mature protein. A more recent study compared caprine milks from the indigenous Greek breed and the international breeds Saanen and Alpine (Moatsou et al., 2008). This study identified a wide range of protein polymorphisms and phosphorylations. Phosphorylation of αs1-casein was 7P, 8P, and 9P, with lesser amounts of 6P and, rarely, 10P; phosphorylation of αs2-casein ranged from 6P to 11P; phosphorylation of β-casein was 5P, 6P, and 7P, with small amounts of lesser phosphorylation; κ-casein was phosphorylated with mostly 2P but some lesser amounts. A recent study using nanoscale liquid chromatography coupled with tandem electrospray MS has identified the majority of the phosphorylation sites of the goats’ milk caseins (Olumee-Shabon and Boehmer, 2013). These are summarized in Table 5.2 (Mercier et al., 1977). TABLE 5.2 Phosphorylation positions for the caprine caseins; data from (Olumee-Shabon and Boehmer, 2013) Casein

Number of phosphorylations

Phosphorylated residues

αs1

9

S61, S63, S130 (six others not detected)

αs2

10

S23, S24, S25, S72, S73, S74, S77, S145, S147, S159

β

5

S32, S33, S34, S37, S50

Caseins from other species

197

Other species A detailed proteomic analysis of water buffalo milk (D’Ambrosio et al., 2008) showed high consistency with the bovine counterpart proteins (residue numbers here based on the mature protein sequence). The study identified phosphopeptides from the following: αSl-casein, where phosphorylation occurred at Ser 41, 46, 48, 64, 66, 67, 68, and 75 in the mature protein, and αS2-casein, where phosphorylation occurred at the same sites as those of the bovine counterpart (Ser 129, 131, and 143); β-casein phosphorylation sites were consistent with the bovine counterpart sites (Ser 15, 17, 18, 19, and 35); the main and secondary sites of phosphorylation in buffalo κ-casein were Ser149 and Ser127, as observed for the bovine protein. In donkey milk “caseome,” Chianese et al. (2010) identified 11 phosphorylated components for κ-casein, six phosphorylated components for β- and αs1-casein, and three main phosphorylated components for αs2-casein. Up to five isoforms of β-casein are present in elephant milk (Madende et al., 2018). Of these isoforms, only a nonphosphorylated isoform and a singly phosphorylated (at Ser9) isoform in the mature protein were confirmed using high-resolution MS analysis.

Glycosylation The full amino acid sequences of κ-casein from more than 60 species are currently in the UniProt database (Table 5.3), with another 114 entries covering subspecies and incomplete sequences. Although the amino acid sequences are known for these species, the variation in the glycosylation of κ-casein is known for only a handful of species, and there is conclusive evidence for the positions of the modification sites for even fewer species. Least is known about the variation in carbohydrate moieties and the structures of κ-casein between species, which has been examined in detail only for bovine, human, and caprine milks. An overview of the variation in glycosylation of different species is provided in Table 5.4. Human In human milk, indirect evidence (the absence of Thr residues during Edman sequencing) showed that up to 10 of the residues were glycosylated (Fiat et al., 1980). Additional experiments confirmed that human milk seemed to carry a greater amount of glycosylated groups than bovine milk. N-acetylgalactosamine, galactose, N-acetylneuraminic acid, fucose, and N-acetylglucosamine were detected in human milk (Azuma et al., 1984). The first three are also common in cows’ milk, whereas N-acetylgalactosamine and fucose have been found only in bovine colostrum. As the content of total glycosylated groups was approximately 40% of the total κ-casein weight (Yamauchi et al., 1981), the amount of glycosylated groups in human milk is much higher than in bovine milk, which is estimated at around 4% (Table 5.4). With TABLE 5.3

Species with complete κ-casein entries in the UniProt database (December 2018)

Ailuropoda melanoleuca (giant panda) Aotus nancymaae (Ma’s night monkey) Balaenoptera acutorostrata scammoni (North Pacific minke whale) (Balaenoptera davidsoni) Continued

198

5. Posttranslational modifications of caseins

TABLE 5.3 Species with complete κ-casein entries in the UniProt database (December 2018)—cont’d Bos mutus grunniens (wild yak) (Bos grunniens) Bos taurus (Bovine) Bubalus bubalis (domestic water buffalo) Callithrix jacchus (white-tufted-ear marmoset) Camelus bactrianus (Bactrian camel) Camelus dromedarius (dromedary) (Arabian camel) Camelus ferus (wild Bactrian camel) (Camelus bactrianus ferus) Canis lupus familiaris (dog) (Canis familiaris) Capra hircus (goat) Capricornis crispus (Japanese serow) (Naemorhedus crispus) Capricornis sumatraensis (Sumatran serow) Capricornis swinhoei (Taiwan serow) (Naemorhedus swinhoei) Cavia porcellus (guinea pig) Cercocebus atys (sooty mangabey) (Cercocebus torquatus atys) Cervus nippon (sika deer) Chlorocebus sabaeus (green monkey) (Cercopithecus sabaeus) Colobus angolensis palliatus (Peters’ Angolan colobus) Delphinapterus leucas (beluga whale) Dipodomys ordii (Ord’s kangaroo rat) Enhydra lutris kenyoni Equus asinus africanus Equus caballus (horse) Erinaceus europaeus (Western European hedgehog) Felis catus (cat) (Felis silvestris catus) Fukomys damarensis (Damaraland mole rat) (Cryptomys damarensis) Gorilla gorilla gorilla (Western lowland gorilla) Heterocephalus glaber (naked mole rat) Homo sapiens (human) Lama glama (llama) Leptonychotes weddellii (Weddell seal) Lipotes vexillifer (Yangtze river dolphin) Macaca fascicularis (crab-eating macaque) (Cynomolgus monkey)

Caseins from other species

TABLE 5.3

Species with complete κ-casein entries in the UniProt database (December 2018)—cont’d

Macaca mulatta (Rhesus macaque) Macaca nemestrina (pig-tailed macaque) Mandrillus leucophaeus (drill) (Papio leucophaeus) Mus musculus (Mouse) Myotis brandtii (Brandt’s bat) Naemorhedus goral (Himalayan goral) Neomonachus schauinslandi (Hawaiian monk seal) (Monachus schauinslandi) Neophocaena asiaeorientalis asiaeorientalis (Yangtze finless porpoise) Nomascus leucogenys (Northern white-cheeked gibbon) (Hylobates leucogenys) Odobenus rosmarus divergens (Pacific walrus) Oreamnos americanus (mountain goat) Ornithorhynchus anatinus (duck-billed platypus) Oryctolagus cuniculus (rabbit) Ovis aries (sheep) Pan paniscus (pygmy chimpanzee) (bonobo) Pan troglodytes (chimpanzee) Papio anubis (olive baboon) Physeter catodon (sperm whale) (Physeter macrocephalus) Pongo abelii (Sumatran orangutan) (Pongo pygmaeus abelii) Pteropus alecto (black flying fox) Rattus norvegicus (rat) Rhinopithecus roxellana (golden snub-nosed monkey) (Pygathrix roxellana) Rupicapra rupicapra (chamois) Saiga tatarica (saiga antelope) Sus scrofa (pig) Tachyglossus aculeatus (Australian echidna) Tarsius syrichta (Philippine tarsier) Trichechus manatus latirostris (Florida manatee) Trichosurus vulpecula (brush-tailed possum) Tursiops truncatus (Atlantic bottle-nosed dolphin) (Delphinus truncatus) Ursus maritimus (polar bear) (Thalarctos maritimus)

199

Scientific name of organism

Length (number of amino acids)

Bovine

Bos taurus

190

142, 152, 153, 154, 157, 163, 170, 186

Dromedary (Arabian camel)

Camelus dromedarius

182

125, 129, 169, 172, 173, 154, 161, 178

Caprine (goat)

Capra hircus

192

152, 155, 156, 159, 165, 172, 188

Horse

Equus caballus

185

143, 161, 178

Human

Homo sapiens

182

133, 143, 148, 151, 157, 167, 169, 178

Fucose, N-acetylglucosamine, galactose, N-acetylgalactosamine, N-acetylneuraminic acid

Ovine (sheep)

Ovis aries

192

152, 155, 156, 159, 172, 188

N-acetylgalactosamine, galactose, N-acetylneuraminic acid, N-glycolylneuraminic acid

Amino acid glycosylation site

Carbohydrate moieties

Structure

Amount

N-acetylgalactosamine, galactose, N-acetylneuraminic acid, N-acetylglucosamine (colostrum only), Fucose (colostrum only)

Mono-, di-, tri(two variants), tetrasaccharides

Estimated to be 4% on total κ-casein weight, assuming 50% of the κ-casein is glycosylated with two oligosaccharide side chains each with a molecular weight of 798 Da on average

N-acetylgalactosamine, galactose, N-acetylneuraminic acid, N-glycolylneuraminic acid

Mono-, di-, tri(four variants), tetrasaccharides (four variants)

The contents of N-acetylneuraminic acid and N-glycolylneuraminic acid were, respectively, 1.07 and 1.17 mg/ 100 mg in caprine caseinomacropeptide Higher amount of carbohydrates and sialic acid residues compared with bovine Estimated to be 40% on total κ-casein weight

Mono-, di-, tri(four variants), tetrasaccharides (four variants)

The contents of N-acetylneuraminic acid and N-glycolylneuraminic acid were, respectively, 0.13 and 0.93 mg/ 100 mg in ovine caseinomacropeptide

5. Posttranslational modifications of caseins

Common name of organism

200

TABLE 5.4 Variation in glycosylation of κ-casein between species in length, glycosylation site, carbohydrate moieties, structure, and amount; organisms and glycosylation sites as stated in the UniProt database; amino acid modifications in italics; modification expected based on similarity to bovine casein sequence or sequence analysis

Caseins from other species

201

respect to changes during lactation, a study on glycosylation in human milk revealed changing patterns of glycosylation for many of the whey proteins, but not for κ-casein (Froehlich et al., 2010). Horse The glycosylation of equine κ-casein has been reviewed by Uniacke-Lowe et al. (2010). Glycosylation of horse milk has been indicated by lectin-binding studies (Iametti et al., 2001). However, the exact positions of the glycosylation sites or the carbohydrate moieties are not conclusive. It has been suggested that the κ-casein of equine milk contains more carbohydrates and more sialic acid residues than that of bovine milk (Egito et al., 2001, 2002). Ovine and caprine The positions of the glycans in ovine milk and caprine milk have not been confirmed. Interestingly, the carbohydrate moieties have been identified and quantified. In the milk of both species, N-acetylgalactosamine, galactose, N-acetylneuraminic acid, and N-glycolylneuraminic acid (NeuGc) were detected (Moreno et al., 2001). The contents of N-acetylneuraminic acid and N-glycolylneuraminic acid were, respectively, 0.13 and 0.93 mg/100 mg in ovine caseinomacropeptide and 1.07 and 1.17 mg/100 mg in caprine caseinomacropeptide. Mono- di-, tri-, and tetrasaccharides were detected in ovine milk (Moreno et al., 2000; Mamone et al., 2003; Casal et al., 2013) and in caprine milk ( Javier Moreno et al., 2001; Casal et al., 2013). The structures of the tri- and tetrasaccharides were elucidated by graphitized carbon liquid chromatographyelectrospray ionization ion trap tandem MS (Casal et al., 2013). Compared with bovine milk and human milk, the presence of N-glycolylneuraminic acid on κ-casein is unique for ovine and caprine species. The presence of this additional carbohydrate moiety results in eight different structures of tri- and tetrasaccharides (Fig. 5.9), whereas only three are present in bovine milk. These eight structures are the trisaccharides (NeuAcα(2–3)Galβ(1-3)GalNAc or Galβ(1–3) [NeuAcα(2–6)]GalNAc), (NeuGcα(2–3)Galβ(1–3)GalNAc or Galβ(1–3)[NeuGcα(2–6)]GalNAc) and the tetrasaccharides NeuAcα(2–3)Galβ(1–3)[NeuAcα(2–6)]GalNAc, NeuGcα(2–3)Galβ (1–3)[NeuAcα(2–6)]GalNAc, NeuAcα(2–3)Galβ(1–3)[NeuGcα(2–6)]GalNAc, NeuGcα(2–3)Galβ (1–3)[NeuGcα(2–6)]GalNAc. FIG. 5.9 Tri- and tetrasaccharides present in ovine and

▪

caprine milk. ( ) GalNAc, (●) Gal, (♦) NeuAc, and ( ) NeuGc.

202

5. Posttranslational modifications of caseins

Other species In water buffalo, diglycosylated forms of κ-casein were identified, but the specific residues modified were not reported; however, there is no reason to believe that they would be different from the bovine sites. In dromedary, also known as Arabian camel, five threonine residues have been proposed to be glycosylated (125, 129, 169, 172, 173) (Kappeler et al., 1998); however, to date, no direct evidence confirms these positions. Using sodium dodecyl sulfate polyacrylamide gel electrophoresis, most of the κ-casein in dromedary milk was found to be of low molecular mass and therefore was a low glycosylated form (Kappeler et al., 1998). This was in disagreement with earlier findings of Mehaia (1987), who found a high content of released sialic acid by neuraminidase (7.35 mg/g camel casein) compared with that for bovine κ-casein (3.02 mg/g bovine casein). The glycan structures of dromedary κ-casein have not been determined as yet (Mati et al., 2017).

Conclusions PTMs such as phosphorylation, glycosylation, and perhaps disulfide bond formation play a critical role in casein micelle formation and stability. It seems somewhat surprising that so much variability occurs in these PTMs on the caseins. Whereas significant functional differences in milk properties have consistently been reported for milks with different genetic variants of the caseins and are well established, the effects reported for variable PTMs have been investigated only more recently, enabled by proteomic methods, and supported by a range of highly sensitive MS techniques. A full understanding of these effects and their implications for dairy processing and products is still at an early stage.

Acknowledgments The authors gratefully acknowledge Hein van Valenberg and Lotte Bach Larsen for helpful discussions and critical reading of this manuscript.

References Addeo, F., Martin, P., Ribadeau-Dumas, B., 1984. Susceptibility of buffalo and cow kappa-caseins to chymosin action. Milchwissenschaft 39, 202–205. Alais, C., Jolles, P., 1961. Comparative study of the caseinoglycopeptides formed by the action of rennin on caseins from the cow, sheep and goat. II. Study of the non-peptide portion. Biochim. Biophys. Acta 51, 315–322. Alexander, L.J., Stewart, A.F., Mackinlay, A.G., Kapelinskaya, T.V., Tkach, T.M., Gorodetsky, S.I., 1988. Isolation and characterization of the bovine kappa-casein gene. Eur. J. Biochem. 178, 395–401. Azuma, N., Kaminogawa, S., Yamauchi, K., 1984. Properties of glycomacropeptide and para-κ-casein derived from human κ-casein and comparison of human and bovine κ-caseins as to susceptibility to chymosin and pepsin. Agric. Biol. Chem. 48, 2025–2031. https://doi.org/10.1080/00021369.1984.10866443. Belloque, J., Villamiel, M., Lopez, F.R., Olano, A., 2001. Release of galactose and N-acetylglucosamine during the storage of UHT milk. Food Chem. 72, 407–412. Bijl, E., de Vries, R., van Valenberg, H., Huppertz, T., van Hooijdonk, T., 2014a. Factors influencing casein micelle size in milk of individual cows: genetic variants and glycosylation of κ-casein. Int. Dairy J. 34, 135–141. https://doi. org/10.1016/j.idairyj.2013.08.001.

References

203

Bijl, E., van Valenberg, H.J.F., Huppertz, T., van Hooijdonk, A.C.M., Bovenhuis, H., 2014b. Phosphorylation of alpha (s1)-casein is regulated by different genes. J. Dairy Sci. 97, 7240–7246. https://doi.org/10.3168/jds.2014-8061. Bijl, E., van Valenberg, H., Sikkes, S., Jumelet, S., Sala, G., Olieman, K., van Hooijdonk, T., Huppertz, T., 2014c. Chymosin-induced hydrolysis of caseins: influence of degree of phosphorylation of alpha-s1-casein and genetic variants of beta-casein. Int. Dairy J. 39, 215–221. https://doi.org/10.1016/j.idairyj.2014.07.005. Bijl, E., Huppertz, T., van Valenberg, H., Holt, C., 2018. A quantitative model of the bovine casein micelle: ion equilibria and calcium phosphate sequestration by individual caseins in bovine milk. Eur. Biophys. J. https://doi.org/ 10.1007/s00249-018-1330-2. Bobe, G., Beitz, D.C., Freeman, A.E., Lindberg, G.L., 1999. Effect of milk protein genotypes on milk protein composition and its genetic parameter estimates. J. Dairy Sci. 82, 2797–2804. Bodenmiller, B., Mueller, L.N., Mueller, M., Domon, B., Aebersold, R., 2007. Reproducible isolation of distinct, overlapping segments of the phosphoproteome. Nat. Methods 4, 231–237. Bonfatti, V., Grigoletto, L., Cecchinato, A., Gallo, L., Carnier, P., 2009. Erratum to “Validation of a new reversed-phase high-performance liquid chromatography method for separation and quantification of bovine milk protein genetic variants” [J. Chromatogr. A, 1195 (2008) 101–106]. J. Chromatogr. A 1216, 1528. https://doi.org/10.1016/j. chroma.2009.01.008. Bonfatti, V., Chiarot, G., Carnier, R., 2014. Glycosylation of kappa-casein: genetic and nongenetic variation and effects on rennet coagulation properties of milk. J. Dairy Sci. 97, 1961–1969. https://doi.org/10.3168/jds.2013-7418. Bouguyon, E., Beauvallet, C., Huet, J.-C., Chanat, E., 2006. Disulphide bonds in casein micelle from milk. Biochem. Biophys. Res. Commun. 343, 450–458. Bouniol, C., Printz, C., Mercier, J.-C., 1993. Bovine αs2-casein D is generated by exon VIII skipping. Gene 128, 289–293. Boutrou, R., Jardin, J., Blais, A., Tome, D., Leonil, J., 2008. Glycosylations of kappa-casein-derived caseinomacropeptide reduce its accessibility to endo- but not exointestinal brush border membrane peptidases. J. Agric. Food Chem. 56, 8166–8173. Bovenhuis, H., Van Arendonk, J.A., Korver, S., 1992. Associations between milk protein polymorphisms and milk production traits. J. Dairy Sci. 75, 2549–2559. Brody, E.P., 2000. Biological activities of bovine glycomacropeptide. Br. J. Nutr. 84 (Suppl. 1), S39–S46. Buitenhuis, B., Poulsen, N.A., Gebreyesus, G., Larsen, L.B., 2016. Estimation of genetic parameters and detection of chromosomal regions affecting the major milk proteins and their post translational modifications in Danish Holstein and Danish Jersey cattle. BMC Genet. 17, 114. https://doi.org/10.1186/s12863-016-0421-2. Casal, E., Lebro´n-Aguilar, R., Chuan-Lee, Y., Noboru, T., Quintanilla-Lo´pez, J.E., 2013. Identification of sialylated oligosaccharides derived from ovine and caprine caseinomacropeptide by graphitized carbon liquid chromatography-electrospray ionization ion trap tandem mass spectrometry. Food Anal. Methods 6, 814–825. https://doi.org/10.1007/s12161-012-9489-0. Cases, E., Vidal, V., Cuq, J.L., 2003. Effect of kappa-casein deglycosylation on the acid coagulability of milk. J. Food Sci. 68, 2406–2410. Chabance, B., Jolles, P., Izquierdo, C., Mazoyer, E., Francoual, C., Drouet, L., Fiat, A.M., 1995. Characterization of an antithrombotic peptide from kappa-casein in newborn plasma after milk ingestion. Br. J. Nutr. 73, 583–590. Chaplin, B., Green, M.L., 1980. Determination of the proportion of kappa-casein hydrolyzed by rennet on coagulation of skim-milk. J. Dairy Res. 47, 351–358. Chianese, L., Calabrese, M.G., Ferranti, P., Mauriello, R., Garro, G., De Simone, C., Quarto, M., Addeo, F., Cosenza, G., Ramunno, L., 2010. Proteomic characterization of donkey milk “caseome”. J. Chromatogr. A 1217, 4834–4840. Choi, J.W., Ng-Kwai-Hang, K.F., 2003. Effects of genetic variants of kappa-casein and beta-lactoglobulin and heat treatment on coagulating properties of milk. Asian Australas. J. Anim. Sci. 16, 1212–1217. Chokhawala, H.A., Chen, X., 2007. Enzymatic approaches to O-glycoside introduction: glycosyltransferases. In: Kamerling, J.P. (Ed.), Comprehensive Glycoscience: From Chemistry to Systems Biology. In: 1. Elsevier, Amsterdam, pp. 415–451. Clare, D.A., Swaisgood, H.E., 2000. Bioactive milk peptides: a prospectus. J. Dairy Sci. 83, 1187–1195. Claverol, S., Burlet-Schiltz, O., Gairin, J.E., Monsarrat, B., 2003. Characterization of protein variants and posttranslational modifications: ESI-MSn analyses of intact proteins eluted from polyacrylamide gels. Mol. Cell. Proteomics 2, 483–493. Collins, M.O., Yu, L., Choudhary, J.S., 2007. Analysis of protein phosphorylation on a proteome-scale. Proteomics 7, 2751–2768.

204

5. Posttranslational modifications of caseins

Corfield, A.P., Berry, M., 2015. Glycan variation and evolution in the eukaryotes. Trends Biochem. Sci. 40, 351–359. https://doi.org/10.1016/j.tibs.2015.04.004. Coolbear, K.P., Elgar, D.F., Ayers, J.S., 1996. Profiling of genetic variants of bovine κ-casein macropeptide by electrophoretic and chromatographic techniques. Int. Dairy J. 6, 1055–1068. Cox, D.M., Zhong, F., Du, M., Duchoslav, E., Sakuma, T., McDermott, J.C., 2005. Multiple reaction monitoring as a method for identifying protein posttranslational modifications. J. Biomol. Tech. 16, 83–90. Creamer, L.K., Richardson, T., Parry, D.A.D., 1981. Secondary structure of bovine αS1- and β-casein in solution. Arch. Biochem. Biophys. 211, 689–696. https://doi.org/10.1016/0003-9861(81)90505-1. Dalgleish, D.G., 1985. Glycosylated kappa-caseins and the sizes of bovine casein micelles. Analysis of the different forms of kappa-casein. Biochim. Biophys. Acta 830, 213–215. Dalgleish, D.G., 1986. Analysis by fast protein liquid chromatography of variants of kappa-casein and their relevance to micellar structure and renneting. J. Dairy Res. 53, 43–51. Dalgleish, D.G., Corredig, M., 2012. The structure of the casein micelle of milk and its changes during processing. Annu. Rev. Food Sci. Technol. 3, 449–467. https://doi.org/10.1146/annurev-food-022811-101214. Dalgleish, D.G., Horne, D.S., Law, A.J.R., 1989. Size-related differences in bovine casein micelles. Biochim. Biophys. Acta 991, 383–387. Dall’Olio, F., Chiricolo, M., 2001. Sialyltransferases in cancer. Glycoconj. J. 18, 841–850. D’Ambrosio, C., Arena, S., Salzano, A.M., Renzone, G., Ledda, L., Scaloni, A., 2008. A proteomic characterization of water buffalo milk fractions describing PTM of major species and the identification of minor components involved in nutrient delivery and defense against pathogens. Proteomics 8, 3657–3666. Darula, Z., Medzihradszky, K.F., 2018. Analysis of mammalian O-glycopeptides—we have made a good start, but there is a long way to go. Mol. Cell. Proteomics 17, 2–17. https://doi.org/10.1074/mcp.MR117.000126. Datta, N., Deeth, H.C., 2001. Age gelation of UHT milk—a review. Food Bioprod. Process. 79, 197–210. Davies, D.T., Law, A.J.R., 1983. Variation in protein composition of bovine casein micelles and serum casein in relation to casein micelle size and milk temperature. J. Dairy Res. 50, 67–75. de Vries, R., van Knegsel, A., Johansson, M., Lindmark-Ma˚nsson, H., van Hooijdonk, T., Holtenius, K., Hettinga, K., 2015. Effect of shortening or omitting the dry period of Holstein-Friesian cows on casein composition of milk. J. Dairy Sci. 98, 8678–8687. https://doi.org/10.3168/jds.2015-9544. Doi, H., Kawaguchi, N., Ibuki, F., Kanamori, M., 1979. Susceptibility of kappa-casein components to various proteases. J. Nutr. Sci. Vitaminol. (Tokyo) 25, 33–41. Doi, H., Kobatake, H., Ibuki, F., Kanamori, M., 1980. Attachment sites of carbohydrate portions to peptide chain of kappa-casein from bovine colostrum. Agric. Biol. Chem. 44, 2605–2611. Donnelly, W.J., McNeill, G.P., Buchheim, W., McGann, T.C., 1984. A comprehensive study of the relationship between size and protein composition in natural bovine casein micelles. Biochim. Biophys. Acta 789, 136–143. Dziuba, J., Minikiewicz, P., 1996. Influence of glycosylation on micelle-stabilising ability and biological properties of C-terminal fragments of cow’s κ-casein. Int. Dairy J. 6, 1017–1044. Egito, A.S., Girardet, J.-M., Miclo, L., Moll
e, D., Humbert, G., Gaillard, J.-L., 2001. Susceptibility of equine κ- and β-caseins to hydrolysis by chymosin. Int. Dairy J. 11, 885–893. https://doi.org/10.1016/S0958-6946(01)00123-6. Egito, A.S., Miclo, L., Lo´pez, C., Adam, A., Girardet, J.-M., Gaillard, J.-L., 2002. Separation and characterization of mares’ milk αs1-, β-, κ-caseins, γ-casein-like, and proteose peptone component 5-like peptides. J. Dairy Sci. 85, 697–706. https://doi.org/10.3168/jds.S0022-0302(02)74126-X. Erhardt, G., Prinzenberg, E.-M., Buchberger, J., Krick-Salek, H., Krause, I., Miller, M., 1997. Bovine kappa-casein G: detection, occurrence, molecular genetic characterisation, genotyping and coagulation properties. In: Milk Protein Polymorphism (IDF Special Issue 9702). International Dairy Federation, Brussels, pp. 328–329. Fang, Z.H., Visker, M.H.P.W., Miranda, G., Delacroix-Buchet, A., Bovenhuis, H., Martin, P., 2016. The relationships among bovine alpha(S)-casein phosphorylation isoforms suggest different phosphorylation pathways. J. Dairy Sci. 99, 8168–8177. https://doi.org/10.3168/jds.2016-11250. Fang, Z.H., Bovenhuis, H., Delacroix-Buchet, A., Miranda, G., Boichard, D., Visker, M.H.P.W., Martin, P., 2017. Genetic and nongenetic factors contributing to differences in alpha(s)-casein phosphorylation isoforms and other major milk proteins. J. Dairy Sci. 100, 5564–5577. https://doi.org/10.3168/jds.2016-12338. Fang, Z.H., Bovenhuis, H., van Valenberg, H.J.F., Martin, P., Huppertz, T., Visker, M.H.P.W., 2018. Genetic parameters for alpha(S1)-casein and alpha(S2)-casein phosphorylation isoforms in Dutch Holstein Friesian. J. Dairy Sci. 101, 1281–1291. https://doi.org/10.3168/jds.2017-13623.

References

205

Farrell Jr., H.M., Cooke, P.H., Wickham, E.D., Piotrowski, E.G., Hoagland, P.D., 2003. Environmental influences on bovine kappa-casein: reduction and conversion to fibrillar (amyloid) structures. J. Protein Chem. 22, 259–273. Farrell Jr., H.M., Jimenez-Flores, R., Bleck, G.T., Brown, E.M., Butler, J.E., Creamer, L.K., Hicks, C.L., Hollar, C.M., NgKwai-Hang, K.F., Swaisgood, H.E., 2004. Nomenclature of the proteins of cows’ milk—sixth revision. J. Dairy Sci. 87, 1641–1674. Farrell, H.M., Malin, E.L., Brown, E.M., Qi, P.X., 2006. Casein micelle structure: what can be learned from milk synthesis and structural biology? Curr. Opin. Colloid Interface Sci. 11, 135–147. https://doi.org/10.1016/J. COCIS.2005.11.005. Ferranti, P., Pizzano, R., Garro, G., Caira, S., Chianese, L., Addeo, F., 2001. Mass spectrometry-based procedure for the identification of ovine casein heterogeneity. J. Dairy Res. 68, 35–51. Fiat, A.M., Jolles, J., Aubert, J.P., Loucheux-Lefebvre, M.H., Jolles, P., 1980. Localisation and importance of the sugar part of human casein. Eur. J. Biochem. 111, 333–339. https://doi.org/10.1111/j.1432-1033.1980.tb04946.x. Fiat, A.M., Jolles, J., Loucheux-Lefebvre, M.H., Alais, C., Jolles, P., 1981. Localisation of the prosthetic sugar groups of bovine colostrum kappa-casein. Hoppe Seylers Zeitsch. Physiol. Chem. 362, 1447–1454. Fiat, A.M., Chevan, J., Jolles, P., De Waard, P., Vliegenthart, J.F., Piller, F., Cartron, J.P., 1988. Structural variability of the neutral carbohydrate moiety of cow colostrum kappa-casein as a function of time after parturition. Identification of a tetrasaccharide with blood group I specificity. Eur. J. Biochem. 173, 253–259. FitzGerald, R.J., Hill, J.P., 1997. The relationship between milk protein polymorphism and the manufacture and functionality of dairy products. In: Milk Protein Polymorphism (IDF Special Issue 9702). International Dairy Federation, Brussels, pp. 355–371. Fournet, B., Fiat, A.M., Montreuil, J., Jolles, P., 1975. The sugar part of kappa-caseins from cow milk and colostrum and its microheterogeneity. Biochimie 57, 161–165. Fournet, B., Fiat, A.M., Alais, C., Jolles, P., 1979. Cow kappa-casein: structure of the carbohydrate portion. Biochim. Biophys. Acta 576, 339–346. Frederiksen, P.D., Andersen, K.K., Hammershoj, M., Poulsen, H.D., Sorensen, J., Bakman, M., Qvist, K.B., Larsen, L.B., 2011. Composition and effect of blending of noncoagulating, poorly coagulating, and well-coagulating bovine milk from individual Danish Holstein cows. J. Dairy Sci. 94, 4787–4799. https://doi.org/10.3168/jds.2011-4343. Froehlich, J.W., Dodds, E.D., Barboza, M., McJimpsey, E.L., Seipert, R.R., Francis, J., An, H.J., Freeman, S., German, J.B., Lebrilla, C.B., 2010. Glycoprotein expression in human milk during lactation. J. Agric. Food Chem. 58, 6440–6448. Ginger, M.R., Grigor, M.R., 1999. Comparative aspects of milk caseins. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 124, 133–145. https://doi.org/10.1016/S0305-0491(99)00110-8. Girardet, J.-M., Miclo, L., Florent, S., Moll
e, D., Gaillard, J.-L., 2006. Determination of the phosphorylation level and deamidation susceptibility of equine β-casein. Proteomics 6, 3707–3717. Glantz, M., Devold, T.G., Vegarud, G.E., Mansson, H.L., Stalhammar, H., Paulsson, M., 2010. Importance of casein micelle size and milk composition for milk gelation. J. Dairy Sci. 93, 1444–1451. https://doi.org/10.3168/ jds.2009-2856. Gorodetskii, S.I., Kershulite, D.R., Korobko, V.G., 1983. Primary structure of cDNA of Bos taurus kappa-casein macropeptide. Bioorg. Khim. 9, 1693–1695. Greenberg, R., Groves, M.L., Dower, H.J., 1984. Human beta-casein. Amino acid sequence and identification of phosphorylation sites. J. Biol. Chem. 259, 5132–5138. Grosclaude, F., Mahe, M.F., Mercier, J.C., Ribadeau-Dumas, B., 1972a. Characterization of genetic variants of alpha-S1 and beta bovine caseins. Eur. J. Biochem. 26, 328–337. Grosclaude, F., Mahe, M.-F., Mercier, J.-C., Ribadeau-Dumas, B., 1972b. Localization of amino-acid substitutions differentiating the A and B variants of kappa-casein in cattle. Ann. Genet. Select. Anim. 4, 515–521. Groves, M.L., Dower, H.J., Farrell Jr., H.M., 1992. Reexamination of the polymeric distributions of kappa-casein isolated from bovine milk. J. Protein Chem. 11, 21–28. Groves, M.L., Wickham, E.D., Farrell Jr., H.M., 1998. Environmental effects on disulfide bonding patterns of bovine kappa-casein. J. Protein Chem. 17, 73–84. Guerin, J., Alais, C., Jolles, J., Jolles, P., 1974. Kappa-casein from bovine colostrum. Biochim. Biophys. Acta 351, 325–332. Guyomarc’h, F., Law, A.J., Dalgleish, D.G., 2003. Formation of soluble and micelle-bound protein aggregates in heated milk. J. Agric. Food Chem. 51, 4652–4660.

206

5. Posttranslational modifications of caseins

Heck, J.M.L., Olieman, C., Schennink, A., van Valenberg, H.J.F., Visker, M.H.P.W., Meuldijk, R.C.R., van Hooijdonk, A.C.M., 2008. Estimation of variation in concentration, phosphorylation and genetic polymorphism of milk proteins using capillary zone electrophoresis. Int. Dairy J. 18, 548–555. Heck, J.M.L., Schennink, A., van Valenberg, H.J.F., Bovenhuis, H., Visker, M.H.P.W., van Arendonk, J.A.M., van Hooijdonk, A.C.M., 2009. Effects of milk protein variants on the protein composition of bovine milk. J. Dairy Sci. 92, 1192–1202. Hernandez-Hernandez, O., Lebron-Aguilar, R., Quintanilla-Lopez, J.E., Luz Sanz, M., Javier Moreno, F., 2011. Detection of two minor phosphorylation sites for bovine ĸ-casein macropeptide by reversed-phase liquid chromatography-tandem mass spectrometry. J. Agric. Food Chem. 59, 10848–10853. Holland, J.W., Deeth, H.C., Alewood, P.F., 2004. Proteomic analysis of kappa-casein micro-heterogeneity. Proteomics 4, 743–752. Holland, J.W., Deeth, H.C., Alewood, P.F., 2005. Analysis of O-glycosylation site occupancy in bovine kappa-casein glycoforms separated by two-dimensional gel electrophoresis. Proteomics 5, 990–1002. Holland, J.W., Deeth, H.C., Alewood, P.F., 2006. Resolution and characterisation of multiple isoforms of bovine kappa-casein by 2-DE following a reversible cysteine-tagging enrichment strategy. Proteomics 6, 3087–3095. Holland, J.W., Deeth, H.C., Alewood, P.F., 2008. Analysis of disulphide linkages in bovine kappa-casein oligomers using two-dimensional electrophoresis. Electrophoresis 29, 2402–2410. Holt, C., 2004. An equilibrium thermodynamic model of the sequestration of calcium phosphate by casein micelles and its application to the calculation of the partition of salts in milk. Eur. Biophys. J. Biophys. Lett. 33, 421–434. https://doi.org/10.1007/s00249-003-0377-9. Holt, C., Horne, D.S., 1996. The hairy casein micelle: evolution of the concept and its implications for dairy technology. Neth. Milk Dairy J. 50, 85–111. Holt, C., Carver, J.A., Ecroyd, H., Thorn, D.C., 2013. Invited review: Caseins and the casein micelle: their biological functions, structures, and behavior in foods. J. Dairy Sci. 96, 6127–6146. https://doi.org/10.3168/jds.2013-6831. Horne, D.S., 1998. Casein interactions: casting light on the black boxes, the structure in dairy products. Int. Dairy J. 8, 171–177. Horne, D.S., 2002. Casein structure, self-assembly and gelation. Curr. Opin. Colloid Interface Sci. 7, 456–461. Iametti, S., Tedeschi, G., Oungre, E., Bonomi, F., 2001. Primary structure of κ-casein isolated from mares’ milk. J. Dairy Res. 68, 53–61. https://doi.org/10.1017/S0022029900004544. Ibeagha-Awemu, E.M., Prinzenberg, E.M., Jann, O.C., Luhken, G., Ibeagha, A.E., Zhao, X., Erhardt, G., 2007. Molecular characterization of bovine CSN1S2*B and extensive distribution of zebu-specific milk protein alleles in European cattle. J. Dairy Sci. 90, 3522–3529. Javier Moreno, F., Recio, I., Olano, A., L€ opez-Fandin˜o, R., 2001. Heterogeneity of caprine κ-casein macropeptide. J. Dairy Res. 68, 197–208. https://doi.org/10.1017/S002202990100471X. Jensen, H.B., Holland, J.W., Poulsen, N.A., Larsen, L.B., 2012a. Milk protein genetic variants and isoforms identified in bovine milk representing extremes in coagulation properties. J. Dairy Sci. 95, 2891–2903. Jensen, H.B., Poulsen, N.A., Andersen, K.K., Hammershoj, M., Poulsen, H.D., Larsen, L.B., 2012b. Distinct composition of bovine milk from Jersey and Holstein-Friesian cows with good, poor, or noncoagulation properties as reflected in protein genetic variants and isoforms. J. Dairy Sci. 95, 6905–6917. https://doi.org/10.3168/ jds.2012-5675. Jensen, H.B., Pedersen, K.S., Johansen, L.B., Poulsen, N.A., Bakman, M., Chatterton, D.E.W., Larsen, L.B., 2015. Genetic variation and posttranslational modification of bovine kappa-casein: effects on caseinomacropeptide release during renneting. J. Dairy Sci. 98, 747–758. https://doi.org/10.3168/jds.2014-8678. Jolles, J., Fiat, A.M., Alais, C., Jolles, P., 1973. Comparative study of cow and sheep kappa-caseinoglycopeptides: determination of the N-terminal sequences with a sequencer and location of the sugars. FEBS Lett. 30, 173–176. Kanamori, M., Kawaguchi, N., Ibuki, F., Doi, H., 1980. Attachment sites of carbohydrate moieties to peptide chain of bovine kappa-casein from normal milk. Agric. Biol. Chem. 44, 1855–1861. Kanamori, M., Doi, H., Ideno, S., Ibuki, F., 1981. Presence of O-glycosidic linkage through serine residue in kappacasein component from bovine mature milk. J. Nutr. Sci. Vitaminol. (Tokyo) 27, 231–241. Kapkova, P., Lattova, E., Perreault, H., 2006. Nonretentive solid-phase extraction of phosphorylated peptides from complex peptide mixtures for detection by matrix-assisted laser desorption/ionization mass spectrometry. Anal. Chem. 78, 7027–7033.

References

207

Kappeler, S., Farah, Z., Puhan, Z., 1998. Sequence analysis of Camelus dromedarius milk caseins. J. Dairy Res. 65, 209–222. https://doi.org/10.1017/S0022029997002847. Keating, A.F., Davoren, P., Smith, T.J., Ross, R.P., Cairns, M.T., 2007. Bovine κ-casein gene promoter haplotypes with potential implications for milk protein expression. J. Dairy Sci. 90, 4092–4099. Kilara, A., Panyam, D., 2003. Peptides from milk proteins and their properties. Crit. Rev. Food Sci. Nutr. 43, 607–633. Kirmiz, N., Robinson, R.C., Shah, I.M., Barile, D., Mills, D.A., 2018. Milk glycans and their interaction with the infant-gut microbiota. Annu. Rev. Food Sci. Technol. 9, 429–450. https://doi.org/10.1146/annurev-food030216-030207. Kreuss, M., Strixner, T., Kulozik, U., 2009a. The effect of glycosylation on the interfacial properties of bovine caseinomacropeptide. Food Hydrocoll. 23, 1818–1826. Kreuss, M., Krause, I., Kulozik, U., 2009b. Influence of glycosylation on foaming properties of bovine caseinomacropeptide. Int. Dairy J. 19, 715–720. Kumosinski, T.F., Brown, E.M., Farrell, H.M., 1991. 3-dimensional molecular modeling of bovine caseinsalpha-s1-casein. J. Dairy Sci. 74, 2889–2895. Lasa-Benito, M., Marin, O., Meggio, F., Pinna, L.A., 1996. Golgi apparatus mammary gland casein kinase: monitoring by a specific peptide substrate and definition of specificity determinants. FEBS Lett. 382, 149–152. https://doi.org/ 10.1016/0014-5793(96)00136-6. Le, T.T., Deeth, H.C., Larsen, L.B., 2017. Proteomics of major bovine milk proteins: novel insights. Int. Dairy J. 67, 2–15. https://doi.org/10.1016/j.idairyj.2016.11.016. Leaver, J., Horne, D.S., 1996. Chymosin-catalysed hydrolysis of glycosylated and nonglycosylated bovine kappacasein adsorbed on latex particles. J. Colloid Interface Sci. 181, 220–224. Ledoux, N., Mah
e, S., Dubarry, M., Bourras, M., Benamouzig, R., Tom
e, D., 1999. Intraluminal immunoreactive caseinomacropeptide after milk protein ingestion in humans. Nahrung 43, 196–200. Li, E.W.Y., Mine, Y., 2004. Immunoenhancing effects of bovine glycomacropeptide and its derivatives on the proliferative response and phagocytic activities of human macrophage-like cells, U937. J. Agric. Food Chem. 52, 2704–2708. Lodes, A., Krause, I., Buchberger, J., Aumann, J., Klostermeyer, H., 1996. The influence of genetic variants of milk proteins on the composition and technological properties of milk. 1. Casein micelle size and the content of non-glycosylated κ-casein. Milchwissenschaft 51, 368–373. Maciel, G.D.M., Mogensen, L., Lehmann, J.O., Kidmose, U., Kristensen, T., Larsen, L.B., Poulsen, N.A., 2017. Impaired milk quality and cheese making properties is not a concern for managing cows for 15 or 18 months calving intervals. Int. Dairy J. 70, 2–11. https://doi.org/10.1016/j.idairyj.2016.12.014. Madende, M., Kemp, G., Stoychev, S., Osthoff, G., 2018. Characterisation of African elephant beta casein and its relevance to the chemistry of caseins and casein micelles. Int. Dairy J. 85, 112–120. https://doi.org/10.1016/j. idairyj.2018.05.013. Mahe, M.F., Miranda, G., Queval, R., Bado, A., Zafindrajaona, P.S., Grosclaude, F., 1999. Genetic polymorphism of milk proteins in African Bos taurus and Bos indicus populations. Characterisation of variants alphaS1-casein H and kappa-casein. Genet. Sel. Evol. 31, 239–253. Malkoski, M., Dashper, S.G., O’Brien-Simpson, N.M., Talbo, G.H., Macris, M., Cross, K.J., Reynolds, E.C., 2001. Kappacin, a novel antibacterial peptide from bovine milk. Antimicrob. Agents Chemother. 45, 2309–2315. Mamone, G., Caira, S., Garro, G., Nicolai, A., Ferranti, P., Picariello, G., Malorni, A., Chianese, L., Addeo, F., 2003. Casein phosphoproteome: identification of phosphoproteins by combined mass spectrometry and twodimensional gel electrophoresis. Electrophoresis 24, 2824–2837. https://doi.org/10.1002/elps.200305545. Manson, W., Carolan, T., Annan, W.D., 1977. Bovine αS0 casein; a phosphorylated homologue of αS1 casein. Eur. J. Biochem. 78, 411–417. Marin, A., Mawhinney, T.P., Marshall, R.T., 1984. Glycosidic activities of Pseudomonas fluorescens on fat-extracted skim milk, buttermilk, and milk fat globule membranes. J. Dairy Sci. 67, 52–59. Martin, P., Szymanowska, M., Zwierzchowski, L., Leroux, C., 2002. The impact of genetic polymorphisms on the protein composition of ruminant milks. Reprod. Nutr. Dev. 42, 433–459. Mateos, A., Girardet, J.-M., Molle, D., Corbier, C., Gaillard, J.-L., Miclo, L., 2010. Identification of phosphorylation sites of equine beta-casein isoforms. Rapid Commun. Mass Spectrom. 24, 1533–1542.

208

5. Posttranslational modifications of caseins

Mati, A., Senoussi-Ghezali, C., Si Ahmed Zennia, S., Almi-Sebbane, D., El-Hatmi, H., Girardet, J.-M., 2017. Dromedary camel milk proteins, a source of peptides having biological activities—a review. Int. Dairy J. 73, 25–37. https://doi. org/10.1016/j.idairyj.2016.12.001. McGann, T.C., Donnelly, W.J., Kearney, R.D., Buchheim, W., 1980. Composition and size distribution of bovine casein micelles. Biochim. Biophys. Acta 630, 261–270. Mehaia, M.A., 1987. Studies of camel casein micelles: treatment with soluble and immobilized neuraminidase. Carbohydr. Polym. 7, 361–369. https://doi.org/10.1016/0144-8617(87)90003-8. Mercier, J.C., 1981. Phosphorylation of caseins, present evidence for an amino acid triplet code post translationally recognized by specific kinases. Biochimie 63, 1–17. Mercier, J.C., Brignon, G., Ribadeau-Dumas, B., 1973. Primary structure of bovine kappa B casein. Complete sequence. Eur. J. Biochem. 35, 222–235. Mercier, J.-C., Addeo, F., P
elissier, J.-P., 1977. Structure primaire du cas
einomacropeptide de la caseine K caprine. Biochimie 58, 1303–1310. https://doi.org/10.1016/S0300-9084(77)80013-8. Miclo, L., Girardet, J.M., Egito, A.S., Molle, D., Martin, P., Gaillard, J.L., 2007. The primary structure of a low-M(r) multiphosphorylated variant of beta-casein in equine milk. Proteomics 7, 1327–1335. Minkiewicz, P., Dziuba, J., Muzinska, B., 1993. The contribution of N-acetylneuraminic acid in the stabilization of micellar casein. Polish J. Food Nutr. Sci. 2, 39–48. Miranda, G., Anglade, P., Mahe, M.F., Erhardt, G., 1993. Biochemical characterization of the bovine genetic kappacasein C and E variants. Anim. Genet. 24, 27–31. Moatsou, G., Moschopoulou, E., Molle, D., Gagnaire, V., Kandarakis, I., Leonil, J., 2008. Comparative study of the protein fraction of goat milk from the Indigenous Greek breed and from international breeds. Food Chem. 106, 509–520. Molinari, C.E., Casadio, Y.S., Hartmann, B.T., Arthur, P.G., Hartmann, P.E., 2013. Longitudinal analysis of protein glycosylation and beta-casein phosphorylation in term and preterm human milk during the first 2 months of lactation. Br. J. Nutr. 110, 105–115. Molle, D., Leonil, J., 1995. Heterogeneity of the bovine κ-casein caseinomacropeptide, resolved by liquid chromatography on-line with electrospray ionization mass spectrometry. J. Chromatogr. A 708, 223–230. Moore, A., Boulton, A.P., Heid, H.W., Jarasch, E.-D., Craig, R.K., 1985. Purification and tissue-specific expression of casein kinase from the lactating guinea-pig mammary gland. Eur. J. Biochem. 152, 729–737. https://doi.org/ 10.1111/j.1432-1033.1985.tb09254.x. Moremen, K.W., Tiemeyer, M., Nairn, A.V., 2012. Vertebrate protein glycosylation: diversity, synthesis and function. Nat. Rev. Mol. Cell Biol. 13, 448–462. https://doi.org/10.1038/nrm3383. Moreno, F.J., Recio, I., Olano, A., Lopez-Fandino, R., 2000. Chromatographic characterization of ovine κ-casein macropeptide. J. Dairy Res. 67, 349–359. https://doi.org/10.1017/S0022029900004222. Moreno, F.J., Olano, A., Villamiel, M., 2001. N-acetyl- and N-glycolyl-neuraminic acid content in kappa-casein macropeptides. Milchwissenschaft 56, 669–671. Mulagapati, S., Koppolu, V., Raju, T.S., 2017. Decoding of O-linked glycosylation by mass spectrometry. Biochemistry 56, 1218–1226. https://doi.org/10.1021/acs.biochem.6b01244. Neveu, C., Molle, D., Moreno, J., Martin, P., Leonil, J., 2002. Heterogeneity of caprine beta-casein elucidated by RP-HPLC/MS: genetic variants and phosphorylations. J. Protein Chem. 21, 557–567. Ng-Kwai-Hang, K.F., 1997. A review of the relationship between milk protein polymorphism and milk composition/ milk production. In: Milk Protein Polymorphism. (IDF Special Issue 9702). International Dairy Federation, Brussels, pp. 22–37. Nwosu, C.C., Strum, J.S., An, H.J., Lebrilla, C.B., 2010. Enhanced detection and identification of glycopeptides in negative ion mode mass spectrometry. Anal. Chem. 82, 9654–9662. https://doi.org/10.1021/ac101856r. O’Connell, J.E., Fox, P.F., 2000. The two-stage coagulation of milk proteins in the minimum of the heat coagulation time-pH profile of milk: effect of casein micelle size. J. Dairy Sci. 83, 378–386. O’Connell, J.E., Fox, P.F., 2003. Heat-induced coagulation of milk. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1, Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, pp. 879–945. O’Donnell, R., Holland, J.W., Deeth, H.C., Alewood, P., 2004. Milk proteomics. Int. Dairy J. 14, 1013–1023. O’Riordan, N., Kane, M., Joshi, L., Hickey, R.M., 2014a. Glycosidase activities in bovine milk over lactation. Int. Dairy J. 35, 116–121.

References

209

O’Riordan, N., Kane, M., Joshi, L., Hickey, R.M., 2014b. Structural and functional characteristics of bovine milk protein glycosylation. Glycobiology 24, 220–236. https://doi.org/10.1093/glycob/cwt162. O’Riordan, N., O’Callaghan, J., Butto`, L.F., Kilcoyne, M., Joshi, L., Hickey, R.M., 2018. Bovine glycomacropeptide promotes the growth of Bifidobacterium longum ssp. infantis and modulates its gene expression. J. Dairy Sci. 101, 6730–6741. https://doi.org/10.3168/jds.2018-14499. Olumee-Shabon, Z., Boehmer, J.L., 2013. Detection of casein phosphopeptides in goat milk. J. Proteome Res. 12, 3034–3041. Peterson, R., Cheah, W.Y., Grinyer, J., Packer, N., 2013. Glycoconjugates in human milk: protecting infants from disease. Glycobiology 23, 1425–1438. https://doi.org/10.1093/glycob/cwt072. Petrat-Melin, B., Kristiansen, G.H., Le, T.T., Poulsen, N.A., Larsen, L.B., Young, J.F., 2016. In vitro gastrointestinal digestion of purified bovine kappa-casein variants A, B, and E: effects on antioxidant and angiotensin 1-converting enzyme inhibitory capacity. Int. Dairy J. 57, 44–51. https://doi.org/10.1016/j.idairyj.2016.02.036. Pisano, A., Packer, N.H., Redmond, J.W., Williams, K.L., Gooley, A.A., 1994. Characterization of O-linked glycosylation motifs in the glycopeptide domain of bovine kappa-casein. Glycobiology 4, 837–844. Poulsen, N.A., Bertelsen, H.P., Jensen, H.B., Gustavsson, F., Glantz, M., Mansson, H.L., Andren, A., Paulsson, M., Bendixen, C., Buitenhuis, A.J., Larsen, L.B., 2013. The occurrence of noncoagulating milk and the association of bovine milk coagulation properties with genetic variants of the caseins in 3 Scandinavian dairy breeds. J. Dairy Sci. 96, 4830–4842. https://doi.org/10.3168/jds.2012-6422. Poulsen, N.A., Jensen, H.B., Larsen, L.B., 2016. Factors influencing degree of glycosylation and phosphorylation of caseins in individual cow milk samples. J. Dairy Sci. 99, 3325–3333. https://doi.org/10.3168/jds.2015-10226. Prinzenberg, E.M., Hiendleder, S., Ikonen, T., Erhardt, G., 1996. Molecular genetic characterization of new bovine kappa-casein alleles CSN3F and CSN3G and genotyping by PCR-RFLP. Anim. Genet. 27, 347–349. Prinzenberg, E.M., Krause, I., Erhardt, G., 1999. SSCP analysis at the bovine CSN3 locus discriminates six alleles corresponding to known protein variants (A, B, C, E, F, G) and three new DNA polymorphisms (H, I, A1). Anim. Biotechnol. 10, 49–62. Rasmussen, L.K., Hojrup, P., Petersen, T.E., 1992. The multimeric structure and disulfide-bonding pattern of bovine kappa-casein. Eur. J. Biochem. 207, 215–222. Rasmussen, L.K., Hojrup, P., Petersen, T.E., 1994. Disulphide arrangement in bovine caseins: localization of intrachain disulphide bridges in monomers of kappa- and alpha s2-casein from bovine milk. J. Dairy Res. 61, 485–493. Rasmussen, L.K., Sorensen, E.S., Petersen, T.E., Nielsen, N.C., Thomsen, J.K., 1997. Characterization of phosphate sites in native ovine, caprine, and bovine casein micelles and their caseinomacropeptides: a solid-state phosphorus-31 nuclear magnetic resonance and sequence and mass spectrometric study. J. Dairy Sci. 80, 607–614. Rasmussen, L.K., Johnsen, L.B., Tsiora, A., Sorensen, E.S., Thomsen, J.K., Nielsen, N.C., Jakobsen, H.J., Petersen, T.E., 1999. Disulphide-linked caseins and casein micelles. Int. Dairy J. 9, 215–218. Recio, M.I., Villamiel, M., Martı´nez-Castro, I., Olano, A., 1998. Changes in free monosaccharides during storage of some UHT milks: a preliminary study. Zeitsch. Lebensmitteluntersuch. Forsch. A 207, 180–181. Riggs, L., Sioma, C., Regnier, F.E., 2001. Automated signature peptide approach for proteomics. J. Chromatogr. A 924, 359–368. Robitaille, G., Ayers, C., 1995. Effects of κ-casein glycosylation on heat stability of milk. Food Res. Int. 28, 17–21. Robitaille, G., Ng-Kwai-Hang, K.F., Monardes, H.G., 1991a. Variation of the N-acetylneuraminic acid content of bovine kappa-casein. J. Dairy Res. 58, 107–114. Robitaille, G., Ng-Kwai-Hang, K.F., Monardes, H.G., 1991b. Association of kappa-casein glycosylation with milk production and composition in Holsteins. J. Dairy Sci. 74, 3314–3317. Robitaille, G., Ng-Kwai-Hang, K.F., Monardes, H.G., 1993. Effect of κ-casein glycosylation on cheese yielding capacity and coagulating properties of milk. Food Res. Int. 26, 365–369. Rollema, H.S., 1992. Casein association and micelle formation. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry. Elsevier Applied Science, London, pp. 111–140. Rowles, J., Slaughter, C., Moomaw, C., Hsu, J., Cobb, M.H., 1991. Purification of casein kinase I and isolation of cDNAs encoding multiple casein kinase I-like enzymes. Proc. Natl. Acad. Sci. U. S. A. 88, 9548–9552. https://doi.org/ 10.1073/pnas.88.21.9548. Saito, T., Itoh, T., 1992. Variations and distributions of O-glycosidically linked sugar chains in bovine kappa-casein. J. Dairy Sci. 75, 1768–1774.

210

5. Posttranslational modifications of caseins

Saito, T., Itoh, T., Adachi, S., 1981a. The chemical structure of a tetrasaccharide containing N-acetylglucosamine obtained from bovine colostrum kappa-casein. Biochim. Biophys. Acta 673, 487–494. Saito, T., Itoh, T., Adachi, S., Suzuki, T., Usui, T., 1981b. The chemical structure of neutral and acidic sugar chains obtained from bovine colostrum kappa-casein. Biochim. Biophys. Acta 678, 257–267. Saito, T., Itoh, T., Adachi, S., Suzuki, T., Usui, T., 1982. A new hexasaccharide chain isolated from bovine colostrum kappa-casein taken at the time of parturition. Biochim. Biophys. Acta 719, 309–317. Sawyer, W.H., 1969. Complex between β-lactoglobulin and κ-casein. A review. J. Dairy Sci. 52, 1347–1355. Schauer, R., 2000. Achievements and challenges of sialic acid research. Glycoconj. J. 17, 485–499. Schild, T.A., Wagner, V., Geldermann, H., 1994. Variants within the 50 -flanking regions of bovine milk-proteinencoding genes. I. Kappa-casein-encoding gene. Theor. Appl. Genet. 89, 116–120. Schlieben, S., Erhardt, G., Senft, B., 1991. Genotyping of bovine kappa-casein (kappa-CNA, kappa-CNB, kappa-CNC, kappa-CNE) following DNA sequence amplification and direct sequencing of kappa-CNE PCR product. Anim. Genet. 22, 333–342. Schopen, G.C.B., Visker, M., Koks, P.D., Mullaart, E., van Arendonk, J.A.M., Bovenhuis, H., 2011. Whole-genome association study for milk protein composition in dairy cattle. J. Dairy Sci. 94, 3148–3158. https://doi.org/10.3168/ jds.2010-4030. Shekar, P.C., Goel, S., Rani, S.D.S., Sarathi, D.P., Alex, J.L., Singh, S., Kumar, S., 2006. κ-Casein-deficient mice fail to lactate. Proc. Natl. Acad. Sci. U. S. A. 103, 8000–8005. Slattery, C.W., 1978. Variation in the glycosylation pattern of bovine kappa-casein with micelle size and its relationship to a micelle model. Biochemistry 17, 1100–1104. Smith, M.H., Hill, J.P., Creamer, L.K., Plowman, J.E., 1997. Towards understanding the variant effect on the rate of cleavage by chymosin on κ-caseins A and C. In: Milk Protein Polymorphism (IDF Special Issue 9702). International Dairy Federation, Brussels, pp. 185–188. Smith, M.H., Edwards, P.J., Palmano, K.P., Creamer, L.K., 2002. Structural features of bovine caseinomacropeptide A and B by 1H nuclear magnetic resonance spectroscopy. J. Dairy Res. 69, 85–94. Sulimova, G.E., Sokolova, S.S., Semikozova, O.P., Nguet, L.M., Berberov, E.M., 1992. Analysis of DNA polymorphism of cluster genes in cattle: casein genes and major histocompatibility complex (MHC) genes. Tsitolog. Genet. 26, 18–26. Sulimova, G.E., Badagueva, I.N., Udina, I.G., 1996. Polymorphism of the kappa-casein gene in populations of the subfamily Bovinae. Genetika 32, 1576–1582. Swaisgood, H.E., 2003. Chemistry of the caseins. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1, Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, pp. 139–202. Swaisgood, H.E., Brunner, J.R., Lillevik, H.A., 1964. Physical parameters of κ-casein from cow’s milk. Biochemistry 20, 1616–1623. Sweet, S.M.M., Creese, A.J., Cooper, H.J., 2006. Strategy for the identification of sites of phosphorylation in proteins: neutral loss triggered electron capture dissociation. Anal. Chem. 78, 7563–7569. Tagliabracci, V.S., Engel, J.L., Wen, J.Z., Wiley, S.E., Worby, C.A., Kinch, L.N., Xiao, J.Y., Grishin, N.V., Dixon, J.E., 2012. Secreted kinase phosphorylates extracellular proteins that regulate biomineralization. Science 336, 1150–1153. https://doi.org/10.1126/science.1217817. Takeuchi, M., Tsuda, E., Yoshikawa, M., Sasaki, R., Chiba, H., 1985. Fractionation and characterization of 9 subcomponents of bovine kappa-casein A. Agric. Biol. Chem. 49, 2269–2276. Talbot, B., Waugh, D.F., 1970. Micelle-forming characteristics of monomeric and covalent polymeric kappa caseins. Biochemistry 9, 2807–2813. Ten Hagen, K.G., Fritz, T.A., Tabak, L.A., 2003. All in the family: the UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferases. Glycobiology 13, 1R–16R. Ubersax, J.A., Ferrell Jr., J.E., 2007. Mechanisms of specificity in protein phosphorylation. Nat. Rev. Mol. Cell Biol. 8, 530–541. Uniacke-Lowe, T., Huppertz, T., Fox, P.F., 2010. Equine milk proteins: chemistry, structure and nutritional significance. Int. Dairy J. 20, 609–629. https://doi.org/10.1016/j.idairyj.2010.02.007. van Halbeek, H., Dorland, L., Vliegenthart, J.F., Fiat, A.M., Jolles, P., 1980. A 360-MHz 1H NMR study of three oligosaccharides isolated from cow kappa-casein. Biochim. Biophys. Acta 623, 295–300. van Halbeek, H., Dorland, L., Vliegenthart, J.F., Fiat, A.M., Jolles, P., 1981. Structural characterization of a novel acidic oligosaccharide unit derived from cow colostrum kappa-casein. A 500 MHz 1H NMR study. FEBS Lett. 133, 45–50.

References

211

van Hooydonk, A.C.M., Olieman, C., Hagedoorn, H.G., 1984. Kinetics of the chymosin-catalysed proteolysis of kappa-casein in milk. Neth. Milk Dairy J. 38, 207–222. van Hooydonk, A.C.M., De Koster, P.G., Boerrigter, I.J., 1987. The renneting properties of heated milk. Neth. Milk Dairy J. 41, 3–18. Vasbinder, A.J., Alting, A.C., Visschers, R.W., De Kruif, C.G., 2003. Texture of acid milk gels: formation of disulfide cross-links during acidification. Int. Dairy J. 13, 29–38. Vincent, D., Elkins, A., Condina, M.R., Ezernieks, V., Rochfort, S., 2016. Quantitation and identification of intact major milk proteins for high-throughput LC-ESI-Q-TOF MS analyses. PLoS One 1. https://doi.org/10.1371/journal. pone.0163471. Vreeman, H.J., Visser, S., Slangen, C.J., Van Riel, J.A., 1986. Characterization of bovine kappa-casein fractions and the kinetics of chymosin-induced macropeptide release from carbohydrate-free and carbohydrate-containing fractions determined by high-performance gel-permeation chromatography. Biochem. J. 240, 87–97. Walsh, C.D., Guinee, T.P., Reville, W.D., Harrington, D., Murphy, J.J., O’Kennedy, B.T., FitzGerald, R.J., 1998. Influence of κ-casein genetic variant on rennet gel microstructure, Cheddar cheesemaking properties and casein micelle size. Int. Dairy J. 8, 707–714. Walstra, P., 1999. Casein sub-micelles: do they exist? Int. Dairy J. 9, 189–192. Wu, H.Y., Tseng, V.S., Liao, P.C., 2007. Mining phosphopeptide signals in liquid chromatography-mass spectrometry data for protein phosphorylation analysis. J. Proteome Res. 6, 1812–1821. Wuhrer, M., Catalina, M.I., Deelder, A.M., Hokke, C.H., 2007. Glycoproteomics based on tandem mass spectrometry of glycopeptides. J. Chromatogr. B 849, 115–128. https://doi.org/10.1016/j.jchromb.2006.09.041. Yamauchi, K., Azuma, N., Kobayashi, H., Kaminogawa, S., 1981. Isolation and properties of human kappa-casein. J. Biochem. 90, 1005–1012. Yoshikawa, M., Takeuchi, M., Sasaki, R., Chiba, H., 1982. Chemical characteristics of bovine casein micelles fractionated by size on CPG-10/300 chromatography. Agric. Biol. Chem. 46, 1043–1048. Zevaco, C., Ribadeau-Dumas, B., 1984. A study on the carbohydrate binding sites of bovine kappa-casein using high performance liquid chromatography. Milchwissenschaft 39, 206–210. Zhang, C., Bijl, E., Hettinga, K., 2018. Destabilization of UHT milk by protease AprX from Pseudomonas fluorescens and plasmin. Food Chem. 263, 127–134. https://doi.org/10.1016/j.foodchem.2018.04.128. Zhou, H., Xu, S., Ye, M., Feng, S., Pan, C., Jiang, X., Li, X., Han, G., Fu, Y., Zou, H., 2006. Zirconium phosphonatemodified porous silicon for highly specific capture of phosphopeptides and MALDI-TOF MS analysis. J. Proteome Res. 5, 2431–2437.

This page intentionally left blank

C H A P T E R

6 Casein micelle structure and stability David S. Horne Wisconsin Center for Dairy Research, University of Wisconsin—Madison, Madison, WI, United States

Introduction What is published in a scientific journal does not necessarily represent the “final truth.” It simply represents a position for debate and discussion, a point of view occasionally based on the outcome of some experiment. Just as we expect in experimental measurement, the language of such expression should also be precise and unequivocal. Any debate arising from such publication should be conducted in a dispassionate manner, addressing directly any criticisms raised. In that spirit, we begin our review of caseins and their interactions. The caseins are a family of phosphoproteins that are found in the milks of all mammals. They exist in these milks generally as complex aggregates or micelles of the proteins and mineral calcium phosphate (Fox and Brodkorb, 2008). Because the caseins utilize the same calcium-sequestering mechanism to regulate the calcium phosphate concentration of their environment, they have recently been identified as members of a wider family of secretory calcium-binding phosphoproteins that are descended from a common ancestor gene (Kawasaki and Weiss, 2003, 2006). These secretory phosphoproteins include enamel matrix proteins, dentine, salivary proteins, bone extracellular matrix proteins, and the caseins. All are descended from early primordial genes by duplication and divergence to serve their specialized adaptive functions. Their genes retain common functional and sequence features, even after this extensive divergence. It is thought that primordial calcium-sensitive casein genes diverged from enamel matrix protein genes before the appearance of monotremes in the Jurassic era (Kawasaki and Weiss, 2003). More controversially, Kawaski et al. (2011) have recently argued that all caseins, both calcium-sensitive and calcium-insensitive, that is, κ-type caseins, evolved from the odontogenic ameloblast-associated (ODAM) gene via two different pathways: the calcium-sensitive genes are postulated to originate directly from secretory calcium-binding phosphoprotein (SCPP) genes, whereas the calcium-insensitive genes are directly differentiated from the follicular dendritic cell-secreted peptide (FDCSP) gene, both genes having a common ancestor in the ODAM gene. This ancestral line for κ-casein genes

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00006-2

213

# 2020 Elsevier Inc. All rights reserved.

214

6. Casein micelle structure and stability

is inconsistent with the hypothesis of Jolle`s et al. (1978), based on high levels of sequence identity, that κ-casein was derived from γ-fibrinogen, a blood coagulation factor. Casein allows milk to appear to be supersaturated with calcium phosphate. Essentially, it transports safely through the mammary gland the mineral calcium phosphate that is essential for the development of bones and teeth in the suckling infant, through the medium of the casein micelle. Locking up the calcium phosphate in this package is one aspect of the biological function of the micelle, and we consider this and the role of the phosphoserine residues in detail in later sections. Ensuring release of this same calcium phosphate in the gastric destination of the milk is a property of the micelle that is not generally given great consideration. More research effort has been put into trying to give mechanistic understanding to the technological behavior of the micelle, an understanding that is necessary to achieve efficient conversion of milk into products such as cheese and yogurt or to explain the behavior of milk components in emulsions or reconstituted dairy products. Many of the physical and technological properties of the casein micelle (diffusion, viscosity, and light scattering) can be described by treating the casein micelles as colloidal hard spheres (Alexander et al., 2002). The initial stages, up to the onset of instability, in processes such as renneting, acid-induced gelation, and flocculation in the presence of ethanol can apparently be well described by allowing these micellar hard spheres to become adhesive, essentially, as described later, treating their κ-casein outer layer as a salted polyelectrolyte brush (De Kruif and Zhulina, 1996; De Kruif and Holt, 2003). Beyond the critical point in all three processes, however, changes in micellar integrity and internal structure render the simple colloidal particle approach inadequate (Horne, 2003a, b; Choi et al., 2007; Oczan-Yilsay et al., 2007). Neither is the adhesive sphere approach at all helpful in furthering our understanding of the processes of micellar assembly, the pathways to dissociation, or the maintenance of micellar integrity. For that, we must turn to a structural model of the casein micelle, bearing in mind that, notwithstanding the inadequacy of the adhesive sphere approach, our micelle model also has to be adaptable enough to explain why that approach has been so successful within its limitations. Various models of casein micelle assembly and structure have appeared over the years and have been subjected to regular review and appraisal, most recently and comprehensively by Fox and Brodkorb (2008). It is not our intention to repeat any of that; rather, we would remind readers that, in considering all of the suggestions made over the years from Waugh (1971) onward, a model has to be more than a picture of how the proposer sees the casein micelle. An acceptable model has to provide a wholly coherent mechanism through which the various components assemble into the final structure. Simply placing them onto the board will not suffice; there has to be a reason for their positional presence. For example, because Waugh (1971) observed the role of κ-casein in rennet-induced coagulation, he reasoned that κ-casein resided on the casein micelle surface and elaborated his coat-core model (Waugh, 1971). However, he did not tell us how the κ-casein got there. He wanted it to be there but did not provide a cause for its presence. For this reason, his model fails the test. More recently, the submicelle model of Slattery and Evard (Slattery and Evard, 1973; Slattery, 1977), subsequently elaborated by Schmidt (1980), requires the formation two different submicelles, one κ-casein rich and the other κ-casein poor, but as the authors do not tell us how this separation is achieved, the model and the subsequent derivations also fail. Schmidt (1980) was perhaps the first to

Primary structure and interactions of caseins

215

recognize the importance of calcium phosphate in linking micellar components, unfortunately in his case, between Slattery-type submicelles. However, in 1992, Holt proposed his first nanocluster model, which used calcium phosphate nanoclusters to cross-link between phosphoserine clusters on the caseins, crucially recognizing the existence of several such clusters on some of those molecules and allowing a three-dimensional network to be built up but omitting to describe how κ-casein enters the picture (Holt, 1992). The dual-binding model of Horne (1998) overcomes this deficiency by invoking a second binding interaction between the hydrophobic regions of the caseins, building a second interlinked network in which κ-casein has a path-terminating role and an assumed position on the micelle surface. There has been considerable debate in more recent years over the suitability and success of these models (Farrell et al., 2006; Horne, 2006; Dalgleish, 2011). Dalgleish (2011) repeated the arguments of Horne (1998) in dismissing the submicelle model and proposed minor but dubiously founded amendments to the dual-binding model, as did McMahon and Oomen (2008). The Holt nanocluster model has continued to evolve (De Kruif et al., 2012; Thorn et al., 2015) and, in its latest form, is virtually mechanistically indistinguishable from the dualbinding model, an observation that has been groundlessly disputed by Carver et al. (2017). The Thorn et al. (2015) and Horne (1998, 2006) models continue to differ in the nature of the linking interactions, which have their origin in the chemistry of the caseins and of calcium phosphate, and these are highlighted and discussed in later sections. As Thorn et al. (2015) remind us, the caseins may act as chaperones, may be identified as intrinsically disordered proteins (IDPs), or, as individual proteins, may be capable of amyloid fibril formation. In whatever situation the caseins find themselves, however, their behavior is dictated by their primary structure and the interactions to which this can give rise. As we demonstrate in the following, the model is only an aid to visualizing how the casein micelle responds in the production of dairy products such as yogurts or rennet curds or in stability situations with ethanol or heat. We first summarize the physicochemical properties and interactions of the caseins and show how these lead naturally to the dual-binding picture.

Primary structure and interactions of caseins Just as casein micelles are aggregates of all of the casein proteins and micellar calcium phosphate, the dual-binding model involves the properties and interactions of all of the caseins. Central to this argument are those features of the proteins that are conserved across species and through millennia. The caseins were identified as members of the wider secretory calcium phosphate-binding family by their possession of functional and sequence features common to that family (Kawasaki and Weiss, 2003, 2006). Among the conserved motifs is the SXE peptide (Ser-Xaa-Glu) where Xaa may be any amino acid. In the caseins, this peptide provides a recognition template for posttranslational phosphorylation of the serine in the mammary gland by a casein kinase (Mercier, 1981). Moreover, in the caseins, the serine residues are often found clustered in groups of two, three, or four. Such clusters in the αS- and βcaseins are highly conserved (Martin et al., 2003), and their numbers attest to the significance of the calcium phosphate requirement for postnatal growth in mammals, even more so when it is noted, by reference to their sequences (Swaisgood, 2003), that the αS-caseins, for example,

216

6. Casein micelle structure and stability

the αS1- and αS2-caseins of bovine milk, themselves possess two or more such clusters. As we move to the discussion of the micelle model, the importance of these phosphoserine clusters in locking up micellar calcium phosphate will become apparent, but for the moment, we confine ourselves to casein sequence properties. Associated with these phosphoserine clusters are significantly high densities of negative charge at the normal milk pH. There is a charge density of
9e within the span of residues 65–72 of αS1-casein and a further
6e along the sequence 48
53 of the same protein. A similarly high charge density of
9e is found between residues 16 and 23 of β-casein, encompassing the phosphoserine cluster there. Similar high densities are found around the phosphoserine clusters of αS2-casein. All of these estimates of charge density assume a contribution of
1.5e from each phosphoserine residue at or near the natural pH of milk, 6.7. Away from the phosphoserine clusters, the casein molecules are distinctly hydrophobic. This segregation of hydrophilic and hydrophobic residues, first noted by Swaisgood (1973), confers on the caseins a definite amphipathic nature, which contributes to their ability to function successfully as stabilizers in oil-in-water emulsions. The topography of β-casein adsorbed at the oil-water interface was probed by testing the accessibility of the reactive sites to the proteolytic enzyme trypsin (Leaver and Dalgleish, 1990). In aqueous solution, the reactive sites of β-casein were hydrolyzed randomly at no preferential rate. With β-caseinstabilized emulsions, however, the peptides released showed the lysines at positions 25 and 28 of the sequence to be readily accessible to trypsin, whereas all other possible hydrolytic sites were less so (Leaver and Dalgleish, 1990). These residues lie in the center of the highly charged, hydrophilic N-terminal region containing the four-phosphoserine cluster in β-casein. Measured by dynamic light scattering, a decrease of approximately 13 nm in hydrodynamic radius of the emulsion droplets also accompanied the scission of these peptides, indicating the extent to which they stretch out into the aqueous phase from the emulsion droplet surface (Dalgleish and Leaver, 1991). The remaining hydrophobic portion of the molecule was speculated to lie along the droplet surface, shielded from trypsin attack. Similar changes in hydrodynamic radius were observed when β-casein was adsorbed from aqueous buffers onto the surface of polystyrene latex particles, indicating a similar adsorption pattern (Dalgleish, 1990; Brooksbank et al., 1993), a pattern that was replicated at the air-water interface, as observed by neutron reflectivity (Dickinson et al., 1993). The combined experimental evidence was therefore consistent with the view that much of the hydrophobic end of the adsorbed β-casein was directly associated with the hydrophobic interface, with the hydrophilic N-terminal tail extending significantly out into the aqueous phase. Self-consistent-field calculations of the conformation of β-casein adsorbed at a planar hydrophobic interface confirmed this picture of a tail-train structure and also predicted a train-loop-train structure for adsorbed αS1-casein, with anchor points at both ends of the molecule (Leermakers et al., 1996; Dickinson et al., 1997a, b). In the light of these observations, it seemed reasonable to suggest that these hydrophobic regions should be involved in casein aggregation and micelle formation (Horne, 1998). Although the Holt group (Holt et al., 2013; Thorn et al., 2015) accepted a role for the hydrophobic regions (renamed by them as polar tracts), they rejected the proposition that hydrophobic interactions play a part. Indeed, Thorn et al. (2015) went as far as to write “the caseins are not hydrophobic and do not take part in hydrophobic interactions.” In a challenge to this dismissal, Horne and Lucey (2017) showed that the initial arguments of Holt et al. (2013) were without foundation. Without

Primary structure and interactions of caseins

217

responding to these criticisms, Carver et al. (2017) challenged the work of Horne by suggesting that he had been selective in identifying certain amino acids as hydrophobic for an analysis that located hydrophobic amino acid clusters in the caseins (Horne, 2017). As a brief perusal of Horne (2017) will show, he did not select the list of amino acids; he adopted the list used by the authors of the hydrophobic cluster analysis (Gabouriaud et al., 1987), but in a spark of prescience, he verified that listing by analyzing data on the hydrophobicity rankings of amino acids collated by Biswas et al. (2003). Carver et al. (2017) based their assertions on the hydrophobicity scale of Kyte and Doolittle (1982), citing its usage in predicting the caseins as IDPs. However, another of their supporting citations on the hydrophobicity status of proline, from Huang et al. (2014), noted that Kyte and Doolittle (1982) provided anomalous values for the hydrophobicities of tyrosine and tryptophan and went on to suggest an optimized IDP hydrophobicity scale, which lists as hydrophobic those amino acids identified as such by Horne (2017). In reality, a hydrophobicity scale should reflect the chemistry of the side chains. Aliphatic chains and aromatic or heterocyclic rings will always favor a hydrophobic environment. How far they succeed will be dependent on their neighbors and the solvent conditions. Calculating a global ranking, as Horne (2017) attempted, is an attempt to accommodate those unknowns but really does nothing more than acknowledge that chemistry. Lucey and Horne (2018), using the approach of Callebaut et al. (1997) and their software package, available online at http://mobyle.rpbs.univ-paris-diderot.fr/cgi-bin/portal. py#forms::HCA, have identified the hydrophobic clusters of the bovine caseins, shown in the two-dimensional hydrophobic cluster analysis (2D-HCA) plots of Fig. 6.1. These are essentially the same results as those of Horne (2017), now wrapped around a cylinder and then opened out, the aim being simply to visualize those peptide segments that are capable of involvement in hydrophobic interactions within or between casein molecules. They help to give a better idea of how hydrophobic residues in the identified cluster might arrange themselves; how they might clump together, as it were; and how they might provide a grouping that could interact with another such cluster in another region of the same molecule or of an incoming partner. Although they are arranged on a cylinder by the 2D-HCA methodology, this does not necessarily mean that they are locked into that structure when coming into the hydrophobic interaction or for the hydrophobic interaction to occur. Examination of these pictures shows the clumped hydrophobic clusters congregating in the hydrophobic regions away from the phosphoserine clusters; all of this is in line with the schematic structures of Fig 6.2, which are based on the results of the self-consistent-field calculations of the conformations of these proteins adsorbed at a hydrophobic interface (Leermakers et al., 1996; Dickinson et al., 1997a, b). This, of course, is in agreement with the behavior of β-casein self-association into detergent-like micelles, in a mechanism proposed by Payens and Van Markwijk (1963) and Payens et al. (1969). In this picture, the hydrophobic trains are buried inside, and the charged hydrophilic tails extend from the surface into the solution. Neutron and magnetic resonance spectroscopy measurements by De Kruif et al. (2002) suggest that the associated state of βcasein is a highly hydrated assembly with open, flexible polypeptide chains, in contrast to the closely packed interiors of surfactant micelles. More recent small-angle neutron scattering (SANS) and light scattering measurements by Ossowski et al. (2012) suggest a core solvent (water) content of 63% for the ellipsoidal β-casein micelles. Given that Holt is a coauthor

218

6. Casein micelle structure and stability

(A)

(B)

(C)

(D) FIG. 6.1 HCA of (A) αs1-casein, (B) αs2-casein, (C) β-casein, and (D) κ-casein. Shown as 2D-HCA plots. In the 2DHCA plot, the sequence is written on an α-helical net that is unrolled and duplicated. The contours of hydrophobic amino acids are joined together to indicate clusters. Special symbols are used for proline (star), glycine (diamond), serine (dotted square), and threonine (square), as described by Rebehmed et al. (2016). Plots created with the website: http://mobyle.rpbs.univ-paris-diderot.fr/cgi-bin/portal.py?form¼HCA#forms::HCA.

in both the De Kruif et al. (2002) paper and the Ossowski et al. (2012) paper, his rejection of the micellar model for β-casein in Holt et al. (2013) and in particular his insistence that this model demands a close-packed core is quite baffling. The physics of the model are quite simple. The heavily negatively charged N-termini of the β-casein molecules form a corona around

Primary structure and interactions of caseins

αS1-Casein

219

FIG. 6.2 Schematic structures of αS1-casein and β-casein, based on self-consistent-field calculations of the proteins adsorbed on to a hydrophobic interface, illustrating the train-loop-train structure of αS1-casein and the loop-train structure of β-casein. The dashed circles give an idea of the range of the interaction potential components, the larger circle around the loops being the electrostatic repulsion arising from the negative charge centers thereon, and the smaller circle being the regions of hydrophobic attraction in the train.

β-Casein

the micelle, seeking to keep as far from one another as possible, but are prevented from flying into the solution by the attraction between their tails in the core of the micelle. That this attraction is hydrophobic is attested by the temperature dependence of their formation and the observation that the scission of a few hydrophobic residues at the C-terminus tail removes the micelle-forming ability (Qi et al., 2005). Analyses of differential scanning calorimetry studies (De Kruif and Grinberg, 2002; Mikheeva et al., 2003; O’Connell et al., 2003; Portnaya et al., 2006) have concluded that the micelles form as a stepwise addition of monomer to a growing micelle, as suggested in the shell model of Kegeles (1979). The first stage would be the formation of the dimer, perhaps as depicted in Fig. 6.3. Portnaya et al. (2006) confirmed the Kegeles model at lower temperatures, up to 30°C, but proposed that the micellization becomes more cooperative at higher temperatures. None of their results detract from or diminish the role of hydrophobic interactions, but the effects of ionic strength on the measured ionization free energy confirm the importance of the temperature dependence of ionization and its effect on the net charge of the protein and intermolecular repulsions, a factor that is often neglected in such studies.

220

6. Casein micelle structure and stability

FIG. 6.3 Diagrammatic representations of the polymeric structures generated when the hydrophobic chains of the caseins interact: (A) the wormlike chain of αS1-casein and (B) the micelle of β-casein, in which only two molecules have been included to simplify the diagram.

β-Casein micelle

αS1-Casein polymer

In a like fashion, with its hydrophobic chains at opposite ends of the molecule and a central section containing the highly charged phosphoserine clusters, αS1-casein self-associates to produce a wormlike chain polymer (Payens and Schmidt, 1966; Schmidt, 1970a, b). Kumosinski et al. (1994) proposed that the initial dimerization takes place via an antiparallel docking of the hydrophobic clusters in the N-terminal hydrophobic region of two αS1-casein molecules. This hydrophobic interaction is reinforced and stabilized by two intermolecular ion pairs between opposing chains. Although driven by hydrophobic interactions, electrostatic repulsive interactions are also very important in the self-association of these caseins. In particular, note how the equilibrium structures adopted by their polymers place the centers of charge as far apart as possible while still permitting the self-association to take place. Compared with hydrophobic interaction, electrostatic repulsion is a long-range force, an important factor that is now being recognized in studies of protein-protein interactions (Kegel and Van der Schoot, 2004; Piazza, 2004; Stradner et al., 2004), and certainly manifests itself in these reactions of the caseins. These electrostatic interactions define the degree of polymerization and limit further growth. Thus, increasing the pH, which increases the protein charge, decreases the polymer size in both αS1-casein and β-casein solutions, whereas increasing the ionic strength, which decreases the range of the electrostatic repulsion component, allows the formation of larger polymers for both casein species (Payens et al., 1969; Schmidt, 1970a, b). High net charge combined with relatively low mean hydrophobicity leads to the caseins displaying highly disordered structures, firmly identifying them within the class of IDPs. For a review of caseins in this context, the reader is directed to Redwan et al. (2015). Mineralization of IDPs is included in a review of their behavior and properties (Uversky, 2011). This is particularly relevant for the caseins, where the importance of charge in controlling the extent of aggregation of the caseins cannot be stressed too highly. Precipitation of the caseins can be achieved by lowering the pH and titrating away sufficient of the charge of the phosphoseryl and carboxyl groups to reach the isoelectric points of the proteins. αS2-Casein,

Casein micelle properties

221

αS1-casein, and β-casein are termed the calcium-sensitive caseins because they can be precipitated in the presence of ionic calcium, the order of sensitivity being as given (Swaisgood, 2003). The most extensive dataset is available for αS1-casein. Here, the aggregation shows a lag phase with little change in molecular weight with time until a critical time, beyond which rapid aggregation occurs. Horne and Dalgleish (1980) demonstrated that the logarithm of this critical coagulation time was a linear function of Q2, where Q is the net negative charge of the protein. Thus, Q is the algebraic sum of the negative and positive charges of the protein, reduced by twice the number of calcium ions bound to the protein, each calcium carrying two positive charges. Furthermore, this relationship held when changes in the net negative charge were produced by chemical modification, whether by conversion of positively charged lysine residues to neutral or negatively charged derivatives or even by the introduction of new negatively charged sites by iodination of tyrosine residues to the di-iodoform (Horne, 1983; Horne and Moir, 1984). Each of these modifications effectively increases the net negative charge of the protein, thereby reducing its propensity for calcium-induced precipitation and slowing down the rate of aggregation. However, once the protein charge is corrected for the measured extent of modification, the logarithm of the rate of precipitation has been shown to remain linear in Q2, all points lying on the same line as those obtained with the unmodified protein and all other reaction conditions being the same (Horne, 1983; Horne and Moir, 1984). The net charge of the protein therefore dominates its precipitation behavior. Farrell et al. (2006) have suggested that positively charged residues in the N-terminal hydrophobic chain of αS1-casein could participate in binding to the phosphate groups of phosphoseryl residues but such  bridging does nothing to reduce the net negative charge of the protein, having already been accounted for in the algebraic summation leading to Q, which, as we have demonstrated, controls the level of aggregation in these proteins. It is only when the local balance of electrostatic repulsion and hydrophobic attractive interaction is in favor of attraction that hydrophobic bonds are formed. Sequentially, the major centers of electrostatic repulsion, the phosphoserine clusters, are remote from the hydrophobic regions in the trains of Fig. 6.2, although, depending on the adopted conformation, they may not be remote spatially. Fig. 6.2 is a depiction of a possible conformation of each protein at a hydrophobic interface, the puckering displaying the tendency of the hydrophobic chain to form a multitude of weak, short-range bonds. Lowering the temperature weakens hydrophobic bonds; hence, there is a tendency to find monomeric β-casein at low temperatures but a micellar aggregate at room temperature and above. The aggregation of β-casein induced by calcium also shows a marked temperature dependence with no precipitation observed at 4°C (Parker and Dalgleish, 1981). At higher temperature, these hydrophobic bonds are in general stronger, but because they are relatively weak overall, statistically individual bonds are readily ruptured by the increased thermal energy available, leading to more mobile, labile interactions between molecules.

Casein micelle properties Almost all of the casein proteins present in bovine milk expressed at 37°C are incorporated into the casein micelles, together with a high proportion of the available calcium and inorganic phosphate. The calcium and phosphate within the micelle form low-molecular-mass

222

6. Casein micelle structure and stability

species collectively known variously as colloidal calcium phosphate, micellar calcium phosphate, and, latterly, calcium phosphate nanoclusters. The micelles are very open, highly hydrated structures, with typical hydration values of 2–3 g H2O/g protein, depending on the method of measurement. Electron microscopy shows that casein micelles are generally spherical in shape, with diameters ranging from 50 to 500 nm (average  150 nm) and a molecular weight ranging from 106 to >109 Da (average  108 Da) (Fox and Brodkorb, 2008). For a casein content of 2.5 g/100 mL milk, there are some 1014–1016 micelles/mL milk, which implies a relatively close packing with intersurface separations of less than one micelle diameter. Milk is white largely because the colloidal dimensions of the casein micelles are such that they scatter significant amounts of light, an effect that is compounded by their high number density. Scattering of shorter wavelength radiation (neutrons and X-rays) reveals the internal structure to be heterogeneous, with a correlation length for variations in scattering length density within the particle of approximately 18 nm. This scattering behavior has been interpreted by Stothart and Cebula (1982) as being due to a structure that is composed of closely packed spherical subunits of this diameter, a picture that mirrors the raspberrylike appearance in early electron micrographs of the casein micelle (Schmidt, 1982). More recent small-angle X-ray scattering (SAXS) and SANS studies (De Kruif et al., 2012; De Kruif, 2014; Ingham et al., 2015, 2016; Day et al., 2017) interpret this feature of the scattering spectrum, a shoulder at  0.35 nm
1, as the correlation length separating the calcium phosphate nanoclusters, the scattering of which gives rise to a second shoulder in the SANS/SAXS spectra at  0.8 nm
1. To further interpret the SANS/SAXS spectra, authors have relied on guidance on the internal micellar structure that has been gleaned from electron microscopy studies. McMahon and McManus (1998), supported by McMahon and Oomen (2008), proposed that the casein micelle was a homogeneous mesh of protein interspersed with small regions of high electron density, 2–3 nm in diameter, that is, a granular picture. They suggested that the well-defined structures seen in earlier studies were probably artefacts of the fixation procedures then employed. Marchin et al. (2007), using cryo-transmission electron microscopy (cryo-TEM), supported the conclusions of McMahon and coworkers and, in a pH study, found that the granular aspect diminished as the pH was reduced from 6.7 to 5.2. Paralleling this disappearance, the shoulder in the measured SAXS spectra at  0.35 nm
1 was also lost (Marchin et al., 2007). For their fitting of their SAXS and SANS data, De Kruif et al. (2012) adopted a model that considered the casein micelle to be mainly homogeneous but with calcium phosphate nanoclusters dispersed throughout, like currants in a bun. The shoulder at 0.35 nm
1 was interpreted as the correlation length between the nanoclusters, giving a separation distance of 18.6 nm. For their “average micelle size,” based on this spacing, De Kruif et al. (2012) calculated that there were 285 nanoclusters per micelle. In his later review of micellar SANS/ SAXS data published to that date, De Kruif (2014) concluded that there was insufficient scattering from this number of small nanoclusters (2–3 nm in size) to account for the high qshoulder at 0.8–1.0 nm
1 and suggested that this was due to clumps or nodules of interacting casein proteins. Meanwhile, other studies on casein micelles were suggesting that the assumption of a homogeneous protein distribution was perhaps less reliable than first thought. Micrographs obtained by field emission scanning electron microscopy showed a complex surface structure

Casein micelle properties

223

of cylindrical or tubular, but not spherical, protrusions between 10 and 20 nm in diameter, extending from the surface of the micelle (Dalgleish et al., 2004). These samples were not metal coated, although they were of necessity subjected to a fixation and dehydration process, which might have introduced some collapse of more loosely bound protein onto a denser skeleton. Cryo-TEM was also employed by Trejo et al. (2011) to study the internal micellar structure. By varying the angle of incidence of the electron beam onto the same micelle, they constructed tomographic images that demonstrated the presence of water-filled channels and cavities within the micelle and mapped the size, number, and location of the calcium phosphate nanoclusters, which were significantly different from the seemingly homogeneous networks apparent in earlier cryo-TEM studies (Marchin et al., 2007; McMahon and Oomen, 2008). Day et al. (2017) presented similar tomographic images of the casein micelle, laying bare a higher-density skeleton structure within the micelle. Although these interpretations chime with observations on the internal accessibility of the micelles to enzymes and the ready release of β-casein, a word of caution should be injected. The analysis by Trejo et al. (2011) is essentially a paint-by-numbers exercise. If the electron density is less than x, color it black and call it a channel or cavity; if the density is greater than y, then color it purple and call it a calcium phosphate nanocluster. As there is no chemical identification of the scattering species, both the number and the size of the nanoclusters are likely to be an overestimate and similarly for the channels and cavities. The tomographic images could be regarded as a clinical x-ray of the casein micelle, revealing a skeleton but not the lower-density soft tissue (protein) surrounding it. A similar picture emerged from the experiments of Bouchoux et al. (2010), who combined the osmotic stress technique with SAXS to study the structural response of the micelle to an increase in concentration. Their SAXS results indicated that, as the micelles were compressed, they lost water and shrank to a smaller volume, but this compression was nonaffine; that is, some parts of the micelle readily lost water and collapsed, whereas other harder parts resisted deformation but were pushed closer together. The structure that Bouchoux et al. (2010) proposed for the casein micelle bears a strong resemblance to the tomographic images of Trejo et al. (2011) and Day et al. (2017), with the calcium phosphate nanoclusters distributed within the hard skeletal regions. In the analyses of their own SANS/SAXS data, Ingham et al. (2016) attempted to reproduce such pictures, by considering four populations of contributors to the scattering. These are the form factors of the casein micelle itself, of a number of spheres representing the hard skeletal region, of spherical calcium phosphate nanoclusters over which a much larger hard sphere radius is applied—this reproduces the shoulder seen at 0.35 nm
1—and of a distribution of polymeric (protein) scattering regions to encompass the wide angle shoulder at  1 nm
1. There is no indication of how the hard skeletal spheres are distributed and no correlated contribution from these or the hard sphere-“protected” nanoclusters. Instead, the radius of these hard spheres keeps the nanoclusters apart at the same separation as in the De Kruif et al. (2012) calculations. Hence, Ingham et al. (2016) also restricted themselves to a maximum, of around 300 nanoclusters in their “typical” micelle. They obtained excellent fits, using these functions, to both scattering profiles and the Kratky plots (ln(I(q)  q2) vs. ln q2) derived from these functions, but there are still some questions relating to their modeling of the internal micelle structure. They also measured scattering spectra following the addition of aliquots of ethylenediaminetetracetic acid (EDTA) and a time series following the addition

224

6. Casein micelle structure and stability

of glucono-δ-lactone (GDL), the latter hydrolyzing and producing a gradual acidification of their milk sample over time. Additions of 5- and 10-mM EDTA were made, sufficient to sequester a maximum of about 50% of the Ca2+ in the nanoclusters. This produced little spectral change in the shoulder at 0.35 nm
1, meaning that the nanocluster core factor, its major contributor, remained unaltered. Instead, the shoulder at 0.8–1.0 nm
1, that is, the shoulder ascribed by Ingham et al. (2016) to protein inhomogeneities, was diminished by approximately 50%. Ingham et al. (2016) interpreted this as an indication that the protein structure became more uniform as casein proteins were released from nanocluster involvement, which is a possibility because the serine phosphate negative charge is restored by the sequestration of the Ca2+. However, this shoulder rapidly disappeared on the acidification in their GDL series of measurements. The acidification also disrupted the nanoclusters, but the protein charge was also titrated away, so that there was no cause for immediate breakdown of the protein clusters or nodules. Although the protein interactions are weak and rearrangements are to be anticipated, they perhaps do not occur on the timescale of this study. An alternative interpretation is presented later, after discussion of the origin and structure of the nanoclusters themselves. To conclude this discussion, casein micelle structure is not fixed; rather, it is dynamic. In various ways, it responds to changes in the micellar environment, temperature, and pressure. Cooling milk on release from the udder at 37°C to storage at refrigeration temperatures brings about significant solubilization of β-casein, some κ-casein, and much lower amounts of αS1and αS2-casein from the micelles (Dalgleish and Law, 1988). Raising the temperature back to 37°C reverses the process. None of this movement of β-casein does anything to disrupt the internal structure of the micelle, as observed by cryo-TEM and SAXS (Marchin et al., 2007). Almost complete disruption of the micelles, manifested by a loss of their scattering power and removal of the white color of milk, can be achieved by the addition of sufficient of a strong calcium sequestrant such as EDTA (Griffin et al., 1988), by the addition of urea (McGann and Fox, 1974), by dialysis against a phosphate-free buffer (Holt et al., 1986), by increasing the pH, by exposure to high pressure (Huppertz et al., 2006), or by the addition of ethanol at  70°C (O’Connell et al., 2001). Significantly, the calcium phosphate nanoclusters can also be solubilized by lowering the pH but, as confirmed by Marchin et al. (2007) using cryo-TEM, without substantial disruption of the micelle structure. Fractionation of the casein micelles according to size can be realized by a stepwise centrifugation protocol. The proportions of αS1- and αS2-caseins remain constant with the micelle size, but the κ-casein content increases inversely with that size (Donnelly et al., 1984; Dalgleish et al., 1989). For a solid sphere, the surface-to-volume ratio is inversely proportional to the radius of the sphere, and these results imply that κ-casein resides on the micellar surface, where its content controls the total micellar surface area and hence the micelle size. A surface location for the κ-casein component may also be inferred from the requirement that this protein be readily accessible for rapid and specific hydrolysis by chymosin and similar proteases, a reaction that destabilizes the micelles and leads to clot formation, which is exploited in cheese manufacture. A surface location is also required to enable the κ-casein to interact with the β-lactoglobulin in milk to form a complex on heating, the formation of which modifies the rennet and acid coagulation properties of the micelles. It is evident that a principal requirement, which must be met by any micelle model, is that it should cause κ-casein to be located on the micelle surface.

Dual-binding model for micelle assembly and structure

225

Models of casein micelle structure Casein micelle structure and casein micelle models have been extensively reviewed (Schmidt, 1982; Walstra, 1990, 1998; Holt, 1992; Rollema, 1992; Horne, 1992, 1998, 2006; Fox, 2003; Farrell et al., 2006; Fox and Brodkorb, 2008). One reviewer of a recent submission made the comment that all micelle models were simply concepts and all had merit in some way or another. This could not be further from the truth. An acceptable model for the casein micelle must provide an adequate mechanism for its formation, represent its internal structure in the manner informed by available experimental data, and map or predict the behavior encountered in the many dairy technology situations in which milk or its derivatives might be used. This is a challenging remit for any model. Should the model fail in any way at any stage, drastic rethinking is required. Based on the biochemical and physical properties of the micelles and the casein proteins already outlined, three main models have been proposed: the submicelle model (Slattery and Evard, 1973; Schmidt, 1982; Walstra, 1998), the nanocluster model of Holt (Holt, 1992; De Kruif and Holt, 2003), and the dual-binding model (Horne, 1998, 2002). In the submicelle model, the casein micelles are composed of smaller proteinaceous subunits, the submicelles, linked together via calcium phosphate nanoclusters. In the second model, that of Holt (1992), the calcium phosphate nanoclusters are randomly distributed, cross-linking a threedimensional web of casein molecules. Both these models have been severely criticized (Farrell et al., 2006; Horne, 2006), and the dual-binding model (Horne, 1998) arose first as an attempt to overcome their deficiencies. It is significant that the most recent description of the nanocluster model (De Kruif et al., 2012) adopts tacitly an essentially dual-binding approach, replacing the terminology “hydrophobic interactions” with “nonspecific interactions.” Carver et al. (2017) are quite adamant that the De Kruif and Holt (2003) variant is not a dual-binding model, but the description in Holt et al. (2013) is difficult to interpret in any other terms. This being the case, we present first a summary of the dual-binding model as providing a rational mechanism for micelle assembly and structure and demonstrate how this model may be exploited to explain various observations of micellar properties and behavior.

Dual-binding model for micelle assembly and structure The description here largely follows that found in Horne (2002), with minor refinements highlighted. In the dual-binding model, micellar assembly and growth take place by polymerization processes involving, as the name suggests, two distinct forms of bonding, namely, crosslinking through hydrophobic regions of the caseins and bridging across calcium phosphate nanoclusters. The locations of the hydrophobic clusters have been identified earlier in this chapter, and their mechanism of interaction is summarized in Horne (2017). We cannot be certain that hydrophobic interactions are the only source of attraction contributing to the second binding pathway. De Kruif and Holt (2003) echoed this in placing hydrophobic interactions, hydrogen bonding, and other forces under the umbrella of nonspecific interactions

226

6. Casein micelle structure and stability

when they adopted their version of the dual-binding model. In part, this was justified because, when Horne first published the original model in 1998, the recognized partial dissociation of micelles in urea solutions was thought to be through disruption of hydrogen bonds. Urea (Shimizu, 2011; Guilherme et al., 2015) is now considered to be a disrupter of hydrophobic interactions, providing further evidence of their involvement in casein micelle formation and stability. Holt et al. (2013) still dispute the domination of hydrophobic interactions, insisting on putative main chain-to-main chain interactions. Recently, Holt has serially renamed historically accepted terms and concepts. Thus, the hydrophobic regions of the caseins became sticky P.Q-rich regions (Holt et al., 2013) and then polar tracts (Holt, 2016), despite the original definition being a sequence devoid of hydrophobic residues and proline. The nonspecific interactions of De Kruif and Holt (2003) became entropic interactions in Holt et al. (2013), and the interaction of side chains in a hydrophobic interaction became the formation of a dry interface (Holt et al., 2013; Supplementary Information). Although Holt has generally gone from historic to new in these trends, the identity of the main chain-to-main chain interactions now suggested remains undefined. Horne (2017) made the assumption that these were hydrogen bonds and argued against these interactions, but should that choice be a misinterpretation of Holt’s ideas, there still remains the role of the residue side chains, and those that are hydrophobic will act according to their nature. For that reason and because of the significant amount of evidence in their favor, we continue to assert the dominance of hydrophobic interactions in the second binding pathway of the dual-binding model and note that bond formation through this pathway is facilitated and hence micellar integrity and stability are maintained, by a local excess of hydrophobic attraction over electrostatic repulsion, bearing in mind the quite different ranges of these interaction components. The individual casein molecules behave and interact as they do in their selfassociation equilibria, as described previously. Each casein molecule effectively functions as a block copolymer, as detailed in Fig. 6.2, with the hydrophobic region(s) offering the opportunity for a multitude of individual, weak, hydrophobic interactions. The hydrophilic regions of the casein molecules contain the phosphoserine cluster (or clusters), with the exception of κ-casein, which has no such cluster, each offering multiple functionality for cross-linking. Thus, as we have seen, αS1-casein can polymerize (self-associate) through the hydrophobic blocks, giving the wormlike chain of Fig. 6.3. Further growth is limited by the strong electrostatic repulsion of the hydrophilic regions, but, in the casein micelle situation, the negative charges of the phosphoserine clusters are neutralized by intercalating their phosphate groups into a facet of the calcium phosphate nanocluster. This has two very important implications for the micelle. Firstly, by the removal of a major electrostatic repulsion component, it increases the propensity for hydrophobic bonding upstream and downstream of the nanocluster link. It effectively permits and strengthens those bonds. Secondly, it allows for multiple protein binding to each nanocluster, allowing a different network to be built up. β-Casein, with only two blocks, a hydrophilic region containing its phosphoserine cluster and the hydrophobic C-terminal tail, can form polymer links into the network through both, allowing further chain extension through both. αS2Casein is envisaged in this model as having two of each block, two (possibly three, see later) phosphoserine clusters and two hydrophobic regions. It is only a small fraction of the total bovine casein, but by being able to sustain growth through all its blocks, it is likely to be bound tightly into the network. κ-Casein is the most important of the caseins in the dual-binding

Calcium phosphate nanoclusters

227

model of micellar assembly and structure. It can link into the growing chains through its hydrophobic N-terminal block, but its C-terminal block is hydrophilic and cannot sustain growth by linking hydrophobically to another casein molecule. Neither does κ-casein possess a phosphoserine cluster, and therefore, it cannot extend the polymer cluster through a nanocluster link. Thus, chain and network growth are terminated wherever κ-casein joins the chain. This leaves the network with an outer layer of κ-casein, satisfying the prime requirement recognized earlier. This assembly process is, in its essential requirements, identical to the model proposed by McMahon and Oomen (2008). Their main contribution is to describe the structure resulting from the assembly processes as an interlocked lattice, which seems to be an excellent description of a complex structure. The nanocluster bridging pathway through the phosphoserine clusters was the only pathway allowed in the original nanocluster micelle model of Holt (Holt, 1992; De Kruif and Holt, 2003). Some consideration must also be given to what constitutes a phosphoserine cluster that is capable of linking into the calcium phosphate nanocluster. Aoki et al. (1992) suggested a minimum of three phosphoserine residues, but De Kruif and Holt (2003) argued that two might be sufficient. This would allow the phosphoserine pair at positions 46 and 48 of αS1casein or those at positions 129 and 131 of αS2-casein, to function as nanocluster linkage sites, particularly if the carboxyls of the neighboring glutamate residues acted as pseudophosphate groups. This would give αS1-casein two linkage sites and αS2-casein three linkage sites. This level of functionality in these caseins is absolutely essential to the Holt model to build the required three-dimensional network, as, without them, an αS2-casein molecule with only two linkage sites and with such a low percentage of the total casein would probably prove to be insufficient. Although they are not essential to the dual-binding model, these miniclusters of pairs of phosphoserines may provide for a weaker bridging link to the calcium phosphate nanocluster, allowing a range of nanocluster bond strengths to prevail.

Calcium phosphate nanoclusters From the mineral viewpoint, casein micelle assembly is a frustrated crystallization of calcium phosphate. Milk is supersaturated in calcium and phosphate, and were it not for the presence and intervention of the highly phosphorylated caseins, a precipitation of calcium phosphate and potentially painful calcification of the mammary gland duct system would occur. There is considerable controversy over the possible (crystalline) structure of the nanocluster. Thorn et al. (2015) believe that the radius and mass of the calcium phosphate nanocluster should be the same as that of the particle created by the controlled precipitation of calcium phosphate in the presence of β-casein phosphopeptide (Holt et al., 1996, 1998; Little and Holt, 2004). This particle has an inner core of amorphous calcium phosphate of radius 2.7 nm and is surrounded by a shell of around 50 phosphopeptides. Thorn et al. (2015) prepared these particles from a mixture of calcium and phosphate salts and β-casein phosphopeptide, with the salts at levels equivalent to their final milk concentrations and the peptide at a sufficient level to provide phosphate centers equivalent to the total casein levels in bovine milk. Instant precipitation on mixing was prevented by mixing at a low pH,  5.0. The solution pH was then raised in a controlled fashion by the action of the enzyme urease on the included

228

6. Casein micelle structure and stability

urea. The latter reagents control the rate of increase in pH and hence the size of the nanoclusters produced (Holt et al., 1996). Little and Holt (2004) reported that all preparations adjust to the same particle size over a period of hours, but is such a time available in the mammary gland? This renders the size postulated for the micellar nanoclusters to be somewhat arbitrary. However, of further concern is the behavior of the SAXS profiles of these nanoclusters (Little and Holt, 2004). When redrawn as a Kratky plot, their data show a maximum at q ¼ 0.8 nm
1, suggesting that these exemplary nanoclusters are ellipsoidal or rodlike. If such proves to be the case, how appropriate are the spherical-based functions used by De Kruif et al. (2012) and De Kruif (2014) in fitting micellar SAXS and SANS profiles, and how valid are the conclusions drawn? It is also unlikely that the conditions employed in the preparation of the Holt et al. nanoclusters, that is, low pH, extant concentrations of salts and protein, and urea and urease availability, are the conditions present in the mammary gland and the Golgi vesicles. The pathways for the transport of calcium and phosphate have yet to be fully elucidated (Shennan and Peaker, 2000; Farrell et al., 2006), but none of those postulated involve a low pH scenario, and all propose a transporting partner. However, there is an alternative route to the creation of micellar calcium phosphate nanoclusters, which exploits the biomineralization properties of the serine phosphate centers of the caseins. The Holt mechanism invokes the classic nucleation and growth mechanism of crystallization, with the phosphate centers binding to the growing nanocluster nucleus (De Kruif et al., 2012). Horne (1982) found that the inclusion of phosphate in a solution of αS1-casein, to which calcium was then added, induced rates of precipitation at those ion product levels that were orders of magnitude greater than would be anticipated when protein is not present. The rates were also higher than those encountered with the same calcium concentration but in the absence of phosphate. In other words, the phosphate was taking an active primary role in inducing precipitate formation. This would appear to preclude the phosphoserines of the caseins binding to a preformed calcium phosphate nanocluster nucleus, as required by the Holt scenario, but would offer the potential for the phosphoserine cluster to act as a template to initiate nanocluster growth, to moderate that growth by closing off at a particular facet, and finally to terminate that growth at the final free facet. Pursuing these ideas to their logical conclusion and invoking only knowledge of the gross composition of the bovine casein micelle, of both proteins and minerals, Horne et al. (2007) developed a picture of the micellar calcium phosphate nanocluster, predicting both molecular weight and size. Previously known facts about the bovine micelles show that the ratio of calcium to inorganic phosphate is approximately 1.5, close to that of hydroxyapatite salt and closer still to that of amorphous calcium phosphate. When the organic ester phosphate is included in this mix, the Ca/P ratio drops to about 1.0, that is, close to the dicalcium salt figure (Holt, 1997). The molecular weight calculation is then as follows. For every cluster motif of 4 ester phosphates, there are associated 8 inorganic phosphates and 12 matching calcium ions. This inorganic portion has a molecular weight of 1.24 kDa. Horne et al. (2007) envisaged that all phosphoserine clusters would be involved in the production and stabilization of the nanoclusters. They calculated that there would be six such clusters on the three calciumsensitive caseins per 100 kDa of the casein micelle. In their scenario, these six phosphoserine clusters were incorporated into planes of crystalline mineral phosphate, forming the faces of a

Calcium phosphate nanoclusters

229

hexagonal bipyramid. Each phosphate of the organic phosphoserine took a surface position in the mineral lattice. With six phosphoserine cluster motifs, this would mean a nanocluster core molecular weight of 7344 Da, that is, close to the 6%–8% mineral content of the casein micelle. Choi et al. (2011) have measured the molecular weights of peptide-stabilized nanoclusters excised from casein micelle suspensions using either trypsin or papain. The trypsin-excised material had a molecular weight of 17,450 Da; the papain-hydrolyzed material had a molecular weight of 23,940 Da, with this enzyme cleaving at different scission points in the protein sequence from those of trypsin. Estimating the molecular weight of each stabilizing phosphopeptide to be about 2500 Da would leave the calcium phosphate nanocluster inorganic core at around 2500 Da for trypsin scission or 9000 Da for papain scission, if six peptides were incorporated. If these were reduced to four peptides, then the core molecular weight could be 7500–14,000 Da, and stabilization by four peptides could easily be accommodated as arising as three nanoclusters from 200 kDa of micelle. The experimental figure for the core would move closer to the theoretical estimate when it is recalled that this would also include water of crystallization, driven off in ashing to determine the mineral content, which is 2 moles of water per mole of calcium phosphate in brushite. Not all of the phosphoserine clusters consist of four serine phosphates. In fact, Aoki et al. (1992) suggested that at least three serine phosphates would suffice and Holt concurred with this. In later work, Holt (2016) went further and suggested that all serine phosphates should be considered to be active in calcium phosphate nanocluster formation, and we accept this view. However, the most important requirement is that brushite stoichiometry be maintained in the calcium phosphate nanocluster. Evidence that this is maintained across the milks of different species comes from the work of Jenness (1979), who observed linear correlations in plots of total milk calcium and total milk phosphorus versus the casein contents of the milks of 33 species. Assuming a casein monomer in these milks to have a mean molecular weight of 22,500 Da, the slope values of these plots corresponded to 20 Ca2+ ions/monomer and 18 moles of micellar phosphate (ester bound + inorganic P)/monomer, that is, close to the 1:1 stoichiometry of dicalcium phosphate. The average phosphoserine content of bovine casein is about 6 moles/monomer, which again realizes a 1:2 ratio of PSer:Pi in the micelle, this apparently occurring over a wide range of species with their inherent variation in overall milk composition. On the opposite side of the coin, determining the inorganic phosphate content of micellar casein will reveal the degree of phosphorylation of the casein but not the distribution of those sites across the various casein proteins. As these phosphate nanoclusters will also incorporate balancing calcium ion levels, the micellar content of calcium will by subtraction give the number of calcium ions bound to other sites within the micelle. Evidence for the presence of crystalline brushite in the nanoclusters comes from Holt et al. (1982), who found that the spectrum of milk calcium phosphate obtained by X-ray fine structure absorption was close to that of a brushite sample, but only the earliest peaks in the radial distribution functions could be obtained. This was taken to mean that the nanocrystals did not have any long-range extended structure and were therefore amorphous; this conclusion was supported by the calculation of X-ray diffraction patterns as a function of crystal size by Lenton et al. (2016), who found peaks appearing only beyond a few unit cells in dimension. This is a prime example of how we see things differently from Holt; with the same evidence, a

230

6. Casein micelle structure and stability

lack of long-range features simply means that the nanoclusters are small, not more than a few unit cells, just as the calculation of their molecular weight reveals. There are also strong thermodynamic arguments in favor of this small size and our proposed mechanism of nanocluster assembly. It gives the nanocluster an inorganic crystalline core, among the more stable structures envisaged for calcium phosphate (e.g., see Holt, 2013). When coupled with the serine phosphate clusters of the caseins, this structure is extraordinarily stable. There is no migration from the casein micelles or maturation/transformation into other forms of phosphate over extended storage of milk, although heat treatment may induce some further changes. If the incorporation of the serine phosphates is thermodynamically beneficial, then the micellar system actually confirms this by creating the largest surface area and hence the smallest size for its nanoclusters. Lenton et al. (2015) contended that stronger peptide bonding favors larger nanoclusters, but this argument is fallacious, as it dilutes the energy benefit derived from the more stable phosphoserine binding. In summary, we see the formation of nanoclusters via a biomineralization mechanism with the phosphate centers of the caseins first initiating the reaction through the binding of calcium. This is stronger than binding to an isolated phosphate anion and is therefore favored (Mekmene and Gaucheron, 2011). This structure then accumulates further phosphate and calcium ions in a brushite lattice framework, which is closed off to further growth and completed by other phosphoserine clusters. These clusters can come from the same casein molecule or from other casein molecules in the vicinity. These, in turn, may be part of other phosphate nanoclusters in an extended interlinked network. A minimum of four phosphate centers is envisaged for each nanocluster. If these come from four β-casein molecules and this is possible, then the only way in which this nanocluster can be incorporated into the micelle is through hydrophobic interactions between its hydrophobic clusters and those of other molecules. Such interactions are weaker at lower temperatures and allow the release of β-casein into the serum phase at lower temperatures. This is despite the involvement of the phosphoserine clusters approaching saturation levels. There is some release of micellar calcium and phosphate at low temperatures at the natural milk pH (Dalgleish and Law, 1989), but no one has attempted relating the quantities or tracing their origin. As to the other phosphate nanoclusters locked up in the protein matrix, their spatial separation is determined by the lengths of the peptides between the phosphate centers on those nanoclusters and their extension. There are less than 20 residues between the phosphate centers on αS1-casein, the most abundant of the multi-P-clustered caseins. At approximately 0.3 nm/residue, this peptide will stretch at most to about 7 nm, nowhere near the 18 nm required in the De Kruif et al. (2012) picture. It is almost a physical impossibility to achieve that separation and no other, because the randomness of the cross-linking will undoubtedly result in a spread of distances. Thus, a picture of much smaller nanoclusters and many more of them is emerging; that is, several thousands, as opposed to low hundreds, separated spatially within the micelle by much smaller distances. How can this be reconciled with the scattering data? A plausible model might start by considering how a clumping of these small nanoclusters might appear as a much larger, lower-density entity, postulating a range of sizes for these clusters and allowing lower-density regions between them. Such a model might look like the structures proposed by Bouchoux et al. (2010) and might behave in scattering terms like the data of Ingham et al. (2016) or Day et al. (2017) when EDTA is introduced.

Application of the dual-binding model

231

Application of the dual-binding model Predictions of casein micelle properties Size and appearance A major failing of the earlier micelle models was their lack of a plausible mechanism for assembly, growth, and, more importantly, termination of growth. All such elements are in place in the dual-binding model. Furthermore, the product of the dual-binding model satisfactorily represents the appearance of and can be tested against the scattering behavior of the native casein micelle. Network growth is envisaged as a random process, and its termination along any particular pathway depends on the serendipitous arrival of a κ-casein molecule. Micelle size will therefore depend on the proportion of κ-casein in the mix but will also present a range of sizes, dependent as it is on random events. Until this point, we have been emphasizing the interactions involved in casein micelle assembly, but the final structure of the micelle is dictated by the kinetics of the aggregation processes. Simulations of particle aggregation reactions (Meakin, 1999) demonstrated that open, ramified structures result when weak repulsion exists between particles. Essentially, every collision between particles creates a bond, and chains and clusters grow from their extremities. When reaction barriers are higher, more collisions between particles are required before a favorable encounter produces a bond. This extensive sampling of phase space produces denser, more closely packed aggregates. These computer simulations were confirmed experimentally (Weitz et al., 1991). There is no doubt that casein micelles are open, ramified, highly hydrated structures with extensive waterlogged channels and cavities, most recently pictured in the cryo-TEM study of Trejo et al. (2011). The creation of such structures is dictated by the kinetics of the assembly process, implying that such reactions are rapid, occurring on almost every encounter, and demonstrate low or insignificant energy barriers to bond formation. The dual-binding model presents the casein micelle as a dynamic, “living” entity. The hydrophobic interactions are individually weak and capable of breaking and recombining on an almost continuous basis. To some extent, the molecular movements this allows will be restricted by the potentially stronger nanocluster linkages and the low probability of rupturing simultaneously all hydrophobic bonds involving any particular molecule. However, molecular movement has several consequences. The micelle may have an outer layer of individual κ-casein molecules when initially constructed, but conditions within the Golgi vesicle are suitable for disulfide bond formation; otherwise, β-lactoglobulin and the other whey proteins would not fold properly. Movements within that outer “hairy” layer may bring those κcaseins into proximity and allow their polymerization through disulfide bridging, the size of the polymer depending on when the chain closes into a loop but giving rise to the polymeric κ-casein entities observed on micellar dissociation. Moreover, the greater the degree of polymerization, the more difficult it is to dissociate the polymer from the micelle surface when the milk is cooled and the hydrophobic bonds are weakened; thus, this provides an explanation for the lack of solubilization of κ-casein on cooling, compared with β-casein. Another consequence of the “living” nature of these hydrophobic bonds is that the dehydration of the micelle required in the preparation of a sample for some forms of electron microscopy would also tend to be accompanied by the collapse of the more mobile, weaker, and less multitudinously bonded regions onto those more strongly cross-linked; hence, this

232

6. Casein micelle structure and stability

perhaps gives rise to the raspberrylike (Schmidt, 1982) or tubular (Dalgleish et al., 2004) structures seen in some electron micrographs. Even the putative caps suggested for those tubules by Dalgleish et al. (2004) can be provided by the dual-binding model, as the disulfide bridging between the κ-casein molecules would enhance the Velcro effect of the weak hydrophobic bonding of an individual molecule to many such molecules in the chain. Effects of urea, pH, sequestrants, and temperature The concept of a local excess of hydrophobic attraction over electrostatic repulsion and permitting the visualization of micellar growth successfully accommodates the response of the micelle to changes in pH, temperature, urea addition, or removal of calcium phosphate by sequestrants, all in accordance with experimental observations. Urea disrupts hydrophobic bonds, and high concentrations will bring about micellar disintegration. In some regions of the micelle, this may be only partial because the nanocluster cross-links through the phosphoserines remain unaffected by this reagent. Micellar fragments in a range of sizes may be produced, even some as large as some of the original micelles, although perhaps more open and swollen from their own starting state before urea treatment. Extensive disruption does occur, however, as is observed by the loss of the white appearance of skim milk (McGann and Fox, 1974). The dual-binding model fully accounts for these observations. Removal of calcium from the calcium phosphate nanocluster by sequestrant addition, whether EDTA, citrate, or oxalate, restores the negative charge of the hydrophilic region, if the pH is maintained at the native milk pH. This shifts the hydrophobic attraction/electrostatic repulsion balance in favor of repulsion, and the micelle breaks up. Decreasing the milk pH solubilizes the colloidal calcium phosphate (Dalgleish and Law, 1989), but the negative charges associated with the cross-linking phosphoseryl groups are also titrated away. The strength of the hydrophobic bonds remains unaffected or may be enhanced if other carboxyl charges are also titrated away. The integrity of the micelles is maintained, but their scattering behavior and their appearance in cryo-TEM micrographs reflect the loss of the nanoclusters (Marchin et al., 2007; Moitzi et al., 2011). The loss of the nanocluster linkages also produces a smoothing and spreading of the micelles as seen by atomic force microscopy (AFM) on lowering the pH around an adsorbed micelle, as internal bonding becomes dominated by the multitudinous but ephemeral hydrophobic bonds (Ouanezar et al., 2012). Increasing the pH may be expected to be the reverse of the dissolution process and to favor the formation of calcium phosphate species. Fox (2003) noted that raising the pH to >9.0 does not dissolve colloidal calcium phosphate but rather increases its level. However, increasing the milk pH to these levels does lead to the dissociation of the micelles and the creation of a translucent solution. Dialyzing these high pH solutions against excess of the original milk restores the milk pH and produces milks with enhanced levels of colloidal calcium phosphate. It is argued (Oczan et al., 2011) that this is incorporated into larger nanoclusters rather than increasing the number of nanoclusters. There are also shifts in the soluble calcium phosphate equilibria in milk that are associated with temperature change. Ultrafiltration permeate is a clear, straw-yellow liquid when prepared at 4°C but becomes turbid when heated to room temperature and above because of the precipitation of calcium phosphate. Even permeate collected at room temperature clouds on heating but reverts to clarity on cooling.

Application of the dual-binding model

233

Hilgemann and Jenness (1951) noted that calcium phosphate also precipitated in milk. However, the precipitate resolubilized only slowly ( Jenness and Patton, 1959). Weakening the calcium phosphate “solution” equilibrium would favor preservation of the nanoclusters, but anything that pushes that “solution” equilibrium to the solid side could have an effect on the continuing existence of the nanoclusters. There are indications that heating milks in the temperature range 50°C–90°C brings about increasing mobility in the micelle (Rollema and Branches, 1989), which would be in line with partial disruption. Thachepan et al. (2010) have also found that prolonged heating at 60°C at pH 7 for weeks in the case of micellar casein and for days for β-casein/calcium phosphate nanocluster constructs yields mesocrystals of hydroxyapatite and products of dissociated micelles. The behavior of casein micelles in this temperature range merits further scrutiny, particularly as so many processes in the dairy industry are conducted in this range.

Dual-binding model and micellar interactions The ideas outlined earlier in this chapter allow us to schematically describe in Fig. 6.4 how the casein micelle might appear as an interacting species at the various pH values indicated. Internally, at pH 6.7, the micellar matrix is closely interlinked through a combination of nanocluster bridging bonds (the small black circles) and hydrophobic interactions, occurring randomly along any selected polymer chain. The hydrophobic interactions at this pH (indicated as crossover points in the tangled protein network in the diagrams in Fig. 6.4) are many but relatively weak, being counterbalanced by the repulsive energies between the negative charges present on ionized carboxyl groups, dispersed along the chains and throughout the network. The micellar outer reaches are mainly κ-casein molecules, which have terminated polymer extension and limited micellar growth in the dual-binding model. The negative charges from the ionized carboxyls and sialic acid groups on the κ-casein macropeptides provide the electrostatic repulsion component in the intermicellar interaction potential, which inhibits micellar aggregation. Its longer range, illustrated by the thickness of the red shell around the micelle, prevents close approach of the hydrophobic regions buried beneath the shell and amply fulfills the requirements of a hard sphere model colloid at this pH, 6.7. At the lower intermediate pH of 5.6 in this series of illustrations, the same shell continues to prevent close approach of the micelles. The pK values of the acidic groups giving rise to the negative charge are generally lower than 5.5 and have yet to be titrated away. Internally, however, most of the micellar calcium phosphate nanoclusters have been solubilized, and the bridges between phosphoserine cluster motifs have been lost, weakening the overall network structure of the micelle. The bond strengths of hydrophobic interactions remain relatively weak, still being counterbalanced by ionized carboxyl groups dispersed through the micelle. However, the relatively weak bonding allows for rapid interchange and restructuring of the micelle in this range of pH, smoothing out gross structural features apparent in AFM pictures at pH 6.7 (Ouanezar et al., 2012). By pH 5.1, the surface charges are being titrated away, the red shell in the Fig. 6.4 depiction is much thinner, and aggregation begins. Internally, the hydrophobic interactions are effectively being strengthened (indicated by a deepening of the color of the chains) because the counterbalancing electrostatic repulsions are also being removed from the equation, leading to reduced mobility within the micellar particles.

234

6. Casein micelle structure and stability

pH 6.7

pH 5.6

pH 5.1

FIG. 6.4 Representations of casein micelle structures at various pH values, as indicated. The pale chains indicate protein molecules, where they cross being a hydrophobic interaction junction, with the depth of color indicating the intensity of attraction at that pH. The small black circles are the calcium phosphate nanoclusters that are solubilized when the pH is lowered. The outer circle is indicative of the range of steric repulsion generated between micelles and preventing interaction of the surface protein chains.

Concentrated micellar dispersions In milk, as produced from the cow at its natural pH of 6.7 and at temperatures from ambient to blood heat, casein micelles closely follow the behavior of hard sphere colloids (De Kruif, 1998; Alexander et al., 2002). Justification for this assertion comes from studies utilizing light and neutron scattering to measure micelle size and polydispersity (Hansen et al., 1996), from sedimentation behavior (De Kruif, 1998), and from measurements of micellar voluminosity (De Kruif, 1998), diffusivity (De Kruif, 1992), and viscosity of micellar suspensions (Griffin et al., 1989). Paralleling colloidal hard sphere behavior holds only for a limited range of concentrations, and above a critical concentration, micellar suspensions show marked deviations from expected hard sphere behavior (Mezzenga et al., 2005). The viscosity continues to increase

Application of the dual-binding model

235

but at a slower rate than that expected for hard spheres. This is accompanied by a transition from Newtonian viscosity behavior at the natural milk concentration to non-Newtonian viscoelastic behavior in the high-concentration regime. More enlightening demonstrations of the departure from hard sphere behavior come from studies of the rheology of highconcentration micellar suspensions produced by ultrafiltration (Karlsson et al., 2005), by evaporation to 45% total solids (Bienvenue et al., 2003), by centrifugal sedimentation and pelleting (Horne, 1998), and by osmotic compression (Bouchoux et al., 2010). In these instances, concentrated micellar suspensions are close packed and show gel-like behavior, which can be interpreted with the assistance of the dual-binding model. Karlsson et al. (2005) concentrated skim milk by ultrafiltration to produce a micellar suspension with 19.5% casein and studied the effects of pH and ionic strength on its viscoelastic properties. Their suspensions exhibited Newtonian viscosities at very low (Brownian) and very high (hydrodynamic) shear rates, with shear thinning at intermediate shear rates and stresses. The concentration of the micelles by ultrafiltration forced the micelles to interact, jamming them together at this high volume fraction and producing a honeycomb-like structure in freeze-fracture electron micrographs. The elastic modulus of these gels decreased as the pH was lowered from the value achieved in the ultrafiltration retentate. Addition of NaCl at levels of 0.33 and 0.66 mol/kg prior to ultrafiltration increased the elasticity of the gels but shifted their pHs to more acid values. Thereafter in the added-salt systems, lowering the pH produced a decrease in elasticity that paralleled the behavior of the untreated suspension, the higher salt level giving greater elasticity throughout. Karlsson et al. (2005) also measured the phase angle, that is, the partitioning between viscous and elastic components in these gels, as the pH was reduced. In the no-added-salt system, they found that the phase angle increased, indicating a greater mobility within the system, through a maximum close to 45 degrees and thereafter decreased with decreasing pH. In the presence of added salt, a maximum in the phase angle was again observed but was shifted to much lower pH values: in the case of the higher salt level, to a pH value lower than for acid gel formation in milk of normal concentration, and, in both cases, where elasticity had been observed to increase again in these salted concentrated suspensions. The dual-binding model explains this behavior with reference to the schematic of the micellar interaction potential depicted in Fig. 6.5. The increase in micellar concentration in the no-added-salt case forces the micelles together and into the secondary minimum generated by hydrophobic interactions. This is the source of the attractive interaction giving rise to the viscoelasticity observed. The micelles are also in a jammed structure, and their internal bonding contributes to the measured elasticity. On lowering the pH, the loss of the calcium phosphate nanocluster bridges weakens this structure, and the elasticity decreases, as observed. The bonding caused by hydrophobic interactions is relatively weak, and the loss of the nanocluster bridges further contributes to the mobility in the gel, as evidenced by the observed increase in phase angle. Decreasing the pH further titrates away carboxyl groups; however, it reduces the counterbalancing electrostatic component and thereby strengthens hydrophobic bonds in the matrix, reducing mobility and producing the subsequent decrease in the phase angle. The major effect of the addition of salt is to reduce the Debye-H€ uckel parameter and shorten the range of the electrostatic repulsion between micelles. This makes it easier to enter the secondary minimum in the interaction potential and increases the gel elasticity, as

236

6. Casein micelle structure and stability

Pot. energy

Separation

FIG. 6.5 Repulsive intermicellar interaction potential with inner hydrophobic interaction minimum. The dashed line shows the effect of salt addition on the range of the electrostatic repulsion component.

observed, with more salt producing the stronger gel. Again, however, the calcium phosphate nanocluster bridges contribute stress-carrying bonds, and their removal by lowering the pH leads to the observed decrease in the elasticity of the gel. Throughout this titration, the bonds in the system are relatively stronger than in the no-salt case—the salt also contributes to decreasing the effectiveness of intramicellar electrostatic repulsion—and the phase angles are lower in comparison. Karlsson et al. (2005) suggested that a significant effect of the salt addition is to exchange bound calcium within the micelle for monovalent ions, which would imply that no nanoclusters in the system would be solubilized on decreasing the pH and thereby would void the earlier explanation for the decrease in elasticity with pH. Huppertz and Fox (2006) did indeed find increased levels of calcium in serum when 600-mM NaCl was added to a two-times-concentrated milk, but they found no increase in serum inorganic phosphate, suggesting that the increase in calcium came from the displacement of casein-bound calcium rather than a salt-induced dissociation of the calcium phosphate nanoclusters, leaving these to be solubilized on acidification. The evaporated milks produced by Bienvenue et al. (2003) had 45% total solids or were approximately concentrated from normal by a factor of four, rather than the eight times concentration of the micelles in the ultrafiltration retentates of Karlsson et al. (2005). The milks of Bienvenue et al. (2003) increased in viscosity on storage at 50°C, with salt addition accelerating the increase. Such behavior is in line with the predictions of the dual-binding model. The collision rate increased by concentration will be further increased by raising the temperature, bringing about a higher frequency of micelles attempting to enter the secondary minimum. A higher success rate and flocculation because of more thermal energy will give rise to the observed increase in viscosity. The weak flocs can be disrupted by higher shear stresses, giving the observed shear-thinning behavior. The effect of salt would be to render it easier to enter the secondary minimum and promote the flocculation reaction.

Application of the dual-binding model

237

In another study of the rheological behavior of concentrated micellar systems, casein micelle pellets were produced by the centrifugation of skim milk at 19,000 g for 60 min, giving protein concentrations of approximately 20% (Horne, 1998). At high temperatures (40°C), this pellet flowed freely. Its viscosity was Newtonian, independent of shear rate or frequency. At low temperature (5°C), however, this micellar suspension exhibited all the properties of a classical viscoelastic gel, with elastic moduli independent of frequency and phase angles of less than 45°. At intermediate temperatures, there was a crossover between viscous and elastic behavior. The behavior here is dominated by that of the hydrophobic interactions. At low temperatures, the strength of these interactions is low. Both β-casein and κ-casein are known to depart from the micelle under such conditions (Dalgleish and Law, 1988); however, in the close-packed conditions prevailing in the pellet, they are liable to migrate or link to neighboring micelles or to become entangled with proteins loosened from those micelles, leading to the gel-like behavior. As the temperature is increased, the strength of the hydrophobic interaction increases, but the ability to break bonds is also enhanced, and more mobility is allowed. The strengthening of the bonding may also lead to a tightening up of the micelles, and their becoming more compact may allow the suspension to flow more freely. Finally, Bouchoux et al. (2010) combined the osmotic stress technique with SAXS to study the structural response of the casein micelle to an increase in concentration. Their SAXS results indicate that, as the micelles are compressed, they lose water and shrink to a smaller volume, but this compression is nonaffine, that is, some soft parts of the micelle lose water and readily collapse, whereas other hard parts resist deformation and are pushed closer together. Bouchoux et al. (2010) argued that existing models of the casein micelle fail to reproduce this behavior and suggested a physical model based on hard regions that assume the structure of Voronoi tessellations but without providing any mechanism for the creation of this structure. Moreover, the structure that they believed corresponded to the behavior deduced from their SAXS spectra, with hard regions containing the nanoclusters and major water-filled channels and cavities, is nothing more than the structure deduced by Trejo et al. (2011) from their tomographic analysis of their cryo-TEM pictures. As we have seen earlier in this chapter, such a structure can be developed within the dual-binding model provided due consideration is given to the kinetics of protein aggregation and micellar assembly. The apparently harder regions of the micelle might also contain the “clumps” of nanoclusters or be enriched in those and their cross-linking peptides between the casein phosphoserine clusters.

Dual-binding model and micellar destabilization The concept of the casein micelle being electrosterically stabilized by a “hairy layer” coat of κ-casein appears to enjoy universal acceptance (Holt, 1975; Walstra, 1979; Holt and Horne, 1996). Because the dual-binding model of the casein micelle naturally provides a surface location for κ-casein in a growth-limiting role, it readily explains the destabilization of the casein micelle system on the proteolysis of κ-casein by chymosin and the loss of the sterically stabilizing hairs. Such proteolysis also leads to a significant drop in the micellar zeta potential (Dalgleish, 1984) and a consequent reduction in the electrostatic repulsion between micelles.

238

6. Casein micelle structure and stability

Further confirmation of the importance of electrostatic repulsion in intermicellar interactions is evinced by the necessary presence of ionic calcium to bring about/promote the aggregation of the chymosin-treated micelles. Notwithstanding the importance of electrostatics, hydrophobic interactions also play an important part, as evidenced by the fact that fully renneted micelles show no signs of aggregation at low temperatures ( 500

161,000 (G1)b

Many isoforms

Lactoferrin (Lf )

< 0.1

689

76,110

8.95

a

Passive transfer of immunities 17

Bacteriostatic role; glycoprotein

a

Molar mass for the A variant. G1 is the major Ig; two other classes, IgM and IgA, are present in much lower abundance. (Based on Farrell, H.M., Jr., Jimenez-Flores, R., Bleck, G.T., Brown, E.M., Butler, J.E., Creamer, L.K., Hicks, C.L., Hollar, C.M., Ng-Kwai-Hang, K.F., Swaisgood, H.E., 2004. Nomenclature of the proteins of cows’ milk—sixth revision. J. Dairy Sci. 87 (6), 1641–1674.) b

Bovine β-lactoglobulin β-Lactoglobulin (β-Lg) contains 162 amino acids and has a molecular weight of 18.3 kDa (Hambling et al., 1992). It is a member of the lipocalin (a contraction of the Greek lipos, “fat, grease,” and calyx, “cup”) family of proteins (Banaszak et al., 1994; Flower, 1996), so called because of their ability to bind small hydrophobic molecules into a hydrophobic cavity. This led to the proposal that β-Lg functions as a transport protein for retinoid species, such as vitamin A (Papiz et al., 1986). β-Lg is the most abundant whey protein in the milk of most mammals (about 10% of the total protein or about 50% of the whey protein) but has not been detected in the milk of humans, rodents, or lagomorphs. In the case of human milk, α-La (see later) is the dominant whey protein. Bovine β-Lg is by far the most commonly studied milk protein. There are ten known genetic variants of bovine β-Lg. The most abundant variants are labeled β-Lg A and β-Lg B (Farrell et al., 2004) and differ by two amino acid substitutions, Asp64Gly and Val118Ala, respectively. The quaternary structure of the protein varies among monomers, dimers, or oligomers depending on the pH, temperature, and ionic strength, with the dimer being the prevalent form under physiological conditions (Kumosinski and Timasheff, 1966; McKenzie and Sawyer, 1967; Gottschalk et al., 2003). This variable state of association is likely to be the result of a delicate balance among hydrophobic, electrostatic, and hydrogen-bond interactions (Sakurai et al., 2001; Sakurai and Goto, 2002).

Molecular structure of bovine β-Lg β-Lg was an early target of x-ray diffraction, as newly applied at the Royal Institution to protein crystals. This was due to its high abundance and relatively easy purification from

Bovine β-lactoglobulin

253

milk and to its propensity to form suitable crystals. In retrospect, this was a very ambitious project because β-Lg was not the easiest protein to analyze (Green et al., 1979), partly because of its multiple crystal forms. Nevertheless, this study established that the protein monomer was nearly spherical with a block of electron density with a rodlike structure across one face. The next attempt (Creamer et al., 1983) to determine the structure was by calculation using sequence data and structural probabilities to estimate which portions of the amino acid sequence might form into helices, strands, and sheets. The secondary structure of β-Lg was predicted to comprise
15% α-helix,
50% β-sheet, and 15%–20% reverse turns (Creamer et al., 1983). It is interesting to note that many of the residues that reside in the extended structures of the native protein have been shown to have a nascent propensity to form α-helical structures in the presence of trifluoroethanol or amphiphiles (Hamada et al., 1995; Kuroda et al., 1996; Chamani et al., 2006). In 1986, the first medium resolution structure of β-Lg was published (Papiz et al., 1986). Structural similarity to a seemingly different type of protein, plasma retinol-binding protein, has given rise to much speculation as to the role of β-Lg in bovine milk. Higherresolution structures subsequently revealed the now familiar eight-stranded β-barrel (calyx), flanked by a three-turn α-helix. A final ninth strand forms the greater part of the dimer interface at neutral pH (Papiz et al., 1986; Bewley et al., 1997; Brownlow et al., 1997). The β-barrel is formed by two β-sheets, where strands A–D form one sheet and strands E–H form the other (with some participation from the A strand, facilitated by a 90 degrees bend at Ser21). Two disulfide bonds link Cys66 on loop CD (which, as its name suggests, connects strands C and D) with Cys160 near the C-terminus and Cys106 on strand G with Cys119 on strand H, leaving Cys121 as a free, but largely unexposed, thiol. The loops connecting strands BC, DE, and FG are relatively short, whereas those at the open end of the barrel, AB, CD, EF, and GH, are longer and more flexible. These features are illustrated in Fig. 7.1. The structures of the A and B variants are very similar. However, the Asp64Gly substitution results in the CD loop adopting differing conformations (Qin et al., 1999). The Val118Ala substitution causes no detectable change to the structures, but the void created by substituting the bulky isopropyl substituent with the smaller methyl group results in the hydrophobic core of the B variant being less well packed and may account for its lower thermal stability under some measurement conditions (Qin et al., 1999). Very careful titrimetric and thermodynamic measurements in the late 1950s (Tanford and Nozaki, 1959; Tanford et al., 1959) established the presence of a carboxylic acid residue with an anomalously high pKa value of 7.3. This was attributed to a pH-dependent conformational change, a conclusion that rationalized earlier measurements of pH-dependent sedimentation coefficients (Pedersen, 1936) and specific optical rotation data (Groves et al., 1951). Much later, x-ray structure analyses (Qin et al., 1998b) at pH values above and below this so-named Tanford transition established that, at pH 6.2, the EF loop is closed over the top of the barrel, burying Glu89 (the carboxylic acid with the anomalous pKa) inside the calyx. At pH 8.1, this loop is articulated away from the barrel such that the formerly buried glutamic acid becomes exposed in the carboxylate form (Qin et al., 1998b). An early structure of bovine β-Lg crystallized in the presence of retinol appeared to show retinol bound externally to the protein (Monaco et al., 1987); this was apparently later confirmed by a body of fluorescence data (Dufour et al., 1994; Lange et al., 1998;

254

7. Structure and stability of whey proteins

FIG. 7.1 Diagram of the dimeric structure of bovine β-Lg A looking down the twofold axis. The coordinates are taken from the structure of β-Lg A in the trigonal Z lattice with 12-bromododecanoic acid bound (PDB code/1bso). The strands that form the β-barrel are labeled A–H. The I strand, together with part of the AB loop, forms the dimer interface at neutral pH. The locations of the sites of difference between the A and B variants are also shown. The structure is rainbow colored, beginning with blue at the N-terminus and ending with red at the C-terminus. Ser21, which shows conformational flexibility, and the 12-bromododecanoate anion are shown as spheres. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this chapter.) Figure drawn with PyMOL. Delano, W.L., 2002. PyMOL. Delano Scientific, Palo Alto, CA.

Narayan and Berliner, 1998). However, subsequent structural analyses have shown that fatty acids, retinoid species (including vitamin A), and cholesterol (including vitamin D) all bind inside the calyx (Kontopidis et al., 2004). Induced circular dichroism (CD) measurements and NMR measurements confirm the x-ray crystallographic observations. Ligand binding is discussed in more detail later, because it relates to the probable physiological function of β-Lg and to the stability of this molecule and the current technological interest in the role of protein–ligand interactions in flavor perception (see Chapter 14). At this stage, there is no evidence for the binding of fatty acids or retinoid species outside the calyx. Except for very bulky ligands (see later), ligands bind inside the calyx of β-Lg at pH
7. At about the same time as the Tanford transition and the ligand-binding modes were elucidated by high resolution x-ray crystallography (Qin et al., 1998b,b), NMR studies of the structure of β-Lg in low pH (
2.5), very low ionic strength (0.1)

Fast (>0.1)

>25,000

>40,000

11  3

92

Fast (>0.1)

Fast (>0.1)

>9200

>12,000

pH 2.5

e

pH 7.5 a b c d e f

f

f

d

Calculated from global fitting of sedimentation velocity (SV) and sedimentation equilibrium (SE) data. Calculated from global fitting of SV data (SE data contain no kinetic signal). Calculated error ranges represent the sensitivity of the values to changes in other fitting parameters. Indicative value, as no error range could be determined; kon not calculated. Citrate buffer. 3-(N-morpholino)propanesulfonic acid buffer.

constants KD are about 10 μM. At a pH around the isoelectric point, KD decreases markedly and falls outside the range of standard AUC detection. The dissociation rate constant is quite slow (
0.008 s1), and the association rate constant is well removed from the diffusioncontrolled limit. This was attributed (Mercadante et al., 2012) to the considerable restructuring of the cloud of counterions on dimerization: at low pH, where β-Lg is strongly positively charged, there is substantial restructuring of counteranions; at pH > 7, where β-Lg is mildly negatively charged, there is restructuring of countercations and sharp distance dependence for the optimization of hydrophobic contacts. Bello et al. (2011) also observed by isothermal dilution calorimetry that the dimer dissociation constant of variant B of β-Lg is smaller than that of variant A at temperatures up to 35°C, consistent with AUC data (Mercadante et al., 2012). Indeed, the dissociation constant of 14.5 (1) μM measured at pH 7.0 in 50-mM phosphate buffer and 100-mM NaCl is very similar to that determined by AUC measurements (Table 7.2).

Studies of bovine β-Lg by NMR at neutral pH The large size of the bovine β-Lg dimer at pH 7 is expected to cause some broadening of the peaks in its 1H NMR spectrum because of slower molecular reorientation. However, this problem is exacerbated by chemical exchange broadening of peaks in the vicinity of the dimer interface by the dynamic equilibrium of molecules between the associated (dimeric) and unassociated (monomeric) states. These factors render the resulting spectra unsuitable for structure determination. Several methods have been employed to allow NMR studies at neutral pH. The most straightforward of these has been to use a nonruminant β-Lg that is intrinsically monomeric, yet with the same overall tertiary structure as the bovine protein, in this case equine β-Lg (Kobayashi et al., 2000). Alternatively, the dimer interface may be disrupted by producing bovine β-Lg mutants with amino acid substitutions carefully chosen to disrupt the intermolecular interactions between either the I strands or the AB loops (Sakurai and Goto, 2002) (see Fig. 7.1).

Bovine β-lactoglobulin

257

It is worth noting that an attempt to form dimeric equine β-Lg by producing a mutant with amino acid substitutions chosen to mimic those of the bovine protein at the interface was not successful (Kobayashi et al., 2002), indicating that subtle features in the β-Lg conformation remote from the interface have an impact on successful dimer formation. Indeed, reaction of the free thiol of Cys121 (located away from the interface in the H strand and covered by the main α-helix, Asp129-Lys141, see Fig. 7.2) with 2-nitro-5-thiobenzoic acid produces a monomeric species with the native structure at pH 2 and a monomeric but unfolded structure at

FIG. 7.2 Ligand-binding sites on β-Lg as inferred from NMR measurements of the binding of small (5.4 A the putative vitamin D3. Arg148 is noted as buried but is buried only by the putative vitamin D3; otherwise, it is well exposed to solvent. In brief, there remains no unequivocal crystallographic evidence for the binding of (largely) hydrophobic molecules to an external site on bovine β-Lg. Although ligands such as palmitic acid appear to be released at acidic pH (Ragona et al., 2000), NMR evidence (based on perturbations of backbone chemical shifts) for the binding of the flavor compounds γ-decalactone and β-ionone at pH 2 has been reported (Luebke et al., 2002; Tromelin and Guichard, 2006). Thus, there is evidence for three binding sites to β-Lg: the canonical site inside the calyx; a second site involving the perturbation of residues Trp19, Tyr20, Tyr42, Glu44, Gln59, Gln68, Leu156, Glu157, Glu158, and His161; and a third site involving the perturbation of residues Tyr102, Leu104, and Asp129. These regions are illustrated in Fig. 7.2. Initially, it was thought that porcine β-Lg did not bind fatty acids (Frapin et al., 1993). However, NMR studies have shown that the pH for 50% uptake of ligand shifted by nearly 4 pH units from
5.8 for bovine β-Lg to 9.7 for porcine β-Lg, whereupon the EF loop underwent a structural change analogous to that of its bovine counterpart (Ragona et al., 2003). Indeed, this loop with its absolutely key Glu residue (Glu89 in bovine β-Lg) was predicted to be pH gated for all β-Lgs (Ragona et al., 2003).

Effect of temperature on bovine β-Lg The thermal properties of β-Lg variants are of considerable commercial relevance because of their role in the fouling of processing equipment and the functional qualities that can be imparted to dairy products by thermally induced β-Lg aggregation. Consequently, this aspect of the protein’s behavior has been the focus of extensive experimental work. At neutral pH, the midpoint of the thermal unfolding transitions, as determined by differential scanning calorimetry (DSC), is
70°C (de Wit and Swinkels, 1980), whereupon the protein dimer dissociates and the constituent molecules begin to unfold. This reveals the free thiol of Cys121 (located at the C-terminal end of the H strand—see Fig. 7.1) and a patch of hydrophobic residues, leading to the possibility of both covalent and hydrophobic intermolecular association (Qi et al., 1995; Iametti et al., 1996). The ensuing disulfide interchange reactions lead to the formation of a variety of mixed disulfide-bonded polymeric species (Creamer et al., 2004). Genetically engineered mutants with an extra cysteine positioned to allow a third disulfide bond to be formed to Cys121 have been shown to both retard thermal denaturation by 8–10°C and to resist heat-induced aggregation (Cho et al., 1994). In mixtures of bovine β-Lg, α-La, and BSA or of β-Lg and one or other of α-La and BSA, at pH 6.8 and

Bovine β-lactoglobulin

265

subjected to high temperatures, homo- and heteropolymeric disulfide-bridged species were observed (Havea et al., 2001). The formation of α-La-α-La disulfide links (α-La has no free cysteine; see later) was attributed to catalysis by BSA or β-Lg (Havea et al., 2001). At low pH, where the protein is monomeric, denaturation is largely reversible at temperatures below 70°C (Pace and Tanford, 1968; Alexander and Pace, 1971; Mills, 1976; Edwards et al., 2002). Heating above this temperature leads to the formation of large aggregates, but, in contrast to the behavior at neutral pH, the species are predominantly noncovalently bonded (Schokker et al., 2000). The precise denaturation process is complex and is influenced by factors such as pH, protein concentration, ionic environment, genetic variant, and the presence of ligands. Both lowering of the pH (Kella and Kinsella, 1988; Relkin et al., 1992) and the addition of calyx-bound ligands (Puyol et al., 1994; Considine et al., 2005a; Busti et al., 2006) make the protein more resistant to thermal unfolding. The stability of the genetic variants (at pH 6.7) appears to decrease in the order C > A > B, with the A variant showing the least cooperative unfolding transition (Manderson et al., 1997). The protein’s susceptibility to thermal denaturation at pH 6.7–8 is strongly concentration dependent up to about 6 mM, being most susceptible to unfolding at a concentration of
1.4 mM (Qi et al., 1995). It is possible that, at high protein concentration (
6 mM), tertiary structure is lost directly from the native dimer state (Qi et al., 1995). There is some evidence that the thermal unfolding occurs in more than one step. Kaminogawa et al. (1989) used antibody-binding affinities to propose that thermal unfolding of variant A of β-Lg occurs in at least two stages, starting with conformational changes near the N-terminus followed by changes in the region of the three-turn α-helix. Fourier transform infrared (FTIR) measurements by Casal et al. (1988) have also indicated a loss of helical content early in the denaturation process (using variant B of β-Lg in 50-mM phosphate buffer at pH 7). Qi et al. (1997) used FTIR and CD measurements to propose that variant A of β-Lg forms a molten globule with reduced β structure when heated above 65°C in 30–60 mM NaCl at pH 6.5. NMR studies, observing hydrogen/deuterium (H–D) exchange of the backbone amide protons of β-Lg A at pH 2–3, have revealed a stable core comprising the FG and H strands, possibly stabilized by the Cys106-Cys119 disulfide bond between strands F and G (Belloque and Smith, 1998; Edwards et al., 2002). Significant secondary structure, even at a temperature as high as 90°C, has been reported (Casal et al., 1988; Qi et al., 1997; Bhattacharjee et al., 2005).

Effect of pressure on bovine β-Lg High-pressure treatment of food is of increasing commercial importance because of increasing consumer demand for products that have been subjected to minimal processing damage. Pressure treatment as part of the processing regime has the potential to produce dairy products with improved functional and organoleptic properties compared with those produced by thermal treatment alone (Messens et al., 2003). Of the major whey proteins, β-Lg is the most susceptible to pressure-induced change (Stapelfeldt et al., 1996; Patel et al., 2005). Presumably, this is due to its relatively inefficient packing, which is caused by the presence of the β-barrel with its large solvent-exposed hydrophobic pocket and the lower number of disulfide bonds (two compared with four in, e.g., the

266

7. Structure and stability of whey proteins

similar-sized α-La). A reduction in the molar volume of bovine β-Lg has been detected at pressures as low as 10 MPa, possibly because of a contraction of the calyx (Vant et al., 2002). A number of studies have shown that β-Lg becomes more susceptible to enzymatic cleavage when exposed to pressure, possibly because of pressure-induced conformational change. The free cysteine has been shown to become exposed at between 50 and 100 MPa (Stapelfeldt et al., 1996; Tanaka and Kunugi, 1996; Møller et al., 1998). Exposure of the protein to pressures in excess of
300 MPa causes irreversible changes to β-Lg’s tertiary and quaternary structure. A combination of CD and fluorescence spectroscopy of β-Lg at neutral pH and exposed to pressures as high as 900 MPa indicated that pressure induces monomer formation with subsequent aggregation but with only small irreversible effects on the β-Lg tertiary structure (Iametti et al., 1997). However, more recent results from tryptic hydrolysis suggest that, whereas exposure to pressures below 150 MPa has no detectable permanent effect on the conformation of β-Lg A, pressures above 300 MPa lead to the detachment of strands D and G from the β-barrel together with the formation of disulfide-bonded oligomers (Knudsen et al., 2002). In mixtures of bovine β-Lg with either α-La or BSA at pH 6.6 that are subjected to high pressures, intermolecular disulfide-bridged aggregates form only between β-Lg and itself. No β-Lg-α-La or β-Lg-BSA disulfide-bridged species are detected (Patel et al., 2005), in contrast to heat-treated mixtures, in which such species are observed (Havea et al., 2001). To correlate the pressure-induced conformational changes with the protein’s primary sequence, Belloque et al. (2000) made NMR amide H–D exchange observations of β-Lg A and B following exposure of solutions at neutral pH to pressures of up to 400 MPa. Little change in H–D exchange was reported at 100 MPa, which indicates that any conformational change occurring does not increase the exposure of most amide protons to the solvent compared with their exposure in the native conformation at ambient pressure. A large increase in the extent of H–D substitution at 200 MPa and above indicated increased conformational flexibility, but the similarity of the spectra of control samples recorded in H2O rather than D2O before and after pressurization demonstrated that any pressure-induced conformational changes were largely reversible up to 400 MPa. The authors proposed that the structure of the A variant was more sensitive to changes in pressure than that of the B variant and that the F, G, and H strands of the protein’s β-barrel were the most resistant to conformational change, the latter conclusion paralleling the effects of temperature (Belloque and Smith, 1998; Edwards et al., 2002). FTIR and SAXS experiments suggest that, even at 1 GPa, the unfolded state contains significant secondary structure (Panick et al., 1999). Combined application of pressure and heat has shown that changing the temperature over the range 5–37°C has negligible effect on the susceptibility of β-Lg to pressures up to 200 MPa (Skibsted et al., 2007). However, combined application of pressure and moderate temperature at 600 MPa/50°C (Yang et al., 2001) and 294 MPa/62°C (Aouzelleg et al., 2004) has indicated the formation of a molten globule with an α-helical structure, on the basis of results obtained from CD spectroscopy. Therefore, it should be noted that the potential for temperature increases induced by rapid pressurization of the sample needs to be considered when studying the effects of pressure on protein conformation and stability. Enzymatic proteolysis observations indicate that β-Lg is less susceptible to pressure-induced change at acidic pH (Dufour et al., 1995) than at neutral or basic pH, but this result may be confounded by pressure-induced changes in the proteases, thermolysin, and pepsin. Nevertheless, NMR measurements of monomeric β-Lg at pH 2 while under pressure at up to 200 MPa have shown

Bovine β-lactoglobulin

267

that the two β-sheets unfold independently to form two intermediates to an unfolded state that still appears to contain significant secondary structure (Kuwata et al., 2001). A three-step mechanism has been proposed for the denaturation of β-Lg at neutral pH and ambient temperature, which broadly encompasses the above observations. A pressure of 50 MPa causes partial collapse of the calyx (with concomitant reduction in ligand-binding capacity) together with exposure of Cys121. Increasing the pressure to 200 MPa causes further (partially reversible) disruption to the hydrophobic structure, together with a decrease in the molecular volume. Higher pressures cause irreversible aggregation reactions involving disulfide interchange reactions (Stapelfeldt and Skibsted, 1999; Considine et al., 2005b). The very recent crystal structure determinations of bovine β-Lg at high pressures now provide a more precise structural underpinning to these studies, notwithstanding that a crystal is a highly crowded environment compared with these studies on more dilute solutions. The ˚ resolution and at 430 MPa (Kurpiewska structures of bovine β-Lg with dodecane at 2.85 A ˚ et al., 2018) and with myristic acid at 2.50 A resolution (an alarmingly high Rfree of 0.397) and at 550 MPa (Kurpiewska et al., 2019) reveal significant differences compared with the corresponding structures at ambient temperatures. For dodecane-bound β-Lg, shrinkage of the binding pocket, weakening of the dimer interface, and partial unraveling of secondary structure elements are observed. However, in these and, it appears, also the solution-state studies cited earlier, the pH was not corrected for the significant changes in pH that occur at high pressures (>1 pH unit for citrate buffer between 0.1 and 500 MPa and nearly double this for phosphate buffer [El’yanov and Hamann, 1975; Kitamura and Itoh, 1987]); thus, the effects observed may be due, in part at least, to changes in pH.

Effect of chemical denaturants on bovine β-Lg Chemical denaturants are often used to unfold proteins and to characterize mechanisms and transition states of protein-folding processes. Commonly used denaturants include alcohols, particularly 2,2,2-trifluoroethanol (TFE), urea, and guanidinium chloride (GdmCl). Theoretical calculations predict a significantly higher amount of α-helical secondary structure than is actually observed in native β-Lg (Creamer et al., 1983; Nishikawa and Noguchi, 1991). That is, the native structure is the result of competition between α-helix favoring local interactions and β-sheet forming long range interactions. However, the addition of alcohols such as TFE can disturb this balance by weakening the hydrophobic interactions and strengthening the helical propensity of the peptide chain (Thomas and Dill, 1993). The ability to increase the α-helical content of bovine β-Lg by the addition of alcohols (ethanol, 1-propanol, and 2-chloroethanol) was first demonstrated by Tanford using optical rotary dispersion measurements (Tanford et al., 1960). Contemporary studies tend to favor the use of TFE, where the β-Lg β-sheet-to-α-helix transition has been shown to be highly cooperative, occurring over the range
15%–20% v/v of cosolvent (Shiraki et al., 1995; Hamada and Goto, 1997; Kuwata et al., 1998). The higher proportion of α-helical structure in the so-called TFE state is found in the N-terminal half of the molecule (Kuwata et al., 1998). Magnetic relaxation dispersion measurements of the solvent nuclei have shown that this state is an open, solventpermeated structure (unlike the collapsed state of a molten globule) and that its formation is accompanied by a progressive swelling of the protein with increasing TFE concentration (Kumar et al., 2003). High protein and TFE concentrations (8% v/w and 50% v/v, respectively)

268

7. Structure and stability of whey proteins

can lead to fibrillar aggregation (see later) and gel formation of bovine β-Lg at both acidic and neutral pH (Gosal et al., 2002). The ability of urea to induce protein unfolding is thought to be via a combination of hydrogen-bond formation with the protein backbone and a reduction in the magnitude of the hydrophobic effect (Bennion and Daggett, 2003). Therefore, in contrast to TFE, both the helical propensity and the hydrophobic effect are reduced. Urea-induced unfolding of bovine β-Lg at acidic pH was first reported as a two-state process (Pace and Tanford, 1968). Subsequent NMR H–D exchange measurements of bovine β-Lg B at pH 2.1 also allowed the ureainduced unfolding to be well approximated as a two-state transition between the folded protein and the unfolded state via a cooperative unfolding of the β-barrel and the C-terminus of the major α-helix (Ragona et al., 1999). However, Dar et al. (2007) have provided evidence that urea also causes unfolding via an intermediate, albeit with structural properties between those of the native and unfolded states. The addition of anionic amphiphiles, that is, sodium dodecyl sulfate or palmitate, causes β-Lg to resist urea-induced unfolding because of binding inside the calyx (Creamer, 1995). GdmCl is often used as an alternative to urea in studies of protein stability. At the neutral or acidic pH of most stability studies, GdmCl will be fully dissociated. At low GdmCl concentration (below about 1 M), chloride ions screen the electrostatic repulsion between positively charged groups of the protein (Hagihara et al., 1993). The result is that the additional electrostatic interactions of GdmCl compared with the neutral urea molecule have the potential to both stabilize and destabilize the protein structure depending on the concentration of GdmCl (Hagihara et al., 1993). D’Alfonso et al. (2002) have compared the denaturations of bovine β-Lg B with both GdmCl and urea between pH 2 and pH 8, as monitored by CD, UV differential absorption, and fluorescence measurements. Discrepancies between the unfolding free energies obtained using the two denaturants could be reconciled if GdmCl denaturation was assumed to occur via an intermediate state. The secondary structure of this state is similar to that of the native protein, but with greater rigidity in the vicinity of the Trp residues, consistent with the screening of electrostatic repulsion between charged residues (D’Alfonso et al., 2002). The GdmCl-induced unfolding intermediate of bovine β-Lg A at pH 2 has been reported to have increased α-helical structure (Dar et al., 2007). Porcine β-Lg has also been shown to unfold via an intermediate state on the addition of GdmCl. The stability of the porcine protein was lower than that of its bovine counterpart, and the intermediate state was richer in α-helical structure. Most of the hydrophobichydrophobic interactions of the buried core of the native state are conserved between bovine β-Lg and porcine β-Lg. However, four pairwise interactions of the Phe105 side chain of bovine β-Lg are lost on the change to Leu in the porcine protein. This indicates that the presence of the aromatic residue may play an important role in the increased stability of the bovine protein (D’Alfonso et al., 2004). It is interesting to note that this residue is particularly resistant to H–D exchange in heated β-Lg solutions (Edwards et al., 2002).

Fibrillar formation from bovine β-Lg At high temperatures (>80°C), low pH (typically pH
2), and low ionic strength (less than about 20 mM), β-Lg self-assembles into fibrils (Aymard et al., 1999; Gosal et al., 2002; Arnaudov et al., 2003). These structures have some characteristics of classic β-amyloid fibrils,

α-Lactalbumin

269

such as binding the fluorophores thioflavin-T and Congo red and showing x-ray diffraction consistent with a cross-β-sheet structure (Bromley et al., 2005). However, they differ substantially in that they are easily dissociated (Akkermans et al., 2008; Jordens et al., 2011). The fibrils have contour lengths of the order of microns (Aymard et al., 1999) and diameters in the range 4–8 nm (Gosal et al., 2004). Results from the combined use of atomic force microscopy, polyacrylamide gel electrophoresis, and in situ FTIR suggest that the formation of heat-induced β-Lg fibrils involves the partial unfolding of β-Lg, self-hydrolysis, and self-assembly of some of the resulting species (Oboroceanu et al., 2010). It has been proposed that each fibril typically comprises several protofilaments that grow to a length of approximately 0.5–1 μm before aligning and twisting together (Adamcik et al., 2010; Bolisetty et al., 2011; Adamcik and Mezzenga, 2012). The hydrolysis of β-Lg appears to play an important role in fibril formation (Akkermans et al., 2008), but it is also likely to be responsible for their disintegration if the temperature is raised to 120° C for extended periods (Loveday et al., 2012). Fibrils have been shown to contain peptide fragments in the range 2–8 kDa, with sequences weighted toward β-Lg’s termini (Akkermans et al., 2008) and with those toward the N-terminus being the more abundant (Dave et al., 2013). Two-dimensional gel electrophoresis indicates the presence of disulfide-linked fragments, possibly involving the Cys66-Cys160 linkage found in the native protein (Dave et al., 2013). Irradiation with microwaves followed by sample storage has been shown to be a particularly efficient method of fibril formation and, notably, the resulting fibrils contain not only peptides but also sequences that are consistent with intact β-Lg monomers (Hettiarachchi et al., 2012). Although fibril formation has yet to be utilized in food products, there is potential for their application as functional ingredients to control such properties as viscosity, gelation propensity, and emulsion and foam stability (Kroes-Nijboer et al., 2012; Nicolai and Durand, 2013).

α-Lactalbumin α-Lactalbumin (α-La) is a 123-amino-acid, 14.2-kDa globular protein that is found in the milk of all mammals. The bovine protein binds calcium ions, with the holo form being the more abundant form in milk. Within the Golgi apparatus of the mammary epithelial cell, α-La is the regulatory component of the lactose synthase complex (in which it combines with N-acetyl galactosamine synthase, now named β-1,4-galactosyltransferase-I), the role of which is to transfer galactose from UDP-galactose to glucose (Brew, 2003). The structure of human α-La in a 1:1 complex with β-1,4-galactosyltransferase-I has been determined (Ramakrishnan and Qasba, 2001; Ramakrishnan et al., 2001, 2006), and a 2:1 structure has recently become available (Ramakrishnan and Qasba, 2013). In the absence of α-La and in the presence of a transition metal ion such as manganese(II), the catalytic domain of bovine β-1,4-galactosyltransferase-I (residues 130–402) transfers galactose (Gal) to N-acetylglucosamine (GlcNAc), which may be either free or linked to an oligosaccharide, generating a disaccharide unit, Gal-β-1,4-GlcNAc (N-acetyllactosamine). The calcium-ion binding site is remote to the active site of the α-La-β-1,4-galactosyltransferase-I complex (Brew, 2003). α-La has been studied extensively, largely because of its formation of a molten globule state under mild denaturing conditions (Dolgikh et al., 1981).

270

7. Structure and stability of whey proteins

Molecular structure of bovine α-La The tertiary structure of bovine α-La is typical of that of the protein from other mammalian species (Acharya et al., 1991; Calderone et al., 1996; Pike et al., 1996) and is similar to that of lysozyme, with which it shares significant homology. As illustrated in Fig. 7.3A, α-La is made up of two lobes: the α-lobe contains residues 1–34 and 86–123 and the smaller β-lobe spans residues 35–85. The α-lobe contains three α-helices (residues 5–11, 23–34, and 86–98) and two short 310-helices (residues 18–20 and 115–118). A small, three-stranded β-sheet (residues 41–44, 47–50, and 55–56) and a short 310-helix (residues 77–80) make up the β-lobe (Calderone et al., 1996; Pike et al., 1996). The structure is stabilized by four disulfide bonds (Cys6-Cys120 and Cys28-Cys111 in the α-lobe, Cys60-Cys77 in the β-sheet, and Cys73-Cys90 tethering the two lobes together) (Brew, 2003). Unlike β-Lg, α-La has no free thiol. A calcium ion binds with a submicromolar dissociation constant at the so-called “binding elbow” formed by residues 79–88 located in a cleft between the two lobes, with the metal ion coordinated in a distorted pentagonal bipyramidal configuration by the side-chain carboxylate groups of Asp82, Asp87, and Asp88, the carbonyl oxygens of Lys79 and Asp84, and the oxygen atoms of two water molecules (Calderone et al., 1996; Pike et al., 1996). The structure of the apo form of bovine α-La is similar to that of the holo protein. The largest changes involve the movement of the Tyr103 side chain in the interlobe cleft with little change in the vicinity of the calcium-ion binding site (Chrysina et al., 2000). The salient features of the structure of the holo protein are depicted in Fig. 7.3, together with its complex

FIG. 7.3 (A) Structure of bovine α-La showing the calcium-ion binding site (PDB code: 1f6s). The peptide chain is rainbow colored, beginning at the N-terminus in blue and progressing to the C-terminus in red, to show the assembly of the subdomains. The calcium ion is seven coordinate. Loop 79–84 provides three ligands, two from main-chain carbonyl oxygen atoms of Lys79 and Asp84 and one from the side chain of Asp82. Coordination about the calcium ion is completed by carboxylato oxygen atoms from Asp87 and Asp88 at the N-terminal end of the main four-turn helix and by two water molecules. The four disulfide bonds are shown in ball-and-stick representation (one in the helical domain is obscured, and the two linking the helical domain and calcium-ion binding loop to the β domain are on the left half of the panel). (B) The lactose synthase complex formed from bovine α-La (yellow) with β-1,4galactosyltransferase (gray) (PDB code: 2fyd). Several substrate molecules are observed, together with the cleaved nucleotide-sugar moiety (cyan sticks). The MnII ion is shown as a pink sphere, and the calcium ion is shown as a gray sphere. For clarity, loop regions are given a smoothed representation. β-1,4-Galactosyltransferase is not present in milk, and the lactose synthase complex is present only in the milk-producing cells of the alveoli. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this chapter.) Figure drawn with PyMOL. Delano, W.L., 2002. PyMOL. Delano Scientific, Palo Alto, CA.

α-Lactalbumin

271

with β-1,4-galactosyltransferase-I with bound substrates and, interestingly, a trapped intermediate species. The structures of α-La from several other species, including goat, baboon, human, guinea pig, and buffalo, have also been characterized by x-ray diffraction methods (Acharya et al., 1989, 1991; Calderone et al., 1996; Pike et al., 1996; Makabe et al., 2012). Consistent with high sequence identity, there are no significant differences among these structures, except for a flexible loop at residues 105–110, which is implicated in the formation of the lactose synthase complex (Acharya et al., 1989; Calderone et al., 1996; Pike et al., 1996). It is worth noting that the recombinant goat protein, which has an added methionine at the N-terminus, is markedly less stable, by
14 kJ mol1, than the native protein, mostly the result of an increased rate of unfolding (but a preserved rate of refolding) (Chaudhuri et al., 1999). Similar observations have been made on recombinant bovine α-La (Acharya et al., 1989). Thus, native protein functionality and stability should not in general be inferred from measurements of recombinant proteins heterologously expressed in bacterial systems (which generally add an N-terminal methionine residue).

Effect of temperature on bovine α-La In general, holo-α-La undergoes a thermal unfolding at a lower temperature than β-Lg (Ruegg et al., 1977). The role of bound calcium ions appears to be to confer stability to the tertiary structure: with less than equimolar amounts of bound calcium, the thermal unfolding transition is lowered substantially, decreasing to about 35°C for the apo form (Relkin, 1996; Ishikawa et al., 1998). The presence of calcium ions also accelerates the rate of refolding of α-La by more than two orders of magnitude (Wehbi et al., 2005). The presence of calcium ions also aids in the refolding and formation of the correct disulfide linkages of the denatured reduced protein (Wehbi et al., 2005). The structurally closely related, but functionally unrelated, enzyme lysozyme can be subdivided into two classes: a noncalcium-binding subclass, typified by egg-white lysozyme (Grobler et al., 1994; Steinrauf, 1998), and a calcium-binding subclass, including equine and echidna lysozyme (Tsuge et al., 1992; Guss et al., 1997). The thermal denaturation behavior of bovine α-La from three different sources has been studied; significant differences have been reported (McGuffey et al., 2005), which has provided some resolution of the apparently discordant denaturation data from different groups. In the presence of β-Lg or BSA, each of which has an unpaired cysteine, β-Lg-α-La, α-La-BSA, and even α-La-α-La oligomers form at high temperature. Because α-La (which lacks a free thiol) by itself fails under similar conditions to form disulfide-linked oligomers, intermolecular disulfide-sulfhydryl interchange reactions appear to play a role in forming α-La-α-La oligomers (Havea et al., 2001; Hong and Creamer, 2002).

Effect of pressure on bovine α-La Again, the absence of free thiol groups renders α-La intrinsically less susceptible to irreversible structural and functional change induced by high pressure. Reversible unfolding to a molten globule state begins at 200 MPa, and loss of native structure becomes irreversible

272

7. Structure and stability of whey proteins

beyond 400 MPa (McGuffey et al., 2005) (corresponding numbers for β-Lg are 50 and 150 MPa [Stapelfeldt and Skibsted, 1999; Considine et al., 2005b]). In the presence of calcium ions, the denaturation pressure increases by 200 MPa for α-La (Dzwolak et al., 1999). Only in the presence of thiol reducers does oligomerization of α-La occur at high pressures ( Jegouic et al., 1996).

Effect of denaturants on bovine α-La At neutral pH, the calcium-depleted, or apo, form of α-La reversibly denatures to a variety of partially folded or molten globule states upon moderate heating (45°C) or, at room temperature, by dissolving the protein in aqueous TFE (15% TFE) or by adding oleic acid (7.5 equivalents) (Svensson et al., 2000; de Laureto et al., 2002). Under these various conditions, the UV-CD spectra of apo-α-La are essentially identical to those of the most studied molten globule form of α-La, that is, the A state found at pH 2.0 (Kuwajima, 1996). At 4°C and pH 8.3, proteolysis of apo-α-La by proteinase-K occurs slowly and nonspecifically, leading to small peptides only. In contrast, at 37°C, preferential cleavage by proteinase-K is observed at peptide bonds located in loop regions of the β-sheet subdomain of the β-domain of the protein (residues 35–85), creating peptides in which disulfide bridges link N-terminal residues 1–34 to C-terminal fragments, residues 54–123 or 57–123. Preferential cleavage at similar sites and similar disulfide-bridged fragments has also been observed for proteolysis of the molten globule states induced by TFE and oleic acid. Polypeptides formed from the molten globule A state of α-La comprise, therefore, a well-structured native-like conformation of the α-domain and a disordered conformation of the β-subdomain, residues 34–57 (de Laureto et al., 2002). Oleic acid treatment leads to a kinetically trapped folding variant of the protein, which can also bind calcium ions, called human α-lactalbumin made lethal to tumor cells (HAMLET and its bovine analog BAMLET), which has been shown to induce apoptosis in tumor cells (Svensson et al., 2000). Under conditions at which the thermal denaturation of α-La is reversible, the thermal denaturation of HAMLET is irreversible, with respect to the loss of its apoptotic effect on tumor cells (Fast et al., 2005). Human α-La and bovine α-La also weakly bind a second calcium ion, structurally characterized for human α-La (Chandra et al., 1998). In addition, zinc-ion binding to possibly structurally inequivalent sites has been characterized for human α-La and bovine α-La by fluorescence spectroscopic techniques (Permyakov and Berliner, 2000). The binding of zinc ions to calcium-loaded α-La has been shown to destabilize the native structure to heat denaturation (Permyakov and Berliner, 2000). However, the weak binding of zinc (submillimolarity dissociation constant) means that this binding is probably not physiologically relevant.

Serum albumin Serum albumin (SA) is an approximately 580-residue protein that is found in both the blood serum and milk of all mammals and appears to function as a promiscuous transporter of hydrophobic molecules. However, as with many proteins, this transport role appears not to

Serum albumin

273

be the only physiological function for SAs. The structure of human serum albumin (HSA), described in more detail in the next subsection, is notable also for its number of disulfide bridges, 17 in total. There is one unpaired cysteine, Cys34, in HSA (highlighted in Fig. 7.4), and also Cys34 in BSA. This cysteine is part of a highly conserved QQCP(F/Y) motif. It is susceptible to various oxidations, including a two-electron oxidation to sulfenic acid (dSOH) (Carballal et al., 2007). There is evidence that, at least in blood serum, this cysteine is involved in HSA’s role in the control of redox properties (Kawakami et al., 2006). A similar role can be postulated for BSA in blood serum, but this redox role has not been established (or even investigated) in both human milk and bovine milk. In terms of milk flavor and the flavor of milk products, control of the redox states of milk components is obviously of importance. It appears also that, in blood serum, in which HSA is the major protein component, present at a concentration of 0.6 mM,

FIG. 7.4 Structure of HSA complexed with halothane (slate/purple), partially occupying seven distinct sites, and myristic acid (yellow/red), fully occupying five distinct sites (PDB code: 1e7c). Domain IA (residues 5–107) is shown in blue; domain IB (residues 108–196) is shown in light blue; domain IIA (residues 197–297) is shown in green; domain IIB (residues 297–383) is shown in light green; domain IIIA (residues 384–497) is shown in red; domain IIIB (residues 498–582) is shown in light red. The single cysteine, Cys34, is arrowed (Bhattacharya et al., 2000). The 17 disulfide bonds, which tie together individual subdomains, are represented in stick format. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this chapter.) Figure drawn with PyMOL. Delano, W.L., 2002. PyMOL. Delano Scientific, Palo Alto, CA.

274

7. Structure and stability of whey proteins

HSA is the first line of defense against radicals, including reactive oxygen species and nitric oxide. In vitro studies that show the reactivity of Cys34 in HSA have been complemented by in vivo studies that show that, in primary nephrotic syndrome, Cys34 is oxidized to sulfonate, dSO3 (Musante et al., 2006). Again, in milk, defenses against reactive oxygen species are essential to preserve the quality of milk, but HSA and BSA are at much lower concentrations in milk than in blood. HSA has also been shown to have esterase activity (Sakurai et al., 2004). Whether this is physiologically important in blood (or in milk) has not been established for either BSA or HSA. Finally, an active role for HSA in the transport of fatty acids across membranes has been characterized (Cupp et al., 2004). It is in this process that SAs are introduced into mammalian milks.

Structure of SAs The structure of HSA, with which BSA shares 75% sequence identity, has been well characterized for the apo protein and for a variety of complexes with a variety of long-chain fatty acids and other more compact hydrophobic molecules. The structure of HSA complexed with the anesthetic halothane (C2F3Cl2Br) and myristic acid (CH3(CH2)12COOH) is shown in Fig. 7.4. The structure of HSA comprises three structurally homologous domains, each of just under 200 residues, denoted I–III (Curry et al., 1998) and involving residues 5–196, 197–383, and 384–582, respectively. Each domain has two subdomains, each of
100 residues, denoted A and B. The structure lacks β-strands and is predominantly (68%) α-helical, with several lengthy loops connecting the A and B subdomains. On ligand binding, there is substantial movement of the domains with respect to each other, but the tertiary structure of each domain undergoes only small changes (Curry et al., 1998). Medium- and long-chain fatty acids occupy five distinct sites (dissociation constants 0.05–1 μM) (Spector, 1975), one in domain I, a second between domains I and II, and the remaining three in domain III, as characterized by x-ray techniques for HSA. In the case of halothane binding, two sites are located in domain I, and five are located in domain II. A comprehensive study of the binding of 17 distinct drugs to HSA, in the presence and absence of myristate, has been published (Ghuman et al., 2005). Whereas the binding of steroids to BSA is influenced by the binding of fatty acids, for HSA, there is much less influence (Watanabe and Sato, 1996). NMR titrations have shown that BSA, like HSA, binds five myristates; four of the five sites appear to be in structurally homologous sites to those identified crystallographically for HSA (Hamilton et al., 1984; Cistola et al., 1987; Hamilton et al., 1991; Simard et al., 2005). The x-ray structure of equine serum albumin (ESA) is very similar to that of HSA, consistent with
75% sequence identity between these two proteins (Ho et al., 1993). The structures of bovine, equine, and leporine (rabbit) SAs have been published by two independent groups (Bujacz, 2012; Majorek et al., 2012; Sekula et al., 2013). HSA has also been heterologously expressed in rice and structurally characterized (He et al., 2011), finessing problems associated with availability and disease risks when sourced for clinical and cellculture applications from human blood sources.

Serum albumin

275

Effect of temperature on SAs Careful DSC measurements have been made on defatted HSA and its binding from shortto medium-chain fatty acids. In the absence of fatty acids, a single sharp endotherm is observed, yielding a midpoint temperature for denaturation, Tm, of 64.7°C, consistent with a concerted unfolding. In the presence of fatty acids, the endotherm broadens, and there is a steady increase in Tm as the chain length of the fatty acid increases from n-butanoate (Tm ¼ 77.6°C) to n-octanoate (Tm ¼ 87.2°C); the Tms for n-nonanoate and n-decanoate are very similar to that for n-octanoate. The short-chain fatty acids, formate, acetate, and n-propionate, show evidence for inducing increased stability in HSA through the binding of the fatty acids at secondary sites that are inaccessible to the longer-chain fatty acids. The concentration of fatty acid at which maximum stability of HSA is achieved decreases from >2900 mM for formate to 350 MPa, where solubilization of MCP and micellar disruption were nearly complete. Furthermore, this process was more extensive when HPP was performed at a higher temperature (Orlien et al., 2006) and at a lower pH (Huppertz and De Kruif, 2006). This process was not influenced by the presence of whey proteins (Huppertz and De Kruif, 2007a,b), suggesting that it is driven primarily by casein-casein interactions. When the pressure was released, there was a subsequent reduction in light transmission, indicating further reformation of the casein particles. This reduction in light transmission coincided with the reversal in the solubility of MCP upon pressure release (Hubbard et al., 2002), to the extent that the distribution of calcium and inorganic

298

8. Effect of nonthermal processing on milk protein interactions and functionality

phosphate between the micellar and serum phases of milk was identical to that in untreated samples (Regnault et al., 2006). However, when treatment was carried out at pressures >300 MPa, the initial light transmission values were not restored, indicating that the casein micelles did not reform in their native state (Keenan et al., 2001; Huppertz et al., 2006c, 2011). Another factor to take into consideration is the pressure-release rate, which effectively controls the rate of reformation of the casein micelles on pressure release (Merel-Rausch et al., 2006). Because of these changes, the properties of the casein micelles in high-pressure-treated milk can differ markedly from those in untreated milk, for example, in terms of particle size, turbidity, and size distribution. It has been reported that HPP treatment of milk at 250 MPa for up to 10 min causes a reduction in particle size, whereas treatment at this pressure for >15 min causes an increase in particle size (Huppertz et al., 2004a). Casein micelle size and turbidity were not strongly affected at pressures
200 MPa but were reduced considerably at pressures >300 MPa (Gaucheron et al., 1997; Huppertz et al., 2004a,b; Regnault et al., 2004; Anema et al., 2005; Anema, 2008). Transmission electron micrographs of skim milk samples treated at up to 900 MPa for 5 min showed that the smallest casein micelles were formed at around 450 MPa with no further variation at higher pressures (Fig. 8.2). These changes in micellar size correlated with the concentration of soluble casein, because treatment at 250 MPa significantly enhanced the level of nonsedimentable casein, whereas, compared with lower pressures, there was no further increase between 700 and 900 MPa (Bravo et al., 2015). Such HPP-induced changes were also dependent on the pressurization temperature (Huppertz et al., 2004a; Anema et al., 2005), the pH (Huppertz et al., 2004a), and the solids content of the milk (Anema, 2008). However, such changes were independent of treatment time. The increase in the particle size was accompanied by reductions in turbidity and light scattering intensity, suggesting that this increase in the average particle size was probably due to the presence of a small proportion of large particles, as was also apparent from the electron micrographs of Knudsen and Skibsted (2009). The increase in particle size at 250 MPa was reversible on subsequent storage of the milk, with greater reversibility upon storage at 20°C than at 5°C. In contrast, HPP-induced reductions in particle size were largely irreversible on subsequent storage (Huppertz et al., 2004a).

Effects of HPP on whey proteins Considerable differences in the sensitivities of the different whey proteins to heat (lactoferrin > immunoglobulins > bovine serum albumin (BSA) > β-lactoglobulin (β-LG) > α-lactalbumin (α-LA)) and pressure (β-LG > lactoferrin > immunoglobulins > BSA > α-LA) have been reported (Patel et al., 2006), showing that β-LG is the most pressure sensitive among all the whey proteins. About 70%–80% denaturation of β-LG occurs at 400 MPa (Lo´pez-Fandin˜o et al., 1996; Lo´pez-Fandin˜o and Olano, 1998; Arias et al., 2000; Garcı´a-Risco et al., 2000; Scollard et al., 2000). Relatively, little further denaturation of β-LG occurs at 400–800 MPa (Scollard et al., 2000). Compared with β-LG, α-LA is stable to pressures up to about 400–500 MPa in the milk environment at ambient temperature (Hinrichs et al., 1996a,b; Lo´pez-Fandin˜o et al., 1996; Felipe et al., 1997; Gaucheron et al.,

High-pressure processing

299

FIG. 8.2 Transmission electron micrographs of unpressurized milk (0.1 MPa) and milk pressurized at 250, 450, 700, 800, and 900 MPa for 5 min. The bar corresponds to 500 nm. Reproduced from Bravo, F.I., Felipe, X., Lo´pez-Fandin˜o, R., Molina, E., 2015. Skim milk protein distribution as a result of very high hydrostatic pressure. Food Res. Int. 72, 74–79, with permission.

1997; Lo´pez-Fandin˜o and Olano, 1998; Arias et al., 2000; Garcı´a-Risco et al., 2000; Needs et al., 2000; Scollard et al., 2000; Huppertz et al., 2002, 2004b). It has also been reported that the molecular structure of α-LA is more stable than that of β-LG (Lo´pez-Fandin˜o et al., 1996; Funtenberger et al., 1997; Gaucheron et al., 1997), and that oligomerization takes place only if, during unfolding, free sulfhydryl groups from other molecules are available (Hinrichs et al., 1996b; Lo´pez-Fandin˜o et al., 1996; Gaucheron et al., 1997; Jegouic et al., 1997). This difference in pressure sensitivity can also be explained by the types of bonds stabilizing the conformational structures of β-LG and α-LA (Hinrichs et al., 1996a,b; Gaucheron et al., 1997; Messens et al., 1997). BSA has also been found to be resistant to pressures up to 400 MPa in raw milk (Hinrichs et al., 1996b; Lo´pez-Fandin˜o et al., 1996) and up to 600 MPa (Hayakawa et al., 1992). The high

300

8. Effect of nonthermal processing on milk protein interactions and functionality

stability of BSA can be explained by the fact that BSA has one sulfhydryl group and 17 disulfide bonds. The energy received under pressure treatment is too small to disrupt all the disulfide bonds and to change the molecular structure of BSA. Immunoglobulin G in caprine milk (Felipe et al., 1997) and bovine milk (Carroll et al., 2006; Patel et al., 2006) has been reported to be more resistant to pressure denaturation than to heat denaturation. Different extents of denaturation of β-LG following HPP at 600 MPa of pasteurized milk (Needs et al., 2000) and reconstituted skim milk powder (Gaucheron et al., 1997) have been reported; this may be attributed to the extent of denaturation that has been caused by treatments before pressurization, which may influence the amount of denaturation measured subsequently. The pressure intensity and the holding time have also been reported to affect the extent of denaturation of the whey proteins in milk (Lo´pez-Fandin˜o and Olano, 1998; Huppertz et al., 2004a; Anema et al., 2005; Hinrichs and Rademacher, 2005). The reaction order of the HPP-induced denaturation of β-LG is 2.5 (Hinrichs et al., 1996b), indicating that the denaturation process is concentration dependent and that a lower initial concentration of native β-LG should reduce its extent of denaturation under pressure. Also, β-LG and α-LA are reported to be comparatively more pressure resistant in whey than in milk, which may be attributed to the absence of casein micelles and colloidal calcium phosphate in whey (Huppertz et al., 2004b).

High pressure-induced interactions between whey proteins and casein micelles For functionality reasons, one of the major reactions of interest in heat-treated and pressure-treated milk systems is the interaction between the denatured whey proteins and the casein micelles. On the high-pressure treatment of milk at 300–600 MPa, β-LG may form small aggregates (Felipe et al., 1997) or may interact with the casein micelles (Needs et al., 2000; Scollard et al., 2000; Huppertz et al., 2004c). Lo´pez-Fandin˜o et al. (1997) reported that, when mixtures of κ-casein and β-LG were pressure treated at 400 MPa, the presence of β-LG reduced the susceptibility of κ-casein to subsequent hydrolysis by chymosin, indicating interactions between the proteins. SDS-polyacrylamide gel electrophoresis (PAGE) analysis of pressure-treated and untreated milks or solutions containing κ-casein, or β-LG, or both in the presence or the absence of denaturing agents showed evidence of the formation of aggregates linked by intermolecular disulfide bonds (Lo´pez-Fandin˜o et al., 1997). Interestingly, αs2-casein occurs at the same concentration as κ-casein and has one disulfide bond but has not normally been reported to interact with β-LG in milk systems heated at 85–90°C. In contrast, Patel et al. (2006) reported that the effects of heat treatment and high-pressure treatment on the interactions of the caseins and whey proteins in milk were significantly different, by demonstrating the formation of disulfide-linked complexes involving αs2-casein, κ-casein, and whey proteins in heat- and pressure-treated milks. The results have been explained using modified twodimensional SDS-PAGE and then reduced SDS-PAGE and by proposing possible reactions of the caseins and whey proteins in heat- and pressure-treated milks. The virtual absence of αs2-casein from the heat-induced aggregates formed at 85–90°C in milk, as reported in previous studies, might be because αs2-casein is not a surface component of the micelle

High-pressure processing

301

and therefore its disulfide bond(s) are inaccessible to the denatured β-LG. In contrast, κ-casein is on the surface of the micelles, and its disulfide bond(s) could be readily accessible to a thiol group of β-LG. Moreover, it has been reported that large quantities of very large aggregates that cannot enter the gel are present to a greater extent in heat-treated milk than in pressuretreated milk, indicating that the sizes of the aggregates are comparatively smaller in pressuretreated milks than in heat-treated milks. Such differences can be attributed to different effects of heat treatment and pressure treatment on the structure of the proteins, which may ultimately lead to different textures of the final products. Upon HPP treatment of milk serum that was depleted of casein micelles, no sedimentable whey proteins were observed despite high levels of whey protein denaturation, indicating that the sedimentable whey proteins in HPP-treated milk are mostly associated with the casein micelles (Huppertz et al., 2004a). The level of denatured β-LG associated with the casein micelles increased with an increase in the pressure intensity, the treatment time, and the pressurization temperature (Huppertz et al., 2004a; Zobrist et al., 2005; Anema, 2008, 2010). There was no effect of the β-LG content or the solids content of the milk (Anema, 2008). However, the association of β-LG with the casein micelles increased with an increase in the pH of the milk before HPP treatment (Huppertz et al., 2004a; Anema, 2010), whereas the addition of KIO3 to the milk prior to HPP treatment resulted in a lower level of denatured β-LG being associated with the casein micelles in HPP-treated milk, probably because the formation of disulfide bridges through thiol-thiol interactions rather than thiol-disulfide interchange reactions is favored in the oxidizing environment (Zobrist et al., 2005). In contrast to β-LG, most denatured α-LA was found in the serum phase of HPP-treated milk. The two-dimensional PAGE also showed that pressure treatment of milk at 200 MPa caused β-LG to form disulfide-bonded dimers and incorporated β-LG into aggregates, probably disulfide bonded to κ-casein, suggesting that this was the preferential reaction at this pressure (Patel et al., 2006). The other whey proteins appeared to be less affected at 200 MPa. In contrast, pressure treatment at 800 MPa incorporated β-LG and most of the minor whey proteins (including immunoglobulins and lactoferrin), as well as κ-casein and much of the αs2-casein, into large aggregates. However, only a proportion of the α-LA was denatured or incorporated into the large aggregates. The relatively lower reactivity of α-LA at high pressures is probably related to the relative stability of this protein compared with β-LG, as discussed earlier, and is based on the unusual pressure-dependent behavior of α-LA (Kuwajima et al., 1990; Kobashigawa et al., 1999; Lassalle et al., 2003). At higher pressures (>400 MPa), the polymerization of β-LG becomes the norm, and pressure-induced β-LG aggregation becomes similar to heat-induced β-LG aggregation. The β-LG in whey protein concentrate (WPC) or in milk is not significantly modified by the other components, that is, β-LG dominates the denaturation and aggregation pathway during pressure (>400 MPa) treatment and dominates the reaction under heat treatment at high temperature. All these results show that the differences between the stabilities of the proteins and the accessibilities of the disulfide bonds of the proteins at high temperature or high pressure affect the formation pathways, result in differences among the compositions of the resultant aggregation or interaction products (including their sizes), and may ultimately affect product functionalities.

302

8. Effect of nonthermal processing on milk protein interactions and functionality

US processing Among nonthermal processing techniques, US processing has received significant attention in recent years. US refers to sound waves of frequency above 20 kHz and is undetectable by the human ear. Its functional mechanism is based on passing waves that create regions of high and low pressure; this variation in acoustic pressure is directly proportional to the amount of energy applied to the system. US can be classified into two categories: low intensity (1 W/cm2) with a frequency of 5–10 MHz and high intensity (10–1000 W/cm2) with a frequency of 20–100 kHz (Gallego-Jua´rez et al., 2010; Awad et al., 2012). The major advantage of US is its benign, nontoxic, and environmentally friendly nature (Kentish and Ashokkumar, 2011; Shanmugam et al., 2012). US has a wide range of applications in food processing, such as dehydration, drying, freezing, thawing, emulsification, tenderization of meat, crystallization of lactose, and enzyme inactivation, in cleaning operations, and as an analytical tool (Bhaskaracharya et al., 2009). It is considered to be an effective processing aid, based on the physical effects that it generates in a liquid. The physical effects of US include cavitation (growth and collapse of microbubbles), which can produce high localized temperatures, pressures, and turbulence (Ashokkumar and Grieser, 2007). The US processing technique can also produce some changes in the components of food. Proteins have important functional properties from the point of view of food technology; their amphiphilic character and their ability to form interfacial films create stabilizing systems, such as emulsions and foams. They can also interact to create networks and to form gels to develop edible films. There is evidence to support that US can improve the physical and functional properties of different proteins but, given that US technology is relatively new in food processing, there is limited published information on the effect of US processing on individual proteins and their functionality. The effects of US on caseins (including micellar casein) and whey proteins are reviewed here.

Effect of US treatment on the structure and functionality of milk proteins Even though US treatment has been shown to reduce the size of protein particles in aqueous solution and to enhance dissolution, it does not appear to cause scission of the primary structure for a large number of proteins, including milk protein concentrate (MPC) (Yanjun et al., 2014) and whey protein suspensions (Martini et al., 2010). Although the energy associated with the disulfide bond is less than that associated with the peptide bonds that maintain the primary structure of proteins, most ultrasonic energy is utilized in the disruption of the associative noncovalent interactions maintaining the protein structure, rather than in the disruption of covalent linkages. The acoustic energy employed provides sufficient energy to disrupt hydrogen bonding, hydrophobic and electrostatic interactions (Arzeni et al., 2012; Chandrapala et al., 2011), reducing the aggregate size, but insufficient energy to achieve the scission of covalent linkages (O’Sullivan et al., 2017b). Chandrapala et al. (2012) studied the effect of ultrasonication on the integrity of the casein micelles in fresh skim milk, reconstituted micellar casein, and casein powder samples and reported that there was no change in the size of the casein micelles in skim milk samples sonicated for up to 60 min at 20 kHz. They observed no measurable changes in the free casein

US processing

303

content or the soluble calcium concentration in either fresh skim milk or reconstituted micellar casein, which confirms that the casein micelles were stable during the US treatment. Shanmugam et al. (2012) studied the effect on pasteurized homogenized skim milk of US treatment of 20 kHz at 20 and 41 W for up to 60 min. They observed that the whey proteins and whey protein-whey protein aggregates present in the milk were denatured and formed soluble whey protein-whey protein/whey protein-casein aggregates, which further interacted with the casein micelles to form micellar aggregates during the initial 30 min of sonication. Prolonged sonication resulted in the partial disruption of some whey proteins from these aggregates. The physical forces of acoustic cavitation did not affect the intact structure of the casein micelles. The minor changes to the proteins, caused by sonication, did not alter the viscosity of the milk, although there was an overall particle size reduction of the fat globules and soluble particles in the supernatant, which could be attributed to the shear forces generated by acoustic cavitation. The effects of US treatment on the structural, physical, and emulsifying properties of three dairy proteins (sodium caseinate, whey protein isolate (WPI), and milk protein isolate (MPI)) were studied by O’Sullivan et al. (2014); they found that US treatment (20 kHz, 34 W/cm2 for 2 min) caused a significant (P < 0.05) reduction in the micelle size and the hydrodynamic volume of the proteins. This effect was attributed to the high shear forces resulting from ultrasonic cavitation. However, there were no differences in molecular weight between untreated and US-treated sodium caseinate, WPI, and MPI. Whereas the ultrasonic treatment resulted in no change in the emulsifying capacities of sodium caseinate and WPI, the emulsifying capacity of MPI increased significantly, the droplet size was reduced, and the stability of the emulsions was prolonged, compared with untreated MPI. Zhang et al. (2018) recently reported improved functional properties, such as solubility, emulsification, and gelation, of micellar casein concentrate (MCC) as a result of high-intensity ultrasound (HIUS) pretreatment. SDS-PAGE analysis showed that there were no significant differences in the protein profiles between untreated and US-treated samples (Fig. 8.3), indicating that the HIUS pretreatment did not change the individual subunits of the proteins and that there was no aggregation of the proteins. Similar observations were also reported by Yanjun et al. (2014), who found that MPC solutions sonicated for 0, 0.5, 1, 2, and 5 min at an output power of 600 W displayed similar peptide profiles on SDS-PAGE. The structural characteristics of HIUS-treated MCC samples indicated an increase in surface hydrophobicity and a reduction in particle size compared with the control (without HIUS pretreatment). These changes were closely related to the exposure of hydrophobic regions from the interior to the surface of the molecules, as revealed by the increase in surface hydrophobicity (Ho et al., 1997). Fourier transform infrared spectroscopy showed that HIUS pretreatment increased the β-sheet and random coil contents and decreased the α-helix and β-turn contents of MCC, indicating changes in the secondary structure (Table 8.1). The increase in β-sheet structure contributed to the exposure of the hydrophobic regions of the protein (Wang et al., 2014) and therefore led to an increase in surface hydrophobicity. Furthermore, β-sheets are important in the formation of gels because the intensity of water hydration in a β-sheet is greater than that in an α-helix, therefore resulting in gels with higher elasticity. The reduction in α-helix content might correspond to the partial unfolding of the α-helical region that is caused by ultrasonic cavitation (Kang et al., 2016). Moreover, the increase in random coils can be

304

8. Effect of nonthermal processing on milk protein interactions and functionality

FIG. 8.3 Reducing SDS-PAGE profile of micellar casein concentrates treated with US at 20 kHz frequency with different pretreatment times. Reproduced from Zhang, R., Pang, X., Lu, J., Liu, L., Zhang, S., Lv, J., 2018. Effect of high intensity ultrasound pretreatment on functional and structural properties of micellar casein concentrates. Ultrason. Sonochem. 47, 10–16, with permission. TABLE 8.1 Secondary structure analysis of micellar casein concentrates with high-intensity ultrasound frequency at 20 kHz (power density of 58 W/L) for different pretreatment times Secondary structure composition (%) Ultrasound time (min)

α-Helix

β-Sheet

0.0

26.39  0.34

0.5

β-Turn

Random coil

31.19  0.32

a

20.30  0.58

22.11  0.24c

24.68  0.50b

32.23  0.48b

16.57  0.34b

25.87  0.69c

1.0

24.65  0.63b

32.58  0.38b

16.29  0.31b

26.45  0.59a

2.0

23.96  0.13b

34.09  0.33a

16.46  0.13b

25.46  0.43b

5.0

22.82  0.79c

34.56  0.66a

13.94  0.75c

27.22  0.52a

a

c

Values are the mean  standard deviation. Different letters in the same column indicate significant differences at P < 0.05 (n ¼ 3). Reproduced from Zhang, R., Pang, X., Lu, J., Liu, L., Zhang, S., Lv, J., 2018. Effect of high intensity ultrasound pretreatment on functional and structural properties of micellar casein concentrates. Ultrason. Sonochem. 47, 10–16, with permission.

attributed to the transformation of β-turns into random coils because the US treatment can reduce the number of intramolecular hydrogen bonds (Yong et al., 2006). Tammineedi et al. (2013) observed no change in the allergenicity of the major milk proteins (i.e., α-casein, β-LG, and α-LA) as a result of HIUS treatment. There was no noticeable change in the band intensities on SDS-PAGE for α-casein, β-LG, and α-LA, indicating no change in protein concentration with US treatment. Competitive inhibition enzyme-linked immunosorbent assay analysis further confirmed that there were no significant differences (P > 0.05) in immunoglobulin E binding values for control milk and US-treated milk.

US processing

305

In a study on the effect of US on the renneting properties of milk, Liu et al. (2014) reported an improvement in the renneting properties of milk sonicated at pH 6.7 compared with those of the nonsonicated control milk. However, better sonication results, in terms of rennet gelation time, curd firming rate, curd firmness, and the connectivity of the rennet gel network, were observed in milk samples that were ultrasonicated at pH 8.0 and readjusted back to pH 6.7 than in milk samples that were sonicated at pH 6.7. These superior renneting properties were attributed to the reduction in the size of the particles in the milk and to possible changes in the protein hydrophobicity, caused by the physical effects of cavitation when applying low-frequency US. The altered renneting properties may have benefits in cheese manufacture, such as decreasing the time to produce the cheese curd or increasing the firmness of the cheese curd. Zisu et al. (2011) evaluated the effect of US on the physicochemical properties of reconstituted whey proteins and observed an improvement in their gelling property and obtained gels with higher strength and reduced syneresis. Jambrak et al. (2008) reported an increase in solubility and foam formation. Arzeni et al. (2012) performed a comparative evaluation of high-energy ultrasonic effects on the functionality of food proteins and found that gelation, viscosity, and solubility are related to molecular modification and increased hydrophobicity and to the particle size variation. Recently, high-energy US has been demonstrated to be an alternative in improving the solubility of proteins from different food sources; various studies have reported increased solubility, decreased particle size, and improved functional properties ( Jambrak et al., 2008, 2009; Arzeni et al., 2012; Morales et al., 2015). In a study of changes to the functional properties of proteins, Jambrak et al. (2008) evaluated the effect of high-energy US (20–40 kHz) at low intensity for 15–30 min on whey proteins. US was applied to solutions of WPI, WPC, and hydrolyzed whey protein at 10% (w/w). It resulted in increased solubility in all samples, except for WPC. This exception was attributed to its different sample compositions because the WPC contained a considerable amount of lactose (25%), which possibly protected the proteins and prevented a change in solubility. Conversely, the improvement in solubility of the WPI and the hydrolyzed whey protein was attributed to conformational changes in the globular structure of the proteins, exposing the hydrophilic regions to the water. This was confirmed by an increase in the electrical conductivity, allowing greater water-protein interactions (Higuera-Barraza et al., 2016). MPCs represent enriched forms of the naturally occurring casein and whey protein complement in defatted (skimmed) milk. The rehydration of MPC powders is a time-consuming process and is significantly affected by several parameters, such as the heat treatment of the skim milk, the protein and mineral contents, the spray drying temperatures, the storage conditions of the subsequent powder (i.e., time and temperature), and the rehydration process (i.e., temperature of reconstitution and mixing conditions) (Havea, 2006; Gaiani et al., 2007, 2009, 2010; Mimouni et al., 2010; Richard et al., 2013). McCarthy et al. (2014) studied the effects of US on the dissolution properties of MPC powders. They reported that ultrasonication can deagglomerate and disperse primary MPC powder particles more rapidly than conventional mixing, while surpassing the latter in terms of overall levels of protein solubilization. Ultrasonic processing is capable of producing a smaller particle size distribution after 1 min (energy density, 21.1 J/mL) of treatment and therefore within a fraction of the time compared with stirring at 50°C.

306

8. Effect of nonthermal processing on milk protein interactions and functionality

Several researchers have also studied the effects of US on individual proteins such as lysozyme (Cavalieri et al., 2008), BSA (Stathopulos et al., 2004; G€ ulseren et al., 2007), casein (Madadlou et al., 2009), WPI (Gordon and Pilosof, 2010), WPC (Chandrapala et al., 2011; Arzeni et al., 2012), and whey protein aggregates (Ashokkumar et al., 2009; Saffon et al., 2011).

PEF processing PEF technology was first reported in the 1960s in Germany by H. Doevenspeck, who can rightfully be called the father of this processing method (Misra et al., 2017), as he introduced different PEF equipment and techniques in his first patent (DE 1237541 B) (Doevenspeck, 1960). Among other nonthermal techniques, PEF processing is one of the most explored technologies in the food processing sector. It involves the treatment of a pumpable food with very short pulses (microseconds) at very high electric field strengths. Typically, it is carried out at electric field strengths of 10–50 kV/cm in multiple short pulses (typically 1–5 μs) at frequencies of 200–400 Hz. The total treatment time, which is the product of the pulse width and the number of pulses, is usually much β-LG. Perez and Pilosof (2004) reported that β-LG concentrates that were subjected to high-intensity electric fields (12.5 kV/cm, 1–10 number of pulses) were partially denatured and that the thermal stability was greatly reduced. They further reported that PEF processing can induce changes such as polarization, dissociation of noncovalently linked subunits, change in the protein conformation, attraction between protein structures by electrostatic forces, and, if the electric field pulses are induced for a sufficiently long period, hydrophobic interactions or covalent bonds, resulting in the formation of aggregates. However, contrary to previous findings, Barsotti et al. (2001) reported that PEF treatment of β-LG-stabilized oil-in-water emulsions and β-LG solution at 21–36 kV/cm in a continuous mode of operation did not result in any noticeable unfolding or aggregation of this protein. The observed differences in the stability of the whey proteins could have been due to the different pH values and other conditions used during the measurements (Sharma et al., 2014).

UV irradiation processing UV radiation is electromagnetic energy that is located in the electromagnetic spectrum (100–400 nm) at wavelengths between those of X-rays and visible light (Harm, 1980; Miller et al., 1999). The UV spectrum exists in three forms: (a) UV-A, which extends from 320 to 400 nm; (b) UV-B, which extends between 290 and 320 nm; and (c) UV-C, which extends between 100 and 290 nm (Miller et al., 1999). It has been known for almost a century that UV light can kill bacteria (McDonald et al., 2000a). The germicidal effectiveness of UV radiation depends on its wavelength, with light at 180–320 nm being able to inactivate bacteria and viruses, with an optimum effect at 265 nm (Blatchley and Peel, 2001). There are a number of sources of UV light, including low-pressure mercury glow discharges, medium-pressure mercury discharges, pulsed xenon arc discharge, xenon excimer, and submerged arc, all of which function similarly. An electric discharge ionizes the gas, which radiates photons. Approximately 95% of the UV light radiated by low-pressure mercury arcs is at a wavelength of 253.7 nm and is considered to be the most effective source of UV light for germicidal applications (Meulemans, 1987). Because of its germicidal effect, UV radiation has been used

UV irradiation processing

313

extensively in many sterilization applications, such as drinking water disinfection (McDonald et al., 2000b), medical device sterilization (Khomich et al., 1998), and polymeric material sterilization (Fischbach et al., 2001). UV light technology has been successfully commercialized for the disinfection of water and the pasteurization of fruit beverages such as juices (Basaran et al., 2004). However, the limited ability of UV light to penetrate turbid liquids is considered to be one of the main bottlenecks in its use as a nonthermal technology for milk (Kristo et al., 2012). A loss of radiation intensity of up to 30% can be achieved at 40 cm from the surface in distilled water, but at only 5 cm in a 10% sucrose solution (Snowball and Hornsey, 1988). In fruit juices, 90% of UV light is absorbed in the first 1 mm from the surface (Sizer and Balasubramaniam, 1999). Short wave UV light (UV-C; 200–280 nm) offers one of the most promising nonthermal continuous flow treatments; it could provide milk processors with a safe, energy-efficient, and cost-effective method of achieving an added measure of quality and an extended shelf life, because of its lethal activity against most microorganisms, including bacteria, viruses, and parasites (Bintsis et al., 2000). Modern UV treatment equipment typically involves a UV-penetrable tube, through which the liquid product is pumped. Ideally, the flow in the tube is turbulent. Turbulent flow continuously renews the surface and ensures that all parts of the liquid come into contact with the UV light. This is essential for an opaque product such as milk and can be achieved with the use of static mixers (Altic et al., 2007; Kristo et al., 2012) or with a complex swirling flow such as is achieved through a corrugated spiral tube in the commercial SurePure Turbulator system (Cilliers et al., 2014). An alternative arrangement to turbulent flow is laminar flow in a very thin film (Mahmoud and Ghaly, 2004). Research on the UV irradiation of milk has been conducted over many decades (Burton, 1951). UV-C has been shown to reduce bacterial numbers, but the dose required to achieve a sufficient reduction for pasteurization (a 5-log reduction) causes unpleasant, light-induced flavors in milk. In the dairy industry, UV radiation is used mainly for the sterilization of cheese whey, as an alternative to pasteurization, because whey often has to be stored for some time before being processed into WPC or WPI and cannot be thermally pasteurized. A reduction in the bacterial load by a nonthermal process such as UV irradiation to improve its keeping quality is therefore attractive: (1) it shortens the sterilization time, (2) it is considered to be an inexpensive technique, and (3) it removes the need to repump the medium in continuous mode fermentations. However, the problem encountered with the use of UV radiation in the sterilization of cheese whey is the high turbidity of the whey, which reduces the efficiency of the sterilization by shielding the microorganisms. Corbitt (1990) and Blatchley and Peel (2001) reported that turbidity, color, and/or organic compounds reduce the effectiveness of the UV disinfection process because of the limited penetration ability of the UV radiation. Therefore, the whey must be clarified before the UV treatment to ensure that it achieves an acceptable level of bacterial destruction. In recent years, UV radiation technology has also been used for the preservation of milk (Matak et al., 2007; Semagoto et al., 2014; Cappozzo et al., 2015). However, as for other nonthermal processing technologies, to date, most of the available information on UV processing is focused on preservation and microbial food safety aspects. There is limited published information on the effect of UV applications on milk proteins and their interactions.

314

8. Effect of nonthermal processing on milk protein interactions and functionality

Effect of UV irradiation on milk proteins Kuan et al. (2011) investigated the structural changes in sodium caseinate upon UV irradiation for 6 h, with observations being recorded at intervals of 30 min. Although there was no significant change in the major SDS-PAGE bands at exposure times of 30–120 min, the intensities of the bands around molecular weights of 75–150 kDa were slightly reduced on extended exposure for 4 h, compared with the control sample, and all of the major bands disappeared on extended exposure for 6 h, indicating UV-induced cross-linking. Fourier transform infrared spectral analysis of UV-treated sodium caseinate samples gave similar results (Fig. 8.4). The presence of primary or secondary amines and amides can be detected by absorption because of the stretching of NH2 or NH groups between 3350 and 3200 cm1. Fig. 8.4 shows no major differences for any of the sodium caseinate samples, except for those treated for an extended duration (over 4 h). The band at 1240 cm1, indicating C–N stretching in aromatic amines, disappeared. A significant improvement in the emulsifying and foaming

FIG. 8.4 Fourier transform infrared spectra of control and ultraviolet (UV)-irradiated sodium caseinate samples. Reproduced from Kuan, Y.H., Bhat, R., Karim, A.A., 2011. Emulsifying and foaming properties of ultraviolet-irradiated egg white protein and sodium caseinate. J. Agric. Food Chem. 59(8), 4111–4118, with permission.

315

UV irradiation processing

properties of sodium caseinate was reported, which was attributed to possible cross-linking of the proteins as a result of prolonged UV irradiation. Scheidegger et al. (2010) reported on the photooxidation of the proteins in skim milk and whole milk as a result of irradiation under UV light and fluorescent light. They found the presence of carbonyl moieties as early as after 1 h of UV treatment and 4 h of fluorescent lighting exposure, probably because of the oxidation of tryptophan, histidine, and methionine, and the formation of dityrosine (Dalsgaard et al., 2007). Fluorescent irradiation for 4 h produced several changes in addition to carbonyl formation, such as the appearance of dityrosine bonds and alteration of the whey protein to casein balance. Aggregation of proteins was observed only in milk samples exposed to UV irradiation and not in those exposed to fluorescent light. It has also been suggested that the aggregation and cleavage of proteins that is produced by light exposure would not have any detrimental effect on the enzymatic (pepsin and chymosin) proteolysis during cheese making. Pattison et al. (2012) reviewed the various mechanisms for the photooxidation of proteins. From structural and functional perspectives, the two most important consequences of the photo-induced oxidation of milk proteins are their unfolding and aggregation. The photooxidation of some proteins can result in an increased extent of, or susceptibility to, unfolding and conformational changes, enhanced exposure of hydrophobic residues, and altered light scattering and optical rotation properties (Silva et al., 2000; Dalsgaard et al., 2007; Redecke et al., 2009). Studies on milk proteins have demonstrated that random coil proteins (such as caseins) are more susceptible to individual amino acid damage than more densely packed globular proteins (e.g., α-LA and β-LG). Table 8.3 shows that the proteins that displayed the greatest changes in secondary and/or tertiary TABLE 8.3 Summary of changes introduced in the protein structures after photooxidation of six different protein solutionsa αs-Casein

β-Casein

κ-Casein

α-Lactalbumin

β-Lactoglobulin

Lactoferrin

Carbonyl formation

High

High

Medium

Medium

Medium

Low

Dityrosine formation

High

High

Medium

Medium

Low

Low

Loss of tryptophan

High

High

Low

Medium

Low

Low

Change in secondary structure

No changes

No changes

No changes

Loss of α-helix

Loss of β-sheets

No changes

Change in tertiary structure

Not detected

Not detected

Not detected

Unfolding (10%)

Unfolding (3%)

More compact structure (12%)

Polymer formation

Dimers and polymers

Dimers and polymers

No changes

No changes

No changes

Dimers and polymers

a

The numbers in parentheses indicate the level of measured decrease or increase in tryptophan emission. Reproduced from Dalsgaard, T.K., Otzen, D., Nielsen, J.H., Larsen, L.B., 2007. Changes in structures of milk proteins upon photo-oxidation. J. Agric. Food Chem. 55(26), 10968–10976, with permission.

316

8. Effect of nonthermal processing on milk protein interactions and functionality

structure were globular proteins, which exhibited less measurable damage at the individual amino acid level. In contrast, the photooxidation of lactoferrin resulted in a more compact protein structure, rather than protein unfolding (Dalsgaard et al., 2007). These data are consistent with low levels of damage to particular residues in globular proteins having major effects on the protein structure, particularly if these residues are buried within a compact structure. In contrast, damage to exposed residues (e.g., in random coil structures) appears to have a much less marked structural impact. Although the chemical modifications and the resulting structural ramifications of the effect of UV/visible radiation on proteins are being increasingly well characterized, much remains to be done in unraveling the biological and functional consequences of these changes.

Concluding remarks Nonthermal processes can be used in a wide range of food applications, especially because of their potential for microbial destruction, shelf-life extension, and modification of the physicochemical properties of food ingredients. However, among all the potential food applications of these novel technologies, the prime focus to date has remained either on shelf-life extension or on food safety. Because of the positive effect of these nonthermal processing techniques on product quality and taste, compared with their conventional thermal counterparts, consumer interest in such products has grown significantly in recent years. However, there has been less focus on other effects of these nonthermal technologies on the structure and functionality of milk proteins. In this chapter, the effects of some novel nonthermal processing techniques, primarily HPP, US, PEF processing, and UV radiation, on the structure of milk proteins, their interactions, and their functionalities have been reviewed. It is expected that this review will help readers to understand the effects of these processing techniques on the milk proteins, including an understanding of the complex process-structure-function relationships, and to explore the possibility of a science-based development of tailor-made foods.

References Altic, L.C., Rowe, M.T., Grant, I.R., 2007. UV light inactivation of Mycobacterium avium subsp. paratuberculosis in milk as assessed by FASTPlaqueTB phage assay and culture. Appl. Environ. Microbiol. 73 (11), 3728–3733. Anema, S.G., 2008. Effect of milk solids concentration on whey protein denaturation, particle size changes and solubilization of casein in high-pressure-treated skim milk. Int. Dairy J. 18, 228–235. Anema, S.G., 2010. Effect of pH at pressure treatment on the acid gelation of skim milk. Innov. Food Sci. Emerg. Technol. 11, 265–273. Anema, S.G., Lowe, E.K., Stockmann, R., 2005. Particle size changes and casein solubilisation in high-pressure-treated skim milk. Food Hydrocoll. 19, 257–267. Arias, M., Lo´pez-Fandin˜o, R., Olano, A., 2000. Influence of pH on the effects of high pressure on milk proteins. Milchwissenschaft 55, 191–194. Arzeni, C., Martı´nez, K., Zema, P., Arias, A., Perez, O.E., Pilosof, A.M.R., 2012. Comparative study of high intensity ultrasound effects on food proteins functionality. J. Food Eng. 108 (3), 463–472. Ashokkumar, M., Grieser, F., 2007. The effect of surface active solutes on bubbles in an acoustic field. Phys. Chem. Chem. Phys. 9 (42), 5631–5643. Ashokkumar, M., Lee, J., Zisu, B., Bhaskarcharya, R., Palmer, M., Kentish, S., 2009. Hot topic: sonication increases the heat stability of whey proteins. J. Dairy Sci. 92 (11), 5353–5356.

References

317

Awad, T.S., Moharram, H.A., Shaltout, O.E., Asker, D., Youssef, M.M., 2012. Applications of ultrasound in analysis, processing and quality control of food: a review. Food Res. Int. 48 (2), 410–427. Barbosa-Ca´novas, G., Bermu´dez-Aguirre, D., 2010. Other novel milk preservation technologies: ultrasound, irradiation, microwave, radio frequency, ohmic heating, ultraviolet light and bacteriocins. In: Improving the Safety and Quality of Milk. Woodhead Publishing, Cambridge, UK, pp. 420–450. Barbosa-Ca´novas, G.V., Go´ngora-Nieto, M.M., Pothakamury, U.R., Swanson, B.G., 1999. PEF inactivation of vegetative cells, spores, and enzymes in foods. In: Preservation of Foods With Pulsed Electric Fields. Academic Press, San Diego, CA, pp. 108–152. Barsotti, L., Dumay, E., Mu, T.H., Diaz, M.D.F., Cheftel, J.C., 2001. Effects of high voltage electric pulses on proteinbased food constituents and structures. Trends Food Sci. Technol. 12 (3–4), 136–144. Basaran, N., Quintero-Ramos, A., Moake, M.M., Churey, J.J., Worobo, R.W., 2004. Influence of apple cultivars on inactivation of different strains of Escherichia coli O157:H7 in apple cider by UV irradiation. Appl. Environ. Microbiol. 70 (10), 6061–6065. Bermu´dez-Aguirre, D., Barbosa-Ca´novas, G.V., 2010. Recent advances in emerging nonthermal technologies. In: Food Engineering Interfaces. Springer, New York, NY, pp. 285–323. Bermu´dez-Aguirre, D., Ferna´ndez, S., Esquivel, H., Dunne, P.C., Barbosa-Ca´novas, G.V., 2011. Milk processed by pulsed electric fields: evaluation of microbial quality, physicochemical characteristics, and selected nutrients at different storage conditions. J. Food Sci. 76 (5), S289–S299. Beverung, C.J., Radke, C.J., Blanch, H.W., 1999. Protein adsorption at the oil/water interface: characterization of adsorption kinetics by dynamic interfacial tension measurements. Biophys. Chem. 81 (1), 59–80. Bhaskaracharya, R.K., Kentish, S., Ashokkumar, M., 2009. Selected applications of ultrasonics in food processing. Food Eng. Rev. 1 (1), 31–49. Bintsis, T., Litopoulou-Tzanetaki, E., Robinson, R.K., 2000. Existing and potential applications of ultraviolet light in the food industry—a critical review. J. Sci. Food Agric. 80 (6), 637–645. Blatchley III, E.R., Peel, M.M., 2001. Disinfection by ultraviolet irradiation. In: Block, S.S. (Ed.), Disinfection, Sterilization and Preservation, fifth ed. Lippincott Williams & Wilkins, Philadelphia, PA, pp. 823–851. Bouyer, E., Mekhloufi, G., Rosilio, V., Grossiord, J.L., Agnely, F., 2012. Proteins, polysaccharides, and their complexes used as stabilizers for emulsions: alternatives to synthetic surfactants in the pharmaceutical field? Int. J. Pharm. 436 (1–2), 359–378. Bravo, F.I., Felipe, X., Lo´pez-Fandin˜o, R., Molina, E., 2015. Skim milk protein distribution as a result of very high hydrostatic pressure. Food Res. Int. 72, 74–79. Buckow, R., Chandry, P.S., Ng, S.Y., McAuley, C.M., Swanson, B.G., 2014. Opportunities and challenges in pulsed electric field processing of dairy products. Int. Dairy J. 34 (2), 199–212. Burton, H., 1951. Ultra-violet irradiation of milk. Dairy Sci. Abstr. 13 (3), 229–244. Cappozzo, J.C., Koutchma, T., Barnes, G., 2015. Chemical characterization of milk after treatment with thermal (HTST and UHT) and nonthermal (turbulent flow ultraviolet) processing technologies. J. Dairy Sci. 98 (8), 5068–5079. Carroll, T.J., Patel, H.A., Gonzalez, M.A., Dekker, J.W., Collett, M.A., Lubbers, M.W., 2006. High Pressure Processing of Metal Ion Lactoferrin. International Patent Publication Number WO 2006/096073 A1. Cavalieri, F., Ashokkumar, M., Grieser, F., Caruso, F., 2008. Ultrasonic synthesis of stable, functional lysozyme microbubbles. Langmuir 24 (18), 10078–10083. Chandrapala, J., Zisu, B., Palmer, M., Kentish, S., Ashokkumar, M., 2011. Effects of ultrasound on the thermal and structural characteristics of proteins in reconstituted whey protein concentrate. Ultrason. Sonochem. 18 (5), 951–957. Chandrapala, J., Oliver, C., Kentish, S., Ashokkumar, M., 2012. Ultrasonics in food processing. Ultrason. Sonochem. 19 (5), 975–983. Cilliers, F.P., Gouws, P.A., Koutchma, T., Engelbrecht, Y., Adriaanse, C., Swart, P., 2014. A microbiological, biochemical and sensory characterisation of bovine milk treated by heat and ultraviolet (UV) light for manufacturing Cheddar cheese. Innov. Food Sci. Emerg. Technol. 23, 94–106. Considine, T., Patel, H.A., Anema, S.G., Singh, H., Creamer, L.K., 2007. Interactions of milk proteins during heat and high hydrostatic pressure treatments - a review. Innov. Food Sci. Emerg. Technol. 8, 1–23. Corbitt, R.A., 1990. Standard Handbook of Environmental Engineering. McGraw-Hill, New York, NY. Craven, H.M., Swiergon, P., Ng, S., Midgely, J., Versteeg, C., Coventry, M.J., Wan, J., 2008. Evaluation of pulsed electric field and minimal heat treatments for inactivation of pseudomonads and enhancement of milk shelf-life. Innov. Food Sci. Emerg. Technol. 9 (2), 211–216.

318

8. Effect of nonthermal processing on milk protein interactions and functionality

Dalgleish, D.G., 1990. Denaturation and aggregation of serum proteins and caseins in heated milk. J. Agric. Food Chem. 38 (11), 1995–1999. Dalsgaard, T.K., Otzen, D., Nielsen, J.H., Larsen, L.B., 2007. Changes in structures of milk proteins upon photooxidation. J. Agric. Food Chem. 55 (26), 10968–10976. Damodaran, S., 1996. Functional properties. In: Nakai, S., Modler, H.W. (Eds.), Food Proteins: Properties and Characterization. VCH Publishers, New York, NY, pp. 167–234. de Luis, R., Arias, O., Puertolas, E., Benede, S., Sanchez, L., Calvo, M., Perez, M.D., 2009. Effect of high-intensity pulse electric fields on denaturation of bovine whey proteins. Milchwissenschaft 64 (4), 422–426. Deeth, H.C., Datta, N., 2011. Heat treatment of milk j non-thermal technologies: high pressure processing. In: Fuquay, J.W., McSweeney, P.L.H., Fox, P.F. (Eds.), Encyclopedia of Dairy Sciences. Academic Press, New York, NY, pp. 732–737. Devi, A.F., Buckow, R., Hemar, Y., Kasapis, S., 2013. Structuring dairy systems through high pressure processing. J. Food Eng. 114 (1), 106–122. Doevenspeck, H., 1960. Verfahren und Vorrichtung zur Gewinnung der einzelnen Phasen aus dispersen Systemen. German Patent DE 1237541. Donovan, M., Mulvihill, D.M., 1987. Thermal denaturation and aggregation of whey proteins. Irish J. Food Sci. Technol. 11, 87–100. Drapala, K.P., Auty, M.A., Mulvihill, D.M., O’Mahony, J.A., 2015. Influence of lecithin on the processing stability of model whey protein hydrolysate-based infant formula emulsions. Int. J. Dairy Technol. 68 (3), 322–333. Dunn, J., 1996. Pulsed light and pulsed electric field for foods and eggs. Poult. Sci. 75 (9), 1133–1136. Eberhard, P., Sieber, R., 2006. Treatment of milk with pulsed electric fields–an alternative heat treatment. Mitt. Lebensmittelunters. Hyg. 97 (6), 407–432. Falguera, V., Paga´n, J., Garza, S., Garvı´n, A., Ibarz, A., 2011. Ultraviolet processing of liquid food: a review: Part 2: Effects on microorganisms and on food components and properties. Food Res. Int. 44 (6), 1580–1588. Felipe, X., Capellas, M., Law, A.J.R., 1997. Comparison of the effects of high-pressure treatments and heat pasteurization on the whey proteins in goat’s milk. J. Agric. Food Chem. 45, 627–631. Fischbach, C., Tessmar, J., Lucke, A., Schnell, E., Schmeer, G., Blunk, T., Gopferich, A., 2001. Does UV irradiation affect polymer properties relevant to tissue engineering? Surf. Sci. 491, 333–345. Floury, J., Grosset, N., Leconte, N., Pasco, M., Madec, M.N., Jeantet, R., 2006. Continuous raw skim milk processing by pulsed electric field at non-lethal temperature: effect on microbial inactivation and functional properties. Lait 86 (1), 43–57. Funtenberger, S., Dumay, E., Cheftel, J.C., 1997. High pressure promotes β-lactoglobulin aggregation through SH/S-S interchange reactions. J. Agric. Food Chem. 45, 912–921. Gaiani, C., Schuck, P., Scher, J., Desobry, S., Banon, S., 2007. Dairy powder rehydration: influence of protein state, incorporation mode, and agglomeration. J. Dairy Sci. 90 (2), 570–581. Gaiani, C., Scher, J., Schuck, P., Desobry, S., Banon, S., 2009. Use of a turbidity sensor to determine dairy powder rehydration properties. Powder Technol. 190 (1–2), 2–5. Gaiani, C., Morand, M., Sanchez, C., Tehrany, E.A., Jacquot, M., Schuck, P., Jeantet, R., Scher, J., 2010. How surface composition of high milk proteins powders is influenced by spray-drying temperature. Colloids Surf. B: Biointerfaces 75 (1), 377–384. Gallego-Jua´rez, J.A., Rodriguez, G., Acosta, V., Riera, E., 2010. Power ultrasonic transducers with extensive radiators for industrial processing. Ultrason. Sonochem. 17 (6), 953–964. Garcı´a-Risco, M.R., Olano, A., Ramos, M., Lo´pez-Fandin˜o, R., 2000. Micellar changes induced by high pressure. Influence in the proteolytic activity and organoleptic properties of milk. J. Dairy Sci. 83, 2184–2189. Gaucheron, F., Famelart, M.H., Mariette, F., Raulot, K., Michel, F., Le Graet, Y., 1997. Combined effects of temperature and high pressure treatments on physicochemical characteristics of skim milk. Food Chem. 59, 439–447. Gebhart, R., Doster, W., Kulozik, U., 2005. Pressure-induced dissociation of casein micelles: size distribution and effects of temperature. Braz. J. Med. Biol. Res. 38, 1209–1214. Giteru, S.G., Oey, I., Ali, M.A., 2018. Feasibility of using pulsed electric fields to modify biomacromolecules: a review. Trends Food Sci. Technol. 72, 91–113. Go´mez-Lo´pez, V.M., Koutchma, T., Linden, K., 2012. Ultraviolet and pulsed light processing of fluid foods. In: Novel Thermal and Non-thermal Technologies for Fluid Foods. Academic Press, New York, NY, pp. 185–223. Gordon, L., Pilosof, A.M., 2010. Application of high-intensity ultrasounds to control the size of whey proteins particles. Food Biophys. 5 (3), 203–210.

References

319

Grahl, T., M€ arkl, H., 1996. Killing of microorganisms by pulsed electric fields. Appl. Microbiol. Biotechnol. 45 (1–2), 148–157. Grigorovich, N.V., Moiseenko, D.V., Antipova, A.S., Anokhina, M.S., Belyakova, L.E., Polikarpov, Y.N., Korica, N., Semenova, M.G., Baranov, B.A., 2012. Structural and thermodynamic features of covalent conjugates of sodium caseinate with maltodextrins underlying their functionality. Food Funct. 3 (3), 283–289. Guerrero-Beltra´n, J.A., Barbosa-Ca´novas, G.V., 2004. Advantages and limitations on processing foods by UV light. Food Sci. Technol. Int. 10, 137–147. _ G€ G€ ulseren, l., uzey, D., Bruce, B.D., Weiss, J., 2007. Structural and functional changes in ultrasonicated bovine serum albumin solutions. Ultrason. Sonochem. 14 (2), 173–183. Hamilton, W., Sale, J., 1967. Effects of high electric fields on microorganisms. Biochim. Biophys. Acta 148, 789–800. Harm, W., 1980. Biological Effects of Ultraviolet Radiation. Cambridge University Press, New York, NY. Haskard, C.A., Li-Chan, E.C., 1998. Hydrophobicity of bovine serum albumin and ovalbumin determined using uncharged (PRODAN) and anionic (ANS-) fluorescent probes. J. Agric. Food Chem. 46 (7), 2671–2677. Havea, P., 2006. Protein interactions in milk protein concentrate powders. Int. Dairy J. 16 (5), 415–422. Hayakawa, I., Kajihara, J., Morikawa, K., Oda, M., Fujio, Y., 1992. Denaturation of bovine serum albumin (BSA) and ovalbumin by high pressure, heat and chemicals. J. Food Sci. 57, 288–292. Hemar, Y., Augustin, M.A., Cheng, L.J., Sanguansri, P., Swiergon, P., Wan, J., 2011. The effect of pulsed electric field processing on particle size and viscosity of milk and milk concentrates. Milchwissenschaft 66, 126–128. Higuera-Barraza, O.A., Del Toro-Sanchez, C.L., Ruiz-Cruz, S., Ma´rquez-Rı´os, E., 2016. Effects of high-energy ultrasound on the functional properties of proteins. Ultrason. Sonochem. 31, 558–562. Hinrichs, J., Rademacher, B., 2005. Kinetics of combined thermal and pressure-induced whey protein denaturation in bovine skim milk. Int. Dairy J. 15, 315–323. Hinrichs, J., Rademacher, B., Kessler, H.G., 1996a. Food processing of milk products with ultra high pressure. In: Heat Treatments and Alternative Methods IDF Special Issue No. 9602. International Dairy Federation, Brussels, Belgium, pp. 185–201. Hinrichs, J., Rademacher, B., Kessler, H.G., 1996b. Reaction kinetics of pressure-induced denaturation of whey proteins. Milchwissenschaft 51, 504–509. Ho, S.Y., Mittal, G.S., Cross, J.D., 1997. Effects of high field electric pulses on the activity of selected enzymes. J. Food Eng. 31, 69–84. Hubbard, C.D., Caswell, D., L€ udemann, H.D., Arnold, M., 2002. Characterization of pressure-treated skimmed milk powder dispersions: application of NMR spectroscopy. J. Sci. Food Agric. 82, 1107–1114. Huppertz, T., 2010. High-pressure-processing of milk. In: Griffiths, M. (Ed.), Improving the Safety and Quality of Milk: Milk Production and Processing. Woodhead Publishing, Cambridge, UK, pp. 373–399. Huppertz, T., De Kruif, C.G., 2006. Disruption and reassociation of casein micelles under high pressure: influence of milk serum composition and casein micelle concentration. J. Agric. Food Chem. 54, 5903–5909. Huppertz, T., De Kruif, C.G., 2007a. High pressure-induced solubilisation of micellar calcium phosphate from crosslinked casein micelles. Colloids Surf. A 295, 264–268. Huppertz, T., De Kruif, C.G., 2007b. Disruption and reassociation of casein micelles during high pressure treatment: influence of whey proteins. J. Dairy Res. 74, 194–197. Huppertz, T., Kelly, A.L., Fox, P.F., 2002. Effects of high pressure on constituents and properties of milk. Int. Dairy J. 12, 561–572. Huppertz, T., Fox, P.F., Kelly, A.L., 2004a. High pressure treatment of bovine milk: effects of casein micelles and whey proteins. J. Dairy Res. 71, 97–106. Huppertz, T., Fox, P.F., Kelly, A.L., 2004b. High pressure-induced denaturation of α-lactalbumin and β-lactoglobulin in bovine milk and whey: a possible mechanism. J. Dairy Res. 71, 489–495. Huppertz, T., Fox, P.F., Kelly, A.L., 2004c. Properties of casein micelles in high pressure-treated bovine milk. Food Chem. 87, 103–110. Huppertz, T., Fox, P.F., de Kruif, K.G., Kelly, A.L., 2006a. High pressure-induced changes in bovine milk proteins: a review. Biochim. Biophys. Acta 1764, 593–598. Huppertz, T., Smiddy, M.A., Upadhyay, V.K., Kelly, A.L., 2006b. High pressure-induced changes in bovine milk: a review. Int. J. Dairy Technol. 59, 58–66. Huppertz, T., Kelly, A.L., de Kruif, C.G., 2006c. Disruption and reassociation of casein micelles under high pressure. J. Dairy Res. 73, 294–298. Huppertz, T., Smiddy, M.A., Goff, H.D., Kelly, A.L., 2011. Effects of high pressure treatment of mix on ice cream manufacture. Int. Dairy J. 21, 718–726.

320

8. Effect of nonthermal processing on milk protein interactions and functionality

Huppertz, T., 2013. Chemistry of the caseins. In: Advanced Dairy Chemistry. Springer, Boston, MA. pp. 135–160. Huppertz, T., Vasiljevic, T., Zisu, B., Deeth, H., 2019. Novel processing technologies: effects on whey protein structure and functionality. In: Deeth, H., Bansal, N. (Eds.), Whey Proteins: From Milk to Medicine. Academic Press, London, UK, pp. 281–334. Jambrak, A.R., Mason, T.J., Lelas, V., Herceg, Z., Herceg, I.L., 2008. Effect of ultrasound treatment on solubility and foaming properties of whey protein suspensions. J. Food Eng. 86 (2), 281–287. Jambrak, A.R., Lelas, V., Mason, T.J., Kresˇic, G., Badanjak, M., 2009. Physical properties of ultrasound treated soy proteins. J. Food Eng. 93 (4), 386–393. Jegouic, M., Grinberg, V.Y., Guingant, A., Haertle, T., 1997. Baric oligomerization in α-lactalbumin/β-lactoglobulin mixtures. J. Agric. Food Chem. 45, 19–22. Jelen, P., Rattray, W., 1995. Thermal denaturation of whey proteins. In: Fox, P.F. (Ed.), Heat-Induced Changes in Milk. International Dairy Federation, Brussels, Belgium, pp. 66–85. IDF Special Issue No. 9501. Kang, D.C., Zou, Y.H., Cheng, Y.P., Xing, L.J., Zhou, G.H., Zhang, W.G., 2016. Effects of power ultrasound on oxidation and structure of beef proteins during curing processing. Ultrason. Sonochem. 33, 47–53. Keenan, R.D., Young, D.J., Tier, C.M., Jones, A.D., Underdown, J., 2001. Mechanism of pressure-induced gelation of milk. J. Agric. Food Chem. 49, 3394–3402. Kentish, S., Ashokkumar, M., 2011. The physical and chemical effects of ultrasound. In: Ultrasound Technologies for Food and Bioprocessing. Springer, New York, NY, pp. 1–12. Khomich, V.A., Soloshenko, I.A., Tsiolko, V.V., Mikhno, I.L., 1998. Investigation of principal factors of the sterilization by plasma DC glow discharge. In: Proceedings of the International Congress on Plasma Physics, Prague, pp. 2745–2748. Klompong, V., Benjakul, S., Kantachote, D., Shahidi, F., 2007. Antioxidative activity and functional properties of protein hydrolysate of yellow stripe trevally (Selaroides leptolepis) as influenced by the degree of hydrolysis and enzyme type. Food Chem. 102, 1317–1327. Knudsen, J.C., Skibsted, L.H., 2009. High pressure effects on the structure of casein micelles in milk as studied by cryotransmission electron microscopy. Food Chem. 119, 202–208. Kobashigawa, Y., Sakurai, M., Nitta, K., 1999. Effect of hydrostatic pressure on unfolding of α-lactalbumin: volumetric equivalence of the molten globule and unfolded state. Protein Sci. 8, 2765–2772. Kristo, E., Hazizaj, A., Corredig, M., 2012. Structural changes imposed on whey proteins by UV irradiation in a continuous UV light reactor. J. Agric. Food Chem. 60 (24), 6204–6209. Kuan, Y.H., Bhat, R., Karim, A.A., 2011. Emulsifying and foaming properties of ultraviolet-irradiated egg white protein and sodium caseinate. J. Agric. Food Chem. 59 (8), 4111–4118. Kurukji, D., Norton, I., Spyropoulos, F., 2016. Fabrication of sub-micron protein-chitosan electrostatic complexes for encapsulation and pH-modulated delivery of model hydrophilic active compounds. Food Hydrocoll. 53, 249–260. Kuwajima, K., Ikeguchi, M., Sugawara, T., Hiraoka, Y., Sugai, S., 1990. Kinetics of disulfide bond reduction in α-lactalbumin by dithiothreitol and molecular basis of superreactivity of the Cys6-Cys120 disulfide bond. Biochemistry 29, 8240–8249. Lassalle, M.W., Li, H., Yamada, H., Akasaka, K., Redfield, C., 2003. Pressure-induced unfolding of the molten globule of all-Ala α-lactalbumin. Protein Sci. 12, 66–72. Li, S.Q., Zhang, Q.H., Lee, Y.Z., Pham, T.V., 2003. Effects of pulsed electric fields and thermal processing on the stability of bovine immunoglobulin G (IgG) in enriched soymilk. J. Food Sci. 68 (4), 1201–1207. Li, S.Q., Bomser, J.A., Zhang, Q.H., 2005. Effects of pulsed electric fields and heat treatment on stability and secondary structure of bovine immunoglobulin G. J. Agric. Food Chem. 53 (3), 663–670. Li-Chan, E., 1999. Hydrophobicity in food protein systems. In: Francis, F.J. (Ed.), Encyclopedia of Food Science and Technology, second ed. Wiley, New York, NY, pp. 1429–1439. Liu, Z., Juliano, P., Williams, R.P., Niere, J., Augustin, M.A., 2014. Ultrasound improves the renneting properties of milk. Ultrason. Sonochem. 21 (6), 2131–2137. Liu, Z., Hemar, Y., Tan, S., Sanguansri, P., Niere, J., Buckow, R., Augustin, M.A., 2015. Pulsed electric field treatment of reconstituted skim milks at alkaline pH or with added EDTA. J. Food Eng. 144, 112–118. Loey, A.V., Verachtert, B., Hendrickx, M., 2002. Effects of high electric field pulses on enzymes. Trends Food Sci. Technol. 12, 94–102. Lo´pez-Fandin˜o, R., 2006a. High pressure-induced changes in milk proteins and possible applications in dairy technology. Int. Dairy J. 16, 1119–1131.

References

321

Lo´pez-Fandin˜o, R., 2006b. Functional improvement of milk whey proteins induced by high hydrostatic pressure. Crit. Rev. Food Sci. Nutr. 48, 351–363. Lo´pez-Fandin˜o, R., Olano, A., 1998. Effects of high pressures combined with moderate temperatures on the rennet coagulation properties of milk. Int. Dairy J. 8, 623–627. Lo´pez-Fandin˜o, R., Carrascosa, A.V., Olano, A., 1996. The effects of high pressure on whey protein denaturation and cheese-making properties of raw milk. J. Dairy Sci. 79, 929–936. Lo´pez-Fandin˜o, R., Ramos, M., Olano, A., 1997. Rennet coagulation of milk subjected to high pressures. J. Agric. Food Chem. 45, 3233–3237. Madadlou, A., Mousavi, M.E., Emam-djomeh, Z., Ehsani, M., Sheehan, D., 2009. Sonodisruption of re-assembled casein micelles at different pH values. Ultrason. Sonochem. 16 (5), 644–648. Mahmoud, N.S., Ghaly, A.E., 2004. On-line sterilization of cheese whey using ultraviolet radiation. Biotechnol. Prog. 20 (2), 550–560. Malaki Nik, A., Wright, A.J., Corredig, M., 2011. Micellization of beta-carotene from soy-protein stabilized oil-inwater emulsions under in vitro conditions of lipolysis. J. Am. Oil Chem. Soc. 88 (9), 1397–1407. Martini, S., Potter, R., Walsh, M.K., 2010. Optimizing the use of power ultrasound to decrease turbidity in whey protein suspensions. Food Res. Int. 43 (10), 2444–2451. Matak, K.E., Sumner, S.S., Duncan, S.E., Hovingh, E., Worobo, R.W., Hackney, C.R., Pierson, M.D., 2007. Effects of ultraviolet irradiation on chemical and sensory properties of goat milk. J. Dairy Sci. 90 (7), 3178–3186. McCarthy, N.A., Kelly, P.M., Maher, P.G., Fenelon, M.A., 2014. Dissolution of milk protein concentrate (MPC) powders by ultrasonication. J. Food Eng. 126, 142–148. McDonald, K.F., Curry, R.D., Clevenger, T.E., Brazos, B.J., Unklesbay, K., Eisenstark, A., Baker, S., Golden, J., Morgan, R., 2000a. The development of photosensitized pulsed and continuous ultraviolet decontamination techniques for surfaces and solutions. IEEE Trans. Plasma Sci. 28 (1), 89–96. McDonald, K., Clevenger, T., Curry, R., Golden, J., 2000b. Ultraviolet light technologies for water treatment. In: Environmental and Pipeline Engineering. ASCE, Reston, VA, pp. 299–311. Merel-Rausch, E., Duma, I., Hinrichs, J., 2006. Pressure-induced modification of casein micelles. Influence of pressure built-up rate, pressure level, release rate and temperature on viscosity and particle size. Michwissenschaft 61, 255–259. Messens, W., Van Camp, J., Huyghebaert, A., 1997. The use of high pressure to modify the functionality of food proteins. Trends Food Sci. Technol. 8, 107–112. Meulemans, C.C.E., 1987. The basic principles of UV–disinfection of water. Ozone Sci. Eng. 9, 299–314. Michalac, S., Alvarez, V., Ji, T., Zhang, Q.H., 2003. Inactivation of selected micro-organisms and properties of pulsed electric field processed milk. J. Food Process. Preserv. 27 (2), 137–151. Miller, R.V., Jeffrey, W., Mitchell, D., Elasri, M., 1999. Bacterial responses to ultraviolet light. ASM News—American Society for Microbiology 65, 535–541. Mimouni, A., Deeth, H.C., Whittaker, A.K., Gidley, M.J., Bhandari, B.R., 2010. Investigation of the microstructure of milk protein concentrate powders during rehydration: alterations during storage. J. Dairy Sci. 93 (2), 463–472. Mirmoghtadaie, L., Aliabadi, S.S., Hosseini, S.M., 2016. Recent approaches in physical modification of protein functionality. Food Chem. 199, 619–627. Misra, N.N., Koubaa, M., Roohinejad, S., Juliano, P., Alpas, H., Ina´cio, R.S., Saraiva, J.A., Barba, F.J., 2017. Landmarks in the historical development of twenty first century food processing technologies. Food Res. Int. 97, 318–339. Morales, R., Martı´nez, K.D., Ruiz-Henestrosa, V.M.P., Pilosof, A.M., 2015. Modification of foaming properties of soy protein isolate by high ultrasound intensity: particle size effect. Ultrason. Sonochem. 26, 48–55. Morild, E., 1981. The theory of pressure effects on enzymes. In: Advances in Protein Chemistry. vol. 34. Academic Press, pp. 93–166. Moro, A., Gatti, C., Delorenzi, N., 2001. Hydrophobicity of whey protein concentrates measured by fluorescence quenching and its relation with surface functional properties. J. Agric. Food Chem. 49 (10), 4784–4789. Mu´jica-Paz, H., Valdez-Fragoso, A., Samson, C.T., Welti-Chanes, J., Torres, J.A., 2011. High-pressure processing technologies for the pasteurization and sterilization of foods. Food Bioprocess Technol. 4 (6), 969–985. Mukhopadhyay, S., Tomasula, P.M., Luchansky, J.B., Porto-Fett, A., Call, J.E., 2010. Removal of Salmonella enteritidis from commercial unpasteurized liquid egg white using pilot scale cross flow tangential microfiltration. Int. J. Food Microbiol. 142 (3), 309–317. Mulcahy, E.M., Mulvihill, D.M., O’Mahony, J.A., 2016. Physicochemical properties of whey protein conjugated with starch hydrolysis products of different dextrose equivalent values. Int. Dairy J. 53, 20–28.

322

8. Effect of nonthermal processing on milk protein interactions and functionality

Needs, E.C., Capellas, M., Bland, A.P., Manoj, P., MacDougal, D., Paul, G., 2000. Comparison of heat and pressure treatments of skim milk, fortified with whey protein concentrate, for set yogurt preparation: effects of milk proteins and gel structure. J. Dairy Res. 67, 329–348. Ngadi, M.O., Latheef, M.B., Kassama, L., 2012. Emerging technologies for microbial control in food processing. In: Green Technologies in Food Production and Processing. Springer, Boston, MA, pp. 363–411. O’Connell, J.E., Flynn, C., 2007. The manufacture and applications of casein-derived ingredients. In: Hui, Y.H. (Ed.), Handbook of Food Products Manufacturing. first ed. John Wiley & Sons, New York, NY, pp. 557–591. Odriozola-Serrano, I., Bendicho-Porta, S., Martı´n-Belloso, O., 2006. Comparative study on shelf life of whole milk processed by high-intensity pulsed electric field or heat treatment. J. Dairy Sci. 89 (3), 905–911. Oldfield, D.J., Singh, H., Taylor, M.W., Pearce, K.N., 1998. Kinetics of denaturation and aggregation of whey proteins in skim milk heated in an ultra-high temperature (UHT) pilot plant. Int. Dairy J. 8, 311–318. Oldfield, D.J., Singh, H., Taylor, M.W., Pearce, K.N., 2000. Heat-induced interactions of β-lactoglobulin and α-lactalbumin with the casein micelle in pH-adjusted skim milk. Int. Dairy J. 10 (8), 509–518. Orlien, V., Knudsen, J.C., Colon, M., Skibsted, L.H., 2006. Dynamics of casein micelles in skim milk during and after high pressure treatment. Food Chem. 98, 513–521. O’Sullivan, J.J., O’Mahony, J.A., 2016. Food ingredients. In: Reference Module in Food Science. Elsevier, Amsterdam, Netherlands, pp. 1–3. O’Sullivan, J., Arellano, M., Pichot, R., Norton, I., 2014. The effect of ultrasound treatment on the structural, physical and emulsifying properties of dairy proteins. Food Hydrocoll. 42 (3), 386–396. O’Sullivan, J.J., Kurukji, D., Norton, I.T., Spyropoulos, F., 2017a. Investigation of the fabrication and subsequent emulsifying capacity of potato protein isolate/κ-carrageenan electrostatic complexes. Food Hydrocoll. 71, 282–289. O’Sullivan, J.J., Park, M., Beevers, J., Greenwood, R.W., Norton, I.T., 2017b. Applications of ultrasound for the functional modification of proteins and nanoemulsion formation: a review. Food Hydrocoll. 71, 299–310. Patel, H.A., Huppertz, T., 2014. Effects of high-pressure processing on structure and interactions of milk proteins. In: Boland, M., Singh, H., Thompson, A. (Eds.), Milk Proteins: From Expression to Food, second ed. Academic Press, New York, NY, pp. 243–267. Patel, H.A., Singh, H., Anema, S.G., Creamer, L.K., 2006. Effects of heat and high hydrostatic pressure treatments on disulfide bonding interchanges among the proteins in skim milk. J. Agric. Food Chem. 54, 3409–3420. Pattison, D.I., Rahmanto, A.S., Davies, M.J., 2012. Photo-oxidation of proteins. Photochem. Photobiol. Sci. 11 (1), 38–53. Pereira, R.N., Vicente, A.A., 2010. Environmental impact of novel thermal and non-thermal technologies in food processing. Food Res. Int. 43 (7), 1936–1943. Perez, O.E., Pilosof, A.M.R., 2004. Pulsed electric fields effects on the molecular structure and gelation of β-lactoglobulin concentrate and egg white. Food Res. Int. 37, 102–110. Payens, T.A.J., Heremans, K., 1969. Effect of pressure on the temperature-dependent association of β-casein. Biopolymers 8 (3), 335–345. Qin, B., Zhang, Q., Barbosa-Ca´novas, G.V., Swanson, B.G., Pedrow, P.D., 1995. Pulsed electric field treatment chamber design for liquid food pasteurization using a finite element method. Trans. ASAE 38 (2), 557–565. Qin, B., Pothakamury, U., Barbosa-Ca´novas, G.V., Swanson, B., 1996. Nonthermal pasteurization of liquid foods using high-intensity pulsed electric fields. Crit. Rev. Food Sci. Nutr. 36 (6), 603–627. Quass, D.W., 1997. Pulsed Electric Field Processing in the Food Industry. A Status Report on Pulsed Electric Field. CR-109742, Electric Power Research Institute, Palo Alto, CA, pp. 23–35. Redecke, L., Binder, S., Elmallah, M.I., Broadbent, R., Tilkorn, C., Schulz, B., May, P., Goos, A., Eich, A., R€ ubhausen, M., Betzel, C., 2009. UV-light-induced conversion and aggregation of prion proteins. Free Radic. Biol. Med. 46 (10), 1353–1361. Regnault, S., Thiebaud, M., Dumay, E., Cheftel, J.C., 2004. Pressurisation of raw skim milk and of a dispersion of phosphocaseinate at 9°C or 20°C: effects on casein micelle size distribution. Int. Dairy J. 14, 55–68. Regnault, S., Dumay, E., Cheftel, J.C., 2006. Pressurisation of raw skim milk and of a dispersion of phosphocaseinate at 9°C or 20°C: effects on the distribution of minerals and proteins between colloidal and soluble phases. J. Dairy Res. 73, 91–100. Richard, B., Le Page, J.F., Schuck, P., Andre, C., Jeantet, R., Delaplace, G., 2013. Towards a better control of dairy powder rehydration processes. Int. Dairy J. 31 (1), 18–28.

References

323

Saffon, M., Britten, M., Pouliot, Y., 2011. Thermal aggregation of whey proteins in the presence of buttermilk concentrate. J. Food Eng. 103 (3), 244–250. Sale, A., Hamilton, W., 1967. Effect of high electric fields on microorganisms. I. Killing of bacteria and yeast. Biochim. Biophys. Acta 148, 781–788. Scheidegger, D., Pecora, R.P., Radici, P.M., Kivatinitz, S.C., 2010. Protein oxidative changes in whole and skim milk after ultraviolet or fluorescent light exposure. J. Dairy Sci. 93 (11), 5101–5109. Schmidt, D.G., Payens, T.A.J., 1972. The evaluation of positive and negative contributions to the second virial coefficient of some milk proteins. J. Colloid Interface Sci. 39 (3), 655–662. Scollard, P.G., Beresford, T.P., Needs, E.C., Murphy, P.M., Kelly, A.L., 2000. Plasmin activity, β-lactoglobulin denaturation and proteolysis in high pressure treated milk. Int. Dairy J. 10, 835–841. Semagoto, H.M., Liu, D., Koboyatau, K., Hu, J., Lu, N., Liu, X., Regenstein, J., Zhou, P., 2014. Effects of UV induced photo-oxidation on the physicochemical properties of milk protein concentrate. Food Res. Int. 62, 580–588. Shamsi, K., Versteeg, C., Sherkat, F., Wan, J., 2008. Alkaline phosphatase and microbial inactivation by pulsed electric field in bovine milk. Innov. Food Sci. Emerg. Technol. 9, 217–223. Shanmugam, A., Chandrapala, J., Ashokkumar, M., 2012. The effect of ultrasound on the physical and functional properties of skim milk. Innov. Food Sci. Emerg. Technol. 16, 251–258. Sharma, P., Oey, I., Everett, D.W., 2014. Effect of pulsed electric field processing on the functional properties of bovine milk. Trends Food Sci. Technol. 35 (2), 87–101. Sharma, P., Oey, I., Everett, D.W., 2016. Thermal properties of milk fat, xanthine oxidase, caseins and whey proteins in pulsed electric field-treated bovine whole milk. Food Chem. 207, 34–42. Sharma, P., Oey, I., Bremer, P., Everett, D.W., 2018. Microbiological and enzymatic activity of bovine whole milk treated by pulsed electric fields. Int. J. Dairy Technol. 71 (1), 10–19. Silva, E., De Landea, C., Edwards, A.M., Lissi, E., 2000. Lysozyme photo-oxidation by singlet oxygen: properties of the partially inactivated enzyme. J. Photochem. Photobiol. B Biol. 55 (2–3), 196–200. Singh, H., 1995. Heat-induced changes in casein, including interactions with whey proteins. In: Fox, P.F. (Ed.), Heat Induced Changes in Milk IDF Special Issue No. 9501. International Dairy Federation, Brussels, Belgium, pp. 86–104. Singh, H., Waungana, A., 2001. Influence of heat treatment of milk on cheesemaking properties. Int. Dairy J. 11 (4–7), 543–551. Sizer, C.E., Balasubramaniam, V.M., 1999. New intervention processes for minimally processed juices. Food Technol. 53 (10), 64–67. Snowball, M.R., Hornsey, I.S., 1988. Purification of water supplies using ultraviolet light. In: Developments in Food Microbiology. vol. 3. Elsevier Applied Science Publishers, London, UK, pp. 171–191. Speroni, F., Beaumal, V., De Lamballerie, M., Anton, M., An˜o´n, M.C., Puppo, M.C., 2009. Gelation of soybean proteins induced by sequential high-pressure and thermal treatments. Food Hydrocoll. 23 (5), 1433–1442. Stathopulos, P.B., Scholz, G.A., Hwang, Y.M., Rumfeldt, J.A., Lepock, J.R., Meiering, E.M., 2004. Sonication of proteins causes formation of aggregates that resemble amyloid. Protein Sci. 13 (11), 3017–3027. Stippl, V.M., Delgado, A., Becker, T.M., 2005. Ionization equilibria at high pressure. Eur. Food Res. Technol. 221, 151–156. Sui, Q., Roginski, H., Williams, R.P., Versteeg, C., Wan, J., 2011. Effect of pulsed electric field and thermal treatment on the physicochemical and functional properties of whey protein isolate. Int. Dairy J. 21 (4), 206–213. Tammineedi, C.V., Choudhary, R., Perez-Alvarado, G.C., Watson, D.G., 2013. Determining the effect of UV-C, high intensity ultrasound and nonthermal atmospheric plasma treatments on reducing the allergenicity of α-casein and whey proteins. LWT–Food Sci. Technol. 54 (1), 35–41. Toepfl, S., Mathys, A., Heinz, V., Knorr, D., 2006. Potential of emerging technologies for energy efficient and environmentally friendly food processing. Food Rev. Int. 22 (4), 405–423. Toepfl, S., Heinz, V., Knorr, D., 2007. High intensity pulsed electric fields applied for food preservation. Chem. Eng. Process. Process Intensif. 46 (6), 537–546. Torregrosa, F., Cortes, C., Esteve, M.J., Frı´gola, A., 2005. Effect of high-intensity pulsed electric fields processing and conventional heat treatment on orange–carrot juice carotenoids. J. Agric. Food Chem. 53 (24), 9519–9525. Ugarte-Romero, E., Feng, H., Martin, S.E., 2007. Inactivation of Shigella boydii 18 IDPH and Listeria monocytogenes Scott A with power ultrasound at different acoustic energy densities and temperatures. J. Food Sci. 72 (4), M103–M107.

324

8. Effect of nonthermal processing on milk protein interactions and functionality

Vega-Mercado, H.M., Martı´n-Belloso, O., Qin, B.L., Chang, F.J., Go´ngora-Nieto, M.M., Barbosa-Ca´novas, G.V., Swanson, B.G., 1997. Non-thermal food preservation: pulsed electric fields. Trends Food Sci. Technol. 8 (5), 151–157. Walstra, P., van Vliet, T., 2003. Functional properties. Prog. Biotechnol. 23, 9–30. Wang, Z., Li, Y., Jiang, L., Qi, B., Zhou, L., 2014. Relationship between secondary structure and surface hydrophobicity of soybean protein isolate subjected to heat treatment. J. Chem. 2014, 1–10. Xiang, B.Y., Ngadi, M.O., Ochoa-Martinez, L.A., Simpson, M.V., 2011a. Pulsed electric field-induced structural modification of whey protein isolate. Food Bioprocess Technol. 4 (8), 1341–1348. Xiang, B.Y., Simpson, M.V., Ngadi, M.O., Simpson, B.K., 2011b. Effect of pulsed electric field on the rheological and colour properties of soy milk. Int. J. Food Sci. Nutr. 62 (8), 787–793. Xiang, B.Y., Simpson, M.V., Ngadi, M.O., Simpson, B.K., 2011c. Flow behaviour and viscosity of reconstituted skimmed milk treated with pulsed electric field. Biosyst. Eng. 109 (3), 228–234. Yanjun, S., Jianhang, C., Shuwen, Z., Hongjuan, L., Jing, L., Lu, L., Uluko, H., Yanling, S., Wenming, C., Wupeng, G., Jiaping, L., 2014. Effect of power ultrasound pre-treatment on the physical and functional properties of reconstituted milk protein concentrate. J. Food Eng. 124, 11–18. Yong, Y.H., Yamaguchi, S., Matsumura, Y., 2006. Effects of enzymatic deamidation by protein-glutaminase on structure and functional properties of wheat gluten. J. Agric. Food Chem. 54 (16), 6034–6040. Yu, L.J., Ngadi, M., Raghavan, G.S.V., 2009. Effect of temperature and pulsed electric field treatment on rennet coagulation properties of milk. J. Food Eng. 95 (1), 115–118. Zhang, R., Pang, X., Lu, J., Liu, L., Zhang, S., Lv, J., 2018. Effect of high intensity ultrasound pretreatment on functional and structural properties of micellar casein concentrates. Ultrason. Sonochem. 47, 10–16. Zhao, W., Yang, R., Zhang, H.Q., 2012. Recent advances in the action of pulsed electric fields on enzymes and food component proteins. Trends Food Sci. Technol. 27 (2), 83–96. Zhong, K., Hu, X., Chen, F., Wu, J., Liao, X., 2005. Inactivation and conformational change of pectinesterase induced by pulsed electric field. Trans. CSAE 21 (2), 149–152. Zisu, B., Lee, J., Chandrapala, J., Bhaskaracharya, R., Palmer, M., Kentish, S., Ashokkumar, M., 2011. Effect of ultrasound on the physical and functional properties of reconstituted whey protein powders. J. Dairy Res. 78 (2), 226–232. Zobrist, M.R., Huppertz, T., Uniacke, T., Fox, P.F., Kelly, A., 2005. High pressure-induced changes in rennetcoagulation properties of bovine milk. Int. Dairy J. 15, 655–662.

Further reading Buhler, S., Solari, F., Gasparini, A., Montanari, R., Sforza, S., Tedeschi, T., 2019. UV irradiation as a comparable method to thermal treatment for producing high quality stabilized milk whey. LWT–Food Sci. Technol. 105, 127–134. Zhao, W., Tang, Y., Lu, L., Chen, X., Li, C., 2014. Pulsed electric fields processing of protein-based foods. Food Bioprocess Technol. 7 (1), 114–125.

C H A P T E R

9 The whey proteins in milk: Thermal denaturation, physical interactions, and effects on the functional properties of milk Skelte G. Anema Fonterra Research and Development Centre, Palmerston North, New Zealand

Introduction Milk is produced in the mammary gland of female mammals and is intended for the feeding of the neonate from birth to weaning. Milk is a highly nutritious, readily digested food that is rich in protein, minerals, and energy in an aqueous solution. It also provides the neonate with many other essential compounds such as protective agents, hormones, and growth factors. Milk is a highly perishable fluid and was intended by nature to be consumed soon after production. Humans have used milk and dairy-derived foods to supplement their diet for centuries, and dairy products are still a major food source. Because of the commercial and nutritional significance of dairy products, manufacturing processes to preserve the food value of milk long after its initial production have been developed. Over the last century, modern dairy milk processing has been transformed from an art into a science. Traditional products such as cheeses and yogurts combine century-old knowledge with modern science, technology, and processing techniques. In contrast, more recently developed products (such as spray-dried milk products, milk protein concentrates, micellar casein, and whey protein concentrates) have been based on modern technologies of the time. The majority of the milk processed is of bovine (cow) origin; however, significant quantities of buffalo, goat, and sheep milk are also manufactured into dairy products (Fox, 2003). In milk, the lactose, some of the mineral components, and the native whey proteins are in true molecular solution. However, the casein and most of the calcium and phosphate are

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00009-8

325

# 2020 Elsevier Inc. All rights reserved.

326

9. The whey proteins in milk

found in large macromolecular assemblies called casein micelles. The colloidal suspension of casein micelles in milk serum is a remarkably stable food protein system. Milk can be subjected to high temperatures and pressures, high shear, and variations in concentration without appreciable damage to the casein micelle system. Even the extreme action of drying milk to a powder does not significantly alter the milk system, as milk powders can be reconstituted to produce liquid milks that have many properties similar to those of the milk from which they were derived (Singh and Newstead, 1992; Kelly et al., 2003; Nieuwenhuijse and van Boekel, 2003; O’Connell and Fox, 2003).

The casein micelle To understand and rationalize any changes to the properties and stability of milk, it is necessary to have some knowledge of the casein micelle structure. Despite extensive research efforts, the detailed structure and assembly of the casein micelle has not been unequivocally established. Several models have been proposed over the years, and these models have been progressively updated or modified as more information on the casein micelle has become available (Fig. 9.1; Schmidt, 1982; Walstra, 1990; Holt, 1992; Holt and Horne, 1996; Horne, 1998; Walstra, 1999; Dalgleish, 2011; Huppertz et al., 2017). Although there is some agreement over some aspects of the various models, there is still considerable debate over the detailed structure of the casein micelle, as evidenced from the numerous recent papers and reviews devoted to this subject (Farrell et al., 2006; Qi, 2007, 2009; Dalgleish, 2011; Trejo et al., 2011; Dalgleish and Corredig, 2012; de Kruif et al., 2012; Holt et al., 2013; Horne, 2017; Huppertz et al., 2017; Lucey and Horne, 2018; also see Chapter 6). Evidence from early electron microscopy and light scattering studies suggested that the casein micelle was assembled from smaller subunits, and as a consequence, submicelle models of the casein micelle structure were developed. In the later iterations of these submicelle models, the casein proteins were hydrophobically aggregated to form the submicelle units, and these submicelle units were linked by colloidal calcium phosphate (CCP) via serine phosphate clusters at the submicelle surfaces to form the casein micelle. The distribution of κ-casein between submicelles was heterogeneous, and submicelles with high levels of κ-casein were located at the micelle surface, whereas those with low levels of κ-casein were in the interior, thus giving a surface location to κ-casein that was consistent with experiment (Fig. 9.1A; Schmidt, 1982; Walstra, 1990). Subsequent experimental evidence did not support the existence of submicelles, and therefore, the validity of the submicelle model of the casein micelles was questioned (Holt, 1992; Holt and Horne, 1996; Horne, 1998; Walstra, 1999; Horne, 2002, 2006). In particular, there was evidence to show that the CCP was uniformly distributed through the casein micelle, which precluded submicelles being linked by CCP to form the micelle. In addition, it was considered to be unlikely that there would be heterogeneous populations of casein submicelles with different levels of κ-casein or that assembly into casein micelles via calcium phosphate would occur only after the casein submicelles had been formed. In addition, electron micrograph images of casein micelles using modern techniques did not display the internal substructure expected for casein submicelles, and it was considered

(A)

(B)

(C)

(D)

(E)

(F)

FIG. 9.1 Recent models of the casein micelle. (A) Original “hairy” submicelle model of the casein micelle. (B) Modified hairy submicelle model of the casein micelle. (C) Nanocluster model of the casein micelle. (D) Dualbinding model of the casein micelle. (E) Dalgleish model of the casein micelle. (F) Huppertz model of the casein micelle (Huppertz et al., 2017). (A) Reproduced with permission from Walstra, P., Jenness, R., 1984. Dairy chemistry and physics. In: Dairy Chemistry and Physics. John Wiley and Sons, New York and Walstra, P., 1990. On the stability of casein micelles. J. Dairy Sci. 73, 1965–1979. Copyright (1984) John Wiley & Sons. (B) Reproduced with permission from Walstra, P., 1999. Casein sub-micelles: do they exist? Int. Dairy J. 9, 189–192. Copyright (1999) Elsevier. (C) Reproduced with permission from Holt, C. (1992). Structure and stability of bovine casein micelles. Adv. Protein Chem. 43, 63–151. Copyright (1992) Elsevier. (D) Reproduced with permission from Horne, D.S., 1998. Casein interactions: casting light on the black boxes, the structure in dairy products. Int. Dairy J. 8, 171–177. Copyright (1998) Elsevier. (E) Reproduced with permission from Dalgleish, D.G., 2011. On the structural models of bovine casein micelles-review and possible improvements. Soft Matter, 7, 2265–2272. Copyright (2011) Royal Society of Chemistry. (F) Used with permission.

328

9. The whey proteins in milk

that the appearance of submicelles in earlier micrographs was an artifact of the early preparation techniques for electron microscopy (McMahon and McManus, 1998; Horne, 2006). In an attempt to reconcile this new evidence, the submicelle model of the casein micelle was refined to change the role of CCP from that of linking the submicelles to a charge-neutralizing agent to allow for a uniform distribution of CCP, and the submicelles were now linked together via hydrophobic interactions (Fig. 9.1B; Walstra, 1999). However, new models for the casein micelle that did not rely on the formation of submicelles were developed (Holt, 1992; Holt and Horne, 1996; Horne, 1998; de Kruif and Holt, 2003; Dalgleish, 2011). Nonsubmicelle models included the nanocluster model (Fig. 9.1C; Holt, 1992; Holt and Horne, 1996; de Kruif and Holt, 2003) and the dual-binding model (Fig. 9.1D; Horne, 1998). There has been some convergence of these two models as the combined results of detailed experiments on micelle structure using small angle x-ray/ neutron scattering and static light scattering were consistent with calculation only if weak interactions (hydrophobic interactions, hydrogen bonding, ion pairing, etc.) were incorporated into the nanocluster model (de Kruif et al., 2012). The dual-binding model proposed by Horne (1998) describes the types of interactions involved in the assembly of casein micelles and demonstrates that micelles with a consistent arrangement of casein proteins and CCP can be achieved. However, this dual-binding model does not give a detailed description of the surface or the internal structure of the casein micelle. A recent model of the casein micelle proposed by Dalgleish (2011) and Dalgleish and Corredig (2012) (Fig. 9.1E) has a relatively sparse hairy layer of κ-casein on the surface, dense enough to stabilize against approach by other casein micelles or other large colloidal particles but sufficiently diffuse to allow denatured whey proteins to interact with the para-κ-casein region of κ-casein or to allow β-casein to dissociate from the micelles on cooling and to reassociate with the micelles on subsequent warming. In addition, this model attempts to reconcile the structural arrangement of the interior of the micelles with the high hydration of the micelle by giving a specific role to β-casein. It is proposed that some of the β-casein acts as a surfactant in stabilizing the hydrated internal channels of the micelle. This β-casein may be loosely bound to the micelles so that it is able to dissociate and reassociate during cold and warm temperature cycling. Very recently, Duerasch et al. (2018) presented a casein micelle model with a similar structure to that proposed by Dalgleish (2011) and Dalgleish and Corredig (2012); however, they suggest that “αS2-casein must be the major compound of calcium phosphate nanoclusters forming the inner structure of casein micelles” (Duerasch et al., 2018). It is highly unlikely that αS2-casein has greater importance than αS1-casein in casein micelle structure as, in bovine milk, αS1-casein is more dominant than αS2-casein, and αS1-casein can bind high levels of calcium phosphate (Swaisgood, 1982, 1992, 1993). In addition, the low level of αS2-casein in bovine milk would require this protein to bind large amounts of calcium phosphate if it was the major casein of the calcium phosphate nanoclusters. The fact that bovine milk and caprine milk may have vastly different αS1-casein and αS2-casein levels yet similar sized casein micelles (Richardson et al., 1974; Anema and Stanley, 1998; Raynal and Remeuf, 1998; Tziboula and Horne, 1999) suggests that αS1-casein and αS2-casein play similar roles in casein micelle structure and, in particular, similar roles in stabilizing the colloidal calcium phosphate nanoclusters.

The casein micelle

329

Another very recent model of the casein micelle has a structure that is composed of nonspherical subparticles, referred to as “primary casein particles” (PCPs) rather than submicelles (Fig. 9.1F; Huppertz et al., 2017). In many respects, this model is similar to that described by Schmidt (1982) and Walstra (1990), with the core of the micelle being composed of PCPs depleted in κ-casein (and enriched in αS2-casein) and the surface of the micelle being composed of PCPs enriched in κ-casein (and depleted in αS2-casein). The PCPs are cross-linked into a three-dimensional structure by the CCP through the phosphoserine centers on the αS1-casein, αS2-casein, and β-casein. The primary differences between the model by Huppertz et al. (2017) and the Schmidt/Walstra models (Schmidt, 1982; Walstra, 1990) are that the PCPs are nonspherical whereas the submicelles in the earlier models were spherical and that the PCPs are somewhat smaller, containing about 9–11 casein molecules, whereas the submicelles were considered to contain about 15–25 casein molecules. In a review on submicelle models of the casein micelle, Walstra (1999) indicated that experimental evidence was strongly against the concept of submicelles being linked via calcium phosphate bridges; however, this concept is being reintroduced in this latest casein micelle model. From the regular emergence of new casein micelle models and the continuing debate on casein protein interactions with respect to casein micelles (e.g., Carver et al., 2017; Horne and Lucey, 2017), it is clear that there is still not universal agreement on the casein micelle structure. There are proponents of submicelle models, and convincing arguments based on a structural biology perspective have been presented (Farrell et al., 2006; Qi, 2007). Even within the groups that support the models without submicelles, there are diverse views on the structural arrangements and the relative importance of different types of bonding (Horne, 2006; McMahon and Oommen, 2008; Dalgleish, 2011; Dalgleish and Corredig, 2012; de Kruif et al., 2012; Carver et al., 2017; Horne and Lucey, 2017). Although various models of the casein micelles have been proposed, they have largely been derived from the same pool of research data and are therefore different depictions or interpretations of similar information. As a consequence, many of the salient features of the structure, assembly, and stability of the different models are similar (Schmidt, 1982; Walstra, 1990; Holt and Horne, 1996; Horne, 1998; Walstra, 1999; Dalgleish, 2011; Dalgleish and Corredig, 2012; de Kruif et al., 2012; Huppertz et al., 2017). Hydrophobic interactions and CCP are important in maintaining micelle integrity. Therefore, micelle integrity can be modified or destroyed by disruption to hydrophobic interactions or by the dissolution of the CCP. In all recent models, κ-casein has a preferential surface location, with the C-terminal region protruding from the surface layer as a flexible hair. These models of the casein micelles, with the surface layers of κ-casein and an internal structure maintained by hydrophobic interactions and CCP, have been used to explain micelle stability and the destabilization by the enzymes in rennet, by acidification, or by the addition of alcohol (Walstra, 1990; Holt and Horne, 1996; Horne, 1998, 2003). However, less studied and less well understood are the mechanisms responsible for the changes occurring to the casein micelles during the heating of milk, in particular the interactions with the denatured whey proteins, the heat-induced, pH-dependent dissociation of the casein (especially κ-casein) from the micelles, and the eventual heat-induced coagulation of the casein micelles.

330

9. The whey proteins in milk

The heat treatment of milk The effect of heat on the milk system is an important consideration in dairy chemistry as a heat treatment is involved in the manufacture of almost all milk products. The heat treatment may range from thermization (about 65°C for 15 s) to sterilization (about 120°C for 10–20 min) or ultrahigh-temperature (UHT) treatment (typically 138–142°C for several seconds). As the thermal history of milk influences its behavior in subsequent applications, the effects of heat on milk have been the subject of intensive, if somewhat intermittent, research, and many reviews and books on the subject are available (Singh and Creamer, 1992; International-DairyFederation, 1995, 1996; O’Connell and Fox, 2003; Singh, 2004). When milk is heated, a number of competitive and often interdependent reactions occur, and the importance of each reaction is determined by the heating conditions and by factors such as milk composition or concentration. When considering the protein components of milk, reactions of particular importance are whey protein denaturation, the interactions of denatured whey proteins with other proteins (including those of the casein micelles), and casein micelle dissociation. These three reaction processes can markedly modify the physicochemical properties of milk and may play a major role in determining the stability of milk and the functional performance of heated milk products.

Whey protein denaturation The whey proteins are typical globular proteins with well-defined secondary and tertiary structures. They (especially α-lactalbumin and β-lactoglobulin) retain their native conformations only within relatively limited temperature ranges. Exposing the whey proteins to extremes of temperature results in the denaturation and aggregation of the proteins; this process can be expressed using the simple reaction scheme as shown in Eq. (9.1): ðPN Þn ⇆nPN

(9.1a)

PN ⇆PU

(9.1b)

PU + A ! ðP  AÞ

(9.1c)

For protein species where the native protein is in the form of noncovalently linked oligomers (such as dimeric β-lactoglobulin), the first step in the denaturation process is the reversible dissociation of the oligomer into monomeric species (Eq. 9.1a). The monomeric protein can then unfold, disrupting the native conformation (Eq. 9.1b). In principle, this unfolding step is reversible; however, in complex mixtures such as milk, the unfolding process is accompanied by the exposure of reactive amino side chain groups, which allows irreversible aggregation reactions to occur. The unfolded whey protein can undergo aggregation reactions with other (unfolded) whey proteins, with aggregates, or with the casein micelles (represented by A in Eq. 9.1c). At a fundamental level, protein denaturation is often defined as any noncovalent change to the secondary or tertiary structure of the protein molecule (Eq. 9.1b). From this denatured state, the protein can revert to its native state (refold) or interact with other components in the system (aggregate). Under this definition, α-lactalbumin is generally regarded as one

The heat treatment of milk

331

of the most heat-labile whey proteins, whereas β-lactoglobulin is one of the most heat-stable whey proteins (Ruegg et al., 1977). However, for the dairy industry, it is the irreversible aggregation processes that largely determine the functional properties of dairy products. Hence, it is common practice to define whey protein denaturation as the formation of irreversibly denatured and aggregated whey proteins (Sanderson, 1970a; Singh and Newstead, 1992; Kelly et al., 2003), and therefore, this encompasses only the irreversible process shown in Eq. (9.1c). Unless otherwise stated, the irreversible denaturation process is the definition used in this chapter. Using this definition for the denaturation of the whey proteins in milk, the immunoglobulins are the most heat labile, and α-lactalbumin is the most heat stable of the whey proteins, with β-lactoglobulin and bovine serum albumin being intermediate (Larson and Rolleri, 1955). In general, significant denaturation of the major whey proteins, α-lactalbumin and β-lactoglobulin, occurs only on heating milk at temperatures above about 70°C. Assessment of the denaturation of whey proteins in milk A considerable amount of research has been directed toward determining and understanding the denaturation processes of the major whey proteins when milk is heated. In early studies, the casein and denatured whey proteins were precipitated by the adjustment of the pH to the isoelectric point of the casein (about pH 4.6). The supernatants from the unheated and heated milks were analyzed for protein nitrogen, which gave estimates of initial native whey protein levels and levels after heat treatment (Rowland, 1933). However, a rapid method for determining the native whey protein levels was required for assessing milk powders for suitability in applications in the bakery industry (Harland and Ashworth, 1947) and also for categorizing milk powders based on the heat treatments received during the powder manufacturing process (Sanderson, 1970b,c). From these requirements, the whey protein nitrogen index (WPNI) method was developed. In the WPNI method, the casein and the denatured whey proteins were precipitated and separated from the native whey proteins by saturation of the milk with common salt, and the supernatant containing the native whey proteins was analyzed for protein content. This was originally achieved by dilution and pH adjustment of the supernatant to produce a turbid solution, with the turbidity proportional to the level of native whey protein present in the milk (Harland and Ashworth, 1947; Kuramoto et al., 1959; Leighton, 1962). However, the WPNI method displayed considerable variability in the degree of turbidity developed for samples with similar levels of whey protein denaturation. To overcome this, Sanderson (1970a) combined a dye-binding method for determining the total protein content of milk with the original WPNI method, thus giving a more accurate and reliable method for determining the WPNI. The original WPNI method or one of its variants is still the industry standard for determining the native whey protein levels of milk powder products and is still widely used to classify milk powders according to the heat treatment received (Singh and Newstead, 1992; Kelly et al., 2003). However, a recent report indicates that Fourier transform near-infrared spectroscopy may have potential as a rapid method for the determination of the WPNI of milk powders in the dry state, eliminating the necessity for reconstitution, precipitation, and filtration (Patel et al., 2007).

332

9. The whey proteins in milk

Although the WPNI method can give an estimate of the level of whey protein denaturation, research into the denaturation and interactions of the individual whey proteins requires more accurate separation and analysis procedures. There are numerous quantitative methods for separating and determining the level of the individual whey proteins in milk, and these can be used directly or can be adapted to determine the level of denaturation after defined heat treatments. Methods that have been used include polyacrylamide gel electrophoresis (PAGE; e.g., Hillier and Lyster, 1979; Dannenberg and Kessler, 1988b; Kessler and Beyer, 1991; Anema and McKenna, 1996), capillary electrophoresis (e.g., Fairise and Cayot, 1998; Butikofer et al., 2006), differential scanning calorimetry (e.g., Ruegg et al., 1977; Manji and Kakuda, 1987), highperformance liquid chromatography (HPLC; e.g., Kessler and Beyer, 1991), and various immuno-based assays (e.g., Lyster, 1970). In recent years, miniaturized and/or rapid automated methods have been adapted for determining whey protein denaturation and protein compositions, such as PAGE techniques on microfluidic chips (Wu et al., 2008; Anema, 2009; Costa et al., 2014; Gubic et al., 2016) or optical biosensor-based assays (Dupont and MullerRenaud, 2006; Indyk, 2009). In general, good correlations have been observed when the various methods for determining whey protein denaturation have been compared (Manji and Kakuda, 1987; Kessler and Beyer, 1991; Anema and Lloyd, 1999; Patel et al., 2007; Anema, 2009; Indyk, 2009). These methods are more time consuming than the traditional WPNI methods and therefore cannot be used for the routine analysis and classification of milk products. However, they have higher accuracy and reproducibility and can be used to determine the denaturation behavior of the individual whey proteins. In addition, variations of the techniques or coupling to additional detection devices can provide further information on the interactions of the denatured whey proteins with other components in the milk (e.g., Lowe et al., 2004; Patel et al., 2006, 2007; Donato and Guyomarc’h, 2009). Kinetic evaluation and modeling of whey protein denaturation Early studies showed that the denaturation of whey proteins was a kinetic phenomenon and therefore dependent on both the temperature and the duration of the heat treatment (Rowland, 1933; Harland and Ashworth, 1945). Although these early studies considered the whey protein components as a single entity, it was noted that the denaturation process did not follow a simple exponential law and was not a first-order (unimolecular) process. In addition, there was a change in temperature dependence above about 80°C, which was probably the first indication of the complex nature of the irreversible denaturation of the whey proteins in milk (Rowland, 1933). Although early studies on the effect of temperature and heating time on the denaturation of the individual whey proteins had been performed (e.g., Harland and Ashworth, 1945; Gough and Jenness, 1962), it was the kinetic study of Lyster (1970) over a wide temperature range (68–155°C) that conclusively demonstrated the complexity of the denaturation process of the whey proteins. Lyster (1970) found that the denaturation of α-lactalbumin appeared to follow first-order kinetics and that the denaturation of β-lactoglobulin was second order. Arrhenius plots for both α-lactalbumin and β-lactoglobulin indicated that the irreversible denaturation reaction was not a simple process, as a change in temperature dependence was observed at about 80–90°C for both α-lactalbumin and β-lactoglobulin (Fig. 9.2). The rate

333

The heat treatment of milk

FIG. 9.2 Effect of milk concentration on the Arrhenius plot for the thermal denaturation of β-lactoglobulin (A) and α-lactalbumin (B) over the temperature range 75–100°C. , 9.6% total solid (TS) milk; ●, 19.2% TS milk; □, 28.8% TS milk; and , 38.4% TS milk. (A) Reproduced with permission from Anema, S.G., 2000. Effect of milk concentration on the irreversible thermal denaturation and disulfide aggregation of β-lactoglobulin. J. Agric. Food Chem. 48, 4168–4175. Copyright (2000) American Chemical Society. (B) Reproduced with permission from Anema, S.G., 2001. Kinetics of the irreversible thermal denaturation and disulfide aggregation of α-lactalbumin in milk samples of various concentrations. J. Food Sci. 66, 2–9. Copyright (2001) Blackwell Publishing.

▪

constants increased more rapidly with an increase in temperature in the low-temperature ranges than at higher temperatures. Further studies confirmed the complexity of the denaturation process and provided relationships between compositional aspects and the rate of denaturation (Lyster, 1970; Hillier and Lyster, 1979; Manji and Kakuda, 1986). The kinetic and thermodynamic studies of Dannenberg and Kessler (1988b) provided insights into the possible mechanisms responsible for the complex temperature dependences of the denaturations of α-lactalbumin and β-lactoglobulin. Dannenberg and Kessler (1988b) found, in milk, that the denaturation of β-lactoglobulin had an order of about 1.5, which is now generally accepted, and that the denaturation of α-lactalbumin was pseudo first order. From thermodynamic evaluations of the denaturation reactions of β-lactoglobulin and α-lactalbumin in the two temperature ranges (i.e., at temperatures above and below the marked change in temperature dependence for the denaturation reactions; Fig. 9.2), information on the possible rate-determining steps in the denaturation reactions was obtained. At temperatures below about 90°C for β-lactoglobulin and 80°C for α-lactalbumin, the high values for the activation energies and enthalpies indicated that a large number of bonds were disrupted, and the positive activation entropies indicated a lower state of order of the reaction

334

9. The whey proteins in milk

products. These kinetic and thermodynamic parameters were interpreted as indicating that the unfolding (reversible denaturation) of the whey proteins was the rate-determining step in the lower-temperature ranges. At higher temperatures, above 80°C for α-lactalbumin and above 90°C for β-lactoglobulin, the considerably lower activation energies and enthalpies were typical of chemical reactions, and the negative activation entropies indicated a higher state of order. These parameters suggested that chemical (aggregation) reactions were the rate-determining step in the higher temperature ranges. Subsequent studies have supported these interpretations in skim milk and whole milk under industrial processing conditions (Anema and McKenna, 1996; Oldfield et al., 1998a). In a recent study, kinetic and thermodynamic evaluations were conducted by treating the whey proteins (α-lactalbumin and β-lactoglobulin) in milk as a single “total whey protein” entity (Anema, 2017b). When considered as total whey protein, the denaturation followed second-order reaction kinetics, with the temperature dependence changing at about 85°C. As with the denaturation of the individual proteins, the thermodynamic parameters of total whey protein denaturation were consistent with unfolding reactions as the rate-determining step at temperatures below 85°C and aggregation reactions as the rate-determining step at higher temperatures. From the kinetic evaluation, equations that could predict the denaturation levels of the total whey proteins after any defined heat treatment were developed, and these predicted denaturation levels were highly correlated with the experimental WPNI of the milk samples (Anema, 2017b). When considering compositional changes of the milk, the denaturation reactions of both β-lactoglobulin and α-lactalbumin are enhanced when the pH of the milk is increased from the natural pH and are retarded when the pH is decreased (Law and Leaver, 2000). The denaturation of β-lactoglobulin was retarded when all components in the milk were concentrated, although the effect was less pronounced as the temperature was increased (Fig. 9.2A; Anema, 2000). In contrast, the denaturation of α-lactalbumin was hardly affected by the milk concentration, with similar rates of denaturation at all milk concentrations regardless of the heating temperature (Fig. 9.2B; Anema, 2001). The seemingly contrasting effects of milk concentration on the denaturation of α-lactalbumin and β-lactoglobulin have been explained through detailed studies on the effect of the concentrations of the individual components of milk on the denaturation reactions. Increasing the protein concentration of milk while maintaining essentially constant concentrations of nonprotein soluble components increased the rate of denaturation of both α-lactalbumin and β-lactoglobulin (Fig. 9.3; Law and Leaver, 1997; Anema et al., 2006), with a similar effect at all temperatures (Anema et al., 2006). Increasing the concentration of nonprotein-soluble components while maintaining constant protein concentrations retarded the denaturation of both β-lactoglobulin and α-lactalbumin; however, the effects on these two proteins were somewhat different (Fig. 9.3; Anema et al., 2006). For β-lactoglobulin, increasing the nonprotein-soluble components caused a substantial retardation of denaturation in the lower-temperature range, and this effect became less pronounced at higher temperatures. In contrast, the effect of increasing the nonprotein-soluble components on α-lactalbumin denaturation was less pronounced than for β-lactoglobulin denaturation and was similar at all temperatures investigated (Fig. 9.3). The increase in lactose concentration, the major component of the nonprotein-soluble

The heat treatment of milk

335

FIG. 9.3 Comparison of the effects of the concentrations of (●, ) protein, ( , □) nonprotein-soluble components, (▲, 4) lactose, and (⧫, ◇) TS on the rate constants for the denaturation of (A) β-lactoglobulin (A) and (B) α-lactalbumin at 80°C (filled symbols) and 95°C (open symbols). Reproduced with permission from Anema, S.G., Lee, S.K., Klostermeyer, H., 2006. Effect of protein, nonprotein-soluble components, and lactose concentrations on the irreversible thermal denaturation of β-lactoglobulin and α-lactalbumin in skim milk. J. Agric. Food Chem. 54, 7339–7348. Copyright (2006) American Chemical Society.

▪

components, explained much of the effect of increasing nonprotein-soluble components; however, clearly, other compositional factors such as pH and mineral components also have an effect (Fig. 9.3; Anema et al., 2006). From these results, it was possible to explain the effects of milk concentration on the denaturation of β-lactoglobulin and α-lactalbumin. For α-lactalbumin, on increasing the total solid (TS) concentration of the milk (both protein- and nonprotein-soluble components), the retardation of the reaction rate by increasing the nonprotein-soluble components concentration was almost exactly offset by the increase in the denaturation rate on increasing the protein concentration. As this effect was similar at all temperatures, increasing the TS of the milk appeared to have no effect on the rate of denaturation of α-lactalbumin (Figs. 9.2B and 9.3B; Anema, 2001; Anema et al., 2006). For β-lactoglobulin, the retardation of the rate of denaturation on increasing the concentration of the nonprotein soluble components was not completely offset by the increasing rate of denaturation on increasing the protein concentration; therefore β-lactoglobulin denaturation was retarded by increasing the TS concentration of the milk. However, the nonprotein-soluble components were less effective in retarding the denaturation of β-lactoglobulin at higher temperatures, and as a consequence, the increase in TS concentration appeared to have a smaller effect on the denaturation of β-lactoglobulin at the higher temperatures and particularly

336

9. The whey proteins in milk

above about 90°C (Figs. 9.2A and 9.3A; Anema, 2000; Anema et al., 2006). The effects of the nonprotein-soluble components concentration or the lactose concentration on the denaturation of β-lactoglobulin and α-lactalbumin have been discussed in terms of the preferential hydration theory (Anema, 2000; Anema et al., 2006).

Interactions between denatured whey proteins and κ-casein/casein micelles An understanding of the denaturation reactions of the whey proteins provides information on the initial steps of a complex series of aggregation reactions that can occur when milk is heated. This aggregation process can involve other milk protein components and may involve numerous reaction pathways or interaction processes. Although the reactions of the denatured whey proteins with other milk protein components are important, these types of reactions are considerably more difficult to measure than the irreversible denaturation processes, particularly in a complex mixture of components such as found in (skim) milk. Interactions between denatured whey proteins and κ-casein in model systems One of the major reactions of interest is the interaction between the denatured whey proteins and the casein micelles, particularly interactions of denatured β-lactoglobulin with κ-casein at the micelle surface. Early studies on model systems indicated that there was an interaction between β-lactoglobulin and κ-casein when these components were heated together (Zittle et al., 1962; Long et al., 1963; Sawyer et al., 1963). These conclusions were drawn from electrophoretic studies, which showed that the discrete bands assigned to κ-casein and β-lactoglobulin observed in unheated solutions produced species of intermediate mobility when the solutions were mixed and heated together. Sedimentation velocity experiments also confirmed complex formation, as the β-lactoglobulin-κ-casein complex formed on heating had markedly higher sedimentation coefficients than did the individual proteins when heated separately (Zittle et al., 1962). Once interactions between κ-casein and denatured β-lactoglobulin had been confirmed, subsequent investigations in heated model systems were aimed at determining the types of bonds involved in complex formation, the stoichiometry of the complexes formed, and the involvement of other whey proteins (particularly α-lactalbumin) in the complexes. It was shown that reducing agents dissociated the heat-induced complexes and that thiolblocking agents prevented the formation of the complexes (Sawyer et al., 1963). These results supported earlier suggestions that the free thiol group of β-lactoglobulin was involved in the interactions (Trautman and Swanson, 1958; Zittle et al., 1962), and it was suggested that intermolecular disulfide bonds were formed between κ-casein and denatured β-lactoglobulin (Sawyer et al., 1963). This has been corroborated by numerous subsequent studies (e.g., Grindrod and Nickerson, 1967; Purkayastha et al., 1967; Sawyer, 1969; Tessier et al., 1969). Some studies indicated that the heat-induced self-aggregation of β-lactoglobulin was limited when κ-casein was present, which suggested that κ-casein formed complexes with intermediate species of aggregated β-lactoglobulin (Sawyer, 1969; McKenzie et al., 1971). In contrast, other studies indicated that the aggregation of β-lactoglobulin was not a prerequisite for its interaction with κ-casein (Euber and Brunner, 1982). The reason for these apparently

The heat treatment of milk

337

conflicting observations may have been resolved through the detailed study of Cho et al. (2003), in which many of the possible pathways involved in the aggregation of β-lactoglobulin with κ-casein in heated model systems were elucidated. Cho et al. (2003) proposed that, when mixtures of β-lactoglobulin and κ-casein were heated, the free thiol of β-lactoglobulin was exposed and this initiated a series of thiol-disulfide exchange reactions of β-lactoglobulin with other denatured β-lactoglobulin molecules or with κ-casein. The products formed ranged from 1:1 β-lactoglobulin-κ-casein complexes to large heterogeneous aggregates, and the product mix was dependent on the ratio of κ-casein to β-lactoglobulin. The aggregate species were held together by either or both disulfide bonds and hydrophobic interactions. Although there have been some indications of interactions between α-lactalbumin and κ-casein on heating (Shalabi and Wheelock, 1976; Doi et al., 1983), others have reported that interactions between these proteins do not occur (Baer et al., 1976; Elfagm and Wheelock, 1978). It is now generally believed that interactions between α-lactalbumin and κ-casein will occur only if β-lactoglobulin (or another whey protein with a free thiol) is present during heating, and this may require the initial formation of a β-lactoglobulin-α-lactalbumin complex, which subsequently interacts with κ-casein (Baer et al., 1976; Elfagm and Wheelock, 1978). A recent study has shown the importance of thiol-disulfide interactions in the irreversible denaturation and aggregation of α-lactalbumin in milk and whey protein solutions (Nguyen et al., 2018). The irreversible thermal denaturation of α-lactalbumin was promoted when low levels of cysteine (or other thiol-bearing reagents) were added to milk, with significant denaturation occurring at temperatures as low as 61°C. At higher temperatures (above about 70°C), the thermal denaturation of β-lactoglobulin was also enhanced on addition of these thiol reagents. It was proposed that, at temperatures between 61 and 70°C, the added cysteine initiated thiol-disulfide interchange reactions with the disulfide bonds of the unfolded α-lactalbumin, thus promoting irreversible aggregation reactions. At temperatures above 70°C, the cysteine, along with the free thiol of unfolded β-lactoglobulin, initiated thiol-disulfide interchange reactions with the disulfide bonds of both α-lactalbumin and β-lactoglobulin, thus promoting the irreversible aggregation of both these proteins (Nguyen et al., 2013; Nguyen et al., 2018). The addition of thiol reagents to milk prior to heating markedly increased the level of α-lactalbumin interacting with the casein micelles; however, only a small increase in β-lactoglobulin interacting with the micelles was observed (Nguyen et al., 2013). There is considerable evidence to show that disulfide bonds are involved in the aggregates formed between the denatured whey proteins and κ-casein; however, there are reports that suggest that noncovalent bonding may be important in these interactions, particularly in the early stages of heating and at lower heating temperatures (Sawyer, 1969; Haque et al., 1987; Haque and Kinsella, 1988; Hill, 1989). Other studies have shown that, although a substantial part of the denatured whey proteins in heated milk are involved in disulfide-bonded aggregates, there is a significant proportion that can be recovered as monomeric protein under dissociating but nonreducing conditions, indicating that noncovalent interactions are also involved (Oldfield et al., 1998b; Anema, 2000). As Cho et al. (2003) have suggested, it is likely that both hydrophobic and disulfide interactions are important in the early stages of aggregate formation, with the interaction mechanism being dependent on the composition of the system and the conditions of heating.

338

9. The whey proteins in milk

Interactions between denatured whey proteins and κ-casein/casein micelles in milk systems Most of the early studies examining the heat-induced interactions between denatured whey proteins and κ-casein were in model systems using purified proteins in buffer solutions. Milk is considerably more complex, with numerous protein species that could potentially interact on heating. A number of the milk proteins have free thiol groups and/or disulfide bonds. Although β-lactoglobulin is the major whey protein component, denatured α-lactalbumin and bovine serum albumin can also be involved in thiol-disulfide exchange reactions and therefore can be incorporated in the aggregated products. For the caseins, as both κ-casein and αS2-casein have disulfide bonds, both could participate in thiol-disulfide exchange reactions with denatured β-lactoglobulin or other denatured thiol-bearing whey proteins. As a consequence of this complexity, there are numerous potential thiol-disulfide interaction pathways, as well as noncovalent interactions, and therefore, the separation and analysis of the reaction products can be difficult. Studies on the interactions between the proteins in heated milk suggest that, despite the complexity of the system, the reactions between β-lactoglobulin and κ-casein may be similar to those occurring in the model systems. In early electrophoretic studies on heated milk, it was noted that the bands corresponding to β-lactoglobulin disappeared, along with a reduction in the intensity of the bands corresponding to casein. This was accompanied by the formation of bands corresponding to new (heterogeneous) components (Slatter and van Winkle, 1952; Tobias et al., 1952). When thiol-blocking agents were added, the band pattern was comparable with that of the original skim milk, indicating that thiol-disulfide exchange reactions were involved in the interaction mechanisms (Trautman and Swanson, 1958). Subsequent studies confirmed that an interaction between denatured β-lactoglobulin and κ-casein on the casein micelles occurred on heating milk, although, as expected, the other denatured whey proteins were also involved in the interactions (Snoeren and van der Spek, 1977; Elfagm and Wheelock, 1978; Smits and van Brouwershaven, 1980; Noh et al., 1989b; Noh et al., 1989a; Corredig and Dalgleish, 1996a,b, 1999; Oldfield et al., 1998b). Unlike κ-casein, αS2-casein does not readily interact with denatured whey proteins when milk is heated, although some interactions in UHT milks have been reported (Snoeren and van der Spek, 1977; Patel et al., 2006). This low reactivity may be due to the location of αS2-casein in the interior of the casein micelles, which makes it less accessible for interaction, whereas κ-casein is located at the casein micelle surface and is therefore more accessible for interaction (Walstra, 1990; Horne, 1998; Huppertz et al., 2017). Interestingly, in pressuretreated skim milk, disulfide-bonded aggregates between αS2-casein and the denatured whey proteins are observed, suggesting that the disulfide bonds of αS2-casein may become accessible to thiol groups of the denatured whey proteins when the casein micelle structure is disrupted under pressure (Patel et al., 2006). The degree of interaction of the denatured whey proteins with the casein micelles is dependent on many variables including the time, temperature, and rate of heating; the milk and individual protein concentrations; the milk pH; and the concentration of the milk salts (Smits and van Brouwershaven, 1980; Corredig and Dalgleish, 1996a,b; Oldfield et al., 2000; Anema and Li, 2003b; Oldfield et al., 2005). For example, when the temperature of milk is gradually increased above 70°C, as in indirect heating systems, most of the denatured

The heat treatment of milk

339

β-lactoglobulin and α-lactalbumin associates with the casein micelles, presumably as disulfidebonded complexes with κ-casein at the micelle surface (Smits and van Brouwershaven, 1980; Corredig and Dalgleish, 1996b). In contrast, when milk is heated rapidly, as in direct heating systems, only about half of the denatured β-lactoglobulin and α-lactalbumin associates with the casein micelles, with the rest remaining in the milk serum (Singh and Creamer, 1991a; Corredig and Dalgleish, 1996a; Oldfield et al., 1998b). Corredig and Dalgleish (1999) suggested that, on heating milk, α-lactalbumin and β-lactoglobulin initially aggregate in the serum phase at a ratio that is dependent on the initial individual whey protein concentrations. These complexes subsequently associate with κ-casein at the casein micelle surface on prolonged heating. However, Oldfield et al. (1998b) proposed that, under rapid heating rates, β-lactoglobulin forms aggregates in the serum before interacting with the casein micelles and this limits the level of association with the casein micelles, whereas at slower heating rates, monomers or smaller aggregates of β-lactoglobulin may interact with the micelles, and this may allow greater association with the casein micelles. The pH of the milk at heating is important in determining the level of interaction between the denatured whey proteins and the casein micelles. When milk is heated at high temperatures (about 140°C), the heat coagulation time/pH profiles of most milks show increasing heat stability with increasing pH to a maximum at about pH 6.7, followed by decreasing stability to a minimum at about pH 6.9, and increasing stability again as the pH is increased further (Rose, 1961). Considerable research has been undertaken over decades in an attempt to explain this unusual pH-dependent heat stability of milk, and numerous factors are known to influence the heat stability behavior. Many review papers on the heat stability of milk are available (Singh and Creamer, 1992; International-Dairy-Federation, 1995; O’Connell and Fox, 2003; Singh, 2004). The results from the studies on the heat stability of milk have influenced the direction of the future research on the effects of heat on milk and, in particular, on the interactions between denatured whey proteins and κ-casein/casein micelles. Therefore, it is appropriate to briefly review aspects of the pH dependence of heat stability that are relevant to understanding the interactions between denatured whey proteins and κ-casein/casein micelles. Electron microscopic studies showed that, when milk was heated at high temperatures (90–140°C) for long times (30 min) at pHs below 6.7, the denatured whey proteins complexed on to the micelle surfaces as filamentous appendages. However, when the milk was heated at higher pH, the denatured whey proteins were found in the serum phase as aggregated complexes (Creamer et al., 1978; Creamer and Matheson, 1980). These were the first indications that the pH at which the milk is heated may influence the interactions between the denatured whey proteins and the casein micelles. Kudo (1980) showed that the amount of nonsedimentable protein in milk heated at pH 6.5 was lower than that in unheated milk; however, the level of nonsedimentable protein increased with the pH at heating, so, at above pH 6.7, the level was markedly higher than in the unheated milk and increased with increasing pH. Kudo (1980) concluded that the denatured whey proteins cosedimented with the casein micelles at low pH (about pH 6.5), whereas most of the denatured whey proteins along with some casein (particularly κ-casein) was released from the casein micelles at pHs above 6.8. It was also proposed that the transition from

340

9. The whey proteins in milk

whey protein-coated casein micelles to protein-depleted forms with changing pH at heating could explain the pH dependence of the heat stability of milk at high temperatures. Singh and Fox (1985a,b 1986, 1987a,b,c), in a series of extensive studies, showed that the dissociation of κ-casein-rich protein on heating was dependent on the pH at heating. At pHs below about 6.8, little dissociation of micellar κ-casein occurred, whereas at higher pH, particularly above pH 6.9, high levels of κ-casein dissociated from the micelles, with the level increasing proportionally with increased pH. The whey proteins, particularly β-lactoglobulin, played an important role in the heat-induced, pH-dependent dissociation of κ-casein (Singh and Fox, 1987a,b), as did mineral components such as calcium and phosphate (Singh and Fox, 1987c). The results from these studies have been used to develop detailed mechanisms for the pH-dependent heat stability of milk, including concentrated milk systems (Singh and Creamer, 1992; O’Connell and Fox, 2003; Singh, 2004). Initially, it was reported that the dissociation of κ-casein from the casein micelles occurred only when milk at high pH (above about pH 6.8) was heated at high temperatures, particularly 90°C or above (Singh and Fox, 1985a). However, subsequent studies demonstrated that, at these pH values, the dissociation of κ-casein occurred as soon as the temperature was raised above ambient, with the level of dissociated κ-casein increasing proportionally with temperature up to 90°C. In these studies, the dissociation of αS-casein (αS1-casein and αS2-casein combined) and β-casein showed unusual temperature dependence. Increasing levels of these caseins dissociated as the temperature was increased up to about 70°C, with the levels then decreasing again at higher temperatures (Fig. 9.4; Anema and Klostermeyer, 1997; Anema, 1998). The unusual temperature dependence of αS-casein and β-casein dissociation was a consequence of the whey proteins, particularly β-lactoglobulin. When whey protein-depleted milk was heated, the levels of αS-casein and β-casein dissociating from the casein micelles increased with increasing temperature up to 90°C. When compared with heating standard milk, this indicated that higher levels of αS-casein and β-casein dissociated from the micelles in the whey protein-depleted milks at temperatures above about 70°C (Anema and Li, 2000). It was postulated that all the caseins dissociated from the micelles on heating. On subsequent cooling, the dissociated κ-casein stabilized the dissociated αS-casein and β-casein as small serum-phase aggregates if the heating temperature was below about 70°C. However, above about 70°C, κ-casein associated with denatured whey proteins. It was already known that the complex formed between κ-casein and denatured β-lactoglobulin was less effective than uncomplexed κ-casein at stabilizing αS-casein and β-casein in the presence of calcium ions (Zittle et al., 1962); therefore, this interaction may have prevented κ-casein from stabilizing the other caseins, and they either reassociated with the casein micelles or formed larger aggregates on subsequent cooling (Anema and Li, 2000). Early studies on the effect of the pH at heating on the interaction of denatured whey proteins with the casein micelles tended to use relatively large pH steps. In a model milk system containing casein micelles and β-lactoglobulin, about 80% of the denatured β-lactoglobulin associated with the casein micelles when the milk was heated at pH 5.8 or pH 6.3, whereas only about 20% associated with the casein micelles at pH 6.8 or pH 7.1 (Smits and van Brouwershaven, 1980). The studies on the heat-induced, pH-dependent dissociation of κ-casein from the casein micelles showed that this dissociation was accompanied by increases in the levels of

341

The heat treatment of milk

FIG. 9.4 Effect of temperature and pH on the level of protein in the supernatants obtained from 10% TS reconstituted skim milk samples heated for 30 min: (A) κ-casein, (B) αs-casein, and (C) β-casein. , pH 6.3; ●, pH 6.5; □, pH 6.7; , pH 6.9; and 4, pH 7.1. Reproduced with permission from Anema, S.G., Klostermeyer, H., 1997. Heat-induced, pH-dependent dissociation of casein micelles on heating reconstituted skim milk at temperatures below 100°C. J. Agric. Food Chem. 45, 1108–1115. Copyright (1997) American Chemical Society.

▪

denatured whey proteins remaining in the serum (Singh and Creamer, 1991b), and this was confirmed by Anema and Klostermeyer (1997) and Oldfield et al. (2000), who reported that 80%–90% of the denatured whey proteins associated with the casein micelles when milk was heated at pH below 6.7, whereas only about 20% associated with the casein micelles at pH above 6.8. Corredig and Dalgleish (1996b) measured the ratio of β-lactoglobulin or

342

9. The whey proteins in milk

α-lactalbumin to κ-casein in the colloidal phase obtained from heated milk adjusted to pH 5.8, 6.2, or 6.8. Although the denatured whey proteins interacted with the casein micelles at a faster rate at lower pH and at higher temperatures, the ratios of denatured whey proteins to κ-casein on the casein micelles were not markedly different under the different heating conditions. Further studies demonstrated the extreme importance of pH in the association of denatured whey proteins (α-lactalbumin and β-lactoglobulin) with the casein micelles when milk was heated above 70°C, particularly at pH 6.7 or below, where differences in association behavior could be measured at pH differences as small as 0.05 pH units (Anema and Li, 2003a,b; Vasbinder and de Kruif, 2003). From these studies, it was shown that, on heating milk at 90°C for 30 min, about 80% of the denatured whey protein associated with the casein micelles at pH 6.5 and that this level of association decreased linearly as the pH at heating was increased, so that only about 30% was associated at pH 6.7. At higher pH (above pH 6.7), very low levels of denatured whey proteins associated with the casein micelles on heating the milk (Fig. 9.5). Although the heat-induced, pH-dependent dissociation of κ-casein from the casein micelles could explain the low levels of denatured whey proteins interacting with the casein micelles at pH above 6.8, it had been reported that very little κ-casein dissociated from the casein micelles at pH below 6.8 (Singh and Fox, 1985a; Nieuwenhuijse et al., 1991; Singh, 2004). Therefore, it was initially unknown why small shifts in pH between 6.5 and 6.7 affected the association of denatured whey proteins with the casein micelles when milk was heated. The level of κ-casein in the serum phase was low; therefore, it was initially believed that κ-casein was not involved in this partition of the whey proteins between the serum and colloidal phases (Oldfield et al., 1998b; Anema and Li, 2003b; Vasbinder and de Kruif, 2003).

FIG. 9.5 Level of whey proteins associated with the casein micelles/nonsedimentable whey proteins in skim milk samples that were heated at 90°C for various times. The pH values of the milk samples prior to heating were ●, pH 6.5; , pH 6.55; ., pH 6.6; 4, pH 6.65; , pH 6.7; □, pH 6.9; and ⧫, pH 7.1. Reproduced with permission from Anema, S.G., Lee, S.K., Lowe, E.K., Klostermeyer, H., 2004a. Rheological properties of acid gels prepared from heated pH-adjusted skim milk. J. Agric. Food Chem. 52, 337–343. Copyright (2004) American Chemical Society.

▪

The heat treatment of milk

343

However, subsequent studies showed that the heat-induced dissociation of κ-casein was pH dependent from pH 6.5 to pH 7.1, with a linear increase in serum-phase κ-casein as the pH was increased throughout the pH range from 6.5 to 7.1 (Fig. 9.6A), and that the level of serum-phase κ-casein was correlated with the level of serum-phase denatured whey protein (Fig. 9.6B; Anema, 2007). At the natural pH, about 30%–40% of the κ-casein dissociated on heating the milk at 90°C for 20–30 min (Fig. 9.6A; Anema, 2007), and recent studies showed that, when milk at its natural pH was heated under commercial direct or indirect UHT processing conditions (approximately 140°C/3 s), high levels of κ-casein (up to 60%) dissociated from the casein micelles. Under these UHT heating conditions, most of the denatured whey proteins remained in the serum phase (Anema, 2017a; Gaur et al., 2018). The differences in the level of dissociated κ-casein between the earlier studies and the later studies may be related to the conditions of centrifugation. Early studies used very high centrifugal forces (about 50,000  g), which may have masked the effects at the lower pH, especially under

FIG. 9.6 (A) Effect of the pH at heating on the level of nonsedimentable κ-casein in milk. ●, serum-phase κ-casein in unheated milk; , serum-phase κ-casein in milk heated at 90°C for 20 min; ., serum-phase κ-casein in milk heated at 90°C for 25 min; 4, and serum-phase κ-casein in milk heated at 90°C for 30 min. (B) Relationship between the serum-phase denatured whey protein and the level of serum-phase κ-casein for the heated milk samples. ●, milk heated at 90°C for 20 min; , milk heated at 90° C for 25 min; and ., milk heated at 90°C for 30 min. Reproduced with permission from Anema, S.G., 2007. Role of κ-casein in the association of denatured whey proteins with casein micelles in heated reconstituted skim milk. J. Agric. Food Chem. 55, 3635–3642. Copyright (2007) American Chemical Society.

344

9. The whey proteins in milk

conditions where the particles are less hydrated and more readily deposited. Later studies used lower centrifugal forces (about 25,000  g), which allowed smaller dissociated particles to be observed (Parker et al., 2005; Rodriguez del Angel and Dalgleish, 2006; Anema, 2007). Although the level of κ-casein in the serum phase at pH below 6.7 was relatively low (less than about 30% of the total κ-casein), the ratio of denatured whey protein to κ-casein was high and relatively constant (about 2.5 whey protein molecules to each monomeric κ-casein molecule) for the serum-phase proteins at all pH values. In contrast, the ratio of denatured whey protein to κ-casein was only about 1:1 for the whey protein associated with the casein micelles at pH 6.5, and this decreased to about 0.5:1 at pH 7.1 (Anema, 2007). Intensive studies on the soluble whey protein-κ-casein complexes formed when milk was heated at the natural pH also showed that κ-casein was intimately involved in the serum-phase aggregates and that a high ratio of denatured whey proteins to κ-casein was observed (Guyomarc’h et al., 2003). Electron micrographs of the serum-phase whey protein-κ-casein aggregates indicated that these particles were roughly spherical with a relatively uniform size of about 20–50 nm (Parker et al., 2005; Rodriguez del Angel and Dalgleish, 2006). A similar pH-dependent dissociation of κ-casein and the formation of serum-phase denatured whey proteins was observed when concentrated milk was heated at different pH values (Anema, 1998, 2008a). For milk heated at 80°C/30 min, at each milk concentration, the level of nonsedimentable whey protein increased markedly with increasing pH, and at any given pH, the level of nonsedimentable whey protein increased as the milk concentration increased. The level of κ-casein dissociating from the casein micelles when the milk was heated (80°C/30 min) increased markedly with increasing pH at each milk concentration and with increasing milk concentration at any given pH (Anema, 2008a). When milk samples of different concentrations from 10% to 25% TS and at pHs from 6.3 to 7.1 were heated at temperatures of 20–120°C for 10 min, the pH-dependent dissociation of κ-casein from the casein micelles occurred as soon as the milk was heated at temperatures above ambient conditions. At any given pH and milk concentration, the level of κ-casein dissociating from the casein micelles increased with increasing temperature. Similarly, at any given pH and temperature, the level of κ-casein dissociating from the casein micelles increased with increasing milk concentration (Anema, 1998). When the whey protein level in skim milk was increased by up to 1% by the addition of whey protein concentrate/isolate, increased levels of κ-casein were dissociated from the casein micelles, and increased levels of denatured whey proteins remained in the serum phase when the milk was heated (80°C/30 min) at its natural pH. In pH-adjusted milk, at pH 6.5, low levels of κ-casein dissociated from the casein micelles and low levels of denatured whey proteins remained in the serum phase after heating the milk. The levels of κ-casein and denatured whey protein in the serum phase increased as the pH of the milk increased and, at any given pH, higher levels of κ-casein and denatured whey proteins were in the serum phase for the milk with 0.75% added whey protein (Anema, 2018a). There is still some debate over the sequence of events in the interaction reactions between the denatured whey proteins and κ-casein. Some reports suggest that κ-casein dissociates from the micelles early in the heating process and that the denatured whey proteins subsequently interact with the κ-casein either in the serum phase or on the micelles, with a preferential serum-phase reaction (Anema and Li, 2000; Anema, 2007, 2008b). This proposal was supported by the observation that the dissociation of κ-casein is a rapid process and that significant dissociation of κ-casein can occur at temperatures below those at which the

The heat treatment of milk

345

denaturation of whey proteins occurs (Anema and Klostermeyer, 1997). In addition, significant dissociation of κ-casein occurs in systems that have been depleted of whey proteins (Anema and Li, 2000). The higher ratio of denatured whey protein to κ-casein for the serum-phase protein aggregates regardless of the pH at heating or the level of dissociated κ-casein may also suggest a preferential serum-phase reaction between the denatured whey proteins and κ-casein (Anema, 2007). However, other reports suggest that, on heating milk, the denatured whey proteins first interact with the casein micelles and that the whey protein-κ-casein complexes subsequently dissociate from the casein micelles (Parker et al., 2005; Donato and Dalgleish, 2006; Donato et al., 2007b; Donato and Guyomarc’h, 2009). This proposal was supported by the observation that the addition of sodium caseinate to milk did not increase the level of serum-phase complexes between the denatured whey proteins and κ-casein, which was interpreted as indicating that the complexes between the denatured whey proteins and κ-casein were formed on the casein micelle surface regardless of the pH at heating (Parker et al., 2005). Milks with added κ-casein were analyzed by size exclusion chromatography (Donato et al., 2007b). The difference profiles between the sera from unheated milk and heated milk with added κ-casein produced a negative peak in the region of the native whey proteins and a positive peak in the region of the whey protein-κ-casein aggregates. As the difference spectra were the same as for milks without added κ-casein, this was also interpreted as indicating that the added κ-casein was not involved in the formation of the serum-phase aggregates and therefore that the denatured whey proteins interacted only with micelle-bound κ-casein, with the whey protein-κ-casein complexes subsequently dissociating from the casein micelles (Donato et al., 2007b; Donato and Guyomarc’h, 2009). The partial hydrolysis of κ-casein by chymosin prevented the dissociation of unhydrolyzed κ-casein from the casein micelles. This in turn prevented the formation of serum-phase whey protein-κ-casein complexes and therefore increased the level of denatured whey proteins associating with the casein micelles (Renan et al., 2007). This observation was also used as evidence to support the initial interaction of denatured whey proteins with the casein micelles, with the subsequent dissociation of whey protein-κ-casein complexes, as it was suggested that the unhydrolyzed κ-casein should still be able to dissociate and interact with serum-phase denatured whey proteins if this was the preferential reaction pathway (Renan et al., 2007). However, this proposal did not take into account the polymeric nature of κ-casein (Holland et al., 2008), and therefore, the partial hydrolysis of κ-casein would substantially increase the hydrophobicity of the polymer even when it contained some unhydrolyzed κ-casein. This increased hydrophobicity may account for the reduced dissociation of κ-casein from the micelles and the increased interaction of denatured whey proteins with the casein micelles (Donato and Guyomarc’h, 2009). It was also suggested that two mechanisms occur depending on the pH at heating. This was based on observations that the protein composition of the serum phase appeared to vary markedly depending on whether the milk was heated at a pH above or below the natural pH of the milk (Donato and Dalgleish, 2006). However, other studies did not display a marked difference in composition of the serum-phase proteins with pH (Anema, 2007). A detailed study that was specifically aimed at elucidating the sequence of events occurring when denatured whey proteins interacted with κ-casein in heated milks was conducted (Anema, 2008b). It was shown that κ-casein could dissociate from the casein micelles at temperatures that were below those at which the whey proteins denatured (Fig. 9.7A). When

346

9. The whey proteins in milk

FIG. 9.7 (A) Level of serum-phase whey protein (filled symbols) and κ-casein (open symbols) in milk samples heated for 15 min at different temperatures. (B) Level of denatured (filled symbols) and micelle-bound (open symbols) whey proteins in milk samples heated at 90°C for different times. (C) Level of serum-phase κ-casein in milk samples heated at 90°C for different times. The skim milks were adjusted to (, ●), pH 6.5; (4, ▲), pH 6.7; and (□, ), pH 6.9 before heating. Reproduced with permission from Anema, S.G., 2008b. On heating milk, the dissociation of κ-casein from the casein micelles can precede interactions with the denatured whey proteins. J. Dairy Res. 75, 415–421. Copyright (2008) Cambridge University Press.

▪

heated at temperatures at which the whey proteins could denature, it was found that κ-casein dissociated from the casein micelles in the early stages of heating and before significant levels of whey proteins were denatured. In addition, the maximum level of serum-phase κ-casein was obtained when less than half the whey proteins were denatured, and once this maximum

The heat treatment of milk

347

level of κ-casein was dissociated, any additional denatured whey proteins formed on prolonged heating were found predominantly in the serum phase (Fig. 9.7B), indicating a preferential interaction of the denatured whey proteins with the serum-phase whey protein-κ-casein complexes. When κ-casein was added to the milk prior to heating, the denatured whey proteins preferentially interacted with the added serum-phase κ-casein, regardless of the pH at heating (Table 9.1). Taken together, these results provide unequivocal evidence that κ-casein dissociation from the micelles can precede the interaction of denatured whey proteins with the κ-casein and that denatured whey proteins will preferentially interact with serum-phase κ-casein (Anema, 2008b). The pH-dependent changes in the association of the denatured whey proteins with the casein micelles, and the dissociation of κ-casein from the micelles, can have an effect on some of the physical properties of the milk. A marked increase in casein micelle size was observed when high levels of denatured whey protein were associated with the colloidal phase, as is observed on heating milk at pH 6.5. This change in size was less pronounced as the pH at heating the milk was increased to pH 6.7, and a decrease in casein micelle size was observed when significant levels of κ-casein were dissociated from the colloidal phase, as is observed on heating milk at pH values above 6.7 (Fig. 9.8A; Anema and Li, 2003a,b). Similar changes in viscosity (Fig. 9.8B) and turbidity (Fig. 9.8C) with the pH at heating were also observed (Anema et al., 2004c).

TABLE 9.1 Serum-phase denatured whey proteins from skim milk with different levels of added κ-casein that were adjusted to pH 6.5, 6.7, and 6.9 before heating at 90°C for 15 min; the numbers represent the average and standard deviation of triplicate measurements Added κ-casein (%)

pH at heating

Serum-phase denatured whey protein (% of total)

0

6.5

35  1a

0.1

6.5

67  2b

0.2

6.5

84  3c

0

6.7

71  2a

0.1

6.7

78  3b

0.2

6.7

92  3c

0

6.9

86  3a

0.1

6.9

91  3a,b

0.2

6.9

95  3b

Data at a given pH with the same letters are not significantly different from each other at P < .05. (Reproduced with permission from Anema, S.G., 2008b. On heating milk, the dissociation of κ-casein from the casein micelles can precede interactions with the denatured whey proteins. J. Dairy Res. 75, 415–421. Copyright (2008) Cambridge University Press.)

348

9. The whey proteins in milk

FIG. 9.8 Effects of the pH at heating on the changes in (A) the size of the casein micelles, (B) the viscosity of the milk, and (C) the turbidity of the milk. The milk samples were heated at 90°C for various times, and the pH values of the milk samples prior to heating were ●, pH 6.5; , pH 6.55; ., pH 6.6; 4, pH 6.65; □, pH 6.7; □, pH 6.9; and ⧫, pH 7.1. Some of the particle size and viscosity results were reproduced with permission from Anema, S.G., Lowe, E.K., Li, Y., 2004c. Effect of pH on the viscosity of heated reconstituted skim milk. Int. Dairy J. 14, 541–548. Copyright (2004) Elsevier.

The difficulty in interpreting these changes in size, viscosity, and turbidity is determining whether the association of the denatured whey proteins with the casein micelles is directly responsible for the change in size/volume of the casein micelles by increasing the diameters of the individual particles as the proteins interact or whether there is some associated phenomenon, such as aggregation of the casein micelles, that is related to the level of whey

The heat treatment of milk

349

protein or κ-casein that is in the serum phase or associated with the casein micelles. The strong relationship between the level of whey protein associating with the colloidal phase and the size/volume of the casein micelles, the observation that the size change plateaus on prolonged heating, and the relationships between the protein composition of the micelles and size, viscosity, and turbidity seem to suggest that the size changes are a direct consequence of the distribution of protein between the colloidal and serum phases, rather than an associated aggregation reaction (Anema and Li, 2003a,b; Anema et al., 2004c). Determination of the specific disulfide bonds formed between κ-casein and β-lactoglobulin Although many types of bonding may be involved in the early stages of interactions between the denatured whey proteins and κ-casein, there is clear evidence that disulfide bonds are involved in complex formation when model systems and milk are heated. Studies have focused on determining the specific thiol groups of κ-casein and, in particular, β-lactoglobulin that are involved in the disulfide bonding between these two protein species when they are heated in model systems or milk (Henry et al., 2002; Creamer et al., 2004; Livney and Dalgleish, 2004; Lowe et al., 2004). Understanding the specific disulfide bonds involved in the interaction process may provide useful insights into the mechanisms for the denaturation and subsequent aggregation reactions of the whey proteins in milk. Native β-lactoglobulin has two disulfide bonds and one free thiol group at Cys121 (Qin et al., 1999), whereas κ-casein is found as a heterogeneous polymeric protein that is cross-linked in a random manner by disulfide bonding via the two Cys groups in the monomer protein (Rasmussen et al., 1999; Holland et al., 2008). It was believed that the formation of disulfide bonds between the denatured β-lactoglobulin and other milk proteins, including κ-casein, during heating first involved the dissociation of the β-lactoglobulin dimer to monomer species and followed by the unfolding of the native structure, exposing the buried side groups including the free thiol at Cys121. The exposure of this free Cys121 then initiated a series of intermolecular thiol-disulfide exchange reactions with other denatured whey proteins and with κ-casein on the casein micelle surface (Snoeren and van der Spek, 1977; Iametti et al., 1996; Hoffmann and van Mil, 1997; Verheul et al., 1998; Vasbinder and de Kruif, 2003; Creamer et al., 2004). However, studies on pure β-lactoglobulin indicated that, in the early stages of the denaturation process, nonnative monomeric β-lactoglobulin species, which may be intermediates in the intermolecular aggregation processes, are formed. These nonnative monomeric species are stable on subsequent cooling and can be separated by alkaline PAGE techniques (Manderson et al., 1998; Hong and Creamer, 2002) or by gel permeation chromatography (Iametti et al., 1996; Croguennec et al., 2003, 2004). It was hypothesized that the nonnative monomers were formed from intramolecular thiol-disulfide exchange reactions between the free Cys121 of β-lactoglobulin and the Cys106-Cys119 and/or Cys66-Cys160 disulfide bonds within the same β-lactoglobulin monomer (Iametti et al., 1996; Manderson et al., 1998; Hong and Creamer, 2002; Croguennec et al., 2003, 2004). With the advent of sensitive mass spectrometric techniques, the identification of the Cys residues involved in disulfide-bonded protein species was possible. The strategy for the identification of specific disulfide bonds involves a number of steps (Gillece Castro and Stults, 1996; Gorman et al., 2002; Lowe et al., 2004; Kehoe et al., 2007). For the sample under analysis,

350

9. The whey proteins in milk

the protein species are first hydrolyzed under conditions where no further thiol-disulfide bond exchange reactions are likely to occur. The peptides formed from this hydrolysis are separated, usually by reverse phase HPLC, and the mass of the individual peptides is determined by mass spectrometry (MS). The identification of individual peptides can be achieved by comparing the measured masses with those of expected peptides for the hydrolysis of the protein under study. For the disulfide-bonded peptides, usually, other criteria also need to be satisfied, such as the peptides being present in nonreduced hydrolyzates but being absent in the reduced system. Further confirmation can be gained by the use of tandem MS, where single molecular ions are isolated and analyzed in the first mass analyzer and are then passed into a collision cell where fragmentation of the peptide is induced by collision with an inert gas (collisioninduced dissociation [CID]) and the fragments are characterized in the second mass analyzer. From the mass of the fragments, the sequence of the amino acids in the peptides can be determined, providing conclusive characterization of the peptides (Gorman et al., 2002; Lowe et al., 2004; Kehoe et al., 2007). Using these types of mass spectrometric techniques, a stable nonnative β-lactoglobulin monomer with a free thiol group at position Cys119 rather than the natural position of Cys121 was found in heated β-lactoglobulin solutions (Croguennec et al., 2003), confirming that intramolecular thiol-disulfide exchange within monomeric β-lactoglobulin could occur. It was suggested that this β-lactoglobulin with the free thiol at Cys119 may be the activated monomer that was proposed as the starting point for intermolecular aggregation reactions leading to large polymers, although it was equally possible that unfolded protein with a free thiol at the natural position of Cys121 was that activated monomer (Croguennec et al., 2003, 2004; Creamer et al., 2004). A more recent investigation on the disulfide-bonding patterns in heated β-lactoglobulin found that a significant proportion of Cys160 was in the reduced form after heating β-lactoglobulin in solution (Creamer et al., 2004; Kehoe et al., 2007), indicating that the Cys66-Cys160 disulfide bond was broken during the early stages of heating. This may occur concurrently with the interchange of the free thiol from Cys121 to Cys119. It was suggested that a monomeric β-lactoglobulin species with a free thiol at Cys160 may be (one of ) the reactive species involved in the intermolecular thiol-disulfide bonding that is responsible for crosslinking in heat-induced whey protein aggregates because of its position near the C-terminal end of the protein. Attempts have been made to identify the specific Cys residues involved in disulfide bonds formed between κ-casein and β-lactoglobulin when these proteins are heated together. Livney and Dalgleish (2004) compared masses of peptides from tryptic digests of heated κ-casein/ β-lactoglobulin mixtures with theoretical values and concluded that Cys106/119/121 of β-lactoglobulin were involved in disulfide bonds with both Cys11 and Cys88 of κ-casein (note that the hydrolysis pattern does not allow the separation of the three Cys106/119/121 residues of β-lactoglobulin unless CID is used for sequencing). Although some peptides involving Cys66 and Cys160 of β-lactoglobulin and the two Cys residues of κ-casein were also identified based on mass comparisons, the high abundance of disulfide-bonded peptides containing Cys106/119/121 led these authors to conclude that β-lactoglobulin with a free thiol at Cys119/121 was the predominant species that was involved in intermolecular disulfide bonding. The potential disulfide-bonded species were characterized based on mass analysis alone;

The heat treatment of milk

351

no confirmatory experiments were performed, such as comparing reduced with nonreduced systems to ensure that the proposed intermolecular peptides were disulfide bonded, or confirming sequences by CID-MS to preclude misidentification of similarly massed peptides. In a novel study, Lowe et al. (2004) used an activated monomeric κ-casein with reduced thiol groups that were blocked with thionitrobenzoate (TNB). β-Lactoglobulin was added to the mixture, and the system was heated under very mild conditions (60°C). The TNB groups on the thiols of κ-casein are good leaving groups, and when a reactive thiol from β-lactoglobulin is exposed, it is capable of interacting with the activated TNB groups on κ-casein in a specific 1:1 oxidative reaction, forming a disulfide-bonded complex and releasing the TNB as a brightly colored compound. This approach allowed the formation of specific disulfide bonds between κ-casein and β-lactoglobulin under mild heating conditions. Because of the chemical nature of the reaction, it limited further thiol-disulfide exchange reactions, which allowed specific interactions between β-lactoglobulin and κ-casein to be monitored during the early stages of the denaturation of β-lactoglobulin. The interacted β-lactoglobulin-κ-casein complexes were hydrolyzed with trypsin and were separated by reverse phase HPLC followed by MS. In addition, disulfide bonding was confirmed by comparing the HPLC traces of nonreduced systems with those of reduced systems, and the identities of some peptides were confirmed by sequencing using CID-MS. Although it was possible to identify disulfide bonds between Cys106/119/121 of β-lactoglobulin and Cys88 of κ-casein, Cys160 of β-lactoglobulin was found to have formed disulfide bonds with both Cys11 and Cys88 of κ-casein as major products (Fig. 9.9). This supported the earlier findings on pure β-lactoglobulin that intramolecular thiol-disulfide exchange may precede the intermolecular reactions and that the nonnative monomeric form of β-lactoglobulin species with a free thiol at Cys160 is likely to be (one of ) the reactive monomer species that initiates intermolecular thiol-disulfide exchange reactions (Creamer et al., 2004; Kehoe et al., 2007). Lowe et al. (2004), using the techniques developed for the model system, expanded the study to examine the specific disulfide bonds involved in aggregation between β-lactoglobulin and κ-casein in heated milk systems. Interestingly, no disulfide bonds between Cys106/119/121 of β-lactoglobulin and either of the Cys residues of κ-casein could be found, even though the disulfide bond between Cys88 of κ-casein and Cys106/119/121 of β-lactoglobulin was readily identified in the model system. In the heated milk system, it was found that Cys160 of β-lactoglobulin formed disulfide bonds with both Cys88 and Cys11 of κ-casein, as was found in the model system (Fig. 9.9). In independent studies, a similar disulfide bond between Cys160 of β-lactoglobulin and Cys88 of κ-casein was identified in a heated model goat milk system consisting of isolated casein micelles and β-lactoglobulin suspended in milk ultrafiltrate, although it appears that no attempts were made to isolate and characterize other intermolecular disulfide bonds between these protein species (Henry et al., 2002). From these observations, it was concluded that, in the model system of β-lactoglobulin and activated κ-casein, nonnative monomeric β-lactoglobulin species with a free thiol at either Cys119 or Cys121 (but probably not Cys106) could be the reactive monomer that is involved in intermolecular thiol-disulfide exchange reactions, as Cys119/121 was involved in disulfide bonds with κ-casein. However, further intramolecular thiol-disulfide exchange reactions in β-lactoglobulin must precede or occur concurrently with the intermolecular reactions, as disulfide bonds between Cys160 of β-lactoglobulin and the two Cys residues of κ-casein were also observed as major products in the model system (Fig. 9.9).

352

9. The whey proteins in milk

S-S

p

p

p

k-Casein Confirmed tryptic peptides CID-MS confirmed peptides Cys11

Cys88

Found in Fou n

Fo un d

heated m ilk and m

odel syst

d in

hea

ted

Confirmed by CID MS Not confirmed

Foun d in

ated

m

od el

em

in he

sy

milk

and

st

milk

Cys66

mod

el sy

stem

em

Cys106

Found in model system Cys119/121

Cys160

Peptides only in reduced hydrolyzate Hydrolyzate peptides from model system and heated milk

b-Lactoglobulin SS SS

1

20

40

60

80

100

120

140

160

170

Residue number

FIG. 9.9 Diagram indicating the identified peptides on the linear sequences of κ-casein and β-lactoglobulin and the

intermolecular disulfide bonds formed between κ-casein and β-lactoglobulin on heating model systems and milk. The horizontal box lines represent the protein sequence, the lines over the boxes represent the peptides, and SdS indicates the presence of a disulfide bond. Arrows indicate the major proteolytic sites, the chymosin site for κ-casein, and the rapid tryptic sites for β-lactoglobulin. Potential glycosylation and phosphorylation sites are indicated for κ-casein. Reproduced with permission from Lowe, E.K., Anema, S.G., Bienvenue, A., Boland, M.J., Creamer, L.K., Jimenez-Flores, R., 2004. Heat-induced redistribution of disulfide bonds in milk proteins. 2. Disulfide bonding patterns between bovine β-lactoglobulin and κ-casein. J. Agric. Food Chem. 52, 7669–7680. Copyright (2004) American Chemical Society.

In the heated milk system, no peptides involving Cys106/119/121 and the two Cys residues of κ-casein were isolated. Only peptides involving Cys160 and Cys66 with both Cys residues of κ-casein were found (Fig. 9.9). As Cys160 and Cys66 are involved in a disulfide bond in native β-lactoglobulin, this indicates that intramolecular thiol-disulfide exchange reactions in β-lactoglobulin precede the intermolecular thiol-disulfide exchange reactions and that a β-lactoglobulin (monomeric) species with a free thiol group at Cys160 may play a significant role in the interprotein disulfide bonding that occurs in heated milk or whey protein systems. The differences in reaction products between the model systems and milk may be a consequence of factors such as the heating conditions, the nature of the reactions (oxidative interaction compared with thiol-disulfide interchange reactions), and the fact that the κ-casein in milk is found on the casein micelles whereas in the model system, it is not (Lowe et al., 2004). Because of the C-terminal location of Cys160 in β-lactoglobulin, when this Cys is in the free thiol form and not linked to Cys66, it may be able to productively react with the disulfide bonds of κ-casein to give stable whey protein-κ-casein aggregates (Lowe et al., 2004).

Relationships between denaturation/interactions and functional properties

353

Relationships between denaturation/interactions of the whey proteins in heated milk and the functional properties of milk products When milk is heated, there are numerous changes, including changes to the proteins, the milk salts (such as changes in the mineral equilibria between the colloidal and serum phases), and lactose. Many of the changes can involve more than one of the milk constituents (International Dairy Federation, 1995). The changes can be irreversible or reversible to various extents depending on the changes being monitored and the conditions of the heat treatment. Although the changes to the protein system are an important determinant of the functional properties of milk products, all other changes to the milk system should also be considered to obtain a full understanding of the relationships between heat treatments, interactions, and functional performance. However, as there are limited examples of changes to components other than the proteins and the functional behavior of milk products, this review is restricted to some examples of the relationships between the changes in the milk protein system and the functional performance of the milk.

Examples of the relationships between whey protein denaturation and the functional properties of milk In the early days of milk powder manufacture, it was recognized that the level of whey protein denaturation could be used as an index for the extent of heat treatment the milk had received during the manufacture of the milk powders. The functional properties of the milk products were related to some extent to the heat treatment that the milk had received during processing and therefore the level of whey protein denaturation (Harland and Ashworth, 1947; Larson et al., 1951; Harland et al., 1952). Even as early as 1952, the concept of “tailor-made” milk powders was discussed, where powders were processed to provide specific requirements, such as low heat powders for beverage applications and cottage cheese manufacture, and high heat powders for bakery applications (Harland et al., 1952). Although there were no standards of quality or processing at this time, it was recognized that the proper control of processing conditions, particularly preheating of the milk, was necessary to produce satisfactory products. Measurement of the level of whey protein denaturation could be used as an objective method for determining the suitability of milk (powder) products for particular commercial and functional applications. The heat treatment of milk, whether in liquid milk applications or prior to drying for milk powder manufacture, remains one of the major processes for manipulating the functional properties of milk products. Products such as milk powders are still generally classified according to the heat treatments received during manufacture using one of the derivatives of the WPNI test (Singh and Newstead, 1992; Kelly et al., 2003). With the extensive research on the denaturation of the whey proteins and the ability to predict the denaturation levels after defined heat treatments, it could be envisaged that the level of denaturation of the whey proteins could be used as an indicator of the functional properties of milk products. In a broad sense, this is true. For example, certain heat classifications of milk powders will give improved functionality over other classes of milk powders for particular applications.

354

9. The whey proteins in milk

Some of the general applications of different heat-classified milks and their functional uses, in particular for milk powder products, have been summarized in numerous publications (Singh and Newstead, 1992; International-Dairy-Federation, 1995, 1996; Kelly et al., 2003). However, an infinite number of temperature and heating time combinations to denature the whey proteins when milk is heated are available. As a consequence, specific correlations between the level of whey protein denaturation and the functional properties of milk across all possible heating conditions and milk sources do not exist. For example, the WPNI method was developed for assessing the suitability of milk powders for use in bakery applications; however, it was noted that a powder with a low WPNI did not always correspond to good baking qualities (Harland and Ashworth, 1947). Some of these variations are due to factors such as natural variations in the initial whey protein levels in the milk (Harland et al., 1955; Sanderson, 1970b); however, others are due to the methods of heat treatment during milk processing. As such, the WPNI or level of whey protein denaturation is, at best, a guide for the suitability of powders for specific applications or an in-factory guide on processing conditions. Many manufacturers impose additional specifications to the milk powders to ensure suitability in their specific applications (Sanderson, 1970c; Singh and Creamer, 1991a). Some of the most detailed studies on the relationship between the functional performance of milk and the heat treatment conditions or whey protein denaturation levels have been reported for acid gel or yogurt systems. Parnell-Clunies et al. (1986) showed correlations between the level of whey protein denaturation and the firmness and apparent viscosity of yogurt, regardless of the method used to heat the milk (batch [85°C], high temperature, short time [98°C], and UHT [140°C] heating systems for different holding times). However, other properties, such as water-holding capacity/syneresis, were more dependent on the heating system used, and it was concluded that high levels of whey protein denaturation in milk were not necessarily associated with an improved water-holding capacity in yogurt. In extensive studies, Dannenberg and Kessler (1988a,c) examined the relationship between the denaturation level of whey proteins in the milk and the functional performance (firmness, flow properties, and syneresis) of the milk in set yogurt applications. There was a clear relationship between whey protein denaturation and yogurt firmness, with a higher firmness at higher levels of whey protein denaturation. Similar results were obtained for the flow properties of the yogurt. However, very high levels of whey protein denaturation appeared to be detrimental, with a decrease in the firmness and flow properties at denaturation levels above about 95% (Dannenberg and Kessler, 1988c). For syneresis of the yogurt, a negative relationship between whey protein denaturation and the level of serum expelled from the yogurt was observed (Dannenberg and Kessler, 1988a). Despite the apparent correlations between denaturation and firmness, flow properties, or syneresis, there were significant variations at each denaturation level, indicating that the temperature of heating used to denature the whey proteins, rather than just the whey protein denaturation level, may be an important factor in determining the functional performance in acid gels. In a study on reconstituted whole milk, McKenna and Anema (1993) also observed a positive correlation between the denaturation of the whey proteins in the milk and the firmness of the yogurt made from the milk regardless of whether the heat treatment was performed before or after powder manufacture (Fig. 9.10A). However, when individual heating conditions were examined, it was also noted that excessive heat treatment of the milk could be

Relationships between denaturation/interactions and functional properties

355

FIG. 9.10 Relationship between the level of whey protein denaturation in reconstituted whole milk and (A) the firmness and (B) the syneresis of acid gels prepared from the heated milks. The milks were (●) heated only before powder manufacture, (.) heated only after reconstitution, or () heated both before powder manufacture and after reconstitution. Plotted from the data of McKenna, A.B., Anema, S.G., 1993. The effect of thermal processing during whole milk powder manufacture and after its reconstitution on set-yoghurt properties. In: International Dairy Federation Special Issue 9303. Protein & Fat Globule Modifications by Heat Treatment, Homogenization & Other Technological Means for High Quality Dairy Products. International Dairy Federation, Brussels, Belgium, pp. 307–316.

detrimental to the firmness of the set yogurt (McKenna and Anema, 1993). A less clear relationship between syneresis and the level of whey protein denaturation was observed, with the level of syneresis appearing to have a greater dependence on the heating conditions (temperature, time, and before/after reconstitution) than on the level of denaturation itself (Fig. 9.10B; McKenna and Anema, 1993), which supports the findings of Parnell-Clunies et al. (1986).

Examples of the relationships between the level of interactions of whey proteins with κ-casein/casein micelles and the functional properties of milk A major limitation in using whey protein denaturation as an index of the functional properties of milk is that it does not consider the subsequent interaction reactions of the denatured whey proteins. These interactions will be dependent on the conditions of denaturation such as temperature and time as well as on the properties of the milk such as pH, concentration, and composition. It is more complex to investigate these aggregation reactions, as there are potentially numerous pathways and there is great difficulty in isolating and characterizing the specific reaction products. However, in recent years, some effort has been made in identifying the interaction reactions of the denatured whey proteins with other components in milk and, in some cases, their effects on the functional properties of the milk.

356

9. The whey proteins in milk

Acid-induced aggregation/gelation of heated milk Anema et al. (2004a) showed that small changes in the pH of milk from the natural value at the time of heating markedly affected the properties of acid gels prepared from these heated milks. During acidification of the heated pH-adjusted milks, the acid gelation profiles were progressively shifted to higher firmness as the pH at heating was increased (Fig. 9.11A) so that the final firmness of the acid gels (measured as the storage modulus, G0 , at pH 4.2) was almost doubled as the pH at heating was increased from pH 6.5 to pH 7.1 (Fig. 9.11). The effect was particularly pronounced in the milks that were heated for times sufficient to fully denature the whey proteins (Fig. 9.11B). This effect of small changes in the pH of the milk at the time of heating on the firmness of acid gels prepared from heated milk has been independently confirmed (Lakemond and van Vliet, 2005; Rodriguez del Angel and Dalgleish, 2006; Lakemond and van Vliet, 2008a,b). Large-strain deformation experiments were also conducted on the set gels. In these experiments, the strain was increased at a constant rate, and the stress was monitored until the gel structure yielded and the stress decreased. The maximum in the strain versus stress curves FIG. 9.11 (A) Changes in firmness (G0 ) with time after the addition of glucono-δ-lactone (GDL) for heated (90°C/30 min) skim milk samples. (B) Changes in the final firmness (final G0 ) for acid gels prepared from milk samples heated for various times at 90°C. The pH values of the milk samples prior to heating were ●, pH 6.5; , pH 6.55; ., pH 6.6; 4, pH 6.65; , pH 6.7; □, pH 6.9; and ⧫, pH 7.1. In all samples, the pH was readjusted back to pH 6.7 before addition of GDL, and this pH was reduced to about pH 4.2 after 5.5 h. Reproduced with permission from Anema, S.G., Lee, S.K., Lowe, E.K., Klostermeyer, H., 2004a. Rheological properties of acid gels prepared from heated pH-adjusted skim milk. J. Agric. Food Chem. 52, 337–343. Copyright (2004) American Chemical Society.

▪

Relationships between denaturation/interactions and functional properties

357

FIG. 9.12 Stress versus strain curves for acid gels prepared from heated (90°C/15 min) skim milk samples. The pH values of the milk samples prior to heating were ●, pH 6.5; , pH 6.6; ., pH 6.7; 4, pH 6.9; and , pH 7.1. The maximum in the stress represents the breaking point of the gel, and the stress and the strain at this point are considered to be the breaking stress and the breaking strain, respectively.

▪

was considered to be the point at which the gel structure broke (Fig. 9.12). The yield properties of the final set acid gels were affected by the pH at which the milk was heated. The yield stresses of the gels increased markedly, and the yield strains of the gels decreased slightly as the pH at which the milk was heated was increased (Fig. 9.12; Anema, 2008a; Lakemond and van Vliet, 2008a). In addition to influencing the final firmness of the acid gels, the pH at the heat treatment of the milk also influenced the pH at which the milk started to gel/aggregate during acidification (Vasbinder and de Kruif, 2003; Anema et al., 2004a; Anema et al., 2004b; Lakemond and van Vliet, 2005; Rodriguez del Angel and Dalgleish, 2006; Lakemond and van Vliet, 2008a,b). On subsequent acidification of milk samples that were heated over a pH range from 6.5 to 7.1, samples heated at higher pH (pH 7.1) started to gel at significantly higher pH than samples heated at lower pH. These effects were very dependent on the temperature at which the milks were acidified (Anema et al., 2004b). In further detailed studies on the properties of the acid gels prepared from milks heated at different pHs (Lakemond and van Vliet, 2008a), it was shown that the permeability coefficients of the gels increased as the pH of the milks at heating increased. In addition, adding thiol-blocking agents to the milks during acid gelation did not affect the firmness of the gels prepared from milks heated at low pH (about pH 6.20) but markedly reduced the firmness of the gels prepared from milks heated at higher pH (about pH 6.90). Based on the permeability, the effects of thiol-blocking agents, and the small and large-strain rheological results of the acid gels prepared from pH-adjusted heated milks, it was concluded that acid gels from milks heated at low pH had a finer stranded structure with a higher strand curvature and that the gels contained fewer (intermolecular) disulfide bonds. These differences in gel properties accounted for the lower firmness of the gels prepared from milks heated at lower pH (Lakemond and van Vliet, 2008a). In concentrated milk, a similar effect of the pH of the milk at heating on acid gel firmness and yield stress was observed, with a marked increase in final gel firmness (Fig. 9.13A and B) and yield stress (Fig. 9.14A and B) as the pH of the milk at heating was increased from the natural pH and a marked decrease in firmness when the pH of the milk at heating was

358

9. The whey proteins in milk

FIG. 9.13 (A) Changes in storage modulus, G0 , with time after GDL addition for heated (80°C/30 min) 10% TS skim milk samples. ●, pH 6.48; , pH 6.55; ., pH 6.59; 4, pH 6.67; , pH 6.90; and □, pH 7.10. (B) Changes in storage modulus, G0 , with time after GDL addition for heated (80°C/30 min) 20% TS skim milk samples. ●, pH 6.28; , pH 6.39; ., pH 6.48; 4, pH 6.75; and , pH 6.95. (C) Percentage change in final G0 versus change in pH at heating from the natural pH of the milk. ●,  10% TS milk samples; ., 4 15% TS milk samples; , □ 20% TS milk samples; and ⧫, ◇ 25% TS milk samples. Open symbols, samples at the natural pH; filled symbols, samples that were pH adjusted before heating. Reproduced with permission from Anema, S.G., 2008a. Effect of milk solids concentration on the gels formed by the acidification of heated pH-adjusted skim milk. Food Chem. 108, 110–118. Copyright (2008) Elsevier.

▪

▪

▪

Relationships between denaturation/interactions and functional properties

359

FIG. 9.14 (A) Stress versus strain curves for acid gels prepared from heated (80°C/30 min) 10% TS skim milk samples. ●, pH 6.48; , pH 6.55; ., pH 6.59; 4, pH 6.67; , pH 6.90; and □, pH 7.10. (B) Stress versus strain curves for acid gels prepared from heated (80°C/30 min) 20% TS skim milk samples. ●, pH 6.28; , pH 6.39; ., pH 6.48; 4, pH 6.75; and , pH 6.95. (C) Percentage change in breaking stress versus change in pH at heating from the natural pH of the milk. ●, 10% TS milk samples; ., 15% TS milk samples; , 20% TS milk samples; and ⧫, 25% TS milk samples. Reproduced with permission from Anema, S.G., 2008a. Effect of milk solids concentration on the gels formed by the acidification of heated pH-adjusted skim milk. Food Chem. 108, 110–118. Copyright (2008) Elsevier.

▪

▪

▪

360

9. The whey proteins in milk

decreased (Anema, 2008a). As the natural pH of milk decreases when milk is concentrated, the percentage changes in firmness or yield stress of the gels from that obtained at the natural pH were plotted against the relative change in pH; it was found that the firmnesses (Fig. 9.13C) or yield stresses (Fig. 9.14C) of the gels at all milk concentrations fell on a single curve. This indicates that changing the pH of milk at heating had the same effect on the final firmnesses or yield stresses of the acid gels at all milk concentrations (Anema, 2008a). When low levels of starch (up to 1% w/w) were added to milks prior to heating and acidification, the effect of pH on gel firmness was similar to that of milk without starch addition; however, the firmness of the gels increased as the starch level increased (Oh et al., 2007). The gelatinized starch absorbed the aqueous phase in the milk and consequently increased the density of the protein network in the acid gels, increasing the firmness in a similar fashion to increasing the concentration of the milk. However, when higher levels of starch were added to the milk (>1% w/w), the pH effect diminished, although the gels were still firmer as the starch level increased (Oh et al., 2007). At these high levels of starch, the viscosity of the milk increased markedly after heating because of the gelatinization of the starch and the leaching of amylose into the continuous phase. It was proposed that the high viscosity of the continuous phase affected the diffusion of protein components during heating and subsequent acidification and that this in turn changed the network structure formed in the acid gel, diminishing the importance of serum-phase components (Oh et al., 2007). The changes in acid gel firmness or yield stress on changing the pH at heating of the milk could not be related solely to the level of whey protein denaturation (Fig. 9.11A). Small changes in the pH of the milk before heating markedly affect the distribution of the denatured whey proteins and κ-casein between the colloidal and serum phases in milk at its natural concentration (Figs. 9.4 and 9.5; Anema and Klostermeyer, 1997; Anema and Li, 2003a,b; Vasbinder and de Kruif, 2003; Rodriguez del Angel and Dalgleish, 2006; Lakemond and van Vliet, 2008b), and similar effects were observed in concentrated milks (Anema, 2008a; Chandrapala et al., 2010). Although heating milk prior to acidification markedly increased the firmness of the acid gels (i.e., acid gels prepared from heated milks always had a considerably higher firmness than acid gels prepared from unheated milks (Dannenberg and Kessler, 1988c; Lucey et al., 1997; Lucey and Singh, 1998)), the distribution of the denatured whey proteins and κ-casein between the colloidal and serum phases also appeared to influence the firmness of the acid gels. When the final firmness of the acid gels was plotted against the level of nonsedimentable denatured whey proteins in the milk, the results for all pH values fell on a single curve for milk at its natural concentration (Fig. 9.15). Similarly for concentrated milks, when the percentage change in final gel firmness (Fig. 9.16A) or yield stress (Fig. 9.16B) was plotted against the level of nonsedimentable denatured whey proteins in the milk, the results for all pHs and milk concentrations fell on a single curve (Anema et al., 2004a; Anema, 2008a). From these results, it was concluded that, although the denatured whey proteins that associate with the micelles have a significant effect on the final firmness of acid gels, the denatured whey proteins that remain in the serum appear to have a more dominant influence over the final firmness than those associated with the casein micelles (Anema et al., 2004a; Anema, 2008a). There was no relationship between the final gel strength and level of whey protein denaturation in the milk used to prepare the acid gel. For example, in samples where virtually all the whey proteins were denatured, there was a very large range of final gel strengths

Relationships between denaturation/interactions and functional properties

361

FIG. 9.15 Comparison between the final firmness (final G0 ) and the level of denatured whey protein (open symbols) and the level of nonsedimentable denatured whey protein (gray-filled symbols) for acid gels prepared from heated (90°C/30 min) skim milk samples. The pH values at heating of the milks were (●, ), pH 6.5; ( , □), pH 6.55; (▲, 4), pH 6.6; (., 5), pH 6.65; (⧫, ◇), pH 6.7; ( , ⊙), pH 6.9; and ( , &  ), pH 7.1. Reproduced with permission from Anema, S.G., Lee, S.K., Lowe, E.K., Klostermeyer, H., 2004a. Rheological properties of acid gels prepared from heated pH-adjusted skim milk. J. Agric. Food Chem. 52, 337–343. Copyright (2004) American Chemical Society.

▪

FIG. 9.16 (A) Relationship between the change

in final G0 for acid skim milk gels and the level of nonsedimentable denatured whey protein in heated (80°C/30 min), pH-adjusted skim milk. (B) Relationship between the change in breaking stress for acid skim milk gels and the level of nonsedimentable denatured whey protein in heated (80°C/30 min), pH-adjusted skim milk. ●, 10% TS milk samples; ., 15% TS milk samples; , 20% TS milk samples; and ⧫, 25% TS milk samples. Reproduced with permission from Anema, S.G., 2008a. Effect of milk solids concentration on the gels formed by the acidification of heated pH-adjusted skim milk. Food Chem. 108, 110–118. Copyright (2008) Elsevier.

▪

362

9. The whey proteins in milk

(Fig. 9.15). However, the final gel strength for acid gels prepared from milks in which all the denatured whey proteins were in the serum phase (i.e., 100% nonsedimentable denatured whey proteins) was found to be essentially a factor of two higher than that for acid gels prepared from milks in which all the whey proteins were associated with the casein micelles (i.e., 0% nonsedimentable denatured whey proteins; Fig. 9.15; Anema et al., 2004a; Anema, 2008a). This effect was observed over a wide range of milk concentrations (Anema, 2008a). Rodriguez del Angel and Dalgleish (2006) separated the nonsedimentable whey protein-κcasein aggregates from milks heated at different pHs using size exclusion chromatography and related the peak area of these aggregates to the firmness of the acid gels. They also concluded that the gel firmness appeared to be strongly dependent on the formation of soluble complexes in the milks and that there appeared to be a linear relationship between the level of soluble aggregates in the heated milk and the final strength of the acid gels. Based on these results, a hypothesis on the roles of the nonsedimentable and micelle-bound denatured whey protein-κ-casein aggregates has been developed (Anema et al., 2004a,b; Rodriguez del Angel and Dalgleish, 2006; Donato et al., 2007a; Anema, 2008a; Lakemond and van Vliet, 2008a,b). The increased pH of gelation and the increased acid gel strength of heated milk when compared with unheated milk have been attributed to the incorporation of the whey proteins and casein (micelles) in the acid gel structure during the acidification of milk (Lucey et al., 1997; Lucey, 2002; Graveland-Bikker and Anema, 2003). In milk, the casein is insoluble at its isoelectric point (about pH 4.6), whereas the native whey proteins remain soluble at all pHs. Therefore, unheated milk starts to aggregate when the milk pH approaches the isoelectric point of casein, and visible gelation is observed at about pH 4.9. However, for heated milk, the denatured whey proteins are insoluble at their isoelectric points (about pH 5.3 for β-lactoglobulin, the major whey protein). Therefore, on the acidification of heated milk, the proteins will start to aggregate at a much higher pH, closer to the isoelectric points of the whey proteins. As a consequence, the contribution of the denatured whey proteins to the acid gel structure and the firmness of the acid gels is markedly higher than that observed for unheated milk (Lucey et al., 1997; Graveland-Bikker and Anema, 2003). The pH at heating the milk will produce casein micelle particles with markedly different compositions (Figs. 9.5 and 9.6; Anema and Li, 2003b; Vasbinder and de Kruif, 2003; Anema, 2008a). Therefore, on the acidification of milks heated at different pHs, different aggregation and gelation behaviors are observed. Originally, it was hypothesized that, for milks heated at high pH, the serum-phase denatured whey protein-κ-casein complexes may aggregate separately and at a higher pH than the casein micelles because of the higher isoelectric points of the serum-phase protein components when compared with the κ-casein-depleted casein micelles. Based on this, it was proposed that the pH at which aggregation occurs will be progressively shifted to higher pH as the heating pH and the concentration of the serum-phase denatured whey protein-κ-casein complexes is increased (Anema et al., 2004a; Rodriguez del Angel and Dalgleish, 2006; Guyomarc’h et al., 2009). However, subsequent experimental evidence does not support this hypothesis. In a study in which the serum-phase and colloidal-phase protein aggregates were labeled with different fluorescent dyes before remixing and acidification, it was not possible to identify separate aggregation stages of the different fractions at the early stages of gelation. In addition, the final gel had colocalized serum-phase and colloidal-phase aggregates (Guyomarc’h et al., 2009).

Relationships between denaturation/interactions and functional properties

363

A further detailed study examined the reassociation behavior of the caseins and whey proteins during the slow acidification of skim milk that was adjusted to pHs from 6.5 to 7.1 before heating at 20–120°C for 10 min (Anema and Li, 2015). Prior to acidification, these skim milk samples had various levels of denatured whey protein and different levels of κ-casein and denatured whey proteins distributed between the serum and colloidal phases. On slow acidification, the dissociated caseins along with the denatured whey proteins aggregated with the κ-casein-depleted casein micelles once the pH was below about 5.4. There was no evidence for the separate aggregation of the denatured whey protein-κ-casein complexes or the κ-caseindepleted casein micelles (Fig. 9.17). As both serum-phase and colloidal-phase components are involved in the aggregation/gelation process at pHs below 5.4, it was unlikely that the earlier gelation of milks with high levels of serum-phase denatured whey protein-κ-casein complexes was due to a two-step gelation process. These two studies indicate that the earlier gelation of milk samples heated at high pH was a consequence of the destabilization of the entire milk protein system (Guyomarc’h et al., 2009; Anema and Li, 2015). Thus, the serum-phase denatured whey protein-κ-casein complexes aggregate at a higher pH because of the higher isoelectric points of these serum-phase complexes. In addition, the dissociation of κ-casein from the casein micelles will reduce the density of the surface hairy layer. This may cause the surface hairy layer to collapse at a higher pH, or this layer may have a reduced efficiency in stabilizing the casein micelles as the pH is decreased. As a consequence, the serum-phase denatured whey protein-κ-casein complexes and the κ-casein-depleted casein micelles aggregate at a pH that is markedly higher than that observed for the native casein micelles or for casein micelles in milk heated at a lower pH and therefore with low levels of serum-phase denatured whey protein-κ-casein complexes (Fig. 9.17). The firmness of acid gels can be related to the number and the properties of the contact points between the protein components in the acid gel (van Vliet and Keetals, 1995; Lucey et al., 1997; Mellema et al., 2002; Lakemond and van Vliet, 2008a). As the pH at heating of the milk is increased, the level of serum-phase denatured whey protein-κ-casein complexes increases, and therefore, there are a greater number of particles to aggregate during the subsequent acidification to form the acid gels. There is also the potential for the formation of a more complex acid gel structure when the milk is heated at high pH, at which there are high levels of serum-phase denatured whey protein-κ-casein complexes, than when the milk is heated at low pH, when most of the denatured whey protein and κ-casein are associated with the casein micelles. In the latter case, the acid gel process will probably involve only entire whey protein-casein micelle complexes. Therefore, there may be fewer contact points in the acid gels formed from milk with the denatured whey proteins associated with the micelles than in those formed from milk with soluble denatured whey proteins, and hence, a gel with a lower firmness is observed (Anema et al., 2004a,b; Lakemond and van Vliet, 2008a). The large-strain deformation properties also give some indication of the types of bonds involved in the acid gel network. As the pH at heating was increased, the breaking stress of the acid gels prepared from the heated milks was found to increase markedly; however, the breaking strain was virtually unchanged (Figs. 9.12 and 9.14; Anema, 2008a; Lakemond and van Vliet, 2008a). For a gel to break on increasing the strain, the strands within the gel network are first straightened and then stretched until rupture of the strands or the bonds within the strands (van Vliet and Walstra, 1995; Mellema et al., 2002; Lakemond and van Vliet, 2008a). Therefore, the breaking strain is dependent on factors such as the degree of curvature of the

364

9. The whey proteins in milk

FIG. 9.17 Levels of casein and whey proteins in sera of treated milks. The milk samples had an initial pH of (A) 6.5 or (B) pH7.1. The proteins measured were (i) αS-casein, (ii) β-casein, (iii) κ-casein, (iv) β-lactoglobulin, and (v) α-lactalbumin. The sera were obtained from milks that were treated at (●) 20°C, () 60°C, (.) 80°C, (4) 100°C, and ( ) 120°C for 10min before acidification. Reproduced with permission from Anema, S.G., Li, Y., 2015. Reassociation of dissociated caseins upon acidification of heated pH-adjusted skim milk. Food Chem. 174, 339–347. Copyright (2015) Elsevier.

▪

Relationships between denaturation/interactions and functional properties

365

strands, with a higher breaking strain when the strands have greater curvature. As the breaking strain of the acid gels changed only slightly with the pH at heating of the milk, despite the marked change in final firmness (Figs. 9.12 and 9.14), this indicates that the relative curvature of the individual strands within the gel network was similar for all acid gel samples. The types of bonds involved in the acid gel network will have an influence on the breaking stress (van Vliet and Walstra, 1995; van Vliet, 1996; Mellema et al., 2002; Lakemond and van Vliet, 2008a). The breaking of strands containing covalent bonds requires a greater force than the breaking of strands held together by noncovalent bonds, as covalent bonds have higher bond energies. Therefore, a change in the number or distribution of covalent bonds within the gel network may explain the differences in breaking stress as the pH of the milk at heating was changed (Figs. 9.12 and 9.14). It seems unlikely that the difference in breaking stress can be due to a greater degree of disulfide bonding within the gelled sample; although thiol-disulfide exchange reactions may continue during acidification (Vasbinder et al., 2003b), the physical number of disulfide bonds is unlikely to be markedly different between the samples. The denatured whey proteins, along with some of the κ-casein, are progressively transferred to the serum phase when the pH of the milk is increased before heating (Figs. 9.5 and 9.6). As these interactions involve disulfide bonding, this indicates that the interaction between the denatured whey proteins and κ-casein is transferred from the colloidal phase (casein micelle) to the serum phase as the pH of the milk at heating is increased. On subsequent acidification, both nonsedimentable and colloidal-phase denatured whey proteins are incorporated in the acid gel structure. The nonsedimentable denatured whey protein-κ-casein complexes can form strands that may be involved in interconnecting the colloidal particles. As the nonsedimentable aggregates are disulfide bonded, those samples heated at high pH and with high levels of nonsedimentable whey protein-κ-casein aggregates will have a greater number of these strands interconnecting the residual casein micelles. In contrast, samples heated at lower pH will have the denatured whey proteins predominantly associated with the casein micelles and therefore will have fewer of the whey proteinκ-casein aggregates interconnecting the colloidal particles. Therefore, samples heated at higher pH may have a greater number of disulfide bonds interconnecting the colloidal particles and therefore a higher breaking stress, whereas for samples heated at lower pH, most of the disulfide bonds are on the colloidal particles, and fewer disulfide bonds interconnect the colloidal particles, and this may explain the lower breaking stress (Figs. 9.12 and 9.14). Chymosin-induced aggregation/gelation of heated milk It should be noted that the pH of milk has a marked effect on the action of chymosin in destabilizing the system (Walstra and Jenness, 1984; Walstra et al., 1999). In all studies in which the pH of the milk at heating is discussed in relation to chymosin treatment, after the heat treatment, the pHs of the milks were readjusted back to the natural pH so that effects of pH on the enzyme activity were eliminated. Changing the pH of milk at heating changed the interactions between denatured whey proteins and the casein micelles; however, this did not appear to influence the chymosininduced gelation behavior of heated milk to any great extent (Anema et al., 2007, 2011; Kethireddipalli et al., 2010, 2011). The rate of release of glycomacropeptide (GMP) was similar in all milks regardless of whether they were unheated or heated, and the pH at heating had almost no effect on GMP release (Vasbinder et al., 2003a; Anema et al., 2007; Kethireddipalli

366

9. The whey proteins in milk

et al., 2011). When profiles of chymosin-induced gelation were considered, all heated milks had very long gelation times and formed very weak gels compared with unheated milks (Fig. 9.18; Anema et al., 2007, 2011; Kethireddipalli et al., 2010). These results indicated that there was a retardation in the chymosin-induced gelation regardless of whether the denatured whey proteins were associated with the caseins micelles (as observed on heating milk at low pH) or in the serum phase and associated with κ-casein that had dissociated from the casein micelles (as observed on heating milk at high pH). The inhibition of the gelation process was therefore not due to steric or charge effects of the denatured whey proteins that are associated with κ-casein at the casein micelle surface. The denatured whey proteins, whether as serum-phase complexes with κ-casein or associated with κ-casein at the casein micelle surface, interfere with the aggregation process and therefore increase the gelation time (Anema et al., 2007; Kethireddipalli et al., 2010). This was confirmed by experiments in which serum and colloidal phases from unheated and heated milks were exchanged, as it was shown that denatured whey proteins, whether in the serum phase or associated with casein micelles, inhibited the chymosin-induced gelation of milk. However, this inhibition may be more complex as it was also shown that heated casein micelles in the absence of whey proteins and nonprotein serum components from heated milks also inhibited gelation (Kethireddipalli et al., 2010). From these observations, it was concluded that the inhibition of chymosin-induced gelation was complex for heated milks and may have been due to combined effects of heat on casein micelles, denatured whey proteins (both serum phase and colloidal phase), and nonprotein serum components (Kethireddipalli et al., 2010). The chymosin-induced gelation of pH-adjusted, heated milks did not reveal differences; however, this is not an objective measure of the destabilization of the system as it requires the formation of an interconnected network structure, and it is possible for the casein micelles to be destabilized without forming a gel. By monitoring the particle size changes of the milk during chymosin treatment, it was possible to examine the early stages of the aggregation process (Anema et al., 2011). For all milk samples, after adding chymosin, the size initially decreased slightly because of the cleavage of GMP from κ-casein (lag phase) and then increased as the destabilized particles aggregated (aggregation phase; Fig. 9.19). FIG. 9.18 Changes in storage modulus (G0 ) with time after addition of rennet to unheated milks (solid symbols) and skim milk samples heated at 90°C for 30 min (open symbols). The pH values at heating of the milks were (●, ) pH 6.5; (.,4) pH 6.7; ( , □) pH 6.9; and (⧫, ◇) pH 7.1. All samples were readjusted back to the natural pH (pH 6.67) before the addition of rennet (40 μL of 1:3 diluted rennet per 1.3 mL of milk). Reproduced with permission from Anema, S.G., Lee, S.K., Klostermeyer, H., 2011. Rennet-induced aggregation of heated pH-adjusted skim milk. J. Agric. Food Chem. 59, 8413–8422. Copyright (2011) American Chemical Society.

▪

Relationships between denaturation/interactions and functional properties

367

FIG. 9.19 Changes in particle size on the rennet treatment of (A) unheated skim milk, (B) skim milk heated at 90°C for 2.5 min, and (C) skim milk heated at 90°C for 30 min. Milk samples (20 μL) were diluted in Ca-imidazole buffer (1 mL) before the addition of rennet (10 μL of 1:120 diluted rennet). The pHs of the milk samples were (●) 6.5, () 6.6, (.) 6.7, (4) 6.9, and ( ) 7.1. (D) Effect of pH on the time taken for the particle size to increase by 50 nm from the initial size (T50) after the addition of rennet to the milk samples. The milk samples were unheated (●) or heated at 90°C for () 1 min, (.) 2.5 min, (4) 5 min, ( ) 10 min, (□) 15 min, and (⧫) 30 min. Reproduced with permission from Anema, S.G., Lee, S.K., Klostermeyer, H., 2011. Rennet-induced aggregation of heated pH-adjusted skim milk. J. Agric. Food Chem. 59, 8413–8422. Copyright (2011) American Chemical Society.

▪

▪

368

9. The whey proteins in milk

Milks heated at temperatures below those at which whey proteins denature (60°C) had short lag phases and rapid aggregation phases; these were similar to those from unheated milks and were only slightly affected by the pH of the milk at heating. However, milks heated at higher temperatures (>60°C) and at low pH (e.g., pH 6.5) had extremely long lag phases, and once destabilized, the particles aggregated at a slow rate. Increasing the pH at heating shortened the lag phase and increased the rate at which the particles aggregated. In fact, samples heated at pH 7.1 had lag phases and aggregation rates not dissimilar to those of unheated milks (Anema et al., 2011). Similarly, for milks heated at 90°C for different times, the lag phase decreased, and the rate of aggregation increased markedly as the pH of the milk at heating was increased from pH 6.5 to pH 7.1. In fact, the lag phases and aggregation rates for the milks heated at pH 7.1 were not substantially different from those for unheated milks (Fig. 9.19). The effect of pH was greater as the heating time was increased, predominantly as a result of an increased lag phase and a decreased aggregation rate of the samples heated at lower pH when compared with those heated at higher pH (Anema et al., 2011; Fig. 9.19). For milk samples heated at high temperatures, there were positive correlations between the aggregation time and the level of whey protein or κ-casein that was associated with the casein micelles (Fig. 9.20). It was proposed that casein micelles that were substantially coated in denatured whey proteins and κ-casein (obtained on heating milks at low pH) were more resistant to chymosin-induced aggregation than casein micelles with low levels of denatured whey proteins and κ-casein on the casein micelle surface (as observed on heating milks at high pH; Anema et al., 2011). It was also proposed that, although the casein micelles in heated milks aggregated on chymosin treatment, the denatured whey proteins, whether associated with the micelles or in the serum phase, inhibited the gelation of the milk by stabilizing the surface of aggregated particles and preventing or slowing the formation of an interconnected network structure (Anema et al., 2011). Sedimentation and age gelation of UHT milk samples UHT milk is microbiologically sterile and can usually be stored for long periods; however, many chemical and physical changes can occur over time, and these can cause instability such as excessive sedimentation or age gelation. Excessive sedimentation is the formation of a thick layer of proteinaceous material at the bottom of the pack, and this can occur relatively soon after the manufacture of the milk. For age gelation, the milk is usually stable for a period of several months, after which the contents can gel spontaneously to form a three-dimensional proteinaceous network within the pack. Both excessive sedimentation and age gelation are unacceptable and effectively end the shelf life of the UHT milk. A large number of factors are known to affect the sedimentation propensity of UHT milk, such as milk composition, milk concentration, the type (indirect or direct) and severity of the UHT treatment, the pH and ionic calcium levels of the milk, and whether additives such as calcium chelators have been added to the milk (Deeth and Lewis, 2016, 2017; Anema, 2018b). Recent studies have shown that both indirect and direct UHT processing of milk cause significant levels of κ-casein to dissociate from the casein micelles, and this dissociated κ-casein associates with the denatured whey proteins in the serum phase. The sediments from these UHT milk samples had a similar composition regardless of the pH or ionic calcium level of the milk or whether the sediment level was high or low. The sediment was formed from the

369

Relationships between denaturation/interactions and functional properties

FIG. 9.20 Relationship between the time taken for the particle size to increase by 50 nm (T50) and (A) the percentage of micelle-bound β-lactoglobulin, (B) the percentage of micelle-bound α-lactalbumin, (C) the percentage of micelle-bound κ-casein (all data), and (D) the percentage of micelle-bound κ-casein (selected data with high levels of denatured whey proteins). The milk samples were heated either at 90°C for various times (open symbols) or at different temperatures for 30 min (filled symbols). The pHs of the samples were (, ●) 6.5, (◇) 6.55, () 6.6, (⊠) 6.65, (□, ) 6.7, (4, ▲) 6.9, and (5, .) 7.1. Reproduced with permission from Anema, S.G., Lee, S.K., Klostermeyer, H., 2011. Rennet-induced aggregation of heated pH-adjusted skim milk. J. Agric. Food Chem. 59, 8413–8422. Copyright (2011) American Chemical Society.

▪

κ-casein-depleted casein micelles and had only low levels of denatured whey proteins, although the level was higher for sediment from indirect UHT milk than for that from direct UHT milk (Fig. 9.21; Gaur, 2017; Gaur et al., 2018).

370

9. The whey proteins in milk

FIG. 9.21 Electrophoretic traces of (A) commercial indirect UHT milk samples, their supernatants, and their sediments and (B) commercial direct UHT milk samples, their supernatants, and their sediments. The red lines represent the milk samples, and the blue lines (dark gray in print version) represent their respective supernatants (i) or sediments (ii). The peaks correspond to (a) α lactalbumin, (b) β-lactoglobulin, (c) β-casein, (d) αS1-casein, and (e) κ-casein. (C) Percentages of αs-casein + β-casein combined (gray bar), κ-casein (red bar; dark gray in print version), and β-lactoglobulin + α lactalbumin combined (green bar; light gray in print version) in commercial indirect UHT (CI-1, CI-2, and CI-3) and direct UHT (CD-1, CD-2, and CD-3) milk (M), supernatant (SUP), and sediment (SED) samples. Reproduced with permission from Gaur, V., Schalk, J., Anema, S.G., 2018. Sedimentation in UHT milk. Int. Dairy J. 78, 92–102. Copyright (2018) Elsevier.

Relationships between denaturation/interactions and functional properties

371

The ionic calcium levels and the pH of the milk have a major impact on sedimentation, with increased sedimentation when the ionic calcium exceeds a certain critical level and/or when the pH is below a certain critical level (Fig. 9.22; Lewis et al., 2011; Anema, 2018b; Gaur et al., 2018). These results suggest that the κ-casein-depleted micelles are metastable and that, at sufficiently high ionic calcium levels or low pH values, the metastable micelles associate in the serum via calcium-mediated aggregation reactions, increasing their size and therefore their propensity to sediment. A number of mechanisms are known to cause age gelation (Datta and Deeth, 2001; Deeth and Lewis, 2016, 2017; Anema, 2018b). Two mechanisms involve proteolytic degradation of the proteins by either plasmin or heat-stable exogenous enzymes prior to the onset of gelation. These mechanisms have been studied in some detail and are reasonably well understood (Datta and Deeth, 2001; Deeth and Lewis, 2016, 2017; Anema, 2018b). In another “physicochemical” mechanism, age gelation can occur in the complete absence of proteolysis. The factors leading to this type of gelation are less well understood. A mechanism for age FIG. 9.22

Relationship between (A) pH and sediment level and (B) ionic calcium and sediment level for UHT milk samples: (●) direct UHT milk samples, () indirect UHT milk samples, (.) direct UHT milk sample with added sodium hexametaphosphate. The inserts show the indirect UHT milk samples on an expanded scale. Reproduced with permission from Gaur, V., Schalk, J., Anema, S.G., 2018. Sedimentation in UHT milk. Int. Dairy J. 78, 92–102. Copyright (2018) Elsevier.

372

9. The whey proteins in milk

gelation proposed by McMahon (1996), McMahon and Alleyne (1994) suggested that the β-lactoglobulin-κ-casein complex, initially on the micelle surface after UHT treatment, progressively dissociated from the casein micelles into the serum phase during storage. These β-lactoglobulin-κ-casein complexes formed the gel network on prolonged storage, entrapping the remaining κ-casein-depleted casein micelles when extensive gelation had occurred. However, a recent study has indicated that this mechanism may not be correct (Anema, 2017a). High levels of the β-lactoglobulin-κ-casein complex were in the serum phase immediately after the UHT treatment of milk, and these levels hardly changed during storage, even at the point of gelation or beyond. In addition, the gelled material was composed of κ-caseindepleted casein micelles, with little κ-casein or denatured whey proteins associated with this gelled phase. A new hypothesis was proposed, in which the UHT treatment forms metastable κ-casein-depleted casein micelles immediately after processing. If the ionic calcium levels are high and/or the pH levels are low at the time of processing, then sediments form rapidly, as detailed earlier (Anema, 2017a, 2018b). However, if the calcium activity and/or the pH are outside certain limits at the time of processing, the milk will appear to be stable for long periods. On storage, changes to the milk occur, such as a pH decline, changes in the ionic calcium levels, and a slow settling of the κ-casein-depleted casein micelles to the bottom of the pack. The high concentration of κ-casein-depleted casein micelles at the bottom along with the changes in pH and ionic calcium levels lead to the aggregation and eventual gelation of the milk, starting at the bottom of the pack and progressing upward as more κ-caseindepleted casein micelles settle. All milk samples will eventually gel by this mechanism; however, in most cases, the gelation will occur beyond the shelf life of the milk. It is the age gelation that occurs within the shelf life of the milk that causes complaint. From the studies on sedimentation (Gaur, 2017; Gaur et al., 2018) and the study on the physicochemical age gelation (Anema, 2017a) of UHT milk, it is evident that a self-consistent mechanism can be proposed to explain these two phenomena, as outlined in Fig. 9.23. The propensity for a UHT milk to sediment or gel is dependent on the pH and/or ionic calcium level at the time of manufacture. If the pH is below a critical level and/or the ionic calcium is above a critical level at the time of manufacture, the κ-casein-depleted casein micelles will aggregate in the milk serum through calcium-mediated bridging reactions, and these aggregates will sediment fairly rapidly during storage. However, if the pH and/or the ionic calcium levels are outside these critical levels after manufacture, the milk will not sediment significantly on storage. However, the slow settling of the κ-casein-depleted casein micelles along with changes to the pH and/or ionic calcium levels during storage will, at some stage, cause aggregation and gelation, starting at the bottom of the pack and progressing upward. Concentrate viscosity Skim milk samples at low pH (approximately pH 6.5) and heated at 90°C for up to 30 min had high levels of denatured whey proteins associated with the casein micelles and had a markedly higher viscosity than similarly heated milk samples at higher pH (Fig. 9.8B, Anema et al., 2004c). If similar viscosity differences could be achieved at higher milk solid concentrations, this may allow for more efficient production of milk powders as milk samples that maintain a lower viscosity could be evaporated to higher total solids before spray drying. A detailed study examined the rheological properties of concentrates (up to about 45% total solids) that were prepared by evaporating skim milk samples that were heated (90°C/30 min)

Relationships between denaturation/interactions and functional properties

FIG. 9.23

373

A self-consistent mechanism for the age gelation and sedimentation of UHT milk during storage. (1) Casein micelles and native whey proteins in unheated milk, with κ-casein on the casein micelle surface. (2) On UHT processing, κ-casein dissociates from the casein micelles, forming whey protein-κ-casein complexes in the serum and κ-caseindepleted casein micelles. (3) If the calcium activity is low and/or the pH is high during processing, the milk is initially stable. During storage, the environment changes so that the pH and the calcium activity move into an unstable range. The κ-casein-depleted casein micelles aggregate to form a gel. The whey protein-κ-casein complexes remain in the serum on initial gelation but may be trapped in the gel during extensive gelation. (4) If the calcium activity is high and/or the pH is low during processing, the milk is unstable. Aggregation of the κ-casein-depleted casein micelles occurs over days/weeks to form large particles that rapidly sediment. The whey protein-κ-casein complexes remain in the serum. These mechanisms were developed, with permission, from the results of Anema, S.G., 2017a. Storage stability and age gelation of reconstituted ultra-high temperature skim milk. Int. Dairy J. 75, 56–67. Gaur, V., 2017. Sedimentation Reduction in UHT Milk. PhD Thesis, University of Canterbury, Christchurch, New Zealand, and Gaur, V., Schalk, J., Anema, S.G., 2018. Sedimentation in UHT milk. Int. Dairy J. 78, 92–102. Copyright (2018) Elsevier.

374

9. The whey proteins in milk

at different pHs (pH 6.5–7.1). The viscosities of the concentrates followed power law behavior and therefore could be described by consistency coefficients and flow behavior indices (Anema et al., 2014). Milk concentrates of up to 40% total solids did have higher consistency coefficients than concentrates from unheated skim milk, whereas those from skim milk heated at pH 7.1 had lower consistency coefficients than those from the unheated milk. The concentrates from milk heated at intermediate pH had consistency coefficients between those from the milks heated at pH 6.5 and pH 7.1. These effects were observed for concentrates produced from the skim milk at the pH of heating and for those from the milk samples that were readjusted back to the natural pH after heating and before concentration. However, if the milk was concentrated to above 40% total solids, the consistency coefficients of the concentrates from heated milk samples converged so that they were similar regardless of the pH at heating, and they were significantly higher than those of the concentrates prepared from unheated milk (Fig. 9.24). It is proposed that, when the denatured whey proteins associate with the casein micelles, they increase the voluminosity of the micelles, leading to an increased viscosity and therefore consistency coefficient at all milk concentrations. In contrast, when the denatured whey proteins and κ-casein are in the serum, they are part of the continuous phase and do not contribute to the viscosity and therefore consistency coefficient until the milk is at very high concentrations (>40% total solids).

Examples of the effect of denaturing whey proteins separately from casein micelles on the functional properties of milk Compared with heated milks, different effects on functional properties can be observed when the whey proteins are denatured and aggregated separately from the casein micelles and then remixed. However, the effects of adding predenatured whey proteins to casein micelle suspensions or milk on the functional properties are dependent on the conditions under which the whey proteins are denatured. Lucey et al. (1998) showed that acid gels prepared FIG. 9.24 Consistency coefficients of skim milk concentrates measured at 20°C (filled symbols) and 40°C (open symbols). Concentrates were from (⧫) unheated skim milk or skim milk heated at (●, ) pH 6.5, (▲, 4) pH 6.7, and ( , □) pH 7.1 and 90°C for 30 min before being concentrated to different TS levels. Reproduced with permission from Anema, S.G., Lowe, E.K., Lee, S.K., Klostermeyer, H., 2014. Effect of the pH of skim milk at heating on milk concentrate viscosity. Int. Dairy J. 39, 336–343. Copyright (2014) Elsevier.

▪

Relationships between denaturation/interactions and functional properties

375

from milk samples in which the whey proteins were denatured in the presence of casein micelles had a markedly higher firmness than acid gels prepared from milk samples in which the whey proteins in the milk serum were predenatured and then added back to the casein micelle suspensions. In fact, in many cases, the samples with denatured whey proteins added back to the casein micelles produced acid gels with firmness similar to or only slightly higher than those of acid gels prepared from unheated milks. In another study, Schorsch et al. (2001) prepared model milk systems in which the whey proteins in a simulated milk serum either were heated in the presence of casein micelles or were heated separately and added back to the casein micelles. Acid gels were prepared from these model milk systems. It was shown that the acid-induced gelation occurred at a higher pH and in a shorter time when the whey proteins were denatured separately from the casein micelles than when the whey proteins were heated in the presence of the casein micelles. However, the gels formed were weaker and more heterogeneous because of the particulate nature of the denatured whey proteins. It was suggested that the large denatured whey protein aggregates, as formed when the whey proteins in the milk serum were heated separately from the casein micelles, hinder the formation of a casein gel network when the milk is subsequently acidified and that a weak acid gel with a heterogeneous structure results. When the whey proteins are heated in the presence of the casein micelles, the denatured whey proteins interact with the κ-casein at the casein micelle surface, and on subsequent acidification, the denatured whey proteincasein micelle complexes aggregate to form a firmer acid gel with a more homogeneous structure (Schorsch et al., 2001). This proposal is supported by early studies, which showed that the aggregation of the denatured whey proteins, and in particular β-lactoglobulin, formed large aggregate species when heated in the absence of κ-casein, whereas aggregation was limited when the whey proteins were heated in the presence of κ-casein (McKenzie et al., 1971). In contrast, if the whey proteins are denatured at relatively low protein concentrations, at low ionic strengths, and at a pH far from the isoelectric point (pH > 6.5), then soluble denatured whey protein polymers can be formed. The polymers are linear and can be induced to gel when salt is added or the pH is reduced (Britten and Giroux, 2001; Gustaw et al., 2006, 2009). When these whey protein polymers were added to heated skim milk and the preparations were acidified, the acid gels formed had markedly higher firmness and water-holding capacities than those from the skim milk or from milks heated with equivalent levels of native whey proteins (Britten and Giroux, 2001; Gustaw et al., 2006, 2009). The firmnesses and waterholding capacities were markedly higher than when the whey proteins were denatured in the milk serum and then added back to casein micelle suspensions or milk before acidification (Lucey et al., 1998; Schorsch et al., 2001). The isoelectric points of denatured whey protein complexes were chemically modified through succinylation or methylation of carboxyl groups (Morand et al., 2012a). These modified denatured whey protein complexes were added to whey protein-free milk suspensions to produce model heated milk systems, and the milks were subsequently acidified to form gels. The pH of gelation of these milks increased markedly as the isoelectric point of the whey protein complexes increased, supporting the proposition that it is the higher isoelectric point of the whey proteins that causes heated milks to gel at markedly higher pH than unheated milks (Lucey et al., 1997; Anema et al., 2004a). Interestingly, the final firmness of the gels was not markedly affected by the isoelectric point of the complexes (Morand et al., 2012a).

376

9. The whey proteins in milk

In similar studies, the hydrophobicities of denatured whey protein complexes were altered by acylation of lysine amino acids using anhydrides of various carbon chain lengths (Morand et al., 2012b). These complexes were also added to whey protein-free milk suspensions, and the milks were acidified to form gels. Increasing the hydrophobicity of the whey protein complexes also increased the pH at which the milks gelled and also increased the maximum firmness of the gels, although this maximum was not always at the same pH. Taken together, these results indicate that both the isoelectric properties and the hydrophobicity of the serumphase complexes influence the acid gelation properties of heated milk systems (Morand et al., 2011, 2012a,b).

Conclusions A considerable amount of work has gone into understanding the irreversible denaturation reactions of the whey proteins in heated milk systems. These detailed studies have produced models that allow reasonably accurate prediction of the level of whey protein denaturation in milks under a wide range of heating conditions, even in milk samples with markedly modified concentrations and compositions. However, with a few exceptions, monitoring of the whey protein denaturation levels provides only a crude indication of the functionality of the milk system. As a consequence, more recent research efforts have focused on trying to understand the specific interaction reactions of the denatured whey proteins with other proteins in the milk system. Early indications suggest that these types of studies on the interactions of denatured whey proteins may provide greater insights into the functional properties of heated milk products than can be obtained by monitoring just whey protein denaturation levels. These initial studies on protein interactions have been conducted under relatively narrowly defined conditions (temperatures, heating times, pH, milk concentrations, and milk compositions). It is likely that changes to these variables will markedly influence the interaction behavior and will explain the changes in functional behavior when the heating conditions are changed (even though the whey protein denaturation levels may be similar). Although studies on understanding the specific interactions between milk proteins, particularly in complex systems such as milk, are extremely difficult, these types of studies should continue to give useful insights into the behavior of milk proteins during heating and the functional behavior of the heated milk products.

References Anema, S.G., 1998. Effect of milk concentration on heat-induced, pH-dependent dissociation of casein from micelles in reconstituted skim milk at temperatures between 20 and 120°C. J. Agric. Food Chem. 46, 2299–2305. Anema, S.G., 2000. Effect of milk concentration on the irreversible thermal denaturation and disulfide aggregation of β-lactoglobulin. J. Agric. Food Chem. 48, 4168–4175. Anema, S.G., 2001. Kinetics of the irreversible thermal denaturation and disulfide aggregation of α-lactalbumin in milk samples of various concentrations. J. Food Sci. 66, 2–9. Anema, S.G., 2007. Role of κ-casein in the association of denatured whey proteins with casein micelles in heated reconstituted skim milk. J. Agric. Food Chem. 55, 3635–3642. Anema, S.G., 2008a. Effect of milk solids concentration on the gels formed by the acidification of heated pH-adjusted skim milk. Food Chem. 108, 110–118.

References

377

Anema, S.G., 2008b. On heating milk, the dissociation of κ-casein from the casein micelles can precede interactions with the denatured whey proteins. J. Dairy Res. 75, 415–421. Anema, S.G., 2009. The use of “lab-on-a-chip” microfluidic SDS electrophoresis technology for the separation and quantification of milk proteins. Int. Dairy J. 19, 198–204. Anema, S.G., 2017a. Storage stability and age gelation of reconstituted ultra-high temperature skim milk. Int. Dairy J. 75, 56–67. Anema, S.G., 2017b. The thermal denaturation of the total whey protein in reconstituted whole milk. Int. J. Dairy Technol. 70, 332–338. Anema, S.G., 2018a. Effect of whey protein addition and pH on the acid gelation of heated skim milk. Int. Dairy J. 79, 5–14. Anema, S.G., 2018b. Age gelation, sedimentation, and creaming in UHT milk: a review. Compr. Rev. Food Sci. Food Saf. 18, 140–166. Anema, S.G., Klostermeyer, H., 1997. Heat-induced, pH-dependent dissociation of casein micelles on heating reconstituted skim milk at temperatures below 100°C. J. Agric. Food Chem. 45, 1108–1115. Anema, S.G., Lee, S.K., Klostermeyer, H., 2006. Effect of protein, nonprotein-soluble components, and lactose concentrations on the irreversible thermal denaturation of β-lactoglobulin and α-lactalbumin in skim milk. J. Agric. Food Chem. 54, 7339–7348. Anema, S.G., Lee, S.K., Klostermeyer, H., 2007. Effect of pH at heat treatment on the hydrolysis of κ-casein and the gelation of skim milk by chymosin. LWT Food Sci. Technol. 40, 99–106. Anema, S.G., Lee, S.K., Klostermeyer, H., 2011. Rennet-induced aggregation of heated pH-adjusted skim milk. J. Agric. Food Chem. 59, 8413–8422. Anema, S.G., Lee, S.K., Lowe, E.K., Klostermeyer, H., 2004a. Rheological properties of acid gels prepared from heated pH-adjusted skim milk. J. Agric. Food Chem. 52, 337–343. Anema, S.G., Li, Y., 2000. Further studies on the heat-induced, pH-dependent dissociation of casein from the micelles in reconstituted skim milk. Lebensm. Wiss. Technol. 33, 335–343. Anema, S.G., Li, Y., 2003a. Association of denatured whey proteins with casein micelles in heated reconstituted skim milk and its effect on casein micelle size. J. Dairy Res. 70, 73–83. Anema, S.G., Li, Y., 2003b. Effect of pH on the association of denatured whey proteins with casein micelles in heated reconstituted skim milk. J. Agric. Food Chem. 51, 1640–1646. Anema, S.G., Li, Y., 2015. Reassociation of dissociated caseins upon acidification of heated pH-adjusted skim milk. Food Chem. 174, 339–347. Anema, S.G., Lloyd, R.J., 1999. Analysis of whey protein denaturation: a comparative study of alternative methods. Milchwissenschaft 54, 206–210. Anema, S.G., Lowe, E.K., Lee, S.K., 2004b. Effect of pH at heating on the acid-induced aggregation of casein micelles in reconstituted skim milk. Lebensm. Wiss. Technol. 37, 779–787. Anema, S.G., Lowe, E.K., Lee, S.K., Klostermeyer, H., 2014. Effect of the pH of skim milk at heating on milk concentrate viscosity. Int. Dairy J. 39, 336–343. Anema, S.G., Lowe, E.K., Li, Y., 2004c. Effect of pH on the viscosity of heated reconstituted skim milk. Int. Dairy J. 14, 541–548. Anema, S.G., McKenna, A.B., 1996. Reaction kinetics of thermal denaturation of whey proteins in heated reconstituted whole milk. J. Agric. Food Chem. 44, 422–428. Anema, S.G., Stanley, D.J., 1998. Heat-induced, pH-dependent behaviour of protein in caprine milk. Int. Dairy J. 8, 917–923. Baer, A., Oroz, M., Blanc, B., 1976. Serological studies on heat-induced interactions of α-lactalbumin and milk-proteins. J. Dairy Res. 43, 419–432. Britten, M., Giroux, H.J., 2001. Acid-induced gelation of whey protein polymers: effects of pH and calcium concentration during polymerization. Food Hydrocoll. 15, 609–617. Butikofer, U., Meyer, J., Rehberger, B., 2006. Determination of the percentage of α-lactalbumin and β-lactoglobulin of total milk protein in raw and heat treated skim milk. Milchwissenschaft 61, 263–266. Carver, J.A., Thorn, D.C., Ecroyd, H., Holt, C., 2017. Letter to the editor: a response to Horne and Lucey (2017). J. Dairy Sci. 100, 5121–5124. Chandrapala, J., Augustin, M.A., McKinnon, I., Udabage, P., 2010. Effects of pH, calcium-complexing agents and milk solids concentration on formation of soluble protein aggregates in heated reconstituted skim milk. Int. Dairy J. 20, 777–784.

378

9. The whey proteins in milk

Cho, Y., Singh, H., Creamer, L.K., 2003. Heat-induced interactions of β-lactoglobulin A and κ-casein B in a model system. J. Dairy Res. 70, 61–71. Corredig, M., Dalgleish, D.G., 1996a. The binding of α-lactalbumin and β-lactoglobulin to casein micelles in milk treated by different heating systems. Milchwissenschaft 51, 123–127. Corredig, M., Dalgleish, D.G., 1996b. Effect of temperature and pH on the interactions of whey proteins with casein micelles in skim milk. Food Res. Int. 29, 49–55. Corredig, M., Dalgleish, D.G., 1999. The mechanisms of the heat-induced interaction of whey proteins with casein micelles in milk. Int. Dairy J. 9, 233–236. Costa, F.F., Vasconcelos Paiva Brito, M.A., Moreira Furtado, M.A., Martins, M.F., Leal de Oliveira, M.A., Mendonc¸a de Castro Barra, P., Amigo Garrido, L., Siqueira de Oliveira dos Santos, A., 2014. Microfluidic chip electrophoresis investigation of major milk proteins: study of buffer effects and quantitative approaching. Anal. Methods 6, 1666–1673. Creamer, L.K., Berry, G.P., Matheson, A.R., 1978. Effect of pH on protein aggregation in heated skim milk. New Zeal. J. Dairy Sci. Technol. 13, 9–15. Creamer, L.K., Bienvenue, A., Nilsson, H., Paulsson, M., van Wanroij, M., Lowe, E.K., Anema, S.G., Boland, M.J., Jimenez-Flores, R., 2004. Heat-induced redistribution of disulfide bonds in milk proteins. 1. Bovine β-lactoglobulin. J. Agric. Food Chem. 52, 7660–7668. Creamer, L.K., Matheson, A.R., 1980. Effect of heat-treatment on the proteins of pasteurized skim milk. New Zeal. J. Dairy Sci. Technol. 15, 37–49. Croguennec, T., Bouhallab, S., Molle, D., O’Kennedy, B.T., Mehra, R., 2003. Stable monomeric intermediate with exposed Cys-119 is formed during heat denaturation of β-lactoglobulin. Biochem. Biophys. Res. Commun. 301, 465–471. Croguennec, T., Molle, D., Mehra, R., Bouhallab, S., 2004. Spectroscopic characterization of heat-induced nonnative β-lactoglobulin monomers. Protein Sci. 13, 1340–1346. Dalgleish, D.G., 2011. On the structural models of bovine casein micelles-review and possible improvements. Soft Matter 7, 2265–2272. Dalgleish, D.G., Corredig, M., 2012. The structure of the casein micelle of milk and its changes during processing. Annu. Rev. Food Sci. Technol. 3, 449–467. Dannenberg, F., Kessler, H.G., 1988a. Effect of denaturation of β-lactoglobulin on texture properties of set-style nonfat yoghurt. 1. Syneresis. Milchwissenschaft 43, 632–635. Dannenberg, F., Kessler, H.G., 1988b. Reaction kinetics of the denaturation of whey proteins in milk. J. Food Sci. 53, 258–263. Dannenberg, F., Kessler, H.G., 1988c. Effect of denaturation of β-lactoglobulin on texture properties of set-style nonfat yoghurt. 2. Firmness and flow properties. Milchwissenschaft 43, 700–704. Datta, N., Deeth, H.C., 2001. Age gelation of UHT milk—a review. Food Bioprod. Process. 79, 197–210. Deeth, H.C., Lewis, M.J., 2016. Protein stability in sterilized milk and milk products. In: McSweeney, P.L.H., O’Mahony, J.A. (Eds.), Advanced Dairy Chemistry-1B: Proteins: Applied Aspects, fourth ed. Springer Science and Business Media, New York, NY, USA, pp. 247–286. Deeth, H.C., Lewis, M.J., 2017. Changes during storage of UHT milk. In: High Temperature Processing of Milk and Milk Products. Wiley-Blackwell, Chichester, UK, pp. 261–319. Doi, H., Tokuyama, T., Kuo, F.H., Ibuki, F., Kanamori, M., 1983. Heat-induced complex-formation between κ-casein and α-lactalbumin. Agric. Biol. Chem. 47, 2817–2824. Donato, L., Alexander, M., Dalgleish, D.G., 2007a. Acid gelation in heated and unheated milks: interactions between serum protein complexes and the surfaces of casein micelles. J. Agric. Food Chem. 55, 4160–4168. Donato, L., Dalgleish, D.G., 2006. Effect of the pH of heating on the qualitative and quantitative compositions of the sera of reconstituted skim milks and on the mechanisms of formation of soluble aggregates. J. Agric. Food Chem. 54, 7804–7811. Donato, L., Guyomarc’h, F., 2009. Formation and properties of the whey protein/κ-casein complexes in heated skim milk—a review. Dairy Sci. Technol. 89, 3–29. Donato, L., Guyomarc’h, F., Amiot, S., Dalgleish, D.G., 2007b. Formation of whey protein/κ-casein complexes in heated milk: preferential reaction of whey protein with κ-casein in the casein micelles. Int. Dairy J. 17, 1161–1167. Duerasch, A., Wissel, J., Henle, T., 2018. Reassembling of alkali-treated casein micelles by microbial transglutaminase. J. Agric. Food Chem. 66, 11748–11756. Dupont, D., Muller-Renaud, S., 2006. Quantification of proteins in dairy products using an optical biosensor. J. AOAC Int. 89, 843–848.

References

379

Elfagm, A.A., Wheelock, J.V., 1978. Heat interaction between α-lactalbumin, β-lactoglobulin and casein in bovine milk. J. Dairy Sci. 61, 159–163. Euber, J.R., Brunner, J.R., 1982. Interaction of κ-casein with immobilized β-lactoglobulin. J. Dairy Sci. 65, 2384–2387. Fairise, J.F., Cayot, P., 1998. New ultrarapid method for the separation of milk proteins by capillary electrophoresis. J. Agric. Food Chem. 46, 2628–2633. Farrell, H.M., Malin, E.L., Brown, E.M., Qi, P.X., 2006. Casein micelle structure: what can be learned from milk synthesis and structural biology? Curr. Opin. Colloid Interface Sci. 11, 135–147. Fox, P.F., 2003. Milk proteins: general and historical aspects. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1: Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, pp. 1–48. Gaur, V., 2017. Sedimentation Reduction in UHT Milk. PhD Thesis, University of Canterbury, Christchurch, New Zealand. Gaur, V., Schalk, J., Anema, S.G., 2018. Sedimentation in UHT milk. Int. Dairy J. 78, 92–102. Gillece Castro, B.L., Stults, J.T., 1996. Peptide characterization by mass spectrometry. In: High Resolution Separation and Analysis of Biological Macromolecules, Part B.vol. 271, pp. 427–448. Gorman, J.J., Wallis, T.P., Pitt, J.J., 2002. Protein disulfide bond determination by mass spectrometry. Mass Spectrom. Rev. 21, 183–216. Gough, P., Jenness, R.J., 1962. Heat denaturation of β-lactoglobulins A and B. J. Dairy Sci. 45, 1033–1039. Graveland-Bikker, J.F., Anema, S.G., 2003. Effect of individual whey proteins on the rheological properties of acid gels prepared from heated skim milk. Int. Dairy J. 13, 401–408. Grindrod, J., Nickerson, T.A., 1967. Changes in milk proteins treated with hydrogen peroxide. J. Dairy Sci. 50, 142–146. Gubic, J., Tomic, J., Torbica, A., Ilicic, M., Tasic, T., Sˇaric, L., Popovic, S., 2016. Characterization of several milk proteins in domestic Balkan donkey breed during lactation, using lab-on-a-chip capillary electrophoresis. Chem. Ind. Chem. Eng. Q. 22, 9–15. Gustaw, W., Glibowski, P., Mleko, S., 2006. The rheological properties of yoghurt with incorporated whey protein aggregates/polymers. Milchwissenschaft 61, 415–419. Gustaw, W., Szwajgier, D., Mleko, S., 2009. The rheological properties of yoghurt with the addition of lyophilized polymerized whey protein. Milchwissenschaft 64, 60–64. Guyomarc’h, F., Jemin, M., Le Tilly, V., Madec, M.N., Famelart, M.H., 2009. Role of the heat-induced whey proteinκ-casein complexes in the formation of acid milk gels: a kinetic study using rheology and confocal microscopy. J. Agric. Food Chem. 57, 5910–5917. Guyomarc’h, F., Law, A.J.R., Dalgleish, D.G., 2003. Formation of soluble and micelle-bound protein aggregates in heated milk. J. Agric. Food Chem. 51, 4652–4660. Haque, Z., Kinsella, J.E., 1988. Interaction between heated κ-Casein and β-lactoglobulin: predominance of hydrophobic interactions in the initial stages of complex formation. J. Dairy Res. 55, 67–80. Haque, Z., Kristjansson, M.M., Kinsella, J.E., 1987. Interaction between κ-casein and β-lactoglobulin: possible mechanism. J. Agric. Food Chem. 35, 644–649. Harland, H.A., Ashworth, U.S., 1945. The preparation and effect of heat treatment on the whey proteins of milk. J. Dairy Sci. 28, 879–886. Harland, H.A., Ashworth, U.S., 1947. A rapid method for estimation of whey proteins as an indication of baking quality of nonfat dry-milk solids. Food Res. 12, 247–251. Harland, H.A., Coulter, S.T., Jenness, R., 1952. The effect of the various steps in the manufacture on the extent of serum protein denaturation in nonfat dry milk solids. J. Dairy Sci. 35, 363–368. Harland, H.A., Coulter, S.T., Jenness, R., 1955. Natural variation of milk serum proteins as a limitation of their use in evaluating the heat treatment of milk. J. Dairy Sci. 38, 858–869. Henry, G., Molle, D., Morgan, F., Fauquant, J., Bouhallab, S., 2002. Heat-induced covalent complex between casein micelles and β-lactoglobulin from goat’s milk: identification of an involved disulfide bond. J. Agric. Food Chem. 50, 185–191. Hill, A.R., 1989. The β-lactoglobulin-κ-casein complex. Can. Inst. Food Sci. Technol. J. 22, 120–123. Hillier, R.M., Lyster, R.L.J., 1979. Whey protein denaturation in heated milk and cheese whey. J. Dairy Res. 46, 95–102. Hoffmann, M.A.M., van Mil, P.J.J.M., 1997. Heat-induced aggregation of β-lactoglobulin: role of the free thiol group and disulfide bonds. J. Agric. Food Chem. 45, 2942–2948. Holland, J.W., Deeth, H.C., Alewood, P.F., 2008. Analysis of disulphide linkages in bovine κ-casein oligomers using two-dimensional electrophoresis. Electrophoresis 29, 2402–2410. Holt, C., 1992. Structure and stability of bovine casein micelles. Adv. Protein Chem. 43, 63–151.

380

9. The whey proteins in milk

Holt, C., Carver, J.A., Ecroyd, H., Thorn, D.C., 2013. Caseins and the casein micelle: their biological functions, structures, and behavior in foods. J. Dairy Sci. 96, 6127–6146. Holt, C., Horne, D.S., 1996. The hairy casein micelle: evolution of the concept and its implications for dairy technology. Neth. Milk Dairy J. 50, 85–111. Hong, Y.H., Creamer, L.K., 2002. Changed protein structures of bovine β-lactoglobulin B and α-lactalbumin as a consequence of heat treatment. Int. Dairy J. 12, 345–359. Horne, D.S., 1998. Casein interactions: casting light on the black boxes, the structure in dairy products. Int. Dairy J. 8, 171–177. Horne, D.S., 2002. Casein structure, self-assembly and gelation. Curr. Opin. Colloid Interface Sci. 7, 456–461. Horne, D.S., 2003. Casein micelles as hard spheres: limitations of the model in acidified gel formation. Colloids Surf. A Physicochem. Eng. Asp. 213, 255–263. Horne, D.S., 2006. Casein micelle structure: models and muddles. Curr. Opin. Colloid Interface Sci. 11, 148–153. Horne, D.S., 2017. A balanced view of casein interactions. Curr. Opin. Colloid Interface Sci. 28, 74–86. Horne, D., Lucey, J.A., 2017. Letter to the editor: hydrophobic interactions in the caseins: challenging their dismissal by Holt et al. (2013). J. Dairy Sci. 100, 5119–5120. Huppertz, T., Gazi, I., Luyten, H., Nieuwenhuijse, H., Alting, A., Schokker, E., 2017. Hydration of casein micelles and caseinates: implications for casein micelle structure. Int. Dairy J. 74, 1–11. Iametti, S., Gregori, B., Vecchio, G., Bonomi, F., 1996. Modifications occur at different structural levels during the heat denaturation of β-lactoglobulin. Eur. J. Biochem. 237, 106–112. Indyk, H.E., 2009. Development and application of an optical biosensor immunoassay for α-lactalbumin in bovine milk. Int. Dairy J. 19, 36–42. International-Dairy-Federation, 1995. Heat-induced changes in milk. In: Fox, P.F. (Ed.), International Dairy Federation Special Issue 9501, second ed. International Dairy Federation, Brussels, Belgium, p. 455. International-Dairy-Federation, 1996. Heat treatments and alternative methods. In: International Dairy Federation (Ed.), International Dairy Federation Special Issue 9602. International Dairy Federation, Brussels, Belgium. Kehoe, J.J., Brodkorb, A., Molle, D., Yokoyama, E., Famelart, M.H., Bouhallab, S., Morris, E.R., Croguennec, T., 2007. Determination of exposed sulfhydryl groups in heated β-lactoglobulin A using IAEDANS and mass spectrometry. J. Agric. Food Chem. 55, 7107–7113. Kelly, A.L., O’Connell, J.E., Fox, P.F., 2003. Manufacture and properties of milk powders. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1: Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, USA, pp. 1027–1061. Kessler, H.G., Beyer, H.J., 1991. Thermal denaturation of whey proteins and its effect in dairy technology. Int. J. Biol. Macromol. 13, 165–173. Kethireddipalli, P., Hill, A.R., Dalgleish, D.G., 2010. Protein interactions in heat-treated milk and effect on rennet coagulation. Int. Dairy J. 20, 838–843. Kethireddipalli, P., Hill, A.R., Dalgleish, D.G., 2011. Interaction between casein micelles and whey protein/κ-casein complexes during renneting of heat-treated reconstituted skim milk powder and casein micelle/serum mixtures. J. Agric. Food Chem. 59, 1442–1448. de Kruif, C.G., Holt, C., 2003. Casein micelle structure, functions and interactions. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1: Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, pp. 233–276. de Kruif, C.G., Huppertz, T., Urban, V.S., Petukhov, A.V., 2012. Casein micelles and their internal structure. Adv. Colloid Interf. Sci. 171–172, 36–52. Kudo, S., 1980. The heat stability of milk: formation of soluble proteins and protein-depleted micelles at elevated temperatures. New Zeal. J. Dairy Sci. Technol. 15, 255–263. Kuramoto, S., Jenness, R., Coulter, S.T., Choi, R.P., 1959. Standardization of the Harland-Ashworth test for whey protein nitrogen. J. Dairy Sci. 42, 28–38. Lakemond, C.M.M., van Vliet, T., 2005. Rheology of acid skim milk gels. In: Dickinson, E. (Ed.), Food Colloids: Interactions, Microstructure and Processing. In: Vol. Special Publication 298. The Royal Society of Chemistry, Cambridge, pp. 26–36. Lakemond, C.M.M., van Vliet, T., 2008a. Rheological properties of acid skim milk gels as affected by the spatial distribution of the structural elements and the interaction forces between them. Int. Dairy J. 18, 585–593. Lakemond, C.M.M., van Vliet, T., 2008b. Acid skim milk gels: the gelation process as affected by preheating pH. Int. Dairy J. 18, 574–584.

References

381

Larson, B.L., Jenness, R., Geddes, W.F., Coulter, S.T., 1951. An evaluation of the methods used for determining the baking quality of nonfat dry milk solids. Cereal Chem. 28, 351–370. Larson, B.L., Rolleri, G.D., 1955. Heat denaturation of the specific serum proteins in milk. J. Dairy Sci. 38, 351–360. Law, A.J.R., Leaver, J., 1997. Effect of protein concentration on rates of thermal denaturation of whey proteins in milk. J. Agric. Food Chem. 45, 4255–4261. Law, A.J.R., Leaver, J., 2000. Effect of pH on the thermal denaturation of whey proteins in milk. J. Agric. Food Chem. 48, 672–679. Leighton, F.R., 1962. Determination of the whey protein index of skim milk powder. Aust. J. Dairy Technol. 17, 166–169. Lewis, M., Grandison, A., Lin, M.-J., Tsioulpas, A., 2011. Ionic calcium and pH as predictors of stability of milk to UHT processing. Milchwissenschaft 66, 197–200. Livney, Y.D., Dalgleish, D.G., 2004. Specificity of disulfide bond formation during thermal aggregation in solutions of β-lactoglobulin B and κ-casein A. J. Agric. Food Chem. 52, 5527–5532. Long, J.E., Gould, I.A., van Winkle, Q., 1963. Heat-Induced interaction between crude κ-casein and β-lactoglobulin. J. Dairy Sci. 46, 1329–1334. Lowe, E.K., Anema, S.G., Bienvenue, A., Boland, M.J., Creamer, L.K., Jimenez-Flores, R., 2004. Heat-induced redistribution of disulfide bonds in milk proteins. 2. Disulfide bonding patterns between bovine β-lactoglobulin and κ-casein. J. Agric. Food Chem. 52, 7669–7680. Lucey, J.A., 2002. Formation and physical properties of milk protein gels. J. Dairy Sci. 85, 281–294. Lucey, J.A., Horne, D.S., 2018. Perspectives on casein interactions. Int. Dairy J. 85, 56–65. Lucey, J.A., Singh, H., 1998. Formation and physical properties of acid milk gels: a review. Food Res. Int. 30, 529–542. Lucey, J.A., Tamehana, M., Singh, H., Munro, P.A., 1998. Effect of interactions between denatured whey proteins and casein micelles on the formation and rheological properties of acid skim milk gels. J. Dairy Res. 65, 555–567. Lucey, J.A., Teo, C.T., Munro, P.A., Singh, H., 1997. Rheological properties at small (dynamic) and large (yield) deformations of acid gels made from heated milk. J. Dairy Res. 64, 591–600. Lyster, L.J., 1970. The denaturation of α-lactalbumin and β-lactoglobulin in heated milk. J. Dairy Res. 37, 233–243. Manderson, G.A., Hardman, M.J., Creamer, L.K., 1998. Effect of heat treatment on the conformation and aggregation of β-lactoglobulin A, B, and C. J. Agric. Food Chem. 46, 5052–5061. Manji, B., Kakuda, Y., 1986. Thermal denaturation of whey proteins in skim milk. Can. Inst. Food Sci. Technol. J. 19, 161–166. Manji, B., Kakuda, Y., 1987. Determination of whey protein denaturation in heat-processed milks: comparison of three methods. J. Dairy Sci. 70, 1355–1361. McKenna, A.B., Anema, S.G., 1993. The effect of thermal processing during whole milk powder manufacture and after its reconstitution on set-yoghurt properties. In: International Dairy Federation Special Issue 9303. Protein & Fat Globule Modifications by Heat Treatment, Homogenization & Other Technological Means for High Quality Dairy Products. International Dairy Federation, Brussels, Belgium, pp. 307–316. McKenzie, G.H., Norton, R.S., Sawyer, W.H., 1971. Heat-induced interaction of β-lactoglobulin and κ-casein. J. Dairy Res. 38, 343–351. McMahon, D.J., 1996. Age-gelation of UHT milk: changes that occur during storage, their effect on shelf life and the mechanism by which age-gelation occurs. In: IDF Special Issue 9602: Heat Treatments & Alternative Methods. International Dairy Federation, Brussels, Belgium, pp. 315–325. McMahon, D.J., Alleyne, M.C., 1994. A mechanism for the age-gelation of milk. In: Proceedings of the 24th International Dairy Congress, Melbourne, Australia. International Dairy Federation, Brussels, Belgium, p. 188. McMahon, D.J., McManus, W.R., 1998. Rethinking casein micelle structure using electron microscopy. J. Dairy Sci. 81, 2985–2993. McMahon, D.J., Oommen, B.S., 2008. Supramolecular structure of the casein micelle. J. Dairy Sci. 91, 1709–1721. Mellema, M., van Opheusden, J.H.J., van Vliet, T., 2002. Categorization of rheological scaling models for particle gels applied to casein gels. J. Rheol. 46, 11–29. Morand, M., Dekkari, A., Guyomarc’h, F., Famelart, M.-H., 2012b. Increasing the hydrophobicity of the heat-induced whey protein complexes improves the acid gelation of skim milk. Int. Dairy J. 25, 103–111. Morand, M., Guyomarc’h, F., Famelart, M.-H., 2011. How to tailor heat-induced whey protein/κ-casein complexes as a means to investigate the acid gelation of milk-a review. Dairy Sci. Technol. 91, 97–126. Morand, M., Guyomarc’h, F., Legland, D., Famelart, M.-H., 2012a. Changing the isoelectric point of the heat-induced whey protein complexes affects the acid gelation of skim milk. Int. Dairy J. 23, 9–17.

382

9. The whey proteins in milk

Nguyen, N.H.A., Streicher, C., Anema, S.G., 2018. The effect of thiol reagents on the denaturation of the whey protein in milk and whey protein concentrate solutions. Int. Dairy J. 85, 285–293. Nguyen, N.H.A., Wong, M., Havea, P., Guyomarc’h, F., Anema, S.G., 2013. The protein interactions and rheological properties of skim milk heated in the presence of low levels of reducing agent. Food Chem. 138, 1604–1609. Nieuwenhuijse, J.A., van Boekel, M.A.J.S., 2003. Protein stability in sterilised milk and milk products. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1: Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, USA, pp. 947–974. Nieuwenhuijse, J.A., van Boekel, M.A.J.S., Walstra, P., 1991. On the heat-induced association and dissociation of proteins in concentrated skim milk. Neth. Milk Dairy J. 45, 3–22. Noh, B.S., Creamer, L.K., Richardson, T., 1989a. Thermally induced complex formation in an artificial milk system. J. Agric. Food Chem. 37, 1395–1400. Noh, B., Richardson, T., Creamer, L.K., 1989b. Radiolabelling study of the heat-induced interactions between α-lactalbumin, β-lactoglubulin and κ-casein in milk and in buffer solutions. J. Food Sci. 54, 889–893. O’Connell, J.E., Fox, P.F., 2003. Heat-induced coagulation of milk. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1: Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, USA, pp. 879–945. Oh, H.E., Wong, M., Pinder, D.N., Hemar, Y., Anema, S.G., 2007. Effect of pH adjustment at heating on the rheological properties of acid skim milk gels with added potato starch. Int. Dairy J. 17, 1384–1392. Oldfield, D.J., Singh, H., Taylor, M.W., 1998b. Association of β-lactoglobulin and α-lactalbumin with the casein micelles in skim milk heated in an ultra-high temperature plant. Int. Dairy J. 8, 765–770. Oldfield, D.J., Singh, H., Taylor, M.W., 2005. Kinetics of heat-induced whey protein denaturation and aggregation in skim milks with adjusted whey protein concentration. J. Dairy Res. 72, 369–378. Oldfield, D.J., Singh, H., Taylor, M., Pearce, K., 1998a. Kinetics of denaturation and aggregation of whey proteins in skim milk heated in an ultra-high temperature (UHT) pilot plant. Int. Dairy J. 8, 311–318. Oldfield, D.J., Singh, H., Taylor, M.W., Pearce, K.N., 2000. Heat-induced interactions of β-lactoglobulin and α-lactalbumin with the casein micelle in pH-adjusted skim milk. Int. Dairy J. 10, 509–518. Parker, E.A., Donato, L., Dalgleish, D.G., 2005. Effects of added sodium caseinate on the formation of particles in heated milk. J. Agric. Food Chem. 53, 8265–8272. Parnell-Clunies, E.M., Kakuda, Y., Mullen, K., Arnott, D.R., deMan, J.M., 1986. Physical properties of yogurt: a comparison of vat versus continuous heating systems of milk. J. Dairy Sci. 69, 2593–2603. Patel, H.A., Anema, S.G., Holroyd, S.E., Singh, H., Creamer, L.K., 2007. Methods to determine denaturation and aggregation of proteins in low-, medium- and high-heat skim milk powders. Lait 87, 251–268. Patel, H.A., Singh, H., Anema, S.G., Creamer, L.K., 2006. Effects of heat and high hydrostatic pressure treatments on disulfide bonding interchanges among the proteins in skim milk. J. Agric. Food Chem. 54, 3409–3420. Purkayastha, R., Tessier, H., Rose, D., 1967. Thiol-disulfide interchange in formation of β-lactoglobulin-κ-casein complex. J. Dairy Sci. 50, 764. Qi, P.X., 2007. Studies of casein micelle structure: the past and the present. Lait 87, 363–383. Qi, P.X., 2009. Studies of the biological function and structure of casein micelles, and future implications. In: Corredig, M. (Ed.), Dairy-Derived Ingredients: Food and Nutraceutical Uses. Woodhead Publishing, pp. 147–169. Qin, B.Y., Bewley, M.C., Creamer, L.K., Baker, E.N., Jameson, G.B., 1999. Functional implications of structural differences between variants A and B of bovine β-lactoglobulin. Protein Sci. 8, 75–83. Rasmussen, L.K., Johnsen, L.B., Tsiora, A., Sorensen, E.S., Thomsen, J.K., Nielsen, N.C., Jakobsen, H.J., Petersen, T.E., 1999. Disulphide-linked caseins and casein micelles. Int. Dairy J. 9, 215–218. Raynal, K., Remeuf, F., 1998. The effect of heating on physicochemical and renneting properties of milk: a comparison between caprine, ovine and bovine milk. Int. Dairy J. 8, 695–706. Renan, M., Guyomarc’h, F., Chatriot, M., Gamerre, V., Famelart, M.H., 2007. Limited enzymatic treatment of skim milk using chymosin affects the micelle/serum distribution of the heat-induced whey protein/κ-casein aggregates. J. Agric. Food Chem. 55, 6736–6745. Richardson, B.C., Creamer, L.K., Pearce, K.N., Munford, R.E., 1974. Comparative micelle structure. 2. Structure and composition of casein micelles in ovine and caprine milk as compared with those in bovine milk. J. Dairy Res. 41, 239–247. Rodriguez del Angel, C., Dalgleish, D.G., 2006. Structure and some properties of soluble protein complexes formed by the heating of reconstituted skim milk powder. Food Res. Int. 39, 472–479.

References

383

Rose, D., 1961. Variations in the heat stability and composition of milk from individual cows during lactation. J. Dairy Sci. 44, 430–441. Rowland, S.J., 1933. The heat denaturation of albumin and globulin in milk. J. Dairy Res. 5, 46–53. Ruegg, M., Moor, U., Blanc, B., 1977. A calorimetric study of thermal denaturation of whey proteins in simulated milk ultrafiltrate. J. Dairy Res. 44, 509–520. Sanderson, W.B., 1970a. Determination of undenatured whey protein nitrogen in skim milk powder by dye binding. New Zeal. J. Dairy Sci. Technol. 5, 46–48. Sanderson, W.B., 1970b. Seasonal variations affecting the determination of the whey protein nitrogen index of skim milk powder. New Zeal. J. Dairy Sci. Technol. 5 (2), 48–52. Sanderson, W.B., 1970c. Reconstituted and recombined dairy products. New Zeal. J. Dairy Sci. Technol. 5, 139–143. Sawyer, W.H., 1969. Complex between β-lactoglobulin and κ-casein. A review. J. Dairy Sci. 52, 1347–1355. Sawyer, W.H., Coulter, S.T., Jenness, R., 1963. Role of sulfhydryl groups in the interaction of κ-casein and β-lactoglobulin. J. Dairy Sci. 46, 564–565. Schmidt, D.G., 1982. Association of caseins and casein micelle structure. In: Fox, P.F. (Ed.), Developments in Dairy Chemistry-1. Proteins. Elsevier Applied Science Publishers, London, England, pp. 61–86. Schorsch, C., Wilkins, D.K., Jones, M.G., Norton, I.T., 2001. Gelation of casein-whey mixtures: effects of heating whey proteins alone or in the presence of casein micelles. J. Dairy Res. 68, 471–481. Shalabi, S.I., Wheelock, J.V., 1976. Role of α-lactalbumin in the primary phase of chymosin action on heated casein micelles. J. Dairy Res. 43, 331–335. Singh, H., 2004. Heat stability of milk. Int. J. Dairy Technol. 57, 111–119. Singh, H., Creamer, L.K., 1991a. Denaturation, aggregation and heat-stability of milk protein during the manufacture of skim milk powder. J. Dairy Res. 58, 269–283. Singh, H., Creamer, L.K., 1991b. Influence of concentration of milk solids on the dissociation of micellar κ-casein on heating reconstituted milk at 120°C. J. Dairy Res. 58, 99–105. Singh, H., Creamer, L.K., 1992. Heat stability of milk. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry, Volume 1: Proteins. Elsevier Applied Science, London, U.K, pp. 621–656. Singh, H., Fox, P.F., 1985a. Heat stability of milk: pH-dependent dissociation of micellar κ-casein on heating milk at ultra high temperatures. J. Dairy Res. 52, 529–538. Singh, H., Fox, P.F., 1985b. Heat stability of milk: the mechanism of stabilization by formaldehyde. J. Dairy Res. 52, 65–76. Singh, H., Fox, P.F., 1986. Heat stability of milk: further studies on the pH-dependent dissociation of micellar κ-casein. J. Dairy Res. 53, 237–248. Singh, H., Fox, P.F., 1987a. Heat stability of milk: role of β-lactoglobulin in the pH-dependent dissociation of micellar κ-casein. J. Dairy Res. 54, 509–521. Singh, H., Fox, P.F., 1987b. Heat stability of milk: influence of modifying sulfhydryl-disulfide interactions on the heat coagulation time-pH profile. J. Dairy Res. 54, 347–359. Singh, H., Fox, P.F., 1987c. Heat stability of milk: influence of colloidal and soluble salts and protein modification on the pH-dependent dissociation of micellar κ-casein. J. Dairy Res. 54, 523–534. Singh, H., Newstead, D.F., 1992. Aspects of proteins in milk powder manufacture. In: Advanced Dairy Chemistry, Volume 1: Proteins. Elsevier Applied Science, London, pp. 735–765. Slatter, W.L., van Winkle, Q., 1952. An electrophoretic study of the protein in skimmilk. J. Dairy Sci. 35, 1083–1093. Smits, P., van Brouwershaven, J.H., 1980. Heat-induced association of β-lactoglobulin and casein micelles. J. Dairy Res. 47, 313–325. Snoeren, T.H.M., van der Spek, C.A., 1977. Isolation of a heat-induced complex from UHTST milk. Neth. Milk Dairy J. 31, 352–355. Swaisgood, H.E., 1982. Chemistry of milk protein. In: Fox, P.F. (Ed.), Developments in Dairy Chemistry-1. Elsevier Applied Science, London, pp. 1–59. Swaisgood, H.E., 1992. Chemistry of the caseins. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry: Proteins, third ed. In: vol. 1. Elsevier Applied Science, London, U.K, pp. 63–110. Swaisgood, H.E., 1993. Review and update of casein chemistry. J. Dairy Sci. 76, 3054–3061. Tessier, H., Yaguchi, M., Rose, D., 1969. Zonal ultracentrifugation of β-lactoglobulin and κ-casein complexes induced by heat. J. Dairy Sci. 52, 139–145. Tobias, J., Whitney, R.M., Tracy, P.H., 1952. Electrophoretic properties of milk proteins, II. Effect of heating to 300°F by means of the Mallory small-tube heat exchanger on skimmilk proteins. J. Dairy Sci. 35, 1036–1045.

384

9. The whey proteins in milk

Trautman, J.C., Swanson, A.M., 1958. Additional evidence of a stable complex between β-lactoglobulin and α-casein. J. Dairy Sci. 41, 715. Trejo, R., Dokland, T., Jurat-Fuentes, J., Harte, F., 2011. Cryo-transmission electron tomography of native casein micelles from bovine milk. J. Dairy Sci. 94, 5770–5775. Tziboula, A., Horne, D.S., 1999. The role of αs1-casein in the structure of caprine casein micelles. Int. Dairy J. 9, 173–178. Vasbinder, A.J., Alting, A.C., Visschers, R.W., de Kruif, C.G., 2003b. Texture of acid milk gels: formation of disulfide cross-links during acidification. Int. Dairy J. 13, 29–38. Vasbinder, A.J., de Kruif, C.G., 2003. Casein-whey protein interactions in heated milk: the influence of pH. Int. Dairy J. 13, 669–677. Vasbinder, A.J., Rollema, H.S., de Kruif, C.G., 2003a. Impaired rennetability of heated milk; study of enzymatic hydrolysis and gelation kinetics. J. Dairy Sci. 86, 1548–1555. Verheul, M., Roefs, S.P.F.M., de Kruif, K.G., 1998. Kinetics of heat-induced aggregation of β-lactoglobulin. J. Agric. Food Chem. 46, 896–903. van Vliet, T., 1996. Large deformation and fracture behaviour of gels. Curr. Opin. Colloid Interface Sci. 1, 740–745. van Vliet, T., Keetals, C.J.A.M., 1995. Effect of preheating of milk on the structure of acidified milk gels. Neth. Milk Dairy J. 49, 27–35. van Vliet, T., Walstra, P., 1995. Large deformation and fracture behaviour of gels. Faraday Discuss. 101, 359–370. Walstra, P., 1990. On the stability of casein micelles. J. Dairy Sci. 73, 1965–1979. Walstra, P., 1999. Casein sub-micelles: do they exist? Int. Dairy J. 9, 189–192. Walstra, P., Geurts, T.J., Noomen, A., Jellema, A., van Boekel, M.A.J.S., 1999. Dairy Technology: Principles of Milk Properties and Processes. Marcel Dekker, Inc., New York. Walstra, P., Jenness, R., 1984. Dairy Chemistry and Physics. John Wiley and Sons, New York. Wu, D., Qin, J., Lin, B., 2008. Electrophoretic separations on microfluidic chips. J. Chromatogr. A 1184, 542–559. Zittle, C.A., Custer, J.H., Cerbulis, J., Thompson, M.P., 1962. κ-Casein -β-lactoglobulin interaction in solution when heated. J. Dairy Sci. 45, 807–810.

C H A P T E R

10 The effect of UHT processing and storage on milk proteins Hilton C. Deeth School of Agriculture and Food Sciences, University of Queensland, Brisbane, QLD, Australia

Introduction The high processing temperatures and long storage times at ambient temperature of ultrahigh-temperature (UHT)-processed milk put huge demands on the stability of a reasonably fragile biological product. The temperature-time conditions of both processing and storage vary considerably. Ambient temperatures of storage (and transport) can vary from well below 0°C in cold countries to
50°C in tropical zones and some storage facilities. When these variable conditions are considered in conjunction with the variability in the raw material involved, it is not surprising that the proteins in the product can undergo several changes before the product is consumed. It is essential that the nature of these changes is understood so that those detrimental to the stability and nutritive value of the product can be minimized. The changes in proteins during processing and storage are caused by chemical reactions, physicochemical interactions, and enzymatic activity. However, changes can also be caused by microbiological contamination; this source of change is not addressed in this chapter. UHT milk has a shelf life of 6–9 months although some companies are placing a 12-month use-by date on their product. Given the numerous ways in which the product can deteriorate during storage, even under ideal conditions, as outlined in this chapter, it is apparent that the product is outside its comfort zone when stored for more than perhaps 6, and certainly 9, months. However, it is noted that the legal expiry date for UHT milk in some countries is less than 6 months [e.g., 90 days (Pizzano et al., 2012)]. Several different UHT-processed milk products are now produced. This chapter primarily concerns UHT milk, but other products are mentioned as some of the changes in the milk proteins are more relevant to some of these products. Good examples of these are UHT concentrated milks, flavored milks, and lactose-hydrolyzed milks.

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00010-4

385

# 2020 Elsevier Inc. All rights reserved.

386

10. The effect of UHT processing and storage on milk proteins

The UHT process The UHT process, that is, the heating of milk at high temperature for a short period of time followed by aseptic packaging, was introduced to produce a shelf-stable milk that had superior properties to in-container sterilized milk. The thermal treatment is in the range 135–150°C for holding times of 1–10 s (Deeth and Smithers, 2018). These conditions have been developed to achieve what is known as “commercial sterility,” meaning that no microorganisms will grow in the product under the normal conditions of storage (Burton, 1988; Lewis and Heppell, 2000; Deeth and Lewis, 2017). The minimum heating conditions for producing a commercially sterile product are generally recognized to be 135°C for 10.1 s or equivalent, which is sufficient to cause a 9-log reduction of thermophilic bacterial spores. These conditions are equal to a B* of 1, where B* is a bacteriological index (Kessler, 1981). UHT processes should be equivalent to a B* of
1.0. A related bacteriological index is the F0, which is sometimes used in UHT processing but is more commonly used in canning technology. It refers to a 12-log reduction of spores of Clostridium botulinum and is equivalent to heating at 121.1°C for 1 min. A UHT process should be equivalent to an F0 of
 3.0. The bacteriological indices differ from the chemical index, C*, whereby a C* of 1.0 relates to the heat input that would cause a 3% reduction in the vitamin thiamine. It is more of a theoretical construct than a practical guide as virtually no one measures the loss of thiamine. The heating conditions for a C* of 1.0 are 135°C for 30.5 s or equivalent. A general recommendation is that the C* of a UHT process should be  1.0. Thus, it is recommended that any UHT heat treatment should be equivalent to a B* of
1.0 and a C* of  1.0; however, most commercial UHT plants operate at B* and C* values that are considerably greater than 1.0 (Tran et al., 2008), with the conditions of many considered to be excessive (Cattaneo et al., 2008). The chemical index, C*, is useful for juxtaposing bacterial and chemical changes, but it provides only a guide to the specific chemical changes caused by a heat treatment. As each chemical change follows its own kinetics, the changes in individual components, such as the denaturation of the whey proteins, can be estimated only by taking into account the individual kinetics. There are two main broad types of UHT heating: direct and indirect. Direct heating involves the direct mixing of the milk with saturated steam (free of entrained water droplets) (Lewis and Heppell, 2000) and the subsequent removal of the added water in a vacuum flash vessel. In indirect processes, the heating medium, steam or hot water, heats the milk indirectly by conduction and convection through a stainless steel barrier in the form of a heat exchanger. Indirect processes, using either plate or tubular heat exchangers, are most commonly used. The main difference between the two processes that is relevant to their effects on milk proteins can be readily seen in their temperature-time profiles (Fig. 10.1A–C); direct processes heat and cool the milk to and from the sterilization temperature very quickly ( 6.8, most of the whey proteins are found in the serum attached to κ-casein (Kudo, 1980). The amount of attached whey proteins is very sensitive to small changes in pH. Oldfield et al. (1998) reported a maximum association of whey proteins with casein micelles of  55% during heating in a direct steam injection UHT pilot plant at temperatures from 70°C to 130°C. However, this appears to conflict with an earlier study by Corredig and Dalgleish (1996b), which reported that almost all of the whey proteins bound to casein micelles during heating in a water bath to 70–90°C at the natural pH of milk, 6.75–6.8. Oldfield et al. (1998) considered that the different results from those in their work could have been due to the much slower rate of heating. This explanation is consistent with the report that more whey proteins associate with the casein micelles during indirect UHT heating than during direct UHT heating (Corredig and Dalgleish, 1996b). Almost all of the β-Lg attached to the casein micelles in milk processed on an indirect plant but much less associated with the micelles in milk directly processed at the same nominal holding temperature and time. The heating method also influences the relative amounts of β-Lg and α-La that become attached to the micelle. Rapid heating, such as in direct UHT processing, results in a high β-Lg:α-La ratio, whereas slower heating, such as in indirect UHT treatments, results in a lower ratio, that is, more α-La and less β-Lg become attached (Mottar et al., 1989). The ratio is also influenced by the temperature of heating. Oldfield et al. (1998) found that, in the range 80–130°C, β-Lg was the first to associate with the casein micelles and was followed by α-La on prolonged heating. Curiously, at temperatures < 80°C, β-Lg and α-La attached to the micelles simultaneously. β-Lg is denatured to a lesser extent but α-La is denatured to a similar extent in concentrated milk compared with single-strength milk (McKenna and O’Sullivan, 1971). The denaturation rates of both β-Lg and α-La decrease with an increase in the level of nonprotein solids but increase with an increase in the protein content. It has been suggested that these effects cancel each other out for α-La but that the effect of the nonprotein components is greater than that of the proteins for β-Lg (Anema, 2009).

391

Protein changes during processing and storage

Whey protein denaturation levels in UHT milk Table 10.1 shows some reported levels of undenatured whey proteins in commercial UHT milk produced in direct and indirect UHT plants. It is clear that, whereas a wide range of values has been reported, the values for direct milks are consistently lower than those for indirect milks (Morales et al., 2000; Elliott et al., 2003, 2005). The variation in the results largely reflects the different UHT processing conditions, but different methodologies may also contribute to the variability. The other inference from these data is that the levels of residual α-La are comparable with and sometimes higher than those of β-Lg. This reflects the greater stability of α-La, given that its concentration in milk is less than half that of β-Lg. The kinetics of the denaturation of β-Lg and α-La have been reported by several authors (Lyster, 1970; Dannenberg and Kessler, 1988; Anema et al., 1996; Reddy et al., 1999). Using the kinetic data of Lyster (1970) and Dannenberg and Kessler (1988) and the temperature-time profiles of the plants, Tran et al. (2008) calculated the denaturation percentages of β-Lg and α-La in milk processed by 22 commercial UHT plants. The percentages calculated using the data of Dannenberg and Kessler (1988) were higher than those based on the data of Lyster (1970). The denaturation of β-Lg was calculated to be close to 100% in the commercial indirect plants but to be 74%–92% in the direct plants. For α-La, the calculated denaturation percentages were 25%–90% for indirect plants and 27%–58% for direct plants. The denaturation percentages for β-Lg and α-La, calculated using the Lyster (1970) kinetics data, for the indirect and direct profiles shown in Fig. 10.1A and B, were 94% and 58% (indirect) and 82% and 25% (direct), respectively. Denaturation in the preheat and high-temperature sections of UHT plants During UHT processing, the extent of irreversible denaturation and the associated proteinprotein interactions of the whey proteins depend on the severity of processing in terms of temperature, the duration of heating, and the temperature-time profile of the UHT plant. As the denaturation of β-Lg commences at  70°C, much of the denaturation occurs in the preheat section of indirect UHT plants and in direct plants if a preheat holding step is included. This is illustrated in Table 10.2, which shows the calculated denaturation percentages of β-Lg and α-La in milk at three points (after the preheat section, after the sterilization holding tube, and after cooling) during processing on an indirect UHT plant with preheating at 90°C for 60 s, a direct plant with the same preheating, and a direct plant with no preheat hold. The temperature-time profiles of the plants are shown in Fig. 10.1A–C. The calculated B* of the three plants is the same, that is, 2.95. This shows that, where such preheating is used, TABLE 10.1 Acid-soluble whey proteins (mg/L) in commercial UHT milks UHT process type Analyte

Direct

Indirect

References

α-Lactalbumin

356  176 63–1210

1131  340 220–1036

Elliott et al. (2005) Andreini et al. (1990)

β-Lactoglobulin

190  52 142–154

1404  516 736–792

Elliott et al. (2005) Recio et al. (1996)

392

10. The effect of UHT processing and storage on milk proteins

TABLE 10.2 Calculated denaturation levels (%)a of β-lactoglobulin and α-lactalbumin during UHT processing on indirect and direct UHT plantsb β-Lactoglobulin

α-Lactalbumin

Indirect (with preheat hold)

Direct (with preheat hold)

Direct (no preheat hold)

Indirect (with preheat hold)

Direct (with preheat hold)

Direct (no preheat hold)

At end of preheating including preheat hold (immediately before sterilization holding tube)

91

72

13

44

11

2

At end of sterilization holding tube

93

81

66

50

23

14

At end of cooling

93

82

67

58

25

16

Position in UHT plant

a

Based on kinetics of Lyster (1970). The temperature-time profiles are shown in Fig. 10.1A–C; B* for all plants, 2.95. Sterilization holding tube conditions: indirect, 140°C for 4 s; direct, 144°C for 4.05 s. Preheat hold conditions: 90°C for 60 s. b

most of the denaturation occurs prior to the sterilization holding tube; however, in the direct plant in which no preheat holding is used, most of the denaturation occurs in the sterilization holding tube. On the basis of the data in Table 10.2, the extent of the denaturation of β-Lg is not necessarily a good indicator of the severity of the overall heat treatment. This is significant as the Aschaffenberg turbidity test (Aschaffenburg, 1950), which gives an estimate of undenatured whey protein, is sometimes used an indicator of the total heat treatment that a UHT milk has received. Furthermore, the extent of the denaturation of β-Lg in some UHT plants, especially indirect plants, is often
95%, and hence, discrimination between plants on the basis of their heat input using this criterion could be difficult. Table 10.2 shows that the calculated extent of the denaturation of α-La is much lower than that of β-Lg and that, even under quite severe heating conditions, as in an indirect UHT plant with preheating at 90°C for 60 s, the overall level is not excessive. Because of this, the denaturation percentage of α-La has been suggested to be a better index of heat treatment than that of β-Lg (Reddy et al., 1999; Tran et al., 2008; Pizzano et al., 2012). The preheat conditions used in Table 10.2 reflect those that have been recommended for the inactivation of plasmin (Newstead et al., 2006; Anema, 2017). They may also be suitable for minimizing heat exchanger fouling. It should be noted that a preheat holding section is not normally included for the direct UHT processing of milk but is included in indirect plants (Tran et al., 2008; SPX, 2013). However, as discussed in the section on The plasmin system, a preheat holding section in all UHT plants could minimize or even prevent the development of plasmin-related defects. Measurement of whey protein denaturation The extent of whey protein denaturation is typically mentioned in terms of the percentage of the whey proteins that precipitate at pH 4.6 (Dannenberg and Kessler, 1988; Resmini et al., 1989). However, in practice, as the level of whey proteins before heat treatment is seldom

Protein changes during processing and storage

393

known, the percentage denaturation cannot be determined. Consequently, the concentration of whey proteins that remain soluble at pH 4.6, that is, acid-soluble whey proteins, is usually measured. These proteins are determined by a range of methods with HPLC and electrophoresis-densitometry being the most widely used. Some authors have measured the concentration of whey proteins remaining in the supernatant after ultracentrifugation to remove the casein micelles (Oldfield et al., 1998; Anema, 2017). These whey proteins include native and native-like whey proteins and those complexed to κ-casein, which reassociate with the casein micelles and precipitate when milk is acidified to pH 4.6 (Anema and Li, 2015). The extent of denaturation can be determined from the concentration of the native and native-like whey proteins in the supernatant, which can be analyzed by nonreducing polyacrylamide gel electrophoresis (Oldfield et al., 1998). An alternative method for separating the casein from the whey proteins to assess the extent of denaturation is to precipitate the casein by saturating the milk with salt. This is the basis of the rapid method for determining the extent of whey protein denaturation in the whey protein nitrogen index (WPNI) test. This test was developed by the American Dairy Products Institute (2009) for grading skim milk powders into low-, medium-, and high-heat powders. The native whey proteins, which remain in the supernatant after removal of the casein, can be quantified by a protein analytical method such as the dye-binding method (Sanderson, 1970). Unlike the HPLC and electrophoresis methods, the total amount of undenatured whey proteins, not the amounts of the individual whey proteins, is measured. The test has not generally been applied to UHT milk; however, Oliveira et al. (2015) used the WPNI as a heat treatment indicator for UHT milk in Brazil. On the basis of the test, they rated 90% of 60 commercial UHT samples analyzed to be equivalent to medium heat and 10% to be high heat. Role in fouling of heat exchangers The denaturation of whey proteins, particularly β-Lg, during UHT processing plays a major role in the formation of fouling deposits in the first stages of heating; deposits formed in the later stages of heating, that is, in high-temperature sections, are largely mineral in nature (Burton, 1988). Several authors have reported good correlations between fouling deposit formation and β-Lg denaturation during the thermal processing of milk and cream (Hiddink et al., 1986; de Jong et al., 1992). For example, Grandison (1988) found a significant correlation (r ¼ 0.792) between plant run times and the weight of the deposit formed in the regenerative section of a UHT plant operating at 80–110°C. Interestingly, the deposit in the high-heat section was not correlated with run time. Several models of fouling have been developed, and many are based on the kinetics of the denaturation of β-Lg (de Jong et al., 2002; Grijspeerdt et al., 2004; Schutyser et al., 2008). de Jong et al. (1993) reported that such a model was valid up to 115°C. Others have reported the models to be valid up to 90°C (Grijspeerdt et al., 2004). The denatured/unfolded monomer of β-Lg appears to be largely responsible for the deposit formation ( Jeurnink et al., 1996; Mounsey and O’Kennedy, 2007). This is because this molten-state form of β-Lg is unable to associate with casein micelles or other whey proteins before the protein moves toward, and deposits on, the wall of the heat exchanger. Therefore, the longer is the time for which the unfolded state of β-Lg is present in the UHT plant, the greater is the deposit formation, or conversely, the shorter is the time for which the unfolded form is present, the longer is the run time of the plant (Grijspeerdt et al., 2004). This suggests

394

10. The effect of UHT processing and storage on milk proteins

that the unfolded form is “sticky” and readily attaches to the walls of the stainless steel tube or plate and forms a base for further deposit formation. For this reason, it has been proposed that the preheat conditions in a UHT plant should be designed to minimize the time spent by this nonaggregated, “sticky” form of β-Lg in the UHT plant. This accords with the reports of a reduction in fouling when relatively severe preheating conditions are used (Prakash et al., 2015). Role in flavor production in UHT milk Volatile sulfur-containing compounds have long been associated with the cooked flavor of UHT milk. The major compounds responsible are hydrogen sulfide, methane thiol, and dimethyl disulfide, the last being considered to be an oxidation product of methane thiol. The major source of these compounds in UHT milk is the denaturation of β-Lg although, in milk containing fat globules, the proteins of the milk fat globule also contribute. Higher levels of hydrogen sulfide and methane thiol occur in UHT whole milk than in UHT skim milk soon after processing (Al-Attabi et al., 2014). The mechanisms for the formation of these sulfur volatiles have been discussed by several authors including Walstra and Jenness (1984), Al-Attabi et al. (2009), and Zabbia et al. (2012). A major source of hydrogen sulfide is β-elimination of cysteine with the formation of dehydroalanine (Walstra and Jenness, 1984) and of methane thiol from a Strecker degradation of methionine via methional (Zabbia et al., 2012). The dehydroalanine formed from cysteine can take part in protein cross-linking. In relation to the role of the denaturation of β-Lg in cooked flavor production, Gaafar (1987) concluded that the threshold of cooked flavor corresponds to about 60% denaturation. This was reported to correspond to a hydrogen sulfide concentration of 3.4 μg/L and a reactive sulfhydryl concentration of 0.037 mmol/L. Reducing the effects of whey protein denaturation The addition of potassium iodate at 10–20 mg/L reduces fouling and increases UHT run times and also reduces cooked flavor development (Skudder et al., 1981). The potassium iodate oxidizes the heat-activated sulfhydryl groups of β-Lg and reduces or prevents the interaction of the “sticky” intermediate form with other proteins and the heated surfaces of heat exchangers. The use of iodate is not a practical solution to the fouling and flavor problems; the processed milk develops a bitter flavor on storage because the plasmin is not inactivated. A similar effect to that of iodate can be achieved with hydrogen peroxide. Al-Attabi (2009) found that hydrogen peroxide at 50 mg/L reduced fouling and increased the run time during the UHT processing of concentrated skim milk (15% total solids). Furthermore, low concentrations (10 or 50 mg/L) of hydrogen peroxide significantly reduced the level of sulfur volatiles in UHT milk. When hydrogen peroxide was added at 50 mg/L before or after processing, the UHT milk contained no detectable hydrogen sulfide or methane thiol by the second day after manufacture (Al-Attabi, 2009). Without the peroxide addition, these compounds are detectable for up to 10 days. The denaturation of whey proteins and the consequent fouling make the UHT processing of drinks containing high levels of whey proteins difficult. However, this problem can be alleviated by the addition of sodium caseinate, because of the chaperoning effect of the caseins on the whey proteins (O’Kennedy and Mounsey, 2006; Mounsey and O’Kennedy, 2010). In the author’s laboratory, the processing of a 2% whey protein concentrate (WPC) solution

Protein changes during processing and storage

395

caused severe fouling, but the addition of 2% sodium caseinate to the 2% WPC enabled it to be successfully UHT processed on a laboratory UHT plant. Furthermore, sodium caseinate + WPC mixtures up to 5% + 5% were successfully processed. Another approach to heat processing whey proteins is to reduce the pH, as whey proteins are stable at low pH ( 3) (Kelly, 2018b). Hence, it is possible to produce shelf-stable, highprotein acidic beverages based on whey protein isolate (WPI) at these pH values. (WPI is better to use than WPC because, as the lipid content is very low, clear products can be obtained.) These products, which are popular with athletes and health-conscious consumers, are shelf stable for up to 6 months with processing at 120°C for 20 s being reported to provide longterm stability (Villumsen et al., 2015a). These conditions prevented the formation of visible aggregates during storage, which are largely caused by the presence of caseinomacropeptide (CMP) in WPI made from cheese whey (Villumsen et al., 2015b). A further approach to stabilizing whey protein is through microparticulation (Kelly, 2018a). This has been utilized in the production of a commercial shelf-stable acidic whey protein drink, Upbeat, containing 8% protein. Ultrasonication has also been used to increase the heat stability of whey proteins (Ashokkumar et al., 2009; Gordon and Pilosof, 2010). After a whey protein solution (4%– 15% protein) had been partially denatured at  90°C and then ultrasonicated for 5 s at low frequency (20 kHz), it was stable to heating at 130°C (Huppertz et al., 2018). The acoustic cavitation caused by the ultrasonication treatment disrupted aggregates and thereby reduced the particle size.

Proteolysis Proteolysis occurs in many UHT products; however, with good control of the quality of the raw milk and judicious selection of the processing conditions, it can be minimized or prevented. Proteolysis is mostly caused by proteases although some nonenzymatic proteolysis can occur during the heat treatment of milk. Enzymatic proteolysis can be caused by indigenous or exogenous proteases. The alkaline protease, plasmin, is the most important indigenous protease, but others such as cathepsins (Gaucher et al., 2009) have been implicated. Of the exogenous enzymes, the heat-resistant proteases produced in raw milk by psychrotrophic bacteria, particularly Pseudomonas species, are the most significant, but proteases produced by spore-forming bacteria such as Bacillus and Bacillus-like species can also be involved. Plasmin The plasmin system

The plasmin system in milk is a complex system that includes plasmin and its precursor plasminogen, as well as a plasmin inhibitor, a plasminogen activator, and a plasminogen activator inhibitor (Ismail and Nielsen, 2010). Plasmin, plasminogen, and plasminogen activator are associated with the casein micelles, whereas plasmin inhibitor and plasminogen activator inhibitor are present in the serum. This distribution is of relevance when products such as milk protein concentrate and micellar casein, which contain plasmin, plasminogen, and

396

10. The effect of UHT processing and storage on milk proteins

plasminogen activator but no inhibitors, are used in long shelf life products such as UHT products. The components of the plasmin system have different heat stabilities, which makes the interpretation of the effect of heat treatments interesting. For example, heating at  75°C for 15 s completely inactivates plasminogen activator inhibitor and causes 36% inactivation of plasmin inhibitor. The sensitivity of the plasmin inhibitor is exemplified in the work of Newstead et al. (2006), in which the plasmin activity was greater in UHT-processed (direct steam injection to 140°C for 4 s) reconstituted skim milk when a preheat treatment of 80°C for 30 s was used than when a preheat treatment of 75°C for 15 s was used. Plasminogen activator has a similar or higher heat stability than plasmin and plasminogen, which explains why plasminogen is converted to plasmin during the storage of UHT milk if some remains in the milk after processing (Kelly and Foley, 1997; Prado et al., 2007). In fresh unheated milk, plasminogen is normally in higher concentration than plasmin. The plasminogen to plasmin ratio has been reported to be between 50:1 and 2:1. For example, Richardson (1983) reported up to nine times more plasminogen than plasmin in unheated milk, with a total level of plasmin of  0.3 mg/L. As plasminogen is converted to plasmin during the storage of heat-treated milks by the plasminogen activator, its level in milk and its heat stability are highly significant. Plasminogen, which can be activated to plasmin by plasminogen activator through the cleavage of a single peptide bond, is more heat stable than plasmin. Manji and Kakuda (1988) reported that 19% plasmin activity and 37% plasminogen remained in directly (steam infusion) processed (142°C for 5 s) milk, whereas no residual plasmin activity and 19% of the original plasminogen remained in indirectly (plate heat exchanger) processed milk (145°C for 3 s). Similarly, Cauvin et al. (1999) found that commercial UHT milks had virtually no remaining plasmin and attributed the significant proteolysis during storage to activated plasminogen. Plasminogen activators are even more heat stable than plasmin and plasminogen. Therefore, to prevent plasmin-induced proteolysis during the storage of UHT milk, plasminogen should be inactivated in the heat process. Milk with a high somatic cell count (SCC), from cows with mastitis, has an elevated level of plasmin (Bastian and Brown, 1996). In fact, Bikker et al. (2014) proposed the use of a plasmin test based on specific substrates as a potential diagnostic tool for detecting mastitic milk. As the SCC increases from 1,000,000/mL, the concentrations of both plasmin and plasminogen approximately double (Politis et al., 1989). The stage of lactation also affects the level of plasmin in milk; late-lactation milk tends to have a higher level than early- and midlactation milks (Auldist et al., 1996). The order of susceptibility of the caseins to hydrolysis by plasmin is β ¼ αs2 > αs1 > κ. The peptides formed are largely gamma-caseins (γ1 and γ2). The specificity of plasmin and the peptides it releases differ from those of bacterial proteases; this is useful for diagnostic purposes (Datta and Deeth, 2003). Proteolysis induced by plasmin and its effects in UHT milk

Plasmin-induced proteolysis in UHT milk has been reported by several authors (Kelly and Foley, 1997; Cauvin et al., 1999; Huijs et al., 2004; Newstead et al., 2006; Rauh et al., 2014a,b; Anema, 2017; Malmgren et al., 2017). From these reports, it is now clear that such proteolysis occurs only if the heat treatment conditions are insufficient to inactivate plasmin. In addition, it is much more likely to occur in directly processed UHT milk than in indirectly processed

Protein changes during processing and storage

397

UHT milk. The following examples illustrate the importance of appropriate heating conditions. • Newstead et al. (2006) used direct steam injection heating at 140°C for 4 s with preheat conditions ranging from 75°C for 15 s to 90°C for 60 s. With the lower intensity heat treatments (75°C for 15 s and 80°C for 15 and 30 s), extensive proteolysis occurred on storage, but, with preheat treatments of 90°C for 30 and 60 s, proteolysis was minimal during storage. • Rauh et al. (2014b) used preheating at 72°C for 180 s and direct steam infusion heating to > 150°C for < 0.2 s and found extensive proteolysis on storage; adjustment of the preheat conditions to 95°C for 180 s prevented proteolysis. • Anema (2017) processed reconstituted skim milk using preheating at 75°C for 15 s, or 80°C for 30 s, and direct steam injection heating to 143°C for 3 s and found extensive proteolysis on storage; adjustment of the preheat conditions to 90°C for 60 s prevented proteolysis. Furthermore, adjusting the high-heat conditions to 143°C for 30 s with a preheat of 75°C for 3 s also prevented subsequent proteolysis; however, heating at 143°C for 30 s is excessive as it is equivalent to a B* of 17 and an F0 of 78 for a direct process. • An interesting case was reported by Huijs et al. (2004). UHT milk was produced with innovative steam injection heating at 150–180°C for < 0.1 s. Although this process had an excellent bactericidal effect, it did not inactivate the plasmin, and the milk became bitter. To overcome this problem, a pretreatment of 80°C for 300 s was devised, which inactivated the plasmin and prevented proteolysis in the UHT milk (van Asselt et al., 2008). Because of the higher level of plasmin in mastitic milk, UHT milk made from high SCC milk shows more proteolysis than that made from low SCC milk (Auldist et al., 1996). Furthermore, it has been reported that indirectly processed UHT milk made from raw milk with high and low SCCs exhibited very low residual plasmin activity; however, when plasminogen was added, the high SCC milk showed the greatest tendency to form a plasmin-induced gel. This was attributed to an elevated level of plasminogen activator in the high SCC milks (Kelly and Foley, 1997) converting residual plasminogen to plasmin during storage. The major physical effects associated with plasmin-induced proteolysis are bitterness and age gelation, but its roles in sedimentation and fat separation have also been suggested. Evidence for the role of plasmin in age gelation is strong. Numerous papers have linked the level of residual plasmin in UHT milk and the tendency to gel during storage (e.g., Rauh et al., 2014a,b; Malmgren et al., 2017). Furthermore, the addition of plasmin or plasminogen has been shown to lead to gelation (Kohlmann et al., 1991). Proteolysis is considered to be a major factor in the development of age gelation in UHT milks. In some early reports of proteolysis-induced gelation, it is not entirely clear whether the proteolysis was due to plasmin or bacterial proteases. Therefore, it is essential to be able to eliminate the possibility of bacterial proteolysis in describing the gelation caused by plasmin. Several reports in which this condition is satisfied are available. • Newstead et al. (2006) found a strong correlation between plasmin-induced proteolysis and age gelation. No gelation occurred in milk samples that had undergone a preheat treatment of 90°C for 30 or 60 s. • Another example is that reported by Rauh et al. (2014b), in which the milk used was 1 day old, which was unlikely to contain bacterial proteases. The milk in which active plasmin

398

10. The effect of UHT processing and storage on milk proteins

remained after manufacture was processed with a preheat of 72°C for 180 s and high heat, using steam infusion, at > 150°C for < 0.2 s. It was stored at 20°C. The residual plasmin and plasminogen activities were 30.9% and 14%, respectively, of their activities in the original milk. Several points in this paper are typical of plasmin-induced proteolysis. Plasminogen was converted to plasmin during storage of the milk, resulting in an increase in both plasmin activity and the rate of release of peptides. Bitterness occurred after 7 weeks and was intense after 8 weeks. Viscosity and opacity increased from 10 weeks, at which time a gel commenced to form on the bottom of the container. The gel increased in thickness and floating gel particles were visible at 11 weeks. The gel occupied the whole of the container at 16 weeks. It was soft and fragile, and syneresis was observed. Extensive proteolysis of β- and αs1-casein occurred. A creamy layer formed on top of the gelled milk at 11 weeks. Although such fat separation is seldom linked to proteolysis, there are now several reports indicating a relationship (e.g., Kohlmann et al., 1991; Hardham, 1998). This is consistent with a report by Zhang et al. (2007) that the inactivation of proteases reduces fat separation. Along with this milk, which gelled, a corresponding milk in which the preheat conditions were changed to 95°C for 180 s was processed. This milk showed no active plasmin and did not gel or develop proteolysis or bitterness during storage (up to 16 weeks). • A report by Malmgren et al. (2017) also concerned UHT milk in which plasmin was active and no bacterial proteases were present. The processing conditions were preheating at 74°C for 20 s and high-temperature heating using steam infusion or injection, at 140°C for 4 s. The UHT milks were stored for 6 months at 5°C, 22°C, 30°C, and 40°C. Virtually, no proteolysis was observed at 5°C as plasmin has very low activity at this temperature (Grufferty and Fox, 1988), but proteolysis, particularly of β-casein and, to a lesser extent, αs1-casein, was observed at higher temperatures. Gelation occurred in milks stored at 22°C after 5 months and in milks stored at 30°C after 4 months. No gelation occurred in milks stored at 5°C or 40°C, a result that was consistent with previous reports (Kocak and Zadow, 1985a; Kohlmann et al., 1988). Heat inactivation of plasmin

It is evident from this discussion that proteolysis can cause major quality defects in UHT milk, including bitterness, gelation, and fat separation. All these defects can be overcome by selecting appropriate heating conditions during UHT processing. This was mentioned for the denaturation of β-Lg because, during its denaturation, it interacts with plasmin via disulfide bonds and causes loss of proteolytic activity. Therefore, it is possible to inactive plasmin in the preheat section under certain conditions. Conditions to achieve this are discussed in the section on The plasmin system; a summary of the conditions that have been found to inactivate plasmin and to prevent plasmin-induced proteolysis is as follows: 90–95°C for 30–60 s (Newstead et al., 2006), 80°C for 300 s (van Asselt et al., 2008), 95°C for 180 s (Rauh et al., 2014b), and 90°C for 60 s (Anema, 2017). Bacterial proteases The interest in bacterial proteases is largely due to their heat stability and their ability to survive, at least partially, UHT heat processes. Given that UHT milk is a product with a long shelf life (up to 12 months) at ambient temperature, which can be > 30°C, only trace amounts

Protein changes during processing and storage

399

of active protease can cause considerable proteolysis in UHT milk during storage. The major two effects of this proteolysis, as for plasmin-induced proteolysis, are age gelation and bitter flavor production; bacterial proteases can also cause casein micelle aggregation and sedimentation. The major proteases of interest are those produced by psychrotrophic bacteria, particularly Pseudomonas species. The nature of these proteases and their heat stabilities have been previously reviewed (Griffiths et al., 1981; Fairbairn and Law, 1986; Sørhaug and Stepaniak, 1997; Datta and Deeth, 2001) and are not covered in detail here. The main relevance of these enzymes is the changes they make to the proteins in UHT milk and the related changes to the properties of the milk. Production and characteristics of proteases of psychrotrophic bacteria

Extracellular proteases are produced by psychrotrophic bacteria such as Pseudomonads during the late log phase of growth. Most are alkaline metalloproteases of the AprX family with molecular weights of around 50,000 Da (Bagliniere et al., 2013; Mat
eos et al., 2015; Zhang et al., 2018). It is generally considered that they are produced after the total bacterial count reaches  105 cfu/mL although some strains have been reported to produce protease at 104 cfu/mL (Adams et al., 1976). However, as many authors have reported, there is considerable variation in the counts at which the production of protease occurs. Therefore, the total bacterial count is not always a good guide to the presence of extracellular enzymes as different bacteria have different propensities to produce these enzymes. Haryani et al. (2003) showed that significant quantities of enzymes were produced in milk with total bacterial counts of 105 cfu/mL, whereas some milks with a total plate count of
107 cfu/mL did not contain significant quantities. As a rule of thumb, raw milk for UHT processing should have a total bacterial count of less than 106 cfu/mL and preferably less than 105 cfu/mL. The specificity of the bacterial proteases toward milk proteins differs from that of plasmin, which has a strong preference for β-casein. In general, the proteases of psychrotrophic bacteria such as Pseudomonads have a preference for κ-casein, followed by β-casein, followed by αs1-casein. Most show little activity toward whey proteins but may hydrolyze them with prolonged exposure to high concentrations of the enzymes. The hydrolysis of κ-casein by these enzymes is similar to that of rennin, in that the first break is around the Phe105Met106 bond to produce a glycomacropeptide (GMP), sometimes referred to as caseinomacropeptide (CMP). However, the specificity of the bacterial proteases on κ-casein is not as defined as that of rennin. Miralles et al. (2003) reported that peptides released by Pseudomonas fluorescens protease included fragments 1–103, 1–104, 1–106, and 1–107. However, different specificities of bacterial proteases have been reported. This may have been due to the different bacterial species and strains tested, the presence of some active plasmin, or the different experimental conditions used (Nicodeme et al., 2005). Some authors have reported that β-casein was the most preferred: for example, Bagliniere et al. (2012) reported the order of preference of proteases from five strains of P. fluorescens to be β- > αs1- > κ- > αs2-casein. Proteolysis induced by bacterial proteases and its effects in UHT milk

Knowledge of the effects of bacterial proteases has been acquired through two main experimental approaches: studies of the effects on UHT milk of the growth of normal psychrotrophic bacteria in raw milk before processing; addition of cultures of proteolytic

400

10. The effect of UHT processing and storage on milk proteins

bacteria previously isolated from milk or of proteases purified from such cultures. Some examples of the former are as follows: • Law et al. (1977) grew a proteolytic strain of P. fluorescens (B11) in raw milk until the total plate count (TPC) reached three different levels (8  105, 8  106, and 5  107 cfu/mL), and then, UHT processed the milks at 140°C for 3.5 s. The UHT samples produced from milks with 8  106 and 5  107 cfu/mL gelled after 8–10 weeks and 10–14 days, respectively. These samples showed extensive hydrolysis of κ-casein with the formation of “para-κcasein,” considerable hydrolysis of β-casein, and some hydrolysis of αs1-casein. The UHT sample from the raw milk with 8  105 cfu/mL did not gel but showed some degradation of κ-casein; neither β-casein nor αs1-casein was affected in this milk. • Richardson and Newstead (1979) added crude preparations of heat-stable proteases produced by P. fluorescens isolates B12 and B52 to raw milk. Culture supernatants were diluted from 1 in 1000 to 1 in 100,000 in the raw milk before UHT processing on an indirect plant at 138–141°C for 2.8 s. After 3 months of storage, proteolysis and bitterness were detected in all treatment milks, even those with the lowest concentration of added protease. It was concluded that, for UHT milk to have a shelf life of more than 3 months, it should contain no more than 1–2 ng protease/mL. • Snoeren et al. (1979) and Snoeren and Both (1981) UHT processed “good” milk (TPC ¼ 2.7  103 cfu/mL) and “bad” milk (TPC ¼ 2.07  106 cfu/mL) on direct and indirect UHT plants with holding tube conditions of 142°C for 4 s. The results, summarized in Table 10.3, show the much higher level of proteolysis in the “bad” milk than in the “good” milk and the resultant effects on the time to gelation. They further illustrate that indirect processing reduces the proteolysis and increases the time to gelation by increasing inactivation of the proteases present. There is also clear evidence in these data of plasmin-induced proteolysis, particularly in the directly processed milk. • Button (2007) stored raw milk at 4°C for 4 days (control) and 7 days (treatment) before UHT processing on an indirect plant at 140°C for 4 s. The TPCs of the control and treatment milks were 1.0  107 and 8.0  107 cfu/mL, respectively. Rapid proteolysis occurred in the treatment milk and it gelled at 15–19 days, whereas the control milk showed no proteolysis or gelation. It should be noted that this is an example of the TPC not giving a true indication of the proteolysis potential; it would normally be expected that milk with a TPC of 1.0  107 cfu/mL would contain significant protease activity.

TABLE 10.3 Effects of raw milk quality and processing mode on proteolysis and gelation in UHT milk Total proteolysis in 90 days at 28°C (increase in nonprotein nitrogen, mg/100 g)

Days to gelation at 28°C

Milk quality (TPC)

Direct

Indirect

Direct

Indirect

Good (2.7  10 cfu/mL)

34

7

91

>>91

Bad (2.07  10 cfu/mL)

40 (in only 20 days)

30

20

91

3

6

Data from Snoeren, T.H.M., van Riel, J.A.M., Both, P., 1979. Proteolysis during storage of UHT-sterilised whole milk. 1. Experiments with milk heated by the indirect system for 4 seconds at 142°C. Neth. Milk Dairy J. 33, 31–39; Snoeren, T.H.M., Both, P., 1981. Proteolysis during storage of UHT-sterilised whole milk. 2. Experiments with milk heated by the direct system for 4 seconds at 142°C. Neth. Milk Dairy J. 35, 113–119.

Protein changes during processing and storage

401

• Bagliniere et al. (2012) grew nine Pseudomonas isolates “selected because of their differences in terms of proteolytic activity quantified using azocasein as substrate” in raw milk to 2.5  106 cfu/mL. These milks and an uninoculated control were UHT processed on an indirect plant at 140°C for 4 s. The control milk and four of the inoculated milks showed no sign of destabilization, whereas the other five inoculated milks were destabilized with the production of a “white sediment.” Particle size analysis showed that all stable milks had a monomodal size distribution with an average size of  0.2 μm (which corresponded to unmodified casein micelles), whereas the unstable milks showed populations of particles with sizes up to 700 μm, which the authors assumed were aggregates of casein micelles. The unstable milks also showed low phosphate test results; a similar finding was made by Gaucher et al. (2011). • Pinto et al. (2016) produced UHT milk from raw milk containing 3.0  106 cfu/mL of psychrotrophic bacteria and 8.0  105 cfu/mL of proteolytic psychrotrophs. Although the UHT treatment reduced the protease activity by 93.2%, proteolysis and associated sedimentation still occurred during storage of the UHT milk. • Stoeckel et al. (2016) UHT processed raw milk to which aliquots of milk cultures of one of three Pseudomonas species were added to give appropriate protease levels (apparent protease activities in the milks were from 0.01 to 0.35 pkat/mL, which relate to activities measured against the substrate azocasein). The UHT processing conditions used were indirect (tubular) heating with preheating at 90°C for 40 s followed by heating to 140°C and holding for 8.4 s (as discussed earlier, such conditions are excellent for ensuring that there is no residual plasmin in the final products; unfortunately, this has not been the case for several trials and hence the role of plasmin in some of those trials cannot be ruled out). The UHT milk samples were stored at 20°C for up to 4 months. The authors summarized their results by stating that the defects in the milks “were detected in the order: bitterness > formation of particles > creaming > sediment (> 5%) > gelation”; the order was chronological not severity. The control milk samples, which contained no added Pseudomonas cultures, showed no negative product changes. Some examples of the second approach are as follows: • Gebre-Egziabher et al. (1980) added various dilutions of a filter-sterilized supernatant of a Pseudomonas culture to commercial UHT milk. At an addition level of 2 protease units/mL, bitterness developed within 7 days at 21°C (1 protease unit was defined as the amount of enzyme producing 1 μg of acid-soluble tyrosine/mL of enzyme solution per 24 h at 40°C). Addition of higher enzyme concentrations resulted in the development of an unclean flavor and bitterness, followed by coagulation, within 3 days at 21°C. • Mitchell and Ewings (1985) aseptically added eight purified proteases, six of P. fluorescens and two of Serratia marcesens, to UHT milk at concentrations from 0.3 to 30 ng/mL, and studied the effects on proteolysis, flavor, and gelation. A control milk with no added protease, held under the same conditions as the trial milk samples, showed no protein breakdown at up to 22 weeks of storage. All milks with added protease showed proteolysis and gelled. The time to gelation varied from 2 to 21 weeks. Bitterness developed before gelation in all cases; however, the flavor progressed through the following stages: lacks freshness, slightly stale, unclean, mild off-flavor, slightly bitter, and bitter. There was no statistical relationship between time of onset of gelation and proteolysis, as measured by

402

10. The effect of UHT processing and storage on milk proteins

nonprotein nitrogen or noncasein nitrogen; however, there was a (negative) relationship between the time to gelation and the protease activity in the milk. A threshold activity value of 30 U/mL, below which UHT milk could be expected to be stable at 23°C for 5 months, was proposed. Such levels of protease activity in milk are beyond the sensitivity level of existing assay methods. All eight proteases showed a distinct preference for κ-casein, followed by β-casein, and then αs-caseins in milk. • Button et al. (2011) added a cell-free culture of a proteolytic P. fluorescens to UHT milk at from 0.0003% to 0.006% and stored the milk at 25°C for 12 months. The level of proteolysis, as measured by the concentration of free amino groups, increased in all samples during storage. The authors determined that a concentration of free amino groups of at least 12 mM Leu-Gly equivalents was required before the onset of age gelation. In this trial, this level was reached in the milk sample containing the lowest level (0.0003%) of added crude protease after 5 months of storage. Distinguishing between proteolysis induced by plasmin and by bacterial proteases Plasmin and bacterial proteases cause similar effects in milk. Both cause age gelation and bitterness, and there is growing evidence that both can be associated with sediment formation and creaming or fat separation. However, to be able to remedy problems associated with proteolysis, it is necessary to be able to diagnose the cause. If the cause is plasmin, then adjusting the processing conditions, as discussed earlier, is essential; however, if the cause is bacterial protease, then the quality of the raw milk needs to be addressed. An alternative for the latter is increasing the severity of the sterilization conditions, particularly increasing the holding time, that is, the length of the holding tube; however, such heating conditions exacerbate the level of cooked flavor. In terms of gelation, plasmin gels are soft and voluminous, whereas gels initiated by bacterial proteases tend be firm and more compact, and resemble rennet gels (Zhang et al., 2018). However, it is sometimes difficult to distinguish between the causes when the milk presents with some gel-like material or sediment in the bottom of the container and possibly some creaming. Such changes could be the beginning of more complete gelation caused by either plasmin or bacterial protease. In addition, it is quite possible to have both active plasmin and bacterial proteases present; this can occur if the raw milk quality is poor and the UHT processing conditions are insufficient to inactivate plasmin. Laboratory-based tests for distinguishing between proteolysis by plasmin and by bacterial proteases are now available (Lo´pez Fandin˜o et al., 1993; Datta and Deeth, 2003). They are based on the fact that plasmin preferentially acts on β-casein, whereas bacterial proteases act primarily on κ-casein. These tests have proved to be valuable in determining the cause of proteolysis in practice (e.g., Topc¸u et al., 2006). This is despite some reports indicating that some bacterial proteases preferentially attack β-casein (Gebre-Egziabher et al., 1980; Bagliniere et al., 2012). UHT milk containing such proteases would give misleading results by the tests described here. The tests are based on the properties of the peptides produced. Namely, plasmin tends to produce large hydrophobic peptides and bacterial proteases tend to produce small hydrophilic peptides. Therefore, by RP-HPLC, the peptides produced by bacterial protease will elute before those produced by plasmin. Furthermore, the peptides from bacterial proteases

403

Protein changes during processing and storage

TABLE 10.4 Interpretation of peptide analyses by RP-HPLC and total amino groupsa for diagnosing the cause of proteolysis in UHT milk 4% Trichloroacetic acid filtrate

pH 4.6 filtrate

Peptide peaks by RP-HPLC

Proteolysis— amino groups

Peptide peaks by RP-HPLC

Proteolysis— amino groups

Interpretation

Virtually no peaks after the initial solvent peak

Little or none

Virtually no peaks after the initial solvent peak

Low

Good quality milk

Few peaks after the initial solvent peak

Little or none

Significant late peaks

Significant

Plasmin action only

Significant early peaks but few late peaksb

Significant

Significant early peaks but few late peaks

Significant

Bacterial proteinase action only

Significant early peaks but few late peaks

Significant

Significant early and late peaks

Significant

Both bacterial proteinase and plasmin action

a

Determined by a method such as fluorescamine. In the original paper (Datta and Deeth, 2003), early and late peaks were defined as eluting before and after 20 min, respectively; however, elution times are dependent on the conditions of analysis and hence specific times have not been included here.

b

are soluble in both pH 4.6 filtrate and 4% trichloroacetic acid, whereas those produced by plasmin are soluble in pH 4.6 filtrate but insoluble in 4% trichloroacetic acid. The peptides in the pH 4.6 filtrate and the 4% trichloroacetic acid can be quantified by methods such as fluorescamine and ortho-phthalaldehyde (OPA), which detect total amino groups (Vaghela et al., 2018). Table 10.4 outlines the possible results of peptide analyses by RP-HPLC and total amino group analysis and their interpretation for diagnosing the cause of proteolysis in UHT milk. Reducing enzymatic proteolysis in UHT milk In summary, plasmin-induced proteolysis can be greatly reduced, if not prevented, by the selection of appropriate preheat UHT conditions. Bacterial proteolysis can be prevented by using raw milk with low bacterial counts. If bacterial proteases are present in the raw milk, their effect can be minimized by the use of quite severe sterilization conditions. A further method to reduce the effect of proteases is to employ what is known as low temperature inactivation (LTI). This has been shown by several authors to be effective in reducing protease action (Hill, 1988; Reddy et al., 1991). Interestingly, Zhang et al. (2007) found that LTI of proteases reduced fat separation, a further indication of the role of proteolysis in the fat separation or creaming of UHT milk during storage. LTI involves a mild heat treatment, which can be performed before or after UHT processing. Typical heating conditions for LTI are  55°C for 30–60 min (Barach et al., 1976) although different proteases may require different conditions for optimum effect (Kocak and Zadow, 1985c). The mechanism of LTI is unclear but has been suggested to be due to autodigestion of the enzyme or changes in the conformation of the enzyme through interaction with milk proteins (Schokker and van Boekel, 1999).

404

10. The effect of UHT processing and storage on milk proteins

Proteolysis in UHT hydrolyzed-lactose milk Several authors have reported higher levels of proteolysis in UHT hydrolyzed-lactose milk than in unmodified milk (Tossavainen and Kallioinen, 2007; Jansson et al., 2014). The proteolysis has been attributed to the proteolytic side activity in the β-galactosidase preparations. Jansson et al. (2014) reported that β-casein and αs1-casein are the most susceptible milk proteins to this proteolysis. Heat-induced proteolysis Several peptides are produced from β- and αs1-caseins when milk is heated and some have been proposed as markers for distinguishing UHT milk from other heat-treated milk. For example, Dalabasmaz et al. (2017) found 12 peptides that increased with time of heating (at 99°C) and could be used to distinguish UHT milk from mildly heated milk (pasteurized and extended shelf life). However, the peptide with the highest discriminatory power was found to be pyroGln-β-casein 194–209 with an m/z of 1710. It originated from cleavage at Tyr193-Gln194 at the C-terminal region of β-casein and the concomitant formation of pyroGln. Formation of the pyroGln modification was associated with a mass reduction of 17 Da, because of the loss of ammonia, compared with the parent peptide with an m/z of 1718.0 Da. Interestingly, this latter peptide is also formed by cleavage at Tyr193-Gln194 by cathepsin B, which can be active in UHT milk. However, Dalabasmaz et al. (2017) believed that cathepsin-B-induced formation of β-casein 194–209 could occur in parallel with the heat-induced formation of pyroGln-β-casein 194–209. Interestingly, Gaucher et al. (2008) identified eight peptides from αs1-casein in UHT milk that they concluded could have been produced by heat treatment or cathepsin B; many of these were not detected in freshly processed UHT milk, but only after storage, which casts doubt on their origin being from heat treatment. The authors suggested that long-term storage could have a similar effect to heat treatment. Ebner et al. (2016) identified nine peptides that increased with time of heating (at 120°C) and had higher intensity in UHT milk than in high-temperature-short-time (HTST)-pasteurized milk when analyzed by matrix-assisted laser desorption/ionization time-of-flight (MALDITOF) mass spectrometry. Of these, the peptide β-casein 196–209 (m/z 1460.9 Da) was most influenced by the intensity of the heat treatment and was suggested by these authors to be a possible heat treatment marker. Practical effects of heat-induced proteolysis have not been reported. However, it may contribute to destabilization of the casein micelle, which leads to sedimentation and/or gelation.

Age gelation Age gelation is one of the major causes of a reduction in the shelf life of UHT milk. It is now well established that proteolysis by plasmin or bacterial proteases is the major initiator of gelation. However, gelation has also been observed in milk in which proteolysis did not appear to be a factor; the cause of this type of gelation has been designated to be physicochemical (Datta and Deeth, 2001; Nieuwenhuijse and van Boekel, 2003; Deeth and Lewis, 2016, 2017; Anema, 2017). Physicochemical gelation tends to occur in UHT milk with high solids content, that is, concentrated milks, although cases involving nonconcentrated milk have also

Protein changes during processing and storage

405

been reported (Auldist et al., 1996; Anema, 2017). According to Anema (2017), it is more likely to occur in milk after long storage. As indicated earlier, several studies have indicated the involvement of proteolysis in age gelation. Proteolysis caused by both plasmin and bacterial proteases has been proved to be involved. However, curiously, many authors have not been able to find a good correlation between the time to onset of gelation and the extent of proteolysis. There has been more success in relating the protease activity to the onset of gelation (e.g., Mitchell and Ewings, 1985). However, most protease assays are not sensitive enough to quantify the levels that can cause gelation after storage of the milk. Assays involving long incubation times have been shown to detect low protease activities (Button et al., 2011). However, scientists in Australia have developed a biosensor-based method, the Cybertongue method, that can detect “proteases in milk at concentrations relevant to industry within a few minutes” (Anon, 2014). The biosensor uses a Bioluminescence Resonance Energy Transfer (BRET) transduction system. Factors that influence the time of onset of age gelation in UHT milk Temperature of storage

Several authors have shown that the susceptibility of UHT milk to age gelation is low at low temperatures, reaches a maximum in the 25–28°C range, and then decreases at temperatures above 30°C (Kocak and Zadow, 1985a; Manji et al., 1986). The reason for this is not entirely clear. However, one hypothesis is that, at the high temperatures, proteolysis proceeds at such a rate as to prevent the formation of the ordered protein network that is required for a gel. A second hypothesis is that the higher temperature favors protein cross-linking, which stabilizes the casein micelles and prevents release of the β-Lg-κ-casein complex into the milk serum, which would otherwise form into a gel (McMahon, 1996). The cross-linking can occur either through dehydroalanine forming linkages between lysine and alanine or histidine on adjacent molecules (Al-Saadi and Deeth, 2008; Holland et al., 2011), or via advanced Maillard reaction products such as the dicarbonyls glyoxal and methylglyoxal (Le et al., 2013). Other reactions that may retard gelation at temperatures >30°C are lactosylation of lysine residues and deamidation of asparagine and glutamine residues of proteins. The involvement of lactosylation was indicated by Malmgren et al. (2017), who showed that high degrees of lactosylation were associated with the absence of gelation and long tendrils. Deamidation alters the charge on proteins and can cause changes in their physical properties. An implication of the phenomenon of decreased gelation at higher temperatures of storage is that accelerated shelf life trials to assess the susceptibility of a product to age gelation at temperatures higher than  30°C will provide misleading results (Grewal et al., 2017a). However, such trials can provide valuable information on the susceptibility of the product to proteolysis (Deeth and Lewis, 2017). They have also been shown to be useful for assessing susceptibility to sedimentation (Grewal et al., 2017c). Temperature-time conditions of heat treatment

The importance of selecting preheat conditions that will inactivate plasmin and prevent plasmin-related proteolysis and gelation was mentioned in the section on Heat inactivation of plasmin. However, the temperature and time in the sterilization holding tube can also be manipulated to minimize proteolysis and age gelation. For example, Topc¸u et al. (2006)

406

10. The effect of UHT processing and storage on milk proteins

reported that increasing the conditions from 145°C for 4 s to 150°C for 4 s substantially reduced proteolysis and bitterness in UHT milk. However, increasing the severity of the high-heat step has a major effect on the flavor of the product. In other words, there is a trade-off between flavor retention and physical stability of the product. Mechanism of age gelation As indicated earlier, proteolysis by either plasmin or bacterial proteases is accepted as the major cause of age gelation, but it can also occur by physicochemical means in the absence of proteolysis. Proteolysis-induced gelation is generally accepted to occur via a three-stage process. The first stage occurs during heat treatment and involves the interaction of β-Lg with κ-casein on the micelle surface and with κ-casein in the serum phase, which is released from the micelle during heating to form a β-Lg-κ-casein complex (Hillbrick et al., 1999; Anema, 2017). Electron microscopy has been used to show tendrils of β-Lg-κ-casein complex attached to casein micelles (McMahon, 1996; Malmgren et al., 2017). The tendrils have been shown to grow in length prior to gelation; when prevented from growing because of, for example, extensive lactosylation, gelation does not occur (Malmgren et al., 2017). The second stage is proteolysis of the caseins within the casein micelle, which cleaves the peptide bonds that anchor the κ-casein to the casein micelle, facilitating the release of the β-Lg-κ-casein complex tendrils into the serum. The third stage is the formation of a gel from the β-Lg-κ-casein complexes in the serum when their concentration exceeds a critical value (McMahon, 1996). For the physicochemical type of age gelation, Anema (2017) proposed a mechanism involving the aggregation of the κ-casein-depleted casein micelles that are formed during heat treatment. The aggregation was proposed to be mediated by calcium bridging. Retardation of age gelation with polyphosphate The most common additive that is used commercially to delay age gelation is sodium hexametaphosphate (SHMP), sometimes referred to as “polyphosphate.” Some UHT milk manufacturers add it routinely but this may not be allowed in some jurisdictions. When added to milk at  0.1% (w/w) before UHT treatment, it extends the time to gelation (Kocak and Zadow, 1985b). In terms of its mode of action, SHMP does not prevent proteolysis but appears to prevent the subsequent aggregation of the milk proteins and the formation of the gel. In fact, milk with added SHMP may undergo considerable proteolysis without gelling. The effect of SHMP is probably due to its association with casein, thereby altering the net charge on the casein micelles (increasing their negative charge). The addition of a low level of SHMP would facilitate bridging between ionized groups of casein micelles, which would not otherwise form an ionic bond. This would hold the κ-casein more tightly to the micelle and delay release of the β-Lg-κ-casein complex, retarding gelation of the UHT milk during storage. Another possible role of SHMP is to bind ionic calcium and reduce its activity in the milk. However, this role seems to be unlikely as EDTA and citrate, which also bind ionic calcium, accelerate rather than retard gelation (Kocak and Zadow, 1985b). Fat separation or creaming related to age gelation Some creaming or fat separation occurs in almost all fat-containing UHT milks. Some of this is undoubtedly related to poor homogenizer performance and maintenance, but there

Protein changes during processing and storage

407

is increasing evidence that it is related to age gelation and the often-associated proteolysis, because the defects tend to occur at the same time (Kohlmann et al., 1991; Hardham, 1998; Stoeckel et al., 2016). As homogenization causes the formation of a casein-dominant membrane around the fat globules (Lee and Sherbon, 2002), bacterial proteases can hydrolyze the casein in this secondary milk fat globule membrane during the storage of UHT milk and lead to aggregation of the fat globules in a similar manner to the formation of a proteolysis-induced gel.

Sedimentation Sedimentation is one of the major defects that develop in UHT milk. It is defined as the material that settles to the bottom of the container in a fairly compact manner. Although there is some confusion in the literature between a gel and a sediment, it is generally accepted that a gel is softer and more voluminous than a sediment. The confusion sometimes arises when gelation commences in the bottom of the container (as mentioned in the section on The plasmin system) before it spreads to the rest of the container. Furthermore, as noted in that section, the formation of a sediment can be followed by gel formation (Malmgren et al., 2017). Most UHT milk contains some sediment. Lewis et al. (2011) reported 0.29% sediment in UHT cows’ milk. In general, the amount of sediment resulting from the UHT processing of good quality cows’ milk should be < 0.5% (On-Nom et al., 2012). The amount of sediment increases with the time of storage (Blanc and Odet, 1981; Ramsey and Swartzel, 1984; Stoeckel et al., 2016; Malmgren et al., 2017). In most cases, the sediment is composed of milk proteins, specifically caseins. Gaur et al. (2018) reported that the sediment consisted largely (>  85%) of aggregated κ-casein-depleted casein micelles with small amounts of whey proteins (β-Lg and α-La) and κ-casein; a similar predominance of β- and αs-caseins in the sediment of directly processed UHT milk was reported by Malmgren et al. (2017). An exception to the dominance of protein in the sediment is the presence of fat in the sediment of fat-containing milk. For example, Grewal et al. (2017b) reported changes to Fourier transform infrared (FTIR) spectra that were associated with the conformation of the fat in whole milk samples stored for 14 days at 40°C and 50°C. They indicated the formation of an intermolecular beta sheet of proteins, which was attributed to protein-lipid interactions and aggregation. The sediment from these samples contained fat, which confirmed the involvement of protein-lipid interaction in the sedimentation. Boumpa et al. (2008) showed that the sediment from whole goats’ milk that was processed indirectly at 140°C for 2 s with upstream homogenization contained  40% fat. Another exception is the sediment in some UHT mineral-fortified milks such as those fortified with insoluble calcium salts, for example, calcium carbonate or calcium citrate. These sediments contain a high proportion of mineral salts. In contrast, Boumpa et al. (2008) reported 40% of the total solids in the milk, in UHT goats’ milk. The heat instability of goats’ milk is attributable to a high ionic calcium level. Hence to enable goats’ milk to be processed without excessive fouling and sedimentation, the ionic calcium level has to be reduced by stabilizers, such as phosphate and citrate salts, or to be removed by, for example, a cation exchange resin (Prakash et al., 2007). Boumpa et al. (2008) used three added stabilizers, disodium hydrogen phosphate, trisodium citrate, and sodium dihydrogen

Protein changes during processing and storage

409

phosphate, to reduce ionic calcium and found that, for a given reduced ionic calcium content, the amount of sediment following UHT processing was greatest for sodium dihydrogen phosphate and least for disodium hydrogen phosphate. Sediment levels as low as 0.13 g/100 mL were achieved in UHT goats’ milk when stabilizers were added. As well as ionic calcium, pH has a great effect on the heat stability of milk and hence its propensity to cause fouling and sedimentation (Lewis et al., 2011). Either a decrease in pH or an increase in ionic calcium decreases the heat stability of milk and reduces its ability to be successfully UHT processed and produce minimal sediment. Both of these changes reduce the ethanol stability, an indication of its suitability for UHT processing; the ethanol stability should be
74% for UHT processing (Deeth and Lewis, 2017, pp. 220–222). Gaur et al. (2018) attributed sedimentation in New Zealand UHT milks to high ionic calcium (
1.5 mM) and/ or low pH ( 6.65) and showed that sedimentation could be reduced by reducing the ionic calcium using chelators or by increasing the pH with alkali. Adding trisodium citrate could affect both of these modifications. Although such an addition may reduce sedimentation, it may have an adverse effect on age gelation. There have been numerous indications in the literature that sedimentation in UHT milk is linked with proteolysis. Newstead et al. (2006) observed a strong correlation between plasmin-induced proteolysis in directly processed UHT milk and sedimentation. Pinto et al. (2016) found a good correlation between sediment formation and proteolysis during the storage of UHT milk at 37°C and concluded that sedimentation could be prevented by the use of raw milk containing low levels of proteolytic psychrotrophic bacteria. Similarly, Bagliniere et al. (2012, 2017) reported that residual proteolytic activity of strains of P. fluorescens and Serratia liquifaciens L53 led to the destabilization of UHT milk, with sedimentation and the formation of aggregates. Likewise, Gaucher et al. (2009), Mat
eos et al. (2015), and Stoeckel et al. (2016) reported that casein micelles are destabilized by the action of Pseudomonas peptidases and that this results in the development of sediment. However, some reports suggest that there is not always a strong link between proteolysis and sediment formation. Richardson and Newstead (1979) added cultures of two proteolytic P. fluorescens strains, B12 and B52, to raw milk before UHT processing and found that, on storage of the UHT milk, one strain, B12, formed a “granular precipitate” (a sediment), whereas the other, B52, formed a “gel similar to that obtained by rennin action.” In a 16-week storage study by Rauh et al. (2014b) of UHT milk containing active plasmin, proteolysis but no sedimentation was observed. Malmgren et al. (2017) considered the link between proteolysis and sedimentation to be unlikely as, in their study, sediment occurred in milk stored at 5°C, but virtually no proteolysis occurred at this temperature.

Maillard reaction The following quote from Fennema (1996) indicates the importance of the Maillard reaction in food science: “Among the various processing-induced chemical reactions in proteins, the Maillard reaction (nonenzymic browning) has the greatest impact on sensory and nutritional properties.” Aspects of the Maillard reaction have been reviewed by O’Brien (1995), van Boekel (1998), and Mehta and Deeth (2016).

410

10. The effect of UHT processing and storage on milk proteins

Lactosylation The Maillard reaction commences with a reaction between a reducing sugar and a protein. In milk, this initial reaction is usually lactosylation, that is, the reaction between lactose and the ε-amino group of lysine and, to lesser extent, arginine, methionine, tryptophan, and histidine. The first major product formed with lysine is the protein-bound Schiff’s base lactosyl lysine, but this readily undergoes an Amadori rearrangement to produce the more stable protein-bound lactulosyl lysine (ε-N-deoxylactulosyl-L-lysine). This rearrangement is highly significant because the lysine in the Amadori product is biologically unavailable; the lactosylated lysine hinders proteolysis of the protein by digestive proteases. This is often referred to as lysine blockage (Mehta and Deeth, 2016). Therefore, the greater is the extent of lactosylation, the greater is the negative effect on the nutritional quality of the milk. However, the reduction in nutritive value in this manner is not considered to be a significant issue in UHT milk (Dehn-M€ uller et al., 1991). It is sometimes raised as a concern with infant formulas, including UHT-processed formulas, where up to one-third of the lysine can be blocked (Mehta and Deeth, 2016). However, infant formulas generally have a high ratio of whey proteins to casein (Fenelon et al., 2018), and whey proteins are rich in lysine ( 12%); therefore, the product will still contain an abundance of available lysine. The Maillard reaction in UHT milk is initiated during heat treatment but continues during storage. It is very dependent on temperature, and hence, the more severe is the heating process and the higher is the temperature of storage, the more the reaction will progress. Because of this, the extent of lactosylation is sometimes used as an index of heat treatment. It is usually measured indirectly as furosine (furosine does not occur in milk), which is formed in the analysis by acid digestion. However, because lactosylation continues during storage, often to a greater extent than during heating, furosine is a good indicator of thermal treatment only in freshly processed UHT milk. Lactosylation can also be observed by methods based on mass spectrometry (Holland et al., 2011; Siciliano et al., 2013). They have been used to observe the various lactosylated forms of whey proteins and caseins as each lactose adds 324 Da to the molecular weight. These forms can be observed not only in the mass spectrographs but also in two-dimensional electrophoresis gels, in which they appear as a series of parallel spots for each protein; they are most easily seen for β-Lg and α-La (Holland et al., 2011). More molecules of lactose attach to the milk proteins with increased severity of heat treatment and time and temperature of storage. Holland et al. (2011) showed five parallel spots for β-Lg in UHT milk that had been stored at 40°C for 2 months; the five bands represented β-Lg with zero, one, two, three, and four attached lactose molecules. By analyzing tryptic digests of lactosylated proteins, it is possible to locate the positions of the lactosylated lysines. Interestingly, the lactose in a monolactosylated molecule may be attached to different lysines in different molecules. Fogliano et al. (1998) reported that the lactosylation of the caseins in UHT milks seemed to be nonspecific, with 7 of the 14 lysines in αs1-casein and 5 of the 11 lysines in β-casein being lactosylated. In monolactosylated forms of β-Lg, Lys100 was preferentially lactosylated. Lactosylation of the milk proteins during storage is one reason why new peaks for the proteins appear in HPLC chromatograms of the proteins in stored UHT milk. Furthermore, it is also a reason for existing peaks to broaden and for the chromatograms to contain an increasing number of protein species (Gaucher et al., 2008).

411

Protein changes during processing and storage

Browning A brown discoloration of UHT milk is an indication of the final stage of the Maillard reaction. This is due to the formation of melanoidins, a complex mixture of largely highmolecular-weight brown pigments. The melanoidins formed by reaction with amino acids bound to proteins have a protein skeleton (Wang et al., 2011). The heating conditions in UHT processing are normally insufficient to cause a noticeable brown coloration in milk. For example, the conditions represented by the temperature-time profiles in Fig. 10.1A and B are equivalent to 106 and 38 s at 121°C (based on z ¼ 26.3°C; Browning et al., 2001); according to Fink and Kessler (1988), a brown coloration is noticeable at the equivalent of
400 s at 121°C. Although little browning occurs in unmodified UHT milk during heating, it occurs during storage, particularly if the storage temperature is
30°C. This is particularly significant for UHT hydrolyzed-lactose milk, in which both the products of the hydrolysis, glucose and galactose, are more Maillard reactive than lactose. The data in Table 10.5 illustrate the increase in color in UHT whole milk, skim milk, and hydrolyzed-lactose milk after storage for 4 months at 35°C and 50°C; greater color development occurred in all milks at 50°C than at 35°C and more occurred in hydrolyzed-lactose milk than in unmodified milks (data derived from Deeth and Lewis, 2017, p. 295). The ΔE value for visual perception of a brown color in milk has been reported to be 3.8 (Pagliarini et al., 1990). Thus, all milk such as those shown in Table 10.5 would be noticeably brown. When Maillard browning occurs in UHT milk, there is also a significant decrease in pH. This is largely due to the concomitant formation of formic and acetic acids, which are also products of the Maillard reaction (Adhikari and Singhal, 1991; van Boekel, 1998).

Deamidation Deamidation is the hydrolysis of the amide groups in asparagine (Asn) and glutamine (Gln) residues to aspartic acid (Asp) and glutamic acid (Glu), respectively, and ammonia. Nonenzymatic deamidation occurs to only a small degree during normal UHT processing but occurs significantly during the storage of UHT milk (Holland et al., 2011). The level of deamidation increases with temperature and time of storage. Asparagine residues adjacent to glycines show the highest rates of deamidation, but asparagine residues adjacent to serines TABLE 10.5 Browning, as measured by color difference (ΔE) values, for UHT whole milk, skim milk, and hydrolyzed-lactose milk stored at 35°C and 50°C for 4 months; color comparisons were made with the same product stored at 4°C Product

Storage at 35°C

Storage at 50°C

Whole milk (average of 3)

4.8

21.1

Skim milk (average of 2)

10.6

33.7

Hydrolyzed-lactose milk

20.3

43

Data from Deeth, H.C., Lewis, M.J., 2017. High Temperature Processing of Milk and Milk Products, Wiley Blackwell, Oxford, UK, p. 295.

412

10. The effect of UHT processing and storage on milk proteins

and histidines are also deamidated. Deamidation of asparagine occurs more readily than that of glutamine (van Boekel, 1999). Holland et al. (2012) examined deamidation of αs1-casein. Two-dimensional electrophoresis showed predominantly one spot for this casein at manufacture; however, during storage, this spot split into multiple spots with approximately the same molecular weights. Each new spot represented the loss of ammonia and one negative charge. Up to four new spots appeared for samples stored at 40°C for 2 months. Each spot could represent more than one molecular species with the same charge. Several sites of deamidation were identified. These included Gln24, Asn32, Asn129, and Asn205. Both Holland et al. (2011) and Le et al. (2016) reported that Asn63 was the site for deamidation of β-Lg A and B. Although deamidation can be clearly detected by two-dimensional electrophoresis (Holland et al., 2011, 2012; Le et al., 2016), it can also be detected by measuring the ammonia released (van Boekel, 1999). Gaucheron and Le Graet (2000) reported a sensitive method (to < 0.5 mg/kg) based on separation of the ammonium ion by cation exchange chromatography and detection by suppressed conductivity. The effects of deamidation during storage on the functional and sensory qualities of UHT milk have not been determined. However, Le et al. (2016) showed that the proteins in a shelfstable acidic whey protein beverage produced by heating at 120°C for 20 s were deamidated during 12 months of storage. They suggested that deamidation, in which an Asn is converted to an Asp, may increase the susceptibility of proteins and peptides to acid hydrolysis by making an extra hydrolysis site available. Acid hydrolysis during storage of these beverages was beneficial as it reduced the formation of CMP-rich aggregates. Furthermore, it is also now known that the deamidation of proteins can increase their heat stability, reduce the extent of fouling deposit formation, preserve their nutritional value through reducing blockage of lysines, and improve some physical functional properties (Miwa et al., 2010; Timmer-Keetels et al., 2011). Miwa et al. (2010) have shown that enzymatic deamidation using protein glutaminase improves the solubility, viscosity, and emulsification capacity of skim milk solutions.

Cross-linking Cross-linking is a major change that occurs in the proteins in UHT milk. The nature of the cross-linking includes disulfide bonding and linking via dehydroalanine and Maillard reaction products. Other forms of cross-linking such as via lipid oxidation products may also be involved (Singh, 1991; Grewal et al., 2017a). Disulfide bonding occurs mainly during the heating process and occurs when the whey proteins are denatured, as discussed in the section on whey protein denaturation. Nondisulfide cross-linking can occur during thermal processing but occurs mainly during storage of the product. Nondisulfide cross-linking can be distinguished from disulfide cross-linking, as the latter links are broken by reduction with sulfhydryl reagents such as 2-mercaptoethanol and dithiothreitol. Therefore, high-molecular-weight bands remaining in reducing electrophoresis gels are due to nondisulfide cross-linking. These were clearly shown in two-dimensional electrophoresis gels by Holland et al. (2011). The extent of cross-linking increases with both temperature and duration of storage. Reports of the extent of cross-linking occurring during UHT processing have varied. By electrophoresis, little if any was detectable (Al-Saadi and Deeth, 2008; Holland et al., 2011).

Protein changes during processing and storage

413

In contrast, by high-performance gel permeation chromatography, 5% of the protein in indirectly processed (138°C for 2 min) UHT milk was found to be cross-linked (Zin El-Din et al., 1991); by size-exclusion chromatography under denaturing reducing conditions, up to 14% cross-linking was found in freshly processed milk (Andrews, 1975; Lauber et al., 2001). However, much higher percentages of cross-linked protein occur in stored UHT milk, particularly if the milk is stored at elevated temperatures. Andrews (1975) reported that more than 50% of the casein was polymerized in UHT milk after storage at 37°C for 9 months; 23% was polymerized after storage at 4°C for 9 months. Al-Saadi and Deeth (2008) reported that the storage of UHT milk at 45°C for 12 weeks caused a 36% increase in high-molecular-weight proteins, but little increase occurred during storage at 5°C and 20°C. This cross-linking is a major cause of the marked change in the HPLC chromatograms of the proteins that is observed during the storage of UHT milk (Al-Saadi and Deeth, 2008; Gaucher et al., 2008). Using a proteomic approach, Holland et al. (2011) showed that the high-molecular-weight bands on two-dimensional electrophoresis gels of stored UHT milk contained mostly αs1-casein together with some β-casein and αS2-casein. Interestingly, this is consistent with the suggestion by Andrews (1975) that αs1-casein was most involved and that β-casein was involved to a lesser extent. A major pathway for nondisulfide cross-linking is via (protein-linked) dehydroalanine, which is a reactive intermediate formed by the elimination of phosphate from O-phosphorylserine, of a saccharide from O-glycosylserine, or of hydrogen sulfide from cysteine. It can react with lysine, histidine, or cysteine to produce lysinoalanine (LAL), histidinoalanine, and lanthionine cross-links (Friedman, 1999). These isodipeptides can be detected by amino acid analysis, gas chromatography/mass spectrometry, HPLC, and other methods after digestion with acid or enzymatic proteolysis, as the bonding between the amino acids is not a classical amide peptide bond. LAL is the most commonly measured isodipeptide. Fritsch et al. (1983) reported up to 50 mg LAL/kg protein in indirectly processed UHT milk, which was much less than in in-container sterilized milk, that is, 110–710 mg LAL/kg protein. Al-Saadi and Deeth (2008) found indirectly processed UHT milk to contain 62 mg LAL/kg protein immediately after manufacture. After storage at 5°C and 20°C, the LAL had increased to 64 and 70 mg/kg protein, respectively; after storage at 37°C and 45°C, it had increased to 130 and 170 mg/kg protein, respectively. Protein cross-linking can also occur via several Maillard reaction products. These include formaldehyde and dicarbonyls such as glyoxal and methylglyoxal. Such cross-linking in UHT milk was first proposed by Andrews and Cheeseman (1971). Le et al. (2013) demonstrated the involvement of Maillard reaction products in cross-linking by incubating milk proteins with methylglyoxal and producing similar polymerized proteins to those observed in UHT milk during storage. In milk products containing reducing sugars such as lactose and the lactose hydrolysis products, glucose and galactose, cross-linking via these reactive Maillard products can occur concomitantly with cross-linking via dehydroalanine (Al-Saadi et al., 2013). Protein cross-linking in milk may have some practical consequences. It can reduce the digestibility of the proteins and decrease the availability of some essential amino acids such as lysine (Friedman et al., 1981). However, it may have some advantages. As discussed in the section on the effect of storage temperature on age gelation, McMahon (1996) suggested that intramicellar cross-linking of caseins may be the reason for the decreased susceptibility of UHT milk to gelation when stored at temperatures >30°C.

414

10. The effect of UHT processing and storage on milk proteins

Conclusions Several chemical and physical changes to the proteins occur during the processing and storage of UHT milk. Many of the changes can be minimized by storage at low temperature, but this negates a major advantage of UHT milk, that is, it does not require refrigeration. Therefore, other strategies are required to minimize the changes. At this stage, some of the changes, such as deamidation, are mainly of academic interest but others have major practical and ultimately economic implications. Although the causes of some defects are not completely understood, other defects such as those caused by proteolysis are now well understood. Current knowledge enables defects caused by proteolysis by plasmin to be effectively prevented and those caused by bacterial proteases to be identified and managed through attention to raw milk quality.

References Adams, D.M., Barach, J.T., Speck, M.L., 1976. Effect of psychrotrophic bacteria from raw milk on milk proteins and stability of milk proteins to ultrahigh temperature treatment. J. Dairy Sci. 59, 823–827. Adhikari, A.K., Singhal, O.P., 1991. Changes in the flavour profile of indirectly heated UHT milk during storage: effect of Maillard browning and some factors affecting it. Indian J. Dairy Sci. 44, 442–448. Al-Attabi, Z.H.R., 2009. Volatile Sulphur Compounds in UHT Milk (PhD thesis). University of Queensland, Brisbane, QLD, Australia. Al-Attabi, Z., D’Arcy, B.R., Deeth, H.C., 2009. Volatile sulphur compounds in UHT milk. Crit. Rev. Food Sci. Nutr. 49, 28–47. Al-Attabi, Z., D’Arcy, B.R., Deeth, H.C., 2014. Volatile sulfur compounds in pasteurised and UHT milk during storage. Dairy Sci. Technol. 94, 241–253. Al-Saadi, J.M.S., Deeth, H.C., 2008. Cross-linking of proteins and other changes in UHT milk during storage at different temperatures. Aust. J. Dairy Technol. 63, 93–99. Al-Saadi, J.M.S., Easa, A.M., Deeth, H.C., 2013. Effect of lactose on cross-linking of milk proteins during heat treatments. Int. J. Dairy Technol. 66, 1–6. American Dairy Products Institute (ADPI), 2009. Standards for Grades of Dry Milks Including Methods of Analysis, third ed. ADPI Bulletin 916. ADPI, Elmhurst, IL. Andreini, R., Choidi, J., De Noni, I., Resmini, P., Battelli, G., Cecchi, L., Todesco, R., Cattaneo, T.M.P., Rampilli, M., Foschino, R., 1990. Evaluation of heat damage in UHT and in-bottle sterilized milk samples treated in Italy. Scienza e Tecnica Lattier-Caseria 41, 472–492 (in Italian). Andrews, A.T., 1975. Properties of aseptically packed ultra-high-temperature milk. 3. Formation of polymerized protein during storage at various temperatures. J. Dairy Res. 42, 89–99. Andrews, A.T., Cheeseman, G.C., 1971. Properties of aseptically packed UHT milk: casein modification during storage and studies with model systems. J. Dairy Res. 38, 193–207. Anema, S.G., 2008. On heating milk, the dissociation of kappa-casein from the casein micelles can precede interactions with the denatured whey proteins. J. Dairy Res. 75, 415–421. Anema, S.G., 2009. The whey proteins in milk: thermal denaturation, physical interactions and effects on the functional properties of milk. In: Thompson, A., Boland, M., Singh, H. (Eds.), Milk Proteins: From Expression to Food. Academic Press/Elsevier, Amsterdam, The Netherlands, pp. 230–281. Anema, S.G., 2017. Storage stability and age gelation of reconstituted ultra-high temperature skim milk. Int. Dairy J. 75, 56–67. Anema, S.G., Li, Y.M., 2003. Effect of pH on the association of denatured whey proteins with casein micelles in heated reconstituted skim milk. J. Agric. Food Chem. 51, 1640–1646. Anema, S.G., Li, Y.M., 2015. Reassociation of dissociated caseins upon acidification of heated pH-adjusted skim milk. Food Chem. 174, 339–347.

References

415

Anema, S.G., Hill, J.P., Boland, M.J., McKenna, A.B., 1996. Kinetic and thermodynamic study of the denaturation of α-lactalbumin, β-lactoglobulin A and β-lactoglobulin B in heated reconstituted whole milk. In: Heat Treatments & Alternative Methods (Document 9602). International Dairy Federation, Brussels, Belgium, pp. 367–372. Anon, 2014. CSIRO Develops New Biosensors for Dairy Safety and Quality. http://www.ausfoodnews.com.au/ 2014/05/28/csiro-develops-new-biosensors-for-dairy-safety-and-quality.html. (Accessed 13 December 2018). Aschaffenburg, R., 1950. A simple test for sterilised milk. J. Soc. Dairy Technol. 3, 236–237. Ashokkumar, M., Lee, J., Zisu, B., Bhaskharacharya, R., Palmer, M., Kentish, S., 2009. Hot topic: sonication increases the heat stability of whey proteins. J. Dairy Sci. 92, 5353–5356. Auldist, M.J., Coats, S.J., Sutherland, B.J., Hardham, J.F., McDowell, G.H., Rogers, G.L., 1996. Effect of somatic cell count and stage of lactation on the quality and storage life of ultra high temperature milk. J. Dairy Res. 63, 377–386. Bagliniere, F., Tanguy, G., Jardin, J., Mat
eos, A., Briard, V., Rousseau, F., Robert, B., Beaucher, E., Humbert, G., Dary, A., Amiel, C., Gaucheron, F., 2012. Quantitative and qualitative variability of the caseinolytic potential of different strains of Pseudomonas fluorescens: implications for the stability of casein micelles of UHT milks during their storage. Food Chem. 135, 2593–2603. Bagliniere, F., Mat
eos, A., Tanguy, G., Jardin, J., Briard-Bion, V., Rousseau, F., Robert, B., Beaucher, E., Gaillard, J.L., Amiel, C., Humbert, G., Dary, A., Gaucheron, F., 2013. Proteolysis of ultra high temperature-treated casein micelles by AprX enzyme from Pseudomonas fluorescens F induces their destabilisation. Int. Dairy J. 31, 55–61. Bagliniere, F., Jardin, J., Gaucheron, F., de Carvalho, A.F., Vanetti, M.C.D., 2017. Proteolysis of casein micelles by heatstable protease secreted by Serratia liquefaciens leads to the destabilisation of UHT milk during its storage. Int. Dairy J. 68, 38–45. Barach, J.T., Adams, D.M., Speck, M.L., 1976. Low temperature inactivation in milk of heat-resistant proteases from psychrotrophic bacteria. J. Dairy Sci. 59, 391–395. Bastian, E.D., Brown, R.J., 1996. Plasmin in milk and dairy products: an update. Int. Dairy J. 6, 435–457. Bikker, F.J., Koop, G., Leusink, N.B., Nazmi, K., Kaman, W.E., Brand, H.S., Veerman, E.C.I., 2014. Tailor made plasmin substrates as potential diagnostic tool to test for mastitis. Vet. Res. Commun. 38, 271–277. Blanc, B., Odet, G., 1981. Appearance, flavour and texture aspects: recent developments. In: New Monograph on UHT Milk (Document 133). International Dairy Federation, Brussels, Belgium, pp. 25–48. Boumpa, T., Tsioulpas, A., Grandison, A.S., Lewis, M.J., 2008. Effects of phosphates and citrate on sediment formation in UHT goats’ milk. J. Dairy Res. 75, 160–166. Browning, E., Lewis, M., MacDougall, D., 2001. Predicting safety and quality parameters for UHT-processed milks. Int. J. Dairy Technol. 54, 111–120. Burton, H., 1968. Reviews of progress of dairy science, section G, deposits from whole milk in heat treatment plant—a review and discussion. J. Dairy Res. 35, 317–330. Burton, H., 1988. Ultra-High Temperature Processing of Milk and Milk Products. Elsevier Applied Science, London, UK. Button, P.D., 2007. UHT Milk Spoilage Owing to Lipase & Protease From Psychrotrophic Pseudomonas Species: Characteristics of Enzymatic Breakdown, Improved Assays for Enzyme Detection and Sources of the Bacteria (PhD thesis). University of Melbourne, Melbourne, VIC, Australia. Button, P.D., Roginski, H., Deeth, H.C., Craven, H.M., 2011. Improved shelf-life estimation of UHT milk by prediction of proteolysis. J. Food Qual. 34, 229–235. Cattaneo, S., Masotti, F., Pellegrino, L., 2008. Effects of overprocessing on heat damage of UHT milk. Eur. Food Res. Technol. 226, 1099–1106. Cauvin, E., Sacchi, P., Rasero, R., Turi, R.M., 1999. Proteolytic activity during storage of UHT milk. Ind. Aliment. 38, 825–829. Chaplin, L.C., Lyster, R.L., 1986. Irreversible heat denaturation of bovine α-lactalbumin. J. Dairy Res. 53, 249–258. Corredig, M., Dalgleish, D.G., 1996a. Effect of temperature and pH on the interactions of whey proteins with casein micelles in skim milk. Food Res. Int. 29, 49–55. Corredig, M., Dalgleish, D.G., 1996b. The binding of alpha-lactalbumin and beta-lactoglobulin to casein micelles in milk treated by different heating systems. Milchwissenschaft 51, 123–127. Corredig, M., Dalgleish, D.G., 1999. The mechanisms of the heat-induced interaction of whey proteins with casein micelles in milk. Int. Dairy J. 9, 233–236. Dalabasmaz, S., Ebner, J., Pischetsrieder, M., 2017. Identification of the peptide pyroQ-beta-casein (194-209) as a highly specific and sensitive marker to differentiate between ultrahigh-temperature processed (UHT) milk and mildly heated milk. J. Agric. Food Chem. 65 (49), 10781–10791.

416

10. The effect of UHT processing and storage on milk proteins

Dalgleish, D.G., 1992. Sedimentation of casein micelles during the storage of ultra-high temperature milk products—a calculation. J. Dairy Sci. 75, 371–379. Dannenberg, F., Kessler, H., 1988. Reaction kinetics of the denaturation of whey proteins in milk. J. Food Sci. 53, 258–263. Datta, N., Deeth, H.C., 2001. Age gelation of UHT milk—a review. Food Bioprod. Process. 79, 197–210. Datta, N., Deeth, H.C., 2003. Diagnosing the cause of proteolysis in UHT milk. Lebensm. Wiss. Technol. 36, 173–182. de Jong, P., Bouman, S., van der Linden, H., 1992. Fouling of heat treatment equipment in relation to the denaturation of beta-lactoglobulin. J. Soc. Dairy Technol. 45, 3–8. de Jong, P., Waalewijn, R., Van Der Linden, H., 1993. Validity of a kinetic fouling model for heat-treatment of whole milk. Lait 73, 293–302. de Jong, P., te Giffel, M.C., Straatsma, H., Vissers, M.M.M., 2002. Reduction of fouling and contamination by predictive kinetic models. Int. Dairy J. 12, 285–292. Deeth, H.C., Lewis, M.J., 2016. Protein stability in sterilised milk and milk products. In: McSweeney, P.L.H., O’Mahony, J.A. (Eds.), Advanced Dairy Chemistry, Volume 2, Proteins. Springer, New York, NY, pp. 247–286. Deeth, H.C., Lewis, M.J., 2017. High Temperature Processing of Milk and Milk Products. Wiley Blackwell, Oxford, UK. Deeth, H.C., Smithers, G., 2018. Heat Treatment of Milk—Overview. IDF Factsheet 001/2018/02, International Dairy Federation, Brussels, Belgium. Dehn-M€ uller, B., M€ uller, B., Erbersdobler, H.F., 1991. Studies on protein damage in UHT milk. Milchwissenschaft 46, 431–434. Donovan, M., Mulvihill, D.M., 1987. Thermal-denaturation and aggregation of whey Proteins. Ir. J. Food Sci. Tech. 11, 87–100. Ebner, J., Baum, F., Pischetsrieder, M., 2016. Identification of sixteen peptides reflecting heat and/or storage induced processes by profiling of commercial milk samples. J. Proteome 147, 66–75. Elliott, A.J., Dhakal, A., Datta, N., Deeth, H.C., 2003. Heat-induced changes in UHT milks—part 1. Aust. J. Dairy Technol. 58, 3–10. Elliott, A.J., Datta, N., Amenu, B., Deeth, H.C., 2005. Heat-induced and other chemical changes in commercial UHT milks. J. Dairy Res. 72, 442–446. Fairbairn, D.J., Law, B.A., 1986. Proteinases of psychrotrophic bacteria: their production, properties, effects and control. J. Dairy Res. 53, 139–177. Fenelon, M.A., Hickey, R.M., Buggy, A., McCarthy, N., Murphy, E.G., 2018. Whey proteins in infant formula. In: Deeth, H.C., Bansal, N. (Eds.), Whey Proteins; From Milk to Medicine. Academic Press/Elsevier, London, UK, pp. 439–494. Fennema, O.R., 1996. Food Chemistry, third ed. Marcel Dekker, New York, NY. Fink, R., Kessler, H.G., 1988. Comparison of methods for distinguishing between UHT treatment and sterilisation of milk. Milchwissenschaft 43, 275–279. Fogliano, V., Monti, S.M., Visconti, A., Randazzo, G., Facchiano, A.M., Colonna, G., Ritieni, A., 1998. Identification of a β-lactoglobulin lactosylation site. Biochim. Biophys. Acta 1388, 295–304. Friedman, M., 1999. Chemistry, biochemistry, nutrition, and microbiology of lysinoalanine, lanthionine and histidinoalanine in food and other proteins. J. Agric. Food Chem. 47, 1295–1349. Friedman, M., Zahnley, J., Masters, P., 1981. Relationship between in vitro digestibility of casein and its content of lysinoalanine and D-amino acids. J. Food Sci. 46, 127–131. Fritsch, R.J., Hoffmann, H., Klostermeyer, H., 1983. Formation of lysinoalanine during heat treatment of milk. Z. Lebensm. Unters. Forsch. 176, 341–345. Gaafar, A.M.M., 1987. Investigation Into the Cooked Flavour in Heat-Treated Milk (PhD thesis). University of Reading, Reading, UK. Garcia-Risco, M., Villamiel, M., Lopez-Fandino, R., 2002. Effect of homogenisation on protein distribution and proteolysis during storage of indirectly heated UHT milk. Lait 82 (5), 589–599. Gaucher, I., Molle, D., Gagnaire, V., Gaucheron, F., 2008. Effects of storage temperature on physico-chemical characteristics of semi-skimmed UHT milk. Food Hydrocoll. 22, 130–143. Gaucher, I., Moll
e, D., Gagnaire, V., L
eonil, J., Rousseau, F., Gaucheron, F., 2009. Destabilisation of commercial UHT milks: proteolysis and changes in milk particles. Milchwissenchaft 64, 43–47.

References

417

Gaucher, I., Tanguy, G., Fauquant, J., Jardin, J., Rousseau, F., Robert, B., Madec, M.-N., Gaucheron, F., 2011. Proteolysis of casein micelles by Pseudomonas fluorescens CNRZ 798 contributes to the destabilisation of UHT milk during its storage. Dairy Sci. Technol. 91, 413–429. Gaucheron, F., Le Graet, Y., 2000. Determination of ammonium in milk and dairy products by ion chromatography. J. Chromatogr. A 893, 133–142. Gaur, V., Schalk, J., Anema, S.G., 2018. Sedimentation in UHT milk. Int. Dairy J. 78, 92–102. Gebre-Egziabher, A., Humbert, E.S., Blankenagel, G., 1980. Hydrolysis of milk proteins by microbial enzymes. J. Food Prot. 43, 709–712. Gordon, L., Pilosof, A., 2010. Application of high-intensity ultrasound to control the size of whey protein particles. Food Biophys. 5, 201–210. Grandison, A.S., 1988. UHT processing of milk: seasonal variation in deposit formation in heat exchanger. J. Soc. Dairy Technol. 41, 43–49. Grewal, M.K., Chandrapala, J., Donkor, O., Apostolopoulos, V., Vasiljevic, T., 2017a. Electrophoretic characterization of protein interactions suggesting limited feasibility of accelerated shelf-life testing of ultra-high temperature milk. J. Dairy Sci. 100, 76–88. Grewal, M.K., Chandrapala, J., Donkor, O., Apostolopoulos, V., Stojanovska, L., Vasiljevic, T., 2017b. Fourier transform infrared spectroscopy analysis of physicochemical changes in UHT milk during accelerated storage. Int. Dairy J. 66, 99–107. Grewal, M.K., Chandrapala, J., Donkor, O., Apostolopoulos, V., Vasiljevic, T., 2017c. Predicting sediment formation in ultra-high temperature-treated whole and skim milk using attenuated total reflectance-Fourier transform infrared spectroscopy. Int. Dairy J. 74, 39–48. Griffiths, M.W., Phillips, J.D., Muir, D.D., 1981. Thermostability of proteases and lipases from a number of species of psychrotrophic bacteria of dairy origin. J. Appl. Bacteriol. 50, 289–303. Grijspeerdt, K., Mortier, L., De Block, J., van Renterghem, R., 2004. Applications of modelling to optimise ultra-high temperature milk heat exchangers with respect to fouling. Food Control 15, 117–130. Grufferty, M.B., Fox, P.F., 1988. Factors affecting the release of plasmin activity from casein micelles. N. Z. J. Dairy Sci. Technol. 23, 153–163. Hardham, J.F., 1998. Effect of protein standardisation of milk by addition of UF milk permeate on the composition and storage stability of UHT processed milk. Aust. J. Dairy Technol. 53, 22–27. Harwalkar, V.R., 1992. Age gelation of sterilised milks. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry, Volume 1, Proteins, second ed. Elsevier Applied Science, London, UK, pp. 691–734. Haryani, S., Datta, N., Elliott, A.J., Deeth, H.C., 2003. Production of proteinases by psychrotrophic bacteria in raw milk stored at low temperature. Aust. J. Dairy Technol. 58, 15–20. Havea, P., Singh, H., Creamer, L.K., 2001. Characterization of heat-induced aggregates of β-lactoglobulin, α-lactalbumin and bovine serum albumin in a whey protein concentrate environment. J. Dairy Res. 68, 483–497. Hawran, L.J., Jones, V.A., Swartzel, K.R., 1985. Sediment formation in aseptically processed and packaged milk. J. Food Process. Preserv. 9, 189–207. Hiddink, J., Lalande, M., Maas, A.J.R., Streuper, A., 1986. Heat-treatment of whipping cream. 1. Fouling of the pasteurization equipment. Milchwissenschaft 41, 542–546. Hill, A.R., 1988. Quality of ultra-high-temperature processed milk. Food Technol. 42, 92–97. Hillbrick, G.C., McMahon, D.J., McManus, W.R., 1999. Microstructure of indirectly and directly heated ultra-hightemperature (UHT) processed milk examined using transmission electron microscopy and immunogold labelling. LWT-Food Sci. Technol. 32, 486–494. Holland, J., Gupta, R., Deeth, H.C., Alewood, P.F., 2011. Proteomic analysis of temperature-dependent changes in UHT milk. J. Agric. Food Chem. 59, 1837–1846. Holland, J.W., Gupta, R., Deeth, H.C., Alewood, P.F., 2012. UHT milk contains multiple forms of alpha(S1)-casein that undergo degradative changes during storage. Food Chem. 133, 689–696. Huijs, G., van Asselt, A., Verdurmen, R., de Jong, P., 2004. High speed milk. Dairy Ind. Int. 69 (11), 30–32. Huppertz, T., Vasiljevic, T., Zisu, B., Deeth, H., 2018. Novel processing technologies: effects on whey protein structure and functionality. In: Deeth, H.C., Bansal, N. (Eds.), Whey Proteins; From Milk to Medicine. Academic Press/ Elsevier, London, UK, pp. 281–334. Ismail, B., Nielsen, S.S., 2010. Invited review: plasmin protease in milk: current knowledge and relevance to dairy industry. J. Dairy Sci. 93, 4999–5009.

418

10. The effect of UHT processing and storage on milk proteins

Jansson, T., Jensen, H.B., Sundekilde, U.K., Clausen, M.R., Eggers, N., Larsen, L.B., Ray, C., Andersen, H.J., Bertram, H.C., 2014. Chemical and proteolysis-derived changes during long-term storage of lactose-hydrolyzed ultra-high-temperature (UHT) milk. J. Agric. Food Chem. 62, 11270–11278. Jeurnink, T.J.M., Verheul, M., Cohen Stuart, M.A., de Kruif, C.G., 1996. Relation between the denaturation and fouling behaviour of β-lactoglobulin. In: Heat Treatments and Alternative Methods (Document 9602). International Dairy Federation, Brussels, Belgium, pp. 73–82. Kelly, P., 2018a. Manufacture of whey protein products: concentrates, isolate, whey protein fractions and microparticulated. In: Deeth, H.C., Bansal, N. (Eds.), Whey Proteins; From Milk to Medicine. Academic Press/Elsevier, London, UK, pp. 97–122. Kelly, P., 2018b. Whey protein ingredient applications. In: Deeth, H.C., Bansal, N. (Eds.), Whey Proteins; From Milk to Medicine. Academic Press/Elsevier, London, UK, pp. 335–375. Kelly, A.L., Foley, J., 1997. Proteolysis and storage stability of UHT milk as influenced by milk plasmin activity, plasmin/beta-lactoglobulin complexation, plasminogen activation and somatic cell count. Int. Dairy J. 7, 411–420. Kessler, H.G., 1981. Food Engineering and Dairy Technology. Verlag A. Kessler, Freising, Federal Republic of Germany. Kocak, H.R., Zadow, J.G., 1985a. Age gelation of UHT whole milk as influenced by storage temperature. Aust. J. Dairy Technol. 40, 14–21. Kocak, H.R., Zadow, J.G., 1985b. Controlling age gelation of UHT milk with additives. Aust. J. Dairy Technol. 40, 58–64. Kocak, H.R., Zadow, J.G., 1985c. The effect of low-temperature-inactivation treatment on age gelation of UHT whole milk. Aust. J. Dairy Technol. 40, 53–58. Kohlmann, K.L., Nielsen, S.S., Ladisch, M.R., 1988. Effect of serine proteolytic enzymes (trypsin and plasmin), trypsin inhibitor, and plasminogen activator addition to ultra-high temperature processed milk. J. Dairy Sci. 71, 1728–1739. Kohlmann, K.L., Nielsen, S.S., Ladisch, M.R., 1991. Effects of a low concentration of added plasmin on ultra-high temperature processed milk. J. Dairy Sci. 74, 1151–1156. Kudo, S., 1980. The heat stability of milk: formation of soluble proteins and protein-depleted micelles at elevated temperatures. N. Z. J. Dairy Sci. Technol. 15, 255–263. Lauber, S., Klostermeyer, H., Henle, T., 2001. On the influence of non-enzymatic crosslinking of caseins on the gel strength of yogurt. Nahrung/Food 45, 215–217. Law, B.A., Andrews, A.T., Sharpe, M.E., 1977. Gelation of ultra-high-temperature-sterilized milk by proteases from a strain of Pseudomonas-fluorescens isolated from raw-milk. J. Dairy Res. 44, 145–148. Le, T.T., Holland, J.W., Bhandari, B., Alewood, P.F., Deeth, H.C., 2013. Direct evidence for the role of Maillard reaction products in protein cross-linking in milk powder during storage. Int. Dairy J. 31, 83–91. Le, T.T., Nielsen, S.D., Villumsen, N.S., Kristiansen, G.H., Nielsen, L.R., Nielsen, S.B., Hammershøj, M., Larsen, L.B., 2016. Using proteomics to characterise storage-induced aggregates in acidic whey protein isolate drinks. Int. Dairy J. 60, 39–46. Lee, S.J., Sherbon, J.W., 2002. Chemical changes in bovine milk fat globule membrane caused by heat treatment and homogenization of whole milk. J. Dairy Res. 69, 555–567. Lewis, M.J., Heppell, N., 2000. Continuous Thermal Processing of Foods: Pasteurization and UHT Sterilization. Aspen Publishers, Gaithersburg, MD. Lewis, M., Grandison, A., Lin, M.-J., Tsioulpas, A., 2011. Ionic calcium and pH as predictors of stability of milk to UHT processing. Milchwissenschaft 66, 197–200. Lo´pez Fandin˜o, R., Olano, A., San Joso´, C., Ramos, M., 1993. Application of reversed phase HPLC to the study of proteolysis in UHT milk. J. Dairy Res. 60, 111–116. Lyster, R.L.J., 1970. The denaturation of α-lactalbumin and β-lactoglobulin in heated milk. J. Dairy Res. 37, 233–243. Malmgren, B., Ardo, Y., Langton, M., Altskar, A., Bremer, M., Dejmek, P., Paulsson, M., 2017. Changes in proteins, physical stability and structure in directly heated UHT milk during storage at different temperatures. Int. Dairy J. 71, 60–75. Manji, B., Kakuda, Y., 1988. The role of protein denaturation, extent of proteolysis, and storage temperature on the mechanism of age gelation in a model system. J. Dairy Sci. 71, 1455–1463. Manji, B., Kakuda, Y., Arnott, D.R., 1986. Effect of storage temperature on age gelation of ultra-high temperature milk processed by direct and indirect heating systems. J. Dairy Sci. 69, 2994–3001.

References

419

Mat
eos, A., Guyard-Nicodeme, M., Bagliniere, F., Jardin, J., Gaucheron, F., Dary, A., Humbert, J., Gaillard, J.L., 2015. Proteolysis of milk proteins by AprX, an extracellular protease identified in Pseudomonas LBSA1 isolated from bulk raw milk, and implications for the stability of UHT milk. Int. Dairy J. 49, 78–88. McKenna, B.M., O’Sullivan, A.C., 1971. Whey protein denaturation in concentrated skimmilks. J. Dairy Sci. 54, 1075–1077. McMahon, D.J., 1996. Age-gelation of UHT milk: changes that occur during storage, their effect on shelf-life and the mechanism by which age-gelation occurs. In: Heat Treatments and Alternative Methods (Document 9602). International Dairy Federation, Brussels, Belgium, pp. 315–326. Mehta, B.M., Deeth, H.C., 2016. Blocked lysine in dairy products—formation, occurrence, analysis and nutritional implications. Compr. Rev. Food Sci. Food Saf. 15, 206–218. Miralles, B., Amigo, L., Ramos, M., Recio, I., 2003. Analysing para-kappa-casein and related peptides as indicators of milk proteolysis. Milchwissenschaft 58, 412–415. Mitchell, G.E., Ewings, K.N., 1985. Quantification of bacterial proteolysis causing gelation in UHT-treated milk. N. Z. J. Dairy Sci. Technol. 20, 65–76. Miwa, N., Yokoyama, K., Wakabayashi, H., Nio, N., 2010. Effect of deamidation by protein-glutaminase on physicochemical and functional properties of skim milk. Int. Dairy J. 20, 393–399. Morales, F.J., Romero, C., Jimenez-Perez, S., 2000. Characterization of industrial processed milk by analysis of heatinduced changes. Int. J. Food Sci. Technol. 35, 193–200. Mottar, J., Bassier, A., Joniau, M., Baert, J., 1989. Effect of heat-induced association of whey proteins and casein micelles on yogurt texture. J. Dairy Sci. 72, 2247–2256. Mounsey, J.S., O’Kennedy, B.T., 2010. Heat-stabilisation of β-lactoglobulin through interaction with sodium caseinate. Milchwissenschaft 65 (1), 79–83. Mounsey, J.S., O’Kennedy, B.T., 2007. Conditions limiting the influence of thiol-disulphide interchange reactions on the heat-induced aggregation kinetics of beta-lactoglobulin. Int. Dairy J. 17, 1034–1042. Newstead, D.F., Paterson, G., Anema, S.G., Coker, C.J., Wewala, A.R., 2006. Plasmin activity in direct-steam-injection UHT-processed reconstituted milk: effects of preheat treatment. Int. Dairy J. 16, 573–579. Nicodeme, M., Grill, J.P., Humbert, G., Gaillard, J.L., 2005. Extracellular protease activity of different Pseudomonas strains: dependence of proteolytic activity on culture conditions. J. Appl. Microbiol. 99, 641–648. Nieuwenhuijse, J.A., van Boekel, M.A.J.S., 2003. Protein stability in sterilised milk and milk products. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry, Volume 1, Proteins, third ed. Kluwer Academic/Plenum Press, New York, NY, pp. 947–974. O’Brien, J., 1995. Heat-induced changes in lactose: isomerisation, degradation, Maillard browning. In: Fox, P.F. (Ed.), Heat-Induced Changes in Milk (Document 9501). International Dairy Federation, Brussels, Belgium, pp. 134–170. O’Kennedy, B.T., Mounsey, J.S., 2006. Control of heat-induced aggregation of whey proteins using casein. J. Agric. Food Chem. 54, 5637–5642. Oldfield, D.J., Singh, H., Taylor, M.W., 1998. Association of β-lactoglobulin and α-lactalbumin with the casein micelles in skim milk heated in an ultra-high temperature plant. Int. Dairy J. 8, 765–770. Oliveira, L.N., Marinho, V.T., Zamagno, M.V., Lauro, M.A., Barboza, J.A.N., da Silva, P.H.F., 2015. Assessment of whey protein nitrogen index as an indicator of heat treatment for UHT milk and milk powder. J. Candido Tostes Dairy Inst. 70, 78–84. On-Nom, N., Grandison, A.S., Lewis, M.J., 2012. Heat stability of milk supplemented with calcium chloride. J. Dairy Sci. 95, 1623–1631. Pagliarini, E., Vernille, M., Peri, C., 1990. Kinetic study on colour changes in milk due to heat. J. Food Sci. 55, 1766–1767. Perkin, A.G., Henschel, M.J., Burton, H., 1973. Comparison of milks processed by the direct and indirect methods of ultra-high-temperature sterilization. VI. Effects on sediment formation and clotting with enzymes. J. Dairy Res. 40, 215–220. Pinto, C.L.D., Machado, S.G., Cardoso, R.R., Vanetti, M.C.D., 2016. Sedimentation, proteolytic activity and proteolysis of whole UHT milk during storage. J. Candido Tostes Dairy Inst. 71, 197–205. Pizzano, R., Manzo, C., Nicolai, M.A., Addeo, F., 2012. Occurrence of major whey proteins in the pH 4.6 insoluble protein fraction from UHT-treated milk. J. Agric. Food Chem. 60, 8044–8050. Politis, I., Lachance, E., Block, E., Turner, J.D., 1989. Plasmin and plasminogen in bovine milk: a relationship with involution. J. Dairy Sci. 72, 900–926.

420

10. The effect of UHT processing and storage on milk proteins

Prado, B.M., Ismail, B., Ramos, O., Hayes, K.D., 2007. Thermal stability of plasminogen activators and plasminogen activation in heated milk. Int. Dairy J. 17, 1028–1033. Prakash, S., Datta, N., Lewis, M.J., Deeth, H.C., 2007. Reducing fouling during UHT treatment of goat’s milk. Milchwissenschaft 62, 16–19. Prakash, S., Kravchuk, O., Deeth, H.C., 2015. Influence of pre-heat temperature, pre-heat holding time and high-heat temperature on fouling of reconstituted skim milk during UHT processing. J. Food Eng. 153, 45–52. Ramsey, J.A., Swartzel, K.R., 1984. Effect of ultra-high temperature processing and storage conditions on rates of sedimentation and fat separation of aseptically packaged milk. J. Food Sci. 49 (1), 257–262. Rauh, V.M., Johansen, L.B., Ipsen, R., Paulsson, M., Larsen, L.B., Hammershoj, M., 2014a. Plasmin activity in UHT milk: relationship between proteolysis, age gelation, and bitterness. J. Agric. Food Chem. 62, 6852–6860. Rauh, V.M., Sundgren, A., Bakman, M., Ipsen, R., Paulsson, M., Larsen, L.B., Hammershoj, M., 2014b. Plasmin activity as a possible cause for age gelation in UHT milk produced by direct steam infusion. Int. Dairy J. 38, 199–207. Recio, I., de Frutos, M., Olano, A., Ramos, M., 1996. Protein changes in stored ultra-high-temperature-treated milks studied by capillary electrophoresis and high-performance liquid chromatography. J. Agric. Food Chem. 44, 3955–3959. Reddy, K.K., Nguyen, M.H., Kailasapathy, K., Zadow, J.G., 1991. The effects of some treatments and storage temperatures on UHT whole milk. Aust. J. Dairy Technol. 46, 57–63. Reddy, G.K., Nguyen, M.H., Kailasapathy, K., Zadow, J.G., Hardham, J.F., 1999. Kinetic study of whey protein denaturation to assess the degree of heat treatment in UHT milk. J. Food Sci. Tech. Mys. 36, 305–309. Resmini, P.L., Pellegrino, L., Hogenboom, J.A., Andreini, R., 1989. Thermal denaturation of whey protein in pasteurized milk. Fast evaluation by HPLC. Ital. J. Food Sci. 3, 51–62. Richardson, B.C., 1983. The proteases of bovine milk and the effect of pasteurisation on their activity. N. Z. J. Dairy Sci. Technol. 18, 233–245. Richardson, B.C., Newstead, D.F., 1979. Effect of heat-stable proteases on the storage life of UHT milk. N. Z. J. Dairy Sci. Technol. 14, 273–279. Robertson, G., 2002. The paper beverage carton: past and future. Food Technol. 56 (7), 46–48. 50–52. Robertson, G.L., 2011. Heat treatment of milk: ultra-high temperature treatment (UHT): aseptic packaging. In: Fuquay, J.W., Fox, P.F., McSweeney, P.L.H. (Eds.), Encyclopedia of Dairy Sciences, second ed. In: vol. 2. Academic Press, San Diego, CA, pp. 708–713. Robertson, G.L., 2013. Aseptic packaging of foods. In: Food Packaging: Principles and Practice, third ed. CRC Press, Boca Raton, FL, pp. 367–382. Sanderson, W.B., 1970. Determination of undenatured whey protein nitrogen in skim milk powder by dye binding. N. Z. J. Dairy Sci. Technol. 5, 46–48. Schokker, E.P., van Boekel, M., 1999. Influence of pH on low-temperature inactivation of the extracellular proteinase from Pseudomonas fluorescens 22F. Milchwissenschaft 54, 377–379. Schutyser, M.A.I., Straatsma, J., Keijzer, P.M., Verschueren, M., de Jong, P., 2008. A new web-based modelling tool (Websim-MILQ) aimed at optimisation of thermal treatments in the dairy industry. Int. J. Food Microbiol. 128, 153–157. Sharma, D.K., Prasad, D.N., 1990. Changes in the physical properties of high temperature processed buffalo milk during storage. J. Dairy Res. 57, 187–196. Siciliano, R.A., Mazzeo, M.F., Arena, S., Renzone, G., Scaloni, A., 2013. Mass spectrometry for the analysis of protein lactosylation in milk products. Food Res. Int. 54, 988–1000. Singh, H., 1991. Modification of food proteins by covalent crosslinking. Trends Food Sci. Technol. 2, 196–200. Skudder, P.J., Thomas, E.L., Pavey, J.A., Perkin, A.G., 1981. Effects of adding potassium iodate to milk before UHT treatment. J. Dairy Res. 48, 99–113. Smet, K., De Block, J., De Campeneere, S., De Brabander, D., Herman, L., Raes, K., Dewettinck, K., Coudijzer, K., 2009. Oxidative stability of UHT milk as influenced by fatty acid composition and packaging. Int. Dairy J. 19, 372–379. Snoeren, T.H.M., Both, P., 1981. Proteolysis during storage of UHT-sterilised whole milk. 2. Experiments with milk heated by the direct system for 4 seconds at 142°C. Neth. Milk Dairy J. 35, 113–119. Snoeren, T.H.M., van Riel, J.A.M., Both, P., 1979. Proteolysis during storage of UHT-sterilised whole milk. 1. Experiments with milk heated by the indirect system for 4 seconds at 142°C. Neth. Milk Dairy J. 33, 31–39. Sørhaug, T., Stepaniak, L., 1997. Psychrotrophs and their enzymes in milk and dairy products: quality aspects. Trends Food Sci. Technol. 8, 35–40.

References

421

SPX, 2013. Long Life Dairy, Food and Beverage Products. https://www.spxflow.com/en/assets/pdf/Long_Life_ Dairy_Food_22000_06_01_2013_GB_tcm11-7664.pdf. (Accessed 13 December 2018). Stoeckel, M., Lidolt, M., Achberger, V., Gluck, C., Krewinkel, M., Stressler, T., von Neubeck, M., Wenning, M., Scherer, S., Fischer, L., Hinrichs, J., 2016. Growth of Pseudomonas weihenstephanensis, Pseudomonas proteolytica and Pseudomonas sp in raw milk: impact of residual heat-stable enzyme activity on stability of UHT milk during shelf-life. Int. Dairy J. 59, 20–28. Swartzel, K.R., 1983. The role of heat exchanger fouling in the formation of sediment in aseptically processed and packaged milk. J. Food Process. Preserv. 7, 247–257. Timmer-Keetels, C.J., Nieuwenhuijse, J.A., Zijtveld-Van Der Wiel, J., Nieuwenhuijse, H.A., Zijtveld-Van Der Wie, J.H., Keetels, C.J.A., Van Der Wiel, J.H., Zijtveld-Van, D.W.J., 2011. Producing Food Product Having Milk Protein Involves Subjecting Milk Protein to Enzymatic Deamidation Procedure, Formulating Protein Into Liquid Product, Followed by Heat Sterilisation of Food Product and Spray Drying of Product Into Powder. Patents: WO2011034418-A2; WO2011034418-A3; NL2003494-C; WO2011034418-A4; EP2477506-A2; CN102595925-A; US2012231117-A1. Topc¸u, A., Numanoglu, E., Saldamli, I., 2006. Proteolysis and storage stability of UHT milk produced in Turkey. Int. Dairy J. 16, 633–638. Tossavainen, O., Kallioinen, H., 2007. Proteolytic changes in lactose hydrolysed UHT milks during storage. Milchwissenschaft 62, 410–415. Tran, H., Datta, N., Lewis, M.J., Deeth, H.C., 2008. Processing parameters and predicted product properties of industrial UHT milk processing plants in Australia. Int. Dairy J. 18, 939–944. Vaghela, K.D., Chaudhary, B.N., Mehta, B.M., 2018. Review on proteolysis rate in UHT milk: its mechanism, pattern, assessment and enzymatic changes during storage. Res. Rev. J. Dairy Sci. Technol. 6 (3), 1–16. van Asselt, A.J., Sweere, A.P.J., Rollema, H.S., de Jong, P., 2008. Extreme high-temperature treatment of milk with respect to plasmin inactivation. Int. Dairy J. 18, 531–538. van Boekel, M.A.J.S., 1998. Effect of heating on Maillard reactions in milk. Food Chem. 62, 403–414. van Boekel, M.A.J.S., 1999. Heat-induced deamidation, dephosphorylation and breakdown of caseinate. Int. Dairy J. 9, 237–241. Vesconsi, C.N., Valduga, A.T., Cichoski, A.J., 2012. Particle sedimentation in semi-skimmed, skimmed on whole milk UHT, during storage. Cienc. Rural 42, 730–736. Villumsen, N.S., Hammershoj, M., Nielsen, L.R., Poulsen, K.R., Sorensen, J., Larsen, L.B., 2015a. Control of heat treatment and storage temperature prevents the formation of visible aggregates in acidic whey dispersions over a 6-month storage period. LWT-Food Sci. Technol. 64, 164–170. Villumsen, N.S., Jensen, H.B., Le, T.T.T., Moller, H.S., Nordvang, R.T., Nielsen, L.R., Nielsen, S.B., Sorensen, J., Hammershoj, M., Larsen, L.B., 2015b. Self-assembly of caseinomacropeptide as a potential key mechanism in the formation of visible storage induced aggregates in acidic whey protein isolate dispersions. Int. Dairy J. 49, 8–15. Walstra, P., Jenness, R., 1984. Dairy Chemistry and Physics. John Wiley and Sons, New York, NY, pp. 173–177. Wang, H.-Y., Qian, H., Yao, W.-R., 2011. Melanoidins produced by the Maillard reaction: structure and biological activity. Food Chem. 128, 573–584. Wijayanti, H.B., Bansal, N., Deeth, H.C., 2014. Stability of whey proteins during thermal processing: a review. Compr. Rev. Food Sci. Food Saf. 13, 1235–1251. Wijayanti, H.B., Brodkorb, A., Hogan, S.A., Murphy, E.G., 2018. Thermal denaturation, aggregation, and methods of prevention. In: Deeth, H.C., Bansal, N. (Eds.), Whey Proteins; From Milk to Medicine. Academic Press/Elsevier, London, UK, pp. 185–247. Zabbia, A., Buys, E.M., De Kock, H.L., 2012. Undesirable sulphur and carbonyl flavor compounds in UHT milk: a review. Crit. Rev. Food Sci. Nutr. 52, 21–30. Zhang, S., Jin, Y., Yun, Z., 2007. Effect of low temperature heat treatment on the activity of lipase VIII from Pseudomonas species in UHT milk. Trans. Chin. Soc. Agric. Eng. 23, 231–235. Zhang, C.Y., Bijl, E., Hettinga, K., 2018. Destabilization of UHT milk by protease AprX from Pseudomonas fluorescens and plasmin. Food Chem. 263, 127–134. Zin El-Din, M., Aoki, T., Kako, Y., 1991. Polymerization and degradation of casein in UHT milk during storage. Milchwissenschaft 46, 284–287.

This page intentionally left blank

C H A P T E R

11 Effects of drying and storage on milk proteins Alan Baldwina, Kerianne Higgsa, Mike Bolandb, Pierre Schuckc a

Formerly Fonterra Research Centre, Palmerston North, New Zealand bRiddet Institute, Massey University, Palmerston North, New Zealand cINRA, Rennes, France

Introduction Milk is an unstable foodstuff, prone in particular not only to microbiological degradation but also to long-term chemical change. A large part of the world’s dairy production occurs in areas remote from the markets in which it is consumed, and production is often seasonal, requiring storage to smooth out supply. The production of milk powders and other dried milk protein-containing products has been the method of choice for over 100 years for the storage and shipping of milk over long distances and/or times, as it confers stability and massively reduces weight and bulk. Milk powders were known to the Chinese and were described by Marco Polo. The production of milk powders was described by Nicolas Appert in the early 19th century, and commercial processes for the spray drying of milk were patented in the United States in 1872 and 1905. This opened the way for large-scale industrial production of milk powders throughout the 20th century. The purpose of the dehydration of milk and whey is to stabilize these products for their storage and later use. Milk and whey powders in Europe were used mostly in animal feed. With changes in agricultural policies (such as the implementation of the quota system and the dissolution of the price support system in the European Union), the dairy industry was forced to look for better uses for the dairy surplus and for the by-products of cheese (whey) produced from milk and buttermilk produced from cream. Studies on the reuse of protein fractions with their nutritional qualities and functional properties led to a range of applications in food ingredients. With the emergence of filtration technology (microfiltration, ultrafiltration, nanofiltration, and reverse osmosis), in the past 50 years, the dairy industry has developed new technological processes for extracting and purifying proteins (casein, caseinates, whey proteins, etc.) (Kjaergaard et al., 1987; Maubois, 1991). Such products include

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00011-6

423

# 2020 Elsevier Inc. All rights reserved.

424

11. Effects of drying and storage on milk proteins

• dairy proteins and whey concentrates (Le Grae¨t and Maubois, 1979; Goud
edranche et al., 1980; Madsen and Bjerre, 1981; Maubois et al., 1987; Caron et al., 1997); • micellar casein concentrates (Fauquant et al., 1988; Schuck et al., 1994a); • micellar casein (MC) (Pierre et al., 1992; Schuck et al., 1994b); • whey concentrates, selectively demineralized concentrates ( Jeantet et al., 1996); • super clean skim milk concentrates (Piot et al., 1987; Vincens and Tabard, 1988; Trouv
e et al., 1991; Schuck et al., 1994a). Used as either nutritional or functional ingredients, most of these proteins are marketed in dehydrated form (Fig. 11.1). The application of different processing steps allows the production of a wide range of different dried and stable intermediate dairy products. Many new uses for these constituents have emerged with the manufacture of formula products, substitutes, and adapted raw materials. The most frequently used technique for the dehydration of dairy products is spray drying. A number of spray-drying plants were installed from the 1930s onward, and much was learnt about operating conditions by practical experience and product investigations (Hall and Hedrick, 1971). Milk powder became a popular product in the dairy industry in the 1970s, which attracted increased scientific interest in spray drying particularly into the effects of

Cream

Cream separation

Standardization Whole milk

Standardized milk

Skim milk

VE VE

+ Fat (option)

MF / DF

[Milk] MCC/I VE Milk powder

Casein

Caseinate Spray drying Caseinate powder

MPC / I

[MCC / I]

VE + C°

MPC / I powder

UF / DF

Demineralized whey / Mfate

Spray drying

VE + C° [Demineralized whey / MFat]

Demineralized whey / MFate powder

Fractionation of milk.

Ultrafiltrate*

WPC / I VE [WPC / I]

C° [Ultrafiltrate]

Spray drying

Spray drying

FIG. 11.1

Spray drying

Spray drying Buttermilk powder

Legend

[Whey] / [MFate]

Whey / MFate powder

[MPC / I]

MCC / I powder NF, ED, IE

[Buttermilk]

VE

Spray drying

Whey / microfiltrate

+ Alcalis

Buttermilk VE

Spray drying

Coagulation

Butter

UF / DF

WPC / I powder

Ultrafiltrate powder

[ ]: Concentrate DF: Diafiltration ED: Electrodialysis I: Isolate IE: Ions Exchange MCC: Micellar casein concentrate MF: Microfiltration 0.1 µm MFate: Microfiltrat MPC: Milk protein concentrate NF: Nanofiltration UF: Ultrafiltration VE: Vacuum evaporation WPC: Whey protein concentrate *: Std°with Milk UF / MF (not Whey) or Lactose

425

World dairy powder situation

spray-drying parameters on the physicochemical composition and on the microbiology of the concentrates on powder quality. Because of the variety and complexity of the mixes to be dried, a more rigorous method based on physicochemical and thermodynamic properties has become necessary. Greater understanding of the biochemical properties of milk products before drying, water transfer during spray drying, and the properties of powders and influencing factors is now essential in the production of milk powders. A lack of technical and economic information and of scientific methods prevents the manufacturer from optimizing powder plant in terms of energy costs and powder quality.

World dairy powder situation Because dairy protein products are widely produced and traded, it is pertinent to review the situation with respect to dairy powders (see also Chapter 1; for trends in previous decades, refer to the first and second editions). Table 11.1 summarizes the biggest exporters and importers of milk powders in 2017. In the same period, about 849 million tons of milk powder was produced globally (IDF, 2018). In addition to milk powders, dried dairy protein products that are traded on the world market, largely as food ingredients, include casein and caseinate, whey powders, whey protein concentrates, whey protein isolates, milk protein concentrates, milk protein isolates, and specialist nutritional powders and blends, which may also contain hydrolyzed dairy proteins. These products, especially those from New Zealand and Australia, often have long storage times because of geographic distance to market and seasonal production. Research from our laboratories and that of others has shown that a range of reactions can occur in the dry powders and that these are known to affect powder functionality and nutritional value. The quality of dried milk protein products deteriorates on storage at ambient temperatures mainly because of two reactions: the Maillard reaction and isopeptide bond formation. There are also some minor reactions, which are covered later.

TABLE 11.1

Top six exporters and importers of milk powders in 2017 (000 tons)

Export

New Zealand

European Union

USA

Australia

Belarus

Uruguay

SMP

404

781

606

158

114

WMP

1351

394

27

55

29

108

Total

1755

1175

633

213

143

108

Import

China

Algeria

Mexico

Indonesia

Philippines

Russia

SMP

247

161

331

162

154

123

WMP

470

262

52

23

49

Total

717

423

214

177

172

331

Source: International Dairy Federation, 2018. The World Dairy Situation, 2018 (Bulletin of the International Dairy Federation, 494/2018). International Dairy Federation, Brussels.

426

11. Effects of drying and storage on milk proteins

Standard abbreviations for many of the dried milk products are used throughout the world and are used in this chapter: SMP, skim milk powder; WMP, whole milk powder; WPC, whey protein concentrate; WPI, whey protein isolate (usually >85% protein); MPC, milk protein concentrate; and MPI, milk protein isolate. Milk powders sold internationally must conform to standard protein levels and fat levels for WMP. A number following the abbreviation for the other products is the percentage of protein by weight in the dry powder. Caseins and caseinates are not usually abbreviated and are typically >90% protein by dry weight. Milk powders are used primarily for making reconstituted and recombined milks, usually sold to consumers in UHT format, although substantial amounts are used to make other dairy products such as yoghurt and ice cream as well as being minor ingredients in a wide range of nondairy foods. The main uses for dried milk protein products are nutritional, and products include infant formulas, medical foods, dietary foods for the elderly, specialist foods for weight management, and foods for muscle building (where high protein and high levels of branched-chain amino acids are desirable). They are also used in nonnutritional applications including desserts, confectionery, toppings, imitation cheeses, sauces and dressings, and coffee whiteners, where their functional properties are important.

Whole milk powder (WMP) According to the International Dairy Federation (2018), world production of WMP is estimated at around 4.8 million tons with New Zealand and China together producing about half this total and the EU a further 18% and New Zealand exporting almost all of its production and accounting for 55% of all WMP exports. The main markets for WMP in 2016 were China, Algeria, and Saudi Arabia, but these account for only about 35% of all importers, with wide distribution of importing countries. The main trade flows for WMP are shown in Fig. 11.2. For further details of production trends, refer to International Dairy Federation (2018).

Skim milk powder (SMP) World production of SMP is estimated at 4.9 million tons (International Dairy Federation, 2018) with the EU producing one-third of this and significant production by the United States, India, New Zealand, Australia, and Brazil. The principal exporters were the EU, the United States, New Zealand, and Australia. World trade in SMP in 2017 was 2.4 million tons, with the EU, the United States, and New Zealand accounting for 75% of exports. The principal importers were Mexico and China, accounting for 25% of imports. The main trade flows for SMP are shown in Fig. 11.3.

Whey products and casein Liquid whey production results mainly from the production of cheese, which generates more than 80% of the total whey available and secondarily from casein production. The major processors of whey are therefore located in Europe, North America, and the Oceania, corresponding to the major cheese production areas.

World dairy powder situation

FIG. 11.2 Trade flows of WMP. Adapted from figures supplied by the International Dairy Federation.

FIG. 11.3 Trade flows of SMP. Adapted from figures supplied by the International Dairy Federation.

427

428

11. Effects of drying and storage on milk proteins

Production of whey powders in 2017 is estimated at 3 million tons, a figure that has been fairly constant for several years. The major producers are the EU, accounting for two-thirds of this production, and the United States (International Dairy Federation, 2018). In 2011, casein production in the EU was estimated at around 145,000 tons (International Dairy Federation, 2012); more recent data and data for the rest of the world are not regularly reported. There is also a range of dairy protein mixes and blends that are produced and traded, including, notably, infant formula mixes. While this may be substantial, there is a dearth of reliable data. FAOSTAT (http://www.fao.org/faostat/en/#data) reports exports of about 1.3 million tons of infant food in 2016, but it is unclear how much of this is based on milk proteins.

Properties of spray-dried milk products A dairy powder is characterized not only by its composition (proteins, carbohydrates, fats, minerals, and water) but also by its microbiological and physical properties (bulk and particle density, instant characteristics, flowability, floodability, hygroscopicity, degree of caking, whey protein nitrogen index, thermostability, insolubility index, dispersibility index, wettability index, sinkability index, free fat, occluded air, interstitial air, and particle size), which form the basic elements of quality specifications. There are well-defined test methods for their determination according to international standards (Pisecky, 1986, 1990, 1997; American Dairy Products Institute, 1990; Masters, 1991). These characteristics depend on drying parameters (type of spray-drier tower; operation of atomizers, either nozzles or disc; fine return agglomeration; and thermodynamic conditions of the air, such as temperature, relative humidity, and velocity) and the characteristics of the concentrate before spraying (heat treatment, composition/physicochemical characteristics, viscosity, thermosensitivity, and availability of water). Several scientific papers have been published on the effects of technological parameters on these properties (Hall and Hedrick, 1966; De Vilder et al., 1979; Baldwin et al., 1980; Pisecky, 1980, 1981, 1986; Kessler, 1981; Bloore and Boag, 1982; De Vilder, 1986; Tuohy, 1989; Ilari and Loisel, 1991; Masters, 1991; Mahaut et al., 2000). Water content, water dynamics, and water availability are among the most important properties of spray-dried milk products (Fig. 11.4).

Principles of spray drying According to Pisecky (1997), spray drying is an industrial process for the dehydration of a liquid containing dissolved and/or dispersed solids (e.g., dairy products) by transforming the liquid into a spray of small droplets and exposing these droplets to a flow of hot air. The very large surface area of the spray droplets causes evaporation of the water to take place very quickly, converting the droplets into dry powder particles.

429

Principles of spray drying

Properties of concentrate Physical, biochemical, and microbiological Drying conditions Parameters, type of spray drying installations

Powder

Properties/qualities

Storage conditions

Physical biochemical microbiological

Water transfer

Rehydration conditions FIG. 11.4 Properties and qualities of powders.

Indeed, when a wet droplet is exposed to hot dry gas, variations in the temperature and the partial pressure of the water vapor are established spontaneously between the droplet and the air: • Heat transfer from the air to the droplet occurs under the influence of the temperature difference. • Water transfer occurs in the opposite direction, explained by the difference in the partial pressure of water vapor between the air and the droplet surface. Air is thus used both for fluid heating and as a carrier gas for the removal of water. The air enters the spray drier hot and dry and leaves wet and cool. Spray drying is a phenomenon of surface water evaporation maintained by the movement of capillary water from the interior to the surface of the droplet. As long as the average moisture is sufficient to feed the surface regularly, the evaporation rate is constant. If not, it decreases. The drying kinetics are related to three factors: • Evaporation surface created by the diameter of the particles. Spraying increases the exchange surface: 1 L of liquid sprayed as particles of 100-μm diameter develops a surface area of 60 m2, whereas the surface area is only approximately 0.05 m2 for one sphere of the same volume. A consequence is that large particles dry more slowly; drying time is proportional to the square of the particle diameter. • Difference in the partial pressure of water vapor between the particle and the drying air. A decrease in the absolute humidity of the air and/or an increase in the air temperature tend to increase the difference in the partial pressure of water vapor between the particle and the drying air. • Rate of water migration from the center of the particle toward its surface. This parameter is essential for the quality of dairy powders. Indeed, it is important that there is always water on the surface of the product so that the powder surface remains at the wet bulb

430

11. Effects of drying and storage on milk proteins

17 4

5

1 Concentrate

2

Outlet air

15 3

10

9

8

Legend:

7

1- Feed pump

11 17 4

5

2- Feed Flow

12

16

3- Sprayer/disperser

16

15

5- Air heater

16

Inlet air

6- Air cooler 7- Drying chamber

14 17 13

4- Inlet air fan

8- Primary cyclone

5

9- Secondary cyclone

Powder

10- Outlet air fan 11- Integrated fluid bed

17 13 6

5

12- Pressure conveyer system 13- Vibro-fluidizer air fan 14- Vibro-fluidizer 15- Reincorporating fines 16- Rotary valve 17- Air filtration

FIG. 11.5

Multiple effect spray drier.

temperature for as long as possible. The rate of water migration depends on the water diffusion coefficient, which varies according to the biochemical composition, water content, and droplet temperature. Calculation of this coefficient is therefore complex, and the mathematical models suggested are not easily exploitable by the dairy industry. To define the components of a spray-drying installation according to Masters (1991) and Pisecky (1997), the main components of the spray drier shown in Fig. 11.5 are as follows: • A drying chamber (Fig. 11.5, point 7). The chamber can be horizontal (box drier), although in the dairy industry, the chamber design is generally vertical with a conical or flat base. • An air disperser with a hot air supply system such as a main air filter, supply fan, air heater, and air disperser (Fig. 11.5, point 3). The air aspiration is performed through filters, the type depending on the local conditions, and the nature of the product to be treated. The air can be heated in two different ways: by direct heating (gas) and/or by indirect heating (vapor, gas, oil, or electricity). The air flow in the chamber can be in cocurrent, countercurrent, or mixed mode. • An atomizing device with a feed supply system such as feed tank, feed pump, water tank, concentrate heater, and atomizing device. There are three types of atomizing device: rotary atomizer (wheel or disc), nozzle atomizer (pressure, pneumatic, or sonic), and combined (rotary and pneumatic) (Fig. 11.5, point 3).

Principles of spray drying

431

• A powder recovery system. Separation of the dried product can be achieved by a primary discharge from the drying chamber followed by a secondary discharge from a particulate collector (using a cyclone, bag filter, or electrostatic precipitation), followed by total discharge from the particulate collector and finishing with final exhaust air cleaning in a wet scrubber and dry filter (Fig. 11.5, points 8 and 9). According to Sougnez (1983), Masters (1991), and Pisecky (1997), the simplest types of installation are single-stage systems with a very short residence time (20–60 s). Thus, there is no real balance between the relative humidity of the air and the moisture content of the powder. The outlet temperature of the air must therefore be higher, and the thermal efficiency of the single-stage spray drier is then reduced. This type of drying chamber was the standard equipment for drying milk in the 1960s. Space requirements were small, and building costs were low. Generally, installations without any posttreatment system are suitable only for nonagglomerated powders not requiring cooling. If necessary, a pneumatic conveying system could be added to cool the powder while transporting the chamber fraction and the cyclone fraction to a single discharge point. The two-stage drying system consists of limiting the spray-drying process to a process with a longer residence time (several minutes) to provide a better thermodynamic balance. This involves a considerable reduction in the outlet air temperature and also allows an increase in the inlet air temperature. A second final drying stage is necessary to optimize the moisture content by using an integrated fluid bed (static) or an external fluid bed (vibrating), the air temperatures of which are 15–25°C lower than with a single-stage system to improve and/ or preserve the quality of the dairy powder (Fig. 11.5, points 11 and 14). Consequently, the surrounding air temperature in the chamber at the critical drying stage and the particle temperature are also correspondingly lower, thus contributing to further economic improvement. The integrated fluid bed can be either circular [e.g., multi stage drier (MSD) chamber] or annular [e.g., compact drier (CD) chamber]. Two-stage drying has its limitations, but it can be applied to products such as skim milk, whole milk, precrystallized whey, caseinates, whey proteins, and derivatives. The moisture content of the powder leaving the first stage is limited by the thermoplasticity of the wet powder, that is, by its stickiness in relation to the aw and the glass transition temperature (Roos, 2002). The moisture content must be close to 7%–8%, 9%–10%, and 2%–3% for skim or whole milk powders, caseinate or whey protein powders, and precrystallized whey powders, respectively. The two-stage drying techniques can be applied to the production of both nonagglomerated and agglomerated powders, but this technique is very suitable for the production of agglomerated powders, by separating the nonagglomerated particles from the agglomerates [i.e., collecting the cyclone fractions and reintroducing these fine fractions (called fines) into the wet zone around the atomizer of the chamber]. The three-stage drying systems, with an internal fluid bed as a second stage in combination with an external vibrating fluid bed as a third-stage drier, first appeared at the beginning of the 1980s and were called compact drier instantization (CDI) or MSD. Today, they dominate the dairy powder industry (Fig. 11.5). Three-stage systems combine all the advantages of extended two-stage drying, using spray drying as the primary stage, fluid bed drying of a static fluid as the second drying stage, and drying on an external vibrating fluid bed as the third drying stage. The final drying stage terminates with cooling to under the glass transition

432

11. Effects of drying and storage on milk proteins

temperature. Evaporation performed at each stage can be optimized to achieve both gentle drying conditions and good thermal economy. The compact drier (CD) is suitable for producing both nonagglomerated and agglomerated powders of practically any kind of dried dairy product. It can also cope successfully with whey powders, fat-filled milk, and whey products as well as caseinates, both nonagglomerated and agglomerated. It has a fat content limit of about 50% fat in total solids. Powder quality and appearance are comparable with those of products from two-stage drying systems, but they have considerably better flowability, and the process is more economical. In comparison with the CD, the MSD can process an even wider range of products and can handle an even higher fat content. The main characteristics of MSD powder are the very good agglomeration, mechanical stability, and low particle size fractions (below 125 μm). Optimization of the process has allowed considerable improvement in the drying efficiency, and the quality of the product obtained is generally better. The various advantages are: • improved thermal efficiency: significant reduction in the outlet air temperature, permitting an increase in the inlet air temperature; • reduction in material obstruction: the capacity in one volume is two or three times higher than for a traditional unit; • considerable reduction in powder emission to the atmosphere: a reduction in the drying air flow and an increase in powder moisture content reduce the loss of fine particles in the outlet air; • improved powder quality in relation to the agglomeration level, solubility, dispersibility, wettability, particle size, density, etc. There are other examples of drying equipment such as the “tall-form drier,” the “Filtermat drier”, the “Paraflash drier,” and the “Tixotherm drier.” All these towers have characteristics related to the specific properties of the product being dried (e.g., high fat content, starch, maltodextrin, egg, and hygroscopic products).

Process improvement This section describes the use of a thermohygrometric sensor with some examples of such measurements [temperature, absolute humidity (AH) and relative humidity (RH), and dry air flow rate, aw], through calculation of mass and absolute humidity to prevent sticking in the drying chamber and to optimize powder moisture and aw in relation to the relative humidity of the outlet air. It was demonstrated by Schuck et al. (2005) that a thermohygrometer can be used to avoid sticking and to optimize water content and aw in dairy powders. These results demonstrate that the calculated AH is systematically higher than the measured AH, because the calculated AH corresponds to the maximum theoretical value that can be reached. Calculation of AH by means of the mass balance is based on the hypothesis that the air circulating in the spray drier removes all the water from the concentrate. Thus, if the difference between the calculated and the measured absolute humidity of the outlet air is below 2-g water kg
1 dry air (depending on the spray drier with regard to measurement accuracy), there is no problem of sticking in

Drying of proteins

433

the spray-drier chambers, whatever the dairy concentrate used. On the other hand, sticking was observed in this study for differential AH above 2-g water kg
1 dry air, corresponding to lower water removal and consequently to favorable sticking conditions. The operator can follow the absolute humidity and anticipate a variation in drying parameters according to the differences between the calculated and the measured absolute humidity. The operator can also follow the relative humidity in the outlet air. To achieve a dairy powder with the same aw and moisture content, the same relative humidity in the outlet air using the previous equations according to each dairy product, whatever the spray-drying conditions (inlet air temperature and relative and absolute humidity) must always be maintained. The changes in relative and absolute humidity (resulting from variations in absolute humidity of inlet air, total solid content of concentrate, crystallization rate, outlet air temperature, etc.) can be rapidly observed in the outlet air using a thermohygrometer before such changes significantly affect powder moisture, aw, and powder behavior with regard to sticking.

Drying of proteins The native properties of milk components are substantially unaffected by moderate drying conditions. Depending on the preheating conditions, drier design, and temperature of operation, the properties of spray-dried powder may vary significantly. An evaporating milk droplet in a spray drier in cocurrent air flow does not initially appreciably exceed the wet bulb temperature and can be held effectively at temperatures below 60°C. As the falling temperature period is approached in the course of further evaporation, the temperature rises to a final value determined by the final temperature of the drying gas and the residence time in the drier. Under properly controlled spray-drying conditions, the changes in milk protein structure and solubility are minor. Spray drying does not denature the whey protein significantly, and the levels of denatured whey protein in dairy powders are set by the preheating of the milk. An excellent example is in relation to the whey protein nitrogen index (WPNI). According to Pisecky (1997), the WPNI expresses the amount of undenatured whey protein (milligrams of whey protein nitrogen per gram of powder). It is a measure of the sum of heat treatments to which the milk has been subjected prior to evaporation and spray drying. The heat treatment of a concentrate and subsequently of a powder has only a negligible effect on the WPNI. The main operation to achieve the required value is the preheating time/temperature combination. However, there are other factors that influence the WPNI, including the total amount of whey protein and the overall composition of the processed milk as influenced by animal breed and lactational and seasonal variations. The individual design of the heating equipment, whether direct or indirect heating, is also important. The WPNI test is best used to classify powders into the categories low, medium, and high heat; a guide to heating conditions is given by Oldfield and Singh (2005). One of the purposes of heat pretreatment is obviously to ensure the microbiological quality of dairy products. The influence of the heat treatment on the denaturation of whey proteins to achieve the desired properties of the final products is just as important in milk powder production. SMP for cheese manufacture should have as much undenatured protein as possible, that

434

11. Effects of drying and storage on milk proteins

is, it should be low heat produced (WPNI > 6), whereas, for bakery product, high heatproduced powder with high denaturation is required (WPNI < 1.5). For ice cream, chocolate, and confectionery, medium heat-produced powder is required. According to Schuck et al. (1994a), the use of microfiltration (pore diameter, 1.4 μm), combined with low heat treatment during vacuum evaporation, allows the production of a low low heat SMP with a WPNI close to 9 mg of whey protein nitrogen/g of powder, bacterial count 99.5%, dispersibility index >98.5%, and wettability index 80 products) have indicated that this method could be applied to a wide range of food products and spray-drier types. Combined with knowledge of the temperature, the software provides analysis of the desorption curve (measured relative humidity vs time), total solids, density and specific heat capacity of the concentrate, air flow rates, water content, RH of the outlet air, the current weather conditions, cost per kWh, and the percentage of drying in the integrated fluid allows determination of enthalpy H, T, RH (including ΔE) for each inlet air, concentrate and powder flow rate, specific energy consumption, energy and mass balance, yield of the dryer, and cost (in € or in $) to remove 1 kg of water or to produce 1 kg of powder (summarized in Fig. 11.6). This figure is a representation of the software delivery: – Air characteristics at the dryer (with or without integrated fluid bed) inlet and outlet (upper part) – Flow, energy, and cost calculations (lower part) (Schuck et al., 2009) Thus, the interest of the desorption curves was in evaluating water transfer during spray drying of various dairy concentrates using thermodynamic and biochemical approaches. Whey protein concentrates and isolates (WPC35, WPC50, WPC70, and WPI90) with or without heat denaturation, MC, sodium caseinate (NaCas), and milk with and without whey protein enrichment were dried in a three-stage pilot plant spray drier. When the concentrate temperature, air flow rate, concentrate flow rate, total solids content of the concentrate, inlet air temperature absolute humidity, inlet air temperature before and after heating, and outlet air temperature after drying are known, it is possible to determine the specific energy consumption (SEC), that is, the ratio of the energy consumed to the evaporation of 1 kg of water (measured in kJ kg
1 water) (Schuck et al., 1998a; Bimbenet et al., 2002). Thus, if you spray-dry only free water, the energy used in terms of SEC would be close to 2500-kJ kg
1 water. In the concentrate amounts of bound water to free water increase, the SEC increases (e.g., up to 10,000-kJ kg
1 water). The significance of a very high SEC relates to the decreasing availability of the water, limiting water transfer, and thus increasing the surface temperature of the droplet, and hence increasing the risk of protein denaturation of the powder.

Whey proteins The results presented in Table 11.3 show that water transfer during spray drying decreased when the whey proteins were native proteins. For the same moisture content, the SEC for

Drying of proteins

439

FIG. 11.6 Parameters of spray drying calculated by SD2P software.

drying was higher when (a) the native whey protein content increased in WPC and in milk and (b) the whey proteins were heat denatured in WPC35. However, the SEC was lower when (c) the whey proteins were heat denatured in WPC50, WPC70, and WPI90. These results can be explained by the availability of the water (bound or unbound) in the concentrate in relation to the nature and the content of the whey proteins.

Caseins The results presented in Table 11.3 show that water transfer during spray drying decreased when the micellar casein content increased. For the same moisture content, the SEC for drying

440

11. Effects of drying and storage on milk proteins

TABLE 11.3 Specific energy consumption at 4% moisture content for the drying of dairy proteins SM

SM + WPI

WPC

Protein content (%)

34

50

Heat treatment (72°C/4 min)

N

N

N

Y

N

Y

N

Y

N

SEC (3%) (kJ kg
1 water)

5900

6400

5950

7700

6800

6550

7050

6600

7200

35

50

WPI

MC

NaCas

90

90

90

Y

N

N

6500

6900

5900

70

MC, micellar casein; NaCas, sodium caseinate; N, no heat treatment; SEC, specific energetic consumption at 4% of moisture content; SM, skim milk; WPC, whey protein concentrate; WPI, whey isolate; Y, heat treatment.

was higher when (a) the micellar casein content increased in MC compared with skim milk and (b) casein remained in a micellar state (as in MC) compared with a soluble state (e.g., in NaCas). These results can be explained by the availability of the water in the concentrate in relation to the content and the structure of the caseins. Water is more available when the caseins are soluble than when they are in a micellar state. All these results also show that water transfer depends on the relationship between the water and the protein components and that these components should be taken into account when optimizing spray-drying parameters for proteins.

Characterization of insolubility The first sections of this chapter have dealt primarily with isolated protein components; however, commercial products are usually mixtures of the various milk components, manufactured to specified compositions. Drying of milk products in dairy factory facilities began to develop rapidly at the beginning of the 20th century. At this time, a lot of progress was made by engineering developments, but the developing equipment also attracted researchers interested in product quality. One of the topics of interest was the effect of drying on the reconstitution properties of the powder; it was observed that the drying equipment and processing conditions could have a major effect on this phenomenon. One obvious difference was the insolubility of powder produced by roller driers compared with spray driers; even within spray driers, different designs could give markedly different performance. Early workers who investigated the chemistry of this phenomenon found by investigation of heat treatments of product at moistures up to 13%, and milk concentrates, gave a total solid zone of sensitivity of about 60%–87% total solids. Thus, the rate of degradation was greatest at intermediate moisture contents in the dehydration process (Wright, 1933). It was also established that the insolubility formed in drying was associated with the casein protein (Lampitt and Bushill, 1931). These results were discussed in later literature, for example, Hall and Hedrick (1971) and Parry (1974), but little work was done following this. A technique was developed by Kudo et al. (1990) to apply heat treatments to skim milk powders at different moisture contents using preheated copper tubes. Later, Straatsma et al. (1999b) adapted the process to products with total solids between concentrate and

441

Characterization of insolubility

powder, but the heat treatments were too severe; only very insoluble products were obtained with no meaningful reaction rate data. In this case, the reaction rate was modeled purely as a step function, over an assumed moisture range of 10%–30%. Insolubility data were obtained from drying trials on a pilot scale drier and deriving an insolubility equation by fitting exponents to an exponential equation. The solution to this experimental issue was to reduce the treatment temperatures so that the heating-up times were a smaller fraction of the treatment time. Baldwin and Truong (2007) applied this approach and found it to be practical; treatment temperatures in the range 5–55°C were applied to skim milk powder materials with moisture contents of 10–55 g moisture per 100-g product (moisture on a wet basis). One difficulty encountered was handling the materials manufactured for the heating trials, which ranged from thick concentrate to sticky paste, to moist particulates. It was also found that to be stable, the moist materials had to be held at minus 85°C. The zone of most sensitivity was approximately 15%–38% moisture (Fig. 11.7), in broad agreement with the Wright (1933) hypothesis. In the sensitive moisture range, the time to 6.0 mL of insoluble material at a product temperature of 50°C was about 8 min. The insolubility versus temperature data were found to fit Arrhenius equations (rate versus 1000/K) with a high degree of correlation (r 2 > 0.98) (Baldwin and Truong, 2007) and were used to estimate rates of reaction at higher temperatures. For example with temperatures of drying of 50°C, 70°C, and 90°C the times to reach 1.0-mL sediment in skim milk powder, a common specification figure, were calculated to be 60, 8, and 1 s, respectively. This demonstrates the need for short drying times to achieve high solubility of skim milk powders in cold water. MPC powders have even greater sensitivity to the insolubility reaction. A MPC 70 product with very good solubility was made up at different moisture contents and held at low temperatures (5°C). The MPC times at maximum sensitivity were 1/10 that of skim milk powder, and the moisture zone of sensitivity was displaced with respect to skim milk powder (Fig. 11.8); interestingly, the moisture/protein ratio was similar to skim milk powder, suggesting that the association of water with the protein is a key factor. The insolubility results to 1.0-mL insolubility were described mathematically and used in a computational fluid dynamics (CFD) simulation of spray drying, to identify drier variables

Time to 6.0 mL insolubility 1,000,000 100,000

Time (min)

FIG. 11.7 Moist skim milk products held at 50°C. Time (log10 scale in minutes) to 6.0 mL of sediment, against moisture (%). Reproduced from Baldwin, A.J., Truong G.N.T., 2007. Development of insolubility in dehydration of dairy milk powders. Food Bioprod. Process. 85(C3), 202–208.

10,000

50°C

1000

100 10 1

0

10

20

30

40

Moisture (%)

50

60

11. Effects of drying and storage on milk proteins

FIG. 11.8 Moist products held at 5°C, comparing MPC with SMP. Time (log10 scale in minutes) to 6.0 mL of sediment, against moisture (%). Reproduced from Baldwin, A.J., Truong G.N.T., 2007. Development of insolubility in dehydration of dairy milk powders. Food Bioprod. Process. 85(C3), 202–208.

10,000 SMP and MPC, 5°C

Time (min)

442

1000 SMP B SMP A 100 MPC70 10 0

10

20

30

40

50

60

Moisture (%)

and to investigate the drying of MPC in a pilot scale and commercial drier (unpublished results). A large number of papers have been published on the use of CFD to model the drying of droplets in spray driers. A few of these have included an insolubility reaction calculation such as Straatsma et al. (1999a,b) and a laboratory drier with an atomizing device that gave monosized droplets (Rogers et al., 2012). The different methods of CFD modeling of spray drying and the development of surfaces in droplet drying have been reviewed (Wu and Langrish, 2009); this reveals the complexity of modeling the spray drying of food materials. Studies based on small-scale driers should be used to identify variables for further investigations on larger driers because of the assumptions made in modeling and the different operating conditions between laboratory and commercial driers. For instance, commercial driers produce a wide range of particle diameters; as drying time is proportional to the square of the particle size, this will have a major effect on overall results. The conditions for drying may need to be a compromise between the optimum for solubility and the optimum for powder flowability and reconstitution properties.

Casein micelle insolubility A number of different protein molecules make up the casein profile of milk. A small proportion of the caseins exist in the milk serum, but the majority exists in micelles as submicron entities in colloidal solution. The micelles carry calcium and phosphate, which are involved in the structure. For the discussion of insolubility, two features of the micelle are important: (i) the large quantity of water associated with the micelle, estimated to be in the order of 2–3-g water/g dry solids, depending on the method of measurement (Horne, 2014) and (ii) the flexibility of the protein chain. The casein micelle structure in milk has been described as “loose and spongy” (Schmidt and Payens, 1976). The water will contain molecular species such as lactose that may be linked to the protein by hydrogen bonding, ionic species linked by ionic bonds, which contribute to the stability in water. If we follow the drying pathway, the ionic environment within the micelle will become more concentrated, and the proteins will reorientate to adapt to the modified environment.

Characterization of insolubility

443

However, if the drying time is short, the amount of adaptation will be limited, and the properties of the powder will be little changed when reintroduced to water. It can be presumed that when the powder is below a critical moisture, the protein structure becomes more rigid, and the changes in molecular structure become much slower.

Effects of the dehydration process Most parts of the process for manufacturing high protein products have been linked to the underlying cause of the slow rate of hydration of the dried material. However, the drying conditions are key: products with low or high solubility with a given solubility test can be produced by manipulation of spray-drying conditions, as noted previously. The key controlling factor of the rehydration of protein powders is the dispersion of the micelles. Transmission electron microscopy revealed close packing of micelles at the surface in high protein products (McKenna, 2000). The slow rate of dissolution has often been attributed to intermicellar bonding forming a hydrophobic “skin” on the surface of the particles. Compositional analysis by electron microscopy has shown that this concentration of protein at surfaces is due to diffusion of lactose in the initial stages of drying, arising from the moisture gradient in the droplet, and is common across a range of powders (Kim et al., 2009). This reduction in lactose probably slows the initial ingress of water during reconstitution of the products. However, it is unlikely to explain all of the slow solubility observed in particle solubility studies. A thorough study of dissolving of stored MPC products, using particle size measurement, micrographs, and analysis of rates of solution of components, has been undertaken (Mimouni et al., 2009, 2010a,b). The results of this work tend to discount many of the theories proposed to date for long hydration times (Mimouni et al., 2010b). Given the structure of the micelle, there are two environments where hindrance of hydrophilic forces could operate, one at the surface of the micelle and the other internally. It is well known that the casein micelle is stabilized in solution by κ-casein, which is concentrated on the surface. Drying may turn the micelle surface into a hydrophobic form leading to the aggregation of micelles. Another alternative (Baldwin, 2010) is that the resistance to rehydration of the micelle arises from reconfiguration of the protein chains throughout the micelle to a more hydrophobic form. The casein molecules in concentrate will be orientated so that more hydrophilic groups are associated with these water molecules. If the water is removed slowly, the molecular backbone will be able to reconfigure to the lower moisture environment, and hydrophobic portions of the amino acid chain will be exposed. During drying, a higher particle temperature considerably raises the rate of the configuration change reaction. In addition, a longer residence time allows more time for the protein to react to the withdrawal of moisture. The former hydrogen bonds with water will be satisfied with bonds within the protein molecule. The effects of both time and temperature are well documented in experience with commercial plants and in laboratory characterization. Once the protein is reorientated, it will resist hydration when placed in a water environment; it appears to be difficult for it to adopt its previous orientation when reconstituted in cold water. However, warm temperature water provides more energy and can speed up the process considerably. Finding the explanation for the cause of the slow hydration is made more difficult by the uncertainty surrounding the structure of the casein micelle (Anema, 2014).

444

11. Effects of drying and storage on milk proteins

Various bonding mechanisms can be postulated, and reagents can be used in the insolubility test to determine which bonds are involved in holding the particles together (Baldwin, 2010). With SMP and moderate insolubility, the bonding is dominated by ionic and hydrogen bonds; this reflects the changes occurring during the drying process. Polyacrylamide gel electrophoresis (PAGE) indicated that noncovalent hydrophobic and hydrogen bonding was implicated in changes in solubility with storage (Havea, 2006). Modifications of composition can be effective in overcoming the phenomenon of slow hydration. It has been demonstrated in studies of protein dehydration that proteins can be stabilized during dehydration with certain carbohydrates (Prestreleski et al., 1993). It is thus possible that in products with lower protein content and higher lactose content, the lactose molecules can act as water replacement molecules. The exchange of sodium for calcium by ion exchange (Bhaskar et al., 2001); the addition of monovalent salts, especially sodium salts (Schuck et al., 2002); or the addition of sodium caseinate (Schokker et al., 2011) will disrupt some of the casein micelles leading to more free casein and improvement in rates of hydration These ions may also alter the charge environment of the protein molecules in the casein micelle. A few studies have succeeded in measuring changes in the configuration of proteins by NMR (Allison et al., 1996), as a consequence of dehydration. Prestreleski et al. (1993) investigated a few different proteins including α-casein and deduced changes in orientation from unstructured in solution to an arrangement in powder that had elements of β-sheet. Single casein units are much less complex than the casein micelle structure, but these studies demonstrate in principle the development of hydrophobic forms of protein from dehydration.

Changes in milk proteins during storage of dry powders Changes in solubility The casein protein products become increasingly resistant to rehydration with storage (Anema et al., 2006; Burgaina et al., 2016; Havea, 2006; Haque et al., 2011, 2015; Mimouni et al., 2009, 2010a,b; Richard et al., 2013; Schokker et al., 2011). Cao et al. (2011) manufactured MPCs by either evaporation or nanofiltration and compared composition and properties during storage for 24 weeks. Studies by Haque et al. (2011, 2015) of MPC 85 using carbon NMR found changed backbone rigidity indicating evidence of conformational changes during storage, but the changes were small, and their significance for solubility was not certain. Syll et al. (2012) used isothermal calorimetry to follow dissolution of stored dairy products. Using atomic force microscopy (AFM), very large changes in surface rigidity of micellar casein particles were measured with storage (Burgaina et al., 2016). However, the storage conditions, 40°C for 10 months, were very extreme. The changes on storage are a hindrance to the distribution of high protein products and warrant more investigation of stabilizing treatments and most importantly on the efficacy of reconstitution techniques involving increased water temperatures.

Chemical changes A range of chemical reactions that modify proteins can occur in dried milk products, particularly at elevated temperature. Both of the most important of these involve lysyl side chains

445

Changes in milk proteins during storage of dry powders

and are the formation of Maillard and pre-Maillard compounds, in the presence of sugars, and the formation of isopeptide bonds, particularly in products containing phosphoseryl residues, such as casein.

The formation of Maillard and pre-Maillard compounds The Maillard reaction occurs when lysine-containing proteins interact with reducing sugars. The first stable compound formed during the Maillard reaction is the Amadori product (so-called because it is the result of a class of reaction called the Amadori rearrangement), shown in Fig. 11.9. These compounds block the ε-amino groups of lysine residues, reducing the bioavailability of that essential amino acid. This reaction is dependent on a reducing sugar, usually lactose, being present as the coreactant (Erbersdobler, 1986). Highly pure protein products such as casein and caseinate do not suffer significantly from this reaction, as not enough lactose is present. The reaction is particularly important in WMPs, SMPs, WPCs, and MPCs and can occur to a limited extent in some MPIs and WPIs. The rate of reaction is critically dependent on the level of moisture (aw) and the temperature as well as the lactose content. Advanced Maillard reaction products are partly responsible for the development of aromas and color during food processing and preparation. Protein

Protein

NH2

NH

O OH

H

OH

H

OH

OH

OH OH

HO

Addition product

Advanced maillard products

HO

Schiff base

O

O

OH O

OH O

HO

OH

HO OH

OH

OH

Lactulosyl lysine (Amadori product)

Lactose

HCL heat

Lysine

+

C O

CH2NH-Lys

O

Furosine

FIG. 11.9 Formation of lactulosyl lysine.

+

Pyridosine

446

11. Effects of drying and storage on milk proteins

Measuring lactulosyl lysine levels Lactulosyl lysine formation is conveniently monitored by measuring furosine concentrations (Erbersdobler et al., 1987) and can be indirectly measured by monitoring the amount of available amine (Herna´ndez et al., 1991). Lactulosyl lysine in WPCs can be measured directly by mass spectrometry where the addition of one lactose molecule increases the molecular weight of the protein by 324 Da. Fig. 11.10A shows the mass spectrum of a mixture of the native A and B variants of β-lactoglobulin and their mono-, di-, and trilactosylated derivatives. The results from all three methods have been shown to correlate well (Fig. 11.10). A recent review (Ritota et al., 2017) on the evaluation of heat treatments of milk discusses other methods that may be used to monitor lactulosyl lysine levels and other thermal treatment indicators.

Rates of formation of lactulosyl lysine We studied the rate of lactosylation for WPC products as dry powders. The rates were found to be dependent on the lactose concentration (a consequence of processing to reach desired protein levels), aw, and temperature (T). The trials lasted only a few months, and extrapolation beyond that timeframe cannot be done with confidence. Detailed kinetics developed using WPC80 for a range of aW and T values allowed the rates of lactosylation to be predicted for periods up to 4 months. For the kinetic evaluation, Fig. 11.9 can be simplified to k1

k2

reactants ƒƒƒ! lactulosyl lysine ƒƒƒ! advanced Maillard products where k1 and k2 are the rate constants for the formation and degradation of lactulosyl lysine. The rate equation for the formation of lactulosyl lysine is dL ¼ k1 ½reactantsn1
k2 ½Ln2 dt

(11.1)

where L represents lactulosyl lysine and n1 and n2 are the rate orders for the formation and degradation of lactulosyl lysine, respectively. Because the formation of lactulosyl lysine in dairy powders uses only a fraction of the available reactants, the first reaction can be considered to be zero order, and the degradation of lactulosyl lysine is a first-order process. This simplifies the equation to dL ¼ k1
k2 ½L dt Rearrangement of this equation and solving for [L]t gives     k1 k1

½L0 Þ exp ð
k2 tÞ ½Lt ¼ k2 k2

(11.2)

(11.3)

Furosine is a hydrolysis product of lactulosyl lysine and gives a direct measure of its concentration (Fig. 11.10). Therefore, furosine can be substituted for lactulosyl lysine in the rate equations. The samples used were individual samples that were removed from controlled storage at seven time points (2, 6, 12, 24, 40, 78, and 116 days). They were analyzed for furosine, and the

2500

y = –8.5x + 3202 r 2 = 0.956 Furosine (mg.g of sample–1)

2000

1500

1000

500

0 100

150

250

200

300

350

400

Available amine (µmol g of sample–1)

(A) 12

y = 0.0048x – 0.5 r 2 = 0.984

Average lactose bound

10

8

6

0.33 30°C 0.33 35°C 0.33 40°C 0.54 30°C 0.54 35°C 0.54 40°C 0.80 30°C 0.80 35°C 0.80 40°C

4

2

0 0

(B) FIG. 11.10

500

1000

1500

2000

2500

Furosine (mg/g of powder)

(A) Correlation between available amine and furosine values for stored samples of a WPC56. (B) Average number of lactose molecules bound, determined by mass spectrometry, against furosine for a WPC56 (Higgs, unpublished data).

448

11. Effects of drying and storage on milk proteins

values were plotted against time. The rates were determined using nonlinear regression with SigmaPlot 8.0. At 40°C; the samples at time points after 40 days for water activities of 0.54 and 0.80 showed advanced Maillard browning and were not included in the regression analysis. R2 values for the regressions were between 0.95 and 0.99. The lactosylation rate constants were greater at higher temperatures and at higher water activities. The rates were low at low water aw (0.33), with small increases with temperature; however, at higher aw, the rates increased substantially with increasing temperature (Fig. 11.11). It was observed for a number of WPC80 specifications that lactosylation appeared to stop when only 2.5 (average) or 3 lactose molecules had been bound per β-lactoglobulin molecule. This corresponded to about 20% of the total lysine being blocked. This condition was specific to the 80% protein products, which contained about 12% lactose. WPC56 products showed a much greater degree of lactosylation. However, it should be noted that much higher levels of lactosylation have reportedly been seen in overseas laboratories in WPC80 samples stored for long periods (Harper, Ohio State University, 2003, personal communication). Similar kinetic work has been reported for SMPs stored at a range of temperatures (37°C, 50°C, and 60°C) or with a series of initial water activities (0.33, 0.43, 0.52, 0.69, 0.85, and 0.98) (Pereyra Gonzales et al., 2010) and temperatures of 30°C, 32.5°C, and 35°C with water activities of 0.43 and 0.52 (Aalaei et al., 2018). Pereyra Gonzales et al. (2010) found significant increases in Maillard reaction rates with temperature but no increase related to the initial aw of the powder, whereas Aalaei et al. found significant increases in Maillard reaction rates with both temperature and aw. 250 30°C 35°C 40°C

Rate for formation of furosine

200

150

100

50

0 0.3

0.4

0.5

0.6 Water activity

FIG. 11.11

Rate constants for the formation of furosine in a WPC80.

0.7

0.8

0.9

449

Changes in milk proteins during storage of dry powders

Formation of isopeptide bonds Isopeptide bonds are formed largely by the breakdown of the phosphoseryl side chains that are present in products containing casein during processing or storage, to form dehydroalanyl side chains (Friedman, 1999). The latter are reactive and will form cross-links, mainly with adjacent lysyl (but also with histidinyl or cysteinyl) side chains to form lysinoalanyl, histidinoalanyl, or lanthionyl isopeptides, respectively. The reaction during milk powder processing is considered to be an alkali-catalyzed reaction that is accelerated by heat treatment. This reaction is not known to be significant in whey products, which do not contain significant amounts of phosphoseryl residue (Fig. 11.12). The main isopeptide product on both acid and gastrointestinal digestion gives lysinoalanine, which renders the lysine nonbioavailable. Additional minor reactions form histidinoalanine and lanthionine on digestion. The latter compound, although only a minor component, is also important because it renders cysteine partially nonbioavailable, and the sulfur amino acids are often nutritionally limiting in milk proteins. (Note that lanthionine formation blocks the bioavailability of cysteine, which is not normally considered to be an essential amino acid, because it can be synthesized from methionine; however, methionine is itself a nutritionally limiting amino acid in casein.) Studies have indicated that, although lanthionine and histidinoalanine linkages are formed under the alkaline conditions encountered during processing, it is only lysinoalanine that is formed in the neutral conditions encountered in powders. Lysinoalanine formation in β-casein has been found to be enhanced by pressure treatment under alkaline conditions (Schwarzenbolz and Henle, 2010). It has not been investigated if pressure treatment also affects the rate at which lysinoalanine is formed during subsequent storage. Lysinoalanine is usually measured directly in protein hydrolysates as part of an extended amino acid analysis (Henle et al., 1991). In casein products, measurement of available amine is a useful and simpler, though less specific, alternative.

Rates of formation of lysinoalanine The rates of formation of lysinoalanine in caseinates and proteinates (TMPates) have been investigated by W. Thresher (personal communication, 1996, 1997), for a range of temperatures and water activities. The rate constants for lysinoalanine formation as a function of temperature are shown in Fig. 11.13. The rate of lysinoalanine formation increased with increasing temperature and aw for both caseinates and TMPates. Calcium TMPates and caseinates were found to have lower rates of lysinoalanine formation than potassium or sodium TMPates and caseinates. Amino acids other than lysine Cysteine, methionine, and tryptophan are other essential amino acids that could be rendered nonbioavailable by reacting during processing and/or storage. Cysteine undergoes β-elimination to give dehydroalanine when treated with alkali. This can then react with a FIG. 11.12 Formation of lysinoalanine.

Phosphate Phosphoserine

Dehydroalanine

Lysine

Lysinoalanine

450

11. Effects of drying and storage on milk proteins

FIG. 11.13 Rate constants for lysinoalanine formation in caseinates stored at various temperatures (Thresher, 1996, personal communication).

500

Rate of LAL formation (ppm/month)

400

300

Potassium Rep 2 Potassium Rep 1 Calcium Rep 2 Calcium Rep 1 Sodium Rep 2 Sodium Rep 1

200

100

0

–90 –85 –80

20

25

30

35

40

45

50

55

60

Storage temperature (°C)

lysine residue to give lysinoalanine. Cysteine can also react with dehydroalanine to give lanthionine (Friedman, 1999). Chemically determined values for cysteine and lysine availability have been found to correlate well with rat protein efficiency ratios for heat- and alkalitreated caseinates (Thresher, 1996, personal communication). Tryptophan residues are relatively stable during processing and storage. They are not easily oxidized and have been found to be relatively resistant to oxidizing lipids, alkali, quinones, and reducing sugars (Nielsen et al., 1985). Any losses are small and not significant when compared with losses of other amino acids such as methionine and lysine. However, we have seen small amounts of oxidized tryptophan residues in digests of skim milk purchased from the supermarket (Fig. 11.14). Methionine is relatively easily oxidized to the sulfoxide, but methionine in the sulfoxide form is still bioavailable (Nielsen et al., 1985). Table 11.4 indicates levels of key essential amino acids following either alkali treatment of casein or extensive lactosylation of WPC. Note that the losses of amino acids other than lysine are considerably lower than the losses of lysine. The conditions used for the casein are well beyond any normal exposure during processing or storage. Implications for nutritional value of milk proteins Milk protein is rich in essential amino acids, with many, including lysine, well exceeding recommended requirements (Table 11.5). Use of these proteins as the predominant nutritional source thus poses few problems if some of the lysine is nonbioavailable.

451

Changes in milk proteins during storage of dry powders

FIG. 11.14 Time-of-flight mass spectrometry of the M2+ ion from the tryptophan-containing κ-casein peptide SPAQILQWQVLSNTVPAK. The chemical structures in the figure show the various levels of oxidation found. TABLE 11.4 Amino acid concentrations of a control and an alkali-treated casein and a control and a lactosylated WPC56 Concentration (mg/g crude protein)

Concentration (mg/g sample)

Amino acid

a

Control casein

Alkali-treated casein

Control WPC

Lactosylated WPCb

Tryptophan

13.4

12.1 (90%)

8.4

8.1 (96%)

Lysine

91.0

60.8 (67%)

50.3

44.5 (88%)

Methionine

31.4

25.3 (80%)

12.3

12.5 (102%)

Cysteine

4.4

3.3 (75%)

14.9

14.8 (99%)

a

Casein was heated for 4 h at 80°C in 0.15 M NaOH (Nielsen et al., 1985). b WPC was heated at 40°C, aw 0.75 for 100 h. The median number of lactosyl groups on β-lactoglobulin was 5 (Higgs, unpublished results). Sources: Nielsen, H.K., de Weck, D., Finot, P.A., Liardon, R., Hurrell, R.F., 1985. Stability of tryptophan during food processing and storage. 1. Comparative losses of tryptophan, lysine and methionine in different model systems. Br. J. Nutr. 53, 281–292; Higgs (unpublished results).

Lysine The main concern during the storage of most nutritional proteins is the loss of lysine as a result of Maillard reactions or isopeptide bond formation. Lactulosyl lysine renders lysine nonbioavailable. The protein efficiency ratio (PER) was found to be decreased in a WPC56, with an average of five lactulosyl lysine residues per protein molecule (Fig. 11.13). A more detailed study using skim milk diets with pigs confirmed that lactulosyl lysine was nonbioavailable (Rerat et al., 2002). That study also found a decrease

452

11. Effects of drying and storage on milk proteins

TABLE 11.5

Essential amino acid content of milk and other proteins

Essential amino acid (AA)

Recommended requirementa (mg/g protein)

Caseinate

TMP, MPI

WPC

Soy

Wheat

Isoleucine

28

46

44

54

47

33

Leucine

66

91

103

119

85

68

58

77

81

94

63

27

Sulfur AA

25

33

39

52

24

39

Aromatic AA

63

106

102

68

97

78

Threonine

34

43

45

66

8

29

Tryptophan

11

12

14

20

11

11

Valine

35

57

57

51

49

43

Histidine

19

29

27

21

29

–

Lysine b

a b

For 2–5 year olds. Includes cysteine, cystine, and methionine.

in the digestibility of lysine, phenylalanine, valine, cystine, aspartic acid, glycine, and methionine residues. This decrease suggests that lactulosyl lysine residues hinder the release and therefore utilization of adjacent amino acids (Fig. 11.15). Limited human studies have been conducted on Maillard reaction products, which include lactulosyl lysine. One study looked at adolescent males on diets containing different levels of 3.8 3.6 3.4

PER

3.2 3.0 2.8 2.6 2.4 2.2 2.0 0

1

2

3

4

5

6

Lactosylation level (average lactose bound)

FIG. 11.15

Plot of lactosylation level against PER. The PER value for ANRC casein was used for zero lactosylation.

Changes in milk proteins during storage of dry powders

453

Maillard reaction products, which principally showed that Maillard reaction reduces protein digestibility (Gilani et al., 2012). Studies on infant formula have shown that products with an increased amount of Maillard reaction products had decreased PER values. Those formulas with lower PER values also resulted in the rats having lower levels of plasma lysine. A recent study (Hillmana et al., 2019) found that skim milk powders with a significant content of Maillard products caused intestinal inflammation and reduced initial weight in young rats. Lysinoalanine is not a bioavailable source of lysine (Robbins et al., 1980; Friedman, 1999; Gilani et al., 2012). Alkali treatment of casein with 0.2 N NaOH at 80°C for 1 h reduced the PER of casein from 3.09 to 0.02 for a diet containing 10% casein (Possompes et al., 1989; cited in Friedman, 1999). Milk proteins are unusually rich in lysine (Table 11.5) and can stand to lose a significant proportion of the lysine before it becomes limiting (around 50% in the case of whey proteins and 25% in the case of casein). The real concern arises when milk proteins, and particularly whey proteins, are being added to a mixture to provide a source of lysine supplementation. When this is the case, it is particularly important to ensure that the lysine content has not been compromised after long storage. Steps to ensure this include keeping the product at temperatures 90% in milk powder that had been stored at 60°C for 5 weeks, which resulted in a loss of 80% of available amine. This powder was considered to show “advanced” Maillard browning.

455

Changes in milk proteins during storage of dry powders

Product-specific storage trials Samples of WPC80 powders that had earlier been shipped from New Zealand to the United States or Europe were obtained and analyzed for lactosylation levels. The powders were at the time 80% protein powders to determine changes if any in nutritional properties, but these powders were kept at constant temperature and were

Market samples

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Average lactose

2000/2001 Production

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Average lactose

FIG. 11.17 Levels of lactosylation in market samples of a WPC80 compared with levels in freshly produced WPC80 in a subsequent season.

456

11. Effects of drying and storage on milk proteins

not exposed to any of the temporal variations possible during shipping and storage in overseas warehouses. A decrease in available amine of 5% was seen in MPC85 when stored at 20°C for 2 years; this increased to 10% when the storage temperature was increased to 30°C. A single MPI stored for 2 years gave consistent available amine results over the storage period. The caseinates and caseins in the 2-year study showed no definitive trend in available amine values. Most values remained consistent over the 2 years. This was expected as a previous 3-month study showed that storage temperatures in excess of 30°C were required for significant levels of lysinoalanine formation.

Rehydration of protein powders Most food additives are prepared in powder form and need to be dissolved before use. Water interactions in dehydrated products and dissolution are thus important factors in food development and formulation (Hardy et al., 2002). Dissolution is an essential quality attribute of a dairy powder as a food ingredient (King, 1966). Many sensors and analytical methods such as the insolubility index (International Dairy Federation, 1988; American Dairy Products Institute, 1990), nuclear magnetic resonance (NMR) spectroscopy (Davenel et al., 1997), turbidity, viscosity, and particle size distribution (Gaiani et al., 2006a,b) can now be used to study water transfer in dairy protein concentrates during rehydration. Using combinations of these methods, it is very easy to determine the different stages of the rehydration process, that is, wettability, swellability, sinkability, dispersibility, and solubility. The insolubility index (ISI) described by the IDF standard (International Dairy Federation, 1988; Schuck et al., 2012) for skim milk is the volume of sediment (mL for 50 mL of solution) after rehydration using 10 g of powder in 100 mL of distilled water, at 24°C (temperature of 50°C for roller dried powders). Mixing is for 90 s, at 4000 rev/min, and centrifugation for 300 s, at 160 g. With this method, the quantity of insoluble (more correctly slow to dissolve) material can be determined. NMR spectroscopy is a technique for determining the rate of solution, the time required for complete reconstitution of powders and the transverse relaxation rate of reconstituted solutions. The method was first described by Davenel et al. (1997). A 40-mm diameter glass tube filled with 20 mL of water at 40°C was placed on the gap of the magnet of a Minispec Bruker PC 10 NMR spectrometer operating at a resonance frequency of 10 MHz. A suitably designed funnel and an electric stirrer (glass spatula) were inserted into the tube. They showed that the solubilization rate was independent of the quantity of powder poured (up to 20-g powder/100-mL water); the solubilization rate increased with the stirring rate. In subsequent experiments, the rotation rate of the stirrer was adjusted after starting to 1150 rev/min for spray-dried powders, and 1 g of powder was poured into the water. The NMR measurements were generally continued until the solution was completely reconstituted, except if insoluble material was formed. Each decay curve was obtained by sampling a maximum of 845 spin echoes of a Carr-Purcell-Meiboom-Gill (CPMG) sequence every 20 s during the reconstitution period. Interpulse spacing between 180-degree pulses was fixed at 2 ms to limit the diffusion effect caused by stirring. The NMR kinetics

Rehydration of protein powders

457

method was used in triplicate. The CPMG curves were well approximated by the sum of two exponential curves to determine the protons attributed to water protons in fast exchange with exchangeable protons of nondissolved powder particles, as well as the protons attributed to water protons and exchangeable protons in the reconstituted phase (Davenel et al., 1997). With this method, it is possible to differentiate between the truly insoluble material and the falsely insoluble material. The falsely insoluble material can be explained by the low water transfer during rehydration in contrast to denatured protein, which is truly insoluble (Schuck et al., 1994b). For viscosity measurement, a rheometer can be used to obtain viscosity profiles. In the studies of Gaiani et al. (2005, 2006a,b), the blades were placed at right angles to each other to provide good homogenization. Industrial dissolution processes usually include stirring at a constant speed, and the experiments were therefore designed to provide a constant shear rate (100 s
1). MCP was added to the rheometer cup manually. The aqueous phase used was distilled water at a volume of 18 mL. The powder was dispersed in the rheometer cup 50 s after starting the rheometer. Dissolution is highly dependent on temperature and concentration. The total nitrogen concentration employed to study these effects was about 5% (w/v), and the temperature was about 24°C. The experiments to provide the turbidity profiles were carried out in a 2-L vessel equipped with a four-blade 45-degree impeller rotating at 400 rev/min. A double-walled jacket vessel maintained the temperature at 24°C. The turbidity sensor was placed 3 cm below the surface of the water and was positioned through the vessel wall to avoid disturbance during stirring. Turbidity changes accompanying powder rehydration were followed using a turbidity meter. The apparatus used light in the near-infrared region (860 nm), the incident beam being reflected back at 180 degrees by any particle in suspension in the fluid to a sensitive electronic receptor (Gaiani et al., 2005). A laser light diffraction apparatus with a 5-mW He-Ne laser operating at a wavelength of 632.8 nm can be used to record particle size distributions. The particle size distribution of dried particles was determined using a dry powder feeder attachment, and the standard optical model presentation for particles dispersed in air was used. To measure the particle size distribution of micellar casein in concentrates, 0.5 mL of suspension was taken from the rheometer cup and introduced into 100 mL of prefiltered distilled water (membrane diameter, 0.22 μm) to reach the correct obscuration. The results obtained corresponded to average diameters calculated according to the Mie theory. The criterion selected was d(50), meaning that 50% of the particles had diameters lower than this criterion (midpoint of cumulative volume distribution) (Gaiani et al., 2005, 2006a,b). By using this combination of three methods, it was possible to follow the water transfer during rehydration and obtain the wetting time, determined using the first peak of increased viscosity and turbidity, and the swelling time, determined using the second peak of viscosity in relation to the increase in particle size. The rehydration time was then determined according to stabilization of the viscosity, turbidity, and particle size values. The results in Table 11.6 show that the rehydration of MCP occurs in different stages: first, there is wetting and swelling of the particles, followed by slow dispersion to reach a homogeneous fluid, in agreement with Gaiani et al. (2005, 2006a,b, 2007). Using an NMR method, Davenel et al. (1997) also demonstrated two stages during MC rehydration, attributed to water absorption by powder and solubilization of particles (i.e., swelling and dispersion stages).

458

11. Effects of drying and storage on milk proteins

TABLE 11.6 Reconstitution period, insolubility index, and rehydration time of dairy protein powders Powders

RP using NMR (min)

ISI using IDF standard (mL)

WT (min)

ST (min)

DT + SolT (min)

RT (min)

MCP G

22

14.5

1

2

804

807

MCP NG

8

3.5

3

17

551

571

MCP + carbohydrate, G

18

5.0

1

nm

nm

116

MCP + carbohydrate, NG

nm

nm

2

0

95

97

MCP + NaCl, G

9.5

0.9

nm

nm

nm

nm

MCP + CaCl2, G

∞

14.5

nm

nm

nm

nm

MCP + SCS / SPS, G

6 /5

α-lactalbumin > αs-caseins ¼ κ-casein > β-lactoglobulin (Mulvihill and Fox, 1989). Once a protein is adsorbed at an interface, it undergoes unfolding and rearrangement to form a stabilizing adsorbed layer (Dickinson, 1992; Dalgleish, 1996), and the extent of unfolding depends on the flexibility of the protein molecule, that is, on the strength of the forces maintaining the secondary and tertiary structures. Because the caseins have rather flexible structures, they unfold rapidly at the interface and may form extended layers up to about 10 nm thick (Dalgleish, 1990). Dalgleish (1999) suggested that casein molecules are stretched to their maximum extent when their overall surface coverage is less than about 1 mg/m2. Conversely, the presence of excess casein increases the monolayer coverage to a maximum value of 3 mg/m2, the parts of the molecules in contact with the interface adopt a more compact conformation, and the hydrophilic moieties protrude further from the interface. Whey proteins (such as β-lactoglobulin), which give adsorbed layers that are only about 2 nm thick, change conformation and unfold their structure to some extent at the surface (Dalgleish and Leaver, 1993; Mackie et al., 1993; Dalgleish, 1995, 1996; Dickinson and McClements, 1995; Fang and Dalgleish, 1998). The adsorbed whey protein structure lies somewhere in between the native structure and the fully denatured state, which may have a native-like secondary structure and an unfolded tertiary structure (Dickinson, 1998a). Additionally, the partial unfolding of the globular whey protein structure following adsorption causes exposure of the reactive sulfhydryl group, leading to slow polymerization of the adsorbed protein in the aged layer via sulfhydryl-disulfide interchange (Dickinson and Matsumura, 1991; McClements et al., 1993). The amount of protein adsorbed on the interface of an emulsion droplet suggests the state of the protein adsorbed at the interface. If the protein load is 20% degree of hydrolysis), because of the production of many short peptides, has been found to be detrimental to the emulsifying and stabilizing properties of whey proteins (Singh and Dalgleish, 1998). The main form of instability in emulsions formed with highly hydrolyzed whey proteins is the coalescence that arises because of the inability of the predominantly short peptides to adequately stabilize the large oil surface generated during homogenization (Agboola et al., 1998a,b; Singh and Dalgleish, 1998). Nevertheless, it seems to be possible to make a fairly stable emulsion using highly hydrolyzed whey proteins at high peptide concentrations (protein-to-oil ratio about 1:1) and at low homogenization pressures, as the sole emulsifier (Agboola et al., 1998a,b). Under these conditions, there is a sufficient amount of high-molecular weight peptides (>5000 Da) in the emulsion to cover and stabilize the emulsion droplets. The addition of calcium or magnesium at above 20 mM has been shown to reduce the stability of emulsions formed with whey protein hydrolyzates (Ramkumar et al., 2000). This instability arises mainly from the binding of calcium to the adsorbed peptides, leading to a reduction in the charge density at the droplet surface, which would reduce the interdroplet repulsion and enhance the likelihood of droplet flocculation. The formation of calcium bridges between peptides present on two different emulsion droplets would also enhance flocculation. All these results confirm that, under a given set of homogenization conditions, the surface composition is largely dependent on the protein-to-oil ratio and the aggregation state of the proteins in solution. It appears that the structure of the interfacial layer in emulsions can be manipulated by controlling the protein concentration, the protein type, and the ionic environment. Because of their different interfacial structures, these droplets would be expected to exhibit different reactivities, which could be exploited to develop new food textures. Further studies are required for an understanding of the relationship between droplet surface structures and the sensitivity of the droplets to different environments and processing conditions.

476

12. Interactions and functionality of milk proteins in food emulsions

Recently, Ye et al. (2013) developed a novel oil-in-water emulsion system, which was stabilized by nanoscale emulsion droplets that were coated with casein micelles. Firstly, the nanoemulsion droplets (average size about 150 nm) were prepared by mixing oil (n-hexadecane) and MPC in a microfluidizer. In the second step, these nanoemulsion droplets were mixed with another oil and then subjected to high shear mixing or homogenization. It was found that these nanodroplets had the ability to adsorb at the oil-water interface and generate stable emulsions; the size of the droplet-stabilized emulsions was dependent on the concentration of nanodroplets in the dispersions. Confocal micrographs clearly showed that the nanodroplets adsorbed at the interface of the droplets of the final emulsion system (Fig. 12.3A). In the emulsion system containing 2% nanodroplets, the emulsion droplets were uniform and large, although the surface of the droplets was not totally covered by the nanodroplets. In emulsions containing higher concentrations of nanodroplets, the emulsion droplets were apparently smaller in size, and more of the surface was covered by the nanodroplets. The nanodroplets formed an integral droplet surface layer, such that some interfacial droplets assembled with other droplets to stretch into the aqueous phase. Transmission electron microscopy showed the adsorption of the nanodroplets onto the surface of larger droplets of the final emulsion (Fig. 12.3B). The protein was clearly located between the interfaces of two different kinds of droplets. The nanodroplets appeared to be connected with each other to form a layer along the interface of the large droplet. It was interesting to note that the nanodroplets directly adsorbed onto the interface of the large droplet had changed in shape from a sphere to an ellipse or even a rectangle. The flattening of the nanodroplets became less pronounced toward the aqueous phase, and their shape gradually returned to a sphere. This assembly resulted in the formation of a multidroplet layer with a three-dimensional network of droplets on the interface surrounding the large droplet.

Stability of milk protein–based emulsions The term “emulsion stability” refers to the ability of an emulsion to resist any alteration in its properties over the timescale of observation (McClements, 1999, 2005; Dickinson, 2003). An emulsion is thermodynamically unstable as the free energy of mixing is positive because of the large interfacial area between the oil and the aqueous phase. Therefore, the kinetic stability, that is, the time period for which the emulsion is stable, is important (Damodaran, 1997; McClements, 1999, 2005; Dickinson, 2003). For instance, an emulsion can be considered to be “stable” if the inevitable process of separation has been slowed to an extent that it is not of practical importance during the shelf life of the product. An emulsion may become unstable because of a number of different types of physical and chemical processes. Physical instability refers to the change in spatial arrangement or size distribution of emulsion droplets, such as creaming, flocculation, or coalescence, whereas chemical instability includes change in the composition of the emulsion droplet itself, such as oxidation and hydrolysis (McClements and Decker, 2000; McClements, 2005). Creaming is the movement of oil droplets, under gravity or in a centrifuge, to form a concentrated layer at the top of an oilin-water emulsion sample, with no accompanying change in the droplet size distribution.

Stability of milk protein–based emulsions

477

FIG. 12.3 (A) Confocal micrographs of individual emulsion droplets formed with various concentrations of casein micelle–coated nanodroplets: (a) 2%, (b) 6%, (c) 12%, and (d) 16%. The arrow indicates a surface without coating by nanodroplets. (B) Transmission electron micrographs of droplets on the interfacial layer of emulsions stabilized with 16wt% casein micelle–coated nanodroplets. Reproduced from Ye, A., Zhu, X., Singh, H., 2013. Oil-in-water emulsion system stabilized by protein-coated nanoemulsion droplets. Langmuir 29(47), 14403–14410, with permission from ACS Publications.

Creaming is reversible, and the original uniform distribution of droplets can usually be obtained by gentle mixing. The creaming process can be explained by Stokes’ law (Hunter, 1986; McClements, 2005): νstokes ¼

2γ 2 ðρ1  ρ2 Þ 9η

(12.1)

where νstokes is the velocity of creaming; γ is the emulsion droplet radius; ρ1 and ρ2 are the density of the continuous phase and the dispersed phase, respectively; and η is the shear

478

12. Interactions and functionality of milk proteins in food emulsions

viscosity of the continuous phase. The creaming rate can be reduced by lowering the droplet radius, increasing the continuous phase viscosity, or decreasing the difference in density between the two phases. However, this law often fails to define the rate of creaming caused by flocculation or coalescence. Coalescence, that is, an increase in droplet size by accretion, gradually results in the separation of the oil and the aqueous phase and is always irreversible. Coalescence requires rupture of the stabilizing film at the oil-water interface, but this occurs only when the layer of continuous phase between the droplets has thinned to a certain critical thickness (Dickinson and Stainsby, 1988; Britten and Giroux, 1991; Walstra, 1993). Flocculation has been defined as the reversible aggregation mechanism that arises when droplets associate as a result of unbalanced attractive and repulsive forces (Dalgleish, 1997). Generally, two types of flocculation are distinguished, that is, depletion flocculation and bridging flocculation (Dickinson, 2003). The type of mechanism prevailing depends upon the interaction between the interfacial layer and the emulsion droplets. Bridging flocculation normally occurs when a high-molecular weight biopolymer at a significantly low concentration adsorbs to two or more emulsion droplets, forming bridges (Dickinson, 1998b, 2003; McClements, 1999, 2005; Fellows and Doherty, 2006). Depletion flocculation occurs as a result of the presence of unadsorbing biopolymer in the continuous phase, which can promote association of the oil droplets by inducing an osmotic pressure gradient within the continuous phase surrounding the droplets (de Hek and Vrij, 1981; Dickinson, 1999; Tuinier and de Kruif, 1999; McClements, 2005). Essentially, if the added biopolymer is either unadsorbed or poorly adsorbed, the biopolymer is squeezed out of the area between two approaching emulsion droplets. The concentration of biopolymer between the emulsion droplets becomes less than its overall solution concentration, resulting in an osmotic imbalance. The net effect is that the particles are attracted toward each other, resulting in flocculation. The attraction energy is determined by the concentration of the polymer, and the range of interaction depends on the radius of gyration of the polymer molecule. The bonds formed through the depletion flocculation mechanism are generally weak, flexible, and reversible. The ability of proteins to stabilize emulsions is the most important criterion besides the emulsion formation in most food applications. The forces involved in stabilizing and destabilizing emulsions include van der Waals attractive forces, electrostatic interactions, and steric factors. At pH values away from their isoelectric point, as proteins are electrically charged, there is an electrostatic repulsion, which prevents dispersed droplets from closely approaching one another. With the possible exception of highly charged proteins, a predominant contribution to emulsion stabilization by protein comes from the steric stabilization mechanism. Interactions between droplets stabilized by proteins may be influenced by the presence of certain ions, particularly calcium, as proteins are capable of binding ions. As long as sufficient protein is present during homogenization to cover the oil droplets, emulsions stabilized by milk proteins are generally very stable to coalescence over prolonged storage. However, these emulsions are susceptible to different types of flocculation, which in turn leads to enhanced creaming or serum separation. At low protein-to-oil ratios, there is insufficient protein to fully cover the oil-water interface during homogenization, and this results in bridging flocculation. Another consequence of insufficient protein is coalescence of droplets during or immediately after emulsion formation. Bridging flocculation is commonly observed

479

Stability of milk protein–based emulsions

in emulsions formed with aggregated milk protein products, such as calcium caseinate or MPC, in which the droplets are bridged by casein aggregates or micelles. Optimum stability can generally be attained at protein concentrations high enough to allow full saturation coverage at the oil-water interface. However, at very high protein-to-oil ratios, the presence of excess, unadsorbed protein may lead to depletion flocculation in some emulsions. Both depletion flocculation and bridging flocculation cause an emulsion to cream more rapidly. Depletion flocculation has been observed in sodium caseinate–based emulsions but not in emulsions formed with calcium caseinate, MPC, or whey proteins (Dickinson and Golding, 1997; Euston and Hirst, 1999; Srinivasan et al., 2001; Singh, 2005) (Fig. 12.4). In sodium caseinate–based emulsions, it was shown that, at a protein content of nearly 2.0 wt%, the emulsion droplets were protected from flocculation by a thick steric-stabilizing layer of sodium caseinate. The emulsion was stable against flocculation, coalescence, and creaming for several weeks. However, when the protein content was increased to above 3.0 wt%, unadsorbed protein gave rise to depletion flocculation. Because of this depletion flocculation, the effective diameter of the droplets increased, resulting in a marked decrease in creaming stability with an increase in the caseinate concentration from 3 to 5 wt%. Further increasing the protein content to 6.0 wt% and above resulted in very high depletion flocculation, leading to a strong emulsion droplet network that was stable to creaming. The differences in the creaming stabilities of emulsions made with different kinds of milk protein products are largely related to depletion flocculation effects (Singh, 2005). The depletion interaction free energy (ΔGDEP), of the order of a few kT, can be estimated using Eq. (12.2) (Walstra, 1993): Y ΔGDEP ¼ 2πγ 2m ðγ d  2γ m =3Þ (12.2)

(3%)

(0.5%)

80

Na-CN

Stability rating (%)

70 60 50 40

Na-CN

30 20 10 0 0

WPC

1 2 3 4 Protein concentration (%)

5

WPC

FIG. 12.4 Creaming stability and microstructure of emulsions made with (●) sodium caseinate or (▪) WPC (30% oil). The scale bar represents 10 μm. Reproduced from Singh, H., 2005. Milk protein functionality in food colloids. In: Dickinson, E. (Ed.), Food Colloids: Interactions, Microstructure and Processing. The Royal Society of Chemistry, Cambridge, UK, pp. 179–193, with permission from The Royal Society of Chemistry.

480

12. Interactions and functionality of milk proteins in food emulsions

where Π is the osmotic pressure of the polymer solution, represented as a fluid of hard spheres of radius γ m, and γ d is the mean droplet radius. The osmotic pressure under ideal conditions is given by the following equation: Y ¼ cRT=M (12.3) where R is the molar gas constant, T is the temperature, M is the molecular mass of the polymer, and c is the number concentration of the polymer. For depletion flocculation to occur, the polymer has to have a fairly high M, so that the γ m is relatively large. However, at a given c, M is inversely proportional to Π. Therefore, an increase in the polymer molecular mass will reduce the osmotic pressure driving the depletion interaction. Hence, at a given concentration, the depletion interaction free energy is low for a polymer of low molecular mass, increases with an increase in molecular mass until it reaches a maximum, and then decreases with a further increase in molecular mass. Similarly, a reduction in the polymer number concentration will reduce the osmotic pressure. Although the exact state of the casein molecules in concentrated sodium caseinate solutions is unknown, a sodium caseinate solution has been reported to have a radius of gyration of about 20–30 nm, as determined by static light scattering (Lucey et al., 2000). It is likely that depletion flocculation in sodium caseinate emulsions is caused by the presence of these casein aggregates formed from self-assembly of sodium caseinate in the aqueous phase of the emulsion at concentrations above 2 wt% (Dickinson and Golding, 1998). Theoretically, a casein aggregate with a radius of approximately 20 nm causes the strongest depletion flocculation of fine emulsion droplets (mean diameter 0.4 μm), that is, corresponding to a size ratio of about 10:1 (Radford et al., 2004). The estimated optimum size of casein particles for inducing depletion flocculation is similar to the size of the small casein aggregates actually found in sodium caseinate dispersions at low ionic strength (Lucey et al., 2000). Emulsions formed with whey proteins, MPC, and calcium caseinate do not show depletion flocculation, probably because there are no suitably sized protein particles at the required concentrations in the aqueous phase. The molecular size of whey proteins is less than the optimum, whereas the casein micelles in MPC are too large to induce depletion flocculation. Calcium caseinate consists of mixtures of casein aggregates of different sizes, but the concentration of aggregates capable of inducing depletion flocculation is probably too low. The extent of creaming in these emulsions is largely determined by the particle size of the droplets. Generally, in these emulsion systems, the creaming stability increases with increasing protein concentration up to a certain concentration and then remains almost constant (Euston and Hirst, 1999; Srinivasan et al., 2001). However, the creaming stability of emulsions formed with calcium caseinate or MPC at relatively high protein concentrations tends to be higher than that of whey protein–stabilized emulsions. This can be attributed to an increase in the droplet density as a result of the presence of a much thicker and denser adsorbed protein layer at the droplet surface. The addition of moderate amounts of CaCl2 to emulsions containing excess sodium caseinate has been shown to eliminate depletion flocculation and to improve the creaming stability (Ye and Singh, 2001). This effect appears to be due to an increase in the average size of the casein aggregates in the aqueous phase, resulting in a large increase in the effective molecular mass of the caseins (Dickinson et al., 2001). In addition, there is a reduction in the concentration of unadsorbed caseinate. Both these effects are expected to cause a substantial reduction

Process-induced changes in milk protein–based emulsions

481

in the concentration of small particles, which are assumed to be the main depleting species responsible for inducing reversible flocculation in the calcium-free systems. In contrast, depletion of calcium from MPC causes dissociation of casein micelles, resulting in much smaller casein aggregates, which induce depletion flocculation in MPC-stabilized emulsions (Ye, 2011). Presumably, the substantial reduction in osmotic pressure makes the magnitude of ΔGDEP predicted from Eq. (12.2) too small to cause depletion flocculation. Similarly, the addition of NaCl at above a certain concentration reduces the extent of depletion flocculation of sodium caseinate emulsions and improves their creaming stability (Srinivasan et al., 2000). This effect is due to increased adsorption of protein at the droplet surface and hence a lower concentration of unadsorbed protein remaining in the solution. Decreasing the pH of emulsions formed with excess sodium caseinate also gradually eliminates depletion flocculation, through aggregation of adsorbed protein and a transfer of more protein to the droplet surface (Singh, 2005). Therefore, it seems to be possible to switch depletion flocculation off and on by controlling the concentration and the aggregation state of the casein molecules in the aqueous phase. It has been reported that the multilayered protein emulsions are more stable against various environmental conditions, such as Ca2+, high ionic strength, and heat treatment, than the standard protein emulsions (Schmelz et al., 2011; Ye et al., 2012). For example, the addition of Ca2+ to casein- or whey protein–stabilized emulsions at neutral pH causes droplet aggregation (Agboola and Dalgleish, 1995; Dickinson and Davies, 1999; Ye and Singh, 2000, 2001). The addition of lactoferrin to caseinate- or whey protein–stabilized emulsions results in the association of lactoferrin with adsorbed caseins or whey proteins via electrostatic interactions (Ye et al., 2012). This multisurface layer significantly reduces the calcium-induced destabilization of the emulsions, even when a small amount of lactoferrin is involved in the surface layer. Steric repulsion interactions produced by the large lactoferrin molecules on the surface were considered to contribute to this stabilizing effect (Ye et al., 2012).

Process-induced changes in milk protein–based emulsions Liquid food emulsions are often subjected to heat treatment at relatively high temperatures (e.g., retort or ultrahigh-temperature processing) to provide a long shelf life to the product via microbial sterility. These heat treatments can cause denaturation and aggregation of adsorbed and unadsorbed proteins, resulting in aggregation or coalescence of droplets and gel formation (Liang et al., 2017). In some cases, emulsion destabilization may be desirable if the final application is in a gelled food product, whereas, in other cases, such as in beverages, dispersion stability cannot be compromised. Emulsions formed with whey proteins at neutral pH are considered to be stable against heating when the ionic strength and/or the concentration of protein in the emulsions are low. High ionic strength (e.g., addition of KCl at 100 mM or above) has been shown to cause destabilization of whey protein emulsions, leading to gel formation (Hunt and Dalgleish, 1995). Both unadsorbed protein and adsorbed protein are necessary for the heat-induced aggregation of whey protein–stabilized emulsions. Aggregation of emulsion droplets is more extensive and proceeds more rapidly as the concentration of protein in the emulsion is

482

12. Interactions and functionality of milk proteins in food emulsions

increased, whereas the removal of unadsorbed protein from the emulsion decreases the rate of droplet aggregation (Euston et al., 2000). During heat treatment, the protein-covered droplet appears to interact more readily with the unadsorbed protein than with another emulsion droplet. This has been explained by assuming that the relative surface hydrophobicities of the emulsion droplet and the unadsorbed denatured whey proteins are different. Interaction of two emulsion droplets through their respective adsorbed protein layers will have a relatively low probability because the surface hydrophobicity is likely to be relatively low. When an emulsion droplet and an unadsorbed protein molecule aggregate, at least one of them (the denatured protein molecule) has a relatively high surface hydrophobicity, and this will increase the probability of interaction and aggregation (Euston et al., 2000). In emulsions made with 3.0% WPI and 25% soya oil, the amount of adsorbed protein was shown to increase from 2.9 to 3.7 mg/m2 within the first 10 min of heating at 75°C, but further heating had no effect (Sliwinski et al., 2003). At 90°C, a plateau value of about 4 mg/m2 was reached within 5 min of heating. Studies on the effects of heating temperature in the range of 50–90°C on WPI emulsions (pH 7.0) (Monahan et al., 1996; Demetriades and McClements, 1998) show that droplet aggregation occurs on heating in the range of 75–80°C, which causes an increase in viscosity and a loss of creaming stability, but the degree of aggregation and the susceptibility to creaming decrease on heating at temperatures above 80°C. It has been suggested that, in the temperature range of 75–80°C, the whey protein molecules at the droplet surface are only partly unfolded and that not all of the hydrophobic amino acid residues are directed toward the oil phase. Consequently, the surface of the droplet is more hydrophobic, making it susceptible to droplet aggregation. At higher temperatures, the proteins become fully unfolded with all of the hydrophobic residues being directed into the oil phase, which makes the droplets less prone to aggregation. The role of sulfhydryl-disulfide interchange reactions in droplet aggregation is not clear. It has been suggested that disulfidemediated interactions during heat treatment are not critical during the initial stages of aggregation but that they tend to strengthen the aggregates (Demetriades and McClements, 1998). Dickinson and Parkinson (2004) and Parkinson and Dickinson (2004) reported that the addition of a very small proportion of casein (0.03%–0.15% of the total emulsion) can stabilize a whey protein–based emulsion against heat treatment. The magnitude of the effect is dependent on the type of casein, with the order of effectiveness being β-casein > sodium caseinate > αs1-casein. The stabilizing effect of the casein in these mixed milk protein systems is strongly synergistic. The casein polymer appears to act in a colloidal stabilizing capacity at a surface concentration that is very much lower than that at which it could be used as an emulsifying or stabilizing agent simply on its own. It has been suggested that adsorbed casein molecules keep the emulsion droplet surfaces sufficiently far apart to prevent the “normal” crosslinking processes that occur between whey protein–coated droplets during heat-induced aggregation and gelation, because of the steric hindrance from the loops and tails of the disordered casein polymers (Parkinson and Dickinson, 2004). In contrast to whey proteins, emulsions formed with sodium caseinate (2 wt% protein and 20% soya oil) are stable to heating at 90°C for 30 min or 121°C for 15 min, as determined by droplet size analysis (Hunt and Dalgleish, 1995; Srinivasan et al., 2002). However, the protein coverage and the adsorbed casein composition change upon heat treatment, indicating that interactions between unadsorbed caseinate molecules and caseinate at the droplet surface may occur during heating (Srinivasan et al., 2002). Analysis of adsorbed caseins isolated from

Process-induced changes in milk protein–based emulsions

483

emulsions heated at 121°C for 15 min has shown that a substantial proportion of the adsorbed caseinate is polymerized to form high-molecular weight aggregates (Srinivasan et al., 2002), which are held together through covalent bonds other than disulfide bonds. These covalent bonds appear to form mainly between caseinate molecules at the surface of the same droplet because of the higher local concentrations of casein molecules at the droplet surface. Interestingly, the adsorbed caseins also appear to undergo thermal degradation, resulting in the formation of low-molecular weight species. Relatively high proportions of casein degradation products present at the droplet surface indicate that the adsorbed caseinate molecules are more susceptible to fragmentation during heating than those in solution and that these peptides remain adsorbed. This is probably due to different structures and conformations of the caseins at the droplet surface compared with those in the solution. The creaming stability of sodium caseinate emulsions has been found to improve upon heating, with the onset of depletion flocculation occurring at higher protein concentration than in unheated emulsions (Srinivasan et al., 2002). This can be attributed to a reduction in the number of unadsorbed caseinate molecules/aggregates in the aqueous phase as a result of increased surface coverage and heat-induced polymerization and degradation of the casein molecules. The improvement in the creaming stability in heated emulsions at low protein concentrations may be attributed to an increase in droplet density because of the presence of greater amounts of polymerized protein at the droplet surface. The surface protein composition of emulsion droplets may also change during heat treatment in emulsions formed with whey proteins. For WPI-stabilized emulsions, the amount of β-lactoglobulin at the droplet surface was found to increase during heat treatment, whereas the amount of adsorbed α-lactalbumin decreased markedly (Ye and Singh, 2006a; Ye, 2010). It seems that β-lactoglobulin displaces α-lactalbumin from the interface on heating at temperatures up to 90°C, but the reason for this is not clear. Similar phenomena were observed in studies of exchanges of caseins and whey proteins at the interfaces of oil-in-water emulsion droplets (Dalgleish et al., 2002). It was found that, at temperatures above 40°C, the addition of WPI to the aqueous phase of caseinate-stabilized emulsions caused a displacement of adsorbed caseins. As the β-lactoglobulin and α-lactalbumin adsorbed, the αs1- and β-caseins were desorbed, principally the former, whereas the αs2- and κ-caseins were not displaced. The rate of the displacement or exchange reaction was temperature dependent, being almost undetectable at room temperature, but complete within 2 min at 80°C. The displacement reaction was not affected by ionic strength; neither were any of the reactions apparently dependent on sulfhydryl exchange reactions (Dalgleish et al., 2002). However, no exchange of proteins occurred when an emulsion prepared with WPI was treated with caseinate and heat treated at 80°C for 2 min (Brun and Dalgleish, 1999). This was surprising in view of the known interactions of whey proteins with αs2- and κ-caseins that involve sulfhydryl-disulfide interchange reactions. Liang et al. (2013) investigated the effect of heat treatment (120°C for 10 min) at neutral pH on the physical stability of MPC-stabilized emulsions. Emulsions made with unheated MPC solutions had significantly larger particle sizes than emulsions made using preheated MPC solutions (95°C for 5 min). Interestingly, in both emulsion systems, some droplet clusters were formed by extensive covalent protein cross-linking, and these clusters could not be dissolved by dissociating agents.

484

12. Interactions and functionality of milk proteins in food emulsions

Heat treatment of emulsions stabilized by highly hydrolyzed whey proteins at 121°C for 16 min resulted in destabilization of the emulsions, which appeared to occur mainly via a coalescence mechanism (Agboola et al., 1998b). As the adsorbed peptide layers in these emulsions lacked the cohesiveness of the parent proteins and had poor ability to provide steric or charge stabilization, increased collisions between the droplets during heating would cause droplet aggregation, leading to coalescence. It is also possible that desorption of some loosely adsorbed peptides occurred during heating, as indicated by the decrease in the amount of peptides associated with the oil surface after heating, which would also enhance coalescence. The effect of ultrahigh pressure (100–1000 MPa) on the structures of milk proteins in aqueous solution has received considerable attention over many years (see Chapter 8). High pressure can disrupt the quaternary and tertiary structures of globular proteins with relatively little influence on their secondary structure. In addition, the proteinaceous colloidal aggregates (e.g., casein micelles), which are held together by ionic and hydrophobic interactions, can be dissociated by high-pressure treatment (Gaucheron et al., 1997; Huppertz et al., 2004). Whey proteins are sensitive to high-pressure treatments (Lo´pez-Fandino´ et al., 1996; Anema et al., 2005). Solution studies (Patel et al., 2005) of native β-lactoglobulin and whey protein products have shown that high-pressure treatment has a marked effect on the protein’s conformation and consequently its aggregation behavior; the aggregation is more extensive at high protein concentrations (Patel et al., 2005). The formation of aggregates is most probably due to the generation of intermolecular disulfide bridges through sulfhydryl-disulfide interchange reactions (Patel et al., 2006). In model oil-in-water emulsions, high-pressure treatment has been shown to have no effect on the droplet size distribution or the emulsion viscosity of sodium caseinate–based emulsions at pH 7 (Dumay et al., 1996). However, high-pressure treatment significantly induced flocculation of emulsion droplets and increased the emulsion viscosity of oil-in-water emulsions stabilized by β-lactoglobulin or WPC at neutral pH (Dumay et al., 1996; Dickinson and James, 1998). The unfolded unadsorbed whey proteins in the emulsion treated by high pressure appeared to be the major contributor to the cross-linking or flocculation of emulsion droplets because greater emulsion flocculation was observed in emulsions with higher proportions of unadsorbed protein in the aqueous phase. As in the case of emulsions treated by heat processing, whey protein–stabilized emulsions are more sensitive to pressure and temperature at pH values closer to the isoelectric point and at high ionic strength. In terms of the change in emulsion rheology, severe high-pressure treatment (800 MPa for 60 min) is equivalent to relatively mild thermal treatment (65°C for 5 min) (Dickinson and James, 1998). In a concentrated emulsion formed with β-lactoglobulin (1% protein and 40 vol% n-tetradecane), an emulsion gel was produced following high-pressure treatment. When β-lactoglobulin or WPC solution was treated by high pressure before emulsion formation, the emulsions had larger droplet sizes than emulsions made with the native protein (Galazka et al., 1995). The results indicated a modification of protein structure, leading to the loss of emulsifying efficiency as a result of protein aggregation, despite an increase in surface hydrophobicity. After adsorption on the surface, the protein probably became partially unfolded at the interface, and subsequent pressure treatment caused no further conformational change. No studies on the behavior of emulsions formed with aggregated milk proteins, such as micellar casein, upon high-pressure treatment have been reported.

Process-induced changes in milk protein–based emulsions

485

Lipid oxidation in milk protein–based emulsions In addition to the physical properties of emulsions in foods, lipid oxidation is one of the major issues in food storage and consumption; it greatly influences the flavor, odor, and color of foods. Similar to the physical properties of emulsions, lipid oxidation in oil-in-water emulsions is influenced by the droplet size, the interfacial characteristics of the lipid droplets, and the type of emulsifying agent (Dickinson and Stainsby, 1982; McClements and Decker, 2000). In addition to their remarkable emulsifying properties, both WPI and caseinate have shown to inhibit the oxidative deterioration of unsaturated fatty acids, either as part of triacylglycerols or in free form (Hu et al., 2003; Djordjevic et al., 2004; Kiokias et al., 2006; Ries et al., 2010). WPI and caseinate therefore appear to be useful for the design of emulsions that serve as delivery systems for omega-3 fatty acids because of their dual functionality as emulsifiers and antioxidants (Singh et al., 2006). Such emulsions may be incorporated into real food emulsion systems, for example, milk, yogurt, mayonnaise, ice cream, and cheese (Ye et al., 2009). Ries et al. (2010) evaluated the oxidative stability of emulsions made with various milk protein products and linoleic acid by determining the formation of lipid hydroperoxides and hexanal. The oxidative stability of both WPI- and sodium caseinate–stabilized linoleic acid emulsions with smaller droplet size was greater than that of emulsions with larger droplet size. Other studies have reported contradictory results; some found greater lipid oxidation in emulsions with small droplets (Gohtani et al., 1999; Jacobsen et al., 2000; Lethuaut et al., 2002), whereas others found greater lipid oxidation in emulsions with large droplets (Nakaya et al., 2005; Let et al., 2007). Caseinate appears to be a better antioxidant than WPI in emulsions (Hu et al., 2003; Djordjevic et al., 2004; Kiokias et al., 2006; Ries et al., 2010). The inhibition of lipid oxidation by proteins in emulsions is considered to be mostly due to metal ion chelation and free radical scavenging (Benjelloun et al., 1991). In general, the specific antioxidative activity of caseinate appears to be due to its chelating capacity because of its phosphoseryl groups (Gaucheron et al., 1996; Bennett et al., 2000; Sugiarto et al., 2010), and that of whey protein appears to be due to its free radical–scavenging activity because of its free sulfhydryl groups (McClements and Decker 2000; Tong et al., 2000; Hu et al., 2003; Kiokias et al., 2007). As caseins do not possess any free sulfhydryl group, their free radical–scavenging activity would be expected to be lower than that of whey proteins. In contrast, whey proteins have a limited ability to chelate metal ions, because of their lack of phosphoseryl groups. However, phosphoseryl groups and free sulfhydryl groups do not contribute solely to the total antioxidative capacity of the respective protein. It has been shown that the dephosphorylation of αs1- and β-caseins only partially suppressed their antioxidative activity in a liposome system (Cervato et al., 1999) and the blocking of the sulfhydryl groups of whey protein with N-ethylmaleimide in aqueous solution reduced its free radical–scavenging activity by only 20% (Tong et al., 2000). Ries et al. (2010) reported that the extent of lipid oxidation decreased with an increase in the protein concentration. Furthermore, an increase in protein concentration led to a decrease in the difference in lipid hydroperoxide production between large- and small-droplet-sized emulsions. At high protein concentrations, the antioxidative effect of the protein in the emulsions appeared to offset the effects of emulsion droplet size and protein type. In addition to the physical barrier of the interfacial protein layer and the antioxidative effect of protein on the interface of emulsion droplets, unadsorbed protein in the continuous phase played an important role in the oxidative stability of emulsions.

486

12. Interactions and functionality of milk proteins in food emulsions

Experiments involving the replacement of the continuous phase of the emulsions with water or protein solution showed that, compared with the control emulsion, the replacement of the continuous phase with water increased the production of lipid hydroperoxides. Replacement of the continuous phase with protein solution decreased the production of lipid hydroperoxides (Ries et al., 2010). In addition, the lipid hydroperoxide concentration was lower in the aqueous phase of emulsions containing caseinate than in the aqueous phase of emulsions containing WPI solution. Furthermore, it has been shown that the oxidative stability increased with increasing protein concentration in the continuous phase. This suggests that the antioxidative mechanism of protein at the interfacial region, such as binding trace metal ions from the lipid phase and free radical–scavenging activity, may involve a dynamic exchange process with protein molecules from the continuous phase. The antioxidative properties of the milk proteins are also influenced by the processing and environmental conditions, that is, heat treatment and change in the pH. When whey proteins that had been heated at temperatures higher than 80°C were added to fish oil emulsions, the oxidative stability of the fish oil improved significantly compared with the control samples, as assessed by lipid hydroperoxide formation and thiobarbituric acid reactive substances (Tong et al., 2000; Elias et al., 2007). It was suggested that the increased oxidative stability could have been due to a greater exposure of free radical–scavenging amino acid residues (e.g., tryptophan, tyrosine, phenylalanine, or methionine) on whey proteins and/or a greater interfacial contact of the protein because of increased hydrophobicity after heat treatment, both leading to improved effectiveness of the free radical–scavenging process. Recently Okubanjo et al. (2019), using the nanoemulsion droplet system developed by Ye et al. (2013), investigated the oxidation resistance of polyunsaturated fatty acid oil (PUFA oil) within droplet-stabilized emulsions, using shell lipids (nanoemulsion droplets) with a range of melting points: olive oil (low melting point), trimyristin (high melting point), and palm olein oil (intermediate melting point). The rate of oxidation of PUFA oil, monitored via conjugated dienes, lipid hydroperoxides, and hexanal levels, was found to be slower in droplet-stabilized emulsions than in conventional emulsions or control emulsions of the same composition as droplet-stabilized emulsions, and trimyristin gave the greatest oxidation resistance. It was suggested that the structured interface of droplet-stabilized emulsions limits contact between prooxidants and oxidation-sensitive bioactives encapsulated within, and this antioxidative effect is greatly enhanced with solid surface lipid.

Behavior of milk protein–stabilized emulsions under physiological conditions In recent years, there has been considerable research on the physicochemical and structural changes in food emulsions during oral and gastrointestinal processing. These studies have focused mainly on understanding the role of emulsion structure in the lipolysis of emulsified triacylglycerols, with a view to developing emulsion systems with a controlled rate of lipid digestion and the delivery of lipid-soluble nutrients. Several studies have shown that the ability of lipases to digest emulsified oil droplets is affected by the composition of the interfacial layer and the droplet size of the emulsions (see recent reviews, Singh, 2011; Singh and Ye, 2013). The process of digestion involves complex mechanical, physicochemical, and

Behavior of milk protein–stabilized emulsions under physiological conditions

487

physiological processes. Different parts of the human gastrointestinal tract, which includes the mouth, stomach, small intestine, and large intestine, work in a highly cooperative manner in the overall digestion and absorption of food. Here, we provide a brief overview of the basic physicochemical and physiological processes that occur during the digestion of milk protein– based emulsions. The behavior of milk protein–based emulsions in the oral cavity is largely driven by the interactions of saliva with the adsorbed layer on the emulsion droplets (van Aken et al., 2005; Vingerhoeds et al., 2008). Emulsions formed with WPI or sodium caseinate showed flocculation of droplets when mixed with human saliva. This flocculation was considered to be driven by depletion, van der Waals forces, and/or electrostatic interactions between emulsion droplets and salivary mucins and was largely dependent on the initial charge of the emulsion droplets (Vingerhoeds et al., 2005; Silletti et al., 2007a,b; Sarkar et al., 2009). For example, negatively charged protein-stabilized emulsions, that is, β-lactoglobulin emulsions at neutral pH, did not interact with the artificial saliva because of strong repulsive forces between anionic mucin and the anionic β-lactoglobulin interfacial layer at neutral pH but underwent depletion flocculation on the addition of higher concentrations of mucin (1.0 wt%). In contrast, positively charged lactoferrin-stabilized emulsions interacted with mucin via electrostatic interaction and resulted in the formation of a secondary layer around the lactoferrin-stabilized droplets, with some bridging-type flocculation (Sarkar et al., 2009). These interactions could have an impact on sensorial and textural perceptions of food emulsions in vivo. The biochemical conditions prevailing in the stomach have a major impact on the structure and stability of protein-based emulsions. The stomach has a highly acidic pH (between 1 and 3 for a fasted stomach) and contains various minerals and both proteolytic and lipolytic enzymes. There is also some mechanical agitation because of peristalsis in the stomach (Ekmekcioglu, 2002; Kalantzi et al., 2006; Pal et al., 2007). As most protein-based emulsions are negatively charged at neutral pH, the decrease in the pH to below 2.0 causes substantial changes in the emulsion droplet charge and some droplet aggregation around the isoelectric point of the protein. The action of pepsin is most critical as it causes major changes in the adsorbed protein layers and thus the droplet characteristics, affecting the stability of the emulsion and the digestibility of its components. Because of their highly disordered structures, caseins undergo rapid hydrolysis by pepsin, but β-lactoglobulin shows some resistance to gastric digestion because of its highly folded conformation (Reddy et al., 1988; Schmidt and van Markwijk, 1993). However, the rate of hydrolysis of β-lactoglobulin by pepsin increases when this protein is present as the adsorbed layer in an emulsion (Macierzanka et al., 2009; Sarkar et al., 2009), possibly because of the change in the conformation of the β-lactoglobulin molecules upon adsorption at the oil-water interface (Macierzanka et al., 2009). Surprisingly, adsorbed α-lactalbumin in oil-in-water emulsions appears to be more resistant to hydrolysis by pepsin, compared with native α-lactalbumin in solution (Nik et al., 2010). Casein and bovine serum albumin adsorbed at the surface of emulsion droplets have been shown to be readily hydrolyzed by pepsin after mixing with the gastric fluid (Li et al., 2012; Kenmogne-Domguia et al., 2013). Because of the hydrolysis of interfacial protein by pepsin (Macierzanka et al., 2009; Sarkar et al., 2009; Nik et al., 2010; Li et al., 2012; Kenmogne-Domguia et al., 2013), the emulsions stabilized by milk proteins (such as WPI, sodium caseinate, β-lactoglobulin, or β-casein)

488

12. Interactions and functionality of milk proteins in food emulsions

undergo flocculation and coalescence of the droplets. In WPI-stabilized emulsions, the presence of excess unadsorbed protein appears to significantly improve the stability of the oil droplets during gastric digestion (Nik et al., 2010). All the studies mentioned earlier have been carried out using static in vitro gastric digestion methods. Recently, Wang et al. (2019) compared the behaviors of different milk protein–based emulsions during digestion, using an in vitro dynamic gastric digestion model—the human gastric simulator (HGS). This HGS closely mimics many relevant features of gastric physiology, such as progressive acidification, peristaltic movements, mixing dynamics, and emptying rates (Kong and Singh, 2008). Emulsions containing 4.0% (w/w) protein and 20.0% soya oil were prepared using MPC, calcium-depleted MPC, or sodium caseinate as the protein source. The MPC-stabilized emulsion showed extensive flocculation of the droplets within 20 min of digestion (at pH > 6) in the HGS, leading to a flocculated network of oil droplets; some of the droplets within this network appeared to coalesce to form larger droplets (Fig. 12.5). This droplet flocculation was due to destabilization of adsorbed casein micelles (from the MPC) by the proteolytic action of pepsin on κ-casein. As a result, casein micelles adsorbed onto different droplets began to interact with each other and with the unadsorbed casein micelles, resulting in a cohesive casein network containing entrapped oil droplets. As expected, the structure of the flocculated network changed with an increase in the digestion time in the HGS, with the network becoming more open and porous, with more voids. However, some flocs were still visible at the end of gastric hydrolysis at 220 min. The proportion of coalesced oil droplets within the network appeared to increase with the digestion time (Fig. 12.5), because of more extensive hydrolysis of the adsorbed protein layers by pepsin. The remaining peptides on the interface were unable to provide sufficient stabilization of the oil droplets, resulting in coalescence. Only a few small flocs were seen in the calcium-depleted-MPC-stabilized emulsion (Fig. 12.5, 0 min) after 20 min of digestion, but they became larger at 60 min; these flocs appeared to be more open and porous than those in the MPC-stabilized emulsions. This difference in the microstructures of emulsions made with normal MPC and calcium-depleted MPC was attributed to the presence of fewer intact casein micelles in calcium-depleted MPC. As discussed earlier, the casein micelles are dissociated to some extent by a reduction in the calcium content in the MPC, which results in lower surface protein concentrations and different compositions of the individual caseins at the droplet surface during emulsion formation. The sodium caseinate–stabilized emulsion showed considerable flocculation (Fig. 12.5, 0 min), because of depletion flocculation of the oil droplets, as discussed earlier. Interestingly, this flocculation disappeared during digestion (20 min), probably because of hydrolysis of the protein in the aqueous phase by pepsin, which would lead to a change in the size of the selfassociated casein particles in the aqueous phase, as discussed earlier. Droplet flocculation was seen again at a digestion time of 60 min, when the pH approached the isoelectric point of casein, possibly because of electrostatic interactions between the oil droplets. The cluster of flocs disappeared completely at a digestion time of 120 min, and at the end of the gastric digestion, numerous tiny, evenly dispersed flocs and some free oil droplets were visible. Flocculation of the emulsions in this system was driven mainly by low pH rather than by pepsin action (Wang et al., 2019). It is clear that the different types of emulsion structures formed in the gastric environment may lead to different rates of protein hydrolysis, release of lipids, and gastric emptying. It appears that the flocculated droplet structures formed during the gastric digestion of the

Behavior of milk protein–stabilized emulsions under physiological conditions

489

FIG. 12.5 Confocal micrographs of digestion residues of emulsions stabilized by different milk protein ingredients in an HGS at different times during gastric digestion. Samples were stained with Nile Red (for oil) and Fast Green (for protein). The scale bar in all images is 50 μm. Reproduced from Wang, X., Lin, Q., Ye, A., Han, J., Singh, H., 2019. Flocculation of oil-in-water emulsions stabilised by milk protein ingredients under gastric conditions: impact on in vitro intestinal lipid digestion. Food Hydrocoll. 88, 272–282, with permission from Elsevier Inc.

MPC-stabilized emulsion were disintegrated slowly compared with those of emulsions stabilized with other milk protein ingredients. This would translate into a lower gastric emptying rate and a delay in the delivery of oil droplets to the small intestine for further digestion. The emulsions stabilized by calcium-depleted MPC and sodium caseinate formed flocs with relatively smaller sizes, and they dissociated faster in the gastric environment. This would

490

12. Interactions and functionality of milk proteins in food emulsions

result in a faster stomach emptying rate and a more rapid delivery of oil droplets to the small intestine. It must be pointed out that the human stomach also contains gastric lipase, which is able to penetrate the adsorbed layer and act on the triglyceride core, preferentially cleaving at sn-3 ester bonds of the triglycerides. This lipolysis leads to the accumulation of protonated free fatty acids at the oil-water interface, which competitively displace the proteins and peptides from the interface (Armand et al., 1994, 1996; Pafumi et al., 2002). However, the effects of gastric lipase on the stability of protein-stabilized emulsions have not been studied in any detail, partly because of the nonavailability of commercial gastric lipase for in vitro experiments. Rabbit gastric lipase has recently become commercially available and has been suggested to a good substitute for human gastric lipase. Several other materials, such as mucins and phospholipids, that could alter the physicochemical properties of emulsions are present in the stomach. The small intestine is the main site for digestion and absorption; it contains various salts, pancreatic enzymes, coenzymes, bile salts, and phospholipids. The pH of the partly digested/ modified emulsions entering the small intestine increases (to between 6 and 7), because of the secretion of sodium bicarbonate, which allows pancreatic enzymes to act efficiently (Krondahl et al., 1997; Bauer et al., 2005). The changes in pH and ionic strength affect the stability of the protein-stabilized oil droplets via electrostatic interactions. The pancreatic proteases (i.e., trypsin and chymotrypsin) cause further hydrolysis of the adsorbed and unadsorbed proteins/peptides, although the mechanisms of the complex interactions of these proteases with the adsorbed proteins/peptides are not known. Pancreatic lipase adsorbs to the droplet interface, usually by a complexation with colipase and/or bile salts, and then cleaves triglycerides to form 2-monoglycerides and free fatty acids. Bile salts are highly surface active and are able to displace any protein or peptide material remaining on the droplet surface. Partial or complete displacement of protein from the droplet surface after the introduction of bile salts into the simulated intestinal fluid for emulsions formed with caseins, WPI, and lactoferrin has been demonstrated (Mun et al., 2007; Maldonado-Valderrama et al., 2008; Hur et al., 2009; Klinkerson and McClements, 2010; Sarkar et al., 2010; Torcello-Gomez et al., 2011). Whey proteins appeared to be more readily displaced than caseinates from the emulsion droplet surface (Mun et al., 2007). β-Lactoglobulin was rapidly displaced from the oil-water interface compared with lactoferrin, possibly because of a difference in droplet charges. Bile salts appeared to bind to the positively charged lactoferrin emulsion droplets, forming a mixed lactoferrin/bile salt interfacial layer (Sarkar et al., 2010). The exposure of protein-stabilized emulsions to in vitro intestinal conditions has been shown to cause coalescence of some droplets initially (Golding and Wooster, 2010; Sarkar et al., 2010), but all aggregated/flocculated droplets are broken down, resulting in uniform dispersions of small droplets. The negative charge imparted by adsorbing bile salts/other surface-active molecules is thought to provide electrostatic repulsions between the droplets and to prevent their further aggregation. It must be pointed out that lipid digestion is highly efficient, and over 95% of ingested lipids are absorbed by the healthy human gut. Thus, it may be very difficult to design protein-based emulsions to influence the total uptake of fatty acids. However, as discussed earlier, the kinetics of lipid hydrolysis in emulsions could be controlled to some extent through manipulating emulsion structures in the gastric phase, the interfacial layer structures, the oil droplet size, and the physical state of the lipids. The rate of fatty acid absorption

References

491

into the blood (i.e., postprandial triglyceride levels) is considered to be important for human health; high postprandial triglyceride levels are associated with the activation of various inflammatory pathways and are recognized as risk factors for cardiovascular disease and diabetes (Kolovou et al., 2005). Moreover, as the products of lipolysis (fatty acids and monoglycerides) are associated with signaling pathways for the control of food intake, emulsion design offers possibilities for developing foods to control satiety and weight management (Singh and Ye, 2013).

Conclusions Milk proteins in soluble and dispersed forms have excellent surface-active and emulsionstabilizing properties. Differences in the emulsifying abilities of milk proteins arise largely from differences in the structure, flexibility, state of aggregation, and composition of the proteins. These attributes of milk proteins (and hence their emulsifying abilities) are modified through various interactions that occur during the processing required to isolate the protein components and during the manufacture of prepared foods. Emulsions with different surface compositions and structures can be made using different kinds of milk proteins, and these emulsions exhibit different sensitivities to solution conditions, such as pH and ionic strength, and processing conditions, such as heat and high-pressure treatments. This could offer possibilities for the formation of emulsions with a range of functionalities for different food applications. Most of the research during the last 20 years has been performed on oil-in-water emulsions using purified or simple mixtures of caseins and whey proteins. A great deal of information is now available on the conformation of proteins at oil-water interfaces, competitive exchange reactions between adsorbed and unadsorbed proteins, protein-polysaccharide interactions, and factors controlling the rheology and stability of emulsions. In addition, some understanding of how processing conditions (heat treatments and high-pressure treatments) influence interfacial structures and emulsion properties has been achieved. There is much less understanding of the behavior of more complex mixtures of proteins in emulsions and of the stability behavior of emulsions under processing environments that are commonly encountered in the food industry. In addition, there is a lack of understanding of the behavior of emulsions during oral processing in the mouth and during digestion processes. It is critical to understand the oral behavior of emulsions, as common sensorial attributes (e.g., creaminess and smoothness) and the release of fat-soluble flavors are based on interfacial structures and rheological parameters. There is increasing evidence to show that the behavior of emulsions in the gastrointestinal tract is affected by their physicochemical properties and that the properties of the interface modulate fat digestion and consequently influence the bioavailability of lipid nutrients. This is the emerging area of emulsion science and the knowledge has the potential to contribute to the development of novel products with health and sensory attributes.

References Agboola, S.O., Dalgleish, D.G., 1995. Calcium-induced destabilization of oil-in-water emulsions stabilized by caseinate or by β-lactoglobulin. J. Food Sci. 60, 399–403. Agboola, S.O., Singh, H., Munro, P.A., Dalgleish, D.G., Singh, A.M., 1998a. Destabilization of oil-in-water emulsions formed using highly hydrolyzed whey proteins. J. Agric. Food Chem. 46, 84–90.

492

12. Interactions and functionality of milk proteins in food emulsions

Agboola, S.O., Singh, H., Munro, P.A., Dalgleish, D.G., Singh, A.M., 1998b. Stability of emulsions formed using whey protein hydrolysate: effects of lecithin addition and retorting. J. Agric. Food Chem. 46, 1814–1819. Anema, S.G., Stockmann, R., Lowe, E.K., 2005. Denaturation of β-lactoglobulin in pressure treated skim milk. J. Agric. Food Chem. 53, 7783–7791. Armand, M., Borel, P., Dubois, C., Senft, M., Peyrot, J., Salducci, J., 1994. Characterization of emulsions and lipolysis of dietary lipids in the human stomach. Am. J. Physiol. Gastrointest. Liver Physiol. 266, 372–381. Armand, M., Borel, P., Pasquier, B., Dubois, C., Senft, M., Andre, M., 1996. Physicochemical characteristics of emulsions during fat digestion in human stomach and duodenum. Am. J. Physiol. Gastrointest. Liver Physiol. 271, 172–183. Baker, E.N., Baker, H.M., 2005. Molecular structure, binding properties and dynamics of lactoferrin. Cell. Mol. Life Sci. 62, 2531–2539. Bauer, E., Jakob, S., Mosenthin, R., 2005. Principles of physiology of lipid digestion. Asian Austral. J. Anim. Sci. 18, 282–295. Baumy, J.J., Brule, G., 1988. Binding of bivalent cations to α-lactalbumin and β-lactoglobulin: effect of pH and ionic strength. Lait 68, 33–48. Benjelloun, B., Talou, T., Delmas, M., Gaset, A., 1991. Oxidation of rapeseed oil – effect of metal traces. J. Am. Oil Chem. Soc. 68 (3), 210–211. Bennett, T., Desmond, A., Harrington, M., McDonagh, D., FitzGerald, R., Flynn, A., Cashman, K.D., 2000. The effect of high intakes of casein and casein phosphopeptide on calcium absorption in the rat. Br. J. Nutr. 83 (6), 673–680. Britten, M., Giroux, H.J., 1991. Emulsifying properties of whey protein and casein composite blends. J. Dairy Sci. 74, 3318–3325. Brun, J.M., Dalgleish, D.G., 1999. Some effects of heat on the competitive adsorption of caseins and whey proteins in oil-in-water emulsions. Int. Dairy J. 9, 323–327. Cervato, G., Cazzola, R., Cestaro, B., 1999. Studies on the antioxidant activity of milk caseins. Int. J. Food Sci. Nutr. 50 (4), 291–296. Dalgleish, D.G., 1990. The conformation of proteins on solid/water interfaces—caseins and phosvitin on polystyrene lattices. Colloids Surf. 46, 141–155. Dalgleish, D.G., 1995. Structures and properties of adsorbed layers in emulsions containing milk proteins. In: Dickinson, E., Lorient, D. (Eds.), Food Macromolecules and Colloids. The Royal Society of Chemistry, Cambridge, UK, pp. 23–33. Dalgleish, D.G., 1996. Conformations and structures of milk proteins adsorbed to oil-water interfaces. Food Res. Int. 29, 541–547. Dalgleish, D.G., 1997. Adsorption of protein and the stability of emulsions. Trends Food Sci. Technol. 8, 1–6. Dalgleish, D.G., 1999. Interfacial structures and colloidal interactions in protein-stabilised emulsions. In: Dickinson, E., Rodriguez-Patino, J.M. (Eds.), Food Emulsions and Foams: Interfaces, Interactions and Stability. The Royal Society of Chemistry, Cambridge, UK, pp. 1–16. Dalgleish, D.G., Leaver, J., 1993. Dimensions and possible structures of milk proteins at oil-water interfaces. In: Dickinson, E., Walstra, P. (Eds.), Food Colloids and Polymers: Stability and Mechanical Properties. The Royal Society of Chemistry, Cambridge, UK, pp. 113–122. Dalgleish, D.G., Goff, H.D., Brun, J.M., Luan, B.B., 2002. Exchange reactions between whey proteins and caseins in heated soya oil-in-water emulsion systems – overall aspects of the reaction. Food Hydrocoll. 16, 303–311. Dam, B.V., Watts, K., Campbell, I., Lips, A., 1995. On the stability of milk protein-stabilized concentrated oil-in-water food emulsions. In: Dickinson, E., Lorient, D. (Eds.), Foods Macromolecules and Colloids. The Royal Society of Chemistry, Cambridge, UK, pp. 215–222. Damodaran, S., 1997. Food proteins: an overview. In: Damodaran, S. (Ed.), Food Proteins and Their Applications. CRC Press, London, UK, pp. 1–24. de Hek, H., Vrij, A., 1981. Interactions in mixtures of colloidal silica spheres and polystyrene molecules in cyclohexane. J. Colloid Interface Sci. 84, 409–422. Demetriades, K., McClements, D.J., 1998. Influence of pH and heating on physicochemical properties of whey proteinstabilized emulsions containing a nonionic surfactant. J. Agric. Food Chem. 46, 3936–3942. Dickinson, E., 1992. Structure and composition of adsorbed protein layers and the relationship to emulsion stability. J. Chem. Soc. Faraday Trans. 88, 2973–2983. Dickinson, E., 1998a. Proteins at interfaces and in emulsions: stability, rheology and interactions. J. Chem. Soc. Faraday Trans. 94, 1657–1669.

References

493

Dickinson, E., 1998b. Stability and rheological implications of electrostatic milk protein-polysaccharide interactions. Trends Food Sci. Technol. 9, 347–354. Dickinson, E., 1999. Adsorbed protein layers at fluid interfaces: interactions, structure and surface rheology. Colloids Surf. B: Biointerfaces 15, 161–176. Dickinson, E., 2003. Hydrocolloids at interfaces and the influence on the properties of dispersed systems. Food Hydrocoll. 17, 25–39. Dickinson, E., Davies, E., 1999. Influence of ionic calcium on stability of sodium caseinate emulsions. Colloids Surf. B Biointerfaces 2, 203–212. Dickinson, E., Golding, M., 1997. Depletion flocculation of emulsions containing unadsorbed sodium caseinate. Food Hydrocoll. 11, 13–18. Dickinson, E., Golding, M., 1998. Influence of alcohol on stability of oil-in-water emulsions containing sodium caseinate. J. Colloid Interface Sci. 197, 133–141. Dickinson, E., James, J.D., 1998. Rheology and flocculation of high pressure-treated β-lactoglobulin-stabilized emulsions: comparison with thermal treatment. J. Agric. Food Chem. 46, 2565–2571. Dickinson, E., Matsumura, Y., 1991. Time-dependent polymerization of β-lactoglobulin through disulphide bonds at the oil-water interface in emulsions. Int. J. Biol. Macromol. 13, 26–30. Dickinson, E., McClements, D.J. (Eds.), 1995. Advances in Food Colloids. Blackie Academic & Professional, London, UK. Dickinson, E., Parkinson, E.L., 2004. Heat-induced aggregation of milk protein-stabilized emulsions: sensitivity to processing and composition. Int. Dairy J. 14, 635–645. Dickinson, E., Stainsby, G., 1982. Colloids in Food. Applied Science Publishers, London, UK. Dickinson, E., Stainsby, G., 1988. Emulsion stability. In: Dickinson, E., Stainsby, G. (Eds.), Advances in Food Emulsion and Foams. Elsevier Applied Science, London, UK, pp. 1–44. Dickinson, E., Murray, B.S., Stainsby, G., 1988a. Protein adsorption at air-water and oil-water interfaces. In: Dickinson, E., Stainsby, G. (Eds.), Advances in Food Emulsion and Foams. Elsevier Applied Science, London, UK, pp. 123–162. Dickinson, E., Rolfe, S., Dalgleish, D.G., 1988b. Competitive adsorption of αs1-casein and β-casein in oil-in water emulsions. Food Hydrocoll. 2, 193–203. Dickinson, E., Semenova, M.G., Belyakova, L.E., Antipova, A.S., Il’in, M.M., Tsapkina, E.N., Ritzoulis, C., 2001. Analysis of light scattering data on the calcium ion sensitivity of caseinate solution thermodynamics: relationship to emulsion flocculation. J. Colloid Interface Sci. 239, 87–97. Djordjevic, D., McClements, D.J., Decker, E.A., 2004. Oxidative stability of whey protein-stabilized oil-in-water emulsions at pH 3: potential omega-3 fatty acid delivery systems (part B). J. Food Sci. 69, C356–C362. Dumay, E., Lambert, C., Funtenberger, S., Cheftel, J.C., 1996. Effects of high pressure on the physico-chemical characteristics of dairy creams and model oil/water emulsions. Lebensm. Wiss. Technol. 29, 606–625. Ekmekcioglu, C., 2002. A physiological approach for preparing and conducting intestinal bioavailability studies using experimental systems. Food Chem. 76, 225–230. Elias, R.J., McClements, D.J., Decker, E.A., 2007. Impact of thermal processing on the antioxidant mechanisms of continuous phase β-lactoglobulin in oil-in-water emulsions. Food Chem. 104 (4), 1402–1409. Euston, S.R., Hirst, R.L., 1999. Comparison of the concentration-dependent emulsifying properties of protein products containing aggregated and non-aggregated milk protein. Int. Dairy J. 9, 693–701. Euston, S.E., Singh, H., Munro, P.A., Dalgleish, D.G., 1995. Competitive adsorption between sodium caseinate and oilsoluble and water-soluble surfactants in oil-in-water emulsions. J. Food Sci. 60, 1124–1131. Euston, S.E., Singh, H., Munro, P.A., Dalgleish, D.G., 1996. Oil-in-water emulsions stabilized by sodium caseinate or whey protein isolate as influenced by glycerol monostearate. J. Food Sci. 61, 916–920. Euston, S.R., Finnigan, S.R., Hirst, R.L., 2000. Aggregation kinetics of heated whey protein-stabilized emulsions. Food Hydrocoll. 14, 155–161. Fang, Y., Dalgleish, D.G., 1998. The conformation of α-lactalbumin as a function of pH, heat treatment and adsorption at hydrophobic surfaces studied by FTIR. Food Hydrocoll. 12, 121–126. Fellows, C.M., Doherty, W.O.S., 2006. Insights into bridging flocculation. Macromol. Symp. 231, 1–10. Galazka, V.B., Ledward, D.A., Dickinson, E., Langley, K.R., 1995. High pressure effects on emulsifying behaviour of whey protein concentrate. J. Food Sci. 60, 1341–1343. Gaucheron, F., Famelart, M.H., LeGraet, Y., 1996. Iron-supplemented caseins: preparation, physicochemical characterization and stability. J. Dairy Res. 63 (2), 233–243.

494

12. Interactions and functionality of milk proteins in food emulsions

Gaucheron, F., Famelart, M.H., Mariette, F., Raulok, K., Michel, F., Le Graet, Y., 1997. Combined effects of temperature and high-pressure treatments on physicochemical characteristics of skim milk. Food Chem. 59, 439–447. Gohtani, S., Sirendi, M., Yamamoto, N., Kajikawa, K., Yamano, Y., 1999. Effect of droplet size on oxidation of docosahexaenoic acid in emulsion system. J. Dispers. Sci. Technol. 20 (5), 1319–1325. Golding, M., Wooster, J., 2010. The influence of emulsion structure and stability on lipid digestion. Curr. Opin. Colloid Interface Sci. 15, 90–101. Hu, M., McClements, D.J., Decker, E.A., 2003. Impact of whey protein emulsifiers on the oxidative stability of salmon oil-in-water emulsions. J. Agric. Food Chem. 51, 1435–1439. Hunt, J.A., Dalgleish, D.G., 1994. Adsorption behaviour of whey protein isolate and caseinate in soya oil-water emulsions. Food Hydrocoll. 8, 175–187. Hunt, J.A., Dalgleish, D.G., 1995. Heat stability of oil-in-water emulsions containing milk proteins: effect of ionic strength and pH. J. Food Sci. 60, 1120–1123. Hunter, R.J. (Ed.), 1986. Foundations of Colloid Science. In: vol. 1. Oxford University Press, Oxford, UK, p. 674. Huppertz, T., Fox, P.F., Kelly, A.L., 2004. Dissociation of caseins in high-pressure treated bovine milk. Int. Dairy J. 14, 675–680. Hur, S.J., Decker, E.A., McClements, D.J., 2009. Influence of initial emulsifier type on microstructural changes occurring in emulsified lipids during in vitro digestion. Food Chem. 114, 253–262. Jacobsen, C., Hartvigsen, K., Lund, P., Thomsen, M.K., Skibsted, L.H., Adler-Nissen, J., Holmer, G., Meyer, A.S., 2000. Oxidation in fish oil-enriched mayonnaise 3. Assessment of the influence of the emulsion structure on oxidation by discriminant partial least squares regression analysis. Eur. Food Res. Technol. 211 (2), 86–98. Kalantzi, L., Goumas, K., Kalioras, V., Abrahamsson, B., Dressman, J., Reppas, C., 2006. Characterization of the human upper gastrointestinal contents under conditions simulating bioavailability/bioequivalence studies. Pharm. Res. 23, 165–176. Kenmogne-Domguia, H.B., Meynier, A., Viau, M., Llamas, G., Genot, C., 2013. Gastric conditions control both the evolution of the organization of protein-stabilized emulsions and the kinetic of lipolysis during in vitro digestion. J. Funct. Foods 3, 1302–1309. Kiokias, S.N., Dimakou, C.P., Tsaprouni, I.V., Oreopoulou, V., 2006. Effect of compositional factors against the thermal oxidative deterioration of novel food emulsions. Food Biophys. 1, 115–123. Kiokias, S., Dimakou, C., Oreopoulou, V., 2007. Effect of heat treatment and droplet size on the oxidative stability of whey protein emulsions. Food Chem. 105 (1), 94–100. Klinkerson, U., McClements, D.J., 2010. Impact of lipase, bile salts, and polysaccharides on properties and digestibility of tuna oil multilayer emulsions stabilized by lecithin–chitosan. Food Biophys. 5, 73–81. Kolovou, G.D., Anagnostopoulou, K.K., Pavlidis, A.N., Salpea, K.D., Iraklianou, S.A., Tsarpalis, K., Mandis, A., Cokkinos, D.V., 2005. Postprandial lipemia in men with metabolic syndrome, hypertensives and healthy subjects. Lipids Health Dis. 4, 21. Kong, F., Singh, R.P., 2008. A model stomach system to investigate disintegration kinetics of solid foods during gastric digestion. J. Food Sci. 73 (5), E202–E210. Krondahl, E., Orzechowski, A., Ekstrom, G., Lennernas, H., 1997. Rat jejunal permeability and metabolism of mu-selective tetrapeptides in gastrointestinal fluids from humans and rats. Pharm. Res. 14, 1780–1785. Lazzaro, F., Saint-Jalmes, A., Violleau, F., Lopez, C., Gaucher-Delmas, M., Madec, M.N., Beacher, E., Gaucheron, F., 2017. Gradual disaggregation of the casein micelle improves its emulsifying capacity and decreases the stability of dairy emulsions. Food Hydrocoll. 63, 189–200. Le Reverend, B.J.D., Norton, I.T., Cox, P.W., Spyropoulos, F., 2010. Colloidal aspects of eating. Curr. Opin. Colloid Interface Sci. 15, 84–89. Let, M.B., Jacobsen, C., Meyer, A.S., 2007. Lipid oxidation in milk, yoghurt, and salad dressing enriched with neat fish oil or pre-emulsified fish oil. J. Agric. Food Chem. 55 (19), 7802–7809. Lethuaut, L., Metro, F., Genot, C., 2002. Effect of droplet size on lipid oxidation rates of oil-in-water emulsions stabilized by protein. J. Am. Oil Chem. Soc. 79 (5), 425–430. Li, J., Ye, A., Lee, S.J., Singh, H., 2012. Influence of gastric digestive reaction on subsequent in vitro intestinal digestion of sodium caseinate-stabilized emulsions. J. Funct. Foods 3, 320–326. Liang, Y., Patel, H., Matia-Merino, L., Gillies, G., Ye, A., Golding, M., 2013. Structure and stability of heat-treated concentrated dairy-protein-stabilized oil-in-water emulsions: a stability map characterization approach. Food Hydrocoll. 33, 297–308.

References

495

Liang, Y., Matia-Merino, L., Gillies, G., Patel, H., Ye, A., Golding, M., 2017. The heat stability of milk protein-stabilized oil-in-water emulsions. Curr. Opin. Colloid Interface Sci. 28, 63–73. Liu, F., Zhang, S., Li, J., McClements, D.J., Liu, X., 2018. Recent development of lactoferrin-based vehicles for the delivery of bioactive compounds: complexes, emulsions, and nanoparticles. Trends Food Sci. Technol. 79, 67–77. Lo´pez-Fandino´, R., Carrascosa, A.V., Olano, A., 1996. The effects of high pressure on whey protein denaturation and cheese-making properties of raw milk. J. Dairy Sci. 79, 929–936. Lucey, J.A., Srinivasan, M., Singh, H., Munro, P., 2000. Characterization of commercial and experimental sodium caseinates by multiangle laser light scattering and size-exclusion chromatography. J. Agric. Food Chem. 48, 1610–1616. Macierzanka, A., Sancho, A.I., Mills, E.N.C., Rigby, N.M., Mackie, A.R., 2009. Emulsification alters simulated gastrointestinal proteolysis of β-casein and β-lactoglobulin. Soft Matter 5, 538–550. Mackie, A.R., Mingins, J., Dann, R., 1993. Preliminary studies of β-lactoglobulin adsorbed on polystyrene latex. In: Dickinson, E., Walstra, P. (Eds.), Food Colloids and Polymers: Stability and Mechanical Properties. The Royal Society of Chemistry, Cambridge, UK, pp. 96–112. Maldonado-Valderrama, J., Woodward, N.C., Gunning, A.P., Ridout, M.J., Husband, F.A., Mackie, A.R., 2008. Interfacial characterization of beta-lactoglobulin networks: displacement by bile salts. Langmuir 24, 6759–6767. McClements, D.J. (Ed.), 1999. Food Emulsions: Principles, Practice, and Techniques. CRC Press, London, UK. McClements, D.J., 2005. Theoretical analysis of factors affecting the formation and stability of multilayered colloidal dispersions. Langmuir 21, 9777–9785. McClements, D.J., Decker, E.A., 2000. Lipid oxidation in oil-in-water emulsions: impact of molecular environment on chemical reactions in heterogeneous food systems. J. Food Sci. 65 (8), 1270–1282. McClements, D.J., Monahan, F.J., Kinsella, J.E., 1993. Disulfide bond formation affects stability of whey protein isolate emulsions. J. Food Sci. 58, 1036–1039. McClements, D.J., Decker, E.A., Park, Y., 2009. Controlling lipid bioavailability through physicochemical and structural approaches. Crit. Rev. Food Sci. Nutr. 49, 48–67. Monahan, F.J., McClements, D.J., German, J.B., 1996. Disulfide-mediated polymerization reactions and physical properties of heated WPI-stabilized emulsions. J. Food Sci. 61, 504–509. Mulvihill, D.M., 1992. Production, functional properties and utilization of milk protein products. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry – 1. Proteins. London, UK, Elsevier Applied Science, pp. 369–405. Mulvihill, D.M., Fox, P.F., 1989. Physico-chemical and functional properties of milk proteins. In: Fox, P.F. (Ed.), Developments in Dairy Chemistry—4. Functional Milk Proteins. London, UK, Elsevier Applied Science, pp. 131–172. Mun, S., Decker, E.A., McClements, D.J., 2007. Influence of emulsifier type on in vitro digestibility of lipid droplets by pancreatic lipase. Food Res. Int. 40, 770–781. Nakaya, K., Ushio, H., Matsukawa, S., Shimizu, M., Ohshima, T., 2005. Effects of droplet size on the oxidative stability of oil-in-water emulsions. Lipids 40 (5), 501–507. Nik, A.M., Wright, A.J., Corredig, M., 2010. Surface adsorption alters the susceptibility of whey proteins to pepsindigestion. J. Colloid Interface Sci. 344, 372–381. Okubanjo, S., Loveday, S.M., Ye, A., Wilde, P.J., Singh, H., 2019. Droplet-stabilized oil-in-water emulsions protect unsaturated lipids from oxidation. J. Agric. Food Chem. 67, 2626–2636. Pafumi, Y., Lairon, D., de la Porte, P.L., Juhel, C., Storch, J., Hamosh, M., 2002. Mechanisms of inhibition of triacylglycerol hydrolysis by human gastric lipase. J. Biol. Chem. 277, 28070–28079. Pal, A., Brasseur, J.G., Abrahamsson, B., 2007. A stomach road or “magenstrasse” for gastric emptying. J. Biomech. 40, 1202–1210. Pappas, C.P., Rothwell, J., 1991. The effect of heating, alone or in the presence of calcium or lactose, on calcium binding to milk proteins. Food Chem. 42, 183–201. Parkinson, E.L., Dickinson, E., 2004. Inhibition of heat-induced aggregation of a beta-lactoglobulin-stabilized emulsion by very small additions of casein. Colloids Surf. B: Biointerfaces 39, 23–30. Patel, H.A., Singh, H., Havea, P., Considine, T., Creamer, L.K., 2005. Pressure-induced unfolding and aggregation of the protein in whey protein concentrate solutions. J. Agric. Food Chem. 53, 9590–9601. Patel, H.A., Singh, H., Anema, S.G., Creamer, L.K., 2006. Effects of heat and high hydrostatic pressure treatments on disulfide bonding interchanges among the proteins in skim milk. J. Agric. Food Chem. 54, 3409–3420. Phillips, M.C., 1981. Protein conformation at liquid interfaces and its role in stabilizing emulsions and foams. Food Technol. 35, 424–427.

496

12. Interactions and functionality of milk proteins in food emulsions

Radford, S.J., Dickinson, E., Golding, M., 2004. Stability and rheology of emulsions containing sodium caseinate: combined effects of ionic calcium and alcohol. J. Colloid Interface Sci. 274, 673–686. Ramkumar, C., Singh, H., Munro, P.A., Dalgleish, D.G., Singh, A.M., 2000. Influence of calcium, magnesium, or potassium ions on the formation and stability of emulsions prepared using highly hydrolyzed whey proteins. J. Agric. Food Chem. 48, 1598–1604. Reddy, M., Kella, N.K.D., Kinsella, J.E., 1988. Structural and conformational basis of the resistance of β-lactoglobulin to peptic and chymotryptic digestion. J. Agric. Food Chem. 36, 737–741. Ries, D., Ye, A., Haisman, D., Singh, H., 2010. Antioxidant properties of caseins and whey proteins in model oil-inwater emulsions. Int. Dairy J. 20, 72–78. Rollema, H.S., 1992. Casein association and micelle formation. In: Fox, P.F. (Ed.), Advanced Dairy Chemistry—1: Proteins. Elsevier Applied Science, London, UK, pp. 111–140. Sarkar, A., Goh, K.K.T., Singh, R.P., Singh, H., 2009. Behaviour of an oil-in-water emulsion stabilized by β-lactoglobulin in an in vitro gastric model. Food Hydrocoll. 23, 1563–1569. Sarkar, A., Horne, D.S., Singh, H., 2010. Pancreatin-induced coalescence of oil-in-water emulsions in an in vitro duodenal model. Int. Dairy J. 20, 589–597. Schmelz, T., Lesmes, U., Weiss, J., McClements, D.J., 2011. Modulation of physicochemical properties of lipid droplets using beta-lactoglobulin and/or lactoferrin interfacial coatings. Food Hydrocoll. 25, 1181–1189. Schmidt, D.G., van Markwijk, B.W., 1993. Enzymatic hydrolysis of whey proteins. Influence of heat treatment of α-lactalbumin and β-lactoglobulin on their proteolysis by pepsin and papain. Neth. Milk Dairy J. 47, 15–22. Silletti, E., Vingerhoeds, M.H., Norde, W., van Aken, G.A., 2007a. Complex formation in mixtures of lysozymestabilized emulsions and human saliva. J. Colloid Interface Sci. 313, 485. Silletti, E., Vingerhoeds, M.H., Norde, W., van Aken, G.A., 2007b. The role of electrostatics in saliva-induced emulsion flocculation. Food Hydrocoll. 21, 596. Singh, H., 2005. Milk protein functionality in food colloids. In: Dickinson, E. (Ed.), Food Colloids: Interactions, Microstructure and Processing. The Royal Society of Chemistry, Cambridge, UK, pp. 179–193. Singh, H., 2011. Aspects of milk-protein-stabilised emulsions. Food Hydrocoll. 25, 1938–1944. Singh, A.M., Dalgleish, D.G., 1998. The emulsifying properties of hydrolyzates of whey proteins. J. Dairy Sci. 81, 918–924. Singh, H., Sarkar, A., 2011. Behaviour of protein-stabilised emulsions under various physiological conditions. Adv. Colloid Interf. Sci. 165, 47–57. Singh, H., Ye, A., 2013. Structural and biochemical factors affecting the digestion of protein-stabilized emulsions. Curr. Opin. Colloid Interface Sci. 18 (4), 360–370. Singh, H., Zhu, X., Ye, A., 2006. Lipid Encapsulation. W.I.P.O. Patent No. WO2006/115420. Singh, H., Ye, A., Horne, D., 2009. Structuring food emulsions in the gastrointestinal tract to modify lipid digestion. Prog. Lipid Res. 48, 92–100. Singh, H., Ye, A., Ferrua, M.J., 2015. Aspects of food structures in the digestive tract. Curr. Opin. Food Sci. 3, 85–93. Sliwinski, E.L., Roubos, P.J., Zoet, F.D., van Boekel, M.A.J.S., Wouters, J.T.M., 2003. Effects of heat on physicochemical properties of whey protein-stabilised emulsions. Colloids Surf. B: Biointerfaces 31, 231–242. Srinivasan, M., Singh, H., Munro, P.A., 1996. Sodium caseinate-stabilized emulsions: factors affecting coverage and composition of surface proteins. J. Agric. Food Chem. 44, 3807–3811. Srinivasan, M., Singh, H., Munro, P.A., 1999. Adsorption behaviour of sodium and calcium caseinates in oil-in-water emulsions. Int. Dairy J. 9, 337–341. Srinivasan, M., Singh, H., Munro, P.A., 2000. The effect of sodium chloride on the formation and stability of sodium caseinate emulsions. Food Hydrocoll. 14, 497–507. Srinivasan, M., Singh, H., Munro, P.A., 2001. Creaming stability of oil-in-water emulsions formed with sodium and calcium caseinates. J. Food Sci. 66, 441–446. Srinivasan, M., Singh, H., Munro, P.A., 2002. Formation and stability of sodium caseinate emulsions: influence of retorting (121 degrees C for 15 min) before or after emulsification. Food Hydrocoll. 16, 153–160. Sugiarto, M., Ye, A., Taylor, M.W., Singh, H., 2010. Milk protein-iron complexes: Inhibition of lipid oxidation in an emulsion. Dairy Sci. Technol. 90, 87–98. Tong, L.M., Sasaki, S., McClements, D.J., Decker, E.A., 2000. Antioxidant activity of whey in a salmon oil emulsion. J. Food Sci. 65 (8), 1325–1329. Torcello-Gomez, A., Maldonado-Valderrama, J., de Vicente, J., 2011. Investigating the effect of surfactants on lipase interfacial behaviour in the presence of bile salts. Food Hydrocoll. 25, 809–816.

References

497

Tuinier, R., de Kruif, C.G., 1999. Phase behavior of casein micelles/exocellular polysaccharide mixtures: experiment and theory. J. Colloid Interface Sci. 110, 9296–9304. van Aken, G.A., Vingerhoeds, M.H., De Hoog, E.H.A., 2005. Colloidal behaviour of food emulsions under oral conditions. In: Dickinson, E. (Ed.), Food Colloids: Interactions, Microstructure and Processing. The Royal Society of Chemistry, Cambridge, UK, pp. 356–366. Vingerhoeds, M.H., Blijdenstein, T.B.J., Zoet, F.D., van Aken, G.A., 2005. Emulsion flocculation induced by saliva and mucin. Food Hydrocoll. 19, 915–922. Vingerhoeds, M.H., Silletti, E., de Groot, J., Schipper, R.G., van Aken, G.A., 2008. Relating the effect of saliva-induced emulsion flocculation on rheological properties and retention on the tongue surface with sensory perception. Food Hydrocoll. 23, 773–785. Wahlgren, M.C., Arnebrant, T., Paulsson, M.A., 1993. The adsorption from solutions of beta-lactoglobulin mixed with lactoferrin or lysozyme onto silica and methylated silica surfaces. J. Colloid Interface Sci. 158, 46–53. Walstra, P., 1993. Introduction to aggregation phenomena in food colloids. In: Dickinson, E., Walstra, P. (Eds.), Food Colloids and Polymers: Stability and Mechanical Properties. The Royal Society of Chemistry, Cambridge, UK, pp. 3–15. Wang, X., Lin, Q., Ye, A., Han, J., Singh, H., 2019. Flocculation of oil-in-water emulsions stabilised by milk protein ingredients under gastric conditions: impact on in vitro intestinal lipid digestion. Food Hydrocoll. 88, 272–282. Ye, A., 2010. Surface protein composition and concentration of whey-protein-isolate-stabilized oil-in-water emulsions: effect of heat treatment. Colloids Surf. B: Biointerfaces 78, 24–29. Ye, A., 2011. Functional properties of milk protein concentrates: emulsifying properties, adsorption and stability of emulsions. Int. Dairy J. 21, 14–20. Ye, A., Singh, H., 2000. Influence of calcium chloride addition on the properties of emulsions stabilized by whey protein concentrate. Food Hydrocoll. 14, 337–346. Ye, A., Singh, H., 2001. Interfacial composition and stability of sodium caseinate emulsions as influenced by calcium ions. Food Hydrocoll. 15, 195–207. Ye, A., Singh, H., 2006a. Heat stability of oil-in-water emulsions formed with intact or hydrolysed whey proteins: influence of polysaccharides. Food Hydrocoll. 20, 269–276. Ye, A., Singh, H., 2006b. Adsorption behaviour of lactoferrin in oil-in-water emulsions as influenced by interaction with β-lactoglobulin. J. Colloid Interface Sci. 295, 249–254. Ye, A., Singh, H., 2007. Formation of multilayers at the interface of oil-in-water emulsion via interactions between lactoferrin and β-lactoglobulin. Food Biophys. 2, 125–132. Ye, A., Cui, J., Taneja, A., Zhu, X., Singh, H., 2009. Evaluation of processed cheese fortified with fish oil emulsion. Food Res. Int. 42, 1093–1098. Ye, A., Lo, J., Singh, H., 2012. Formation of interfacial milk protein complexation to stabilize oil-in-water emulsions against calcium. J. Colloid Interface Sci. 378, 184–190. Ye, A., Zhu, X., Singh, H., 2013. Oil-in-water emulsion system stabilized by protein-coated nanoemulsion droplets. Langmuir 29 (47), 14403–14410.

This page intentionally left blank

C H A P T E R

13 Milk protein-polysaccharide interactions Kelvin K.T. Goha, Anges Teob, Anwesha Sarkarc, Harjinder Singhd a

School of Food and Advanced Technology, Massey University, Palmerston North, New Zealand Abbott Nutrition Research and Development, Singapore, Singapore cSchool of Food Science and Nutrition, University of Leeds, Leeds, United Kingdom dThe Riddet Institute, Massey University, Palmerston North, New Zealand

b

Introduction Proteins and polysaccharides are broadly classified as biopolymers because of their large molecular structures. These macromolecules are known to possess important physicochemical roles, such as imparting thickening, stabilizing, gelling, and emulsifying properties in food products (Hemar et al., 2001a,b; Dickinson, 2003; Dickinson et al., 2003). The physicochemical properties of proteins and polysaccharides individually have been studied extensively over the last several decades. It is well established that the factors influencing these properties of these macromolecules in solution include molar mass, molecular conformation, polydispersity, charge density, concentration, pH, ionic strength, temperature, solvent quality, and the nature of molecular (intra-/inter)interactions (Tolstoguzov, 1997a; Doublier et al., 2000; de Kruif and Tuinier, 2001). The physical properties of many food systems become more complex as both proteins and polysaccharides are present (either naturally or added as ingredients) in the complex multicomponent mixtures. The overall stability and the microstructure of food systems depend not only on the physicochemical properties of proteins or polysaccharides alone but also on the nature and strength of interactions between the proteins and polysaccharides (Dickinson, 1998; Dickinson et al., 1998). This chapter reviews a number of recent studies carried out in the field of protein-polysaccharide interactions, with a particular focus on milk proteins and a diverse range of polysaccharides in aqueous systems.

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00013-X

499

# 2020 Elsevier Inc. All rights reserved.

500

13. Milk protein-polysaccharide interactions

Mixing behavior of biopolymers When aqueous solutions of proteins and polysaccharides are mixed, one of four phenomena can arise: (1) cosolubility, (2) thermodynamic incompatibility, (3) depletion interaction, and (4) complex coacervation (Fig. 13.1) (Tolstoguzov, 1991, 1997a, 2003; Schmitt et al., 1998; Syrbe et al., 1998; de Kruif and Tuinier, 2001; Benichou et al., 2002; Dickinson, 2003; de Kruif et al., 2004; Martinez et al., 2005). These phenomena can be explained as follows. (1) Cosolubility refers to the creation of a stable homogeneous solution, that is, the generation of one phase in which two macromolecular species either do not interact or exist as soluble complexes in the aqueous medium. When intermolecular attraction is absent, macromolecules are cosoluble only in dilute solutions where the entropy of mixing favors more randomness in the system (Tolstoguzov, 2003). To achieve cosolubility from a thermodynamic angle, the Gibbs free energy of mixing (ΔGmixing) given in Eq. (13.1) must be negative. This means that the entropy of mixing should favorably exceed the enthalpy term (note that the highest level of entropy is achieved when the different kinds of molecules are randomly distributed throughout the system) (McClements, 2005). The expression for the Gibbs free energy accompanying mixing under standard conditions is given by ΔGmixing ¼ ΔHmixing  TΔSmixing

(13.1)

where ΔGmixing, ΔHmixing, and TΔSmixing are the free energy, enthalpy (interaction energy), and entropy changes between the mixed and unmixed states, respectively.

FIG. 13.1 Different types of interactions between proteins and polysaccharides in aqueous solutions.

Protein

Polysaccharide

Mixing

Interacting

Soluble complexes

Insoluble complexes

Ionic strength, pH

Co-soluble (usually in dilute system)

Noninteracting

Thermodynamic incompatibility

Depletion flocculation

Mixing behavior of biopolymers

501

When small molecules are mixed, as in the case of monomer sugars and hydrophilic amino acids, a cosoluble system is formed between the two species. However, with increasing molecular weight and concentration of the polymers, the system tends to become less cosoluble as a result of thermodynamic incompatibility (Tolstoguzov, 1991, 1997a), because the entropy of mixing of the polymers is significantly lower than that of the monomers. The bulky size and the rigid structure of polymer molecules decrease the entropy of mixing, resulting in a higher free energy. For a mixed polymer solution, the enthalpyentropy balance generally results in mutual exclusion of one biopolymer from the local vicinity of the other. This means that the biopolymers in a mixed solution show a preference to be surrounded by their own type; otherwise, their mixtures separate into liquid phases (Grinberg and Tolstoguzov, 1972, 1997; Tolstoguzov et al., 1985; Tolstoguzov, 1988, 1991; Polyakov et al., 1997). (2) Thermodynamic incompatibility occurs when two dissimilar noninteracting macromolecular species separate into two different phases when the enthalpy of mixing exceeds the entropy difference (Grinberg and Tolstoguzov, 1997; Schmitt et al., 1998; Benichou et al., 2002; Tolstoguzov, 2002). The driving force to segregation is the enthalpic advantage of molecules being surrounded by others of the same type. For small molecules, this is normally outweighed by the entropic advantage of both species being free to move throughout the entire volume. However, for polymer solutions, where there are far fewer individual molecules, the entropy of mixing is much smaller, which can allow phase separation to occur. For the two distinct immiscible aqueous phases formed, each phase is loaded with mainly only one polymer species, that is, a protein-rich phase and a polysaccharide-rich phase. Phase separation caused by incompatibility can also occur if each biopolymer shows varying affinity toward the solvent (Tolstoguzov, 1991; Piculell and Lindman, 1992). In this case, solvent-protein (or solvent-polysaccharide) interactions are favored over protein-polysaccharide interactions and solvent-solvent interactions, leading to two phases, one enriched in protein and the other enriched in polysaccharide (Doublier et al., 2000). Thermodynamic incompatibility can also arise among different types of polysaccharides or proteins. Some examples include polysaccharides with different structures, proteins of different classes such as water-soluble albumins and salt-soluble globulins, native and denatured forms of the same protein, and aggregated and nonaggregated forms of the same protein (Tolstoguzov, 2002). Thermodynamic incompatibility of protein and polysaccharide is highly dependent on pH and ionic strength and is prevalent when both polymers carry the same negative charge at neutral pH (Doublier et al., 2000). Although thermodynamic incompatibility is prevalent in mixed polymer systems, some of these systems do not achieve thermodynamic equilibrium within a limited timescale because of the presence of kinetic energy barriers. When the kinetic energy exceeds the thermal energy of the system, the molecules become “trapped” in a metastable state (McClements, 2005). Some examples of kinetic energy barriers include the formation of a gel network within an incompatible system or a highly viscous continuous phase that slows down the phase separation process. The choice of which phase to gel and the component used to promote gelation depends on the type of biopolymers used in the system (Bryant and McClements, 2000; Norton and Frith, 2001; Kim et al., 2006).

502

13. Milk protein-polysaccharide interactions

(3) Depletion flocculation usually involves spherical particles in the presence of macromolecules (Asakura and Oosawa, 1954, 1958; Bourriot et al., 1999a). Phase separation of a particulate suspension is enhanced by the addition of a polymer. This phenomenon usually occurs in the colloidal dispersion in the presence of noninteracting polymers (e.g., polysaccharides in an emulsion, polysaccharides, and colloidal casein micelles). The higher osmotic pressure of the polymer molecules surrounding the colloidal particles (compared with the interparticle region) causes an additional attractive force between the particles, leading to flocculation. The attractive force depends on the size, shape, and concentration of the polymer molecules and the colloidal particles. When colloidal particles approach each other, the excluded (or depleted) layers start to overlap, allowing more space for the polymer molecules. The increase in volume causes the total entropy of the system to increase (i.e., a free energy decrease), which, in turn, encourages attraction between the colloidal particles (de Bont et al., 2002). In mixed proteinpolysaccharide systems containing casein micelles, phase separations are often attributed to depletion flocculation (Bourriot et al., 1999a; Tuinier and De Kruif, 1999; Tuinier et al., 2000), because of the large colloidal particle size of the casein micelles and because increasing the concentration of polysaccharides results in greater attraction between the casein micelles (Doublier et al., 2000). (4) Complex coacervation is the formation of electrostatic complexes between the protein and polysaccharide molecules, leading to a two-phase system. One phase has both biopolymers in a complex matrix, and the other phase contains mainly the solvent water or is depleted in both biopolymers. Complex coacervation commonly occurs between oppositely charged biopolymers. Complex coacervation between oppositely charged proteins and polysaccharides was first reported when gelatin and gum arabic were mixed in an acetic acid solution (Tiebackx, 1911). The term “coacervation” was first introduced in 1929 to describe a process in which aqueous colloidal solutions separate into two liquid phases, one rich in colloid, that is, the coacervate, and the other containing little colloid (Bungenberg de Jong and Kruyt, 1929). If the two biopolymers are present in equal proportion by weight at a certain pH, whereby they carry net equal opposite charges, the yield of coacervates will be at its maximum (Schmitt et al., 1998). Depending on the charge density between the two polymers, these coacervates may exist as stable suspensions (soluble complexes) or may phase separate as insoluble complexes, as shown in Fig. 13.1. The size and morphology of these structures may be exploited to bring about new functionalities and textural changes in processed foods.

Phase diagram Mixing two aqueous solutions of protein and polysaccharide may give rise to a one-phase system or a two-phase system depending on the solution composition and the environmental conditions, as depicted in Fig. 13.1 (Tolstoguzov, 1991, 1997a, 2002; Schmitt et al., 1998; Syrbe et al., 1998; de Kruif and Tuinier, 2001; Benichou et al., 2002; Dickinson, 2003; de Kruif et al., 2004; Martinez et al., 2005). In a one-phase system, protein and polysaccharide can exist either as individual molecules or as soluble complexes that are uniformly

503

Phase diagram

Polysaccharide (%w/w)

D

Tie-lines

Rectilinear diameter

+ B

+

O

O1

Binodal curve

E F A Protein (%w/w)

C

FIG. 13.2 A typical phase diagram showing a protein-polysaccharide solution with water as the solvent at a particular pH, temperature, and ionic strength. A sample of composition O (which was initially made with A% protein and B% polysaccharide) separates out into two bulk polymer-rich phases. The protein-enriched phase will have a composition of C% protein, whereas the polysaccharide-enriched phase will have a composition of D% polysaccharide. The binodal (solid) curve separates the single-phase region from the two-phase domain (obtained by direct observation of the phase separation in test tubes). The percentage of protein in the polysaccharide phase will be negligible and vice versa. The tie line is obtained by joining C and D. The ratio DO/OC represents the volume ratio of the protein-rich phase C and the polysaccharide-rich phase D, respectively, using the inverse-lever rule. If O is shifted along the tie line to O1, the new phase volume ratio will be DO1/O1C. Although any composition lying on the same tie line results in the same effective concentration as the enriched phases, the phase volume varies. The line obtained by joining the midpoints (+) of two or more tie lines gives the rectilinear diameter. The coordinates of the critical point E (obtained from the intersection of the binodal curve and the rectilinear diameter) show the composition of a system separating into two phases of the same volume and composition. In other words, the separated phase systems will have 50% protein and 50% polysaccharide in the same phase volume ratio. Point F represents the separation threshold, which is the minimum critical concentration required for the biopolymers to separate into two phases.

dispersed throughout the entire system. However, with increasing molecular weight and concentration of the biopolymers, the system tends to become less cosoluble and to give rise to a two-phase system in which two distinct phases with different biopolymer concentrations are formed. For a system with relatively strong net repulsion between protein and polysaccharide in aqueous solution, the additional electrostatic forces assist the two biopolymers to move into two different phases, that is, thermodynamic incompatibility. Two distinct immiscible aqueous phases are formed, and each of them is mainly loaded with only one biopolymer species, that is, one phase protein rich and the other phase polysaccharide rich. A typical phase diagram for a segregating biopolymer system is shown in Fig. 13.2 (Grinberg and Tolstoguzov, 1972, 1997; Polyakov et al., 1980; Antonov et al., 1982; Tolstoguzov et al., 1985; Bourriot et al., 1999a; Closs et al., 1999; Clark, 2000; Lundin et al., 2003; Thaiudom and Goff, 2003; Tolstoguzov, 2003; Ercelebi and Ibanoglu, 2007).

504

13. Milk protein-polysaccharide interactions

The phase diagram consists of a typical binodal curve (the solid line curve), which divides the single-phase miscible region (below the curve) from the two-phase immiscible region (shaded region). The binodal branches exhibit the points of limited cosolubility. The points of the binodal curve connected by the tie line represent the composition of the coexisting equilibrium phases. In general, the part of the binodal curve closer to the x-axis constitutes the protein-rich biopolymers, whereas the polysaccharide-rich biopolymers are found in the part closer to the y-axis. From the phase diagram, it is possible to determine the effective concentrations of biopolymers in the two phases and the concentrations at which maximal cosolubility of the biopolymers is achieved. The phase diagram is useful for establishing which of the two biopolymers forms the continuous phase.

Nature of interactions in protein-polysaccharide systems The interactions responsible for complex formation between biopolymers can be classified as weak or strong, specific or nonspecific, and attractive or repulsive. The overall interaction between protein and polysaccharide is the average of the following different intermolecular forces arising between the various segments and chains of the two biopolymers (Dickinson, 1998; Schmitt et al., 1998).

Repulsive interactions Repulsive interactions are always nonspecific and of transient duration. They usually arise from excluded volume effects and/or electrostatic interactions and tend to be weak, except at very close range or very low ionic strength. The excluded volume or steric exclusion effects are the nonspecific and transient interactions that can be found in nonionic, nonpenetrable polysaccharides and polypeptides that cannot occupy the same solution volume (Tolstoguzov, 1991, 2002; Polyakov et al., 1997; Schmitt et al., 1998). Excluded volume effects exhibit mutual spatial restrictions and competition between the biopolymers for solution space, that is, there is a reduction in the mixing entropy of the system because of the reduction in the volume available for occupation by the biopolymer molecules. Net repulsive interactions, caused by electrostatic effects, depend largely on the pH and ionic strength of the background electrolyte. Electrostatic repulsive interactions are commonly found in mixtures of proteins and anionic polysaccharides under conditions where both biopolymers carry the same net charge, for example, at pH above the isoelectric point (pI) of the protein.

Attractive interactions Attractive interactions between proteins and polysaccharides may be weak or strong and either specific or nonspecific. Nonspecific attractive interactions, such as electrostatic, van der Waals, hydrogen bonding, and hydrophobic interactions, arise as a result of a multitude of weak interactions between groups on the biopolymers. Hydrogen bonding and hydrophobic

Nature of interactions in protein-polysaccharide systems

505

interactions are actually collective interactions (e.g., electrostatic, van der Waals, and steric overlap) including some entropy effects (McClements, 2005). Electrostatic interactions are the most important forces involved in complex formation between proteins and ionic polysaccharides. These interactions between charged biopolymers lead to a decrease in the electrostatic free energy of the system. Moreover, the enthalpy contribution, caused by the interactions of oppositely charged biopolymers and the liberation of counterions along with water molecules, often compensates for the loss of configurational entropy of the mixing of rigid biopolymers (Piculell and Lindman, 1992; Tolstoguzov, 1997a). Strong electrostatic attractive interactions occur between positively charged proteins (pH < pI) and anionic polysaccharides, especially at low ionic strengths. Generally, two types of complexes are formed by electrostatic interactions (Tolstoguzov, 1997a, 2002; Schmitt et al., 1998): soluble complexes that are obtained when the opposite charges carried by the two biopolymers are not equal in number and insoluble complexes that result when the net charge on the complex is close to zero. van der Waals forces are extremely weak electrical attractions that arise because of temporary dipole interactions (Sherony and Kintner, 1971; Stainsby, 1980; Dickinson, 1998). Basically, every atom has an electron cloud that can yield a temporary electric dipole. The dipole in one atom can induce a corresponding dipole in another atom. This is possible only if the atoms are close. However, if they are too close, repulsive forces between the adjacent negatively charged electron clouds may not allow this van der Waals attraction. Although these transient electrical attractive forces are very weak, they influence macromolecular interactions together with other noncovalent forces described earlier (Damodaran, 1997). Hydrophobic bonding is an entropy-driven long-range interaction between nonpolar groups and is promoted by conformational and structural modifications of biopolymers, mostly by the unfolding of polymeric chains and the exposure of hydrophobic groups. These kinds of interactions are promoted by an increase in temperature (Stainsby, 1980; Piculell and Lindman, 1992; Samant et al., 1993; Antonov et al., 1996; Tolstoguzov, 1997a). Hydrogen bonds are moderately strong bonds (dOdHδ+…δ O 100 Pa) were formed with pectin concentrations of 0.6% at one particular degree of methylation (31%) and amidation (17%) in the presence of Ca2+ ions (1.8 mM) at pH 3.6 (Matia-Merino et al., 2004).

Milk protein-polysaccharide interactions in the aqueous phase and at the interface Milk proteins together with polysaccharides dissolved in the aqueous phase form a pseudo-ternary system comprising milk protein-polysaccharide-water. Various interactions in these systems could lead to complex formation or bulk phase separation. Extensive studies in areas of protein-polysaccharide interactions have been carried out, particularly using well-studied milk proteins and commercially available polysaccharides (Dickinson, 1998). A compilation (nonexhaustive) of the conditions and interactions between various milk proteins (casein and/or whey proteins) and polysaccharide mixtures in aqueous systems is presented in Table 13.1. Considerable work on the use of protein-polysaccharide complexes at oil-water and airwater interfaces, to improve properties such as emulsification and foaming, has been carried

507

Milk protein-polysaccharide interactions in the aqueous phase and at the interface

out. Various methods that are used to form emulsions with different interfacial structures using proteins and polysaccharides are illustrated in Fig. 13.3. The protein-polysaccharide complexes formed can be directly emulsified with the oil to form mixed-layer emulsions. Alternatively, the proteins and polysaccharides can be added together under conditions where there is no interaction between them and then emulsified with the oil to form emulsions with bilayer or multilayer structures at the interface. During emulsification, the proteins adsorb rapidly on to the droplet surface to form a primary coating (or a single layer), whereas the polysaccharides remain in the aqueous phase. Subsequently, the emulsion mixture can be manipulated to favor interaction with the polysaccharide by adjusting the solution conditions (e.g., pH changes, heating, and addition of salts) to form a bilayer. This process can be repeated several times to obtain multiple layers of protein-polysaccharide coatings on the droplet surface. In another method, bilayer emulsions can be produced by first forming a primary emulsion using the protein and then adding the polysaccharide to the emulsion to form a bilayer. Subsequent layers at the interface can be formed by the deposition of proteinpolysaccharide complexes.

TABLE 13.1

No.

Milk protein-polysaccharide interactions in aqueous systems

Milk proteinpolysaccharide in aqueous systems

Conditions

Interactions

References

1.

β-Lactoglobulin + carboxymethyl cellulose

• 60°C • pH 2.5–7.0 • 0.05–0.2 M NaCl

Insoluble electrostatic complex

(Hidalgo and Hansen, 1969; Hansen et al., 1974)

2.

Casein micelle + alginate

• 25°C • pH 7.2

Thermodynamic incompatibility

(Suchkov et al., 1981, 1988)

3.

Milk proteins + high methoxyl (HM) pectin (62.7% methylated)

• 20°C • pH 6.0–10.5 • 0–0.5 M NaCl

Thermodynamic incompatibility

(Antonov et al., 1982)

Milk proteins + gum arabic 4.

β-Lactoglobulin + carboxymethyl dextran

• 4 and 25°C • pH 4.75 and 5.5 • 1:1 and 7:2 ratio

Covalent conjugate

(Hattori et al., 1994)

5.

Whey protein isolate (WPI) + carboxymethyl potato starch

• 24°C • pH 7

Covalent conjugate

(Hattori et al., 1995)

6.

Bovine serum albumin + sulfated polysaccharides (ι-, κ-carrageenan, dextran sulfate)

• pH 6.5–8.0 • High pressure

Complex coacervation

(Galazka et al., 1996, 1997, 1999)

Continued

508

13. Milk protein-polysaccharide interactions

TABLE 13.1 Milk protein-polysaccharide interactions in aqueous systems—cont’d

No.

Milk proteinpolysaccharide in aqueous systems

Conditions

Interactions

References

7.

Casein micelle + carrageenan (ι-, κ-, λforms)

• • • • •

20, 50 and 60°C pH 6.7 and 7.0 0.25 M NaCl 0.01 M KCl 0.12% carrageenan

Depletion interaction

(Dalgleish and Morris, 1988; Langendorff et al., 1997, 1999, 2000; Bourriot et al., 1999b)

8.

β-Lactoglobulin + κcarrageenan

• • • •

45–80°C 0.1 M NaCl 0.01 M CaCl2 1:2, 5:1 and 10:1 ratio

Phase-separated bicontinuous gel formation

(Capron et al., 1999; Eleya and Turgeon, 2000)

9.

Sodium caseinate + maltodextrin

• 60°C • 2–4 days • 2:1, 1:1 and 1:4 ratio

Covalent conjugate via Maillard reaction; no phase separation

(Shepherd et al., 2000; Morris et al., 2004)

10.

Casein micelle + pectin (35 and 73% methylated)

• 60°C • pH 6.7 and 5.3 • 0.1%–0.2% pectin

Depletion interaction at pH 6.7

(Maroziene and de Kruif, 2000)

Casein micelle + low methoxyl (LM) amidated pectin (25 and 35% methylated) 11.

Sodium caseinate + sodium alginate

• 23°C • pH 7.0

Thermodynamic incompatibility

(Guido et al., 2002; Simeone et al., 2002)

12.

WPI + locust bean gum

• pH 5.0–7.0

Biphasic gel

(Tavares and Lopes da Silva, 2003)

13.

β-Lactoglobulin + pectin (28.3, 42.6, 71.3, 73.4% methylated)

• 4–40 and 25, 87°C • pH 4.0–7.5 and 6.5 • 0.11 and 0.1–1.0 M NaCl • High pressure

Complex coacervation

(Dumay et al., 1999; Wang and Qvist, 2000; Girard et al., 2002, 2003a,b, 2004; Kazmiersi et al., 2003)

14.

β-Lactoglobulin + pullulan

• 4°C • 0.01 M NaCl

Depletion interaction

(Wang et al., 2001)

15.

WPI + exopolysaccharide (Lactococcus lactis subsp. cremoris B40)

• 25°C • 0–0.1 M NaCl or CaCl2 • Heat treatment of WPI • 2:1 ratio

Depletion interaction

(de Kruif and Tuinier, 1999; Tuinier and De Kruif, 1999; Weinbreck et al., 2003)

509

Milk protein-polysaccharide interactions in the aqueous phase and at the interface

TABLE 13.1

No.

Milk protein-polysaccharide interactions in aqueous systems—cont’d

Milk proteinpolysaccharide in aqueous systems

Conditions

Interactions

References

Covalent conjugate

(Mishra et al., 2001; Neirynck et al., 2004)

16.

WPI + pectin Whey protein concentrate (WPC) + pectin

• • • •

60°C 14 days pH 7.0 4:1, 2:1, 1:1 and 1:2 ratio

17.

WPI + gum arabic

• pH 4.0–7.0 • 0–0.1 M NaCl • 2:1 ratio

Complex coacervation (glassy state)

(Weinbreck et al., 2004)

18.

WPI + λ-carrageenan

• pH 4–9 • 0–0.1 M NaCl or CaCl2 • 1:1 to 150:1 ratio

Electrostatic complexation

(Weinbreck et al., 2004)

19.

Heat-denatured WPI + xanthan gum

• • • •

Thermodynamic incompatibility

(Bryant and McClements, 2000; Li et al., 2006)

20.

Heat-denatured WPI + pectin (28%, 35%, 40%, 47%, and 65% methylation)

• 80 and 85°C • pH 6.0 and 7.0 • 5 and 10 mM CaCl2

Thermodynamic incompatibility

(Beaulieu et al., 2001; Kim et al., 2006)

21.

Sodium caseinate + gum arabic

• 0.5 M NaCl • pH 2.0–7.0 • 0.01%–5% gum arabic

Soluble electrostatic complexes

(Ye et al., 2006)

22.

β-Lactoglobulin + chitosan (85% deacetylation)

• pH 3.0–7.0 • 5 mM phosphate buffer • 0–0.1% chitosan

pH-soluble/insoluble complex coacervation

(Guzey and McClements, 2006a)

23.

WPI + chitosan

• pH 4.0–6.0 • Chitosan:WPI ratios – 0.007:1, 0.005:1, 0.2:1

Properties of complexes formed depend on pH, ionic strength, and protein: polysaccharide ratio.

(Bastos et al., 2010)

24.

Lactoferrin + HM pectin

• pH 2.0–7.0 • 30–90°C • 0.25% w/w lactoferrin • 0.005–0.15% w/w HM pectin

Soluble complexes pH 3.5–7 Improved stability against aggregation

(Bengoechea et al., 2011)

25–90°C pH 5.4 and 7.0 0.2 M NaCl High pressure

Continued

510

13. Milk protein-polysaccharide interactions

TABLE 13.1 Milk protein-polysaccharide interactions in aqueous systems—cont’d

No.

Milk proteinpolysaccharide in aqueous systems

Conditions

Interactions

References

25.

β-Lactoglobulin + soluble soy polysaccharide

• 50°C • 79% RH • 24 h

Conjugates formed via Maillard reaction with improved emulsifying properties

(Hayashi et al., 2015)

26.

Lactoferrin + N-succinyl chitosan Lactoferrin + galactomannan

• 68–80°C • pH 6.5 • 0.15 M NaCl

Soluble electrostatic complex

(Il’ina et al., 2016)

27.

Zein-caseinate-pectin

• • • •

pH 4.6 pH 6.5 Heating sequence Simulated gastric conditions

Zein/NaCas/pectin complex nanoparticles by modulating the biopolymer concentration, pH conditions, and heating sequence

(Chang et al., 2017)

28.

Lactoferrin + sodium alginate

• pH 2.0–8.0 • 1:4, 1:1, 4:1, 8:1, 16:1 and 24:1 ratio • Addition of transglutaminase

Complex coacervation

(Wang et al., 2017)

29.

Sodium caseinate + exopolysaccharides (Leuconostoc citreum, Leuconostoc mesenteroides, Pediococcus pentosaceus, Bacillus tequilensis)

• pH 3.0 and 3.5 • 0–1 g/L exopolysaccharides

Electrostatic complexation

(Abid et al., 2018)

30.

β-Lactoglobulin/bovine serum albumin + Na alginate

• pH 1.0–7.0 • 0.2% w/w protein solution • 0.1% Na alginate solutions • Ratios 1:1, 1.5:1, 2:1

Formation of insoluble to soluble complexes, dependent on concentration and pH; pH 4.2 shows maximum coacervation

(Gorji et al., 2018)

As seen in most emulsion systems containing both milk protein and polysaccharide, protein generally forms the primary interfacial layer by directly adsorbing to the oil surface. The hydrophilic polysaccharide possibly forms a thick secondary steric-stabilizing layer on the outside of the protein-adsorbed emulsion droplets, provided the protein-polysaccharide interaction is satisfactorily attractive (Dickinson, 1994). Generally, strong electrostatic interaction between the mutually oppositely charged adsorbed protein and added polysaccharide leads to the formation of multilayered interfacial membranes stabilizing the emulsion droplets by steric repulsion (Dickinson and James, 2000; Moreau et al., 2003; Guzey et al., 2004; Laplante et al.,

Milk protein-polysaccharide interactions in the aqueous phase and at the interface

511

FIG. 13.3 Schematic illustrations of various methods for the preparation of oil-in-water emulsions with different interfacial structures (mixed layer, bilayer, and multilayer) using protein and polysaccharide complexes.

2005; Guzey and McClements, 2006b; Mun et al., 2006; Hong and McClements, 2007; Bouyer et al., 2012; Nobuhara et al., 2014; Cai et al., 2017; Liang et al., 2017; Anal et al., 2019). To date, there has been much research on protein-polysaccharide interactions in emulsion systems under different conditions of temperature, pH, ionic strength, protein and polysaccharide concentrations, pressure treatment, etc. In most of these cases, the presence of polysaccharide creates flocculation by bridging interaction at lower concentrations, followed by emulsion stabilization at sufficiently high concentrations to saturate the protein-adsorbed emulsion droplets (Fig. 13.4B and C). However, in systems where there is excess polysaccharide that does not interact with the proteins, depletion flocculation can be expected (Fig. 13.4D). Protein-polysaccharide interactions are sensitive to the details of the protein structure and to the charge density on the biopolymers. For example, partial denaturation of globular proteins can result in increased complexation with hydrocolloids at the interface compared with an aqueous solution at the same pH and ionic strength (Dickinson, 2003). Generally, strong electrostatic interaction between the adsorbed protein and the added polysaccharide leads to the formation of a stabilizing layer. At sufficiently high hydrocolloid concentrations, the emulsion stability is increased by immobilizing the emulsion droplets in the gelled protein-polysaccharide network (Fig. 13.4E). However, the same polysaccharide can induce irreversible bridging flocculation of the protein-coated emulsion droplets if there is an insufficient amount of polysaccharide for surface coverage. Published investigations on milk protein-polysaccharide interactions in emulsion systems (nonexhaustive) are summarized and presented in Table 13.2. The interaction of polysaccharides with proteins is not limited to electrostatic interactions only. In addition to the use of transglutaminase as a cross-linker between proteins, covalent conjugates between milk proteins and polysaccharides that are formed via Maillard reactions have gained much interest because of their improved emulsification abilities compared with

512

13. Milk protein-polysaccharide interactions

FIG. 13.4 Schematic illustrations of interactions of protein-coated droplets with increasing polysaccharide concentrations (not drawn to scale): (A) In the absence of polysaccharides, there is a stable emulsion of protein-coated droplets; (B) at low concentration, the polysaccharides are insufficient to completely cover the droplet surface and result in the bridging of droplets; (C) the amount of polysaccharide in the system is sufficient to interact with the proteincoated droplets and produces a stable emulsion because of an increase in steric repulsion; (D) however, further increasing the concentration renders the emulsion unstable because of a depletion mechanism; (E) at sufficiently high concentration, the emulsion droplets are immobilized because of thickening of the continuous phase by the polysaccharides.

the individual biopolymers (Akhtar and Ding, 2017; Abd El-Salam and El-Shibiny, 2018). These conjugates are stable over a wide range of temperature, pH, and ionic strength. The conjugates with high molecular weight possess both the properties of a hydrophobic protein being adsorbed to the surface of the oil droplet and the properties of a hydrophilic polysaccharide being highly hydrated by the aqueous phase. Although these conjugates possess hydrophobic and hydrophilic groups and are effective as surface-active polymers, the presence of excess unreacted hydrocolloid may lead to depletion effects (Syrbe et al., 1998). Consequently, interfacial layers made up of different structures, thicknesses, compositions, and charges require knowledge of the functionality of different protein-polysaccharide combinations to meet the structural demands, environmental challenges, and stability of emulsions in real food systems. Emulsions stabilized by different structures of protein-polysaccharide complexes have promising applications in foods, with many superior functionalities expected. Zhao et al. (2015) investigated the stability of orange oil emulsions prepared using lactoferrin combined with soybean-soluble polysaccharide (SSPS) or beet pectin (BP). Bilayer emulsions were formed by electrostatic deposition of SPSS or BP on to the primary droplets containing lactoferrin. Bilayer emulsions with polysaccharide coatings were more physically stable during storage, whereas lactoferrin emulsions showed an increase in particle size after 2 weeks when stored at 55°C. There was also better retention of volatile compounds (decanal, octanal, geranial, and limonene) in the anionic bilayer emulsions. In a further study, the same type of bilayer emulsion was produced using lactoferrin and anionic polysaccharides and was shown to have improved stability with respect to pH and ionic strength over a wide pH range of 3–10 and at salt concentrations of less than 0.3 M NaCl (Zhao et al., 2014). Not only protein-polysaccharide complexes are useful for improving the environmental stability of emulsions, but also other functionalities are anticipated. There have been a number of recent studies on the use of protein-polysaccharide complexes to control the lipid digestion of emulsified oils in the gastrointestinal tract. For example, Chang and McClements (2016) used

Milk protein-polysaccharide interactions in the aqueous phase and at the interface

TABLE 13.2

513

Milk protein-polysaccharide interactions in emulsion systems

Milk protein-polysaccharide in emulsion systems

Oil phase

Conclusion

References

1.

Bovine serum albumin (BSA) + Carrageenan (ι- and κ- forms)

nTetradecane (20, 40 vol%)

• Strong bridging flocculation • BSA-κ-carrageenan interaction weaker than BSA-ι-carrageenan interaction at same pH and ionic strength

(Dickinson and Pawlowsky, 1997, 1998)

2.

Caseins (αs1-, β-casein and caseinate) + HM pectin

Sunflower oil (11, 40 vol%)

• Depletion flocculation at low pH • Salt-induced destabilization in αs1-casein-stabilized emulsion

(Dickinson et al., 1998)

3.

Sodium caseinate + Portulaca oleracea gum

MCT oil (5 wt %)

• Caseinate-Portulaca gum electrostatic complex • Better stability

(Garti et al., 1999)

4.

Casein + maltodextrin

Soybean oil (30 wt%)

• Covalent conjugates formed with better emulsion stability

(Shepherd et al., 2000)

5.

Sodium caseinate + Xanthan gum

Soybean oil (30 wt%)

• Depletion flocculation

(Hemar et al., 2001a)

6.

WPC + hydrocolloids xanthan gum, polypropylene glycol alginate ([GA], carrageenan)

Soybean oil (20 wt%)

• Heat-induced droplet aggregation by depletion mechanism • WPC–PGA complex showed better creaming stability

(Euston et al., 2002)

7.

WPI + Dextran sulfate

MCT oil, Silicone oil, Orange oil and nTetradecane (20 vol%)

• Covalent conjugates formed • Long-term emulsion stability

(Akhtar and Dickinson, 2003)

8.

β-Lactoglobulin + Dextran

Sunflower oil (10 wt%)

• Depletion flocculation at > 20% sucrose content

(Blijdenstein et al., 2004)

9.

WPI (85%β-lactoglobulin) + LM pectin

Soybean oil (20 wt%)

• Covalent conjugates formed with better emulsion stability at pH 5.5

(Neirynck et al., 2004)

10.

WPI + Chitosan

Canola oil (10 vol%)

• Electrostatic interaction at pH > 5 resulted in interfacial coadsorption to form stable emulsion • Depletion flocculation at pH  5

(Laplante et al., 2005)

No.

Continued

514

13. Milk protein-polysaccharide interactions

TABLE 13.2 Milk protein-polysaccharide interactions in emulsion systems—cont’d No.

Milk protein-polysaccharide in emulsion systems

11.

Oil phase

Conclusion

References

β-Lactoglobulin + carrageenan

Corn oil (5 wt%)

• Thermal processing at ionic strength (500 mM NaCl, 2 mM CaCl2) • Stable emulsion

(Gu et al., 2005,b)

12.

Hydrolyzed WPI + hydrocolloid (xanthan gum, guar gum, κcarrageenan)

Corn oil (4 wt%)

• Coalescence rate: guar gum > xanthan gum > κ-carrageenan • Depletion interaction after retort treatment

(Ye et al., 2004; Ye and Singh, 2006)

13.

β-Lactoglobulin + pectin (59% DE) + chitosan

Corn oil (2.5 wt%)

• Stable tertiary emulsion at low pH (3–5) and 0.1-M NaCl

(Guzey and McClements, 2006a)

14.

Lactoferrin + pectin (LM or HM) or alginate

Corn oil (5 wt%)

• Stable secondary emulsions with improved heat stability

(Tokle et al., 2010)

15.

Sodium caseinate + κ-carrageenan

Soybean oil (10 wt%)

• Bridging flocculation at lower κcarrageenan concentration (0.05– 0.25 wt%) and depletion flocculation at 0.5 wt%

(Perrechil and Cunha, 2013)

16.

WPI + tragacanth

Sunflower oil (1, 4, 8, 16 wt %)

• Layer-by-layer emulsions were more stable than mixed emulsions

(Azarikia and Abbasi, 2016a,b)

17.

Milk proteins (α-lactalbumin or sodium caseinate) + chitosan-(2)epigallocatechnin-3-gallate (EGCG)

MCT oil (5 wt%) + β-carotene

• Better emulsion stability than those formed by monolayer proteins

(Wei and Gao, 2016)

18.

Lactoferrin + alginate + E-polyL-lysine (PLL)

Corn oil (5 wt %) + resveratrol (0.01wt%)

• Stable tertiary emulsions (0.0036% PLL) with good antioxidant activity compared with unencapsulated resveratrol

(AcevedoFani et al., 2017)

19.

Xanthan gum (XG)-locust bean gum (LBG) mixtures + WPIstabilized oil-in-water emulsions

Menhaden oil (20% v/v)

• Mechanism: reduced depletion flocculation by XG:LBG mixtures • Addition of XG:LBG mixtures greatly decreased the creaming of the emulsions and improved oxidative stability

(Khouryieh et al., 2015)

20.

(Lactoferrin, α-lactalbumin, βlactoglobulin) and beet pectin

Citral oil (10 wt%)

• Bilayer emulsions by layer-bylayer deposition technique • Electrostatic interactions • Secondary emulsions coated with milk proteins and beet

(Xiang et al., 2015)

Milk protein-polysaccharide interactions in the aqueous phase and at the interface

TABLE 13.2 No.

515

Milk protein-polysaccharide interactions in emulsion systems—cont’d

Milk protein-polysaccharide in emulsion systems

Oil phase

Conclusion

References

pectin were stable at pH 3.5–10 and showed better physical stability over a wider range of NaCl concentrations and temperature than the primary emulsions 21.

Corn fiber gum-bovine serum albumin (CFG-BSA) conjugates

Soybean oil (20 wt%)

• CFG-BSA conjugates formed by enzymatic cross-linking • Better emulsion stability than those stabilized by CFG or BSA alone

(Liu et al., 2018)

22.

Whey protein hydrolyzate (WPH) + Cacao pod husk pectin (CPHP)

Canola oil (10 wt%)

• Soluble complexes of WPH and CPHP (5:1) at pH 3.25 formed stable and smaller droplets

(TrujilloRamı´rez et al., 2018)

fucoidan (an anionic marine polysaccharide) to form bilayer emulsions with different types of emulsifiers (viz., sodium caseinate, whey protein, lecithin, and Tween 80). They found that the addition of fucoidan increased the lipid digestion rate of the protein-stabilized emulsions by retarding droplet aggregation. Malinauskyt_e et al. (2014) undertook similar studies but found that the addition of carboxymethylcellulose (CMC) decreased the lipid digestion of whey protein emulsions. In this study, the whey protein emulsion and the CMC were mixed under conditions where there was no interaction between them and no complexes were formed. The authors explained that the CMC created a thickening effect on the emulsion and limited the access of lipase to the oil droplet surface, resulting in a slower lipid digestion compared with the original emulsion without CMC. From these studies, protein and polysaccharide interactions play an important role in the gastrointestinal fate of emulsions, and this is thought to have implications in the design of novel foods for controlling satiety or delivering encapsulated bioactive substances or compounds at specific sites of the body. As well as structuring emulsions with multiple layers of biopolymers at the interface, solid particles derived from protein and polysaccharide complexes can be used to obtain highly stable emulsions by Pickering stabilization. During emulsification, the solid particles adsorb on to the oil-water interface and form a protective shell around the droplets. Depending on the contact angle of the particles, the emulsions formed can be oil in water (90 degrees) (Yang et al., 2017). These particle-stabilized emulsions have been increasingly researched for their applications in food formulations, which have to date been limited to a few food materials, such as fat crystals (Rousseau, 2013), protein particles (Wu et al., 2015), protein-polysaccharide complexes (Santos et al., 2018), chitin nanocrystals, chitosan nanoparticles (Wei et al., 2012), cellulose nanocrystals (Sarkar et al., 2018), and nanofibrillated cellulose (Winuprasith et al., 2018). In a recent study, lactoferrin and inulin nanoparticles were used for the stabilization of emulsions by the Pickering mechanism (Sarkar et al., 2018). The emulsions were initially stabilized by lactoferrin nanogel particles

516

13. Milk protein-polysaccharide interactions

and then cross-linked with inulin to form a secondary layer. The gastric proteolysis was reduced for emulsions stabilized by lactoferrin nanoparticles compared with conventional emulsions stabilized by lactoferrin monolayers, and the addition of inulin further decreased the hydrolysis of lactoferrin in the system containing nanogels. This type of Pickering emulsion, stabilized by protein-polysaccharide complexes, is useful for the delivery of bioactive materials that are sensitive to harsh conditions in the gastrointestinal tract, as the particles provide a strong mechanical barrier to protect the droplets from degradation and coalescence.

Rheological properties and microstructures of protein-polysaccharide systems The rheological properties of a solution containing only protein are expected to be different from those of a pure polysaccharide solution. Polysaccharide molecules generally have a greater effect than proteins in causing a significant increase in the solution viscosity, because polysaccharide molecules are usually much larger and more extended (5.0  105– 2.0  106 Da) than globular proteins (1.0  104–1.0  105 Da). Hence, polysaccharide molecules generally occupy larger hydrodynamic volumes, which give rise to higher solution viscosities. This assumes that intermolecular interactions are absent or negligible (e.g., in dilute solution). When intermolecular interactions are present among neighboring polymer molecules (i.e., polysaccharide-polysaccharide or protein-protein), the rheological properties of many systems are expected to change significantly. The changes in rheological properties may arise as a result of an increase in the size of the particles (e.g., protein-polysaccharide complexes), when depletion interactions occur in the mixed system, or if one or more polymer species forms continuous network structures. The overall effect results in the formation of different microstructures. Schematic illustrations of some possible microstructures formed from mixtures of proteins and polysaccharides under some specific conditions (e.g., pH, ionic strength, and heat treatment) are shown in Figs. 13.5A and 13.5B. Various rheological techniques have been employed to characterize the physicochemical properties of protein-polysaccharide systems. Generally, if the mixtures are liquid-like, rotational viscometers are commonly used to obtain viscosity measurements and hence steadystate viscosity curves, yield stresses, etc. Other simpler methods include the use of a kinematic viscometer (e.g., Ubbelohde capillary viscometer) to obtain a single point relative viscosity measurement. If the samples are viscoelastic (e.g., gels), rheometers are widely used to obtain rheological data (e.g., loss and storage moduli obtained within the linear viscoelastic region) by performing small deformation oscillatory measurements. The rheological data yield information on the viscosity and the viscoelastic properties of the mixed systems. Knowledge of the rheological properties of mixed protein-polysaccharide systems is essential to gain insights into the nature of the interactions and the resulting microstructure of the system. A fundamental understanding of the interaction at the molecular and colloidal levels will provide a strong foundation for exploiting the physical functionality of such complex systems in different applications (e.g., microencapsulation technology, imparting specific sensory characteristics, time/temperature/pH/ionic control release, and emulsion stability). In the following sections, we provide various examples of mixed systems involving different milk proteins and polysaccharides. An attempt is made to classify these mixed systems

Rheological properties and microstructures of protein-polysaccharide systems

517

into two broad categories (i.e., interacting and noninteracting). Under each of these headings, the systems are further grouped according to whether they do or do not form gels (i.e., gelling or nongelling). The discussion focuses mainly on the techniques used and the rheological properties of the systems.

Noninteracting protein-polysaccharide mixtures Noninteracting protein-polysaccharide mixtures existing as one phase are rare but may occur when the two different molecular species have hydrophilic surfaces that are chemically similar (Tolstoguzov, 1991, 2006). However, many polymer mixtures are thermodynamically incompatible, and segregative interactions often occur in the absence of electrostatic Noninteracting mixtures/Nongelling

(A)

(B)

Noninteracting mixtures/Gelling system

(C)

(D)

(F)

(E)

(G)

FIG. 13.5A Schematic diagrams of some possible microstructures formed between noninteracting proteinpolysaccharide mixtures. Circle (●) represents protein molecules; coil structure represents polysaccharide molecules. (A) A flocculated protein network formed with polysaccharide filling the space in the network; (B) polysaccharide molecules overlap and form a continuous “network” with protein filling the space; (C) a particulate protein gel network formed with polysaccharide filling the space; (D) a polysaccharide gel network formed with protein filling the space; (E) a bicontinuous network formed from protein and polysaccharide; (F) polysaccharide gels dispersed among a weakly flocculated protein network; (G) protein gels dispersed among entangled polysaccharide molecules.

518

13. Milk protein-polysaccharide interactions

FIG. 13.5B Schematic diagrams of some possible microstructures formed between interacting protein-polysaccharide mixtures. Circle (●) represents protein molecules; coil structure represents polysaccharide molecules. (A) Protein-polysaccharide complexes formed, (B) protein interacting with gelling polysaccharide helices, (C) polysaccharide interacting with protein particulate gel network, and (D) polysaccharide gel helices interacting with protein particulate gel network.

Interacting mixtures/ Nongelling

(A) Interacting mixtures/ Gelling

(B)

(C)

(D)

interactions or in the presence of electrostatic repulsions (Neiser et al., 1998). Proteinpolysaccharide mixtures that commonly exist as two separate phases are the result of either thermodynamic incompatibility or a depletion phenomenon (Doublier et al., 2000). Nongelling phase-separated systems The following are examples of noninteracting protein and polysaccharide mixtures. Both proteins and polysaccharides were mixed under conditions where the mixtures did not form gels. The rheological properties of these systems are discussed in relation to their interactions and the microstructures formed. Casein micelles and galactomannans

An example of a noninteracting protein-polysaccharide mixture in which phase separation occurs is a mixed system consisting of micellar casein (3%) and guar gum (0.2%) at pH 7 (Bourriot et al., 1999c). Rheological examination showed a significant change in the flow and viscoelastic properties compared with the individual biopolymer system. With the mixed system, there was an increase in the apparent viscosity. Furthermore, the mechanical spectra (elastic modulus G’ and viscous modulus G”) of the frequency sweeps showed slightly higher values of the moduli, which were less frequency dependent. The results suggested the formation of a weak network structure within the system, caused by flocculation of the casein micelles as the polysaccharide molecules were excluded from the protein phase. The appearance of slightly thixotropic behavior indicated that the network could be easily broken under shear; the network formed by the micellar casein was weakly flocculated and reversible, which presumably could be attributed to the depletion flocculation mechanism. The study also showed that the lower was the intrinsic viscosity of the polysaccharide and the higher was the concentration of polysaccharide required before phase separation occurred

Rheological properties and microstructures of protein-polysaccharide systems

519

(Bourriot et al., 1999c). An increase in the concentration of polysaccharide resulted in stronger flocculation of the casein micelles as the volume occupied and the osmotic pressure from the surrounding polysaccharides increased. Similar thixotropic behavior has been reported for a ternary solution consisting of micellar casein/locust bean gum (LBG)/sucrose (Schorsch et al., 1999). At pH 6.8, the casein micelles and the LBG were thermodynamically incompatible, behaving as a water-in-water emulsion. The presence of sucrose, even at high concentration (40%), did not significantly improve the compatibility of the biopolymers (Schorsch et al., 1999). Milk proteins and xanthan gum

Hemar et al. (2001a) investigated the interaction between xanthan gum (0%–1% w/w, a polysaccharide with known “weak gel” properties) and different types of milk protein (5% w/w, sodium caseinate [Na-CN], skim milk powder [SMP], whey protein isolate [WPI], and milk protein concentrate [MPC]) in aqueous solution at neutral pH. Depending on the xanthan gum concentrations and the protein type, the microstructures of the mixtures were different. For xanthan gum mixtures with either MPC or SMP, depletion flocculation of the casein micelles took place. The size of the depleted protein aggregates decreased with increasing xanthan gum concentration (the microstructure resembled a particulate network). For xanthan gum solutions containing either Na-CN or ultracentrifuged WPI, no phase separation occurred within the timescale of the experiment. This was attributed to the larger size of the casein micelles (average diameter 0.2 μm) compared with the nanometer-size scale of WPI and Na-CN (0.05 μm) (Lucey et al., 2000). However, the rheological behavior of the mixtures was found to be very similar to that of xanthan gum. The differences in the microstructures of the mixtures, observed using confocal laser scanning microscopy (CLSM), were not detected by viscosity measurements, probably because the weakly flocculated proteins were easily redispersed by the shearing action of the viscometer during measurement. Gelling phase-separated systems In a system where two biopolymer species (e.g., proteins and polysaccharides) do not interact, gelation of one or more of the components in a thermodynamically incompatible system will cause competition between phase separation and gelation (Neiser et al., 1998). Gelation basically means the formation of a three-dimensional aggregated network structure, which is generally induced by heating, cooling, acidification, enzymatic treatments, highpressure processing, etc. Generally, heating enhances hydrophobic and covalent interactions. Unfolded globular whey proteins interact to give rise to aggregates (Kinsella, 1984; Boye et al., 1997). In mixed systems, the microstructure will depend on the rates of phase separation and gel formation (Tavares et al., 2005). The gel may appear to be homogeneous at a macroscopic level but heterogeneous at the microscopic level. However, the rheological properties of such gels depend on the concentration and arrangement of each species in the different phases. The gel strength is higher if the gelling species is in the continuous phase than if it is in the dispersed phase where the network is disrupted (Neiser et al., 1998). Whey protein and galactomannans

Tavares and Lopes da Silva (2003) studied a mixture of LBG (a nongelling neutral polysaccharide) and whey protein at neutral pH and pH 5 (close to the pI of whey proteins). At

520

13. Milk protein-polysaccharide interactions

neutral pH, it is known that whey protein forms clear fine-stranded gels (protein aggregation is hindered by electrostatic repulsion), whereas, at lower pH (e.g., pH 5), an opaque coarse particulate gel is formed (Langton and Hermansson, 1992; Aguilera, 1995). Rheological measurements have shown that a whey protein isolate (WPI) gel (13% w/w) has a stronger and more elastic character at pH 5 than at pH 7 because of the thick particulate network formed (Stading et al., 1993; Bertrand and Turgeon, 2007). For the protein gels at pH 7, increasing the LBG concentration (>0.25%) decreased the onset temperature for gelation and decreased the gelation time. The presence of LBG was also found to increase the rigidity of the gel. Tavares and Lopes da Silva (2003) attributed this to a decrease in macromolecular mobility within the network in the presence of LBG because of segregative interactions and the “local” concentration of each polymer species. The LBG molecules acted as fillers in the continuous protein network. At pH 5, the elastic character of the particulate gel network was shown to decrease in the presence of LBG, especially at low protein concentration (5%). It was suggested that LBG chains hampered protein-protein interactions and were detrimental to the development of the protein gel. However, at a higher protein concentration (13%), at which sufficient particulate gel network was formed, the LBG molecules acted as fillers within the network, hence improving the gel strength. In a subsequent study using WPI and guar gum at pH 7, an increase in protein gel strength was found with a decreasing degree of branching of the galactomannans (Tavares et al., 2005). Like LBG, the guar gum was dispersed as droplets among the whey protein network at low concentration (0.2%). However, at higher guar gum concentration (0.6%), the dispersed droplets joined to form a continuous polysaccharide-rich phase. Despite the different microstructures observed, the linear viscoelastic profiles were rather similar, indicating that the viscoelasticity was fairly insensitive to microstructural changes of this nature. WPI and xanthan gum

A very similar trend was observed for whey protein and xanthan gum mixtures after heat treatment (Bertrand and Turgeon, 2007). The microstructures and rheological properties of the gels were highly dependent on the pH and the salt concentration. At pH 6.5, the presence of xanthan gum improved the elastic modulus of the WPI gel. This was attributed to segregative phase separation, where xanthan gum was dispersed among the protein gel network. However, lowering the pH decreased the elastic character of the gel. At pH 5.5 (close to the pI of WPI), the addition of xanthan gum decreased the elastic modulus of the gel. It was suggested that WPI-xanthan gum complexes were possibly formed and decreased the protein-protein interactions, producing a weaker gel network. β-Lactoglobulin and pectin

A different type of network was formed in mixtures of β-lactoglobulin (8% w/w) and low methoxyl (LM) pectin (0.85% w/w) after thermal treatment at pH 6.8. The storage modulus of the mixed gel system was significantly lower than that of the protein gel alone. The microstructure observed by CLSM revealed phase separation, with β-lactoglobulin appearing as spherical colloidal particles distributed in a continuous pectin network (Donato et al., 2005). A similar type of protein depletion-induced phase separation was reported for a mixed system containing aggregated whey protein and an exopolysaccharide (EPS) from lactic acid bacteria (Tuinier et al., 2000).

Rheological properties and microstructures of protein-polysaccharide systems

521

κ-Carrageenan and β-lactoglobulin

If two gelling species are present in a binary system, the mixed gels may form interpenetrating, coupled, or phase-separated networks (Morris, 1986). Interpenetrating networks are the result of two independent continuous networks formed throughout the gel, with only topological interactions existing between the networks. Coupled networks (ordered into junction zones like those of a polysaccharide gel) are formed when favorable interactions between the two molecular species exist. However, such systems involving proteinpolysaccharide interactions are uncommon (Rao, 1999). Phase-separated networks are formed when one polymer species is incompatible with the other, forming phase-separated regions within the network gel (Piculell and Lindman, 1992; Turgeon and Beaulieu, 2001). An example of a phase-separated gel is κ-carrageenan and βlactoglobulin (Capron et al., 1999). The mixed polymer formed a weaker gel than the carrageenan gel alone when the protein was in its native state. Upon heating the mixture to 90°C, holding for 30 min, and then cooling to 20°C, the gel rheology indicated the melting of κcarrageenan and the gelation of β-lactoglobulin above 65°C. There was no aggregation of κ-carrageenan with β-lactoglobulin upon heating. The gelation time of the β-lactoglobulin was reduced in the presence of κ-carrageenan, which was attributed to micro phase separation causing an increase in local concentration of the β-lactoglobulin (Capron et al., 1999). Upon cooling, the mixed gel system formed a phase-separated bicontinuous network (Eleya and Turgeon, 2000).

Interacting protein-polysaccharide mixtures Another phase separation phenomenon is associative phase separation, in which associative interactions are present. Associative interactions between proteins and polysaccharides can occur as a result of electrostatic interactions, hydrogen bonding, hydrophobic interactions, or poor solvent conditions (Antonov et al., 1996; Gao and Dubin, 1999; Doublier et al., 2000; de Kruif et al., 2004). In some cases, complexes are formed via electrostatic interactions (known as coacervates). Coacervates of protein and polysaccharide can occur when the pH of the mixture is lower than the isoelectric point of the protein. At this pH, proteins possess a net positive charge, whereas polysaccharides still possess a negative charge. The result of the complexation is the formation of a solvent-rich phase and a coacervate-rich phase (Doublier et al., 2000; Eleya and Turgeon, 2000). The rheological properties of milk protein-polysaccharide complexes are related to the interaction between the complexes and the water molecules, which form soluble (or liquid coacervate phase) or insoluble (or precipitate) complexes. The solubility of the complexes is based on the energetic difference between biopolymer-biopolymer and biopolymersolvent interactions (Damodaran, 1997). The main parameters that affect the solubility of biopolymer complexes are charge density, pH, ionic strength, and protein/polysaccharide (PP/PS) ratio (Schmitt et al., 1998). It has been suggested that a complex involving a strong polyelectrolyte will form a precipitate rather than a liquid coacervate. A number of proteinpolysaccharide systems in which complex coacervations occur have been reviewed (Schmitt et al., 1998; Turgeon et al., 2003; de Kruif et al., 2004; Xiao et al., 2014; Azarikia and Abbasi, 2016b; Gorji et al., 2018). The following presents some examples of interacting polymers in mixed systems and their effect on the rheological properties.

522

13. Milk protein-polysaccharide interactions

Nongelling phase-separated systems β-Lactoglobulin and chitosan

It has been reported that the solubility of protein increases below its isoelectric pH when it complexes with an anionic polysaccharide (Tolstoguzov et al., 1985; Tolstogusov, 1986). A study on a β-lactoglobulin-chitosan complex showed that, depending on the pH, the complex was either soluble or insoluble (Guzey and McClements, 2006a). The interaction of soluble chitosan (MW 15,000 Da, degree of acetylation 85%, 0–0.1 wt%, and 5-mM phosphate buffer) with β-lactoglobulin (0.5 wt% β-lactoglobulin and 5-mM phosphate buffer) in aqueous solutions studied at pH 3–7 showed that at pH 3, 4, and 5, the majority of the β-lactoglobulinchitosan complex in the solutions was soluble, but that, at pH 6 and 7, a significant fraction of the two biopolymers was insoluble. Whey proteins and exopolysaccharides

“Soluble complexes” formed via electrostatic interactions were reported for EPS B40 (an exopolysaccharide from Lactococcus lactis subsp. cremoris NIZO B40) and whey proteins (PP/PS ¼ 2:1) under specific pH and ionic conditions (with no macroscopic phase separation) (Weinbreck et al., 2003). Decreasing the pH of the mixtures further increased aggregation of the complexes, which led to phase separation. In addition, increasing the ionic strength of the solution caused a shift to lower pH values for the onset of complexation. In this study, complexation in this system led to a decrease in the solution viscosity, as intramolecular repulsion of the EPS was reduced in the presence of whey proteins. The decrease in viscosity was attributed to a reduction in the quantity of dispersed phase, that is, water present within the complexes. Consequently, it was suggested that viscosity measurement in dilute solution (which is related to the size of the complexes) could be used to determine the optimum conditions for complexation (Weinbreck et al., 2003). A potential benefit of this complexation is that it protects the protein from loss of solubility caused by aggregation during thermal or high-pressure treatments (Imeson, 1977; Galazka et al., 1997). Whey proteins and gum arabic

Viscosity curves were obtained to evaluate the “strength” of electrostatic interactions of whey protein/gum arabic coacervates (Weinbreck and Wientjes, 2004). This study showed that the stronger was the interaction, the greater was the shear-thinning behavior, and the slower was the reformation of the complexes after shearing. The highly viscous coacervates at pH 4 were considered to be due to electrostatic interactions. At pH above the isoelectric point (without electrostatic interactions), the mixtures appeared to be more elastic than viscous. Sodium caseinate and gum arabic

In contrast to whey proteins, sodium caseinate and gum arabic mixtures showed some peculiar behavior (Ye et al., 2006) as no coacervation was observed in these systems. Below a certain pH (pH 5.4), electrostatic interactions between sodium caseinate and gum arabic led to the formation of stable composite particles in the size range from 100 to 200 nm. These particle complexes remained constant in size and were stable and soluble over a defined pH range (pH 3.2–5.4). This pH range was dependent on the ratio of sodium caseinate to gum

Rheological properties and microstructures of protein-polysaccharide systems

523

arabic in the mixtures and also on the ionic strength. The sodium caseinate/gum arabic particles associated to form large particles, which resulted in phase separation when the pH was lower than 3.0. A mechanism for the formation of these particles, based around the selfaggregation of casein and the electrostatic interaction between the aggregated particles of casein and gum arabic molecules, has been proposed. As the pH of the mixture decreases below pH 5.4, the caseinate molecules tend toward small-scale aggregation prior to large-scale aggregation and precipitation at pH values closer to their pI (pH 4.6). In this case, the gum arabic molecules may attach to the outside of these small-scale aggregates in the early stages of aggregation through electrostatic interactions between negatively charged gum arabic and exposed positive patches on the surface of the caseinate aggregates. The presence of hydrophilic gum arabic molecules on the outside of the caseinate aggregate may be enough to sterically stabilize these nanoparticles and consequently prevent self-aggregation. As the charge on these particles is quite low, for example, 15 mV at pH 4.0, steric stabilization is probably important. In an another study, the formation of sodium caseinate-gum arabic complexes was reported to occur at temperatures above 60°C at a certain mass ratio of protein to gum arabic (e.g., 1:5) and pH (maximum complexation at pH 6.5) (Ye et al., 2012). Interestingly, the complex formation was reversible when the temperature was decreased below 60°C (although not in the case of pH 5.0). The temperature-dependent complexation between sodium caseinate and gum arabic was attributed to hydrophobic interaction between the two polymer molecules. These unique complexes can potentially be used to form interfacial layers of emulsion droplets that can be altered by temperature. Casein micelles and pectin

Protein-polysaccharide interactions were shown to be pH dependent, as exemplified by pectin and casein micelles (Ambjerg and Jørgensen, 1991; Maroziene and de Kruif, 2000). At pH 6.7, pectin did not adsorb on to the casein micelles. With sufficient pectin present (0.1%–0.2%), phase separation occurred because of depletion interactions of the casein micelles (0.1%). However, adsorption of pectin on to the casein micelles occurred at pH 5.3. Viscosity measurements were employed to study the changes that occurred at different polymer concentrations. At low pectin concentrations (0.1%) and at pH 5.3, bridging flocculation occurred. A maximum viscosity at this pectin concentration was attributed to bridging flocculation, as bridging among the casein particles was translated as having a larger effective volume. As the pectin concentration increased (>0.1%), the casein micelles became fully covered, and interactions between the casein particles were reduced. When the protein was fully covered by the pectin, the viscosity decreased to a certain extent but remained higher than that of the pure milk samples (without pectin). The amount of pectin required for full coverage of the casein micelles differed depending on the type of pectin: HM pectin < low methoxyl amidated (LMA) pectin < LM pectin. Adding more pectin beyond the concentration for full coverage led to phase separation because of depletion interactions. A further increase in pectin reduced the thickness of the casein-depleted layer, as the viscosity of the continuous phase became very high and gelled polymer networks were formed (Maroziene and de Kruif, 2000). When the pH of the mixture was increased from 5.3 to 6.7, pectin desorbed from the casein but over a much longer timescale (10–15 min) than the adsorption process (Maroziene and de Kruif, 2000).

524

13. Milk protein-polysaccharide interactions

Gelling phase-separated systems Sodium caseinate and pectin

The dynamic rheological properties of glucono-δ-lactone (GDL)-acidified protein gels (2% w/v sodium caseinate) were studied in the presence of LMA pectin (0.01%–1% w/v) at pH 4 (Matia-Merino et al., 2004). The presence of pectin (0.01%–0.05% w/v) decreased the storage modulus and increased the gelation time as pectin adsorbed on to the casein particles. At pectin concentrations > 0.08% w/v, acid-induced gelation appeared to be completely inhibited over a timescale of 9 h at 25°C. Casein micelles and ι-carrageenan

In casein-carrageenan mixed systems, the negatively charged sulfated groups of the polysaccharide and the positive “patches” between residues 97 and 112 of ĸ-casein were involved in attractive interactions (Snoeren, 1975) despite a pH above the isoelectric point and an overall net negative charge of the casein micelles. The interaction between ι-carrageenan (0.5%) and skim milk (based on 3.3% protein) mixtures was studied above and below carrageenan’s coil-helix transition temperature (Langendorff et al., 1999). At temperatures above the coilhelix transition temperature, carrageenan did not adsorb to casein micelles, resulting in depletion flocculation. In contrast, at temperatures below the coil-helix transition temperature, attractive interactions between carrageenan and casein micelles occurred. The higher charge density of the double-helix form compared with the coil conformation of carrageenan probably explained the stronger attractive interaction between casein micelles and carrageenan. The presence of casein micelles increased the gel strength (indicated by higher G’ and G”) and the gelation temperature (from 39 to 47°C) when the mixtures were heated to 65°C and cooled to 25°C. Depending on the concentration of carrageenan, different types of gel network were deduced from the frequency sweep. At low carrageenan concentrations ( αS1- > β- > κ-casein (see Chapter 6). Increased binding of calcium to the caseins results in reduced negative charges on the casein molecule, resulting in diminished electrostatic repulsion and consequently inducing precipitation. Caseins with high numbers of phosphoserine residues, such as αS1-casein B, αS1-casein C, and the αS2-caseins, are insoluble at Ca2+ concentrations above about 4 mM (Singh and Flanagan, 2005). However, β-casein is soluble at high Ca2+ concentrations (0.4 M) at temperatures below 18°C but is very insoluble above 18°C, even in the presence of low Ca2+ concentrations (4 mM). κ-Casein, with only one phosphoserine, binds little calcium and remains soluble at all Ca2+ concentrations. Although κ-casein does not bind calcium to any great extent, its ability to stabilize αS1-, αS2-, and β-caseins against precipitation by Ca2+ is well known, and κ-casein plays a large part in the stabilization of the casein micelle. This is discussed in more detail in Chapter 6. It is worth noting the use of more advanced techniques to probe the casein micelle. Recent activity in this area has seen the application of small angle neutron and X-ray scattering to further probe the casein micelle and in particular the interaction between the casein moieties and the minerals in milk (Smialowska et al., 2017). Small angle X-ray scattering analysis found that three scattering objects on a range of length scales were present and depended on the level of calcium present in the system. Overall, this showed the significant effect that the interaction of calcium had on the size of the casein aggregates formed and the overall structuring of caseinate. Sugiarto et al. (2009) tested whether sodium caseinate and/or WPI could bind and solubilize iron (ferrous sulfate) for food fortification. Caseinate had more binding sites than WPI, and iron was bound more strongly to caseinate, but caseinate was increasingly precipitated at >4 mM Fe. Caseinate-iron complexes with 2-mM Fe remained soluble as the pH was decreased from 7 to 5.5, whereas the solubility of WPI-iron complexes decreased with decreasing pH. Chelation of iron with milk proteins mitigated iron-catalyzed oxidation in emulsions, although some contribution from antioxidant amino acid side chains was also postulated (Sugiarto et al., 2010). More recent work in this area found that prior depletion of calcium from milk systems can dramatically improve the binding of iron to the casein proteins present, allowing much higher levels of iron to be stabilized for food fortification (Das et al., 2013; Mittal et al., 2015). Further studies by the same researchers found that, with a caseinate-based system, a three-way complex between casein, iron, and phosphate was formed (Fig. 14.2), with the ratio of the three being critical to the overall stability of the complex, as was confirmed by 31P-nuclear magnetic resonance (Mittal et al., 2016, 2018). By targeting hydrolyzates with iron-chelating abilities, it was found that the iron solubility could be enhanced by binding ferrous iron to whey protein hydrolyzates, which were fractionated using cascade membrane filtration (O’Loughlin et al., 2015). Intact casein will bind zinc and calcium, but tryptic hydrolyzates of αS1-, αS2-, β-, and κ-caseins also display mineral-binding properties. Termed caseinophosphopeptides (CPPs), these peptides can bind and solubilize high concentrations of calcium because of their highly polar acidic domain. Calcium-binding CPPs can have an anticariogenic effect,

544

14. Interaction between milk proteins and micronutrients

FIG. 14.2 Schematic diagram of the three-way iron-phosphate-casein complex, which imparts increased stability and solubility to the iron within the system.

in that they inhibit caries lesions through recalcification of the dental enamel (FitzGerald, 1998). This effect has been exploited in CPP-fortified chewing gums (Recaldent and Trident brands) using ingredients developed by CSIRO Australia. CPPs have also been reported to improve the intestinal absorption of zinc, as studied using an isolated perfused rat intestinal loop system (Peres et al., 1998). The amount of iron bound to CPPs produced with the enzyme Alcalase depends on the degree of hydrolysis and the temperature and pH during the binding reaction (Wang et al., 2011). Binding of iron to CPPs reduces iron-induced peroxidation in CaCo-2 cells, suggesting that CPPs could help to mitigate against unintended side effects of iron fortification (Kibangou et al., 2008). Other studies have also examined the calcium-binding affinity of CCPs, with the actual peptides responsible for the binding of calcium and magnesium purified from the CPP fractions. From this work, five phosphopeptides from CPPs were identified as having excellent mineral-binding properties (Cao et al., 2019). Enzymatic hydrolysis of β-LG dramatically increases its iron-binding capacity, which may be due to improved contact between iron and aromatic amino acids (Zhou et al., 2012). Lactoferrin has the ability to bind iron very strongly. In vivo, the ferric (III) form of iron is bound to lactoferrin (Anderson et al., 1989). Considerable interest has been expressed in supplementing bovine milk-based infant formulas with lactoferrin, as bovine milk contains much lower levels of lactoferrin than human milk and lactoferrin, isolated from human milk, can bind 2 mol of iron per mole of protein (Bezwoda and Mansoor, 1986). Nagasako et al. (1993) reported that lactoferrin can bind iron at sites other than its chelate-binding sites, probably on the surface of the molecule. The thermal stability of lactoferrin-iron complexes is enhanced by soluble soybean polysaccharide, which was apparently due to enhanced electrostatic repulsion (Ueno et al., 2012). Other studies involving the interactions of minerals/ions and milk proteins are listed in Table 14.1.

545

Interaction between milk proteins and micronutrients

TABLE 14.1 Interaction of milk proteins and minerals Milk protein

Mineral

Reference

β-LG A

Chromium

Divsalar et al. (2006b)

β-LG A and B

Lead

Divsalar and Saboury (2005)

Caseins and β-LG

Mercury

Mata et al. (1997)

Sodium caseinate

Iron

Shilpashree et al. (2015)

Casein phosphopeptides

Iron

Delshadian et al. (2018)

α-LA

Copper

Permyakov et al. (1988)

Whey protein concentrate

Iron; zinc

Shilpashree et al. (2016)

Whey protein hydrolyzates

Iron

Caetano-Silva et al. (2015)

Fatty acids Most of the fatty acids present in milk are found as triglycerides, which form the fat globule. P
erez et al. (1992) proposed that ruminant β-LG, because of its activity to bind fatty acids, might play a role in the activity of pregastric lipases. P
erez et al. (1989) demonstrated that two types of lipid, namely, free fatty acids and triglycerides, bound to β-LG. The total amount of fatty acids extracted from β-LG was 0.71 mol/mol of monomer protein. The predominant fatty acids were palmitic (31%–35%), oleic (22%–23%), and myristic (14%–17%) acids, which combined account for 66%–75% of the total fatty acids bound to β-LG. The unsaturated fatty acids extracted from β-LG were α-LA is consistent. Table 14.2 illustrates the interactions between 2-nonanone and various milk proteins (K€ uhn et al., 2006).

Other micronutrients Some of the milk proteins, most particularly the whey proteins, have been used as model proteins in studies involving a range of other micronutrients. The interaction of small heatshock proteins, such as alpha-crystallin, prevents the precipitation of α-LA when in the molten globule state (Lindner et al., 1997). This finding was confirmed by Sreelakshmi and Sharma (2001), who found that the active site of alpha-crystallin by itself can maintain a significantly denatured and unfolded protein in soluble form. Zhang et al. (2005) reported on the chaperone-like activity of β- and α-caseins. β-Casein was able to suppress the thermal and chemical aggregation of insulin, lysozyme, and catalase. A similar chaperone-like effect is seen with β-LG, α-LA, and BSA (Kehoe and Foegeding, 2011). The use of milk proteins as chaperones for drugs has also been studied. The interaction of chlorpromazine with β-LG and αs-casein affected the proteins in different ways. Far UV circular dichroism studies revealed that chlorpromazine increased the secondary structure of β-LG, whereas the structure of casein became further disordered (Bhattacharyya and Das, 2001). Divsalar et al. (2006a) also reported on the interaction between genetic variants of β-LG and an anticancer component. A number of the most recent studies that have examined the interactions between milk proteins and various bioactive compounds are listed in Table 14.3. Most studies were carried out with a view to using milk proteins as carrier molecules or particles for protecting and/or delivering bioactive compounds, thereby increasing their bioaccessibility to the body. The potential application of milk proteins as delivery systems or carriers for a range of

TABLE 14.2 Binding data for the interactions between 2-nonanone and milk proteins (25°C): n, number of binding sites per monomer; K, intrinsic binding constant Protein a

WPC

b

n

K (M21)

Method

Reference

61

1,920,000

Equilibrium dialysis

Jasinski and Kilara (1985)

0.2

53,000,000

Fluorescence spectroscopy

Liu et al. (2005)

c

WPI

1

2059

Headspace SPME

Zhu (2003)

Sodium caseinate

0.3

1858

Headspace SPME

Zhu (2003)

β-LG

1

2439

Equilibrium dialysis

O’Neill and Kinsella (1987)

0.2 0.5

6250 (40 ppm) 1667 (45 ppm)

Static headspace analysis

Charles et al. (1996)

14

122

Equilibrium dialysis

Jasinski and Kilara (1985)

α-LA

33

11

Equilibrium dialysis

Jasinski and Kilara (1985)

BSA

5–6

1800

Liquid-liquid partitioning

Damodaran and Kinsella (1980)

15

14,100

Equilibrium dialysis

Jasinski and Kilara (1985)

7

833

d

PFG-NMR spectroscopy

Jung et al. (2002)

a

WPC, whey protein concentrate. WPI, whey protein isolate. c SPME, solid-phase microextraction. d PFG-NMR, pulsed-field gradient nuclear magnetic resonance. € J., Considine, T., Singh, H., 2006. Interactions of milk proteins and volatile compounds: implications in Reproduced with the permission of Kuhn, the development of protein foods. J. Food Sci. 71, R72–R82; copyright 2006 Journal of Food Science, Institute of Food Technologists. b

TABLE 14.3 Recent studies of milk protein interactions with miscellaneous biologically active compounds Protein

Active agent

Notes

Reference

Epigallocatechin

Binding study

Wu et al. (2011)

Epigallocatechin-3-gallate

Binding study

Wu et al. (2013)

Epigallocatechin-3-gallate, chlorogenic acid, ferulic acid

Binding study

Jia et al. (2017)

Tea polyphenols

Binding study

Kanakis et al. (2011)

Polyphenol extracts (tea, coffee, cocoa)

Binding study

Stojadinovic et al. (2013)

Curcumin

Binding; encapsulation

Sneharani et al. (2010)

Apigenin, naringenin kaempferol, genistein

Binding study

Li et al. (2018)

Trans-resveratrol; curcumin

Binding study

Mohammadi and Moeeni (2015)

Epigallocatechin-3-gallate

Binding study

Al-Hanish et al. (2016); Radibratovic et al. (2019)

Polyphenolic compounds β-LG

α-LA

Continued

554

14. Interaction between milk proteins and micronutrients

TABLE 14.3 Recent studies of milk protein interactions with miscellaneous biologically active compounds—cont’d Protein

Active agent

Notes

Reference

Casein

Quercetin

Chitosan-casein nanoparticles

Ha et al. (2013)

Polymethoxyflavones

Binding study

He et al. (2013)

Curcumin (turmeric)

Binding study

Benzaria et al. (2013) Rahimi Yazdi and Corredig (2012)

Tannic acid

Binding study

Zhan et al. (2018)

Naringenin

Binding study

Moeiniafshari et al. (2015)

Eriocitrin

Binding study

Cao et al. (2019)

Flavonoids

β-Casein showed strongest interactions

Bohin et al. (2012)

Pelargonidin

Strongest interaction with casein and β-LG

Arroyo-Maya et al. (2016)

Polyphenols Polyphenols

Protein sequence influence on noncovalent binding Binding studies and structural changes

Nagy et al. (2012) Yildirim-Elikoglu and Erdem (2018)

Green tea catechins

Effect of milk proteins on bioaccessibility

Xie et al. (2013)

Cyanidin-3-O-glucoside

Binding study

Cheng et al. (2017)

Trans- and cis-resveratrol Tea polyphenols

Complexation study Review article

Cheng et al. (2018) Chanphai et al. (2018)

β-Casein

Various milk proteins

Various bioactive compounds α- and β-Caseins

Folic acid

Various milk proteins

Norbixin

Reassembled casein micelles

n-3 Polyunsaturated fatty acids

β-LG

Piperine (pepper alkaloid)

Bourassa and Tajmir-Riahi (2012) Cheese coloring agent

Zhang and Zhong (2013a,b) Zimet et al. (2011)

Binding study

Zsila et al. (2005)

Pharmaceutical compounds β-LG A

Bioactive peptides

Antihypertensive peptide

Roufik et al. (2006)

β-LG

Doxorubicin

Antibiotic

Agudelo et al. (2012)

Oxaliplatin

Anticancer drug

Ghalandari et al. (2015)

Casein

Alfuzosin

Prostate cancer drug in genipincrosslinked casein nanoparticles

Elzoghby et al. (2013)

β-Casein nanoparticles

Paclitaxel, vinblastine, mitoxantrone

Antitumor drugs

Shapira et al. (2010)

Lactoferrin

Gambogic acid

Antitumor compound

Zhang et al. (2013)

Effect of processing on milk protein structure

555

compounds has been examined in several in-depth reviews (Livney, 2010; Elzoghby et al., 2011; Abd El-Salam and El-Shibiny, 2012; Chanphai et al., 2018). The number of studies into the binding of polyphenols by milk proteins is especially notable. This area of research has seen an explosion of studies over the past decade and is by far the area of most significant development. A comprehensive review on the advances in the methods used to investigate these protein-phenolic compound interactions that were also carried out recently (Czubinski and Dwiecki, 2017).

Effect of processing on milk protein structure Heat has been used extensively in food processing for centuries and is a widely applied treatment in food production, primarily for the control of microbial populations. Fields of application are pasteurization under mild temperatures and sterilization under higher temperatures. However, heating may also affect texture and taste development and may result in flavor and color changes. The latter effects are often described as disadvantages of heat treatment. Changes in the organoleptic properties are generally as a result of structural changes occurring within the constituents of the food, namely, the proteins, polysaccharides, or fats. Another technology that is similar in its control of the microbial population of food products is high-pressure treatment. In contrast to heat processing, foods are preserved with minor changes in texture, flavor, or color, and high pressure can be considered to be a cold preservation technology. High pressure is a long-used technique in Japan and has become increasingly popular worldwide. However, high-pressure treatment may cause some conformational and structural changes to the individual constituents of the food, possibly resulting in altered functional and organoleptic properties. Both heat treatment and high-pressure treatment may cause the denaturation of globular whey proteins such as β-LG; although there may be differences in the mechanisms behind the denaturation process, the general process appears to be similar. Nonthermal processing is examined in detail in Chapter 8. The denaturation of whey proteins during the heat treatment of milk, the interactions of the denatured whey proteins with other milk components, and the effect of these reactions on the physical and functional properties of milk products have been extensively reported and reviewed in great detail (O’Connell and Fox, 2003; Singh and Havea, 2003). Studies have shown that heat-induced aggregation and gelation occur along detailed pathways and are influenced by the types of proteins and forces (disulfide bonding and hydrophobic interactions) present (Schokker et al., 1999; Havea et al., 2001; Abbasi and Dickinson, 2002). The use of heat to induce self-assembly and coassembly of milk proteins into micro-/nanoparticles is discussed in Loveday et al. (2012). The effect of high pressure on whole milk and individual constituents has become a subject of much recent activity, particularly regarding the effect of pressure treatment on the physical and functional properties of milk products (Anema, 2010) and the pressureinduced changes to individual proteins (Anema, 2012). Interested readers are referred to the reviews of Huppertz et al. (2006) and Considine et al. (2007a). The mechanistic effects of high-pressure processing and several other novel processing technologies were reviewed recently in the context of the industrial potential of these technologies in yogurt

556

14. Interaction between milk proteins and micronutrients

manufacture (Loveday et al., 2013). The use of high pressure in other dairy systems, such as whey or casein gels, has also been reviewed (Devi et al., 2013).

Protein denaturation by thermal and pressure treatments and effect of micronutrients The caseins have not been suitable candidates for observing changes in protein denaturation, because of their lack of defined secondary and tertiary structure. In contrast, the whey proteins have been studied widely as model globular proteins because of their well-defined secondary and tertiary structures, as outlined earlier. The interactions between whey proteins and other species that are induced by either heat treatment or pressure treatment may be divided into two separate classes: covalent interactions and noncovalent interactions. The most important covalent interaction involving whey proteins upon storage is their reaction with reducing sugars via the Maillard reaction to form discolored protein powders, which also have reduced solubilities and diminished nutritional properties. Noncovalent interactions can also occur; these may also lead to a loss of protein solubility after association of the proteins with polysaccharides, and these noncovalent interactions are driven primarily by reversible electrostatic interactions. In this chapter, the effects of noninteracting species on the unfolding and structural transitions of whey proteins are of specific interest. The marked increase in the thermal and conformational stability of globular proteins in aqueous media in the presence of sugars is well known and has been extensively studied.

Processing treatments involving ligands Several studies have shown that ligands can retard the heat or pressure denaturation of β-LG, and the type of ligand has an impact on this process. For example, during the heat denaturation of β-LG, both SDS and palmitate stabilized the native structure of β-LG against heat-induced structural flexibility, subsequent unfolding, and denaturation up to approximately 70°C, whereas both retinol and ANS provided very little stabilization (Considine et al., 2005a). When a similar range of ligands was used during pressure denaturation, a similar effect was noted, that is, higher pressures were required to cause unfolding of β-LG when a ligand was present (Considine et al., 2005b). It was noted in these studies and in the comparison study of heat and pressure using myristate and conjugated linoleic acid as ligands that β-LG unfolds slightly differently with respect to the type of treatment (Fig. 14.5) (Considine et al., 2007b). Barbiroli et al. (2011) have shown that endogenous ligands (mostly palmitic acid and stearic acid) bound to β-LG stabilize the tertiary structure against denaturation by urea or heat. They reported evidence that the binding of palmitic acid in the calyx enhanced the thermal stability of both the calyx region and the helix held against the outside of the β-barrel (the helix conceals the free thiol at Cys121). They believed that fatty acid binding in the calyx made the whole structure “tighter,” and inhibited the movement of the helix region and the exposure of Cys121, which is crucially involved in disulfide-bonded aggregation. In related work with synthetic ligands, Busti et al. (2006) reported that alkyl sulfonates with a chain length of >10 increased the denaturation temperature of β-LG at pH 6.8 by up to 13°C. In a recent study,

557

Effect of processing on milk protein structure

0.1 MPa

150 MPa

Native Stage I P SDS

Stage I P

ANS

Stage I P

Retinol

Stage I P

800 MPa

450 MPa

Stage II P

Stage III P Stage II P

Stage II P Stage II P

Stage III P Stage III P Stage III P

FIG. 14.5 Proposed three-stage model of the pressure denaturation of β-LG B and of β-LG B with added ANS, retinol, or SDS. Reproduced with the permission of Considine, T., Singh, H., Patel, H.A., Creamer, L.K., 2005b. Influence of binding of sodium dodecyl sulfate, all-trans-retinol, and 8-anilino-1-naphthalenesulfonate on the high-pressure-induced unfolding and aggregation of β-lactoglobulin B. J. Agric. Food Chem. 53, 8010–8018. Copyright 2005 Journal of Agricultural and Food Chemistry, American Chemical Society.

the conjugation of WPI with sugar beet pectin was found to increase the thermal stability of the secondary and tertiary structures of the protein. The conjugation was brought about through controlled dry heating (60°C, 79% relative humidity). It was surmised that the interactions (studied using circular dichroism and UV-visible spectroscopy alongside steady-state fluorescence spectroscopy) were covalent and noncovalent in nature (Qi et al., 2017). Hansted et al. (2011) conducted a detailed investigation of how surfactants affect thermally induced unfolding and aggregation of β-LG, using homologous series of cationic (alky trimethyl ammonium chlorides, xTAC), anionic (sodium alkyl sulfates, SxS), and nonionic (alkyl maltopyranosides, xM) surfactants. SxS inhibited thermal unfolding and aggregation at concentrations well below the critical micelle concentration, indicating that surfactant monomers were responsible for the effect. xM also inhibited aggregation, although only at above the critical micelle concentration, and smaller xM promoted unfolding at such concentrations. xTAC strongly promoted aggregation at subcritical micelle concentrations. The findings highlight the effect of the surfactant charge on aggregation at pH 6.5: anionic SxS and nonionic xM reduced aggregation, whereas cationic xTAC promoted aggregation. Hansted et al. (2011) also showed how the concentration of surfactants strongly modifies their effects, and they postulated surface interactions between β-LG and micelles of nonionic or cationic surfactants. Celej et al. (2005) compared the effects of the binding of two ANS derivatives, namely, 1,8-ANS and 2,6-ANS, on the thermostability of BSA. They reported that 1,8-ANS had a stronger effect on the thermal stability of BSA and that the binding parameters of the two ANS derivatives were quite different. This was thought to indicate that stereochemistry is an important factor in determining protein-ligand interactions. Thus, electrostatic interactions should also be considered, along with hydrophobic interactions. The authors emphasized the importance of the free ligand concentration rather than the ligand-to-protein mole ratio when determining protein stability. As discussed earlier, the binding of retinol to casein is through hydrophobic interactions (Poiffait and Adrian, 1991). β-Casein is the most hydrophobic casein and has a highly charged N-terminal domain, containing an anionic phosphoserine cluster that is clearly distinct from a very hydrophobic C-terminal domain (Swaisgood, 2003). There has been little work on the ability of the caseins to bind retinol, although Poiffait and Adrian (1991) reported that casein

558

14. Interaction between milk proteins and micronutrients

plays an important role in stabilizing retinol over time or during heat treatment. However, information in this area is limited.

Processing treatments involving sugars or polyols The effect of up to 70% (w/w) glycerol or sorbitol on the properties and functionality of β-LG was examined in several studies by Chanasattru and coworkers. Sorbitol strongly increased the thermal denaturation temperature of β-LG at pH 7, whereas glycerol had a very minor effect (Chanasattru et al., 2007b). This translated too much stronger gels with glycerol when 10% β-LG solutions were heated to 90°C for 70 min. Both polyols increased the complex modulus (G*) relative to controls, which was attributed to the strengthening of proteinprotein interactions, but the inhibitory effect of sorbitol on denaturation was thought to explain the low G* with this polyol. Later studies noted that glycerol decreases the surface tension at hexadecane-water interfaces, whereas sorbitol slightly increases it (Chanasattru et al., 2007c). The authors proposed that glycerol could interact with nonpolar regions on the surface of proteins in a way that counterbalanced steric exclusion effects, leading to small net effects on the denaturation temperature (Chanasattru et al., 2008). More recently, it was shown that, although these cosolutes tended to impart similar stabilization to the protein overall, they were actually stabilizing specific regions on β-LG. This finding could be of significant importance, given that some of these cosolutes interact with residues that may be important for the final functionality of the protein (Barbiroli et al., 2017). This group also studied the effects of polyols in β-LG-stabilized emulsions (Chanasattru et al., 2007a). Glycerol and sorbitol improved the emulsion stability against salt-induced flocculation to an approximately equal extent on a % (w/w) basis. This effect was attributed partly not only to increased viscosity (especially for sorbitol) but also to a predicted reduction in attractive van der Waals’ and hydrophobic interactions that was large enough to overcome a slight weakening of electrostatic repulsion. Similar studies on β-LG- and casein-stabilized emulsions were discussed by Dickinson (2010). The effect of small mono- and polyhydroxy alcohols on the thermal stability of β-LG at pH 5.5 was examined in more detail by Romero et al. (2007), using a homologous series of four-carbon alcohols with from one to four hydroxyl groups. All alcohols destabilized β-LG but to an extent that decreased as the number of hydroxyl groups increased. The authors proposed that 1-butanol was hydrophobic enough to interact with nonpolar regions on the surface of β-LG, whereas more hydroxylated (and therefore more hydrophilic) alcohols interacted preferentially with water instead of protein, and thereby had less destabilizing effect. This theory aligns well with the proposal from Chanasattru et al. (2008) that glycerol (1,2,3-propanetriol) interacts with nonpolar regions on the surface of the protein. Boye and Alli (2000) reported on the thermal denaturation of 1:1 mixtures of α-LA and β-LG in the presence of a range of sugars, using differential scanning calorimetry (DSC). Sugars protected against heat-induced denaturation and the protection offered (i.e., the size of the increase in the thermal transition temperature of β-LG) were in the order galactose ¼ glucose > fructose ¼ lactose > sucrose > sugar-free control. No significant effects of sugar were observed with apo-α-LA. Interestingly, an earlier study by the same authors solely on α-LA found an increase in the thermal transition temperature of both the apo form and the holo form of α-LA when either 50% sucrose or 50% glucose was added

Effect of processing on milk protein structure

559

(Boye et al., 1997). This increase was fully reversible in the holo form but only partly reversible in the apo form. The thermal transition temperature of β-LG was found to be increased in the presence of sucrose, lactose, and glucose at 10%–50% (Boye et al., 1996b). Jou and Harper (1996) found an increase in the DSC thermal transition temperature of whey protein concentrates following the addition of sugars, and the protection offered by the sugars was in the order maltose > trehalose > sucrose. Lactose was also found to provide some protection against heat-induced denaturation. Dierckx and Huyghebaert (2002) followed the heat-induced gelation of a WPI solution using DSC and small amplitude oscillatory rheometry. They found that, by adding increasing concentrations of sucrose or sorbitol, both the thermal transition temperature of the protein denaturation process and the gelation temperature were increased, with a linear relationship existing between the transition and gelation temperatures. They suggested that, because of the differences in the gelation mechanisms observed at different pH values, sucrose and sorbitol affected protein-protein interactions in gels through the enhancement of hydrophobic interactions. Kulmyrzaev et al. (2000) had previously conducted a study on the effect of sucrose on the thermal denaturation, gelation, and emulsion stabilization of WPI. They also observed increases in the thermal transition temperatures on the addition of increasing concentrations of sucrose and improved gel formation and enhanced emulsification flocculation. They postulated that sucrose played different roles in a predenatured (improved heat stability) and a postdenatured (enhanced protein-protein interactions) whey protein solution system. In a study on the effects of different lactose concentrations (within a naturally occurring range) on the formation of whey protein microparticulates, Spiegel (1999) put forward a two-stage process in the aggregation of whey proteins: up to approximately 85°C, the aggregation of whey proteins is limited by the slow unfolding of the individual proteins; above 100° C, aggregation is the rate-limiting step, as the rate of unfolding is high. Lactose (at 500 mM) was also found to increase the temperature of the denaturation of WPI at pH 9.0 by approximately 3°C. However, the authors realized the effect that the Maillard reaction was having in these systems a factor that some reports seem to ignore. Baier and McClements (2001) found that increased concentrations of sucrose (up to 40%) could increase the thermal stability of BSA. These systems had a higher gelation temperature and produced gels with a lower complex shear modulus. Similar effects were found in a subsequent study (Baier and McClements, 2003). A further study by the same group (Baier et al., 2004) showed that 40% glycerol increased the temperature of gelation of BSA, but no change in the temperature of denaturation of BSA with an increasing concentration of glycerol was detected. Some early DSC work (Dumay et al., 1994) showed that the presence of 5% sucrose was sufficient to reduce the extent of β-LG unfolding by 22% following high-pressure treatment at 450 MPa for 15 min. In a later study, Dumay et al. (1998) found that adding sucrose to β-LG solutions prior to pressure-induced gelation resulted in gels with decreased pore size and strand thickness. They attributed this to a reduction in the number of protein-protein interactions occurring under the influence of pressure. Keenan et al. (2001) reported that low concentrations of sucrose aided in the pressureinduced gel formation of a range of milk protein-containing systems, but that gel formation was reduced at higher sucrose concentrations. In another group of studies, the pressureinduced gelation properties of skim milk powder were found to be improved by adding

560

14. Interaction between milk proteins and micronutrients

low concentrations of sucrose, glucose, or fructose, whereas high (45%–50%) sugar concentrations inhibited gel formation (Abbasi and Dickinson, 2001). Boye et al. (1996a) described how lactose, sucrose, and glucose increased the temperature of denaturation of BSA, with 50% glucose having a greater stabilizing effect than 50% sucrose. Wendorf et al. (2004) studied the ability of different proteins (ribonuclease A, BSA, and egg white lysozyme) to adsorb to a liquid-solid interface in the presence of a range of sugars. They found that the ability of sugars to reduce protein adsorption followed the trend trisaccharides > disaccharides > six-carbon polyols > monosaccharides, and this was explained by the stabilization of the protein in the native state in solution. Other studies have also shown the beneficial effects of sugars in protecting against denaturation induced by freeze drying, spray drying, and chemicals. At low temperatures, high concentrations of sugar cause a substantial increase in the solution viscosity and can thus affect protein denaturation. Tang and Pikal (2005) showed that, by negating the thermal stabilizing effects of sucrose by adding denaturants, the increased stability of β-LG in the freeze drying process could be directly attributed to a viscosity effect. Murray and Liang (1999) explored the addition of sucrose, trehalose, lactose, and lactitol to whey protein concentrate solutions prior to spray drying and found that the foaming properties of the spray-dried powders were dramatically decreased when sugars were absent. Trehalose was particularly successful in retaining the original foaming properties of both whey protein concentrate and β-LG but did not perform as well in spray-dried BSA powders (Murray and Liang, 1999).

Conclusions The interaction of milk proteins with various micronutrients is primarily governed by the physicochemical properties of the proteins. The whey proteins, with extensive secondary and tertiary structure and significant hydrophobicity (albeit largely shielded in the native form), tend toward hydrophobic interactions with ligands. Preferential exclusion effects govern the interaction of sugars and polyols with proteins, thus affecting their denaturing properties in the presence of pressure or heat. Electrostatic interactions drive the association of minerals and proteins. In the food industry, an increasing emphasis is being placed on foods that will have a physiologically functional benefit, in addition to the nutritional benefit of the food. This is being driven by consumers who are becoming increasingly more health aware and health responsible. The challenge now for the food scientist is to deliver the required physiologically functional activities into the final food product, while retaining product quality and shelf life. Knowledge of the interactions of these micronutrients with milk proteins, a major component in many food products, is necessary to achieve this aim. Relevant examples of this concept are detailed in patents concerning the delivery of micronutrients in complexes with β-LG (Swaisgood, 2001) or casein micelles (Livney and Dalgleish, 2007). An explosion of studies in the area of polyphenol-protein interactions in particular has been evident over the past decade. Although the concept of using milk proteins as nutrient carriers has been explored in a range of protein-nutrient combinations, there is still relatively little knowledge about how

References

561

nutrient interactions can be used to manipulate the functionality of proteins during processing. The binding of ligands to certain whey proteins increases their resistance to thermal denaturation, and noninteracting solutes such as sugars can also stabilize proteins against heat processing. Mineral binding to caseins affects their solubility, which has obvious implications for beverage products. Greater knowledge of protein-micronutrient interactions will enable the use of milk proteins as nutrient carriers and will allow the use of micronutrients as processing aids; perhaps both objectives could even be achieved simultaneously.

References Abbasi, S., Dickinson, E., 2001. Influence of sugars on high-pressure induced gelation of skim milk dispersions. Food Hydrocoll. 15, 315–319. Abbasi, S., Dickinson, E., 2002. Influence of high-pressure treatment on gelation of skim milk powder + low methoxyl pectin dispersions. High Pressure Res. 22, 643–647. Abd El-Salam, M.H., El-Shibiny, S., 2012. Formation and potential uses of milk proteins as nano delivery vehicles for nutraceuticals: a review. Int. J. Dairy Technol. 65, 13–21. Adrian, J., Frangne, R., Rabache, M., 1984. Effect of milk-proteins on retinol efficiency. Sci. Aliment. 4, 305–308. Agudelo, D., Beauregard, M., B
erub
e, G., Tajmir-Riahi, H.A., 2012. Antibiotic doxorubicin and its derivative bind milk β-lactoglobulin. J. Photochem. Photobiol. B 117, 185–192. Al-Hanish, A., Stanic-Vucinic, D., Mihailovic, J., Prodic, I., Minic, S., Stojadinovic, M., Radibratovic, M., Milcic, M., Cirkovic Velickovic, T., 2016. Noncovalent interactions of bovine α-lactalbumin with green tea polyphenol, epigalocatechin-3-gallate. Food Hydrocoll. 61, 241–250. Allahdad, Z., Varidi, M., Zadmard, R., Saboury, A.A., Haertle, T., 2019. Binding of beta-carotene to whey proteins: multi-spectroscopic techniques and docking studies. Food Chem. 277, 96–106. Anderson, B.F., Baker, H.M., Norris, G.E., Rice, D.W., Baker, E.N., 1989. Structure of human lactoferrin: crystallo˚ resolution. J. Mol. Biol. 209, 711–734. graphic structure analysis and refinement at 2.8 A Andersson, L.O., Brandt, J., Johansson, S., 1971. The use of trinitrobenzenesulfonic acid in studies on the binding of fatty acid anions to bovine serum albumin. Arch. Biochem. Biophys. 146, 428–440. Anel, A., Calvo, M., Naval, J., Iturralde, M., Alava, M.A., Pineiro, A., 1989. Interaction of rat α-fetoprotein and albumin with polyunsaturated and other fatty acids: determination of apparent association constants. FEBS Lett. 250, 22–24. Anema, S.G., 2010. Instability of pressure-treated reconstituted skim milk to acidification. Food Biophys. 5, 321–329. Anema, S.G., 2012. Pressure-induced denaturation of β-lactoglobulin in skim milk: effect of milk concentration. J. Agric. Food Chem. 60, 6565–6570. Arakawa, T., Timasheff, S.N., 1982. Stabilization of protein structure by sugars. Biochemistry 21, 6536–6544. Arroyo-Maya, I.J., Campos-Tera´n, J., Herna´ndez-Arana, A., McClements, D.J., 2016. Characterization of flavonoidprotein interactions using fluorescence spectroscopy: binding of pelargonidin to dairy proteins. Food Chem. 213, 431–439. Aymard, P., Durand, D., Nicolai, T., 1996. The effect of temperature and ionic strength on the dimerization of β-lactoglobulin. Int. J. Biol. Macromol. 19, 213–221. Baier, S., McClements, D.J., 2001. Impact of preferential interactions on thermal stability and gelation of bovine serum albumin in aqueous sucrose solutions. J. Agric. Food Chem. 49, 2600–2608. Baier, S.K., McClements, D.J., 2003. Combined influence of NaCl and sucrose on heat-induced gelation of bovine serum albumin. J. Agric. Food Chem. 51, 8107–8112. Baier, S.K., Decker, E.A., McClements, D.J., 2004. Impact of glycerol on thermostability and heat-induced gelation of bovine serum albumin. Food Hydrocoll. 18, 91–100. Ball, G.F.M., 1988. Fat-Soluble Vitamin Assays in Food Analysis, A Comprehensive Review. Elsevier Applied Science, London, UK. Barbana, C., P
erez, M.D., Sa´nchez, L., Dalgalarrondo, M., Chobert, J.M., Haertl
e, T., Calvo, M., 2006. Interaction of bovine α-lactalbumin with fatty acids as determined by partition equilibrium and fluorescence spectroscopy. Int. Dairy J. 16, 18–25.

562

14. Interaction between milk proteins and micronutrients

Barbiroli, A., Bonomi, F., Ferranti, P., Fessas, D., Nasi, A., Rasmussen, P., Iametti, S., 2011. Bound fatty acids modulate the sensitivity of bovine β-lactoglobulin to chemical and physical denaturation. J. Agric. Food Chem. 59, 5729–5737. Barbiroli, A., Marengo, M., Fessas, D., Ragg, E., Renzetti, S., Bonomi, F., Iametti, S., 2017. Stabilization of betalactoglobulin by polyols and sugars against temperature-induced denaturation involves diverse and specific structural regions of the protein. Food Chem. 234, 155–162. Bello, M., Portillo-T
ellez, M.D.C., Garcı´a-Herna´ndez, E., 2011. Energetics of ligand recognition and self-association of bovine β-lactoglobulin: differences between variants A and B. Biochemistry 50, 151–161. Benzaria, A., Maresca, M., Taieb, N., Dumay, E., 2013. Interaction of curcumin with phosphocasein micelles processed or not by dynamic high-pressure. Food Chem. 138, 2327–2337. Bezwoda, W.R., Mansoor, N., 1986. Isolation and characterization of lactoferrin separated from human whey by adsorption chromatography using Cibacron Blue F3G-A linked affinity adsorbent. Clin. Chim. Acta 157, 89–93. Bhattacharyya, J., Das, K.P., 2001. Interactions of chlorpromazine with milk proteins. Mol. Cell. Biochem. 221, 11–15. Bohin, M.C., Vincken, J.P., Van Der Hijden, H.T.W.M., Gruppen, H., 2012. Efficacy of food proteins as carriers for flavonoids. J. Agric. Food Chem. 60, 4136–4143. Bourassa, P., Tajmir-Riahi, H.A., 2012. Locating the binding sites of folic acid with milk α- and β-caseins. J. Phys. Chem. B 116, 513–519. Boye, J.I., Alli, I., 2000. Thermal denaturation of mixtures of α-lactalbumin and β-lactoglobulin: a differential scanning calorimetric study. Food Res. Int. 33, 673–682. Boye, J.I., Alli, I., Ismail, A.A., 1996a. Interactions involved in the gelation of bovine serum albumin. J. Agric. Food Chem. 44, 996–1004. Boye, J.I., Ismail, A.A., Alli, I., 1996b. Effects of physicochemical factors on the secondary structure of β-lactoglobulin. J. Dairy Res. 63, 97–109. Boye, J.I., Alli, I., Ismail, A.A., 1997. Use of differential scanning calorimetry and infrared spectroscopy in the study of thermal and structural stability of α-lactalbumin. J. Agric. Food Chem. 45, 1116–1125. Brinkmann, C.R., Thiel, S., Larsen, M.K., Petersen, T.E., Jensenius, J.C., Heegaard, C.W., 2011. Preparation and comparison of cytotoxic complexes formed between oleic acid and either bovine or human α-lactalbumin. J. Dairy Sci. 94, 2159–2170. Busti, P., Scarpeci, S., Gatti, C.A., Delorenzi, N.J., 2005. Binding of alkylsulfonate ligands to bovine β-lactoglobulin: effects on protein denaturation by urea. Food Hydrocoll. 19, 249–255. Busti, P., Gatti, C.A., Delorenzi, N.J., 2006. Binding of alkylsulfonate ligands to bovine β-lactoglobulin: effects on protein thermal unfolding. Food Res. Int. 39, 503–509. Caetano-Silva, M.E., Bertoldo-Pacheco, M.T., Paes-Leme, A.F., Netto, F.M., 2015. Iron-binding peptides from whey protein hydrolysates: evaluation, isolation and sequencing by LC–MS/MS. Food Res. Int. 71, 132–139. Cao, X., He, Y., Kong, Y., Mei, X., Huo, Y., He, Y., Liu, J., 2019. Elucidating the interaction mechanism of eriocitrin with β-casein by multi-spectroscopic and molecular simulation methods. Food Hydrocoll. 94, 63–70. Cawthern, K.M., Narayan, M., Chaudhuri, D., Permyakov, E.A., Berliner, L.J., 1997. Interactions of α-lactalbumin with fatty acids and spin label analogs. J. Biol. Chem. 272, 30812–30816. Celej, M.S., Dassie, S.A., Freire, E., Bianconi, M.L., Fidelio, G.D., 2005. Ligand-induced thermostability in proteins: thermodynamic analysis of ANS-albumin interaction. Biochim. Biophys. Acta 1750, 122–133. Chanasattru, W., Decker, E.A., McClements, D.J., 2007a. Inhibition of droplet flocculation in globular-protein stabilized oil-in-water emulsions by polyols. Food Res. Int. 40, 1161–1169. Chanasattru, W., Decker, E.A., McClements, D.J., 2007b. Modulation of thermal stability and heat-induced gelation of beta-lactoglobulin by high glycerol and sorbitol levels. Food Chem. 103, 512–520. Chanasattru, W., Decker, E.A., McClements, D.J., 2007c. Physicochemical basis for cosolvent modulation of betalactoglobulin functionality: interfacial tension study. Food Res. Int. 40, 1098–1105. Chanasattru, W., Decker, E.A., McClements, D.J., 2008. Impact of cosolvents (polyols) on globular protein functionality: ultrasonic velocity, density, surface tension and solubility study. Food Hydrocoll. 22, 1475–1484. Chanphai, P., Bourassa, P., Kanakis, C.D., Tarantilis, P.A., Polissiou, M.G., Tajmir-Riahi, H.A., 2018. Review on the loading efficacy of dietary tea polyphenols with milk proteins. Food Hydrocoll. 77, 322–328. Charles, M., Bernal, B., Guichard, E., 1996. Interaction of β-lactoglobulin with flavour compounds. In: Taylor, A.J., Mottram, D.S. (Eds.), Flavour Science: Recent Developments. Royal Society of Chemistry, Cambridge, UK, pp. 433–436.

References

563

Chavoshpour-Natanzi, Z., Sahihi, M., Gharaghani, S., 2018. Structural stability of β-lactoglobulin in the presence of cetylpyridinum bromide: spectroscopic and molecular docking studies. J. Biomol. Struct. Dyn. 36, 753–760. Cheng, J., Liu, J.-H., Prasanna, G., Jing, P., 2017. Spectrofluorimetric and molecular docking studies on the interaction of cyanidin-3-O-glucoside with whey protein, β-lactoglobulin. Int. J. Biol. Macromol. 105, 965–972. Cheng, H., Fang, Z., Wusigale, Bakry, A.M., Chen, Y., Liang, L., 2018. Complexation of trans- and cis-resveratrol with bovine serum albumin, β-lactoglobulin or α-lactalbumin. Food Hydrocoll. 81, 242–252. Cho, Y.J., Batt, C.A., Sawyer, L., 1994. Probing the retinol-binding site of bovine β-lactoglobulin. J. Biol. Chem. 269, 11102–11107. Clark, D.C., Wilde, P.J., Wilson, D.R., Wustneck, R., 1992. The interaction of sucrose esters with β-lactoglobulin and β-casein from bovine-milk. Food Hydrocoll. 6, 173–186. Considine, T., Patel, H.A., Singh, H., Creamer, L.K., 2005a. Influence of binding of sodium dodecyl sulfate, all-transretinol, palmitate, and 8-anilino-1-naphthalenesulfonate on the heat-induced unfolding and aggregation of β-lactoglobulin B. J. Agric. Food Chem. 53, 3197–3205. Considine, T., Singh, H., Patel, H.A., Creamer, L.K., 2005b. Influence of binding of sodium dodecyl sulfate, all-transretinol, and 8-anilino-1-naphthalenesulfonate on the high-pressure-induced unfolding and aggregation of β-lactoglobulin B. J. Agric. Food Chem. 53, 8010–8018. Considine, T., Patel, H.A., Anema, S.G., Singh, H., Creamer, L.K., 2007a. Interactions of milk proteins during heat and high hydrostatic pressure treatments – a review. Innov. Food Sci. Emerg. Technol. 8, 1–23. Considine, T., Patel, H.A., Singh, H., Creamer, L.K., 2007b. Influence of binding conjugated linoleic acid and myristic acid on the heat- and high-pressure-induced unfolding and aggregation of β-lactoglobulin B. Food Chem. 102, 1270–1280. Creamer, L., 1980. A study of the effect of sodium dodecyl sulfate on bovine β-casein self-association. Arch. Biochem. Biophys. 199, 172–178. Creamer, L.K., 1995. Effect of sodium dodecyl sulfate and palmitic acid on the equilibrium unfolding of bovine β-lactoglobulin. Biochemistry 34, 7170–7176. Creamer, L.K., Loveday, S.M., Sawyer, L., 2011. Milk proteins j β-lactoglobulin. In: Encyclopedia of Dairy Sciences, second ed. Elsevier, Oxford, UK. Czubinski, J., Dwiecki, K., 2017. A review of methods used for investigation of protein–phenolic compound interactions. Int. J. Food Sci. Technol. 52, 573–585. Dai-Dong, J.X., Novak, G., Hardy, J., 1990. Stabilization of vitamin C by β-lactoglobulin during heat treatment. Sci. Aliment. 10, 393–401. Damodaran, S., Kinsella, J.E., 1980. Flavor protein interactions. Binding of carbonyls to bovine serum albumin: thermodynamic and conformational effects. J. Agric. Food Chem. 28, 567–571. Damodaran, S., Kinsella, J.E., 1981. The effect of neutral salts on the stability of macromolecules—a new approach using a protein-ligand binding system. J. Biol. Chem. 256, 3394–3398. Das, S., Ellis, A., Mittal, V.A., Singh, H., Ye, A., 2013. Micronutrient fortification process and its uses. PCT Application No. PCT/NZ2013/000109. Delavari, B., Saboury, A.A., Atri, M.S., Ghasemi, A., Bigdeli, B., Khammari, A., Maghami, P., MoosaviMovahedi, A.A., Haertl
e, T., Goliaei, B., 2015. Alpha-lactalbumin: a new carrier for vitamin D3 food enrichment. Food Hydrocoll. 45, 124–131. Delshadian, Z., Mortazavian, A.M., Tabarzad, M., Hosseini, S.M., Mohammadi, R., Rouhi, M., Salami, M., KhosraviDarani, K., 2018. Optimisation of experimental conditions for binding of divalent iron to bioactive casein phosphopeptides. Int. J. Food Sci. Technol. 53, 784–793. Devi, A.F., Buckow, R., Hemar, Y., Kasapis, S., 2013. Structuring dairy systems through high pressure processing. J. Food Eng. 114, 106–122. Diarrassouba, F., Remondetto, G., Liang, L., Garrait, G., Beyssac, E., Subirade, M., 2013. Effects of gastrointestinal pH conditions on the stability of the β-lactoglobulin/vitamin D3 complex and on the solubility of vitamin D3. Food Res. Int. 52, 515–521. Dickinson, E., 2010. Flocculation of protein-stabilized oil-in-water emulsions. Colloids Surf. B 81, 130–140. Dierckx, S., Huyghebaert, A., 2002. Effects of sucrose and sorbitol on the gel formation of a whey protein isolate. Food Hydrocoll. 16, 489–497. Divsalar, A., Saboury, A.A., 2005. Comparative structural and conformational studies on two forms of β-lactoglobulin (A and B) upon interaction with lead ion. FEBS J. 272, 340.

564

14. Interaction between milk proteins and micronutrients

Divsalar, A., Saboury, A.A., Mansoori-Torshizi, H., Hemmatinejad, B., 2006a. Comparative and structural analysis of the interaction between β-lactoglobulin type A and B with a new anticancer component (2,20 -bipyridin n-hexyl dithiocarbamato Pd(II) nitrate). Bull. Korean Chem. Soc. 27, 1801–1808. Divsalar, A., Saboury, A.A., Moosavi-Movahedi, A.A., 2006b. A study on the interaction between β-lactoglobulin a and a new antitumor reagent (2,2-bipyridinglycinato Pd(II) chloride). FEBS J. 273, 333. Domı´nguez-Ramı´rez, L., Del Moral-Ramirez, E., Cortes-Hernandez, P., Garcia-Garibay, M., Jimenez-Guzman, J., 2013. Beta-lactoglobulin’s conformational requirements for ligand binding at the calyx and the dimer interphase: a flexible docking study. PLoS One 8. e79530. Dufour, E., Haertle, T., 1990. Alcohol induced changes of β-lactoglobulin-retinol-binding stoichiometry. Protein Eng. 4, 185–190. Dufour, E., Marden, M.C., Haertle, T., 1991. Binding of retinoids and β-carotene to β-lactoglobulin. Influence of protein modifications. Biochim. Biophys. Acta 1079, 316–320. Dufour, E., Hui Bon Hoa, G., Haertle, T., 1994. High-pressure effects on β-lactoglobulin interactions with ligands studied by fluorescence. Biochim. Biophys. Acta 1206, 166–172. Dumay, E.M., Kalichevsky, M.T., Cheftel, J.C., 1994. High-pressure unfolding and aggregation of β-lactoglobulin and the baroprotective effects of sucrose. J. Agric. Food Chem. 42, 1861–1868. Dumay, E.M., Kalichevsky, M.T., Cheftel, J.C., 1998. Characteristics of pressure-induced gels of β-lactoglobulin at various times after pressure release. LWT—Food Sci. Technol. 31, 10–19. Ebel, C., Eisenberg, H., Ghirlando, R., 2000. Probing protein-sugar interactions. Biophys. J. 78, 385–393. Elzoghby, A.O., Abo El-Fotoh, W.S., Elgindy, N.A., 2011. Casein-based formulations as promising controlled release drug delivery systems. J. Control. Release 153, 206–216. Elzoghby, A.O., Samy, W.M., Elgindy, N.A., 2013. Novel spray-dried genipin-crosslinked casein nanoparticles for prolonged release of alfuzosin hydrochloride. Pharm. Res. 30, 512–522. FitzGerald, R.J., 1998. Potential uses of caseinophosphopeptides. Int. Dairy J. 8, 451–457. Fogolari, F., Ragona, L., Licciardi, S., Romagnoli, S., Michelutti, R., Ugolini, R., Molinari, H., 2000. Electrostatic properties of bovine β-lactoglobulin. Proteins Struct. Funct. Genet. 39, 317–330. Fontecha, J., Swaisgood, H.E., 1995. Interaction of sucrose ester with casein micelles as characterized by size-exclusion chromatography. J. Dairy Sci. 78, 2660–2665. Ford, J.E., 1974. Some observations on possible nutritional significance of vitamin-B12-binding and folate-binding proteins in milk. Br. J. Nutr. 31, 243–257. Forrest, S.A., Yada, R.Y., Rousseau, D., 2005. Interactions of vitamin D3 with bovine β-lactoglobulin A and β-casein. J. Agric. Food Chem. 53, 8003–8009. Franzen, K.L., Kinsella, J.E., 1974. Parameters affecting binding of volatile flavor compounds in model food systems 1. Proteins. J. Agric. Food Chem. 22, 675–678. Frapin, D., Dufour, E., Haertle, T., 1993. Probing the fatty acid binding site of β-lactoglobulins. J. Protein Chem. 12, 443–449. Futterman, S., Heller, J., 1972. The enhancement of fluorescence and the decreased susceptibility to enzymatic oxidation of retinol complexed with bovine serum albumin, β-lactoglobulin, and the retinol-binding protein of human plasma. J. Biol. Chem. 247, 5168–5172. Ghalandari, B., Divsalar, A., Eslami-Moghadam, M., Saboury, A.A., Haertl
e, T., Amanlou, M., Parivar, K., 2015. Probing of the interaction between β-lactoglobulin and the anticancer drug oxaliplatin. Appl. Biochem. Biotechnol. 175, 974–987. Guichard, E., Langourieux, S., 2000. Interactions between b-lactoglobulin and flavour compounds. Food Chem. 71, 301–308. Gupta, C., Arora, S., Syama, M.A., Sharma, A., 2017. Preparation of milk protein-vitamin A complexes and their evaluation for vitamin A binding ability. Food Chem. 237, 141–149. Guth, H., Fritzler, R., 2004. Binding studies and computer-aided modelling of macromolecule/odorant interactions. Chem. Biodivers. 1, 2001–2023. Ha, H.K., Kim, J.W., Lee, M.R., Lee, W.J., 2013. Formation and characterization of quercetin-loaded chitosan oligosaccharide/β-lactoglobulin nanoparticle. Food Res. Int. 52, 82–90. Haham, M., Ish-Shalom, S., Nodelman, M., Duek, I., Segal, E., Kustanovich, M., Livney, Y.D., 2012. Stability and bioavailability of vitamin D nanoencapsulated in casein micelles. Food Funct. 3, 737–744.

References

565

Hammou, H.O., Plaza del Pino, I.M., Sanchez-Ruiz, J.M., 1998. Hydration changes upon protein unfolding: cosolvent effect analysis. New J. Chem. 22, 1453–1461. Hansted, J.G., Wejse, P.L., Bertelsen, H., Otzen, D.E., 2011. Effect of protein-surfactant interactions on aggregation of β-lactoglobulin. Biochim. Biophys. Acta Proteins Proteomics 1814, 713–723. Havea, P., Singh, H., Creamer, L.K., 2001. Characterization of heat-induced aggregates of β-lactoglobulin, α-lactalbumin and bovine serum albumin in a whey protein concentrate environment. J. Dairy Res. 68, 483–497. He, L., Zheng, J., Labuza, T.P., Xiao, H., 2013. A surface enhanced Raman spectroscopic study of interactions between casein and polymethoxyflavones. J. Raman Spectrosc. 44, 531–535. Housaindokht, M.R., Chamani, J., Saboury, A.A., Moosavi-Movahedi, A.A., Bahrololoom, M., 2001. Three binding sets analysis of α-lactalbumin by interaction of tetradecyl trimethyl ammonium bromide. Bull. Korean Chem. Soc. 22, 145–148. Huppertz, T., Fox, P.F., de Kruif, K.G., Kelly, A.L., 2006. High pressure-induced changes in bovine milk proteins: a review. Biochim. Biophys. Acta Proteins Proteomics 1764, 593–598. Jasinski, E., Kilara, A., 1985. Flavor binding by whey proteins. Milchwissenschaft 40, 596–599. Jia, J., Gao, X., Hao, M., Tang, L., 2017. Comparison of binding interaction between β-lactoglobulin and three common polyphenols using multi-spectroscopy and modeling methods. Food Chem. 228, 143–151. Jones, M.N., Wilkinson, A., 1976. The interaction between β-lactoglobulin and sodium n-dodecylsulphate. Biochem. J. 153, 713–718. Jou, K.D., Harper, W.J., 1996. Effect of di-saccharides on the thermal properties of whey proteins determined by differential scanning calorimetry (DSC). Milchwissenschaft 51, 509–511. Jouenne, E., Crouzet, J., 2000. Determination of apparent binding constants for aroma compounds with β-lactoglobulin by dynamic coupled column liquid chromatography. J. Agric. Food Chem. 48, 5396–5400. Jung, D.M., Ebeler, S.E., 2003. Investigation of binding behaviour of α- and β-ionones to β-lactoglobulin at different pH values using a diffusion-based NOE pumping technique. J. Agric. Food Chem. 51, 1988–1993. Jung, D.M., de Ropp, J.S., Ebeler, S.E., 2002. Application of pulse field gradient NMR techniques for investigating binding of flavor compounds to macromolecules. J. Agric. Food Chem. 50, 4262–4269. Kamiyama, T., Sadahide, Y., Nogusa, Y., Gekko, K., 1999. Polyol-induced molten globule of cytochrome c: an evidence for stabilization by hydrophobic interaction. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1434, 44–57. Kanakis, C.D., Hasni, I., Bourassa, P., Tarantilis, P.A., Polissiou, M.G., Tajmir-Riahi, H.A., 2011. Milk β-lactoglobulin complexes with tea polyphenols. Food Chem. 127, 1046–1055. Kanno, C., Kanehara, N., Shirafuji, K., Tanji, R., Imai, T., 1991. Binding form of vitamin-B2 in bovine-milk—its concentration, distribution and binding linkage. J. Nutr. Sci. Vitaminol. 37, 15–27. Kaushik, J.K., Bhat, R., 1998. Thermal stability of proteins in aqueous polyol solutions: role of the surface tension of water in the stabilizing effect of polyols. J. Phys. Chem. B 102, 7058–7066. Keenan, R.D., Young, D.J., Tier, C.M., Jones, A.D., Underdown, J., 2001. Mechanism of pressure-induced gelation of milk. J. Agric. Food Chem. 49, 3394–3402. Kehoe, J.J., Foegeding, E.A., 2011. Interaction between beta-casein and whey proteins as a function of pH and salt concentration. J. Agric. Food Chem. 59, 349–355. Kibangou, I., Bouhallab, S., Bureau, F., Allouche, S., Thouvenin, G., Bougl
e, D., 2008. Caseinophosphopeptide-bound iron: protective effect against gut peroxidation. Ann. Nutr. Metab. 52, 177–180. Klaassen, I., Brakenhoff, R.H., Smeets, S.J., Snow, G.B., Braakhuis, B.J.M., 1999. Considerations for in vitro retinoid experiments: importance of protein interaction. Biochim. Biophys. Acta, Gen. Subj. 1427, 265–275. Kontopidis, G., Holt, C., Sawyer, L., 2004. Invited review: β-lactoglobulin: binding properties, structure, and function. J. Dairy Sci. 87, 785–796. Konuma, T., Sakurai, M., Goto, Y., 2007. Promiscuous binding of ligands by β-lactoglobulin involves hydrophobic interactions and plasticity. J. Mol. Biol. 368, 209–218. K€ uhn, J., Considine, T., Singh, H., 2006. Interactions of milk proteins and volatile compounds: implications in the development of protein foods. J. Food Sci. 71, R72–R82. Kulmyrzaev, A., Bryant, C., McClements, D.J., 2000. Influence of sucrose on the thermal denaturation, gelation, and emulsion stabilization of whey proteins. J. Agric. Food Chem. 48, 1593–1597. Lamiot, E., Dufour, E., Haertle, T., 1994. Insect sex pheromone binding by bovine β-lactoglobulin. J. Agric. Food Chem. 42, 695–699.

566

14. Interaction between milk proteins and micronutrients

Lee, J.C., Timasheff, S.N., 1981. The stabilization of proteins by sucrose. J. Biol. Chem. 256, 7193–7201. Lerbret, E., Bordat, P., Affouard, F., Deschamps, M., Migliardo, F., 2005. How homogeneous are the trehalose, maltose and sucrose water solutions? An insight from molecular dynamics simulations. J. Phys. Chem. B 109, 11046–11057. Li, T., Hu, P., Dai, T., Li, P., Ye, X., Chen, J., Liu, C., 2018. Comparing the binding interaction between β-lactoglobulin and flavonoids with different structure by multi-spectroscopy analysis and molecular docking. Spectrochim. Acta A: Mol. Biomol. Spectrosc. 201, 197–206. Liang, L., Subirade, M., 2012. Study of the acid and thermal stability of β-lactoglobulin-ligand complexes using fluorescence quenching. Food Chem. 132 (4), 2023–2029. Lietaer, E., Poiffait, A., Adrian, J., 1991. Studies on the interaction between casein and vitamin A. LWT—Food Sci. Technol. 24, 39–45. Lin, T.Y., Timasheff, S.N., 1996. On the role of surface tension in the stabilization of globular proteins. Protein Sci. 5, 372–381. Lindner, R.A., Kapur, A., Carver, J.A., 1997. The interaction of the molecular chaperone, alpha-crystallin, with molten globule states of bovine-lactalbumin. J. Biol. Chem. 272, 27722–27729. Lisˇkova´, K., Auty, M.A.E., Chaurin, V., Min, S., Mok, K.H., O’Brien, N., Kelly, A.L., Brodkorb, A., 2011. Cytotoxic complexes of sodium oleate with β-lactoglobulin. Eur. J. Lipid Sci. Technol. 113, 1207–1218. Liu, X., Powers, J.R., Swanson, B.G., Hill, H.H., Clark, S., 2005. Modification of whey protein concentrate hydrophobicity by high hydrostatic pressure. Innov. Food Sci. Emerg. Technol. 6, 310–317. Livney, Y.D., 2010. Milk proteins as vehicles for bioactives. Curr. Opin. Colloid Interface Sci. 15, 73–83. Livney, Y.D., Dalgleish, D.G., 2007. Casein micelles for nanoencapsulation of hydrophobic compounds. Patent Application WO2007122613 A1. Loch, J.I., Polit, A., Bonarek, P., Olszewska, D., Kurpiewska, K., Dziedzicka-Wasylewska, M., Lewi
nski, K., 2012. Structural and thermodynamic studies of binding saturated fatty acids to bovine β-lactoglobulin. Int. J. Biol. Macromol. 50, 1095–1102. Loewen, A., Chan, B., Li-Chan, E.C.Y., 2018. Optimization of vitamins A and D3 loading in re-assembled casein micelles and effect of loading on stability of vitamin D3 during storage. Food Chem. 240, 472–481. Loveday, S.M., Singh, H., 2008. Recent advances in technologies for vitamin A protection in foods. Trends Food Sci. Technol. 19, 657–668. Loveday, S.M., Rao, M.A., Singh, H., 2012. Food protein nanoparticles: formation, properties and applications. In: Bhandari, B., Roos, Y.H. (Eds.), Food Materials Science and Engineering. Blackwell Publishing, Oxford, UK, pp. 263–294. Loveday, S.M., Sarkar, A., Singh, H., 2013. Innovative yoghurts: novel processing technologies for improving acid milk gel texture. Trends Food Sci. Technol. 33, 5–20. Lu, R.-C., Cao, A.-N., Lai, L.-H., Xiao, J.-X., 2006. Effect of cationic surfactants on the interaction between sodium perfluorooctanoate and β-lactoglobulin. J. Colloid Interface Sci. 293, 61–68. Mata, L., Sanchez, L., Calvo, M., 1997. Interaction of mercury with human and bovine milk proteins. Biosci. Biotechnol. Biochem. 61, 1641–1645. Maulik, S., Moulik, S.P., Chattoraj, D.K., 1996. Biopolymer–surfactant interaction: 4. Kinetics of binding of cetyltrimethyl ammonium bromide with gelatin, hemoglobin, β-lactoglobulin and lysozyme. J. Biomol. Struct. Dyn. 13, 771–780. McClements, D.J., 2002. Modulation of globular protein functionality by weakly interacting cosolvents. Crit. Rev. Food Sci. Nutr. 42, 417–471. McNeill, V.L., Schmidt, K.A., 1993. Vanillin interaction with milk protein isolates in sweetened drinks. J. Food Sci. 58 (1142–1144), 1147. Mensi, A., Choiset, Y., Rabesona, H., Haertl
e, T., Borel, P., Chobert, J.-M., 2013. Interactions of β-lactoglobulin variants A and B with vitamin A. Competitive binding of retinoids and carotenoids. J. Agric. Food Chem. 61, 4114–4119. Mercadante, D., Melton, L.D., Norris, G.E., Loo, T.S., Williams, M.A.K., Dobson, R.C.J., Jameson, G.B., 2012. Bovine β-lactoglobulin is dimeric under imitative physiological conditions: dissociation equilibrium and rate constants over the pH range of 2.5-7.5. Biophys. J. 103, 303–312. Mittal, V.A., Ellis, A., Ye, A., Das, S., Singh, H., 2015. Influence of calcium depletion on iron-binding properties of milk. J. Dairy Sci. 98, 2103–2113. Mittal, V.A., Ellis, A., Ye, A., Edwards, P.J.B., Das, S., Singh, H., 2016. Iron binding to caseins in the presence of orthophosphate. Food Chem. 190, 128–134.

References

567

Mittal, V.A., Ellis, A., Ye, A., Edwards, P.J.B., Singh, H., 2018. The adsorption of orthophosphate onto casein-iron precipitates. Food Chem. 239, 17–22. Moeiniafshari, A.-A., Zarrabi, A., Bordbar, A.-K., 2015. Exploring the interaction of naringenin with bovine betacasein nanoparticles using spectroscopy. Food Hydrocoll. 51, 1–6. Mohammadi, F., Moeeni, M., 2015. Study on the interactions of trans-resveratrol and curcumin with bovine α-lactalbumin by spectroscopic analysis and molecular docking. Mater. Sci. Eng. C 50, 358–366. Mohammadzadeh, K.A., Smith, G.M., Feeney, R.E., 1969. Hydrophobic binding of hydrocarbons by proteins II. Relationship of protein structure. Biochim. Biophys. Acta 194, 256–264. Mohan, M.S., Jurat-Fuentes, J.L., Harte, F., 2013. Binding of vitamin A by casein micelles in commercial skim milk. J. Dairy Sci. 96, 790–798. Monaco, H.L., Zanotti, G., Spadon, P., Bolognesi, M., Sawyer, L., Eliopoulos, E.E., 1987. Crystal structure of the ˚ resolution. J. Mol. Biol. trigonal form of bovine β-lactoglobulin and of its complex with retinol at 2.5 A 197, 695–706. Mora-Gutierrez, A., Farrell, H.M., 2000. Sugar–casein interaction in deuterated solutions of bovine and caprine casein as determined by oxygen-17 and carbon-13 nuclear magnetic resonance: a case of preferential interactions. J. Agric. Food Chem. 48, 3245–3255. Muresan, S., van der Bent, A., de Wolf, F.A., 2001. Interaction of β-lactoglobulin with small hydrophobic ligands as monitored by fluorometry and equilibrium dialysis: nonlinear quenching effects related to protein–protein association. J. Agric. Food Chem. 49, 2609–2618. Murray, B.S., Liang, H.J., 1999. Enhancement of the foaming properties of protein dried in the presence of trehalose. J. Agric. Food Chem. 47, 4984–4991. Nagasako, Y., Saito, H., Tamura, Y., Shimamura, S., Tomita, M., 1993. Iron-binding properties of bovine lactoferrin in iron-rich solution. J. Dairy Sci. 76, 1876–1881. Nagy, K., Courtet-Compondu, M.C., Williamson, G., Rezzi, S., Kussmann, M., Rytz, A., 2012. Non-covalent binding of proteins to polyphenols correlates with their amino acid sequence. Food Chem. 132, 1333–1339. Narayan, M., Berliner, L.J., 1997. Fatty acids and retinoids bind independently and simultaneously to β-lactoglobulin. Biochemistry 36, 1906–1911. Narayan, M., Berliner, L.J., 1998. Mapping fatty acid binding to β-lactoglobulin: ligand binding is restricted by modification of Cys 121. Protein Sci. 7, 150–157. Nixon, P.F., Jones, M., Winzor, D.J., 2004. Quantitative description of the interaction between folate and the folatebinding protein from cow’s milk. Biochem. J. 382, 215–221. O’Connell, J.E., Fox, P.F., 2003. Heat-induced coagulation of milk. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry—Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, pp. 879–945. Oelrichs, B.A., Kratzing, C.C., Kelly, J.D., Winzor, D.J., 1984. The binding of ascorbate to bovine serum-albumin. Int. J. Vitam. Nutr. Res. 54, 61–64. O’Loughlin, I.B., Kelly, P.M., Murray, B.A., Fitzgerald, R.J., Brodkorb, A., 2015. Molecular characterization of whey protein hydrolysate fractions with ferrous chelating and enhanced iron solubility capabilities. J. Agric. Food Chem. 63 (10), 270–2714. O’Neill, T.E., Kinsella, J.E., 1987. Binding of alkanone flavors to β-lactoglobulin—effects of conformational and chemical modification. J. Agric. Food Chem. 35, 770–774. Papiz, M.Z., Sawyer, L., Eliopoulos, E.E., North, A.C.T., Findlay, J.B.C., Sivaprasadarao, R., Jones, T.A., Newcomer, M.E., Kraulis, P.J., 1986. The structure of β-lactoglobulin and its similarity to plasma retinol-binding protein. Nature 324, 383–385. Park, Y.J., Kim, K.H., Lim, D.W., Lee, E.K., 2015. Effects of pH and protein conformation on in-solution complexation between bovine α-lactalbumin and oleic acid: binding trend analysis by using SPR and ITC. Process Biochem. 50, 1379–1387. Peres, J.M., Bouhallab, S., Petit, C., Bureau, F., Maubois, J.L., Arhan, P., Bougle, D., 1998. Improvement of zinc intestinal absorption and reduction of zinc/iron interaction using metal bound to the caseinophosphopeptide 1–25 of β-casein. Reprod. Nutr. Dev. 38, 465–472. P
erez, M.D., Calvo, M., 1995. Interaction of β-lactoglobulin with retinol and fatty acids and its role as a possible biological function for this protein: a review. J. Dairy Sci. 78, 978–988. P
erez, M.D., Diaz de Villegas, C., Sanchez, L., Aranda, P., Ena, J.M., Calvo, M., 1989. Interaction of fatty acids with β-lactoglobulin and albumin from ruminant milk. J. Biochem. 106, 1094–1097.

568

14. Interaction between milk proteins and micronutrients

P
erez, M.D., Sanchez, L., Aranda, P., Ena, J.M., Calvo, M., 1990. Synthesis and evolution of concentration of β-lactoglobulin and α-lactalbumin from cow and sheep colostrums and milk throughout early lactation. Cell. Mol. Biol. 36, 205–212. P
erez, M.D., Sanchez, L., Aranda, P., Ena, J.M., Oria, R., Calvo, M., 1992. Effect of β-lactoglobulin on the activity of pregastric lipase. A possible role for this protein in ruminant milk. Biochim. Biophys. Acta 1123, 151–155. P
erez, M.D., Puyol, P., Ena, J.M., Calvo, M., 1993. Comparison of the ability to bind lipids of β-lactoglobulin and serum-albumin of milk from ruminant and non-ruminant species. J. Dairy Res. 60, 55–63. Permyakov, E.A., Morozova, L.A., Kahnichenko, L.P., Derezhkov, V.Y., 1988. Interaction of α-lactalbumin with Cu2+. Biophys. Chem. 32, 37–42. Poiffait, A., Adrian, J., 1991. Interaction between casein and vitamin A during food processing. Adv. Exp. Med. Biol. 289, 61–73. Politi, R., Harries, D., 2010. Enthalpically driven peptide stabilization by protective osmolytes. Chem. Commun. 46, 6449–6451. Puyol, P., Perez, M.D., Peiro, J.M., Calvo, M., 1991. Interaction of bovine β-lactoglobulin and other bovine and human whey proteins with retinol and fatty acids. Agric. Biol. Chem. 55, 2515–2520. Puyol, P., Perez, M.D., Mata, L., Calvo, M., 1994. Study on interaction between β-lactoglobulin and other bovine whey proteins with ascorbic acid. Milchwissenschaft 49, 25–27. Qi, P.X., Xiao, Y., Wickham, E.D., 2017. Stabilization of whey protein isolate (WPI) through interactions with sugar beet pectin (SBP) induced by controlled dry-heating. Food Hydrocoll. 67, 1–13. Qin, B.Y., Creamer, L.K., Baker, E.N., Jameson, G.B., 1998. 12-Bromododecanoic acid binds inside the calyx of bovine β-lactoglobulin. FEBS Lett. 438, 272–278. Radibratovic, M., Al-Hanish, A., Minic, S., Radomirovic, M., Milcic, M., Stanic-Vucinic, D., Cirkovic Velickovic, T., 2019. Stabilization of apo α-lactalbumin by binding of epigallocatechin-3-gallate: experimental and molecular dynamics study. Food Chem. 278, 388–395. Ragona, L., Fogolari, F., Zetta, L., Perez, D.M., Puyol, P., De Kruif, K., Lohr, F., Ruterjans, H., Molinari, H., 2000. Bovine β-lactoglobulin: interaction studies with palmitic acid. Protein Sci. 9, 1347–1356. Rahimi Yazdi, S., Corredig, M., 2012. Heating of milk alters the binding of curcumin to casein micelles. A fluorescence spectroscopy study. Food Chem. 132, 1143–1149. Raica Jr., N., Vavich, M.G., Kemmerer, A.R., 1959. The effects of several milk components and similar compounds on the utilization of carotene by the rat. Arch. Biochem. Biophys. 83, 376–380. Reineccius, G.A., Coulter, S.T., 1969. Flavour retention during drying. J. Dairy Sci. 52, 1219–1223. Rodrı´guez Nin˜o, M.R., Rodrı´guez Patino, J.M., 2002. Effect of the aqueous phase composition on the adsorption of bovine serum albumin to the air–water interface. Ind. Eng. Chem. Res. 41, 1489–1495. Romero, C.M., Lozano, J.M., Sancho, J., Giraldo, G.I., 2007. Thermal stability of β-lactoglobulin in the presence of aqueous solution of alcohols and polyols. Int. J. Biol. Macromol. 40, 423–428. R€ osgen, J., 2007. Molecular basis of osmolyte effects on protein and metabolites. Methods Enzymol. 428, 459–486. Roufik, S., Gauthier, S.F., Leng, X., Turgeon, S.L., 2006. Thermodynamics of binding interactions between bovine β-lactoglobulin A and the antihypertensive peptide β-Lg f142-148. Biomacromolecules 7, 419–426. Sahihi, M., Ghayeb, Y., Bordbar, A.-K., 2012. Fluorescence spectroscopic study on interaction of retinol with β-lactoglobulin in the presence of cetylpyridinium chloride. Spectroscopy 27, 27–34. Sa´iz-Abajo, M.J., Gonza´lez-Ferrero, C., Moreno-Ruiz, A., Romo-Hualde, A., Gonza´lez-Navarro, C.J., 2013. Thermal protection of β-carotene in re-assembled casein micelles during different processing technologies applied in food industry. Food Chem. 138, 1581–1587. Salter, D.N., Mowlem, A., 1983. Neonatal role of milk folate-binding protein—studies on the course of digestion of goat’s milk folate binder in the 6-d-old kid. Br. J. Nutr. 50, 589–596. Schokker, E.P., Singh, H., Pinder, D.N., Norris, G.E., Creamer, L.K., 1999. Characterization of intermediates formed during heat-induced aggregation of β-lactoglobulin AB at neutral pH. Int. Dairy J. 9, 791–800. Semenova, M.G., Antipova, A.S., Belyakova, L.E., 2002. Food protein interactions in sugar solutions. Curr. Opin. Colloid Interface Sci. 7 (5), 438–444. Shahraki, S., Heydari, A., Saeidifar, M., Gomroki, M., 2018. Biophysical and computational comparison on the binding affinity of three important nutrients to beta-lactoglobulin: folic acid, ascorbic acid and vitamin K3. J. Biomol. Struct. Dyn. 36, 3651–3665.

References

569

Shapira, A., Assaraf, Y.G., Epstein, D., Livney, Y.D., 2010. β-Casein nanoparticles as an oral delivery system for chemotherapeutic drugs: impact of drug structure and properties on co-assembly. Pharm. Res. 27, 2175–2186. Shilpashree, B.G., Arora, S., Sharma, V., Singh, A.K., 2015. Preparation of succinylated sodium caseinate–iron complex by adopting ultrafiltration technology: a novel food fortificant. Innov. Food Sci. Emerg. Technol. 32, 165–171. Shilpashree, B.G., Arora, S., Sharma, V., Bajaj, R.K., Tomar, S.K., 2016. Preparation of iron bound succinylated milk protein concentrate and evaluation of its stability. Food Chem. 196, 800–807. Shimoyamada, M., Yoshimura, H., Tomida, K., Watanabe, K., 1996. Stabilities of bovine β-lactoglobulin/retinol or retinoic acid complexes against tryptic hydrolysis, heating and light-induced oxidation. Lebensm. Wiss. Technol. 29, 763–766. Singh, H., Flanagan, J., 2005. Milk proteins. In: Hui, Y.H. (Ed.), Handbook of Food Science, Technology and Engineering. CRC Press, New York, NY, pp. 1–23. Singh, H., Havea, P., 2003. Thermal denaturation, aggregation and gelation of whey proteins. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry—Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, pp. 1261–1287. Singh, L.R., Poddar, N.K., Dar, T.A., Kumar, R., Ahmad, F., 2011. Protein and DNA destabilization by osmolytes: the other side of the coin. Life Sci. 88, 117–125. Smialowska, A., Matia-Merino, L., Ingham, B., Carr, A.J., 2017. Effect of calcium on the aggregation behaviour of caseinates. Colloids Surf. A: Physicochem. Eng. Asp. 522, 113–123. Sneharani, A.H., Karakkat, J.V., Singh, S.A., Rao, A.G.A., 2010. Interaction of curcumin with β-lactoglobulin; stability, spectroscopic analysis, and molecular modeling of the complex. J. Agric. Food Chem. 58, 11130–11139. Sostmann, K., Guichard, E., 1998. Immobilized β-lactoglobulin on a HPLC-column: a rapid way to determine proteinflavour interactions. Food Chem. 62, 509–513. Spector, A.A., 1975. Fatty acid binding to plasma albumin. J. Lipid Res. 16, 165–179. Spector, A.A., John, K., Fletcher, J.E., 1969. Binding of long-chain fatty acids to bovine serum albumin. J. Lipid Res. 10, 56–67. Spiegel, T., 1999. Whey protein aggregation under shear conditions—effects of lactose and heating temperature on aggregate size and structure. Int. J. Food Sci. Technol. 34, 523–531. Sreelakshmi, Y., Sharma, K.K., 2001. Interaction of α-lactalbumin with mini-αA-crystallin. J. Protein Chem. 20, 123–130. Stojadinovic, M., Radosavljevic, J., Ognjenovic, J., Vesic, J., Prodic, I., Stanic-Vucinic, D., Cirkovic Velickovic, T., 2013. Binding affinity between dietary polyphenols and β-lactoglobulin negatively correlates with the protein susceptibility to digestion and total antioxidant activity of complexes formed. Food Chem. 136, 1263–1271. Sugiarto, M., Ye, A., Singh, H., 2009. Characterisation of binding of iron to sodium caseinate and whey protein isolate. Food Chem. 114, 1007–1013. Sugiarto, M., Ye, A., Taylor, M.W., Singh, H., 2010. Milk protein-iron complexes: inhibition of lipid oxidation in an emulsion. Dairy Sci. Technol. 90, 87–98. Swaisgood, H.E., 2001. Protein ingredient for carrying lipophilic nutrients. United States Patent 6,290,274. Swaisgood, H.E., 2003. Chemistry of the caseins. In: Fox, P.F., McSweeney, P.L.H. (Eds.), Advanced Dairy Chemistry—Proteins, third ed. Kluwer Academic/Plenum Publishers, New York, NY, pp. 139–201. Swaney, J.B., Klotz, I.M., 1970. Amino acid sequence adjoining the lone tryptophan of human serum albumin. A binding site of the protein. Biochem. J. 9, 2570–2574. Taheri-Kafrani, A., Bordbar, A.-K., Mousavi, S.H.-A., Haertle, T., 2008. β-Lactoglobulin structure and retinol binding changes in presence of anionic and neutral detergents. J. Agric. Food Chem. 56, 7528–7534. Tang, X., Pikal, M.J., 2005. Measurement of the kinetics of protein unfolding in viscous systems and implications for protein stability in freeze-drying. Pharm. Res. 22, 1176–1185. Timasheff, S.N., 1993. The control of protein stability and association by weak-interactions with water—how do solvents affect these processes. Annu. Rev. Biophys. Biomol. Struct. 22, 67–97. Timasheff, S.N., 1998. Control of protein stability and reactions by weakly interacting cosolvents: the simplicity of the complicated. Adv. Protein Chem. 51, 355–432. Timasheff, S.N., 2002. Protein hydration, thermodynamic binding, and preferential hydration. Biochemistry 41, 13473–13482.

570

14. Interaction between milk proteins and micronutrients

Tiwari, A., Bhat, R., 2006. Stabilization of yeast hexokinase A by polyol osmolytes: correlation with the physicochemical properties of aqueous solutions. Biophys. Chem. 124, 90–99. Toyosaki, T., Mineshita, T., 1988. Antioxidant effects of protein-bound riboflavin and free riboflavin. J. Food Sci. 53, 1851–1853. Tukamoto, T., Ozeki, S., Hattori, F., Ishida, T., 1974. Drug interactions. 1. Binding of ascorbic-acid and fatty-acid ascorbyl esters to bovine serum-albumin. Chem. Pharm. Bull. 22, 385–389. Ueno, H.M., Ueda, N., Morita, M., Kakehi, Y., Kobayashi, T., 2012. Thermal stability of the iron-lactoferrin complex in aqueous solution is improved by soluble soybean polysaccharide. Food Biophys. 7, 183–189. van Ruth, S.M., de Vries, G., Geary, M., Giannouli, P., 2002. Influence of composition and structure of oil-in-water emulsions on retention of aroma compounds. J. Sci. Food Agric. 82, 1028–1035. Vegarud, G.E., Langsrud, T., Svenning, C., 2000. Mineral-binding milk proteins and peptides; occurrence, biochemical and technological characteristics. Br. J. Nutr. 84 (Suppl. 1), S91–S98. Viseu, M.I., Melo, E.P., Carvalho, T.I., Correia, R.F., Costa, S.M.B., 2007. Unfolding kinetics of β-lactoglobulin induced by surfactant and denaturant: a stopped-flow/fluorescence study. Biophys. J. 93, 3601–3612. Voilley, A., Beghin, V., Charpentier, C., Peyron, D., 1991. Interaction between aroma substances and macromolecules in model wine. Lebensm. Wiss. Technol. 24, 469–472. Wang, Q.W., Allen, J.C., Swaisgood, H.E., 1997a. Binding of retinoids to β-lactoglobulin isolated by bioselective adsorption. J. Dairy Sci. 80, 1047–1053. Wang, Q.W., Allen, J.C., Swaisgood, H.E., 1997b. Binding of vitamin D and cholesterol to β-lactoglobulin. J. Dairy Sci. 82, 257–264. Wang, Q., Allen, J.C., Swaisgood, H.E., 1999. Binding of lipophilic nutrients to β-lactoglobulin prepared by bioselective adsorption. J. Dairy Sci. 82, 257–264. Wang, X., Li, M., Li, M., Mao, X., Zhou, J., Ren, F., 2011. Preparation and characteristics of yak casein hydrolysate-iron complex. Int. J. Food Sci. Technol. 46, 1705–1710. Waninge, R., Paulsson, M., Nylander, T., Ninham, B., Sellers, P., 1998. Binding of sodium dodecyl sulphate and dodecyl trimethyl ammonium chloride to β-lactoglobulin: a calorimetric study. Int. Dairy J. 8, 141–148. Weiser, H., Somorjai, G., 1992. Bioactivity of cis and dicis isomers of vitamin A esters. Int. J. Vitam. Nutr. Res. 62, 201–208. Wendorf, J.R., Radke, C.J., Blanch, H.W., 2004. Reduced protein adsorption at solid interfaces by sugar excipients. Biotechnol. Bioeng. 87, 565–573. Wu, S.Y., Perez, M.D., Puyol, P., Sawyer, L., 1999. β-Lactoglobulin binds palmitate within its central cavity. J. Biol. Chem. 274, 170–174. Wu, X., Wu, H., Liu, M., Liu, Z., Xu, H., Lai, F., 2011. Analysis of binding interaction between (
)-epigallocatechin (EGC) and β-lactoglobulin by multi-spectroscopic method. Spectrochim. Acta A 82, 164–168. Wu, X., Dey, R., Wu, H., Liu, Z., He, Q., Zeng, X., 2013. Studies on the interaction of -epigallocatechin-3-gallate from green tea with bovine β-lactoglobulin by spectroscopic methods and docking. Int. J. Dairy Technol. 66, 7–13. Xie, G., Timasheff, S.N., 1997a. The thermodynamic mechanism of protein stabilization by trehalose. Biophys. Chem. 64, 25–43. Xie, G.F., Timasheff, S.N., 1997b. Mechanism of the stabilization of ribonuclease A by sorbitol: preferential hydration is greater for the denatured than for the native protein. Protein Sci. 6, 211–221. Xie, Y., Kosi
nska, A., Xu, H., Andlauer, W., 2013. Milk enhances intestinal absorption of green tea catechins in in vitro digestion/Caco-2 cells model. Food Res. Int. 53, 793–800. Yang, M.C., Guan, H.H., Liu, M.Y., Lin, Y.H., Yang, J.M., Chen, W.L., Chen, C.J., Mao, S.J.T., 2008. Crystal structure of a secondary vitamin D3 binding site of milk β-lactoglobulin. Proteins: Struct., Funct., Genet. 71, 1197–1210. Yi, C., Wambo, T.O., 2015. Factors affecting the interactions between beta-lactoglobulin and fatty acids as revealed in molecular dynamics simulations. Phys. Chem. Chem. Phys. 17, 23074–23080. Yildirim-Elikoglu, S., Erdem, Y.K., 2018. Interactions between milk proteins and polyphenols: binding mechanisms, related changes, and the future trends in the dairy industry. Food Rev. Intl. 34, 665–697. Zhan, F., Yang, J., Li, J., Wang, Y., Li, B., 2018. Characteristics of the interaction mechanism between tannic acid and sodium caseinate using multispectroscopic and thermodynamics methods. Food Hydrocoll. 75, 81–87. Zhang, Y., Zhong, Q., 2013a. Encapsulation of bixin in sodium caseinate to deliver the colorant in transparent dispersions. Food Hydrocoll. 33, 1–9.

References

571

Zhang, Y., Zhong, Q., 2013b. Probing the binding between norbixin and dairy proteins by spectroscopy methods. Food Chem. 139, 611–616. Zhang, X., Fu, X., Zhang, H., Liu, C., Jiao, W., Chang, Z., 2005. Chaperone-like activity of β-casein. Int. J. Biochem. Cell Biol. 37, 1232–1240. Zhang, Z.H., Wang, X.P., Ayman, W.Y., Munyendo, W.L., Lv, H.X., Zhou, J.P., 2013. Studies on lactoferrin nanoparticles of gambogic acid for oral delivery. Drug Deliv. 20, 86–93. Zhou, J., Wang, X., Ai, T., Cheng, X., Guo, H.Y., Teng, G.X., Mao, X.Y., 2012. Preparation and characterization of β-lactoglobulin hydrolysate-iron complexes. J. Dairy Sci. 95, 4230–4236. Zhu, X.Q., 2003. Interactions Between Flavour Compounds and Milk Proteins. Institute of Food, Nutrition and Human Health, Massey University, Palmerston North, New Zealand. Zimet, P., Rosenberg, D., Livney, Y.D., 2011. Re-assembled casein micelles and casein nanoparticles as nano-vehicles for ω-3 polyunsaturated fatty acids. Food Hydrocoll. 25, 1270–1276. Zsila, F., Imre, T., Szabo´, P.T., Bika´di, Z., Simonyi, M., 2002. Induced chirality upon binding of cis-parinaric acid to bovine β-lactoglobulin: spectroscopic characterization of the complex. FEBS Lett. 520, 81–87. Zsila, F., Hazai, E., Sawyer, L., 2005. Binding of the pepper alkaloid piperine to bovine β-lactoglobulin: circular dichroism spectroscopy and molecular modeling study. J. Agric. Food Chem. 53, 10179–10185.

This page intentionally left blank

C H A P T E R

15 Model food systems and protein functionality W. James Harper†, Sheelagh A. Hewitta, Lee M. Huffmanb a

Fonterra Research and Development Centre, Fonterra Co-operative Group Ltd, Palmerston North, New Zealand bNew Zealand Institute for Plant & Food Research Ltd, Palmerston North, New Zealand

Introduction The utilization of proteins in food for nutritional and functional purposes goes back many centuries, but the relationship between structure and function has been given close attention only during the past 30–40 years (Owusu-Apenten, 2004; Foegeding, 2015). Numerous studies and many reviews have contributed to gaining an understanding of precisely how proteins act in a complex food system and how their structure and function are altered by the other ingredients in the food, its intrinsic properties, and its processing. These include Anfinsen (1972), Pour-El (1981), Kinsella (1982), Nakai (1983), Mulvihill and Fox (1987), Mangino et al. (1994), Zayas (1996), Li-Chan (2004), Luyten et al. (2004), Owusu-Apenten (2004), Turgeon et al. (2007), Ghoush et al. (2008), Foegeding and Davis (2011), Singh (2011), Dickinson (2012), and Foegeding (2015). There are two broad ways of gaining knowledge of the structure and function of protein systems: (1) study of pure proteins in simple systems and (2) study of commercial proteins in the food systems in which they are used. These methods are entirely different (Luyten et al., 2004; Owusu-Apenten, 2004; Foegeding, 2015) and provide quite different information. Functionality tests can be very useful in obtaining reproducible functional properties, even though such tests cannot be used to predict the final characteristics in a real food system (de Wit, 1984, 1989; Harper, 1984; Owusu-Apenten, 2004). Some differences include the following: • In pure structure/function studies, pure proteins are generally used and are used at concentrations much lower than those used in food systems (Owusu-Apenten, 2004). †

Deceased.

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00015-3

573

# 2020 Elsevier Inc. All rights reserved.

574

15. Model food systems and protein functionality

• In food systems, proteins are seldom pure and may actually involve complex mixtures of proteins from a given food source (such as milk proteins, egg proteins, and soy proteins) or proteins from multiple food sources (i.e., meat, soy, milk, and gluten) or proteins that have been selectively denatured to provide the desired functionality (Mangino et al., 1994; Wagoner et al., 2015). • In structure/function studies, care is taken to avoid interactions with other components and to avoid modifying the secondary and tertiary structure during the experiments. The proteins are fully hydrated (Kinsella, 1982; Owusu-Apenten, 2004). • In food systems, the proteins are constantly exposed to other ingredients, which can modify the structure and hence function and can be modified by processes that often include pH, heat, and shear (Lee et al., 1992; Kilara, 1994; Dickinson, 2012). Competition for water can also modify functionality (Hogan et al., 2012) as can changes in intrinsic properties (Zayas, 1996). • In structure/function studies, outcome is generally measured for a specific and single response (Owusu-Apenten, 2004; Ako et al., 2009; Nicolai and Durand, 2013). • In food systems, the ingredients can influence product functionality at different points in the process, or functionality can be expressed in more than one outcome with respect to the characteristics of the food (de Wit, 1984, 1989; Harper, 1984). Unquestionably, proteins and other ingredients are important and are required to give the food desirable characteristics. However, our knowledge remains incomplete today because we still cannot fully predict the characteristics of a formulated food on the basis of our knowledge of the structure and function of pure proteins and other ingredients under strictly controlled conditions (Kinsella, 1982; de Wit, 1984, 1989; Harper, 1984; Owusu-Apenten, 2004; Zayas, 1996; Foegeding, 2015). Testing for functionality using simplified systems has been reviewed in depth by numerous investigators, including Kinsella (1982), Kilara (1984), Modler and Jones (1987), Mulvihill and Fox (1987), Patel and Fry (1987), Hall (1996), Zayas (1996), Owusu-Apenten (2004), and Foegeding (2015). The continued need to develop standardized testing for protein solubility, viscosity, water absorption, gelation, emulsification, and foaming properties has been emphasized by Mulvihill and Fox (1987), Patel and Fry (1987), German and Phillips (1994), Kilara (1994), Hall (1996), Zayas (1996), Luyten et al. (2004), and Foegeding (2015).

Protein functionality in foods Proteins used in foods include plant proteins (soy, wheat, rice, pea, potato, corn, canola, and other plant sources), milk proteins (micellar casein, caseins, caseinates, whey proteins, and milk protein concentrates and isolates—both caseins and whey proteins), egg proteins (egg white and egg yolk proteins), meat proteins, fish proteins, algal proteins, and insect proteins. Each type of protein exhibits different functional properties and has application in different types of food products (Inglett and Inglett, 1982; Kinsella, 1982; Lee et al., 1992; Kilara, 1994; Mangino et al., 1994; Owusu-Apenten, 2004). The major functionalities of food proteins include solubility, emulsification, gelation, viscosity enhancement, foaming, water binding, and heat stability. As shown in Table 15.1, different types of foods have different functional requirements and may require multiple functionalities.

Role of interactions in determining food characteristics

575

TABLE 15.1 Multiple functionalities in selected food products Food type

Multiple functionalities

Beverages

Solubility, heat stability, pH stability, color, or transparency

Baked goods

Emulsification, foaming, gelation

Dairy analogs

Gelation, foaming, emulsification, water holding

Egg substitutes

Foaming, gelation, emulsification

Meat emulsions

Emulsification, foaming, gelation, adhesion/cohesion

Soups and sauces

Viscosity, emulsification, water absorption

Infant formulas

Emulsification, heat stability

Whipped toppings

Foaming, emulsification

Frozen desserts

Foaming, gelation, emulsification

Source: Adapted from Kinsella, J.N., 1982. Relationship between structure and functional properties of food proteins. In: Fox, P.F., Condon, J.J. (Eds.), Food Proteins. Applied Science Publishers, New York, NY, pp. 51–103; de Wit, J.N., 1984. Functional properties of whey proteins in food systems. Neth. Milk Dairy J. 38, 71–89; Kilara, A., 1994. Whey protein functionality. In: Hettiarachchy, N.S., Ziegler, G.R. (Eds.), Protein Functionality in Food Systems. Marcel Dekker, New York, NY, pp. 325–355; Owusu-Apenten, R.K., 2004. Testing protein functionality. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 217–244.

Factors that may modify the protein during processing and hence its effect on the product characteristics include heat, shear, salts, pH, pressure, and enzyme modification (de Wit, 1984; Mangino et al., 1987; Yada, 2004; Banach et al., 2016b; Ho et al., 2018).

Role of interactions in determining food characteristics Interactions between ingredients and modifications caused by processing are the primary reasons why the functionality of proteins and other ingredients cannot be predicted in food systems (de Wit, 1984, 1989; Harper, 1984; Zayas, 1996; Owusu-Apenten, 2004; Yada, 2004; Foegeding, 2015). The following diagram provides an overview of the potential interactions that can occur in a food product (adapted from Harper, 1984): Surfactants

Fats and oils Proteins

Polysaccharides

Salts

Essentially, almost everything can modify the functionality of everything else. Salts are somewhat unique in that they do not in themselves affect product characteristics but can act on proteins, surfactants, polysaccharides, and, to some extent, polar lipids to modify the functionality of each in the food system. The nature and extent of these interactions will

576

15. Model food systems and protein functionality

be modified by pH, ionic strength, ingredient concentrations, and process-induced modifications. Some examples include the following: • Surfactants and proteins can interact competitively at the surface of an oil droplet to modify the interfacial properties and thus the characteristics of the emulsion, such as stability, size distribution, and light scattering. The extent to which a given component will dominate the characteristics will depend on the relative concentrations (Fig. 15.1) and chemical natures of the surfactant and the protein. • Starch is frequently used to provide texture in food products. However, the viscosity during processing and the final viscosity can be greatly altered by interactions with other components. Recent reviews of the use of simple model systems to obtain a better understanding of protein-food ingredient interactions and the mechanisms involved include Dickinson (2011, 2012), Foegeding and Davis (2011), Singh (2011), and Foegeding (2015).

A case study To further expand on protein interactions, the following presents a case study of the effect of various milk proteins on the modification of the pasting characteristics of potato starch, including the effect of different types of milk proteins and of differences in the concentrations of protein and/or starch. The texture of formulated foods is strongly modified by interactions. This section gives results of a simplified model system that provides a better understanding of food formulations that involve protein–food component interactions. The information provides an understanding of how milk proteins can modify the pasting properties of potato starch as investigated by Harper and Illingworth (unpublished data), Harper and Hemar (unpublished data), Doublier et al. (2001), and Bertolini et al. (2005). Starch is important in a number of food products in which the texture can be modified by protein-starch interactions to change the textural properties of the food.

Serum separation (mL per 10 mL sample)

4 3.5 3 2.5

0 Protein 2% Protein 4% Protein

2 1.5 1 0.5 0 0 EM

FIG. 15.1

0.05 EM

0.1 EM

0.2 EM

Effect of emulsifier EM (mono- and diglycerides) and protein concentration on phase separation of an oil-in-water emulsion (coffee whitener). Adapted from Harper, W.J., 1984. Model food system approaches for evaluating whey protein functionality. J. Dairy Sci. 67, 2745–2756.

Role of interactions in determining food characteristics

577

Effect of sodium caseinate on the pasting characteristics of different starches The Rapid Visco Analyzer (RVA) was used to evaluate the pasting properties of starch. Fig. 15.2 shows a typical starch pasting curve for potato starch. Generally, an 8% starch paste was utilized, and the various characteristics are shown as a function of time. The two most important characteristics are peak height and final viscosity. Peak height shows the maximum viscosity during manufacture, and the final viscosity gives a measure of the final texture. The addition of sodium caseinate to six different starches (corn, rice, wheat, potato, cassava, and waxy maize) influenced the peak (pasting) temperature, the time to reach peak viscosity, the peak viscosity, and the viscosity after cooling. Sodium caseinate had different effects on these parameters for the different starches. The percentage change in peak viscosity of the six sodium caseinate-starch mixtures is shown in Fig. 15.3. All starches, except potato starch, showed an increase in peak viscosity, whereas potato starch showed a dramatic decrease in peak viscosity. There was no statistically significant change in the final viscosities. The differences in the pasting characteristics will be important with respect to both processing and the characteristics of different food products using starch and caseinate. Based on the results of the marked difference between the pasting characteristics of potato starch and those for all the other starches, attention was given only to potato starch for further studies.

100

8000

B Typical starch gelation and pasting curve Temperature profile for starch gelation and pasting

90

80

4000 70

D C

2000

60

Temperature (°C)

Viscosity (cP)

6000

A = Initiation B = Peak viscosity C = Trough D = Final viscosity B-A = Peak time B-C = Breakdown D-C = Setback

50

A

0 0

200

400

600

800

Time (s)

FIG. 15.2 Diagram of the steps in starch pasting, including time to initiate gelation, peak viscosity, trough, final viscosity, breakdown, and setback.

578

15. Model food systems and protein functionality

FIG. 15.3 Peak viscosity during pasting of six different starches with sodium caseinate. Change in peak viscosity

200 CO = Corn

100

R = Rice 0

W = Wheat P = Potato

–100

CA = Cassava –200

WA = Waxy maize

–300 –400 CO

R

W

P CA WA

Starch

Effect of milk proteins on the pasting characteristics of potato starch, with emphasis on peak viscosity Proteins and polysaccharides, including starch, are frequently used together in food systems to import specific attributes to the final product. Hardacre et al. (2004) showed that sodium caseinate decreased the peak viscosity of potato starch. Different milk proteins were used to determine their effect on the pasting characteristics of potato starch. These included individual milk casein fractions, sodium caseinate with different concentrations of protein and starch, sodium caseinate, milk protein concentrate, and milk protein isolate. Individual casein fractions

Casein was fractionated into αs (αs1 + αs2) and β-caseins, and these were in turn converted to their sodium and calcium salts. The peak viscosity, final viscosity, and pasting time of starch, starch + 1% calcium caseinate, starch + 1% calcium αs-caseinate, and starch + 1% calcium β-caseinate are shown in Table 15.2 and Fig. 15.4 using standard starch concentrations (8%). The time to reach maximum viscosity increased with the added calcium caseinates. Peak viscosity decreased with the calcium caseinates, whereas there was no significant change final TABLE 15.2 RVA results showing the effect of the calcium salts of caseinate fractions on the viscosity and pasting of potato starch Viscosity (cP) Peak

Final

Pasting time (s)

Starch

7315

2952

196

Starch + 1% calcium αs-caseinate

4487

2975

232

Starch + 1% calcium β-caseinate

3806

2834

288

Starch + 1% calcium caseinate

3467

2543

336

579

Role of interactions in determining food characteristics 100

8000

90

6000

4000

70

Temperature (°C)

Viscosity (cP)

80

Starch 1% Ca αs-caseinate 1% Ca β -caseinate 1% Ca caseinate Temperature profile

60

2000

50

0

40 0

200

400

600

800

Time (s)

FIG. 15.4 Effect of calcium caseinate fractions and calcium caseinate on the peak viscosity of potato starch.

viscosity. The data suggest that the addition of caseinate would put less stress on processing, through a decrease in peak viscosity, without a significant effect on the final viscosity. The peak viscosity, final viscosity, and pasting time of starch, starch + 2.5% sodium caseinate, starch + 2.5% sodium αs-caseinate, starch + 2.5% sodium β-caseinate, starch + 5% sodium αs-caseinate, and starch + 5% sodium β-caseinate, are shown in Table 15.3. Effect of the concentration of starch and milk proteins on the pasting characteristics of potato starch The concentration of starch used for determining the effects of starch on pasting characteristics has generally been 6%–8%. However, the use level in most food applications ranges from 2% to 4%. Therefore, from a practical viewpoint, effects on the starch pasting characteristics at the food use concentration would be useful. The effect of protein concentration and starch concentration on the loss of peak viscosity and the final viscosity was determined for sodium caseinate, milk protein concentrate, and milk protein isolate. All of the different proteins gave similar results, and only the effects of sodium caseinate and milk protein isolate are presented in this investigation, for illustrative purposes. The loss in peak viscosity increased as the starch concentration increased and the sodium caseinate protein concentration decreased, as illustrated in Fig. 15.5. The effect of protein concentration was significant and linear up to 90% peak viscosity loss, from 0.02% to 1% sodium

580

15. Model food systems and protein functionality

TABLE 15.3 RVA results for the effect of sodium salts of caseinate fractions on the viscosity and pasting of potato starch

Peak viscosity loss (%)

Viscosity (cP) Peak

Final

Pasting time (s)

Starch

7361

2968

200

Starch + 2.5% sodium αs-caseinate

5885

3106

228

Starch + 2.5% sodium β-caseinate

4359

2712

228

Starch + 2.5% sodium caseinate

2113

2671

342

Starch + 5% sodium αs-caseinate

5676

3329

220

Starch + 5% sodium β-caseinate

4727

2947

212

100 90 80 70 60 50 40 30 20 10 0

6% Starch 4% Starch 3% Starch 2% Starch

0.02

0.1

1

2

3

4

NaCN (%)

FIG. 15.5 Effect of sodium caseinate (NaCN) concentration on the loss of peak viscosity as a function of potato starch concentration.

caseinate. However, once the protein concentration reached 1%, the percentage loss was similar for higher protein concentrations, peak viscosity losses ranging from 80% to 90%. Under standard test conditions with the RVA (8% starch), there was little to no loss in final viscosity with the addition of milk protein (data not shown). However, as the starch content was decreased below 6%, the loss in final viscosity markedly decreased, as shown in Fig. 15.6. This would limit the advantage of using potato starch to increase viscosity in foods containing milk protein, especially at starch concentrations below 4%.

Processing effects The functionality of commercial food proteins and other ingredients can be modified both during their production and during the processing of the food product itself. An overview of the conversion of a raw protein source to a functional food ingredient and the subsequent further processing during food manufacture is outlined in Fig. 15.7.

581

% Final viscosity loss

Uses of model food systems

100 90 80 70 60 50 40 30 20 10 0

6% Starch 4% Starch 3% Starch 2% Starch

0.02

0.1

1

2

3

% MPI

FIG. 15.6 Effect of milk protein isolate (MPI) concentration on the loss of final viscosity as a function of potato starch concentration.

During the production of commercial food proteins for use as food ingredients, the proteins may be exposed to a wide range of processing steps that can include thermal processes (pasteurization and sterilization, e.g., UHT and retorting), shear (pumping, mixing, and homogenization), pressure (high pressure processing), concentration (membrane processing, evaporation, and drying), precipitation (heat, acid, salts, and solvents), and enzyme modification. Each of these steps will modify the functional properties of the protein and thus will affect the final characteristics of the food (Kinsella, 1982; de Wit, 1984, 1989; Harper, 1984; Dybing and Smith, 1991; Kilara, 1994; Zayas, 1996; Owusu-Apenten, 2004; Ho et al., 2018). Such processes can alter functionality in food through a number of different modifications of the protein, including changes in sulfhydryl interactions, modification of secondary and quaternary structure, and shifts in the hydrophilic/lipophilic balance (Kinsella, 1982). Subsequent processing during use of the protein as a functional ingredient in food will bring further changes in the system, especially those occurring in the presence of other interacting ingredients. Generally, such changes in the characteristics of the food cannot be predicted; thus, there is a need for the use of model food systems as an intermediate step in product development (Franzen and Kinsella, 1974; Owusu-Apenten, 2004; Considine et al., 2010).

Uses of model food systems Model food systems can be used in a variety of ways (de Wit, 1984; Harper, 1984; OwusuApenten, 2004; Purwanti et al., 2010; Foegeding, 2015; Liu et al., 2016), including the following: • Determining the relative significance of the main effects of ingredients • Studying factors in food that affect chemical and physical changes (Maillard reaction, lipid oxidation, changes in water activity, viscosity, protein-protein interactions during processing and on storage, etc.)

582

15. Model food systems and protein functionality

Raw protein source “Native structure”

Ingredient manufacturer Biological properties

Processing (unit) operations

Commercial protein (Functional structure)

Food manufacturer Components and interactions (lipids, carbohydrates, salts)

Processing factors (temp, salt, pH, etc.)

Functionality

Fabricated food

FIG. 15.7 Steps in the manufacture of food proteins and the subsequent processes during food manufacture. From Owusu-Apenten, R.K., 2004. Testing protein functionality. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 217–244.

• • • • • • •

Evaluating the sensitivity of the food to alterations in formulation and processing Defining ingredient interactions Optimizing the formulation for robustness Determining critical steps in the processing of the product Determining interrelationships between ingredients and the process Tailor-making ingredients for a specific food application Evaluating and minimizing the sensitivity of product attributes to the formulation, to the process, and to subsequent storage • Determining the impact of ingredients on sensory and mouthfeel characteristics • Developing examples for patent applications Owusu-Apenten (2004) stated that the advantages of model food systems over standard functionality tests included (1) their ease of use, (2) the lack of a need for specialized equipment and methodologies, (3) the ability to aid in product optimization, and (4) the ability to test for multiple factors and interactions with respect to formulation and processing.

Uses of model food systems

583

Initial steps to developing model food systems The approach to developing a model food system is the same, regardless of whether the ingredient being investigated is a protein, lipid, emulsifier, starch, gum, etc. The development of a model food system begins by reviewing as many formulations as can be found and selecting those ingredients that are common to all formulations at a concentration that is at the central point of the various formulas (Harper, 1984). Next, a small-scale process for making the products is developed using processing steps and conditions as close to the commercial process as possible. When more than four or five ingredients are involved, it is often necessary to do a screening experiment to eliminate ingredients that do not have a main effect on important characteristics. Each different food will have different characteristics, which may include taste, color, and texture, that can be modified by the formulation and the process. Key attributes and methods for their evaluation need to be selected. Generally, the evaluation methods are different from those that are used in research (Owusu-Apenten, 2004).

Statistical design Statistical design is an essential component of model food systems because of the information it provides on ingredient and processing interactions (Dziezak, 1990; Earle et al., 2001; Hanrahan and Lu, 2006). Most fabricated food products have from 5 to 25 variables when both the ingredients and the processing steps are taken into consideration. This makes full factorial designs, which would exceed several hundred experiments, an impractical choice. Thus, fractional factorial screening designs are generally required. For most food products, the experimental design is a stepwise process, starting with screening experiments to minimize the variables that do not have major effects on the characteristics of the products. One of the most common screening designs is the Plackett-Burman, which can be used with up to 36 variables (Mullen and Ennis, 1985; Hanrahan and Lu, 2006). The screening experiments allow determination of the main effects that can be used in further fractional factorial designs; these designs will provide a better understanding of ingredient and process interactions and will generate response surfaces that give an understanding of the interactions (Hanrahan and Lu, 2006). In developing a fractional factorial experimental design in model food systems, it is necessary to know (1) the critical factors associated with the ingredients and the process, (2) the region of interest where the factor levels influencing the product characteristics are known, (3) that the factors vary continuously throughout the experimental range tested, (4) that a mathematical function relates the variable factors to the measured response, and (5) that the response defined by the function is a smooth curve. Mixture designs such as simplex centroid can be used to determine the effect of ingredient interactions, whether synergistic or antagonistic, on the properties of proteins in mixtures (Arteaga et al., 1993; Imtiaz et al., 2012). This obviates the assumption that the functional properties of a mixture are the weighted average of the properties of these individual ingredients. Numerous studies have used statistical design and response surface methodology to determine the effect of interactions on product characteristics and to optimize specific characteristics in a food (Dziezak, 1990; Zhen et al., 2016; Yolmeh and Jafari, 2017).

584

15. Model food systems and protein functionality

In developing an experimental design, consultation with a statistician who is familiar with the factors that affect the outcomes of the specific design is needed to avoid common pitfalls. Among these pitfalls are the following: (1) incorrectly defined or specified critical factors; (2) too narrow or too broad a range of factors selected, so that the optimum cannot be defined; (3) the lack of the use of good statistical practices; (4) too large a variation in the range of the factors utilized, introducing bias and error; (5) overreliance on computer-generated results; and (6) failure to ensure that the results make good sense and are practically significant.

Applications of model food systems Initially, model food systems were applied to milk proteins to gain a better understanding of what was required to get desired characteristics in complex food products that could not be predicted from standard functionality tests. de Wit (1998) has stated: “Information obtained from functional characterization tests in model systems is more suitable to explain retroactively protein behaviors in complex food systems than to predict functionality.” What has been learned using milk proteins in model food systems has been shown to be equally applicable to other food proteins. In addition to understanding the protein being used, there is a need to know the functionality of other ingredients in the food, the probability of how they will interact and modify the function of the food protein, and the use of statistical design to gain the full potential of the model system approach. Studies of model food systems, used to assess their performance in foods, have included a large number of different types of foods, as shown in Table 15.4. The examples of the model food systems used to illustrate applications in this chapter are primarily from the first-generation and second-generation categories. These include bakery products, dairy products, infant formula, meat products, sauces and dressings, whipped toppings, cheeses, fermented foods such as yogurt, protein nutrition bars, and high-protein beverages.

Bakery products Bread represents a system in which the methods of the evaluation of the ingredients have been standardized and covered by American Association of Cereal Chemists (AACC)–approved methods (AACC methods 10-9, 10-10 and 10-11). Details of the procedures and evaluation techniques have been given by various investigators (Lindblom, 1977; Pomeranz et al., 1984; Ranhortra et al., 1992; Fenn et al., 1994; Cauvain and Young, 2006). In general, the substitution or addition of other proteins (milk, whey proteins, etc.) can lead to a loss of loaf volume (Harper et al., 1980; de Wit, 1984). Harper and Zadow (1984) found that heat treatments that prevented loss of loaf volume in bread made with milk powder were ineffective in preventing loss of loaf volume in bread made with whey protein concentrates. Model food systems have been used widely in cake systems including pound cake, which consists of equal proportions of egg, flour, and sugar (Lee, 1999; Wilderjans et al., 2010; Deleu et al., 2017; Lambrecht et al., 2018); Madeira cake (de Wit, 1984); white cake (Harper et al., 1980);

Applications of model food systems

585

TABLE 15.4 Model food systems used to assess functionality in foods First-generation model foodsa

Second-generation model foodsb,c

Bread

Beef batters and beef patties

Cakes (angel food, high ratio cake, pound)

Cheese

Coffee whitener

Cream

Drinks

Gravies

Ham

Low fat spreads

Ice cream

Meat emulsions

Infant formula

Milk

Meringues

Nutritional beverage emulsions

Restructured meats

Processed cheese

Salad dressing

Protein nutrition bars

Sausage

Soups and sauces

Starch pudding

Surimi

Whipped topping

Yogurt-acid milk gels

a

Adapted from Harper, W.J., 1984. Model food system approaches for evaluating whey protein functionality. J. Dairy Sci. 67, 2745–2756; de Wit, J.N., 1984. Functional properties of whey proteins in food systems. Neth. Milk Dairy J., 38, 71–89. b Adapted from Owusu-Apenten, R.K., 2004. Testing protein functionality. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 217–244. c Drapala, K.P., Mulvihill, D.M., O’Mahony, J.A., 2018. A review of the analytical approaches used for studying the structure interactions and stability of emulsions in nutritional beverage systems. Food Struct. 16, 27–42.

and angel food cake (Kissell and Bean, 1978). Factors that should be considered in developing a model food system for bakery products are outlined briefly in Table 15.5. Of these, angel food cake has received the most attention (Lowe et al., 1969; DeVilbiss et al., 1974; Cunningham, 1976; Regenstein et al., 1978; Johnson and Zabik, 1981a,b; Ball and Winn, 1982; Froning et al., 1987; Froning, 1988; Martinez et al., 1995; Berry et al., 2009). Angel food cake consists of only three main ingredients: egg whites, sugar, and flour (42:42:15) (Pyler, 1988). An enriched wet foam of egg white-sugar is formed, followed by the addition of the rest of the sugar and the flour while maintaining the foam structure (Yang and Foegeding, 2010). The primary protein evaluated has been egg white, for which cake height and texture can be related to the individual egg white proteins ( Johnson and Zabik, 1981a,b; Ball and Winn, 1982). Attempts to replace egg white with whey proteins have never been completely successful (DeVilbiss et al., 1974; Harper et al., 1980; Pernell et al., 2002; Berry et al., 2009). Arunepanlop et al. (1996) were able to replace 25%–50% of the egg white with whey protein and could achieve greater replacement by the addition of xanthan gum, which effectively increases the viscosity of the batter phase. The cake volume was essentially the same as it is with egg white, but the cakes collapsed upon baking. This emphasizes the requirement for both foaming and gelation (Owusu-Apenten, 2004). This is due in part to the lower gelation temperature for foams made with egg white (Pernell et al., 2002) and in part to the shear-induced denaturation of egg white with mixing (DeVilbiss et al., 1974); whey proteins are less sensitive to shear.

586

15. Model food systems and protein functionality

TABLE 15.5 Factors affecting the functionality of protein in bakery products Product type

Functional requirement of protein

Ingredient modifying functionality

Processing factors affecting functionality

Bread

Dough formation, water binding, gelation, elasticity of dough

Protein source, polar lipids, oxidizing and reducing agents, other proteins with sulfhydryl groups

Mixing, method of breadmaking (sponge dough versus mechanical dough development)

Cakes (fatted)

Fat binding, foaming, gelation

Protein type and concentration, gums, fat, sugar concentration, emulsifier, starch

Mixing speed and time, preemulsification, batter specific gravity, time and temperature of baking

Other proteins that have been evaluated for angel food cake include blood plasma protein (Kahn et al., 1979; Raeker and Johnson, 1995) and dried beef plasma (Duxbury, 1988). AbuGhoush et al. (2010) investigated an even wider range of egg alternatives that included different types of collagen and gelatin, whey protein concentrate, whey protein isolate, casein, dairy protein hydrolyzates, and vegetable proteins from rice, soy, and corn. Only the whey protein isolate formed and maintained a stable foam to the baking process; the other foams collapsed. Egg white replacement with lentil protein was studied by Jarpa-Parra et al. (2017). They too found that only partial replacement ( 50%) could be used without adversely affecting crumb structure and color and firming of the cake on storage.

Dairy analogs Dairy analogs include coffee whiteners, whipped toppings, and processed cheese products. Coffee whiteners Generally, coffee whiteners, first developed in the 1950s, are protein-stabilized oil-in water emulsions, with vegetable oil as the dispersed phase. A model system developed by Harper and Raman (1979) and Harper et al. (1980) utilized caseinate, soy bean oil, carbohydrate, phosphate, emulsifier, and a gum (xanthan gum or carrageenan). The role of the ingredients has been reviewed by Knightly (1969) and Patel et al. (1992), and the process has been reviewed by Owusu-Apenten (2004). This is summarized in Table 15.6. Patented processes include using milk protein retentate (Kosikowski and Jimenez-Flores 1987), reformed casein micelles (McKenna et al., 1992), phosphate-modified milk protein (Melachouris et al., 1994), modified milk protein concentrate (Bhaskar, 2008), and soy proteins (Melmychyn, 1973). Alternative proteins that have been suggested to replace caseinate include milk protein concentrate (Euston and Hirst, 2000), whey protein (Hlavacek et al., 1970; Gruetzmacher and Bradley, 1991; Euston and Hirst, 2000), wheat protein (Golde and Schmidt, 2005; Patil et al., 2006), soy protein (Hlavacek et al., 1970; Golde and Schmidt, 2005), peanut protein (Malundo et al., 1992), and cottonseed protein (Choi et al., 1982).

587

Applications of model food systems

TABLE 15.6

Factors affecting the functionality of protein in coffee whiteners

Product type

Functional requirement of protein

Ingredient modifying functionality

Processing factors affecting functionality

Coffee whiteners

Emulsification, whitening ability, stability to the pH and temperature of coffee

Emulsifiers, gums, phosphate or citrate buffers and chelators, calcium

Homogenization pressures, pasteurization or sterilization time and temperature, temperature and pH of coffee

Coffee whiteners are evaluated to ensure that they provide an emulsion with a small fat globule size to maximize whiteness and reduce fat rise, minimize astringency of the coffee by binding with the coffee tannins, maintain stability in hot coffee under acidic conditions, minimize feathering in the presence of hard water salts, and readily disperse in the coffee (Pearce and Harper, 1982; Tran and Einerson, 1987; Kneifel et al., 1992; Kelly et al., 1999). Golde and Schmidt (2005) compared coffee whiteners made from sodium caseinate, soy protein isolate, and wheat protein isolate and found that they gave similar whiteness (L*) to the coffee. However, the liquid coffee whiteners made with wheat protein tended to separate upon storage, and the whiteners made with soy protein isolate tended to show feathering. Whipped toppings Most commercial whipped toppings contain sodium caseinate as the protein of choice (Knightly, 1968). Other proteins used for whipped toppings include whey protein concentrate (Peltonen-Shalaby and Mangino, 1986; Liao and Mangino, 1987), modified milk protein concentrates (Bhaskar, 2008), and soy protein isolates (Kolar et al., 1979; Lah et al., 1980; Chow et al., 1988; Abdullah et al., 1993; Shurtleff and Aoyagi, 1994). Whipped toppings are high fat (25%–30%), foamed emulsions with about 40% total solids, and model food systems generally also contain sugars, gums, and low-molecular-weight emulsifiers (Knightly, 1968; Harper et al., 1980). A brief summary of key factors that affect the functionality of protein in whipped toppings is given in Table 15.7. The model system differs from whipping or foaming tests with respect to both composition and a much lower fat content in foaming tests (Owusu-Apenten, 2004). Min and Thomas (1977) found that calcium addition to a 15% fat-containing whipped topping stabilized with sodium caseinate gave improved stability to the system. Peltonen-Shalaby and Mangino (1986) showed that pasteurization also improved the overrun of the topping. Liao and Mangino (1987) used whey TABLE 15.7 Product type Whipped toppings

Factors affecting the functionality of protein in whipped toppings

Functional requirement of protein Emulsification, whipping to 200% overrun in the presence of high fat and high solids and whipped foam stability

Ingredient modifying functionality

Processing factors affecting functionality

Emulsifiers, gums, phosphate, calcium

Homogenization pressures, pasteurization or sterilization time and temperature, equipment used to produce the whipped product

588

15. Model food systems and protein functionality

proteins to make a model whipped topping and found a positive correlation between exposed hydrophobicity and overrun. Other factors that affect overrun and stability include the hardness of the fat, the type and percentage of emulsifier, homogenization pressures to form the emulsion, and the equipment used for mixing the whipped topping (Harper, 1984; Stanley et al., 1996).

Salad dressings Salad dressings are high fat emulsions that are frequently stabilized by high shear in the presence of egg yolk as the primary emulsifier (Parker et al., 1995). Mayonnaise, a spoonable dressing, contains 75% oil by definition and egg yolks. Subsequently, starch pastes were used to make a spoonable dressing with about 40% oil. Today, the most common dressings are pourable, with a wide range of oil contents, and are stabilized primarily by xanthan gum (Franco et al., 1995). Model oil-in-water emulsion food systems have been used to gain a better understanding of both ingredients (Smith, 1977; Paredes et al., 1988; Chung et al., 2014) and processing (Parker et al., 1995). Smith (1977), using a central composite statistical design, found that the coefficients of the regression analysis were larger for the interaction terms than for the main effect terms in pourable salad dressing with 40% oil and containing egg, vinegar, xanthan gum, and mustard powder. The order of addition was also found to be important to the viscous properties of the pourable salad dressing. Zhen et al. (2016) also used a central composite design to examine the effects of pulse (pea, lentil, and chickpea) protein isolates, egg yolk, and different oil contents on the rheological behavior of salad dressing. Response surface methodology was subsequently used to optimize the formulations to give the desired textural and emulsion stability properties. The use of microparticulated whey protein with polysaccharides as a fat mimetic in a low calorie dressing was reported by Chung et al. (2014). Changing the ionic strength by calcium chloride addition and altering the pH (2–8) had little impact on the dressing system, whereas increasing the protein content increased its viscosity.

Meat products Model meat products, including beef, pork, lamb, poultry, and fish, have been utilized for recombined meats (ham, steaks, etc.) and meat macroemulsions (comminuted meats) that include bologna, sausages, liver sausages, frankfurters, and meat loaves. Nonmeat proteins have been injected into beef and ham, together with water, followed by tumbling to maintain nutritionally equivalent protein levels, increase yield, and improve texture (Zayas, 1996; Yada, 2004; Szerman et al., 2007). Szerman et al. (2007) found whey protein isolates to be superior to vegetable proteins on the basis of flavor. Meat emulsions generally have particle size distributions of between 0.1 and 50 mm, and many investigators suggest that they are three-dimensional gel networks with entrapped fat (Regenstein, 1989; Krishnan and Sharma, 1990; Xiong et al., 1992; Correia and Mittal, 1993;

Applications of model food systems

589

Barbut, 1995). However, most reviewers continue to classify them as meat emulsions (Gordon, 1969; Webb, 1974; Owusu-Apenten, 2004). The factors that affect the functionality of the proteins in these products include meat extraction temperature, emulsification temperature, shear during emulsification, fat melting point, pH, ionic strength, ratios of ingredients, salt, soluble protein concentration, and type of salt (salt, phosphates, citrates, etc.). Achievement of functionality has been determined by a number of different methods, including • emulsification capacity (EC) (Swift et al., 1961; Swift and Sulzbacher, 1963), • emulsion activity (EA) (Acton and Saffle, 1972), • emulsion stability (ES) (Carpenter and Saffle, 1964; Townsend et al., 1968; Marshall et al., 1975). Although the tests for EC and ES are widely used for comminuted meat products, there does not appear to be much collaborative testing of the different methods (Owusu-Apenten, 2004). The type of protein affects the EC of meat emulsions, with isolated muscle proteins giving different EC values. In general, the EC was in the order of myosin > actinomyosin > actin for different types of meat (Tsai et al., 1972; Galluzzo and Regenstein, 1978; Li-Chan et al., 1984). Substitution of the meat protein with other proteins in meat emulsions, as measured by large deformation rheological testing, showed that • gluten, soy protein isolate, or egg white increased the yield after the cooking of meat emulsions (Randall et al., 1976); • corn germ protein at 2% substitution reduced the shear force and reduced cooking losses (Mittal and Usborne, 1985); • partial substitution of meat protein with sodium caseinate, soy protein isolate, whey protein concentrate, or wheat germ protein all increased the cook yield, increased the protein level, and decreased the fat in frankfurters, without affecting quality (Atughonu et al., 1988); • sodium caseinate, milk protein isolate, whey protein concentrate, and whey protein isolate addition to an extensively washed chicken breast meat model showed that whey protein concentrate and whey protein isolate improved the ES, whereas the other dairy protein sources inhibited actin functionality in the emulsion and therefore decreased stability (Imm and Regenstein, 1998); • addition of bovine blood plasma to meat emulsion products improved the ES and yield, and the contents of protein, phenylalanine, and valine (Marquez et al., 1997).

Protein nutrition bars For the sports person, the elderly, and those in hospital care, there is a market for proteinrich foods (Drapala et al., 2018), and these bring inherent issues in texture and stability (Purwanti et al., 2010). Model foods can be used to study these effects, and an example is protein nutrition bars (e.g., McMahon et al., 2009; Loveday et al., 2010; Hogan et al., 2012; Imtiaz et al., 2012; Banach et al., 2017). Protein nutrition bars are microbiologically stable as they are intermediate moisture foods, with a water activity of generally 0.6. However, they are

590

15. Model food systems and protein functionality

heterogeneous systems containing protein, fat, sugar syrups, and polyols and undergo deteriorative reactions such as hardening and Maillard browning during storage, limiting consumer acceptability. The dairy proteins used in nutrition bars include milk protein concentrates and isolates, whey proteins, and caseinates. The effect of physical, chemical, and enzymatic modifications of the proteins in model protein bars has been evaluated. Physical properties include the powder particle size, which is altered by, for example, freeze drying, extrusion, and jet milling of milk protein concentrates (Banach et al., 2016a, 2017). These processes also affect particle density, occluded air, hydrophobicity, and water-holding capacity. Hogan et al. (2016) took a soft matter approach and determined that the onset of hardening in whey protein–based bars was a function of powder type, volume fraction, and particle interaction energy. Heat treatment of milk protein concentrate during its manufacture (Imtiaz et al., 2012) and calcium reduction (Banach et al., 2016b) have also been studied as a means to alter textural properties. Enzymatic modification by cross-linking of the casein micelle improved cohesiveness and reduced hardening (Banach et al., 2016b), and hydrolysis of whey protein softened bars and slowed the progression of hardening by suppressing the glass transition temperature and plasticizing the system (Hogan et al., 2016). Chemical reactions such as disulfide bonding, Maillard browning, and oxidation have also been implicated in bar hardening (e.g., Rao et al., 2013). As protein bars are multicomponent systems, the interaction between cosolvents and the protein surface has been implicated in textural changes (Hassan and McMahon, 2016) as has the migration of small molecules (water, polyols) into protein particles, resulting in changes in microstructure (Lu et al., 2016).

Use of model food systems for other food components In addition to evaluating the performance of proteins in food systems, a wide range of other applications have been utilized. During the past several years, many papers have been published on other uses of model food systems. A full review of such uses is outside the scope of this chapter. However, selected applications from studies over the past several years are cited both to present a basis for understanding the scope of the use of model food systems in the food industry and as patent examples to provide a starting point for obtaining more detailed information. Applications include • factors affecting flavor release in foods (Bylaite et al., 2005; Heinemann et al., 2005; CondePetit et al., 2006; Nongonierma et al., 2006; Seuvre et al., 2007), • factors affecting D values in food (Rodriguez et al., 2007), • lipid oxidation ( Jaswir et al., 2004; Sakanaka and Tachibana, 2006; Wijeratne et al., 2006; Rao et al., 2013; Kobayashi et al., 2017), • water migration in foods (Guignon et al., 2005; Doona and Moo, 2007; Loveday et al., 2010; Hogan et al., 2012), • Maillard reaction investigations (Severini et al., 2003; Miao and Roos, 2005; Acevedo et al., 2006; Casal et al., 2006; Potes et al., 2014), • stability of antioxidants in beverages (Wegrzyn et al., 2008),

Conclusions

591

• effects of high pressure processing of food (Severini et al., 2003; Sila et al., 2007; Venir et al., 2010), • phase separation phenomena in concentrated protein food systems (McMahon et al., 2009; Loveday et al., 2010), • migration of small molecules implicated in nutrition bar hardening (Lu et al., 2016), • sensory and creamy mouthfeel perception (Liu et al., 2016). Patent examples similar to model foods include • different proteins for analog cheeses and dairy products (Fitzsimons et al., 2003; Ye and Taylor, 2008a; Boursier et al., 2015); • different proteins and processing treatments of beverage models (Fletcher et al., 2007; Hofman et al., 2010; Boursier et al., 2015; Wearly et al., 2016); • lipid-protein emulsions with different processes for salad dressings, spreads, and yogurts (Bhaskar, 2008; Ye and Taylor, 2008a,b).

Limitations Model food systems can tell you “what,” but they cannot tell you the mechanism(s) by which the effects occur. Frequently, the results with a model system cannot be scaled up to full commercial practice because of differences in equipment and processes. However, they do provide insight into directions to take to overcome scale-up problems. Generally, the results are valid only within the parameters that have been established. Optimization of a food system can sometimes be outside the limits of either the processing equipment or the functionality of a specific ingredient.

Conclusions Historically, model food systems were used first to improve the functionality of milk proteins in food systems. Currently, there are limited publications on the use of milk proteins for these purposes, although it is known that a number of dairy food companies use model food systems in their product development programs. Today, the publications concerning model food systems have a much broader usage, with attention being given to a better understanding of how complex food systems affect such factors as oxidation, Maillard reactions, and shelf-life. Model food systems can be a valuable tool in product development with respect to developments of both formulations and manufacturing processes and have a role in the development of ingredients for new foods. Model food systems do not provide information on why interactions occur, but they can provide insights into which interactions need basic study to provide a more robust product. In the future, model food systems can be expected to continue to provide a better understanding of how interactions modify the functionality of proteins in complex food systems and to give insight into how to use this information to interface with studies on the basis of protein structure/function.

592

15. Model food systems and protein functionality

References Abdullah, A., Resurreccion, A.V.A., Beuchat, L.R., 1993. Formulation and evaluation of a peanut milk based whipped topping using response surface methodology. Lebensm. Wiss. Technol. 26, 167–170. Abu-Ghoush, M., Herald, T.J., Aramouni, F.M., 2010. Comparative study of egg white protein and egg alternatives used in an angel food cake system. J. Food Process. Pres. 34 (Suppl. 2), 411–425. Acevedo, N., Schebor, C., Buera, M.P., 2006. Water solids interactions: matrix, structural properties and the rate of non-enzymatic browning. J. Food Eng. 77 (4), 1108–1115. Acton, J.C., Saffle, R.L., 1972. Emulsifying capacity of muscle protein: phase volumes at emulsion collapse. J. Food Sci. 37, 904–906. Ako, K., Nicolai, T., Durand, D., Brontons, G., 2009. Micro-phase separation explains the abrupt structural change of globular protein gels on varying ionic strength or the pH. Soft Matter 5, 4033–4041. Anfinsen, C.B., 1972. Studies on the Principles That Govern the Folding of Protein Chains. Nobel Lecture, Nobel Foundation. Arteaga, G.E., Li-Chan, E., Nakai, S., Cofrades, S., Jimenez-Colmenero, F., 1993. Ingredient interaction effects on protein functionality: mixture design approach. J. Food Sci. 58 (3), 656–662. Arunepanlop, B., Morr, C.V., Karleskind, D., Laye, I., 1996. Partial replacement of egg white proteins with whey proteins in angel food cakes. J. Food Sci. 61, 1085–1093. Atughonu, A.-G., Zayas, J.-F., Herald, T.-J., Harbers, L.-H., 1988. Thermo-rheology, quality characteristics, and microstructure of frankfurters prepared with selected plant and milk additives. J. Food Qual. 21, 223–238. Ball Jr., H.R., Winn, S.E., 1982. Acylation of egg white proteins with acetic anhydride and succinic anhydride. Poult. Sci. 61, 1041–1046. Banach, J.C., Clark, S., Lamsal, B.P., 2016a. Instrumental and sensory attributes of high protein nutrition bars formulated with extruded milk protein concentrate. J. Food Sci. 81 (5), S1254–S1262. Banach, J.C., Clark, S., Metzger, L.E., Lamsal, B.P., 2016b. Textural performance of cross-linked or reduced-calcium milk protein ingredients in model high-protein nutrition bars. J. Dairy Sci. 99 (8), 6061–6070. Banach, J.C., Clark, S., Lamsal, B.P., 2017. Particle size of milk protein concentrate powder affects the texture of highprotein nutrition bars during storage. J. Food Sci. 82 (4), 913–921. Barbut, S., 1995. Importance of fat emulsification and protein matrix characteristics in meat batter stability. J. Muscle Foods 6, 161–177. Berry, T.K., Yang, X., Foegeding, E.A., 2009. Foams prepared from whey protein isolate and egg white protein. 2. Changes associated with angel food cake functionality. J. Food Sci. 74, E269–E277. Bertolini, A.C., Creamer, L., Eppink, M., Boland, M., 2005. Some rheological properties of sodium caseinate starch gels. J. Agric. Food Chem. 53, 2248–2254. Bhaskar, G.V., 2008. Calcium Depleted Milk Protein Concentrates for Stabilizing Foods. Assignee Fonterra Co-operative Group Ltd. PCT Patent Application WO 2008/206940 A1. Boursier, B., Moretti, E., Ribadeau-Dumas, G., Belaid, S., Riaublac, A., Gueguen, J., Lepoudere, A., Snapper, J.J., Colin, I., 2015. Assembly of At Least One Vegetable Protein and At Least One Dairy Protein. Assignee Roquette Freres, Ingredia, Institut National de la Recherche Agronomique. United States Patent Application US2015/0237885 A1. Bylaite, E., Adler-Nissen, J., Meyer, A.S., 2005. Effect of xanthan on flavor release from thickened food model systems. J. Agric. Food Chem. 53, 3577–3583. Carpenter, J.A., Saffle, R.L., 1964. A simple method of estimating the emulsifying capacity of various sausage meats. J. Food Sci. 29, 774–781. Casal, E., Ramirez, P., Ibanez, E., Corzo, N., Olano, A., 2006. Effect of supercritical carbon dioxide treatment on the Maillard reaction in model food systems. Food Chem. 97 (2), 272–276. Cauvain, S., Young, L., 2006. Baked Products. Blackwell Publishing Ltd., Oxford, UK. Choi, Y.R., Lucas, E.W., Rhee, K.C., 1982. Formulation of non-dairy coffee whiteners with cottonseed protein isolates. J. Am. Oil Chem. Soc. 59, 564–567. Chow, E.T.S., Wei, L.S., DeVor, R.E., Steinberg, M.P., 1988. Performance of ingredients in a soybean whipped topping: a response surface analysis. J. Food Sci. 53, 1761–1765. Chung, C., Degner, B., McClements, D.J., 2014. Reduced calorie emulsion-based foods: protein microparticles and dietary fiber as fat replacers. Food Res. Int. 64, 664–670. Conde-Petit, B., Escher, F., Nuessli, J., 2006. Structural features of starch-flavor complexation in food model systems. Trends Food Sci. Technol. 17 (5), 227–235.

References

593

Considine, T., Noisuwan, A., Hemar, Y., Wilkinson, B., Bronlund, J., Kasapis, S., 2010. Rheological investigations of the interactions between starch and milk proteins in model dairy systems: a review. Food Hydrocoll. 45, 327–336. Correia, L.R., Mittal, G.S., 1993. Selection of fillers based on zp values of meat emulsion properties during smokehouse cooking. Int. J. Food Sci. Technol. 28 (5), 443–451. Cunningham, F.E., 1976. Properties of egg white foam drainage. Poult. Sci. 55, 738–743. de Wit, J.N., 1984. Functional properties of whey proteins in food systems. Neth. Milk Dairy J. 38, 71–89. de Wit, J.N., 1989. The use of whey protein products. In: Fox, P.F. (Ed.), Developments in Dairy Chemistry. In: vol. 3. Elsevier Science Publishers, London, UK, pp. 323–345. de Wit, J.N., 1998. Nutrition and functional characteristics of whey protein in foods. J. Dairy Sci. 81, 597–608. Deleu, L.J., Melis, S., Wilderjans, E., van Haesendonck, I., Brijs, K., Delcour, J.A., 2017. Protein network formation during pound cake baking: the role of egg yolk and its fractions. Food Hydrocoll. 63, 226–232. DeVilbiss, E.D., Holsinger, V.H., Posati, L.P., Pallansch, M.J., 1974. Properties of whey protein concentrates. Food Technol. 28 (3), 40–42. 46, 48. Dickinson, E., 2011. Mixed biopolymers at interfaces: competitive absorption and multilayer structures. Food Hydrocoll. 25 (8), 1966–1983. Dickinson, E., 2012. Emulsion gels: the structuring of soft gels with protein-stabilized oil droplets. Food Hydrocoll. 28 (1), 224–241. Doona, C.J., Moo, Y.B., 2007. Molecular mobility in model dough systems studied by time-domain nuclear magnetic resonance spectroscopy. J. Cereal Sci. 45 (3), 257–262. Doublier, J.-L., Garnier, C., Renard, D., Sanchez, C., 2001. Protein-polysaccharide interactions. Curr. Opin. Colloid Interface Sci. 5 (3/4), 202–214. Drapala, K.P., Mulvihill, D.M., O’Mahony, J.A., 2018. A review of the analytical approaches used for studying the structure interactions and stability of emulsions in nutritional beverage systems. Food Struct. 16, 27–42. Duxbury, D.D., 1988. Powdered beef plasma replaces eggs in cakes. Food Processing 49 (2), 73–74. Dybing, S.T., Smith, D.E., 1991. Relation of chemistry and processing procedures to whey protein functionality. Cult. Dairy Prod. J. 26 (1), 4–12. Dziezak, J.D., 1990. Taking the gamble out of food product development. Food Technol. 44 (6), 109–117. Earle, M.D., Earle, R.L., Anderson, A., 2001. Food Product Development. Woodhead Publishing, Cambridge, UK. Euston, S.R., Hirst, R.L., 2000. The emulsifying properties of commercial milk protein products in simple oil-in-water emulsions and in a model system. J. Food Sci. 65, 934–940. Fenn, D., Lukow, O.M., Bushuk, W., Depauw, R.M., 1994. Milling and baking quality of 1BL/1RS translocation wheats. I. Effects of genotype and environment. Cereal Chem. 71, 189–195. Fitzsimons, W., Hewitt, S., Gregory, S., Carr, A., Mills, O., 2003. Method of Making High Solids, High Protein DairyBased Food. Assignee Fonterra Co-Operative Group Ltd. PCT Patent Application WO 03/051130 A2. Fletcher, K., Chen, P., Carroll, T., 2007. Enhancing Clarity and/or Stability Properties of Protein-Containing Liquids and Gels. Assignee Fonterra Co-operative Group Ltd. United States Patent Application US 20070822118 (A1). Foegeding, E.A., 2015. Food protein functionality—a new model. J. Food Sci. 80, C2670–C2677. Foegeding, E.A., Davis, J.P., 2011. Food protein functionality: a comprehensive approach. Food Hydrocoll. 25, 1853–1864. Franco, J.M., Guerrero, A., Gallegos, C., 1995. Rheology and processing of salad dressing emulsions. Rheol. Acta 34, 513–524. Franzen, K.I., Kinsella, J.E., 1974. Parameters affecting the binding of volatile flavor compounds in model food systems. I. Proteins. J. Agric. Food Chem. 22, 675–678. Froning, G.W., 1988. Nutritional and functional properties of egg proteins. In: Hudson, B.J.F. (Ed.), Developments in Food Proteins. In: vol. 6. Elsevier Applied Science Publishers, London, UK, pp. 1–34. Froning, G.W., Wehling, R.L., Ball, H.R., Hill, R.M., 1987. Effect of ultrafiltration and reverse osmosis on the composition and functional properties of egg white. Poult. Sci. 66, 1168–1173. Galluzzo, S.J., Regenstein, J.M., 1978. Role of chicken breast muscle proteins in meat emulsion formation: myosin, actin and synthetic actomyosin. J. Food Sci. 43 (6), 1761–1765. German, J.B., Phillips, L., 1994. Protein interactions in foams: protein-gas interactions. In: Hettiarachchy, N.S., Ziegler, G.R. (Eds.), Protein Functionality in Food Systems. Marcel Dekker, New York, NY, pp. 181–208. Ghoush, M.A., Samhouri, M., Al-Holy, M., Hera, T., 2008. Formulation and fuzzy modeling of emulsion stability and viscosity of a gum–protein emulsifier in a model mayonnaise system. J. Food Eng. 84 (2), 348–357.

594

15. Model food systems and protein functionality

Golde, A.E., Schmidt, K.A., 2005. Quality of coffee creamers as a function of protein sources. J. Food Qual. 28 (1), 46–61. Gordon, A., 1969. Emulsification in comminuted meat systems. II. Food Process. Ind. 38 (459), 50–52. 54. Gruetzmacher, T.J., Bradley Jr., R.L., 1991. Acid whey as a replacement for sodium caseinate in spray-dried coffee whiteners. J. Dairy Sci. 74, 2838–2849. Guignon, B., Otero, L., Molina-Garcia, A.D., Sang, P.D., 2005. Liquid water-ice phase diagrams under high pressure: sodium chloride and suggested model food systems. Biotechnol. Prog. 21, 439–445. Hall, G.M. (Ed.), 1996. Methods of Testing Protein Functionality. Blackie Academic and Professional, London, UK. Hanrahan, G., Lu, K., 2006. Application of factorial and response surface methodology in modern experimental design and optimization. Crit. Rev. Anal. Chem. 36, 141–151. Hardacre, A., Clark, S.N., Harper, J., Hemar, Y., Illingworth, D., Boland, M., 2004. Variation in the rheological properties of potato starches and interaction with casein and phosphorus. Report 2. Crop and Foods Research Report. Harper, W.J., 1984. Model food system approaches for evaluating whey protein functionality. J. Dairy Sci. 67, 2745–2756. Harper, W.J., Raman, R.V., 1979. Protein interactions in whey protein concentrates and functionality in model food systems. New Zealand J. Dairy Sci. Technol. 14, 198–199. Harper, W.J., Zadow, G., 1984. Heat-induced changes in whey protein concentrates as related to bread manufacture. New Zeal. J. Dairy Sci. Technol. 19, 229–237. Harper, W.J., Peltonen, R., Hayes, J., 1980. Model food systems yield clearer utility evaluations of whey protein. Food Product Develop. 14 (10). 52, 54, 56. Hassan, S.K., McMahon, D.J., 2016. Hardening and microstructure of high protein nutrition bars made using whey protein isolate or milk protein concentrate. J. Anim. Sci. 94 (Suppl. 5), 439–440. Heinemann, C., Zinsli, M., Renggli, A., Escher, F., Conde-Petit, B., 2005. Influence of amylose-flavor complexation on build-up and breakdown of starch structures in aqueous food model systems. Lebensm. Wiss. Technol. 38, 885–894. Hlavacek, R.G., Robe, K., Stinson, W.S., Belshaw, F., 1970. Reduce fat at 1/3 in coffee whiteners, emulsifiers— stabilizer also improves flavor. Food Processing 31 (5), 18. Ho, Q.T., Murphy, K.M., Drapala, K.P., O’Callaghan, T.P., Fenelon, M.A., O’Mahony, J.A., McCarthy, N.A., 2018. Effect of pH and heat treatment on viscosity and heat coagulation properties of milk protein concentrate. Int. Dairy J. 85, 219–224. Hofman, Z., Van Anhold, R.D., Kiers, W.H.A., 2010. Pea-Based Protein Mixture and Use Thereof in a Liquid Nutritional Composition Suitable for Enteral Feeding. Assignee Nutricia. PCT Patent Application WO 2010/126353 A1. Hogan, S.A., Chaurin, V., O’Kennedy, B.T., Kelly, P.M., 2012. Influence of dairy proteins on textural changes in highprotein bars. Int. Dairy J. 26, 58–65. Hogan, S.A., O’Loughlin, I.B., Kelly, P.M., 2016. Soft matter characterisation of whey protein powder system. Int. Dairy J. 52, 1–9. Imm, J.Y., Regenstein, J.H., 1998. Emulsification of commercial dairy proteins with exhaustively washed muscle. J. Food Sci. 63 (6), 1012–1074. Imtiaz, S.R., Kuhn-Sherlock, B., Campbell, M., 2012. Effect of dairy protein blends on texture of high protein bars. J. Texture Stud. 43, 275–286. Inglett, M.J., Inglett, G.E., 1982. Food Products Formulary, vol. 4, Fabricated Foods. AVI Publishing Co., Westport, CT. Jarpa-Parra, M., Wong, L., Wismer, W., Temelli, F., Han, J., Huang, W., Eckhart, E., Tian, Z., Shi, K., Sun, T., Chen, L., 2017. Quality characteristics of angel food cake and muffin using lentil protein as egg/milk replacer. Int. J. Food Sci. Technol. 52 (7), 1604–1613. Jaswir, I., Hassan, T.H., Said, M.Z.M., 2004. Effect of Malaysian plant extracts in preventing peroxidation reactions in model food oil systems. J. Oleo Sci. 53, 525–529. Johnson, T.M., Zabik, M.E., 1981a. Response surface methodology for analysis of protein interactions in angel cake food system. J. Food Sci. 46, 1226–1230. Johnson, T.M., Zabik, M.E., 1981b. Egg albumen protein interactions in an angel food system. J. Food Sci. 46, 1231–1236. Kahn, M.N., Rooney, L.W., Dill, C.W., 1979. Baking properties of plasma protein isolate. J. Food Sci. 44, 274–276. Kelly, P.M., Oldfield, D.J., O’Kennedy, B.T., 1999. The thermostability of spray dried coffee whiteners. Int. J. Dairy Technol. 52 (3), 107–113.

References

595

Kilara, A., 1984. Standardization of methodology for evaluating whey proteins. J. Dairy Sci. 67, 2734–2744. Kilara, A., 1994. Whey protein functionality. In: Hettiarachchy, N.S., Ziegler, G.R. (Eds.), Protein Functionality in Food Systems. Marcel Dekker, New York, NY, pp. 325–355. Kinsella, J.N., 1982. Relationship between structure and functional properties of food proteins. In: Fox, P.F., Condon, J.J. (Eds.), Food Proteins. Applied Science Publishers, New York, NY, pp. 51–103. Kissell, L.T., Bean, M.M., 1978. AACC technical committee report: development of a method for angel food cake. Cereal Foods World 23 (3), 136–142. Kneifel, W., Ulbeth, F., Shaffer, E., 1992. Evaluation of coffee whitening ability of dairy products and coffee whiteners by means of reflectance colorimetry. Milchwissenschaft 47, 567–569. Knightly, W.H., 1968. The role of ingredients in the formulation of whipped toppings. Food Technol. 22 (6). 73–74, 77–78, 81, 85–86. Knightly, W.H., 1969. The role of ingredients in the formulation of coffee whiteners. Food Technol. 32 (2), 37–48. Kobayashi, H., Sasahara, R., Yoda, S., Kotake-Nara, E., 2017. Egg white hydrolysate inhibits oxidation in mayonnaise. Biosci. Biotechnol. Biochem. 81 (6), 1206–1215. Kolar, C.W., Cho, I.C., Watrous, W.L., 1979. Vegetable protein applications in yogurt, coffee creamers and whip toppings. J. Am. Oil Chem. Soc. 56, 389–391. Kosikowski, F., Jimenez-Flores, R., 1987. Low-Fat Dairy Coffee Whitener. United States Patent US4689245. Krishnan, K.R., Sharma, N., 1990. Studies on emulsion-type buffalo meat sausages incorporating skeletal and offal meat with different levels of pork fat. Meat Sci. 28 (1), 51–60. Lah, C.L., Cheryan, M., DeVor, R.E., 1980. A response surface methodology approach to the optimization of whipping properties of an ultrafiltered soy product. J. Food Sci. 45, 1720–1726. 1731. Lambrecht, M.A., Deleu, L.J., Rombouts, I., Delcour, J.A., 2018. Heat-induced network formation between proteins of different sources in model systems, wheat-based noodles and pound cakes. Food Hydrocoll. 79, 352–370. Lee, K.M., 1999. Functionality of 34% Whey Protein Concentrate (WPC) and Its Application in Selected Model Food Systems (PhD Dissertation). The Ohio State University, Columbus, OH. Lee, S.Y., Morr, C.V., Ha, E.Y.W., 1992. Structural and functional properties of caseinate and whey protein isolate as affected by temperature and pH. J. Food Sci. 57, 1210–1213. Liao, S.Y., Mangino, M.E., 1987. Characterization of the composition, physicochemical and functional properties of acid whey protein concentrates. J. Food Sci. 52, 1033–1037. Li-Chan, E.C.Y., 2004. Properties of proteins in food systems: an introduction. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 2–29. Li-Chan, E., Nakai, S., Wood, D.F., 1984. Hydrophobicity and solubility of meat proteins and their relationship to emulsifying properties. J. Food Sci. 49 (2), 345–350. Lindblom, M., 1977. Bread baking properties of yeast protein concentrates. Lebensm. Wiss. Tecnol. 10, 341–345. Liu, K., Tian, Y., Stieger, H., van der Linden, E., van de Velde, F., 2016. Evidence of ball-bearing mechanism of microparticulated whey protein as fat replacer in liquid and semi-solid multicomponent model foods. Food Hydrocoll. 52, 403–419. Loveday, S.M., Hindmarsh, J.P., Creamer, L.K., Singh, H., 2010. Physico-chemical changes in intermediate moisture protein bars made with whey protein or calcium caseinate. Food Res. Int. 43, 1321–1328. Lowe, E., Durkee, E.L., Mersen, R.L., Ijichi, K., Cimino, S.L., 1969. Egg white concentrated by reverse osmosis. Food Technol. 23 (6). 753, 757–759, 760–762. Lu, N., Zhang, L., Zhang, X., Li, J., Labuza, T.P., Zhou, P., 2016. Molecular migration in high protein intermediatemoisture foods during the early stage of storage: variations between dairy and soy proteins and effects on texture. Food Res. Int. 82, 34–43. Luyten, H., Vereijken, J., Buecking, M., 2004. Using proteins as additives in foods: an introduction. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 421–437. Malundo, T. M. M., Resurreccio´n, A. V. A., & Koehler, P. E. (1992). Sensory quality and performance of spray-dried coffee whitener from peanuts. J. Food Sci., 57, 222–225, 226. Mangino, M.E., Liao, Y.Y., Harper, N.J., Morr, C.V., Zadow, J.G., 1987. Effects of heat processing on the functionality of whey protein concentrates. J. Food Sci. 52, 1522–1524. Mangino, M.E., Hettiarachchy, N.S., Ziegler, G.R., 1994. Protein Functionality in Food Systems. Marcel Dekker, New York, NY. Marquez, E., de Barboza, M.Y., Izquierdo, P., Torres, G., 1997. Studies on the incorporation of bovine plasma in emulsion type of meat product. J. Food Sci. Technol. 34 (4), 337–339.

596

15. Model food systems and protein functionality

Marshall, W.H., Dutson, T.R., Carpenter, Z.L., Smith, G.C., 1975. A simple method for emulsion end-point determinations. J. Food Sci. 40, 896–897. Martinez, R.M., Dawson, P.L., Ball Jr., H.R., Swartzel, K.R., Winn, S.E., Giesbrecht, F.G., 1995. The effects of ultrapasteurization with and without homogenization on the chemical, physical, and functional properties of aseptically packaged liquid whole egg. Poult. Sci. 74 (4), 742–752. McKenna, R.J., Keller, D.J., Andersen, D.L., 1992. Process for Preparing an Alternative Protein Source for Coffee Whiteners and Other Products. United States Patent US5128156, US616909. McMahon, D.J., Adams, S.I., McManus, W.R., 2009. Hardening of high-protein nutrition bars and sugar/polyolprotein phase separation. J. Food Sci. 74 (6), E312–E321. Melachouris, N., Moffitt, K.R., Rasilowicz, C.E., Tonner, G.F., 1994. Powdered Coffee Whitener Containing Re-formed Casein Micelles. United States Patent US5318793, US937574. Melmychyn, P., 1973. Acetylated Protein for Coffee Whitener Formulations. United States Patent US3764711. Miao, S., Roos, Y.H., 2005. Nonenzymatic browning kinetics in low-moisture food system as affected by matrix composition and crystallization. J. Food Sci. 70 (2), 69–77. Min, D.B.S., Thomas, E.L., 1977. A study of physical properties of dairy whipped topping mixtures. J. Food Sci. 42, 221–224. Mittal, G.S., Usborne, W.R., 1985. Meat emulsion extenders. Food Technol. 39 (4), 121–130. Modler, H., Jones, J.D., 1987. Selected processes to improve the functionality of dairy ingredients. Food Technol. 41 (10), 114–117. 129. Mullen, K., Ennis, D., 1985. Fractional factorials in food product development. Food Technol. 39 (5). 90, 92, 94, 97–98, 100, 102–103. Mulvihill, D.M., Fox, P.F., 1987. Assessment of the functional properties of milk protein products. Bull. Int. Dairy Fed. 209, 3–11. Nakai, S., 1983. Structure-function relationships of food proteins: with an emphasis on the importance of protein hydrophobicity. J. Agric. Food Chem. 31, 876–883. Nicolai, T., Durand, D., 2013. Controlled food protein aggregation for new functionality. Curr. Opin. Colloid Interface Sci. 18, 249–256. Nongonierma, A.B., Springett, M., Le Quere, J.-L., Cavot, P., Voilley, A., 2006. Flavor release at gastric/matrix interfaces of stirred yogurt models. Int. Dairy J. 16, 102–110. Owusu-Apenten, R.K., 2004. Testing protein functionality. In: Yada, R.Y. (Ed.), Proteins in Food Processing. Woodhead Publishing, Cambridge, UK, pp. 217–244. Paredes, M.D.C., Rao, M.A., Bourne, M.C., 1988. Rheological characterization of salad dressings. 1. Steady shear, thixotropy and effect of temperature. J. Texture Stud. 19, 247–258. Parker, A., Gunning, P.A., Ng, K., Robins, M.M., 1995. How does xanthan stabilize salad dressing. Food Hydrocoll. 9, 333–342. Patel, P.D., Fry, J.C., 1987. The search for standardized methods for assessing protein functionality. In: Hudson, B.J.F. (Ed.), Developments in Food Proteins. In: vol. 5. Elsevier Applied Science, London, UK, pp. 299–333. Patel, J.R., Dave, R.I., Joshi, N.S., Thakar, P.N., 1992. Coffee whiteners. J. Dairy Sci. Assoc. 44, 18–25. Patil, S.K., Baczynsk, I.M., McCurry, T., 2006. Wheat protein isolates—alternative to sodium caseinate. Cereal Foods World 51 (5), 279–281. Pearce, R.J., Harper, W.J., 1982. A method for the quantitative evaluation of emulsion stability. J. Food Sci. 47, 680–681. Peltonen-Shalaby, R., Mangino, M.E., 1986. Compositional factors that affect the emulsifying and foaming properties of whey protein concentrate. J. Food Sci. 51, 91–95. Pernell, C.W., Luck, P.J., Foegeding, E.A., Daubert, C.R., 2002. Heat-induced changes in angel food cakes containing egg-white protein or whey protein isolate. J. Food Sci. 67, 2945–2951. Pomeranz, Y., Bolling, H., Zwingelberg, H., 1984. Wheat hardness and baking properties of wheat flours. J. Cereal Sci. 2 (3), 137–143. Potes, N., Kerry, J.P., Roos, Y.H., 2014. Protein modifications in high protein-oil and protein-oil-sugar systems at low water activity. Food Biophys. 9, 49–60. Pour-El, A., 1981. Protein functionality: classification, definition, and methodology. In: Cherry, J.P. (Ed.), Protein Functionality in Foods. American Chemical Society, Washington, DC, pp. 1–19. Purwanti, N., van der Goot, A.J., Boom, R., Vereijken, J., 2010. New directions towards structure formation and stability of protein-rich foods from globular proteins. Trends Food Sci. Technol. 21, 89–94.

References

597

Pyler, E.J., 1988. Cake baking technology. In: Bakery Science and Technology, third ed. vol. II. Sosland Publishing Co, Kansas City, MO. Chapter 21. Raeker, M.O., Johnson, L.A., 1995. Cake-baking (high-ratio white layer) properties of egg white bovine blood plasma, and their protein fractions. Cereal Chem. 72, 299–303. Randall, C.J., Raymond, D.P., Voisey, P.W., 1976. Effects of various animal and vegetable materials on replacing the beef component in a meat emulsion system. Can. Inst. Food Sci. Technol. J. 9 (4), 216–221. Ranhortra, G.S., Gelroth, J.A., Eisenbraun, G.J., 1992. Gluten index and bread-making quality of commercial dry glutens. Cereal Foods World 37 (3), 261–263. Rao, C., Fisher, M.C., Guo, M., Labuza, T.P., 2013. Storage stability of a commercial hen yolk egg powder in dry and intermediate-moisture food matrices. J. Agric. Food Chem. 61 (36), 8676–8686. Regenstein, J.M., 1989. Are comminuted meat products emulsions or a gel matrix? In: Proceedings of Food Processing and Preservation Meeting. Cornell University. Regenstein, J.M., Grunden, L.P., Baker, P.C., 1978. Proteolytic enzymes and the functionality of chicken egg albumin. J. Food Sci. 43, 279–280. Rodriguez, O., Castell-Perez, M.E., Moreira, R.G., 2007. Effect of sugar and storage temperature on the survival and recovery of irradiated Escherichia coli K-12 MG 1655. Food Sci. Technol. 40 (4), 690–696. Sakanaka, S., Tachibana, Y., 2006. Active oxygen scavenging activity of egg-yolk protein hydrolysates and their effects on lipid oxidation in beef and tuna homogenates. Food Chem. 95, 243–249. Seuvre, A.-M., Philippe, E., Rochard, S., Voilley, A., 2007. Kinetic study of the release of aroma compounds in different model food systems. Food Res. Int. 40 (4), 480–492. Severini, C., Baiano, A., Rovere, P., Dall’Aglio, G., 2003. Effect of high pressure on olive oil oxidation and the Maillard reaction in model and food systems. Italian Food Bever. Technol. 31, 5–10. Shurtleff, W., Aoyagi, A., 1994. Non-Dairy Whip Toppings (With or Without Soy Protein)—Bibliography and Source Book: 1944 to 1994. Soy Foods Centre, Lafayette, CA. Sila, D.N., Smout, C., Satara, Y., Vu-Truong, L., Hendricloc, M., 2007. Combined thermal and high pressure effect on carrot pectin methylesterase stability and catalytic activity. J. Food Eng. 78 (3), 755–764. Singh, H., 2011. Aspects of milk-protein stabilized emulsions. Food Hydrocoll. 24, 1938–1944. Smith, R. A. (1977). Emulsion Formation and Stability in Salad Dressings. (M.S. Thesis). The Ohio State University, Columbus, OH. Stanley, D.W., Goff, H.D., Smith, A.K., 1996. Texture-structure relationships in foamed dairy emulsions. Food Res. Int. 29 (1), 1–13. Swift, C.E., Sulzbacher, W.L., 1963. Comminuted meat emulsions: factors affecting meat proteins as emulsion stabilizers. Food Technol. 17, 224–226. Swift, C.E., Lockett, C., Fryar, A.S., 1961. Comminuted meat emulsions. The capacity of meats for emulsifying fat. Food Technol. 15 (11), 468–473. Szerman, N., Gonzalez, C.B., Sancho, A.M., Grigioni, G., Carduza, F., Vaudagna, S.R., 2007. Effect of whey protein concentrate and sodium chloride addition plus tumbling procedures on technological parameters, physical properties and visual appearance of sous vide cooked beef. Meat Sci. 76, 463–473. Townsend, W.E., Witnauer, L.P., Riloff, J.A., Swift, C.E., 1968. Comminuted meat emulsions: differential thermal analysis of fat transitions. Food Technol. 22, 319–323. Tran, K.H., Einerson, M.A., 1987. A rapid method for the evaluation of emulsion stability of non-dairy creamers. J. Food Sci. 52, 1109–1110. Tsai, R., Cassens, R.G., Briskey, E.J., 1972. The emulsifying properties of purified muscle proteins. J. Food Sci. 37, 286–288. Turgeon, S.L., Schmitt, C., Sanchez, C., 2007. Protein–polysaccharide complexes and coacervates. Curr. Opin. Colloid Interface Sci. 12 (4–5), 166–178. Venir, E., Marchesini, G., Biasutti, M., Innocente, N., 2010. Dynamic high pressure-induced gelation in milk protein model systems. J. Dairy Sci. 93, 483–494. Wagoner, T.B., Ward, L., Foegeding, E.A., 2015. Using state diagrams for predicting colloidal stability of whey protein beverages. J. Agric. Food Chem. 63, 4335–4344. Wearly, D., Johns, P., Dewille, N., Blumbert, R., Baxter, J., 2016. Nutritional Compositions Containing Rice Protein Together With Pea and/or Potato Proteins. Abbott Laboratories. US 2016/0309755 A1. Webb, N.B., 1974. Emulsion technology. In: Proceedings of the Meat Industry Research Conference, pp. 1–15.

598

15. Model food systems and protein functionality

Wegrzyn, T.F., Farr, J.M., Hunter, D.C., Au, J., Wohlers, M.W., Skinner, M.A., Stanley, R.A., Sun-Waterhouse, D., 2008. Stability of antioxidants in an apple polyphenol-milk model system. Food Chem. 109, 310–318. Wijeratne, S.S.K., Amarowicz, R., Shahidi, F., 2006. Antioxidant activity of almonds and their by-products in food model systems. J. Am. Oil Chem. Soc. 83, 223–230. Wilderjans, E., Luyts, A., Goesaert, H., Brijs, S.K., Delcour, J.A., 2010. A model approach to starch and protein functionality in a pound cake system. Food Chem. 120, 44–51. Xiong, Y.L., Blanchard, S.P., Means, W.J., 1992. Properties of broiler myofibril gels containing emulsified lipids. Poult. Sci. 71 (9), 1548–1555. Yada, R.Y., 2004. Proteins in Food Processing. Woodhead Publishing, Cambridge, UK. Yang, X., Foegeding, E.A., 2010. Effects of sucrose on egg white protein and whey protein isolate foams: determining properties of wet and dry foams. Food Hydrocoll. 24, 227–238. Ye, A., Taylor, S.M., 2008a. Dairy Product and Process. Assignee Fonterra Co-operative Group Ltd. PCT Patent Application WO2008130253 A1. Ye, A., Taylor, S.M., 2008b. Dairy Product and Process. Assignee Fonterra Co-operative Group Ltd. PCT Patent Application WO2008130252 A1. Yolmeh, Y., Jafari, S.M., 2017. Applications of response surface methodology in the food industry processes. Food Bioprocess Technol. 10 (3), 413–433. Zayas, J.F., 1996. Functionality of Proteins in Food. Springer-Verlag, Berlin, Germany. Zhen, M., Boye, J.I., Simpson, B.K., 2016. Preparation of salad dressing emulsions using lentil, chickpea and pea protein isolates: a response surface methodology. J. Food Qual. 39 (4), 274–291.

C H A P T E R

16 Milk protein gels John A. Lucey Wisconsin Center for Dairy Research, University of Wisconsin-Madison, Madison, WI, United States

Introduction The gelation of the proteins in milk is the basis for the manufacture of cheese and fermented milk products. Various different approaches, including heat (whey proteins), use of rennet enzyme (caseins), and acidification (caseins and denatured whey proteins), can be used to destabilize the milk proteins. Combinations of these approaches can also be used to form dairy products, for example, the use of a low concentration of rennet in cottage cheese (or quarg), which is primarily a cultured product. Yogurt is a cultured product in which caseins and denatured whey proteins are responsible for the gelation properties. Milk protein gels are irreversible, in contrast to many polysaccharide gels, which are thermoreversible. Milk protein gels are often classified as particle gels, although it is now recognized that they are not simple particle gels, as the internal structure of the casein particle plays an important role in their rheological properties (Horne, 2001, 2003). The properties of milk protein gels have been reviewed (Green, 1980; de Kruif et al., 1995; Lucey, 2002; van Vliet et al., 2004). The casein micelles are the building blocks for rennet and yogurt gels. Casein micelles are assembled because of the concerted action of two major types of interactions, namely, hydrophobic interactions and the formation of calcium phosphate nanoclusters across phosphoserine clusters (Lucey and Horne, 2018). The casein particles (formed from the aggregation of many individual casein micelles) in rennet gels undergo rearrangement, fusion, and syneresis in the process of forming cheese curd; thus, they are inherently dynamic in nature, and the rearrangement processes involved have been studied (e.g., see the review by Dejmek and Walstra, 2004). Recent processing strategies for increasing the cross-linking within milk protein gels include the use of oxidase enzymes and transglutaminases (Isaschar-Ovdat and Fishman, 2018). The gelation methods and processing methods such as heat treatment, impact digestion, and the rate of release of amino acids in milk gels (Barb
e et al., 2013). The microstructure of a milk gel such as yogurt can be designed by altering the processing conditions (Loveday et al., 2013) or by adding a low concentration of a polysaccharide that can induce (controlled) phase separation (Rohart and Michon, 2014).

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00016-5

599

# 2020 Elsevier Inc. All rights reserved.

600

16. Milk protein gels

Rennet-induced gels Introduction The coagulation of milk by rennet probably occurred initially by accident, as warm milk was stored in sacks, which were made from the stomachs of ruminant animals that contained some residual proteinase enzymes. Crude extracts, prepared from the fourth stomach of young calves (called rennets, which are a type of aspartic proteinase), have been used for cheesemaking for thousands of years. Pepsin is the predominant proteinase in adult mammals. Naturally produced calf chymosin (EC 3.4.23.4) may contain up to six molecular species, which have slight differences in their amino acid residues (Crabbe, 2004). Chymosin has been cloned into several genetically modified organisms to produce fermentation-derived chymosin, which is widely used in many countries around the world (Crabbe, 2004). Kappeler et al. (2006) expressed the gene for camel (Camelus dromedarius) chymosin in Aspergillus niger and produced camel chymosin by fermentation. As camel rennet is less proteolytic during cheese ripening than calf rennet, it has found applications in extending the shelf life of cheese and reducing bitterness. The rennet coagulation of milk has been extensively reviewed (Dalgleish, 1987, 1993; Hyslop, 2003; Horne and Banks, 2004; Horne and Lucey, 2017).

Primary phase of rennet coagulation The basic building blocks of rennet-induced gels are the casein micelles. Both αs-casein (αs1- and αs2-) and β-casein are sensitive to precipitation by the Ca2+ in milk and are protected by association with κ-casein, which is one reason for the formation of micelles. κ-Casein molecules have a predominantly surface position on micelles, where the hydrophilic C-terminal apparently acts as a “hairy” layer, providing steric stabilization and a barrier against association with other micelles (Walstra, 1990). The two stages of the rennet coagulation of milk are shown in Fig. 16.1. In the primary phase of rennet coagulation, the C-terminal part (residues 106–169) of the κ-casein molecule is hydrolyzed, and this hydrophilic peptide diffuses away from the micelle (called paracasein) into the serum phase. This macropeptide is called caseinomacropeptide (CMP) or, if it is highly glycosylated, glycomacropeptide (GMP). Most microbial coagulants, including those derived from Rhizomucor miehei, hydrolyze the same Phe105-Met106 bond as chymosin; however, Cryphonectria parasitica hydrolyzes the Ser104-Phe105 bond (Drøhse and Foltmann, 1989). The proteolysis by chymosin of other proteins in milk occurs at a much slower rate (Crabbe, 2004). The enzymatic reaction in milk appears to obey first-order kinetics. The proteolysis of κ-casein is usually described by standard Michaelis-Menten kinetics, although Hyslop (2003) questioned whether this was truly appropriate. It should be noted that the primary phase and the secondary phase of clotting overlap as the aggregation begins before the enzymatic reaction is complete.

Rennet-induced gels

601

Enzymatic stage, dependent on: [rennet], temperature, [casein], pH

Micelle

Para-casein + macropeptide 100% k-casein hydrolysis

0%

Aggregation stage, dependent on: [Ca2+ ], temperature, [casein], pH, [denatured whey protein]

Gelation at lower degree of hydrolysis with low pH or high protein level

Rennet-induced gel FIG. 16.1 The two stages of the rennet coagulation of milk.

Secondary phase of rennet coagulation The stability of the casein micelles in milk is attributed to their net negative charge and to steric repulsion by the flexible macropeptide region of κ-casein (the so-called hairs that extend out into the solution), calcium-induced interactions between protein molecules, hydrogen bonding, and electrostatic and hydrophobic interactions. The release of the CMP (or GMP), which diffuses away from the micelles, leads to a decrease in the zeta potential, by
5–7 mV (
50%), which reduces the electrostatic repulsion between rennet-altered micelles. Removal of the “hairs” results in a decrease in the hydrodynamic diameter by
5 nm and a loss of steric stabilization and causes a slight minimum in the viscosity during the initial lag phase of renneting. Various attempts have been made to model the aggregation reaction (see the review by Horne and Banks, 2004). The nature of the attractive forces during the aggregation of casein micelles is still not completely clear, although calcium bridges, van der Waals forces, and hydrophobic interactions are probably involved. Destabilized micelles will aggregate only in the presence of free Ca2+. Rennet acts on casein at temperatures as low as 0°C, but milk does not clot at temperatures below 15°C, whereas aggregation is very rapid at high temperature (e.g., 55°C). When milk is clotted under normal conditions of pH and protein content, the viscosity does not increase until the enzymatic phase is mostly complete, that is, at >60% of the (visual) rennet coagulation time. Coagulation does not occur until the enzymatic phase is at least
87% complete. Sandra et al. (2007) studied the rennet gelation process using diffusing wave spectroscopy, which allowed gelation to be monitored without the need for dilution. They suggested that partially renneted casein micelles do not begin to approach one another until the extent of breakdown of the κ-casein hairs has reached about 70%; above this point, they interact increasingly strongly with an increase in the extent of proteolysis. This interaction initially restricts the diffusive motion of the particles rather than causing true aggregation. Only after more extensive removal of the protective κ-casein hairs does true aggregation

602

16. Milk protein gels

occur, with the appearance of a space-filling gel (as defined by rheology terms, such as having a loss tangent value 80% β-lactoglobulin denaturation). As a result, β-lactoglobulin becomes mostly attached to the κ-casein of the casein micelles or forms soluble complexes (with serum casein), depending on the heating conditions (i.e., pH) (Lucey et al., 1998b).

Acid-induced milk gels

613

Denatured whey proteins (DWP) that are attached to the surface of casein micelles during heating (i.e., bound DWP) are a critical factor in the increased stiffness of yogurt gels made from heated milk. DWP cause micelles to aggregate at higher pH because of the higher isoelectric pH (
5.3) of the main whey protein, β-lactoglobulin, than that of caseins (Lucey et al., 1997; Guyomarc’h et al., 2003; Morand et al., 2012). An alternative view is that the DWP associated with the micelles alter the hydrophobic interactions between heated micelles, which facilitate gelation at higher pH values (although there is greater electrostatic repulsion at higher pH) ( Jean et al., 2006). More cross-linking of gels by bound DWP increases the gel strength. Soluble DWP are not able to increase the gel stiffness of milk in which there are no bound DWP present, that is, the micelle surface does not contain any “bound” DWP, which can be created experimentally (Lucey et al., 1998b). In industrial practice, heating milk always creates some bound DWP, which allows soluble DWP to become attached to the micelles and to contribute to the gel strength. The pH at heating influences the association of DWP with the casein micelles. At pH 6.5, most DWP are associated with micelles (e.g., >70% for milk heated at 90°C for 30 min). At higher pH (e.g., 7.0), fewer DWP are associated with micelles as more κ-casein dissociates from the micelles to interact with β-lactoglobulin during heating. The gel strength of acid gels made from milk heated at high pH is greater than that of acid gels made from milk heated at low pH (Lucey et al., 1998b; Anema et al., 2004). Milk heated at pH 6.2 produced yogurt gels that had much lower storage modulus values than those heated at pH 6.7 or 7.2 (Ozcan et al., 2015) (Fig. 16.3). A similar trend was observed in milks fortified with whey protein (Mahomud et al., 2016). At high pH values, there is an increase in the concentration of CCP (additional crosslinking) in milk (McCann and Pyne, 1960), which could potentially increase the stiffness of acid gels made from milk of high pH. Increasing the pH of heat treatment of the milk to 7.0 should also alter protein unfolding and disulfide bond formation, involving β-lactoglobulin, as the pK value of its free thiol group is 9.35 (Kella and Kinsella, 1988a). The creation of additional covalent disulfide bonds that involve whey protein and caseins should increase the strength of the yogurt gel. Morand et al. (2011) indicated that increasing intermolecular FIG. 16.3 Storage modulus at pH 4.6 for yogurt gels made from milk heated at pH 6.2, 6.7, or 7.2. Milk samples were heated at 85°C for 30 min prior to yogurt fermentation at 40°C. Adapted from Ozcan, T., Horne, D.S., Lucey, J.A., 2015. Yogurt made from milk heated at different pH values. J. Dairy Sci. 98, 6749–6758.

614

16. Milk protein gels

disulfide bonding between the proteins in milk, by increasing the number of free thiol groups in the reactant proteins, improves the firmness of acid gels. Various studies have shown some conflicting results about the relative importance of the soluble and bound DWP fractions to the texture of acid milk gels (Lucey et al., 1998b; Guyomarc’h et al., 2003; Anema et al., 2004). Differences in the proportions of soluble and bound DWP fractions in these studies could have contributed to these conflicting results. Guyomarc’h et al. (2003) had only a small proportion (10%–15%) of β-lactoglobulin in the bound DWP fraction, whereas Lucey et al. (1998b) had around 80% β-lactoglobulin in the bound DWP fraction. Guyomarc’h et al. (2003) suggested that differences in the quantitative amounts of aggregates (and the total amount of DWP) present in the systems, independently of whether they were soluble or not, could be the reason for some of the conflicting results reported by the different groups. Irrespective of the pH of the milk at heating, DWP (i.e., those designated as “soluble” and “bound” at the pH of heat treatment) are insoluble at low pH and should associate with casein at the pH values involved in yogurt fermentation. As the pH decreases during fermentation, virtually all the residual soluble complexes become attached to caseins via the bound DWP. The rate of acidification and the gelation temperature may also influence how these complexes associate with the caseins during acidification. The extent of denaturation of the whey proteins is often determined by their loss of solubility at pH 4.6 (de Wit, 1981), so that all the DWP should precipitate with the caseins as the pH approaches pH 4.6. The addition of WPC to milk that was then given a high heat treatment resulted in an increase in the pH of gelation, an increase in gel stiffness, and a reduction in the fracture strain compared with gels made from heated milk without added WPC (Lucey et al., 1999). When WPC was added to heated milk and this mixture was not given any further heat treatment, the acid gels formed after acidification were weaker than those made from heated milk without WPC. This suggests that any added whey proteins must be denatured to reinforce the network, even when DWP are already present in the milk. Schorsch et al. (2001) examined the effect of heating whey proteins in the presence or absence of casein micelles on the subsequent acid gelation properties of milk. The acid-induced gelation occurred at a higher pH (around pH 6.0) and in a shorter time when the whey proteins (concentration of 1-g whey protein/kg) were denatured separately from the casein micelles than when the whey proteins were heated in the presence of the casein micelles. However, the gels formed were very weak, probably because of the formation of a weak network in which whey proteins entrapped caseins. In gels made from heated milk, because of the high gelation pH, the gel goes through a period of solubilization of the CCP that is present within the casein particles that are already part of the gel network (this event is responsible for the maximum in the loss tangent during gelation) (Lucey et al., 1997). This process loosens the interactions between caseins in the gel network, and the higher bond mobility in yogurt gels during this period has been associated with whey separation (Lucey, 2001). The rheological changes during the acid-induced gelation (with GDL) of unheated and heated milks at 30°C are shown in Fig. 16.4. The much shorter gelation time, the large increase in the storage modulus, and the maximum in the loss tangent (as indicated by the hatched region between the two arrows, region A) in the heated milk sample should be noted. As the low gelation pH (4.8) of the unheated milk gel occurs after most or all of the CCP is already solubilized, there is no maximum in the loss tangent

615

Acid-induced milk gels

Gelation pH = 4.8

0.6

Gelation pH = 5.2

0.5

10 0.4

1

A

Loss tangent

Storage modulus, G¢ (Pa)

100

0.3

0.1

0.2 0

10

20

30

40

50

Time (ks) FIG. 16.4 Storage modulus (solid lines) and loss tangent (dashed lines) of acid gels made from heated milk (●) and unheated milk (). Heat treatment was at 80°C for 30 min, and acidification was at 30°C with 1.3% GDL. The area marked by the letter A indicates the region in which the loss tangent increases after gelation because of solubilization of CCP in casein particles that are already part of the gel network. Reproduced with the permission of Cambridge University Press, from Lucey, J.A., Tamehana, M., Singh, H., Munro, P.A., 1998b. Effect of interactions between denatured whey proteins and casein micelles on the formation and rheological properties of acid skim milk gels. J. Dairy Res. 65, 555–567.

in this type of gel. When acid-induced gelation of heated milk occurs rapidly at high temperature, a plateau in the storage modulus, which corresponds to the region where there is a maximum in the loss tangent, can be observed (Horne, 2001). Bikker et al. (2000) reported that the addition of β-lactoglobulin variant B or variant C to the milk prior to heating and acidification caused a larger increase in the storage modulus of acid gels than the addition of β-lactoglobulin variant A. Soluble whey protein polymers have been used as ingredients for yogurt applications (Britten and Giroux, 2001). The use of whey protein polymers to standardize the protein content of milk increased the viscosity of the yogurt to about twice that obtained using skim milk powder at the same protein concentration. The water-holding capacity of yogurt standardized with whey protein polymers was considerably higher than that of yogurt standardized with skim milk powder (Britten and Giroux, 2001). Incubation temperature Although 42°C is a commonly used fermentation temperature for yogurt, the use of slightly lower incubation temperatures (e.g., 40°C) leads to slightly longer gelation times but to the formation of firmer and more viscous gels that are less prone to whey syneresis (Lee and Lucey, 2004). At a lower incubation temperature, there is an increase in the size of the casein particles because of a reduction in hydrophobic interactions, which, in turn, leads to an increased contact area between the casein particles (Lee and Lucey, 2004); a similar

616

16. Milk protein gels

trend occurs when the gels are cooled. A high incubation temperature also makes the gel network more prone to rearrangements (more flexible) during gelation, and these changes can lead to greater whey separation (Lucey, 2001; Mellema et al., 2002). Peng et al. (2010) investigated the effect of altering the temperature immediately after gel formation. Cooling after gelation resulted in an increase in gel stiffness and greater intercluster strand formation, whereas heating of gels may promote intracluster fusion and the breakage of strands between clusters. Production of EPS Some yogurt starter cultures produce EPS during the fermentation process. It can be viewed as a naturally produced thickener. This EPS can be produced as a capsular layer around the bacterial cell or can be excreted into the medium to produce an effect that is sometimes called “ropy” or “stringy” (Hassan, 2008); popular examples of ropy yogurt are Viili and La˚ngfil from Scandinavia. Capsular EPS has little impact on the gelation or texture of yogurt. Ropy EPS can be either charged or uncharged. It is possible that charged EPS may associate electrostatically with the caseins, depending on the pH of the milk, whereas uncharged EPS may influence gelation via a depletion flocculation-type mechanism (Girard and Schaffer-Lequart, 2007; Pachekrepapol et al., 2015). The molar mass, the chain flexibility, the concentration of EPS, and the exact period during fermentation (before, during, or after gelation) when EPS is produced may all play a critical role in determining the impact of EPS on yogurt gels. The surface properties of bacteria can influence the texture of fermented milk gels (Tarazanova et al., 2018). Casein conjugates Proteins can be conjugated (glycated) with reducing sugars via the Maillard reaction, which modifies the functional properties of the protein (O’Mahony et al., 2017). Conjugation of sodium caseinate with maltodextrin resulted in weaker acid gels and slightly lower gelation pH values (Zhang et al., 2017). One likely explanation for this impact is that the covalent attachment of polysaccharides to the casein would inhibit the particle fusion and protein rearrangements needed to increase the gel strength (Zhang et al., 2017). Glycation of casein with lactose resulted in gels with greater firmness, which might have been due to protein cross-linking caused by the formation of advanced Maillard products (Hannß et al., 2018). With the use of polysaccharides (compared with simple sugars) and the “wet” heating process employed by Zhang et al. (2017), less advanced Maillard products were probably produced.

Mixed gels made with rennet and acid Milk coagulation can be induced by the combined action of acid and enzyme (i.e., mixed gels). Mixed milk coagulation has received very little attention when compared with rennetor acid-induced coagulation (Roefs et al., 1990; Lucey et al., 2000, 2001; Tranchant et al., 2001; Liu et al., 2014). Cottage cheese is generally manufactured by acid coagulation of pasteurized skim milk, and a small concentration of rennet is sometimes added after the starter has been

Whey protein gels

617

allowed to develop some acidity (i.e., at pH around 5.5) (Castillo et al., 2006). The use of rennet in combination with acid development initiates gelation at a high pH, and the gel can undergo a “weakening” stage (as indicated by a decrease/plateau of the storage modulus, a decrease in the light backscatter ratio, or an increase in the loss tangent). This weakening is more pronounced with unheated milk gels and when there has been a very high degree of κ-casein hydrolysis prior to acidification (Li and Dalgleish, 2006). This “weakening” stage is related to rearrangements caused by CCP demineralization of the casein particles in the gel network because this CCP solubilization occurs after gelation (gelation is initiated at a high pH in mixed gels) (Lucey et al., 2000). The final storage modulus of mixed gels can be considerably higher than that of acid gels made without rennet. Mixed gels made from heated milk formed firmer gels, as they were cross-linked by denatured whey proteins and underwent fewer large-scale rearrangements (Lucey et al., 2000). The rheological and microstructural properties of mixed gels are complex, and these properties can be adjusted by varying the rennet level or the acidification rate (Tranchant et al., 2001). The use of low rennet levels during the fermentation of milk resulted in a coarser acid gel network and higher syneresis (Aichinger et al., 2003). Micelle fusion was faster in gels with added rennet because of the removal of the κ-casein hairs (Aichinger et al., 2003). Gastaldi et al. (2003) studied the acid-induced gelation of milk samples in which chymosin was used to vary the degree of κ-casein hydrolysis prior to acidification (further chymosin activity during acidification was blocked using an inhibitor). The gelation pH increased, and the gelation time decreased with an increasing degree of κ-casein hydrolysis. Gels with much higher storage moduli were formed as a result of partial κ-casein hydrolysis prior to gelation, although the loss tangent and the serum-holding capacity were lower (Gastaldi et al., 2003). Presumably, partial κ-casein hydrolysis prior to acid gelation facilitated greater rearrangement/fusion of casein, which not only was responsible for the increase in the storage modulus but also increased the serum separation (Lucey et al., 2001).

Whey protein gels Whey is usually obtained as a by-product of cheesemaking (although recent developments in membrane technology mean that, in future, “whey” will come not necessarily from a cheese vat but as “native” whey directly from milk prior to cheesemaking) (Coppola et al., 2014). The composition of whey depends on the cheesemaking conditions; for example, acid whey derived from cottage cheese has different mineral (ash), lactic acid, and pH values from whey derived from rennet-coagulated cheeses such as Cheddar (Table 16.1). Whey products are widely used as food ingredients because of their excellent functional and nutritional properties. Various types of whey products are made commercially, ranging from dried whey to WPC (WPC has protein contents ranging from
35% to 80%) to whey protein isolate (WPI) (protein content 90%) (Table 16.2). Membrane filtration, that is, ultrafiltration (UF) and diafiltration (DF), is used to concentrate the protein fraction before spray drying into WPC. Two different approaches are used to produce WPI: (a) membrane filtration (microfiltration, UF, and DF) and (b) ion-exchange chromatography coupled with UF/DF. These two approaches result in WPIs with different protein profiles (Table 16.3; Wang and

618

16. Milk protein gels

TABLE 16.1

Composition of rennet and acid wheys Average composition Rennet whey

Acid whey

Total whey protein (g/L)

6.7

5.8

Glycomacropeptide (g/L)

1.0

–

Lactose (g/100 g)

5.0

4.4

Ash (g/100 g)

0.5

0.6

Na (mg/100 g)

35

40

K (mg/100 g)

109

133

Ca (mg/100 g)

22

86

Mg (mg/100 g)

6

9

P (mg/100 g)

42

63

Adapted from Oakenfull, D., Pearce, J., Burley, R.W., 1997. Protein gelation. In: Damodaran, S., Paraf, A. (Eds.), Food Proteins and Their Applications. Marcel Dekker, New York, NY, pp. 111–142.

TABLE 16.2 Typical composition of some whey powders (approximate, wet, or as-is basis) Whey ingredient

Moisture (%)

Fat (%)

Protein (%)

Lactose (%)

Ash (%)

Sweet whey

3–5

1.1–1.5

11–14.5

75

8–10

Acid whey

3.5

0.5–1.5

11–13.5

70

10–12

WPC35

3–4.5

3–4.5

34–36

48–52

6.5–8

WPC80

3.5–4.5

6–8

80–82

4–8

3–4

WPI

4–5

500-mL bolus) was also associated with a trend toward decreased appetite and desire to eat and significantly increased satiety. Despite recent publication in 2017, it is not clear why the authors omitted

Milk proteins, obesity, and weight control

655

multiple trials conducted from 2014 onward (including Poppitt et al., 2011, 2013; Chungchunlam et al., 2014, 2016, 2017; Wiessing et al., 2015), and further meta-analysis may be warranted. To date, there have been few long-term studies investigating the effect of milk proteins on weight loss, in the absence of other lifestyle interventions. A 6-month study of overweight and obese individuals reported that 56 g WPC/day resulted in significant loss of body weight and fat mass and waist circumference compared with a CHO control (Baer et al., 2011), whereas a shorter 12-week study found no effect of 54 g WPI/day (Pal et al., 2010a). Interestingly, fasting blood lipid and insulin levels improved in this study, supporting the hypothesis that whey protein may improve metabolic health even in the absence of weight loss (Pal et al., 2010a). Energy-restricted diets are a common way to successfully decrease body weight, at least in the short term, but characteristic of this is loss of both fat mass and lean mass. Loss of lean mass results in a concomitant reduction in basal metabolic requirements. A return to habitual dietary habits may then result in rapid weight regain. Traditional energy-restricted diets for weight loss in middle-aged or older individuals with the age-related sarcopenia and obesity known as “sarcobesity” (Parr et al., 2013) may be even more susceptible to rapid weight regain because of already low muscle mass. The ingestion of whey protein as part of an energy-restricted diet has been proposed as a strategy for decreasing fat mass while preserving lean mass. There are some data to support this. In a study of obese individuals undertaking a severely energy-restricted diet of 2 MJ/day for 12 weeks, supplementation twice daily with a milk protein successfully led to greater weight loss, greater fat loss, and yet maintenance of lean muscle mass (Frestedt et al., 2008). Another similar study, supplementing with a high-protein meal replacement comprising whey protein, soy protein, and free AAs, led to greater loss of body fat (Treyzon et al., 2008). Also, in a study of diet- and exercise-induced weight loss, a higher-protein/higher-dairy product (30 energy%, 15 energy%) intervention promoted greater total and visceral fat loss and lean mass gain as assessed by dual-energy X-ray absorptiometry and magnetic resonance imaging in a group of obese premenopausal women ( Josse et al., 2011). A high whey protein, high leucine, and high vitamin D supplement was also shown to preserve muscle mass during intentional weight loss in a group of obese older adults (Verreijen et al., 2015). Older adults represent a special category. A study of 60 older obese adults undergoing 3 months of intentional weight loss, notably also accompanied by resistance exercise, led to recommendations of 1.2 g/kg body weight (or 1.9 g/kg fat-free mass) as the optimal daily protein intake in this group when undertaking periods of intentional weight loss, to optimize preservation of muscle mass (Weijs and Wolfe, 2016). Conversely, recent data from a small study of 29 obese adults aged 18–55 years failed to show a significant effect of whey protein supplementation during a 4-week low-energy diet LED, when accompanied by walking exercise. The authors reported no improvement in the quality of weight loss with respect to changes in body composition when compared with nonsupplemented controls (Larsen et al., 2018). Milk proteins have also been hypothesized as beneficial for maintaining a lower body weight after energy restriction and fat loss. In an early study of whey protein and casein supplementation, there was significantly better weight loss maintenance over a 12-week period compared with a CHO control (Claessens et al., 2009a,b). However, more recent data have failed to confirm this effect, with a well-conducted trial of 220 overweight and obese adults in Copenhagen showing that whey protein supplementation did not result in improved

656

18. Milk proteins and human health

postweight loss weight maintenance over a 6-month follow-up period compared with normal levels of dietary protein intake of 0.8–1.0 g/kg/day (Kjølbæk et al., 2017). Whether whey protein or casein supplements may be more effective for preservation of lean muscle mass during periods of weight loss induced by an energy-restricted diet has also been addressed (Adechian et al., 2012). Obese individuals underwent a 6-week energy-restricted diet with whey protein or casein supplementation. Although all individuals showed weight loss after 6 weeks, there were no differential effects of whey protein or casein for weight loss, fat loss, or preservation of lean muscle mass. Intriguingly, an assessment of whole-body protein synthesis and whole-body protein breakdown showed casein to cause greater inhibition of protein breakdown, whereas whey protein increased protein synthesis (Adechian et al., 2012). Preservation of postprandial myofibrillar protein synthesis by whey protein supplementation during short-term energy restriction in obese adults has more recently been confirmed (Hector et al., 2015). The implications are that casein supplementation may be optimal for preservation of skeletal muscle mass during energy restriction but needs to be verified. Long-term trials comparing milk proteins with other protein sources for long-term weight loss and maintenance are also required.

Milk proteins, muscle wasting, and sarcopenia Advancing age and a sedentary lifestyle are associated with a gradual decline in skeletal muscle mass, function, and strength, which in the extreme form is termed sarcopenia or muscle wasting. Loss of skeletal muscle mass or function of the muscle has major implications for quality of life because activities of daily living such as walking upstairs become difficult or are no longer possible. At the extreme end, patients with chronic or end-stage diseases including cancer, heart failure, AIDS, and chronic obstructive lung disease and sepsis are also often susceptible to muscle wasting (Tan and Fearon, 2008; Fearon et al., 2011). In addition to these mobility issues, skeletal muscle also has a significant impact on metabolic health. As skeletal muscle is one of the major organs that is responsible for insulinstimulated glucose uptake, loss of skeletal muscle mass is often associated with insulin resistance (Evans, 2010). In addition to the insulin stimulatory effect of milk protein, it is the high level of BCAAs in whey protein and casein that prevents the loss of lean body mass through an increase in skeletal MPS and/or a decrease in breakdown, as demonstrated by the Adechian group (Adechian et al., 2012). There is evidence that the anabolic effect of milk protein is decreased during aging; this has been termed “anabolic resistance” (Volpi et al., 2000). Whether this is an anabolic resistance to the dietary intake of BCAAs or simply a reflection of underutilization of the major muscle groups in older people as exercise levels decline is a matter of considerable debate. Whether milk proteins can prevent the development of anabolic resistance or overcome established anabolic resistance in older individuals is of great interest. Both circulating insulin and AAs are important for the activation of MPS. Whey protein and casein are high-quality proteins, based on both the protein digestibility corrected AA score (PDCAAS) (Boye et al., 2012) and the more recently developed digestible indispensable AA score (DIAAS) (FAO, 2013; Mathai et al., 2017), both of which take into account human AA requirements and protein digestibility. However, whey protein contains a greater amount

Milk proteins, muscle wasting, and sarcopenia

657

of the BCAAs leucine, isoleucine, and valine than does casein, and there is a body of literature showing greater MPS (Devries and Phillips, 2015). Of the BCAAs, leucine is thought to be the most potent activator of protein synthesis (Katsanos et al., 2006; van Loon, 2012), although there is evidence that high levels of nonleucine BCAAs can induce equivalent protein synthesis when given with a whey protein supplement (Churchward-Venne et al., 2012). More recently, it is notable that data from our laboratory on healthy middle-aged men also showed that the consumption of a low leucine milk protein supplement resulted in a similar increase in MPS to that of a high leucine whey protein supplement (Mitchell et al., 2015). This was despite the evidence of an increased rate of digestion and leucine availability following the ingestion of a 20 g whey protein bolus. More is also known on efficacious doses of milk proteins to promote a muscle anabolic signaling response, with another recent study by our research team demonstrating 9 g to be an effective dose following a bout of resistance exercise, again in healthy middle-aged men (Mitchell et al., 2017). Casein contains several essential amino acids (EAAs) including histidine, methionine, and phenylalanine in a greater amount than whey protein and also contains a greater amount of the non-EAAs arginine, glutamic acid, proline, serine, and tyrosine (Hall et al., 2003). EAAs have been demonstrated to be primarily responsible for the activation of MPS (Volpi et al., 2003), although they are not necessarily effective for the inhibition of protein breakdown. As noted earlier, studies have demonstrated greater whole-body protein synthesis following whey protein supplementation than following casein supplementation, whereas protein catabolism was greater following casein supplementation (Boirie et al., 1997; Adechian et al., 2012). The faster digestion rate of whey protein compared with casein, which, because of its micellar structure, tends to clot in the stomach, has been commonly believed to lead to more rapid delivery of AAs into plasma following whey protein ingestion and longer more sustained delivery following casein ingestion. Although there are data that show greater muscle protein accretion over the 6 h following whey protein supplementation than following casein or casein hydrolyzate supplementation (Pennings et al., 2011), this certainly is not a universal effect, and leucine blood kinetics may be unreliable markers of MPS (Mitchell et al., 2015) and therefore of improvements in muscle function. Conversely, a sophisticated study of intrinsically labeled whey protein and casein, coingested as milk, showed the absorption and retention of AAs from whey protein and casein to be similar after 2 h, with AAs derived from casein showing greater absorption and retention rates beyond 3 h (Soop et al., 2012). Some discrepancies in outcome may possibly have been due to differential effects of aging on the response to these two dairy proteins. It may be worthwhile to increase the typically low whey protein content of milk, which then theoretically would provide both an early (whey protein) and a sustained (casein) stimulation of MPS and an inhibition of muscle protein breakdown (Reitelseder et al., 2011). Plant-based proteins, including the ubiquitous soy protein, which also provide an alternative option as dietary components aimed at preserving or increasing skeletal muscle mass, are proposed to result in a lower muscle protein synthetic response than animal-based proteins, although interesting propositions including plant breeding and fortification with the AAs methionine, lysine, and/or leucine may represent a novel way forward (van Vliet et al., 2015). Whether milk protein-enhanced MPS is effective for the prevention of muscle wasting and the promotion of muscle protein accretion, which may result in function improvements in strength and/or mobility in healthy aging populations, elderly populations, or patients with

658

18. Milk proteins and human health

chronic end-stage disease, still needs to be determined. A recent study from our laboratory (Mitchell et al., 2018) in healthy middle-aged men who underwent 2 weeks of unilateral lower limb immobilization as a model of accelerated aging, showed that, although there was an expected rapid decline in strength, function, and muscle size through immobilization, the ingestion of a supplement of 20 g milk protein/day (1.2 g total protein/day) surprisingly did not attenuate this decline of any measures of muscle size or function. Despite observing an increase in the rate of synthesis of contractile proteins during recovery, the supplemental milk protein did not enhance the recovery of muscle size or function from disuse. This was an aggressive muscle “aging” model, in which both function and size were not fully recovered after 2 weeks of normal activity; they required a further 2 weeks of supervised resistance training (Mitchell et al., 2018). A recent similar study in a cohort of older men also failed to show a benefit of protein supplementation in protecting lean leg mass during a period of both inactivity and energy restriction; notably, however, whey protein supplementation did augment muscle mass and MPS during the recovery phase (Oikawa et al., 2018). Regular resistance-type exercise transiently activates MPS and can, over time, lead to increases in skeletal muscle mass. Milk protein supplements are popular among recreational gym users seeking to increase muscle mass, but there is also considerable interest in whether milk protein ingestion in conjunction with resistance exercise may be beneficial for individuals with sarcopenia, with growing evidence that mitigation of this muscle wasting can be achieved by the consumption of higher-quality protein in sufficient quantities, probably at higher levels than the recommended intakes of 0.8 g/kg bodyweight/day (Phillips and Martinson, 2018). After resistance exercise, whey protein and casein have been shown to cause comparable increases in both net protein balance (Tipton et al., 2004) and myofibrillar protein synthesis rate (Reitelseder et al., 2011), whereas other data show whey protein ingestion to cause a greater increase in the MPS rate during the early 3-h period (Tang et al., 2009). To date, longer-term studies have produced mixed findings. In obese postmenopausal women, WPI, in combination with an energy-restricted diet plus exercise over 6 months, led to 4% greater weight loss than in the control group and notably greater loss of subcutaneous adipose tissue and greater increase in leg muscle mass (Mojtahedi et al., 2011). Conversely, in elderly men who undertook resistance training for 12 weeks, protein supplementation did not improve muscle hypertrophy (Verdijk et al., 2009). Whereas effects on muscle anabolism and catabolism are clear, the evidence underpinning clinically significant gains in lean body mass remain equivocal (Cermak et al., 2013; Mitchell et al., 2018).

Milk proteins and heart health Atherosclerosis Atherosclerosis is a common cause of myocardial infarction, stroke, and peripheral vascular disease and represents the progressive damage to the vascular endothelium that is caused by buildup of lipids, immune cell infiltration, and plaque formation, leading to impaired endothelial function and reduced blood flow. Milk protein has been suggested to aid in the prevention of atherosclerosis (Marcone et al., 2017), potentially through the amelioration of adverse circulating lipid levels, one of the primary metabolic risk factors for CVD.

Milk proteins and heart health

659

Most studies have focused on effects on postmeal (postprandial) lipemia, based on the premise that rapid clearance of blood lipids after a meal decreases arterial exposure to these circulating lipids. Animal models of chronic whey protein supplementation in ApoE mice have shown improvements in atherosclerosis through the regulation of circulating lipid and inflammatory cytokines and an increase in the expression of the cholesterol transporters ABCA1 and ABCG1 (Zhang et al., 2018). Clinical studies of overweight postmenopausal women given a high fat meal have reported that both WPI and sodium caseinate decrease circulating TGs and the TG:ApoB48 ratio in comparison with glucose (Pal et al., 2010b). Another study of obesity compared the effects of different milk protein fractions, including WPI, whey protein hydrolyzate, α-lactalbumin, and casein GMP, on postprandial lipemia after a high fat meal, but found no significant differences (Holmer-Jensen et al., 2012). Whey protein has also been shown to suppress the appearance of postprandial circulating TGs, free fatty acids, and chylomicron-rich lipoprotein in diabetic patients, following a high fat meal, compared with controls of casein, fish, or plant protein sources (Mortensen et al., 2009). In contrast, another casein study failed to show any effects on postprandial TGs in T2D patients (Brader et al., 2010). There have been few long-term clinical studies. A 3-month trial of the effects of a fermented whey product on metabolic syndrome found some improvements in metabolic markers, although confounding effects on body weight make this trial difficult to interpret (Gouni-Berthold et al., 2012).

Blood pressure and vascular reactivity Observational and clinical studies have shown that the consumption of dairy products is associated with decreased risk of hypertension (Soedamah-Muthu et al., 2012; Fekete et al., 2016a,b). Much work has focused on identifying and isolating the bioactive peptides that may be responsible. The discovery that milk-derived peptides inhibit angiotensin converting enzyme (ACE) activity and hence alter vasoconstriction, vasodilation, and blood pressure (BP) in vitro led to a plethora of animal and subsequently human trials. Despite promising evidence that lactokinins or caseinkinins can reduce BP in spontaneously hypertensive animals, these findings are yet to be translated. The most well-studied milk protein-derived peptides are isoleucine-proline-proline (IPP) and valine-proline-proline (VPP), which are generated by the fermentation of milk (Saito, 2008; Boelsma and Kloek, 2009). IPP and VPP have been shown to be weak ACE inhibitors based on in vitro experiments (FitzGerald and Meisel, 2000). Several meta-analyses of clinical trials of the effects of IPP and VPP on BP, which suggest that these milk-derived peptides may have antihypertensive effects in humans, have been published (Xu et al., 2008; Cicero et al., 2013; Turpeinen et al., 2013). For example, a recent meta-analysis of 19 clinical trials, over the last 15 years, of the effects of a daily dose of 6 but will lose its milk-clotting ability at pH > 6.8 (Brown et al., 1988).

pH and ionic strength of small intestine The small intestine is the major region in which nutrients are digested and converted to an absorbable form. The human intestinal tract is a complex environment consisting of bile salts, pancreatic enzymes, coenzymes, various salts, phospholipids, yeasts, and various bacteria (Singh and Ye, 2013). When partially digested food products pass into the small intestine, the pH undergoes a rapid increase from the highly acidic environment (pH 1–3) in the stomach to the neutral environment (pH 6–7.5) in the duodenum, because of the secretion of sodium bicarbonate. This neutral pH provides an optimal environment for the action of pancreatic enzymes (McClements et al., 2008; Hur et al., 2009; Golding and Wooster, 2010). The increase in pH may cause some critical changes in the physicochemical properties of the proteins, for example, a reversal of the protein charge. It has been found that the osmolality of the duodenum contents is 180 mOsm/kg during the fasting state and the ionic strength is 140 mM (Lindahl et al., 1997; Kalantzi et al., 2006). However, because of the presence of various ions and solutes in ingested food, the postprandial osmolality and ionic strength may increase appreciably. For example, Kalantzi et al. (2006) have reported that, after the ingestion of a nutrition beverage, the osmolality of the duodenum increased rapidly to 290 mOsm/kg. In addition, the ionic strength in the small intestine is known to be a particularly significant factor that affects the electrostatic interactions among food components. The multivalent cations (e.g., Ca2+ and Mg2+) may influence the digestibility of long-chain saturated fatty acids and proteins through the formation of precipitates (Vaskonen, 2003; Reid, 2004; Karupaiah and Sundram, 2007).

Bile salts Bile salts play a significant role in both the digestion and the adsorption of lipids because of their high surface activity (Sarkar et al., 2016). Bile salts are present in the small intestine and originate from the liver through the gall bladder (Singh and Ye, 2013). Bile salt is a native biosurfactant (Golding and Wooster, 2010) but, unlike other surfactants, does not contain a hydrophobic head group and a hydrophilic tail group. Instead, its amphiphilic nature is mainly because of its flat steroidal structure, with methyl groups on the convex side and polar hydroxyl groups on the concave side (Euston et al., 2013; Galantini et al., 2015). Bile salts can adsorb readily at the oil-water interface in an emulsion and can displace the initial surfactants

674

19. Structural changes to milk protein products during gastrointestinal digestion

at the oil droplet surface when introduced to a simulated intestinal fluid (Hur et al., 2009). Therefore, bile salts facilitate the digestion of lipids by providing accessibility of the lipase/colipase complexes to the bile-coated lipid droplets (Sarkar et al., 2016). In addition, bile salts can facilitate the deformation of oil droplets under mechanical agitation, which enhances stability against aggregation and transports the hydrophobic substance by forming micelles (McClements et al., 2008). It has been reported that bile salts can improve the digestibility of both adsorbed and unadsorbed proteins in an emulsion, for example, β-lg-, myoglobin-, and bovine serum albumin (BSA)-stabilized emulsions (Gass et al., 2007).

Pancreatic lipase In healthy human adults, the digestion of lipids by the gastric lipases is very limited. Lipid digestion takes place mainly (70%–90%) in the small intestine and is catalyzed by pancreatic lipases (Bauer et al., 2005; Mun et al., 2006; Singh et al., 2009). Pancreatic lipases can work efficiently at about pH 6.5 (Singh and Ye, 2013), although their optimum pH is 8–9 (Patton and Carey, 1981). When the partially digested lipid droplets pass into the small intestine, the pancreatic lipase tends to adsorb at the droplet surface as a complex with colipase and/or bile salts (Bauer et al., 2005). Triglycerides and diacylglycerol are then broken down to FFAs and 2-monoglycerides. It has been reported that pancreatic lipase has no obvious specificity for the chain length of fatty acids but preferentially cleaves their sn-1 and sn-3 positions (Mu and Høy, 2004). This complexation reaction of pancreatic lipase at the lipid droplet interface requires the presence of bile salts, colipase, and calcium (Hur et al., 2011). Its degree of binding appears to depend on the electric charge of the interface and competitive adsorption with bile salts, digestion products, or other surfactants (McClements et al., 2008). Colipase is a coenzyme, which is essential to the action of lipase. It interacts with lipase to form a stoichiometric complex that adsorbs at the oil-water interface and thereby facilitates accessibility to the lipid substrate. Colipase consists of a hydrophilic group that combines with lipase and a hydrophobic part that connects with the interfacial layers (Bauer et al., 2005). Interestingly, the influence of the presence of bile salts on the activity of pancreatic lipase is complex. When bile salts are present in a relatively low concentration, they tend to solubilize the products of lipid digestion, such as FFAs and 2-monoglycerides, and remove them from the interfacial layers. In that situation, they accelerate the activity of pancreatic lipase. In contrast, a relatively high concentration of bile salts restrains the digestive ability of pancreatic lipase, which is mainly due to the competitive adsorption between bile salts and lipases (Gargouri et al., 1983).

Coagulation of milk protein under gastric conditions The casein micelles in milk are considered to be very stable to high temperature, compaction, commercial homogenization, and high concentrations of calcium ions (O’Mahony and Fox, 2013). However, acidic pH, specific proteolytic enzymes, and ethanol can result in destabilization of the casein micelles (see Chapter 6). Lowering the pH to the isoelectric point

Coagulation of milk during gastric digestion

675

of casein (pI ¼ 4.6) causes precipitation, and micellar calcium phosphate is fully removed when the pH is 4.9 (see Chapter 6). In addition, some milk-clotting enzymes, for example, chymosin and pepsin, can specifically catalyze the hydrolysis of κ-casein, which is split into para-κ-casein and a macropeptide. As a result, the casein micelles will coagulate and form a curd-like gel in the presence of calcium or other divalent ions (Lucey, 2011). These properties are also significant in the digestion of milk, as both acidic conditions and proteases (e.g., pepsin) exist in the gastric environment. Milk clotting is believed to be a complex process, involving a primary enzymatic hydrolysis (first stage) that causes the alteration of κ-casein and a loss of its ability to stabilize the casein micelle. Aggregation of the altered casein micelles takes place in the second stage, which is a nonenzymatic step. Then, the aggregates of caseins further form a firm milk gel, and curd syneresis probably occurs (McMahon and Brown, 1984). In the first stage (i.e., enzymatic hydrolysis) of the coagulation of milk, the proteases cleave the κ-casein molecules that are on the surface of the casein micelles into para-κ-casein and a macropeptide and hence initiate the milk-clotting process: κ
casein ! para
κ
casein + macropeptide The rate of hydrolysis is proportional to the enzyme concentration and depends on the pH (with an optimal pH at 5.6) (Carlson et al., 1987). Chymosin is capable of hydrolyzing κ-casein (Holt and Horne, 1996), and uniquely cleaves the Phe105-Met106 peptide bond, whereas other milk-clotting enzymes (e.g., pepsin) are less specific. However, all milkclotting enzymes have the same general functions (Fox, 1981). During the second stage, aggregation of the altered micelles is due mainly to the loss of electrostatic repulsion and steric repulsion of the κ-casein. The presence of calcium ions facilitates the formation of the coagulum by connecting micelles as a bridge and inducing an isoelectric condition (Horne, 2014). For further details on milk coagulation, see Chapters 6 and 16.

Coagulation of milk during gastric digestion Previous studies have shown that the digestion behavior of protein in the GI tract can be affected by its structure (Schmidt and van Marwijk, 1993; Kitabatake and Kinekawa, 1998; Zeece et al., 2008; Tunick et al., 2016). The structure of a protein in food is dependent on its source and the processing treatment. As different proteins have different structures and conformational properties, the nature of the protein may markedly affect its susceptibility to proteolysis by pepsin. Casein has a highly flexible, disordered conformation (Modler, 1985), which is more exposed to gastric hydrolysis by pepsin in the stomach (Mahe et al., 1996). However, the casein in milk appeared to cause a delayed delivery of amino acids to the small intestine in an in vivo digestion (Mahe et al., 1996), indicating slower hydrolysis of the caseins. Recently, it was shown that unheated skim milk formed a firm, dense, cheese-like clot with a porous network structure at about 10 min when the pH was >6.2 (Fig. 19.1). The clot structures became tighter and less permeable at longer gastric digestion times, as demonstrated with a dynamic digestion model—a human gastric simulator (HGS) (Ye et al., 2016b). Such

676

19. Structural changes to milk protein products during gastrointestinal digestion

10 min

80 min

220 min

FIG. 19.1 Images (left) and confocal micrographs (right) of clots formed at different digestion times in the in vitro

dynamic gastric digestion of unheated skim milk. Scale bars represent 10 μm. Reproduced from Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016. Formation of a structured clot during the gastric digestion of milk: impact on the rate of protein hydrolysis. Food Hydrocoll. 52, 478–486, with permission from Elsevier Inc.

a structure prevented accessibility of the pepsin to the interior of the clot, consequently influencing the digestion of casein by pepsin, which was shown by the different rates of pepsin hydrolysis between the caseins at the surface of the clot and those in the inner part of the clot. The HGS developed at the Riddet Institute and UC Davis (Kong and Singh, 2010) is a sophisticated model that can closely mimic human gastric behavior and mimics many relevant factors of gastric physiology, such as progressive acidification and emptying, that might significantly affect the bioaccessibility of nutrients. The formation of the clot is considered to be the result of the hydrolysis of κ-casein by pepsin leading to destabilization of the casein micelles, rather than the effect of low pH on the stability of the casein micelles because the coagulation occurs at pH >6.2. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Fig. 19.2) showed that the intensity of the κ-casein band decreased markedly even after 20 min of digestion and was virtually absent after 40 min. In contrast, the intensities of the αS-casein (αS1- + αS2-casein) and β-casein bands did not appear to change with the digestion time, which confirms the hypothesis that the pepsin in the simulated gastric fluid (SGF) was involved in converting the milk into a clot. However, β-lg and α-lactalbumin (α-la) were not involved in the formation of clot in the unheated milk. Few soluble proteins were found in the clot; they had emptied out from the stomach (Ye et al., 2016b). The formation of structured clots was also observed in whole milk under gastric conditions (Ye et al., 2016a). The fat globules in the milk appeared to be embedded in the clots when they formed (Fig. 19.3, images). The release of the fat globules was faster in the clots from the

Coagulation of milk during gastric digestion

677

(A)

(B) FIG. 19.2 SDS-PAGE patterns under reducing conditions of (A) the clot and (B) the emptied digesta obtained during the gastric digestion of unheated skim milk in an HGS at different digestion times. M, unheated milk. Reproduced from Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016. Formation of a structured clot during the gastric digestion of milk: impact on the rate of protein hydrolysis. Food Hydrocoll. 52, 478–486, with permission from Elsevier Inc.

heated whole milk. A linear correlation between the fat content and the protein content in the clots was observed. The slope of the regression line in this linear correlation was close to 1 (Fig. 19.3C), suggesting that, in the coagula, the fat was evenly distributed and was released when the local protein matrix was broken down. The clot from unheated whole milk had a close-knit network, with the fat globules evenly distributed in the matrix of the clot. Some fat globules appeared to have coalesced within the matrix of the clot, as evidenced by the large area of fat seen in the samples. With an increase in the digestion time, the knitted network became more dense, and the fat globules within the matrix tended to aggregate and coalesce. These images show that the fat globules were important in the building of structure during the formation of the clots, as the globules were not part of the clot matrix, and appeared to have been trapped within the protein matrix of the clot. In unheated milk, there were no associative interactions between the milk fat globule membrane (MFGM) and the proteins, so the only way in which the fat globules were contained within the protein gel was by entrapment.

19. Structural changes to milk protein products during gastrointestinal digestion

Fresh whole milk Milk before digestion

Heated whole milk 6.0

Fat in clot (g /100 g milk)

678

5.0 4.0 3.0 2.0 Fresh milk 1.0

20 min

0.0

(A)

0

50

100

150

200

250

Digestion time (min)

Weight of fat-free matter in clots (g /100 g)

80 min

Heated milk

6.0 5.0 4.0 3.0 2.0 Fresh milk 1.0

Heated milk

0.0 0

(B)

50

100

150

200

250

Digestion time (min)

100 y = 0.9617x + 1.9481 R2 = 0.9208

Fat in clot (%)

90

220 min

80 70 60 50 40

(C)

40

50

70 80 60 Fat-free matter in clot (%)

90

100

FIG. 19.3 Confocal microscopy images (left) of unheated whole milk and heated whole milk at different times (from 0 to 220 min) during gastric digestion. Red (dark gray in print version) shows the fat, and green (light gray in print version) shows the protein. The scale bar in all images is 20 μm. Change in (A) fat content and (B) dried fat-free matter weight (g/100-g milk) of (●) unheated whole milk and (●) heated whole milk (90°C for 20min) during digestion in SGF containing pepsin. (C) Relationship between the fat content and the fat-free matter content of the clots of (●) unheated whole milk and (●) heated whole milk during gastric digestion from 20 to 220 min. The regression line for all points is drawn. Reproduced from Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016. The formation and breakdown of structured clots from whole milk during gastric digestion. Food Funct. 7, 4259–4266, with permission from Royal Society of Chemistry.

Effect of processing treatments Fresh milk is usually processed to be safe for human consumption and to extend its shelf life. Thermal treatments, such as pasteurization and ultrahigh temperature (UHT) processing, lead to denaturation of the whey proteins and MFGM proteins, interactions between the

Effect of processing treatments

679

MFGM proteins and the milk serum proteins (Ye et al., 2004), and the association of the whey proteins with the casein micelles by interaction with κ-casein (Anema and Li, 2003a,b). The influence of the processing treatment on the digestibility of the milk proteins has been demonstrated. Wada and L€ onnerdal (2014) showed that the protein bands in SDS-PAGE gels of in-can sterilized milk underwent a rapid decrease in intensity during in vitro digestion, and this was true to a lesser extent in UHT milk. Proteomic analysis revealed that the level of lactulosyllysine, which reflects a decrease in the digestibility of protein, in α-la, β-lg, and caseins was higher in in-can sterilized milk than in UHT milk. Thus, industrial heating may improve the digestibility of milk proteins by denaturation, but the improvement is likely to be offset by heat-derived protein modifications, which may decrease the digestibility of the protein, for example, through the formation of cross-links such as lysinoalanine and lanthionine (Wada and L€ onnerdal, 2014). Miranda and Pelissier (1987) reported that heat treatment of milk resulted in accelerated gastric emptying and an increase in the rate of hydrolysis of the caseins in vivo. Barb
e et al. (2013) also reported a significant influence of heat treatment on the resistance and sensitivity of both casein and β-lg to hydrolysis in the human stomach in an in vivo experiment. Tunick et al. (2016) recently investigated the effect of heat and homogenization on the in vitro digestion of milk protein and fat and showed a faster release of the FFAs from homogenized milk in the intestinal fluid. However, these studies used in vivo experiments or simple in vitro models, and there is limited understanding of why there are differences in the digestibility of milk components after different processing treatments. Recent work showed that skim milk heated at 90°C for 20 min, in which the whey protein had been denatured, formed a looser, fragmented network–structured clot with numerous larger voids (Fig. 19.4) than unheated milk, which formed a firm, dense-structured clot (Fig. 19.1) (Ye et al., 2016b). This loose and fragmented structure resulted in altered hydrolysis by pepsin of the caseins and whey proteins under gastric conditions (Fig. 19.4). The hydrolysis of the caseins was much faster in heated milk than in unheated milk, essentially because the open, fragmented structure increased the effective contact area with the SGF, which contained pepsin. Pepsin was thus able to diffuse into and act on the proteins within the clot. In contrast to unheated milk, a large amount of whey protein was observed in the clot of heated milk; both β-lg and α-la were also hydrolyzed gradually, along with the caseins, into peptides during digestion (Fig. 19.4). This indicates that the whey proteins were involved in the formation of the clot during the gastric digestion of heated milk, partly because of the association of the whey proteins with the casein micelles during heat treatment (Anema and Li, 2003a,b; Donato et al., 2007b) and partly because the nonmicelle bound whey protein-κ-casein complexes formed during heating associated with the micelles during renneting (Kethireddipalli et al., 2011) and at pH values below 5.3 (Vasbinder et al., 2001; Donato et al., 2007a). However, the emptied digesta from the heated milk did not contain intact casein and whey proteins (Fig. 19.4), suggesting that the denatured whey proteins in heated milk were hydrolyzed rapidly in the digesta. In contrast, both β-lg and α-la remained intact during the whole digestion period in unheated milk (Fig. 19.2). Commercially available milk is commonly homogenized to increase the stability of the milk fat by reducing the size of the milk fat globules. This process leads to the caseins and whey proteins adsorbing onto the surface of the fat globules and reducing the amount of MFGM at the fat globule surface (Ye et al., 2008). Homogenization of whole milk followed by heat treatment causes the denatured whey proteins to attach to the adsorbed caseins

680

19. Structural changes to milk protein products during gastrointestinal digestion

(A)

(B) FIG. 19.4 Image and confocal micrograph (left) of clots at 10 and 220 min in the in vitro dynamic gastric digestion of heated (90°C, 20 min) skim milk. SDS-PAGE patterns of (A) the clot and (B) the emptied digesta obtained during the gastric digestion of heated (90°C for 20 min) skim milk in the HGS at different times. M, unheated milk. Reproduced from Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016. Formation of a structured clot during the gastric digestion of milk: impact on the rate of protein hydrolysis. Food Hydrocoll. 52, 478–486, with permission from Elsevier Inc.

and MFGM proteins via disulfide bonds (Michalski and Januel, 2006). These interactions can lead to changes in the interactions of the proteins under gastric conditions and hence the formation of clots and structures of clots (Ye et al., 2017; Mulet-Cabero et al., 2019). Ye et al. (2017) observed a greater reduction in the intensity of the protein bands at an earlier gastric digestion time in homogenized milk, suggesting greater hydrolysis of the proteins in the homogenized milk than in the untreated milk, in agreement with the more rapid breakdown of the clot. Very faint β-lg and α-la bands were observed in the clots from both untreated milk and homogenized milk, indicating that the whey proteins were not significantly involved in the formation of the clots and remained in the serum phase (digesta) (Ye et al., 2017). Using both an in vivo rat study and an in vitro dynamic gastric digestion model, UHT (140°C for 4 s) homogenized milk exhibited a more rapid rate of protein hydrolysis than unheated milk and pasteurized (72°C for 15 s or 85°C for 15 s) homogenized milk during gastric digestion (Ye et al., 2019). UHT milk also showed changes in the composition of the emptied digesta from the stomach. The faster rate of protein digestion has been suggested to be due to the looser structure, with more and larger voids within the curds, because of the UHT treatment. These results indicate that the heat intensity of the UHT treatment of milk is strong enough to cause a significant change in the digestion of the milk proteins. Mulet-Cabero et al. (2019) also showed a similar result for the digestion behavior of whole milk with homogenization and heat treatments of pasteurization at 72°C for 15 s and UHT treatment at 140°C for 3 s using a semidynamic digestion model. These authors reported that protein aggregation

Impact of milk coagulation on the release of fat globules during digestion

681

and coagulum formation occurred within the first 15 min of gastric digestion, at which time the pH of the whole milk ranged from 5.5 to 6. The homogenized samples creamed regardless of the heat treatment, whereas all nonhomogenized samples exhibited sedimentation. The coagulum of the heated samples was more fragmented than that of the nonheated samples. Rheological analysis showed that the higher was the temperature of the heat treatment, the softer was the coagulum obtained and the greater was the protein hydrolysis at the end of digestion. Although the mechanism behind this phenomenon is not clear, it is probably related to the involvement of the whey proteins with the casein micelles during heat treatment (Dalgleish, 1983). Heat treatment leads to the association of denatured whey proteins with the casein micelles and with the surface of the fat globules (Anema and Li, 2003a,b; Ye et al., 2004), which could reduce casein-casein interactions and casein-fat globule interactions during coagulation. This may also alter the orientation of the casein strands and provide a less complete fusion rearrangement, which may result in a change in the structure of the curds to a more open and looser network. This impairment of the coagulum properties has been reported in the rennet clotting of heat-treated milk (Singh and Waungana, 2001); an increase in moisture retention in the clot resulted in clots that were soggy and ragged in appearance, with poor matting ability. It has been suggested that this is due mainly to casein micelle–bound whey proteins, which are assumed to sterically hinder aggregation of the casein micelles even after the removal of nearly all of the κ-casein and a decrease in the concentration of ionic calcium in heated milk (Kethireddipalli and Hill, 2015). A schematic diagram showing a possible mechanism for the events during the formation of protein curds from unheated and heated milk in the gastric environment is given in Fig. 19.5.

Impact of milk coagulation on the release of fat globules during digestion In whole milk, the structure and the breakdown of the clots during gastric digestion influence the release of fat from the stomach to the small intestine. Comparison of the release of fat and the disintegration of the clot in fresh milk, homogenized milk, and heated homogenized milk showed that the fat was liberated in proportion to the digestion of the protein and consequently to the breaking down of the clots (the slope was close to 1 for all three milk samples) (Ye et al., 2017). The rate of release of the fat globules from the clots during gastric digestion showed a linear correlation with the disintegration of the clots under gastric conditions. This result suggests that, perhaps surprisingly, the manner in which the fat globules are incorporated within the clot matrix does not control their release. In untreated milk, the fat globules are trapped within the casein clot matrix, but there are no interactions between proteins and fat globules. In comparison, in homogenized milk, the surfaces of the fat globules are covered with casein, so that the globules themselves become part of the matrix as the clot is formed. In heated homogenized milk, the fat globule surfaces contain both denatured whey proteins and caseins and again participate in the formation of the clot matrix. Confocal microscopy showed similar close-knit networks of protein for untreated whole milk and unheated homogenized milk at the initial formation of the clot (Fig. 19.6). Some fat globules appeared to have coalesced within the matrix of the clot, as evidenced by the large

682

19. Structural changes to milk protein products during gastrointestinal digestion -SH

Whey protein (b-lactoglobulin) -SH

-SH

-SH

Heat treatment

-SH

-SH

Casein micelle

-SH

k-casein Pepsin

Denatured (b-lactoglobulin) Pepsin

Gastric conditions

-SH

-SH -SH

Dense clot

Slow hydrolysis

-SH

Loose clot

Quick hydrolysis

FIG. 19.5 A schematic diagram of the possible mechanism for the events during the formation of protein curds from unheated milk and heated milk in a gastric environment. Reproduced from Ye, A., Liu, W., Cui, J., Kong, X., Roy, D., Kong, Y., Han, J., Singh, H., 2019. Coagulation behaviour of milk under gastric digestion: effect of pasteurization and ultra-high temperature treatment. Food Chem. 286, 216–225, with permission from Elsevier Inc.

area of fat seen in the samples; in contrast, other fat globules were within the aqueous phase. In heated homogenized milk, the structure of the clots was much more open, with blocks of aggregated protein. Within the clot, the fat globules appeared to be clumped in separate areas and did not seem to be contained within the protein matrix, although they seemed to remain associated with it. These fat globules in the aqueous phase were apparently larger than the fat globules in the homogenized milk before digestion, suggesting that coalescence of the fat globules was taking place as the MFGM proteins surrounding the fat globules were hydrolyzed by pepsin (Ye et al., 2011). These results suggest that differences in the structure of the clot or coagulum formed from milks with various treatments also affect the delivery of the fat globules to the next digestion step.

Behavior of milk fat globule membrane proteins during digestion The influence of gastric proteolysis on the physicochemical characteristics of milk fat globules and the MFGM proteins in milk and cream has also been examined in an in vitro static

Behavior of milk fat globule membrane proteins during digestion

Fresh whole milk

683

Homogenized milk Heated (90°C, 20 min) homogenized milk

Milk before digestion

20 min

80 min

220 min

FIG. 19.6 Confocal microscopy images of untreated whole milk, homogenized milk, and heated homogenized milk (90°C for 20 min) at different times (from 0 to 220 min) during gastric digestion. Red (dark gray in print version) shows the fat, and green (light gray in print version) shows the protein. The scale bar in all images is 20 μm. Reproduced from Ye, A., Cui, J., Dalgleish, D., Singh, H., 2017. Effect of homogenization and heat treatment on the behavior of protein and fat globules during gastric digestion of milk. J. Dairy Sci. 100, 36–47, with permission from Elsevier Inc.

gastric digestion model (Ye et al., 2011; Gallier et al., 2012). As demonstrated in an SDS-PAGE analysis of the milk fat globules isolated from raw milk digested in an in vitro simulated gastric system, all the MFGM proteins were hydrolyzed by pepsin into a range of peptides, but the rate of hydrolysis was different for different MFGM proteins. For instance, xanthine oxidase was hydrolyzed much faster than butyrophilin; PAS 6 and PAS 7 were hydrolyzed faster than butyrophilin; PAS 7 was more sensitive than PAS 6 (Fig. 19.7). The hydrolysis of the interfacial MFGM proteins was followed by an apparent flocculation of the fat globules, as indicated by an increase in the average size (d32) of the milk fat globules with an increase in the gastric digestion time. The flocculation of the fat globules in the gastric environment may be attributed to the hydrolysis of charged MFGM proteins, resulting in a decrease in electrostatic repulsion. Examination of the gastric digesta by confocal laser scanning microscopy

684

19. Structural changes to milk protein products during gastrointestinal digestion

MUC1 Xanthine oxidase

Butyrophilin PAS 6 PAS 7

150 100 75 50 37 25 15 Peptides

10

RM

pH1.6

0.5

1.5

3

5

10

15

20

30 min

FIG. 19.7 SDS-PAGE patterns of the MFGM proteins/peptides obtained from the cream material of raw milk samples incubated in SGF containing pepsin at 0.1 mg/mL. Reproduced from Ye, A., Cui, J., Singh, H., 2011. Proteolysis of milk fat globule membrane proteins during in vitro gastric digestion of milk. J. Dairy Sci. 94, 2762–2770, with permission from Elsevier Inc.

showed that flocculation of the fat globules started at the early stages of digestion and seemed to be enhanced at long digestion times (30 min) (Ye et al., 2011). The fat globules were linked through some protein/peptide material between the globules. A closer examination of the microstructure revealed that the fat globules were physically entrapped in the casein aggregate network and that they remained essentially intact during gastric digestion. It is known that the casein micelles in the serum phase of milk will aggregate with a reduction in the pH and hydrolysis by pepsin. Gallier et al. (2013a) studied the in vivo gastric digestion of bovine milk fat globules in cream derived from either raw milk or heated milk (63°C for 30 min). Fasted rats were orally gavaged once with one of the cream preparations, and stomach chyme samples were collected from the rats posteuthanasia after 30 min, 2 h, and 3 h postgavage. Cream was used to minimize casein coagulation in the stomach. The two cream samples presented similar initial protein profiles and showed rapid hydrolysis of the MFGM proteins during digestion. Several peptides appeared after 30 min of gastric digestion. These results were generally similar to the observations made in the in vitro experiments (Gallier et al., 2012). They suggest that some of the peptides hydrolyzed from the MFGM proteins remained at the globule surface and provided sufficient electrostatic repulsion and steric barriers to prevent coalescence of the fat globules. In addition, the phospholipid in the MFGM would be expected to remain at the surface and prevent the coalescence of the fat globules even in the absence of MFGM proteins (Singh, 2019). It has been reported that lipid in homogenized human milk was more rapidly hydrolyzed by pancreatic lipase than in native human milk, which shows a lag phase in the fatty acid release profile (Berton et al., 2009; Ye et al., 2010; Liang et al., 2017). This is probably because of the inhibitory effect of components in the MFGM on the activity of lipase. Pancreatic lipase may take longer to penetrate through the dense membrane (with 40% phospholipids) to reach the core fat. In contrast, in recombined milk, casein micelles and serum proteins make up the

Milk protein ingredients

685

surface layer of the fat globules. Pancreatic lipase may have higher affinity for the micellar casein surface of the fat globules, leading to more lipase binding with the interface of the fat globules. It has been reported that virtually, all intrinsic milk lipoprotein lipase is associated with casein, with about 70% being bound to micellar casein and the remainder existing in the form of soluble casein-enzyme complexes in the milk serum (Downey and Andrews, 1966; Anderson, 1982). The promotion of lipolysis in milk that is agitated or homogenized may be attributed to disruption of the MFGM and adsorption of caseins onto the surface of the fat globules (Cartier and Chilliard, 1989), such that the casein brings more lipase to the interface of the fat globules. In the presence of bile extract (5.0 mg/mL), the rate of lipid digestion by pancreatic lipase was much faster and almost similar in both milks. However, the total amount of fatty acids released was slightly lower in the raw milk sample than in the recombined milk sample. It appeared that the bile salts either rapidly adsorbed or rapidly displaced (or both) the exiting biomaterials from the fat globules and largely eliminated the different properties of the original surfaces required for the adsorption of pancreatic lipase. It has been reported that the presence of bile salts at the fat globule surface promotes greater binding of pancreatic lipase to the interface by facilitating the formation of substrate clusters and thus enhancing lipid digestion (Wickham et al., 1998; Sarkar et al., 2016).

Milk protein ingredients The physicochemical behavior and the coagulation behavior of milk proteins in different commercial dairy protein ingredients during in vitro gastric digestion have also been investigated recently using the dynamic digestion model (HGS) (Wang et al., 2018). At the same protein concentration, skim milk reconstituted from skim milk powder (SMP) and milk protein concentrate (MPC) solution (both of which contain intact casein micelles) formed a cheese-like curd with a close-knit network at an early digestion time (10 min), and a long time was required to disintegrate this structure fully. There were some slight differences in the structures of the curds derived from SMP and MPC. The SMP curd contained a much larger number of fragmented, aggregated protein blocks than the MPC curd. The curd obtained from SMP contained numerous loose, crumbled fragments, but the curd obtained from MPC was a more integrated ball-like pellet. This difference in the structure of the clots is probably due to the difference in the preprocessing treatment of these two ingredients. The SMP had a whey protein nitrogen index of between 1.51 and 5.99 mg/g, which indicates that the whey proteins had been partly denatured by preheating during the manufacturing process. Thus, denatured β-lg was involved in the formation of the SMP curd, because of the association of β-lg with casein micelles via micellar κ-casein, as shown for heated milk (Ye et al., 2016b). However, membrane separation, that is, a combined ultrafiltration/ diafiltration process, is mainly applied in the manufacture of MPC, for which the intensity of the heat treatment is generally lower (Carr and Golding, 2016). In contrast, sodium caseinate produced a white precipitate rather than a structured curd at a later digestion time (after 40 min) when the pH was about 5. The amount of this loose, fragmented protein matrix reduced during digestion. Interestingly, the coagulation behavior

686

19. Structural changes to milk protein products during gastrointestinal digestion

of a calcium-depleted MPC ingredient was more close to that of sodium caseinate; the structured curd was not observed during gastric digestion, and the protein coagulated at a later stage of digestion and at a lower pH. This may be attributed to the disintegration of the casein micelles because of the depletion of calcium in this calcium-reduced MPC. Caseins without micellar structure or caseins dissociated from casein micelles are not coagulated by pepsin under gastric conditions but coagulate only because of a low pH. Furthermore, the differences in the structures between the coagulum formed from casein micelles and that formed from individual caseins led to various disintegration rates of the protein coagulum and various hydrolysis rates of the protein by pepsin. The hydrolysis rate of the curds derived from SMP and MPC was much slower than that of the curds derived from calcium-depleted MPC and sodium caseinate. The composition and the rate of protein delivered from the stomach to the intestine are apparently different for ingredients containing casein micelles (SMP and MPC) and for ingredients containing no micellar casein (caseinate and calcium-depleted MPC) during digestion (Fig. 19.8). No intact proteins were observed in the digesta of SMP and MPC after 20 min but occurred after 40 min for caseinate and calcium-depleted MPC. This was attributed to the formation of clots at earlier digestion times for SMP and MPC, which delayed the delivery of protein to the next step of digestion. Of the whey proteins, native β-lg is resistant to some proteases, particularly pepsin, because of its unique structural stability at low pH (Miranda and Pelissier, 1983; Reddy et al., 1988). Heating, high-pressure treatment, addition of alcohols, and esterification have been reported to increase the susceptibility of β-lg to hydrolysis by pepsin (Chobert et al., 1995; Dalgalarrondo et al., 1995; Guo et al., 1995; Zeece et al., 2008). These treatments induce conformational changes in β-lg, resulting in increased exposure of peptic cleavage sites and thus increased susceptibility to pepsin action. In addition, several different kinds of intermediates and aggregates are formed during heat treatment, and it has been shown that different aggregates and modified monomers may behave differently under GI conditions, particularly in the presence of pepsin. Other whey proteins, α-la, BSA, and lactoferrin, are more susceptible to hydrolysis by pepsin (Nik et al., 2010; Furlund et al., 2013). Although the hydrolysis of β-lg, α-la, and BSA by pepsin has been studied extensively (Chobert et al., 1995; Dalgalarrondo et al., 1995; Guo et al., 1995; Zeece et al., 2008; Peram et al., 2013), the interactions of these proteins in whey protein isolate (WPI) in a dynamic gastric digestion have been studied only recently (Wang et al., 2018). Under dynamic gastric conditions, the pH decreases from pH 6 to pH 2 at 240 min, and β-lg in the WPI remains intact during the whole digestion process, releasing gradually to the intestine. α-La and BSA remain intact during the early digestion period when the pH is >5 but are hydrolyzed at pH < 4 (Fig. 19.8). For heated WPI during gastric digestion, the proteins form aggregates with large particle sizes and precipitate when the pH reduces to about pH 5, but the aggregates disappear with further digestion when the pH is cow > buffalo, indicating better digestion of goat milk and camel milk because of their smaller fat globule sizes. Digestion of the fat globules will also be influenced by the structure in which the fat globules are embedded when milk is consumed and coagulated in the stomach. The capture and subsequent liberation of the fat globules from the protein matrix formed by the casein micelles will define their availability for lipolysis by gastric and pancreatic lipases (Ye et al., 2016a, 2017). As the milks from different species vary in protein composition, there may be differences in the structure of the protein clot formed from the whole milks of different species, leading to differences in the liberation of the fat globules from the protein matrix of the milks of different species.

Concluding remarks We have reviewed recent studies on the formation of the structures induced by the interactions between the milk components and physiological components within the GI tract after the ingestion of milk and dairy ingredients, to which relatively little attention has been given to date. These recent studies have confirmed the coagulation of the casein micelles that is induced by pepsin proteolysis and the formation of structural clots containing protein and fat globules during digestion in the stomach. The formation of structural clots in milk during

694

19. Structural changes to milk protein products during gastrointestinal digestion

gastric digestion and its impact on digestion are considered to be a key factor in the digestion behaviors of the milk components, including protein and fat. These studies have revealed that the formation and the structure of the clots are largely dependent on the composition of the milk, previous processing treatments, and the milk source. Homogenization and heat treatment of the milk before ingestion in the body markedly influence the coagulation of the protein and the formation and the structure of the clots. The curds formed by homogenized and heat-treated milk have a more crumbled and porous structure, which leads to greater rates of protein hydrolysis by pepsin, and the fat globules are released into the small intestine more rapidly. Furthermore, any changes in the composition and structure of the milk proteins and the milk fat resulting from treatments during the manufacture of dairy products and dairy ingredients or resulting from different species as the milk source could modify the interactions between milk components and physiological components and consequently could result in variations in the formation and structure of the clots and the digestion of the milk components. One key interaction is the coagulation of the casein micelles that is induced by pepsin hydrolysis under gastric conditions, in which the milk proteins not only are hydrolyzed to provide peptides and amino acids for nutrition but also provide the structural aspects for manipulating the digestion rates of nutrients in milk and dairy products. More understanding of the interactions and the factors that influence the interactions and the formation of structure is required. This will potentially contribute to the development of novel dairy products with superior healthy functions.

References Alf
erez, M.J.M., Barrionuevo, M., Lo´pez Aliaga, I., Lisbona, F., Robles, J.C., Campos, M.S., Sanz-Sampelayo, M.R., 2001. Digestive utilization of goat and cow milk fat in malabsorption syndrome. J. Dairy Res. 68, 451–461. Almaas, H., Cases, A.-L., Devold, T.G., Holm, H., Langsrud, T., Aabakken, L., Aadnoey, T., Vegarud, G.E., 2006. In vitro digestion of bovine and caprine milk by human gastric and duodenal enzymes. Int. Dairy J. 16, 961–968. Anderson, M., 1982. Factors affecting the distribution of lipoprotein lipase activity between serum and casein micelles in bovine milk. J. Dairy Res. 49, 51–59. Anema, S.G., Li, Y., 2003a. Association of denatured whey proteins with casein micelles in heated reconstituted skim milk and its effect on casein micelle size. J. Dairy Res. 70, 73–83. Anema, S.G., Li, Y., 2003b. Effect of pH on the association of denatured whey proteins with casein micelles in heated reconstituted skim milk. J. Agric. Food Chem. 51, 1640–1646. Ballard, O., Morrow, A.L., 2013. Human milk composition: nutrients and bioactive factors. Pediatr. Clin. N. Am. 60, 49–74. Barb
e, F., M
enard, O., Le Gouar, Y., Buffie`re, C., Famelart, M.-H., Laroche, B., Le Feunteun, S., Dupont, D., R
emond, D., 2013. The heat treatment and the gelation are strong determinants of the kinetics of milk proteins digestion and of the peripheral availability of amino acids. Food Chem. 136, 1203–1212. Barlowska, J., Szwajkowska, M., Litwinczuk, Z., Krol, J., 2011. Nutritional value and technological suitability of milk from various animal species used for dairy production. Compr. Rev. Food Sci. Food Safety 10, 291. Bauer, E., Jakob, S., Mosenthin, R., 2005. Principles of physiology of lipid digestion. Asian Australas. J. Anim. Sci. 18, 282–295. Bell, A.E., Vlahopouloua, I., 1995. Preliminary studies on the gelation processes of fermented and GDL-acidified bovine and caprine milk systems. Int. J. Dairy Technol. 48, 112–116. Berton, A., Sebban-Kreuzer, C., Rouvellac, S., Lopez, C., Crenon, I., 2009. Individual and combined action of pancreatic lipase and pancreatic lipase-related proteins 1 and 2 on native versus homogenized milk fat globules. Mol. Nutr. Food Res. 53, 1592–1602. Blakeborough, P., Gurr, M.I., Salter, D.N., 1986. Digestion of the zinc in human milk, cow’s milk and a commercial babyfood: some implications for human infant nutrition. Br. J. Nutr. 55, 209–217.

References

695

Boirie, Y., Dangin, M., Gachon, P., Vasson, M.-P., Maubois, J.-L., Beaufre`re, B., 1997. Slow and fast dietary proteins differently modulate postprandial protein accretion. Proc. Natl. Acad. Sci. U S A 94, 14930–14935. Brown, R.J., Ernstrom, C., Johnson, M.E., 1988. Milk-clotting enzymes and cheese chemistry. In: Wong, N.P. (Ed.), Fundamentals of Dairy Chemistry. Springer, New York, NY, pp. 609–654. Carlson, A., Hill, C.G., Olson, N.F., 1987. Kinetics of milk coagulation: I. The kinetics of kappa casein hydrolysis in the presence of enzyme deactivation. Biotechnol. Bioeng. 29, 582–589. Carr, A., Golding, M., 2016. Functional milk proteins production and utilization: casein-based ingredients. In: McSweeney, P.L.H., O’Mahony, J.A. (Eds.), Advanced Dairy Chemistry—1B, Proteins: Applied Aspects. Springer, New York, NY, pp. 35–66. Cartier, P., Chilliard, Y., 1989. Lipase redistribution in cows’ milk during induced lipolysis: I. Activation by agitation, temperature change, blood serum and heparin. J. Dairy Res. 56, 699–709. Ceballos, L.S., Morales, E.R., Martı´nez, L.P., Extremera, F.G., Sampelayo, M.R.S., 2009. Utilization of nitrogen and energy from diets containing protein and fat derived from either goat milk or cow milk. J. Dairy Res. 76, 497–504. Chobert, J.-M., Briand, L., Grinberg, V., Haertl
e, T., 1995. Impact of esterification on the folding and the susceptibility to peptic proteolysis of β-lactoglobulin. Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1248, 170–176. Claeys, W.L., Verraes, C., Cardoen, S., De Block, J., Huyghebaert, A., Raes, K., Dewettinck, K., Herman, L., 2014. Consumption of raw or heated milk from different species: an evaluation of the nutritional and potential health benefits. Food Control 42, 188–201. Clark, S., Sherbon, J., 2000. Alpha s1-casein, milk composition and coagulation properties of goat milk. Small Rumin. Res. 38, 123–134. Crowley, S.V., Kelly, A.L., Lucey, J.A., O’Mahony, J.A., 2017. Potential applications of non-bovine mammalian milk in infant nutrition. In: Park, Y.W., Haenlein, G.F.W., Wendorff, W.L. (Eds.), Handbook of Milk of Non-Bovine Mammals, second ed. John Wiley & Sons, Oxford, UK, pp. 625–654. Cunsolo, V., Saletti, R., Muccilli, V., Foti, S., 2007. Characterization of the protein profile of donkey’s milk whey fraction. J. Mass Spectrom. 42, 1162–1174. Cunsolo, V., Saletti, R., Muccilli, V., Gallina, S., Di Francesco, A., Foti, S., 2017. Proteins and bioactive peptides from donkey milk: the molecular basis for its reduced allergenic properties. Food Res. Int. 99, 41–57. Dalgalarrondo, M., Dufour, E., Chobert, J.-M., Bertrand-Harb, C., Haertl
e, T., 1995. Proteolysis of β-lactoglobulin and β-casein by pepsin in ethanolic media. Int. Dairy J. 5, 1–14. Dalgleish, D.G., 1983. Coagulation of renneted bovine casein micelles: dependence on temperature, calcium ion concentration and ionic strength. J. Dairy Res. 50, 331–340. Dalgleish, D.G., Corredig, M., 2012. The structure of the casein micelle of milk and its changes during processing. Annu. Rev. Food Sci. Technol. 3, 449–467. Dalziel, J.E., Smolenski, G.A., McKenzie, C.M., Haines, S.R., Day, L., 2018. Differential effects of sheep and cow skim milk before and after fermentation on gastrointestinal transit of solids in a rat model. J. Funct. Foods 47, 116–126. Dangin, M., Boirie, Y., Garcia-Rodenas, C., Gachon, P., Fauquant, J., Callier, P., Balle`vre, O., Beaufre`re, B., 2001. The digestion rate of protein is an independent regulating factor of postprandial protein retention. Am. J. Physiol. Endocrinol. Metab. 280, E340–E348. Domagała, J., 2009. Instrumental texture, syneresis and microstructure of yoghurts prepared from goat, cow and sheep milk. Int. J. Food Prop. 12, 605–615. Donato, L., Alexander, M., Dalgleish, D.G., 2007a. Acid gelation in heated and unheated milks: interactions between serum protein complexes and the surfaces of casein micelles. J. Agric. Food Chem. 55, 4160–4168. Donato, L., Guyomarc’h, F., Amiot, S., Dalgleish, D.G., 2007b. Formation of whey protein/κ-casein complexes in heated milk: preferential reaction of whey protein with κ-casein in the casein micelles. Int. Dairy J. 17, 1161–1167. Downey, W., Andrews, P., 1966. Studies on the properties of cow’s-milk tributyrinases and their interaction with milk proteins. Biochem. J. 101, 651–660. Dupont, D., Le Feunteun, S., Marze, S., Souchon, I., 2017. Structuring food to control its disintegration in the gastrointestinal tract and optimize nutrient bioavailability. Innov. Food Sci. Emerg. Technol. 46, 83–90. Ehlayel, M.S., Hazeima, K.A., Al-Mesaifri, F., Bener, A., 2011. Camel milk: an alternative for cow’s milk allergy in children. Allergy Asthma Proc. 32, 255–258. El-Agamy, E.I., 2007. The challenge of cow milk protein allergy. Small Rumin. Res. 68, 64–72. El-Agamy, E.I., Nawar, M., Shamsia, S.M., Awad, S., Haenlein, G.F.W., 2009. Are camel milk proteins convenient to the nutrition of cow milk allergic children? Small Rumin. Res. 82, 1–6.

696

19. Structural changes to milk protein products during gastrointestinal digestion

El-Zeini, H.M., 2006. Microstructure, rheological and geometrical properties of fat globules of milk from different animal species. Pol. J. Food Nutr. Sci. 15, 147. Euston, S.R., Baird, W.G., Campbell, L., Kuhns, M., 2013. Competitive adsorption of dihydroxy and trihydroxy bile salts with whey protein and casein in oil-in-water emulsions. Biomacromolecules 14, 1850–1858. Farah, Z., 1993. Composition and characteristics of camel milk. J. Dairy Res. 60, 603–626. Fardet, A., Dupont, D., Rioux, L.-E., Turgeon, S.L., 2019. Influence of food structure on dairy protein, lipid and calcium bioavailability: a narrative review of evidence. Crit. Rev. Food Sci. Nutr., 59, 1987–2010. Fox, P.F., 1981. Proteinases in dairy technology. Neth. Milk Dairy J. 35, 233–253. Furlund, C., Ulleberg, E., Devold, T., Flengsrud, R., Jacobsen, M., Sekse, C., Holm, H., Vegarud, G., 2013. Identification of lactoferrin peptides generated by digestion with human gastrointestinal enzymes. J. Dairy Sci. 96, 75–88. Galantini, L., di Gregorio, M.C., Gubitosi, M., Travaglini, L., Tato, J.V., Jover, A., Meijide, F., Tellini, V.H.S., Pavel, N.V., 2015. Bile salts and derivatives: rigid unconventional amphiphiles as dispersants, carriers and superstructure building blocks. Curr. Opin. Colloid Interface Sci. 20, 170–182. Gallier, S., Ye, A., Singh, H., 2012. Structural changes of bovine milk fat globules during in vitro digestion. J. Dairy Sci. 95, 3579–3592. Gallier, S., Cui, J., Olson, T.D., Rutherfurd, S.M., Ye, A., Moughan, P.J., Singh, H., 2013a. In vivo digestion of bovine milk fat globules: effect of processing and interfacial structural changes. I. Gastric digestion. Food Chem. 141, 3273–3281. Gallier, S., Zhu, X.Q., Rutherfurd, S.M., Ye, A., Moughan, P.J., Singh, H., 2013b. In vivo digestion of bovine milk fat globules: effect of processing and interfacial structural changes. II. Upper digestive tract digestion. Food Chem. 141, 3215–3223. Gantner, V., Mijic, P., Baban, M., Skrtic, Z., Turalija, A., 2015. The overall and fat composition of milk of various species. Mljekarstvo 65, 223–231. Gargouri, Y., Julien, R., Bois, A., Verger, R., Sarda, L., 1983. Studies on the detergent inhibition of pancreatic lipase activity. J. Lipid Res. 24, 1336–1342. Gass, J., Vora, H., Hofmann, A.F., Gray, G.M., Khosla, C., 2007. Enhancement of dietary protein digestion by conjugated bile acids. Gastroenterology 133, 16–23. Golding, M., Wooster, T.J., 2010. The influence of emulsion structure and stability on lipid digestion. Curr. Opin. Colloid Interface Sci. 15, 90–101. Guo, M., Fox, P., Flynn, A., Kindstedt, P., 1995. Susceptibility of β-lactoglobulin and sodium caseinate to proteolysis by pepsin and trypsin. J. Dairy Sci. 78, 2336–2344. Guo, Q., Ye, A., Bellissimo, N., Singh, H., Rousseau, D., 2017. Modulating fat digestion through food structure design. Prog. Lipid Res. 68, 109–118. Gurr, M.I., 1981. Human and artificial milks for infant feeding. J. Dairy Res. 48, 519–554. Haenlein, G.F.W., 1996. Nutritional value of dairy products of ewe and goat milk. In: Proceedings of the IDF/CIRVAL Seminar Production and Utilization of Ewe and Goat Milk, Grete, Greece, October 19–21, 1995. International Dairy Federation, Brussels, Belgium, pp. 159–178. Haenlein, G.F.W., 2004. Goat milk in human nutrition. Small Rumin. Res. 51, 155–163. Haenlein, G.F.W., Wendorff, W.L., 2008. Sheep milk. In: Park, Y.W., Haenlein, G.F.W. (Eds.), Handbook of Milk of Non-Bovine Mammals. Blackwell Publishing Professional, Oxford, UK, pp. 137–194. Hinz, K., O’Connor, P.M., Huppertz, T., Ross, R.P., Kelly, A.L., 2012. Comparison of the principal proteins in bovine, caprine, buffalo, equine and camel milk. J. Dairy Res. 79, 185–191. Hodgkinson, A.J., Wallace, O.A.M., Boggs, I., Broadhurst, M., Prosser, C.G., 2018. Gastric digestion of cow and goat milk: impact of infant and young child in vitro digestion conditions. Food Chem. 245, 275–281. Hodgkinson, A.J., Wallace, O.A.M., Smolenski, G., Prosser, C.G., 2019. Gastric digestion of cow and goat milk: peptides derived from simulated conditions of infant digestion. Food Chem. 276, 619–625. Hollebeeck, S., Borlon, F., Schneider, Y.-J., Larondelle, Y., Rogez, H., 2013. Development of a standardised human in vitro digestion protocol based on macronutrient digestion using response surface methodology. Food Chem. 138, 1936–1944. Holt, C., Horne, D.S., 1996. The hairy casein micelle: evolution of the concept and its implications for dairy technology. Neth. Milk Dairy J. 50, 85–111. Horne, D.S., 2014. Casein micelle structure and stability. In: Thompson, A., Boland, M., Singh, H. (Eds.), Milk Proteins: From Expression to Food, second ed. Elsevier, Amsterdam, pp. 169–200. Hur, S.J., Decker, E.A., McClements, D.J., 2009. Influence of initial emulsifier type on microstructural changes occurring in emulsified lipids during in vitro digestion. Food Chem. 114, 253–262.

References

697

Hur, S.J., Lim, B.O., Decker, E.A., McClements, D.J., 2011. In vitro human digestion models for food applications. Food Chem. 125, 1–12. Inglingstad, R.A., Devold, T.G., Eriksen, E.K., Holm, H., Jacobsen, M., Liland, K.H., Rukke, E.O., Vegarud, G.E., 2010. Comparison of the digestion of caseins and whey proteins in equine, bovine, caprine and human milks by human gastrointestinal enzymes. Dairy Sci. Technol. 90, 549–563. Jandal, J., 1996. Comparative aspects of goat and sheep milk. Small Rumin. Res. 22, 177–185. Jasi
nska, B., 1995. The comparison of pepsin and trypsin action on goat, cow, mare and human caseins. Rocz. Akad. Med. Bialymst. 40, 486–493. Jenness, R., 1974. Biosynthesis and composition of milk. J. Investig. Dermatol. 63, 109–118. Jenness, R., 1980. Composition and characteristics of goat milk: review 1968
1979. J. Dairy Sci. 63, 1605–1630. Kalantzi, L., Goumas, K., Kalioras, V., Abrahamsson, B., Dressman, J.B., Reppas, C., 2006. Characterization of the human upper gastrointestinal contents under conditions simulating bioavailability/bioequivalence studies. Pharm. Res. 23, 165–176. Kamal, M., Foukani, M., Karoui, R., 2017. Rheological and physical properties of camel and cow milk gels enriched with phosphate and calcium during acid-induced gelation. J. Food Sci. Technol. 54, 439–446. Karupaiah, T., Sundram, K., 2007. Effects of stereospecific positioning of fatty acids in triacylglycerol structures in native and randomized fats: a review of their nutritional implications. Nutr. Metab. 4, 16. Kethireddipalli, P., Hill, A.R., 2015. Rennet coagulation and cheesemaking properties of thermally processed milk: overview and recent developments. J. Agric. Food Chem. 63, 9389–9403. Kethireddipalli, P., Hill, A.R., Dalgleish, D.G., 2011. Interaction between casein micelles and whey protein/κ-casein complexes during renneting of heat-treated reconstituted skim milk powder and casein micelle/serum mixtures. J. Agric. Food Chem. 59, 1442–1448. Kitabatake, N., Kinekawa, Y.-I., 1998. Digestibility of bovine milk whey protein and β-lactoglobulin in vitro and in vivo. J. Agric. Food Chem. 46, 4917–4923. Kong, F., Singh, R.P., 2010. A human gastric simulator (HGS) to study food digestion in human stomach. J. Food Sci. 75, E627–E635. Li, J., Ye, A., Lee, S.J., Singh, H., 2013. Physicochemical behaviour of WPI-stabilized emulsions in in vitro gastric and intestinal conditions. Colloids Surf. B: Biointerfaces 111, 80–87. Liang, L., Qi, C., Wang, X., Jin, Q., McClements, D.J., 2017. Influence of homogenization and thermal processing on the gastrointestinal fate of bovine milk fat: in vitro digestion study. J. Agric. Food Chem. 65, 11109–11117. Lindahl, A., Ungell, A.-L., Knutson, L., Lennern€as, H., 1997. Characterization of fluids from the stomach and proximal jejunum in men and women. Pharm. Res. 14, 497–502. Lucey, J.A., 2011. Rennet-induced coagulation of milk. In: Fuquay, J.W., Fox, P.F., McSweeney, P.L. (Eds.), Encyclopedia of Dairy Sciences. In: vol. 1. Academic Press, San Diego, CA, pp. 579–584. Maathuis, A., Havenaar, R., He, T., Bellmann, S., 2017. Protein digestion and quality of goat and cow milk infant formula and human milk under simulated infant conditions. J. Pediatr. Gastroenterol. Nutr. 65, 661. Mah
e, S., Benamouzig, R., Gaudichon, C., Huneau, J., De Cruz, I., Rautureau, J., Tom
e, D., 1995. Nitrogen movements in the upper jejunum lumen in humans fed low amounts of casein or beta-lactoglobulin. Gastroenterol. Clin. Biol. 19, 20–26. Mahe, S., Roos, N., Benamouzig, R., Davin, L., Luengo, C., Gagnon, L., Gausserges, N., Rautureau, J., Tom
e, D., 1996. Gastrojejunal kinetics and the digestion of [15N] beta-lactoglobulin and casein in humans: the influence of the nature and quantity of the protein. Am. J. Clin. Nutr. 63, 546–552. Mandalari, G., Mackie, A.M., Rigby, N.M., Wickham, M.S., Mills, E.C., 2009. Physiological phosphatidylcholine protects bovine β-lactoglobulin from simulated gastrointestinal proteolysis. Mol. Nutr. Food Res. 53, S131–S139. Mao, L., Miao, S., 2015. Structuring food emulsions to improve nutrient delivery during digestion. Food Eng. Rev. 7, 439–451. McClements, D.J., Decker, E.A., Park, Y., 2008. Controlling lipid bioavailability through physicochemical and structural approaches. Crit. Rev. Food Sci. Nutr. 49, 48–67. McClements, D.J., Decker, E.A., Park, Y., Weiss, J., 2009. Structural design principles for delivery of bioactive components in nutraceuticals and functional foods. Crit. Rev. Food Sci. Nutr. 49, 577–606. McMahon, D.J., Brown, R.J., 1984. Composition, structure, and integrity of casein micelles: a review. J. Dairy Sci. 67, 499–512. Meena, S., Rajput, Y., Sharma, R., 2014. Comparative fat digestibility of goat, camel, cow and buffalo milk. Int. Dairy J. 35, 153–156.

698

19. Structural changes to milk protein products during gastrointestinal digestion

Michalski, M.-C., Januel, C., 2006. Does homogenization affect the human health properties of cow’s milk? Trends Food Sci. Technol. 17, 423–437. Miranda, G., Pelissier, J.-P., 1983. Kinetic studies of in vivo digestion of bovine unheated skim-milk proteins in the rat stomach. J. Dairy Res. 50, 27–36. Miranda, G., Pelissier, J.-P., 1987. Influence of heat treatment of bovine skim-milk on in vivo digestion in rat stomach. Lait 67, 365–377. Modler, H., 1985. Functional properties of nonfat dairy ingredients—a review. Modification of products containing casein. J. Dairy Sci. 68, 2195–2205. Mros, S., Carne, A., Ha, M., Bekhit, A.E.-D., Young, W., McConnell, M., 2017. Comparison of the bioactivity of whole and skimmed digested sheep milk with that of digested goat and cow milk in functional cell culture assays. Small Rumin. Res. 149, 202–208. Mu, H., Høy, C.-E., 2004. The digestion of dietary triacylglycerols. Prog. Lipid Res. 43, 105–133. Mulet-Cabero, A.-I., Mackie, A.R., Wilde, P.J., Fenelon, M.A., Brodkorb, A., 2019. Structural mechanism and kinetics of in vitro gastric digestion are affected by process-induced changes in bovine milk. Food Hydrocoll. 86, 172–183. Mun, S., Decker, E.A., Park, Y., Weiss, J., McClements, D.J., 2006. Influence of interfacial composition on in vitro digestibility of emulsified lipids: potential mechanism for chitosan’s ability to inhibit fat digestion. Food Biophys. 1, 21–29. Murgia, A., Scano, P., Contu, M., Ibba, I., Altea, M., Bussu, M., Demuru, M., Porcu, A., Caboni, P., 2016. Characterization of donkey milk and metabolite profile comparison with human milk and formula milk. LWT–Food Sci. Technol. 74, 427–433. Nakai, S., Li-Chan, E., 1987. Effect of clotting in stomachs of infants on protein digestibility of milk. Food Microstruct. 6, 161–170. N’Goma, J.-C.B., Amara, S., Dridi, K., Jannin, V., Carrie`re, F., 2012. Understanding the lipid-digestion processes in the GI tract before designing lipid-based drug-delivery systems. Ther. Deliv. 3, 105–124. Nguyen, H.T.H., Afsar, S., Day, L., 2018. Differences in the microstructure and rheological properties of low-fat yoghurts from goat, sheep and cow milk. Food Res. Int. 108, 423–429. Nik, A.M., Corredig, M., Wright, A.J., 2010. Changes in WPI-stabilized emulsion interfacial properties in relation to lipolysis and ß-carotene transfer during exposure to simulated gastric–duodenal fluids of variable composition. Food Dig. 1, 14–27. Nyemb, K., Gu
erin-Dubiard, C., Dupont, D., Jardin, J., Rutherfurd, S.M., Nau, F., 2014. The extent of ovalbumin in vitro digestion and the nature of generated peptides are modulated by the morphology of protein aggregates. Food Chem. 157, 429–438. Nyemb, K., Gu
erin-Dubiard, C., P
ezennec, S., Jardin, J., Briard-Bion, V., Cauty, C., Rutherfurd, S.M., Dupont, D., Nau, F., 2016. The structural properties of egg white gels impact the extent of in vitro protein digestion and the nature of peptides generated. Food Hydrocoll. 54, 315–327. O’Mahony, J., Fox, P.F., 2013. Milk proteins: introduction and historical aspects. In: McSweeney, P.L.H., Fox, P.F. (Eds.), Advanced Dairy Chemistry–1A, Proteins: Basic Aspects, fourth ed. Springer, New York, NY, pp. 43–85. Ould Eleya, M.M., Desobry Banon, S., Hardy, J., 1995. A comparative study of pH and temperature effects on the acidic coagulation of milks from cows, goats, and sheep. J. Dairy Sci. 78, 2675–2682. Park, Y.W., 2008. Goat milk – chemistry and nutrition. In: Park, Y.W., Haenlein, G.F.W. (Eds.), Handbook of Milk of Non-Bovine Mammals. Blackwell Publishing Professional, Oxford, UK, pp. 34–58. Park, Y.W., Haenlein, G.F.W., 2008a. Overview of milk of non-bovine mammals. In: Park, Y.W., Haenlein, G.F.W. (Eds.), Handbook of Milk of Non-Bovine Mammals. Blackwell Publishing Professional, Oxford, UK, pp. 3–9. Park, Y.W., Haenlein, G.F.W., 2008b. Therapeutic and hypoallergenic values of goat milk and implication of food allergy. In: Park, Y.W., Haenlein, G.F.W. (Eds.), Handbook of Milk of Non-Bovine Mammals. Blackwell Publishing Professional, Oxford, UK, pp. 121–135. Park, Y.W., Juarez, M., Ramos, M., Haenlein, G.F.W., 2007. Physico-chemical characteristics of goat and sheep milk. Small Rumin. Res. 68, 88–113. Patton, J.S., Carey, M.C., 1981. Inhibition of human pancreatic lipase-colipase activity by mixed bile salt-phospholipid micelles. Am. J. Physiol. Gastrointest. Liver Physiol. 241, G328–G336. Peram, M.R., Loveday, S.M., Ye, A., Singh, H., 2013. In vitro gastric digestion of heat-induced aggregates of β-lactoglobulin. J. Dairy Sci. 96, 63–74. Raynal-Ljutovac, K., Lagriffoul, G., Paccard, P., Guillet, I., Chilliard, Y., 2008. Composition of goat and sheep milk products: an update. Small Rumin. Res. 79, 57–72.

References

699

Reddy, I.M., Kella, N.K., Kinsella, J.E., 1988. Structural and conformational basis of the resistance of beta-lactoglobulin to peptic and chymotryptic digestion. J. Agric. Food Chem. 36, 737–741. Reid, I.R., 2004. Effects of calcium supplementation on circulating lipids. Drugs Aging 21, 7–17. Ribeiro, A.C., Ribeiro, S.D.A., 2010. Specialty products made from goat milk. Small Rumin. Res. 89, 225–233. Rutherfurd, S.M., Darragh, A.J., Hendriks, W.H., Prosser, C.G., Lowry, D., 2006. True ileal amino acid digestibility of goat and cow milk infant formulas. J. Dairy Sci. 89, 2408–2413. Salami, M., Yousefi, R., Ehsani, M.R., Dalgalarrondo, M., Chobert, J.-M., Haertl
e, T., Razavi, S.H., Saboury, A.A., Niasari-Naslaji, A., Moosavi-Movahedi, A.A., 2008. Kinetic characterization of hydrolysis of camel and bovine milk proteins by pancreatic enzymes. Int. Dairy J. 18, 1097–1102. Sarkar, A., Ye, A., Singh, H., 2016. On the role of bile salts in the digestion of emulsified lipids. Food Hydrocoll. 60, 77–84. Schmidt, D., van Marwijk, B., 1993. Enzymatic hydrolysis of whey proteins. Influence of heat treatment of alphalactalbumin and beta-lactoglobulin on their proteolysis by pepsin and papain. Neth. Milk Dairy J. 47, 15–22. Schuman, A.J., 2003. A concise history of infant formula (twists and turns included). Contemp. Pediatr. 20, 91–98. Sheehan, W.J., Phipatanakul, W., 2009. Tolerance to water buffalo milk in a child with cow milk allergy. Ann. Allergy Asthma Immunol. 102, 349. Singh, H., 2019. Symposium review: fat globules in milk and their structural modifications during gastrointestinal digestion. J. Dairy Sci. 102, 2749–2759. Singh, H., Waungana, A., 2001. Influence of heat treatment of milk on cheesemaking properties. Int. Dairy J. 11, 543–551. Singh, H., Ye, A., 2013. Structural and biochemical factors affecting the digestion of protein-stabilized emulsions. Curr. Opin. Colloid Interface Sci. 18, 360–370. Singh, H., Ye, A., Horne, D., 2009. Structuring food emulsions in the gastrointestinal tract to modify lipid digestion. Prog. Lipid Res. 48, 92–100. Stevens, E.E., Patrick, T.E., Pickler, R., 2009. A history of infant feeding. J. Perinat. Educ. 18, 32. Tagliazucchi, D., Martini, S., Shamsia, S., Helal, A., Conte, A., 2018. Biological activities and peptidomic profile of in vitro-digested cow, camel, goat and sheep milk. Int. Dairy J. 81, 19–27. Tam, J.J., Whitaker, J.R., 1972. Rates and extents of hydrolysis of several caseins by pepsin, rennin, endothia parasitica protease and mucor pusillus protease. J. Dairy Sci. 55, 1523–1531. Tan, H., Zhao, L., Tian, S., Wen, H., Gou, X., Ngai, T., 2017. Gelatin particle-stabilized high internal phase emulsions for use in oral delivery systems: protection effect and in vitro digestion study. J. Agric. Food Chem. 65, 900–907. Tari, N.R., Fan, M.Z., Archbold, T., Kristo, E., Guri, A., Arranz, E., Corredig, M., 2018. Effect of milk protein composition of a model infant formula on the physicochemical properties of in vivo gastric digestates. J. Dairy Sci. 101, 2851–2861. Tunick, M.H., Ren, D.X., Van Hekken, D.L., Bonnaillie, L., Paul, M., Kwoczak, R., Tomasula, P.M., 2016. Effect of heat and homogenization on in vitro digestion of milk. J. Dairy Sci. 99, 4124–4139. Turgeon, S.L., Rioux, L.-E., 2011. Food matrix impact on macronutrients nutritional properties. Food Hydrocoll. 25, 1915–1924. Ul Haq, M.R., Kapila, R., Shandilya, U.K., Kapila, S., 2014. Impact of milk derived β-casomorphins on physiological functions and trends in research: a review. Int. J. Food Prop. 17, 1726–1741. Uniacke-Lowe, T., Huppertz, T., Fox, P.F., 2010. Equine milk proteins: chemistry, structure and nutritional significance. Int. Dairy J. 20, 609–629. Vaisman, N., Reuven, Y., Uzi, M., Georgi, G., Boehm, G., 2006. Camel’s milk and gastric emptying. Clin. Nutr. 25, 622–625. van Aken, G.A., 2010. Relating food emulsion structure and composition to the way it is processed in the gastrointestinal tract and physiological responses: what are the opportunities? Food Biophys. 5, 258–283. Vasbinder, A.J., Van Mil, P.J., Bot, A., De Kruif, K.G., 2001. Acid-induced gelation of heat-treated milk studied by diffusing wave spectroscopy. Colloids Surf. B: Biointerfaces 21, 245–250. Vaskonen, T., 2003. Dietary minerals and modification of cardiovascular risk factors. J. Nutr. Biochem. 14, 492–506. Vithana, N.L.O., Mason, S.L., Morton, J.D., Bekhit, A.E.A., 2012. In vitro digestion of red deer (Cervus elaphus) and cow (Bos taurus) milk. Int. Food Res. J. 19, 1367–1374. Wada, Y., L€ onnerdal, B., 2014. Effects of different industrial heating processes of milk on site-specific protein modifications and their relationship to in vitro and in vivo digestibility. J. Agric. Food Chem. 62, 4175–4185.

700

19. Structural changes to milk protein products during gastrointestinal digestion

Wang, X., Ye, A., Lin, Q., Han, J., Singh, H., 2018. Gastric digestion of milk protein ingredients: study using an in vitro dynamic model. J. Dairy Sci. 101, 6842–6852. Wickham, M., Garrood, M., Leney, J., Wilson, P.D., Fillery-Travis, A., 1998. Modification of a phospholipid stabilized emulsion interface by bile salt: effect on pancreatic lipase activity. J. Lipid Res. 39, 623–632. Ye, A., Anema, S.G., Singh, H., 2004. High-pressure–induced interactions between milk fat globule membrane proteins and skim milk proteins in whole milk. J. Dairy Sci. 87, 4013–4022. Ye, A., Anema, S.G., Singh, H., 2008. Changes in the surface protein of the fat globules during homogenization and heat treatment of concentrated milk. J. Dairy Res. 75, 347–353. Ye, A., Cui, J., Singh, H., 2010. Effect of the fat globule membrane on in vitro digestion of milk fat globules with pancreatic lipase. Int. Dairy J. 20, 822–829. Ye, A., Cui, J., Singh, H., 2011. Proteolysis of milk fat globule membrane proteins during in vitro gastric digestion of milk. J. Dairy Sci. 94, 2762–2770. Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016a. The formation and breakdown of structured clots from whole milk during gastric digestion. Food Funct. 7, 4259–4266. Ye, A., Cui, J., Dalgleish, D., Singh, H., 2016b. Formation of a structured clot during the gastric digestion of milk: impact on the rate of protein hydrolysis. Food Hydrocoll. 52, 478–486. Ye, A., Cui, J., Dalgleish, D., Singh, H., 2017. Effect of homogenization and heat treatment on the behavior of protein and fat globules during gastric digestion of milk. J. Dairy Sci. 100, 36–47. Ye, A., Liu, W., Cui, J., Kong, X., Roy, D., Kong, Y., Han, J., Singh, H., 2019. Coagulation behaviour of milk under gastric digestion: effect of pasteurization and ultra-high temperature treatment. Food Chem. 286, 216–225. Zeece, M., Huppertz, T., Kelly, A., 2008. Effect of high-pressure treatment on in-vitro digestibility of β-lactoglobulin. Innov. Food Sci. Emerg. Technol. 9, 62–69. Zenebe, T., Ahmed, N., Kabeta, T., Kebede, G., 2014. Review on medicinal and nutritional values of goat milk. Acad. J. Nutr. 3, 30–39.

C H A P T E R

20 Milk proteins: Digestion and absorption in the gastrointestinal tract Didier Duponta, Daniel Tomeb a

STLO, INRA, Agrocampus Ouest, Rennes, France bPNCA, INRA, AgroParisTech, Paris, France

Introduction In the industrialized world, dairy products constitute an important part of the diet, especially in northern Europe and North America. In these regions, milk products contribute around 30% of the total dietary protein supply and represent about 65% of the intake of animal protein. The protein content of cow’s milk ranges from 32 to 35 g/L. There are two major types of milk protein: the caseins (80%), which are represented by four distinct proteins (αs1-, αs2-, β-, and κ-caseins), and the whey proteins (20%), which are represented by proteins such as β-lactoglobulin, α-lactalbumin, and lactoferrin. These two families of proteins are opposite in terms of structure. Caseins exhibit a loose and highly flexible structure and are associated into a supramolecular structure called the micelle, whereas whey proteins have a globular, well-defined three-dimensional structure. These structural differences between the two families markedly affect the behavior of these proteins in the gastrointestinal tract and particularly their susceptibility to hydrolysis by the digestive enzymes. The nutritive value of proteins, including milk proteins, is generally associated with their capacity to provide two components: nitrogen (related to protein quantity) and essential amino acids (related to protein quality). The overall nutritional efficiency of protein is most commonly measured via nitrogen retention, which assesses protein retention. In terms of protein quality, the nutritive value is related to the amino acid composition and the bioavailability of these amino acids. The content and the bioavailability of indispensable amino acids, that is, those that cannot be synthesized in the body and consequently must be supplied through the diet, are of particular concern.

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00020-7

701

# 2020 Elsevier Inc. All rights reserved.

702

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

TABLE 20.1 Fecal versus ileal digestibility (%) of milk proteins in humans Ileal

Protein

Fecal True

Apparent

True

References

Milk protein

96.6

91

95

Bos et al. (2003), Gaudichon et al. (2002), Mahe et al. (1994)

Fermented milk

–

90

–

Mahe et al. (1994)

Casein

–

–

94.1

Deglaire et al. (2009)

Digestion of milk proteins In the evaluation of the nutritive value of dietary proteins, nitrogen and individual amino acid digestibility, ileal and fecal digestibility, and apparent and true digestibility should be considered (Fuller and Tome, 2005). The true digestibility of milk protein, as measured in the ileum, averages 95%, which corresponds to one of the highest digestibilities for dietary proteins (Table 20.1). The ileal digestibility of caseins has been estimated to be around 93% in pigs and 94% in humans; that of whey proteins appears to be even higher (97%–98%) (Gilani and Sepehr, 2003; Rutherfurd and Moughan, 2003; Lacroix et al., 2006) but has never been precisely assessed in humans. Measurement of the true digestibility values of dietary nitrogen and amino acids in healthy human volunteers after the ingestion of milk indicated that ileal digestibility values for the individual amino acids ranged from 92% for serine to 99% for tyrosine, with an average amino acid digestibility of 95.3% (Gaudichon et al., 2002), that is, the same value as for nitrogen digestibility. In vitro measurement of protein nitrogen and amino acid digestion showed that the digestibility was not different between a goat milk–based infant formula, a cow milk– based infant formula, and human milk (78.3%
3.7%, 73.4%
2.7%, and 77.9%
4.1%, respectively) (Maathuis et al., 2017). In the context of digestion, although caseins exhibit a structure that makes them highly sensitive to hydrolysis by digestive enzymes, they are considered to be “slow proteins” (Boirie et al., 1997) because they cause a slow postprandial release of amino acids in the plasma. This contrasts with whey proteins, which rapidly give rise to an intense peak of amino acids in the plasma. This property of caseins has been attributed to their ability to form a coagulum in the stomach through the joint action of acidic secretions and digestive enzymes (Wang et al., 2018). This coagulum of caseins is retained in the stomach for longer than the whey proteins, which remain soluble and are rapidly delivered from the stomach into the small intestine. These differences in gastric emptying lead to differences in the rate at which dietary amino acids enter the bloodstream (Mahe et al., 1996; Lacroix et al., 2006). The longer retention of caseins in the stomach leads to a lower level but a longer persistence of amino acids in the plasma than is observed for amino acids from whey proteins (Fig. 20.1). The type of protein can also specifically influence postmeal aminoacidemia. The chemical composition of whey protein is characterized by high leucine and isoleucine contents, and its ingestion is followed by a peripheral plasma elevation of these amino acids, which are known to be poorly oxidized in the liver. Similarly, the higher plasma proline concentration observed after the ingestion of casein is due to the higher proline content of this fraction (Lacroix et al., 2006).

Milk protein hydrolysis in the intestinal lumen

703

FIG. 20.1 Mean (
standard deviation) changes from baseline in serum total amino acid (AA), indispensable AA (IAA), branched-chain AA (BCAA), and dispensable AA (DAA) concentrations in subjects after the ingestion of total milk protein (TMP; n ¼ 8), micellar casein (MC; n ¼ 8), and milk-soluble protein isolate (MSPI; n ¼ 7). A significant effect of time (P ¼ 0.0001) and a significant meal-by-time interaction (P ¼ 0.01) were observed for all variables, as tested on the crude values using a mixed-model analysis of variance (ANOVA) with time as a repeated measure. *, **, *** Significantly different from baseline: *P ¼ 0.05, **P ¼ 0.01, ***P ¼ 0.005 (Lacroix et al., 2006).

Milk protein hydrolysis in the intestinal lumen Caseins Caseins are extensively degraded during the gastric phase of digestion, an observation that is consistent with the fact that pepsin has a preference for mobile, loosely structured polypeptides. All in vitro studies on purified proteins have clearly demonstrated that caseins are hydrolyzed within the first minutes of pepsin hydrolysis, in adults and in infants (Fig. 20.2). More recently, an animal trial conducted on minipigs fed skim milk and yogurt also showed a rapid and extensive hydrolysis of caseins during the first minutes of digestion, with intact caseins being detected for only 20 min after the meal intake (Barbe et al., 2013).

704

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

FIG. 20.2 Evolution of residual immunoreactivity of (A) β-lactoglobulin and (B) β-casein during in vitro gastric (left) and duodenal (right) digestion using an infant model (□) and an adult model (▪), as determined by inhibition enzyme-linked immunosorbent assay (data are the results of three independent determinations made in duplicate).

Similarly, in piglets fed milk-based infant formulas, caseins were shown to be more rapidly hydrolyzed than whey proteins, with only 23% of intact caseins being present in the stomach 30 min after ingestion (Bouzerzour et al., 2012). An in vivo study on human volunteers fed caseins showed extensive release of mediumsized peptides (750–1050 kDa) in the jejunum during the first 6 h after the meal intake. Most of the identified peptides originated from the two major caseins, that is, β-casein (61%) and αs1-casein (25%), and most contained two or more proline residues; the largest contained seven proline residues out of the 26 residues in its sequence (Boutrou et al., 2013). This is in agreement with the generally reported resistance of proline-containing peptides to gastric and pancreatic digestive enzymes (Vanhoof et al., 1995; Agudelo et al., 2004) and to epithelial proteases (Bauchart et al., 2007). Protein degradation in human jejuna after oral ingestion of casein was also compared with the digests of the same substrate using a standardized in vitro protocol; no intact casein was detected, neither in the jejuna nor in the in vitro samples taken during the intestinal phase (Sanchon et al., 2018). The digestions of cow milk proteins and goat milk proteins were compared in vitro by varying pH, enzyme concentrations, and incubation times to simulate infant and young child gastric conditions (Hodgkinson et al., 2018). Caseins reacted to pH changes differently from whey proteins, with less digestion of casein at pH 3.0 than at pH 5.0. Caseins from goat milk tended to be more efficiently digested than caseins from cow milk, and the peptide profiles from goat milk were distinct from those from cow milk.

Peptides released during digestion

705

Whey proteins In contrast to caseins, whey proteins, because of their globular structure, are known to be extremely resistant to proteolysis. This is particularly the case for β-lactoglobulin, which is not affected during gastric digestion, being virtually unaltered after 60 min of simulated digestion (Fig. 20.2A) (Schmidt et al., 1995; Mandalari et al., 2009; Dupont et al., 2010a). It has also been shown that the molecular interaction of β-lactoglobulin with phosphatidylcholine from the gastric mucosa protects the protein from duodenal digestion by trypsin and chymotrypsin (Mandalari et al., 2009). However, β-lactoglobulin has been found to be more sensitive to pepsinolysis when located at the interface of a lipid droplet than when in solution, because of drastic conformational changes (Macierzanka et al., 2009). A comparison of the standardized in vitro digestion model (Infogest) with in vivo digestion data from human jejuna showed that intact β-lactoglobulin was visible both in the samples taken at 1 h in human digestion and in the in vitro gastrointestinal simulation (Sanchon et al., 2018). When the digestions of cow and goat milk proteins were compared in vitro to simulate infant and young child gastric conditions, digestion of higher molecular weight whey proteins increased with decreasing pH and higher enzyme concentrations of the young child gastric digestion conditions compared with the infant conditions, and β-lactoglobulin was poorly digested under all gastric digestion conditions (Hodgkinson et al., 2018). Conflicting results have been published for the second major bovine whey protein, that is, α-lactalbumin. Whereas some have found that α-lactalbumin is even more resistant to simulated digestion than β-lactoglobulin (Inglingstad et al., 2010), others have found that α-lactalbumin is susceptible to hydrolysis in solution (Nik et al., 2010). In contrast to β-lactoglobulin, α-lactalbumin appears to be more resistant to digestion when located at the oil-water interface than when in solution. The same protective effect of phospholipids from the gastric mucosa on the susceptibility of α-lactalbumin to hydrolysis by pancreatic enzymes has been described (Moreno et al., 2005). More recently, an in vivo study on human milk–fed preterm neonates showed that α-lactalbumin was the human milk protein that was most resistant to gastric digestion in aspirates collected through a nasogastric probe (de Oliveira et al., 2017). In contrast to α-lactalbumin and β-lactoglobulin, lactoferrin has been shown to be extensively degraded during simulated gastric digestion (Furlund et al., 2013). Multiple sequence analyses of the identified peptides indicated a motif consisting of proline and neighboring hydrophobic residues that could restrict proteolytic processing. Further structure analysis showed that almost all proteolytic cleavage sites were located on the surface and mainly on the nonglycosylated half of lactoferrin.

Peptides released during digestion The hydrolysis of milk proteins in the gastrointestinal tract will result in the production of a myriad of peptides ( Jahan-Mihan et al., 2011), some of which have been shown to exert biological activities such as antihypertensive (Martinez-Maqueda et al., 2012), antiatherogenic (Ricci-Cabello et al., 2012), antimicrobiological, and immunomodulatory (Agyei and Danquah, 2012). Mass spectrometry is the best tool for tracking peptides released during digestion, and the concept of nutritional peptidomics has recently been proposed (Panchaud et al., 2012).

706

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

To date, milk peptides have been identified by submitting food to simulated in vitro digestion (Dupont et al., 2010b; Picariello et al., 2010; Kopf-Bolanz et al., 2012). In the in vitro situation, identifying and quantifying the peptides in digested samples is rather easy because most of the proteins in the samples originate from the food itself. However, it is still questionable whether mimicking digestion with in vitro models perfectly reflects the physiological reality. Only a few in vivo studies have been conducted; detection of dietary peptides is made more difficult by the presence of endogenous proteins secreted in the different compartments of the gut. In a pioneer work, milk caseinomacropeptide, that is, κ-casein (f106–169), was detected in the jejunum of humans fed 15N-labeled casein, whey protein, and yogurt (Ledoux et al., 1999). Similarly, caseinophosphopeptides were identified in the effluent collected from milk-fed humans (Meisel et al., 2003). In 2013, the peptidome of jejunal effluents collected from milk protein–fed humans was established (Boutrou et al., 2013). Totals of 356 and 146 peptides were detected and sequenced in the jejunum following casein and whey protein ingestion, respectively. Protein degradation and peptide release in human jejuna after oral ingestion of casein were compared with the digests of the same substrate using a standardized in vitro protocol (Sanchon et al., 2018). In vivo and in vitro digests showed comparable peptide profiles and a high number of common sequences. Most of the sequences found in the jejunum, some of them not previously described, were also identified in the simulated digests. Common regions that were resistant to digestion were identified, revealing that the in vitro protocol constitutes a good approximation to the physiological gastrointestinal digestion of milk proteins. However, the subjects were fed pure protein fractions, and the possible effect of the food structure was not investigated. Indeed, the structure of the chyme, as affected by food structure, could limit or modify the accessibility of digestive enzymes to some cleavage sites. When the digestions of cow and goat milk proteins were compared in vitro to simulate the infant and young child gastric conditions, the caseins from goat milk tended to be more efficiently digested than the caseins from cow milk, and the peptide profiles from goat milk were distinct from those from cow milk (Hodgkinson et al., 2018). Generated peptides, identified using liquid chromatography coupled to a mass spectrometer, showed both similarities and differences in the cow milk and goat milk postdigestion profiles (Hodgkinson et al., 2019). The majority of peptides were from casein proteins, 50% representing β-casein, with many peptides unique to each species. Low and no peptides from β-lactoglobulin and α-lactalbumin, respectively, suggest that these proteins were highly resistant to infant gastric digestion, as reported by others. Minor milk proteins, comprising 5% peptides, were represented by different proteins from the cow and the goat. Peptides with known bioactivities were also observed, both in common and unique to each species. More recently, the gastric and intestinal peptidomes of raw or pasteurized human milk, digested using a preterm neonate dynamic digestion model, were characterized exhaustively (Deglaire et al., 2019); 1531 peptides were clearly identified and arose from 27 different proteins. β-Casein was shown to be the protein that elicited the highest number of peptides. Surprisingly, the most major human whey proteins, that is, α-lactalbumin and lactoferrin, led to the release of a small number of peptides, confirming their resistance to gastric digestion. Similar work was conducted in vivo, that is, by analyzing the content of gastric aspirates collected from infants (Dallas et al., 2014). The results showed that around 200 peptides were already present in human milk before ingestion, indicating that proteolysis events caused by

Impact of processing on milk protein digestion and absorption

707

indigenous proteases such as plasmin were occurring in milk during expression and storage. Then 649 peptides were identified during digestion in the stomach, and as for the in vitro experiment, β-casein was shown to be the major source of peptides. Finally, bioactive peptides known to carry various biological activities, such as immunomodulating, opioid, antihypertensive, and antibacterial activities, were shown to be released during digestion and were present in the stomachs of the infants. In all the previous examples, peptides released from foods of different compositions were identified but not peptides released from foods of identical compositions but different structures. In recent work, the impact of the structure of the dairy matrix on the number and nature of milk peptides released in the duodenum was investigated using cannulated minipigs fed a dairy liquid, an acid dairy gel, or a rennet dairy gel, all of identical composition. The formation of peptides in vivo was followed by tandem mass spectrometry over a postprandial period of 5 h after the ingestion of the dairy matrices by the minipigs. The effect of the meal structure was investigated at two levels: the microstructure, as modified by thermal treatment, and the macrostructure, as modified by milk gelation. More than 16,000 peptides were sequenced and unambiguously identified. The results obtained showed that the structure of the dairy products had only little influence on the location of the cleavage sites on the protein sequences (Barbe et al., 2014a). However, the structure markedly impacted the number of peptides identified, especially for the rennet dairy gel; about three times fewer peptides were detected than for the other matrices. This effect was attributed to greater extents of dilution by digestive secretions associated with longer gastric retentions for the rennet gel. Potential bioactive peptides were also produced over time, and their identification has increased our knowledge of the peptides present in the lumen in vivo. Our results indicate that the structure of dairy matrices markedly affects the kinetics of milk protein digestion in vivo, more than the mechanism of proteolysis itself.

Impact of processing on milk protein digestion and absorption Milk proteins are introduced into the human diet as processed milk products. It is therefore critical to determine the impact of the major processing technologies on milk protein digestion.

Heat treatment of milk One of the most common processes applied to milk is heat treatment to ensure product safety. Because of the structural differences already mentioned, heat treatment affects caseins and whey proteins quite differently. Heat treatment modifies the three-dimensional structure of the whey proteins markedly, resulting in an “opening” of the globular structure and making the whey proteins more sensitive to the action of digestive enzymes, as demonstrated for β-lactoglobulin (Barbe et al., 2013) and α-lactalbumin (Inglingstad et al., 2010). In contrast, caseins, with their loose and highly flexible structure, are not strongly modified by heat treatment. Heat treatment at high temperature results in an increased resistance of the caseins to simulated digestion (Almaas et al., 2006; Dupont et al., 2010b; Barbe et al., 2013), which has

708

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

been attributed to the formation of thermally induced aggregates between caseins and between caseins and whey proteins.

Homogenization of milk Homogenization of milk results in the disruption of the milk fat globule membrane. Lipids are present as smaller droplets that are stabilized by milk proteins covering the oil-water interface. β-Lactoglobulin and β-casein have been shown to be more susceptible to pepsinolysis when they are adsorbed to an oil-water interface than when they are in solution (Macierzanka et al., 2009; Sarkar et al., 2009). This has been attributed to the unfolding of the proteins at the droplet surface, which improves their accessibility to pepsin. It has been found that the rate of gastric digestion of β-casein is twice as fast when it is adsorbed to the oil-water interface than when it is in solution. In the small intestine, proteins are displaced from the interface by bile salts (Sarkar et al., 2010), making triglycerides more accessible to the pancreatic lipase. Recently, Ye et al. (2017) investigated the combined effect of both homogenization and heat treatment on the formation and the breakdown of clots during gastric digestion of whole milk using a dynamic digestion simulator. They showed that these processing conditions led to the formation of a coagulum with fragmented and crumbled structures compared with the coagulum formed from raw whole milk. The combination of homogenization and heat treatment resulted in a greater incorporation of protein and fat globules in the coagulum, leading to the formation of more pores. These pores allowed a better diffusion of the digestive enzymes during a simulated digestion, leading to a greater rate of proteolysis. This can result in an increased bioavailability of amino acids. Change in the structure of the coagulum during gastric digestion is also probably of major importance for the regulation of gastric emptying and the transit of food in the gastrointestinal tract.

Physicochemical modifications of proteins Milk protein modification with cross-linking enzymes such as transglutaminase (TG) has been used extensively to change the functionality of proteins and thereby to improve the textural quality and the stability of protein-based food products. In dairy products, TG-induced cross-linking can increase the firmness and water-holding capacity of acid-induced gels in products with low solids and fat contents or can improve the stability of emulsions and foams. The effect of the TG-induced cross-linking of sodium caseinate on postprandial metabolic and appetite responses was recently investigated in 13 healthy individuals ( Juvonen et al., 2012). The results indicated that enzymatically cross-linked sodium caseinate and native sodium caseinate had comparable metabolic responses in a liquid matrix, suggesting similar digestion and absorption rates and first pass metabolism despite the structural modification of the cross-linked sodium caseinate. The hydrolysis of milk proteins has been widely used to reduce their allergenicity properties in infant nutrition. However, hydrolysis could also be considered to be a “predigestion” of proteins, facilitating their digestion and absorption in the gastrointestinal tract. This was confirmed in a study on 10 elderly subjects who received either intact or hydrolyzed caseins (Koopman et al., 2009). The plasma amino acid concentrations increased extensively (25%–50%) after the ingestion of the hydrolyzed casein, compared with the intact casein (P < 0.01).

Impact of processing on milk protein digestion and absorption

709

Coagulation (liquid/gel/solid transition) of milk Milk coagulation is used extensively in the dairy industry, especially for yogurt and cheese manufacture, even though the mechanisms of milk clotting for these two types of product are quite different. Studies on the digestion of dairy matrices (yogurt and cheese) are scarce, compared with studies on purified fractions of casein or whey protein. Gaudichon et al. (1994) showed, using minipigs, that the half gastric emptying time of the liquid phase was not different between milk and yogurt. However, the intestinal deliveries of both the liquid phase and the nitrogenous fraction of the chyme were more delayed in pigs fed yogurt than in pigs fed milk (Fig. 20.3). The kinetics of exogenous nitrogen delivery into the intestine were correlated with the kinetics of exogenous nitrogen absorption. These results suggest that milk proteins are rapidly absorbed after they reach the intestine and that gastric emptying is a major factor controlling the kinetics of milk nitrogen absorption. Rychen et al. (2002) examined the postprandial portal absorption of 15N in the growing pig after the ingestion of milk, yogurt, and heat-treated yogurt. Although the total portal absorption was similar between the three products, yogurt nitrogen was absorbed more

FIG. 20.3

Remaining fraction of exogenous nitrogen in the stomach of minipigs after the ingestion of 500 mL of milk or 500 g of yogurt. The ingested milk and yogurt contained 17 and 18 g of nitrogen, respectively. Values are means
standard error of the mean (SEM) for three or four pigs. No significant differences were found by ANOVA, P < 0.05.

Milk Yogurt 100

Remaining fraction (% of ingested)

80

60

40

20

0

0

1

2

3

4

5

6

7

Time (h)

8

9

10 11 12

710

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

slowly than milk nitrogen, with significant differences being observed after 30, 60, and 180 min. Heat-treated yogurt showed similar behavior to milk; it was hypothesized that heat treatment of the gel was responsible for destroying the natural body and viscosity of the yogurt. These effects were therefore attributed to different emptying rates between milk, yogurt, and heat-treated yogurt. More recently, a determination of the kinetics of milk protein digestion and amino acid absorption after the ingestion of liquid or gelled (acid and rennet gels) dairy matrices by six minipigs showed that the gelation of milk slowed down the outflow of the meal from the stomach, slowed down the subsequent absorption of amino acids, and decreased their bioavailability in peripheral blood (Fig. 20.4) (Barbe et al., 2014b). The nature of the matrix seemed to affect the release of the gastrointestinal hormones involved in appetite regulation, with the gel matrices appearing to be potentially more satiating. It was also shown that two gels with the same composition and similar rheological and structural properties, but differing in their mode of coagulation (acidification/renneting), exhibited different behaviors during digestion. Indeed, ingestion of the rennet gel resulted in lower levels of both proteins in the duodenum and lower levels of amino acids in the plasma, compared with the ingestion of the acid gel. This was probably due to the formation of a coagulum with high stiffness after the ingestion of the rennet gel, under the simultaneous action of the stomach acidity and the rennet, leading to a very long retention of the rennet matrix in the stomach (Barbe et al., 2014b).

FIG. 20.4 Plasma leucine concentration (μmol/L) in minipigs over a 7-h period after the ingestion of liquid (L) and

gel (G) matrices, from unheated (R) and heated (H) milk products. Values are means
SEM calculated for four minipigs (n ¼ 4). The data were analyzed using a mixed-model ANOVA. The time effect was significant (P < 0.001), and the lines at the bottom of the figure indicate a significant difference (P < 0.05) from baseline for each curve. The time-by-matrix interaction was significant (P < 0.001), and at a given time, differences between matrices are indicated by different letters, a and b (P < 0.05).

Conclusions

711

The plasma cholecystokinin and ghrelin concentrations suggested a potentially more satiating effect of the rennet gel than the acid gel. Studies on the digestion of cheese are scarce. A recent study compared the kinetics of the matrix degradation of different cheeses in a gastrointestinal environment (Lamothe et al., 2012). The relationship between the physical characteristics of the cheeses (rheological properties and microstructure) and their digestion patterns was also studied. Rheological measurements and compositional and microstructural analyses were performed on mild cheddar, aged cheddar, light cheddar, and mozzarella cheeses. Mozzarella cheese showed the highest rate of matrix degradation. Aged cheddar cheese showed rapid degradation during the gastric phase but was more resistant to the duodenal environment. Light cheddar cheese showed the opposite behavior, being highly resistant to the gastric environment; however, it underwent extensive degradation at the end of the duodenal phase. The extent of matrix degradation for mild cheddar cheese was similar to that for mozzarella cheese in the gastric phase but was much lower than that for the other cheeses in the duodenal phase. The results suggest that degradation of the cheeses was driven mainly by their physical characteristics. The production of Parmigiano-Reggiano cheese is closely related to the nutritional quality of the final product; in particular, the high digestibility of its proteins is claimed to be proportional to the ripening stage of the cheese. The effect of the aging of cheese on the kinetics of protein digestion was recently investigated. Two different kinds of Parmigiano-Reggiano, young (aged 15 months) and old (aged 30 months), were separately digested using an in vitro system that simulated digestive processes in the mouth, stomach, and small intestine (Bordoni et al., 2011). The results indicated that the digestion of cheeses with different aging times, although starting from different initial compositions, concluded in similar ways, in terms of free amino acids and small organic compounds, but evolved with different kinetics of hydrolysis and peptide formation, discriminating the young cheese from the old cheese.

Conclusions The digestion and the absorption of milk proteins have been extensively studied, and the mechanisms involved are well described. However, many of the studies have been performed either in vitro or with purified protein fractions, and more work is needed to better understand the disintegration of real dairy products in the human gastrointestinal tract. Nevertheless, it appears that casein and whey protein exhibit different behaviors in the gastrointestinal tract because of differences in their structure and physicochemical properties and that processing has a significant impact on the kinetics of protein digestion by modifying the residence time of the products in the stomach. In the context of the nutritional properties of food, it appears that the micro- and macrostructures of a meal, resulting from technological processes used in the food industry, markedly affect the different steps of milk protein digestion. Thus, the design of food matrices at the technological level is of particular interest in the control of the delivery of nutrients, especially for specific subpopulations, such as the elderly or overweight people.

712

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

References Agudelo, R.A., Gauthier, S.F., Pouliot, Y., Marin, J., Savoie, L., 2004. Kinetics of peptide fraction release during in vitro digestion of casein. J. Sci. Food Agric. 84 (4), 325–332. Agyei, D., Danquah, M.K., 2012. Rethinking food-derived bioactive peptides for antimicrobial and immunomodulatory activities. Trends Food Sci. Technol. 23 (2), 62–69. Almaas, H., Cases, A.L., Devold, T.G., Holm, H., Langsrud, T., Aabakken, L., Aadnoey, T., Vegarud, G.E., 2006. In vitro digestion of bovine and caprine milk by human gastric and duodenal enzymes. Int. Dairy J. 16 (9), 961–968. Barbe, F., Menard, O., Le Gouar, Y., Buffie`re, C., Famelart, M.-H., Laroche, B., Le Feunteun, S., Dupont, D., Remond, D., 2013. The heat treatment and the gelation are strong determinants of the kinetics of milk proteins digestion and of the peripheral availability of amino acids. Food Chem. 136, 1203–1212. Barbe, F., Le Feunteun, S., Remond, D., Menard, O., Jardin, J., Henry, G., Laroche, B., Dupont, D., 2014a. Tracking the in vivo release of bioactive peptides in the gut during digestion: mass spectrometry peptidomic characterization of effluents collected in the gut of dairy matrix fed mini-pigs. Food Res. Int. 63, 147–156. Barbe, F., Menard, O., Gouar, Y.L., Buffie`re, C., Famelart, M.-H., Laroche, B., Feunteun, S.L., Remond, D., Dupont, D., 2014b. Acid and rennet gels exhibit strong differences in the kinetics of milk protein digestion and amino acid bioavailability. Food Chem. 143, 1–8. Bauchart, C., Morzel, M., Chambon, C., Mirand, P.P., Reynes, C., Buffiere, C., Remond, D., 2007. Peptides reproducibly released by in vivo digestion of beef meat and trout flesh in pigs. Br. J. Nutr. 98 (6), 1187–1195. Boirie, Y., Dangin, M., Gachon, P., Vasson, M.P., Maubois, J.L., Beaufrere, B., 1997. Slow and fast dietary proteins differently modulate postprandial protein accretion. Proc. Natl. Acad. Sci. U S A 94 (26), 14930–14935. Bordoni, A., Picone, G., Babini, E., Vignali, M., Danesi, F., Valli, V., Di Nunzio, M., Laghi, L., Capozzi, F., 2011. NMR comparison of in vitro digestion of Parmigiano Reggiano cheese aged 15 and 30 months. Magn. Reson. Chem. 49, S61–S70. Bos, C., Metges, C.C., Gaudichon, C., Petze, K.J., Pueyo, M.E., Morens, C., Everwand, J., Benamouzig, R., Tome, D., 2003. Postprandial kinetics of dietary amino acids are the main determinant of their metabolism after soy or milk protein ingestion in humans. J. Nutr. 133 (5), 1308–1315. Boutrou, R., Gaudichon, C., Dupont, D., Jardin, J., Airinei, G., Marsset-Baglieri, A., Benamouzig, R., Tome, D., Leonil, J., 2013. Sequential release of milk protein-derived bioactive peptides in the jejunum in healthy humans. Am. J. Clin. Nutr. 97 (6), 1314–1323. Bouzerzour, K., Morgan, F., Cuinet, I., Bonhomme, C., Jardin, J., Le Huerou-Luron, I., Dupont, D., 2012. In vivo digestion of infant formula in piglets: protein digestion kinetics and release of bioactive peptides. Br. J. Nutr. 108 (12), 1–10. Dallas, D.C., Guerrero, A., Khaldi, N., Borghese, R., Bhandari, A., Underwood, M.A., Lebrilla, C.B., German, J.B., Barile, D., 2014. A peptidomic analysis of human milk digestion in the infant stomach reveals protein-specific degradation patterns. J. Nutr. 144, 815–820. de Oliveira, S.C., Bellanger, A., Menard, O., Pladys, P., Le Gouar, Y., Dirson, E., Kroell, F., Dupont, D., Deglaire, A., Bourlieu, C., 2017. Impact of human milk pasteurization on gastric digestion in preterm infants: a randomized controlled trial. Am. J. Clin. Nutr. 105 (2), 379–390. Deglaire, A., Bos, C., Tome, D., Moughan, P.J., 2009. Ileal digestibility of dietary protein in the growing pig and adult human. Br. J. Nutr. 102 (12), 1752–1759. Deglaire, A., De Oliveira, S., Jardin, J., Briard-Bion, V., Kroell, F., Emily, M., Menard, O., Bourlieu, C., Dupont, D., 2019. Impact of human milk pasteurization on the kinetics of peptide release during in vitro dynamic digestion at the preterm newborn stage. Food Chem. 281, 294–303. Dupont, D., Mandalari, G., Molle, D., Jardin, J., Leonil, J., Faulks, R.M., Wickham, M.S.J., Mills, E.N.C., Mackie, A.R., 2010a. Comparative resistance of food proteins to adult and infant in vitro digestion models. Mol. Nutr. Food Res. 54 (6), 767–780. Dupont, D., Mandalari, G., Molle, D., Jardin, J., Rolet-Repecaud, O., Duboz, G., Leonil, J., Mills, E.N.C., Mackie, A.R., 2010b. Food processing increases casein resistance to simulated infant digestion. Mol. Nutr. Food Res. 54 (11), 1677–1689. Fuller, M.F., Tome, D., 2005. In vivo determination of amino acid bioavailability in humans and model animals. J. AOAC Int. 88 (3), 923–934. Furlund, C.B., Ulleberg, E.K., Devold, T.G., Flengsrud, R., Jacobsen, M., Sekse, C., Holm, H., Vegarud, G.E., 2013. Identification of lactoferrin peptides generated by digestion with human gastrointestinal enzymes. J. Dairy Sci. 96 (1), 75–88.

References

713

Gaudichon, C., Roos, N., Mahe, S., Sick, H., Bouley, C., Tome, D., 1994. Gastric-emptying regulates the kinetics of nitrogen absorption from N-15-labeled milk and N-15-labeled yogurt in miniature pigs. J. Nutr. 124 (10), 1970–1977. Gaudichon, C., Bos, C., Morens, C., Petzke, K.J., Mariotti, F., Everwand, J., Benamouzig, R., Dare, S., Tome, D., Metges, C.C., 2002. Ileal losses of nitrogen and amino acids in humans and their importance to the assessment of amino acid requirements. Gastroenterology 123 (1), 50–59. Gilani, G.S., Sepehr, E., 2003. Protein digestibility and quality in products containing antinutritional factors are adversely affected by old age in rats. J. Nutr. 133 (1), 220–225. Hodgkinson, A.J., Wallace, O.A.M., Boggs, I., Broadhurst, M., Prosser, C.G., 2018. Gastric digestion of cow and goat milk: impact of infant and young child in vitro digestion conditions. Food Chem. 245, 275–281. Hodgkinson, A.J., Wallace, O.A.M., Smolenski, G., Prosser, C.G., 2019. Gastric digestion of cow and goat milk: peptides derived from simulated conditions of infant digestion. Food Chem. 276, 619–625. Inglingstad, R.A., Devold, T.G., Eriksen, E.K., Holm, H., Jacobsen, M., Liland, K.H., Rukke, E.O., Vegarud, G.E., 2010. Comparison of the digestion of caseins and whey proteins in equine, bovine, caprine and human milks by human gastrointestinal enzymes. Dairy Sci. Technol. 90 (5), 549–563. Jahan-Mihan, A., Luhovyy, B.L., El Khoury, D., Anderson, G.H., 2011. Dietary proteins as determinants of metabolic and physiologic functions of the gastrointestinal tract. Nutrients 3 (5), 574–603. Juvonen, K.R., Lille, M.E., Laaksonen, D.E., Mykkanen, H.M., Niskanen, L.K., Herzig, K.H., Poutanen, K.S., Karhunen, L.J., 2012. Crosslinking with transglutaminase does not change metabolic effects of sodium caseinate in model beverage in healthy young individuals. Nutr. J. 11, 35. https://doi.org/10.1186/1475-2891-11-35. Koopman, R., Crombach, N., Gijsen, A.P., Walrand, S., Fauquant, J., Kies, A.K., Lemosquet, S., Saris, W.H.M., Boirie, Y., Van Loon, L.J.C., 2009. Ingestion of a protein hydrolysate is accompanied by an accelerated in vivo digestion and absorption rate when compared with its intact protein. Am. J. Clin. Nutr. 90 (1), 106–115. Kopf-Bolanz, K.A., Schwander, F., Gijs, M., Vergeres, G., Portmann, R., Egger, L., 2012. Validation of an in vitro digestive system for studying macronutrient decomposition in humans. J. Nutr. 142 (2), 245–250. Lacroix, M., Bos, C., Leonil, J., Airinei, G., Luengo, C., Dare, S., Benamouzig, R., Fouillet, H., Fauquant, J., Tome, D., Gaudichon, C., 2006. Compared with casein or total milk protein, digestion of milk soluble proteins is too rapid to sustain the anabolic postprandial amino acid requirement. Am. J. Clin. Nutr. 84 (5), 1070–1079. Lamothe, S., Corbeil, M.M., Turgeon, S.L., Britten, M., 2012. Influence of cheese matrix on lipid digestion in a simulated gastro-intestinal environment. Food Funct. 3 (7), 724–731. Ledoux, N., Mahe, S., Dubarry, M., Bourras, M., Benamouzig, R., Tome, D., 1999. Intraluminal immunoreactive caseinomacropeptide after milk protein ingestion in humans. Nahrung/Food 43 (3), 196–200. Maathuis, A., Havenaar, R., He, T., Bellmann, S., 2017. Protein digestion and quality of goat and cow milk infant formula and human milk under simulated infant conditions. J. Pediatr. Gastroenterol. Nutr. 65 (6), 661–666. Macierzanka, A., Sancho, A.I., Mills, E.N.C., Rigby, N.M., Mackie, A.R., 2009. Emulsification alters simulated gastrointestinal proteolysis of beta-casein and beta-lactoglobulin. Soft Matter 5 (3), 538–550. Mahe, S., Roos, N., Benamouzig, R., Davin, L., Luengo, C., Gagnon, L., Gausserges, N., Rautureau, J., Tome, D., 1996. Gastrojejunal kinetics and the digestion of N-15 beta-lactoglobulin and casein in humans: the influence of the nature and quantity of the protein. Am. J. Clin. Nutr. 63 (4), 546–552. Mandalari, G., Mackie, A.M., Rigby, N.M., Wickham, M.S., Mills, E., 2009. Physiological phosphatidylcholine protects bovine beta-lactoglobulin from simulated gastrointestinal proteolysis. Mol. Nutr. Food Res. 53, S131–S139. Martinez-Maqueda, D., Miralles, B., Recio, I., Hernandez-Ledesma, B., 2012. Antihypertensive peptides from food proteins: a review. Food Funct. 3 (4), 350–361. Meisel, H., Bernard, H., Fairweather-Tait, S., FitzGerald, R.J., Hartmann, R., Lane, C.N., McDonagh, D., Teucher, B., Wal, J.M., 2003. Detection of caseinophosphopeptides in the distal ileostomy fluid of human subjects. Br. J. Nutr. 89 (3), 351–358. Moreno, F.J., Mackie, A.R., Mills, E.N.C., 2005. Phospholipid interactions protect the milk allergen α-lactalbumin from proteolysis during in vitro digestion. J. Agric. Food Chem. 53 (25), 9810–9816. Nik, A.M., Wright, A.J., Corredig, M., 2010. Surface adsorption alters the susceptibility of whey proteins to pepsindigestion. J. Colloid Interface Sci. 344 (2), 372–381. Panchaud, A., Affolter, M., Kussmann, M., 2012. Mass spectrometry for nutritional peptidomics: how to analyze food bioactives and their health effects. J. Proteome 75 (12), 3546–3559. Picariello, G., Ferranti, P., Fierro, O., Mamone, G., Caira, S., Di Luccia, A., Monica, S., Addeo, F., 2010. Peptides surviving the simulated gastrointestinal digestion of milk proteins: biological and toxicological implications. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 878 (3–4), 295–308.

714

20. Milk proteins: Digestion and absorption in the gastrointestinal tract

Ricci-Cabello, I., Herrera, M.O., Artacho, R., 2012. Possible role of milk-derived bioactive peptides in the treatment and prevention of metabolic syndrome. Nutr. Rev. 70 (4), 241–255. Rutherfurd, S.M., Moughan, P.J., 2003. The rat as a model animal for the growing pig in determining ileal amino acid digestibility in soya and milk proteins. J. Anim. Physiol. Anim. Nutr. 87 (7–8), 292–300. Rychen, G., Mpassi, D., Jurjanz, S., Mertes, M., Lenoir-Wijnkoop, I., Antoine, J.M., Laurent, F., 2002. N-15 as a marker to assess portal absorption of nitrogen from milk, yogurt and heat-treated yogurt in the growing pig. J. Dairy Res. 69 (1), 95–101. Sanchon, J., Fernandez-Tome, S., Miralles, B., Hernandez-Ledesma, B., Tome, D., Gaudichon, C., Recio, I., 2018. Protein degradation and peptide release from milk proteins in human jejunum. Comparison with in vitro gastrointestinal simulation. Food Chem. 239, 486–494. Sarkar, A., Goh, K.K.T., Singh, R.P., Singh, H., 2009. Behaviour of an oil-in-water emulsion stabilized by betalactoglobulin in an in vitro gastric model. Food Hydrocoll. 23 (6), 1563–1569. Sarkar, A., Horne, D.S., Singh, H., 2010. Interactions of milk protein-stabilized oil-in-water emulsions with bile salts in a simulated upper intestinal model. Food Hydrocoll. 24 (2–3), 24. Schmidt, D.G., Meijer, R., Slangen, C.J., Vanberesteijn, E.C.H., 1995. Raising the pH of the pepsin-calayzed hydrolysis of bovine whey proteins increases the antigenicity of the hydrolysates. Clin. Exp. Allergy 25 (10), 1007–1017. Vanhoof, G., Goossens, F., Demeester, I., Hendriks, D., Scharpe, S., 1995. Proline motifs in peptides and their biological processing. FASEB J. 9 (9), 736–744. Wang, X., Ye, A., Lin, Q., Han, J., Singh, H., 2018. Gastric digestion of milk protein ingredients: study using an in vitro dynamic model. J. Dairy Sci. 101, 6842–6852. Ye, A., Cui, J., Dalgleish, D., Singh, H., 2017. Effect of homogenization and heat treatment on the behavior of protein and fat globules during gastric digestion of milk. J. Dairy Sci. 100, 36–47.

Further reading Mahe, S., Gaudichon, C., Roos, N., Benamouzig, R., Luengo, C., Bouley, C., Tome, D., 1994a. N-15-labeled milk and yogurt digestion and absorption in the human jejunum. FASEB J. 8 (5), A714. Mahe, S., Roos, N., Benamouzig, R., Sick, H., Baglieri, A., Huneau, J.F., Tome, D., 1994b. True exogenous and endogenous nitrogen fractions in the human jejunum after ingestion of small amounts of N-15-labeled casein. J. Nutr. 124 (4), 548–555.

C H A P T E R

21 Milk proteins: The future Mike Boland The Riddet Institute, Massey University, Palmerston North, New Zealand

Introduction As a finale in the transit from the expression of milk proteins to their uses in food, this chapter takes a high level look at the possible future of food, especially as it relates to milk proteins. We consider global macroenvironmental factors first and then examine consumer demands and trends and the likely impact of new technologies.

Global issues for food The demand for high-quality protein for nutrition, and for dairy protein in particular, has been discussed in the first chapter of this volume. The Food and Agriculture Organization (FAO) has predicted a demand for dairy production by 2050 of 843 million tonnes, more than double that of today. To achieve this sustainably, the industry will need to address a range of resource constraints. Global resource issues that are expected to have a major impact on future food production, including the production of milk proteins, are • competition for land use, • effects of global warming on production, • increasing cost of energy—primarily not only because of the greenhouse gas implications of energy use but also because of the rising cost of energy production, • the water economy, • methane emissions from cows and the effect on global warming. The position of animal production in a constrained world has been discussed in Chapter 1 of this volume and is covered in more depth by Steinfeld et al. (2006), FAO (2009), and Godfray et al. (2010).

Milk Proteins https://doi.org/10.1016/B978-0-12-815251-5.00021-9

715

# 2020 Elsevier Inc. All rights reserved.

716

21. Milk proteins: The future

Competition for land use An important consideration when discussing competition for land use is the opportunity cost incurred in land used for milk production, including land used for growing crops to feed cows. If cows are fed using arable crops, the equation is simple: the opportunity cost is the value of the crop that has been fed to cows that could otherwise have been used to feed humans. In the case of pasture grazing, the consideration changes to one of the value of competing uses for land (e.g., Godfray et al., 2010). In some cases, when the land is unsuitable for arable farming or horticulture, for example, because the terrain is not suitable, the opportunity cost is low. In cases where the land is suitable for cropping, a comparison has to be made between the value of the crop and the value of the equivalent milk production from that land. This comes down to the purpose of production. If the purpose is simply production of energy, a cereal crop will always come out best. The global need for protein nutrition has been addressed in Chapter 1. If the purpose of farming is to produce protein, our calculations based on product yields and composition suggest that, under the right circumstances, the yield of bioavailable protein per hectare per annum can be at least as high for milk as it is for a cereal crop such as wheat, once both are converted into an edible format. Pastoral farming can be relatively efficient, partly because of the perennial nature and low maintenance costs of pasture and partly because the cow is doing the work of harvesting and refertilizing the pasture.

Effects of global warming on production Modern dairy cows (Bos taurus) have evolved to function best in cool to temperate climates, and most farming of these cows occurs in regions with such climates. It is generally accepted that, above a critical temperature of 25–26°C, cows eat less, production decreases, and fertility is lower (Kadzere et al., 2002). In areas where temperatures are marginally high, responses include changed calving patterns, night feeding, use of shade, and, in some cases, cooling by spraying cows with water on hot days. None of these would be expected to have a major effect on the milk proteins. A longer-term solution is either to crossbreed or to change the breed to buffalo (Bos indicus) strains, which are more heat tolerant (Hansen, 2004). Buffalo milk has significant differences from cows’ milk, including less-stable casein micelles and different cheesemaking and buttermaking properties (Ganguli, 1978, 1979).

Milk and energy Milk is energetically very expensive. Milk is an animal product: to produce it requires the cow to eat vegetable material that has already been produced in a nutritional format— although not in a format that is necessarily edible by humans. However, milk is one of the most efficiently produced of the animal-produced foods—largely because the animal itself is not consumed. It has been estimated that the production of 50 kg of milk protein in the United States requires 7  106 kcal of feed energy (i.e., 585 kJ/kg), an energy efficiency of 30:1 (Pimentel and Pimentel, 1979). In contrast, the total energy input per kilogram for the production of corn or soy protein in the United States is estimated to be 58 kJ/kg (calculated from data in Pimentel and Pimentel, 1979), that is, one-tenth of the energy. These figures do not take into account the uptake of direct solar energy through photosynthesis as the crops

717

Global issues for food

grow or the opportunity cost in energy for other use of the land used to grow these products. It is generally accepted that, for grain-fed beef, about 10 kg of grain is required to produce 1 kg of meat. Included in this calculation is the requirement to replace the animal and a loss of around 50% of carcass mass as waste (or by-products not included in the calculation). It has been estimated that the energy efficiency for milk is 4:1 and that the protein efficiency is 4.75:1 (Council for Agricultural Science and Technology, quoted in Fairlie, 2010). More recent calculated values of the efficiency of the conversion of grain and forage into animal protein have been provided by Pimentel (2006), and selected values are given in Table 21.1. A significant difference between the input for grass-fed beef production and that for grain-fed beef production is evident; presumably, a similar difference can be expected for dairy production, although the conversion factor presented is a hybrid. Energy sensitivity in some markets, particularly in Europe, has resulted in the use of “food miles”—an inappropriately named descriptor of the carbon footprint (i.e., the energy cost) that is expended in producing, distributing, and consuming foods. The methods used in calculating carbon footprints have varied, sometimes leading to inappropriate comparisons. Calculations are becoming more accurate and more meaningful as the methodology becomes standardized but are also open to misuse as nontariff barriers in some jurisdictions. Most food products are shipped by sea and the greenhouse gas component of shipping is small compared with production costs, even when food is shipped long distances. For example, in shipping from New Zealand to the United Kingdom, the contribution of CO2 from the shipping was estimated to be 125 kg CO2/tonne milk solids out of a total of 1422 kg CO2/tonne for the life cycle footprint, which in turn compared favorably with the equivalent figure of 2921 kg CO2/tonne milk solids for the locally produced equivalent in the United Kingdom (Saunders et al., 2006). It is becoming increasingly recognized that sea freight is a minor part of the carbon footprint of a food product, typically 5%–10% and that low carbon footprints through more efficient production systems can more than offset this.

Milk and the water economy Increasingly, international attention is being paid to the “water economy” as water becomes a limiting resource in many regions because of the effects of population growth and of climate change. The amount of “virtual water” in a product is the amount of water that TABLE 21.1 Energy efficiency of conversion of animal feed to animal protein Livestock

Grain (kg)

Forage (kg)

Energy input-energy protein

Lamb

21

30

57:1

Beef cattle

13

30

40:1

Grass-fed beef cattle

–

200

20:1

Swine

5.9

–

14:1

Dairy (milk protein)

0.7

1

14:1

Data from Pimentel, D., 2006. Impacts of organic farming on the efficiency of energy use in agriculture. An Organic Center State of Science Review. https://organic-center.org/reportfiles/EnergyReport.pdf (Accessed September 2018).

718

21. Milk proteins: The future

TABLE 21.2

Virtual water content of dairy and related products

Product

Virtual water (m3/ton)

References

Milk

990

Hoekstra and Chapagain (2007)

Milk powder

4602

Hoekstra and Chapagain (2007)

Milk protein powders

18,400

Calculated from earlier

Soybeans

1789

Hoekstra and Chapagain (2007)

Soy protein

5400

Calculated from earlier

is required to produce it throughout the production chain. The amounts of virtual water in a range of products are given in Table 21.2. Most of the virtual water in these products arises from on-farm activities, with processing water being a minor component. Hence, the protein product values were calculated here by simply adjusting for the amount of protein in the parent product, without adjustment for processing water or credit for the water value of any coproducts. The key point is that, as with energy, the cost of water for producing milk-origin products is several-fold higher than that for producing similar plant-origin products. This means that only countries that are water rich can sustainably produce animal-based products for export. This will have an effect in the future as water distribution not only changes with climate change but also threatens production in some parts of the world where existing water use is unsustainable, such as parts of Australia, where water offtake has led to saline ingress into soils (Anderies et al., 2006). The use of water for dairy production must be considered in relation to the availability of water and competition for that water. In countries and regions such as New Zealand and parts of Brazil, there is an abundance of fresh water; as water not used for dairying is unlikely to be used for another more beneficial purpose, there is no real opportunity cost. As the availability of water becomes limiting in different parts of the world because of the effects of population growth and of climate change, the locus of dairy production may shift to water-rich regions. An interesting development is the emergence of “seawater greenhouse” farming based around using seawater as the primary source of water, with various methods for the production of fresh water to support plant growth. This method is already in operation in a number of countries for horticulture. Whether this will extend significantly to dairy production remains to be seen.

Implications of dairy methane production Methane merits special mention as a greenhouse gas because emissions from cows contribute substantially to the global greenhouse gas load as a by-product of rumen digestion. Methane is recognized as a greenhouse gas and is rated as having a global warming potential 21 times that of the equivalent amount of CO2, based on a 100-year timescale. It has been estimated that, of all the greenhouse gases, methane is second in effect only to CO2 and is responsible for around 10%–15% of the present greenhouse gas effect in the atmosphere

Global issues for food

719

(noting that this analysis excludes water vapor as a greenhouse gas). Globally, ruminant livestock produce about 28% of methane emissions from human-related activities. A single adult cow is a relatively minor contributor, emitting only 80–110 kg of methane; however, with about 95 million cattle in the United States alone (of which about 10% are milk cows) and 1.2 billion large ruminants in the world, ruminants are one of the largest sources of methane. Although the greenhouse gas effect of methane is a matter of concern, to put it in perspective, it has been estimated that the methane emissions from the entire dairy herd in the United States in 2007 were 112 billion kg CO2 equivalent, less than half of the emissions calculated for the buffalo (American bison) herd in 1860, that is, 228 billion kg CO2 equivalent (Capper, 2011). It is further recognized that increasing efficiencies in farming have a substantial effect in reducing the carbon footprint of milk. In the United States, it has been estimated that the carbon footprint of milk was reduced by 63% between 1944 and 2007 (Capper et al., 2009). In New Zealand, a strong negative correlation with greenhouse gas emissions has been shown for both kilograms of milk solids per hectare per year and kilograms of milk solids per cow, indicating that more efficient production systems have a lower footprint in terms of their product ( Judge et al., 2010). Most governments recognize the need to limit greenhouse gases, although the United States has recently revised its commitment to reducing greenhouse gas production. International negotiations following the Kyoto protocol were expected to impose penalties (carbon taxes) on greenhouse gas producers in many countries. To date, governments have shied away from imposing carbon taxes on pastoral farming, and it remains to be seen whether this will occur. Methane generation represents both a source of pollution and a waste of energy, and research efforts to specifically target the removal of methanogenic organisms from the rumen are important for the future economic viability of the industry. Current research into ways to control the growth of methanogenic bacteria in the rumen is expected to reduce methane production substantially with a concomitant increase in animal production (for a recent review, see Llonch et al., 2017). The results of a meta-analysis of the CO2 footprint from 52 life cycle assessment studies for a range of protein sources, including 12 for milk, show that milk, poultry, and eggs have substantially lower carbon footprints than other animal protein sources, as shown in Fig. 21.1 (Nijdam et al., 2012). We note that these analyses are silent with respect to the digestibility and the essential amino acid content of the protein, both of which would further favor milk proteins as a source of protein nutrition in terms of its carbon footprint. An analysis of the nutrient density and the greenhouse gas footprint of nearly 500 foods (Drewnowski et al., 2015) found that the more nutrient-dense animal products had higher greenhouse gas emissions per 100 g than other foods, but much lower values per 100 kcal, and identified a need for more work to identify the point at which the higher carbon footprint of nutrient-dense foods is offset by their higher nutritional value. In support of this, a recent analysis comparing the carbon footprints of cows’ milk and soy milk found that, although the carbon footprint of cows’ milk was more than 30% higher on a per volume basis, when corrected for a range of nutritional factors including composition, biological value, and energy, the adjusted carbon footprint of soy milk was 18% greater than that for cows’ milk (Grossi et al., 2017).

720

21. Milk proteins: The future

640

Beef extensive Beef intensive

= data point

Beef from dairy cows Pig meat Poultry meat

750

Sheep meat Milk Eggs

540

Seafood from fisheries Seafood from aquaculture Vegetal protein Other meat substitutes 0

50

100

150

200

250

kg CO2 eq per kg protein

FIG. 21.1 Carbon footprint of a range of protein foods. Reproduced from Nijdam, D., Rood, G.A., Westhoek, H., 2012. The price of protein: review of land use and carbon footprints from life cycle assessments of animal food products and their substitutes. Food Policy 37, 760–770, with permission.

Consumer demands and trends for food and ingredients “Localism” Over recent years, there has been a trend in many countries of supporting local food production, often termed “localism.” This not only will undoubtedly continue, nurtured by local politics, but also because of the often-mistaken belief that locally produced foods will have a smaller environmental impact. In the case of processed foods such as dairy products, this is generally not the case. In a recent study from Serbia (Djekic et al., 2018), a life cycle model was developed to investigate the impact on the environment of the transportation of dairy products. The model was validated for two dairy products from data presented by four dairy plants representing 32% of the total raw milk processed annually in Serbia. The results suggested that big dairy plants with a developed distribution system, together with social and economic indicators, have better results in terms of transportation sustainability. They concluded that ideas hidden behind the term “localism” in food systems, in relation to transportation impacts, may need to be reconsidered. Other studies have produced similar results, the consensus being that efficiencies in production and manufacture far outweigh any gains from local production.

Plant-derived milk substitutes In recent times, there has been an increase in so-called alternative milks, derived from plant sources. This was first led by soy milk, but other forms have emerged, principally almond milk, rice milk, coconut milk, and, more recently, cashew milk and hemp milk. These products are not milk, although they are often labeled as such and generally have

Consumer demands and trends for food and ingredients

721

inferior nutritional properties, particularly with respect to protein. For a recent review of their nutritional properties, see Vanga and Raghavan (2018). Early drivers for the consumption of these milks included lactose intolerance and cows’ milk allergies, but times and consumer attitudes have changed. According to a recent study of consumer attitudes in the United States: “A distinguishing characteristic of those who only drank nondairy plant-based alternatives was that plant-based beverages contributed to a goal to consume less animal products, beliefs about animal mistreatment, and perceived lesser effect on the environment than fluid milk” (McCarthy et al., 2017). This highlights the need for better education concerning the nutritional value of cows’ milk compared with plant-based substitutes and for dispelling some of the myths about the ethical and environmental costs of the dairy industry.

Food safety and traceability Throughout the world, the awareness of foodborne disease has risen in response to the high level of publicity that such outbreaks receive. The toll exacted in human and economic terms is considerable. Notable dairy outbreaks in recent years include Salmonella in ice cream (United States, 1994: 224,000 cases of illness) and staphylococcal enterotoxin in milk (Japan, 2000: 15,000 cases). Contaminated soft cheeses and raw milk are often in the news. Most dairy products, processed to modern standards of hygiene, have an excellent safety record, but consumers are demanding increased surveillance and control of all foods, including dairy. The contamination of animal feed with dioxin in Belgium in 1999 highlighted the importance that consumers place on the absence of toxic chemicals in their food. The deliberate adulteration of infant formula with melamine in China in 2008 resulted in illnesses and deaths of infants. There will be no lessening in the demands on food producers to control risks and deliver assurances of safety. The increased costs from providing this assurance through effective process control will become the norm for dairy businesses in the future. In recent times, increasing attention has been paid to traceability, so that any food safety issue can be quickly traced to its origin and other food from the same batch can be quickly quarantined. Traceability can also be important because of consumers’ desire for products that are sustainably produced or have other connotations of quality (such as organically produced products). Traceability is generally managed through the labeling and tracking of products throughout manufacture and distribution, usually by means of labels on the packaging. This is usually handled well by most food manufacturers and distributors. However, there have been attempts at “false-flagging” products in the past, and this will no doubt continue. In this context, blockchain technology, originally designed to protect cryptocurrencies, will be useful. It has been claimed authoritatively that blockchain technology has the potential to transform the food industry and to herald a food safety and antifraud revolution (Crew, 2018). A full discussion of blockchain is beyond the scope of this chapter, but it appears to be an unbreakable way of labeling the origin and tracing the provenance of food products, and dairy companies are actively considering its use. It has recently been reported that the supermarket chain Carrefour is to use blockchain for meat and dairy products (Chu, 2018). For products containing milk proteins, it is often possible to obtain an internal check on the origin of the product: dairy herds in different countries and regions tend to have a rather unique mixture of breeds and genetics. This is reflected in the distribution of polymorphisms

722

21. Milk proteins: The future

of the proteins, which can be relatively simply analyzed using gel electrophoresis and/or mass spectrometry. Additional information about processing can be gained from mass spectrometric analysis of postproduction changes in the chemistry of milk proteins (see Chapter 11). Concern arose in the past because of the deliberate adulteration of milk used for infant formula in China with melamine, leading to deaths of babies in the summer of 2008 (Tyan et al., 2009). Melamine has a high nitrogen content (66% w/w) and is relatively inexpensive; when milk is paid for on the basis of nitrogen content as a proxy for protein content, melamine can be a cheap way to fraudulently boost the apparent protein content. An unfortunate side effect of the addition of melamine can be the presence of cyanuric acid, which is a common contaminant and hydrolysis product of melamine. Although neither melamine nor cyanuric acid on its own is particularly toxic, the combination of the two can result in the formation of melamine cyanurate crystals, which are very insoluble and tend to form in the kidneys. This can lead to kidney failure. Protein measurements using Fourier transform infrared spectroscopy, which is now a common standard method for routine milk testing in factories, can detect low levels of melamine in dairy products (Balabin and Smirnov, 2011), and melamine itself is relatively easy to test for in other ways (Tyan et al., 2009); such tests are now widely used, coupled with severe penalties for noncompliance.

Food and health: Nutrigenomics and personalized nutrition Consumers are being increasingly sensitized to the effects of diet on health (and appearance). The success of diet clinics attests to this. The occurrence of (and attention being paid to) current high levels of obesity in affluent societies is spurring interest in diet at all levels of society, from individual to government. Food products on supermarket shelves are increasingly differentiated by the presence of (omega-3 fats and antioxidants) or absence of (fat free and gluten free) components that are believed to affect health. Individuals can easily obtain information about their own genetic profile. It is now possible for anyone to have their DNA genotyped for under US$100 (23andMe DNA Saliva Collection Kit or AncestryDNA), with several companies offering sequencing for under $2000, although these services vary according to their stated purpose and only some provide information with respect to known genetic polymorphisms related to health and metabolism—many are focused on heritage or legal issues such as paternity or identity. The combination of the availability of individual genetic data on a large scale and a detailed understanding of nutrition has enabled the field of “nutrigenomics”: the study of the relationship between a person’s genetic makeup and their individual nutritional needs. Although early studies in nutrigenomics suggested quick wins, with simple gene differences (single-nucleotide polymorphisms or SNPs) indicating specific dietary effects, the situation has been found to be much more complex. The focus is now on understanding “nutritional phenotypes,” which take into account not only SNPs but also other genetic differences such as copy number variants (CNVs) of genes and differences in expression of those genes that can be caused by a range of factors, including epigenetic effects, health status, lifestyle, and medications (van Ommen et al., 2010). “Personalized nutrition” is a nutritional response to differences between individuals, whether from a nutrigenomics input or through other identified needs and preferences

Consumer demands and trends for food and ingredients

723

and attempts to balance an individual’s diet to their specific individual (nutritional phenotype) and situational needs. Nutrition today is not just about the balance of macro- and micronutrients; a plethora of “functional” (bioactive) food components is also known to affect health in ways that extend far beyond the simple supply of nutrients. They can also be modifiers of nutrient uptake and usage, thus modifying the effect of nutritional balance as seen by the body’s metabolism. The kinetics of nutrient uptake can be just as important as overall absolute uptakes of nutrients. In the case of carbohydrates, this has translated into the “glycemic index”—an indicator of the rapidity of glucose uptake and thus the effect of a food on insulin production in the body. In the case of proteins, “fast” and “slow” proteins that exit the stomach either rapidly or slowly following ingestion have been identified (Boirie et al., 1997), and these may have significant effects on hormone levels and satiety. Personalized nutrition attempts to take this into account, to provide optimal customized nutrition for the individual. In sophisticated markets today, there is increasing acceptance that nutrition has a profound effect on health and wellness, and as individuals become more aware of their specific nutritional needs, the demand for personalized nutrition is set to increase. To date, the impact of all this on milk proteins has been minimal. However, four aspects are notable. • Allergies to milk. Such allergies, particularly in infants, have mostly been attributed to the β-lactoglobulin in cows’ milk. This protein is not produced in human milk and is the dominant whey protein in bovine milk and the milk of other ruminants (see Chapter 2). Whey proteins are important nutritionally, as they are a valuable source of essential amino acids. β-Lactoglobulin is a particularly important source of branched-chain amino acids. So-called hypoallergenic products are therefore produced by hydrolyzing milk proteins, more particularly whey proteins, so that fragments are sufficiently small to be nonallergenic. • A2 milk. There is some literature that suggests a weak correlation between the consumption of milk containing the β-casein A1 variant and some diseases, notably type 1 diabetes (Elliott et al., 1997) and ischemic heart disease (McLachlan, 2001). Further studies on diabetes proved to be inconclusive (Beales et al., 2002) and the heart disease data do not stand up to scrutiny; furthermore, other epidemiological data show that the A2 hypothesis does not hold up (Truswell, 2005). Notwithstanding this, a New Zealand–based company, the A2 Corporation, markets a niche milk product from cows that do not carry the A1 gene. The milk is sold in the United Kingdom and the United States but has its longest history in New Zealand and in Australia, where the company is now very careful not to make claims about any specific health benefits after being prosecuted and fined $15,000 in Queensland in 2004 for making such claims. A report on A2 milk was released by the European Food Safety Authority (De Noni et al., 2009); it concludes: “Based on this review, EFSA concluded that a cause and effect relationship is not established between the dietary intake of BCM7, related peptides or their possible protein precursors and non-communicable diseases. Consequently, a formal EFSA risk assessment is not recommended.” More recently, some studies have indicated that there may be a group of consumers, particularly among Chinese, who may be generally milk intolerant but tolerant of A2 milk (He et al., 2017). Other studies using in vitro digestion with human-derived enzymes have

724

21. Milk proteins: The future

identified that the release of the bioactive peptide β-casomorphin-7 does not seem to be linked solely to the A1 variant as had been previously claimed (Asledottir et al., 2017). • Bioactive peptides. There is increasing evidence that some milk proteins, and more particularly peptides, have physiological functionality. Effects on cardiovascular health, immune modulation, health of bones and teeth, and cancer have been reported (see Chapters 17 and 18). Although the validity of these effects and the efficacy of functional foods based on bioactive peptides remain to be fully proven, in time, they may lead to new functional foods based on milk proteins and their products. • Milk proteins and the gut microbiome. There has been considerable recent interest in the gut microbiome, its effects on health, and how it can be affected by diet. Because most milk proteins are relatively easily digested, especially the caseins, they would not be expected to reach the colon and have much effect on the microbiome. β-Lactoglobulin can be more resistant to digestion than some proteins, and there is some suggestion that the milk proteins in a high-protein diet can affect the microbiome. A clinical trial by Beaumont et al. (2017) concluded that high-protein diets did not alter the composition of the microbiota but induced a shift in bacterial metabolism toward amino acid degradation, with different metabolite profiles according to the protein source, and that casein-based and soy protein– based high-protein diets did not induce inflammation but differentially modified the expression of genes playing key roles in homeostatic processes in the rectal mucosa, such as cell cycle or cell death. An ex vivo study by Sa´nchez-Moya et al. (2017) using whey protein concluded that whey milk has a potent prebiotic effect and that it can selectively stimulate desirable bacteria and the short-chain fatty acid profile, in both obese and normal weight donors, contributing to improved intestinal health and reducing obesity.

New technologies and their possible effect on milk protein ingredients and products A number of new technologies have the potential to affect dairy production and processing in the near future. They include gene technologies that could lead to new, different milk proteins, new kinds of processing that can produce novel milk protein materials and products containing them, and new analytical techniques that have the potential to improve processing and place ever more stringent requirements on product quality.

Genetic modification Milk proteins have been genetically modified and expressed in nonbovine animals (e.g., Bleck et al., 1998) and in cows (Brophy et al., 2003). However, it seems to be unlikely that the transgenic modification of milk proteins for functional or nutritional purposes will occur widely in the foreseeable future. There are several reasons for this (Creamer et al., 2002):

New technologies and their possible effect on milk protein ingredients and products

725

• Consumer acceptance of genetically modified (GM) foods is still variable, throughout the world, with most countries now having strict labeling requirements. Because milk is a liquid product that is handled in large volumes during processing, maintenance of batch identity and keeping GM milk separate are more problematic than with discrete products, although recent efforts with organic milk have proven that this is possible. • Milk is an animal product that is strongly targeted at the health of babies and young people. This has been identified in consumer surveys as a very sensitive area (compared with, e.g., the acceptability of GM fruit and vegetables), and milk will probably be one of the last foods for which genetic modification is acceptable to most consumers. • The cost of producing herds of GM cows will be very high, and developing herds will be very slow unless expensive cloning and embryo transfer methods are used. This is not justified by a small premium for improved nutrition or functionality arising from genetic modification. • More importantly, a switch to genetic modification will severely limit genetic gain, because the gene pool will be restricted to the genetics of the donor animals for the original GM parents. This segregation from the global bovine gene pool will prevent, or severely limit, participation in the ongoing genetic improvement of the species, currently occurring at about 2% per annum. Notwithstanding these points, if a strong nutriceutical or pharmaceutical component was to be identified, enhanced expression through genetic modification would not be out of the question. However, much-touted “gene pharming” in dairy animals has not been notably successful commercially, and most GM protein products are produced more cheaply and efficiently using fermentation technology. The recent announcement of the development of a female calf that does not produce β-lactoglobulin in its milk is an interesting development (Wei et al., 2018). Because β-lactoglobulin does not occur in human milk, it is often regarded as an undesirable component in milk for infant formula, and it has been implicated in milk allergies (as discussed earlier). Although the composition of the milk from this experimental animal has been shown to be quite unusual, with no detectable β-lactoglobulin and elevated casein levels, this composition must be viewed with caution, as it was produced artificially from a female induced by hormones rather than from natural lactation. It remains to be seen if this animal is able to breed. The fact that this construct removes a protein from milk, rather than adding a foreign protein to it, may lead to a product that has less consumer resistance. In contrast to milk proteins, competitive plant-origin proteins are well advanced in improvement using genetic modification. GM soybeans are now predominant in world soybean crops, with more than 90% of the soybeans produced in the United States being genetically modified for the past decade, according to the US Department of Agriculture (2018). Most of this is accounted for by herbicide (glyphosate)-tolerant varieties. Soybeans can also be genetically modified to remove undesirable proteins such as trypsin inhibitors, soy hemagglutinin, and allergens (e.g., Friedman et al., 1991); at the same time, soy proteins can be modified to provide a more favorable nutritional balance of essential amino acids (Mandal and Mandal, 2000). The more efficient production of soy proteins in terms of energy and water, coupled with these improvements from genetic modification, means that soy proteins will increasingly outcompete dairy proteins as generic nutritional and functional food ingredients.

726

21. Milk proteins: The future

However, as there is expected to be increasing demand for high-quality protein as a result of population growth and emerging economies, there is expected to be a continuing strong market demand for all forms of food protein (Boland et al., 2013).

Novel processing High-pressure processing was originally developed in the late 19th century (Hite, 1899) but did not find application in food processing until the 1990s, when new materials enabled the development of production-scale processing equipment. High-pressure processing has been used commercially as an alternative method for preservation, particularly for acidic foods (Dunne and Kluter, 2001). When milk is subjected to high pressure, the casein micelle undergoes dramatic nonreversible changes, leading to a smaller micelle that is less opaque (Chapter 8). It has also been reported that high-pressure processing can alter the functionality of whey proteins (Patel et al., 2005). Thus far, no commercial products using this technology have been seen. A number of other “nonthermal” processes for the preservation of food have been developed. These include pulsed electric field, irradiation, and cold plasma. Although some have been successful in the processing of acidic foods and solid foods (in the case of cold plasma), commercial application of these processes to dairy products is not an immediate prospect.

New analytical methods Recent years have seen a number of new and improved analytical methods that have the potential to improve process control and tighten product specification. Particularly important are methods that can control product safety (particularly microbiology) and nutritional and functional properties. One of the weaknesses in product safety in the past has been the need to grow up samples on Petri dishes to test for the presence of undesirable microbiological species. This process is time consuming and laborious and can identify issues with process and product only well after they have occurred. A number of rapid microbiological detection methods that allow at-line detection of microbiological problems and can provide early assurance of food safety are now available. For example, the use of flow cytometry was able to reduce the time for measuring bacterial numbers from the 3 days required for the traditional plate count to 2 h (Flint et al., 2006). This is particularly important for proteins manufactured using ultrafiltration, such as whey protein concentrates and milk protein concentrates, because the ultrafiltration step coconcentrates any microbiological contaminants that may be present. However, traditional plate-based methods continue to be used for the enumeration of bacteria, albeit made more efficient with robotic systems, because of their lower cost and their fit with existing regulatory systems. For identification, molecular methods (polymerase chain reaction and DNA sequencing) and matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry are widely used. The latter has become more widely used in the last 5 years. Modern electrospray mass spectrometric analysis has enhanced our ability to understand and control processing effects that can alter the nutritional value of milk proteins, particularly

Conclusions

727

the loss of bioavailable lysine that is caused by processing and storage effects (see Chapter 11). Similarly, once the relationship between functionality and protein chemistry is well understood, the same techniques will allow better management of functional properties. New in-line and at-line methods are now possible through a range of techniques, including nuclear magnetic resonance, Fourier transform infrared spectroscopy, selected ion flow tube mass spectrometry, and surface plasmon resonance analysis. These can, in principle, measure attributes such as the water activity inside packaging and can be used in the prediction of the flavor characteristics of cheeses at maturation (e.g., Langford et al., 2012).

Materials science and nanotechnology At all dimensional scales, structure is important for the sensory properties of food (including texture, mouthfeel, and flavor release) and may have an important effect on nutrient release and bioavailability (Parada and Aguilera, 2007). Increasingly, attention is being paid to materials science approaches to understanding and potentially managing these effects. An example is the physics of soft materials being applied to food (Mezzenga et al., 2005; Ubbink et al., 2008). The potential application of nanotechnology and nanoscience to food can be expected to become an important area. Much of the higher dimensional structure of food is a consequence of nanostructures. It is unlikely that nanorobotics will be applied to food in the foreseeable future: regardless of the considerable technical challenges, public acceptance can be expected to be a major barrier. Notwithstanding this, nanotechnology is having a considerable impact on food science, in part through the use of new improved instrumentation becoming available to support nanotechnology research (Foegeding, 2006; Weiss et al., 2006). One of the important features in nanotechnology is the occurrence of self-assembling molecular superstructures (nanostructures). It turns out that foods naturally contain many such systems, examples being the actin–myosin complex in the muscle fibers of meat, starch granules in plant foods, and the casein micelle in milk. Whey proteins have been shown to form self-assembling systems under a variety of conditions—as whey protein (a mixture), as β-lactoglobulin, and as α-lactalbumin. Early publications (Bolder et al., 2006; Graveland-Bikker and de Kruif, 2006) have been followed by a wide range of published work, reviewed by Jones and Mezzenga (2012) and Loveday et al. (2012), that shows how to control the generation and properties of such structures and their physical and functional properties. More recent work has investigated food safety, digestibility, and possible food applications of nanofibrils (Loveday et al., 2017), although there is no evidence of commercial uptake as yet. A number of nonfood uses for β-lactoglobulin fibrils have been suggested, including biosensors, photovoltaics, and nanocomposites (Knowles and Mezzenga, 2016). A recent publication shows promise for a nanofibril system for the delivery of bioavailable iron (Shen et al., 2017).

Conclusions This brief chapter has provided a glimpse of some of the global issues and new technologies that may influence the future development and use of dairy protein products.

728

21. Milk proteins: The future

Global trends such as rising energy costs, the scarcity of water, and the effect of greenhouse gases will increase the cost of production of milk proteins and will restrict the land areas on which they can be produced sustainably, especially to areas that are water rich. Milk proteins are relatively expensive nutritional and functional food ingredients; although they are nutritionally superior to plant-origin competitors, they are likely to be increasingly restricted to niche applications as less expensive plant-based alternatives become more widely available.

References Anderies, J.M., Ryan, P., Walker, B.H., 2006. Loss of resilience, crisis, and institutional change: lessons from an intensive agricultural system in southeastern Australia. Ecosystems 9, 865–878. Asledottir, T., Le, T.T., Petrat-Melin, B., Devold, T.G., Larsen, L.B., Vegarud, G.E., 2017. Identification of bioactive peptides and quantification of β-casomorphin-7 from bovine β-casein A1, A2 and I after ex vivo gastrointestinal digestion. Int. Dairy J. 71, 98–106. Balabin, R.M., Smirnov, S.V., 2011. Melamine detection by mid- and near-infrared (MIR/NIR) spectroscopy: a quick and sensitive method for dairy products analysis including liquid milk, infant formula, and milk powder. Talanta 85, 562–568. Beales, P.E., Elliott, R.B., Floh
e, S., Hill, J.P., Kolb, H., Pozzilli, P., Wang, G.-S., Wasmuth, H., Scott, F.W., 2002. A multicentre, blinded international trial of the effect of A1 and A2 ß-casein variants on diabetes incidence in two rodent models of spontaneous Type I diabetes. Diabetologia 45, 1240–1246. Beaumont, M., Portune, K.J., Steuer, N., Lan, A., Cerrudo, V., Audebert, M., Dumont, F., Mancano, G., Khodorova, N., Andriamihaja, M., Airinei, G., Tom
e, D., Benamouzig, R., Davila, A.M., Claus, S.P., Sanz, Y., Blachier, F., 2017. Quantity and source of dietary protein influence metabolite production by gut microbiota and rectal mucosa gene expression: a randomized, parallel, double-blind trial in overweight humans. Am. J. Clin. Nutr. 106, 1005–1019. Bleck, G.T., White, B.R., Miller, D.J., Wheeler, M.B., 1998. Production of bovine α-lactalbumin in the milk of transgenic pigs. J. Anim. Sci. 76, 3072–3078. Boirie, Y., Dangin, M., Gachon, P., Vasson, M.P., Maubois, J.L., Beaufre`re, B., 1997. Slow and fast dietary proteins differently modulate postprandial accretion. Proc. Natl. Acad. Sci. U S A 94, 14930–14935. Boland, M.J., Rae, A.N., Vereijken, J.M., Meuwissen, M.P.M., Fischer, A.R.H., van Boekel, M.A.J.S., Rutherfurd, S.M., Gruppen, H., Moughan, P.J., Hendriks, W.H., 2013. The future supply of animal-derived protein for human consumption. Trends Food Sci. Technol. 29, 62–73. Bolder, S.G., Hendrickx, H., Sagis, L.M.G., van der Linden, E., 2006. Fibril assemblies in aqueous whey protein mixtures. J. Agric. Food Chem. 54, 4229–4234. Brophy, B., Smolenski, G., Wheeler, T., Wells, D., L’Huillier, P., Laible, G., 2003. Cloned transgenic cattle produce milk with higher levels of β-casein and κ-casein. Nat. Biotechnol. 21, 157–162. Capper, J.L., 2011. Replacing rose-tinted spectacles with a high-powered microscope: the historical versus modern carbon footprint of animal agriculture. Anim. Front. 1, 26–32. Capper, J.L., Cady, R.A., Bauman, D.E., 2009. The environmental impact of dairy production: 1944 compared with 2007. J. Anim. Sci. 87, 2160–2167. Chu, W., 2018. Carrefour extend blockchain use to dairy and meat product ranges. https://www.foodnavigator. com/Article/2018/03/08/Carrefour-extend-Blockchain-use-to-dairy-and-meat-product-ranges. Accessed 11 December 2018. Creamer, L.K., Pearce, L.E., Hill, J.P., Boland, M.J., 2002. Milk and dairy products in the 21st century. J. Agric. Food Chem. 50, 7187–7193. Crew, S., 2018. The potential of blockchain. Food Sci. Technol. 32, 54–56. De Noni, I., FitzGerald, R.J., Korhonen, H.J.T., Le Roux, Y., Livesey, C.T., Thorsdottir, I., Tom
e, D., Witkamp, R., 2009. Review of the potential health impact of β-casomorphins and related peptides. Report of the DATEX working group on β-casomorphins. EFSA Sci. Rep. 231, 1–107. Djekic, I., Smigic, N., Glavan, R., Miocinovic, J., Tomasevic, I., 2018. Transportation sustainability index in dairy industry – fuzzy logic approach. J. Clean. Prod. 180, 107–115. Drewnowski, A., Rehm, C.D., Martin, A., Verger, E.O., Voinnesson, M., Imbert, P., 2015. Energy and nutrient density of foods in relation to their carbon footprint. Am. J. Clin. Nutr. 101, 184–191.

References

729

Dunne, C.P., Kluter, R.A., 2001. Emerging non-thermal processing technologies: criteria for success. Aust. J. Dairy Technol. 56, 109–112. Elliott, R.B., Wasmuth, H.E., Bibby, N.J., Hill, J.P., 1997. The role of beta-casein variants in the induction of insulindependent diabetes in the non-obese diabetic mouse and humans. In: Milk Protein Polymorphism. International Dairy Federation, Brussels, Belgium, pp. 445–453. Fairlie, S., 2010. Meat: A Benign Extravagance. Chelsea Green Publishing, White River Junction, VT, USA. FAO, 2009. The State of Food and Agriculture: Livestock in the Balance. Food and Agriculture Organisation, Rome, Italy. Flint, S., Drocourt, J.-L., Walker, K., Stevenson, B., Dwyer, M., Clarke, I., McGill, D., 2006. A rapid, two-hour method for the enumeration of total viable bacteria in samples from commercial milk powder and whey protein concentrate powder manufacturing plants. Int. Dairy J. 16, 379–384. Foegeding, E.A., 2006. Food biophysics of protein gels: a challenge of nano and macroscopic proportions. Food Biophys. 1, 41–50. Friedman, M., Brandon, D.L., Bates, A.H., Hymowitz, T., 1991. Comparison of a commercial soybean cultivar and an isoline lacking the Kunitz trypsin inhibitor: composition, nutritional value, and effects of heating. J. Agric. Food Chem. 39, 327–335. Ganguli, N.C., 1978. Realities in buffalo milk processing. Indian Dairyman 30, 165–175. Ganguli, N.C., 1979. Stability of buffalo casein micelles. J. Dairy Res. 46, 401–405. Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., Pretty, J., Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the challenge of feeding 9 billion people. Science 327, 812–818. Graveland-Bikker, J.F., de Kruif, C.G., 2006. Unique milk protein based nanotubes: food and nanotechnology meet. Trends Food Sci. Technol. 17, 196–203. Grossi, G., Vitali, A., Lacetera, N., Nardone, A., Bernabucci, U., 2017. Comparison between cow milk and soy milk combining nutritional values and GHG data. Ital. J. Anim. Sci. 16 (Suppl. 1), 177. Hansen, P.J., 2004. Physiological and cellular adaptations of zebu cattle to thermal stress. Anim. Reprod. Sci. 82–83, 349–360. He, M., Sun, J., Jiang, Z.Q., Yang, Y.X., 2017. Effects of cow’s milk beta-casein variants on symptoms of milk intolerance in Chinese adults: a multicentre, randomised controlled study. Nutr. J. 16, 72. Hite, B.H., 1899. The Effect of Pressure in the Preservation of Milk: A Preliminary Report. Bulletin of the West Virginia University Agricultural Experiment Station, vol. 58, pp. 15–35. Hoekstra, A.Y., Chapagain, A.K., 2007. Water footprints of nations: water use by people as a function of their consumption pattern. Water Resour. Manage. 21, 35–48. Jones, O.G., Mezzenga, R., 2012. Inhibiting, promoting, and preserving stability of functional protein fibrils. Soft Matter 8, 876–895. Judge, A., Ledgard, S., Smeaton, D., Boyes, M., 2010. Greenhouse gas emissions from Rotorua dairy farms. Report to MAF, AgResearch, Hamilton, New Zealand. Kadzere, C.T., Murphy, M.R., Silanikove, N., Maltz, E., 2002. Heat stress in lactating dairy cows: a review. Livest. Prod. Sci. 77, 59–91. Knowles, T.P.J., Mezzenga, R., 2016. Amyloid fibrils as building blocks for natural and artificial functional materials. Adv. Mater. 28, 6546–6561. Langford, V.S., Reed, C.J., Milligan, D.B., McEwan, M.J., Barringer, S.A., Harper, W.J., 2012. Headspace analysis of Italian and New Zealand parmesan cheeses. J. Food Sci. 77, C719–C726. Llonch, P., Haskell, M.J., Dewhurst, R.J., Turner, S.P., 2017. Current available strategies to mitigate greenhouse gas emissions in livestock systems: an animal welfare perspective. Animal 11, 274–284. Loveday, S.M., Su, J., Rao, M.A., Anema, S.G., Singh, H., 2012. Whey protein nanofibrils: the environment–morphology–functionality relationship in lyophilization, rehydration, and seeding. J. Agric. Food Chem. 60, 5229–5236. Loveday, S.M., Anema, S.G., Singh, H., 2017. β-Lactoglobulin nanofibrils: the long and the short of it. Int. Dairy J. 67, 35–45. Mandal, S., Mandal, R.K., 2000. Seed storage proteins and approaches for improvement of their nutritional quality by genetic engineering. Curr. Sci. 79, 576–589. McCarthy, K.S., Parker, M., Ameerally, A., Drake, S.L., Drake, M.A., 2017. Drivers of choice for fluid milk versus plant-based alternatives: what are consumer perceptions of fluid milk? J. Dairy Sci. 100, 6125–6138. McLachlan, C.N., 2001. β-Casein A1, ischaemic heart disease mortality, and other illnesses. Med. Hypotheses 56, 262–272.

730

21. Milk proteins: The future

Mezzenga, R., Schurtenberger, P., Burbridge, A., Michel, M., 2005. Understanding foods as soft materials. Nat. Mater. 4, 729–740. Nijdam, D., Rood, G.A., Westhoek, H., 2012. The price of protein: review of land use and carbon footprints from life cycle assessments of animal food products and their substitutes. Food Policy 37, 760–770. Parada, J., Aguilera, J.M., 2007. Food microstructure affects the bioavailability of several nutrients. J. Food Sci. 72, R21–R32. Patel, H.A., Singh, H., Havea, P., Considine, T., Creamer, L.K., 2005. Pressure-induced unfolding and aggregation of the proteins in whey protein concentrate solutions. J. Agric. Food Chem. 53, 9590–9601. Pimentel, D., 2006. Impacts of organic farming on the efficiency of energy use in agriculture. An Organic Center State of Science Review. https://organic-center.org/reportfiles/EnergyReport.pdf. (Accessed September 2018). Pimentel, D., Pimentel, M., 1979. Food, Energy and Society. John Wiley & Sons, New York, NY. Sa´nchez-Moya, T., Lo´pez-Nicola´s, R., Planes, D., Gonza´lez-Bermu´dez, C.A., Ros-Berruezo, G., Frontela-Saseta, C., 2017. In vitro modulation of gut microbiota by whey protein to preserve intestinal health. Food Funct. 8, 3053–3063. Saunders, C., Barber, A., Taylor, G., 2006. Food miles – comparative energy/emissions performance of New Zealand’s agriculture industry. Research Report No. 285, Lincoln University, Lincoln, New Zealand. Shen, Y., Posavec, L., Bolisetty, S., Hilty, F.M., Nystr€ om, G., Kohlbrecher, J., Hilbe, M., Rossi, A., Baumgartner, J., Zimmermann, M.B., Mezzenga, R., 2017. Amyloid fibril systems reduce, stabilize and deliver bioavailable nanosized iron. Nat. Nanotechnol. 12, 642–647. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M., & de Haan, C. (2006). Livestock’s Long Shadow: Environmental Issues and Options. Rome, Italy: Food and Agriculture Organisation. Truswell, A.S., 2005. The A2 milk case: a critical review. Eur. J. Clin. Nutr. 59, 623–631. Tyan, Y.-C., Yang, M.-H., Jong, S.-B., Wang, C.-K., Shiea, J., 2009. Melamine contamination. Anal. Bioanal. Chem. 395, 729–735. Ubbink, J., Burbidge, A., Mezzenga, R., 2008. Food structure and functionality: a soft matter perspective. Soft Matter 4, 1569–1581. US Department of Agriculture, 2018. Adoption of genetically engineered crops in the U.S. https://www.ers.usda. gov/data-products/adoption-of-genetically-engineered-crops-in-the-us/. (Accessed 11.12.18). van Ommen, B., Bouwman, J., Dragsted, L.O., Drevon, C.A., Elliott, R., de Groot, P., Kaput, J., Mathers, J.C., M€ uller, M., Pepping, F., Saito, J., Scalbert, A., Radonjic, M., Rocca-Sera, P., Travis, A., Woperis, S., Evelo, C.T., 2010. Challenges of molecular nutrition research 6: the nutritional phenotype database to store, share and evaluate nutritional systems biology studies. Genes Nutr. 5, 189–203. Vanga, S.K., Raghavan, V., 2018. How well do plant based alternatives fare nutritionally compared to cow’s milk? J. Food Sci. Technol. 55, 10–20. Wei, J.W., Wagner, S., Maclean, P., Brophy, B., Cole, S., Smolenski, G., Carlson, D.F., Fahrenkrug, S.C., Wells, D.N., Laible, G., 2018. Cattle with a precise, zygote-mediated deletion safely eliminate the major milk allergen betalactoglobulin. Sci. Rep. 8, 7661. Weiss, J., Takhistov, P., McClements, D.J., 2006. Functional materials in food nanotechnology. J. Food Sci. 71, R107–R116.

Index Note: Page numbers followed by f indicate figures and t indicate tables.

A Acid casein, 76 Acid coagulation, 192 Acid gel formation acidification rate, 241 aggregation rate, 241, 243 complex modulus, 241–242, 242f forewarming, 241–242 gelation pH, 241–242, 242f gelation profile, 241–242, 242f hi-pH gel, 243 Acid-induced aggregation/gelation breaking stress, 365 casein micelle particles production, 362 final gel firmness, 357–360, 358f firmness, 356, 356f isoelectric point, 362 nonsedimentable denatured whey proteins, 360–362, 361f reassociation behavior, 363, 364f serum-phase and colloidal-phase aggregates, 362–363 soluble aggregates, 362 starch addition, 360 strain deformation properties, 363–365 stress vs. strain curves, 356–357, 357f thiol-blocking agents, 357 yield stress, 357–360, 359f Acid-induced milk gels acid impact on casein micelles, 609–610 texture casein conjugates, 616 denatured whey proteins (DWP), 613–614 enzymatic modification of proteins, 612 exopolysaccharides (EPS) production, 616 factors influencing, 610 fat globules, 611 heat treatment, 612–615, 613f, 615f high hydrostatic pressure treatment, 611 homogenization, 611 incubation temperature, 615–616 postprocessing operations, 610

TGase treatment, 612 ultrasound processing, 611 Acid-precipitated protein. See Casein Adulteration, 52–53 Age gelation, 371–372, 373f causes, 404–405 fat separation/creaming, 406–407 mechanism, 406 onset protease activity, 405 temperature of storage, 405 temperature-time conditions of heat treatment, 406 retardation, 406 Aging populations dietary essential amino acids, 11–12, 12f essential amino acids in nutrition, 13–14 need for quality protein, 11–12, 12f protein nutritional needs of, 12–13 recommended dietary allowance (RDA), 13 sarcopenia, 12–13 Alpha-crystallin, 552 Amino acid score, 68 Amphiphysin, 111 Anabolic resistance, 656–657 Anabolism, 13–14 Analytical method protein quality assessment biological methods, 67 chemical methods, 68 digestibility and bioavailability, 67 indispensable amino acids (IAAs), 67 mixed assays, 68–69, 70t nitrogen balance in human nutrition, 66–67, 67f protein separation, purification, identification, and quantification (see Chromatography methods; Electrophoresis methods) protein structural analysis circular dichroism (CD), 63–64 Fourier transform infrared (FTIR) spectroscopy, 62–63, 62t, 63f nuclear magnetic resonance (NMR) spectroscopy, 65–66

731

732

Index

Analytical method (Continued) structural levels, 61–62 X-ray crystallography, 64–65, 64f total protein determination (see Total protein determination) Angel food cake, 585–586 Angiogenins, 49 Animal protein consumption, 2 Arginine, 640 Aseptic packaging, 388 Associative interactions, 521 Atherosclerosis, 658–659

Bovine milk composition lactose, 24–25 lipids, 23–24 oligosaccharides, 26 principal constituents, 22, 23t salts in milk, 26–27 vitamins, 27 water, 22 Branched chain amino acid (BCAA) leucine, 653, 656–657 Bridging flocculation, 478–479 Browning, 411, 411t Butyrivibrio fibrisolvens, 23

B

C

Bacterial proteases micelle aggregation and sedimentation, 398–399 production and characteristics, 399 proteolysis addition of cultures, 401–402 enzymatic proteolysis, 403 heat-induced proteolysis, 404 by plasmin, 402–403, 403t raw milk before processing, 399–401, 400t in ultrahigh-temperature hydrolyzed-lactose milk, 404 Pseudomonas species, 399 specificity of, 399 Bakery products, 584–586 Bar hardening, 590 Bennett’s law, 9 Bioactive molecules, 99–100 Bioactive peptides casein hydrolysate-mediated effect, 645 dietary intakes and its effects, 645–646 endogenous ileal lysine and total nitrogen losses, 645, 645t endogenous lysine loss, 644–645, 644t gut metabolism, 644 natural food-derived peptides, 643–644, 644t nonnutritional physiological functions, 643 physiological functions, 643 Bioavailability, 6–7, 7f Biohydrogenation, 23 Biologically active cryptic peptides, 51 Biological value (BV), 67 Bioluminescence Resonance Energy Transfer (BRET) transduction system, 405 Biotin (B7), 27 Blood pressure and vascular reactivity, 659–660 Blood serum albumin, 41 Bone health, 660–661 Bone mineral density (BMD), 660–661 Bos taurus, 175 Bovine Genome Database, 145

Calcium phosphate nanoclusters biomineralization properties, 228 composition, 228–229 crystalline brushite, 229–230 interactions, 230 molecular weights, 229 precipitation, 227–228 production and stabilization, 228–229 spatial separation, 230 thermodynamics, 230 Calories and satiety, 642–643 Capillary electrophoresis, 56 Carbon footprint, 9–10 Casein. See also Heterogeneity, casein; Posttranslational modifications of caseins expression function and milk quality, 159–160 heterogeneity, 33–34 interspecies comparison, 33 micelles (see also Micelle, casein) properties and structure, 35–36 stability, 36 structure, 37–38 microheterogeneity disulfide bonding, 35 genetic polymorphism, 34 glycosylation, 35 hydrolysis, 35 phosphorylation, 34 origins of gene structure, 151 phosphoproteins, 173 powdered ingredients (see Powdered milk protein ingredients) precipitation, 28 secretory calcium-binding phosphoproteins, 213–214 vs. whey proteins amino acid composition, 45 binding property, 45 coagulability, 45 heat stability, 45 hydrophobicity, 43–44

Index

physical state in milk, 46 preparation, 46–47 principal properties, 43, 44t site of biosynthesis, 46 solubility, 45 specific volume, 44 stability, 44 structures, 43 sulfur content, 46 tendency to associate, 44 Caseinates, 47, 76, 467 Caseinomacropeptide (CMP), 187–188, 193, 600 Caseinophosphopeptides (CPPs), 543–544, 706 Cathelicidin apoptosis, 114 bactericidal activity, 113 BMAP-28, 114 cell proliferation, 114 host-defense peptides, 111–112 immune cells, 114 MaeuCath1–8 gene, 113 PR-39, 114 sequence heterogeneity, 113 structure, 112–113, 112f temporal delivery of, 111 Cereal Price Index, 1–2 Ceruloplasmin, 48 Cheese, 238–239, 600 chymosin action, 190 coagulation properties of milk, 190–191 rennet coagulation time (RCT), 191–192 Chemical score, 68 Child Growth Standards, 3 Childhood eczema, 662–663 Childhood stunting, 3 Chlorpromazine, 552 Chromatography methods ion-exchange chromatography (IEX), 59 mass spectrometry (MS) methods, 59–61 reversed phase-high-performance liquid chromatography (RP-HPLC), 57–58, 58f size exclusion chromatography, 58–59 Chymosin, 600 Chymosin-induced aggregation/gelation aggregation process, 366, 367f, 368 aggregation time vs. level of whey protein, 368, 369f glycomacropeptide (GMP) release, 365–366 heated vs. unheated milks, 365–366, 366f retardation, 366 serum and colloidal phases, 366 Circular dichroism (CD), 63–64 Coalescence, 478 Cobalamin (B12), 27 Coffee whiteners, 586–587, 587t Colloidal calcium phosphate (CCP), 607, 609

733

Commercial sterility, 386 Competition for land use, 716 Complex coacervation, 502 Composition, 6 Concentrated micellar dispersions colloidal hard sphere behavior, 234 flocculation, 236 intramicellar electrostatic repulsion, 235–236, 236f micellar suspensions, 234–235 osmotic stress technique, 237 pellets, 237 solubilize on acidification, 236 Concentrate viscosity, 372–374, 374f Concurrent asynchronous lactation, 107–109, 109f Conditionally essential dietary amino acid, 640 Consumer demands and food trends food safety and traceability, 721–722 localism, 720 nutrigenomics and personalized nutrition, 722–724 plant-derived milk substitutes, 720–721 Cosolubility, 500–501 Cosolute solutions, 550 Covalent bonds, 505–506 Cow milk infant formula (CIF), 692 Creaming, 476–477, 479, 479f Cross-linking, 412–413 Cryoprecipitation, 47

D Dairy Price Index, 1–2 Dairy products, 21, 423–424 Deamidation, 411–412 Dehydration of milk, 423 Denatured whey proteins-κ-casein interactions disulfide bonds aggregation, 351 Cys121, 349 Cys residues identification, 349–351 intermolecular thiol-disulfide exchange reactions, 351–352, 352f nonnative monomeric species, 349–350 reactive monomer, 351 thionitrobenzoate (TNB), 351 in milk systems aggregation, 339 association behavior, 342, 342f, 348–349 band pattern, 338 dissociation, 340 heat stability, 339 κ-casein addition, 347, 347t nonsedimentable protein, 339–340, 343–344, 343f pH, 339–340 pH at heating effects, 347, 348f αS2-casein interaction, 338 sequence of events, 345–347, 346f

734 Denatured whey proteins-κ-casein interactions (Continued) serum-phase proteins, 344–345 temperature and pH vs. protein level, 340, 341f temperature and rate of heating, 338–339 in model systems aggregation, 336–337 complex formation, 336 intermediate mobility, 336 irreversible thermal denaturation, 337 noncovalent bonding, 337 sedimentation coefficients, 336 thiol-disulfide interactions, 337 Depletion flocculation, 478–480, 502 Diacylglycerol acyltransferase (DGAT1) gene, 147 Dietary essential amino acids aging populations, 11–12, 12f countries deficient in, 10, 11t demographic changes, 11–12 food items, 10–11, 11f lysine, 10 recommended daily intakes, 5, 6t Digestibility, 6–7, 7f Digestible indispensable amino acid score (DIAAS), 68–69, 637, 656–657, 692 Digestion of milk proteins fecal vs. ileal digestibility, 702, 702t retention of caseins, 702, 703f Disulfide bonding, 35, 181–183, 183–184f Drying of proteins casein, 439–440 dairy protein concentrates and powders, 434–435 desired properties, 433–434 desorption drying slopes, 435, 436t relative humidity (RH), 435 solubilization, 437 water diffusion, 435–436 water transfer rate, 436–437 preheating conditions, 433 processing implications, 437 stability, 434 water availability, 437–438 whey protein nitrogen index (WPNI), 433 whey proteins, 438–439, 440t whey proteins denaturation, 433–434 Dual-binding model for micelle, 328 and acid gel formation, 241–243, 242f block copolymer, 226–227 concentrated micellar dispersions, 234–237, 236f disrupter of hydrophobic interactions, 225–226 and ethanol stability, 239–241 and micellar destabilization, 237–238 and micellar interactions, 233, 234f

Index

nanocluster bridging pathway, 227 and rennet curd formation, 238–239 residue side chains, 226 second binding pathway, 226 size and appearance, 231–232 urea, pH, sequestrants, and temperature effects, 232–233 Dumas method, 53

E Early lactation protein (ELP), 109 Echidna (Tachyglossus aculeatus), 101–102 Edman sequencing, 180–181, 181f Elderly. See Aging populations Electrophoresis methods capillary electrophoresis, 56 isoelectric focusing, 56 lab on a chip, 57 native gel electrophoresis, 54 sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), 54–56 two-dimensional gel electrophoresis (2D GE), 57 urea-PAGE, 56 Electrospray ionization (ESI), 60–61 Electrospray mass spectrometric analysis, 726–727 Electrostatic interactions, 505 Emulsification, 193 Emulsions adsorption aggregated form, 473–474 homogenization/intense mixing, 469, 475 hydrolysis, 475 interfacial layer composition, 472 nanoemulsion droplets, 476, 477f positive charge, 474 properties of adsorbed layers, 469–470 protein concentration, 471–472, 471f protein load, 470 purified milk protein systems, 470–471 sodium caseinate emulsions, 472–473, 473f surface protein coverage, 474 unfolding and rearrangement, 470 formation, 469 lipid oxidation antioxidative properties, 485–486 continuous phase replacement with water, 486 nanoemulsion droplet system, 486 oxidative stability, 485 physical properties, 485 protein concentration, 485 under physiological conditions biochemical conditions in stomach, 487 digestion process, 486–487 emulsion structures, 488–490

Index

gastric lipase, 490 human gastric simulator (HGS), 488, 489f lipid digestion, 490–491 lipolysis, 486–487 saliva-emulsion droplets interaction, 487 small intestine, 490 process-induced changes adsorbed protein, 482 creaming stability, 483 denaturation and aggregation, 481 heat treatment, 482–483 high-pressure treatment, 484 ionic strength, 481–482 oil-in-water emulsions, 484 stabilizing effect, 482 surface protein composition, 483 ultrahigh pressure effect, 484 proteins-polysaccharide interactions in, 511–512, 512f, 513–515t bilayer emulsions, 512 covalent bonds, 511–512 emulsion stability, 511–512 flocculation, 511 lipid digestion, 512–515 multiple biopolymer layers, 515–516 stability biopolymer, 478 coalescence, 478 creaming, 476–477, 479, 479f definition, 476–477 flocculation, 478, 480 forces, 478 kinetic stability, 476–477 multilayered protein emulsions, 481 oil-water interface, 478–479 sodium caseinate solution, 480 whey proteins, MPC, and calcium caseinate, 480 Energy, 716–717, 717t Essential amino acids (EAAs), 657 Ethanol stability, 239–241 Eutherians, 99–101 Evolution/manipulation of bovine milk proteins bioactive peptides, 157–158 experimental modifications, 161–162 function of protein commercial value, 156, 156t developing gastrointestinal tract, 155 immunological functions, 157 lactoferrin, 157 osteopontin, 157 sensory properties, 155–156 genetic variants A1 and A2 variants of β-casein, 159–160

735

A and B variant of α- and κ-casein, 160 milk production and quality, 158–159 phosphorylation, 152–153 structure and genes β-casein gene (CSN2 casein beta), 154 β-Lg (PAEP progestagen-associated endometrial protein/beta-lactoglobulin), 155 κ-casein gene (CSN3 casein kappa), 154 α-La gene (LALBA lactalbumin alpha), 155 αS1-casein gene (CSN1S1 casein alpha-s1), 154 αS2-casein gene (CSN1S1 casein alpha-s2), 154 value of milk, 162–163 Exporters of dairy products, 15, 16t

F Fat-filled milk powder/enriched milk powder (FFMP/EMP), 74 Fat separation/creaming, 406–407 Fatty acids arachidonic acid, 547 binding site, 545, 546f cytotoxic effects, 548 holo-α-LA, 548 organic anion binding, 547–548 relative binding strengths, 547 retinol, 547 triglycerides, 545 Fecal vs. ileal digestibility, 702, 702t Feedback inhibitor of lactation (FIL), 124 Ferroxidase (EC 1.16.3.1), 48 Flavors, 551–552, 553t Flocculation, 478 Flow cytometry, 726 Foaming, 193–194 Folate, 150 Follicular dendritic cell-secreted peptide (FDCSP) gene, 213–214 Food and Agriculture Organization (FAO), 1–2 Food consumption patterns, 2 Food fortification, 543 Food insecurity, 2 Food Price Index, 1–2 Food safety and traceability, 721–722 Foraging lactation strategy, 122–123, 123f Forewarming, 241–242 Fouling, 388–389, 393–394 Fourier transform infrared (FTIR) spectroscopy biomolecule-specific accessories, 63 IR absorption bands, 62–63, 62t technique, 62 vibrations in peptide bonds, 63, 63f Fractionation of milk, 39, 424, 424f

736

Index

Functional foods, 16–17, 99–100. See also Bioactive peptides amino acids arginine, 640 calories and satiety, 642–643 conditionally essential dietary amino acid, 640 dietary protein quality, 637 glutamine, 641 ileal digestibility, 636–637, 637t lysine bioavailability, 637, 638t in milk proteins and average protein, 641, 641t nutritionals, 638–639 physiological roles, 639–640 plant proteins vs. milk proteins, 637, 639f taurine, 640 in vivo determination, 636 bioactive component, 634 definition, 634 genetic and epigenetic control, 633–634 health and vitality, 635 holistic properties of, 646 nutritional and nonnutritional physiological roles, 634 physiological processes, 633 proof of cause and effect, 634–635 regulatory authorities, 635 Functional genomics, 147–148 Functionality, 100, 294–295, 295f Functional properties of milk in food system, 467, 468t separately denatured whey proteins from casein micelles, 374–376 whey protein denaturation firmness and flow properties, 354–355, 355f heat treatment of milk, 353–354 syneresis, 354 tailor-mademilk powders, 353 temperature and heating time, 354 water-holding capacity, 354 whey proteins with κ-casein/casein micelles acid-induced aggregation/gelation (see Acidinduced aggregation/gelation) chymosin-induced aggregation/gelation of heated milk, 365–366, 366–367f, 368 concentrate viscosity, 372–374, 374f sedimentation and age gelation, 368–369, 370–371f, 371–372 Fur seal family, 122–123 feedback inhibitor of lactation (FIL), 124 foraging lactation strategy, 122–123, 123f local apoptotic milk factors apoptotic staining, 127 gene expression, 128, 129f

immunoglobulins, 130 involution, 126–127 α-lactalbumin protein, 127–128 matrix metalloproteinases (MMPs), 128–130 Na-dependent transporter of taurine (TAUT), 128 proapoptotic genes, 128 tissue inhibitors of matrix metalloproteinases (TIMPs), 128–130 unannotated hypothetical genes, 130 milk production and mammary gland function, 123–124, 125–126f suckling and foraging expression profiles, 124–126, 125–126f lactation cycle, transcript profile of, 124 reduced rate of milk production, 126

G Galactose-P, 24 β-Galactosidase, 25 Gastrointestinal (GI) digestion bile salts, 673–674 caseins and whey proteins, digestion behavior, 671–672 coagulation clot structures, 675–676, 676f fat globules, 676–677, 681–682, 682f human gastric simulator (HGS), 675–676 hydrolysis rate, 675 κ-casein band intensity, 676, 677f milk-clotting enzymes, 674–675 unheated skim milk, 675–676 whole milk, 676–677, 678f gastric fluid enzymes, 672–673 pH and ionic strength, 672 milk fat globule membrane (MFGM) proteins gastric proteolysis, 682–684 hydrolysis, 682–684 inhibitory effect, 684–685 lipid digestion, 685 pancreatic lipase, 684–685 milk protein ingredients dynamic gastric conditions, 686, 687f pepsin action, 686 sodium caseinate, 685–686 nonbovine milks casein contents, 689 clotting behavior, 689 comparative digestive utilization and metabolism, 692–693 comparative fat digestibilities, 693 composition, 688–689 consumer interest, 688

Index

curd formation, 692 deer vs. cow milk, 690 dipeptidyl peptidase-IV (DPP-IV) inhibitory activity, 691–692 enzymatic hydrolysis, 689 fat globules, 692–693 gastric and duodenal juices, 690 immunomodulatory activity, 692 infant and young child gastric conditions, 691 infant formulas, 688 metallic beads, 690–691 physicochemical properties, 688–689 radical scavenging activity, 691–692 raw and heated skim milk, 690 reconstituted milk, 690–691 structure, 688–689 true ileal amino acid digestibility, 692 zinc digestion, 689 pancreatic lipase, 674 processing treatments digestibility, 679 homogenization, 679–680 protein curd formation, 681, 682f protein hydrolysis rate, 680–681 structure of heated and unheated milk, 679, 680f whey-MFGM proteins interaction, 678–679 small intestine, pH and ionic strength, 673 Gastrointestinal (GI) tract digestion and absorption cheese matrix degradation, 711 coagulation (liquid/gel/solid transition) of milk, 709–711, 709–710f exogenous nitrogen delivery, 709, 709f heat treatment of milk, 707–708 homogenization of milk, 708 nutritional quality, 711 physicochemical modifications of proteins, 708 postprandial portal absorption, 709–710 protein digestion and amino acid absorption, 710–711, 710f digestion of milk proteins fecal vs. ileal digestibility, 702, 702t retention of caseins, 702, 703f hydrolysis in intestinal lumen caseins, 703–704, 704f whey proteins, 705 overall nutritional efficiency, 701 peptides during digestion biological activities, 705 caseinophosphopeptides, 706 compositions and structures, 707 gastric and intestinal peptidomes, 706–707 postdigestion profiles, 706

737

Gel filtration/gel permeation chromatography, 47. See also Size exclusion chromatography Gelling phase-separated systems casein micelles and ι-carrageenan assembly of structures at surfaces, 525 attractive interactions, 524 gel strength and temperature, 524 macromolecules, 524 molecular orientation at interface, 525 elastic modulus, 520 gel strength, 519–520 interpenetrating, coupled/phase-separated networks, 521 κ-carrageenan and β-lactoglobulin, 521 storage modulus β-lactoglobulin and pectin, 520 sodium caseinate and pectin, 524 whey protein and galactomannans, 519–520 WPI and xanthan gum, 520 Gels acid-induced milk gels (see Acid-induced milk gels) casein micelles, 599 cheese and fermented milk products, 599 mixed gels, 616–617 rennet-induced gels (see Rennet-induced gels) whey protein gels (see Whey protein gels) Gene pharming, 725 Genetically modified (GM) foods, 724–726 Genetic polymorphism, 34 Genomics. See also Evolution/manipulation of bovine milk proteins comparative milk genomics, 149–150 DNA sequence, 145 functional genomics, 147–148 genome map/structure, 145–146 1000 Genomes Project, 145 genomic tools, 144–145 origins of milk proteins, 150–152, 153f polymorphism, 146–147 transgenic technologies for milk manipulation, 148–149 Gestation, 100–101 Global demand for animal proteins, 9–10 Global Hunger Index (GHI), 4–5, 5f Global issues for food competition for land use, 716 energy, 716–717, 717t global warming on production, 716 methane emissions, 718–719, 720f resource issues, 715 water economy, 717–718, 718t

738 Global trade in proteins global consumption of protein, 14–15 milk products dairy cooperatives, 16 exporters, 15, 16t functional and organic foods, 16–17 human consumption, 15 intraindustry trade, 15 milk powders, 16 regional production, 15 source of protein, 15 standards of living, 16–17 Global warming, 716 Glutamine, 641 Glutathione peroxidase (GTPase), 48 Glycodelin, 260 Glycomacropeptide (GMP), 600 Glycoproteins, 49 Glycoproteomics, 186 Glycosylation of κ-casein, 35 dromedary, 202 Edman sequencing, 180–181 equine, 201 heterogeneity, 180–181, 181f human, 197–201, 200t N-glycosylation, 179–180 O-glycosylation, 179–180 ovine and caprine, 201, 201f tetrasaccharide, 180–181, 182f UniProt database, 197, 197–199t water buffalo, 202 Goat milk infant formula (GIF), 692 Green Revolution, 1 Growth factors, 50 Guanidinium chloride (GdmCl) β-lactoglobulin (β-Lg), 268 serum albumin (SA), 276

H Hairy micelle model, 237–238 Haptocorrin, 111 Heat treatment of milk, 192 Heritability, 189 Heterogeneity, casein κ-casein phosphorylation biological significance, 194–195 functionality, 189–194 sources, 187–189 αs-casein phosphorylation functionality, 186 sources, 183–186, 185f High-pressure processing (HPP) casein micelles calcium phosphate nanoclusters, 297

Index

light transmission, 296 particle size and turbidity, 298 reformation, 297–298 solubilization, 297 transmission electron micrographs, 298, 299f weak interactions, 297 technological properties and quality, 295–296 whey proteins, 298–300 whey proteins-casein micelles interactions aggregate formation, 300–301 disulfide-bonded dimers, 301 sedimentable whey proteins, 301 Hi-pH gel, 243 Holt mechanism, 228 Human α-lactalbumin made lethal to tumor cells (HAMLET), 127–128, 272 Human gastric simulator (HGS), 488, 489f, 675–676 Human health and milk protein atherosclerosis, 658–659 blood pressure and vascular reactivity, 659–660 bone health, 660–661 hyperglycemia, 653 infant health, 661–663 metabolic health, 652–653 muscle wasting and sarcopenia absorption and retention rates, 657 anabolic resistance, 656–657 essential amino acids (EAAs), 657 function improvements, 657–658 insulin stimulatory effect, 656–657 resistance-type exercise, 658 obesity and weight control acute and chronic appetite responses, 654–655 energy-restricted diet, 655 high-protein diets, 653–654 sarcobesity, 655 skeletal muscle, 654 whey protein and casein supplementation, 655–656 outcomes, 652 type 2 diabetes, 653 Hunger childhood stunting, 3 definition, 2 energy supply from food, 2 food insecurity, 2 Global Hunger Index (GHI), 4–5, 5f hunger-reduction targets, 3 World Health Assembly Resolution 65.6, 3 world hunger and undernutrition status, 3–4, 4f Zero Hunger, 3 Hydration values, 221–222 Hyperglycemia, 653

Index

I Immunity through milk, 101 Immunoglobulin (IgG), 41–42, 81 functions, 276 structure, 276, 277f temperature, pressure, and chemical denaturants effects, 277 Indigenous milk enzymes, 50 Indispensable amino acids (IAAs), 67–68 Infant formula mixes, 428 Infant health, 661–663 Infrared (IR) spectroscopy, 53, 54f Ingredients, 21–22 Insolubility casein micelle insolubility, 442–443 commercial driers, 442 computational fluid dynamics (CFD), 442 dehydration process bonding mechanisms, 444 composition, 444 configuration change reaction, 443 intermicellar bonding, 443 moisture/protein ratio, 441, 441–442f technique, 440–441 Insolubility index (ISI), 456 Interspecies comparison of milk proteins caseins, 33 coagulation and gel-forming properties, 30–32 concentrations, 30 minor proteins of milk, 33 protein profile, 30, 31t urea-polyacrylamide gel electrophoretic analysis, 32–33, 32f Intraindustry trade, 15 Intrinsically disordered proteins (IDPs), 214–215, 220–221 Ion-exchange chromatography (IEX), 59 Iron (ferrous sulfate), 543 Isoelectric focusing, 56 Isoelectric precipitation, 46 Isoleucine-proline-proline (IPP), 659–660 Isopeptide bond formation, 425

K

κ-casein amino acid sequence, 177f, 178 biological significance bioactive peptides, 194–195 digestibility and bioavailability, 194 nutritional value of glycans, 194 disulfide bonding, 181–183, 183–184f functional significance acid coagulation, 192 casein micelle size, 189–190 emulsification, 193

739

foaming, 193–194 functionality in cheese, 190–192 heat treatment, 192 ultrahigh-temperature milk, 192–193 genetic variants, 178, 179t glycosylation, 178 (see also Glycosylation of κ-casein) phosphorylation, 178–179 sources of heterogeneity caseinomacropeptide, 187–188 dry period, 188–189 heritability, 189 hexosamine and sialic acid, 187–188 N-acetylglucosamine (GlcNAc) and fucose, 188 neuraminic acid (NeuAc) content, 187 nonglycosylated κ-casein, 187 polymorphism and yield, 187 seasonal factors, 189 Kininogen, 49 Kjeldahl method, 51–53, 52t

L Lab on a chip, 57 α-Lactalbumin, 24, 40–41, 127–128 Lactation, 99–100 Lactation cycle, 100–102, 102f, 107–109, 108f Lactoferrin (Lf), 47–48, 111, 157, 161, 544 commercial applications, 278 structure, 278–279, 279f temperature, pressure, and chemical denaturants effects, 279 α-Lactoglobulin (α-Lg) denaturants effects, 272 molecular structure, 270–271, 270f pressure effects, 271–272 structure, 269 temperature effects, 271 β-Lactoglobulin (β-Lg), 39–40, 151 in aqueous solution dissociation constant, 256 equilibrium and rate constants, 255–256, 256t monomer-dimer equilibrium, 255–256, 256t nuclear magnetic resonance (NMR) studies, 255 structure, 252–255, 254f caprine (goat), 260 chemical denaturants guanidinium chloride (GdmCl), 268 2,2,2-trifluoroethanol (TFE), 267–268 urea, 268 dynamics of protein, 258–259 equine, 259 fibrillar formation, 268–269 genetic variants, 144, 151–152 ligand binding affinity, 263 binding sites, 264

740 β-Lactoglobulin (β-Lg) (Continued) circular dichroism (CD) spectrum, 261 under different ionic strength, 262 dodecanoate and dodecyl sulfate, 262–263, 263t fatty acids and retinoids binding, 261–262 fluorescence spectroscopy, 261 nuclear magnetic resonance (NMR) techniques, 260–261 palmitic acid binding, 261 pH-dependent conformational change, 261 vitamin D3, 263–264 as ligand carrier, 538–539 at neutral pH Ala34Cys mutant, 258 ligand-binding sites, 257–258, 257f monomeric species, 257–258 spectra broadening, 256 ovine (sheep), 260 porcine, 259–260 pressure effect, 265–267 rangiferine (reindeer), 260 thermal properties, 253, 265 for vitamin D, 541 Lactollin, 48 Lactoperoxidase, 80–81 Lactose β-galactosidase, 25 characteristics of, 24–25 composition and concentration, 24 function in milk, 24 galactose-P, 24 α-lactalbumin, 24 Lactose synthetase, 24 Lactosylation, 410 Late lactation protein A (LLP-A), 109 Leucine, 13–14 Lipids biohydrogenation, 23 globules, 24 milk fat globule membrane (MFGM), 24 miscellaneous lipids, 23 neutral lipids, 23 polar lipids, 23 ruminant milk fats, 23 Lipocalin, 260 Liquid chromatography coupled with mass spectrometry (LC-MS), 59 Localism, 720 Low birth weight infants, 662 Lung morphogenesis, tammar, 110 Lysine Maillard reaction, 10 phosphoserine, 10 Lysozyme, 81

Index

M Maillard reaction, 10, 425, 445, 445f browning, 411, 411t cross-linking, 412–413 deamidation, 411–412 lactosylation, 410 Mass spectrometry (MS) analytical capability, 61 description, 60 electrospray ionization (ESI), 61 liquid chromatography coupled with mass spectrometry (LC-MS), 59 matrix-assisted laser desorption ionization (MALDI), 60–61 Mastitis, 116 Matrix-assisted laser desorption ionization (MALDI), 60–61 Meat products, 588–589 Metabolic health, 652–653 Metal-binding proteins, 47–48 Methane emissions, 718–719, 720f Micellar casein concentrate (MCC), 76 Micelle, casein. See also Concentrated micellar dispersions acid gel formation, 241–243, 242f aggregation, 220–221 amphipathic nature, 216 assembly and structure, 214–215 charge density, 216 dual-binding model (see Dual-binding model for micelle) electrostatic repulsion, 220–221 ethanol stability, 239–241 formation and stability, 173–174 gelation, 599 (see also Gels) high-pressure processing (HPP) (see High-pressure processing (HPP)) hydrophobic bonds, 221 hydrophobic clusters, 217, 218f hydrophobic interactions, 217–219, 220f hydrophobicity, 216–217 micellar destabilization, 237–238 models, 327f colloidal calcium phosphate (CCP), 326, 329 dual-binding model, 328 hydrophobic interactions, 329 micelle stability and destabilization, 329 nanocluster model, 328 primary casein particles (PCPs), 329 structural arrangements and bonding, 329 submicelle models, 326 nanocluster model of Holt, 225 phosphoserine clusters, 215–216 physical and technological properties, 214

Index

properties calcium phosphate equilibria, 224, 232–233 fractionation, 224 homogeneous protein distribution, 222–223 hydration values, 221–222 nanoclusters, 232 osmotic stress technique, 223 scattering behavior, 222–224 size and appearance, 231–232 spherical calcium phosphate nanoclusters, 223–224 urea, 232 rennet curd formation, 238–239 schematic structures, 217, 219f size, 189–190 submicelle model, 225 SXE peptide, 215–216 two-dimensional hydrophobic cluster analysis (2D-HCA) plots, 217, 218f Microarrays, 146–147 Microbiological detection methods, 726 β2-Microglobulin, 48 Micronutrients, 538 MicroRNAs (miRNAs), 130–131 Milk fat globule membrane (MFGM), 24, 26, 50, 607 gastric proteolysis, 682–684 hydrolysis, 682–684 inhibitory effect, 684–685 lipid digestion, 685 pancreatic lipase, 684–685 Milk powder, 424–425. See also Spray drying; World dairy powder Milk protein concentrate (MPC), 75 Milk protein isolate (MPI), 75 Minerals, 543–544, 544f, 545t Minor milk proteins angiogenins, 49 application, 80–81 biologically active cryptic peptides, 51 β2-microglobulin, 48 functional properties, 80 glycoproteins, 49 growth factors, 50 indigenous enzymes, 50 kininogen, 49 metal-binding proteins, 47–48 milk fat globule membrane (MFGM) protein, 50 nonprotein nitrogen (NPN), 51 osteopontin, 48 powdered ingredients, 80 vitamin-binding proteins, 49 Miscellaneous lipids, 23 Mixed gels, 616–617 Mixing behavior of biopolymers

741

complex coacervation, 502 cosolubility, 500–501 depletion flocculation, 502 interaction types, 500f thermodynamic incompatibility, 501 Model food systems advantages, 582 applications bakery products, 584–586 coffee whiteners, 586–587, 587t functionality in foods, 584, 585t meat products, 588–589 protein nutrition bars, 589–590 salad dressings, 588 whipped toppings, 587–588, 587t development, 583 food characteristics, 574 food protein manufacture, 582f interactions in food characteristics emulsifier EM, 576f food product diagram, 575, 575f food texture, 576–577 individual milk casein fractions, 578–579, 578t salts, 575–576 sodium caseinate-starch mixtures, 577, 578f sodium caseinate with protein and starch concentrations, 579–580, 580–581f, 580t starch, 576–577 starch pasting curve for potato starch, 577, 577f limitations, 591 processing effects, 580–581 protein functionality, 574–575, 575t protein systems, 573–574 statistical design, 583–584 uses of, 581–582, 590–591 Monotremes composition of milk, 102–103 EchAMP antibacterial assays, 103–105, 106f egg-laying mode of reproduction, 101–102 functional attributes, 103–105 gene expression, 103, 104f hydrophilic and secretory proteins, 103 maternal care, 101–102, 102f MLP antibacterial assays, 103–105, 107f MLP protein structure, 103–105, 105f nipples, 103–105 protective properties of milk, 103 Mozzarella cheese, 711 MPC, 468 mTORpathway, 14 Multidimensional nuclear magnetic resonance spectrometers, 66 Muscle wasting and sarcopenia

742

Index

Muscle wasting and sarcopenia (Continued) absorption and retention rates, 657 anabolic resistance, 656–657 essential amino acids (EAAs), 657 function improvements, 657–658 insulin stimulatory effect, 656–657 resistance-type exercise, 658

N N-acetylglucosamine (GlcNAc), 188, 201 Na-dependent transporter of taurine (TAUT), 128 Nanoemulsion droplets, 476, 477f Nanofibril system, 727 Nanotechnology, 727 National Dairy Development Board (NDDB), 16 Native gel electrophoresis, 54 Neonate nutritional requirements, 21 Net protein utilization (NPU), 67 Neuraminic acid (NeuAc), 187, 194 Neutral lipids, 23 Nomenclature of milk proteins, 29 Nonbovine milks casein contents, 689 clotting behavior, 689 comparative digestive utilization and metabolism, 692–693 comparative fat digestibilities, 693 composition, 688–689 consumer interest, 688 curd formation, 692 deer vs. cow milk, 690 dipeptidyl peptidase-IV (DPP-IV) inhibitory activity, 691–692 enzymatic hydrolysis, 689 fat globules, 692–693 gastric and duodenal juices, 690 GI transit, 690–691 immunomodulatory activity, 692 infant and young child gastric conditions, 691 infant formulas, 688 physicochemical properties, 688–689 radical scavenging activity, 691–692 raw and heated skim milk, 690 reconstituted milk, 690–691 structural properties, 688–689 true ileal amino acid digestibility, 692 zinc digestion in milks and baby food, 689 Nonenzymic browning. See Maillard reaction Nongelling phase-separated systems bridging flocculation, 523 casein micelles and galactomannans, 518–519 and pectin, 523 depletion flocculation, 519

β-lactoglobulin and chitosan, 522 milk proteins and xanthan gum, 519 shear-thinning behavior, 522 sodium caseinate and gum arabic, 522–523 stable composite particles, 522–523 thixotropic behavior, 518–519 viscosity, 518–519 whey proteins and exopolysaccharides, 522 and gum arabic, 522 Nonprotein nitrogen (NPN), 51 Nonspecific steric exclusion, 550 Nonthermal processing. See also High-pressure processing (HPP); Pulsed electric field (PEF) processing; Ultrasound (US) processing; Ultraviolet (UV) irradiation processing concept, 293–294 functionality, 294–295, 295f Nuclear magnetic resonance (NMR) spectroscopy, 65–66 Nutrigenomics allergies to milk, 723 A2 milk, 723 bioactive peptides, 724 milk proteins and gut microbiome, 724 Nutritionals, 638–639 Nutrition bars, 589–590

O Obesity and weight control acute and chronic appetite responses, 654–655 energy-restricted diet, 655 high-protein diets, 653–654 sarcobesity, 655 skeletal muscle, 654 whey protein and casein supplementation, 655–656 Odontogenic ameloblast-associated (ODAM) gene, 213–214 Oleic acid treatment, 272 Oligosaccharides, 26 Organic foods, 16–17 Origins of milk proteins, 150–152, 153f Osteopontin, 48, 81, 157 Overall nutritional efficiency, 701

P Parmigiano-Reggiano cheese, 711 Passive immunity through milk, 101 Personalized nutrition, 722–723 pH and ionic strength of gastric fluid, 672 Phosphocasein. See Micellar casein concentrate (MCC) Phosphorylation β-casein bovine, 176f, 177–178, 195–196 donkey, 197

Index

elephant, 197 equine, 196 human, 196 ovine and caprine, 196, 196t water buffalo, 197 κ-casein, 178–179 αs-casein, 183, 185–186 Phosphoserinereaction with lysine, 10 Pickering stabilization, 515–516 Pinnipeds, 122–123. See also Fur seal Plant and animal protein sources, 8–9 Plant-derived milk substitutes, 720–721 Plasmin bitterness and age gelation, 397 caseins hydrolysis, 35 heat inactivation, 398 proteolysis, 396–398 system, 395–396 Platypus (O. anatinus), 101–102 Polar lipids, 23 Polyacrylamide gel electrophoresis (PAGE), 332 Polymorphism, 146–147 Polyphosphate, 406 Postmeal aminoacidemia, 702 Posttranslational modifications (PTMs) of caseins. See also Heterogeneity, casein amino acid sequences, 175, 176f β-casein, 177–178 κ-casein amino acid sequence, 177f, 178 disulfide bonding, 181–183, 183–184f genetic variants, 178, 179t glycosylation, 178 (see also Glycosylation of κ-casein) phosphorylation, 178–179 micelle formation and stability, 173–174 αs1-casein, 175–176 αs2-casein, 176–177 two-dimensional gels, 174–175, 174f Pouch young (PY) gut development, 110 lung morphogenesis, 110 mammary gland, 107–109 milk composition and milk production, 110 sucking pattern and milk secretion, 110 Powdered milk protein ingredients by-product and coproduct, 71, 72f casein ingredients acid casein, 76 applications, 77 caseinates, 76 functional properties, 76–77 micellar casein concentrate (MCC), 76 milk protein concentrate (MPC), 75 milk protein isolate (MPI), 75 rennet casein, 76

743

composition of, 71, 73t milk-based infant formula, 69 milk protein ingredients applications, 74 fat-filled milk powder/enriched milk powder (FFMP/EMP), 74 functional properties, 74, 75f skim milk powder (SMP), 74 whole milk powder (WMP), 71–74 whey protein ingredients applications, 80 functional properties, 79 whey powder (WP), 78 whey protein concentrate (WPC), 78 whey protein hydrolyzates (WPHs), 79 whey protein isolate (WPI), 78–79 Precipitation by ethanol, 47 Primary casein particles (PCPs), 329 Properties of milk proteins, 29f casein, 27–28 pioneering work, 27–28 proteose peptones and nonprotein nitrogen (NPN), 28–29 whey/serum proteins, 28 Prosaposin, 49 Proteinaceous sediment, 388–389 Protein digestibility corrected amino acid score (PDCAAS), 68, 637, 656–657 Protein efficiency ratio (PER), 67 Protein-micronutrient interaction alpha-crystallin, 552 bioaccessibility and bioavailability, 537–538 bioactive compounds, 552–555, 553–554t chlorpromazine, 552 electrostatic driving forces, 538 fatty acids arachidonic acid, 547 binding site, 545, 546f cytotoxic effects, 548 holo-α-LA, 548 organic anion binding, 547–548 relative binding strengths, 547 retinol, 547 triglycerides, 545 flavor compounds, 551–552, 553t folate-binding proteins, 542 hydrophobic interaction, 538 minerals, 543–544, 544f, 545t processing effect high-pressure treatment, 555 ligands, 556–557, 557f pasteurization, 555 protein denaturation, 555–556 sugars/polyols, 558–560

744 Protein-micronutrient interaction (Continued) protein structure, 537–538 riboflavin-binding protein, 542 sugars and polyols, 550–551, 550f surfactants, 548–549 vitamin A binding constants, 539 cis-trans isomerization, 538 in commercial skim milk, 540 free retinol, 539 β-lactoglobulin (β-LG), 538–539 whey proteins, 539–540 vitamin C, 540 vitamin D, 541–542, 541f Protein-polysaccharide interactions in aqueous phase, 506, 507–510t attractive interactions, 504–505 complex coacervation, 502 cosolubility, 500–501 covalent bonds, 505–506 depletion flocculation, 502 in emulsion systems, 511–512, 512f, 513–515t, 515–516 interaction types, 500f interfacial structures, 506–507, 511f multilayered interfacial membranes, 510–511 phase diagram binodal curve, 504 net repulsion, 503 one-phase and two-phase system, 502–503 segregating biopolymer system, 503, 503f repulsive interactions, 504 rheological properties and microstructures interacting protein-polysaccharide mixtures, 521–525 intermolecular interactions, 516, 518f noninteracting protein-polysaccharide mixtures, 517–521, 517f rheological techniques, 516 thermodynamic incompatibility, 501 Proteose peptone (PP) 3, 42–43 Pulsed electric field (PEF) processing advantages, 307 casein micelles coagulation properties, 311 micelle size, 307–311 equipment and techniques, 306 outcome of, 306–307 preservation, 306–307 structure and functionality of milk proteins, 307, 308–310t whey proteins, 311–312

Index

Q Quality assessment biological methods, 67 chemical methods, 68 digestibility and bioavailability, 67 indispensable amino acids (IAAs), 67 mixed assays, 68–69, 70t nitrogen balance in human nutrition, 66–67, 67f

R Recaldent, 158 Recommended daily intakes of proteins, 6t Recommended dietary allowance (RDA), 13 Reconstitution techniques, 459–460 Rehydration of protein powders dissolution, 456 granulation, 459 insolubility index (ISI), 456 MCP, 452t, 457–458 nuclear magnetic resonance (NMR) spectroscopy, 456–457 particle size distribution, 457 protein structure, 458 reconstitution techniques, 459–460 solubilization, 458 turbidity profiles, 457 viscosity measurement, 457 whey powders, 459 Rennet casein, 76 Rennet coagulation, 47 Rennet coagulation time (RCT), 190–192 Rennet curd formation acidification, 238–239 aggregate growth, 238 elasticity, 238–239 gel strength, 239 microstructure, 239 Rennet-induced gels, 239 monitoring, 603 primary phase, 600, 601f rheological properties, 603–604 secondary phase aggregation reaction, 601 enzymatic phase, 601–602 gel strength, 602–603 micrograph, 601–602, 602f plasmin, 602 steric stabilization, 601 syneresis, 604–606 texture camel vs. calf chymosin, 607 colloidal calcium phosphate (CCP), 607 Cynara sp. flower, 606 enzymatic cross-linking of caseins, 608

Index

high hydrostatic pressure, 608–609 milk fat globule membrane (MFGM), 607 milk heat treatment, 607–608 pH and NaCl concentration, 606 starch addition, 609 Repulsive interactions, 504 Reversed phase-high-performance liquid chromatography (RP-HPLC), 57–58, 58f Rheological properties casein micelles and galactomannans, 518–519 and ι-carrageenan, 524–525 and pectin, 523 β-lactoglobulin and chitosan, 522 and κ-carrageenan, 521 and pectin, 520 sodium caseinate and gum arabic, 522–523 and pectin, 524 whey proteins and exopolysaccharides, 522 and galactomannans, 519–520 and gum arabic, 522 xanthan gum milk proteins and, 519 WPI and, 520 Riboflavin (B2), 27 Ruminant milk fats, 23

S S100A19 lactation cycle, 119–120, 121f stomach development, 119, 120f Salad dressings, 588 Salting-out methods, 46 Salts, milk, 26–27 Sarcobesity, 652, 655 Sarcopenia, 12–13, 658 Secretory calcium-binding phosphoprotein (SCPP) genes, 213–214 Sedimentation, 368–369, 370–371f, 371–372, 373f definition, 407 destabilization, 407–408 in directly processed milks, 408 fat presence, 407 fouling deposits, 408–409 heat stability, 409 proteolysis, 409 time of storage, 407 ultrahigh-temperature mineral-fortified milks, 407 Self-assembling systems, 727 Serum albumin (SA) chemical denaturants, 275–276 Cys34 reactivity, 274

pressure effects, 275 structure, 272–274, 273f temperature effects, 275 Single nucleotide polymorphisms (SNPs), 146–147 Size exclusion chromatography, 58–59 Skim milk powder (SMP), 74, 426 Sodium caseinate emulsions, 472–473, 473f Sodium dodecyl sulfate (SDS), 549 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) apparent molecular mass, 55 disadvantage, 55–56 gel preparation, 55 procedure, 54–55 Sodium hexametaphosphate (SHMP), 406 Sodium perfluorooctanoate, 549 Sorbitol, 558 Specialty foods, 16–17 Spray-dried milk products, 428, 429f Spray drying compact drier (CD), 432 drier components, 430–431, 430f droplet and air, 429 drying equipment, 432 drying kinetics, 429–430 drying of proteins casein, 439–440 dairy protein concentrates and powders, 434–435 desired properties, 433–434 desorption, 435–437, 436t preheating conditions, 433 processing implications, 437 stability, 434 water availability, 437–438 whey protein nitrogen index (WPNI), 433 whey proteins, 438–439, 440t whey proteins denaturation, 433–434 insolubility casein micelle insolubility, 442–443 commercial driers, 442 computational fluid dynamics (CFD), 442 dehydration process, 443–444 moisture/protein ratio, 441, 441–442f technique, 440–441 installation, 431 optimization, process, 432 parameters, 439f process improvement, 432–433 single-stage drying system, 431 spray-dried milk products, 428, 429f three-stage drying system, 431–432 two-stage drying system, 431

745

746

Index

Starch, 576–577 Storage of dry powders chemical reactions, 444–445 essential amino acids cysteine, 449–450 levels, 450, 458t methionine, 450 nutritional value, 450, 451t tryptophan, 450, 451f isopeptide bond formation, 449, 449f lactulosyl lysine formation rates, 446, 447–448f, 448 levels, 446 lysine, 451–453, 452f, 454f lysinoalanine formation, 449, 450f Maillard and pre-Maillard compounds formation, 445, 445f product-specific storage trials, 455–456, 455f solubility, 444 sulfur amino acids, 453 tryptophan, 453–454 Structural analysis circular dichroism (CD), 63–64 Fourier transform infrared (FTIR) spectroscopy, 62–63, 62t, 63f nuclear magnetic resonance (NMR) spectroscopy, 65–66 structural levels, 61–62 X-ray crystallography, 64–65, 64f Stunting, 3 Sugars and polyols, 550–551, 550f, 558–560 Surfactants, 548–549

T Tailor-mademilk powders, 353 Tammar wallaby (Macropus eugenii) bioactives identification, 111 temporal delivery of, 111 cathelicidin apoptosis, 114 bactericidal activity, 113 BMAP-28, 114 cell proliferation, 114 host-defense peptides, 111–112 immune cells, 114 MaeuCath1–8 gene, 113 PR-39, 114 sequence heterogeneity, 113 structure, 112–113, 112f temporal delivery of, 111 composition of milk, 107–110 concurrent asynchronous lactation, 107–109, 109f

gene expression, 109 lactation cycle, 107–109, 108f mammary function control, 121–122 pouch young (PY) gut development, 110 lung morphogenesis, 110 mammary gland, 107–109 milk composition and milk production, 110 sucking pattern and milk secretion, 110 S100A19 lactation cycle, 119–120, 121f stomach development, 119, 120f whey acidic protein (WAP) gene expression, 118 mammary development, 118 proliferative activity, 118, 119f structure, 117–118 whey four-disulfide core (WFDC) proteins antibacterial activity, 115–116, 117f commensal microbial flora in gut, 116–117 gene expression, 115, 116f immune protection, 116 structure of, 114–115, 115f Tamm-Horsfall protein-1 (THP-1), 692 Taurine, 640 Tear lipocalin, 260 Technologies analytical methods, 726–727 genetic modification, 724–726 materials science and nanotechnology, 727 nonthermal processes, 726 Temporal gene knockout, 109 Thermodynamic incompatibility, 501 Thermohygrometer, 432–433 1000 Genomes Project, 145 Timasheff’s group, 551 Total protein determination Dumas method, 53 infrared (IR) spectroscopy, 53, 54f Kjeldahl method, 51–53, 52t 2,2,2-Trifluoroethanol (TFE) α-lactoglobulin (α-Lg), 272 β-lactoglobulin (β-Lg), 267–268 Two-dimensional gel electrophoresis (2D GE), 57 Type 2 diabetes (T2D), 652–653

U UDP-galactosyl transferase, 24 Ultracentrifugation, 46 Ultrafiltration and microfiltration, 46 Ultrahigh-temperature (UHT) processing age gelation

Index

causes, 404–405 fat separation/creaming, 406–407 mechanism, 406 onset, 405–406 retardation, 406 aseptic packaging, 388 bacteriological indices, 386 chemical index, 386 commercial sterility, 386 direct and indirect heating, 386 homogenizer, 388 proteolysis bacterial proteases, 398–402 enzymatic proteolysis, 395 heat-induced proteolysis, 404 plasmin, 395–398 reducing enzymatic proteolysis, 403 UHT hydrolyzed-lactose milk, 404 sedimentation (see Sedimentation) temperature-time profiles, 385–387, 387f UHT milk, 192–193 whey protein denaturation denaturation process, 389–390 flavor production, 394 fouling of heat exchangers, 393–394 hydrogen bonds and hydrophobic interactions, 389 immunoglobulins, 389 measurement, 392–393 in preheat and high-temperature sections, 391–392, 392t reducing effects, 394–395 in UHT milk, 391, 391t Ultrasound (US) processing advantage, 302 categories, 302 stabilizing systems, 302 structure and functionality of milk proteins integrity of casein micelles, 302–303 particle size, 302 rehydration, 305 renneting properties, 305 SDS-PAGE profile, 303, 304f secondary structure, 303–304, 304t solubility, 305 Ultraviolet (UV) irradiation processing disinfection and pasteurization, 313 on milk proteins Fourier transform infrared spectra, 314–315, 314f photooxidation, 315–316 secondary and/or tertiary structure, 315–316, 315t sodium caseinate, 314–315, 314f quality, 313 sterilization of cheese whey, 313

747

technique, 312–313 Undernutrition/undernourishment, 3–5, 4f United Nations 2030 Agenda for Sustainable Development, 3 Urea, 268 Urea-PAGE, 56 Uses for dried milk protein, 426

V Valine-proline-proline (VPP), 659–660 van der Waals forces, 505 Virtual water, 717–718, 718t Vitamin A binding constants, 539 β-lactoglobulin (β-LG), 538–539 cis-trans isomerization, 538 in commercial skim milk, 540 free retinol, 539 whey proteins, 539–540 Vitamin-binding proteins, 49 Vitamin C, 540 Vitamin D, 541–542, 541f Vitamins, 27

W Water economy, 22, 717–718, 718t Whey acidic protein (WAP), 42 origins of gene structure, 152, 153f tammar gene expression, 118 mammary development, 118 proliferative activity, 118, 119f structure, 117–118 Whey four-disulfide core (WFDC) proteins tammar antibacterial activity, 115–116, 117f commensal microbial flora in gut, 116–117 gene expression, 115, 116f immune protection, 116 structure of, 114–115, 115f Whey products and casein, 426, 428 Whey protein denaturation. See also Functional properties of milk assessment polyacrylamide gel electrophoresis (PAGE), 332 whey protein nitrogen index (WPNI) method, 331 definition, 330–331 irreversible aggregation processes, 331 kinetic evaluation and modeling Arrhenius plots, 332–333, 333f concentration effects, 334–335, 335f pH, 334

748

Index

Whey protein denaturation. (Continued) rate-determining step, 333–334 retardation, 335–336 temperature and heating time, 332–334 total whey protein, 334 reversible dissociation, 330 simple reaction scheme, 330 ultrahigh-temperature (UHT) processing flavor production, 394 fouling of heat exchangers, 393–394 hydrogen bonds and hydrophobic interactions, 389 immunoglobulins, 389 level of whey proteins, 391, 391t measurement of denaturation, 392–393 in preheat and high-temperature sections, 391–392, 392t process, 389–390 reducing effects, 394–395 Whey protein gels cold gelation, 623–624 composition, 617–619, 618t enzymatic modification, 624 fibril formation, 622 formation and properties, 619, 619t gelling time, 623 intermolecular repulsion, 621–622 pH and ionic strength, 622 salt effect on, 622–623 storage modulus, 622 thermal denaturation aggregation and formation, 620, 620f fractal aggregation theory, 621 irreversibility, 620 β-lactoglobulin, 620–621 Monte Carlo computer simulations, 621 pH and temperature, 619 types, 621 whey protein conjugates, 624 whey protein isolate (WPI) production, 617–619 Whey protein nitrogen index (WPNI), 331, 354, 433 Whey proteins. See also Immunoglobulin (IgG); Lactoferrin (Lf); α-Lactoglobulin (α-Lg); β-Lactoglobulin (β-Lg); Serum albumin (SA) β-lactoglobulin, 39–40 blood serum albumin, 41 vs. casein amino acid composition, 45 binding properties, 45 coagulability, 45

heat stability, 45 hydrophobicity, 43–44 physical state in milk, 46 preparation, 46–47 principal properties, 43, 44t site of biosynthesis, 46 solubility, 45 specific volume, 44 stability, 44 structures, 43 sulfur content, 46 tendency to associate, 44 commercial product preparation, 39 composition, 252t fractionation, 39 high-pressure processing (HPP), 298–300 immunoglobulins, 41–42 α-lactalbumin, 40–41 powdered ingredients (see Powdered milk protein ingredients) properties, 28 proteose peptone 3 (PP3), 42–43 pulsed electric field (PEF) processing, 311–312 ultrahigh-temperature (UHT) processing (see Whey protein denaturation) whey acidic protein (WAP), 42 Whipped toppings, 587–588, 587t Whole milk powder (WMP), 71–74, 426 World dairy powder exporters and importers, 425, 425t isopeptide bond formation, 425 Maillard reaction, 425 reconstituted and recombined milks, 426 skim milk powder (SMP), 426 storage times, 425 trade flows, 427f whey products and casein, 426, 428 whole milk powder (WMP), 426 World food prices, 1 World protein supply and regional distribution, 7–8, 8f

X X-ray crystallography, 64–65, 64f

Y Yogurt gels, 610–616

Z Zero Hunger, 3