Brewing Yeast Fermentation Performance [2 ed.] 0632064986, 9780632064984

Table of contents :
i-xxv.pdf......Page 1
1-12.pdf......Page 24
13-21.pdf......Page 35
23-38.pdf......Page 44
39-45.pdf......Page 59
46-60.pdf......Page 66
61-73.pdf......Page 81
75-85.pdf......Page 94
86-95.pdf......Page 104
96-109.pdf......Page 114
110-119.pdf......Page 128
120-128.pdf......Page 138
131-137.pdf......Page 147
138-148.pdf......Page 155
149-161.pdf......Page 166
162-168.pdf......Page 179
169-173.pdf......Page 186
174-179.pdf......Page 191
181-195.pdf......Page 197
196-205.pdf......Page 211
206-212.pdf......Page 221
213-233.pdf......Page 228
234-248.pdf......Page 249
249-256.pdf......Page 264
257-271.pdf......Page 271
272-280.pdf......Page 286
281-292.pdf......Page 295
293-301.pdf......Page 307
303-308.pdf......Page 316

Citation preview

Brewing Yeast Fermentation Performance

Brewing Yeast Fermentation Performance Second edition Edited by KATHERINE SMART Oxford Brookes University Oxford, UK

© Blackwell Science 2003 Blackwell Science Ltd, a Blackwell Publishing Company Editorial Offices: 9600 Garsington Road, Oxford OX4 2DQ, UK Tel: +44 (0)1865 776868 ++ Blackwell Science, Inc., 350 Main Street, Malden, MA 02148-5018, USA Tel: 1 1 781 388 8250 Iowa State Press, a Blackwell Publishing Company, 2121 State Avenue, Ames, Iowa 50014-8300, USA Tel: 1 1 515 292 0140 Blackwell Publishing Asia Pty, 550 Swanston Street, South Carlton, Victoria 3053, Australia Tel: 1 61 (0)3 9347 0300 Blackwell Wissenschafts Verlag, Kurf rstendamm 57, 10707 Berlin, Germany Tel: 1 49 (0)30 32 79 060 The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

First edition published 2000 Second edition published 2003

Library of Congress Cataloging-in-Publication Data is available

ISBN 0-632-06498-6 A catalogue record for this title is available from the British Library Typeset and produced by Gray Publishing, Tunbridge Wells, Kent Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall For further information on Blackwell Science, visit our website: www.blackwellpublishing.com

Contributors

A. Aitchison Scottish Courage Brewing Ltd, Technical Centre, 160 Canongate, Edinburgh EH8 8DD, UK P. Attfield Centre for Fluorimetric Applications in Biotechnology, Department of Biological Sciences, Macquarie University, North Ryde, Sydney, NSW 2109, Australia B. Axcell The South African Breweries Ltd, Corporate Technical Centre, PO Box 782178, Sandton 2146, South Africa C.W. Bamforth Department of Food Science & Technology, University of California, Davis, CA 95616-8598, USA F.F. Bauer Department of Microbiology and Institute for Wine Biotechnology, University of Stellenbosch, Stellenbosch 7600, South Africa H. Berg Oy Sinebrychoff Ab, PO Box 87, FI-04201 Kerava, Finland K. Berghof BIOTECON Diagnostics GmbH, Hermannswerder Haus 17, 14473 Potsdam, Germany C. Boulton Bass Brewers Ltd, Technical Centre, PO Box 12, Cross Street, Burton upon Trent DE14 1XH, UK W. Box Bass Brewers Ltd, Technical Centre, PO Box 12, Cross Street, Burton upon Trent DE14 1XH, UK A. Boyd Centre for Fluorimetric Applications in Biotechnology, Department of Biological Sciences, Macquarie University, North Ryde, Sydney, NSW 2109, Australia P. Chambers School of Food Science and Technology, Victoria University, Werribee Campus, PO Box 14428, Melbourne City, Victoria 8001, Australia M. Chandler School of Food Science and Technology, Victoria University, Werribee Campus, PO Box 14428, Melbourne City, Victoria 8001, Australia

vi

CONTRIBUTORS

S. Collin Université Catholique de Louvain, Unité de Brasserie et des Industries Alimentaires, Croix du Sud 2/7, B-1348 Louvain-la-Neuve, Belgium F.R. Dalvaux Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Katholieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium I. Dawes School of Biochemistry and Molecular Genetics, University of New South Wales, Sydney, NSW 2052, Australia A. Debourg Department of Brewing Sciences and Fermentation Technologies, Institut Meurice, 1 Avenue E. Gryson, B-1070 Brussels, Belgium F.R. Delvaux Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Kathalieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium G. Derdelinckx Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Kathalieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium J. R. Dickinson Cardiff School of Biosciences, Cardiff University, PO Box 915, Cardiff CF10 3TL, UK M. Dillemans Department of Brewing Sciences and Fermentation Technologies, Institut Meurice, 1 Avenue E. Gryson, B-1070 Brussels, Belgium J.-P. Dufour Department of Food Science, University of Otago, PO Box 56, Dunedin, New Zealand M. Fandke BIOTECON Diagnostics GmbH, Hermannswerder Haus 17, 14473 Potsdam, Germany L. Gijs Université Catholique de Louvain, Unité de Brasserie et des Industries Alimentaires, Croix du Sud 2/7, B-1348 Louvain-la-Neuve, Belgium X. Green Scottish Courage Brewing Ltd, Technical Centre, 160 Canongate, Edinburgh EH8 8DD, UK

CONTRIBUTORS

vii

S. Gualdoni School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane, Oxford OX3 0BP, UK T. Gunasekera Centre for Fluorimetric Applications in Biotechnology, Department of Biological Sciences, Macquarie University, North Ryde, Sydney, NSW 2109, Australia V. Higgins School of Biochemistry and Molecular Genetics, University of New South Wales, Sydney, NSW 2052, Australia J.A. Hodgson Scottish Courage Brewing Ltd, Technical Centre, Sugarhouse Close, 160 Canongate, Edinburgh EH8 8DD, UK G.A. Hulse The South African Breweries, Beer Division, Brewing Research & Development Department, PO Box 782178, Sandton 2146, South Africa K.J. Hutter Eichbaum Brauereien AG, Käfertaler Straße170, D-68169 Mannheim, Germany C.L. Jenkins School of Biological and Molecular Sciences, Oxford Brookes University, Headington, Oxford OX3 0BP, UK A.I. Kennedy Scottish Courage Brewing Ltd, Technical Centre, Sugarhouse Close, 160 Canongate, Edinburgh EH8 8DD, UK M. Kiehne BIOTECON Diagnostics GmbH, Hermannswerder Haus 17, 14473 Potsdam, Germany O. Kobayashi Kirin Brewery Co., Ltd., Central Laboratories for Key Technology, 1-13-5, Fukuura, Kanazawa-ku, Yokohama-shi, Kanagawa 236-0004, Japan C. Lange Eichbaum Brauereien AG, Käfertaler Straße170, D-68169 Mannheim, Germany A. Lentini Carlton and United Breweries Ltd/Foster’s Group Ltd, 4-6 Southampton Crescent, Abbotsford, Victoria 3067, Australia P. Malcorps Interbrew, Vaarstraat 94, B-3000 Leuven, Belgium V. Martin School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford OX3 0BP, UK

viii

CONTRIBUTORS

D.L. Maskell School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane, Headington, Oxford OX3 0BP, UK N. Moonjai Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Katholuke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium E. Pajunen Oy Sinebrychoff Ab, PO Box 87, FI-04201 Kerava, Finland A. Pardigol BIOTECON Diagnostics GmbH, Hermannswerder Haus 17, 14473 Potsdam, Germany P. Perpète Université Catholique de Louvain, Unité de Brasserie et des Industries Alimentaires, Croix du Sud 2/7, B-1348 Louvain-la-Neuve, Belgium C.D. Powell School of Biological and Molecular Sciences, Oxford Brookes University, Headington, Oxford OX3 0BP, UK I.S. Pretorius Department of Microbiology and Institute for Wine Biotechnology, University of Stellenbosch, Stellenbosch 7600, South Africa D.E. Quain Bass Brewers, Technical Centre, PO Box 12, Cross Street, Burton-upon-Trent DE14 1XH, UK B. Ranta Oy Sinebrychoff Ab, PO Box 87, FI-04201 Kerava, Finland K.E. Richardson White Labs, Inc., 7564 Trade Street, San Diego, CA 92121, USA P. Rogers Carlton and United Breweries Ltd/Foster’s Group Ltd, 4-6 Southampton Crescent, Abbotsford, Victoria 3067, Australia A.J. Schiewe White Labs, Inc., 7564 Trade Street, San Diego, CA 92121, USA P. Silcock Department of Food Science, University of Otago, PO Box 56, Dunedin, New Zealand O. Simal School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane, Oxford OX3 0BP, UK

CONTRIBUTORS

ix

K. Simic Kent Brewery, Carlton and United Breweries, Broadway, Sydney, NSW 2001, Australia K.A. Smart School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford OX3 0BP, UK P. Soininen-Tengvall Oy Sinebrychoff Ab, PO Box 87, FI-04201 Kerava, Finland R.A. Stafford The South African Breweries Ltd, Engineering Development, Corporate Technical Centre, PO Box 782178, Sandton, 2146, South Africa G. Stanley School of Food Science and Technology, Victoria University Werribee Campus, P.O. Box 14428 Melbourne City, Victoria 8001, Australia B. Taidi Scottish Courage Brewing Ltd, Technical Centre, 160 Canongate, Edinburgh EH8 8DD, UK K. Tanaka Kirin Brewery Co., Ltd., Central Laboratories for Key Technology, 1-13-5, Fukuura, Kanazawa-ku, Yokohama-shi, Kanagawa 236-0004, Japan K. Tapani Oy Sinebrychoff Ab, PO Box 87, FI-04201 Kerava, Finland A. Tauschmann BIOTECON Diagnostics GmbH, Hermannswerder Haus 17, 14473 Potsdam, Germany J.M. Thevelein Laboratory of Molecular Cell Biology, Department of Biology, KU Leuven, Kasteelpark Arenberg 31, B-3001 Leuven (Heverlee), Belgium P. Thurston Scottish Courage Brewing Ltd, Berkshire Brewery, Imperial Way, Reading RG2 0PN, UK L. Van Nedervelde Department of Brewing Sciences and Fermentation Technologies, Institut Meurice, 1 Avenue E. Gryson, B-1070 Brussels, Belgium S.M. Van Zandycke SMART Brewing Services, Oxford Brookes Enterprises, School of Biological and Molecular Sciences, Gipsy Lane, Oxford OX3 0BP, UK D. Veal Centre for Fluorimetric Applications in Biotechnology, Department of Biological Sciences, Macquarie University, North Ryde, Sydney, NSW 2109, Australia

x

CONTRIBUTORS

H. Verachtert Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Katholieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium K.J. Verstrepen Centre for Malting and Brewing Science, Faculty of Agricultural and Applied Biological Sciences, Katholieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Heverlee, Belgium S. Vincent Kent Brewery, Carlton and United Breweries, Broadway, Sydney, NSW 2001, Australia C.E. White White Labs, Inc., 7564 Trade Street, San Diego, CA 92121, USA L.R. White White Labs, Inc., 7564 Trade Street, San Diego, CA 92121, USA P.A. White School of Biological and Molecular Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford OX3 0BP, UK J. Winderickx Laboratory of Molecular Cell Biology, Department of Biology, KU Leuven, Kasteelpark Arenberg 31, B-3001 Leuven (Heverlee), Belgium

Contents

Contributors Preface to the second edition Katherine A. Smart Part 1

Molecular Innovations

1 Analysis of karyotypic polymorphisms in a bottom-fermenting yeast strain by polymerase chain reaction K. Tanaka and O. Kobayashi 1.1 Introduction 1.2 Materials and methods 1.2.1 Strains and media 1.2.2 Pulsed field gel electrophoresis and Southern hybridisation of chromosomal DNA 1.2.3 DNA manipulations and sequencing 1.2.4 Polymerase chain reaction procedures 1.3 Results and discussion 1.3.1 Chromosome length polymorphisms in a bottom-fermenting yeast strain 1.3.2 Structure of the 840 kb chromosome 1.3.3 Structure of the 820 kb chromosome 1.3.4 Translocation point in the 960 kb chromosome 1.3.5 Development of the method for detection of the 960 kb chromosome by polymerase chain reaction 1.4 Conclusions References 2

Fast detection of beer spoilage microorganisms by consensus polymerase chain reaction with foodproof® beer screening K. Berghof, M. Fandke, A. Pardigol, A. Tauschmann and M. Kiehne 2.1 Introduction 2.2 Materials and methods 2.2.1 LightCycler™ Technology 2.2.2 Design of the polymerase chain reaction 2.2.3 Analytical procedure 2.2.3.1 Microbiological enrichment 2.2.3.2 Sample preparation 2.2.3.3 Standard protocol for polymerase chain reaction preparation 2.3 Results and discussion 2.3.1 Detection of bacteria 2.3.2 Identification of bacteria

v xxv 1 3 3 4 4 4 4 4 5 5 6 6 8 10 11 11 13 13 14 14 15 16 16 16 17 18 18 19

xii

CONTENTS

2.4 Conclusions References Part 2

Brewing Yeast Stress Responses During Handling

3 The impact of ethanol stress on yeast physiology A. Lentini, P. Rogers, V. Higgins, I. Dawes, M. Chandler, G. Stanley and P. Chambers 3.1 Introduction 3.2 Materials and methods 3.2.1 Yeast storage trials 3.2.1.1 Membrane lipid composition 3.2.1.2 Trehalose content 3.2.1.3 Yeast slurry pH 3.2.1.4 Yeast protease 3.2.1.5 Yeast viability 3.2.1.6 Yeast vitality 3.2.2 Gene array technology 3.3 Results and discussion 3.3.1 Impact of ethanol and temperature on the structure of the yeast cell membrane 3.3.2 Cell-wall trehalose 3.3.3 Yeast slurry pH 3.3.4 Protease release from yeast 3.3.5 Yeast vitality 3.3.5.1 Acidification power test 3.3.5.2 Oxygen uptake rate 3.3.6 Changes in gene expression 3.3.6.1 Observations on using gene array technology 3.4 Conclusions Acknowledgements References 4 Yeast physical (shear) stress: the engineering perspective R.A. Stafford 4.1

4.2

Introduction 4.1.1 Yeast cell response to shear stress 4.1.2 Cell stimuli 4.1.3 Newton’s law of viscosity: a gross deforming force 4.1.4 Yeast rheology 4.1.5 Methods of estimating shear rate of agitated systems 4.1.6 Energy dissipation rate 4.1.7 Kolmogorov turbulence scale 4.1.8 Residence/exposure time Conclusions

20 21 23 25

25 26 26 26 26 26 26 26 26 27 27 27 28 29 30 31 31 32 32 36 36 37 37 39 39 40 40 41 41 42 43 43 43 44

CONTENTS

Acknowledgements References 5 The osmotic stress response of ale and lager brewing yeast strains P.A. White, A.I. Kennedy and K.A. Smart 5.1 Introduction 5.2 Materials and methods 5.2.1 Yeast strains 5.2.2 Media and growth conditions 5.2.3 Osmotic challenge 5.2.4 Viability determinations 5.2.5 Glycerol determination 5.2.6 Preparation of cells for confocal microscopic analysis 5.2.6.1 Staining of vacuole lumen 5.2.6.2 Staining of tonoplast 5.2.6.3 Staining of plasma membrane 5.2.6.4 Visualisation of samples 5.3 Results and discussion 5.3.1 Osmotic stress tolerance of YPD-grown cells 5.3.1.1 Physiological state 5.3.1.2 Strain dependence 5.3.1.3 Solute considerations 5.3.2 Compatible solute accumulation 5.3.2.1 Physiological state 5.3.2.2 Strain dependence and glycerol accumulation 5.3.2.3 Solute considerations of glycerol accumulation 5.3.3 Vacuolar changes 5.3.3.1 Vacuolar morphology of YPD-grown cells 5.3.3.2 Vacuolar morphology of exponential-phase cells 5.3.3.3 Vacuolar fragmentation and osmotic stress 5.4 Conclusions Acknowledgements References

xiii 44 44 46 46 48 48 48 48 48 48 49 49 49 49 49 49 49 49 51 51 53 53 53 53 56 57 57 57 58 59 59

6 Brewing yeast oxidative stress responses: impact of brewery handling V. Martin, D.E. Quain and K.A. Smart

61

6.1 Introduction 6.2 Materials and methods 6.2.1 Yeast strains and growth conditions 6.2.2 Yeast sample collection 6.2.3 Determination of response to oxidative stress 6.2.4 Glutathione concentration 6.2.5 Protein extraction for enzymic assays by glass bead cell lysis method 6.2.6 Catalase activity 6.2.7 Glycogen and trehalose concentration

61 62 62 62 62 62 62 63 63

xiv

CONTENTS

6.3

Results and discussion 6.3.1 Oxidative stress resistance is dependent on growth phase, strain and medium 6.3.2 Defence mechanisms against hydrogen peroxide are dependent on strain and medium 6.3.3 Cellular damage 6.3.4 Oxidative stress during the brewing process 6.3.5 Propagation 6.3.6 Pitching 6.3.7 Storage and acid washing 6.3.8 Serial repitching 6.4 Conclusions Acknowledgements References Part 3 Wort Composition: Impact on Yeast Metabolism and Performance 7 Wort composition and beer quality C.W. Bamforth 7.1 Introduction 7.2 The relationship of wort composition to beer quality 7.3 The key components of wort 7.4 The impact of wort on the production of flavour compounds by yeast 7.5 Models 7.6 Sources of variability in wort composition 7.7 Conclusions Acknowledgements References

63 63 63 66 67 67 68 69 70 71 71 72

75 77 77 78 78 79 81 83 84 84 84

8 Wort substitutes and yeast nutrition B. Taidi, A.I. Kennedy and J.A. Hodgson

86

8.1 Introduction 8.2 Materials and methods 8.2.1 Materials 8.2.2 Fully defined medium 8.2.3 Semi-defined medium 8.2.4 Analytical methods 8.3 Results and discussion 8.3.1 Fully defined medium 8.3.2 Semi-defined medium 8.4 Conclusions Acknowledgements References

86 87 87 87 89 89 90 90 92 95 95 95

CONTENTS

9 Wort supplements: from yeast and for yeast M. Dillemans, L. Van Nedervelde and A. Debourg 9.1 Introduction 9.2 Materials and methods 9.2.1 Yeast strains 9.2.2 Fermentations 9.2.3 Measurement of glucose uptake 9.2.4 Measurement of fructose-2,6-biphosphate 9.2.5 Acidification power test 9.2.6 Determination of enzyme activities 9.2.7 Measurement of glycerol 9.2.8 Protein determination 9.2.9 Lipid extraction 9.2.10 Glycogen determination 9.2.11 Farnesol-induced growth inhibition 9.2.12 Effect of ethanol and osmotic pressure on growth on glucose and maltose 9.2.13 Effect of ethanol and osmotic pressure on fermentation power 9.3 Results and discussion 9.3.1 Influence of yeast peptide complex on fermentation rate 9.3.2 Influence of yeast peptide complex on glucose metabolism 9.3.3 Influence of yeast peptide complex on anabolic enzyme activities 9.3.4 Influence of yeast peptide complex on yeast synthesis 9.3.5 Mode of action of yeast peptide complex 9.3.6 Influence of yeast peptide complex on ethanol and osmotic stresses of growing cells References 10 Unsaturated fatty acid supplementation of stationary-phase brewing yeast and its effects on growth and fermentation ability N. Moonjai, K.J. Verstrepen, F.R. Delvaux, G. Derdelinckx and H. Verachtert 10.1 Introduction 10.2 Materials and methods 10.2.1 Yeast strain and maintenance 10.2.2 Growth medium 10.2.3 Yeast propagation 10.2.4 Preparation of stationary-phase cells and unsaturated fatty acid supplementation 10.2.5 Analysis of pitching yeast 10.2.6 Test fermentations 10.2.7 Monitoring of fermentation 10.2.8 Analysis of volatile esters and higher alcohols

xv 96 96 97 97 97 97 98 98 98 98 99 99 99 100 100 100 100 100 101 103 105 106 107 108 110

110 111 111 111 111 111 112 112 113 113

xvi

CONTENTS

10.3

Results and discussion 10.3.1 Unsaturated fatty acid supplementation of pitching yeast 10.3.2 Fermentation with unsaturated fatty acid-supplemented yeast 10.4 Conclusions References 11

113 113 115 118 118

Impact of wort composition on flocculation B. Axcell

120

11.1 Introduction 11.2 Molecular mechanism of yeast flocculation 11.3 Premature flocculation and beer quality 11.4 The antimicrobial peptide hypothesis 11.5 Possible mechanism for premature flocculation 11.6 Conclusions References

120 121 123 124 125 126 127

Part 4 Yeast Quality Maintenance and Assessment

129

12 Management of multi-strain, multi-site yeast storage and supply A.I. Kennedy, B. Taidi, A. Aitchison and X. Green

131

12.1 12.2

Introduction 12.1.1 Historical perspective Yeast culture management 12.2.1 Aims 12.2.2 Strategies for strain maintenance 12.2.3 Selection of master cultures 12.2.4 Testing procedures 12.2.4.1 Flocculation (Tullo) and adhesion 12.2.4.2 Sedimentation (Helm’s test) 12.2.4.3 Sugar utilisation 12.2.4.4 Head formation 12.2.4.5 Petite stability 12.2.4.6 Fermentation performance 12.2.5 Deposition in liquid nitrogen 12.2.6 Cascade storage system 12.2.7 Retrieval from liquid nitrogen and slope preparation 12.2.8 Quality assurance 12.2.8.1 Freedom from contamination 12.2.8.2 Petite mutants 12.2.8.3 Viability 12.2.8.4 Genetic confirmation of identity 12.2.9 Integrity of supply 12.2.10 Statistics

131 131 132 132 132 133 133 133 133 133 133 134 134 134 134 134 135 135 135 135 135 136 136

CONTENTS

12.3 Conclusions Acknowledgements References 13 Comparison of yeast viability/vitality methods and their relationship to fermentation performance L.R. White, K.E. Richardson, A.J. Schiewe and C.E. White 13.1 Introduction 13.2 Materials and methods 13.2.1 Yeast 13.2.2 Citrate methylene blue 13.2.3 Alkaline methylene blue 13.2.4 Alkaline methylene violet 13.2.5 Acidification power 13.2.6 Standard plate count 13.2.7 Fermentation 13.3 Results and discussion 13.3.1 Citrate methylene blue 13.3.2 Alkaline stains 13.3.2.1 Alkaline methylene blue 13.3.2.2 Alkaline methylene violet 13.3.2.3 Acidification power test 13.3.2.4 Standard plate count 13.3.2.5 Yeast performance 13.4 Conclusions References 14 Yeast quality and fluorophore technologies S.M. Van Zandycke, O. Simal, S. Gualdoni and K.A. Smart 14.1 Introduction 14.2 Materials and methods 14.2.1 Yeast strains and growth conditions 14.2.2 Yeast starvation and heat treatment 14.2.3 Citrate methylene violet 14.2.4 MgANS 14.2.5 Oxonol 14.2.6 Propidium iodide 14.2.7 Sytox orange 14.2.8 Berberine 14.2.9 FUN1 14.2.10 Plate count 14.2.11 Photographs 14.3 Results and discussion

xvii 136 136 136

138 138 139 139 139 139 139 140 140 140 140 140 142 142 142 145 145 145 145 147 149

149 153 153 153 153 154 154 154 154 154 155 155 155 155

xviii

CONTENTS

14.3.1 Can fluorophores differentiate between viable and non-viable populations? 14.3.1.1 Lager strain L138 14.3.1.2 Ale strain 2593 14.3.2 Determination of yeast cell viability of starved populations 14.4 Conclusions Acknowledgements References 15 Vitality assessment using the fluorescent stain FUN1 S.M. Van Zandycke, O. Simal and K.A. Smart

155 156 157 158 160 160 160 162

15.1 Introduction 15.2 Materials and methods 15.2.1 Yeast strains and growth conditions 15.2.2 Starvation and oxidative stress 15.2.3 Acidification power test 15.2.4 Glycogen and trehalose 15.2.5 FUN1 stain for vitality assessment 15.3 Results and discussion 15.3.1 Determination of yeast cell vitality of starved stressed populations 15.3.2 Determination of yeast cell vitality of oxidatively stressed populations 15.4 Conclusions Acknowledgements References

162 164 164 164 164 164 165 165

Flow cytometry: a new tool in brewing technology K.J. Hutter and C. Lange

169

16.1 Introduction 16.2 Materials and methods 16.2.1 Glycogen content 16.2.2 DNA content 16.2.3 Detection of beer spoilage contaminants 16.2.4 Flow cytometry 16.3 Results and discussion Acknowledgement References

169 170 170 170 170 170 171 173 173

17 Comparison of the methylene blue assay with a new flow-cytometric method for determining yeast viability in a brewery A. Boyd, T. Gunasekera, P. Attfield, K. Simic, S. Vincent and D. Veal

174

16

17.1

Introduction

165 166 167 167 168

174

CONTENTS

17.2 Materials and methods 17.2.1 Trial location and yeast analysed 17.2.2 Methylene blue staining and microscopic analysis 17.2.3 Oxonol staining and flow-cytometric analysis 17.2.4 Statistical analyses 17.3 Results and discussion 17.3.1 Comparison of viability assays 17.3.2 Operator error and reproducibility of viability data 17.4 Conclusions Acknowledgements References

xix 175 175 175 175 176 176 176 177 178 179 179

Part 5 The Role of Brewing Yeast in Beer Flavour Development

181

18 Formation and disappearance of diacetyl during lager fermentation C. Boulton and W. Box

183

19

18.1 Introduction 18.2 Materials and methods 18.3 Results and discussion 18.4 Conclusions Acknowledgements References

183 184 184 193 194 194

The formation of higher alcohols J.R. Dickinson

196

19.1 Introduction 19.2 Conclusions References

196 204 205

20 Methionine: a key amino acid for flavour biosynthesis in beer P. Perpète, L. Gijs and S. Collin 20.1 Introduction 20.2 Materials and methods 20.2.1 Reagents 20.2.2 Strains 20.2.3 Culture media and sampling 20.2.4 Methanethiol quantification 20.3 Results and discussion References 21 Control of ester synthesis during brewery fermentation J.-P. Dufour, Ph. Malcorps and P. Silcock 21.1 Introduction 21.2 Ester formation and excretion during fermentation

206 206 207 207 207 208 208 208 211 213 213 215

xx

CONTENTS

21.3 The rate-limiting factors of ester synthesis and the relationship between ester synthesis, lipid metabolism and growth 21.3.1 Synthesis of the acetate esters 21.3.2 Synthesis of the medium-chain fatty acid esters (C6–C10) 21.4 Parameters influencing the synthesis of beer esters 21.5 Influence of the yeast characteristics on the synthesis of esters 21.5.1 Yeast strain 21.5.2 Pitching rate 21.5.3 Genetic and physiological instability of brewing yeast 21.6 Physicochemical and technological parameters affecting the production of esters during brewing fermentation 21.6.1 Influence of lipids on ester synthesis 21.7 Influence of oxygen/air on ester synthesis 21.7.1 Influence of the trace element: zinc 21.8 Influence of fermentation conditions 21.8.1 Stirring 21.8.2 Effect of carbon dioxide pressure 21.8.3 Fermentation in cylindroconical fermenters 21.8.4 Continuous fermentation and maturation 21.8.5 Temperature 21.9 Contribution of esterase activities to beer ester levels 21.10 Conclusions References 22 Genetic regulation of ester synthesis in yeast: new facts, insights and implications for the brewer K.J. Verstrepen, N. Moonjai, F.F. Bauer, G. Derdelinckx, J.-P. Dufour, J. Winderickx, J.M. Thevelein, I.S. Pretorius and F.R. Delvaux 22.1 Introduction 22.2 Materials and methods 22.2.1 Microbial strains, media and culturing conditions 22.2.2 DNA manipulations 22.2.3 Fermentation experiments 22.2.4 Sensory analysis 22.2.5 Headspace analysis for the measurement of acetaldehyde, ethyl acetate, n-propanol, isobutanol, isoamyl alcohol, isoamyl acetate and ethyl caproate 22.2.6 Liquid chromatography for the measurement of wort sugars 22.2.7 Carbon starvation 22.2.8 RNA extraction and Northern analysis

215 216 217 218 219 219 219 219 221 221 222 223 224 224 224 224 225 226 226 227 228

234

234 236 236 237 237 238

238 238 238 239

CONTENTS

xxi

Results and discussion 22.3.1 Activity of ATF1, ATF2 and EHT1 during brewery fermentations 22.3.2 Overexpression of ATF1 and ATF2 in brewing yeast: genetic modification allows management of ester production 22.3.3 ATF1 is regulated by glucose through the cyclic AMP/protein kinase A signalling pathway 22.4 Conclusions Acknowledgements References

239

22.3

239 240 242 245 246 246

Part 6 Yeast Handling: Objectives, Obstacles and Opportunities

249

23 Yeast Propagation G.A. Hulse

251

23.1 Introduction 23.2 Historical perspective 23.3 Current perspective 23.4 Future perspectives 23.5 Conclusions References 24 Serial repitching fermentation performance and functional biomarkers C.L. Jenkins, A.I. Kennedy, P. Thurston, J.A. Hodgson and K.A. Smart 24.1 Introduction 24.2 Materials and methods 24.2.1 Yeast strains and growth conditions 24.2.2 Citrate methylene violet 24.2.3 MgANS 24.2.4 Viability plate counts 24.2.5 Intracellular glycogen and trehalose determination 24.2.6 Determination of frequency of petite mutation 24.2.7 Propensity to form petites 24.2.8 Budding index 24.2.9 Percentage of yeast solids 24.2.10 Flocculation 24.2.11 Cell-surface charge 24.2.12 Hydrophobicity 24.2.13 Vicinal diketone uptake 24.3 Results and discussion 24.3.1 Impact of serial repitching on yeast quality 24.3.2 Impact of serial repitching on petite mutation

251 252 252 255 255 256 257

257 259 259 259 260 260 260 260 260 261 261 261 262 262 262 262 262 265

xxii

CONTENTS

24.3.3

Impact of serial repitching on the fermentation performance of lager brewing yeast 24.3.4 Impact of fermentation on the replicative capacity of lager brewing yeast 24.3.5 Impact of serial repitching on the attenuation of lager brewing yeast 24.3.6 Impact of serial repitching on the flavour development of lager brewing yeast 24.3.7 Impact of serial repitching on the flocculation capacity and cell-surface characteristics of lager brewing yeast 24.4 Conclusions Acknowledgements References 25 The impact of yeast cell age on fermentation, attenuation and flocculation C.D. Powell, D.E. Quain and K.A. Smart 25.1 Introduction 25.2 Materials and methods 25.2.1 Yeast strains 25.2.2 Preparation of aged cell fractions 25.2.3 Sucrose gradients 25.2.3.1 Preparation of virgin cells 25.2.4 Fermentations 25.2.5 Measurement of cell flocculation 25.2.5.1 Helm’s test 25.2.6 Cell-surface hydrophobicity 25.2.7 Cell-surface charge 25.3 Results and discussion 25.3.1 Age synchronisation of yeast 25.3.2 Influence of cell age on the rate of sugar utilisation during fermentation 25.3.3 Impact of age on cell flocculation 25.3.4 Relationship between age and cell hydrophobicity and cell surface charge 25.4 Conclusions Acknowledgements References 26 Chronological and replicative lifespan in lager brewing yeast D.L. Maskell, A.I. Kennedy, J.A. Hodgson and K.A. Smart 26.1 Introduction 26.2 Materials and methods 26.2.1 Yeast strains 26.2.2 Media and growth conditions

266 266 267 267 268 269 269 269 272 272 273 273 273 273 273 273 274 274 274 274 274 274 274 276 276 279 279 279 281

281 283 283 283

CONTENTS

26.2.3

Micromanipulation 26.2.3.1 Data analysis 26.2.4 Extended stationary phase 26.2.5 Production of sucrose gradients 26.2.6 Production of virgin and non-virgin populations 26.2.7 Viability assessment 26.2.7.1 Citrate methylene violet 26.2.7.2 Oxonol 26.2.7.3 Plate counts 26.3 Results and discussion 26.3.1 Replicative lifespan of four strains of lager brewing yeast 26.3.2 Chronological lifespan of four strains of lager brewing yeast 26.3.3 Is there a correlation between replicative and chronological lifespan? 26.3.4 Do chronologically aged brewing yeast cells demonstrate a reduced replicative lifespan? 26.4 Conclusions Acknowledgements References

xxiii 283 284 284 284 284 284 284 285 285 285 285 286 287 288 289 290 290

27 Continuous primary fermentation of beer with immobilised yeast K. Tapani, P. Soininen-Tengvall, H. Berg, B. Ranta and E. Pajunen

293

27.1 Introduction 27.2 Materials and methods 27.2.1 Yeast and wort 27.2.2 Carrier 27.2.3 Pilot plant unit 27.2.4 Start-up procedures 27.2.5 Basis for continuous fermentation 27.2.6 Process conditions 27.2.7 Analytical methods 27.2.7.1 Fermentation analyses 27.2.7.2 Flavour compounds and vicinal diketones 27.2.7.3 Fermentable sugars 27.2.7.4 Microbiological analysis 27.3 Results and discussion 27.3.1 Fermentation 27.3.2 Flavour formation 27.3.3 Vicinal diketones 27.3.4 Free amino nitrogen 27.4 Conclusions Acknowledgements References

293 294 294 294 294 294 295 295 295 295 296 296 296 296 296 296 298 299 299 300 300

Index

303

Preface to the second edition Controlling the impact of stress on brewing biomass, predicting yeast activity and ensuring consistent fermentation performance through successive fermentations remain areas of active interest for the brewing industry. To be able to control and perhaps even manipulate yeast activity, it is necessary to identify factors that affect its functionality during fermentation. Genetic stability and integrity are crucial to maintaining predictable performance. The brewing yeast genome is inherently unstable, leading to the formation of nuclear and mitochondrial variants during yeast handling and fermentation. Although recent molecular innovations may allow rapid detection of such occurrences, the causes and nature of the DNA damage remain to be elucidated. During handling and fermentation the yeast is subjected to a rapidly changing environment. There are many stresses to be considered, including physical stresses such as shear, cold shock and hydrostatic pressure, and those created by the yeast’s own biochemical activity such as oxidative stress, nutrient limitation, anaerobiosis, osmotic stress, low pH, excess carbon dioxide and the formation of toxic metabolites. In addition, wort composition is a critical determinant of yeast performance and final product quality. Batch-to-batch changes in component ratios inevitably contribute to the variability in performance exhibited by a given slurry, yet few extensive studies have been conducted in this area. This very variability in both wort composition and yeast quality is reflected in final beer quality and in particular beer flavour. The role of the yeast cell in flavour attributes is therefore dependent on both intrinsic and extrinsic factors. It is not unreasonable to suggest that the physiological condition of brewing yeast influences fermentation performance, therefore brewers require consistent yeast quality and quantity. Ensuring the correct quality can be achieved by adequate strain selection and maintenance though master culture storage regimes and effective propagation and yeast handling during serial repitching. Preventing slurry deterioration through the use of immobilisation may prove successful but there is still a requirement to identify adequate biomarkers for slurry deterioration and potential to perform. Katherine A. Smart Royal Society Industrial Fellow Scottish Courage Reader in Brewing Science

Part 1

Molecular Innovations

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

1 Analysis of Karyotypic Polymorphisms in a Bottom-fermenting Yeast Strain by Polymerase Chain Reaction K. TANAKA and O. KOBAYASHI

Abstract Chromosomal rearrangement causes karyotypic variation in bottom-fermenting yeast. However, the molecular basis of this phenomenon has not yet been clearly defined. The complete genome sequence of Saccharomyces cerevisiae, which has been published, can be used for genome analysis of bottom-fermenting yeast. The chromosomal organisation of a bottom-fermenting yeast strain is being investigated by pulsed field gel electrophoresis and Southern hybridisation using more than 100 genes from all 16 chromosomes of S. cerevisiae as probes. In this study, the same techniques were used to detect the karyotypic polymorphisms of single colonies isolated from a bottom-fermenting yeast strain. Although the karyotypes of the isolated clones were almost the same, chromosome length polymorphisms were observed in three chromosomes. These chromosomes were investigated in detail and found to be chimeras, constructed from two different chromosomes. In the junction of the chimeric chromosomes, either a retrotransposon Ty or the subtelomeric gene COS was found to exist. This suggested that translocation resulting from homologous recombination produced these chimeric chromosomes. Making use of the sequences of the junction regions, a new method to detect karyotypic changes by polymerase chain reaction was developed. This new method is highly sensitive, and able to detect karyotypic changes within 2 days, from a single colony. This method led to the observation that translocation occurred at a frequency of 105 during yeast cultivation.

1.1 Introduction Genetic changes of bottom-fermenting yeast have been reported.1,2 Such changes may give rise to instabilities and, therefore, affect the performance of the bottomfermenting yeast during fermentation. To control the quality of yeast for fermentation, it is important to know the environmental factors that affect the occurrence of such changes. However, little is known of the mechanism by which genetic changes in bottomfermenting yeasts occur. To investigate the mechanism of karyotypic changes, highly sensitive methods to detect genetic changes are required. Two types of method for the detection of chromosomal rearrangement have been developed. One type uses selectable marker genes on artificial loci.3,4 Although such methods give a rapid and a highly sensitive analysis, naturally occurring chromosomal rearrangements cannot be detected. The other type of method detects chromosome length polymorphisms using pulsed field gel electrophoresis (PFGE).2 However, this latter method requires as long as 8 days to obtain results from the start of culture. Moreover, bottom-fermenting yeasts have many more chromosomes than laboratory yeasts, preventing adequate separation of each chromosome. Bottom-fermenting yeasts are known to have an unusual genomic background.5 Not only are they polyploid strains, but they have at least two different genomic sets.

4

BREWING YEAST FERMENTATION PERFORMANCE

One genomic set is structurally similar to that of Saccharomyces cerevisiae strains, while another is structurally similar to that of S. bayanus. It has been suggested that this complicity makes it difficult to separate each chromosome sufficiently by PFGE. Some chromosomes separated by PFGE have been identified using the Southern hybridisation method.5,6 This method can detect karyotypic changes that cannot be distinguished by PFGE. However, the sensitivity of this method is insufficient. This objective of this study was the development of a novel and highly sensitive method to detect genetic changes in bottom-fermenting yeasts. Chromosome rearrangements were found in a bottom-fermenting yeast that could be detected using PFGE and Southern hybridisation. Chromosomal structure was investigated in detail to identify specific regions for detection by polymerase chain reaction (PCR).

1.2 Materials and methods 1.2.1

Strains and media

KBY011 is a bottom-fermenting strain derived from the authors’ stock culture collection. KY1165, KY1166, and KY1167 are single-colony isolates from KBY011. KT303 and KT334 were derivatives of KY1166 and KY1167, respectively, in which the FLO11 gene was disrupted by insertion of YIp5.7 Cells were grown at 30°C in YPDA (1% yeast extract, 2% peptone, 2% dextrose, 0.02% adenine sulfate) broth or on YPDA plates containing 2% agar. 1.2.2

Pulsed field gel electrophoresis and Southern hybridisation of chromosomal DNA

Preparation of chromosomal DNA, PFGE and Southern hybridisation were carried out by the methods described previously.5 Genes on appropriate loci were selected from the Saccharomyces Genome Database (http://genome-www.stanford.edu/Saccharomyces/) and used as probes. 1.2.3

DNA manipulations and sequencing

Plasmid DNAs were purified using QIAGEN-tip 100 (Qiagen, USA). DNA digestion with restriction enzymes, standard agarose gel electrophoresis and recovery of DNA from agarose gel were carried out according to the methods described by Sambrook et al.8 Small-scale chromosomal DNA extraction from yeast cells was carried out using Dr. GenTLE™ for Yeast and Gram Positive Bacteria Genome (Takara Shuzo, Japan). DNA sequencing was performed by the method described by Sanger et al. using a DNA sequencing System (Applied Biosystems, USA).9 The results of sequencing were analysed using the DNASIS program (Hitachi Software Engineering, Japan). 1.2.4

Polymerase chain reaction procedures

Ex Taq DNA polymerase (Takara Shuzo, Japan) was used for PCR according to the manufacturer’s instructions. Standard PCR was performed with GeneAmp PCR system

ANALYSIS OF KARYOTYPIC POLYMORPHISMS

5

9600 (Applied Biosystems, USA) using a programme consisting of one cycle of 5 min at 94°C followed by 30 cycles of 20 s at 98°C and 5 min at 68°C. Semi-quantitative PCR was carried out as follows: genomic DNAs were isolated from mixture of KY303 and KY334 cells and used for PCR amplification of the target region. YIp5 region inserted in the FLO11 locus in KY303 and KY334 were also amplified and used as an internal standard. PCR products were analysed on agarose gel. The intensity of ethidium bromide stained bands were quantified using FluorImager 595 (Amersham Pharmacia Biotech, UK) and standardised using the intensity of the internal standard. Amounts of the PCR products were calculated using the intensity of the molecular weight marker.

1.3 1.3.1

Results and discussion Chromosome length polymorphisms in a bottom-fermenting yeast strain

Bottom-fermenting yeast is thought to be a natural hybrid between S. cerevisiae and Saccharomyces bayanus. Some chromosomes of S. cerevisiae have been suggested to be rearranged in S. bayanus by reciprocal translocation.10–12 Such S. bayanus-type chromosomes have been also found in bottom-fermenting yeast strains.5,6 However, no reports have appeared concerning the chromosomal organisation of bottom-fermenting yeasts. The chromosomal organisation of a stock of bottom-fermenting yeast strains was investigated in detail using PFGE and Southern hybridisation (K. Tanaka, in preparation). During the investigation, translocations were found in a bottom-fermenting yeast strain which were not found in S. bayanus strains. It is possible that such translocations had occurred after the hybrid cross between S. cerevisiae and S. bayanus. The question has been raised as to whether or not such translocated chromosomes are stable. Single colonies were isolated from a bottom-fermenting yeast, KBY011, and chromosomal organisation of these colony isolates was investigated. Figure 1.1 shows the loci of genes used as probes. Although the chromosomal organisation of these isolates was almost the same, three chromosomes were lost in some of the isolates (Fig. 1.2). The 840 kb chromosome was detected using ERG6 (chromosome XIII) as a probe. The IV VII XII XV XVI XIII II X XIV

VIII IX III VI

XI I

V

Fig. 1.1 Genes used to detect the karyotypic changes in a bottom-fermenting yeast. Arrows indicate the loci in the Saccharomyces cerevisiae chromosome map.

6

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 1.2 (a) Electrophoretic karyotypes and (b–d) results of Southern hybridisation of single colony isolates from a KBY011 with (b) ERG6, (c) DMC1, and (d) SUC2 genes as probes. Lane 1: KY1165; lane 2: KY1166; lane 3: KY1167.

820 kb chromosome was detected using DMC1 (chromosome V). The 960 kb chromosome was detected by SUC2 (chromosome IX). Further research was planned to clarify the structure of these chromosomes. 1.3.2

Structure of the 840 kb chromosome

The 840 kb chromosome may also be detected by some of the genes located on chromosome XI. Therefore, PFGE and Southern analysis was performed using the genes located on chromosome XI or XIII as probes. The results are shown in Fig. 1.3. Concerning the probes from chromosome XI, the URA1 gene hybridised with the 840 kb chromosome but the genes located on the region from ORD1 to SIR1 did not hybridise with it. Concerning the probes from chromosome XIII, the genes located on the region from PHO84 to GUA1 hybridised with the 840 kb chromosome but the ADE4 gene did not. These results suggested that the 840 kb chromosome originated from a translocation of the left end of chromosome XI onto the right end of chromosome XIII. 1.3.3

Structure of the 820 kb chromosome

Figure 1.4 shows the results of PFGE and Southern analysis using the genes located on chromosome V as probes. All of the genes used hybridised with the 610 kb chromosome,

ANALYSIS OF KARYOTYPIC POLYMORPHISMS

7

Fig. 1.3 Chromosomal assignment of a single-colony isolate KY1166 with genes located on Saccharomyces cerevisiae chromosome XIII (a) or chromosome XI (b) as probes. Lanes 1–10: hybridisation patterns with the gene probes PHO84, ERG6, ILV2, GUA1, ADE4, SIR1, PAP1, TOA2, ORD1 and URA1, respectively.

Fig. 1.4 Chromosomal assignment of a single-colony isolate KY1166 with genes located on Saccharomyces cerevisiae chromosome V as probes. Lanes 1–5: hybridisation patterns with the gene probes NPR2, URA3, ILV1, GDI1 and DMC1, respectively.

corresponding to chromosome V. DMC1 also hybridised with the 680 kb chromosome and the 820 kb chromosome, suggesting that these chromosomes were originated from translocation of the right end of chromosome V onto other chromosomes. Ty is inserted in a S. cerevisiae genome in multiple copies, which constitute 3.1% of the genome.13 Ty-mediated chromosomal translocations are believed to cause karyotypic changes in laboratory strains14–16 and wine strains.17 In the region from GDI1 to DMC1, two full-length Ty elements were found from the Saccharomyces Genome Database. PFGE and Southern analysis using the genes flanking these Ty elements as probes were carried out. The genes located on the region from YER139 to SPT2 hybridised with both the 680 kb and the 820 kb chromosomes, but SCR1 did not hybridise with these chromosomes (Fig. 1.5). It was suggested that the translocation points of the 680 kb and the 820 kb chromosome lie between SCR1 and YER139C in the right end of chromosome V, where Ty1 lies. Ty elements or LTRs are able to promote chromosomal

8

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 1.5 Chromosomal assignment of single-colony isolates KY1165 (lane 1), KY1166 (lane 2) and KY1167 (lane 3) with genes located on Saccharomyces cerevisiae chromosome V as probes. (a–d): Hybridisation patterns with the gene probes SCR1, YER139, BUR6 and SPT2, respectively.

translocation by ectopic recombination in S. cerevisiae in the presence of strong selective pressure.14,15 One recent study revealed that Ty-driven genome rearrangement may be common in industrial wine strains of S. cerevisiae.17 The present data suggest that Ty-mediated chromosomal translocations lead to karyotypic changes in bottomfermenting yeasts. The chromosomes onto which translocation of the right end of chromosome V occurred have not been identified. Both of the translocation points of the 680 kb and the 820 kb chromosomes lie between SCR1 and YER139, suggesting that this is the hotspot for translocation. Although colonies that do not have the 680 kb chromosome have not yet been found, it is possible that such cells exist in cultures of KBY011. 1.3.4

Translocation point in the 960 kb chromosome

Except for SUC2, which is an unusual member of the family, all of the SUC genes are located very close to telomeres.18 Telomeric regions of Saccharomyces are highly divergent, suggesting that these regions are highly unstable.19 Therefore, it is possible that the translocation point lies at the region near the SUC gene. Genomic DNA from KY1166 and KY1167 was digested by many kinds of restriction enzyme to carry out Southern analysis using the SUC2 gene as a probe. When the restriction enzymes Asp718 and BamHI were used, a polymorphism was found between KY1166 and KY1167. As shown in Fig. 1.6, an additional 7.5 kb band was found only in KY1167 that has the 960 kb band, suggesting that this fragment came from the 960 kb chromosome.

ANALYSIS OF KARYOTYPIC POLYMORPHISMS

9

Fig. 1.6 Southern blot analysis of genomic DNA from single-colony isolates KY1166 (lane 1) and KY1167 (lane 2) with the SUC gene as a probe. Genomic DNAs were digested with Asp718 and BamHI. Arrow: polymorphic 7.5 kb band.

Fig. 1.7 Polymerase chain reaction (PCR) amplification of the translocation point in the 960 kb chromosome. (a) Diagram of the structure of the translocation point in the 960 kb chromosome. Hatched boxes indicate open reading frames. Arrow: region amplified by PCR. (b) Discrimination of the single colony isolates with karyotypic polymorphism by PCR. Lane 1: molecular weight marker; lane 2: KY1166; lane 3: KY1167.

This fragment was cloned and its DNA sequence determined. The fragment was composed of the SUC gene, COS3 and YML131, suggesting that the 960 kb chromosome had originated from a translocation of the telomeric region containing the SUC gene onto the left end of chromosome XIII (Fig. 1.7a). Since COS is known to lie within the telomeric region in multiple copies, this gene or its flanking region may be the translocation point of the 960 kb chromosome.

10 1.3.5

BREWING YEAST FERMENTATION PERFORMANCE

Development of the method for detection of the 960 kb chromosome by polymerase chain reaction

PCR was used to develop a highly sensitive method for detection of the 960 kb chromosome. The primers were designed to amplify the region from the SUC gene to YML131. The result is shown in Fig. 1.7b. The 4.5 kb fragment was amplified only from KY1167 that has the 960 kb chromosome, suggesting that the specific region for this chromosome was amplified. This result also indicates that the 4.5 kb region composed of SUC, COS3 and YML131 exists only on the 960 kb chromosome. Therefore, it is suggested that the translocation at this point is involved in the occurrence of the 960 kb chromosome. To determine the sensitivity of this method, semi-quantitative PCR was carried out using a mixture of KT303 and KT334. The result is shown in Fig. 1.8. The translocation point could be detected from the cells containing KT334 in the ratio of 107:1. However, the intensity of the signal of the cells containing KT334 in the ratios of 105:1, 106:1 and 107:1 was almost the same. Moreover, a significant level of the signal for the translocation point was detected even from KT303 without KT334 contamination. These results suggest that the translocation of the region containing the SUC gene onto chromosome XIII occurs at a frequency of approximately 105 during cultivation for cell propagation. Hiraoka et al.4 reported that chromosome aberration caused by unequal crossing over or translocation during mitosis occurred at a frequency of 6.3  106. The present results are consistent with their observation. Detection of karyotypic changes requires 11 days using PFGE and Southern analysis. This novel method yields rapid results within 2 days. Moreover, this method is much more sensitive than the traditional one. Genetic instability in flocculation of bottom-fermenting yeasts has been investigated. The gene responsible for flocculation in bottom-fermenting yeast is suggested

Fig. 1.8 Result of the semi-quantitative polymerase chain reaction (PCR) for detection of the translocation point. The relative amount of PCR product was determined as described in Materials and Methods.

ANALYSIS OF KARYOTYPIC POLYMORPHISMS

11

to be Lg-FLO1.20 Methods using PCR to detect mutation of the Lg-FLO1 gene or its homologue, which causes loss of flocculation ability, have been developed.21,22 The present method can detect chromosomal rearrangement very close to the SUC gene, which may affect the brewing performance. It is suggested that a combination of these methods provides a useful tool in the control of yeast quality.

1.4

Conclusions

This study investigated the structure of the chromosomes that revealed polymorphisms in a bottom-fermenting yeast strain. These chromosomes were found to be caused by translocations and the translocation points were investigated in detail. Based on the results, a PCR method was developed to detect translocation of the chromosomal end containing the SUC gene onto chromosome XIII. This method enabled the detection of chromosomal rearrangements within 2 days. The translocation on this locus possibly occurs at a frequency of 105 during cultivation for cell propagation.

References 1. Pedersen, M.B. (1993) Instability of the brewers yeast genome. Proc. Congr. Eur. Brew. Conv. 24, 291–298. 2. Casey, G.P. (1996) Practical applications of pulsed field electrophoresis and yeast chromosome fingerprinting brewing QA and R&D. Tech. Q. Master Brew. Assoc. Am. 33, 1–10. 3. Fassulo, M.T. and Davis, R.W. (1988) Direction of chromosome rearrangements in Saccharomyces cerevisiae by using of his3 recombinational substrates. Mol. Cell. Biol. 8, 4370–4380. 4. Hiraoka, M., Watanabe, K., Umezu, K. and Maki, H. (2000) Spontaneous loss of heterozygosity in diploid Saccharomyces cerevisiae cells. Genetics 156, 1531–1548. 5. Tamai, Y., Momma, T., Yoshimoto, H. and Kaneko, Y. (1998) Co-existence of two types of chromosome in the bottom fermenting yeast, Saccharomyces pastorianus. Yeast 14, 923–933. 6. Yamagishi, H. and Ogata, T. (1999) Chromosomal structure of bottom fermenting yeasts. System. Appl. Microbiol. 22, 341–353. 7. Struhl, K., Stinchcomb, D.T., Scherer, S. and Davis, R.W. (1978) High-frequency transformation of yeast: autonomous replication of hybrid DNA molecules. Proc. Natl Acad. Sci. U.S.A. 76, 1035–1039. 8. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 9. Sanger, F., Nicklen, S. and Coulson, S. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl Acad. Sci. U.S.A. 74, 5463–5467. 10. Ryu, S.-L., Murooka, Y. and Kaneko, Y. (1996) Genomic reorganization between two sibling yeast species, Saccharomyces bayanus and Saccharomyces cerevisiae. Yeast 12, 757–764. 11. Ryu, S.-L., Murooka, Y. and Kaneko, Y. (1998) Reciprocal translocation at duplicated RPL2 loci might cause speciation of Saccharomyces bayanus and Saccharomyces cerevisiae. Curr. Genet. 33, 345–351. 12. Fischer, G., James, S.A., Roberts, I.N. et al. (2000) Chromosomal evolution in Saccharomyces. Nature 405, 451–454. 13. Kim, J.M.,Vanguri, S., Boeke, J.D. et al. (1998) Transposable elements and genome organisation: a comprehensive survey of retrotransposons revealed by the complete Saccharomyces cerevisiae genome sequence. Genome Res. 8, 464–478. 14. Roeder, G.S. and Fink, G.R. (1980) DNA rearrangement associated with a transposable element in yeast. Cell 21, 239–249. 15. Kupiec, M. and Petes, T.D. (1988) Allelic and ectopic recombination between Ty elements in yeast. Genetics 119, 549–559. 16. Casaregola, S., Nguyen, H.V. Lepingle, A. et al. (1998) A family of laboratory strains of Saccharomyces cerevisiae carry rearrangements involving chromosome I and III. Yeast 14, 551–564. 17. Rachidi, N., Barre, P. and Blondin, B. (1999) Multiple Ty-mediated chromosomal translocations lead to karyotype change in a wine strain of Saccharomyces cerevisiae. Mol. Gen. Genet. 261, 841–850.

12

BREWING YEAST FERMENTATION PERFORMANCE

18. Carlson, M., Celenza, J.L. and Eng, F.J. (1985) Evolution of the dispersed SUC gene family of Saccharomyces cerevisiae by rearrangements of chromosome telomeres. Mol. Cell. Biol. 5, 2894–2902. 19. Horowitz, H., Thorburn, P. and Haber, J.E. (1984) Rearrangement of highly polymorphic regions near telomeres of Saccharomyces cerevisiae. Mol. Cell. Biol. 4, 2509–2517. 20. Kobayashi, O., Hayashi, N., Kuroki, R. and Sone, H. (1998) Region of Flo1 proteins responsible for sugar recognition. J. Bacteriol. 180, 6503–6510. 21. Jibiki, M., Ishibashi, T., Yuuki, T. and Kagami, N. (2001) Application of polymerase chain reaction to determine the flocculation properties of brewer’s lager yeast. J. Am. Soc. Brew. Chem. 59, 107–110. 22. Sato, M., Watari, J. and Shinotsuka, K. (2001) Genetic instability in flocculation of bottom-fermenting yeast. J. Am. Soc. Brew. Chem. 59, 130–134.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

2

Fast Detection of Beer Spoilage Microorganisms by Consensus Polymerase Chain Reaction with foodproof® Beer Screening K. BERGHOF, M. FANDKE, A. PARDIGOL, A. TAUSCHMANN and M. KIEHNE

Abstract The use of polymerase chain reaction (PCR) in the brewing industry had been limited for several years to the research laboratory but never found its way into the routine laboratory of the quality assurance department. This was due to the demand for welltrained people to carry out this analysis and to the laborious procedure that was necessary to minimise the high risk of contamination and thus false-positive results, etc. With the new LightCycler™ format of PCR and the ready-to-use kits of BIOTECON Diagnostics it is possible to profit from this very specific and sensitive method in a standard routine laboratory of a brewery. The test, foodproof® Beer Screening, detects 14 different beer spoilage bacteria in one single reaction. After a short enrichment of 24–48 h (recommended for routine use) of the sample (with or without yeast), preparation is very simple and does not take longer than 25 min for 30 samples. The running time of the LightCycler™ is approximately 1 h and offers real-time results that clearly indicate the presence or absence of one or more of the 14 bacteria included in the test. In a second step that does not require any additional hands-on time, the bacteria can be identified in most cases on a species level. This is done by a melting curve analysis that exploits the different behaviour of probes when melted from the DNA. With the newly developed test the detection and identification of beer spoilage organisms is possible within 48 h without laborious biochemical or molecular biological efforts. This allows PCR to be used for the first time in a common laboratory of the brewing industry.

2.1 Introduction The polymerase chain reaction (PCR) is a well-established tool for the fast and specific detection of microorganisms. By this procedure a known and organismspecific piece of DNA is amplified in vitro and then detected in a second step. The selection of the DNA sequence allows very exact differentiation of the organisms. The high speed of this analysis compared with conventional microbiology is due to the amplification of the DNA, which is doubled in minutes, and thus much more rapidly than the growth of complete cells, which normally takes hours, especially in the case of slowly growing beer spoilage bacteria. The PCR is also able to detect low concentrations of spoilage or pathogenic organisms. In the past, the method was very laborious and could only be performed by highly trained people in specially equipped laboratories. By using new technologies many complicated and problematic steps have been eliminated, including the complicated preparation of the target DNA and the detection of the PCR products with gels or by enzyme-linked immunosorbent assay (ELISA).

14

BREWING YEAST FERMENTATION PERFORMANCE

2.2 Materials and methods This chapter describes the technology platform, the design of the PCR reagents and the use of the method in a routine laboratory. Since a commercial product is discussed some details have been omitted. 2.2.1

LightCycler™ Technology

The LightCycler™ of Roche Diagnostics, which is used as the basis for the new kit of BIOTECON Diagnostics, offers a number of advantages over conventional PCR. The time for the reaction is reduced by the new design of the reaction tubes (capillaries with a reaction volume of 20 l) and the tempering unit (heating coil). The capillary offers a very high surface-to-volume ratio, which enables the fast heat changes in the reaction tube that are necessary to carry out the PCR cycles. The combination of the capillaries and air for rapid cycling allows a single PCR cycle to be carried out in less than 1 min. A further increase in speed is achieved by using real-time detection of the resulting reaction products (amplificates). In the reaction mix there are two different types of probe. One is labelled with fluorescin (the donor dye) at the 3-end and the other probe is labelled with LightCycler-Red 640 (the acceptor dye). In the denaturation step the probes are floating in the reaction mix. Only the first probe is excited by the lightemitting diode (LED) from the LightCycler™, but this signal is not detected. In the annealing step both probes find their corresponding regions of the amplified DNA and hybridise to the amplified DNA. The probes are designed to anneal in close proximity. The first probe (donor dye) is excited by the LED and transfers the energy to the second probe (acceptor dye), which emits light at a wavelength of 640 nm. This process is called fluorescence resonance energy transfer (FRET). This signal is detected. The use of two hybridisation probes increase the specificity of the method (Fig. 2.1).

Fig. 2.1

Principle of the probe detection.

DETECTION OF BEER SPOILAGE MICROORGANISMS

15

Another unique tool is the melting curve analysis offered by the LightCycler™. After the PCR is completed the temperature in the capillaries is adjusted to a level where the probes find the optimal annealing conditions and thus the highest signal is achieved. By slowly increasing the temperature stepwise and measuring the signal after each step, the melting behaviour of the probes is detected. With increasing temperature the tension on the hydrogen bonds increases, and at a certain temperature the probes are melted from the DNA. The stronger the bonds between DNA and the probes the higher the melting temperature. The strength of the bonds depends on the length of the probe, the G-C content, the base sequence, etc. Thus, the melting curve analysis allows for differentiation between different DNA–probe combinations by simply measuring their heat sensitivity. The results can be used, for example, to detect by-products of the PCR or to distinguish between different amplificates that were detected by the same probe. 2.2.2

Design of the polymerase chain reaction

PCR in general is used to detect single microorganisms very specifically using unique DNA sequences. BIOTECON Diagnostics developed a PCR test that is capable of detecting several spoilage organisms which are relevant to the industry.1 Given the constraints of standard PCR, a novel approach was required to detect a large spectrum of bacteria in one single step while retaining the traditional advantages of specificity, sensitivity and velocity provided by PCR. BIOTECON Diagnostics included 14 different bacterial species in the test. The bacteria were selected after discussing their importance with several breweries around the world. The PCR is not a specific PCR but a mixture of a consensus and multiplex PCR. It consists of a mix of several primers that amplify the 14 bacteria listed in Table 2.1. The chemicals necessary for the PCR are provided a kit format. The kits include chemicals for the sample preparation (DNA extraction) and for the amplification and detection of the DNA.2 An important aspect of the routine analysis is to control the performance of the method. A negative result, for example, may be due to the absence of the target (true negative) or to a failure in the analysis (false negative). Thus, the reaction mix contains an internal positive control (IPC) which comprises a piece of target DNA that is also amplified with primers to control the performance of the PCR when no DNA of the 14 bacteria (Table 2.1) is present in the sample. The purpose is to prevent

Table 2.1

Bacteria detected by the polymerase chain reaction

Lactobacillus brevis Lactobacillus lindneri Lactobacillus casei Lactobacillus paracasei Lactobacillus coryniformis ssp. coryniformis Lactobacillus coryniformis ssp. torquens Lactobacillus parabuchneri ( frigidus)

Pectinatus cerevisiiphilus Pectinatus spec. DSM 20764 Pectinatus frisingensis Pediococcus damnosus Pediococcus inopinatus Megasphaera cerevisiae Selenomonas lacticifex

16

BREWING YEAST FERMENTATION PERFORMANCE

false-negative results due to inhibition or errors during the preparation of the PCR. The IPC is detected with a different pair of probes, where the acceptor probe is labelled with LightCycler-Red 705, a dye emitting light at 705 nm. Thus, the signal can be detected in a different channel of the LightCycler™. If a result is negative in the wildtype channel F2 (sample DNA) the control channel F3 has to be positive. A further aspect is the prevention of cross-contamination by the amplificates of earlier PCR runs. The PCR is carried out using uracil instead of thymine. Thus, the DNA amplificates produced are artificial and do not occur in nature. In addition, their chemical composition is different from the DNA of the bacteria under investigation. The kits also contain the enzyme uracil-N-glycosylase (UNG), which is used to destroy all old amplificates containing uracil before running a new PCR. The target DNA of the bacteria is not affected. After an incubation period the enzyme UNG itself is destroyed by heat and PCR with uracil can be carried out. This tool prevents falsepositive results caused by cross-contamination of old amplificates. It does not prevent false-positive results from cross-contamination with bacterial DNA. 2.2.3

Analytical procedure

2.2.3.1 Microbiological enrichment. BIOTECON Diagnostics recommends filtration of the sample (e.g. 100–500 ml) and 50 mm cellulose-nitrate filters with 0.2 m pores (Sartorius) are suitable. The drying of the filter during filtration should be avoided. The filter is quickly transferred to a small volume of enrichment broth (such as 10 ml tubes of NBB broth; Döhler, Darmstadt, Germany, Art. No. 4723). The use of alternative media is also possible. In this case it must be ensured that there is no interference with the amplification and/or detection via the LightCycler™. The enrichment should be carried out as a standing culture at 28–30°C under anaerobic conditions for 24–48 h (depending on the species). To detect very slow-growing bacteria such as Lactobacillus lindneri or Pediococcus damnosus it may be necessary to increase the time of enrichment until the density of 1–5  103 colony-forming units (cfu)/ml is achieved. From experience it is known that a smaller number of organisms (100 cfu/ml) can be detected with the system (data not shown), but for routine use it is useful to enrich until an amount of cells is reached which cannot be ‘lost’ during sample preparation (pipetting, diluting, etc.). 2.2.3.2 Sample preparation. The disintegration of the bacterial cells and the extraction of the DNA are essential prerequisites. It is recommended that the samples are not prepared in the microbiological laboratory because of the high amount of bacteria in the environment. The sample preparation should be conducted using the following protocol. 1. The enrichment culture is mixed and transferred (1 ml) to a 1.5 ml reaction tube. 2. The sample is then centrifuged at maximum speed (15 000 g) for 5 min at room temperature in a standard benchtop microcentrifuge. If the media for the enrichment are totally clear, the use of latex beads (Sigma; 10 l of 1:10 in distilled water diluted suspension) is recommended before the centrifugation step to improve sedimentation.

DETECTION OF BEER SPOILAGE MICROORGANISMS

17

3. Immediately after centrifugation the supernatant should be carefully removed by pipetting (not decanting) and discarding. 4. Lysis buffer is added to resuspend the pellet in 50 l and mixed with an orbital mixer (2  5 s). Tapping the reaction tube on the bench will allow particles to be removed from its wall. 5. The suspension is incubated in a unit heater or water bath for 10 min at 95– 100°C. The tubes have to be tightly closed (e.g. using lid clips) to prevent crosscontamination. 6. The lysate is then mixed for 10 s in an orbital mixer, at maximum speed. 7. The lysate is then centrifuged at maximum speed (15 000 g) for 30–60 s at room temperature in a standard benchtop microcentrifuge. 8. The lysate is stored at 4°C or on ice if the PCR is started immediately afterwards. Otherwise, the prepared samples should be stored at 20°C. If the sample contains a high amount of yeast cells (e.g. pitching yeast) an additional centrifugation step should be included in the protocol. 9. The enriched culture is mixed. If the sample has a very low fluid content it should be diluted with one volume of NBB medium or another PCR-approved medium. 10. At this stage 1 ml is removed and transferred into a 1.5 ml Eppendorf cup. 11. The sample is then centrifuged at 100 g (c. 1000 rpm in a standard benchtop centrifuge) for 5 min at room temperature. 12. The supernatant is then transferred to a new Eppendorf cup. 13. Proceed with step 2 of the above protocol. 2.2.3.3 Standard protocol for polymerase chain reaction preparation. The following steps describe the preparation of the PCR. 1. All reagents are thawed, mixed gently (therefore not by vortexing) and centrifuged. 2. The PCR-Master-Mix (enzymes, nucleotides, etc.) is then transferred into a new sterile reaction tube. The ICP-Mix (internal control and primers) is then mixed with the PCR-Master-Mix. 3. 17.5 l of the Mix is transferred into all prepared capillaries. 4. 2.5 l of DNA sample lysate is pipetted into the capillaries prepared for sample reactions and the capillaries are then sealed. 5. 2.5 l Negative Control (PCR water) is pipetted into the appropriate capillary and sealed. 6. 2.5 l Positive Control (DNA of Selenomonas lacticifex) is pipetted into the appropriate capillary and sealed. 7. The capillaries are placed into adapters and centrifuged at 700 g for 5 s to remove air from the bottom of the capillaries. 8. The capillaries are then placed in the rotor of the LightCycler™ instrument. 9. The PCR is run. The LightCycler™ needs approximately 60 min for the PCR and another 20 min for the melting curve analysis. The results can be monitored in real-time on the screen of the device. Figure 2.2 shows a typical temperature profile.

18

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 2.2

2.3 2.3.1

Temperature profile of the foodproof® Beer Screening polymerase chain reaction.

Results and discussion Detection of bacteria

The first step in interpreting the data is the signal in the sample channel F2 of the LightCycler™. If the signal increases, as indicated in Fig. 2.3, one or more of the 14 bacteria in Table 2.1 is present. The signal increases exponentially, indicating the growing number of amplificates in the reaction mix after each cycle. The crossing point, that is the cycle number at which the signal rises above the background, represents the value for the quantification of the bacteria in the sample. The lower the crossing point the higher the amount of target DNA present in the sample. In the case of an enrichment this value can only be used as a rough indicator because the speed of growth during the enrichment would not be known. A positive result does not yield any information about the nature of contamination present, but it clearly indicates the presence of an obligatory beer spoilage bacterium and not only the presence of a lactic acid bacterium. However, it should be noted that only the bacteria in Table 2.1 are detected. 2.3.2

Identification of bacteria

The data of the melting curve may be used to differentiate the bacteria. The melting curve analysis is carried out by the LightCycler™ after the PCR. The melting behaviour of the probes is analysed. The melting curves are specific for bacteria that may be detected by the test. The data can also be exported and transferred to different software for an automated interpretation. A typical result is shown in Fig. 2.4, for Lactobacillus brevis. The upper panel demonstrates the decrease in the signal with increasing temperature. The lower panel demonstrates the first negative derivative of this declining curve (dF/dT), which is easier to interpret. The peaks at 55°C and 46°C and the shape of the curve are used for the identification of the contaminant. Figure 2.5 shows the melting curve of Pectinatus cerevisiiphilus. The curve is clearly different to that of L. brevis. The main peak occurs at a temperature of about 67.5°C and a second peak at 63°C.

DETECTION OF BEER SPOILAGE MICROORGANISMS

Fig. 2.3

Positive results of a polymerase chain reaction run. Bottom line: negative control.

Fig. 2.4

Melting curve of Lactobacillus brevis.

19

20

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 2.5

Melting curve of Pectinatus cerevisiiphilus.

However, the melting curves cannot differentiate among all 14 bacteria (Table 2.1). For example, L. brevis and L. lindneri yield the same melting curve. foodproof® Beer Screening allows the differentiation of nine groups of bacteria, as shown in Table 2.1. Bacteria of the same genus cannot be distinguished, as they exhibit the same melting curve, but differentiation between the genera is always possible. In the case of a mixed contamination (more than one species) the melting curve cannot always be assigned to the bacteria present. In many cases the melting peaks of each species are recognisable, but sometimes this is not the case. Such instances necessitate the use of speciesspecific PCR.3 Another possibility is to isolate the cells by microbiological methods and run the screening PCR again.

2.4

Conclusions

foodproof® Beer Screening enables analysis for the presence and absence of 14 different beer spoilage bacteria in a single reaction. When a positive result occurs, differentiation of the bacteria in groups by melting curve analysis is possible in most cases (see Table 2.1). This analysis requires an additional 15 min following the PCR with no further preparation of the sample, and is conducted using the LightCycler™. To differentiate between the species that cannot be separated by the melting curve, a specific PCR may be required.

DETECTION OF BEER SPOILAGE MICROORGANISMS

21

This new approach to PCR detection of relevant beer spoilage bacteria allows the use of this method as a quality-control tool in the routine laboratory of the brewery. The time savings compared with classical microbiology and the high specificity, sensitivity and quality of the results (detection and identification of groups) offer new opportunities to the user. The ready-to-use and brewery-tailored system reduces the time for analysis drastically, with a single test.

References 1. Kiehne, M. (2001) Schnellnachweis von bierschädlichen Mikroorganismen mittels foodproof® Beer Screening. Brauerei Forum Nr. 3/2001, 72–73. 2. BIOTECON Diagnostics (2001) Fast detection of beer spoilage microorganisms drink. Technol. Market. March, 23–24. 3. Kiehne, M. (2001) Détection rapide des microorganismes de fermentation de la bière. Liquides Conditionnement No. 292, 32e année.

Part 2

Brewing Yeast Stress Responses During Handling

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

3 The Impact of Ethanol Stress on Yeast Physiology A. LENTINI, P. ROGERS, V. HIGGINS, I. DAWES, M. CHANDLER, G. STANLEY and P. CHAMBERS

Abstract The impact of ethanol stress on brewing yeast physiology is an area that is still not fully understood. It has been shown that ethanol, a product of the beer fermentation process, has the greatest impact on yeast performance, by inhibiting cell growth and viability, causing changes in metabolic pathways, cellular structure and function. The degree of ethanol tolerance exhibited by a yeast cell will determine its suitability for fermentation as well as the number of times the yeast could be repitched into subsequent fermentations. This study investigated the impact of ethanol on the physiological status of brewing yeast on a structural and molecular basis. The first part of the study examined the yeast cell and changes that occur within the cell membrane (structure and fluidity), cell-wall structure, protease release and overall vitality of the yeast during prolonged storage under varying conditions of ethanol concentration and temperature. The characteristic physiological signs of cell stress by ethanol are accompanied at the molecular level by the induction of stress response genes. These can then be related to changes in cell-wall composition, trehalose and stress proteins. The last part of this study involved the examination of the technology of yeast genome analysis (Gene Microarrays), to monitor the impact of ethanol stress on gene expression. Yeast genome wide transcription analysis technology allows for the identification of genes with significant and specific differential expression to changes in environmental conditions. The study identified a number of genes that were up-regulated when subjected to ethanol stress. These include genes responsible for sugar metabolism, cell-wall structure, stress responses and transport functions. It is envisaged that using gene expression analysis techniques will provide a process to identify genes that can be monitored for their impact on yeast health and activity and can lead to a greater understanding of the physiological behaviour and structure of brewing yeast under various fermentation conditions.

3.1 Introduction The principal role of brewing yeast during fermentation is to produce ethanol, carbon dioxide and other flavour-active compounds. It is these by-products of fermentation that distinguish a specific beer product from other compatible products. While ethanol is seen as a desirable by-product of the fermentation process, its accumulation during fermentation can result in a significant chemical stress on the physiological status of the yeast cell. The impact of ethanol stress on brewing yeast has been previously reviewed by researchers.1–7 Ethanol as a chemical stressor inhibits cell growth and viability, and causes changes in metabolic pathways, increases in fermentation times, changes in yeast cell wall and membrane structure and function, and modifications in gene expression (i.e. induction of stress response genes). To understand further the impact of ethanol stress on the physiological condition of the yeast cell (activity and health), with particular reference to lager yeast strains, this study concentrated on three objectives:

• to better understand the interaction between environmental conditions and the yeast cell during the brewing process

26

BREWING YEAST FERMENTATION PERFORMANCE

• to determine how one can better diagnose and predict yeast performance in the

presence of ethanol stress factors develop appropriate diagnostic technologies to identify indicators that will assist in the understanding and control of yeast performance in an ethanol-rich environment.

• to

As part of the study to understand better the impact of ethanol stress on brewing yeast, the following investigations were undertaken:

• investigation into the impact of ethanol stress on the physiological structure and function of brewing yeast during storage

• investigation into the yeast molecular responses to ethanol stress using gene array

technology, to identify those genes that are activated by the presence of ethanol and determine how these relate to the structure and function of the yeast cell.

3.2 Materials and methods 3.2.1

Yeast storage trials

Yeast storage trials were undertaken as previously described by Lentini et al.7 In brief, lager yeast was stored in a 2 litre vessel with pressure and temperature control. The vessel maintained a carbon dioxide environment to minimise yeast exposure to oxygen. The vessels were stored at either 4 or 10°C. To determine the impact of ethanol stress, the yeast slurry was washed with non-alcoholic beer to remove any residual ethanol and then dosed with either 5 or 10% (v/v) food-grade ethanol. Samples were collected on a daily basis to monitor cell membrane lipid composition, cell wall trehalose content, yeast slurry pH and protease levels in the slurry (excreted from the yeast cell), and to monitor the viability and vitality of the stored yeast samples. 3.2.1.1 Membrane lipid composition. Saturated and unsaturated lipid composition of the brewing yeast was measured using the method described by Lentini et al.7 3.2.1.2 Trehalose content. Yeast cell wall trehalose was isolated and measured by the method described by Stewart.8 3.2.1.3 Yeast slurry pH. The pH of the storage yeast samples was monitored using a calibrated Orian 520A pH meter and combined electrode. 3.2.1.4 Yeast protease. Yeast protease release into the cell’s environment was measured using the method described by Mochaba et al.9 3.2.1.5 Yeast viability. The viability of the stored yeast samples was monitored using the methylene blue staining method.10 3.2.1.6 Yeast vitality. The vitality of the yeast sampled from the storage vessels was measured using two previously described procedures: the acidification power test (APT)11 and oxygen uptake rate (OUR).12

27

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

20 4°C Control 4°C Ethanol (10%) Total lipids (mg/g)

10°C Control 10°C Ethanol (10%) 15

10 0

1

2

3

4

5

6

7

8

Days Fig. 3.1 Level of lipid material in yeast cell membranes over a 7 day period when exposed to various ethanol and temperature stresses.

3.2.2

Gene array technology

The procedure for measuring the yeast’s molecular response to ethanol stress was that described by Higgins et al.13 and Chandler et al.14 In brief, the method for micro-array analysis involves the isolation of total RNA from the yeast cell (lager brewing yeast strain) using the method described by Ausubel15 and Higgins et al.13 Gene array analysis was performed using Yeast Index Genefilters® (Research Genetics) containing over 6000 yeast genes, according to the manufacturer’s instructions. They were hybridised with complementary DNA (cDNA) produced from total RNA according to the manufacturer’s instructions (Research Genetics). cDNA probes were then synthesised with [-33P] deoxycytidine triphosphate (dCTP) using Superscript™ II (Life Technologies) from extracted RNA. Following hybridisation, the membranes were analysed using either a FLA3000 Phosphor Imaging System (Fujifilm) or a Phosphor Imager (Molecular Dynamics). Filter comparisons were made using either Array Gauge™ Software (Fujifilm) or ImageQuaNT v4.2a software. 3.3 3.3.1

Results and discussion Impact of ethanol and temperature on the structure of the yeast cell membrane

From Fig. 3.1, it is evident that there was little difference between the samples (total lipids) except for the sample containing 10% ethanol and stored at 10°C. In this case, there was a higher lipid content after 2 days of storage, and then a significant increase after 6 days of storage. Figure 3.2 shows that the sample containing 5% ethanol at 4°C maintained a steady ratio of saturated to unsaturated fatty acids over the 7 day storage period. The ratio of saturated to unsaturated fatty acids declined in samples with increased ethanol

28

BREWING YEAST FERMENTATION PERFORMANCE

4.5 4

Ratio

3.5 3 2.5

4°C Control 10°C Control

4°C Ethanol (10%) 10°C Ethanol (10%)

2 0

1

2

3

4 Days

5

6

7

8

Fig. 3.2 Ratio of saturated to unsaturated fatty acids in yeast exposed to various ethanol and temperature effects over a 7 day period.

stress (i.e. samples with increased ethanol concentration and temperature). In these samples the level of unsaturated fatty acids increased significantly as the level of stress increased. The increase occurred mainly with C18:1 (oleic acid) and to a lesser extent C16:1. This observation has been made previously by Beaven et al.16 and Lentini et al.7 Chen found, in a study of highly ethanol-tolerant yeast, that the ethanol-tolerant yeast contained a higher percentage of unsaturated fatty acids, mainly higher acyl chained fatty acids.17 The presence of a higher level of unsaturated fatty acids increases the fluidity of the cell membrane in response to the physical effect of ethanol (i.e. tightening of the membrane). The change in the saturated to unsaturated fatty acid ratio (Fig. 3.2) would suggest that the yeast may be removing saturated fatty acids to enable a greater proportion of unsaturated fatty acids to be accommodated by the membrane (to increase ethanol tolerance). The mainly stable level of fatty acid concentration found in the membrane (Fig. 3.1) would suggest that the yeast may be biosynthesising unsaturated fatty acids. This could only occur in the presence of oxygen, hence the yeast may be picking up oxygen from the environment and selectively producing unsaturated fatty acids. This major increase in unsaturated fatty acids was seen in the sample containing high ethanol levels and stored at high temperatures. However, analysis of the sterol content of these yeast samples did not show a significant increase in ergosterol concentration, as would have been expected if the yeast was exposed to oxygen. The saturated:unsaturated ratio is an indication of the yeast cell’s ability to modify its membrane structure to ensure cellular integrity in response to environmental stresses (i.e. ethanol and temperature). The increase in unsaturated fatty acids is an indication that, under stress conditions, the yeast will actively seek oxygen from the environment and use it to produce a cellular structure, which is fluid, so to tolerate external stressors. 3.3.2

Cell-wall trehalose

The disaccharide trehalose is one of the most effective saccharides in stabilising the yeast cell against osmotic stress.18 Van Dijck et al.19 showed a correlation between cellular

29

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

12

4°C Control 4°C Ethanol (10%) 10°C Control 10°C Ethanol (10%)

Trehalose (% w/w)

10 8 6 4 2 0 0

1

2

3

4 Days

5

6

7

8

Fig. 3.3 Level of trehalose in yeast cell wall when yeast is exposed to various ethanol and temperature effects over a 7 day period.

trehalose and stress resistance within fermenting cells. Trehalose in the cell wall of the yeast cell has been proposed as a stress-protectant molecule.20 Trehalose has been shown to be an osmotic protectant,21 a thermoprotectant22 and a chemical detoxicant.23 Trehalose has been suggested to be an important osmoticprotectant and stress indicator in brewing yeast during fermentation of high- and very high-gravity worts.21 In these environments yeast is expected to produce very high levels of ethanol. From Fig. 3.3, it can be seen that as the degree of stress is increased (i.e. with increasing ethanol concentration and temperature), the level of trehalose isolated from the yeast cell wall increases. This can be seen as an indicator of a stress response by the yeast to increase its survival prospects under non-ideal conditions. It appears that after 2 days of storage under specific stress conditions the level of trehalose stabilised within the yeast cell wall. The yeast sample least subjected to stress did not significantly change its trehalose content during the 6 days of storage. The drop in trehalose content in the yeast sample most subjected to stress (10°C and 10% ethanol) after 5–6 days of storage is not fully understood. This may relate to the significant decrease in yeast viability that occurred during the corresponding period (results not shown). 3.3.3

Yeast slurry pH

As stress conditions increased (i.e. increased temperature and ethanol concentration) a corresponding increase in the pH of the yeast slurry occurred (Fig. 3.4). This increase in pH may be an indication of hydrogen ion uptake from the environment to ensure a membrane influx of essential nutrient uptake by the yeast cell to enable continual survival under harsh conditions.6 The increase in yeast slurry pH can consequently be used as an indicator of the level of stress to which the yeast is exposed during storage.7 The greater the level of stress, the greater the increase in slurry pH. The consequences of this increase in slurry pH can be seen in the viscosity of the slurry. Previous work24 has shown that with some brewing yeast strains an increase in

30

BREWING YEAST FERMENTATION PERFORMANCE

5.6

pH

5.2

4.8

4.4 4°C Control 10°C Control

4°C Ethanol (10%) 10°C Ethanol (10%)

4 0

1

2

3

4 Days

5

6

7

8

Fig. 3.4 pH of the yeast slurry when storage yeast is exposed to various ethanol and temperature effects over a 7 day period.

Protease (Abs)

0.8

0.6

0.4 4°C Control 10°C Control

0.2 0

1

2

3

4 Days

4°C Ethanol (10%) 10°C Ethanol (10%) 5

6

7

8

Fig. 3.5 Level of protease release from yeast held in storage over 7 days under various ethanol and temperature stresses.

pH results in an increase in the slurry’s viscosity. This has been shown to impact significantly on the ability of the yeast to be transported (pumped) through the brewery. Increasing the force required to pump the thick and viscous yeast slurry within the plant further increases the level of stress placed on the yeast, resulting in a further decrease in its activity and health. 3.3.4

Protease release from yeast

It has previously been speculated that under stress conditions, the yeast will excrete protease into its environment.9 This proteolytic enzyme can affect beer quality by decreasing the level of beer foam proteins, resulting in a decrease in beer head retention. Temperature and ethanol stress significantly impact on the level of protease released by the yeast (Fig. 3.5). It appears that the yeast stored at 4°C and 10% ethanol released the equivalent amount of protease to that of yeast stored at 10°C and

31

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

3.0 2.8 Acidification power

2.6 2.4 2.2 2.0

4°C Control 4°C Ethanol (10%) 10°C Control 10°C Ethanol (10%)

1.8 1.6 1.4 1.2 1 0

1

2

3

4

5

6

7

8

Days Fig. 3.6 Acidification power test results for yeast samples stored for 7 days under various ethanol and temperature stress conditions.

containing only 5% ethanol. Increasing the level of stress (both temperature and ethanol concentration) significantly increased the amount of protease released by the yeast cells. The greatest increase was achieved with the accumulative effect of both high temperature and high ethanol concentration (10°C and 10% ethanol). In this sample, the release of protease was evident within 1 day of storage. 3.3.5

Yeast vitality

A series of vitality tests was undertaken to determine the impact of ethanol stress on yeast activity when stored at various temperatures over a specific period. These tests were selected to determine the ability of the yeast to take up glucose (APT) and its ability to take up oxygen (OUR). These tests yield an indication of the yeast population’s ability to uptake essential nutrients (e.g. fermentable sugars and oxygen) and reflect potential fermentation capacity. 3.3.5.1 Acidification power test. Figure 3.6 shows the APT results for the various yeast samples subjected to variable temperature and ethanol concentrations over a 7 day period. The control yeast sample (4°C and 5% ethanol) showed no significant decrease in the rate of glucose uptake over 7 days of storage. However, as the level of stress increased (both ethanol and temperature), the level of glucose uptake decreased with time. The sample stored at 4°C and at 10% ethanol was slightly better at glucose uptake than the sample stored at 10°C and 5% ethanol. Both of these samples showed a gradual decline in the APT value over time. The samples containing the higher concentrations of ethanol showed a decline in glucose uptake within 1 day of storage. The most significant impact on glucose uptake was the extreme stress conditions of prolonged storage at 10°C and 10% ethanol. The synergist effect of high temperature and high ethanol levels enhances the decline in the yeast’s ability to uptake glucose. The decrease in the APT value became more pronounced after 4–5 days of storage.

32

BREWING YEAST FERMENTATION PERFORMANCE

OUR (mg/l/min/106 cells)

0.25

4°C Control 4°C Ethanol (10%) 10°C Control 10°C Ethanol (10%)

0.20 0.15 0.10 0.05 0 0

1

2

3

4 Days

5

6

7

8

Fig. 3.7 Oxygen uptake rate (OUR) results for yeast samples stored for 7 days under various ethanol and temperature stress conditions.

3.3.5.2 Oxygen uptake rate. In Fig. 3.7, it can be seen that the rate at which yeast samples uptake oxygen from the environment is significantly influenced by how the yeast is stored. Even under ideal conditions (4°C and 5% ethanol) there was a gradual decline in the OUR over the 7 days of storage. This rate of oxygen uptake significantly declined as the level of stress (ethanol and temperature) increased. It would appear that temperature has a greater influence than ethanol concentration on the yeast’s ability to uptake oxygen. Within 4 days of storage the yeast samples stored at high temperature (10°C) appeared to have lost the most of their ability to take up oxygen from the environment. At 7 days the OUR for these yeast samples appeared to be negligible. Both of these data sets give an indication of the impact of excess ethanol and higher than normal temperatures during yeast storage on the brewing yeast’s potential fermentation capabilities. Utilisation of these stressed yeast cells will cause variable fermentation performance (e.g. stuck fermentations or slower rates to reach the desired final gravity value) and beer flavour profiles (i.e. variable desirable flavours such as esters or higher levels of undesirable flavours such as diacetyl and hydrogen sulfide). The decline in yeast vitality under stress conditions was matched by the decline in yeast viability. 3.3.6

Changes in gene expression

It has previously been shown that environmental stresses can result in changes at the molecular level of the yeast cell by the induction of stress response genes.8 Several genes are known to be induced during ethanol stress, but limited information is available on the range and number of genes induced. Current procedures to identify stress genes have involved investigating changes in protein profiles, changes in messenger RNA (mRNA) profiles (i.e. Northern analysis and differential display) and gene array technology. To determine the influence of ethanol stress and its impact on gene expression (via gene array technology), initial laboratory experiments were performed with yeast in the absence and presence of 5% (v/v) ethanol. Yeast samples were compared after 2 h of incubation.

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

33

Fig. 3.8 Example of gene array analysis, showed the appearance of expressed HSP26 and GLK1 genes in a yeast sample subjected to ethanol stress.

In the current study, initially only those genes that up-regulated in response to ethanol stress were studied and reported. Work is currently being undertaken to determine the impact of ethanol stress on down-regulating yeast genes. In this study, genes were considered up-regulated when the ratio of stressed: non-stressed mRNA intensity increased by at least a factor of five. It was decided to be conservative in this analysis. It is common to consider a gene to be up-regulated when the ratio is increased more than two to three times (stressed: non-stressed). An example of up-regulated genes is shown in Fig. 3.8. In this example, the up-regulation of the HSP26 and GLK1 genes when the yeast samples were subjected to ethanol stress conditions are shown as dark spots on the gene filters. This compares with an absence of any markers in the non-ethanol-stressed sample (i.e. no indication of mRNA). Analysis of the gene array plates showed a variety of up-regulated genes in response to ethanol stress. These genes were grouped together according to their function. Table 3.1 lists some of the identified up-regulated genes and their function in yeast metabolism. The identified up-regulated genes were found to encode for proteins involved in stress tolerance (i.e. heat stress proteins and cell wall mannoproteins), trehalose metabolism, sugar metabolism (i.e. hexose transporter and glycolysis enzymes), electron transport chain, cell-wall structure (mannoprotein and other structural proteins, lipid metabolism and surface glycoproteins) and transport genes (small peptides and phosphate into vacuoles), as well as other functions such as flocculation and vacuole protein degradation. Various genes of unknown function were also identified as being up-regulated owing to ethanol stress. A closer examination of the up-regulated genes showed that at least 10 stress response genes were up-regulated. From these genes a group of heat stress proteins (HSPs) was highly regulated in response to ethanol stress. HSP26, a small heat shock protein known to be involved in heat stress, was highly expressed in response to ethanol stress.25

34 Table 3.1

BREWING YEAST FERMENTATION PERFORMANCE

Genes up-regulated in the response to ethanol stress

Locus name

Gene name

Fold induction

Stress genes YBR072W YER150W

HSP26

68 24

YBR054W

YR02

23

YDR171W

HSP42

13

YCR021C

HPS30

11

YLL026W

HSP104

9

YBR067C

TIP1

7

YMR251W-A

HOR7

6

YLL039C

UB14

5

YCL035C

5

Trehalose genes YML100W

TSL1

49

YBR126C

TPS1

28

YDR074W

TSP2

10

Metabolism genes (in order through glycolysis to the TCA cycle) YDR342C HXT7 8 YDR343C HXT6 7 YCL042W 54 YCL040W YFR053C YJL052W YGR192C

GLK1 HXK1 TDH1 TDH3

50 35 18 6

YCR012W

PGK1

7

YGR254W

ENO1

5

YHR174W YOR374W

ENO2 ALD4

7 29

YGL062W

PYC1

6

YCR005C

CIT2

11

YNR001C

CIT1

6

Electron transport chain YKL150W YER141W

MCR1 COX15

27 16

Characteristics Stress response, protein folding Unknown function, transcription depends on MSN2/4 Homologue to HSP30, unknown process Stress response, chaparone, cytoskeleton organisation YR01, stress response, plasma membrane Stress response, chaparone, protein folding Cell wall mannoprotein, cold and heat shock induced Stress response, membrane hyperosmolarity Stress response, protein degradation tagging Stress response (oxidative), glutaredoxin Trehalose-6-phosphate, homologous to TSP3 gene Stress response, carbohydrate metabolism, trehalose Stress response, carbohydrate metabolism, trehalose Hexose transporter Hexose transporter ORF next to GLK1, no known function Glycolysis, glucokinase, in cytosol Glycolysis, hexokinase, in cytosol Glycolysis, glucogenesis, in cytosol Glycolysis, glucanogenesis, glyceraldehyde-3-phosphate Glycolysis, glucogenesis, phosphoglycerate kinase Glycolysis, glucanogenesis, enolase Glycolysis, glucogenesis, enolase Aldehyde dehydrogenase, ethanol metabolism, mitochondria Glucanogenesis, pyruvate carboxlase Citrate synthase, glyoxylate cycle, glutamate biosynthesis Citrate metabolism, TCA cycle Cytochrome b 5 reductase Cytochrome c oxidase biogenesis, mitochondria inner membrane (continued)

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

Table 3.1

35

(continued)

Locus name

Gene name

Fold induction

Characteristics

Cell wall-associated genes YBR067C TIP1

7

YDR077W

SED1

5

YER081W YBR177C

ETH1

5 5

Cell wall mannoprotein, cold and heat shock induced Cell wall organisation, putative surface glycoprotein Structural protein of cytoskeleton Alcohol acyl transferase, lipid metabolism

Transport-associated genes YKR093W PTR2

145

YER053C

16

Other up-regulated genes YKL043W PHD1

11

YKL103C

LAP4

11

YHR018C

ARG4

10

YHR211W YKL035W

UGP1

9 9

YNL031C YPL265W YOR185C

HHT2 DIP5 GSP2

6 5 5

Selection of genes with unknown functions YBR214W YGL037C YGR161C YHL021C YMR195W YKL044W YGL117W YDR533C YBR139W YPL092W SSU1

23 13 11 10 10 9 8 8 8 7

Transport of small peptides into cell Phosphate transporter, in vacuole Specific RNA polymerase transcription factor Vacuolar aminopeptidase, vacuolar protein degradation Arginine biosynthesis, argininosuccinate lyase, cytosol Flocculation Uridinephosphoglucose phophory lase Histone H3 Dicarboxylic acid permease GTP binding protein Unknown function Unknown process, nicotinamidase Unknown function Unknown function Unknown function Unknown function Unknown function Unknown function Unknown function Unknown function, sensitive to sulfite

TCA: tricarboxylic acid; ORF: open reading frame.

Twelve genes involved in sugar metabolism were also highly expressed under ethanol stress conditions; in particular, the GLK1 (glycokinase) and HXK1 (hexokinase) genes were highly up-regulated. Both genes are involved in the first stages of glycolysis and may play an important role in facilitating glucose transport into the yeast cell and delivering glucose into the glycolytic pathway. Recently, Alexandre et al.25 reported on the impact of ethanol stress on gene expression within yeast. They found that a large number of the up-regulated genes appeared to be involved in energy metabolism, implying that managing the energy pool [i.e. adenosine triphosphate (ATP)] may improve the yeast’s ability to respond to ethanol stress. The relationship between intensity and activity of the measured mRNA

36

BREWING YEAST FERMENTATION PERFORMANCE

spot cannot be fully understood until the level and specificity of proteins associated with the specific gene have been determined. Many of the observations made here and in other studies on the physiological changes to the yeast cell during various stress conditions (i.e. changes in cell wall trehalose and membrane lipid contents, the ability of the yeast to uptake glucose: acidification power test, cell surface properties, yeast growth rates, etc.) are reflected by the identified genes that were up-regulated in the presence of ethanol stress. This study confirms the yeast’s ability to respond to stress conditions on a molecular basis and the necessary modifications to the yeast’s physiological status and function, to ensure the survival of the cell. 3.3.6.1 Observations on using gene array technology. In the course of using gene array technology to determine the impact of ethanol stress on yeast metabolism, the following observations were made. (i) Industrial yeast strains have a higher ratio of ribosomal RNA to mRNA than laboratory yeast strains. (ii) The quantity of RNA isolated from industrial cells was lower than for laboratory-grown yeast strains. (iii) Industrial yeast strains required at least 10 times the amount of total RNA to generate the cDNA probes. (iv) Current gene array filters are expensive and can only be reused up to three to five times. The development of newer and longer life (microchip-based) filters will significantly decrease the cost of the analysis, thereby increasing the potential of this technique for becoming a standard biochemical analytical tool.

3.4

Conclusions

The purpose of this study was to investigate the impact on the physiological structure and function of brewing yeast when subjected to ethanol stress over time, at both a biochemical and a molecular level. The contributory effect of temperature on ethanol stress was also examined. The results showed that ethanol stress in combination with high storage temperatures significantly impacted on the yeast cell wall and membrane structure, as well as it vitality and viability. It also caused the release of protease into its surrounding environment, which could affect the quality of the final beer product (i.e. beer foam). The study also tried to link changes that occur on a molecular basis with physical characteristics of the yeast cell, by identifying those genes that are responsible for changes made by the yeast to ensure survival under stress conditions (i.e. changes in trehalose levels in the cell wall, the lipid composition in yeast membrane and the yeast vitality). The current study only investigated those genes that up-regulated when exposed to ethanol stress. Work is continuing to investigate further the impact of upregulated as well as down-regulated genes on the yeast’s ability to adapt to chemical and environmental stress conditions. The use of gene expression analysis provides a process to identify genes that can be monitored for their impact on yeast health and activity, thereby achieving a greater understanding of the physiological behaviour and structure of brewing yeast under various fermentation conditions.

THE IMPACT OF ETHANOL STRESS ON YEAST PHYSIOLOGY

37

The advances being made in gene array technology will enable the use of real-time quantitative polymerase chain reaction technology to measure the level of specific identified gene markers within the yeast, to detect conditions that may impact on its health or fermentation performance during storage or fermentation.

Acknowledgements We would like to thank Carlton and United Breweries, The University of New South Wales and Victoria University for their support.

References 1. Casey, G.P. and Ingledew, W.M. (1986) Ethanol tolerance in yeasts. CRC Crit. Rev. Microbiol. 13, 219–280. 2. Jones, R.P. (1987) Factors affecting deactivation of yeast cells exposed to ethanol. J. Appl. Bacteriol. 63, 153–164. 3. Jones, R.P. (1990) Roles for replicative inactivation in yeast ethanol fermentations. Crit. Rev. Biotechnol. 10, 205–222. 4. D’Amore, T. and Stewart, G.G. (1990) Ethanol tolerance of yeast. Enzyme Microb. Technol. 9, 322–330. 5. D’Amore, T., Panchal, C.J., Russell, I. and Stewart, G.G. (1990) A study of ethanol tolerance in yeast. Crit. Rev. Biotechnol. 9, 287–304. 6. Walker, G.M. (1998) Yeast: Physiology and Biotechnology. John Wiley and Sons, Chichester, pp. 163–167. 7. Lentini, A., Mariani, M., Takis, S. et al. (1998) An overview of the physiological changes to the structure and activity of the yeast cell during fermentation, storage and when subjected to successive repitchings. Proc. 25th Conv. Inst. Brew. Asia Pacific Sect., Perth. 8. Stewart, P.R. (1975) Analytical methods for yeast. In Methods in Cell Biology, Vol. XII, Prescott, D.M. (ed.). Academic Press, London, pp. 111–145. 9. Mochaba, F.M., O’Conner-Cox, E.S.C. and Axcell, B.C. (1995) Practical implications of yeast protease activity. Proc. Inst. Brew. Conv. Central and Southern Africa Sect., Victoria Falls, 5, 152–158. 10. Institute of Brewing (1997) Methods of Analysis, Vol. 2, Microbiology, Section 21.33, Assessment of Yeast Viability. IOB, London. 11. Opekarova, M. and Sigler, K. (1982) Acidification power: indicator of metabolic activity and autolytic changes in Saccharomyces cerevisiae. Fol. Microbiol. 27, 395–403. 12. Kara, B.V., Dauod, I. and Searle, B. (1987) Assessment of yeast quality. Proc. 21st Cong. Eur. Brew. Conv., Madrid, pp. 409–416. 13. Higgins, V.J., Oliver, A.D., Day, R.E. et al. (2001) Application of genome-wide transcriptional analysis to identify genetic markers useful in industrial fermentations. Proc. 28th Cong. Eur. Brew. Conv., Budapest. 14. Chandler, M., Stanley, G., Rogers, P. and Chambers, P. (2001) Profiling ethanol stress response genes in Saccharomyces cerevisiae. Proc. XXth Int. Conf. Yeast Genet. Molec. Biol., Prague. 15. Ausubel, F. (1998) Current Protocols in Molecular Biology. Greene, New York. 16. Beaven, M.J., Charpentier, C. and Rose, A.H. (1982) Production and tolerance of ethanol in relation to phospholipid fatty acyl composition of Saccharomyces cerevisiae. J. Gen. Microbiol. 128, 1447–1455. 17. Chen, E.C. (1981) Release of fatty acids as a consequence of yeast autolysis. J. Am. Soc. Brew. Chem. 39, 117–124. 18. Crowe, J.H., Crowe, L.M. and Chapman, D. (1984) Preservation of membranes in anhydrobiotic organisms: the role of trehalose. Science 223, 701–703. 19. Van Dijck, P., Colavizza, D., Smet, P. and Thevelein, J.M. (1995) Differential importance of trehalose in stress resistance in fermenting and non-fermenting Saccharomyces ceravisiae cells. Appl. Environ. Microbiol. 61, 109–115. 20. Wiemken, A. (1990) Trehaole in yeast: stress protectant rather than reserve carbohydrate? Antonie van Leeuwenhoek 58, 209–217.

38

BREWING YEAST FERMENTATION PERFORMANCE

21. Majara, M., O’Conner-Cox, E.S.C. and Axcell, B.C. (1996) Trehalose – an osmoprotectant and stress indicator compound in high and very high gravity brewing. J. Am. Soc. Brew. Chem. 54, 149–154. 22. DeVirgillo, C., Hottiger, T., Dominguez, J. et al. (1994) The role of trehalose systhesis for the acquition of thermotolerance in yeast. 1. Genetic evidence that trehalose is a thermoprotectant. Eur. J. Biochem. 219, 179–186. 23. Attfield, P.V. (1987) Trehalose accumulates in Saccharomyces ceravisae during exposure to agents that induce heat shock response. FEBS Lett. 225, 259–263. 24. Lentini, A., Hawthorne, D.B. and Kavanagh, T.E. (1992) A rheological study of yeast during handling and storage. Proc. 22nd Conv. Inst. Brew. Aust. N.Z. Sect., Melbourne, p. 198. 25. Alexandre, H., Ansanay-Galeote, V., Dequin, S. and Blondin, B. (2001) Global gene expression during short term ethanol stress in Saccharomyces cerevisae. FEBS Lett. 498, 98–103.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

4 Yeast Physical (Shear) Stress: The Engineering Perspective R.A. STAFFORD

Abstract The handling of yeast and the response of the cell to excessive mishandling have received a considerable amount of attention to date. Issues such as cell wall attrition, secretion of intracellular substances and slurry alkalisation owing to cell autolysis have been well documented. It is, however, surprising to note that the cause of such damage, often described as shear, has received little attention within the brewing industry. Within the larger biochemical engineering field, the cause of cell shear damage is a well-documented phenomenon and has been thoroughly investigated, principally because of the often synchronous need to preserve cellular integrity during processing, up until such time that cell disruption is required to harvest the cell contents. The intention of this paper is to review the types of shear damage that have been reported for brewer’s yeast and then to describe some of the possible causal damage mechanisms that may be present within a modern brewery handling circuit.

4.1 Introduction The response of brewing yeast to environmental stresses incurred during yeast handling has received considerable attention within the literature. Stresses produced as a result of elevated levels of osmotic and hydrostatic pressure, temperature and ethanol, to name but a few, have all to varying degrees been investigated by the bioscience community. Comprehensive lists1 of these environmental stresses invariably place hydrodynamic shear stress last and generally little further information is provided. Shear stress is that incurred whenever yeast is physically moved within a brewery, either by its own accord, e.g. by the action of a convection current within a fermenter, or by artificial means, e.g. pumping, agitation or centrifugation. Clearly, an essential part of any modern yeast handling circuit is the need to move the yeast physically from one fermenter, through a handling circuit, to the next fermenter to be pitched (Fig. 4.1). The response of brewing yeast to such physical handling has been investigated by a relatively small number of researchers and these will be reviewed later in this paper. However, for the engineering designer who is interested in designing handling equipment, it is extremely difficult to utilise such experimental data, principally because of the lack of accompanying data on the intensity of the handling conditions, i.e. the stimuli, which induced the cell response. Without such data the designer is faced with the option of using engineering commonsense and developing and applying a heuristic approach to sizing and specifying operating conditions. Examples of these are specifications involve agitator tip speeds, pump rotational speeds2 and line velocities.3 An opportunity exists to gain considerable insight into the nature of cellular shear stress, or more particularly the stimuli, by embracing the more general field of bioengineering, with the often synchronous need to preserve cellular integrity during

40

Fig. 4.1

BREWING YEAST FERMENTATION PERFORMANCE

Schematic of a yeast handling circuit.

processing, up until such time that cell disruption is required to harvest the cell contents. Such insight may help towards the identification and use of more appropriate engineering parameters to reduce the cell stress response during handling. 4.1.1

Yeast cell response to shear stress

The response of yeast to shear has been found, in part at least, to follow what may be termed ‘classical’ stress response, i.e. depleted glycogen and modifications in trehalose,4–6 reduced viabilities,6–10 increased slurry pH7,8 and leakage of intracellular proteases.8,10 Impaired fermentation performance following pitching of sheared yeast has been identified with off-flavours5 and reduced postfermentation viabilities,8 indicating a reduction in the brewing cycle lifespan of the culture. Perhaps less classical and more particular to shear stress is the impact on the cell wall and the functionality that the wall provides. Release of invertase and melibiase11 (cell wall enzymes), together with mannan and glucan8,11,12 (cell wall polysaccharides), has been found in slurry supernatant with increasing exposure times to shear. Haze generation in yeast slurry supernatant8,11 and beer12 caused by the presence of mannan and glucan has been ascribed to cell wall attrition during shearing. Not surprisingly, as a consequence of this cell wall attrition, impaired flocculation performance9 (measured as cell surface charge, hydrophobicity and flocculence) and reduced cells in suspension during fermentation6 have been noted. Viewing these cell responses from a practical brewing aspect, it can be seen that the consequences of shear-induced yeast damage are serious, including changes in flocculation patterns, poor beer clarity, poor head stability and off-profile beer. 4.1.2

Cell stimuli

Having established from the literature that the response of brewing yeast to shear is not beneficial to either yeast health or beer quality, the question needs to be asked: how

YEAST PHYSICAL

( SHEAR )

STRESS : THE ENGINEERING PERSPECTIVE

41

can such cell damage be avoided? The common answer is to ‘reduce shear’. In practice, this is frequently done by adjusting equipment operating parameters such as reducing pump speeds, line velocities and agitator speeds, or reducing the exposure times of yeast to high shear environments, e.g. reducing yeast tank occupancy times. However, this is still very much the empirical approach of changing the operating parameter until yeast damage problems are either eliminated or reduced to an acceptable level. In some cases this approach requires a capital-intensive change of process equipment, with no real guarantee of improved yeast quality, because the underlying cause of yeast physical damage has not been identified. This situation is akin to trying to reduce temperature-induced yeast stress without the means of measuring temperature. A more satisfying solution would be to identify the cause, or stimulus, of yeast physical damage and determine the sensitivity of the cells’ response to this stimulus. Concepts and likely strategies to allow this to be attempted will now be discussed. 4.1.3

Newton’s law of viscosity: a gross deforming force

One of the defining equations of fluid flow is Newton’s law of viscosity. This relates the viscosity of the fluid () to the shear stress ( ) and shear rate ( ) present within the fluid, namely:  

Shear rate can be thought of as an indicator of the degree of agitation within a fluid; in the case of a tank agitator, the shear rate is proportional to the speed of rotation of the agitator. Shear stress is the stretching or deforming force that acts on the cell owing to the shear rate. As shown by Chisti,13 shear stress can be sufficient to deform the cell sufficiently to exceed the bursting strength of the cell, or in less extreme cases, merely temporarily deform the cell. Roberts et al.14 showed, using a micromanipulation technique, that cell bursting strengths were significantly lower during the exponential phase (90 N) than the stationary phase (50 N). Thus, there is a direct relationship between the level of shear rate within a fluid and the deforming force that a yeast cell suspended within the fluid would experience. Unfortunately, Newton’s law is only appropriate for fluids that have a constant viscosity or, more precisely, where the viscosity is not a function of shear rate. Yeast slurries greater than approximately 20% (spun solids) exhibit viscosities that are a strong function of shear rate. This behaviour will now be discussed further. 4.1.4

Yeast rheology

At the typical slurry concentrations encountered in a brewery yeast handling circuit (40–70% wet spun solids), yeast slurries have been shown to be highly non-Newtonian, that is their viscosities are a strong function of shear rate. They have also been shown to be strong functions of consistency and temperature.15–18 Figure 4.2 shows the extent of this dependency for a production lager yeast slurry. As shear rate increases, e.g. the rotational speed of a tank agitator increases, the apparent viscosity of the slurry reduces rapidly from values in excess of 100 Poise (P) to less than 1 P ( 10 to 0.1 N/m2 per second). Such behaviour is termed shear-thinning, and in agitated

42

BREWING YEAST FERMENTATION PERFORMANCE

Apparent viscosity (Poise)

1000 100 10 1 0.1 0

50

100

150

200

250

300

350

400

450

500

Shear rate (/s) Fig. 4.2 Variation in apparent viscosity with shear rate for a production lager yeast slurry of approximately 50% (spun wet solids), showing prominent shear-thinning behaviour.

vessels it accounts for the generally poor mixing performance of conventional yeast agitators.19 Localised high shear regions near the impeller produces low viscosities, which aid mixing. However, in the bulk of the tank, away from the impeller, the shear rates are lower, resulting in a higher viscosity which produces poorer mixing. Not only does this behaviour produce poor mixing, but it results in a split population of cells, which have quite different exposure cycles to high shear conditions. Clearly, specification of a single value of viscosity applicable throughout an entire yeast handling circuit is inappropriate. However, no generic yeast rheology data exist as cell strain, flocculation behaviour, slurry pH20,21 and cell size all influence rheological behaviour; hence, in practice, such arbitrary use of a single viscosity value is commonplace. 4.1.5

Methods of estimating shear rate of agitated systems

Perhaps the most common implicit indicator of shear rate is the concept of agitator tip speed ut. This is calculated by: ut  Nda where N is the agitator rotational speed (revolutions/s) and da is the agitator diameter. In practice, modern yeast tank agitators commonly run with tip speeds between 2 and 3 m/s, with the rotational speed being adjusted for the agitator diameter. The use of tip speed as an indicator of shear rate is also found in industrial bioreactors, where 5.5 m/s is a typical figure across several scales of reactors.22 This approach is a simple one to apply and for agitators of approximately the same size it has a sound basis. However, if substantial variations in agitator diameter are considered, then its use is contradicted by the Otto–Metzner correlation.23 This relates the agitator rotational speed to the mean shear rate ( –) around the agitator by an empirical constant (k),

–  kN Hence, if a constant tip speed condition is applied, then agitators of smaller diameter produce higher values of average shear rate, which may stimulate a greater stress response from yeast cells.

YEAST PHYSICAL

4.1.6

( SHEAR )

STRESS : THE ENGINEERING PERSPECTIVE

43

Energy dissipation rate

As stated earlier, an important aspect of yeast handling is the need to move yeast physically around the brewery. This is done using pumps, agitators, etc., which transfer energy into the yeast slurry which may ultimately be dissipated into the yeast cells with potentially damaging effects. Quantification of energy dissipation could be used to describe the damage potential of a process. In the case of an agitator, for example, the energy delivered can be easily measured, or estimated using the power number concept;24 however, the difficulty is assessing what volume (or mass) of yeast this energy is dissipated within. The most obvious volume to use is the volume of yeast contained within an agitated yeast tank, but this assumes homogeneous energy dissipation. This is far from the case in practice, with energy dissipation rates measured in close proximity to the agitator being up to 15 times the mean value dissipated in the entire tank volume.25 4.1.7

Kolmogorov turbulence scale

As stated above, any form of fluid movement will ultimately dissipate energy into yeast cells that are present within the fluid. The mechanism for energy transfer from fluid to cell is that of eddy dissipation, where energy within the fluid is considered to be transmitted from successive eddies, reducing in size, until viscosity ultimately dissipates the energy as heat (viscous dissipation). Eddies (or vortices) are rotational turbulent structures present within fluids, the size of which can be characterised by the Kolmogorov length scale of turbulence :  3      ε  

0.25

where  is the kinematic viscosity and ε is the energy dissipation rate per unit mass. It is generally believed that for eddy dissipation to be detrimental to cells,  must be smaller than the physical size of the cells. If the eddy is larger, i.e.  cell size, then the cell will be entrained into the eddy and will change its orientation to reduce the shear forces acting upon it. Alternatively, if   cell size, then the eddy has insufficient energy to entrain the cell and instead locally dissipates its kinetic energy, as heat, at the cell wall. This indicator of cell damage potential is complicated by the difficulty in estimating appropriate values of viscosity for a non-Newtonian system together with an energy dissipation rate for such an inhomogeneous system as found within a yeast tank. Furthermore, from the previous discussion of cell deformation due to the action of shear stress, entrainment in an eddy, i.e.  cell size, may be sufficient to deform the cell temporarily, which could induce a cell stress response. 4.1.8

Residence/exposure time

One solution to reducing the potential of a process to damage yeast is to reduce the residence or exposure time of the yeast to that process. Hence, in yeast tanks, for

44

BREWING YEAST FERMENTATION PERFORMANCE

instance, either occupancy time within the tank could be reduced, or intermittent agitation could be used to reduce the effective contact time. In the case of green beer centrifuges, reduced flow rates, which produce longer yeast retention times between desludgings, have been found to produce increased beer haze attributed to cell wall attrition.12 A further complication of considering residence time is illustrated by the case of pipe transfer under laminar or turbulent conditions. In general, turbulent flow conditions would be considered to be more energetic and hence stressful to yeast. However, Robinson26 showed that laminar transfers induced a greater response in yeast than turbulent ones (as measured by increased protease release and reduced fermentation rates). This behaviour was attributed to the interaction between the residence time of the near-wall regions (where the shear stresses are higher) and the average velocity in these regions (lower in the case of laminar than turbulent). Hence, although the near-wall region is more intense in turbulent flow, the residence time in this region is significantly shorter, suggesting that a minimum contact time is required to produce a cellular response.

4.2

Conclusions

The response of brewing yeast to hydrodynamic shear stress has been shown to be deleterious to both yeast health and function, and beer quality. Despite this, the majority of brewing research literature has focused on the response of the cell to shear stress, and the nature and quantification of the shear stress (the stimuli) have been largely ignored. Several concepts have been introduced from the biochemical engineering field, which may offer the opportunity to resolve this. Some of the difficulties associated with the potential use of these concepts have been discussed and it has been proposed that the principal barrier to their incorporation is the lack of understanding of yeast slurry rheology. Studies to date have shown that yeast rheology is complex and strongly multifactorial, which may make its determination for process design or optimisation prohibitively time-consuming or expensive.

Acknowledgements The Directors of The South African Breweries are thanked for permission to publish this work. Thomas Stoupis, ICBD, Heriot-Watt University, UK, is thanked for lengthy discussions on this topic and Professor Graham Stewart, ICBD, is thanked for initiating the author’s interest in this area.

References 1. Walker, G.M. (1998) Yeast Physiology and Biotechnology. John Wiley and Sons, London. 2. Jespen, E. (2000) APV, Denmark. Personal communication. 3. Ball, A. (1994) Brewery yeast routing principles. Brewer February, 53–56.

YEAST PHYSICAL

( SHEAR )

STRESS : THE ENGINEERING PERSPECTIVE

45

4. Sall, C.J., Seipp, J.F. and Pringle, A.T. (1988) Changes in brewer’s yeast during storage and the effect of these changes on subsequent fermentation performance. J. Am. Soc. Brew. Chem. 46, 23–25. 5. Pickerell, A.T.W., Hwang, A. and Axcell, B.C. (1991) Impact of yeast-handling procedures on beer flavour development during fermentation. J. Am. Soc. Brew. Chem. 49, 87–92. 6. McCaig, R. and Bendiak, D.S. (1985) Yeast handling studies. I. Agitation of stored pitching yeast. J. Am. Soc. Brew. Chem. 43, 114–118. 7. Kawamura, K., Makotot, M., Jimbo, E. and Okamato, Y. (1999) The development and successful operation of a sanitary yeast tank with a stirrer capable of uniform mixing without damage to the yeast. Proc. Eur. Brew. Cong. 83, 719–726. 8. Stafford, R.A., Barnes, Z.C. and Stoupis, T. (2001) A comparison of traditional and novel yeast-tank agitator systems. Proc. 8th Conv. Inst. Guild Brew. Africa Sect., Sun City, pp. 150–156. 9. Harrison, S.T.L. and Robinson, A. (2001) Disk stack centrifugation for the recovery of brewers’ yeast: its effect on yeast cell surface, flocculation and fermentation performance. Proc. 8th Conv. Inst. Guild Brew. Africa Sect., Sun City, pp. 157–164. 10. Harrison, S.T.L., Basson, L., Robinson, A. et al. (1997) Mechanical handling of brewer’s yeast during cropping and its effect on yeast quality. Proc. 6th Conv. Inst. Brew. Central and Southern Africa Sect., Durban, pp. 55–60. 11. Lewis, M.J. and Poerwantaro, W.M. (1991) Release of haze material from the cell walls of agitated yeast. J. Am. Soc. Brew. Chem. 49, 43–46. 12. Siebert, K.J., Stenroos, L.E., Reid, D.S. and Grabowski, D. (1987) Filtration difficulties resulting from damage to yeast during centrifugation. Tech. Q. Master Brew. Assoc. Am. 24, 1–8. 13. Chisti, Y. (2001) Hydrodynamic damage to animal cells. Crit. Rev. Biotechnol. 21, 67–110. 14. Roberts, A.D., Zhang, Z., Young, T.W. and Thomas, C.R. (1994) Direct determination of the strength of brewing yeast cells using micromanipulation. Proc. 1994 Inst. Chem. Eng. Res. Symp. pp. 73–75. 15. Reuß, M., Josic´, D., Popovic´, M. and Bronn, W.K. (1979) Viscosity of yeast suspensions. Eur. J. Appl. Microbiol. 8, 167–175. 16. Aiba, S., Kitai, S. and Ishida, N. (1962) Density of yeast cell and viscosity of its suspension. J. Gen. Appl. Microbiol. 8, 103–108. 17. Fatile, I.A. (1985) Rheological behaviour of concentrated yeast suspensions. J. Chem. Technol. Biotechnol. 35B, 94–100. 18. Lenoël, M., Meunier, J-P., Moll, M. and Midoux, N. (1987) Improved system for stabilising yeast fermenting power during storage. Proc. Eur. Brew. Cong. 43, 425–432. 19. McCabe, W.L., Smith, J.C. and Harriott, P. (1985) Unit Operations of Chemical Engineering, 4th edn. McGraw-Hill, New York, p. 229. 20. Wheacroft, R., Lentini, A., Tai, L. et al. (1993) Critical control points analysis for optimising yeast handling. Proc. 4th Conv. Inst. Brew. Central and Southern Africa Sect. pp. 153–160. 21. El-Temtamy, S., Farahat, L., Nour el-din, A. and Gaber, A. (1982) Non-Newtonian behaviour of yeast suspensions. Eur. J. Appl. Microbiol. Technol. 15, 156–160. 22. Einsele, A. (1978) Scaling-up biorectors. Process Biochem. July, 13–14. 23. Metzner, A.B., Feehs, R.H., Ramos, H.L. et al. (1961) Agitation of viscous Newtonian and nonNewtonian fluids. J. Am. Inst. Chem. Eng. 7, 3–9. 24. Cliffe, K. (1988) Bioreactors. In: Biotechnology for Engineers (Biological Systems in Technological Processes), Spragg, A. (ed.). Ellis Horwood, pp. 277–301. 25. Baldyga, J. and Bourne, J.R. (1988) Calculation of micromixing in inhomogeneous stirred tank reactors. Chem. Eng. Res. Dev. 66, 33–38. 26. Robinson, A. (2001) Mechanical handling effects on brewers’ yeast. PhD Thesis, University of Cape Town, South Africa.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

5

The Osmotic Stress Response of Ale and Lager Brewing Yeast Strains P.A. WHITE, A.I. KENNEDY and K.A. SMART

Abstract Brewing yeast is subjected to biological, chemical and physical stress during fermentation and yeast handling. It has been recognised that the transfer of yeast from slurry storage to pitching environments represents a potential source of osmotic stress, exacerbated by the use of high-gravity worts. One general mechanism used as a response to hypertonic conditions is the accumulation of compatible solutes. Compatible solutes can be defined as being those osmolytes that may be accumulated by the cell to increase internal osmolarity, and thus increase the retrieval of water from the environment, without affecting the biochemical or physical processes in the cell. In the brewing context, the accumulation of compatible solutes occurs on inoculation of yeast biomass into the wort. The relative concentration of these osmolytes may therefore indicate brewing yeast stress and subsequently capacity to ferment. This paper addresses the problematic area of determining a possible biomarker of osmotic stress, by examining the potential role of glycerol, a molecule implicated in the osmotic stress response. Glycerol accumulation during osmotic stress is dependent on strain, physiological state of the population and the solute used to elicit the osmotic stress response. The physiological responses to osmotic stress are discussed in terms of viability (as determined by transmembrane potential and intracellular reducing power) and vacuolar changes.

5.1 Introduction It is generally accepted that the physiological state of the yeast directly influences fermentation performance and resulting beer quality. The requirement of the brewing industry to obtain information concerning the quality of brewing yeast has been the driving force for research concerning brewing yeast stress responses.1 To date, the influence of osmotic stress on yeast quality and subsequent fermentation performance has received relatively little attention, although it is an important consideration wherever acid washing and pitching into high- and very high-gravity worts is practised.2–4 Although both hypotonic (where the osmolarity of intracellular fluids is higher than the external medium) and hypertonic (the converse) transitions occur within the brewing process, the former stress predominates. Indeed, there is a positive correlation between the degree of hypertonic stress imposed during pitching and the gravity of the wort used. High- and very high-gravity brewing therefore imparts two forms of stress on brewing yeast,3 albeit at different stages of fermentation, in the form of osmotic stress and ethanol toxicity. Several strategies may be used to combat cellular exposure to hyperosmotic stress (Fig. 5.1). Where the stress imposed is limited by either duration or magnitude, the initial response will involve osmotolerance; in some cases this will involve the accumulation of osmoprotectant molecules. These molecules stabilise cellular components such as membranes, enzymes, other proteins and possibly nucleic acids, with little

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

47

Osmotolerance

Osmotic adjustment

Osmoprotectant

Osmotica

Compatible solutes

Trehalose

Glycerol

Fig. 5.1 Schematic diagram showing the correct hierarchy and relationships between commonly used ‘osmoterminology’.

effect on the intracellular osmotic potential.5 One example of a putative yeast osmoprotectant molecule is trehalose, a dimeric form of D-glucose (-D-glucopyranosyl 6 D-glucopyranoside). Trehalose, originally identified from the fungal pathogen Claviceps purpurea, has been observed in many desiccation-tolerant organisms and so-called resurrection plants.7 It has also been demonstrated to play a role in thermotolerance and a range of other stresses.6 During intense or prolonged exposure to osmotic ‘upshock’, an alternative strategy may be employed in which the cell induces the synthesis of molecules that moderate intracellular osmotic potential. Such molecules, termed osmotica (singular osmoticum) may be accumulated to prevent excessive loss of cellular water. Osmotica comprise several groups of molecules, including compatible solutes. These solutes are sufficiently able to bind water molecules8 and thus either promote retrieval of water from the environment or prevent any further loss of intracellular water. Compatible solutes can be accumulated by the cell in large quantities, without any detrimental impact on cellular functioning, including protein denaturation, membrane degradation or enzyme activity.9 It has therefore been suggested that compatible solutes represent a good biomarker of osmotic stress. The molecular basis of the osmotic stress response has been extensively investigated in haploid strains of Saccharomyces cerevisiae in recent years. Few reports, however, concern the genetically intractable brewing production strains. The elucidation of the high osmolarity glycerol (HOG) synthesis pathway has shown that there is a number of functional genes involved in the production, dissimulation, uptake and export of this solute.10–13 The HOG pathway is complex, involving the stimulated expression of over 100 genes,13 and recent work suggests that this biosynthetic stress response pathway may be induced by other stresses as well as hyperosmotic stress.11 The changes that occur during osmotic stress at the physiological level are relatively unexplored in brewing yeast strains. Vacuoles are implicitly linked with osmotic flux within the yeast cell, and the morphological variations in this organelle have not been typified in brewing yeast strains. New techniques, including confocal laser scanning

48

BREWING YEAST FERMENTATION PERFORMANCE

microscopy, and newly developed fluorochromes specific for the vacuolar lumen and tonoplast, allow visualisation of changes to this organelle during osmotic stress.

5.2 Materials and methods 5.2.1

Yeast strains

Four production strains of lager brewing yeast (designated SCB1–4) and three production strains of ale brewing yeast (designated SCB5–8) were obtained from Scottish Courage Brewing Ltd (Edinburgh, UK). 5.2.2

Media and growth conditions

Strains were stored on beads at 80°C in YPD containing 20% (w/v) glycerol. Stock cultures were maintained and grown on YPD medium (1% w/v yeast extract, 2% w/v bacteriological peptone and 2% D-glucose solidified with 1.2% w/v agar). Yeasts were grown aerobically in 100 ml of YPD at 25°C in 250 ml Erlenmeyer flasks with constant agitation at 120 rpm in YPD medium. Cell growth was monitored using a Shimadzu spectrophotometer at 600 nm. 5.2.3

Osmotic challenge

Yeast cells were harvested at the required phase of growth and washed twice in phosphate buffer, pH 5.9 (89.85 ml 0.1 M NaH2PO4:10.15 ml 0.1 M Na2HPO4 diluted to 200 ml). Cells were resuspended in 100 ml of either sterile deionised water, sorbitol (6, 12, 18, 24 and 30% w/v) or NaCl (6, 12, 18, 24 and 30% w/v) to 1  106 cells/ml, and incubated on an orbital shaker at 25°C and 120 rpm for 48 h. 5.2.4

Viability determinations

Aliquots of cell suspensions (each 1 ml) were washed and resuspended in singlestrength phosphate-buffered saline (PBS) (NaCl 7.650 g/dm3, Na2HPO4 0.724 g/dm3, KH2PO4 0.210 g/dm3, adjusted to pH 7.4 with 1 M NaOH). The bright-field dye citrate methylene violet14 and the fluorescent dyes hemi-MgANS15 and propidium iodide16 were used to distinguish between viable and non-viable cells. One-hundred cells were enumerated for each replicate. Three replicates per sample were analysed. 5.2.5

Glycerol determination

Triplicate cell suspension (1  109 cells total number) were washed twice in phosphate buffer, pH 5.9, and resuspended in 1 ml boiling Tris–HCl, pH 7.0 [50 ml 0.5 M Tris(hydroxymethyl) aminomethane:46.8 ml 0.5 M HCl diluted to a total of 200 ml], and boiled for 10 min. The lysed cells were centrifuged at 4000 rpm for 10 min to remove cellular debris, and the supernatant was assayed for glycerol enzymically using Boehringer-Mannheim kit no. E0148 270, according to the method of Hounsa et al.17

49

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

Table 5.1

Vacuole lumen staining procedures used in this study

Dye C-DCFDA FUN1™ LysoTracker™ Green

Buffer

Incubation time

Cell concentration

Dye final concentration

50 mM sodium citrate, pH 5, 2% glucose 10 mM sodium HEPES, pH 7.2, 2% glucose Single-strength PBS

10 min (RT)

1  106/ml

10 mM

20 min (RT, in dark) 10 min (RT)

1  106/ml

15 M

1  106/ml

100 nM

HEPES: N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid; PBS: phosphate-buffered saline; RT: room temperature.

5.2.6

Preparation of cells for confocal microscopic analysis

5.2.6.1 Staining of vacuole lumen. Cells were washed twice in phosphate buffer, pH 5.9, and stained using the procedures outlined in Table 5.1. 5.2.6.2 Staining of tonoplast. Cells were washed twice in phosphate buffer, pH 5.9, and resuspended in 10 mM HEPES buffer, pH 7.4 (2-hydroxyethylpiperazine-N-2ethanesulfonic acid), 5% glucose to 1  106 cells/ml. The tonoplast marker MDY-64 (Molecular Probes, The Netherlands) was added to a final concentration of 10 mM and incubated at room temperature for 10 min before analysis using fluorescence microscopy. 5.2.6.3 Staining of plasma membrane. Cells were washed twice in phosphate buffer, pH 5.9, and resuspended in 10 mM HEPES buffer, pH 7.4, 5% glucose to 1  106 cells/ml. The membrane marker Cell-Tracker™ blue CMAC (Molecular Probes, The Netherlands) was added to a final concentration of 10 mM and incubated at room temperature for 10 min before analysis using fluorescence microscopy. 5.2.6.4 Visualisation of samples. Cells were immobilised on microscope slides in 2% low gelling temperature (LGT) agarose and examined using a Zeiss® 510 confocal laser scanning microscope. Fluorochromes were excited using three laser lines, Argon at 488 nm and two helium/neon lines at 543 and 633 nm, respectively. Detection of fluorescence was achieved using a 63 plan neofluor objective and the appropriate filter sets. 5.3 5.3.1

Results and discussion Osmotic stress tolerance of YPD-grown cells

5.3.1.1 Physiological state. The response of haploid S. cerevisiae strains to stress has been demonstrated to be dependent on the physiological state of the cells,17 with stationary-phase populations showing a greater resistance to a number of stresses than exponential-phase cells.18,19 For both lager and ale strains, cells grown to stationary phase exhibited a greater resistance to osmotic stress than those at exponential phase,

50

BREWING YEAST FERMENTATION PERFORMANCE

120

Percentage viability

100 80 Methylene violet MgANS

60

Propidium iodide 40 20 0 0

5

(a)

10

15

20

25

30

35

Sorbitol concentration (% w/v) 120

Percentage viability

100

80 Methylene violet MgANS

60

Propidium iodide 40

20

0 0

(b)

5

10

15

20

25

30

35

Sorbitol concentration (% w/v)

Fig. 5.2 Viability profiles during osmotic stress for (a) stationary phase and (b) exponential phase lager yeast cells (SCB1). Viability values represent the mean of triplicate samples. Error bars represent one standard deviation of a normal Poisson distribution.

as demonstrated using citrate methylene violet, Hemi-magnesium 1-anilinonaphthalene-8-sulfonic acid (Hemi MgANS) and propidium iodide (PI) (Fig. 5.2). This increased osmotolerance has been reported to be due in part to the induction of the global STRE (stress-responsive elements)-activated stress response at the onset of the stationary phase.20–22 In addition, Schuller et al.23 demonstrated that induction of the HOG pathway is closely associated with STRE elements, and HOG/STRE interactions regulate stress-induced transcriptional activity.

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

51

Fig. 5.3 Viability profiles during osmotic stress for (a) 12% sorbitol stressed and (b) 30% sorbitol stressed ale (SCB5–7) and lager (SCB1–4) yeast cells in stationary phase of growth. Values represent the mean of triplicate samples of all viabilities. Error bars represent one standard deviation of a normal Poisson distribution.

5.3.1.2 Strain dependence. Many physiological characteristics of brewing yeast are strain dependent. Blomberg24 demonstrated that haploid S. cerevisiae strains exhibit differences in osmotolerance, related directly to strain and also to physiological state. Osmotic stress tolerance in ale and lager brewing yeast strains was also observed to be highly strain dependent (Fig. 5.3). However, ale and lager yeasts did not appear to exhibit greatly differing responses, in contrast to their apparent responses to oxidative25,26 and starvation27 stresses. These findings demonstrate that the selection of strains for high-gravity and very high-gravity brewing should include characterisation of their osmotic tolerance. 5.3.1.3 Solute considerations. It has been previously demonstrated that laboratory haploid S. cerevisiae strains exhibit a greater tolerance to sorbitol stress than NaCl stress,17 although osmotic stress resistance to either solute was a function of the composition of the medium on which the cells were grown. The relationship between osmotic stress tolerance and solute used to induce this stress has not been previously

52

BREWING YEAST FERMENTATION PERFORMANCE

100 90

Methylene violet MgANS Propidium iodide

Percentage viability

80 70 60 50 40 30 20 10 0 0

5

10

(a)

15

20

25

30

Sodium chloride (%w/v) 100 90

Percentage viability

80 70 60 Methylene violet

50

MgANS

40

Propidium iodide

30 20 10 0 0

(b)

5

10

15

20

25

30

Sorbitol (%w/v)

Fig. 5.4 Viability profiles during osmotic stress for (a) salt-stressed and (b) sorbitol-stressed stationary phase lager yeast cells (SCB1). Viability values represent the mean of triplicate samples. Error bars represent one standard deviation of a normal Poisson distribution.

reported for polyploid or brewing yeast strains. The type of solute used to elicit osmotic stress challenge in ale and lager strains greatly affects the tolerance to osmotic stress (Fig. 5.4), indicating a modified response as a function of the solute used. Lager yeast strains were particularly intolerant following exposure to high salt concentration, as opposed to high sorbitol concentrations. In part, this may be explained by a reduced tolerance to salt toxicity in Saccharomyces pastorianus owing to the accumulation of cellular damage as a result of dissociated Na and possibly Cl ions.28,29 The impact of salt toxicity on brewing strains has not been reported previously and merits further investigation. Since solute type influences the degree of tolerance exhibited by brewing strains to osmotic stress, the batch-to-batch variation in wort composition may

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

53

moderate the resistance of pitched yeast to osmotic stress both as a function of gravity and as a concentration of relative solutes. This hypothesis has not been previously suggested and merits further investigation. 5.3.2

Compatible solute accumulation

It has been demonstrated that yeasts appear preferentially to accumulate sugar alcohols during exposure to hyperosmotic stress. In S. cerevisiae glycerol is accumulated; however, other polyols such as mannitol have also been suggested as alternative compatible solutes. Despite this theory other solutes have not been screened for, although their presence has been postulated from nuclear magnetic resonance (NMR) analysis spectra.17 5.3.2.1 Physiological state. The effect of physiological state on glycerol accumulation levels in haploid strains of S. cerevisiae has been reported,30 but the response of brewing yeast strains has not been previously characterised. As with osmotic stress tolerance, a concomitant alteration in glycerol accumulation patterns with physiological state is observed. Cells in exponential phase of growth exhibit a regular pattern of accumulation (Fig. 5.5a). In stationary-phase populations, fluctuations in the levels of intracellular glycerol were also observed, although there was no apparent correlation with sorbitol concentration, and unlike for exponential-phase populations, there was no regular accumulation pattern (Fig. 5.5b). The reason for these growth-dependent but solute concentration-independent fluctuations in glycerol levels remains unclear. One hypothesis is that in both growth phases glycerol is serving as a precursor for another compatible solute. Another is that glycerol is merely a cell storage molecule or osmoprotectant and not a true compatible solute. These hypotheses remain the subject of further investigation. 5.3.2.2 Strain dependence and glycerol accumulation. Glycerol accumulation levels have been examined extensively in haploid yeast strains,18 indicating the importance of glycerol in the survival of these strains during osmotic stress. The level of glycerol accumulation in marine yeast under osmotic stress is also a species- and possibly strain-dependent phenomenon31 (Table 5.2). Indeed, it has also been observed that accumulated intracellular glycerol levels were highly dependent on strain in brewing yeast (Fig. 5.6), but not relative to resistance as a function of cell viability. This information further supports the hypothesis that glycerol is not acting as a true compatible solute, although, as shown previously, it is implicit in the osmotic stress response of ale and lager strains. 5.3.2.3 Solute considerations of glycerol accumulation. In brewing yeast strains it has been previously hypothesised that salt-stressed S. cerevisiae cells may produce higher levels of glycerol than sorbitol- or sugar-stressed cells.31 While this is possibly true of haploid laboratory strains, it was observed that lager strains showed a lower level of glycerol accumulation when stressed with NaCl compared with sorbitol-stressed cells (Fig. 5.7). The reasons for the observed differences in glycerol accumulation with solute type are not known and require elucidation. However, it is possible that the HOG pathway, as a complex signal transduction pathway, is not universally activated

54

BREWING YEAST FERMENTATION PERFORMANCE

0.10

0.08

SCB6

0.07 0.06 0.05 0.04 0.03

Glycerol (g/1  109 cells)

0.09

SCB4

0.02 0.01 0.00 0

5

10

(a)

15

20

25

30

Sorbitol (%w/v)

0.10

SCB6

0.08 0.07 0.06 0.05 0.04 0.03

Glycerol (g/1 1 09 cells)

0.09

SCB4

0.02 0.01 0.00 0

(b)

5

10

15

20

25

30

Sorbitol (%w/v)

Fig. 5.5 Glycerol accumulation levels for (a) exponential-phase and (b) stationary-phase ale (SCB6) and lager (SCB4) yeast exposed to a range of sorbitol concentrations. Values represent the mean of triplicate samples; error bars represent one standard deviation of a normal Poisson distribution from the mean value of triplicates.

for osmotic stress, but may be in part regulated by the nature of the solutes used. This is particularly evident when the osmolyte NaCl is present in low concentrations. The genetic regulation of the HOG signal transduction pathway has been extensively studied and is established in laboratory S. cerevisiae strains. However, the regulation of the

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

55

Table 5.2 Comparison of the percentage total internal osmolarity accounted for by glycerol for five marine strains of yeasta Species Aureobasidium pullulans

Candida sp.

Cryptococcus albidus var. albidus

Debaromyces hanseii (C11)

Rhodotorula rubra

NaCl concentration (M) 0 1.02 2.05 4.11 0 0.51 1.54 0 0.68 1.37 0 0.86 1.71 2.4 0 0.68 1.37 2.05

Intracellular glycerol (% of intracellular osmolarity) 68.5 80.3 84.8 90.9 12.8 60.8 80.4 15.1 73.0 92.3 61.4 83.3 86.9 88.9 10.7 63.7 85.8 95.6

a

Hernandez-Saavedra et al. (1995).

Fig. 5.6 Glycerol accumulation levels for four lager (SCB1–4) and three ale (SCB5–7) strains osmotically challenged with (a) 12% (w/v) and (b) 30% (w/v) sorbitol. Values represent the mean of triplicate samples. Error bars represent one standard deviation of a normal Poisson distribution.

56

BREWING YEAST FERMENTATION PERFORMANCE

0.070 0.060 0.050 0.040 0.030 0.020 0.010

Glycerol (g/1
109 cells)

0.080

0.000 0

5

10

(a)

15

20

25

30

Sorbitol (% w/v)

0.070 0.060 0.050 0.040 0.030 0.020 0.010

Glycerol (g/1
109 cells)

0.080

0.000 0

5

(b)

10

15

20

25

30

NaCl (%w/v)

Fig. 5.7 Glycerol accumulation levels for stationary phase lager (SCB1) cells stressed with (a) sorbitol and (b) NaCl. Values represent the mean of triplicate samples; error bars represent one standard deviation from mean value of triplicates.

pathway in ale polyploid strains of S. cerevisiae and S. pastorianus (lager strains) has not been previously reported. It has also been suggested that a specific calcineurin pathway activated by high salt concentrations exists in S. cerevisiae,32 which is not activated simply by osmotic stress. 5.3.3

Vacuolar changes

The yeast vacuole is a prominent and highly dynamic organelle.33 It has been suggested to play a central role in a number of diverse functions, from sequestration of cationic products and compartmentalisation of hydrolytic enzymes34–36 to protein sorting37,38 and cell division.39 The vacuole is implicitly associated with osmoregulation, and therefore cellular responses to osmotic stress. Changes in this organelle have not been previously shown during osmotic stress in brewing strains. It is thought that vacuolar changes may represent a good biomarker of osmotic stress.

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

57

5.3.3.1 Vacuolar morphology of YPD-grown cells. Previous studies have suggested that the yeast vacuole is homologous to the plant vacuole, in that it occupies a large proportion of the intracellular space,33 but is also analogous with mammalian lysosomes,36 in that it exhibits a low pH and high content of hydrolases. The yeast vacuole has been seen to be more analogous with the mammalian lysosome,36 owing to its low pH and high content of hydrolases. These findings were corroborated using commercially available dyes (Lyso-Tracker™) that become fluorescent dependent on lysosomal conditions and can be clearly shown to localise within yeast vacuoles (Plates 5.1 and 5.2). Indeed, many of the lumen dyes used in this study become fluorescent owing to either the effects of low pH or hydrolase activity. The changes in this organelle in yeast were thought to be minimal in non-stressed populations, although Schwenke suggested in 1977 that yeast vacuolar fragmentation occurred as part of the normal replication cycle.34 Vacuoles of ale and lager yeast strains demonstrate two distinct forms (Plates 5.3 and 5.4), with some cells exhibiting prominent large central vacuoles (Plate 5.3) and others highly fragmented or pro-vacuoles (Plate 5.4). The occurrence of fragmented vacuoles far exceeds that of entire vacuoles even in stationary-phase populations not exposed osmotic stress (Plates 5.5 and 5.6), possibly as a result of normal changes associated with cellular replication. 5.3.3.2 Vacuolar morphology of exponential-phase cells. In line with the work of Schwenke, it would be expected that exponential-phase populations should show a higher degree of vacuolar fragmentation owing to the high proportion of cells at various phases in the replication cycle.34 In this model, budding cells initially contain a large number of small vacuoles, which expand and fuse towards the end of the budding process, giving rise to one or two large vacuoles; just before the emergence of a new bud, vacuoles are observed to fragment again34 (Fig. 5.8 and Plate 5.7). It was observed that more cells of exponential-phase populations of ale and lager strains exhibited vacuolar fragmentation than in stationary-phase populations. The actual degree of fragmentation in individual cells, however, remained approximately equal. To attempt to corroborate the vacuolar fragmentation and cell replication cycle of Schwenke, a sample of previously undivided cells, termed virgin cells, was analysed. These cells were shown frequently to contain a large central vacuole (Plates 5.8 and 5.9), although in many cases some fragmentation was observed (Plate 5.8), and in a proportion of cells entire fragmentation was observed. It is thought that this fragmentation of virgin cell vacuoles represents various stages of ‘readiness’ to divide. Schwenke34 clearly shows that vacuolar fragmentation is required for the division to occur; virgin cells with fragmented vacuoles may simply be in a later stage of the cell cycle, and thus about to divide. The exact mechanism for vacuolar fragmentation is undetermined; however, it has been reported that yeast vacuoles fragment when microtubules are disrupted40 with nocodazole, implying that microtubules are important in the maintenance of vacuolar integrity. 5.3.3.3 Vacuolar fragmentation and osmotic stress. Vacuolar fragmentation and osmotic stress have not been previously correlated; however, fragmentation has been previously associated with oxidative stress (in terms of pro-vacuole formation). Osmotically stressed cells of both exponential- and stationary-phase populations

58

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 5.8 Diagrammatic representation of vacuolar changes during the normal replication cycle. (Adapted from Schwenke, 1977.)

demonstrated a high degree of vacuolar fragmentation (Plate 5.10). It was, however, impossible to differentiate between fragmentation due to osmotic stress and that due to the progression through the cell cycle.

5.4

Conclusions

The viability and vitality of brewing yeast cultures directly affect fermentation performance and final beer quality.1 As yeast is subject to a range of stress factors during brewery handling, it is important to identify the impact of these stresses on yeast quality and thus capacity to perform. Elucidation of the physiological responses to stress will therefore provide information concerning possible biomarkers of exposure to stressful conditions. Identification of a suitable biomarker of osmotic stress will allow informed decisions to be made concerning high-gravity brewing and pitching rates. Brewing yeast osmotolerance was shown to be highly strain dependent, as were the levels of intracellular glycerol accumulated during osmotic stress. Glycerol did not represent a good biomarker of osmotic stress as its intracellular abundance was seen to be dependent on strain, physiological state and the solute used to elicit the osmotic stress response. Vacuolar fragmentation could not be correlated to osmotic stress. Indeed, it was dependent on a number of physiological and environmental conditions, including normal growth and division, and therefore cannot be used as a biomarker of osmotic stress. The exact cytological processes governing fragmentation were not determined, although these processes, along with cellular solute accumulation, remain the subject of further research.

THE OSMOTIC STRESS RESPONSE OF ALE AND LAGER BREWING

59

Acknowledgements Philip White is funded by the J & J Morison educational fund, and the authors would like to thank Mrs Pamela Morison-Inches for her support. The authors are indebted to the directors of Scottish Courage Brewing Ltd for the kind permission to publish this work. Thanks also go to Miss Alexandra Patmanidi for constant help and support and Mrs Dawn Maskell for her help in the preparation and isolation of virgin cells. Katherine Smart is the Scottish Courage Reader in Brewing Science and gratefully acknowledges the support provided by the directors of Scottish Courage Brewing Ltd. Katherine Smart is also a Royal Society Industrial Fellow and gratefully acknowledges the support provided by the Royal Society and the BBSRC.

References 1. Smart, K. (2000) The death of the yeast cell. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 105–113. 2. Panchal, C.J. and Stewart, G.G. (1980) The effect of osmotic pressure on the production and excretion of ethanol and glycerol in brewing yeast strains. J. Inst. Brew. 86, 207–210. 3. Hammond, J., Davis, D., Lee, M. and Storey, K. (2001) Does osmotic pressure affect yeast performance in high gravity fermentation? Proc. Eur. Brew. Conv., Budapest. 4. Nagodawithana, T.W., Castellano, C. and Steinkraus, K.H. (1974) Appl. Microbiol. 28, 383. 5. Reed, H.R. (1984) Use and abuse of osmo-terminology. Plant Cell Environ. 7, 165–170. 6. Wiemken, A. (1990) Trehalose in yeast: stress protectant rather than reserve carbohydrate. Antonie van Leeuwenhoek 58, 209–217. 7. Singer, M.A. and Lindquist, S. (1998) Thermotolerance in Saccharomyces cerevisiae: the Yin and Yang of trehalose. Trends Biotechnol. 16, 460–468. 8. Galinski, E.A., Stein, M., Amendt, B. and Kinder, M. (1997) The kosmotropic (structure-forming) effect of compensatory solutes. Comp. Biochem. Physiol. 117A, 357–365. 9. Gilbert, G.A., Gadush, M.V., Wilson, C. and Madore, M.A. (1998) Amino acid accumulation in sink and source tissues of Coleus blumei Benth. during salinity stress. J. Exp. Bot. 49, 107–114. 10. Raitt, D.C., Posas, F. and Saito, H. (2000) Yeast CDc42 GTPase and Ste20 PAK-like kinase regulate Sho1-dependent activation of the hog1 MAPK pathway. EMBO J. 19, 4623–4631. 11. Garay-Arroyo, A. and Covarrubias, A.A. (1999) Three genes whose expression is induced by stress in Saccharomyces cerevisiae. Yeast 15, 879–892. 12. Kapteyn, J.C., ter Riet, B., Vink, E. et al. (2001) Low external pH induces HOG1-dependent changes in the organization of the Saccharomyces cerevisiae cell wall. Mol. Microbiol. 39, 469–479. 13. Tamás, M.J., Luyten, K., Sutherland, F.C.W. et al. (2000) Fpsp1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087–1104. 14. Smart, K.A., Chambers, K.M., Lambert, I. and Jenkins, C. (1999) Use of methylene violet staining procedures to determine yeast viability and vitality. J. Am. Soc. Brew. Chem. 57, 18–23. 15. McCaig, R. (1990) Evaluation of the fluorescent dye 1-anilino-8-naphthalene sulfonic acid for yeast viability determination. J. Am. Soc. Brew. Chem. 48, 22–25. 16. Deere, D., Shen, J., Vesey, G., Bell, P., Bissinger, P. and Veal, D. (1998) Flow cytometry and cell sorting for yeast viability assessment and cell selection. Yeast 14, 147–160. 17. Hounsa, C.G., Brandt, E.V., Thevelein, J. et al. (1998) Role of trehalose in survival of Saccharomyces cerevisiae under osmotic stress. Microbiology 144, 671–680. 18. Hohmann, S. (1997) Shaping up: the response of yeast to osmotic stress. In: Yeast Stress Responses, Hohmann, S. and Mager, W.H. (eds), R.G. Landes, Austin, TX, pp. 101–146. 19. Jamieson, D.J. (1998) Oxidative stress responses of the yeast Saccharomyces cerevisiae. Yeast 14, 1511–1572. 20. Siderius, M. and Mager, W.H. (1997) General stress response: in search of a common denominator. In: Yeast Stress Responses, Hohmann, S. and Mager, W.H. (eds). R.G. Landes, Austin, TX, pp. 213–230. 21. Mager, W.H. and Moradas-Ferreira, P.M. (1993) Stress response of yeast. Biochem. J. 290, 1–13. 22. Smart, K.A. (2001) The management of yeast stress. Proc. Eur. Brew. Conv., Budapest.

60

BREWING YEAST FERMENTATION PERFORMANCE

23. Schüller, C., Brewster, J.L., Alexander, M.R. et al. (1994) The HOG pathway controls osmotic regulation of transcription via the stress-response element (STRE) of the Saccharomyces cerevisiae CTT1 gene. EMBO J. 13, 4382–4389. 24. Blomberg, A. (1997) The osmotic hypersensitivity of the yeast Saccharomyces cerevisiae is strain and growth media dependent: quantitative aspects of the phenomenon. Yeast 13, 529–539. 25. Martin, V., Quain, D.E. and Smart, K.A. (1999) The oxidative stress response of ale and lager yeast strains. Eur. Brew. Conv. Cong. 27, 679–686. 26. Martin. V., Quain, D.E. and Smart, K.A. (2000) The oxidative stress response of ale and lager yeast strains. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 97–105. 27. Rhymes, M.R. and Smart, K.A. (1996) Effect of starvation on the surface properties of brewing yeast. Proc. 2nd Conv. Ferm. Physiol. Inst. Chem. Eng., Brighton, pp. 34–36. 28. Serrano, R. (1996) Salt tolerance in plants and microorganisms: toxicity targets and defence responses. Int. Rev. Cytol. 165, 1–52. 29. Serrano, R., Márquez, J.A. and Ríos, G. (1997) Crucial factors in salt stress tolerance. In: Yeast Stress Responses, Hohmann, S. and Mager, W.H. (eds). R.G. Landes, Austin, TX, pp. 147–170. 30. Blomberg, A., Larsson, C. and Gustafsson, L. (1988) Microcalorimetric monitoring of growth of Saccharomyces cerevisiae: osmotolerance in relation to physiological state. J. Bacteriol. 170, 4562–4568. 31. Hernandez-Saavedra, N.Y., Ochoa, J.L. and Vazquez-Dulhalt, R. (1995) Osmotic adjustment in marine yeast. J. Plankton Res. 17, 59–69. 32. Eriksson, P., Alipour, H., Adler, L. and Blomberg, A. (2000) Rap1p-binding sites in the Saccharomyces cerevisiae GPD1 promoter are involved in its response to NaCl. J. Biol. Chem. 275, 29368–29376. 33. Klionsky, D.J. (1998) Nonclassical protein sorting to the yeast vacuole. J. Biol. Chem. 273, 10807–10810. 34. Schwenke, J. (1977) Characteristics and integration of the yeast vacuole and cellular functions. Phys. Vég. 15, 491–517. 35. Nishikawa, S., Umemoto, N., Oshumi, Y. et al. (1990) Biogenesis of vacuolar glycoproteins of yeast Saccharomyces cerevisiae. J. Biol. Chem. 265, 7440–7448. 36. Spormann, D.O., Heim, J. and Wolf, D. (1992) Biogenesis of the yeast vacuole (Lysosome). J. Biol. Chem. 267, 8021–8029. 37. Banta, L.M., Robinson, J.S., Klionsky, D. and Emr, S.D. (1988) Organelle assembly in yeast: characterisation of yeast mutants defective in vacuolar biogenesis and protein sorting. J. Cell Biol. 107, 1369–1383. 38. Chiang, H.L. (1995) Protein targeting and degradation in the yeast vacuole. Can. J. Bot. 73 (Suppl. 1), S347–S351. 39. Schwenke, J. (1991) Vacuoles, internal membranous systems and vesicles. In: Rose, A. And Harrison, J.S. eds. The Yeasts, Vol. 4, 2nd edn. Academic Press, New York. 40. Guthrie, B.A. and Wickner, W. (1988) Yeast vacuole fragment when microtubules are disrupted. J. Cell Biol. 107, 115–120.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

6 Brewing Yeast Oxidative Stress Responses: Impact of Brewery Handling V. MARTIN, D.E. QUAIN and K.A. SMART

Abstract Oxidative stress may be defined as the response to cellular damage generated by reactive oxygen species such as superoxide anions and hydrogen peroxide. These compounds are generated as by-products during yeast aerobic metabolism and may damage cellular macromolecules such as lipids, proteins and DNA. Cellular damage incurred as a consequence of exposure to reactive oxygen species may result in lipid peroxidation, protein carbonyl formation and DNA base modifications. Primary defences are provided by enzymes such as superoxide dismutases (SOD1 and SOD2) and catalases (CTT1 and CTA1), although other non-enzymic antioxidants (e.g. glutathione) may also provide protection. During the brewing process, exposure to oxidative stress may occur as a consequence of yeast propagation, storage and fermentation conditions. It is suggested that exposure to oxidants may influence yeast physiological condition and therefore subsequent fermentation performance and beer quality. Production ale and lager strains were grown in YPD (yeast extract, bacteriological peptone and D-glucose) or wort to achieve populations exhibiting exponential or stationary phase and were exposed to oxidants such as menadione (a generator of superoxide anions) and hydrogen peroxide to establish the influence of strain type and media composition on resistance. The relationship between oxidative stress resistance and extent of primary antioxidant defence was investigated for these strains following growth on YPD, wort and defined wort substitute. Resistance to oxidants, oxidative stress defence mechanisms and cellular damage were observed to be strain dependent and affected by media composition and growth phase. The relationship between oxidative stress defence and damage levels is currently under investigation.

6.1 Introduction Oxidative stress is associated with cells responding to and protecting themselves from reactive oxygen species.1 Reactive oxygen species [ROS; superoxide anions (O2·), hydrogen peroxide (H2O2) and hydroxy (OH) radicals] may result in damage to cellular macromolecules such as lipids, proteins and DNA,2,3 and are mainly generated during yeast aerobic metabolism.3,4 Although brewing fermentations are essentially anaerobic, yeast cells are exposed to oxygen during propagation, at pitching5 and during storage if preaerated,6 representing a potential source of oxidative stress. Cell defences against ROS are provided by enzymes (superoxide dismutases and catalases) and other non-enzymic antioxidants (glutathione, metal ions, vitamins C and E).1,3,4 It is suggested that oxidative stress may affect yeast quality and subsequently fermentation performance. The tolerance of production strains of ale and lager yeast to exogenously generated oxidative stress in the form of hydrogen peroxide has been investigated. The relationship between the resistance to this oxidant and the cellular antioxidant defence (total

62

BREWING YEAST FERMENTATION PERFORMANCE

catalase activity and cellular glutathione concentration) has been examined for ale and lager strains grown in YPD and wort. The existence of oxidative stress during yeast handling and its effect on fermentation performance have not been extensively studied. The aim of this work was to investigate the occurrence of oxidative stress and cellular response during yeast handling.

6.2 Materials and methods 6.2.1

Yeast strains and growth conditions

Three lager (BB10, BB11 and BB28) and two ale (BB3 and BB18) production strains were obtained from Bass Brewers (Burton upon Trent, UK). Yeast cells were grown aerobically in 250 ml flasks containing 50 ml of YPD (1% w/v yeast extract, 2% w/v bacteriological peptone and 2% w/v D-glucose) at 25°C on an orbital shaker at 120 rpm. Yeast populations were harvested in the mid-exponential or stationary phase. 6.2.2

Yeast sample collection

Samples were collected in the brewery (Bass Brewers), from propagation vessels, from fermenter vessels at the end of fermentation and from storage vessels, during storage and before and after acid washing. 6.2.3

Determination of response to oxidative stress

Yeast cells were washed three times in sterile deionised water and resuspended in either water, H2O2 (0.01, 0.1 and 1% w/v) or menadione (10, 50, 100, 200 mM) to a final cell concentration of 1  105 cells/ml and incubated at 25°C on an orbital shaker at 120 rpm. Cell viability was assessed using plate counts on YPD and expressed as a percentage of viability. 6.2.4

Glutathione concentration

Yeast cells were harvested by centrifugation and washed three times in cold (4°C) sterile deionised water. Cells were resuspended in 500 l metaphosphoric acid 5% (w/v). Cell suspensions were repeatedly boiled and immersed in liquid nitrogen (six times) to achieve cell lysis. The supernatant was retained for analysis. Glutathione concentration was determined using the assay kit from Calbiochem (La Jolla, CA, USA) and expressed as nmol/cell. 6.2.5 Protein extraction for enzymic assays by glass bead cell lysis method Yeast cells were harvested by centrifugation and washed three times in cold (4°C) sterile deionised water. Cells were resuspended in 100 l fresh buffer [Tris pH 7.5 50 mM, ethylenediaminetetra-acetate (EDTA) 0.5 mM, Triton X100 0.1% (v/v) and the following protease inhibitors suspended in dimethyl sulfoxide (DMSO) 100%,

BREWING YEAST OXIDATIVE STRESS RESPONSES

63

tosylphenylalanine chloromethyl ketone (TPCK) 0.25 mM, tosyllysine chloromethyl ketone (TLCK) 0.25 mM, or in absolute methanol, phenylmethylsulfonyl fluoride (PMSF) 0.5 mM]. To this suspension 100 l of acid washed glass beads (0.5 mm) was added. The samples were vortexed for 3 min and centrifuged at 2100 g for 10 min, and the resulting supernatant retained for analysis. The total protein concentration from each extract was determined using a BioRad protein assay kit (BioRad Laboratories, München, Germany). 6.2.6

Catalase activity

Protein extracts were obtained following the glass bead lysis method. Cellular catalase activity was determined according to the methods of Aebi7 and Izawa et al.8 The catalase activity was expressed as units per gram of protein. 6.2.7

Glycogen and trehalose concentration

Glycogen and trehalose concentrations were determined using the method by Parrou et al.9 The glycogen and trehalose concentrations were expressed as mg equivalent glucose per 108 cells.

6.3 6.3.1

Results and discussion Oxidative stress resistance is dependent on growth phase, strain and medium

For haploid strains of Saccharomyces cerevisiae the response of YPD-grown exponential- and stationary-phase cells to H2O2 has been established,4 with stationaryphase cells exhibiting greater resistance than exponential-phase cells. YPD-grown brewing yeast cells also exhibited an increased resistance for stationaryphase populations and a reduced resistance for exponential-phase cells following exposure to exogenous H2O210 (Fig. 6.1). Viability for all strains decreased with the duration of exposure to H2O2 and the rate of cell death was dependent on the concentration of oxidant. For brewing yeast grown in YPD, H2O2 tolerance was reported to be strain dependent, with the ale strains being more sensitive than the lager strains for both exponential- and stationary-phase populations10 (Fig. 6.2). The tolerance to H2O2 was reported to be medium dependent. Indeed, brewing yeast stationary-phase cells grown in wort exhibited lower tolerance than cells grown in YPD, with the exception of strain BB11 (Fig. 6.3). 6.3.2

Defence mechanisms against hydrogen peroxide are dependent on strain and medium

The primary defences against hydrogen peroxide are provided by catalases (CTT1 and CTA1); however, other non-enzymic antioxidants such as glutathione may also protect the cell.1,3,4

64

BREWING YEAST FERMENTATION PERFORMANCE

Viability (%)

100 80 60 40 20

BB11

0 0.01

0.1

1

Hydrogen peroxide concentration (mM)

(a)

Viability (%)

100 80 60 40 20

BB11

0 0.01

(b)

0.1

1

Hydrogen peroxide concentration (mM)

Fig. 6.1 Effect of oxidative stress on the viability of YPD-grown lager yeast populations (strain BB11) following exposure to hydrogen peroxide (0.01, 0.1 and 1 mM) for 1 h. Viability values represent the mean of three replicate samples. (a) Exponential phase; (b) stationary phase.

Viability (%)

100 80 60 40 20

BB11

0 0.01 (a)

0.1

1

Hydrogen peroxide concentration (mM)

Viability (%)

100 80 60 40 20

BB18 (Ale)

0 0.01 (b)

0.1

1

Hydrogen peroxide concentration (mM)

Fig. 6.2 Effect of oxidative stress on the viability of yeast populations grown in YPD to stationary phase following exposure to hydrogen peroxide (0.01, 0.1 and 1 mM) for 1 h. Viability values represent the mean of three replicate samples. (a) Lager strain BB11; (b) ale strain BB18.

65

BREWING YEAST OXIDATIVE STRESS RESPONSES

Viability (%)

100 80 60 40 20

BB11

0 0.01 (a)

0.1

1

Hydrogen peroxide concentration (mM)

Viability (%)

100 80 60 40 20

BB11

0 0.01 (b)

0.1

1

Hydrogen peroxide concentration (mM)

Fig. 6.3 Effect of oxidative stress on the viability of stationary-phase lager yeast populations (strain BB11) following exposure to hydrogen peroxide (0.01, 0.1 and 1 mM) for 1 h. Viability values represent the mean of three replicate samples. (a) YPD-grown cells; (b) wort-grown cells.

Saccharomyces cerevisiae cells contain two catalase genes, CTA1 and CTT1, which encode the peroxisomal catalase A and the cytosolic catalase T enzymes, respectively. Catalases are primarily expressed during the stationary phase of growth and catalyse the removal of H2O2 (by hydrolysis to form water and oxygen) during this phase.11 Both genes are induced by oxygen; however, catalase A gene expression may be induced by certain fatty acids and growth on non-fermentable carbohydrate, and strongly repressed by glucose. The catalase T gene is negatively regulated by cyclic adenosine monophosphate (cAMP) and has been shown to be induced by stresses including starvation, osmotic and oxidative stress.12 Catalase activity is therefore strain, medium composition and stress dependent.1 Catalase activity for ale (BB1) and lager (BB11) brewing yeast strains has been previously examined following aerobic growth on semi-defined wort at 18 and 12°C, respectively.13 The total catalase activity of stationary-phase populations of the three lager and two ale strains of brewing yeast was reported to be strain dependent in both YPD and wort, although the influence of the growth medium on the levels of this enzymic oxidant defence did not appear to be universal or consistent.10 However, the catalase levels observed (Table 6.1) reflect the relative resistance of each strain to 1% (v/v) exogenous H2O2 (Figs 6.1 and 6.2), but not at the lower concentrations, where the reduction in viability was less pronounced. In addition to enzymic defences, S. cerevisiae cells produce a non-enzymic antioxidant, glutathione, which reduces H2O2 to water and oxygen.8 Glutathione is present

66

BREWING YEAST FERMENTATION PERFORMANCE

Table 6.1 Antioxidant defences for stationary-phase cells grown aerobically in YPD Strain BB3 (ale) BB10 BB11 BB18 (ale) BB28

Catalase (units/g protein)

Glutathione (nmol/cell number)

4.9  0.8 12.1  2.2 35.6  6.1 4.2  1.2 23.6  2.3

98.2  27.1 182.8  83.2 242.4  53.5 214.1  5.0 158.0  67.5

Data are expressed as the mean  SD and represent the mean of at least six replicates. Table 6.2 Antioxidant defences for stationary-phase BB11 cells grown aerobically in YPD or wort

Glutathione Catalase

YPD

Wort

78.6  40.5 27.0  5.3

242.4  53.5 35.6  6.1

Data are expressed as the mean  SD and represent the mean of at least six replicates.

in two forms in yeast cells, the reduced antioxidant form (GSH) and the oxidised form (GSSG), and is important for many biological processes.4 The oxidised form (GSSG) is recycled through a reaction catalysed by the enzyme glutathione reductase.14 The level of total cellular glutathione in YPD-grown stationary-phase populations of lager and ale brewing yeast strains has been previously reported; it was observed that the glutathione concentration exhibited in both YPD- and wort-grown cells was strain dependent.10 Indeed, YPD-grown ale strain cells exhibited the lowest levels of glutathione (Table 6.2). Typically, the levels of glutathione in the cell reflect the level of endogenously generated oxidative stress imposed on the cells or impaired glutathione reductase activity resulting in the necessity to generate higher levels of the antioxidant. 6.3.3

Cellular damage

As highly oxidant compounds, ROS can react with cellular components such as lipids, proteins and DNA. Lipids are highly susceptible to oxygen-derived species and once initiated lipid peroxidation proceeds as a self-perpetuating chain reaction.15 Endproducts of lipid peroxidation are highly reactive and can attack a wide range of biomolecules, such as amino acids and DNA.16 Lipid peroxidation may cause decreased membrane fluidity, inactivation of membrane receptors and enzymes, and nonspecific permeability to ions,17 and therefore results in further cellular damage.18 The membrane function of a haploid yeast mutant not capable of detoxifying superoxide anions (sod1-sod2) was examined.19 The glucose-induced proton efflux (GIPE) of this mutant was lower than the wild-type value, owing to the loss of membrane function (Fig. 6.4).

BREWING YEAST OXIDATIVE STRESS RESPONSES

67

0.8 0.7

GIPE

0.6 0.5 0.4 0.3 0.2 0.1 0 sod1-sod2

Wild-type

Fig. 6.4 Glucose-induced proton efflux (GIPE) for a laboratory wild-type strain and a mutant lacking superoxide dismutase (sod1-sod2). (From Van Zandycke et al.19)

Oxidative damage to proteins may result in peroxide and carbonyl production,15 alteration of molecular weight by protein aggregation, due to TyrˆTyr or ˆSˆSˆ cross-linkage reactions20 or fragmentation (peptide bond cleavage), altered net electric charge, amino acid modification and proteolytic susceptibility.21 ROS may damage DNA at either the sugar or the base.22 The damage caused is generally site specific because of the occurrence of binding sites for metals (bound or not to a protein) or the weakness of interhistone DNA.15 Mitochondrial DNA (mtDNA) is more susceptible to damage than nuclear DNA (nDNA). This can be explained by the fact that mtDNA is unprotected and is located near the site of ROS production, the electron transporter chain and the lipid peroxidation end-products generated in the inner membrane.15,23 If cellular mechanisms are not able to repair efficiently the degradation compounds produced24 (such as ring-opened bases or strand-breaks), DNA damage can lead to point mutation, deletion, insertion, intrachromosomal recombination and sister-chromatid exchange.25 6.3.4

Oxidative stress during the brewing process

Oxidative stress may occur when cells are in contact with oxygen. However, yeast cell growth and replication require the presence of oxygen to synthesise long-chain fatty acids and sterols, particularly during propagation and at the start of fermentation.26 Yeast cells may be in the presence of oxygen during storage, particularly if yeast oxygenation is practised. 6.3.5

Propagation

Propagation is necessary to allow the generation of enough pure yeast in good physiological condition, acclimatised to fermentation temperature and wort sugar composition to conduct fermentations.27 To achieve optimal growth, the synthesis of membrane fatty acids, sterols28 and high levels of reserve carbohydrates29 is necessary and this process is achieved by exposing the yeast to aerobic conditions.

68

BREWING YEAST FERMENTATION PERFORMANCE

Catalase

60

(a)

40 20 0 0

50

100

150

0

50

100

150

0

50

100

150

0

50

100

150

Glutathione

80

(b)

60 40 20 0

Glycogen

2000

Trehalose

(c)

1500 1000 500 0 300 200 100 0

(d)

Time (h)

Fig. 6.5 Antioxidant and reserve carbohydrate levels during propagation. Catalase (a) is expressed as units/g protein, glutathione (b) as nmol/108 cells, and glycogen (c) and trehalose (d) as g equivalent glucose/108 cells. The results represent the mean of three spectrophotometric measurements.

To investigate the occurrence of oxidative stress during propagation, samples were collected from the propagation vessel and assayed for their antioxidant and reserve carbohydrate levels. The levels of catalase, glycogen and trehalose were extremely low 72 h after inoculation. Both catalase activity and glycogen and trehalose concentrations exhibited a significant increase 100 h after inoculation into the propagation vessel (Fig. 6.5a,c,d), whereas glutathione levels remained stable throughout (Fig. 6.5b). Catalase, glycogen and trehalose levels are regulated by stress-responsive elements (STREs), whereas glutathione concentration is under the control of other promoters.30 It is therefore suggested that the antioxidant stress response exhibited during propagation may reflect exposure to oxidative and other stresses that induce STRE-activated responses. 6.3.6

Pitching

Exposure to aeration also occurs at pitching when the yeast cells are inoculated into aerated wort. It was suggested that preoxygenation of yeast slurries instead of wort

BREWING YEAST OXIDATIVE STRESS RESPONSES

69

aeration would decrease the initial lag phase of fermentation,31 since the yeast cells would synthesise sterols at the expense of intracellular glycogen,32 thereby eliminating the necessity to produce these macromolecules during fermentation.5 One disadvantage is that homogeneity of aeration of the yeast slurry is not readily achieved and, since excess oxygen is stressful to the cell, it can result in an excessive depletion of the cells’ energy stores and have adverse effects on fermentation performance,33 highlighted by the accumulation of the stress protectant trehalose.34 Nevertheless, pitching of oxygenated yeast into anaerobic wort increased consistency in fermentation duration, flavour profiles and total yeast growth;35,36 dissimilation of glycogen and synthesis of sterols were observed; however, the trehalose content remained constant.28 In contrast, Majara et al.37 observed an increase in trehalose concentration after oxygenation of a yeast slurry, indicating that the response may be strain dependent. Although no difference in cell growth rate was observed, ethanol production and the enzymic activities of alcohol dehydrogenase and pyruvate decarboxylase were modified by the transition from anaerobiosis to aerobiosis. This alteration in the environmental conditions also significantly enhanced the activity of oxidative stress enzymes such as superoxide dismutase and catalase.13 Oxidative stress would therefore appear to occur during pitching.

6.3.7

Storage and acid washing

Storage of yeast is a critical step in yeast handling38 as it should ensure that yeast cells are maintained in a minimal metabolic state, largely unaffected by environmental stress. Indeed, during storage, yeast cells rely on endogenous reserves to maintain basal cellular functions.28 Depletion of these reserves owing to prolonged storage and exposure to stress such as cold shock or ethanol stress may affect subsequent fermentation performance.33 Yeast viability and glycogen content decreased when storage was prolonged and when temperature was increased.33,39 Glycogen and trehalose levels decreased slightly in yeast cells stored at 5°C, whereas they decreased dramatically when yeast cells were stored at 20°C.40 However, antioxidant levels have not been previously studied during the storage of brewing yeast. To investigate the existence of oxidative stress during storage, samples were collected from storage vessels on a daily basis. The impact of poststorage treatments such as acid washing was also examined. Acid washing enables the elimination of undesirable organisms (mainly bacteria) from the stored yeast before pitching into fresh wort. This procedure is achieved by acidifying the yeast slurry to pH 2–2.5 with food-grade phosphoric acid for 2 h followed by neutralisation.41 The impact of this procedure on the oxidative stress response has not been investigated before. To determine the effect of acid washing on antioxidants and reserve carbohydrate levels, samples were collected before and after this treatment. All samples were assayed for catalase activity, glutathione concentration, glycogen and trehalose levels. The levels of catalase and glutathione remained stable during storage and acid washing (data not shown). No evidence of oxidative stress was found during storage and acid washing.

70 6.3.8

BREWING YEAST FERMENTATION PERFORMANCE

Serial repitching

After fermentation, yeast is cropped, stored and conditioned to perform another fermentation: this is known as serial repitching. From a propagation culture, no more than 15–20 repitching cycles should be performed, to avoid strain genetic drift and to prevent bacterial and wild yeast contaminations which may lead to inconsistent fermentation performance and off-flavour development.6 The effect of serial repitching on various fermentation performance values has not been extensively examined. Extensive serial repitching results in a progressive rise in 50 Catalase

40 30 20 10 0 (a)

1

3

4

4

5

6

7

1

3

4

4

5

6

7

1

3

4

4

5

6

7

1

3

4

7

7

8

Glutathione

30 20 10 0 (b)

7

7

8

Glycogen

1000

500

0 (c)

7

7

8

Trehalose

150 100 50 0 (d)

4

5

6

7

7

7

8

Generation Fig. 6.6 Antioxidant and reserve carbohydrate levels in cropped yeast depending on generation. Catalase (a) is expressed as units/g protein, glutathione (b) as nmol/108 cells, and glycogen (c) and trehalose (d) as g equivalent glucose/108 cells. The results represent the mean of three spectrophotometric measurements.

BREWING YEAST OXIDATIVE STRESS RESPONSES

71

cell surface charge and flocculation capacity.42–44 In addition, trehalose levels increased with generation number, whereas glycogen contents declined from propagation to the second generation, but subsequently returned to propagation levels.44 Samples were collected from yeast containment vessels (YCVs) immediately after cropping and analysed for catalase activity, glutathione concentration, glycogen and trehalose levels. The levels of antioxidant molecules and reserve carbohydrates remained stable throughout serial repitching (Fig. 6.6). However, samples from the fourth generation, which originated from the same yeast slurry and wort batch, exhibited higher concentrations on the four parameters studied. Therefore, it is postulated that the variations observed may be the result of wort composition effects rather than yeast generation number per se. 6.4

Conclusions

The tolerance of brewing yeast strains to exogenous H2O2 stress was dependent on strain and the phase of growth exhibited by the cell population. Lager strains appeared to be more resistant than ale strains, although the reason for this is not known. Cellular catalase activity and glutathione content indicate that the level of defence against H2O2 was strain dependent. While the catalase levels appeared to be directly related to the resistance of the strains to oxidative stress, the glutathione content of the cells was inversely related to the activity of this enzyme; glutathione may therefore compensate for reduced catalase activity and vice versa. Indeed, a recent study suggested that glutathione and catalase represent redundant H2O2 detoxification systems,45 and this implies that other defence systems may be important during this form of stress. Catalase activity, glycogen and trehalose concentrations exhibited a significant increase during propagation, whereas glutathione levels remained stable. The genes involved in catalase, glycogen and trehalose biosynthesis or recycling are under control of promoters called (general) STREs, which mediate the activation of genes after heat shock, high salt or oxidative stresses.46 Glutathione concentration is not regulated by STREs but by the YAP1 pathway.47 Therefore, the observations reported herein may be the consequence of exposure to a combination of stresses rather than to oxidative stress alone during propagation. No evidence of oxidative stress was found during storage, acid washing or cropping. Acknowledgements Veronique Martin and Katherine Smart gratefully acknowledge the support of Bass Brewers Ltd, and would like to thank Wendy Box and David Ruddlesden for their help with the sample collection. Veronique Martin is supported by the Henry Mitchell Scholarship. Katherine Smart is the Scottish Courage Reader in Brewing Science and gratefully acknowledges the support of the Royal Society, BBSRC and EPSRC for the award of her Royal Society Industrial Fellowship. The authors would like to thank the Directors of Bass Brewers for kind permission to publish this work.

72

BREWING YEAST FERMENTATION PERFORMANCE

References 1. Santoro, N. and Thiele, D.J. (1997) Oxidative stress responses in yeast. In: Yeast Stress Responses, Hohmann, S. and Mager, W.H. (eds). Springer, Heidelberg, pp. 171–212. 2. Fridovitch, I. (1978) The biology of free radicals. Science 201, 875–879. 3. Gralla, E.B. and Kosman, D.J. (1992) Molecular genetics of superoxide dismutases in yeasts and related fungi. Adv. Genet. 30, 251–319. 4. Jamieson, D.J. (1998) Oxidative stress responses of the yeast Saccharomyces cerevisiae. Yeast 14, 1511–1527. 5. Boulton, C.A. and Quain, D.E. (1987) Yeast, oxygen and the control of brewery fermentations. Proc. Eur. Brew. Conv. Cong., Madrid, 21, 401–408. 6. Boulton, C.A. (1991) Yeast management and the control of brewery fermentations. Brew. Guardian April, 25–29. 7. Aebi, H. (1984) Catalase in vitro. Methods Enzymol. 105, 121–126. 8. Izawa, S., Inoue, Y. and Kimura, A. (1995) Oxidative stress response in yeast: effect of glutathione on adaptation to hydrogen peroxide stress in Saccharomyces cerevisiae. FEBS Lett. 368, 73–76. 9. Parrou, J.L., Teste, M.A. and Francois, J. (1997) Effects of various types of stress on the metabolism of reserve carbohydrates in Saccharomyces cerevisiae: genetic evidence for a stress-induced recycling of glycogen and trehalose. Microbiology 143, 891–900. 10. Martin, V., Quain, D.E. and Smart, K.A. (1999) The oxidative stress response of ale and lager yeast strains. Proc. Eur. Brew. Conv., Cannes, 1999, pp. 679–686. 11. Izawa, S., Inoue, Y. and Kimura, A. (1996) Importance of catalase in the adaptive response to hydrogen peroxide: analysis of acatalasaemic Saccharomyces cerevisiae. Biochem. J. 320, 61–67. 12. Ruis, H. and Hamilton, B. (1992) Molecular Biology of Free Radical Scavenging Systems. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 153–172. 13. Clarkson, S.P., Large, P.J., Boulton, C.A. and Bamforth, C.W. (1991) Synthesis of superoxide dismutases, catalases and other enzymes and oxygen and superoxide toxicity during changes in oxygen concentration in cultures of brewing yeast. Yeast 7, 91–103. 14. Collinson, L.P. and Dawes, I.W. (1995) Inducibility of the response of yeast cells to peroxide stress. Gene 156, 123–127. 15. Cheeseman, K.H. and Slater, T.F. (1993) An introduction to free radical biochemistry. Br. Med. Bull. 49, 481–493. 16. Bucala, R. (1996) Lipid and lipoprotein oxidation: basic mechanisms and unresolved questions in vivo. Redox Report 2, 291–307. 17. Gutteridge, J.M.C. and Halliwell, B. (1990) The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. 15, 129–135. 18. Moradas-Ferreira, P., Costa, V., Piper, P. and Mager, W. (1996) The molecular defences against reactive oxygen species. Mol. Microbiol. 19, 651–658. 19. Van Zandycke, S.M., Siddique, R. and Smart, K.A. (2001) Carbohydrate utilization and membrane potential. Proc. Eur. Brew. Conv., Budapest. 20. Stadtman, E.R. (1993) Oxidation of free amino acids and amino acid residues in proteins by radiolysis and by metal-catalyzed reactions. Annu. Rev. Biochem. 62, 797–821. 21. Davies, K.J.A. (1987) Protein damage and degradation by oxygen radicals. J. Biol. Chem. 262, 9895–9901. 22. Wiseman, H. and Halliwell, B. (1996) Damage to DNA by reactive oxygen and nitrogen species: role in inflammatory disease and progression of cancer. Biochem. J. 313, 17–29. 23. Halliwell, B. and Gutteridge, J.M.C. (1999) Free Radicals in Biology and Medicine. Clarendon Press, Oxford. 24. Imlay, J.A. and Linn, S. (1988) DNA damage and oxygen radical toxicity. Science 240, 1302–1309. 25. Boiteux, S. and Radicella, J.P. (1999) Base excision repair of 8-hydroxyguanine protects DNA from endogenous oxidative stress. Biochimie 81, 59–67. 26. Quain, D.E. (1986) Differentiation of brewing yeast. J. Inst. Brew. Centenary Rev. 92, 435–438. 27. Masschelein, C.A., Borremans, E. and Van de Winkel, L. (1994) Application of exponentially-fedbatch cultures to the propagation of brewing yeast. Proc. Inst. Brew., Asia Pacific Sect. 23, 104–108. 28. Boulton, C.A. (2000) Trehalose, glycogen and sterols. Proc. Brew. Yeast Ferm. Perform. Cong. 2, 10–19. 29. Lillie, S.H. and Pringle, J.R. (1980) Reserve carbohydrate metabolism in Saccharomyces cerevisiae: responses to nutrient limitation. J. Bacteriol. 143, 1384–1394. 30. Maris, A.F., Kern, A.L., Picada, J.N. et al. (2000) Glutathione, but not transcription factor Yap1, is required for carbon source-dependent resistance to oxidative stress in Saccharomyces cerevisiae. Curr. Genet. 37, 175–182.

BREWING YEAST OXIDATIVE STRESS RESPONSES

73

31. Jakobsen, M. (1982) The effect of yeast handling procedures on yeast oxygen requirement and fermentation. Proc. Conv. Inst. Brew., Aust. N.Z. Sect., Perth, pp. 132–137. 32. Quain, D.E., Thurston, P.A. and Tubb, R.S. (1981) The structural and storage carbohydrates of Saccharomyces cerevisiae: changes during fermentation of wort and a role for glycogen catabolism in lipid biosynthesis. J. Inst. Brew. 87, 108–111. 33. McCaig, R. and Bendiak, D.S. (1985) Yeast handling studies. I. Agitation of stored pitching yeast. J. Am. Soc. Brew. Chem. 43, 114–118. 34. Callaerts, G., Iserentant, D. and Verachtert, H. (1993) Relationship between trehalose and sterol accumulation during oxygenation of cropped yeast. J. Am. Soc. Brew. Chem. 51, 75–77. 35. Boulton, C.A. and Quain, D.E. (1999) A novel system for propagation of brewing yeast. Proc. Eur. Brew. Conv., Cannes, 1999, pp. 647–654. 36. Sasaki, N., Yasuda, Y., Imai, T. et al. (2000) The effect of wort aeration using a high oxygen concentration on fermentation, yeast physiology and the quality of the finished beer. Tech. Q. Master Brew. Assoc. Am. 37, 27–30. 37. Majara, M., O’Connor-Cox, E.S.C. and Axcell, B.C. (1996) Trehalose – a stress protectant and stress indicator compound for yeast exposed to adverse conditions. J. Am. Soc. Brew. Chem. 54, 221–227. 38. O’Connor-Cox, E.S.C. (1998) Improving yeast handling in the brewery. Part 2: Yeast collection. Brew. Guardian 127(2), 22–34. 39. McCaig, R. and Bendiak, D.S. (1985) Yeast handling studies. II. Temperature of storage of pitching yeast. J. Am. Soc. Brew. Chem. 43, 119–122. 40. Morimura, S., Hino, T., Kida, K. and Maemura, H. (1998) Storage of pitching yeast for the production of whisky. J. Inst. Brew. 104, 213–216. 41. Simpson, W.J. and Hammond, J.R.M. (1989) The response of brewing yeasts to acid washing. J. Inst. Brew. 95, 347–354. 42. Smart, K.A. and Whisker, S. (1996) Effect of serial repitching on the fermentations properties and condition of brewing yeast. J. Am. Soc. Brew. Chem. 54, 41–44. 43. Teixeira, J.M., Teixeira, J.A., Mota, M. et al. (1991) The influence of cell wall composition of a brewer’s flocculent lager yeast on sedimentation during successive industrial fermentations. Proc. Eur. Brew. Conv. Cong., Lisbon, 23, 241–248. 44. Jenkins, C., Kennedy, A.I., Thurston, P., Hodgson, J.A. and Smart K.A. (2001). Impact of serial repitching and wort compostion on fermentation performance and organelle integrity of lager brewing yeast. Proc. Eur. Brew. Conv., Budapest. 45. Grant, C.M., Perrone, G. and Dawes, I.W. (1998) Glutathione and catalase provide overlapping defences for protection against hydrogen peroxide in the yeast Saccharomyces cerevisiae. Biochem. Biophys. Res. Commun. 253, 893–898. 46. Mager, H.M. and Hohmann, S. (1997) Stress response mechanisms in the yeast Saccharomyces cerevisiae. In: Yeast Stress Responses, Hohmann, S. and Mager, W.H. (eds). R.G. Landes, Austin, TX, pp. 1–5. 47. Grant, C.M., Collinson, L.P., Roe, J.-H. and Dawes, I.W. (1996) Yeast glutathione reductase is required for protection against oxidative stress and is a target gene for yAP-1 transcriptional regulation. Mol. Microbiol. 21, 171–179.

Part 3 Wort Composition: Impact on Yeast Metabolism and Performance

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

7 Wort Composition and Beer Quality C.W. BAMFORTH

Abstract Whereas much attention has been lavished on ensuring consistency in yeast quality and quantity in pursuit of controlled fermentation, the significance of variations in wort composition has been paid spasmodic attention. Wort has a huge direct influence on beer quality, through its components surviving into the product to determine colour, clarity, foam, safety and wholesomeness and some key aspects of flavour, but also through its impact on yeast performance. While the gross impact that the individual wort components have on yeast physiology and surface behaviour and therefore beer quality is documented in the literature, there is remarkably little appreciation of how much significance can be attached to batch-to-batch fluctuation in wort composition. It is suggested that the most useful index of tolerance is the flavour compound(s) that displays the most sensitive reaction to a change in one or more wort components. Building on this information, there is an urgent need for the construction of models relating wort composition and yeast quality and quantity to flavour compound production. Such models will aid the brewer in decisions concerning how much effort is needed to control wort composition and to what tolerance.

7.1 Introduction Libraries are replete with volumes describing the formulation of media for industrial fermentations (see, for example, Ref. 1). For the most part, however, commercial fermentations are targeted on the production of high yields of either an organism or a single product produced by that organism. In such situations it is entirely feasible systematically to deduce optimised fermentation conditions, for example by changing medium ingredients one at a time to arrive at the appropriate level of each or by adding ingredients to a steady-state continuous culture. In the case of brewery fermentations, however, there is a somewhat different scenario. In this case the product of interest is essentially the spent growth medium. Rather than worrying (for the most part) about the yield of the organism itself or the yield of a single end-product, brewers are concerned with an optimised balance of molecules left behind after the yeast has done its job. The medium, wort, is far less defined than that which is used for most industrial fermentations. Rather than the mixing of a relatively few pure ingredients to produce the medium, wort is derived from vegetative (and therefore inherently variable) sources (malt, adjuncts, hops) in processes controlled by coarse parameters such as time, temperature and solid–liquid ratio. The analytical specifications applied to the wort, too, are relatively broad and by no means is there batch-to-batch measurement of the individual substances (e.g. sugars, amino acids, vitamins, inorganic ions, lipids) on which the yeast depends for its metabolism. Most brewers will rely simply on overall strength of the wort (specific gravity) as an index of quantity, assuming that the relative balance of carbohydrates, nitrogenous constituents and other materials is remaining constant.

78

BREWING YEAST FERMENTATION PERFORMANCE

Just about the only wort components that they will individually specify and quantify are the oxygen dosed in before fermentation and zinc, when added. Is this, or is it not, an excessively simplistic stance? That is the nub of this thesis. 7.2 The relationship of wort composition to beer quality Wort can impinge upon beer quality in two ways: (i) through its impact on yeast, and (ii) directly, in terms of materials passing into the beer without being affected (for the most part) by yeast. Taking the second of these, it is easy to establish that more aspects of beer quality (positive and negative) are a direct consequence of the wort rather than through its being ‘modified’ by yeast. The colour of beer is generally established in the brewhouse or earlier, through either melanoidin formation in malt kilning and roasting and wort boiling, or the oxidation of polyphenolics in mashing and boiling.2 The colloidal susceptibility of beer is very largely established upstream of the fermenter, in terms of sensitive polypeptides, tannoids, glucans, oxalate, etc.3 Foam is a consequence primarily of amphipathic polypeptides derived from cereals and bitter compounds from hops. Indeed, yeast is a nuisance to foam (with the exception of its role in developing carbonation).4 Foaming in the fermenter tends to remove surface-active materials; shorter chain fatty acids released by yeast are foam negative, as too is ethanol itself, and ailing yeast releases proteolytic enzymes that cut away at foaming proteins.4 The most serious episodes of gushing can be directly linked to malt and, in turn, to infected grain.5 The majority of food safety issues have been linked to wort components,6 although equally the wholesome materials (other than alcohol) such as vitamins, silicate, fibre and polyphenols emerge from the wort.7 Finally, a good proportion of the flavour comes from the wort, whether malt8 or hop derived.9 Therefore, regarding the impact of wort on beer quality through the mediacy of yeast, beyond the production of ethanol and carbon dioxide, the primary concern is with matters of flavour (leaving aside considerations of the health of the yeast collected at the end of fermentation and its relationship to the wort that it has just been tackling). The spectrum of flavour-active materials produced by yeast depends on the medium in which it is growing, both via that medium’s impact on the extent of yeast growth and on the metabolic fluxes within the yeast, and also because yeast will convert some materials in the medium directly to flavoursome materials (e.g. dimethyl sulfoxide10). Equally, there may be components in the wort that block these changes taking place (in this specific example, methionine sulfoxide11). Finally, there can be materials present in this medium that influence a yeast’s physical behaviour (e.g. flocculation12) and this will impact on its ability to ferment. As this chapter is part of a book on yeast, it will focus on the wort–yeast interface, leaving aside consideration of the direct impact of wort on beer quality. 7.3

The key components of wort

The general state of knowledge of the key ingredients of wort as they pertain to yeast performance is well understood and was succinctly summarised by John Hammond in

WORT COMPOSITION AND BEER QUALITY

79

Table 7.1 Principal components and parameters of wort relevant to yeast performance Gravity Sugar spectrum – fermentability (Assimilable) free amino nitrogen (FAN) Lipids (clarity) Inorganic ions pH (and buffering) Vitamins Surface modifiers (e.g. flocculation-inducing polysaccharides) Oxygen Flavour precursors

the first edition of this book.13 The present discussion will address the topic of tolerance: how much variation can there be in a wort before perceptible differences can be seen in the quality of the beer that is produced? In particular, the discussion is restricted to consideration of the direct relationship between wort composition and beer quality parameters, leaving aside for the sake of brevity any impact of wort variation on in-brewery performance criteria, such as yield of yeast or filtration performance. Table 7.1 summarises the key components of wort relevant to yeast physiology.

7.4

The impact of wort on the production of flavour compounds by yeast

In practice, then, the discussion perforce focuses on flavour. Put at its simplest: the limit of tolerance on the wort will be that which makes a perceptible difference to flavour. Whichever is the most sensitive flavour compound in terms of change in level (upwards or downwards) sets the goalposts, for it takes only a change in the level of that compound for the beer to be perceptibly different. Again, the thinking must be simplified, because beer flavour is an extremely complex topic, the manifestation of the effects on the naso-olfactory apparatus of a diversity of taste and aroma active materials. For ease, the compounds may be thought of as contributing to taste or smell individually and in isolation. To determine how big a change in level of a flavour substance has to be for it to be detectable, one must consider both the flavour threshold and the relationship between concentration and flavour impact. If the level of a compound is fluctuating at levels below the flavour threshold, e.g. as a consequence of variability in wort composition, then this does not matter at all. If the level of a compound is above the threshold, however, then the relationship between concentration and flavour impact must be considered. Stevens’ power law (Equation 7.1) describes the relationship between a sensory response (e.g. smell) and the concentration of an aroma-active compound. R  kC n

(7.1)

where R is the sensory response, C is concentration, k is a constant and n is the psychophysical constant.

80

BREWING YEAST FERMENTATION PERFORMANCE

(a)

(b)

Identical n

Different n

C

B

Log R

Log R

A

D

Log (conc.) Fig. 7.1

Log (conc.)

Stevens’ power function.

If logarithms are taken, this equation converts to a straight line (Fig. 7.1). The intercept on the y-axis gives k and this depends very much on the units of measurement. The slope of the line is given by n. Consider two substances that give parallel lines in this type of plot (Fig. 7.1a). At a given concentration of compound A it generates a more intense response. However, the extent to which an increase in concentration registers as an impact on the subject’s response is identical. Turning to Fig. 7.1b, it can be seen that the two compounds display different slopes of line (different n values). Compound C shows a much greater slope than does compound D, i.e. smaller increments in level of C will lead to a more perceptible change in sensory response than would be the case for D. These relationships for the majority of flavour-active materials in beer have not been elucidated. One of the problems is how to do this in a multivariate system, where the perception of a given flavour may be due not only to a greater or lesser quantity of a certain compound, but also to the background of fluctuations of other compounds that may reinforce or disguise the perception of the compound of interest. The Stevens’ correlations need to be elucidated for the key flavour components of beer. Only then can the most sensitive marker compound(s) be elucidated. However, turning to dimethyl sulfide (DMS) by way of illustration, it may be inferred from the work of Brown et al.14 that a difference of 5 ppb in DMS (when above the flavour threshold of 28 ppb) is readily detectable. It has also been shown that the perception of DMS is masked by 2-phenylethanol (2-PE),15 and in this case a change in concentration of approximately 5 ppm would seem to be readily detectable as a decrease in perceived DMS character. Taking the example of DMS/2-PE, what changes in wort composition are sufficient to cause a change in these compounds of a magnitude that will be detectable? By lowering the free amino nitrogen (FAN) content of the wort from 160 to 100 ppm, the level of 2-PE produced in fermentation was increased by more than 10 ppm.15 Such a limitation in FAN would also lead to a huge increase in DMS production.16 Such a deficiency in the level of FAN would be unusual indeed. However, it is not easy to determine just how much variation in FAN is tolerable. It is extremely difficult to find reasoned experimental studies where such a parameter has been varied in isolation, such that plots of (say) DMS level versus FAN can be produced. One of the reasons is the difficulty in ‘naturally’ varying a parameter in isolation in a complex

81

WORT COMPOSITION AND BEER QUALITY

245 240 235

FAN

230 225 220 215 210 205 14.2

14.4

14.6

14.8 15 Plato

15.2

15.4

15.6

Fig. 7.2 Commercial data for free amino nitrogen (FAN) levels (mg/l) and degrees Plato from a single company. The data suggest that at a given gravity, FAN may vary over a 15–20 mg/l range. Table 7.2

Effect of lipids on ester production by yeast

Lipid addition a

Control Spent grain lipids (300 ppm)a Controlb Oleic (100 ppm)b

Ethyl acetate 17.6 3.3 111 54

Isoamyl acetate 1.1 0.2 14 8

a

Taylor et al.;20 bAnderson and Kirsop.21

medium such as wort. The best hope is to produce a medium that is to all intents and purposes a ‘mock-up wort’.17 Yet there are substantial indications that relatively modest changes in a parameter such as FAN can make a sizeable contribution to changes in the level of volatiles such as higher alcohols and esters in beer.18 Figure 7.2 shows some commercial data on the range of FAN levels encountered in worts, which suggest that a range of perhaps 20 ppm (10 ppm; or approximately 10%) may be typical. In many ways the wort parameter (other than oxygen) most documented for its effect on yeast performance and hence beer composition is clarity. The impact of ‘dirty worts’ on the production, for example, of sulfur dioxide19 and esters20 is evident. In the latter case this is illustrated in Table 7.2. Table 7.3 illustrates just how much variation can be observed between and within breweries in lipid levels in wort (an index of trub carryover).

7.5

Models

Despite the empiricism of many studies in this area, relatively sophisticated models to describe the relationship between wort components and yeast amount and performance capabilities do exist. For instance, Gee and Ramirez22 derived relatively

82

BREWING YEAST FERMENTATION PERFORMANCE

Table 7.3 Brewery

1 2 3 4 5 6 7

Variation in lipid content of commercial worts Total lipids (C6 to C18:3) (ppm) Median

High

Low

10.1 5.5 9.8 5.1 8.1 6.9 8.0

18.8 9.1 48.3 15.2 83.0 15.5 61.4

4.0 4.1 5.8 3.4 3.1 5.1 4.2

straightforward equations to explain the production of volatiles such as higher alcohols and esters in brewery fermentations. Examples are given in Equations 7.2 and 7.3. Formation of esters: d[IA c ]  YIA c  IA X (7.2) dt where IAc  concentration of isoamyl acetate, YIAc  yield coefficient, moles IAc /moles IA, IA  specific rate of isoamyl alcohol formation, and X  yeast concentration. It is apparent that the rate of production of an ester such as isoamyl acetate is inherently dependent on the production of its precursor isoamyl alcohol (IA), and on the yeast concentration (X). Formation of higher alcohols: d[IA] dt

 YIA /S  x X

K I,L K I,L  L

 YIA /E  L X

(7.3)

where IA  concentration of isoamyl alcohol, YIA/S  yield of isoamyl alcohol by synthetic pathway, YIA/E  yield of isoamyl alcohol by Ehrlich pathway, x  specific yeast growth rate, X  yeast concentration, KI,L  inhibition constant for leucine, L  leucine concentration, and L  specific rate of leucine uptake. The production of isoamyl alcohols is affected by the rate of yeast growth (which will be determined by various factors) and by the presence of inhibitors that block the biosynthetic pathway. It is apparent from these equations that relevant factors are the amount of yeast present and its condition and ability to grow (specific growth rate). The equations illustrate how the amount of ester is a direct consequence of the extent to which the precursor higher alcohol is produced. Most importantly, the equations illustrate how levels of individual amino acids (such as leucine) are important because they inhibit the synthetic pathway of higher alcohol production. In other words, it is a matter of knowing not only how much total FAN is present, but equally what is the balance of amino acids. By inserting real-life numbers into this type of equation it is possible to gain some idea of the extent to which a change in the level of individual amino acids and of the other parameters influences the production of volatiles such as esters. There is a real need for this type of model to be developed (probably using artificial worts of increasing

WORT COMPOSITION AND BEER QUALITY

83

complexity) for all aspects of brewing yeast fermentations and for models to be tested in worts displaying natural variation. The models ought also to take cognisance of the fact that the preferred wort composition can differ for different yeasts and different generations of the same yeast. This was well illustrated by Quilliam et al.,23 who showed that first generation yeast grown in a medium containing 190 mg/l FAN produced the same levels of ethyl acetate as sixth generation yeast presented with 280 mg/l FAN.

7.6

Sources of variability in wort composition

Wort composition may change as a result of variations in water composition, grist composition and brewhouse conditions. Assuming that the last of these variables is controlled (although variations in wort clarity beg the question of whether this is universally the case), then it may be surmised that the main source of variation will be water and grist. The former is more readily regulated, through water treatment protocols. Regarding the grist, it is likely that brewing syrups, produced from relatively pure starches by controlled enzymic and acid hydrolysis, will be more defined than malts. The scenario with regard to the malt is complex. In the conversion of barley to sweet wort thought the mediacy of malting and mashing, a plethora of substrates, enzymic and non-enzymic reactions, specific and non-specific inhibitors and activators is at play. Season upon season, variety by variety, changes occur in the relative balance of substrates in the barley, in its physiological capabilities and therefore in its behaviour in malting and mashing. All that the maltster and brewer can hope to do is apply realistic and meaningful specifications to barley and malt that can be responded to through the adjustment of processing conditions to obtain wort displaying the greatest practical consistency.24 There is no space here to discuss the wealth of knowledge concerning the biochemistry and chemistry of breakdown of barley polymers and their conversion into wort components. It is important to recognise, however, that new discoveries to add to an already complex body of received wisdom are coming to the fore regularly. Consider, for instance, the observations of Stenholm and Home that limit dextrinase is more heat tolerant than was hitherto believed to be the case, and is particularly prevalent if the mash pH is lowered, say to 5.4.25 The implication is that relatively subtle changes in pH can have a substantial effect on fermentability and on the ratio of assimilable sugars to assimilable nitrogen. Most brewers would be comfortable with a pH range of 0.1 either side of the target, and yet that could have a profound impact on wort composition (and more besides26). Another area in which a large quantity of new data is emerging is that of protein degradation. Work in Jones’ laboratory showed, inter alia, that there are more than 40 active proteolytic enzymes in malt,27 and moreover that these can variously be blocked by inhibitors originating in malt which, when released, suppress proteolysis in mashing.28 The extent to which all this varies between varieties, malts with different properties, under different mashing conditions, etc., has not been fully elucidated. Therefore, a clear picture of something as fundamental as the ranges that can be observed in the balance of the various amino acids in malts and worts produced under

84

BREWING YEAST FERMENTATION PERFORMANCE

different conditions is certainly not available, despite the obvious need for this information.

7.7

Conclusions

Much effort has been devoted in recent years to achieving fermentation control in breweries. Through the introduction of robust pitching control systems and attention to the viability and vitality of yeast, the yeast component is surely attended to sufficiently. In contrast, with the exception of control of oxygen, overall strength, absence of starch and perhaps clarity, wort remains largely unspecified, other than perhaps coarse adjustments made on periodic checking of parameters such as free amino nitrogen. While no brewer really questions the fact that yeast performance, both physical and metabolic, is influenced hugely by the wort, the precise extent to which batchto-batch and season-to-season variations in the wort impact on fermentation performance and beer quality is uncertain. No wonder many brewers are now focusing in on the magnitude and significance of variations in the major wort parameters. The most feasible way to tackle this may be to work with increasingly complex model wort systems using yeasts with different characteristics (e.g. strain, age, size). This is a daunting challenge. However, the fact that brewers are already for the most part able to produce beer of rather impressive consistency, without too much paranoia concerning the wort, suggests that the problem is not insurmountable.

Acknowledgements I am very grateful to the commercial brewing companies who volunteered information for this study and am pleased to respect their anonymity.

References 1. Ertola, R.J., Giulietti, A.M. and Castillo, F.J. (1995) Design, formulation and optimization of media. In: Bireactor System Design, Asenjo, J.A. and Merchuk, J.C. (eds). Marcel Dekker, New York, pp. 89–137. 2. Smedley, S.M. (1992) Colour determination of beer using tristimulus values. J. Inst. Brew. 98, 497–504. 3. Bamforth, C.W. (1999) Beer haze. J. Am. Soc. Brew. Chem. 57, 81–90. 4. Bamforth, C.W. (1999) Bringing matters to a head: the status of research on beer foam. Proc. Eur. Brew. Conv. Foam Symp. Amsterdam, pp. 10–23. 5. Munar, M.J. and Sebree, B. (1997) Gushing – a maltster’s view. J. Am. Soc. Brew. Chem. 55, 119–122. 6. Long, D.E. (1999) From cobalt to chloropropanol: de tribulationibus aptis cervisiis imbibendis. J. Inst. Brew. 105, 79–84. 7. Baxter, E.D. (1996) Beer is good for you – discuss. Brewer 92, 63–66. 8. Moir, M. (1989) Effects of raw materials on flavour and aroma. Brew. Guardian 118(9), 64–71. 9. Moir, M. (2000) Hops – a millennium review. J. Am. Soc. Brew. Chem. 58, 131–146. 10. Anness, B.J., Bamforth, C.W. and Wainwright, T. (1979) The measurement of dimethyl sulfoxide in barley and malt and its reduction to dimethyl sulfide by yeast. J. Inst. Brew. 85, 346–349. 11. Gibson, R.M., Large, P.J., Anness, B.J. and Bamforth, C.W. (1983) The identity of an inhibitor in wort of dimethyl sulfoxide reductase from yeast. J. Inst. Brew. 89, 215–218.

WORT COMPOSITION AND BEER QUALITY

85

12. Herrera, V.E. and Axcell, B.C. (1991) Induction of premature yeast flocculation by a polysaccharide fraction isolated from malt husk. J. Inst. Brew. 97, 359–366. 13. Hammond, J. (2000) Yeast growth and nutrition. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 77–85. 14. Brown, D.G.W., Clapperton, J.F., Meilgaard, M.C. and Moll, M. (1978) Flavor thresholds of added substances. J. Am. Soc. Brew. Chem. 36, 73–80. 15. Hegarty, P.K., Parsons, R., Bamforth, C.W. and Molzahn, S.W. (1995) Phenyl ethanol – a factor determining lager character. Proc. Eur. Brew. Conv. Cong., Brussels, pp. 515–522. 16. Gibson, R.M., Large, P.J. and Bamforth, C.W. (1985) The influence of assimilable nitrogen compounds in wort on the ability of yeast to reduce dimethyl sulfoxide. J. Inst. Brew. 91, 401–405. 17. Kennedy, A.I., Taidi, B., Dolan, J.L. and Holdgson, J.A. (1997) Optimisation of a fully defined medium for yeast fermentation studies. Food Technol. Biotechnol. 35, 261–265. 18. Äyräpää, T. (1967) Formation of higher alcohols from 14C-labelled valine and leucine. J. Inst. Brew. 73, 17–33. 19. Dufour, J.-P., Carpentier, B., Kulakumba, M. et al. (1989) Alteration of SO2 production during fermentation. Proc. Eur. Brew. Conv. Cong. Zurich, pp. 331–338. 20. Taylor, G.T., Thurston, P.A. and Kirsop, B.H. (1979) The influence of lipids derived from malt spent grains on yeast metabolism and fermentation. J. Inst. Brew. 85, 219–227. 21. Anderson, R.G. and Kirsop, B.H. (1974) The control of volatile ester synthesis during the fermentation of wort of high specific gravity. J. Inst. Brew. 80, 48–55. 22. Gee, D.A. and Ramirez, W.F. (1994) A flavour model for beer fermentation. J. Inst. Brew. 100, 321–329. 23. Quilliam, W., Hulse, G. and Cameron-Clarke, A. (2000) Yeast management and fermentation performance. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.) Blackwell Science, Oxford, pp. 189–200. 24. Bamforth, C.W. (1999) A critical assessment of malt analysis from the brewer’s perspective. Tech. Q. Master Brew. Assoc. Am. 36, 301–306. 25. Stenholm, K. and Home, S. (1999) A new approach to limit dextrinase and its role in mashing. J. Inst. Brew. 105, 205–210. 26. Bamforth, C.W. (2001) pH in brewing: an overview. Tech. Q. Master Brew. Assoc. Am. 38, 1–8. 27. Zhang, N.Y. and Jones, B.L. (1995) Characterisation of germinated barley endoproteolytic enzymes by 2-dimensional gel electrophoresis. J. Cereal Sci. 21, 145–153. 28. Jones, B.L. and Marinac, L.A. (2000) Purification and partial characterisation of a second cysteine proteinase inhibitor from ungerminated barley (Hordeum vulgare L.). J. Agric. Food Chem. 48, 257–264.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

8 Wort Substitutes and Yeast Nutrition B. TAIDI, A.I. KENNEDY and J.A. HODGSON

Abstract Brewing research and the study of yeast physiology have always been hampered by the batch-to-batch variation in the exact composition of wort. Wort production is a biochemical process, using biological material, naturally leading to variations in the exact chemical composition of wort. Long-term fermentation and physiological studies require the formulation of defined or semi-defined media that mimic wort in their nutritional composition. Any ‘artificial wort’ would require reproducibility of recipe and similarity of composition to brewing wort, including carbohydrate and free amino nitrogen spectra. Unlike most physiological studies where the medium composition is adjusted to provide balanced growth, wort is a medium that causes unbalanced and restricted growth of yeast. Any physiological studies performed in the context of brewing have to reflect this unbalanced composition. Two media, one fully defined and one semi-defined, suitable for use as wort substitutes are discussed in light of yeast nutrition under brewing conditions. Some laboratory studies performed with these media are also outlined, as a demonstration of the potential application of these media.

8.1 Introduction Brewing research and the study of yeast physiology have always been hampered by the variation in the exact composition of wort. Wort production is a biochemical process using biological materials and naturally leads to wort batches with variable composition. Variations in wort composition originate from batch-to-batch variations in raw materials, time-related deterioration of the raw materials in storage and variations in process control during wort production. Normally, all-malt wort can provide all of the nutrients required by brewing yeast, with the exception of unsaturated fatty acids, sterols and possibly zinc.1 Oxygenation of wort overcomes the deficiency of unsaturated fatty acid and sterols. Zinc sulfate is often added to wort as a source of zinc ions. The major soluble components of wort are fermentable carbohydrates, amino acids, peptides, proteins, lipids, metal ions and non-metal ions. Although some of these wort compounds can be measured with relative ease, it is impractical to measure all the components of wort and hence it is not possible to establish the differences between batches of wort. In addition, there are many undefined insoluble and microscopic components in wort. This makes it difficult to perform a long-term study of yeast physiology as the medium used would vary over the course of the investigation. Any attempts to stabilise wort microbiologically will change its composition. Heating processes such as autoclaving or pasteurisation affect the mineral and vitamin levels, colour, carbohydrate composition and particle content of wort. Stabilisation through sterile filtration would remove any particulate matter. In addition, filtration is only practical on a very small scale owing to rapid filter blockage. The particulate matter in

WORT SUBSTITUTES AND YEAST NUTRITION

87

wort (trub) is known to provide nutrients for yeast growth as well as physically increasing the rate of CO2 release.2 Particulate matter in wort generally increases the fermentation rate and trub concentration is highly variable between batches of wort. Various ways of obtaining a constant or totally reproducible ‘wort’ have been attempted. Large batches of wort have been dried, stored and reconstituted in small amounts as required. Reconstitution of wort can be performed reproducibly, but would not necessarily reflect the original composition of wort. Proteins and minerals can precipitate during the drying process and do not necessarily redissolve upon reconstitution with water. The process would also affect the pH and the buffering capacity of the wort. Storage of wort by freezing would have similar problems associated with it. In terms of nutrition, yeast requires macronutrients such as fermentable carbohydrates as a source of carbon, amino acids as a source of nitrogen, and oxygen to provide a source of unsaturated fatty acids and sterols. Many micronutrients such as vitamins, non-metal ions such as phosphate and sulfate ions, and metal ions are also required by yeast. Any artificial wort would have to reflect the concentration of these macronutrients and micronutrients. Preferably, the concentration of wort components should be reflected in terms of their bioavailability to yeast rather than their total concentration in wort. Many of the ions, for instance, are bound by natural chelating agents in wort and are unavailable to the yeast. In this paper, two ‘synthetic wort’ media are described. A fully defined medium3 was developed and tested against wort with ale and lager yeast strains. A semi-defined medium4 is also described which was used to perform studies on ester formation.

8.2 Materials and methods 8.2.1

Materials

Bottom-fermenting ale and lager yeast strains currently used by Scottish Courage Brewing Ltd were chosen for this study. Lager or ale wort from full production scale was used in the laboratory experiments. Bacto yeast nitrogen base without amino acids (YNB w/o aa) and ammonium sulfate was obtained from Difco (0335-15-9) (UK). Amino acids, carbohydrates and other general laboratory reagents were obtained from Sigma-Aldrich (UK). Brewing syrup was obtained from brewing sites and had a carbohydrate composition reflecting that of wort (Table 8.1). Corn steep liquor (CSL) was obtained as a sample and had a free amino nitrogen (FAN) concentration of 30 g/l. 8.2.2

Fully defined medium

The fully defined medium was prepared by supplementing YNB w/o aa and ammonium sulfate with a mixture of fermentable sugars to give an initial gravity of 1055°. The ratio of the various carbohydrates (Table 8.2) was based on those found in brewery worts. The assimilable nitrogen was supplied as a mixture of amino acids and ammonium sulfate (Table 8.3), resulting in an FAN concentration of 153 mg/l in the defined medium. The medium was further supplemented with 0.625 g/l citric acid and

88

BREWING YEAST FERMENTATION PERFORMANCE

Table 8.1

Composition of brewing syrup used in the semi-defined medium

Carbohydrate

Concentration (g/l)

Glucose Fructose Sucrose Maltose Maltotriose

Table 8.2

Ratio with respect to total carbohydrate (%) (% w/v)

15.5 0.0 0.0 613.4 160.8

2 0 0 61 16

2.0 0 0 77.7 20.4

Carbohydrate composition of the defined medium

Carbohydrate

Amount added to medium (g/l)

Glucose Fructose Sucrose Maltose Total

Ratio with respect to total carbohydrate (%)

10.4 4.6 3.5 115.5 134

7.8 3.4 2.6 86.2 100

The grade of maltose used (Sigma grade II) contained up to 7% maltotriose, which was appropriate to a model of brewery wort. Table 8.3 Nitrogenous compounds composition of the defined medium Amino acids L-Aspartic

acid

L-Threonine L-Serine L-Asparagine L-Glutamine L-Glutamic

acid

L-Proline

Glycine L-Alanine L-Valine L-Methionine L-Isoleucine L-Leucine L-Tryosine L-Phenylalanine L-Tryptophane L-Lysine

hydrochloride hydrochloride L-Arginine hydrochloride Ammonium sulfate L-Histidine

Concentration (g/l) 67.5 46.8 37.5 128.6 5.2 77.8 272.9 28.4 88.4 93.6 23.4 49.6 121.8 80.2 95.6 42.3 112.2 50.9 138.4 130.7

Adapted from Thompson et al.5

0.215 g/l CaSO4 · 2H2O. The pH of the medium, before carbohydrate addition, was adjusted to 5.2 using NaOH (1 M) before pasteurisation (60°C, 30 min). A sterile (121°C, 20 min) solution of carbohydrates was added to the pasteurised base medium before inoculation. Wort was stored, sterile (121°C, 20 min), diluted to 1055° and supplemented

WORT SUBSTITUTES AND YEAST NUTRITION

89

with ZnSO4 · 7H2O (0.7 mg/l) before use. Both defined medium and wort were aerated (sterile air, 20 min) through gas distribution tubes before inoculation. Fermentations were carried out using EBC tubes [Institute of Brewing (IOB) Methods of Analysis]. The inoculation rates for both lager and ale yeast strains were 2.0  107 viable cells/ml. Ale fermentations were carried out at 20°C and lager fermentations at 15°C, both followed by chilling to 4°C to separate the yeast. Fermentations were carried out in triplicate. For shake-flask cultures, YNB w/o aa containing ammonium phosphate (3.31 g/l), citric acid (4.20 g/l) and calcium sulfate (0.22 g/l) was prepared in concentrated form without a carbon source. The pH of the media was adjusted to 6.0 before filtersterilisation into sterile conical flasks (250 ml). Carbohydrate solutions of glucose, maltose or fructose were sterilised (121°C, 20 min) and mixed with the contents of the flasks at the appropriate concentrations to a total volume of 100 ml. The total assimilable nitrogen content of the medium was 1.76 g N/l. The flasks were inoculated with 1  106 cells/ml of a production lager strain and incubated at 27°C on a shaker (200 rpm). 8.2.3

Semi-defined medium

The medium was prepared by mixing the appropriate amounts of brewing syrup, tap water and CSL. CSL contains organic acids, of which lactic acid is the major component. The lactic acid was partially removed by adjusting the pH of the medium to 5.5 by the addition of CaCO3. The medium was then autoclaved, cooled to 4°C, allowed to settle overnight and decanted to remove the calcium lactate precipitate. Buffering of the medium was provided by the addition of citric acid (1.0 g/l) and adjusting the pH to 5.2 using NaOH. Fermentations were performed in duplicate at 1 litre scale in a model system consisting of 1 litre measuring cylinders fitted with rubber bungs and a one-way valve to allow escape of CO2 gas from the fermentations. These 1 litre tall tubes were sterilised by autoclaving (121°C, 15 min) before, the addition of medium. The fermentation temperature was a constant 15°C. A production lager strain of yeast at a pitching rate of 5.5 g/l was used. The yeast was acid-washed before pitching. Once target present gravity (PG) had been reached the fermentations were chilled to 4°C to allow yeast settlement and sample processing. 8.2.4

Analytical methods

Samples were withdrawn periodically from fermentations for immediate yeast cell concentration and viability determination (IOB Methods of Analysis). The clarified supernatants (2000 g, 10 min) of these samples were used for specific gravity and pH measurements. The specific gravity of samples was measured using a DMA 55 calculating density meter (Anton Paar, UK). A pH probe (Orion) was used for pH measurement. End-of-fermentation samples, free of the bulk of the yeast population, were stored at 4°C until analysed for headspace components and total vicinal diketones (TVD). Samples for FAN analysis were stored at 18°C. Carbohydrates were measured by high-performance liquid chromatography (HPLC), and volatile headspace components and TVD (diacetyl and 2,3-pentanedione) by headspace gas injection gas

90

BREWING YEAST FERMENTATION PERFORMANCE

chromatography (GC). FAN was measured using IOB Methods of Analysis method 2.12, which is based on a colorimetric measurement using ninhydrin.

8.3

Results and discussion

8.3.1

Fully defined medium

The fully defined medium was tested by performing parallel fermentations with ale and lager yeast in both the defined medium and wort normally associated with each strain.3 Lager yeast fermentation in the defined medium was initially faster than in the lager wort, but both sets of fermentations reached the end-point of 1008° after approximately 65 h of incubation (Fig. 8.1). The TVD concentration profiles obtained with both media were very similar (data not shown) and the viability of yeast in both sets of fermentations remained above 90% throughout the fermentation period. The suspended cell concentration in both sets of fermentations corresponded well, demonstrating satisfactory and realistic flocculation of the lager yeast in the defined medium. Ale yeast fermentation in the defined medium was slower than in the ale wort. Significantly, at the point when cooling was applied (100 h fermentation time) the gravity of the fermented ale wort and the defined medium was 1008° and 1022° respectively (Fig. 8.2). The partial attenuation of the defined medium, compared with wort, was caused by the incomplete uptake and metabolism of maltose, as apparent by the residual maltose in the medium at the end of fermentation (data not shown). Lack of nitrogen can result in the inactivation of sugar transport, and ale strains are more susceptible to this inactivation than lager strains.6 One explanation for the reduced fermentative capability of yeast in the defined medium could have been the relatively low FAN content of this medium. 1060 Defined medium

Lager wort

1050

PG (°)

1040 1030 1020 1010 1000 0

Fig. 8.1

10

20

30 40 Time (h)

50

60

70

Lager fermentation profile in the fully defined medium. PG: present gravity (degrees).

91

WORT SUBSTITUTES AND YEAST NUTRITION

1060 Defined medium

Ale wort

1050

PG (°)

1040 1030 1020 1010 1000

0

20

40

60

80

100

120

Time (h) Fig. 8.2

Table 8.4

Ale fermentation profile in the fully defined medium. PG: present gravity.

Influence of carbon source and concentration on yeast growth

Carbon source

Initial concentration (% w/v)

Final yield (crop) (g dry yeast/l)

Glucose

0.5 1 5 5 10 20 0.5 1 5 5 10 20

1.4 2.9 4.8 5.9 1.4 1.1 1.5 3.4 1.9 4.8 3.8 4.4

Maltose

Final cell concentration (106 cells/ml) 21 36 59 53 16 12 29 42 57 46 36 32

Final ethanol concentration (% v/v) 0 0 0.8 0.3 2.6 4.8 0 0.6 0.5 1.6 5.2 –

A separate study was performed using lager yeast and the fully defined medium with only one carbon source at a time.7 This study aimed to determine the carbon concentration threshold at which yeast respiration would cease in the presence of oxygen. The summary of results is shown in Table 8.4. Inhibition of respiration occurred between 1 and 5% (w/v) glucose, as evident from the formation of ethanol. Maltose had a more severe inhibitory effect on respiration than glucose with ethanol production commencing at 1% (w/v) maltose. The defined medium developed here proved useful for physiological studies of a lager strain. Although the medium mimicked wort in amino acid and carbohydrate composition it is important to bear in mind that the medium still differed from wort considerably. The defined medium is suitable for laboratory work, but scale-up would be costly and impractical.

92

BREWING YEAST FERMENTATION PERFORMANCE

Table 8.5 Experimental set-up for studying the relationship of wort original gravity (OG) and free amino nitrogen (FAN) concentration with production of volatiles

8.3.2

Fermentation

OG (°)

FAN (mg/l)

Volume of CSL added (ml)

1, 2 3, 4 5, 6 7, 8 9, 10 11, 12 13, 14 15, 16

1060 1060 1060 1060 1060 1060 1060 1060

180 150 120 90 180 150 120 90

6 5 4 3 6 5 4 3

Semi-defined medium

A semi-defined medium was developed which was cheap and relatively easy to produce in large batches using brewery equipment.4 The intention was to produce a semidefined medium that could be used in scale-up experiments even at pilot scale. Extensive medium development was performed before the semi-defined formulation was finally reached. As well as CSL, sugar molasses, a combination of CSL and sugar molasses, yeast extract and a number of commercially available yeast foods were tested for their suitability as nutritional supplements. These experiments all demonstrated CSL to be the best supplement in terms of resulting in the fastest fermentation rates (data not shown). A general problem encountered during medium development was the low buffering capacity of media composed chiefly of brewing syrup and water (data not shown). The supplements often also lacked buffering capacity and did not provide pH stabilisation for the medium. CSL, with a typical pH of approximately 2, contains lactic acid and other organic acids. Although lactic acid is a component of some German worts, in this study the removal of lactic acid was desired. Following the removal of lactic acid, buffering was provided with the addition of a small (1.0 g/l) amount of citric acid and adjustment of the pH to 5.2 with NaOH. The medium formulated as a result of the above experiments was used to investigate the independent relationship of wort original gravity (OG) and FAN concentration on ester production by a production lager strain. Fermentations were performed in duplicate, as outlined in Table 8.5. The average PG profiles of the fermentations are shown in Fig. 8.3. Fermentations at higher OG values took longer to reach completion than those starting at a lower OG. Fermentations with higher FAN concentrations reached completion sooner. The combined effect of FAN and OG was that fermentations with the lower OG and higher FAN concentrations were faster than those with high OG and low PG. The fermentations were chilled after they reached target PG, except for the fermentation with an OG of 1050° and a FAN concentration of 180 mg/l, which was stopped before reaching target PG because of time constraints. The fermented wort was analysed for n-propanol, phenylethylalcohol, isobutanol, isoamylalcohol, total higher alcohols, phenylethylacetate, isoamylacetate,

93

WORT SUBSTITUTES AND YEAST NUTRITION

60 (1060°,180 ppm)

50

(1060°,150 ppm)

PG (°)

40

(1060°,120 ppm) (1060°, 90 ppm)

30

(1050°,180 ppm) (1050°,150 ppm)

20

(1040°,180 ppm)

10 0

Fig. 8.3

(1040°,120 ppm)

0

20

40

60 80 Time (h)

100

120

140

Lager fermentation profile in the semi-defined wort. PG: present gravity.

Fig. 8.4 Total biomass response surface in semi-defined media. PG: present gravity; FAN: free amino nitrogen.

ethyloctanoate, ethylhexanoate, ethylacetate, total esters, octanoic acid, isovaleric acid, hexanoic acid, decanoic acid, butyric acid and total organic acids. Yeast biomass at the end of fermentation was also measured (Fig. 8.4). The data were analysed statistically with the response surface function in Minitab™ (release 12). Of all the components measured, a good response was observed for total esters, phenylethylalcohol, ethylacetate, phenylethylacetate, isovaleric acid and biomass production (Fig. 8.5). The biomass produced decreased slightly with an increasing OG, probably owing to the toxic effects of ethanol, but increased with higher FAN concentration. A linear model was obtained for all correlations, as demonstrated in Table 8.6. The model generated a constant, a coefficient for OG, a coefficient for FAN concentration, and a coefficient for the interaction of OG and FAN. A linear model indicated a

94

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 8.5 Total esters response surface in semi-defined media. PG: present gravity; FAN: free amino nitrogen. Table 8.6 Correlation between the formation of some volatile compounds and wort original gravity (OG) and free amino nitrogen (FAN) Analyte Ethylacetate Phenylethylacetate Total esters Phenylethylalcohol Isovaleric acid

Model

p-Valuea with respect to OG

p-Valuea with respect to FAN

Linear with interactions Linear with interactions Linear with interactions Linear (no interactions) Linear (no interactions)

0.001 0.004 0.001 0.012 0.002

0.017 0.031 0.018 0.017 0.018

a The ‘p-value’ represents the level of significance of a correlation. A p-value less than 0.05 would indicate a significant correlation. There would be a 1 in 20 chance of a correlation occurring randomly with a p-value of 0.05. This would be highly unlikely, therefore a correlation with a p-value of 0.05 would be significant and unlikely to occur by chance. A p-value of 0.001 would indicate a highly significant correlation as the chances of such a correlation occurring randomly would be 1 in 100.

proportional change in the analyte concentration with respect to each OG or FAN. Where the model indicated interaction between these two variables a curved response could be expected as a result of the combined interaction of FAN and OG. Linear models with no interaction were obtained for the production of phenylethylalcohol and isovaleric acid, but linear models with interaction were encountered for ethylacetate and total esters production. The concentration of ethylacetate and total esters was predicted to increase linearly with an increase in OG or FAN concentration; however, the model also showed that OG and FAN interact and have an additional combined influence on the concentrations of ethylacetate and total esters. A large proportion of total esters measured in the fermented medium was composed of ethylacetate and it is likely that this may have influenced the good correlation obtained between wort OG and FAN concentration, and total esters concentration. The concentration of isopentylacetate did not show a correlation with OG or FAN concentration. This is what is observed in very high-gravity brewing where a disproportionate amount of esters is produced by the yeast.

WORT SUBSTITUTES AND YEAST NUTRITION

95

The medium based on brewing syrup and CSL is economic to produce at any scale and reflects wort in carbohydrate composition. The results obtained in this study correspond well to current knowledge about ester production. This demonstrated the usefulness of the semi-defined medium as a model for mimicking brewery wort and studying yeast physiology.

8.4

Conclusions

Two media were developed for long term yeast physiology and fermentation studies. One medium was fully defined and reflected wort in terms of amino acid and carbohydrate concentration and profile. This medium was used for studying the repression of respiration in a lager yeast by glucose or maltose. The medium is suitable for smallscale studies. A semi-defined medium based on brewing syrup and CSL was developed as a substitute for wort in physiological studies of brewing yeast. This medium reflected the wort carbohydrate concentration and profile and was suitable for large-scale studies. The medium developed was used to study the production of volatile compounds during fermentation at different OG and FAN concentrations. Using statistical analysis a significant relationship was observed between the OG and FAN concentration and the production of phenylethylalcohol, isovaleric acid, ethylacetate and total esters. The formation of ethylacetate and total esters in the ‘artificial wort’ reflected the pattern known to occur in wort.

Acknowledgements The authors wish to thank the directors of Scottish Courage Brewing for permission to publish this work.

References 1. Taidi, B., Hogenberg, B., Kennedy, A.I. and Hodgson, J. (2000) Pre-treatment of pitching yeast with zinc. Tech. Q. Master Brew. Assoc. Am. 37, 431–434. 2. Nakatani, K., Takahashi, T., Nagami, K. and Kumada, J. (1984) Tech. Q. Master Brew. Assoc. Am. 21, 3. 3. Kennedy, A.I., Taidi, B., Dolan, J.L. and Hodgson, J.A. (1997) Optimisation of a fully defined medium for yeast fermentation studies. Food Technol. Biotechnol. 4, 261–266. 4. Taidi, B., Mina, M. and Hodgson, J. (2001) Development of an artificial wort for yeast fermentation studies. Poster presentation at ASBC Meeting, 24–27 June 2001, Victoria, Canada. 5. Thompson, C.C., Leedham, P.A. and Lawrence, D.R. (1973) Proc. Am. Soc. Brew. Chem., p. 137. 6. Lagunas, R. (1995) International Specialised Symposium on Yeasts, Edinburgh. 7. Taidi, B., Bathgate, F. and Hodgson, J.A. (1998) Regulation of the balance between respirative and fermentative growth of a lager yeast by different wort sugars. Proc. 5th Aviemore Conf. Malting, Brewing & Distilling, 25–28 May, pp. 279–284.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

9 Wort Supplements: From Yeast and For Yeast M. DILLEMANS, L. VAN NEDERVELDE and A. DEBOURG

Abstract Yeast performance and vitality are not constant parameters during successive fermentations, particularly on high-gravity wort. Indeed, yeast is exposed to many stresses, including osmotic, temperature and ethanol shocks. Nevertheless, if yeast has been able to adapt its metabolism to more stressful environmental conditions over the centuries, it must contain some factors allowing resistance to stress. A novel yeast peptide complex (YPC) was partially purified from yeast. It stimulated the growth and fermentation power of brewing yeast and increased ethanol tolerance considerably. The addition of this factor to wort of original gravities of 12–24° Plato enabled an increase in fermentation rate and ethanol production. It also maintained the stability of yeast performance during successive high-gravity wort fermentations. YPC increased the energetic level by stimulating glucose metabolism. Indeed, the results indicate at least three ways in which the YPC controls glycolysis in yeast: it stimulated glucose uptake, increased the intracellular concentration of fructose-2,6-biphosphate and increased pyruvate decarboxylase activity. Moreover, the addition of YPC had a stimulatory effect on many enzymes, starting from the utilisation of pyruvate, the key intermediate in anabolic process, and other important metabolic enzymes such as citrate synthase, acetyl coenzyme A carboxylase or ATPase. The results also showed that YPC exhibited an insulin-mimetic activity not only on glycolysis but also on overall metabolism, stimulating mitochondrial enzymes, confirming the crucial role of mitochondrial energy-generating systems in the biosynthesis of cell material and conferring resistance to stress under fermentation conditions. It has already been established that the level of energy is an essential parameter for inducing resistance mechanisms. Referring to results on fermentation power and yeast performance during successive high-gravity wort fermentations, it could be suggested that YPC would also improve yeast resistance to stress. Indeed, the results show that the yeast factor, when added to the culture, stimulated growth under all conditions and reduced considerably the effect of osmotic or alcoholic stress on yeast growth.

9.1 Introduction An important development in current brewing technology is the fermentation of highgravity worts and subsequent dilution to the desired product strength at the end of fermentation. Although this method is the most cost-effective, it has a negative effect on yeast performance and vitality during successive fermentations.1,2 Yeast performance and vitality are not constant parameters during successive fermentations, particularly on high-gravity wort.1 Indeed, yeast is exposed to many stresses, including osmotic and ethanol shocks, that have a negative effect on yeast performance. Nevertheless, since yeast has been able to adapt its metabolism over centuries to more stressed environmental conditions, the cell must contain some factors allowing resistance to stress. Indeed, for yeast, it has been demonstrated that, apart from heat,

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

97

other stress agents can induce the synthesis of heat shock proteins that are a prerequisite for the development of stress tolerance by yeast.3 Recently, the first results have been presented concerning the partial purification of a yeast factor (termed yeast peptide complex or YPC) that has a positive effect on yeast metabolism. The purified fraction isolated from brewing yeast by a four-step procedure was characterised as a low molecular weight water-soluble substance. Amino acid analysis revealed the presence of aspartic acid, glutamic acid, glycine and cysteine as the main amino acids. Although aromatic compounds were not detected, some ultravioletabsorbing (260 nm) chemical structure was shown to be present. Moreover, analysis by inductively coupled plasma (ICP) spectrometry indicated the presence of potassium [9 mg/g dry weight (DW)], phosphorous (1.1 mg/g DW) and sodium (0.4 mg/g DW).4 Previous work also indicated that YPC is able to stimulate the fermentation rate and improve the attenuation of the wort, and that this influence persists during successive fermentations of high-gravity worts.4 Therefore, the aim of this chapter is to present the effects of this novel peptide factor on yeast metabolism and to summarise the molecular basis of YPC’s stimulating effects on yeast fermentation performance and resistance to stresses.

9.2 Materials and methods 9.2.1

Yeast strains

The strains used in this study were Saccharomyces cerevisiae ale strain no.391 and S. cerevisiae lager strain no.353 for the fermentation experiments, and laboratory strain YF for enzymic assays, all from the collection of the Department of Brewing Sciences and Fermentation Technology of the Institut Meurice (Brussels, Belgium). 9.2.2

Fermentations

Fermentations were conducted in EBC tall tube fermentors. Industrial wort at 12° Plato was adjusted to 19° Plato with high-maltose corn syrup. The fermentations were carried out in duplicate or triplicate at 12 or 21°C with lager and ale strains, respectively. The pitching rate was 7.5 g wet weight/l (2.8 107 cell/ml) for the lager strain and 5 g wet weight/l (2 107 cell/ml) for the ale strain. In the series of consecutive fermentations, the yeast was collected at the end of the fermentation by centrifuging at 5000 rpm for 2 min; the cropped yeast was just resuspended in fresh wort and used immediately to pitch the next fermentation. Samples were taken throughout the fermentations and analysed for the following: ethanol concentration by head-space gas chromatography; viability using the methylene blue stain; wort gravity using the Anton Paar DMA46 densitometer and yeast growth by the measurement of optical density at 660 nm. 9.2.3

Measurement of glucose uptake

After 1.5 h of anaerobic incubation in potassium phosphate buffer with glucose in the presence or absence of YPC, the reaction was started with the addition of a pulse of

98

BREWING YEAST FERMENTATION PERFORMANCE

tritiated 2-deoxyglucose (5 Ci taken from a solution of 10.98 Ci/mmol). Samples (each 500 l) were removed at specified time intervals during the first 12 min, filtered on a filter (Membranefilter 0.45 m, 25 mm, Millipore) and washed with 30 ml of ice-cold water. The filters were added to scintillation liquid and radioactivity was monitored by a scintillation spectrometer (Packard). 9.2.4

Measurement of fructose-2,6-biphosphate

Samples of yeast incubated in glucose 27 mM phosphate buffer (with the necessary tested substances) were removed at specified time intervals during the first 3 h, filtered on a filter (Membranefilter 0.45 m, 45 mm; Millipore), washed rapidly with 30 ml of ice-cold water, weighted and frozen in liquid nitrogen. The extraction was done in 0.25 M Na2CO3 at 90°C as described by François et al.5 The fructose-2,6-biphosphate concentration was measured enzymically according to Bergmeyer6 and related to the cell dry weight. 9.2.5

Acidification power test

Cells were grown in glucose minimal medium until the end of exponential phase, harvested and resuspended in 0.5 M potassium phosphate buffer–glucose 1%, pH 6.4, for 1 h with or without the addition of YPC (1 mg/ml). Washed cells were then resuspended in water at 30°C for the determination of extracellular acidification. The external pH was measured with an Orion pH-meter model 720 A. Before the addition of glucose (100 mM), the pH of the cell suspension, preincubated or not with specific substances was adjusted to about 7 and a baseline was established for 10 min. H efflux from the cells after addition of glucose was measured for another 10 min at 30°C. 9.2.6

Determination of enzyme activities

Cells were grown to stationary phase on minimal glucose medium then harvested by centrifugation, and washed with 0.1 M potassium phosphate buffer, pH 6.4. The yeast was depleted of endogenous substrates by shaking in this buffer at room temperature for 1 h. The cells (2 107 cells/ml) were then centrifuged and resuspended in fresh buffer containing glucose 27 mM in the presence of YPC (1 mg/ml). After 2 h of anaerobic incubation, the yeast suspension was centrifuged and the pellet was resuspended in 0.5 M potassium phosphate buffer, pH 6.4, and crushed in a French pressure cell. The extract was centrifuged at 12 000 rpm for 20 min and the supernatant was collected. The different enzyme activities were measured according to methods described by Dillemans et al.7 9.2.7

Measurement of glycerol

Cells were grown in glucose minimal medium until the end of exponential phase, harvested and resuspended (cell density was 75 mg wet weight/ml) in glucose-minimal medium and incubated anaerobically for 2 h with or without the two specific substances YPC (1 mg/ml) and diacylglycerol (DAG, 0.4 mg/ml).

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

99

Glycerol production was determined by high-performance liquid chromatography (HPLC) using a Waters 600E system equipped with a Waters 410 refractometer detector (34°C) and fitted with a Shodex Ionpack KC-810 P precolumn and two Shodex Ionpack KC-811 columns (60°C). A solution of perchloric acid (2 ml 60%/l of milliQ water) was used as eluent at a flow rate of 0.6 ml/min. 9.2.8

Protein determination

The protein content of whole cells was assayed by a modified biuret method. A fresh 10 ml sample of the culture (2–3 g dry weight/l) was centrifuged, and the yeast was washed twice with distilled water and resuspended in 5 ml of water. The concentrate was boiled in 1 M KOH for 10 min and then cooled on ice. CuSO4  5H2O was then added to a final concentration of 25 mM. After 5 min, the mixture was centrifuged at 13 000 rpm in an Ependorf bench-top centrifuge for 2 min and the absorbance of the supernatant was read at 550 nm with bovine serum albumin as standard. 9.2.9

Lipid extraction

The washed cells (150–500 mg dry weight) were suspended in 5 ml 0.2 M KCl containing 5 mM ethylenediaminetetra-acetate (EDTA), and the lipids extracted as follows. Cells were disrupted by three passages through the French press. The yeast extract obtained was then mixed with 20 ml chloroform–methanol (1:2, v/v). After extensive stirring, 6 ml chloroform and 6 ml 0.2 M KCl containing 5 mM EDTA were added. The resulting two phases were separated by centrifugation at 6000 rpm for 2 min. The lower chloroform phase was evaporated under nitrogen and the pellet was stored in a solution of hexane at 20°C until required. Free fatty acids and ergosterol were analysed by gas chromatography using a Delsi Nermag DI 200 gas chromatograph equipped with a flame ionisation detector (360°C) and fitted with a 30 m  0.25 mm internal diameter WCOT fused silica 0.1 m coating CP-Sil-5CB apolar capillary column (Chrompack). Hydrogen was used as a carrier gas at a flow rate of 1.5 ml/min and the oven temperature programme used was 50°C, followed by an increase of 10°C/min to 340°C and of 1°C/min to 360°C. This temperature was held for 10 min. 9.2.10

Glycogen determination

Cells collected by either centrifugation or membrane filtration were suspended in 1 ml of Na2CO3 solution 0.25 M and heated for 90 min on a boiling water bath. After cooling, 0.2 ml of the well-mixed suspension was acidified with 3 M acetic acid, shifting the pH to 4.6–5.2, and the volume was made up to 1 ml with 0.2 M Na acetate buffer, pH 4.8. For enzymic hydrolysis of the glycogen, 10 l of amyloglucosidase (1.4 U) was added and the suspension was incubated for 2 h at 37°C. The samples were neutralised with buffered KOH (1 M triethanolamine–HCl:10 N KOH:1 M acetic acid, v/v/v 3:0.8:0.2) for glucose determination using Boehringer kit no.716 251.

100

BREWING YEAST FERMENTATION PERFORMANCE

9.2.11

Farnesol-induced growth inhibition

Cells were grown in glucose minimal medium until the end of exponential phase, and inoculated into fresh glucose minimal medium containing 150 M farnesol and test substances. The growth was followed by measuring the optical density at 660 nm as a function of time. 9.2.12

Effect of ethanol and osmotic pressure on growth on glucose and maltose

The growth medium used was synthetic medium (0.67% Difco yeast nitrogen base, 2% glucose or maltose) supplemented with either ethanol 8% (v/v) or sorbitol 25% (w/v). The effect of osmotic pressure or high ethanol concentration was measured by following yeast growth at 30°C in shaking flasks. The survival of yeast grown on synthetic medium in an aqueous ethanolic solution (20% v/v) at 30°C was measured by methylene blue stain. 9.2.13

Effect of ethanol and osmotic pressure on fermentation power

Cells were harvested in exponential growth phase on synthetic medium containing glucose or maltose, washed twice with 0.1 M phosphate buffer (pH 6) and resuspended in the same buffer. Yeast cells at 3  108cell/ml in the 2.5 ml final volume were anaerobically incubated at 30°C in Warburg vessels. The CO2 production rate was measured on glucose or maltose (0.5% w/v) with or without the addition of ethanol 8% (v/v) or sorbitol 25% (w/v).

9.3 9.3.1

Results and discussion Influence of yeast peptide complex on fermentation rate

As mentioned earlier, as the initial wort gravity is increased, the rate of fermentation decreases, as a result of the higher osmotic pressure and ethanol content. As shown in Fig. 9.1 for fermentations conducted in EBC tall tubes at 12 or 21°C with a lager and an ale yeast, respectively, the purified peptide factor is able to stimulate the fermentation rate and to improve the attenuation of the wort.4 These results indicate that this yeast factor has a positive effect on yeast metabolism in lager as well as in ale fermentation. It has also been shown that this influence persists during successive fermentations of high-gravity worts, as illustrated in Fig. 9.2.4 The final gravity after the third fermentation was already 3.1° Plato higher then the first one and ethanol production decreased to 5.7% (v/v). In contrast, when 2 mg/ml YPC factor was added to the wort, no significant variation of the final attenuation could be observed. To determine the capacity of the YPC factor to stimulate the fermentation rate, to achieve final gravity and ethanol production, while also improving the stability of yeast performance during successive high-gravity wort fermentations, it was important to evaluate the effect of this novel peptide factor on yeast metabolism and to identify the mode of action.

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

101

20 18

Wort gravity (°Plato)

16

Lager

14

Lager + YPC (2 mg/ml)

12

Ale

10

Ale + YPC (2 mg/ml)

8 6 4 2 0 0

2

4

6

8

10

12

14

Time (days) Fig. 9.1 Influence of yeast peptide complex (YPC) factor on the fermentation of 19° Plato worts at 12°C with a lager strain and 21°C with an ale strain.4

20 first fermentation first fermentation + YPC second fermentation second fermentation + YPC third fermentation third fermentation + YPC

18

Wort gravity (°Plato)

16 14 12 10 8 6 4 2 0 0

2

4

6 Time (days)

8

10

12

Fig. 9.2 Influence of yeast peptide complex (YPC) factor on successive ale fermentations of 19° Plato worts at 21°C.4

9.3.2

Influence of yeast peptide complex on glucose metabolism

It is well known that the glycolytic pathway in S. cerevisiae is activated by fermentable sugars at several steps. The increased fermentation efficiency observed in the presence of YPC prompted an investigation into whether YPC factor could be involved in the regulatory cascades controlling glucose metabolism.

102

BREWING YEAST FERMENTATION PERFORMANCE

[3H] deoxyglucose uptake (cpm
103)

300 control

250

YPC 200 150 100 50 0 0

5

10

15

Time (min)

Fructose-2,6-biphosphate (nmol/g dry yeast)

Fig. 9.3 Effect of the addition of yeast peptide complex (YPC, 2 mg/ml) on the uptake of [3H]deoxyglucose by Saccharomyces cerevisiae (3 107 cells/ml) in potassium phosphate buffer.8 control

8

+YPC

7 6 5 4 3 2 1 0 0

50

100

150

200

Time after addition of glucose (min) Fig. 9.4 Influence of the addition of yeast peptide complex (YPC) on the intracellular level of fructose-2,6-biphosphate in yeast cells in potassium phosphate buffer.8

Transport of glucose was studied using 2-deoxyglucose. The yeast was depleted of endogenous substrates by shaking in phosphate buffer for 16 h. Figure 9.3 shows that the YPC factor led to an increase in transport of 2-deoxyglucose by 62%.8 Moreover, as illustrated in Fig. 9.4, the addition of YPC increased the intracellular fructose-2,6-biphosphate concentration by a factor of 1.6 after about 2 h of incubation with glucose, explaining partially the observed activation of glucose utilisation.8 Since it was shown that YPC increases the energetic level, the effect of YPC on glucose-induced activation of the H-ATPase was also evaluated. Indeed, the plasma membrane H-ATPase is the major system responsible for cellular proton excretion

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

103

2.5 control (aerobic) YPC (aerobic) control (anaerobic) YPC (anaerobic) DAG (anaerobic)

µequiv.H+/g DW

2.0 1.5

2.26

1.59 1.45

1.0

0.95 0.85

0.5 0 0

5

10

15

20

25

Time (min) Fig. 9.5 Influence of yeast peptide complex (YPC, 1 mg/ml) or diacylglycerol (DAG, 0.4 mg/ml) on the yeast plasma membrane H-ATPase activity.8

in yeast. An application of the measurement of this activity is the acidification power test used in the brewing industry.9 The mechanism by which the ATPase is activated is known as a cascade of reactions: phospholipase C generates diacylglycerol from phosphatidylinositol 4,5-biphosphate, then DAG activates protein kinase C, which phosphorylates and activates the H-ATPase.10 The important role of the plasma membrane H-ATPase in wort fermentation has been illustrated in previous works. Indeed, the inhibition of plasma membrane HATPase by diethylstilbestrol drastically reduced yeast wort fermentation.11 Therefore, the release by yeast cells of H in water after the addition of 100 mM glucose was followed for the acidification power test measurements. As illustrated in Fig. 9.5, preincubations with YPC cause a 1.5–2.5-fold increase in glucose-induced activation of the plasma membrane H-ATPase, whether the preincubation is done aerobically or anaerobically.8 These results confirm that YPC is directly involved in controlling the overall metabolism via the phosphorylation transduction cascade.

9.3.3 Influence of yeast peptide complex on anabolic enzyme activities As the overall glucose metabolism is activated, some anabolic enzyme activities could be modified in the presence of YPC. It is well established that many processes required for cell synthesis take place in the mitochondria or require mitochondrial function. Most of the tricarboxylic acid (TCA) cycle as well as some enzymes involved in sterol biosynthesis and amino acid synthesis are localised inside the mitochondria (Fig. 9.6). Pyruvate is a key intermediate in anabolic processes. Its carboxylation to oxaloacetate, catalysed by pyruvate carboxylase, is an anaplerotic process for the generation of TCA cycle intermediates.

104

BREWING YEAST FERMENTATION PERFORMANCE

Glycolysis

Sugars

Pi ADP ATP

Pyruvate BIOSYNTHESIS

Pyruvate carboxylase CO2

ATP/ADP translocator

ATP

Pyruvate Pyruvate dehydrogenase

BIOSYNTHESIS

OAA

Citrate synthase

Citrate OAA

ATPase H+

Solutes

Citrate Citrate lyase

Acetyl CoA Acetyl CoA carboxylase Fatty acid synthase Lipids

Fig. 9.6 Subcellular localisation of some key anabolic enzymes. Pi: phosphate ions; ADP: adenosine diphosphate; ATP: adenosine triphosphate; OAA: oxaloacetate; acetyl CoA: acetyl coenzyme A. Table 9.1 Influence of yeast peptide complex (YPC, 1 mg/ml) on anabolic enzyme activities7 Enzyme

Pyruvate carboxylase Pyruvate dehydrogenase Citrate synthase Acetyl coenzyme A carboxylase

Enzymic activity (nmol/min per mg protein) Control

YPC (1 mg/ml)

2.7 5.2 71.0 8.2

4.6 8.1 110.0 12.7

The pyruvate dehydrogenase complex has a role in amino acid synthesis; this is further supported by a partial leucine requirement of mutants lacking pyruvate dehydrogenase activity.12 Citrate synthase is a key enzyme of the TCA cycle catalysing the condensation of oxaloacetate and acetyl coenzyme A (CoA) to produce citrate. Citrate carries acetyl groups from mitochondria to the cytosol for fatty acid synthesis. Indeed, acetyl CoA formed in mitochondria must be transferred to the cytosol, but mitochondria are not readily permeable to acetyl CoA. When present at high levels, citrate is transported to the cytosol, where it is cleaved by citrate lyase and transformed in acetyl CoA and oxaloacetate. The biosynthesis of long-chain fatty acids requires four enzymic systems, among which acetyl CoA carboxylase has been shown to be the rate-limiting step.13 Therefore, the influence of YPC on the activity of some key anabolic enzymes was tested under fermentation conditions. As shown in Table 9.1, the addition of YPC to intact S. cerevisiae cells caused a 50–90% increase in key anabolic enzymes after 2 h of anaerobic incubation in phosphate buffer.7

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

105

Table 9.2 Influence of yeast peptide complex (YPC, 1 mg/ml) and diacylglycerol (0.4 mg/ml) on glycerol production during anaerobic incubation in glucose minimal medium7

Glycerol (mg/ml)

Control

YPC (1 mg/ml)

0.023

0.200

Table 9.3 Effect of yeast peptide complex (YPC) on biomass and cell biosyntheses during high-gravity wort fermentation8 Biomass (mg DW/ml) 5 ha Control YPC 28 ha Control YPC

Free fatty acids (mg/g DW)

Ergosterol (mg/g DW)

Proteins (mg/100 g DW)

Glycogen (g/100 g DW)

2.1 2.1

2.8 5.3 (89%)b

0.19 0.93 (400%)

37.0 39.7 (7.2%)

5.1 4.3 (17%)

3.4 4.1 (19.8%)

2.5 3.2 (29%)

0.13 0.18 (35%)

39 43 (10.2%)

9.5 12.7 (33%)

a

Time after pitching. Percentage difference from control. DW: dry weight. b

9.3.4

Influence of yeast peptide complex on yeast synthesis

Under anaerobic conditions, the surplus of reducing equivalents formed in anabolic reactions is balanced by the formation of glycerol. As previous results have shown that YPC has a positive effect on cellular synthesis, it is not surprising that in the presence of YPC about eight times more glycerol was produced during anaerobic incubation in glucose minimal medium compared with the control (Table 9.2).7 Moreover, looking to the influence of YPC addition on the activities of some anabolic enzymes, it could be suggested that YPC would also improve yeast cellular synthesis. Table 9.3 summarises the lipid contents of yeast harvested from the different wort fermentations 5 and 28 h after inoculation.8 After 5 h, the yeast had synthesised important amounts of free fatty acids and ergosterol, but no cell division had occurred at this early stage of fermentation. Yeast from the wort fermentation with YPC contained much higher amounts of free fatty acids (89%) and ergosterol (400%). These results indicate that the addition of YPC strongly activates yeast free fatty acids and sterol synthesis at the beginning of fermentation. The maximal concentration of these compounds stored in yeast cells determines the extent of growth and the degree of wort fermentation.14 It could be assumed that this increased synthesis in the presence of YPC may explain the improved fermentation performances observed. Indeed, after 28 h of fermentation, the total biomass was increased by about 20%, with cells exhibiting higher lipid, protein and glycogen content, although it was intensively used to support synthesis at the beginning of fermentation.

106

BREWING YEAST FERMENTATION PERFORMANCE

ligands, hormones, growth factors Tyrosine-kinase receptors PLASMA MEMBRANE

PIP2 PIK

DAG PKC

PP PP PLC IP3

RAS RAS G T p

? LMW insulin mediators

Ca2+ Calmodulin

MAP kinases

Calmodulin kinases

Kinases C

CELLULAR ACTIVITIES & MITOGENESIS ° Cascade of protein–protein interactions ° DNA transcription & replication ° Mitogenic effects ° Transport ° Anabolic processes ° Stress tolerance Fig. 9.7 Insulin transduction pathways. GTP: guanosine triphosphate; MAP kinase: mitogen-activated protein kinase; LMW: low molecular weight; PLC: phospholipase C; PIP2: phosphatidylinositol 4,5-biphosphate; IP3: inositol 1,4,5-triphosphate; DAG: 1,2-diacylglycerol; PKC: protein kinase C.

9.3.5

Mode of action of yeast peptide complex

Looking at all of these properties, YPC factor seems to have mitogenic and anabolic insulin-like effects. It is therefore interesting to point out that the purified fraction has a structure very similar to insulin mediators, which are low molecular weight, watersoluble peptides isolated from treated insulin-sensitive mammalian cells.15 Indeed, it has been reported that insulin has an effect on fructose-2,6-biphosphate level in human fibroblasts16 and that glucose metabolism in yeast is stimulated by human insulin. Moreover, insulin has been shown to be active in brewing yeast strains.17 Recent findings suggest that an insulin-like signal transduction cascade is somehow involved in the regulation of metabolic biosynthetic pathways in yeast.18 The biological action of insulin is initiated by the binding of the hormone to its receptor tyrosine kinase on the plasma membrane. Three pathways have been identified and are thought to mediate different biological functions of the hormone, as summarised in Fig. 9.7. Among them, the activation of the specific phospholipase C that hydrolyses glycosylphosphatidylinositol lipids (PIP2) in the plasma membrane plays an important role.

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

107

Growth (OD 660 nm) 6 control

5

control + FOH

4

YPC + FOH

3

DAG + FOH

2 1 0 0

10

20

30

Time (h) Fig. 9.8 Influence of yeast peptide complex (YPC, 1 mg/ml) and diacylglycerol (DAG, 0.4 mg/ml) on the growth-inhibitory effect of farnesol (FOH).7

This activity generates two second messengers: inositol 1,4,5-triphosphate (IP3) and DAG. Hydrolysis of phosphatidylinositol-2-phosphate in S. cerevisiae is required for a number of nutritional and stress-related responses. Farnesol, an isoprenoid alcohol, induces in yeast and mammalian cells a significant decrease in the cellular DAG level that leads to a reduction in cell growth. In mammalian and yeast cells alike, the farnesol-induced inhibition of cell proliferation could be restored with exogenously added DAG.19 Therefore, the influence of YPC, compared with that of DAG, on the growth-inhibitory effect of farnesol was investigated.7 The exogenous addition of YPC was effective, like DAG, in preventing the growthinhibitory effect of farnesol (Fig. 9.8). This finding confirms that YPC’s mode of action may be involved in cellular signal transduction.

9.3.6

Influence of yeast peptide complex on ethanol and osmotic stresses of growing cells

It has already been established that the level of energy is an essential parameter for inducing resistance mechanisms. Referring to results on the activation of glucose metabolism and yeast performance during successive high-gravity wort fermentations, it could be suggested that YPC would also improve yeast resistance to stress owing to the higher energetic level. Therefore, the effect of YPC factor on yeast growth and resistance to stresses in synthetic medium was investigated. An increase in osmotic pressure was obtained by the addition of 18% (w/v) sorbitol, and 8% (v/v) ethanol concentration was also used to induce ethanol stress. The time required to reach the end of the exponential phase and the biomass obtained on glucose or maltose synthetic medium were taken as reference. As illustrated in Fig. 9.9, the yeast factor, when added to the culture at a concentration of 2 mg/ml, stimulated growth under all conditions and reduced considerably the effect of osmotic or alcoholic stress on yeast growth and fermentation power.4 Moreover, the survival pattern in an aqueous ethanol solution (20% v/v) of the ale yeast grown on glucose synthetic medium was examined to evaluate the protective

108

BREWING YEAST FERMENTATION PERFORMANCE

Glucose control 140

Glucose + YPC

Maltose control

Maltose + YPC

Growth and fermentation power stimulation of stressed cells (% of non-stressed control)

120 100 80 60 40 20 0

+sorbitol 18% (w/v)

+ethanol 8% (v/v)

+sorbitol 25 %(w/v)

Growth

+ethanol 8% (v/v)

Fermentation power

Fig. 9.9 Influence of yeast peptide complex (YPC) factor on yeast growth and resistance to ethanol or osmotic pressure in synthetic medium.4

% survival

100

control 80

+YPC (2 mg/ml)

60

40

20

0 0

50

100

150

Time (min) Fig. 9.10 Effect of yeast peptide complex (YPC) factor on the survival of ale yeast in aqueous ethanol solution (20% v/v).4

effect of the YPC factor against alcoholic stress. The survival of cells in the presence of 2 mg/ml of YPC factor confirmed its efficient protective effect (Fig. 9.10).4

References 1. Rees, E.M.R. and Stewart, G.G. (1997) The effect of divalent ions magnesium and calcium on yeast fermentation performance in conventional (12°P) and high (20°P) gravity worts in both static and shaking fermentations. Proc. Eur. Brew. Conv. Cong. 26, 461–468.

WORT SUPPLEMENTS : FROM YEAST AND FOR YEAST

109

2. Borthwick, R., Stewart, G.G., Rees, E.M.R. et al. (1997) Very high gravity fermentations with ale and lager yeast strains. Proc. Eur. Brew. Conv. Cong. 26, 493–500. 3. Mager, W.H. and Moradas Ferreira, P. (1993) Stress response in yeast. Biochem. J. 290, 1–13. 4. Dillemans, M., Van Nedervelde, L. and Debourg, A. (1999) Characterization of a novel yeast factor increasing yeast brewing performances. Proc. Eur. Brew. Conv. Cong. 27, 711–718. 5. François, J., Eraso, P. and Gancedo, C. (1987) Changes in the concentration of cAMP, fructose-2, 6-biphosphate and related metabolites and enzymes in Saccharomyces cerevisiae during growth on glucose. Eur. J. Biochem. 164, 369–373. 6. Bergmeyer, H.U. (ed.) (1983) Methods in Enzymatic Analysis, Vol. 3. VCH, Weinheim, pp. 496–501. 7. Dillemans, M., Van Nedervelde, L. and Debourg, A. (2001) Insulin-like anabolic and mitogenic activities of a yeast peptide complex on brewing yeast. Proc. Eur. Brew. Conv. Cong. 28. 8. Dillemans, M., Van Nedervelde, L. and Debourg, A. (2001) An approach to the mode of action of a novel yeast factor increasing yeast brewing performance. J. Am. Soc. Brew. Chem. 59, 101–106. 9. Mathieu Ch., van den Berg, L. and Iserentant, D. (1991) Prediction of yeast fermentation performance using the acidification power test. Proc. Eur. Brew. Conv. Cong. 23, 273–280. 10. Brandao, R.L., de Magalhaes-Rocha, N.M., Alijo, R. et al. (1994) Possible involvement of a phosphatidylinositol-type signaling pathway in glucose-induced activation of plasma membrane H-ATPase and cellular proton extrusion in yeast Saccharomyces cerevisiae. Biochim. Biophys. Acta 1223, 117–124. 11. O’Connor-Cox, E.S.C., Lodolo, E.J. and Axcell, B.C. (1993) Role of oxygen in high gravity fermentations in absence of unsaturated lipid biosynthesis. J. Am. Soc. Brew. Chem. 51, 97–107. 12. Wenzel, T.J., Van Den Berg, M.A., Visser, W. et al. (1992) Characterization of Saccharomyces cerevisiae mutants lacking the E1 a subunit of the pyruvate dehydrogenase complex. Eur. J. Biochem. 209, 697–705. 13. Schneiter R., Hitomi, M., Ivessa, A.S. et al. (1996) A yeast acetyl coenzyme A carboxylase mutant links very-long chain fatty acid synthesis to the structure and function of the nuclear membrane–pore complex. Mol. Cell. Biol. 16, 161–172. 14. Jimenez, J. and Benitez, T. (1988) Temperature and ethanol concentrations depends on the mitochondrial genome. Curr. Genet. 13, 461–469. 15. Gottschalk, W.K., Macaulay, S.L., Macaulay, J.O. et al. (1986) Characterization of mediators of insulin action. Ann. N.Y. Acad. Sci. 488, 385–405. 16. Farnararo, M., Vasta, V., Bruni, P. and D’Alessandro, A. (1984) The effect of insulin on Fru-2,6-P2 levels in human fibroblasts. FEBS Lett. 171, 117–120. 17. Lodolo, E.J., O’Connor-Cox, E.S.C. and Axcell, B.C. (1995) Novel application of glucagon and insulin to alter yeast glycogen concentrations. J. Am. Soc. Brew. Chem. 53, 145–151. 18. Müller, G., Grey, S., Jung, C. and Bandlow, W. (2000) Insulin-like signaling in yeast: modulation of protein phosphatase 2A, protein kinase A, cAMP-specific phosphodiesterase, and glycosylphosphatidylinositol-specific phospholipase C activities. Biochemistry 39, 1475–1488. 19. Machida, K., Tanaka, T., Yano, Y. et al. (1999) Farnesol-induced growth inhibition in Saccharomyces cerevisiae by a cell cycle mechanism. Microbiology 145, 293–299.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

10 Unsaturated Fatty Acid Supplementation of Stationary-phase Brewing Yeast and its Effects on Growth and Fermentation Ability N. MOONJAI, K.J. VERSTREPEN, F.R. DELVAUX, G. DERDELINCKX and H. VERACHTERT

Abstract Supplementation of stationary-phase yeast with an unsaturated fatty acid (UFA), linoleic acid, resulted in an increased intracellular UFA content. However, the glycogen level was decreased, concomitant with an increase in trehalose content compared with unsupplemented cells. The behaviour of supplemented cells during the following fermentation of a synthetic medium (0 ppm dissolved oxygen) containing 8% (w/v) glucose was investigated. The effects of supplementation of the medium with UFA were compared. Growth in the fermentation pitched with unsupplemented cells was restricted and the viability was low, resulting in a slow fermentation. Yeast growth was restored and the fermentation ability was significantly improved when supplemented pitching yeast was used. Moreover, using supplemented pitching yeast did not result in a reduction in acetate esters synthesis compared with UFA supplementation of the medium. From the results achieved under the experimental conditions used here, it suggested that the use of UFA-supplemented pitching yeast in brewery fermentations may be a convenient way to revitalise the cropped yeast used for pitching in the subsequent fermentation.

10.1 Introduction In traditional batch brewing practice, yeast is reused in successive fermentation cycles. However, the physiological condition of cropped yeast is quite poor because of the depletion of sterols and unsaturated fatty acids (UFAs). Insufficient levels lead to an alteration in membrane structure and membrane-linked biochemical processes. Therefore, the yeast taken from a previous fermentation cycle must be revitalised and given the opportunity to synthesise appropriate levels of these essential membrane compounds. To satisfy this requirement the brewer aerates the wort for a short period before pitching. This enables the lipid synthesis necessary for cell growth. However, it is difficult to control wort aeration.1,2 Another possibility for the yeast to regenerate the necessary lipids during fermentation is the uptake of some of these lipids from the surrounding medium.3,4 It has been shown that addition of lipids, especially UFAs, to wort can eliminate the requirement for wort aeration.5,6 Oxygen and UFAs promote yeast growth; however, their presence during wort fermentation drastically decreases the synthesis of volatile esters, which are extremely important for beer flavour.7–11 As a consequence, the brewer constantly has to control the oxygen and fatty acid contents of wort and their respective undesirable side-effects. A general review of the problems was published by Moonjai et al.12 This study investigated the possibility of adding UFAs to the stationary-phase yeast obtained at the end of a fermentation cycle. The effects of UFA supplementation of

UFA SUPPLEMENTATION OF STATIONARY - PHASE BREWING YEAST

111

cropped yeast, before pitching, on yeast growth and fermentation ability in the next fermentation were studied.

10.2 Materials and methods 10.2.1

Yeast strain and maintenance

All experiments were carried out with an industrial brewing strain of yeast, Saccharomyces cerevisiae carlsbergensis KUL-CMBS 12, which was maintained on wort agar (Difco Laboratories, Detroit, MI, USA) plates and stored at 4°C. 10.2.2

Growth medium

Yeast cells were grown in synthetic medium containing (per litre): 80 g of glucose (Sigma Chemical Co., St Louis, MO, USA), 6.5 g of yeast extract (Difco Laboratories), 2.6 g of (NH4)2SO4, 2.72 g of KH2PO4, 0.5 g of MgSO4 · 7H2O and 0.5 g of CaCl2 · 2H2O. Zinc (ZnCl2) was added to a final concentration of 0.2 mg/l. Polypropylene glycol 2000 (Sigma-Aldrich Chemie, Germany) (200 mg/l) was also added as an antifoam reagent. The pH was brought to 5.2 by means of citrate buffer (0.04 M) containing (per litre of medium) 1.5 g of citric acid and 6.0 g of sodium citrate. To avoid excessive Mailard reaction during autoclaving, a glucose solution was prepared and autoclaved separately. After autoclaving at 1.0 kg/cm2 and 121°C for 15 min, portions were combined while still hot. This synthetic medium was used in all propagations and fermentations. 10.2.3

Yeast propagation

A yeast colony was removed from a stock plate and inoculated onto a wort agar slant in a sterile tube. After incubation at 27°C for 48 h, the slant culture was stored at 4°C. When required, 5 ml of synthetic medium was added into the slant culture, which was then gently agitated to disperse yeast and form a suspension. The yeast suspension was inoculated into a cottonwool-stopped 250 ml Erlenmeyer flask containing 150 ml of synthetic medium. Yeast propagation was carried at 20°C for 48 h with orbital shaking at 150 rpm. Yeast cells were harvested by centrifugation at 4000 rpm for 5 min and pitched into fresh medium for a next fermentation at a rate of 15  106 cells/ml. 10.2.4

Preparation of stationary-phase cells and unsaturated fatty acid supplementation

Stationary-phase cells were prepared from a fermentation performed in an airlockstopped 1 litre Erlenmeyer flask containing 500 ml of synthetic medium. The medium was aerated to contain 8 ppm dissolved oxygen at the start of fermentation. After pitching, nitrogen gas was passed through the headspace at a flow rate of 1.5 l/min for 5 min. Fermentation was carried out at 20°C for 72 h with magnetic stirring at 150 rpm to obtain stationary-phase cells. Before harvesting, linoleic acid (Sigma Chemical Co.) was added to the culture. An aliquot solution of linoleic acid in ethanol (0.5 ml)

112

BREWING YEAST FERMENTATION PERFORMANCE

was added into the 500 ml fermentation culture to a final concentration of 60 mg/l. After supplementation with linoleic acid, the yeast culture was incubated for a further 12 h under the same conditions as during the fermentation. Control yeast culture was treated in the same way, but 0.5 ml ethanol without linoleic acid was added to this culture. Subsequently, yeast cells were harvested by centrifugation at 4000 rpm for 5 min and quickly washed twice with cold water (4°C). These cells were pitched in a fresh medium for the test fermentations at a rate of 15  106 cells/ml. The rest of the yeast pellets were weighed and stored at 20°C for further analysis. 10.2.5

Analysis of pitching yeast

After thawing, the yeast pellets were resuspended in cold (4°C) 0.8% NaCl solution to a concentration of 0.1 g wet cells/ml. Cell dry weight (CDW) was determined by drying 1 ml of this cell suspension at 105°C for 2 h. The CDW/ml of cell suspension was calculated for the determination of fatty acids, glycogen and trehalose content of yeast. To determine total fatty acid composition of yeast, i.e. free fatty acids and fatty acyl residues of major classes of lipids (acylglycerols and phospholipids) during fermentation, 0.5 g wet cells (5 ml cell suspension) was disrupted by alkaline saponification carried out at 100°C for 30 min and total fatty acids were extracted with hexane as described by Chen13 with some modification. Heptadecanoic acid (Sigma Chemical Co.) was used as an internal standard. The fatty acids were methylated by boron trifluoride in methanol (14% solution) and the fatty acid methyl esters were analysed by gas chromatography (Varian 3300; Varian Association, Walnut Creek, CA, USA) equipped with a flame ionisation detector. The capillary column was an Alltech Heliflex AT-225 (Alltech Associated, Deerfield, IL, USA) with 30 m length, 0.32 mm internal diameter and 0.25 m film thickness. Chromatography was performed under the following conditions: oven temperature was increased from 150 to 210°C at a rate of 6°C/min with 3 min holding time, injector temperature was 250°C and detector temperature was 230°C. The carrier gas was helium. The glycogen content of yeast was determined using method 4 of Quain.14 The glycogen was extracted from 0.1 g wet cells (1 ml cell suspension). The extract was incubated with 1.4 U/ml amyloglucosidase (Boehringer Mannheim, Germany) at 37°C for 2 h and the resulting glucose was determined spectrophotometrically at wavelength of 505 nm using a glucose oxidase diagnostic kit (Sigma Diagnostics, St Louis, MO, USA). Trehalose was extracted from 0.1 g wet cells (1 ml cell suspension) with cold 0.5 M trichloroacetic acid, as described by Trevelyan and Harrison.15 The extract was assayed for anthrone-positive material as described by Spiro.16 The results were compared against a glucose standard analysed spectrophotometrically at a wavelength of 625 nm. Glycogen and trehalose levels were expressed as equivalent mg glucose/g CDW. 10.2.6

Test fermentations

All test fermentations were performed in a stirred 5 litre jar fermentor (BioFlow III; New Brunswick Scientific, Edison, NJ, USA) containing 3 litres of fresh fermentation medium, and the temperature was kept constant at 15°C with stirring at 100 rpm. Before pitching, the fermentation medium was de-aerated. This medium contained 0 ppm

UFA SUPPLEMENTATION OF STATIONARY - PHASE BREWING YEAST

113

dissolved oxygen, obtained by passing filtered nitrogen gas through the medium for 10 min at 15°C, with agitation at 500 rpm. The final content of dissolved oxygen was measured by means of a dissolved oxygen meter (Oxi 340-A/SET; Weilheim, Germany). Linoleic acid (C18:2)-supplemented medium was prepared by adding a solution of C18:2 in ethanol (3 ml) into de-aerated medium to a final concentration of 15 mg/l. An equal volume of ethanol (without C18:2) was added to the other cultures. Residual oxygen was removed from the headspace by flushing the headspace with nitrogen gas at a flow rate of 1.5 l/min for 5 min after pitching and continuously flushing at a flow rate of 30 ml/min during fermentation to avoid any further entry of oxygen. At regular time intervals 100 ml samples were removed for further analysis.

10.2.7

Monitoring of fermentation

Yeast growth was monitored by measuring the concentration of cells in the fermenting medium by means of optical density at wavelength of 600 nm. Cells were also counted with the Thoma counting chamber and cell viability was assessed using the methylene blue staining technique.17 The percentage of unstained cells was recorded as the viability. Yeast cells were harvested by centrifugation at 4000 rpm for 5 min. The gravity of centrifuged fermenting medium was measured using a digital density meter (Paar DSA 48SP-1; Anton PAAR KG, Graz, Austria) to determine the apparent attenuation and the ethanol content. A 5 ml aliquot of centrifuged fermenting medium was put into a 20 ml vial, which was immediately capped and quickly frozen, after which it was analysed for volatile compounds.

10.2.8

Analysis of volatile esters and higher alcohols

Volatile compounds were determined by headspace gas chromatography (Perkin Elmer Autosystem XL) equipped with a flame ionisation detector. Samples were heated for 16 min at 60°C in the headspace autosampler (Perkin Elmer Headspace Sampler HS-40). Esters and higher alcohols were separated using a 50 m WCOT fused silica capillary column coated with CP-Wax 52CB, with 1.2 m film thickness and 0.32 mm internal diameter. The following conditions were applied: injector temperature was 180°C; initial oven temperature was 75°C and held for 6 min, and increased at 25°C/min to 110°C, held for 3.5 min; detector temperature was 250°C. The carrier gas was helium.

10.3 10.3.1

Results and discussion Unsaturated fatty acid supplementation of pitching yeast

This study investigated the effects of UFA supplementation of yeast cells before pitching. Yeast obtained at the end of a fermentation cycle was supplemented with UFA. To narrow the effects, linoleic acid (C18:2) was used throughout. Unlike some other yeast genera, Saccharomyces yeast cannot synthesise C18:2, which makes it easier to follow the uptake and cellular fate of the added fatty acid. A synthetic medium containing 8%

114

BREWING YEAST FERMENTATION PERFORMANCE

18 Unsupplemented cells 16

Supplemented cells

Fatty acids (mg/g CDW)

14 12 10 8 6 4 2 0 C16:0

C16:1

C18:0

C18:1

C18:2

UFA

Fig. 10.1 Fatty acid content of pitching yeast obtained after a 12 h incubation with or without linoleic acid (C18:2). CDW: cell dry weight.

Table 10.1 Physiological states of pitching yeast recovered at the end of a fermentation cycle with or without linoleic acid (C18:2) supplementation Pitching yeast

Unsupplemented Linoleic acid supplemented

Cellular content (mg/g CDW) in pitching yeast Total fatty acids

Glycogen

Trehalose

21.8 28.0

128.6 101.7

22.1 49.7

CDW: cell dry weight.

(w/v) glucose was used to avoid the variations in wort composition, including its C18:2 content. As the C18:2 supplementation may affect the intracellular metabolism of lipids and storage carbohydrates, the cellular fatty acid, glycogen and trehalose contents were determined. The results show that C18:2 was incorporated in yeast lipids and accounted for approximately 32% of the total UFA within 12 h (Fig. 10.1). Thus, the supplementation of cropped yeast with C18:2 resulted in an increase in the UFA content of pitching yeast from 11.3 to 16.5 mg/g CDW. The data in Table 10.1 show the cellular contents of total fatty acids, glycogen and trehalose in pitching yeast with and without C18:2 supplementation. The total fatty acid content of the supplemented cells was increased by 28% compared with that of the unsupplemented cells. The supplemented cells contained a low level of glycogen, whereas the trehalose content was clearly elevated (more than two-fold) compared with the unsupplemented cells. This may possibly be explained as a response of cells in the resting state to new (stress) conditions imposed on yeast.18 Therefore, the trehalose accumulation occurred as

UFA SUPPLEMENTATION OF STATIONARY - PHASE BREWING YEAST

115

a stress response and the carbon used for trehalose synthesis partially derived from glycogen dissimilation.19 A similar increase in trehalose at the expense of glycogen was also reported by Callaerts et al.20 when cropped yeast was oxygenated. The trehalose content of the pitching yeast had no effect on growth and fermentative ability during the subsequent fermentation; however, a high level of trehalose seemed to sustain cell viability in the first hours of fermentation.21 The level of glycogen in pitching yeast was an important consideration since glycogen was the sole source of energy for lipid synthesis during the initial stage of fermentation.22 Although this may generate a problem for the repitching of supplemented yeast, this was not the case as a high level of cellular UFA could probably compensate for the low glycogen level. 10.3.2

Fermentation with unsaturated fatty acid-supplemented yeast

To study the effects of the UFA linoleic acid, supplementation of pitching yeast on growth and fermentation ability during subsequent fermentation, three different fermentations were compared. All fermentations were pitched with yeast cells harvested from a previous standard fermentation after a further 12 h with or without C18:2 supplementation. The first fermentation was pitched with the C18:2-unsupplemented cells. The second fermentation was carried out with the same cells but the medium was supplemented with C18:2 to a final concentration of 15 mg/l. The third fermentation was pitched with yeast cells that received a 12 h incubation with C18:2 before pitching. Yeast growth was determined by following the changes in optical density of culture medium and in cell numbers during fermentation, as shown in Fig. 10.2a and b, respectively. The results show that the growth of unsupplemented cells was very limited. The maximum cell numbers reached only 37  106 cells/ml. In contrast, supplementation of the medium with linoleic acid led to a large increase of yeast growth as the maximum cell numbers reached were increased to 71  106 cells/ml. These results confirm that C18:2 supplementation to the fermentation significantly promoted yeast growth.9 Surprisingly, when C18:2-supplemented cells were pitched in unsupplemented medium, growth was restored and the maximum cell numbers reached were increased by 42% compared with fermentation with unsupplemented cells. The viability changes during fermentation are shown in Fig. 10.2c. In general, the viability recovered at the beginning of the subsequent fermentation owing to cell replication, but decreased at the end of the fermentation.23 The results of this experiment show that the viability of unsupplemented cells in unsupplemented medium remained low (90%) throughout the fermentation. In contrast, the viability was greatly improved where C18:2 was present, either in the medium or in the pitching yeast, and remained high ( 95%) until the end of the fermentation. Figure 10.3 shows the decreases in gravity of fermented medium throughout the fermentations. The fermentation with the unsupplemented yeast cells in the unsupplemented medium was very slow. It took more than 120 h to reach the maximum attenuation, corresponding to an ethanol concentration of 5.0% (v/v) (data not shown). This could be expected as the yeast cells were depleted in UFA and did not receive those lipids in the medium. Furthermore, there was no oxygen available in the medium (0 ppm dissolved oxygen), so UFA synthesis was impossible. This caused a poor yeast growth and thus a slow fermentation. By supplementation of the medium with C18:2, the

116

BREWING YEAST FERMENTATION PERFORMANCE

Optical density 600 nm (10 times diluted)

1

(a)

0.8

0.6 0.4 0.2 0

0

24

48

72

96

120

24

48

72

96

120

48 72 96 Fermentation time (h)

120

Cell numbers (million cells/ml)

80

(b)

60

40

20

0

0

110

Cell viability (% of total)

100 90 80 70 60 50

(c)

0

24

Fig. 10.2 Changes in optical density of (a) culture medium, (b) cell numbers and (c) viability during fermentation. : Unsupplemented pitching yeast; : supplemented medium; : supplemented pitching yeast.

attenuation was faster, therefore the fermentation time was reduced by approximately 47% compared with the fermentation without C18:2 supplementation. Using C18:2supplemented cells instead of unsupplemented cells in the fermentation without C18:2 supplementation resulted in a decrease in the fermentation time to approximately 75 h. When the formation of volatile compounds was monitored throughout fermentation, it was found for all fermentations that the concentrations of these compounds

UFA SUPPLEMENTATION OF STATIONARY - PHASE BREWING YEAST

117

1.05

Specific gravity

1.04

1.03

1.02

1.01

1.00

0

24

48 72 96 Fermentation time (h)

120

144

50 40 30 20 10 0 0

(a)

20

40

60

Attenuation (%)

80

100

Concentration of higher alcohols (mg/l)

Concentration of acetate esters (mg/l)

Fig. 10.3 Decreases in gravity of fermented medium during fermentation. : Unsupplemented pitching yeast; : supplemented medium; : supplemented pitching yeast.

(b)

300 250 200 150 100 50 0

0

20

40

60

80

100

Attenuation (%)

Fig. 10.4 Formation of (a) acetate esters and (b) higher alcohols, as a function of attenuation, during fermentation. : Unsupplemented pitching yeast; : supplemented medium; : supplemented pitching yeast.

were gradually increased towards the end of fermentation. Since the fermentation rate of the different fermentations was not the same, the concentrations of volatile compounds against apparent attenuation were plotted (Fig. 10.4a,b). It is clear that supplementation of the medium with 15 mg/l of C18:2 resulted in a reduction of acetate esters (ethyl acetate and isoamyl acetate). Consequently, the concentration of acetate esters measured at 80% attenuation in the C18:2-supplemented fermentation was reduced by 30% compared with unsupplemented fermentation (Table 10.2). The

118

BREWING YEAST FERMENTATION PERFORMANCE

Table 10.2 Concentration of volatile compounds in fermented medium at 80% attenuation Volatile compounds

Concentration of volatile compounds (mg/l) measured at 80% attenuation Unsupplemented pitching yeast

Ethyl acetate Isoamyl acetate Total acetate esters Propanol Isobutanol Isoamyl alcohol Total higher alcohols

Supplemented medium

Supplemented pitching yeast

38.7 3.7 42.4

26.5 2.3 28.8

36.8 4.5 41.3

16.3 62.2 152.5 231.0

16.3 45.7 118.8 180.8

19.8 47.4 140.0 207.2

concentrations of higher alcohols (propanol, isobutanol and isoamyl alcohol) were also reduced. In the case where C18:2-supplemented pitching yeast cells were used, the production of acetate esters was as high as in C18:2 unsupplemented fermentation pitched with unsupplemented cells. It seemed likely that the C18:2 content of supplemented pitching yeast did not affect the synthesis of esters during fermentation. This is explained by the lower initial intracellular content in C18:2, owing to the supplementation of the pitching yeast occurring in the resting state in an exhausted fermentation medium. This explanation was excluded where the pitching yeast cells were enriched by C18:2 supplementation during propagation.11 10.4

Conclusions

The possibility was investigated of restoring the yeast cell membrane optimal composition through supplementation of pitching yeast with a UFA as an alternative to wort oxygenation. In particular, the effects on acetate ester synthesis were considered. The results indicate that supplementation of cropped yeast with linoleic acid, before pitching, may be a convenient way to improve the physiological conditions of pitching yeast without affecting the yeast growth and the fermentation rate, while the production of volatile compounds is enhanced. This technique may therefore offer a valuable alternative for the current practice of wort aeration and fatty acid supplementation. In these experiments a synthetic complex medium was used and further research is needed using wort. References 1. Searle, B.A. and Kirsop, B.H. (1979) Sugar utilization by a brewing yeast in relation to the growth and maintenance phase of metabolism. J. Inst. Brew. 85, 342–345. 2. Boulton, C.A. and Quain, D.E. (1987) Yeast, oxygen and the control of brewery fermentations. Proc. Eur. Brew. Conv. Cong. 21, 401–408. 3. Chen, E.C.H. (1980) Utilization of wort fatty acids by yeast during fermentation. J. Am. Soc. Brew. Chem. 38, 148–153.

UFA SUPPLEMENTATION OF STATIONARY - PHASE BREWING YEAST

119

4. DeVries, K. (1990) Determination of free fatty acids in wort and beer. J. Am. Soc. Brew. Chem. 48, 13–17. 5. David, M.H. and Kirsop, B.H. (1972) The varied response of brewing yeasts to oxygen and sterol treatment. J. Am. Soc. Brew. Chem. 30, 14–16. 6. Taylor, G.T., Thurston, P.A. and Kirsop, B.H. (1979) The influence of lipids derived from malt spent grains on yeast metabolism and fermentation. J. Inst. Brew. 85, 219–227. 7. Dufour, J.P. and Malcorp, M. (1994) Ester synthesis during fermentation: enzyme characterization and modulation mechanism. Proc. 4th Aviemore Conference on Malting, Brewing and Distilling, Institute of Brewing, Aviemore, pp. 137–151. 8. Fujji, T., Kobayashi, O., Yoshimoto, H. et al. (1997) Effects of aeration and unsaturated fatty acids on expression of the Saccharomyces cerevisiae alcohol acetyltransferase gene. J. Appl. Environ. Microbiol. 63, 910–915. 9. Quain, D.E. (1988) Studies on yeast physiology – impact on fermentation performance and product quality. J. Inst. Brew. 95, 315–323. 10. Thurston, P.A., Quain, D.E. and Tubb, R.S. (1982) Lipid metabolism and the regulation of volatile ester synthesis in Saccharomyces cerevisiae. J. Inst. Brew. 88, 90–94. 11. Thurston, P.A., Taylor, R. and Ahvenainen, J. (1981) Effects of linoleic acid supplements on the synthesis by yeast of lipids and acetate esters. J. Inst. Brew. 87, 92–95. 12. Moonjai, N., Delvaux, F.R., Derdelinckx, G. and Verachtert, H. (2000) Unsaturated fatty acid supplementation of stationary phase brewing yeast. Cerevisia 3, 37–50. 13. Chen, E.C.-H. (1981) Fatty acid profiles of some cultured and wild yeasts in brewery. J. Am. Soc. Brew. Chem. 39, 117–124. 14. Quain, D.E. (1981) The determination of glycogen in yeasts. J. Inst. Brew. 87, 289–291. 15. Trevelyan, W.E. and Harrison, J.S. (1956) Studies on yeast metabolism. J. Biochem. 63, 23–33. 16. Spiro, R.G. (1966) Analysis of sugar found in glycoprotiens. Methods Enzymol. 8, 3–5. 17. EBC Analytica Microbiologica, Method 3.2.1.1 (1992). 18. Lillie, S.H. and Pringle, J.R. (1980) Reserve carbohydrate metabolism in S. cerevisiae: responses to nutrient limitation. J. Bacteriol. 143, 1384–1394. 19. Wiemken, A. (1990) Trehalose in yeast: stress protectant rather than reserve carbohydrate. Antonie van Leeuwenhoek 58, 209–217. 20. Callaerts, G., Iserentant, D. and Verachtert, H. (1993) Relationship between trehalose and sterol accumulation during oxygenation of cropped yeast. J. Am. Soc. Brew. Chem. 51, 75–77. 21. Guldfeldt, L.U. and Arneborg, N. (1998) The effect of yeast trehalose content at pitching on fermentation performance during brewing fermentations. J. Inst. Brew. 104, 37–39. 22. Quain, D.E. and Tubb, R.S. (1982) The important of glycogen in brewing yeast. Master Brew. Assoc. Am. Tech. Q. 19, 29–33. 23. Cunningham, S. and Stewart, G.G. (2000) Acid washing and serial repitching a brewing ale strain of Saccharomyces cerevisiae in high gravity wort and the role of wort oxygenation conditions. J. Inst. Brew. 106, 389–402.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

11

Impact of Wort Composition on Flocculation B. AXCELL

Abstract Traditionally, wort has been regarded as a source of nutrients for yeasts. However, since the late 1950s several brewers have reported that some worts contain substances that can affect yeast flocculation, particularly by causing this to happen prematurely before fermentation is complete. More recent work has shown that the situation may be more complex than originally thought and that actual damage to the yeast membrane may also occur, which then interferes with the uptake of sugars. The lack of attenuation often seen in such worts may therefore be due to factors modifying the cell membrane as opposed simply to causing the yeast to drop out of suspension. The impact of the malting and brewing process on wort composition is reviewed and a hypothesis for premature flocculation is proposed.

11.1 Introduction Wort is a complex ‘soup’ of carbohydrates, proteins, lipids and their various degradation products, together with polyphenols, hop compounds, and a variety of organic and inorganic molecules. Generally speaking, this spectrum of compounds (plus a little oxygen) will provide yeast with all the nutrients it needs to grow and convert sugar to alcohol. Some brewers believe that zinc and some other trace metals may be limiting and will add supplements before pitching the yeast. In general, the pattern of fermentation is predictable for a given wort composition, and the alcohol yield and yeast growth patterns follow expected trends. Occasionally, however, the yeast growth and flocculation patterns show abnormal characteristics and the amount of carbohydrate converted to alcohol is less than expected.1,2 The brewer often reacts to this scenario by assuming that there is a deficiency in the wort caused by a change in the malt that has been processed. Zinc3 and the combination of zinc and manganese are essential for efficient yeast fermentation. Biotin4 is also an essential cofactor and biotin-deficient worts have resulted in very poor yeast growth. Oxygen deficiencies can also give rise to slow and incomplete fermentation. Another possibility is that the yeast used was of low vitality for some reason. However, rather than a deficiency, several compounds present in wort have also been shown to produce tailing fermentations or impact on yeast flocculation patterns. For example, furfural and hydroxymethyl furfural5 are known to reduce fermentation rates, possibly by acting as sugar mimics and blocking the sites of sugar uptake. In the late 1950s, researchers at Kirin in Japan started to report on the impact of several substances that caused premature flocculation of their yeast.6,7 Kudo6 found that an acid hydrolysate of spent grain that he termed ‘Barmigen’ produced early flocculation of the yeast. Preliminary work indicated that Barmigen was a type of humic acid. The same author7 also isolated a substance called ‘Treberin’ from Japanese six-row barley malt that had a similar effect on yeast. Acid hydrolysis of this substance yielded glucose, xylose and arabinose, indicating that treberin was a gum-like polysaccharide.

IMPACT OF WORT COMPOSITION ON FLOCCULATION

121

In 1975, Morimoto et al.8 isolated an arabinoxylan–protein complex from wort, which produced similar effects. Fujii and Horie9 isolated a factor from wort in the same year, which caused premature flocculation and, when analysed, was shown to be an acidic polysaccharide containing protein. The fact that treatment with Pronase caused loss of activity suggested for the first time that protein played an essential role in this factor. In the late 1970s some of the breweries of South African breweries also experienced problems with the yeast flocculating prematurely and tailing fermentations.1 A factor was isolated from malt husk that was found to generate these problem fermentations when added back to a ‘normal’ all-malt wort in 2 litre laboratory EBC fermentations. This factor appeared to be produced in the steeping process and was very heat stable as well as resistant to wide changes in pH. Further work in this area10,11 identified a high molecular weight polysaccharide containing arabinose, xylose, glucose, mannose, galactose and rhamnose, together with an acidic sugar component. It was postulated through immunoelectron microscopy that this polysaccharide binds to the yeast cell surface and produces its effect by cross-bridging adjacent cells. The binding with the yeast cell was suggested to be via lectin-like proteins, which act as cell-surface receptors. N-Acetylglucosamine was shown to inhibit premature flocculation through hapten-type inhibition and it was possible to unbind the polysaccharide factor from flocculated yeast cells by incubating them with a solution containing glucose, mannose and N-acetylglucosamine. More recently, Axcell et al.2 proposed a hypothesis for premature flocculation invoking the presence of antimicrobial peptides in wort. It was postulated that antimicrobial peptides (4–10 kDa), because of their cationic nature, would bind not only to yeast, but also to high molecular weight polysaccharides which would then crossreact with other yeast cells, generating flocs and causing premature flocculation. The low molecular weights of the peptides would have made them difficult to detect in the presence of high molecular weight polysaccharides, which may explain the focus on polysaccharides in many of the previous papers.

11.2

Molecular mechanism of yeast flocculation

Brewing yeast, Saccharomyces cerevisiae, flocculates spontaneously at the end of fermentation. For bottom-fermenting strains, the flocs settle to the bottom of the fermenter and the majority of cells can then be easily removed from the fermented wort. Flocculation is a reversible, active aggregation of cells into flocs and its timing in the fermentation process is important to achieve good beer quality. Aggregation of microorganisms is not limited to yeast but also occurs among a wide variety of bacteria, filamentous fungi and algae.12 Premature flocculation hampers complete fermentation of the growth medium, whereas failure of the cells to flocculate at the end of the fermentation process necessitates the use of centrifugation or other separation techniques to remove the cells. Early theories on flocculation supposed that cells behaved as negatively charged colloids.13 Subsequent theories postulated calcium bridging, whereby calcium ions linked adjacent cells by coupling to carboxyl groups.14 However, the inhibition of flocculation by specific sugars such as mannose could not be explained by these theories.

122

BREWING YEAST FERMENTATION PERFORMANCE

This led Miki et al.15 to propose a lectin-like theory of flocculation. In this scenario, specific lectin-like components of the cell wall recognise and adhere to mannan on an adjoining cell. Calcium ions were relegated to acting as cofactors to activate the lectins. Cell-surface hydrophobicity has been implicated as being a determinant in flocculence, as this generally increases shortly before the onset of flocculation.16 Treatment of flocculent cells with proteases or shearing forces results in cells that are less hydrophobic and non-flocculent. This ‘hydrophobic factor’ has been isolated and partially characterised, and shown to be a heat-stable protein.17 According to Straver et al., initiation of flocculence under brewing conditions appeared to be triggered after growth limitation by oxygen supply at pitching.18 However, they also suggested that other nutrient limitations could bring about the onset of flocculation. Three molecular/structural factors were found to determine the onset of flocculation: first, the appearance of fimbriae-like structures on the cell surface;19–21 secondly, the synthesis of a flocculin, encoded by a gene homologous to the dominant flocculation gene FL01;17 and thirdly, the release of a mannose-specific agglutinin (lectin) from the cells.22 According to Straver,23 the mannose-specific lectin was found in cell walls of both flocculent and non-flocculent cells, which suggests that its synthesis is not correlated with the initiation of flocculation. Flocculent brewing yeast cells produce a flocculin protein that appears to be associated with fimbriae-like structures but is not an integral part of them. The flocculin did not show agglutination activity, but was susceptible to protease activity. The molecular weight of this glycosylated protein is estimated to be over 400 kDa and to contain 63.5% sugar (mannose, glucose and N-acetylglucosamine). The flocculin possibly acts as an agglutinin ligand. Straver23 proposed the following model for flocculation of brewing yeast (Fig. 11.1). After growth limitation, yeast cells become fimbriated, corresponding with a sharp increase in cell surface hydrophobicity. Agglutinin is released, which gives rise to fimbriae-associated glutinin ligands, and flocs are formed. If agitation is carried out, removal and redistribution of the fimbriae may lead to more compact flocs.

Nutrient Fluffy flocs

Limitation

Lectins

Agitation

Fimbriae Compact flocs Yeast

Fig. 11.1 Model for flocculation of brewing yeast cells during fermentation. (Adapted from Straver et al.23)

123

IMPACT OF WORT COMPOSITION ON FLOCCULATION

However, Speers and co-workers24 suggested that controversy still exists as to the mechanism of yeast flocculation. Javadekar et al.25 claim to have purified a zymolectin to homogeneity from the cell walls of a highly flocculent strain of S. cerevisiae that has a molecular mass of 40 kDa and a pI of 4.0 and contains 44% hydrophobic amino acids. The N-terminal sequence of up to 10 amino acids showed at least 70% homology with the predicted N-terminal sequence of the putative FL01 as well as FL05 gene products. This lectin had a high affinity towards a branched trisaccharide of mannose.

11.3 Premature flocculation and beer quality The production of good-quality beer requires consistent fermentation performance and the timing of yeast flocculation is crucial. When premature flocculation does occur it generally creates low levels of attenuation and produces beers with higher levels of residual fermentable sugars,2 which can result in substantial financial losses for the brewery. Brand identity may be compromised, resulting in consumer reaction. Erratic yeast flocculation may often lead to flavour, microbiological or filtration problems. Malt has generally been implicated when premature flocculation occurs. Standard methods of malt analysis are unable to predict such fermentation problems and these are normally highlighted using a fermentability test.26 Such tests are usually based on all-malt systems, and rely on recording yeast counts and drop in gravity of the wort. The problem with fermentability tests is that they take over a week to carry out and often the results do not predict performance in a brewery. However, the results can give information about potential problem malts. In Fig. 11.2, PYF refers to a fermentation pattern that exhibits premature yeast flocculation. The control is a fermentation producing a typically normal pattern of yeast growth and flocculation. In some situations, for example, where high adjunct levels are used, the classic premature flocculation patterns may not be seen. High residual

70

14 Control 12

PYF

50

10

40

8

30

6

20

4

10

2

0

0 0

Fig. 11.2

1

2

3 4 5 6 Fermentation period (days)

7

Typical fermentability patterns. PYF: premature yeast flocculation.

8

Gravity (°P)

Yeast count (
106/ml)

60

124

BREWING YEAST FERMENTATION PERFORMANCE

sugar concentrations are an indication of problematic malt and this is recorded in the fermentability test as high residual gravities at the end of fermentation. Commercially, such malts produce beers that have high real extract values and poor levels of alcohol conversion. If sucrose is used as the adjunct, this may leave residual fructose in the beer and produce beer with a significant sweet note. Typically, rousing the yeast does not alleviate the problem, although the introduction of fresh yeast will often result in continued fermentation. Previous observations2 have shown that in a commercial situation, some breweries are more prone to these fermentation problems than others. Breweries that run bright worts would appear to be more susceptible than these running cloudier worts. The reduced fermentability is not just due to a lack of nucleation sites in the clearer worts, as the addition of protein-based ‘yeast foods’ did not improve matters.27 If dextrose adjuncts are used, residual glucose is not necessarily a problem, but maltose and maltrotriose can be found in the final beer at excessive levels. One question arising out of the above observation is whether or not premature flocculation is always linked to these low-attenuation fermentations. When premature flocculation does occur, the fermented wort always has a higher residual gravity than normal. This may be due to there being too little yeast left in suspension to finish the fermentation or because sugar uptake is impaired. In those cases where the yeast does not prematurely flocculate, it would seem that sugar uptake is definitely impaired and requires fresh yeast addition to finalise the fermentation.

11.4 The antimicrobial peptide hypothesis In 2000, Axcell et al.2 suggested that the phenomenon of tailing fermentations and premature flocculation may be linked and that the prime causative agents may be antimicrobial peptides of the lipid transfer protein (LTP) or thionin type. As far back as 1970,28 Okado et al. isolated a substance from wheat and barley that was toxic to yeast. In a second paper29 they demonstrated that although this substance was toxic at 4 ppm, it was able to inhibit the uptake of glucose when present at one-tenth of this concentration without causing death of the cell. This ‘toxin’ had a molecular weight of 9.8 kDa and an isoelectric point of greater than 10. It was identified as a peptide that was resistant to proteolysis and heat, but that its effects on yeast could be reduced by the presence of divalent ions such as calcium. These attributes are general properties of some antimicrobial peptide groups such as the thionins and non-specific LTPs. Non-specific LTPs in plants are a family of homologous peptides that have molecular weights between 9 and 10 kDa. Within the brewing literature, LTP is perhaps best known for its putative role in producing a stable foam on beer.30,31 The name LTP was originally proposed as these polypeptides were able to transfer different kinds of lipid between liposomes and mitochondria in vitro.32 More recently, Molina et al.33 and Terras et al.34 demonstrated that LTP2, LTP3 and LTP4 can exhibit potent antimicrobial activity against a range of bacterial and fungal pathogens. Another group of small cysteine-rich basic polypeptides with antimicrobial activity that can be inhibited by calcium ions is the thionins. These have molecular weights in the 4–5 kDa range. Okada’s toxin could therefore be an LTP type of polypeptide or a thionin dimer. Thionin dimers have been shown to bind to yeast through an electrostatic action with

IMPACT OF WORT COMPOSITION ON FLOCCULATION

125

the negatively charged membrane phospholipids.35 Their strong cationic and amphipathic character allows them to interact between hydrophobic aliphatic acyl chains and the polar head groups in contact with the aqueous environment and cause disruption of the membrane function. As stated earlier, much of the initial work carried out on premature flocculation highlighted the role of complex carbohydrates.6–11 These were produced either from extraction of spent grains or by washing yeast that had prematurely flocculated with -methylmannoside. These polysaccharides contained a variety of sugars (xylose, arabinose, glucose, mannose, galactose, rhamnose), but also often contained acidic groups such as uronic acid moieties. However, antibodies produced to these carbohydrates that were isolated in the laboratory failed to discriminate between control malts and those causing premature flocculation, suggesting that these polysaccharides were present in both types of malt (W. Vundla, South African Breweries, personal communication, 2001) at similar levels. Fujii and Horie9 found that an acidic polysaccharide containing protein was responsible for premature flocculation in their breweries and that after treatment with pronase this activity was lost. The diverse nature of the sugars reported in these complexes, however, suggests that they are not specific polysaccharide components derived from the malt but rather a mixture of sugars that have associated together. The research of Fujii and Horie also indicates the essential role of the protein component and would suggest that carbohydrate alone cannot bring about premature flocculation. Earlier work1,10,11 focused on carbohydrate, but there were always traces of protein present, which were interpreted at the time as being contaminants. More recent work2 has concentrated on cationic polypeptides and their potential involvement in abnormal fermentations. Samples of whole malt exhibiting premature flocculation can be washed in distilled water; when this water extract is added back to a malt exhibiting normal fermentation patterns, premature flocculation will result. The factor(s) present in this aqueous extract are heat stable and resistant to a wide range of pH. They have been shown to be antifungal when assayed against Penicillium, and to contain LTPs using an antibody capture enzyme-linked immunosorbent assay (ELISA) system with rabbit anti-LTP1 polyclonal antibodies.2

11.5

Possible mechanism for premature flocculation

Early work by Okada had already demonstrated that the binding of his ‘toxin’ was to negative charges on the yeast, which could be alleviated by high levels of calcium ions. He furthermore demonstrated that flocculent yeast was more susceptible than nonflocculent yeast to this peptide. As the mannose-specific lectin appears to be produced by both flocculent and non-flocculent yeast, this would not seem to be important in mediating the binding. The hydrophobic factor or the flocculin could, however, be involved. As most antimicrobial peptides isolated from plants are cationic and amphipathic it is likely that they would act in a similar manner to Okada’s toxin. They appear to change and disrupt membrane integrity, leading to impairment of sugar uptake and resulting in leakage of cell constituents. However, the present author’s results would support the view that while these peptides, when present, can cause poor attenuation,

126

BREWING YEAST FERMENTATION PERFORMANCE

Wort

Acidic complex carbohydrates

Antimicrobial peptides Poor attenuation

High-dextrose worts Compact flocs Lectins Fimbriae Carbohydrate Antimicrobial peptides

Normal fimbriated yeast

Fig. 11.3

Yeast

Possible mechanism for premature flocculation involving antimicrobial peptides.

they do not necessarily give rise to premature flocculation. This is possibly where the large molecular weight polysaccharides play a role. Acidic residues on these carbohydrates can bind to the cationic peptides and act as ‘pseudo fimbriae’ which then cross-link with other yeast cells, giving rise to premature flocculation. The high molecular weight polysaccharides may be natural material associated with the husk or result from the degradation by bacteria or fungi of the external tissues of the barley. Premature flocculation using this model is then a secondary impact of antimicrobial peptides and not due to their primary action. This may explain the phenomenon in high-dextrose worts where substantial quantities of maltose and maltotriose remain in the fermented wort. Under these conditions, premature yeast flocculation is not normally observed, but worts of low fermentability are produced. Perhaps the residual sugars are blocking the lectins and preventing cell-to-cell aggregation. However, as these fermentations do not attenuate properly, antimicrobial peptides may still be binding to the yeast and interfering with sugar uptake (Fig. 11.3).

11.6

Conclusions

Despite more and more items appearing on modern malt specifications, they are generally inadequate in predicting potential fermentation problems. In general, these problems manifest themselves as either poor attenuation or poor attenuation coupled with abnormal yeast flocculation patterns. The occurrence of these problem fermentations is particularly frustrating for the brewer as they can have major production and quality implications. Although extensive research has been carried out to try to understand the factors in wort that cause these abnormal fermentations, their origins have remained a mystery. It is proposed that under certain conditions such as in wet harvests, or with particular malting regimens, barley produces compounds to protect

IMPACT OF WORT COMPOSITION ON FLOCCULATION

127

itself against microbial attack. These compounds are generally small polypeptides that are heat and pH stable and are located on the outside of the grain. In brewing, they are extracted into the wort, survive the wort boiling stage and remain active during fermentation. The antimicrobial peptides are strongly cationic and amphipathic and will bind to negative groups on the yeast, eventually disrupting membrane integrity and interfering with sugar uptake. Complex, high molecular weight carbohydrates present in the wort and derived from the malt can bind to these polypeptides and form ‘pseudo-fimbriae’, which can then react with lectins on other yeast cells, forming compact flocs and bringing about premature flocculation. Currently, this remains just a hypothesis and further work is required to verify this proposal. However, this hypothesis explains most of the observations reported previously in the literature and goes some way towards explaining why both the malting process and the brewing process may minimise or accentuate these problems. For example, one malting plant may provide more anaerobic conditions than another, and this may lead to the rapid growth of certain microorganisms and generate a response by the germinating barley. In another example, wort produced in one brewery may contain more lipid material than one from another location and this lipid may then be able to ‘titrate’ out the antimicrobial peptides so that they cannot subsequently bind to the yeast. Wort, therefore, cannot simply be looked at as a medium containing appropriate nutrients for the fermentation of yeast. It is a very complex liquid and may contain many compounds that can influence fermentation and flocculation. Within the brewing industry, very few brewers measure much more than wort colour, pH, bitterness and gravity. Ultimately, to improve the prediction of fermentation performance, it may be necessary to define wort composition more accurately and have a few more key definitive measures in place.

References 1. Axcell, B.C., Tulej, R. and Mulder, C.J. (1986) The influence of the malting process on malt fermentability performance. Proc. 19th Conv. Inst. Brew., Aust. N.Z. Sect. pp. 163–169. 2. Axcell, B.C., van Nierop, S. and Vundla, W. (2000) Malt induced premature yeast flocculation. Tech. Q. Master Brew. Assoc. Am. 37, 501–504. 3. Ault, R.G. and Whitehouse, A.G.R. (1952) Determination of zinc in beer and brewing materials. J. Inst. Brew. 58, 136–139. 4. McLeod, A.M. (1979) The Physiology of Malting in Brewing Science, Vol. 1, Pollack, J.R.A. (ed.). Academic Press, London. 5. Ingram, M., Mossel, D.A.A. and de Lange, P. (1955) Factors, produced in sugar-acid browning reactions, which inhibit fermentation. Chem. Ind. 63–64. 6. Kudo, S. (1958) Studies on yeast flocculation. Rep. Res. Lab. Kirin Brew. Co. Ltd 1, 47–51. 7. Kudo, S. and Kijima, M. (1960) Studies of yeast flocculation. Rep. Res. Lab. Kirin Brew. Co. Ltd 3, 33–37. 8. Morimoto, K., Shimazu, T., Fujii, T. and Horie, Y. (1975) Some substances in malt inducing early flocculation of yeast. Rep. Res. Lab. Kirin Brew. Co. Ltd 18, 63–74. 9. Fujii, T. and Horie, Y. (1975) Some substances in malt inducing early flocculation of yeast, Part 2. Rep. Res. Lab. Kirin Brew. Co. Ltd 18, 75–85. 10. Herrera, V.E. and Axcell, B.C. (1991) Induction of premature yeast flocculation by a polysaccharide fraction isolated from malt husk. J. Inst. Brew. 97, 359–366. 11. Herrera, V.E. and Axcell, B.C. (1991) Studies on the binding between yeast and a malt polysaccharide that induces heavy yeast flocculation. J. Inst. Brew. 97, 367–373. 12. Straver, M.H., Kijne, J.W. and Smit, G. (1993) Cause and control of flocculation in yeast. Trends Biotechnol. 11, 228–232.

128

BREWING YEAST FERMENTATION PERFORMANCE

13. Kruyt, H.R. (1952) In: Colloid Science, Vol. 1, Kruyt, H.R. (ed.). Elsevier, Amsterdam, pp. 1–57. 14. Mill, P.J. (1964) Nature of the interactions between flocculent cells in Saccharomyces cerevisiae. J. Gen. Microbiol. 35, 61–68. 15. Miki, B.L.A., Poon, N.H., James, A.P. and Seligy, V.L. (1982) Possible mechanisms for flocculation interaction governed by gene FL01 in Saccharomyces cerevisiae. J. Bacteriol. 150, 887–889. 16. Smit, G., Straver, M.H., Lugtenberg, B.J.J. and Kijne, J.W. (1992) Flocculence of Saccharomyces cerevisiae cells is induced by nutrient limitation, with cell surface hydrophobicity as a major determinant. Appl. Environ. Microbiol. 58, 3709–3714. 17. Straver, M.H., Smit, G. and Kijne, J.W. (1994) Purification and partial characterisation of a flocculin from brewers yeast. Appl. Environ. Microbiol. 60, 2754–2758. 18. Straver, M.H., Smit, G. and Kijne, J.W. (1993) Determinants of flocculence of brewers yeast during fermentation in wort. Yeast 9, 527–532. 19. Day, A.W., Poon, N.H. and Stewart, G.G. (1975) Fungal fimbriae, III, The effect on flocculation in Saccharomyces cerevisiae. Can. J. Microbiol. 21, 558–564. 20. Stewart, G.G. (1981) The genetic manipulation of industrial yeast strains. Can. J. Microbiol. 27, 973–990. 21. Straver, M.H., Smit, G. and Kijne, J.W. (1994) Induced cell surface hydrophobicity influences flocculation of brewer’s yeast in a flocculation assay. In: Colloids Surfaces B: Biointerfaces. Elsevier, Amsterdam. 22. Straver, M.H., Traas, V.M., Smit, G. and Kijne, J.W. (1994) Isolation and partial purification of mannose – specific agglutinin from brewers yeast involved in flocculation. Yeast 10, 1183–1193. 23. Straver, M.H. (1993) Molecular mechanism of yeast flocculation. PhD Thesis, State University of Leiden, The Netherlands. 24. Jin, Y.-L., Ritcey, L.L., Speers, R.A. and Dolphin, P.J. (2001) Effect of cell surface hydrophobicity, charge, and zymolectin density on the flocculation of Saccharomyces cerevisiae. J. Am. Soc. Brew. Chem. 59, 1–9. 25. Javadekar, V.S., Sivaraman, H., Sainkar, S.R. and Khan, M.I. (2000) A mannose-binding protein from the cell surface of flocculent Saccharomyces cerevisiae (NCIM 3528): its role in flocculation. Yeast 16, 99–110. 26. Kruger, L., Ryder, D.S., Alcock, C. and Murray, J. (1982) Malt quality. Prediction of malt fermentability, Part 1. Tech. Qu. Master Brew. Assoc. Am. 19, 45–51. 27. Axcell, B.C., Kruger, L. and Allan, G. (1988) Some investigative studies with yeast foods. Proc. 20th Conv. Inst. Brew., Aust. N.Z. Sect., Brisbane, pp. 201–209. 28. Okada, T., Yoshizumi, H. and Terashima, Y. (1970) A lethal toxic substance for brewing yeast in wheat and barley, Part I. Agric. Biol. Chem. 34, 1084–1088. 29. Okada, T. and Yoshizumi, H. (1970) A lethal toxic substance for brewing yeast in wheat and barley, Part II. Agric. Biol. Chem. 34, 1089–1094. 30. Sorensen, S.B., Bech, L.M., Muldbjerg, M. et al. (1993) Barley lipid transfer protein I is involved in beer foam formation. Tech. Q. Master Brew. Assoc. Am. 30, 136–145. 31. Lusk, L.T., Goldstein, H. and Ryder, D. (1995) Independent role of beer proteins, melanoidins and polysaccharides in foam formation. J. Am. Soc. Brew. Chem. 53, 93–103. 32. Molina, A. and Garcia-Olmedo, F. (1993) Developmental and pathogen-induced expression of three barley genes encoding lipid transfer proteins. Plant J. 4, 983–991. 33. Molina, A., Segura, A. and Garcia-Olmedo, F. (1993) Lipid transfer proteins (nsLTPs) from barley and maize leaves are potent inhibitors of bacterial and fungal plant pathogens. FEBS Lett. 316, 199–122. 34. Terras, F.R.G., Schoofs, H., De Bolle, M.F.C. et al. (1992) Analysis of two novel classes of antifungal proteins from radish (Raphanus sativus L.) seeds. J. Biol. Chem. 267, 15301–15309. 35. Therissen, K., Ghazi, A., De Samblanx, G.W. et al. (1996) Fungal membrane responses induced by plant defensins and thionins. J. Biol. Chem. 271, 15018–15025.

Part 4 Yeast Quality Maintenance and Assessment

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

12 Management of Multi-strain, Multi-site Yeast Storage and Supply A.I. KENNEDY, B. TAIDI, A. AITCHISON and X. GREEN

Abstract Amalgamation of brewing companies and the continued increase in national and international franchise/contract brewing operations have resulted in brewing companies and individual breweries handling an ever-increasing number of yeast strains. Genetically stable, aseptic and confidential storage of a master culture of each yeast strain is essential. It is common practice for brewery yeast cultures to be used to pitch fermentations for a maximum of 10 times, after which each pitching yeast slurry at each brewery is replaced by a new culture of yeast. Centrally held master cultures are used to supply fresh yeast for brewery propagation and subsequent fermentations. The method of choice for long-term yeast strain storage is using liquid nitrogen, a method that has been employed by Scottish Courage Brewing Ltd since 1983. Master cultures were selected from brewery yeast populations after performing a range of microbiological, biochemical and fermentation tests on the isolates. Franchise partners have supplied other yeast strains. This presentation will describe the ‘cascade’ system used by Scottish Courage for the storage of its brewing yeast strains. The master cultures are held in straws placed in close proximity to or within liquid nitrogen. Agar slopes (slants) are made from master cultures and quality assured following ISO 9000 accredited methodology. Batches of approximately 20 slopes are made from the culture in each straw. One slope is always sacrificed to test the batch for microbiological contamination, viability and respiratory-deficient yeast mutants (petites). In addition, the identity of each batch of slopes is confirmed using molecular biology analysis techniques such as polymerase chain reaction. Duplicate cultures of all yeast strains are held confidentially by the National Collection of Yeast Cultures as a back-up. In total, 12 brewing yeast strain master cultures are held at the Scottish Courage Technical Centre in Edinburgh using liquid nitrogen storage. These strains are used to supply 10 breweries with over 600 agar slopes between them during the course of each year. The yeast storage and supply management systems in place have proved to be robust and reliable over a number of years, giving breweries in the group confidence in the quality of the yeast supplied to initiate brewery propagations.

12.1 Introduction 12.1.1

Historical perspective

Traditionally, the majority of breweries produced either ales or lagers, and each brewery would use a single yeast strain (or even a mixed culture) to brew all of its products. In many cases there would be little or no strain maintenance, with the yeast pitched from one fermentation to the next ad infinitum. This situation can be seen to persist to the present day in a few smaller ale breweries in the UK, especially those using mixed strains. Simple procedures had subsequently been introduced into a number of

132

BREWING YEAST FERMENTATION PERFORMANCE

breweries where the brewing strain was maintained by regular subculturing onto slopes made from solidified wort. During the 1970s, significant changes in market forces resulted in a requirement for the production of ales and lagers in the same brewery. Fortunately, the yeast strains involved were usually easily distinguishable using simple microbiological techniques, and most brewers had moved away from mixed cultures to single-strain cultures for production. Also at this time, the microbiological support and techniques available to help the brewer were becoming more and more advanced. The late 1980s onwards saw a dramatic change in the way that many brewing companies operated, especially in the UK. The 1990s was a decade of take-over, amalgamation and consolidation in the brewing industry, and also saw a dramatic rise in the level of contract and franchise brewing being undertaken. Today, this has resulted in a small number of large companies who now focus more on brands rather than generic products, and who must be able to demonstrate brand integrity, be it internally or, more likely, to a franchise partner. All of these factors have resulted in brewing companies and individual breweries handling more brewing yeast strains concurrently than ever before. The need for a reliable and secure system of brewing yeast strain maintenance and starter culture supply has therefore become a priority for many brewing companies.

12.2 Yeast culture management 12.2.1

Aims

A good yeast strain management system in a large, modern, multi-site brewing company must ensure the integrity of strain supply. The stability of desirable characteristics of each yeast strain involved must be assured, and the introduction of undesirable traits minimised, usually by the regular introduction of new yeast cultures. Systems in place must be not only cost-effective, but also secure, as individual brewing yeast strains can be one of the most important assets owned by a brewing company. 12.2.2

Strategies for strain maintenance

Historically, various methods have been used for preserving brewing yeast cultures,1 including the ‘nil option’ (pitch on continually), subculturing (broth to broth, or slope to slope), desiccation, lyophilisation and cryopreservation (either at 70°C, or at 196°C in liquid nitrogen). Several studies have demonstrated that cryopreservation in liquid nitrogen is the method of choice for the maintenance of yeast cultures, both for the optimal storage of the yeast strains2,3 and for a positive impact on subsequent fermentation performance.4 Up until the early 1980s Scottish Courage, like many other brewing companies, maintained their brewing yeast strains by regular subculture onto agar slopes and storage at 4°C. Cryopreservation, using liquid nitrogen, was introduced in 1983 to help overcome the potential problems of genetic variation exacerbated by the subculturing regimen.1 For a number of years after the introduction of cryopreservation,

YEAST STORAGE AND SUPPLY MANAGEMENT

133

each brewing yeast strain was reselected annually from a culture in use in the brewery at that time. This ‘annual master’ was stored in liquid nitrogen and used to supply starter cultures to the breweries for the next 12 months. After several years, when confidence in the stability of the liquid nitrogen system had been built up, it was decided to select a single ‘master culture’ of each strain in use and lay down sufficient stocks of the culture to guarantee supply for a considerable length of time. 12.2.3

Selection of master cultures

A master culture for each production strain was selected from the existing brewery population or previous annual master cultures held in liquid nitrogen, based on the results of a number of tests, as detailed below. One-hundred isolates from each source were tested and the six ‘best’ (from a single source) pooled to produce the master culture. Six isolates were pooled rather than selecting a single isolate, as there was always the possibility that a single clone could be lacking an important attribute that had not been tested for (e.g. enzyme activity), which could cause problems in subsequent fermentations. 12.2.4

Testing procedures

One-hundred isolates, picked from a wort agar plate, were grown in small culture bottles using a rich classification medium and underwent a range of tests. A typical profile for each brewing yeast strain had already been defined. 12.2.4.1 Flocculation (Tullo) and adhesion. After incubation the culture supernatant was carefully decanted and the adhesion of the yeast sediment to the bottle was observed and recorded. The sediment was then suspended and the degree of flocculation assessed visually. The flocculation category of the yeast was classified by the Tullo system. 12.2.4.2 Sedimentation (Helm’s test). Yeast taken from the classification culture bottles was fermented in test-tubes of broth. The yeast was suspended in the incubated sedimentation tubes by vortexing, and an absorptiometer, measuring the transmittance of white light with a 609 nm filter, was used to assess the degree of sedimentation. 12.2.4.3 Sugar utilisation. Yeast from classification culture bottles was inoculated into carbohydrate broths and incubated for 7 days at 27°C. Fermentation-positive yeast isolates were those that fermented the carbohydrate provided, and produced carbon dioxide and sediment (biomass). The carbohydrate sources used were galactose, meliobiose and maltotriose. Atypical isolates were quickly identified. 12.2.4.4 Head formation. Wort in tall test-tubes was inoculated with yeast isolates (again taken from the classification culture bottles) and incubated for 3 days at room temperature. After incubation, the top of the fermented wort was examined for the formation of a yeasty head. Comparison of the head formation characteristics of a number of isolates indicated whether the isolate was normal or atypical.

134

BREWING YEAST FERMENTATION PERFORMANCE

12.2.4.5 Petite stability. Loss of mitochondrial function can occur spontaneously in individual brewing yeast cells. This mutation results in an inability to respire carbon sources and gives a low yield of biomass relative to normal cells under aerobic conditions (e.g. during yeast propagation). A high percentage of petite mutants in a brewing yeast population can give a risk of excessive diacetyl formation. The small colonies produced by these mutant cells on agar plates gives rise to the term ‘petite’. Respiratory-deficient cells are unable to reduce tetrazolium dye to a coloured form. Normal colonies are red–pink, whereas petite colonies are white after exposure to an indicator dye. A ‘healthy’ brewing yeast culture should have less than 2% petite colonies in the population. 12.2.4.6 Fermentation performance. Laboratory-scale (2 litre EBC tall tubes) and pilot brewery-scale (36 hl fermentation vessels) trials were carried out, and a large range of parameters was monitored to ensure that the expected fermentation performance was achieved. 12.2.5

Deposition in liquid nitrogen

Following the fermentation trials, the six pooled isolates were deposited into the liquid nitrogen storage system. The yeast was grown for 3 days in YM broth and a cell count and viability measurement were carried out to confirm successful propagation. The cell count was adjusted to 1  107 cells/ml and mixed with an equal amount of cryoprotectant (glycerol). Small sections of autoclaved drinking straw were filled with the culture and heat sealed. After slow freezing overnight, five straws were placed in each colour-coded cryo-vial and the vials were then deposited into the liquid nitrogen bank. The straws were held just above the liquid level, at a temperature of about 145°C. 12.2.6

Cascade storage system

A ‘cascade’ storage system was used to ensure an almost infinite supply of the designated master culture of each brewery yeast strain. Fifty straws of the yeast were prepared (master culture) and stored in liquid nitrogen. One straw was retrieved and used to prepare 50 more straws (the working master culture). A single straw from the working master culture was then removed from storage and used to prepare 20 slopes, as detailed below. These slopes were then subjected to the same range of tests as was used to select the master culture from the brewery population, ensuring that there had been no drift in desirable characteristics. Only after the yeast propagated from one of these slopes had undergone successful fermentation trials, at laboratory (EBC tall tubes), pilot brewery and full production scale, was the master culture held in liquid nitrogen ‘signed off’ as the true master culture. Once the 50 straws of the working master culture have been exhausted, a second true master straw can be used to prepare 50 more, so providing an almost indefinite supply of yeast. 12.2.7

Retrieval from liquid nitrogen and slope preparation

On removal from liquid nitrogen storage, a straw of yeast was placed in sterile saline at room temperature to thaw gradually. The straw was then surface-sterilised with

YEAST STORAGE AND SUPPLY MANAGEMENT

135

methylated spirits and cut open aseptically. The contents were mixed and 0.05 ml was added to 0.95 ml of YM broth (recovery medium). After 3 days’ incubation at 27°C, this yeast culture was used to prepare a batch of slopes. Pre-prepared YM slopes were used, and 20 were inoculated from each liquid nitrogen straw, then incubated for 3 days at 27°C. 12.2.8

Quality assurance

Two slopes from each batch prepared were sacrificed, and underwent a battery of tests before the rest of the batch could be certified as ready for release to the breweries to initiate new propagations. 12.2.8.1 Freedom from contamination. Yeast growth on a slope was slurried with sterile saline and plated out onto a range of selective agars to examine for microbial contamination:

• • • • • •

YM  copper for Saccharomyces wild yeasts lysine agar for non-Saccharomyces wild yeasts WLN at 37°C for ale/lager yeast cross-contamination WLD for aerobic bacteria Raka Ray agar (anaerobic incubation) for anaerobic, beer spoilage bacteria WLN at 27°C for 7 days for colony morphology check.

12.2.8.2 Petite mutants. The level of petite mutants in the yeast population on the sacrificed slope was determined as before. 12.2.8.3 Viability. The viability of the yeast on the freshly grown slope was measured using the methylene violet stain5 before despatch. Levels close to 100% were expected and achieved routinely. 12.2.8.4 Genetic confirmation of identity. The identity of the yeast on the sacrificed slope was confirmed before dispatch using genetic techniques. With a large number of different yeast strains being held at a central facility, it is imperative that the correct yeast strain is supplied on each occasion. The DNA of Saccharomyces yeast is packaged into 17 chromosomes which range in length from 0.031 to 0.85 nm. The contour-clamped homogeneous electric field (CHEF)6 electophoresis technique can be used to separate the chromosomes on the basis of their lengths to give a specific fingerprint of the yeast strain. Unfortunately, a number of closely related yeast strains, presumably from a very similar genetic origin, cannot be identified using this technique. A second technique must be used to separate these yeasts (usually lager strains). Target nucleotide sequences (specific to each Scottish Courage brewing yeast strain) can be copied repeatedy using the polymerase chain reaction (PCR) technique. Sufficient quantities of the DNA segments can be produced within a few hours so that the presence or absence of the specific sequences can be determined by electrophoresis. Details of the specific DNA primers used in this work have been published previously.7

136

BREWING YEAST FERMENTATION PERFORMANCE

Integrity of supply

12.2.9

A colour-coding system is used for each individual yeast strain, from liquid nitrogen straw through to prepared yeast slopes ready for delivery. After preparation, those slopes under test are physically segregated from those ‘certified’ for release. Each slope is labelled with a unique batch number and ‘use by’ date (16 weeks from the date of preparation). A certificate of analysis is sent out with each batch of slopes, and upon arrival at the brewery a form is returned confirming the identity of the yeast and its condition. The whole yeast management and supply system is the subject of regular audit, both internal and by external bodies such as BSI (as part of ongoing ISO 9001 accreditation). At the beginning of each year, a monthly delivery schedule for yeast slopes is devised. Extra slopes are prepared to allow an early response to ad hoc requests for slopes. Batches of slopes, packed in insulated boxes with ice packs, are sent to the breweries using an overnight courier service and are subjected to standard storage, recovery and laboratory propagation procedures.8 12.2.10

Statistics

In total, 12 brewing yeast strains are involved, and 10 brewing sites across the UK and world-wide are supplied with yeast slopes from the central facility. Approximately 600 slopes are dispatched each year. Capital outlay for the equipment was approximately £10 000, with an ongoing cost of weekly liquid nitrogen delivery. The maintenance of the culture collection and slope preparation takes up approximately 50% of a technician’s time. 12.3

Conclusions

Seventeen years’ experience of liquid nitrogen storage has confirmed it to be the best available method for the maintenance of brewing yeast strains. Six years’ experience of using the cascade system has shown it to be ideal for the robust, reliable and secure storage of yeast master cultures and supply of starter culture slopes. This is evidenced by feedback from the breweries involved, who report reproducibility of propagation and consistency of fermentation performance. The systems in place have proven to be both flexible and cost-effective for a large, multi-site brewing group. Acknowledgements The authors would like to thank the directors of Scottish Courage Brewing Ltd for permission to publish this article. References 1. Kirsop, B.E. and Doyle, A. (1991) Maintenance of Microorganisms and Cells. A Manual of Laboratory Methods. Academic Press, London.

YEAST STORAGE AND SUPPLY MANAGEMENT

137

2. Russell, I. and Stewart, G.G. (1981) Liquid nitrogen storage of yeast cultures compared to more traditional storage methods. J. Am. Soc. Brew. Chem. 39, 19–24. 3. Walker, G.M. (1998) Yeast Physiology and Biotechnology. Wiley, Chichester. 4. Hulse, G., Bihl, G., Morakile, G. and Axcell, B. (2000) Optimisation of storage and propagation for consistent lager fermentations. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 161–169. 5. Smart, K.A., Chambers, K.M., Lambert, I. et al. (1999) Use of methylene violet staining procedures to determine yeast viability and vitality. J. Am. Soc. Brew. Chem. 57, 18–23. 6. Pedersen, M.G. (1994) Molecular analysis of yeast DNA – tools for pure yeast maintenance in the brewery. J. Am. Soc. Brew. Chem. 52, 23–27. 7. Coakley, M., Ross, R.P. and Donnelly, D. (1996) Application of the polymerase chain reaction to the rapid analysis of brewery yeast strains. J. Inst. Brew. 102, 349–354. 8. Kennedy, A.I. (2000) Yeast handling in the brewery. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 129–134.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

13 Comparison of Yeast Viability/Vitality Methods and Their Relationship to Fermentation Performance L.R. WHITE, K.E. RICHARDSON, A.J. SCHIEWE and C.E. WHITE

Abstract In the brewhouse, consistent beer production is the ultimate goal. Experienced brewers implement techniques and methods to curb inconsistencies. Command over raw ingredients, sanitation, fermentation and packaging is desired. Because yeast lends tremendous flavour and character to beer, brewers attempt to determine yeast quality by measurement of its viability and/or its vitality. Ascertaining the physiological condition of yeast is useful information for creating a consistent product. Small-scale microbreweries face an even greater challenge, with limited laboratory equipment and space. In this study, various methods were applied for determining yeast viability and vitality that could be easily replicated in most microbreweries. Recently, the accuracy of the standard methylene blue viability stain has been questioned. Improved reliability and reproducibility with other staining methods, such as methylene violet, have recently emerged within the brewing industry. This laboratory compared citrate methylene blue, alkaline methylene violet, alkaline methylene blue (AMB), acidification power (AP) and standard plate count against fermentation studies to determine the best method for assessing yeast viability and vitality. Laboratory-grown ale and lager yeast were followed over a 6 month storage period, viability and vitality were measured, and small-scale fermentation was performed at various intervals. Vital dye results were unreliable for aged, poor cell wall-defined yeast. AMB and AP were found to have the best correlation between apparent viability/vitality and fermentation performance for yeast with well-defined cell walls. AP gave the most accurate time-line evaluation of yeast and correlation with yeast performance.

13.1 Introduction In the brewhouse, yeast is susceptible to changes in pH, temperature, oxygen, ethanol, CO2, nutrition and gravity, to name a few factors. Stressed yeast may lose its ability to replicate, become unable to ferment or die. Ascertaining the condition of the yeast is useful in repitching a consistent number of ‘live’ yeast cells.1 Breweries have attempted to determine yeast quality by measurement of viability (the percentage of live cells within a population) and/or vitality (metabolically active yeast). Methods for testing the viability and vitality of yeast cells centre around three general principles: loss of replication capability, cell damage and loss of metabolic activity.2 Vital dyes have become the standard for viability testing. Vital dye staining challenges the integrity of the cell wall as well as the ability of the cell to reduce or extrude the dye and remain colourless.3 Methylene blue staining has been the standard for assessing yeast viability since the 1920s.2,4 However, this method has recently been questioned given its poor reproducibility and inaccuracy with apparent viability below 90%.5 Other dyes, such as methylene violet, have recently been introduced as an improved staining

COMPARISON OF YEAST VIABILITY / VITALITY METHODS

139

alternative.6 Fluorescent staining procedures have also been attempted (reviewed elsewhere in this book). Finally, the acidification power test, which measures the ability of the yeast cell to retain proper intracellular and extracellular hydrogen ion concentrations, is a relatively simple measurement of vitality7 and has been previously suggested as an alternative to vital stains.6 This study compared citrate methylene blue (CMB), citrate methylene violet (CMV), alkaline methylene blue (AMB), acidification power (AP) and standard plate count (SPC) against fermentation studies to determine the best method for assessing yeast viability. 13.2 Materials and methods 13.2.1

Yeast

Several White Labs commercial strains of Saccharomyces cerevisiae were used for experimental protocols. Yeast was grown aerobically in a sterile malt media and stored at 4°C until analysis. The age of yeast reflects the date from which the yeast was stored. Yeast was allowed to reach room temperature and then diluted with sterile deionised water to reach a concentration of 1  107 cells/ml. Heat-stressed (HS) yeast was placed in a 87.8°C (190°F) water bath for 3 min. Heat-killed (HK) yeast was subjected to 4 min of microwaving at high power. 13.2.2

Citrate methylene blue

Methylene blue was dissolved in a sodium citrate solution (2%) to a final concentration of 0.01%. Yeast was diluted with sterile deionised water to reach a concentration of 1  107 cells/ml suspension. Then, 0.5 ml of yeast suspension was added to 0.5 ml CMB and gently agitated. The solution was examined microscopically after 2 min. Dark blue cells were counted as dead. The viability was tested in duplicate unless results were greater than 0.2% different. An additional count was performed if such a discrepancy existed. The mean viability counts were represented according to the methods outlined previously.6,8 13.2.3

Alkaline methylene blue

A methylene blue stock solution (0.1%) was diluted 10-fold with a 0.1 M glycine buffer solution, pH 10.6. Then, 0.5 ml of yeast suspension (1  107 cells/ml) was added to 0.5 ml of alkaline methylene blue staining solution, mixed and incubated for 15 min at room temperature. Yeast cells were examined microscopically and medium to dark blue cells were recorded as dead, while pale blue and unstained cells were counted as living. The viability was tested in triplicate. The mean viability counts were represented according to the methods outlined previously.6 13.2.4

Alkaline methylene violet

The same method of preparation was used as with AMB, substituting methylene violet 3 RAX for methylene blue. Cells were considered dead if they displayed any variation

140

BREWING YEAST FERMENTATION PERFORMANCE

of pink colour. The viability was tested in triplicate. The mean viability counts were represented according to the methods outlined previously.6 13.2.5

Acidification power

The pH meter was calibrated using the two-buffer method before each series of assays. Deionised water pH was adjusted to approximately 6.5 pH for AP studies. Sterile deionised water (15 ml) was placed in a 50 ml conical centrifuge tube containing a conical stir bar. The pH of the water was monitored for 5 min with constant stirring. At the end of 5 min, a pH reading was recorded (AP0) and 5 ml of concentrated yeast slurry (1  109 cells/ml) was added to the centrifuge tube. The yeast suspension was allowed to stir for 10 min, after which the pH was recorded (AP10). Immediately after the recording of the AP10, 5 ml of 20% glucose solution was added to the yeast suspension and allowed to incubate for 10 min. At the end of 10 min the final pH reading was recorded (AP20). The acidification power was calculated by subtracting the AP20 from the AP0 reading. The assay was conducted in triplicate and the data were represented as the mean value obtained according to the method of Kara et al.7 13.2.6

Standard plate count

Yeast was diluted with sterile deionised water to achieve a concentration of 1–3  103 cells/ml. Then, 0.1 ml of diluted yeast solution was plated on 100  15 mm nutrient agar plates by spreading. Plates were incubated at 30°C for approximately 42 h. Individual colonies were counted and viability was reported as a mean percentage. 13.2.7

Fermentation

Fermentation to evaluate yeast performance was carried out in 2 litre flasks containing approximately 1500 ml of sterile wort with an original gravity of 1.040. Cultures were pitched with 1  106 cells/ml per degree Plato. Cultures were monitored by recording pH and percentage of sugar.

13.3 13.3.1

Results and discussion Citrate methylene blue

Methylene blue is an autoxidisable dye, whereby entry into the cytoplasm of a living cell results in its oxidation to the colourless leuco-form.6 It is further suggested that living but damaged cell membranes may result in the occurrence of variable cell shading, which is likely to be responsible for inconsistent viability counts. Moreover, viability assessments conducted by different operators can yield data with error margins within the range 10–20%. The accuracy of the methylene blue procedure has been reported by some researchers to be reliable only at viabilities greater than 90%.5 It has also been reported that methylene blue will yield viabilities as high as 30–40% at 0% true viability.

COMPARISON OF YEAST VIABILITY / VITALITY METHODS

141

In this study it was observed that methylene blue was indeed unreliable. The assay appeared to demonstrate accuracy for viabilities over 95%. The HS yeast exhibited viability readings of 89.7–98.1% using CMB compared with 39.5–63.6% obtained when using AMB and AMV, respectively (Fig. 13.1). The HK yeast exhibited viabilities within the range 4.6–15.3% when using CMB, while the AMB and AMV assays both reported 0% viability for the same population.

California-WLP001 100 Viable (%)

80

CMB

60 AMB

40

AMV

20 0 1

2

4

6

8

(a)

12

16

20

24

HS

HK

Week

Edinburgh-WLP028 100 Viable (%)

80

CMB

60

AMB

40 AMV

20 0 1

2

4

6

8

(b)

12

16

20

24

HS

HK

Week

German Lager-WLP830 100 Viable (%)

80

CMB

60 AMB

40

AMV

20 0 1

(c)

2

4

6

8

12

16

20

24

HS

HK

Week

Fig. 13.1 Viability of (a) WLP001, (b) WLP028 and (c) WLP830 following storage at 4°C. Viability was determined using citrate methylene blue (CMB), alkaline methylene blue (AMB) and alkaline methylene violet (AMV). Control cultures of heat-stressed (HS) and heat-killed (HK) cells are also shown.

142

BREWING YEAST FERMENTATION PERFORMANCE

The most challenging aspect of the CMB stain was ascertaining whether cells stained sufficiently to count as dead. According to the American Society of Brewing Chemists (ASBC) methods of analysis, only cells staining dark blue should be considered dead. The distinction between live and dead cells became increasingly difficult to ascertain as the storage time of the yeast increased. This observation is consistent with the hypothesis of Smart et al.6 that the impurities present in methylene blue dye preparations lead to ambiguous staining intensities which render the assay subjective. 13.3.2

Alkaline stains

13.3.2.1 Alkaline methylene blue. Modification of methylene blue to an alkaline pH enhances the uptake rate of the dye molecule into the yeast cell. Consequently, the staining intensity appears modified and more variable in some circumstances. For the purposes of this investigation, yeast cells were recorded as dead if they appeared medium to dark blue and coloured throughout the cytoplasm. Pale blue cells were counted as living. While interpreting the viability status of cells stained with AMB was easier than with CMB, some ambiguous cell coloration remained. Again, viability determination increased in difficulty with increased yeast storage time. AMB consistently gave lower viability values than CMB (Tables 13.1–13.3). HS yeast viability correlated more closely with AMB and AMV than with CMB when compared with fermentation performance (Tables 13.1–13.3). 13.3.2.2 Alkaline methylene violet. As with AMB, the pH of AMV is designed to increase stain uptake. Unlike AMB, AMV has lower levels of impurities and decreased oxidative demethylation,6 thereby decreasing colour variation in staining. Unlike the methylene blue stains, AMV allowed for easy distinction between dead cells (pink)

Table 13.1

Viability and fermentation performance of brewing strain California – WLP001

Time

Week 1 Week 2 Week 4 Week 6 Week 8 Week 12 Week 16 Week 20 Week 24 HS HK

Stain (%)

Fermentation at

CMB

AMB

AMV

AP

24 h

48 h

72 h

% Plating

99.7 98.3 99.4 98.4 96.6 97.1 97.2 96.5 98.4 89.7 4.6

95.8 92.2 88.3 93.4 94.8 91.7 90.2 83.7 83.6 39.5 0

96.2 84.9 96.6 94.8 93.3 91.9 86.0 83.2 75.0 50.9 0

3.22 3.16 3.15 3.18 3.14 2.95 2.95 2.77 2.60 2.49 1.52

1.024 1.028 1.026 1.027 1.026 1.032 1.023 1.038 1.025 1.040 1.040

1.012 1.014 1.011 1.022 1.018 1.012 1.012 1.014 1.013 1.030 1.040

1.008 1.010 1.008 1.012 1.010 1.006 1.006 1.008 1.009 1.014 1.040

100 100 100 100 100 100 100 100 100 29.3 2.3

CMB: citrate methylene blue; AMB: alkaline methylene blue; AMV: alkaline methylene violet; AP: acidification power; HS: heat-stressed (yeast in 87.8°C bath for 3 min); HK: heat-killed (yeast microwaved for 4 min on high power).

COMPARISON OF YEAST VIABILITY / VITALITY METHODS

143

from live cells (colourless). Most cells stained bright violet, while less than 10% stained light pink. However, ambiguity of colour appeared, as with AMB, with increased time of yeast storage. The AMV viability correlated most closely with AP, compared with AMB and CMB (Fig. 13.2). However, the results of the stains were harder to interpret with increased age of the yeast. The reasons for this were not known; however, a gradual thinning of the cell wall was visualised microscopically in all yeast samples after week 12. AMB and AMV showed a high correlation on yeast stored up to 12 weeks, and a declining correlation thereafter (Fig. 13.3).

Table 13.2

Viability and fermentation performance of brewing strain Edinburgh – WLP028

Time

Week 1 Week 2 Week 4 Week 6 Week 8 Week 12 Week 16 Week 20 Week 24 HS HK

Stain (%)

Fermentation at

CMB

AMB

AMV

AP

24 h

48 h

72 h

% Plating

96.25 97.4 97.5 94.8 94.1 93.6 96.8 95.5 97.8 98.1 8.4

82.3 71.5 71.6 77.9 74.5 72.4 71.3 55.8 58.7 55.8 0

90.8 79.7 83.6 86.8 86.0 72.4 60.4 49.1 19.2 58.7 0

3.06 2.95 3.00 3.02 2.90 2.69 2.39 2.25 2.07 2.24 1.84

1.022 1.026 1.024 1.028 1.023 1.022 1.032 1.026 1.028 1.042 1.042

1.010 1.008 1.010 1.018 1.012 1.010 1.018 1.010 1.011 1.038 1.042

1.008 1.006 1.008 1.010 1.008 1.008 1.006 1.006 1.005 1.020 1.042

100 100 100 100 100 100 100 100 14.1 2.5 1.4

CMB: citrate methylene blue; AMB: alkaline methylene blue; AMV: alkaline methylene violet; AP: acidification power; HS: heat-stressed (yeast in 87.8°C bath for 3 min); HK: heat-killed (yeast microwaved for 4 min on high power).

Table 13.3

Viability and fermentation performance of brewing strain German lager – WLP830

Time

Week 1 Week 2 Week 4 Week 6 Week 8 Week 12 Week 16 Week 20 Week 24 HS HK

Stain (%)

Fermentation at

CMB

AMB

AMV

AP

24 h

48 h

72 h

% Plating

99.0 99.5 99.4 97.9 97.1 96.2 98.4 98.0 90.0 98.1 15.3

93.2 95.9 93.4 92.9 91.4 83.7 70.9 15.1 37.0 57.9 0

93.1 97.9 96.7 93.8 94.1 72.6 48.5 9.4 26.9 63.6 0

3 3.21 3.07 2.82 2.55 2.15 2.1 1.84 1.8 1.89 1.36

1.018 1.020 1.030 1.035 1.025 1.028 1.033 1.031 1.034 1.042 1.042

1.008 1.010 1.010 1.012 1.010 1.008 1.012 1.010 1.015 1.042 1.038

1.006 1.006 1.008 1.008 1.010 1.008 1.008 1.006 1.006 1.038 1.038

100 100 100 100 100 100 100 100 100 0 0

CMB: citrate methylene blue; AMB: alkaline methylene blue; AMV: alkaline methylene violet; AP: acidification power; HS: heat-stressed (yeast in 87.8°C bath for 3 min); HK: heat-killed (yeast microwaved for 4 min on high power).

144

Viability (%)

BREWING YEAST FERMENTATION PERFORMANCE

100 80 60 40 20 0 1

1.5

Viability (%)

3.5

100 80 60 40 20 0 1

1.5

2 2.5 Acidification power

1 .5

2 2 .5 Acidification power

(b) Viability (%)

3

2 2.5 Acidification power

(a)

100 80 60 40 20 0 1

(c)

3

3.5

3

3 .5

Fig. 13.2 Relationship between acidification power (AP) and (a) citrate methylene blue (CMB), (b) alkaline methylene blue (AMB), and (c) alkaline methylene violet (AMV).

Alkaline stain comparison 100 AMV

80 60 40 20 (a)

0 0

20

40

20

40

AMB

60

80

100

60

80

100

100 AMV

80 60 40 20 (b)

0 0

AMB

Fig. 13.3 Relationship between alkaline methylene blue (AMB) and alkaline methylene violet (AMV) for (a) storage (weeks 1–8) and (b) prolonged storage (weeks 1–24).

COMPARISON OF YEAST VIABILITY / VITALITY METHODS

145

13.3.2.3 Acidification power test. The AP test is designed to test both the glycolytic activity and endogenous reserves of the yeast cell to maintain a fixed ratio between intracellular and extracellular hydrogen ion concentrations. Water pH variation was adjusted to around 6.5 for all AP trials. Initial water calibration helped to obtain consistent and comparable results. Overall, the AP values displayed a downward trend with increased time of yeast storage (Tables 13.1–13.3). AP values correlated more closely with AMV than with either AMB or CMB (Fig. 13.2). 13.3.2.4 Standard plate count. The data obtained using the SPC technique did not correspond to the vital dye data obtained. In all cases except where the yeast had been heat shocked (HS) or heat killed (HK), the colony number was greater than the number of cells plated. The only instance where the SPC was indicative of the apparent viability was in the cases of the HS and HK yeast (Tables 13.1–13.3). 13.3.2.5 Yeast performance. Yeast performance in this study specifically refers to the ability of the yeast to attenuate malt wort. Measurement of sugar percentage was obtained throughout the 72 h test period (Fig. 13.4). Figure 13.5 compares AP with the specific gravity of fermentation at 24, 48 and 72 h. The most dramatic comparison between yeast performance and apparent viability can be seen with the HS yeast. Fermentation did occur with the HS yeast, although there was a longer lag time to attenuation and higher specific gravity at 72 h. The alkaline stains and AP most closely predicted viability and fermentation performance (Tables 13.1–13.3). Yeast storage had more effect on lager fermentation than on ale fermentation. At 72 h, week 8 yeast had a 12.5% increase in sugar percentage compared with week 4, while having only a 1.22% difference in AMV values, 2.6% difference in AMB values, 0.1% difference in CMB values and 4.36% difference in AP values. In WLP001 fermentation, the final specific gravity readings differed by only 0.6%, despite a range of viability/vitality values (excluding HS and HK yeast) (Fig. 13.4).

13.4

Conclusions

The results of this study indicate that the standard CMB assay is accurate only for yeast with high cell viability ( 95%). SPC is an inaccurate means to assess yeast viability. The AP test appears reliable in reporting vitality, although should not be used as a sole means of interpreting yeast health. While the alkaline stains (AMB, AMV) are easy to use and require approximately 15 min to complete, AMV is preferred for its easier distinction between live and dead cells. However, the results of the stains became harder to interpret with increasing age of the yeast. This is possibly due to the deterioration of the cell wall and/or plasma membrane. A gradual thinning of the cell wall was visualised microscopically in all yeast samples after week 12. The results of this study add evidence to the inaccuracy of the standard CMB assay, and suggest that interpreting yeast health should be a multi-technique approach as the ‘vitality’ or ‘viability’ of yeast is reflected in multiple cell systems.

Fig. 13.4 Fermentation profiles of (a) WLP001, (b) WLP028 and (c) WLP830 yeast strains. Pitching slurries were aged for 2, 4, 6, 12, 16 and 20 weeks as appropriate. Heat-shocked and heat-killed slurries were used as controls. The pitching rate used was 1  106 cells/ml per degree Plato.

COMPARISON OF YEAST VIABILITY / VITALITY METHODS

147

Specific gravity

California-WLP001 Fermentation 1.045 1.040 1.035 1.030 1.025 1.020 1.015 1.010 1.005 1.000

(a)

24h 48h 72h

3.22 3.16 3.15 3.18 3.14 2.95 2.95 2.77 2.60 2.49 1.52 Acidification power

Specific gravity

Edinburgh-WLP028 Fermentation 1.045 1.040 1.035 1.030 1.025 1.020 1.015 1.010 1.005 1.000

(b)

24h 48h 72h

3.06 2.95 3.00 3.02 2.90 2.69 2.39 2.25 2.07 2.24 1.84 Acidification power

Specific gravity

German Lager-WLP830 Fermentation 1.045 1.040 1.035 1.030 1.025 1.020 1.015 1.010 1.005 1.000

(c)

24h 48h 72h

3.00 3.21 3.07 2.82 2.55 2.15 2.10 1.84 1.80 1.89 1.36 Acidification power

Fig. 13.5 Relationship between acidification power and yeast attenuation performance of the brewing strains (a) WLP001, (b) WLP028 and (c) WLP830. Yeast fermentations were conducted in malt wort and analysed at 24, 48 and 72 h.

References 1. Heggart, H.M., Margaritis, A., Pilkington, H. et al. (1999) Factors affecting yeast viability and vitality characteristics: a review. Tech. Q. Master Brew. Assoc. Am. 36, 383–406. 2. Heggart, H.M., Margaritis, A., Stewart, R.J. et al. (2000) Measurement of brewing yeast viability and vitality: a review of methods. Tech. Q. Master Brew. Assoc. Am. 37, 409–430. 3. Jones, R. (1987) Measures of yeast death and deactivation and their meaning: Part I. Proc. Biochem. 22, 118–128.

148

BREWING YEAST FERMENTATION PERFORMANCE

4. American Society of Brewing Chemists (1992) Microscopic yeast cell counting, yeast-4. In: Methods of Analysis of the ASBC, 8th edn. ASBC, St Paul, MN, pp. 1–2. 5. O’Conner-Cox, E., Mochaba, F.M., Lodolo, E.J. et al. (1997) Methylene blue staining: use at your own risk. Tech. Q. Master Brew. Assoc. Am. 34, 306–312. 6. Smart, K.A., Chambers, K.M., Jenkins, C. and Smart, C.A. (1999) Use of methylene violet staining procedures to determine yeast viability and vitality. J. Am. Soc. Brew. Chem. 57, 18–23. 7. Kara, B.V., Simpson, W.J. and Hammond, J.R.M. (1988) Prediction of the fermentation performance of brewing yeast with the acidification power test. J. Inst. Brew. 94, 153–158. 8. American Society of Brewing Chemists (1992) Yeast stains, yeast-3. In: Methods of Analysis of the ASBC, 8th edn. ASBC, St Paul, MN, pp. 1–3.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

14 Yeast Quality and Fluorophore Technologies S.M. VAN ZANDYCKE, O. SIMAL, S. GUALDONI and K.A. SMART

Abstract The assessment of pitching and cropping yeast quality is important in the attainment of adequate fermentation performance. Viability and vitality are common terms used to describe the quality of yeast. Viability corresponds to the percentage of dead cells in a sample, whereas vitality represents the physiological state of the yeast. Methylene blue remains an industry standard for viability assessment, even though the efficiency of this stain is highly controversial. It has been suggested that methylene violet might provide a more accurate and reproducible assessment of viability than methylene blue, owing to the occurrence of impurities in the latter. The objective of this study was to identify an alternative viability assessment to bright-field reductive dye techniques using fluorophores. Two genetically distinct brewing strains were used: a lager strain (L138) and an ale strain (NCYC2593). Viability studies were performed on yeast cell populations exhibiting different viabilities, obtained from mixing heat-treated and stationary-phase cells. In addition, cohorts of cells were submitted to starvation and oxidative stress and were assessed for viability. Viability was determined using the fluorophore dyes oxonol [DiBAC4(3)], MgANS, berberine, sytox orange, propidium iodide and FUN1, and compared with conventional bright-field dyes such as methylene violet. Oxonol successfully distinguished between live and dead cells without ambiguity, regardless of the yeast strain employed. In addition, with the exception of FUN1, fluorophore staining was perceived to be less subjective to the operator than were bright-field dyes, owing to the lack of intermediate colour variations. It is suggested that fluorophore technology may represent a simple reproducible alternative to methylene blue.

14.1 Introduction Viability is defined as a cell’s ability to bud and grow, however slowly,1 and this may be termed replicative potential.2 The ability to assess accurately the viability of a brewing yeast culture is necessary to maintain fermentation performance and produce a standard, uniform product. Long storage periods and physiological stress occurring during the brewing process can seriously impair the relative vigour of yeast slurries, resulting in loss of viability. The industry standard to measure viability remains the citrate methylene blue assay;3 however, it has been demonstrated that this stain may be inaccurate for viabilities lower than 90%.4,5 An alternative to this stain has been suggested by Smart et al.5 with the use of methylene violet, which contains fewer impurities than methylene blue and as a consequence is less subjective to the operator. Since the early 1990s fluorescence has become a common way of assessing viability in a wide variety of cells, including those of animal origin as well as bacteria and yeast.6 However, as early as 1969, Graham and Caiger7 demonstrated the advantage of using fluorophores as an alternative to methylene blue.

150

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 14.1

Dead yeast cell population stained with sytox orange.

Fig. 14.2

Dead yeast cell population stained with propidium iodide.

Fluorescent stains that bind to nucleic acid can either be membrane permeable (syto) or impermeable, such as sytox orange (Fig. 14.1), propidium iodide (Fig. 14.2) or berberine. Membrane-impermeable nucleic acid stains easily penetrate cells with compromised plasma membranes and are therefore useful for detecting dead cell populations (Fig. 14.3), whereas cell-permeable nucleic acid stains can be used as counterstains for detecting live cell populations.8 Levels of fluorescence emitted by nucleic acid binding stains will depend on the affinity of the dyes for nucleic acids. The nucleic acid staining fluorochromes have been used to compare to slide culture techniques and reported to be more accurate than methylene blue.9–11 Potentiometric fluorescent stains can either be positively (rhodamine 123) or negatively charged (oxonol; Fig. 14.4). Live cells exclude anionic dyes owing to the presence

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

151

Nucleic acids

Dead cell

Live cell

Fig. 14.3

Mode of action of nucleic acid-binding membrane-impermeable stain.

Fig. 14.4

(a) Dead and (b) live yeast cell populations stained with oxonol.

+ + +

+ + + + + +

+ + + ++

+ + + Dead cell

Live cell Fig. 14.5

Mode of action of oxonol on live and dead cells.

of a transmembrane potential (Fig. 14.5); however, the dye can freely enter dead cells.12 The fluorescence of these anionic compounds will be significantly enhanced by binding to intracellular lipids and proteins.13 In addition, oxonol [DiBAC4(3)] has been demonstrated to measure successfully the viabilities of cider14,15 and brewing yeast using flow cytometry16 or fluorimetry.17

152

Fig. 14.6

BREWING YEAST FERMENTATION PERFORMANCE

Dead yeast cell population stained with MgANS.

Live cell

Fig. 14.7

Dead cell

Mode of action of MgANS on live and dead cells.

The use of anilinonaphthalenesulfonate, such as the hemimagnesium salt MgANS, has been suggested as a method to determine yeast cell viability, and correlates with slide culture assessment.9,11 MgANS enters non-viable cells and binds to cytoplasmic proteins to generate yellow/green fluorescence (Figs 14.6 and 14.7). A lower concentration of this dye is necessary compared with methylene blue and membrane leakage is therefore less likely to occur.9 FUN1 (2-chloro-4-(2,3-dihydro-3-methyl-(benzo-1,3-thiazol-2-yl)-methyldiene)-1phenylquinolinium iodide) has been recently suggested as a means of assessing viability in yeast.18,19 It is a halogenated asymmetric cyanine dye, which is membrane permeable and nucleic acid binding. It has been observed that this dye gives rise to cylindrical intravacuolar structures (CIVS) in Saccharomyces cerevisiae.19 Biochemical processing of the dye in live cells gives rise to CIVS that emit a red fluorescence, whereas the nucleic acid-bound form of the dye leads to emission of green fluorescence.20 The formation of CIVS is dependent on the temperature, presence of intracellular glutathione and adenosine triphosphate (ATP) production.19 FUN1 has been successfully used to determine stress changes in starved and oxidatively stressed yeast populations, indicating a potential role in determining vitality (see Chapter 15).

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

153

Other dyes such as viablue,20 carboxyfluorescein diacetate (CFDA)21 and fluorescein diacetate are hydrolysed by esterases in living cells, rendering them fluorescent.15,22 These dyes, which are membrane permeable, leak from cells with damaged membranes and, as a result, dead cells appear unstained.23,24 Finally, potentiometric stains can be reduced or oxidised by active cells (resazurin, dihydrorhodamines, dihydrofluoresceins, tetrazolium salts) and will therefore give a good indication of the activity of the cell. These colourless compounds can be either reduced or oxidised to fluorescent products only by live cells, which possess the necessary enzymic activity and reactive oxygen species required for dye conversion.25 Fluorescent dyes may represent useful alternatives to bright-field stains to assess yeast viability.25 Using this method, viabilities of yeast populations have been determined accurately by flow cytometry26 or spectrofluorimetry19 using large concentrations of cells.

14.2 Materials and methods 14.2.1

Yeast strains and growth conditions

A lager strain (L138) from the Oxford Brookes University Yeast Culture Stock (Oxford, UK; original source: Dirk Bendiak, Molson Breweries, Canada) and an ale strain (NCYC2593) from the National Collection of Yeast Culture (Norwich, UK) were used in this study. Stock cultures were maintained on YPD agar (10 g yeast extract, 20 g biological peptone, 20 g glucose, 1.2% agar in 1 litre of distilled and deionised water, autoclaved immediately at 121°C and 15 psi for 15 min) and stored on slopes at 4°C. Single colonies were selected from a YPD plate and inoculated into 250 ml conical flasks, containing 50 ml of YPD media. Flasks were incubated aerobically at 25°C for 48 h on an orbital shaker at 120 rpm to obtain stationary-phase cultures. 14.2.2

Yeast starvation and heat treatment

Yeast cells previously grown in YPD for 48 h shaking at 25°C were harvested by centrifugation, washed three times, resuspended in sterile distilled water (50 ml) in a conical flask (250 ml) and incubated at 25°C in an orbital shaker (120 rpm) for up to 130 days to obtain starved cell populations, or incubated shaking at 65°C in a water bath for 2 h to obtain dead cell populations.

14.2.3

Citrate methylene violet

Viability was determined using citrate methylene violet following the method of Smart et al.5 Methylene violet 3 RAX (Sigma, St Louis, MO, USA) was dissolved in sodium citrate solution (2% w/v) to a final concentration of 0.01% (w/v). Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. Yeast suspension (0.5 ml) was mixed with 0.5 ml of citrate methylene violet and examined microscopically after 5 min. Dead cells stained violet and unstained cells were assumed to be viable.

154 14.2.4

BREWING YEAST FERMENTATION PERFORMANCE

MgANS

Viability was determined using MgANS following the method of McCaig.11 1-Anilino-8naphthalene-sulfonic acid (MgANS; Sigma, St Louis, MO, USA) (0.3 g) was dissolved in 2 ml of absolute ethanol and diluted with 98 ml of sterile water to a final concentration of 0.3%. This stock solution was kept at 4°C in light-protected bottles for up to 6 months. Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. Yeast suspension (0.5 ml) was mixed with 0.5 ml of MgANS solution and examined after 5 min using a fluorescent microscope (Zeiss, Oberkochen, Germany). Dead cells stained yellow–green and unstained cells were assumed to be viable. 14.2.5

Oxonol

Viability was determined using oxonol [DiBAC4(3); Molecular Probes, Eugene, OR, USA] following the method of Lloyd and Dinsdale.15 Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. The yeast cell suspension (900 l) was mixed with 100 l of Oxonol (10 g/ml), incubated at room temperature for 5 min in the dark and examined using a fluorescent microscope (Zeiss). Dead cells stained yellow–green and unstained cells were assumed to be viable. 14.2.6

Propidium iodide

Viability was determined using propidium iodide following the method of Deere et al.26 Propidium iodide solution was prepared by dissolving 1 mg solid dye (Molecular Probes, Eugene, OR, USA) in 1 ml phosphate-buffered saline (PBS). This stock solution was stored at 20°C in the dark for up to 6 months. Cells were washed once in PBS and resuspended to a final concentration of 1  107 cells/ml in PBS. Yeast cell suspension (1 ml) was mixed with 3 l of propidium iodide solution, incubated at room temperature for 20 min in the dark and examined using a fluorescent microscope (Zeiss). Dead cells stained orange–red and unstained cells were assumed to be viable. 14.2.7

Sytox orange

Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. Yeast cell suspension (500 l) was mixed with 500 l of sytox orange (Molecular Probes) 1 M, incubated at room temperature for 15 min with periodic agitation and examined using a fluorescent microscope (Zeiss). Dead cells stained yellow–orange and unstained cells were assumed to be viable. 14.2.8

Berberine

Viability was determined using berberine following the method of Peladan and Leitz.27 Berberine solution was prepared by dissolving 100 mg solid dye (Sigma) in 100 ml sterile distilled water. This stock solution was stored at 4°C in the dark for up to 6 months.

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

155

Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. Yeast cell suspension (1 ml) was mixed with 20 l of berberine solution, incubated at room temperature for 5 min and examined using a fluorescent microscope (Zeiss). Dead cells stained green and unstained cells were assumed to be viable. FUN1

14.2.9

Viability was determined using FUN1 (Molecular Probes) following the method of Millard et al.19 Yeast cells were washed once in sterile distilled water and resuspended to a final concentration of 1  107 cells/ml. Yeast cell suspension (1 ml) was centrifuged for 5 min at 10 000 rpm and the pellet was resuspended in 1 ml of sterile 2% D-()-glucose containing 10 nM Na-N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid (HEPES), pH 7.2 (GH solution). FUN1 (100 l at 12.5 M) was mixed with 100 l of the yeast suspension, incubated for 30 min at 30°C and examined using a fluorescent microscope (Zeiss). Cells exhibiting red–orange intravacuolar structures were assumed to be viable. 14.2.10

Plate count

Cells were washed once in sterile distilled water and diluted to a final concentration of 1  103 cells/ml. This yeast suspension (0.1 ml) was inoculated onto YPD plates and incubated at 25°C for 72 h. 14.2.11

Photographs

Cells were washed once in sterile distilled water and diluted to a final concentration of 1  107 cells/ml. The yeast suspension was mixed with the appropriate dye following the adequate protocol, as described above. Yeast suspension (2 l) and citifluor antifade (2 l) were mixed onto a slide, covered with a coverslip and sealed using clear nail polish. Cells were examined by fluorescence microscopy (Zeiss) and photographed with a Nikon digital camera attachment (DXM 1200) using a 100 or 60 oil immersion lens.

14.3 14.3.1

Results and discussion Can fluorophores differentiate between viable and non-viable populations?

Methylene blue remains the industry standard for viability assessment, even though the efficiency of this stain has been demonstrated to be inaccurate at below 90% viability.4 It has been suggested that methylene violet may provide greater accuracy than methylene blue.5 Alternatives to bright-field reductive dye techniques have been investigated with the use of fluorophores for the determination of viability and compared with methylene violet. Healthy cell populations (100% viable) of lager (L138) and ale (2593) yeast strains were prepared. Dead cells (0% viable) were obtained by heat treatment. Yeast

156

BREWING YEAST FERMENTATION PERFORMANCE

populations of 25, 50 and 75% were obtained by mixing healthy and dead populations in the appropriate proportions. The viability of these populations was determined using the following assays: citrate methylene violet, MgANS, oxonol, propidium iodide, sytox orange, berberine, FUN1 and plate counts where appropriate. Results were compared using the two-tailed Student’s t-test at a 5% confidence level. 14.3.1.1 Lager strain L138. For populations representing 100% viability, it was demonstrated that each of the dyes exhibited an accurate percentage of viability compared with methylene violet (Fig. 14.8). For populations representing 75% viability, FUN1 and sytox orange exhibited significantly lower viabilities than those of the other dyes (p  0.05) (Fig. 14.8). It was observed that the CIVS, which are formed within vacuoles of live cells when using FUN1, were not always clearly visible because of their limited size, resulting in a potential underestimation of viability. Sytox orange was also observed to underestimate viability and this may be explained by the specificity of the dye to stain only the nuclei. Given the levels of fluorescence emitted by each cell, the differentiation between live and dead cells was not always obvious. For populations representing 50% viability, each dye except for berberine exhibited a lower viability than that of methylene violet (Fig. 14.8). Berberine demonstrated a viability that was significantly higher (p  0.05) than that of the other dyes (Fig. 14.8). Viability using MgANS and FUN1 was observed to be significantly lower (p  0.05) than that of the other dyes (Fig. 14.8). The underestimation of viability occurring with MgANS may be due to the difficulty in differentiating between live and dead cells. Indeed, live cells appeared slightly fluorescent owing to the binding of the dye to cell wall proteins. This, in turn, leads to operator subjectivity and indicates that this stain may be inappropriate as an indicator of viability.

Fig. 14.8 Viability of mixed live/dead yeast populations of lager strain L138, determined using methylene violet, MgANS, oxonol, FUN1, sytox orange, propidium iodide and berberine. Values are the mean of three independent experiments and the standard deviation is indicated by error bars.

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

157

For populations representing 25% viability, all dyes exhibited similar viabilities to that of methylene violet (Fig. 14.8). For populations exhibiting 0%, all dyes correctly identified non-viable cells (Fig. 14.8) and can therefore be considered an improvement on the industry standard, methylene blue, which does not consistently achieve this.5,28 14.3.1.2 Ale strain 2593. For populations representing 100% viability, it was observed that each of the dyes gave an accurate percentage of viability except for FUN1, which exhibited a significantly lower value ( p  0.05) than that of the other dyes (Fig. 14.9). It was previously observed with the lager strain L138 that CIVS were not always visible within the cells. This problem was enhanced in the ale strain 2593 owing to its comparatively small cell size. It is suggested that differences in vacuole, and indeed cell size may therefore cause difficulty in the visualisation of CIVS and therefore underestimation of viability. This observation suggests that the stain may be subject to yeast strain differences and as such may be more useful for some strains than others. For populations representing 75% viability, all dyes except for FUN1 exhibited similar viabilities to that of methylene violet (Fig. 14.9). Viability using FUN1 was observed to be significantly lower ( p  0.05) than that of the other dyes (Fig. 14.9). For populations representing 50% viability, sytox orange and propidium iodide exhibited significantly higher ( p  0.05) viabilities than those of the other dyes (Fig. 14.9). Since sytox orange underestimated viability in populations of the lager strain L138 (for the sample representing 75% viability), this would suggest a lack of consistency in staining potential across the viability range. The reasons for these discrepancies are not known, but it is suggested that the low levels of fluorescence emitted by this dye may lead to operator errors. In contrast, viabilities assessed using MgANS and FUN1 were observed to be significantly lower ( p  0.05) than those obtained when using the other dyes (Fig. 14.9). This phenomenon was not strain dependent but was consistent with both the ale and the lager strain.

Fig. 14.9 Viability of mixed live/dead yeast populations of ale strain NCYC 2593, determined using methylene violet, MgANS, oxonol, FUN1, sytox orange, propidium iodide and berberine. Values are the mean of three independent experiments and the standard deviation is indicated by error bars.

158

BREWING YEAST FERMENTATION PERFORMANCE

For populations representing 25% viability, propidium iodide significantly (p  0.05) overestimated the viability compared with the other dyes (Fig. 14.9). These results were surprising, given that propidium iodide has been reported to underestimate viability15,29 owing to its non-specific binding to yeast.16 However, this stain has not been previously compared with the range of fluorescence stains applied in this study. Previous comparisons may have been conducted with dyes that consistently overestimated viability, such as methylene blue. As for L138, all dyes accurately identified 100% non-viable cell cultures (Fig. 14.9). 14.3.2

Determination of yeast cell viability of starved populations

Starvation occurs during yeast storage and causes utilisation of intracellular carbohydrate reserves (mainly glycogen), which in turn may affect subsequent fermentation performance.30–32 Viability and vitality are negatively affected by storage conditions regardless of the preservation method,33 and it is important to be able to assess these two parameters accurately to ensure adequate fermentation performance. During starvation, the metabolic activity of cells is greatly reduced; however, cell survival can occur for prolonged periods.34 Among the fluorescent dyes, FUN1 relies on the metabolic activity of the cell to produce CIVS,19 which indicate that a cell is viable. It is postulated that starvation may have a negative effect on CIVS formation. Therefore, the impact of starvation on the ability of the fluorescent dyes to assess viability, compared with methylene violet, was determined. Populations of lager (L138) and ale (2593) yeast were subjected to starvation. Healthy yeast populations grown in YPD for 48 h were used as a control. The viability of starved and healthy populations was determined using the following assays: methylene violet, MgANS, oxonol, sytox orange, propidium iodide, berberine, FUN1 and plate counts. For populations exposed to 3 days of starvation, the viability of lager strain L138 using FUN1 was observed to be significantly ( p  0.05) lower than that of the other dyes for lager strain L138 (Fig. 14.10). For the ale strain 2593, all dyes and plate counts exhibited similar viabilities to that of methylene violet (Fig. 14.11). For populations starved for 7 days, oxonol, FUN1, sytox orange and propidium iodide were observed to indicate significantly ( p  0.05) lower viabilities than those of the other dyes for lager strain L138 (Fig. 14.10) and ale strain 2593 (Fig. 14.11). In contrast, plate counts, MgANS and berberine yielded similar viabilities to that of methylene violet (Figs 14.10 and 14.11). Cells stained using FUN1 exhibited a low viability after 3 and 7 days of starvation. This result is not surprising given that the metabolic activity of the yeast is reduced, directly influencing CIVS production. Cells stained using sytox orange and propidium iodide exhibited lower viabilities after 7 days of starvation, suggesting that starvation is likely to affect the permeability of the membrane, resulting in live cells being wrongly identified as dead. Oxonol also indicated a low viability, owing to the diminution of the transmembrane potential. It is postulated that the intensity of fluorescence emitted by cells stained with oxonol may reflect their physiological condition. This hypothesis remains the subject of further investigation. For populations starved for 150 days, all dyes and plate counts, except for propidium iodide and sytox orange, demonstrated a viability of 0%, indicating that both replication capacity and metabolic functions had ceased (Figs 14.10 and 14.11). As previously

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

159

Fig. 14.10 Viability of starved and healthy populations of lager yeast L138 determined using methylene violet, MgANS, oxonol, FUN1, sytox orange, propidium iodide, berberine and plate counts. Values are the mean of three independent experiments and the standard deviation is indicated by error bars.

Fig. 14.11 Viability of starved and healthy populations of the ale yeast NCYC2593, determined using methylene violet, MgANS, oxonol, FUN1, sytox orange, propidium iodide, berberine and plate counts. Values are the mean of three independent experiments and the standard deviation is indicated by error bars.

observed with mixed live/dead populations, propidium iodide overestimated viability. The reasons for these inaccuracies are not known, but given that for both strains propidium iodide indicated a viability of 100%, it is suggested that long-term starvation may induce membrane leakage or prevent the dye from binding nucleic acids. The results obtained with the ale strain 2593 were similar to those of lager strain L138, indicating that the response of the fluorophores to starvation may not be a straindependent phenomenon.

160 14.4

BREWING YEAST FERMENTATION PERFORMANCE

Conclusions

Oxonol and berberine exhibited similar viabilities to that of methylene violet for populations of ale and lager yeast representing 0, 25, 50, 75 and 100% viability. Berberine accurately assessed viability for starved populations of lager and ale strains, compared with methylene violet. Under the same conditions, oxonol displayed high levels of background fluorescence, possibly due to a decrease in membrane potential. However, this observation occurred after 7 days of starvation, a situation unlikely to happen in a brewing environment. It is therefore suggested that these two fluorophores may represent a suitable alternative to bright-field dyes for viability assessment. MgANS, sytox orange, FUN1 and propidium iodide demonstrated some inaccuracies in determining viability compared with methylene violet. MgANS appeared to underestimate viability owing to the binding of the dye to extracellular proteins, whereas sytox orange specifically stains the nucleus of the cell and, as a result, differentiation between live and dead cells was very subjective. FUN1 consistently underestimated viability. CIVS were difficult to visualise within the cells examined because of their small size; it is therefore suggested that this phenomenon may be dependent on the strain. Propidium iodide was demonstrated to be inaccurate for both healthy and starved populations of lager and ale strains, indicating that this dye is not suitable for viability assessment. The reasons for this are the subject of further investigations.

Acknowledgements Dr Sylvie Van Zandycke was supported by an Oxford Brookes University studentship and would like to thank the Royal Society for their financial support towards attendance and presentation of this paper. Sara Gualdoni and Olivier Simal were supported by the undergraduate exchange programme ‘Socrates’. Dr Katherine Smart is the Scottish Courage Reader in Brewing Sciences and would like to thank Scottish Courage Ltd for their support. Finally, Katherine Smart is a Royal Society Industrial Fellow and gratefully acknowledges the Royal Society, BBSRC and EPSRC for their collective support.

References 1. Bendiak, D. (2000) Review of metabolic activity and their ability to predict fermentation performance. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 34–43. 2. Powell, C.D, Van Zandycke, S.M., Quain, D.E. and Smart, K.A. (2000) Replicative ageing and senescence in Saccharomyces cerevisiae and the impact on brewing fermentation. Microbiology 146, 1023–1034. 3. Pierce, J. (1970) Institute of Brewing Analysis Committee: measurement of yeast viability. J. Inst. Brew. 34, 306–312. 4. O’Connor-Cox, E.S.C., Mochaba, F.M., Lodolo, E.J. et al. (1997) Methylene blue staining: use at your own risk. Tech. Q. Master Brew. Assoc. Am. 34, 306–312. 5. Smart, K.A., Chambers, K.M., Lambert, I. and Jenkins, C. (1999) Use of methylene violet staining procedures to determine yeast viability and vitality. J. Am. Soc. Brew. Chem. 57, 18–23. 6. Haugland, P.H. (1998) Handbook of Fluorescent Probes and Research Chemicals, 6th edn. Molecular Probes, Eugene, OR. 7. Graham, R.K. and Caiger, P. (1969) Fluorescence staining for the determination of cell viability. Appl. Microbiol. 17, 489–490.

YEAST QUALITY AND FLUOROPHORE TECHNOLOGIES

161

8. Millard, P.J., Roth, B.L. and Kim, C.H. (1997) Fluorescence based methods for microbial characterisation and viability assessment. Biotechnol. Int. 1, 291. 9. King, L.M., Schisler, D.O. and Ruocco, J.J. (1981) Epifluorescent method for the detection of nonviable yeast. J. Am. Soc. Brew. Chem. 39, 52–55. 10. Trevors, J.T., Merrick, R.L., Russell, I. and Stewart, G.G. (1983) A comparison of methods for assessing yeast viability. Biotechnol. Lett. 5, 131–134. 11. McCaig, R. (1990) Evaluation of the fluorescent dye 1-anilino-8-naphthalene sulphonic acid for yeast viability determination. J. Am. Soc. Brew. Chem. 48, 22–25. 12. Wilson, H.A. and Chused, T.M. (1985) Lymphocyte membrane potential and Ca-sensitive potassium channels described by oxonol dye fluorescent measurements. J. Cell. Physiol. 125, 72–81. 13. Epps, D.E., Wolfe, M.L. and Groppi, V. (1994) Characterization of the steady-state and dynamic fluorescence properties of the potential-sensitive dye bis-(1,3-dibutylbarbituric acid) trimethine oxonol (Dibac4(3)) in model systems and cells. Chem. Phys. Lipids 69, 137–150. 14. Dinsdale, M.G., Lloyd, D., McIntyre, P. and Jarvis, B. (1999). Yeast viability during cider fermentation: assessment by energy metabolism. Yeast 15, 285–293. 15. Lloyd, D. and Dinsdale, G. (2000) From bright field to fluorescence and confocal microscopy. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 3–9. 16. Boyd, A., Attfield, P. and Veal, D. (2000) Evaluation of light scattering and autofluorescent properties of brewer’s wort for flow cytometric analysis of yeast viability. J. Inst. Brew. 106, 319–324. 17. Boulton, C.A., Box, W.G., Carvell, J. and Turner, K. (2001) A novel and rapid method for the automatic and simultaneous determination of total and viable cell concentration in pitching yeast slurries. Proc. Eur. Brew. Conv. Cong. 28, 78. 18. Combs, N. and Hatzis, C. (1996) Development of an epi-fluorescence assay for monitoring yeast viability and pre-treatment hydrolysate toxicity in the presence of lignocellulosic solids. Appl. Biochem. Biotechnol. 57/58, 649–657. 19. Millard, P.J., Roth, B.L., Truong Thi, H. et al. (1997) Development of the FUN-1 family of fluorescent probes for vacuole labelling and viability testing of yeasts. Appl. Environ. Microbiol. 63, 2897–2905. 20. Hutcheson, T.C., McKay, T., Farr, L. and Seddon, B. (1988) Evaluation of the stain Viablue for the rapid estimation of viable yeast cells. Lett. Appl. Microbiol. 6, 85–88. 21. Breeuwer, P., Drocourt, J.L., Rombouts, F.M. and Abee, T. (1994) Energy dependent, carrier-mediated extrusion of carboxyfluorescein from yeast Saccharomyces cerevisiae allows rapid assessment of cell viability by flow cytometry. Appl. Environ. Microbiol. 60, 1467–1472. 22. Chilver, M.K., Harrison, J. and Webb, T.J.B. (1978) Use of immunofluorescence and viability stains in quality control. J. Am. Soc. Brew. Chem. 36, 13–18. 23. Rotman, O. and Papermaster, B.W. (1966) Membrane properties of living mammalian cells as studied by enzymatic hydrolysis of fluorogenic esters. Proc. Natl. Acad. Sci. USA 55, 143–141. 24. Betts, R.P., Bankes, P. and Banks, J.G. (1989) Rapid enumeration of viable micro-organisms by staining and direct microscopy. Lett. Appl. Microbiol. 9, 199–202. 25. Lentini, A. (1993) A review of the various methods available for monitoring the physiological status of yeast: yeast viability and vitality. Fermentation 6, 321–327. 26. Deere, D., Shen, J., Vesey, G. et al. (1998) Flow cytometry and cell sorting for yeast viability assessment and cell selection. Yeast 14, 147–160. 27. Peladan, F. and Leitz, R. (1991) New method for the differential staining of dead and living cells of yeasts and bacteria. Proc. Eur. Brew. Conv. Cong., Lisbon, pp. 481–487. 28. Smart, K.A. (2001) Yeast quality: live and let dye. Brew. Guardian 130, 24–26. 29. Willets, J.C., Seward, R., Dinsdale, M.G. et al. (1997) Vitality of cider yeast grown micro-aerobically with added ethanol, butan-1-ol or isobutanol. J. Inst. Brew. 103, 79–84. 30. Murray, C.R., Barich, T. and Taylor, D. (1984) The effect of yeast storage conditions on subsequent fermentations. Tech. Q. Master Brew. Assoc. Am. 21, 189–194. 31. McCaig, R. and Bendiak, D.S. (1985) Yeast handling studies. I. Agitation of stored pitching yeast. II Temperature of storage of pitching yeast. J. Am. Soc. Brew. Chem. 43, 114–118, 119–123. 32. Pickerell, A.T.W., Hwang, A. and Axcell, B.C. (1991) Impact of yeast-handling procedures on beer flavor development during fermentation. J. Am. Soc. Brew. Chem. 49, 87–92. 33. Martens, F.B., Egberts, G.T.C., Kempers, J. et al. (1986) Yeast storage methods and their effects on fermentation. Monograph XII. Eur. Brew. Conv., Symp. Brewer’s Yeast, Vuoranta (Helsinki), Finland, pp. 95–107. 34. Werner-Washburne, M., Braun, E., Johnston, G.C. and Singer, R.A. (1993) Stationary phase in the yeast Saccharomyces cerevisiae. Microbiol. Rev. 57, 383–401.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

15 Vitality Assessment Using the Fluorescent Stain FUN1 S.M. VAN ZANDYCKE, O. SIMAL and K.A. SMART

Abstract The production of high-quality beer is dependent on the condition of the pitching yeast. In many breweries, the only routine test performed to assess this parameter is a viability measurement using methylene blue. However, the physiological state or vitality of the yeast represents a more suitable indicator of fermentation performance. There are many methods available to assess vitality, most of which involve measurement of yeast intracellular compounds (adenosine triphosphate, glycogen, trehalose, sterols), intracellular pH (ICP), extracellular pH (acidification power test) and released products from the cells (magnesium release test, adenylate kinase activity). However, most of these methods lack the reproducibility and sensitivity necessary to measure subtle changes in yeast physiological condition, and none of them allows the examination of individual cells within a heterogeneous population. FUN1 is a fluorophore dye, which has been demonstrated to differentiate between live and dead yeast cells. This dye is membrane permeable and stains live and dead cells green. In live cells, the formation of bright red cylindrical intravacuolar structures (CIVS) occurs in response to metabolic activity, which has not been clearly defined. Dead cells remain bright green. Preliminary experiments revealed that CIVS formation was reduced in stressed cells. It was also observed that FUN1 underestimated viability compared with other fluorescent and bright-field dyes. It is therefore suggested that FUN1 is inaccurate in determining viability, but may represent a potential indicator of vitality. The vitality of starved and oxidatively stressed cells was evaluated using FUN1. The levels of red fluorescence were assessed quantitatively using a fluorimeter with a microplate reader attachment. Vitality was also determined by the acidification power test and measuring levels of intracellular carbohydrate reserves. The intensity of red fluorescence decreased in starved cells and correlated with intracellular levels of glycogen. The response of oxidatively stressed cells appeared to be a strain-dependent phenomenon and correlated with the levels of glucoseinduced proton efflux. Further experiments involving yeast populations subjected to other physiological stresses and an investigation of the metabolic activity responsible for CIVS formation will reveal whether FUN1 is suitable to assess vitality and potentially to predict fermentation performance.

15.1 Introduction The physiological status of yeast slurry can affect final beer quality. Recycling of brewing yeast induces repeated exposure to stress from fermentation and storage conditions. This results in a progressive deterioration of the physiological status of the slurry. It is therefore important to be able to assess the viability and vitality of pitching or cropping yeast before fermentation. Viability corresponds to the percentage of live cells in a sample, whereas vitality represents the physiological state of the yeast that is viable.

VITALITY ASSESSMENT USING THE FLUORESCENT STAIN FUN 1

163

Several quantitative methods have been proposed for the practical evaluation of vitality; among them the improved acidification power test [measuring glucose-induced proton efflux (GIPE)] was observed to be an indicator of yeast physiological state.1 Measurement of glycogen, trehalose and sterol levels may also be used to predict adequate storage of yeast and ensure that the physiological condition is optimised before pitching for increased fermentation performance.2,3 The fluorescent stain FUN1 has been demonstrated to differentiate between live and dead cells.4 FUN1 is a membrane-permeable dye that binds to nucleic acids and generates green fluorescence (Figs 15.1 and 15.2). Biochemical processing of the dye in live cells of Saccharomyces cerevisiae gives rise to cylindrical intravacuolar structures (CIVS), which exhibit bright red fluorescence (Figs 15.1 and 15.2). The formation of CIVS is dependent on the temperature, presence of intracellular glutathione

Dead cells FUN1

Live cells Fig. 15.1

Mode of action of FUN1.

Fig. 15.2

Population of live and dead cells of lager strain L138 stained with FUN1.

164

BREWING YEAST FERMENTATION PERFORMANCE

and adenosine triphosphate (ATP) production.4 CIVS production was also observed to be reduced in the presence of acetic acid.5 FUN1 has been suggested to have a potential for yeast viability analysis using flow cytometry; however, it was observed that the staining procedure requires adjustment and calibration depending on the strain of yeast used.6 FUN1 has been used successfully to test for antimicrobial susceptibility by measuring the fluorescence emitted by the CIVS.7 In a previous study, it was suggested that this dye was inaccurate for assessing yeast viability microscopically (Chapter 14). It was observed that stressed cells stained with FUN1 exhibited reduced levels of CIVS compared with healthy populations and therefore could potentially represent a good indicator of vitality.

15.2 Materials and methods 15.2.1

Yeast strains and growth conditions

A lager (L138) and an ale (2593) strain were grown in sterile YPD broth (1% yeast extract, 0.5% bacteriological peptone, 1% glucose, 1.2% agar) at 25°C with shaking in a conical flask (250 ml) at 120 rpm. 15.2.2

Starvation and oxidative stress

Yeast cells previously grown for 48 h in YPD were resuspended to a final cell concentration of 1  107 cells/ml in water for 1–7 days, generating non-lethal starvation conditions, and in hydrogen peroxide (H2O2, 0.001–0.1%) for 1 h, generating nonlethal oxidative stress conditions. In both cases, flasks (250 ml) were incubated at 25°C with shaking at 120 rpm. 15.2.3

Acidification power test

The acidification power test was conducted according to the method of Siddique and Smart.1 The passive proton efflux was monitored for 10 min (AP10) before the addition of glucose (5 ml of 20.2%, w/v). The GIPE (GAP20) was then monitored for a further 10 min. Glucose acidification power (GAP) was calculated by adding the passive proton efflux to the GIPE (AP10  GAP20). The water acidification power (WAP) test was performed according to the same method described for GAP; however, glucose was replaced by sterile deionised water (5 ml) and WAP was obtained by adding AP10 to WAP20. The GIPE was calculated by subtracting WAP20 from GAP20. 15.2.4

Glycogen and trehalose

Glycogen and trehalose concentrations were determined using the method of Parrou and Francois.8 Yeast cells (1  109) were centrifuged at 4000 rpm for 5 min in 50 ml plastic centrifuge tubes, then washed three times with sterile distilled water, and supernatants were discarded. Then, 250 l of a 0.25 M solution of Na2CO3 was added to the pellet and the solution incubated for 2 h in a 95°C water bath with occasional

VITALITY ASSESSMENT USING THE FLUORESCENT STAIN FUN 1

165

stirring. To each tube, 600 l of a 0.2 M solution of sodium acetate buffered to pH 5.2 and 150 l of 1 M solution of acetic acid were added. Then, 500 l of each sample was transferred to an Eppendorf tube, one for glycogen and one for trehalose; 10 l of trehalase from porcine kidney (Sigma), containing 3 mUnits, was added to the trehalose tubes and 10 l of amyloglucosidase from Aspergillus niger (Sigma), at concentration of 10 mg/ml, was added to the glycogen tubes. Glycogen samples were incubated in a 57°C water bath and trehalose samples in a 37°C water bath for at least 8 h. Samples were then centrifuged at 4000 rpm for 5 min. The glucose concentration of the supernatant was determined using an enzyme-colour reagent solution following the manufacturer’s instructions (Glucose kit-510, Sigma Diagnostic). Levels of glycogen and trehalose were expressed as g glucose/1  108 cells per ml. 15.2.5

FUN1 stain for vitality assessment

FUN1 staining was performed following the procedure of Millard et al.4 Yeast cells were resuspended in sterile 2% D-()-glucose (1 ml) containing 10 nM Na-N-2hydroxyethylpiperazine-N-2-ethanesulfonic acid (HEPES), pH 7.2 (GH solution) to a final concentration of 1  107 cells/ml. Then 250 l of yeast cell suspension was added to 250 l of FUN1 solution (60 M) in a black, 96-well flat-bottomed microplate to a final concentration of 5  106 cells/ml. The intensity of red fluorescence was determined using a spectrofluorimeter equipped with a microplate reader with a 485 nm excitation filter and an emission filter centred at 640 nm.

15.3

Results and discussion

The mechanism by which FUN1 stains CIVS in vital cells is not known; however, the relationship between physiological state and CIVS formation was investigated. The vitality of healthy, starved and oxidatively stressed populations of lager (L138) and ale (2593) yeast strains was determined by monitoring the intracellular levels of glycogen and trehalose, proton efflux and red fluorescence levels from CIVS formation. 15.3.1

Determination of yeast cell vitality of starved stressed populations

It is postulated that starvation may have a negative effect on CIVS formation owing to the decrease in metabolic activity occurring in nutrient-limited cells. The intensity of red fluorescence for the lager strain L138 and the ale strain 2593 was observed to be high for healthy yeast cell populations. Subsequently, the intensity was observed to decrease during the first day of starvation and remained stable for up to 7 days. These results indicate a decrease in metabolic activity occurring with starvation (Fig. 15.3). For the ale strain 2593, a positive correlation (r  0.9717) was identified between the level of glycogen and the intensity of red fluorescence observed ( p  0.05) (Fig. 15.3). The same correlation was observed for the intensity of fluorescence and the concentration of trehalose (data not shown). This indicates that the metabolic activity responsible for CIVS formation is related to the availability of carbohydrate reserves for this strain. It is known that ATP is needed for the cells to

166

BREWING YEAST FERMENTATION PERFORMANCE

Lager strain L138

0

1 2 3 4 5 6 Duration of starvation (days)

Ale strain 2593

7

Intensity of red fluorescence

0

1 2 3 4 5 6 Duration of starvation (days)

7

Glycogen levels

Fig. 15.3 Comparison between the intensity of red fluorescence and the level of glycogen for starved cells of lager (L138) and ale (2593) strains.

produce CIVS.4 Therefore, it is postulated that when levels of glycogen and trehalose are low less ATP is produced, resulting in reduced CIVS formation. However, for the lager strain L138 no correlation (r  0.5512) could be established between the levels of glycogen and the intensity of red fluorescence (Fig. 15.3). The reasons for this observation are not known and merit further investigation. 15.3.2

Determination of yeast cell vitality of oxidatively stressed populations

Oxidative stress occurs during the brewing process when the yeast is in contact with oxygen, mainly during propagation and pitching. Reactive oxygen species are produced during normal cellular metabolism and can cause damage to cell membranes, DNA and proteins.9 Tolerance to oxidative stress is a strain-dependent phenomenon in brewing strains.10 The correct amount of oxygen should be supplied to the yeast for sterol formation, without causing any excessive oxidative damage. Given the decrease in metabolic activity observed with starved cells using FUN1, it was postulated that this dye may also be able to detect changes in the metabolic activity of oxidatively stressed cells. To verify this hypothesis, the occurrence of CIVS formation was followed in oxidatively stressed cells, resuspended in H2O2. For the lager strain, levels of red fluorescence were observed to be lower for healthy populations than for oxidatively stressed cells (Fig. 15.4). Red fluorescence increased with greater H2O2 concentration. In contrast, for the ale strain, levels of red fluorescence were observed to be higher for healthy populations than for oxidatively stressed cells (Fig. 15.4).

VITALITY ASSESSMENT USING THE FLUORESCENT STAIN FUN 1

167

Fig. 15.4 Comparison between the intensity of red fluorescence and glucose-induced proton efflux (GIPE) on oxidatively stressed cell populations for lager (L138) and ale (2593) strains.

A positive correlation (r  0.974) was obtained with the ale strain 2593 (Fig. 15.4) for CIVS formation and GIPE ( p  0.05). However, for the lager strain L138, a negative correlation (r  0.980) between these two parameters was observed ( p  0.05). The reason for this difference between the two strains is not known, but appears to be consistent for both stresses imposed in this study. 15.4

Conclusions

This preliminary study demonstrated that the formation of CIVS and the fluorescence intensity obtained following exposure to FUN1 are strain dependent and may vary depending on the stress applied. The mode of action of FUN1 and CIVS development requires further investigation; in particular, the apparent strain dependence observed requires elucidation. Acknowledgements Sylvie Van Zandycke is supported by Smart Brewing Services and Olivier Simal is supported by the European Undergraduate Exchange Programme ‘Socrates’. Katherine Smart is the Scottish Courage Reader in Brewing Science and a Royal Society Industrial Fellow, and gratefully acknowledges the support provided by Scottish Courage Brewing Limited and the Royal Society. References 1. Siddique, R. and Smart, K.A. (2000) An improved acidification power test. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 46–54.

168

BREWING YEAST FERMENTATION PERFORMANCE

2. Quain, D.E. and Tubb, R.S. (1982) The importance of glycogen in brewing yeast. Tech. Q. Master Brew. Assoc. Am. 19, 29–33. 3. Boulton, C.A., Clutterbuck, V.J. and Durnin, S. (2000) Yeast oxygenation and storage. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 10–18. 4. Millard, P.J., Roth, B.L., Truong Thi, H. et al. (1997) Development of the FUN-1 family of fluorescent probes for vacuole labelling and viability testing of yeasts. Appl. Environ. Microbiol. 63, 2897–2905. 5. Prudencio, C., Sansonetty, F. and Corte-Real, M. (1998) Flow cytometric assessment of cell structural and functional changes induced by acetic acid in the yeasts Zygosaccharomyces bailii and Saccharomyces cerevisiae. Cytometry 31, 307–313. 6. Deere, D., Shen, J., Vesey, G. et al. (1998) Flow cytometry and cell sorting for yeast viability assessment and cell selection. Yeast 14, 147–160. 7. Wenisch, C., Floris Linnau, K., Parschalk, B. et al. (1997) Rapid susceptibility testing of fungi by flow cytometry using vital staining. J. Clin. Microbiol. 35, 5–10. 8. Parrou, J.L. and Francois, J. (1997) A simplified procedure for a rapid and reliable assay of both glycogen and trehalose in whole yeast cells. Anal. Biochem. 248, 186–188. 9. Jamieson, D.J. (1998) Oxidative stress responses of the yeast Saccharomyces cerevisiae. Yeast 14, 1511–1527. 10. Martin, V., Quain, D.E. and Smart, K.A. (2000) The oxidative stress of ale and lager yeast strains. In: Brewing Yeast Fermentation Performance, Smart, K. (ed.). Blackwell Science, Oxford, pp. 97–104.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

16

Flow Cytometry: A New Tool in Brewing Technology K.J. HUTTER and C. LANGE

Abstract As a result of the proliferation of individual yeast cells during the fermentation process, the growth conditions in the culture are changing continuously. Current process control (i.e. temperature, pH, CO2, extract decomposition, ethanol content, etc.) is orientated towards the behaviour of the overall population and not towards individual cells. However, it is the individual cell that is affected by changes in nutrient substrate and growth parameters in a static culture. The empirical assumption that the inoculated yeast cells pass through serial repitching with the same vitality, activity and productivity is basically wrong. Each cell has its own individual cell cycle, dependent on the particular growth conditions. It is therefore necessary to analyse fermentation process continuously, by measuring the direct influence of chemical and physical process variables on yeast at every point during growth. The objective of this study was to use flow cytometry for direct process evaluation. Several flow-cytometric (and image analytical) parameters have been developed to examine industrial yeast strains in the process. Using modern fluorescence optical techniques the biological condition of yeast cells can be evaluated directly, thus allowing immediate intervention in controlling and modelling the fermentation process. Current flow-cytometric (and image analytical) process control is demonstrated with selected examples in this chapter.

16.1 Introduction As a result of the proliferation of individual yeast cells during the fermentation process, growth conditions are never constant, and change with every repitching. Current process control (i.e. temperature, pH, CO2, extract decomposition, ethanol content, etc.) is orientated towards the behaviour of the overall population and not towards individual cells. It is the individual cell, however, that is either stimulated or retarded by changes in nutrient substrate and growth parameters of a static culture. The assumption, based on empirical studies, that inoculated yeast cells pass through serial repitching with unchanged vitality, activity and productivity is basically wrong. Each cell has its own individual cell cycle, depending on its particular growth conditions. Therefore, it is necessary to analyse the fermentation process continuously, i.e. to measure the direct influence of chemical and physical process variables on yeast at any time during growth. The objective of this presentation was to use flow cytometry for direct process evaluation. Many staining procedures have been developed over several decades to monitor flow-cytometric (and image analytical) parameters to determine industrial yeast strains in the fermentation process (Table 16.1). Using modern fluorescence optical techniques the biological condition of yeast cells can be evaluated directly. Therefore, immediate intervention in controlling and modelling the fermentation process is possible. Current flow-cytometric (and image analytical) process control is presented by means of cell-cycle analysis and glycogen content determination. The detection of contaminants is also shown.

170

BREWING YEAST FERMENTATION PERFORMANCE

Table 16.1 No.

Biomonitoring of the fermentation process Flow cytometric application

1 2 3

Ageing of yeast cells Apoptosis of yeast cells Cell cycle analysis

4 5 6

Cell volume/cell count Contaminants Flocculation

7 8

Glycogen content 3-Hydroxysterols

9 10 11 12

Intracellular pH value Mitochondrial fluorescence Neutral lipid content Viability testing

Ref. Count of fluorescing bud scars Detection of micronuclei in yeast cells Characterisation of the cell cycle phase of yeast cells during the process Scatter signal corresponds to cell volume Detection of beer spoilage microorganisms Affinity for special residual sugars on the yeast membrane by lectins Intracellular reserve compounds of yeast cells Information on sparking and bulk membrane functions Viability test Viability test Reserve compounds of yeast cells Esterase activity of yeast cells

1 2 3 4 5 6 7 8 9 9 8 10

16.2 Materials and methods 16.2.1

Glycogen content

After fixation for at least 24 h in 70% ethanol at 4°C the yeast pellet was incubated in 1 ml 1 N HCl for 50 min. HCl was removed by centrifugation and the pellet was stained with acriflavine solution for 1 h (excitation: 488 nm). Further details are described elsewhere.7 16.2.2

DNA content

Fixed yeast cells were incubated in 0.1% RNase for 1 h exactly at 37°C. After removal of the enzyme solution, the pellet was resuspended in 1 ml phosphate-buffered saline (PBS), and 100 l of a propidium iodide staining solution was added. After 1 h of incubation cells were measured by flow-cytometric analysis (excitation: 488 nm).3 16.2.3

Detection of beer spoilage contaminants

Beer samples (each 50 ml) were filtered through a membrane filter (Nuclepore). The filter was placed on a glass slide. The filtration area was then stained with 10 l of a staining solution (kit containing fluorescein diacetate and propidium iodide). After this procedure and excitation at 488 nm, viable cells fluoresce green while dead cells fluoresce red. This spectral differentiation can be detected directly using a fluorescence microscope (equipped with a steerable slide table)5 or an image analysis system. 16.2.4

Flow cytometry

Flow-cytometric analysis was carried out with an EPICS II (Coulter, Krefeld, Germany), equipped with an argon ion laser at 488 nm and/or a PAS from Partec (Münster, Germany). The flow device from Partec was fitted with a mercury high-pressure lamp, HBO 100, and an argon ion laser at 488 nm.

FLOW CYTOMETRY

171

Fig. 16.1 Closed cylindroconical fermentation becomes transparent by flow-cytometric analysis. Representative phases of aerobic and anaerobic growth of industrial yeasts are shown in seven histogams: (1) inoculation (during the start of the lag phase the cells are in a quiescent phase); (2) initial phase (transition from lag to log phase); (3) young krausen (start of exponential growth); (4) high krausen (intensive exponential growth, most yeast cells are in the budding phase); (5) low krausen (exponential growth); (6) retardation of exponential growth; and (7) ready-for-hosing state (beginning of fermentation).

16.3

Results and discussion

Since the early 1990s this group has been using fluorescence–optical methods for controlling fermentation processes as well as detecting beer spoilage microorganisms. From numerous possibilities three examples were chosen for this presentation to demonstrate the efficiency of these techniques: (i) cell cycle analysis to observe growth phases during the running process;3,11 (ii) determination of glycogen content to obtain information on the physiological status of single yeast cells (i and ii were both performed by flow cytometry7); and detection of beer spoilage microorganisms (by fluorescence microscopy).5 DNA content is the most important parameter to control growth during the propagation, proliferation, fermentation and storage of yeast cells. Figure 16.1 shows different histograms of lager yeast, corresponding to characteristic growth phases during lag phase, log phase and stationary proliferation. This quasi on-line analysis makes closed cylindroconical fermentation transparent. Figure 16.2 shows the oscillating course of glycogen synthesis of industrial lager yeasts during proliferation. At the time of inoculation the yeast cells are in the quiescent phase and the glycogen content is relatively low. The beginning of exponential growth is characterised by an increase in glycogen content, while during intensive exponential growth the glycogen content decreases. At the end of fermentation the glycogen content increases. The glycogen content of yeast cells indicates disadvantageous growth conditions during fermentation, such as starvation and low temperature. The rapid production and distribution of beer requires the rapid detection of beer spoilage microorganisms.5 Membrane filtration and a fluorescence staining kit can be used to stain viable and dead cells simultaneously. Lower eukaryotes and prokaryotes can also be detected (Fig. 16.3). Stained microbials can be detected by fluorescence

172

BREWING YEAST FERMENTATION PERFORMANCE

35

Glycogen-Peakmaxima

30 25 20 15 10 5 0 0

1

6 Time

Yeast stored at 5° C

11

19

26

32

Yeast stored at 22° C

Fig. 16.2 Oscillating course of the glycogen content of yeast cells. Storage at low temperature and nutrient limitation affect the utilisation of glycogen.

Fig. 16.3 Simultaneous staining of viable and dead cells with a staining kit containing fluorescein diacetate and propidium iodide. Viable cells fluoresce green while dead cells are fluoresce red after excitation by the blue light of an argon ion laser.

FLOW CYTOMETRY

173

microscopy or by more expensive image analysis. The determination of DNA and glycogen content should be performed before filtration takes place. Flow-cytometric analysis provides rapid and reproducible results during the running process. The propagation and proliferation of yeast can be successfully optimised, thus shortening the fermentation process.

Acknowledgement We thank Wifö of Deutscher Brauerbund e.V. for support of B60.

References 1. Pringle, J.R. (1971) Staining of bud scars and other cell wall chitin with calcofluor. Methods Enzymol. 194, 732–735. 2. Hutter, K.-J. and Lange, C. (2001) Yeast management and control of fermentation process by flow cytometry. Monographs Proc. Brew. Conv., Budapest. 3. Hutter, K.-J. and Eipel, H.E. (1978) DNA determination of yeast by flow cytometry. FEMS Microbiol. Lett. 3, 35–38. 4. Hutter, K.-J. and Schärfe, J. (1997) Zellzahl- und Zellvolumenanalysen. Brauwissenschaft 50, 4–11. 5. Hutter, K.-J. (2000) Detection of beer spoiling contaminants by image analysis. Fermentation 13(4), 57–58. 6. Hutter, K.-J., Remor, M. and Borek, M. (2001) Über das zellzyklusabhängige Flockulationsverhalten unter- und obergäriger Betriebshefen. Brauwissenschaft 54. 7. Hutter, K.-J., Remor, M. and Müller, S. (2000) Bestimmung des Glykogengehaltes der Betriebshefe. Brauwissenschaft 53, 68–76. 8. Hutter, K.-J. and Müller, S. (1996) Zellzyklus und 3 Hydroxysterolgehalt. Brauwissenschaft 49, 234–239. 9. Rothe, G., Oeser, A. and Valet, G. (1988) Dihydrorhodamine 123: a new cytometric indicator for respiratory burst activity in neutrophil granulocytes. Naturwissenschaften 75, 354–355. 10. Rotman, B. and Papermaster, B.W. (1966) Membrane properties of living mammalian cells as studied by enzymatic esters. Proc. Natl Acad. Sci. U.S.A. 55, 134–141. 11. Hutter, K.-J., Herber, M. and Lindemann, B. (1995) DNS-Gehalt und Zellzyklusanalysen verschiedener Betriebshefen. Brauwissenschaft 48, 184–190.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

17 Comparison of the Methylene Blue Assay with a New Flow-cytometric Method for Determining Yeast Viability in a Brewery A. BOYD, T. GUNASEKERA, P. ATTFIELD, K. SIMIC, S. VINCENT and D. VEAL

Abstract The accurate measurement of yeast viability is a key parameter for quality control of brewery fermentations. The most widely used assay for measuring viability involves staining cells with methylene blue and visualising with microscopy. However, this assay is known to overestimate viability, and analysis is subjective and thus subject to operator error. Furthermore, the relatively low number of cells that are analysed using this method (1000) may not be representative of the entire yeast population. A viability assay was developed that involves staining cells with a fluorescent dye, oxonol, which provides an accurate measure of yeast viability. Subsequent analysis of cells by flow cytometry offers a number of advantages. Analysis is automatic, and the cytometer can analyse individual cells at rates of thousands per second. Therefore, viability data for a statistically significant number of cells can be determined rapidly. Flow cytometry can successfully analyse oxonol-stained brewing yeast samples, without interference from background noise (wort particles). The assay was optimised for brewery conditions and then trialled at a brewery for 6 months. Yeast samples (n  179) were analysed using the methylene blue method, and by flow cytometry, and the results compared. The flow-cytometric assay provided highly reproducible viability results (1% variation) compared with methylene blue (6%). Operator error when using methylene blue (5%) was also substantially reduced by flow cytometry (1%). The flow-cytometric assay described enables the analysis of a significant number of cells (typically 50 000), within approximately 1 min. The results indicate that the assay has potential applications in process and quality control in breweries.

17.1 Introduction The brewing industry requires a rapid and reliable yeast viability assay. Current methods are either slow or unreliable. Results from plate counts, for example, take many days, and cells that have lost replicative ability may still possess fermentative activity.1 The methylene blue assay is comparatively rapid. However, it is inconsistent, with many reports of this dye significantly overestimating numbers of live cells1,2 and others finding the opposite.3 Further, analysis is subjective, and it is questionable whether the relatively low number of cells able to be analysed is representative of the entire yeast population. This paper describes a method that involves staining cells with the fluorescent dye oxonol.4–6 Flow-cytometric (FCM) analysis of oxonol-stained cells is objective and a large number of cells can be analysed individually within a short space of time.7 Background noise (i.e. wort particles) can be ignored by the software.8

COMPARISON OF THE METHYLENE BLUE ASSAY

175

This chapter reports the results of a validation trial of the oxonol/flow cytometry yeast viability assay (FCM assay), at a large commercial Australian brewery. This trial compared the viability of samples obtained with both the FCM and methylene blue assays. The various errors involved in both methods were also investigated.

17.2 Materials and methods 17.2.1

Trial location and yeast analysed

The comparison of yeast viability assays was performed at Kent Brewery [Sydney, NSW, Australia – Carlton and United Breweries (CUB)]. The brewing yeasts analysed were lager strains A and J (Saccharomyces carlsbergensis), and ale strain O (Saccharomyces cerevisiae). In total, 179 samples were analysed over a 6 month period. Strains A, J and O made up 79, 15 and 6% of samples, respectively. 17.2.2

Methylene blue staining and microscopic analysis

Methylene blue solution consisted of 0.01% methylene blue (Ajax, L.R. grade) and 2% sodium citrate dihydrate in distilled water. Yeast slurries (1 ml) were diluted in Ringer’s solution (9 ml). Equal volumes of methylene blue and diluted yeast cells were wet-mounted on a glass slide under a coverslip. Analysis was by light microscopy, using a Nikon LabopHot-2 microscope (40  objective, 15  eyepiece). Four-hundred cells were counted and viability was expressed as the percentage of cells that were not blue. 17.2.3

Oxonol staining and flow-cytometric analysis

Flow cytometry was performed at the brewery on a FACSCan cytometer (BectonDickinson, Sydney, NSW, Australia). The threshold was on forward scatter (FSC) at 253 V. FSC and side scatter (SSC) detector voltages were E-1 and 273 V, respectively. Fluorescence detectors were adjusted to 475 V. No compensation was used. Sample analyses in the laboratory were performed on a FACSCalibur cytometer (BectonDickinson). Settings were identical, except that the fluorescence detectors were adjusted to 400 V, to give a similar mean fluorescence of TruCount™ cytometer calibration beads (Becton-Dickinson) to that observed on the FACSCan. Sheath fluid (Osmosol, Lab Aids Pty, Sydney) was passed through a 0.22 M filter on both cytometers. The fluorescent dye used was oxonol [DiBAC4(3); bis-(1,3-dibutylbarbituric acid) trimethine oxonol], obtained from Molecular Probes (Eugene, OR, USA). Yeast slurry samples (100 l) were diluted in distilled water (900 l). For staining, TruCount™ tubes (Becton-Dickinson) were filled with combinations of 10 l of diluted yeast (final concentration of approximately 2  106 cells/ml), and 3 l of oxonol (final dye concentration of 1.5 g/ml), with the tubes being made up to 1 ml with distilled water. Tubes were vortex mixed (5 s), incubated (10 min, 21°C) and further vortex mixed (10 s) before FCM analysis. TruCount™ bead regions were set using an FSC vs FL3 (red fluorescence) dotplot, where the greatest separation between beads and yeast

176

BREWING YEAST FERMENTATION PERFORMANCE

was observed. One-thousand bead events were measured, resulting in approximately 50 000 yeast cell events also being acquired. Data were analysed using CellQuest software (Becton-Dickinson), with viability (FCM viability) being expressed as the percentage of live cells [low fluorescent population on an FSC vs FL1 (green fluorescence) dotplot] compared with total cells (yeast population on a FSC vs SSC dotplot). 17.2.4

Statistical analyses

Linear correlations (r) were calculated with the Microsoft Excel program.

17.3 17.3.1

Results and discussion Comparison of viability assays

Samples of brewery yeast were analysed by both FCM and methylene blue assays. FCM analysis of a typical sample showed two separate yeast populations on an FSC versus FL1 dotplot (Fig. 17.1). Workers in this laboratory have shown previously that the highly fluorescent oxonol-stained population (above approximately 1000 units on

Fig. 17.1 Typical flow-cytometric appearance of a brewery yeast slurry sample. Slurry diluted 1:1000 was stained with 3 l of oxonol in a TruCount™ tube. A plot of forward scatter (FSC) vs FL1 (green fluorescence) enabled the distinction of live yeast from dead yeast.

177

Methylene blue assay viability

COMPARISON OF THE METHYLENE BLUE ASSAY

100 90 80 70 r = 0.87 60 50 50

60

70 80 Flow-cytometry assay viability

90

100

Fig. 17.2 Viability analysis of brewery samples using the methylene blue and flow-cytometric assays (n  179).

this FL1 detector setting) typically cannot be cultured on agar plates after FCM cell sorting, while the low-to-medium fluorescent populations can be cultured.6 The correlation between viability determined by the methylene blue and FCM assays was r  0.87 for all samples (Fig. 17.2). This relationship was somewhat skewed by the low number of samples (15%) that were below 85% viability by both methods. The correlation between both methods with samples above this viability was 0.39. Although methylene blue has been reported to overestimate viability,1,2 only 32% of samples gave a higher viability by methylene blue than by FCM. With samples below 85% viable by FCM, where methylene blue is reported to overestimate viability,2 only 39% were higher in viability by methylene blue than FCM, and 58% vice versa. To investigate this further, a brewery sample stained with both oxonol and methylene blue was analysed microscopically. Analysis of several hundred cells showed that highly green fluorescent cells (when excited by blue light) were also dark blue (when switching to white light for excitation). Therefore, it appeared that methylene blue detected the majority of dead cells in a typical sample of brewing yeast. However, methylene blue does not stain cells that have been heat killed (authors’ unpublished observations). 17.3.2

Operator error and reproducibility of viability data

Both methods were analysed for two types of potential error. The first was reproducibility. Three separate samples were analysed twice by both assays. The average standard deviation from each mean was 6.1% for methylene blue and 0.2% for FCM. Operator error was also investigated, i.e. the viability resulting when the same sample was analysed by two different operators. Operator error for the FCM assay was significantly lower than that for the methylene blue assay (Fig. 17.3). None of the samples analysed by methylene blue, compared with almost two-thirds by FCM, gave the same viability result when analysed by separate operators. Two methylene blue-stained samples gave viability differences of 13%, while no samples analysed by FCM varied by more than 5%. The correlation (r) between viabilities obtained by both operators was 0.689 for methylene blue and 0.998 for FCM.

178

BREWING YEAST FERMENTATION PERFORMANCE

70

% of samples

60 50 40

Me Blue FCM

30 20 10 0

No Difference

1%

1–5%

5–10%

>10%

Sample viability difference between two operators Fig. 17.3 Comparison of operator error using the methylene blue (Me Blue) and flow-cytometric (FCM) viability assays. Samples were analysed by two separate operators. n  36 and 29 for methylene blue and FCM analysis, respectively.

Although the methylene blue result was found to be variable in terms of the operator, the low reproducibility of the assay found when one operator was analysing a sample (6.1%) indicates that the low number of cells analysed was the major source of error. More evidence for this conclusion comes from the fact that most cells were stained dark blue, and that slightly blue cells, which would be prone to subjective analysis, were rarely encountered during the trial. The low reproducibility of the methylene blue assay is also the likely cause of the low correlation (r  0.39) between the two methods, when samples were more than 85% viable by FCM. The large number of cells counted by the FCM assay (typically 50 000) addresses the problem of low sample number encountered with the methylene blue assay. The error involved in microscopic counts can be calculated using the formula: error  (100/n )%.9 Thus, at least 10 000 cells need to be counted to reduce this error to below 1%. This is clearly not practical using microscopic methods, and the more reasonable number of 400 cells gives an error of 5%.

17.4

Conclusions

This chapter has presented results of a brewery trial of a new method for determining yeast viability, using a fluorescent dye, oxonol, which stains dead cells bright green. Combined with FCM, the assay automates the counting of tens of thousands of cells within 1 min. All brewery yeast samples analysed with the oxonol assay gave a clear distinction between live and dead cells. This allowed a viability number to be calculated easily. Both the methylene blue and the FCM assay described here take about 10 min for a result. While FCM analysis takes only about 1 min, 10 min incubation is required for full fluorescence of cells to develop (data not shown). However, the relative fluorescence between the live and dead populations is identical immediately after mixing

COMPARISON OF THE METHYLENE BLUE ASSAY

179

cells and staining. Therefore, an experienced operator would potentially be able to determine sample viability, using the FCM assay, with minimal incubation time. Flow cytometers also have an important advantage in that they are not restricted to one particular assay. For example, the brewing industry is also interested in the development of a rapid assay for yeast fermentation performance, or vitality.10 As most cytometers possess at least three fluorescence detectors,7 one could theoretically stain a cell concurrently with a viability dye that fluoresces green and a vitality dye that fluoresces orange or red. Flow cytometers can also be used to detect contaminating microorganisms in the brewing process.11 It is envisaged that flow cytometry will be used increasingly in breweries in the coming years as their many advantages are realised.

Acknowledgements We thank BrewTech Pty Ltd, Melbourne, Australia, and the Australian Research Council, for funding assistance. We are also grateful for assistance provided by members of the microbiology team at Kent Brewery.

References 1. Jones, R.P. (1987) Measures of yeast death and deactivation and their meaning: Part II. Process Biochem. 8, 118–128. 2. Pierce, J.S. (1970) Institute of brewing: analysis committee. Measurement of yeast viability. J. Inst. Brew. 76, 442–443. 3. Willetts, J.C., Seward, R., Dinsdale, M.G. et al. (1997) Vitality of cider yeast grown micro-aerobically with added ethanol, butan-1-ol or iso-butanol. J. Inst. Brew. 103, 79–84. 4. Lloyd, D. and Hayes, A.J. (1995) Vigour, vitality and viability of microorganisms (Review). FEMS Microbiol. Lett. 133, 1–7. 5. Lloyd, D., Moran, C.A., Suller, M.T.E. et al. (1996) Flow cytometric monitoring of rhodamine 123 and a cyanine dye uptake by yeast during cider fermentation. J. Inst. Brew. 102, 251–259. 6. Deere, D., Shen, J., Vesey, G. et al. (1998) Flow cytometry and cell sorting for yeast viability assessment and cell selection. Yeast 14, 147–160. 7. Shapiro, H.M. (1988) A Practical Guide to Flow Cytometry. Alan R. Liss, New York. 8. Boyd, A.R., Attfield, P., Vincent, S.F. and Veal, D.A. (2000) Evaluation of light scattering and autofluorescent properties of brewer’s worts for flow cytometric analysis of yeast viability. J. Inst. Brew. 106, 319–324. 9. Chang, W.L., Van der Heyde, H.C. and Klein, B.S. (1998) Flow cytometric quantitation of yeast a novel technique for use in animal model work and in vitro immunologic assays. J. Immunol. Methods 211, 51–63. 10. Lentini, A. (1993) A review of the various methods for monitoring the physiological status of yeast: yeast viability and vitality. Fermentation 6, 321–327. 11. Jespersen, L., Lassen, S. and Jakobsen, M. (1993) Flow cytometric detection of wild yeast in lager breweries. Int. J. Food Microbiol. 17, 321–328.

Part 5 The Role of Brewing Yeast in Beer Flavour Development

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

18 Formation and Disappearance of Diacetyl During Lager Fermentation C. BOULTON and W. BOX

Abstract The results are presented of a survey of patterns of vicinal diketone (VDK) formation and disappearance in production- and pilot plant-scale lager fermentations. The influence of major fermentation variables is discussed. Evidence is presented that both wort composition and yeast physiological condition are important parameters. Results are given which confirm that during most of fermentation the concentration of free diacetyl is much lower than that of -acetolactate. This supports the contention that spontaneous decarboxylation of the latter is rate determining, at least, during most of fermentation. In laboratory studies, it has been demonstrated that brewing yeast is capable of rapid reduction of exogenous diacetyl, but this ability decreases during the course of fermentation. Furthermore, reduction proceeds slowly at low initial concentrations of diacetyl. It has been proposed that the deceleration in the rate of decline of total VDK concentration, commonly observed at the end of fermentation, may be due to sedimentation of yeast. Thus, few cells are available in the body of the beer for diacetyl reduction. The evidence presented here contradicts this supposition and suggests that the decrease in rate may be a function of the physiological state of yeast at the end of fermentation. It is proposed that in the latter stages of fermentation, the uptake and/or reduction of free diacetyl by yeast may be rate determining. This suggests that the physiological condition of yeast at the end of fermentation and the conditions pertaining do not promote the efficient removal of diacetyl. This is supportive of those strategies for diacetyl removal that are performed on green beer removed from the fermenter immediately after the achievement of racking gravity.

18.1 Introduction The vicinal diketone (VDK) diacetyl (2,3-butanedione) has a strong aroma and taste of butterscotch or toffee. In lager beers it has a flavour threshold of approximately 0.05 ppm and its presence is considered undesirable. It is generally accepted that it is produced as a result of yeast metabolism during fermentation.1 It derives from pyruvate via the intermediary of -acetolactate, a precursor of valine biosynthesis. -Acetolactate is excreted into wort, where it spontaneously oxidatively decarboxylates to form diacetyl. During the warm phase of conditioning, the latter is assimilated by yeast and reduced to less flavour-active metabolites, acetoin and 2,3-butanediol. It is considered that the spontaneous decarboxylation of -acetolactate is the ratedetermining step in the pathway. For many lager fermentations, total vessel residence time is governed by the time taken for the diacetyl concentration (measured as the sum of -acetolactate and diacetyl, or total VDK) to achieve a minimum specified value. Only then can the beer be chilled, the yeast cropped and the vessel racked. Current regimens of fermentation control aim to produce consistent and minimal vessel residence times. Control is exerted by rigorous regulation of the conditions at the start of fermentation. The tacit assumption of this approach is that wort composition

184

BREWING YEAST FERMENTATION PERFORMANCE

and yeast physiological condition are not significant variables. Furthermore, precise control of the major user variables, wort dissolved oxygen concentration, yeast pitching rate and temperature, exert an equal effect on both primary and secondary fermentation. In this paper, results are presented of a detailed survey of production lager fermentations. In particular, the relationships, if any, that exist between the duration of primary and secondary fermentation are examined. The effect is determined of the major fermentation variables on the duration of primary fermentation and patterns of VDK formation and disappearance. Previous work has shown that variability in the physiological condition of pitching yeast results in inconsistencies in the duration of primary fermentation.2 With regard to the duration of VDK stand times it is generally assumed that wort composition, the total suspended yeast count and temperature are of most influence. Since the uptake and reduction of diacetyl are not considered rate determining in the diacetyl cycle, by inference, the physiological condition of the yeast may be regarded as having little significance. Here, results are presented of an investigation into the ability of brewing yeast to reduce exogenous diacetyl. The results demonstrate that this ability is influenced by the physiological condition of the yeast. The potential impact of this on the duration of diacetyl stand times is discussed.

18.2 Materials and methods Production-scale lager fermentations were performed in cylindroconical vessels containing 1600 hl of 16° Plato wort. Pilot plant fermentations used similar wort and cylindroconical vessels with an operating volume of 8 hl. Fermentation conditions were as described in Section 18.3. The same production lager yeast strain was used throughout. Where appropriate, yeast was obtained from brewery storage vessels. In some experiments, yeast of a defined physiology was cultivated in the laboratory using yeast extract (5% w/v) and peptone (10% w/v) supplemented with a carbon source, as indicated in the legend to the appropriate figure. In some experiments, yeast was oxygenated, as described previously.3 Laboratory fermentations were performed in 1.5 litre stirred glass vessels using brewery wort. Before pitching, at a rate of 15  106 viable cells/ml, the wort was saturated with air. During fermentation, the fermenter headspace was sparged with nitrogen. Total VDK in samples removed from the fermenter was determined using gas–liquid chromatography.4 In some experiments, diacetyl concentration was determined by a spectrophotometric procedure.5 Diacetyl reduction by yeast was assessed using an assay described previously.5

18.3

Results and discussion

Historical data for a number of production-scale lager fermentations were examined to determine the extent of variability in the duration of primary fermentation (Fig. 18.1a)

FORMATION AND DISAPPEARANCE OF DIACETYL

185

50

30 20

(a)

221–230

211–220

201–210

191–200

181–190

171–180

161–170

151–160

141–150

131–140

121–130

111–120

101–110

81–90

91–100

71–80

0

61–70

10

51–60

No. fermentations

40

Time to gravity (h) 50 45

No. fermentations

40 35 30 25 20 15 10

(b)

161–170

151–160

141–150

131–140

121–130

111–120

101–110

91–100

81–90

71–80

61–70

51–60

41–50

31–40

0

21–30

5

VDK stand (h)

Fig. 18.1 Consistency of performance shown as (a) time to achieve racking gravity and (b) time to achieve VDK specification for 1600 hl cyclindroconical lager fermentations (n  272).

VDK stand (h)

200 150 100 50 0 80

130

180

230

280

Time to gravity (h)

Fig. 18.2 Relationship between time to achieve racking gravity and time to achieve VDK specification for 1600 hl cyclindroconical lager fermentations (n  311).

and time to achieve diacetyl specification (Fig. 18.1b). As is readily apparent, there was considerable variability in both of these facets of fermentation performance; however, it was greater in the case of the diacetyl stand time. Similar data are plotted in Fig. 18.2 to show the relationship between primary fermentation and diacetyl stand

BREWING YEAST FERMENTATION PERFORMANCE

Std. deviation (days)

186

2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 1

3

5

7

9 11 13 15 17 19 Month

Fig. 18.3 Consistency of cycle time for 1600 hl cylindroconical lager fermentations over a period of 20 months. Data for each month represent the mean of at least 15 fermentations. At the point indicated by the arrow, an automatic pitching rate control system was introduced.

Table 18.1 Fermentation consistency before and after installation of an automatic pitching rate control system Time to racking gravity (h)

VDK stand time (h)

Before automatic pitching rate control (n  149) Mean (h) 141 95.6 Range (h) 192 167 SD 28.8 32.6 After automatic pitching rate control (n  138) Mean (h) 120 81.6 Range (h) 90 120 SD 14.1 21.5 VDK: vicinal diketone; SD: standard deviation.

times. Here, it may be seen that there is no apparent correlation between these two parameters. It follows that examination of the patterns of primary fermentation would be of little value in terms of predicting overall cycle times. In the period during which these production fermentations were performed, it was suspected that some of the observed inconsistency was due to poor control of yeast pitching rate. Consequently, a manual pitching rate control system was replaced by an automatic in-line procedure.6 This resulted in an immediate reduction in fermentation cycle times and a significant improvement in consistency (Fig. 18.3). However, as shown in Table 18.1, these improvements were largely restricted to changes in primary fermentation. Despite improved control of pitching rate the underlying inconsistency in diacetyl stand times was unchanged. The effects on fermentation performance of other variables were studied at pilot scale. The effect of varying fermentation temperature on the VDK profile is shown in

FORMATION AND DISAPPEARANCE OF DIACETYL

187

Total VDK (ppm)

1.8 20°C

1.6 1.4 1.2

15°C

1.0 0.8 0.6 0.4 0.2

13°C

0 0

50

100

150

200

250

300

Time (h) Fig. 18.4 Effect of temperature on the VDK profile of 8 hl pilot-scale cyclindroconical lager fermentations.

[VDK] (ppm)

2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 G

F

Pi

tc

0

E

hi

ng

ra

50

D te

100

C B

150 200

)

e (h

Tim

Fig. 18.5 Effect of yeast pitching rate on the VDK profile of 8 hl pilot-scale cyclindroconical lager fermentations. Pitching rates were 6, 9, 12, 15, 18 and 24  106 cells/ml in B, C, D, E, F and G, respectively.

Fig. 18.4. Increase in temperature, within the range tested, resulted in a progressive increase in the magnitude of the VDK peak. There was a concomitant increase in the rates of VDK accumulation and dissimilation such that higher fermentation temperatures were associated with shorter VDK stand times. An increase in yeast pitching rate also resulted in an increase in the size of the VDK peak, but in this case the duration of the VDK stand time was relatively unaffected (Fig. 18.5). The effect of varying both pitching rate and the initial wort dissolved oxygen concentration is shown in Fig. 18.6. Predictably, the most rapid rates of primary fermentation were observed where both of these parameters were high (Fig. 18.6a). However, the effects on overall fermentation performance were much less pronounced (Fig. 18.6b).

188

BREWING YEAST FERMENTATION PERFORMANCE

70 60 50 40 Time to half gravity (h) 30 20 10

(a)

6 Pit chi 9 12 ng rat e ( 15 10 6 18 cel 24 ls/m l)

0 8 15 ) p 26 p m C( 32 O rt D Wo

250 200 150 100

Time to VDK spec. (h)

50

(b)

Pit 6 9 chi 12 ng rat e ( 15 18 10 6 cel 24 ls/m l)

0 32

8

26 ) 15 (ppm OC D t r Wo

Fig. 18.6 Effect of varying pitching rate and dissolved oxygen concentration (DOC) on (a) the rate of primary fermentation and (b) overall time to achieve diacetyl specification for 8 hl pilot-scale cyclindroconical lager fermentations.

The results presented so far indicate that the major user variables exert their effects principally on primary fermentation. Only elevated temperature promoted an increase in the rate of primary fermentation and a shortening of VDK stand times. This suggested that other factors were responsible for the observed inconsistencies in VDK metabolism during fermentation. The influence of wort composition and yeast line was examined in a further series of pilot plant fermentations. Duplicate fermentations using identical oxygenated brewery wort were pitched at the same viable pitching rate. Each pair of fermentations was pitched with brewery yeast, of the same type and similar generational age but from different propagations. Vessels were racked at the same time at a point at which previous experience would suggest that the VDK specification

189

FORMATION AND DISAPPEARANCE OF DIACETYL

Table 18.2 Rates of primary fermentation and patterns of yeast growth for serially repitched pilot plant 8 hl cylindroconical lager fermentations

Mean time to racking gravity (h) SD Mean maximum suspended yeast count (106/ml) Mean yeast growth (total crop  pitch) (g/l wet wt)

Yeast line 1

Yeast line 2

86.7 9.2 38.2 (36.1–43.4) 15.5 (14.8–16.0)

84.1 8.8 40.4 (38.7–44.6) 15.6 (15.2–15.8)

0.25

[VDK] (ppm)

0.2 0.15 0.1 0.05 0 0

5

10

15

Generation no. Fig. 18.7 Concentration of total VDK at rack for 12 pairs of serially repitched 8 hl pilot-scale cyclindroconical lager fermentations.

would have been achieved. The cropped yeast from the first pair of fermentations was used to repitch a further duplicate pair using a different batch of brewery wort. This procedure was repeated through 12 generations; on each occasion a different batch of brewery wort was used. The primary fermentations and extent of yeast growth were remarkably consistent. No significant differences between yeast lines were observed and serial repitching had no apparent effect (Table 18.2). The VDK concentration in green beer at rack is shown in Fig. 18.7. This value may be taken as a measure of the duration of the VDK stand time. It may be seen that throughout the 12 serial fermentations, one yeast line was more efficient at VDK removal than the other. In addition, there were apparent differences between individual batches of wort, as indicated by the patterns observed for each paired serial fermentation. The observations that both wort composition and yeast line could influence VDK metabolism prompted a closer examination of production data. This revealed that the decline in VDK concentration was biphasic. Thus, there was an initial rapid fall followed by a second phase of slow decline (Fig. 18.8a). The terminal slow portion of a number of such profiles is shown in Fig. 18.8b. It is evident from these data that small variations in these end-profiles have the potential to produce large differences in the times taken to achieve diacetyl specification. It might be supposed that the reduction in the rate of decrease in VDK concentration at the end of fermentation is due to yeast sedimentation and the concomitant

190

BREWING YEAST FERMENTATION PERFORMANCE

1.4

Total VDK (ppm)

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

50

100

(a)

150

200

250

Time (h) 0.4 0.35

[VDK] (ppm)

0.3 0.25 0.2 0.15 0.1 0.05 0 100

(b)

150

200

250

Time (h)

Fig. 18.8 (a) VDK profiles for six 1600 hl cylindroconical lager fermentations. (b) The terminal phase is highlighted. The diacetyl cooling specification for this beer is 0.1 ppm, as indicated.

lack of suspended yeast to catalyse the reduction of diacetyl. This premise was tested by monitoring total VDK concentration in a stirred laboratory fermentation. In fact, the profile was similar to that seen in production-scale fermentations, with a slow terminal uptake phase being evident (Fig. 18.9). The decline in VDK uptake seen at the end of fermentation was not due to yeast sedimentation. It was possible, therefore, that changes in the physiological condition of yeast during fermentation might be implicated. The decreasing rates of VDK disappearance could be due to either a change in the rate of the spontaneous decarboxylation step and/or a decline in the ability of yeast to metabolise free diacetyl. The latter possibility was probed by spiking stirred laboratory fermentations with free diacetyl. Individual replicate brewery wort fermentations were spiked with 1 ppm diacetyl at those times that equated to the periods of rapid increase, rapid decline, slow decline and no decline in the endogenous VDK profile. Total VDK concentration was monitored throughout each fermentation. In all cases, the free diacetyl rapidly disappeared from the fermenting wort (Fig. 18.10). The underlying VDK profiles for each replicate

[VDK] (ppm)

1.0 0.8 0.6 0.4 0.2 0 0

Fig. 18.9

20

40

60

80

100 120 140 160 180 200 Time (h)

VDK profile for a 1.5 litre stirred laboratory wort fermentation.

[VDK] (ppm)

3 2.5 2 1.5 1 0.5 0

0

50

100

150

200

250

300

0

50

100

150

200

250

300

0

50

100

150

200

250

300

0

50

100

150 200 Time (h)

250

300

[VDK] (ppm)

3 2.5 2 1.5 1 0.5 0

[VDK] (ppm)

3 2.5 2 1.5 1 0.5

[VDK] (ppm)

0 3 2.5 2 1.5 1 0.5 0

Fig. 18.10 VDK profiles for four 1.5 litre stirred laboratory wort fermentations. Diacetyl (1 ppm) was added to individual fermentations at the times indicated by the appearance of the transient peak.

BREWING YEAST FERMENTATION PERFORMANCE

Diacetyl uptake rate (ppm/h)

192

16 14 12 10 8 6 4 2 0 0

50

100

150

200

Cell count (
106/ml)

Initial uptake rate (ppm/h)

Fig. 18.11 Effect of varying yeast concentration on the initial rate of diacetyl uptake. Yeast was obtained from brewery storage vessels and used at the concentrations indicated. 0.3 0.25 0.2 0.15 0.1 0.05 0 0

0.2

0.4 0.6 0.8 [Initial Diacety]l (ppm)

1

Fig. 18.12 Effect of diacetyl concentration on initial rates of diacetyl uptake by brewery pitching yeast (1  108 viable cells/ml).

fermentation were similar and apparently unaffected by the added diacetyl. This supports the contention that yeast is capable of rapid assimilation of free diacetyl and, furthermore, that most of the pool of total VDK is actually the precursor, -acetolactate. However, it may also be seen that the rate of decrease in diacetyl concentration progressively declined the later in fermentation that the addition was made. This supported the view that the ability of yeast to assimilate diacetyl may be affected by physiological condition. This was explored using an in vitro assay which tested the ability of pitching yeast to remove exogenous diacetyl from a buffered medium. Assimilation of exogenous diacetyl was dependent on the presence of yeast and the initial rate was proportional to the yeast concentration (Fig. 18.11). Initial rates of assimilation were dependent on the concentration of exogenous diacetyl within the range tested (Fig. 18.12). This range (0–1 ppm) was of the order that might be expected to occur during fermentation of brewing wort. Observed assimilation rates were most rapid when the yeast was grown under aerobic derepressed conditions, yeast with an anaerobic physiology exhibited the slowest rates of assimilation, whereas aerobic repressed yeast was intermediate between the two (Fig. 18.13). These observations suggested that changes in yeast physiology associated with aerobic/anaerobic transitions would influence its ability to assimilate diacetyl. The expression and repression of specific transporters and/or reductases might be implicated. Previous work3 has shown that when pitching yeast is

FORMATION AND DISAPPEARANCE OF DIACETYL

193

Diacetyl uptake rate (ppm/h)

12 10 8 6 4 2

Aerobic derepressed

Aerobic repressed

Anaerobic

0

Diacetyl uptake rate (ppm/h)

Fig. 18.13 Comparison of diacetyl uptake rates using aerobic repressed (10% w/v glucose), aerobic derepressed (1% glycerol) and anaerobic repressed (10% glucose). In each case the yeast concentration was 1  108 viable cells/ml.

10 8 6 4 2 0 0

1

2 3 Oxygenation time (h)

4

5

Fig. 18.14 Effect of oxygenation of yeast on the ability to assimilate diacetyl. yeast was oxygenated as described by Boulton et al.3 and samples were removed at the times indicated and used in the diacetyl uptake assay.

exposed to oxygen under non-growing conditions there is a concomitant synthesis of sterol. This process allows the development of a competent membrane such that assimilation of nutrients and yeast growth occurs in anaerobic wort. Yeast was oxygenated, as described previously,3 and tested in the in vitro diacetyl assimilation assay. Initial assimilation rates increased in proportion to the duration of oxygen exposure (Fig. 18.14).

18.4

Conclusions

For the particular combination of wort and yeast strain examined in this study there was no correlation between the duration of primary fermentation and the VDK stand time.

194

BREWING YEAST FERMENTATION PERFORMANCE

With the exception of temperature, fermentation control parameters such as pitching rate and wort dissolved oxygen exert their effects on primary fermentation. The duration of the VDK stand appears to be influenced by temperature and wort composition. In addition, some yeast lines (derived from a particular propagation) appear to have an inherently greater ability than others to produce fermentations with shorter VDK stand times. A rapid and simple in vitro test is described which would allow this ability to be screened. Much of the variability in the length of VDK stand times is associated with inconsistencies during a slow terminal phase of VDK assimilation. Using laboratory stirred fermentations it was shown that the deceleration in VDK assimilation was not due to yeast sedimentation. For this yeast and wort combination, the characteristic pattern of VDK formation and disappearance appears to be of physiological significance. Further weight was added to this assertion by the observation that the rate of disappearance of free diacetyl added to a laboratory wort fermentation also declined throughout fermentation. Using the in vitro assay, it was shown that this particular pitching yeast had a poor intrinsic ability to assimilate free diacetyl, especially at the low concentrations that would be expected to occur in a wort fermentation. Manipulation of yeast physiology to promote membrane competence, either by growing under aerobic conditions or by exposure to oxygen under conditions of non-growth, stimulated the assimilation of exogenous diacetyl. These results suggest that the low rates of VDK removal associated with the latter stages of fermentation may be a reflection of the reduced ability of the yeast to assimilate free diacetyl. The correlation with membrane competence suggests that a transport phenomenon could be implicated, although induction and repression of diacetyl reductases cannot be ruled out. Thus, in late fermentation it may be that diacetyl assimilation by yeast is the rate-determining step in the VDK cycle. Yeast sedimentation, especially of flocculent strains, in tall cyclindroconical vessels would probably exacerbate the effect. Manipulation of fermentation conditions to ensure that yeast growth was not limited by sterol depletion might be beneficial in delaying the onset of the slow phase of VDK decline, but this could have counterproductive implications in terms of fermentation efficiency. Alternatively, the evidence presented here suggests that the fermenting vessel is a most unsuitable environment for the removal of VDK and the exploration of alternative strategies is a worthwhile goal. Acknowledgements The authors thank the Directors of Bass Brewers Ltd for permission to publish this paper. References 1. Wainwright, T. (1973) Diacetyl – a review. J. Inst. Brew. 79, 451–470. 2. Boulton, C.A. and Quain, D.E. (1987) Yeast, oxygen and the control of brewery fermentation. Proc. 21st Cong. Eur. Brew. Conv., Madrid, pp. 401–408.

FORMATION AND DISAPPEARANCE OF DIACETYL

195

3. Boulton, C.A., Jones, A.R. and Hinchliffe, E. (1991) Yeast physiological condition and fermentation performance. Proc. 23rd Cong. Eur. Brew. Conv., Lisbon, pp. 385–392. 4. European Brewery Convention (1987) Diacetyl and other vicinal diketones in beer. In: Analytica, 4th ed. Brauerei und Getranke-Rundschau, Zürich, 9.11, E187–E190. 5. Boulton, C.A., Box, W.G., Quain, D.E. and Molzahn, S.W. (1999) Vicinal diketone reduction as a measure of yeast vitality. Proc. 27th Cong. Eur. Brew. Conv., Cannes, pp. 687–694.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

19

The Formation of Higher Alcohols J.R. DICKINSON

Abstract In addition to ethanol, yeast produces long-chain and complex alcohols. These are important flavour and aroma compounds in all yeast-fermented products and have interesting organoleptic properties in their own right, many of which are both concentration and context dependent. Esters derived from these alcohols are also important compounds giving ‘fruity’ flavours and aromas. The genes and enzymes involved in the formation of long-chain and complex alcohols in yeast have been sought. In all cases, the general sequence of biochemical reactions is similar, but for the formation of the individual alcohols there is a complex profile of both specificity and overlapping gene functions in the five-membered family of decarboxylases encoded by PDC1, PDC5, PDC6, YDL080C and YDR380W. In the leucine degradation pathway the major decarboxylase is encoded by YDL080c and the YDR380w gene product accounts for less than 6% of the degradation. Consequently, YDL080c is referred to as KID1, for keto isocaproate decarboxylase 1. In valine degradation any one of the three isoenzymes of pyruvate decarboxylase encoded by PDC1, PDC5 and PDC6 will decarboxylate -ketoisovalerate. In isoleucine catabolism any one of the five possible decarboxylases encoded by PDC1, PDC5, PDC6, YDL080c or YDR380w is sufficient for the catabolism of isoleucine to ‘active’ amyl alcohol. For phenylalanine and tryptophan breakdown a pdc1, pdc5, pdc6, ydr380w quadruple mutant makes zero 2-phenylethanol and tryptophol, respectively. In other words, YDL080c (KID1) is irrelevant: it is not needed in these two pathways. The roles were also examined of the alcohol dehydrogenases encoded by ADH1, ADH2, ADH3, ADH4, ADH5, SFA1, AAD3, AAD4, AAD6, AAD10, AAD14, AAD15 and AAD16 in the final stage of higher alcohol formation in both laboratory and brewing strains. Any one of the first six of these alcohol dehydrogenases will do the job as long as it is operating under the physiological conditions in use.

19.1 Introduction It is common knowledge that yeast converts sugars to ethanol using the glycolytic pathway. In addition, it produces a number of long-chain and complex alcohols such as isoamyl alcohol, ‘active’ amyl alcohol, isobutanol, 2-phenyl ethanol and tryptothol. These are important flavour and aroma compounds in all yeast-fermented products and have interesting organoleptic properties in their own right, many of which are both concentration and context dependent. For example, concentrated isoamyl alcohol has a most unpleasant smell, but at concentrations below about 0.1% it has a pleasant ‘refreshing’ or ‘clean’ note which may account for its use at low concentrations in many deodorants. 2-Phenyl ethanol, when concentrated, smells of rose petals, but when more dilute is described as smelling like ‘Band-aid’. The properties of these compounds are also context dependent, that is, the concentrations required are different in different products. For example, full-bodied ales require higher concentrations of these alcohols than do lagers. Indeed, if the same concentrations were present in a lager as in an ale, then the lager would definitely taste ‘off’. A batch of lager would

THE FORMATION OF HIGHER ALCOHOLS

197

be spoiled by having unsuitably high levels of branched-chain alcohols such as isoamyl alcohol. In other words, controlling the concentrations of these compounds is of prime importance. Esters derived from these alcohols are also important compounds, giving ‘fruity’ flavours and aromas. Just one example would be isoamyl acetate, which imparts a strong flavour of ‘banana’ and is especially important in Glenmorangie whisky and Beaujolais wine. Attempts have been made to discover the genes and enzymes involved in the formation of long-chain and complex alcohols in Saccharomyces cerevisiae. In all cases, the general sequence of biochemical reactions is similar, but for the formation of the individual alcohols the details are surprisingly different. The predominant idea for many years has been that yeasts use the Ehrlich pathway (Fig. 19.1). This scheme envisages first an aminotransferase reaction yielding an -ketoacid and then decarboxylation of the ketoacid to an aldehyde that is then reduced in an NADH-linked reaction producing the appropriate ‘fusel’ alcohol. This scheme has been called the Ehrlich pathway to honour the originator of the ideas, which were first proposed in 19071 and modified later.2 Acceptance of the Ehrlich pathway is problematic for at least four reasons. First, before the present work, the supposed pathway had never been proven to exist. Simply showing that, for example, radioactively labelled leucine is converted into isoamyl alcohol does not prove that the individual steps are those envisaged in the scheme. Other routes are conceivable. The decarboxylase had never been isolated: a few authors had assumed (without proof) that pyruvate decarboxylase was responsible.3 The second problem is that this scheme cannot explain all of the products that yeast makes: some fusel alcohols do not correspond to any known amino acid. Thirdly, even if the pathway did exist as assumed for so long, the production of fusel alcohols in this manner would require a mixing of the synthetic and degradative reactions involving the branched-chain amino acids. This is something that cells normally avoid by co-ordinated regulation. Synthetic enzymes are usually repressed and/or inhibited if the endproduct is present. Conversely, catabolic enzymes are not induced if the compound is being synthesised at rates that are satisfying biosynthetic demand. Furthermore, if this unusual situation did exist, then operation of the metabolic pathways would require some extraordinary channelling of metabolites between different subcellular compartments. Fourthly, the Ehrlich pathway does not explain all of the known facts, as Ehrlich himself was fully aware. Two of the most serious concerns are that the kinetics of amino acid utilisation do not match fusel alcohol formation in complex media, and in media containing low levels of amino acids there is virtually no correlation between the amino acid composition and the composition of the resulting fusel oil.4

Fig. 19.1

The Ehrlich pathway.

198

BREWING YEAST FERMENTATION PERFORMANCE

The first step of branched-chain amino acid catabolism is now known to be transamination. There are at least two distinct aminotransferases: one is mitochondrial (TWT1 gene product) and one is cytosolic (TWT2 gene product). The mitochondrial isozyme is highly expressed during the logarithmic phase and is repressed during the stationary phase, while the cytosolic isozyme has the opposite pattern of expression. In 1993 the present author described a mutation called baa1 (denoting a defect in branched chain amino acid aminotransferase).5 It is now clear from the profiles of enzyme expression and reduced growth that the baa1 mutant is the same as a twt2 mutant. TWT1 expression is controlled by the global regulator of metabolism Gcn4. twt1 twt2 double mutants still possess high levels of branched-chain amino acid aminotransferase activity in the cytosol, indicating that other enzyme(s) exist with this activity. 13 C-Labelled leucine, valine or isoleucine and [13C] nuclear magnetic resonance (NMR) spectroscopy were used to determine the metabolic pathways used in fusel alcohol formation from the three branched-chain amino acids.6–8 Figure 19.2 shows the [13C]NMR spectrum of a culture supernatant of a wild-type strain cultured in a minimal medium in which ethanol was the carbon source and [2-13C]leucine the sole nitrogen source. From the metabolites identified and the positions labelled with 13C, several routes between leucine and isoamyl alcohol are possible (Fig. 19.3). All of these involve initial transamination to -ketoisocaproate. The first is via branched-chain -ketoacid dehydrogenase (route A in Fig. 19.3) to yield isovaleryl coenzyme A (CoA), which would then be converted to isovalerate by acyl CoA hydrolase. The second possibility is via pyruvate decarboxylase (route B). Pathway C envisages -ketoisocaproate reductase to produce -hydroxyisocaproate, followed by decarboxylation to yield isoamyl alcohol. Route D proposes a pyruvate decarboxylase-like enzyme. Pathway E

Fig. 19.2 [13C]Nuclear magnetic resonance spectrum of a culture supernatant of a wild-type strain cultured in a minimal medium in which [2-13C]leucine was the sole nitrogen source. The resonances identified are: L1–L6: C-1 to C-6 of leucine; K: C-2 of -ketoisocaproate; IV: C-1 of isovalerate; H: C-2 of -hydroxyisocaproate; IA: C-1 of isoamyl alcohol; E1 and E2: C-1 and C-2 of ethanol. (Reproduced from Dickinson et al.6 with permission.)

THE FORMATION OF HIGHER ALCOHOLS

199

Fig. 19.3 Potential metabolic routes for the metabolism of leucine to isoamyl alcohol. The asterisks indicate carbon atoms in intermediates that were labelled with 13C in the wild-type strain. Enzyme activities are abbreviated as follows: BCAAT: branched-chain -ketoacid dehydrogenase; ACH: acyl coenzyme A hydrolase; PDC: pyruvate decarboxylase. (Reproduced from Dickinson et al.6 with permission.)

between -ketoisocaproate and isovalerate was included as a theoretical possibility in Fig. 19.3, but cannot be explained using known or potential enzymes and was thus discarded. Using mutants and combined gas chromatography–mass spectrometry (GC-MS) the actual in vivo pathway may be deduced. Thus, route A is not used for the synthesis of isoamyl alcohol because abolition of branched-chain -ketoacid dehydrogenase in an lpd1 disruption mutant does not prevent the formation of isoamyl alcohol. Route B, via pyruvate decarboxylase, is not required either because the complete elimination of pyruvate decarboxylase activity in a pdc1, pdc5, pdc6 triple mutant has no effect on the levels of isoamyl alcohol produced. The third possibility (route C) is via -ketoisocaproate reductase, a novel activity not previously known in S. cerevisiae. The true metabolic significance of this enzyme in yeast is not clear at this time, but it can have no role in the formation of isoamyl alcohol from -hydroxyisocaproate because cell homogenates cannot convert -hydroxyisocaproate to isoamyl alcohol. Route D, a pyruvate decarboxylase-like enzyme encoded by YDL080c, appears to be the major route of decarboxylation of -ketoisocaproate to isoamyl alcohol because strains with disruptions in this gene produce very little isoamyl alcohol. However, ydl080c mutants do not produce zero isoamyl alcohol, so another decarboxylase can

200

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 19.4 [13C]Nuclear magnetic resonance spectrum of a culture supernatant of a wild-type strain cultured in a minimal medium in which [2-13C]valine was the sole nitrogen source. The resonances marked are: V1–V5: C-1 to C-5 of valine; K: C-2 of -ketoisovalerate; IBA: C-1 of isobutyrate; H: C-2 of -hydroxyisovalerate; IBOH: C-1 of isobutanol; E-1 and E-2: C-1 and C-2 of ethanol; IA: C-2 of iosamyl alcohol. (Reproduced from Dickinson et al.7 with permission.)

substitute to a minor extent. More recently it was discovered that this minor decarboxylase is encoded by YDR380w. Hence, if all five decarboxylases are deleted, the yeast makes no isoamyl alcohol. Similar methods ([13C]NMR spectroscopy using [2-13C]valine as substrate combined with GC-MS and specific mutants) were used to examine the catabolism of valine to isobutanol.7 The NMR spectrum (Fig. 19.4) enabled identification of various resonances and led to the realisation that the product of valine transamination, -ketoisovalerate, had four potential routes to isobutanol (Fig. 19.5). The first, via branched-chain -ketoacid dehydrogenase to isobutyryl CoA, is not required for the synthesis of isobutanol because abolition of branched-chain -ketoacid dehydrogenase activity in an lpd1 disruption mutant did not prevent the formation of isobutanol. The second route, via pyruvate decarboxylase, is the one that is used, because elimination of pyruvate decarboxylase activity in a pdc1, pdc5, pdc6 triple mutant virtually abolished isobutanol production. A third potential route involved -ketoisovalerate reductase, but this had no role in the formation of isobutanol from -hydroxyisovalerate because cell homogenates could not convert -hydroxyisovalerate to isobutanol. The final possibility, use of the pyruvate decarboxylase-like enzyme encoded by YDL080c, seemed to be irrelevant, because a strain with a disruption in this gene produced wild-type levels of isobutanol.7 [13C]NMR spectroscopy revealed a further surprise: some 13C from [2-13C]valine appears in C-2 of isoamyl alcohol (see resonance marked IA in Fig. 19.4). The occurrence of isoamyl alcohol labelled at C-2 is readily explained by the pathway of leucine biosynthesis (Fig. 19.6). -Ketoisovaleric acid labelled at C-2 is converted to

THE FORMATION OF HIGHER ALCOHOLS

201

Fig. 19.5 Potential metabolic routes for the metabolism of valine to isobutyl alcohol. The asterisks indicate carbon atoms in intermediates that were labelled with 13C in the wild-type strain. Enzyme activities are abbreviated as in Fig. 19.3. (Reproduced from Dickinson et al.7 with permission.)

-isopropylmalate labelled at C-3 by -isopropylmalate synthase. This is then converted by -isopropylmalate dehydratase to -isopropylmalate labelled at C-3 which is, in turn, converted to -ketoisocaproic acid by -isopropylmalate dehydrogenase. The next step in the leucine biosynthetic pathway is the formation of leucine, which would be labelled at C-3. No signal due to C-3 of leucine was observed, but isoamyl alcohol labelled at C-2 was present. This explanation is confirmed by the fact that a strain which lacks the YDL080c-encoded -ketoisocaproate decarboxylase makes plenty of isobutanol but no isoamyl alcohol when valine is the sole nitrogen source. Hence, the NMR study shows that the pathways of valine catabolism and leucine biosynthesis share a common pool of -ketoisovalerate. This proves that there is a mixing of the valine catabolic and leucine biosynthetic pathways under these conditions, which is an extraordinary feat of metabolic control. The catabolism of isoleucine to active amyl alcohol was also studied. In brief, any one of the decarboxylases encoded by PDC1, PDC5, PDC6, YDL080c or YDR380w must be present to allow yeast to utilise -keto--methyl-valerate. Apparently, any one of this family of decarboxylases is sufficient to allow the catabolism of isoleucine to active amyl alcohol. This was the first demonstration of a role for the gene product of YDR380w. These five decarboxylases are a closely related family (Fig. 19.7). Yeast does have other decarboxylases, but no others are as closely related as this group of five. Very recently, this work on the roles of the five decarboxylases was extended to include

202

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 19.6 The pathway of leucine biosynthesis. The asterisks indicate carbon atoms in intermediates that were labelled with 13C. Enzymes are denoted by the structural genes that encode them. LEU4: -isopropylmalate synthase; LEU1: -isopropylmalate dehydratase; LEU2: -isopropylmalate dehydrogenase; TWT1, TWT2: mitochondrial and cytoplasmic isoenzymes, respectively, of branchedchain amino acid aminotransferases; YDL080C  KID1: -ketoisocaproate decarboxylase. (Reproduced from Dickinson et al.7 with permission.)

catabolism of the aromatic amino acids phenylalanine and tryptophan to 2-phenyl ethanol and tryptothol, respectively. Summarising all the knowledge on the decarboxylase step, there are significant differences in the enzymes used for decarboxylation of the -ketoacids derived from the different amino acid substrates. In the leucine degradation pathway the major decarboxylase is encoded by YDL080c and the

THE FORMATION OF HIGHER ALCOHOLS

203

Fig. 19.7 Alignment of the predicted amino acid sequences of the five decarboxylases involved in long-chain alcohol formation. The figure was produced using the PRETTYBOX program, which displays identical residues in a black box and conservative substitutions in a grey box. (Reproduced from Dickinson et al.8 with permission.)

YDR380w gene product accounts for less than 6% of the degradation.6 Consequently, YDL080c is referred to as KID1, for keto isocaproate decarboxylase 1. In valine degradation any one of the three isoenzymes of pyruvate decarboxylase encoded by PDC1, PDC5 and PDC6 will decarboxylate -ketoisovalerate.7 In isoleucine

204

BREWING YEAST FERMENTATION PERFORMANCE

catabolism any one of these five possible decarboxylases encoded by PDC1, PDC5, PDC6, YDL080c or YDR380w is sufficient for the catabolism of isoleucine to ‘active’ amyl alcohol.8 For phenylalanine and tryptophan breakdown a pdc1, pdc5, pdc6, ydr380w quadruple mutant makes zero 2-phenylethanol and tryptophol, respectively. In other words, YDL080c (KID1) is irrelevant: it is not needed in these two pathways. Finally, what is presumed to be an alcohol dehydrogenase step in the pathways of long-chain alcohol formation was considered. There are a great many potential candidates for this role. An exciting break through came in 1999, when Delneri and Oliver identified a seven-membered family of putative aryl alcohol dehydrogenase genes, a gene family for which there were no known functions among the members.9,10 Their functions look set to remain unknown for a while longer because a mutant strain in which all seven AAD genes had been knocked out was still capable of converting leucine to isoamyl alcohol. Saccharomyces cerevisiae has 13 other alcohol dehydrogenases. Some of these can be eliminated; for example, BDH1 (YAL060W) has recently been shown to be involved in the formation of 2,3-butanediol.11 On the basis of similarity it seems safe to assume that YAL061W could be called ‘BDH2’ and that it has a similar activity. SOR1 (YJR159W), ‘SOR2’ (YDL246C) and XDH1 (YLR070C) are sugar alcohol dehydrogenases, while ‘CDH1’ (YCR105W) and ‘CDH2’ (YMR318C) are NADP-dependent cinnamyl alcohol dehydrogenases. This leaves six other alcohol dehydrogenases, of which perhaps the first four (Adh1–Adh4) are the best known. Adh1 is the main cytosolic enzyme involved in the formation of ethanol during glycolysis. Adh2, which is also cytosolic, is the glucoserepressed enzyme that is needed for growth on ethanol. Adh3 is mitochondrial, it is induced on glucose and its role is not really understood. Adh4 is present at only very low levels in most laboratory strains, but is plentiful in brewing strains.12 Adh5 was discovered by genome sequencing and is a complete mystery. Sfa1 is part of a bifunctional enzyme that has glutathione-dependent formaldehyde dehydrogenase activity (the gene name comes from sensitive to formaldehyde, which is the phenotype of mutants affected in this gene); it is also described as being capable of catalysing the destruction of long-chain alcohols.13 Each has at least one feature that may lead one to expect that it could be responsible for the synthesis of long-chain alcohols. For example, Adh1, Adh2 and Adh3 all require zinc, and brewers have known for years that the concentration of zinc in the wort affects the concentration of long-chain alcohols that are formed. Alternatively since it has been shown that Sfa1 can degrade long-chain alcohols in vitro, its true role in vivo may be the formation of these compounds. Unfortunately, this is all surmise because mutants knocked out in any one of these genes can all produce the long-chain alcohols. Furthermore, having produced combinations of quintuple knockout mutants (e.g. adh1, adh2, adh3, adh4, adh5) it looks as though any one of these alcohol dehydrogenases will fulfil the role as long as it is operating under the physiological conditions in use; for example, Adh2 would not work in high glucose conditions.

19.2

Conclusions

One important conclusion that can be drawn is that the specificity seems to be achieved at the decarboxylation step because the degradation of each amino acid uses

THE FORMATION OF HIGHER ALCOHOLS

205

a different decarboxylase or group of decarboxylases. The current research is aimed at trying to define how this is achieved in each case. In minimal media this is essentially the Ehrlich pathway as proposed nearly 100 years ago, but it is much more complex than Ehrlich ever imagined because yeast uses at least two aminotransferases, five decarboxylases and six alcohol dehydrogenases. The precise combination of enzymes used at a particular time depends on the amino acid, the carbon source and the stage of growth of the culture. In complex media yeast uses branched-chain -ketoacid dehydrogenase. This explains the last of Ehrlich’s conundrums. Ehrlich could not explain why in complex media there was no correlation between the kinetics of amino acid utilisation and fusel alcohol formation. In complex media yeast uses -ketoacid dehydrogenase and fusel alcohols are not formed. The high Michaelis constant of -ketoacid dehydrogenase14 also ensures that metabolic flux will only occur via this enzyme when the -ketoacids are present at extremely high levels.

References 1. Ehrlich, F. (1907) Über die bedingungen der fuselölbildung und uber ihren zusammenhang mit dem eiweissaufbau der hefe. Berichte Dtsch. Chem. Gesellsch. 40, 1027–1047. 2. Neubauer, O. and Fromherz, K. (1911) Über den abbau der aminosären bei der hefegärung. HoppeSeylers Z. Physiol. Chem. 70, 326–350. 3. Derrick, S. and Large, P.J. (1993) Activities of the enzymes of the Ehrlich pathway and formation of branched-chain alcohols in Saccharomyces cerevisiae and Candida utilis grown in continuous culture on valine or ammonium as sole nitrogen source. J. Gen. Microbiol. 139, 2783–2792. 4. Webb, A.D. and Ingraham, J.L. (1965) Fusel oil. Adv. Appl. Microbiol. 5, 317–353. 5. Dickinson, J.R. and Norte, V. (1993) A study of branched-chain amino acid aminotransferase and isolation of mutations affecting the catabolism of branched-chain amino acids in Saccharomyces cerevisiae. FEBS Lett. 326, 29–32. 6. Dickinson, J.R., Lanterman, M.M., Danner, D.J. et al. (1997) A 13C nuclear magnetic resonance investigation of the metabolism of leucine to isoamyl alcohol in Saccharomyces cerevisiae. J. Biol. Chem. 272, 26871–26878. 7. Dickinson, J.R., Harrison, S.J. and Hewlins, M.J.E. (1998) An investigation of the metabolism of valine to isobutyl alcohol in Saccharomyces cerevisiae. J. Biol. Chem. 273, 25751–25756. 8. Dickinson, J.R., Harrison, S.J., Dickinson, J.A. and Hewlins, M.J.E. (2000) An investigation of the metabolism of isoleucine to active amyl alcohol in Saccharomyces cerevisiae. J. Biol. Chem. 275, 10937–10942. 9. Delneri, D., Gardner, D.C.J. and Oliver, S.G. (1999) Analysis of the seven-member AAD gene set demonstrates that genetic redundancy in yeast may be more apparent than real. Genetics 153, 1591–1600. 10. Delneri, D., Gardner, D.C.J., Bruschi, C.V. and Oliver, S.G. (1999) Disruption of seven hypothetical aryl alcohol dehydrogenase genes from Saccharomyces cerevisiae and construction of a multiple knockout strain. Yeast 15, 1681–1689. 11. González, E., Fernández, M.R., Larroy, C. et al. (2000) Characterization of a (2R,3R)-2,3-butanediol dehydrogenase as the Saccharomyces cerevisiae YAL060W gene product. J. Biol. Chem. 275, 35876–35885. 12. Dickinson, J.R. (1999) Carbon metabolism. In: The Metabolism and Molecular Physiology of Saccharomyces cerevisiae, Dickinson, J.R. and Schweizer, M. (eds). Taylor and Francis, London, pp. 23–55. 13. Wehner, E.P., Rao, E. and Brendel, M. (1993) Molecular structure and genetic regulation of SFA, a gene responsible for resistance to formaldehyde in Saccharomyces cerevisiae, and characterization of its protein product. Mol. Gen. Genet. 237, 351–358. 14. Sinclair, D.A., Dawes, I.W. and Dickinson, J.R. (1993) Purification and characterization of the branched chain -ketoacid dehydrogenase complex from Saccharomyces cerevisiae. Biochem. Mol. Biol. Int. 31, 911–922.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

20

Methionine: A Key Amino Acid for Flavour Biosynthesis in Beer P. PERPÈTE, L. GIJS and S. COLLIN

Abstract The level of sulfur flavours such as methional, methionol, dimethyldisulfide, dimethyltrisulfide or methanethiol in fresh beer is mainly dependent on the yeast reduction activity during fermentation. Among these, methional was recently reported to be responsible for the worty aroma of alcohol-free beers and should be considered as a precursor of dimethyltrisulfide which is implied in the staling of beer. Methionine probably also plays a role in methional excretion. However, while the Strecker degradation is a wellknown chemical mechanism transforming amino acids to aldehydes, no genetic evidence of a methionine Ehrlich-like pathway has been reported previously. The efficiency of this pathway will be compared with the C-S lysase activities of Saccharomyces cerevisiae. The aim of this chapter is to demonstrate that this biochemical degradation could lead to the development of sulfurous off-flavours in beer.

20.1 Introduction Sulfur compounds from malt and hops or synthesised during the brewing process are natural components of beer. Individually, sulfur compounds usually impart an aroma or onion, rotted vegetables or cabbage (Table 20.1). Dimethylsulfide (50 ppb in some lager beers) and sulfur dioxide (up to 10 ppm) are major sulfur compounds in beer. The synthesis pathway and parameters influencing their final level have been well documented.4 Often characterised by a very low detection Table 20.1

Organoleptic qualities and detection thresholds of some sulfur compounds

Compound

Structure

Odour

Isopentenylmercaptan 2-Mercapto-3methylbutanol Methional

(CH3)2C ¨CHˆCH2SH CH2OHˆCHSHˆCH(CH3)2

‘Sunstruck’ Onion

CH3ˆSˆCH2ˆCH2ˆCHO

Methionol Methanethiol

CH3ˆSˆCH2ˆCH2ˆCH2OH CH3ˆSH

Boiled potatoes, soup Radish Excrement, putrefaction

DMS DMDS

CH3ˆSˆCH3 CH3ˆSˆSˆCH3

DMTS

CH3ˆSˆSˆSˆCH3

Cooked cabbages, onion, rubber Fresh onion, sulfurous, boiled vegetables

Adapted from Refs 1–3. DMS: dimethylsulfide; DMDS: dimethyldisulfide; DMTS: dimethyltrisulfide.

Detection Concentration threshold (ppb) in beer (ppb) 0.002–0.004 1 ppb 0.1–250

100 – 2–50

1200 0.02–41

2–50 2–12

3–50

0.3–7.5

0.1

0.2–1.8

A KEY AMINO ACID FOR FLAVOUR BIOSYNTHESIS IN BEER

207

threshold, minor sulfur compounds such as thioesters, polysulfides or thiols most probably also modify the overall organoleptic quality of beer. Thioesters (S-methylthioacetate, S-methylthioisovalerate), which originate mainly from hops,5 give rise to rotted vegetable aromas6 in lager beers. With a flavour threshold of 0.1 ppb, dimethyltrisulfide is known to be responsible for the onion off-flavour of aged beers. Recent data show that S-methylcysteine sulfoxide, methional and methionol are its main precursor in beer.7,8 Although thiols are hard to analyse because of their very high reactivity and very low concentration/detection threshold, they are also very relevant compounds for brewers. Among them, isopentenylmercaptan imparts the well-known ‘sunstruck’ flavour9 and methanethiol (detection threshold: 0.1 ppb) is usually characterised with ‘excrement’ or ‘putrefaction’ descriptors. In aged beers, as in wines,10 polyfunctional thiols are probably also responsible for delicate flavours.8 Methional or 3-methylthiopropionaldehyde imparts a boiled potato aroma at high concentrations,2 but can be described as ‘soup’ or ‘hot wort’ at lower concentrations.11 This aldehyde was initially detected in Cheddar cheese,12 corn tortillas13 and boiled trout,14 and has been measured in beer and alcohol-free beers, where it contributes to the worty aroma along with 2-methylbutanal and 3-methylbutanal.11 More recently, methional was also proposed as an additional key compound in aged beers.15 As 2-methylbutanal and 3-methylbutanal can be synthesised by Strecker degradation of amino acids, wort methional may derive from methionine. During wort fermentation, Saccharomyces cerevisiae produces NADPH-dependent enzymes allowing methional reduction.11,16 However, because of the low temperature applied and interaction with polyphenols, this enzymic reduction is incomplete in alcohol-free beer productions, leading to a strong worty flavour in the final product. Methionol is the reduction product of methional. Although not really pleasant, with its cauliflower or radish-like aroma, methionol is not considered an off-flavour in beer (detection threshold close to 1200 ppb). Most of the sulfur compounds described above should be derivable from methionine, so it is very surprising that so little is known about the methionine-degrading pathway in yeast. The aim of this work was to find evidence for a catabolic pathway leading to methanethiol, either with or without a methional intermediate.

20.2 Materials and methods 20.2.1

Reagents

Methional (95%) and methionol (98%) were purchased from Acros Chemika (Brussels, Belgium). Methanethiol (99.5%) was from Aldrich (Brussels, Belgium). 20.2.2

Strains

Saccharomyces cerevisiae BRAS291 (bottom fermentation) and BRAS212 (top fermentation) were provided by the BRAS collection of the Unité de Brasserie et des Industries Alimentaires (Louvain-la-Neuve, Belgium).

208 20.2.3

BREWING YEAST FERMENTATION PERFORMANCE

Culture media and sampling

Precultures were grown in YPS medium (1% yeast extract, 0.5% peptone and 10% sucrose) at 28°C on a rotary shaker and collected in the exponential phase. After collection and washing, the yeast was pitched at 106 cells/ml in model media. Two media were used during these experiments: a glucose-methionine medium (citrate-buffered medium containing 3% glucose and 10 mM methionine) and a glucose-ammonium medium (citrate-buffered medium containing 3% glucose and 10 mM ammonium sulfate). Cultures were grown at 28°C on a rotary shaker. At a given time, samples were collected and centrifuged, and supernatants were immediately frozen in liquid nitrogen for methanethiol quantification. 20.2.4

Methanethiol quantification

Methanethiol was quantified by dynamic headspace gas chromatography. A HewlettPackard model 5890 gas chromatograph equipped with a Chrompack Purge and Trap Injector, a flame ionisation detector and a Shimadzu CR3A integrator was used. Samples were injected into the chromatographic column in the following three steps. 1. Precooling of the trap (CPSil 8 CB capillary column, 0.53 mm internal diameter; film thickness 5 m): the trap was cooled at 95°C for 2 min in a stream of liquid nitrogen; 2. Purging of the sample: the temperature of the purge vessel was set at 50°C. The sample was purged with helium gas (12 ml/min) for 15 min. The gas stream was passed through a condenser kept at 15°C by means of a cryostat (Colora WK 15) to remove water vapour and then through an oven at 200°C. The volatiles were finally concentrated in the cold trap maintained at 95°C (liquid nitrogen); 3. Desorption of the volatiles: cooling was stopped, and the surrounding metal capillary was immediately heated to 220°C for 5 min. The carrier gas swept the trapped compounds into the analytical column. Analysis of samples was carried out on a 50 m  0.32 mm, wall-coated, open tubular (WCOT) CP-Sil5 CB capillary column (Chrompack, Antwerp, Belgium) (film thickness 1.2 m). Oven temperature, initially kept at 36°C for 15 min, was programmed to rise from 36 to 120°C at 2°C/min then to 200°C at 10°C/min, remaining at the maximum temperature for 10 min thereafter. Helium carrier gas was used at a flow rate of 1.0 ml/min. Injection and detection temperatures were 200 and 220°C, respectively. All analyses were done in duplicate. The assessment of the reproducibility of this technique has been described previously (coefficients of variation under 10% for five analyses of the same standard mixture).17

20.3

Results and discussion

During mashing, methionine can be transformed to methional by Strecker degradation (Fig. 20.1) and further oxidation can lead to methanethiol.18,19 Because of its high volatility, this thiol is easily stripped out of the wort in the boiling kettle.

209

A KEY AMINO ACID FOR FLAVOUR BIOSYNTHESIS IN BEER

Mashing – Boiling – Clarification – Staling → chemical synthesis Methionine Glucose Diacetyl Riboflavin, hv

Methional Riboflavin

CH3SH

O2

+ H2S Cu+2

DMDS, DMTS, DMQS Fig. 20.1

Potential mechanisms of methional and methanethiol synthesis during mashing processes.

S

O

O

S

S

S

OH

OH NH2

Methionine

OH

O

O

-keto--methyl thiobutyrate

Methional

SH

OH

Methionol

(a) S

O OH NH2

Methionine

(b)

+

NH4

Methanethiol

+ O

Ammonium, ?ketobutyrate

Fig. 20.2 Hypothetical methionine degradation pathways in Saccharomyces cerevisiae: (a) Ehrlich-like; (b) C-S lyase.

During fermentation, S. cerevisiae can consume both methionine and methional. Methionine uptake by yeast is well documented in the literature. At least five amino acid permeases are involved: MUP1p and MUP3p (for methionine uptake) seem to be the most specific transporters,20 while MUP2p/AGP1p (initially defined as an asparagine glutamine permease) and BAP2 and BAP3 (branched amino acid permease) were initially believed to be dedicated to translocating other amino acids such as isoleucine or leucine.21 Once in the yeast cell, methionine may be further degraded via two enzymic routes (Fig. 20.2). The ‘Ehrlich-like’ pathway would lead to -keto- methylthiobutyrate, methional and methionol (Fig. 20.2a). The end of this biochemical pathway has been demonstrated in S. cerevisiae strains: methional and methionol are detected in beer11 and an NADPH-dependent enzyme can reduce the aldehyde to its primary alcohol.7,11

210

BREWING YEAST FERMENTATION PERFORMANCE

Concentration (ppb)

100 80 60 Ammonium 40

Methionine Without yeast

20 0

(a)

0h 12 h 24h Time (h of fermentation)

Concentration (ppb)

120 100 80 60

Ammonium

40

Methionine Without yeast

20 0

(b)

0h 12 h 24 h Time (h of fermentation)

Fig. 20.3 Methanethiol production by Saccharomyces cerevisiae BRAS291 (bottom yeast) and BRAS212 (top yeast) in glucose-ammonium and methionine-ammonium culture media. Fermentation was carried out at 28°C during 24 h under agitation, with a pitching rate of 106 cells/ml.

A second methionine-degrading pathway hypothetically present in S. cerevisiae is one observed in several fungi, including some yeasts. In Geotrichum candidum, methionine is directly degraded to methanethiol, ketobutyrate and ammonium (Fig. 20.2b). The enzymes responsible for this degradation are C-S-lyases-like enzymes or

-demethiolase.22,23 Some S. cerevisiae enzymes involved in or associated with the sulfydrylation pathway20 should be able to catalyse these reactions. To provide evidence for such a mechanism in S. cerevisiae, glucose-methionine and glucose-ammonium media were pitched with bottom-fermenting (S. cerevisiae BRAS291) or top-fermenting (S. cerevisiae BRAS212) strains. After 24 h of fermentation, a significant increase in methanethiol level was observed in the glucose-methionine medium (Fig. 20.3a,b), resulting in concentrations near 100 ppb for both yeast strains. No trace of this sulfur compound was detected in the same glucose-methionine medium without pitching, showing that methionine was not chemically degraded in this control. Slight methanethiol production (18 ppb after 24 h) was also measured in the glucose-ammonium medium after pitching with the top-fermenting yeast BRAS212. This low level could reasonably be explained by the low initial internal methionine pool. Another possibility to be considered is the chemical degradation of methional or methionol to methanethiol. However, when both model media were spiked with 50 ppm

A KEY AMINO ACID FOR FLAVOUR BIOSYNTHESIS IN BEER

211

Methionine Biochemical degradation

Biochemical degradation

Biochemical degradation

Methional

Methional

Chemical degradation

X X

CH3SH

Chemical degradation  H2S

DMDS, DMTS, DMQS Fig. 20.4

Potential mechanisms of methional and methanethiol biosynthesis during fermentation.

methional or methionol, not even a trace of methanethiol was detected. As depicted in Fig. 20.4, the results suggest that S. cerevisiae catalyses the degradation of methionine into methanethiol. These initial experiments, however, do not exclude the existence of an additional ‘Ehrlich-like’ pathway. To clarify this point, new experiments are being conducted with labelled methionine and intermediates.

References 1. Vermeulen, C., Pellaud, J., Gijs, L. and Collin, S. (2001) Combinatorial synthesis and sensorial properties of polyfunctional thiols. J. Agric. Food Chem. 49, 5445–5449. 2. Meilgaard, M. (1975) Flavour chemistry of beer: Part II: Flavour and threshold of 239 aroma volatiles. Tech. Q. Master Brew. Assoc. Am. 12, 151–168. 3. Olsen, A., Christiansen, B.W. and Madsen, J.O. (1988) Onion-like off-flavour in beer: isolation and identification of the culprits. Carlsberg Res. Commun. 53, 1–9. 4. Anness, B.J. and Bamforth, C.W. (1982) Dimethylsulphide – a review. J. Inst. Brew. 88, 244–252. 5. Suggett, A., Moir, M. and Seaton, J.C. (1979) The role of sulphur compound in hop flavour. Proc. Eur. Brew. Cong. 17, 79–89. 6. Stewart, G.G. and Russel, I. (1981) The influence of yeast on volatile sulphur compounds in beer. Eur. Brew. Conv. Monogr. VII, pp. 173–187. 7. Gijs, L., Perpète, P., Timmermans, A. and Collin, S. (2000) 3-Methylthiopropionaldehyde as precursor of dimethyltrisulphide in aged beers. J. Agric. Food Chem. 48, 6196–6199. 8. Williams, R.S. and Gracey, D.E.F. (1982) Beyond dimethylsulphide: the significance to flavour of thioesters and polysulphides in Canadian beers. J. Am. Soc. Brew. Chem. 40, 68–71. 9. Gunst, F. and Verzele, M. (1978) On the sunstruck flavour in beer. J. Inst. Brew. 84, 291–292. 10. Maga, J.A. (1976) The role of sulphur compounds in food flavour. Part III: Thiols. Crit. Rev. Food Sci. Nutr. January, 147–192. 11. Perpète, P. and Collin, S. (1999) Contribution of 3-methylthiopropionaldehyde to the worty flavour of alcohol-free beers. J. Agric. Food Chem. 47, 2374–2378. 12. Christensen, K.R. and Reineccius, G.A. (1995) Aroma extract dilution analysis of aged Cheddar cheese. J. Food Sci. 60, 218–220.

212

BREWING YEAST FERMENTATION PERFORMANCE

13. Buttery, R.G. and Ling, L.C. (1995) Volatile flavour components of corn tortillas and related products. J. Agric. Food Chem. 43, 1878–1882. 14. Milo, C. and Grosch, W. (1993) Changes in the odorants of boiled trout as affected by the storage of the raw material. J. Agric. Food Chem. 41, 2076–2081. 15. Chevance, F., Gijs, L., Jerkovic, V. et al. (2001) Influence of pH on beer flavour stability during storage. 4th Pangborn Sensory Sci. Symp., 22–26 July, Dijon, France. 16. Laurent, M., Geldorf, B., Van Nedervelde, L. et al. (1995) Characterisation of the aldoketoreductase yeast enzymatic systems involved in the removal of wort carbonyls during fermentation. Proc. Eur. Brew. Conv. 25, 337–344. 17. Collin, S., Osman, M., Delcambre, S. et al. (1993) Investigation of volatile flavour compounds in fresh and ripened Domiati cheeses. J. Agric. Food Chem. 41, 1659–1663. 18. Balance, P.E. (1961) Production of volatile compounds related to the flavour of foods from the Strecker degradation of DL-methionine. J. Sci. Food Agric. 12, 532–536. 19. Yu, T.H. and Ho, C.T. (1995) Volatile compounds generated from thermal reaction of methionine and methionine sulphoxide with or without glucose. J. Agric. Food Chem. 43, 1641–1646. 20. Thomas, D. and Surdin-Kerjan, Y. (1997) Metabolism of sulphur amino acids in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 61, 503–532. 21. Regenberg, B., During-Olsen, L., Kielland-Brandt, M. and Holmberg, S. (1999) Substrate specificity and gene expression of the amino acid permeases in Saccharomyces cerevisiae. Curr. Genet. 36, 317–328. 22. Helinck, S., Spinnler, H.E., Parayre, S. et al. (2000) Enzymatic versus spontaneous S-methyl thioester synthesis in Geotrichum candidum. FEMS Microbiol. Lett. 193, 237–241. 23. Spinnler, H.E., Berger, C., Lapadatescu, C. and Bonnarme, P. (2001) Production of sulfur compounds by several yeast of technological interest for cheese ripening. Int. Dairy J. 11, 245–252.

Brewing Yeast Fermentation Performance: Second edition Edited by: KATHERINE SMART Copyright 0Blackwell Science 2003

21 Control of Ester Synthesis During Brewery Fermentation J.-P. DUFOUR, Ph. MALCORPS and P. SILCOCK

Abstract The synthesis of volatile aliphatic esters by yeast has attracted considerable interest because of their potential contribution to aroma and flavour. In beer, the major esters are ethyl acetate, isobutyl acetate, isoamyl acetate, phenylethyl acetate and the C6–C10 medium-chain fatty acid ethyl esters. The need to understand and control ester synthesis is driven by problems encountered in brewing procedures, such as the production of disproportionate amounts of ethyl acetate and isoamyl acetate during high-gravity brewing, the reduction of ester levels when using large-scale cylindroconical fermenters and the lack of flavour compounds of reduced alcohol beers. Technological parameters that affect the production of esters can be divided into three categories: those related to yeast characteristics (strain, physiological state), those related to wort composition (aeration, lipids, zinc, free amino nitrogen, extract) and those related to fermentation conditions (temperature, pressure, fermenter design, fermentation method). Ester synthesis during fermentation depends mainly on yeast ester synthesis potential, i.e. the amount of available acetyl-coenzyme A/acyl-coenzyme A (an essential building block for yeast cellular components) and the level of ester synthase activities (the enzymes being synthesised during the growth phase). Biochemical evidence suggests that at least five enzymes are involved in the synthesis of esters within yeast. Ester-hydrolysing activities (esterases) may play a determining role in the final beer ester levels of products such as membrane-filtered beer and bottle-refermented beer. Recently, scientists have taken advantage of the completed Saccharomyces cerevisiae genome sequence database and the powerful tools of molecular biology to identify the corresponding genes and investigate the physiological role of ester synthesis. Recent rapid progress has provided insights not only into the regulation of cellular ester synthesis, but also into some general mechanisms of gene regulation. Three distinct alcohol acetyltransferase genes (ATF1, LgATF1 and ATF2), responsible for the production of acetate esters, have been cloned from different yeast backgrounds. A fourth gene, EHT1, described as an ethanol hexanoyltransferase, is involved in the synthesis of the medium-chain fatty acid esters.

21.1 Introduction The synthesis of volatile aliphatic esters by yeast is of major industrial interest because the presence of these compounds determines the fruity aroma of fermented beverages.1 Esters represent the largest group of flavour compounds in alcoholic beverages. In beer, the major esters are ethyl acetate, isobutyl acetate, isoamyl acetate, phenylethyl acetate and the C6–C10 medium-chain fatty acid (MCFA) ethyl esters (Table 21.1). Because of the difference in ester flavour thresholds, it is more important to look at individual beer ester levels than at the total beer ester level. Among these esters, only isoamyl acetate concentrations are above the threshold level in most lager beers (Table 21.1; see average level of isoamyl acetate). Isoamyl acetate is therefore considered a major contributor to beer fruitiness.

214 Table 21.1

BREWING YEAST FERMENTATION PERFORMANCE

Ester levels in European lager beers2

Compound

Concentration (ppm) Average (n  48)

Range Ethyl acetate

8–32

Threshold3 (ppm)

18.4

30

Isoamyl acetate Ethyl caproate

0.30–3.80 0.05–0.30

1.72 0.14

1.2 0.21

Ethyl caprylate Phenylethyl acetate

0.04–0.53 0.10–0.73

0.17 0.54

0.9 3.8

Table 21.2

Description

Light fruity, solvent-like Banana, peardrop Apple-like with note of aniseed Apple-like Roses, honey, apple, sweetish

Synthesis of medium-chain fatty acid esters by ‘eht1’ family mutants

Strain Wild-type eht1 mutant Triple disruptant

Ethyl caproate

Ethyl caprylate

Ethyl caprate

100 87 26

100 72 29

100 87 23

Dufour and Llorente (unpublished).

Synthesis of esters requires two substrates: alcohol and carboxylic acid. Although esters can be formed via a chemical reaction, it is now well established that the esters found in most beers are the result of acyltransferase activities (EC 2.3.1) or ester synthases (for a review see Mason and Dufour4). The carboxylic acid needs to be activated with coenzyme A (CoA) prior to reaction.5,6 Consequently, ester synthesis is an energy-requiring process. Biochemical evidence suggests that several enzymes are involved in the synthesis of esters.7 Recently, scientists have taken advantage of the completed Saccharomyces cerevisiae genome sequence database and the powerful tools of molecular biology to identify the corresponding genes and investigate the physiological role of ester synthesis (for reviews see Mason and Dufour4 and Dufour et al.8). Recent rapid progress has provided insights not only into the regulation of cellular ester synthesis, but also into some general mechanisms of gene regulation. Three distinct alcohol acetyltransferase (AATase; EC 2.3.1.84) genes (ATF1, LgATF1 and ATF2), responsible for the production of acetate esters, have been cloned from different yeast backgrounds. A fourth gene, EHT1, has been described as an ethanol hexanoyl CoA transferase. Sequence comparisons reveal that EHT1 belongs to a three-member gene family.9 A combination of simple, double and triple deletions does not affect growth. Disruption of one of the genes results in decreased levels of medium-chain acid esters, the decrease being accentuated with the increase in the number of disrupted genes (Table 21.2). The results suggest an internal functional redundancy as the phenotype was increased with the number of disrupted genes. Analysis of the relationship between levels of MCFA, and the corresponding ethyl esters strongly suggests that these enzymes may be involved in the removal of the toxic short-chain fatty acids (MCFAs) (Dufour and Llorente, unpublished).

CONTROL OF ESTER SYNTHESIS DURING BREWERY FERMENTATION

2

1 Pyruvate

Acetaldehyde

215

3 Acetate

Acetyl CoA

Fig. 21.1 Synthesis of acetyl coenzyme A (CoA). 1: Pyruvate decarboxylase; 2: acetaldehyde dehydrogenase; 3: acetyl CoA synthase.

It is generally believed that the most abundant acyl CoA, acetyl CoA, arises from the decarboxylation of pyruvate (involving pyruvate dehydrogenase).4 The pyruvate dehydrogenase, however, is subject to glucose repression and almost certainly is inactive under brewing fermentations.10 The evidence suggests that acetyl CoA arises from acetaldehyde produced from pyruvate (involving pyruvate decarboxylase), via the intermediary of acetate (involving acetaldehyde dehydrogenase), followed by direct activation of acetate with adenosine triphosphate (ATP; involving acyl CoA synthase) (Fig. 21.1).11 The other acyl CoAs originate from the metabolism of fatty acids. The need to understand and control ester synthesis is driven by problems encountered in brewing procedures, such as: (i) high-gravity brewing (production of disproportionate amounts of ethyl acetate and isoamyl acetate); (ii) the use of large-scale cylindroconical fermenters (reduction of ester levels); and (iii) the production of reduced alcohol beers (lack of flavour compounds).

21.2 Ester formation and excretion during fermentation Higher alcohols and esters are mainly produced during the primary fermentation. At the end of fermentation, a significant amount of esters is still synthesised, while the production of higher alcohols reaches a plateau. Up to 40% of the final beer ester level can be synthesised during this period.12–16 Ester excretion probably occurs through passive diffusion. Unlike acetate ester excretion, which is rapid and complete, the transfer of fatty acid ethyl esters to the fermenting medium decreases with increasing chain length, from 100% for ethyl caproate, to 54–68% for ethyl caprylate, to 8–17% for ethyl caprate. Longer chain fatty acid ethyl esters are only found in the yeast cells.17,18 Distribution of MCFA esters between yeast and beer is also influenced by the type of yeast used, with a larger proportion of the MCFA esters found in lager yeast.18 An increase in the fermentation and/or maturation temperature releases higher levels of MCFA esters through more efficient excretion and/or autolysis of yeast.19–21

21.3 The rate-limiting factors of ester synthesis and the relationship between ester synthesis, lipid metabolism and growth Ester synthesis requires acyl CoA, the formation of which is a cytosolic process and is dependent on the supply of acetyl CoA as a precursor. Acetyl CoA is also involved in numerous other reactions within the cell, including the biosynthesis of lipids and amino acids, and the tricarboxylic acid (TCA) cycle (Fig. 21.2).

216

BREWING YEAST FERMENTATION PERFORMANCE

Fig. 21.2 Relationship between synthesis of esters and of cellular components. CoA: coenzyme A; TCA: tricarboxylic acid.

21.3.1

Synthesis of the acetate esters

As mentioned earlier, a significant part of the acetate ester synthesis during fermentation occurs after the yeast growth phase. Three hypotheses have been formulated to explain the acetate ester synthesis profile during fermentation, each based on a different ratelimiting factor: the availability of acetyl CoA, the availability of the higher alcohol, and the synthesis of the alcohol acetyltransferase. Because of the strong interrelationship between acetyl CoA and ester synthesis, Nordström22 concluded that any factor influencing the pool of acetyl CoA will affect the formation of esters. Thus, when long-chain saturated fatty acids are added to the fermenting medium, they are transported into the yeast cells and incorporated into the cell membranes. Consequently, they reduce the cellular acetyl CoA utilisation for fatty acid synthesis.23 As a result, there is an increase of the intracellular acetyl CoA level with a resulting increase in ester synthesis. Conversely, any factor lowering the cellular acetyl CoA pool will decelerate ester formation. Thus, excessive growth will lead to a major utilisation of acetyl CoA in the formation of new biomass (lipids, proteins, etc.), leaving little acetyl CoA for ester formation. According to Thurston et al.,24,25 there are two induction phases of ester synthesis during fermentation. The first induction occurs at the beginning of fermentation. Acetyl CoA and oxygen are rapidly utilised for the production of unsaturated fatty acids and sterols. Immediately following this stage, an equilibrium between acetyl CoA consumption for lipid and ester synthesis is established. This corresponds to the first induction of ester synthesis. At the point during fermentation where lipid synthesis ceases, the second induction takes place. Thurston et al.24 suggested that the arrest of the lipid synthesis is responsible for a burst in acetyl CoA available for the ester synthesis. Because of the shift in the acetyl CoA/reduced CoA (CoASH) ratio towards higher values, acetate ester synthesis is stimulated,25 maintaining the equilibrium between acetyl CoA and CoASH. Although short-lived, this second induction is highly significant, and could contribute up to 40% of the final beer ester concentration. Yoshioka and Hashimoto,26 however, suggested that acetyl CoA is not an important factor in determining the formation of acetate esters. Their investigation of the profile of the AATase activity and higher alcohol formation during fermentation led to the conclusion that the slow rate of acetate ester synthesis was attributable to the rapid decrease in AATase activity and insufficient amounts of higher alcohols. The latter was supported

Acetate ester synthesis

CONTROL OF ESTER SYNTHESIS DURING BREWERY FERMENTATION

217

C

A

B

Yeast growth Fig. 21.3

Relationship between yeast growth and acetate ester synthesis.

by Calderbank and Hammond,27 who suggested that the higher alcohol availability determines the rate of ester formation under normal fermentation conditions. They claimed that many of the effects of fermentation conditions on ester synthesis are related to substrate alcohol availability. As a guideline, any condition that stimulates yeast growth will increase the production of higher alcohols during fermentation.28 By comparing the in vitro and in vivo AATase specific activities during fermentation in the presence of high levels of isoamyl alcohol, Malcorps et al.29 concluded that acetyl CoA availability is not the determining factor for the rate of ester synthesis. They related the high specific rate of ester synthesis at the end of the growth phase to the induction of AATase. This was subsequently confirmed by studies on mutants.30 In practice, there is probably an overlap of the effects of the different limiting factors, depending on yeast growth.28 Under reduced yeast growth conditions [limiting amount of oxygen, low free amino nitrogen (FAN), low zinc, immobilised cells], AATase activity will probably be the determining factor owing to low levels of enzyme synthesis (Fig. 21.3, point A). Under excessive yeast growth conditions, the level of acetyl CoA could be considered to be limiting since the major proportion of acetyl CoA will be used for growth, resulting in a deficiency of acetyl CoA for ester synthesis (Fig. 21.3, point B). In between the two growth extremes, a maximum should be observed for ester synthesis (Fig. 21.3, point C).28 21.3.1

Synthesis of the medium-chain fatty acid esters (C6–C10)

The synthesis of the MCFA esters is also influenced by yeast growth.31 To understand how fermentation conditions influence the synthesis of MCFA esters, it is necessary to consider the role of MCFAs as ester precursors. The biosynthesis of fatty acids and its relationship with MCFA ester formation is presented in Fig. 21.4. Because of the close relationship between ester formation and fatty acid biosynthesis (acyl CoA substrate), it is likely that the mechanisms operating to control the ethyl ester levels are also applicable to the control of MCFA in beer. MCFA production appears to be inversely related to yeast growth.32 The key enzyme in the regulation of fatty acid biosynthesis is the acetyl CoA carboxylase.33,34 Under brewing fermentation conditions (limiting amount of oxygen), long-chain

218

BREWING YEAST FERMENTATION PERFORMANCE

Acetyl CoA

Malonyl CoA

>> C6 CoA

C8 CoA

Acetyl CoA Carboxylase Long-chain saturated acyl CoA