Brewing Microbiology

Brewing Microbiology
Second edition

Edited by

F.G. Priest and I. Campbell

International Centre for Brewing and Distilling
Heriot-Watt University
Edinburgh, UK

Springer-Science+Business Media, B.V.

Published by Chapman & Hall, 2-6 Boundary Row, London SE18HN,
UK

ISBN 978-1-4757-4681-5 ISBN 978-1-4757-4679-2 (eBook)

DOI 10.1007/978-1-4757-4679-2

First edition 1987

Second edition 1996
© 1996 Springer Science+Business Media Dordrecht
Originally published by Chapman & Hall in 1996.
Softcover reprint of the hardcover 1st edition 1996
Typeset in 10/12 Palatino by WestKey Ltd, Falmouth, Cornwall
Apart from any fair dealing for the purposes of research or private study,
or criticism or review, as permitted under the UK Copyright Designs and
Patents Act, 1988, this publication may not be reproduced, stored, or
transmitted, in any form or by any means, without the prior permission in
writing of the publishers, or in the case of reprographic reproduction only
in accordance with the terms of the licences issued by the Copyright
Licensing Agency in the UK, or in accordance with the terms of licences
issued by the appropriate Reproduction Rights Organization outside the
UK. Enquiries concerning reproduction outside the terms stated here
should be sent to the publishers at the London address printed on this
page.
The publisher makes no representation, express or implied, with regard
to the accuracy of the information contained in this book and cannot
accept any legal responsibility or liability for any errors otomissions that
maybe made.
A catalogue record for this book is available from the British Library
Library of Congress Catalog Card Number: 95-70016
~ Printed on permanent acid-free text paper, manufactured in
accordance with ANSI/NISO Z39.48-1992 and ANSI/NlSO Z39.48-1984
(Permanence of Paper).
 Contents

List of contributors ix
Preface x
1 Systematics of yeasts 1
1. Campbell
1.1 Classification of yeasts 1
1.2 Nomenclature of yeasts 7
1.3 Properties for identification of yeasts 9
References 10
2 The biochemistry and physiology of yeast growth 13
T.W. Young
2.1 Introduction 13
2.2 Yeast nutrition 14
2.3 Yeast metabolism 15
2.4 Yeast propagation 28
2.5 Brewery fermentation 34
References 40
3 Yeast genetics 43
J.R.M. Hammond
3.1 Introduction 43
3.2 Genetic features of Saccharomyces cerevisiae 44
3.3 The need for new brewing yeasts 48
3.4 Genetic techniques and their application to the analysis
and development of brewing yeast strains 50
3.5 The commercial use of genetically modified brewing
yeasts 73
3.6 Conclusions 74
Acknowledgements 75
References 75
VI Contents
4 The microflora of barley and malt 83
B. Flannigan
4.1 The microflora of barley 83
4.2 The microflora of malt 96
4.3 Effects of microorganisms on malting 102
4.4 Effects of the microflora on beer and distilled
spirit 107
4.5 Health hazards 111
4.6 Assessment of mould contamination 119
References 120
5 Gram-positive brewery bacteria 127
F.G. Priest
5.1 Introduction 127
5.2 Lactic acid bacteria 128
5.3 Lactobacill us 134
5.4 Pediococcus 147
5.5 Leuconostoc 151
5.6 Homofermentative cocci 152
5.7 Micrococcus and Staphylococcus 153
5.8 Endospore-forming bacteria 154
5.9 Identification of genera of Gram-positive bacteria of
brewery origin 154
5.10 Concluding remarks 156
References 157
6 Gram-negative spoilage bacteria 163
H.J.J. Van Vuuren
6.1 Introduction 163
6.2 Acetic acid bacteria 165
6.3 Enterobacteriaceae 169
6.4 Zymomonas 177
6.5 Anaerobic Gram-negative rods 180
6.6 Megasphaera 183
6.7 Miscellaneous non-fermentative bacteria 183
6.8 Detection, enumeration and isolation 184
6.9 Conclusions 186
Acknowledgements 187
References 187
7 Wild yeasts in brewing and distilling 193
1. Campbell
7.1 Introduction 193
7.2 Detection of wild yeasts 193
7.3 Identification of wild yeasts 198
7.4 Effects of wild yeasts in the brewery 201
 Contents vii
7.5 Elimination of wild yeasts 205
References 207
8 Rapid detection of microbial spoilage 209
1. Russell and T.M. Dowhanick
8.1 Introduction 209
8.2 Impedimetric techniques (conductance, capacitance) 211
8.3 Microcalorimetry 214
8.4 Turbidometry 214
8.5 Flow cytometry 215
8.6 ATP bioluminescence 216
8.7 Microcolony method 218
8.8 Direct epifluorescence filter technique 220
8.9 Protein fingerprinting by polyacrylamide gel
electrophoresis 221
8.10 Immunoanalysis 222
8.11 Hybridization using DNA probes 224
8.12 Karyotyping (chromosome fingerprinting) 226
8.13 Polymerase chain reaction 228
8.14 Random amplified polymorphic DNA PCR 230
8.15 Summary 231
References 231
9 Methods for the rapid identification of microorganisms 237
C.S. Gutteridge and F.G. Priest
9.1 What is identification? 237
9.2 Levels of expression of the microbial genome 238
9.3 Identification at the genomic level 241
9.4 Techniques for examining proteins 244
9.5 Methods that examine aspects of cell composition 247
9.6 Developments in techniques for studying morphology and
behaviour 257
9.7 Future trends in rapid identification 264
Acknowledgements 267
References 267
10 Cleaning and disinfection in the brewing industry 271
M. Singh and J. Fisher
10.1 Introduction 271
10.2 Definitions 271
10.3 Standards required within a brewery 272
10.4 Cleaning methods available 275
10.5 Soil composition 280
10.6 Process of detergency 280
10.7 Chemistry of detergents 281
10.8 Caustic and alkaline detergents 282
V111 Contents
10.9 Sequestrants 283
10.10 Acids 286
10.11 Surface active agents 287
10.12 Disinfectants and sanitizers used in breweries 290
10.13 Oxidizing disinfectants 291
10.14 Non-oxidizing disinfectants 294
10.15 Water treabnent 297
10.16 Steam 299
10.17 Summary 299
References 300
Index 301
 Contributors

I. Campbell International Centre for Brewing and Distilling, Heriot-Watt

University, Edinburgh EH14 4AS, UK
T.M. Dowhanick Research Department, Labatt Breweries of Canada,
London, Ontario N6A 4M3, Canada
J. Fisher Diversey (FB) Ltd Technical Centre, Greenhill Lane, Riddings,
Derbyshire DL35 4LQ, UK
B. Flannigan International Centre for Brewing and Distilling, Heriot-Watt
University, Edinburgh EH14 4AS, UK
C.S. Gutteridge Reading Scientific Services Ltd, Lord Zuckerman Research
Centre, The University, Whiteknights, PO Box 234, Reading, Berkshire RG6
2LA, UK
J.R.M. Hammond BRF International, Lyttel Hall, Coopers Hill Road,
Nutfield, Redhill, Surrey RHI 4HY, UK
F.G. Priest International Centre for Brewing and Distilling, Heriot-Watt
University, Edinburgh EH14 4AS, UK
I. Russell Research Department, Labatt Breweries of Canada, London,
Ontario, N6A 4M3, Canada
M. Singh Diversey (FB) Ltd Technical Centre, Greenhill Lane, Riddings,
Derbyshire DL35 4LQ, UK
H.J.J. Van Vuuren Department of Microbiology, University of
Stellenbosch, Stellenbosch 7600, South Africa
T.W. Young Department of Biochemistry, University of Birmingham, PO
Box 363, Birmingham B15 2TT, UK
 Preface

During the latter part of the last century and the early years of this century,
the microbiology of beer and the brewing process played a central role in
the development of modern microbiology. An important advance was
Hansen's development of pure culture yeasts for brewery fermentations
and the recognition of different species of brewing and wild yeasts. The
discovery by Winge of the life cycles of yeasts and the possibilities of
hybridization were among the first steps in yeast genetics with subsequent
far-reaching consequences. Over the same period the contaminant bacteria
of the fermentation industries were also studied, largely influenced by
Shimwell's pioneering research and resulting in the improvement of beer
quality.
Towards the end of the century, the influence of brewing microbiology
within the discipline as a whole is far less important, but it retains an
essential role in quality assurance in the brewing industry. Brewing
microbiology has gained from advances in other aspects of microbiology
and has adopted many of the techniques of biotechnology. Of particular
relevance are the developments in yeast genetics and strain improvement
by recombinant DNA techniques which are rapidly altering the way
brewers view the most important microbiological components of the
process: yeast and fermentation. Moreover, the changing emphasis in
quality control from traditional plating techniques, which essentially pro-
vide a historical account of the process, to new rapid methodologies,
which give up-to-the-minute microbiological status reports of the process
upon which brewers can base decisions, is playing an important role in
the way the modem brewery operates. Developments in other aspects of
brewing, such as packaging under reduced oxygen levels, are affecting
the range of spoilage organisms encountered in beer and have led to
strictly anaerobic bacteria giving spoilage problems. In preparing this
second edition of Brewing Microbiology, we have covered these and other
developments in the microbiology of the brewing process and its products
 Preface xi

while retaining the comprehensive yet specialist treatment of the first

edition.
Once again, we thank the authors for their contributions and Nigel
Balmforth of Chapman & Hall for his help and encouragement in the
preparation of the book.

Fergus G. Priest, lain Campbell

 CHAPTER 1

Systematics of yeasts
1. Campbell

Systematics includes classification, nomenclature and identification.

Routine identification of culture yeasts will be discussed later (Chapter 7);
this present chapter is concerned primarily with classification and
nomenclature of yeasts, and the general principles of yeast identification.

1.1 CLASSIFICATION OF YEASTS

What is a yeast? No satisfactory definition exists, and such features of

commonly encountered yeasts as alcoholic fermentation or growth by
budding are absent from a substantial minority of species. Although yeasts
are generally accepted as fungi which are predominantly unicellular, there
are various borderline 'yeast-like fungi' which are difficult to classify. Also,
many yeasts are capable, under appropriate cultural conditions, of growing
in mycelial form, either as true mycelium or as a pseudomycelium of
branched chains of elongated yeast cells.
Most mycologists accept the series of publications of the Dutch group of
yeast taxonomists as definitive. The current classification is still basically
that of Kreger-van Rij (1984), although updated (Barnett, Payne and
Yarrow, 1990) to recognize the discovery of new species or recent changes
of attitude to the significance of taxonomic tests (Kurtzman, 1984;
Kreger-van Rij, 1987; Kurtzman and Phaff, 1987).
Three of the four groups of fungi include yeasts: Ascomycetes, Basidio-
mycetes and Deuteromycetes. Although Zygomycetes may develop yeast
morphology under unusual cultural conditions, their normal existence is
in the filamentous form, as non-septate hyphae. Classification of fungi
(Table 1.1) is largely on the form of vegetative growth and the nature of the
spores, if formed. Physiological properties, as widely used in bacteriology,
are not used in identification of filamentous fungi; so far as yeasts are

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and I. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412591502
2 Systematics of yeasts
Table 1.1 Classification of yeasts (from Kreger-van Rij, 1984)
1. Group 1: Ascomycetes. The sexual spores are formed endogenously, i.e.
within the cell. In filamentous fungi a specialized spore-bearing ascus is
formed; in yeasts, the spores develop within the former vegetative cell
which is then correctly termed the ascus (but note the exception, Lipomyces).
(a) Spermphthoraceae (needle-shaped spores; Fig. l.1a); principal genera
Metschnikowia, Nematospora.
(b) Saccharomycetaceae (other forms of spore; Fig. 1.1b,c,d,e); four families,
distinguished mainly by method of vegetative growth.
(i) Schizosaccharomycoideae (growth by binary fission, Fig. 1.2a); one
genus, Schizosaccharomyces.
(ii) Nadsoniodeae (growth by polar budding, Fig. 1.2b); principal genera
Hanseniaspora, Nadsonia, Saccharomycodes, distinguished by form of
spores.
(iii) Lipomycoideae (growth by multilateral budding, Fig. 1.2c, but the
principal characteristic is the 'exozygotic ascus', a sac-like ascus with
numerous oval spores, Fig. LIe); one genus, Lipomyces.
(iv) Saccharomycoideae (growth by multilateral budding, Fig. 1.2c);
numerous genera, distinguished mainly by details of the sporulation
cycle and spore morphology. The most important genera in the
brewing and distilling industries are Debaryomyces, Dekkera,
Kluyveromyces, Pichia, Saccharomyces, Schwanniomyces, Torulaspora and
Z ygosaccharomyces.
2. Group 2: Basidiomycetes. The sexual spores are formed exogenously. Yeasts
of this group are unimportant in the fermentation industries, but their
non-sporing, 'imperfect' forms, especially Sporobolomyces, Rhodotorula and
Cryptococcus, are common surface organisms of plant materials, including
barley and malt.
3. Group 3: Blastomycetes, Deuteromycetes or Fungi Imperfecti. No sexual
spores are formed.
(a) Sporobolomycetaceae (forming exogenous asexual spores, i.e.
ballistospores); two genera, Bu/lera (relatively rare) and Sporobolomyces
(very common).
(b) Cryptococcaceae (no exogenous asexual spores). Genera of this family,
which represent those yeasts of groups 1 and 2 which have lost the ability
to form sexual spores, are classified by different characteristics from the
sporing yeasts and therefore do not necessarily coincide with the
equivalent 'perfect' genera. Genera are distinguished by the form of
vegetative growth, fermentative ability and a somewhat haphazard
selection of other tests. Principal genera of importance to the fermentation
industries are Kloeckera (which grows by polar budding), Brettanomyces,
Candida, Cryptococcus and Rhodotorula (growth by multilateral budding).
The genus Trichosporon, often a contaminant of cereal grains, grows both
by multilateral budding and a form of fission. Torulopsis, formerly a
separate genus including various species of spoilage yeasts of the
brewing and related industries, is now incorporated into the genus
Candida (Kreger-van Rij, 1984).

concerned, they are used to distinguish species. Formerly, the ability to

utilize nitrate as a nitrogen source was used to distinguish genera but,
following the demonstration by DNA analysis that species of Hansenula and
 Classification of yeasts 3
Pichia differing in utilization of N03- were otherwise identical (Kurtzman,
1984), the 'nitrate-positive' genus Hansenula has now been abandoned.
Former Hansenula spp. are now incorporated in the genus Pichia, or other
related genera according to life cycles and sporulation characteristics
(Barnett, Payne and Yarrow, 1990).
The definitive property of Ascomycetes is the production of endoge-
nous sexual spores (ascospores). Further subdivision as families and
genera is based on the type of spores (Fig. 1.1), and the nature of the life
cycle by which they are formed. The mode of vegetative multiplication is
also relevant, since growth by binary fission (Fig. 1.2a) or polar budding
(Fig. 1.2b) is sufficiently different from the more common multilateral
budding to be of taxonomic importance. Within the group of yeasts
growing by multilateral budding, the genera Saccharomyces, Kluyveromyces,
Torulaspora and Zygosaccharomyces form a closely related subgroup of
vigorously fermentative yeasts. Indeed, by earlier classification (e.g.
Lodder and Kreger-van Rij, 1952), species at present in these four genera
were all allocated to one genus, Saccharomyces. Subsequently, the differ-
ences in vegetative growth cycles and sporulation were considered suffi-
ciently different to justify four separate genera. S. cerevisiae is of such
industrial importance that its vegetative and sporulation cycles have been
carefully studied, in recent years largely from the viewpoint of applied
genetics. Wild sporulating strains of S. cerevisiae are normally diploid, and
meiosis on sporulation produces asci containing four haploid spores
(Chapter 3). The typical appearance of asci of S. cerevisiae is illustrated in

a b c d e

Fig.t.t Yeast spores. (a) Needle spores. The original vegetative cell, top right, has
been distended to a club shape by development of the spores. Normally either one
or two spores will be formed in Metschnikowia or Nematospora species. (b) Oval
spores, as in the linear ascus of Schizosaccharomyces pombe. (c) 'Hat' spores in the
ascus, and free. The hat appearance is created by a tangential plate or ring forming
the 'brim'. Pichia membranaefaciens is a brewery contaminant forming this type of
spore. (d) 'Saturn' spores, as in Williopsis saturnus. The ring is located equatorially
on the spore, giving the appearance of the planet Saturn. (e) Ascus of Lipomyces.
Although the ascospores are endogenous, i.e. within the structure of the ascus, the
ascus itself is a separate structure (above) from the original vegetative cell (below).
In all other ascosporogenous yeasts the spores develop within the original vegeta-
tive cell which is then, by definition, called the ascus.
4 Systematics of yeasts

~c9
c=J
cO °0
~ ~
xO 0'1-.
0°
C) CJ
b c

Fig. 1.2 Vegetative reproduction of yeasts. (a) Binary fission. The original cell
elongates and after nuclear division a septum separates the two cells, which then
complete the formation of the cell wall, and break apart. (b) Polar budding.
Successive divisions are from either end (pole) of the cell, alternately. The scar
tissue from budding gives the cell a 'lemon' shape. Note also the broad base to
the developing bud, characteristic of this method of growth. In the next generation
the two cells will bud from end X. (c) Multilateral budding, i.e. budding from
many different points on the surface in successive generations. Note the typical
narrow base to the developing bud.

Fig. 1.3a. Often tetrahedral asci are formed, as the most compact
arrangement of the four spores. However, occasional failure of meiotic
nuclear divisions can result in two- or three-spored asci. Spores of
Saccharomyces species are not liberated immediately on maturation; spores
of Kluyveromyces species, although superficially similar in appearance in
some species, are rapidly released.
Kluyveromyces is composed of both homo- and heterothallic species with
spherical or ellipsoidal spores, and homothallic species producing reniform
(kidney-shaped) spores. The number of spores produced by Kluyveromyces
species is often large: up to 60 in K. polysporus.
The sporulation process of Kluyveromyces species, and in particular the
early liberation of mature spores, was judged to be sufficiently distinctive
to justify a separate genus in the previous classification of yeasts (Lodder,
1970). The remaining species of actively fermenting yeasts were, somewhat
illogically, retained in a single genus Saccharomyces, even though there were
substantial differences in methods of sporulation within the genus. This
was subsequently rectified in the more recent classification of Kreger-van
Rij (1984), with the transfer of species to the genera Torulaspora and
Z ygosaccharomyces.
Torulaspora is haploid in the vegetative growth cycle, but sporulation
results from homothallic fusion between the nuclei of parent cell and bud,
under the cultural conditions which promote sporulation. Zygosaccharo-
myces is also haploid in the vegetative state, but fusion leading to sporu-
lation is between independent cells, forming a diplOid zygote which
 Classification of yeasts 5

a b c

Fig. 1.3 Spores of Sacchilromyces, Torulaspora and Zygosacchilromyces. (a)

Saccharomyces. Heterozygous diploid cells undergo two successive meiotic nuclear
divisions to produce four haploid spores in plane (top left) or tetrahedral (bottom
right) configuration. Often only two or three spores are formed, by failure of
nuclear division. (b) Torulaspora. Under conditions which stimulate sporulation,
the haploid cell undergoes nuclear fusion with its bud, which although not yet
fully developed is already, in nuclear terms, a distinct individual. In this way a
transient diploid cell is formed, and then two haploid spores. (Note that in some
species of Debaryomyces, which have a similar method of sporulation, the spores
are formed in the buds. Development in the mother cell, as shown, is more normal,
and invariable practice in Torulaspora.) (c) Zygosaccharomyces. Conjugation of inde-
pendent haploid cells precedes ascus formation. The ascus, composed of the two
former vegetative cells and the conjugation tube, contains one to four haploid
spores; three, as shown, is a common occurrence.

produces two spores, one in each parent, or subsequent meiotic division

may result in two spores in each former cell. Asci of Torulaspora and
Zygosaccharomyces are illustrated in Figs 1.3b and 1.3c. The haploid veg-
etative cycle and sporulation in the manner of Torulaspora and
Zygosaccharomyces spp. are also properties of the common wild yeasts of
the genera Debaryomyces, Hansenula and Pichia. However, these genera
ferment sugars only weakly, or not at all (Kreger-van Rij, 1984).
Yeasts of the genera included in the Basidiomycetes are unlikely to occur
as contaminants of brewery or related fermentations, but are common
surface contaminants of barley and malt (Chapter 4). These yeasts of malt
are killed during mashing and hop boiling, and their poor growth at pH 5
or below prevents later reinfection.
It is reasonable to believe that the imperfect, non-sporing yeasts are
derived from heterothallic sporing yeasts, but in the absence of the opposite
mating type are unable to conjugate and form spores. Single spores germi-
nated in the absence of the opposite mating type grow indefinitely as
haploid cells, provided nutrients are available. Previously the principles of
classification were applied rigidly and yeasts were allocated to perfect or
6 Systematics of yeasts
Table 1.2 Perfect and imperfect states of common yeast contaminants of the
brewing industry (Barnett, Payne and Yarrow, 1990)
Spore-forming yeast Non-spore-forming synonym
Dekkera bruxellensis Brettanomyces bruxellensis
Dek. intermedia B. intermedius
Debaryomyces hansenii Candida famata
Hanseniaspora uvarum Kloeckera apiculata
H. valbyensis K. japonica
H. vineae K. africana
Issatchenkia orientalis C. krusei
Kluyveromyces marxianus c.kefyr
Pichia anomala C. pelliculosa
P·fabianii C.fabianii
P. fermentans C.lambica
P. guilliermondii C. guilliermondii
P.ohmeri
P. membranaefaciens C. valida
Saccharomyces cerevisiae C. robusta
S. exiguus C. holmii
Torulaspora delbrueckii C. colliculosa

imperfect genera according to sporulation. In the most recent classification

this principle has been less rigidly applied. Candida robusta (Lodder, 1970)
had the same properties as S. cerevisiae, except that no spores were formed.
Sporing yeasts were obviously S. cerevisiae, but the logic of the classification
broke down over the nomenclature of the majority of yeasts of the brewing
and related industries, which, in the course of their long history of artificial
cultivation, had lost the ability to form spores. To apply the name
S. cerevisiae was technically wrong, but was nevertheless accepted practice.
In the classification of Kreger-van Rij (1984) a separate species is not listed;
it is now a synonym of S. cerevisiae. Similarly the various other examples of
identical pairs of sporing and non-sporing yeasts (Table 1.2) are now
recognized by the name of the sporing species.
Largely to avoid the complications arising from the use of sporulation
as a fundamental property for 'classical' yeast taxonomy, and the require-
ment to observe spores as a first step in classical identification, various
authors have provided alternative identification schemes which ignore
sporulation. A simple, effective scheme introduced by Beech et al. (1968)
has been superseded by a succession of diagnostic keys by Barnett and
colleagues, who have provided a comprehensive, but unfortunately com-
plex, system for identification by physiological properties (see Barnett,
Payne and Yarrow, 1990). A simplified version of their system is the basis
of the API test kit for identification of yeasts (see Chapter 7).
 Nomenclature of yeasts 7
1.2 NOMENCLATUREOFYEASTS

Hansen, the pioneer yeast taxonomist, applied the name Saccharomyces

cerevisiaeto the traditional top-cropping ale yeast of the British and German
brewing industries. The different properties of lager yeast were acknowl-
edged by allocation to a different species, S. carlsbergensis. Other specific
names were applied to different yeasts according to their fermentation of
sugars, industrial properties or cell morphology, e.g. ale yeast (spherical,
or almost so) and wine yeast S. ellipsoideus (ellipsoidal) were distinguished
by Hansen partly on the basis of shape, and partly on different behaviour
in fermentation (Lodder and Kreger-van Rij, 1952). In the reclassification of
yeasts by Lodder and Kreger-van Rij (1952) the sole difference detectable
in the laboratory, morphology, was judged insufficient for separation of
two species and the wine yeast became S. cerevisiae var. ellipsoideus. How-
ever, another wine yeast, S. uvarum, and lager yeast S. carlsbergensis, despite
their biochemical similarities, remained distinct species. In the next reclas-
sification of yeasts (Lodder, 1970), S. cerevisiae, its ellipsoideus variety and
S. willian us, identical in fermentation properties but of distinctively elon-
gated cell morphology, were merged as a single species S. cerevisiae. An
equivalent series of yeasts, comprising S. carlsbergensis, S. uvarum and
S. logos, were united as S. uvarum. S. cerevisiae and S. uvarum were distin-
guished by the ability of only the latter to ferment melibiose. Subsequently,
extensive research on yeast hybridization demonstrated interbreeding,
with the production of fertile progeny, between S. cerevisiae and
S. carlsbergensis, which suggested that only one species was justified. Also,
the availability of these same hybrids as industrial yeasts completely
blurred any distinction in physiological terms between the species.
Therefore the allocation of both groups to a single species S. cerevisiae by
Kreger-van Rij (1984) was understandable. Indeed, many other former
species were also merged in the new definition of S. cerevisiae, as shown in
Table 1.3.
It was unfortunate that' classical' identification methods failed to distin-
guish the very different industrial properties of these yeasts. Gilliland's
(1971) objection that 'industrial' properties were sufficient to justify sepa-
rate species was confirmed by numerical analyses of data including not
only the standard morphological and physiological properties, but also
such industrially useful properties as tolerance to ethanol, flocculation and
fining ability and optimum growth temperature (Campbell, 1972). More
detailed analysis of the effect of temperature on growth of lager and
non-brewing strains of S. uvarum (i.e. S. carlsbergensis and S. uvarum,
respectively, as defined by Lodder and Kreger-van Rij in 1952) by Walsh
and Martin (1977) confirmed the distinctive nature of lager yeast, however
alike they may have been in the tests of'classical' taxonomy. Unfortunately,
the ability to ferment various sugars, some (galactose, melibiose, raffinose)
8 Systematics of yeasts
Table 1.3 Nomenclature of important yeast species of the brewing and related
industries, 1952-1984
1952 classification 1970 classification 1984 classification
(Lodder and Kreger-van Rij, 1952) (Lodder, 1970) (Kreger-van Rij, 1984)
S"xha"my", boy<>n",'
S.ovifonnis
j S. bayanus*
S. pastorianus*
S. cerevisiae
S. cerevisiae var. elliPsoideuS} S. cerevisiae
S. willianus
S. ""l,.,.g""l'j
S.logos S. uvarum
S. cerevisiae

S. uvarum
S. chevalieri S. chevalieri
S. italicus S. italicus
S. aceti t
S. diastaticus t
S. delbrueckii
S. rosei S. delbrneddlj
S. rosei
Torulaspora
delbrueckii
S. vafer t
S. "'ldlfi"lenj
S. bailii S. bailii
Z ygosaccharomyces
bailii
j
S. elegans
S·fragilis
S. marxianus KiulI""?my",
marxzanus K. marxianus
S. lactis K. lac tis

'5. bayanus and S. pastorianus are valid species again (Barnett, 1992).
tNew species.

unlikely to occur in brewery fermentations, has arbitrarily been given more

importance than industrially useful properties. However, the fundamental
difference between ale, lager and non-brewing strains of S. cerevisiae was
demonstrated again by 'DNA fingerprinting' (Pedersen, 1983; Meaden,
1990). Previous amalgamation of wine yeast species has been reversed on
the basis of similar analyses (Martini and Kurtzman, 1988; Barnett, Payne
and Yarrow, 1990), as shown in Table 1.3, which shows a history of the
renaming over the years of Saccharomyces species of relevance to brewing
and distilling. A more comprehensive history of the nomenclature of the
genus Saccharomyces has recently been provided by Barnett (1992). With so
many types of industrially important culture and wild yeasts all now
named S. cerevisiae, the industrial microbiologist had to continue to distin-
guish these organisms, but without the assistance of distinctive specific
names. Therefore the benefits of a standardized nomenclature were lost, as
each microbiologist developed individual classification schemes for the
different contaminant strains of the large species S. cerevisiae.
 Properties for identification of yeasts 9

1.3 PROPERTIES FOR IDENTIFICATION OF YEASTS

The properties used for yeast taxonomy according to the Dutch mycologists
include both morphology and physiology (Table 1.4). The large number of
growth tests listed by Lodder (1970), which included over 30 carbon or
nitrogen sources, has now been substantially reduced (Kreger-van Rij,
1984) to 18. Barnett, Payne and Yarrow (1983) listed 60 characters based on
sugar fermentation or aerobic growth on carbon or nitrogen compounds,
but suggested substantially fewer for routine identification. There is a
fundamental difference between the identification procedures of the Dutch
school and that of Barnett, Payne and Yarrow (1983, 1990): the former
identify strictly in descending hierarchical order, i.e. family, genus, species,
whereas the latter first subdivide yeasts in general into largely unrelated
groups. The subdivision suggested by Barnett and co-workers was based
on a small number of key tests, designed to subdivide all known yeasts into
groups of approximately equal numbers of species. These groups were in
tum subjected to a small number of tests chosen to identify species in that
group; different tests were required for different groups. The complications
largely negated the biochemical good sense of the plan, and a simpler API
system, although based on the same principles, effectively identifies in a
single set of tests (Anon., 1994). The three systems are compared in
Table 1.4.
Numerous biochemical properties other than fermentation or aerobic
growth tests have been suggested for use in classification, identification or
both (Campbell, 1987). Of these methods, only the determination of the

Table 1.4 Principles of identification of yeasts

Identification by Identification by physiological tests only
'classical' system
(Kreger-van Rij, 1984) (Barnett, Payne and Yarrow, API
1983, 1990) (Anon., 1994)
Isolate Isolate Isolate
J, J, J,
Observe morphology of Aerobic growth tests on Aerobic growth tests on
vegetative cells and spores various C and N various C compounds
(see Table 1.1) compounds
J, J, J,
Family, genus Physiological group Genus, species
J, l
Fermentation and Appropriate tests for that
aerobic growth tests group (usually additional
(see below) growth tests)
J, J,
Species Genus, species
Tests for identification of species (Kreger-van Rij, 1984): fermentation of mono-, di- and
trisaccharides (glucose, galactose, sucrose, maltose, lactose, raffinose); aerobic utilization of
above sugars, and of xylose, ethanol, glycerol, etc.
10 Systematics of yeasts
percentage of guanine + cytosine (% G + C) is still of practical value in
classification (Barnett, Payne and Yarrow, 1990). However, the % G + C is
of no help for identification of new isolates, since different species, and even
genera, may by chance share the same value, e.g. the very different yeasts
Brettanomyces anomaius, Pichia quercuum and S. cerevisiae are all 40.0%
(Kreger-van Rij, 1984). So it is impossible to identify an unknown culture
by determination of % G + C, but it is possible to recognize that an identi-
fication is wrong if % G + C values are widely different. An interesting
example in the genus Pichia was reported by Phaff and Starmer (1987):
several isolates from cacti were identified as P. membranaefaciens by physi-
ological tests but named as different species because of substantially differ-
ent G + C values. More recently, the reintroduction of the species S. bayanus
and S. pastorianus was based partly on different G + C content from other
strains of S. cerevisiae (Martini and Kurtzman, 1988; Barnett, Payne and
Yarrow, 1990); the validity of these restored species names was subse-
quently confirmed by electrophoretic analysis (Martini, Martini and
Cardinali, 1993). Implications of these various aspects of nucleic acid
relatedness have been discussed fully by Kurtzman and Phaff (1987) and
van der Walt (1987) with regard to the concept of species in yeast.
Certainly, however close % G + C values may be, only if reannealing
takes place of separated DNA strands of two different yeast strains can they
be regarded as the same species. DNA reassociation is possible only when
the bases are in essentially the same sequence over the entire DNA mole-
cule. At least 80% reassociation between strains is generally considered to
be a requirement for allocation to the same species (e.g. Price, Fuson and
Phaff, 1978; Kurtzman and Phaff, 1987), although some authors have
suggested that a lower value would be acceptable (van der Walt, 1987).
The ultimate application of this principle is the determination of the
entire sequence of bases of yeast DNA. While that is unlikely to be
achieved for all yeasts, the method of 'fingerprinting' of strains by DNA
fragments has proved useful in distinguishing industrial strains (see
Chapter 3), and also has potential for classification and identification of
yeasts (Meaden, 1990).

REFERENCES

Anonymous (1994) API 20C Analytical Profile Index, BioMerieux UK, Basingstoke.
Barnett, J.A. (1992) Yeast, 8,1.
Barnett, J.A., Payne, RW. and Yarrow, D. (1983) Yeasts, Characteristics and
Identification, Cambridge University Press, Cambridge.
Barnett, J.A., Payne, RW. and Yarrow, D. (1990) Yeasts, Characteristics and
Identification, 2nd edn, Cambridge University Press, Cambridge.
Beech, F. W., Davenport, RR, Goswell, R W. and Burnett, J.K. (1968) In Identification
Methods for Microbiologists, Part B (eds B.M. Gibbs and D.A. Shapton),
Academic Press, London, p. 151.
 References 11
Campbell, I. (1972) Journal of General Microbiology, 73, 279.
Campbell, I. (1987) In Brewing Microbiology (eds F.G. Priest and I. Campbell),
Elsevier, London, pp. 187-205.
Gilliland, R.B. (1971) Journal of the Institute of Brewing, 77, 276.
Kreger-van Rij, N.J.W. (1984) The Yeasts, a Taxonomic Study, 3rd edn,
Elsevier IN orth-Holland, Amsterdam.
Kreger-van Rij, N.J.w. (1987) In The Yeasts, 2nd edn, Vol. 1 (eds A.H. Rose and
J.5. Harrison), Academic Press, London, p. l.
Kurtzman, CP. (1984) Antonie van Leeuwenhoek, 50,209.
Kurtzman, CP. and Phaff, H.J. (1987) In The Yeasts,2nd edn, Vol. 1 (eds A.H. Rose
and J.5. Harrison), Academic Press, London, p. 63.
Lodder, J. (1970) The Yeasts, a Taxonomic Study, 2nd edn, North-Holland,
Amsterdam.
Lodder, J. and Kreger-van Rij (1952) The Yeasts, a Taxonomic Study, North-Holland,
Amsterdam.
Martini, A.V. and Kurtzman, CP. (1988) Mycologia, 80,24l.
Martini, A.V., Martini, A. and Cardinali, G. (1993) Antonie van Leeuwenhoek, 63, 145.
Meaden, P.G. (1990) Journal of the Institute of Brewing, 96, 195.
Pedersen, M.B. (1983) Proceedings of the 19th Congress of the European Brewery
Convention, London, HRL Press, Oxford, p. 457.
Phaff, H.J. and Starmer, W.T. (1987) In The Yeasts, 2nd edn, Vol. 1 (eds A.H. Rose
and J.S. Harrison), Academic Press, London, p. 123.
Price, CW., Fuson, G.B. and Phaff, H.J. (1978) Microbiological Reviews, 42, 16l.
van der Walt, J.P. (1987) In The Yeasts, 2nd edn, Vol. 1 (eds A.H. Rose and
J.5. Harrison), Academic Press, London, p. 95.
Walsh, R.M. and Martin, P.A. (1977) Journal of the Institute of Brewing, 83, 169.
 CHAPTER 2

The biochemistry and

physiology of yeast growth
T.W. Young

2.1 INTRODUCTION

When introduced into a suitable aqueous environment, containing an

adequate supply of nutrients, at moderate temperature and value of pH,
viable yeast cells grow. Growing yeast cells increase in volume and mass
until they reach a critical size when bud formation is initiated (Pringle
and Hartwell, 1981). Once initiated, provided there are sufficient nutrients
to support further growth, buds increase in size and eventually separate
from the parent cell. The growth of a yeast culture therefore encompasses
both an increase in total cell mass (biomass) and an increase in cell number.
Growth thus involves the de novo synthesis of new yeast cells, i.e. the
synthesis of yeast cell constituent macromolecules. The mechanisms by
which these syntheses occur are found in the biochemical reactions of the
metabolism of the yeast cells. These reactions are also responsible for the
production of the ethanol and carbon dioxide of fermentation as well as
for the generation of the many compounds which contribute to the taste
and aroma of beer.
In brewing, yeast growth is observed during the propagation of pure
yeast cultures in specialized propagation vessels and also in the fer-
mentation process to yield beer. Growth cannot be separated from the
fermentation process and is necessary to the production of both beer and
fresh yeast for use in subsequent fermentations. Control of fermentation is
achieved by monitoring the changes in the fermentation medium resulting
from the metabolic activities of the growing yeast cells.

Brewing Microbiology, 2nd edn. Edited by F. C. Priest and 1. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412 59150 2
14 The biochemistry and physiology of yeast growth
2.2 YEAST NUTRITION

Yeasts will grow in simple media which contain fermentable carbohydrates

to supply energy and carbon skeletons' for biosynthesis, adequate nitrogen
J

for protein synthesis, mineral salts and one or more growth factors. It is
usual for some molecular oxygen to be provided although it would seem
that some strains at least of Saccharomyces cerevisiae do not have an absolute
requirement for oxygen (Macy and Miller, 1983).
Sources of carbon include monosaccharides such as D-glucose,
D-mannose, D-fructose, D-galactose and the pentose sugar D-xylulose, but
not other pentoses (Wang, Johnson and Schneider, 1980). Disaccharides
such as sucrose (cane sugar) and maltose are also fermented by brewer's
yeast, which however will not ferment lactose (milk sugar). Trisaccharides
such as maltotriose and raffinose are also fermented by brewers' yeast
although in the case of raffinose some strains conduct only a partial
hydrolysis. The organic compounds glycerol, ethanol and lactate are not
fermented but yeast may grow aerobically by respiration using these
compounds as sources of carbon and energy.
The glucose, fructose and sucrose present in brewers' wort are rapidly
used by the yeast in the early stages of fermentation. Sucrose is hydrolysed
by the enzyme invertase (~-D-fructofuranosidase, EC 3.2.1.26), which is
located externally to the cell membrane and bound within the cell wall. The
glucose and fructose produced, together with that present in wort, are then
transported into the cell. Transport is facilitated possibly by a carrier
molecule (permease) located in the membrane. Glucose has been estimated
to enter the cell some one million times faster than can be accounted for by
simple diffusion across the membrane (Heredia, Sols and De la Fuente,
1968). The uptake of maltose and maltotriose is mediated by specific
inducible permeases (Sols and De la Fuente, 1961; Harris and Thompson,
1960). An a-D-glucosidase (maltase, EC 3.2.1.20) is simultaneously induced
so that on entering the cell both the di- and trisaccharides are hydrolysed
to glucose.
The requirement for nitrogen for protein synthesis may be met by
ammonium ions although, when present, amino acids are a preferred
source. Thus when mixtures of amino acids are available (e.g. in brewers'
wort) growth is more rapid than when ammonium ions are the sole source
of nitrogen (Thome, 1949). Amino acids are taken up in a sequential manner
during fermentation (Jones and Pierce, 1964); this is thought to reflect the
properties and specificities of permeases located in the cell membrane.
Brewing yeasts show a requirement for minerals which resembles that
of other living organisms and in particular a supply of potassium, iron,
magnesium, manganese, calcium, copper and zinc is necessary. In addition,
it is usual for strains to require one or more accessory nutrients for growth
and most, if not all, need biotin. Many other nutrients are known to
 Yeast metabolism 15
stimulate the growth of yeasts although they are not absolutely required
for cell growth. Amongst such compounds are pantothenic acid, meso-
inositol, nicotinic acid, thiamin, p-aminobenzoic acid and pyridoxine. Most
of these materials are necessary cofactors of enzyme activity in yeast
metabolism.
When made wholly from malt, brewers' wort contains all the nutrients
needed by yeast for growth. The carbohydrate requirement is satisfied by
glucose, fructose, sucrose, maltose and maltotriose: the requirement for
nitrogen is met by amino acids and ample quantities of minerals and
accessory nutrients are present. The addition of calcium salts to the
brewing water aids yeast flocculation at the end of fermentation. Worts
made using high levels of carbohydrate material (sugars, cereal starches)
may be deficient in assimilable nitrogen and vitamins (particularly biotin).
Such worts as well as all-malt ones may be supplemented with the
addition of yeast foods based on preparations of yeast extract plus am-
monium, phosphate and zinc ions. More complex formulations may also
be used to ensure that the nutritional requirements of the yeast are met
(Hsu, Vogt and Bernstein, 1980).

2.3 YEAST METABOLISM

2.3.1 General aspects

Metabolism describes all the enzymic reactions which occur within the cell
and the organization and regulation of those reactions. Although from a
biochemical standpoint individual aspects of metabolism are considered as
separate pathways, such pathways do not in reality exist in isolation, but
are merely parts of a whole integrated metabolic process.
Each biochemical pathway consists of a series of chemical reactions
catalysed by enzymes. The catalytic moiety of an enzyme is invariably
protein, although lipid and/ or carbohydrate may be covalently associated
with particular protein molecules. An enzyme increases the rate of a
chemical reaction and enables the reaction to occur at physiological values
of temperature and pH. Extremes of temperature and/or pH inactivate
enzyme molecules by denaturing them. Enzymes show specificity towards
their substrates and often require cofactors to participate in their catalytic
activity.
Biochemical pathways may be described as catabolic, anabolic (bio-
synthetic), amphibolic or anaplerotic. Catabolic pathways are involved in
the degradation (usually by an oxidative process) of simple organic mole-
cules derived from the breakdown of polymers (e.g. amino acids from
proteins) and retain some of the energy released in a 'biologically useful'
form. Anabolic pathways consume energy and synthesize (usually by a
reductive process) the simple molecules which are subsequently assembled
16 The biochemistry and physiology of yeast growth
into macromolecules. Amphibolic pathways have both catabolic and ana-
bolic functions: they are central metabolic pathways which provide, from
catabolic sequences, the intermediates which form the substrates of ana-
bolic reactions. When during the operation of a catabolic sequence of an
amphibolic pathway, intermediates are removed for biosynthetic
purposes, catabolism would cease were it not for the operation of an-
aplerotic reactions. The function of anaplerotic reactions is to replace those
intermediates which link both catabolism and anabolism, thus ensuring the
continued operation of amphibolic pathways.
Biologically, energy produced from the oxidation of carbohydrate and
consumed in biosynthesis is stored in the form of adenosine triphosphate
(ATP). This high-energy compound on hydrolysis to adenosine diphos-
phate (ADP) and inorganic phosphate (Pi) liberates some 30.5 kJ morl
under standard conditions. The magnitude of the free energy of hydrolysis
of ATP is a function of pH, temperature and the concentrations of ATP,
ADP and Pi. In the conditions found inside the yeast cell, the hydrolysis of
1 mole of ATP may yield as much as 52 kJ.
Oxidative reactions in catabolism involve the removal of electrons from
intermediates. This process is controlled by dehydrogenase enzymes and
often involves the participation of the cofactor nicotinamide adenine di-
nucleotide (NAD+). Electrons are transferred to NAD+ in the form of the
hydride ion [Jr] to produce reduced NAD+ (NADH2):
NAD+ + [2H] ~ NADH + H+ (or NADH2)
Biosynthetic reactions are driven by ATP energy and the process of reduc-
tion is mediated by enzymes, most of which use nicotinamide adenine
dinucleotide phosphate (NADP+) as cofactor. The specificity of catabolic
enzymes for NAD+ and those of anabolism for NADP+ is an example of
'chemical compartmentation' and enables some degree of metabolic regu-
lation to be exerted through control of the levels of the two cofactors. The
relative concentration of the oxidized and reduced forms of a particular
cofactor may also serve a regulatory role.
The regulation of metabolism also involves 'biological
compartmentation' whereby biochemical pathways are located within
organelles or at specific points within the cell. In addition, control of
biochemical pathways may occur by a variety of other mechanisms such as:
1. regulating the amount of enzyme synthesized;
2. regulating the degradation of enzymes;
3. modifying the rate of enzyme activity by allosteric inhibition or
activation;
4. using isoenzymes to perform the same reaction for different purposes,
e.g. alcohol dehydrogenase (ADH): ADHI is used when cells grow on
ethanol, converting ethanol to acetaldehyde, whereas ADH2 is used
when cells grow on glucose and convert acetaldehyde to ethanol.
 Yeast metabolism 17
The foregoing is a very limited introduction to the subject of metabolism,
and readers unfamiliar with biochemical principles are advised to consult
a standard text, e.g. Stryer (1995).
The following sections will discuss only the main aspects of yeast
metabolism pertinent to a consideration of brewery fermentation.
Additional information may be found in Rose and Harrison (1969, 1970,
1971), Briggs et ai. (1982) and Nykanen and Suomalainen (1983).

2.3.2 Carbohydrate metabolism during fermentation

(a) Major pathways

The main pathway for the fermentation of glucose is the Embden-
Meyerhof-Pamas route (also named the EMP or glycolytic sequence).
Wort-fermentable carbohydrates enter the pathway either as D-glucose or
D-fructose. Similarly intracellular storage compounds such as glycogen or
trehalose are hydrolysed to o-glucose.
The EMP pathway (Fig. 2.1) commences with the phosphorylation of the
hexose monosaccharide glucose using the high-energy bond of ATP to
produce glucose-6-phosphate (G-6-P) in a reaction mediated by the enzyme
hexokinase (Ee 2.7.1.1). Glycogen degradation by phosphorylase
(Ee 2.4.1.1) yields glucose-I-phosphate which is converted to G-6-P by the
enzyme phosphoglucomutase (Ee 2.7.5.1), which in yeast requires zinc for
activity. The next reaction in the pathway (Fig. 2.1) is the isomerization of
G-6-P to fructose-6-phosphate (F-6-P) by the enzyme glucose-phosphate
isomerase (Ee 5.3.1.9). This reaction is readily reversible. Intracellular
fructose, derived directly from wort fructose or from the action of cell wall
bound invertase, is probably phosphorylated by hexokinase to yield F-6-P.
The third enzyme of the glycolytic sequence, 6-phosphofructokinase
(Ee 2.7.1.11), mediates the formation of fructose-1,6-bisphosphate from
F-6-P and ATP. At this point in the EMP pathway, two molecules of ATP
have been consumed for each molecule of glucose entering the sequence.
Fructose-bisphosphate aldolase (Ee 4.1.2.13), a zinc-containing enzyme,
then catalyses the reversible cleavage of the six-carbon molecule into two
three-carbon molecules, triosephosphates. The triosephosphates are di-
hydroxyacetone phosphate and D-glyceraldehyde-3-phosphate; only the
latter molecule is further processed by the EMP pathway and an equilib-
rium is maintained between the two triosephosphates by the enzyme
triosephosphate isomerase (EC 5.3.1.1).
D-Glyceraldehyde-3-phosphate is converted into 3-phospho-o-
glyceroyl phosphate by the enzyme glyceraldehyde-phosphate de-
hydrogenase (Ee 1.2.1.12) using inorganic phosphate and transferring
electrons to NAD+. This is the first reaction of the pathway which involves
the oxidation of the substrate. Furthermore, it is the first reaction which
involves the formation of a high-energy bond and some of the free energy
18 The biochemistry and physiology of yeast growth

CHzOH CH 20®
AlP ADP CHzO®
Hr.°~ ~~O~ ~O'SHZOH
HO~H CD H~H "®'~H
H OH H OH HO H
«-O-glucose glucose 6-phosphate fructose 6-phosphate
® ~ATP
~AOP
CH 20®
dihydroxy- ~HzOH ~CH20®
acetone C=O -
phosphate ~HzO ® ® H H H OH
CD [
+ HO H
CH 0 (ji', fructose -1-6 -bisphosphate
O-glyceraldehyde CH~H ~,
3-phosphate I
CHO

3-phospho- O-glycerate
Pi- ~ NAOH
NAO
z
®
CH20(P) CHzO®
I I
(HOH
I
<II' ",' ~HOH
COO-
@t
AlP AOP COO®
3-phospho-O-glyceroyl phosphate
(H 20H CH z AOP ATP CH3
I
(HO® . II ~
CO J) ..> L .
<II I
C=O
I I4n\ I
(00- COO- ® COO-
2-phospho-O-glycerate phosphoenol- pyruvate
pyruvate I I
I I
I I
I ~

: BIOSYNTHESES
I
I
~

ETHANOL + (0 2

Fig. 2.1 The Embden-Meyerhof-Parnas or glycolytic pathway. I, Hexokinase

(EC 2.7.1.1); 2, glucose-phosphate isomerase (EC 5.3.1.9); 3, 6-phosphofructokinase
(EC 2.7.1.11); 4, fructose-bisphosphate aldolase (EC 4.1.2.13); 5, triosephosphate
isomerase (EC 5.3.1.1); 6, glyceraldehyde-phosphate dehydrogenase (EC 1.2.1.12);
7, phosphoglycerate kinase (EC 2.7.2.3); 8, phosphoglyceromutase (EC 2.7.5.3); 9,
enolase (EC 4.2.1.11); 10, pyruvate decarboxylase (EC 4.1.1.1).
 Yeast metabolism 19
available from the oxidative process is trapped in this bond. Thus the
phosphate bond formed at position 1 in 3-phospho-o-glyceroyl phosphate
has a standard free energy of hydrolysis of -49.3 kJ mor!. The next step in
the pathway is the transfer of some of the energy in this bond to ADP so
forming ATP and 3-phospho-o-glycerate, a reaction catalysed by the en-
zyme phosphoglycerate kinase (EC 2.7.2.3). At this point in the pathway,
two molecules of ATP and two of NADHz have been produced for each
glucose molecule entering the sequence. Thus the expenditure of ATP
earlier in the process is balanced.
3-Phospho-o-glycerate is converted to 2-phospho-o-glycerate by the
enzyme phosphoglyceromutase (EC 2.7.5.3). Enolase (EC 4.2.1.11) then
mediates the removal of water from 2-phospho-o-glycerate to give
phosphoenolpyruvic acid. The energy-rich bond formed in phospho-
enolpyruvate (,1G'o = -61.9 kJ mor!) is then used to phosphorylate ADP
to yield ATP and pyruvic acid. At this point, for each molecule of glucose
entering the pathway a net gain of two molecules of ATP is achieved.
Also, two molecules of NADHz have been formed. The supply of NAD+
is limited and therefore the continued operation of the pathway requires
the oxidation of the NADHz produced to reform NAD+. Under the condi-
tions operating in a brewery fermentation, the yeast cell decarboxylates
pyruvic acid to acetaldehyde and carbon dioxide using the enzyme pyr-
uvate decarboxylase (EC 4.1.1.1). The carbon dioxide formed is that which
is produced during fermentation and which naturally carbonates beer.
The acetaldehyde produced in the reaction is subsequently reduced to
ethanol, which together with carbon dioxide and yeast represent the major
products of fermentation. The enzyme mediating this reaction is alcohol
dehydrogenase (EC 1.1.1.1) with the participation of NADHz as cofactor
so that the NADHz formed during the oxidation of glucose is finally
oxidized to NAD+, thus permitting the continued operation of the EMP
pathway:

NADHz
By using acetaldehyde as the terminal electron acceptor the cell ensures
the continued operation of the glycolytic pathway and therefore the
continued formation of ATP for use in biosynthetic reactions.
The standard free energy change for fermentation is:
C6H 120 6 - 2COz + 2CaCHzOH ,1G'0 = -234 kJ mor!
The EMP pathway yields a net gain of two ATP molecules per molecule
of glucose, which under standard conditions is equivalent to 61 kJ mor!
free energy and the efficiency of fermentation is therefore 26%. The 74%
of energy not retained by the cell is largely dissipated as heat. Thus it is
mandatory to apply some degree of cooling to all but the smallest
20 The biochemistry and physiology of yeast growth
fermenters (from which heat may be readily lost to the surroundings) in
order to ensure that the temperature of fermentation remains under
control. The fermentation rate is not constant during the time course of
fermentation and the greatest activity occurs during the first 24-36 h
period. For a typical fermentation the cooling required is of the order of
3 x 105 kJ h-1•
The glycolytic sequence is an example of an amphibolic pathway and
the various intermediates are used by the cell in biosynthetic reactions. In
particular, phosphorylated glucose is a precursor for the synthesis of cell
wall carbohydrate polymers and of storage polysaccharide; triose phos-
phates are used in fat synthesis and phosphoenolpyruvate and pyruvate
are precursors of several amino acids. However, growing cells require
many more intermediates for biosynthetic reactions than can be supplied
from the EMP pathway. In respiring yeast cells, i.e. those which in contrast
to yeast in a brewery fermentation use molecular oxygen as a hydrogen
acceptor and completely oxidize glucose, the Krebs cycle (tricarboxylic acid
cycle) is used to supply additional substrates for biosynthesis. Under the
anaerobic conditions of brewery fermentation, the levels of the enzymes of
the Krebs cycle are greatly lowered and the extent (if any) to which the cycle
operates under these circumstances is unclear. The question arises as to
how the cell synthesizes essential intermediates such as succinic acid,
2-oxoglutaric acid and oxaloacetic acid. Two mechanisms have been pro-
posed; the first envisages a limited (although complete) operation of the
citric acid cycle with the formation of succinate, fumarate, malate and
oxaloacetate by an oxidative process (Oura, 1977), whereas the second
involves the synthesis of additional enzymes leading to the formation of
succinate by a reductive pathway (Sols, Gancedo and De la Fuente, 1971).
In this second mechanism it is proposed that the Krebs cycle is inoperative
because of the lack of the key enzyme 2-oxoglutarate dehydrogenase
(EC 1.2.4.2).
In both mechanisms, pyruvate is converted to acetyl coenzyme A (acetyl
CoA) in a reaction involving the multienzyme pyruvate dehydrogenase
complex. This complex contains pyruvate dehydrogenase (lipoamide;
EC 1.2.4.1), which acts on pyruvic acid liberating CO2 and forming acetyl
dihydrolipoamide. Then the enzyme dihydrolipoamide acetyl transferase
(EC 2.3.1.12), also part of the complex, acetylates coenzyme A (CoA), in a
reaction involving thiamin diphosphate, to yield acetyl CoA and leave
dihydrolipoamide. A third member of the complex, a flavoprotein,
dihydrolipoamide reductase (NAD), mediates the reduction of NAD+ to
form NADH2 and lipoamide. The overall reaction is therefore:
CH3CO.COOH + CoASH + NAD+ ~
CO2 + CH3COSCoA + NADH2
Also both mechanisms include the formation of oxaloacetate by the ATP-
dependent carboxylation of pyruvic acid. This single reaction can be seen
 Yeast metabolism 21

Q-Glucose
I
I
I
I
t
, - - - - - - PYRUVATE

CO z
CD
ATP

NADH z
ADP
+
Pi ACETYL CoA

~ ~AD~~II~ONA::~~TE CITRATE

NAD_-lf--NAD+ ® \
®

®F
MALATE

C!l 1l{
HzO / \
@ ISOCITRATE

FUMARATE NADP+

~~ADHZ
\ tP'3,
\'."! NADPH2
FADH2J", FAD ,
2-0XOGLUTARATE

/ >. . .
(8) FAD '
SUCCINATE (j) ® /\~ \
7 ,..::;/ (02
(oASH A SU((INYl (oA .....
f.. I (oASH
BIOSYNTHESIS GTP GOP BIOSYNTHESIS
EX(RETION EX(RETION

Fig.2.2 Formation of metabolic intermediates from pyruvate by fermenting yeast.

See text for details. Continuous arrows indicate the proposed reductive pathway;
broken arrows show the proposed oxidative pathway. 1, Pyruvate dehydrogenase
complex; 2, pyruvate carboxylase (EC 6.4.1.1); 3, citrate (si)-synthase (EC 4.1.3.7);
4, aconitate hydratase (EC 4.2.1.3); 5, isocitrate dehydrogenase (NADP+)
(EC 1.1.1.42); 6, oxoglutarate dehydrogenase (EC 1.2.4.2); 7, succinyl-CoA syn-
thetase (GDP-forming) (EC 6.2.1.4); 8, succinate dehydrogenase (EC 1.3.99.1); 9,
fumarate hydratase (EC 4.2.1.2); 10, 11, malate dehydrogenase (EC 1.1.1.37); 12,
fumarate hydratase (EC 4.2.1.2); 13, fumarate reductase (induced under anaerobic
conditions).
22 The biochemistry and physiology of yeast growth

HO~H
~~, Glucose 6-phosphate

OH
CD V- NADP +

t--NADPH z
CHzOP

~O,-o ~-gluconolactone 6-phosphate

H~

'2 r
COOH
OH

HzO

I
H-C-OH CHzOH
I I
HO-C-H C=O
I I
6-phospho-D- H-C-OH H-C-OH + COz
gluconate I I
H-C-OH H-C-OH
I I
CHzOP CHzOP
ribulose S·phosphate

~ ~:~
~HzOH ~HO
CHO c=o H-C-OH
I I I
H-C-OH HO-C-H H-C-OH
I I I
H-C-OH

H:YE:~: ~';~'::~;"'Ph.t.
I
CHzOP
erythrose
4-phosphate S-phosphate TPP ~

~HzOH CHzOH
I
C=O CHO C=O
I I I
HO-C-H H-C-OH HO-C-H
I I I
H-C-OH CH OP H-( -OH
I I
H-C-OH glyceraldehyde H-C -OH
I I
3·phosphate H-C-OH
CHzOP I
(HzOP
sedoheptulose 7-phosphate

Fig. 2_3 Hexose monophosphate pathway. 1, Glucose-6-phosphate dehydrogen-

ase (EC 1.1.1.49); 2, 6-phosphogluconolactonase (EC 3.1.1.31); 3, phosphogluconate
dehydrogenase (decarboxylating) (EC 1.1.1.44); 4, ribosephosphate isomerase (EC
5.3.1.6); 5, ribulosephosphate 3-epimerase (EC 5.1.3.1); 6, transketolase (EC 2.2.1.1);
7, transaldolase (EC 2.2.1.2).
 Yeast metabolism 23
as an anaplerotic reaction for the operation of the citric acid cycle, replacing
intermediates withdrawn for biosynthesis. The enzyme controlling this
anaplerotic reaction is pyruvate carboxylase (EC 6.4.1.1) a zinc-containing
protein. The alternative mechanisms are shown in Fig. 2.2, where the
reductive pathway is depicted by the solid arrows and the oxidative
pathway by the broken arrows.
The pathways of carbohydrate metabolism so far discussed do not
provide for the synthesis of pentose sugars which are essential to the
biosynthesis of nucleic acid precursors as well as many enzyme cofactors,
e.g. NAD+ and NADP+. Pentose sugars may be synthesized from G-6-P by
an oxidative pathway in which the enzymes mediating the oxidative steps
use NADP+ as cofactor. The pathway is depicted in Fig. 2.3 and is called the
hexose monophosphate pathway (HMP) or the pentose phosphate path-
way. The first reaction is the oxidation of G-6-P by the enzyme glucose-6-
phosphate dehydrogenase (EC 1.1.1.49). The product of this oxidation,
D-gluconolactone-6-phosphate, is acted upon by the enzyme
6-phosphogluconolactonase (EC 3.1.1.31) with the consequent formation of
6-phospho-D-gluconate; this product becomes the substrate for a second
oxidative step with the NADP-enzyme phosphogluconate dehydrogenase
(decarboxylating) (EC 1.1.1.44). The end-products of this reaction are CO2
and the pentose phosphate ribulose-5-phosphate. From ribulose-5-
phosphate the enzymes ribulosephosphate 3-epimerase (EC 5.1.3.1) and
ribosephosphate isomerase (EC 5.3.1.6) produce xylulose-5-phosphate and
ribose-5-phosphate respectively. The enzyme transketolase (EC 2.2.1.1),
with thiamin diphosphate as cofactor, catalyses the transfer of the two-
carbon keto-fragment of xylulose-5-phosphate to the C1 atom of ribose-5-
phosphate, thereby forming sedoheptulose-7-phosphate and
glyceraldehyde-3-phosphate. The enzyme transaldolase (EC 2.2.1.2) cata-
lyses the transfer of the C1, C2 and C3 atoms of a keto-sugar to the C1 atom
of an aldo sugar. Thus from sedoheptulose-7-phosphate and
glyceraldehyde-3-phosphate, fructose-6-phosphate and erythrose-4-
phosphate are formed.
The HMP pathway may be considered to oxidize the glucose molecule
completely by the overall reaction:
G-6-P + 12NADP+ ~ 6C02 + 12NADPH + 12H+ + Pi + 6H20
However, it is probably of greatest significance to the cell as a means of
generating from glucose reduced NADP+ for anabolic reactions and as a
source of pentoses which yeast cannot take up from wort.
In S. cerevisiae during fermentation, the level of glucose-6-phosphate
dehydrogenase is very low (Horecker, 1968) and it is probable that most of
the yeast's requirement for pentose sugar is met by the action of
transketolase on fructose-6-phosphate and glyceraldehyde-3-phosphate
produced by the EMP pathway (see Fig. 2.1).
24 The biochemistry and physiology of yeast growth
(b) Synthesis of carbohydrates
The disaccharide trehalose and the polysaccharide glycogen are both syn-
thesized during fermentation and act as 'food' reserves for the cell. In
addition cells synthesize the mannan and glucan components of the cell
wall.
Both trehalose and glycogen synthesis begin with the formation of
uridine diphosphate glucose (UDPG) catalysed by the enzyme glucose-1-
phosphate uridylyltransferase (Ee 2.7.7.9). Trehalose phosphate is syn-
thesized from UDPG and G-6-P by the enzyme a,a-trehalose-phosphate
synthase (UDP-forming; Ee 2.4.1.15). Trehalose is formed by the action of
trehalose-phosphatase (Ee 3.1.3.12).
Glycogen synthase-D-phosphatase catalyses the sequential addition of
the glucose residue of UDPG to a polysaccharide acceptor. The glucose
units in the polysaccharide are linked 0.(1 ~ 4) in linear chains. The natural
glycogen molecule is a branched structure and short 0.(1 ~ 4) bridging
chains link longer chains together via 0.(1 ~ 6) bonds.
During fermentation, glycogen is accumulated by brewing yeast and
accounts for the consumption of the equivalent of some 0.25% maltose.
When yeast is inoculated into fresh wort, the glycogen is metabolized to
yield energy (and presumably substrates) for the synthesis of sterols essen-
tial to yeast growth (Quain and Tubb, 1982). It is therefore important for
yeast to contain adequate glycogen for this purpose and under normal
circumstances sufficient is present. It has been shown that prolonged
storage of yeast can lead to the depletion of the glycogen reserve without
the concomitant formation of sterols. Under such circumstances the subse-
quent fermentations may be erratic both in terms of rate and final specific
gravity.

(c) Regulation
Knowledge of the control of amphibolic pathways in S. cerevisiae is mainly
derived from studies of the effects of the concentrations of glucose and
oxygen on cell growth and the levels of enzymes of the pathways. Unam-
biguous data are obtained therefore only by using media of defined com-
position and under conditions of full aeration (aerobiosis) or complete lack
of oxygen (anaerobiosis). These conditions are not typical of those obtain-
ing in a brewery fermentation where the medium is complex (and glucose
is not the main fermentable carbohydrate) and is air-saturated at the start
but anaerobic after say the first 12 h (or less) have elapsed. Furthermore,
scientific investigation of metabolic control is usually carried out with
actively growing cells, whereas growth in fermentation is somewhat
restricted.
S. cerevisiae belongs to the so-called glucose-sensitive yeast types. In
these strains respiration is repressed in the presence of a small (::; 0.4%)
 Yeast metabolism 25

concentration of free glucose in the medium. This repression applies

irrespective of the presence or absence of molecular oxygen. The effect of
repressive levels of glucose is to produce changes in enzymic composition
(particularly levels of TCA cycle enzymes) and influence the structure of
the mitochondria of the cell (particularly levels of components of the
respiratory chain). This effect of glucose is referred to as the Crabtree effect,
glucose repression or catabolite repression. The consequence of catabolite
repression in a brewery fermentation is not clear although since brewers'
wort typically contains approximately 1% glucose it is assumed that even
in air- or oxygen-saturated worts the yeast cells are repressed. When the
free glucose has been consumed (probably in the first 16 h), however,
exposure of cells to oxygen may result in derepression since neither
maltose nor maltotriose exhibits a repressive action on respiration. Al-
though exposure to oxygen is avoided during the later stages of fermen-
tation, top-cropping yeast strains are obviously exposed, as is yeast stored
in suspension in cold stores prior to use. The precise mechanism of
catabolite repression is unknown but may be related to intracellular levels
of cyclic adenosine monophosphate (cAMP) (Fiechter, Fuhrmann and
Kappeli, 1981).
In yeast, regulation is also expressed through the inactivation (by enzy-
mic modification or degradation) of enzymes. Such processes are also
initiated in response to free glucose in the medium and are termed
catabolite inactivation (Holzer, 1976; Hagele, Neff and Mecke, 1978).
In the absence of repressing levels of glucose and in the presence of
molecular oxygen, yeast respires the carbohydrate. In this case glucose is
completely oxidized to CO2 and H 20. This is achieved through the opera-
tion of the glycolytic pathway and tricarboxylic acid cycle. The reduced
NADH produced is oxidized to NAD+ by the electron transfer chain using
molecular oxygen as hydrogen acceptor (thus forming H 20) and in the
process synthesizing ATP.1t was observed by Pasteur (1867) that the uptake
of glucose was lower in respiring cells than in fermenting ones. The effect
has been named the Pasteur effect. It is now recognized that this effect is
particularly marked when resting rather than actively growing cells are
analysed (Fiechter, Fuhrmann and Kappeli, 1981). Nevertheless, the
Pasteur effect points to the operation of some regulatory system on the EMP
pathway. The accepted model of this regulatory system is that of Sols (Sols,
Gancedo and De la Fuente, 1971) which recognizes a central role for the
enzyme 6-phosphofructokinase. In this model ATP allosterically inhibits
the enzyme and AMP activates it. Thus in conditions of high energy charge
(respiration) the flux of glucose through the EMP pathway is lowered. In
respiration, ATP synthesis also depletes the intracellular reserve of inor-
ganic phosphate which in turn limits the operation of the EMP pathway
(Lynen, 1963) and lowers the glucose flux. In both cases the overall effect
is for less glucose to enter the pathway than during fermentation.
Further information on carbohydrate metabolism, particularly that of
26 The biochemistry and physiology of yeast growth
aerobic cells, and its regulation may be found in Briggs et al. (1982) and Sols,
Gancedo and De la Fuente (1971).

2.3.3 Nitrogen metabolism

Yeast cells preferentially use the amino acids present in brewers' wort as a
source of nitrogen. The nitrogen assimilated by cells is used to synthesize
amino acids which in tum are used to synthesize proteins. For details of the
protein-synthesizing system Stryer (1995) may be consulted.
Careful analysis of the uptake of isotopically labelled amino acids by
brewers' yeast showed that negligible assimilation of intact amino acids
occurred (Jones, Pragnell and Pierce, 1969). When an amino acid enters
the yeast cell, a transaminase system removes the amino group and the
remaining 'carbon skeleton' is metabolized. It appears that the main
transamination system operating in yeast employs 2-oxoglutarate as ac-
ceptor. The glutamate formed may then be used as an amino donor to
synthesize other amino acids from carbon skeletons formed by anabolic
pathways. The transaminase (aminotransferase) reaction is readily revers-
ible and the enzyme employs pyridoxal phosphate as cofactor. The reac-
tion is shown in Fig. 2.4; the aminated cofactor (pyridoxamine phosphate)
is used by the enzyme as a donor of amino groups. The oxoacid pool
generated by the action of transaminases and anabolic reactions is a
precursor of aldehydes and higher alcohols which contribute to beer
flavour.
The assimilable nitrogen content of a brewers' wort is not specifically
measured. Reliance is placed upon measurement of free amino nitrogen
(FAN) with the assumption that amino acids make a greater contribution
to this analysis than polypeptides. Typical worts contain 100-140 mg r l
FAN and levels below these are found to be inadequate to support fermen-
tation. Fermentable extracts made from other cereals such as sorghum
(Bajomo and Young, 1992) are fermented by brewers' yeast when levels of
FAN of 40 mg r l are present. Similarly wheat extracts containing FAN of
54-58 mg rl fermented successfully (Thomas and Ingledew, 1990). Clearly
more research is needed to evaluate the assimilable nitrogen content and
nutritional status of malt wort.

2.3.4 Lipid metabolism

Yeast lipid comprises the triacylglycerols, phospholipids and sterols. The
triacylglycerols are triesters of long-chain fatty acids and glycerol. Yeast
lipid contains predominantly fatty acids 16 or 18 carbon atoms long al-
though chain lengths of 8-24 are found. The fatty acids may be saturated
or unsaturated (with one or two carbon-to-carbon double bonds). Thus C6
(palmitic), C18 (stearic) are the principal saturated fatty acids and the
unsaturated ones are the monoenoic C6:I (palmitoleic), C8:I (oleic) and the
 Yeast metabolism 27

WORT CARBOHYDRATE

INTRACELLULAR OXOACIO POOL

----------------------------------~

~~ ~~
(Hz CH z

\fl,'
H0t.)CHzOP 70XOGLUTAilATE HOnCHZOP

HldC "N. HlC "N I \

pyrl oxamlne phosphate pyradoxamlne phosphate

TRA SAMIN

H06" OP
H , H

/\ / HD)::",OP
H,C .... N I GLUTAMATE HJ(kN~
pyr,doxal phosphate
1
pyridoxal phosphate

AMINO ACIDS - - - -... PROTEINS

YEAST CElL
WORT AMINO ACIDS

Fig.2.4 Assimilation of wort amino acids by transamination.

dienoic C8:2 (linoleic). Phospholipids are substituted diacylglycero-

phosphates with the most common substituents being choline, ethano-
lamine, serine or inositol. The predominant sterols are ergosterol and
zymosterol; other sterols and sterol esters are also found. All three classes
of lipid are important components of the yeast cell membrane. Since the
cell membrane controls the entry of nutrients into, and the excretion of
metabolites from, the cell its functions are essential to the growth of the cell.
Saturated fatty acids are synthesized from acetyl CoA derived from
fermentable carbohydrate by the enzymes of the fatty acid synthetase
complex. Unsaturated fatty acids are derived from their saturated counter-
parts by an enzymic reaction in which molecular oxygen acts as hydrogen
28 The biochemistry and physiology of yeast growth
acceptor. The mixed-function oxidase employed uses NADP+ as a cofactor
(Bloomfield and Bloch, 1960).
Sterols are also unsaturated molecules and their synthesis also involves
the participation of an NADP+-dependent mixed-function oxidase. The
requirement of yeast for molecular oxygen is greater for sterol synthesis
than for unsaturated fatty acid synthesis (David, 1974).
Failure to provide sufficient molecular oxygen to a fermentation me-
dium restricts yeast growth and viability since the cells cannot produce
unsaturated lipids for membrane biosynthesis. A secondary effect with ale
yeasts is to produce elevated levels of esters in finished beer. Insufficient
oxygenation (or aeration) of wort is not the only means of restricting the
oxygen availability to the yeast. Since the solubility of oxygen is inversely
related to the specific gravity of the wort, high-gravity worts may be unable
to contain sufficient oxygen (particularly if air rather than molecular oxy-
gen is the source) for yeast growth. Brewers therefore need to pay particular
attention to satisfying the requirement of their yeast for molecular oxygen.
This requirement is not fixed but varies with different yeast strains (Van
den Berg, 1978).

2.4 YEAST PROP AGA nON

2.4.1 Measurement of amount, viability and vitality of yeast

In brewing practice, measurement of the amount of viable yeast present is
made in order to control for example the amount of yeast added to a
fermenter (pitching rate) so that reproducible fermentations may be con-
ducted, or the amount of yeast remaining in suspension after fermentation
to 'condition' the beer.
Although the measurement of the dry weight of a suspension may be
used to obtain a direct estimation of the biomass present, this procedure is
generally too slow to be of value in practical brewing. Accordingly, less
direct but more rapid methods are generally employed. Examples of these
techniques are the measurement of packed cell volume following centrifu-
gation under precisely controlled conditions, determination of wet weight
after filtration and the measurement of the opacity or turbidity of a suspen-
sion. Alternatively, where pressed yeast or yeast slurry of consistent quality
is available, the weight or volume of such samples may be measured. The
accuracy and reliability of all these procedures is limited if a significant
amount of non-yeast matter, e.g. precipitated protein (trub), is present in
samples.
Cell number in suspension is most conveniently measured using an
electronic particle counter. This technique does not distinguish between
single cells and budding cells, chains of cells, aggregates or non-yeast
matter, which are all recorded as single counts. Treatment of samples with
 Yeast propagation 29

short bursts of ultrasound prior to analysis may be used to disrupt

aggregates and is of particular value when handling flocculent yeasts.
Cell counts may be obtained microscopically using a counting chamber
and the numbers of cells in chains or aggregates may be estimated. Further-
more, the analyst may readily distinguish between yeast cells and particu-
late matter which is often dissolved by the addition of one or two drops of
20-40% sodium hydroxide or potassium hydroxide solution to the sample
on a microscope slide. Counting cells in this way is considerably more
time-consuming than using a particle counter or one of the indirect meth-
ods. The length of time needed is greatly increased if reproducible and
statistically valid results are to be obtained (Meynell and Meynell, 1965).
The measurements of cell mass or number described cannot by
themselves distinguish between living and dead cells. It is usual therefore
for the brewer to combine measurement of amount of yeast with some
procedure for estimating viability.
The most direct procedure for measuring viability is to prepare a sus-
pension of yeast containing about 1000 cells mr! and spread 0.1-0.2 ml on
the surface of a nutritionally rich medium solidified by the inclusion of
agar. Alternatively, a larger volume of sample is mixed with molten me-
dium at 47-49°C and the mixture poured into a Petri dish. In each case,
isolated viable cells grow to produce colonies which are readily counted.
In general this technique is not favoured for use with brewing yeast cultures
sampled from the brewery since the estimated viability even when cor-
rected for the presence of budding cells, chains and aggregates is invariably
lower than that obtained with other methods (Richards, 1967).
Living cells contain ATP whereas this material is hydrolysed in dead
cells. Detection of ATP therefore forms the basis of a technique for measur-
ing viable cells and the procedure is claimed to be highly reproducible
(Miller et al., 1978) (see Chapter 8).
The most reliable procedure for obtaining accurate estimates of cell
viability is the slide culture technique (Gilliland, 1959; Institute of Brewing
Analysis Committee, 1971). In this method, a suitably diluted suspension
of yeast is mixed with just molten wort gelatin and applied to a microscope
slide or counting chamber. A coverslip is carefully lowered into position
and sealed with sterile paraffin wax. After incubation for 8-16 h in a
humidity chamber at 20°C, microscopic examination reveals the growth of
microcolonies from viable cells. A shorter incubation time and easier
visualization of microcolonies may be obtained by incorporating optical
brighteners into the medium (Harrison and Webb, 1979). These compounds
bind to the cell walls of the yeasts and fluoresce when illuminated with blue
light. The use of a microscope equipped with an epifluorescence illumina-
tion system is therefore required. Fluorescence microscopy may also be
used to estimate viability using the so-called fluorescein-diacetate tech-
nique. This compound is hydrolysed by esterases found in living but not
in dead cells and the fluorescein released inside the cells causes them to
30 The biochemistry and physiology of yeast growth
fluoresce when suitably illuminated. Various fluorescent vital stains (stains
which are absorbed by dead but not living cells) are available. These
procedures, referred to as direct epifluorescence techniques (DEFT), often
employ acridine dyes, e.g. acridine orange (Parkinnen, Oura and
Suomalainen, 1976). (See Chapter 8.)
The most commonly used techniques employ non-fluorescent vital
stains, and in the UK methylene blue in buffered solution (pH 5) is
favoured. Dead cells stain blue whereas living cells are colourless. The
reliability of the procedure is good only if population viabilities are in
excess of 85% (Gilliland, 1959; Institute of Brewing Analysis Committee,
1971). Although microscopic techniques are tedious, when the methylene
blue test is carried out using a counting chamber simultaneous estimates
of cell number and viability are obtained.
The reliability of vital staining techniques has been questioned by many
researchers. Yeast cultures after acid washing or longer than normal storage
often appear viable but fail to ferment satisfactorily. In an attempt to
overcome these problems, techniques aimed at measuring the physio-
logical state of the yeast have been developed. These techniques include
the rate of oxygen uptake (Kara, Daoud and Searle, 1987; Peddie et al., 1991)
and the rate of acidification of a medium after addition of D-glucose (Kara,
Simpson and Hammond, 1988). Each method gives rapid results (minutes).
These procedures are claimed to be more reliable than vital staining and
are much more rapid than slide culture techniques. They are also suited to
measurement in-line and hence offer great advantages for automated pro-
cess control. The two methods however do not agree when used to analyse
the same yeast cultures (Peddie et al., 1991).
It is generally agreed that in the early stages of fermentation the oxygen
consumed is used to produce unsaturated fatty acids and lipids essential
to yeast viability (and vitality). Recent research has been directed at oxy-
genating yeast prior to fermentation in attempts to ensure adequate unsat-
urated lipid levels and thus eliminate wort oxygenation as a possible source
of variability (Boulton, Jones and Hinchliffe, 1991; Devuyst et al., 1991).

2.4.2 Handling pure yeast cultures

Hansen (1896) devised techniques for the isolation of single cells of
brewing yeast. The separation of the different component strains of a
brewing yeast culture was therefore feasible and each component could
be analysed for its brewing properties. Because of the genetic stability of
brewing yeast strains (resulting from their lack of a sexual cycle) a single
isolated cell will grow to produce (excepting any chance mutation) a
population of genetically identical cells, i.e. a clone. Such clones retain
their brewing characteristics; thus, once selected, a fermentation strain
will, in principle, produce consistent fermentations. The brewer, then, may
use a pure culture if there is a means of preserving or maintaining it in a
 Yeast propagation 31
laboratory and a means of culturing it up to the large amount needed to
pitch a brewery fermentation.
In breweries employing pure culture, the strain may be isolated, selected
and kept at each brewery or at a central laboratory . When needed the yeast
may be grown up and transported in slurry or pressed form in refrigerated
containers or propagated from a laboratory culture in a yeast culture
(propagation) plant in the brewery. Some companies rely on commercial
laboratories to isolate and keep their yeast.
Various procedures are used to isolate pure cultures from the pitching
yeast used. These include culturing from a single cell, culturing from a
single colony (isolated on a medium containing agar or gelatin) and cultur-
ing from mixtures of isolated cells or colonies. Some companies isolate two
or more strains which they employ in mixture to conduct fermentations.
All yeast cultures are susceptible to chance mutation; in addition, vari-
ation in mixed cultures may arise by changes in the proportions of the
strains present. Such changes may be induced by use of different raw
materials in the production of wort, a change in fermentation conditions or
even a change in the type of fermentation vessel. On the other hand, it may
be argued that a mixed culture is better able to withstand environmental
changes because its increased genetic diversity enables it to adapt more
readily to changed circumstances.
The successful maintenance of pure cultures requires that the isolates are
given the minimum exposure to conditions which induce mutation, prop-
agate mutant cells or encourage sporulation and must of course keep the
cells in a highly viable condition. It is usual practice therefore to keep
cultures in glass containers in the dark to prevent exposure to the muta-
genic effects of ultraviolet light. Yeasts may be maintained on wort agar
slopes (slants) at ambient temperature (to reduce the possibility of sporu-
lation) or at 4°C, with or without an overlay of sterile mineral oil to maintain
anaerobic conditions and restrict the growth of the cells. Many laboratories
hold stock cultures in liquid medium such as wort or in chemically defined
medium (e.g. YM medium; Haynes, Wickerham and Hesseltine, 1955) at
4°C. Whether sloped or liquid cultures are used, cell growth occurs and
eventually the nutrients are consumed and starvation and cell death ensue.
It is thus necessary to subculture strains regularly on to fresh slants or into
fresh liquid medium every 3--6 months, to preserve the viability of the yeast.
Subculture increases the chance of both mutation and contamination of the
culture.
It is clearly preferable to keep the yeast in a form which removes the
need for subculture, hence saving on labour and materials as well as better
maintaining the integrity of the culture. One means of achieving this is to
freeze-dry (lyophilize) the culture. This technique has not been generally
accepted mainly because some yeasts exhibit a great loss of viability
during the drying process. This phenomenon is however variable and
careful control of the conditions overcomes it to some extent (Barney and
32 The biochemistry and physiology of yeast growth
Helbert, 1976). There has also been some concern about the selection of
mutants by the lyophilization process (Kirsop, 1955; Wynants, 1962); yet
it has been reported that whereas routine subculture over a period of up
to 20 years caused marked changes in the brewing characteristics of ale
yeasts, no such changes were found when the same cultures were
preserved by freeze-drying (Kirsop, 1974).
An alternative to freeze-drying is to store yeast in suspension in glycerol
at the temperature of liquid nitrogen (-196°C). This technique retains the
viability of cultures and is very successful (Mills, 1941; Barney and Helbert,
1976; Russell and Stewart, 1981; Kirsop, 1984).
When needed, yeasts are transferred from the laboratory culture to a
small volume (approximately 10 ml) of liquid medium (wort or YM broth)
and incubated at ambient temperature or 25-30°C for 16-48 h. This culture
is then transferred to a larger volume, usually of sterile wort, then after
further incubation this in tum is inoculated into a yet larger volume. At the
end of this process the laboratory will have cultured up the yeast using
stringent aseptic techniques and sterile media to a volume of approximately
201. The culture from each stage is used as a 5-10% v Iv inoculum for the
succeeding stage. The final culture will often be made in a specially de-
signed stainless steel vessel compatible with making a good aseptic transfer
of the culture to the yeast culture plant. Usually, a propagator with a
working capacity some 5-10-fold the volume of the laboratory culture
would be used. In some breweries however the 10-20 I laboratory culture
is transferred to 4000-5000 I propagators (Van den Berg, 1978).

2.4.3 Operation of yeast culture plant

In modem breweries, where standards of general hygiene are high, re-use
of yeast may often represent the major source of contaminating micro-
organisms. Furthermore, the performance of a brewery yeast will often
decline with repeated use and the yeast is said to become 'weak'. The
propagation of yeast is therefore practised routinely. In most breweries
fresh yeast would be propagated every 8-10 generations (fermentation
cycles) or sooner if particular problems were being experienced. The pro-
duction of clean (uninfected), highly viable culture yeast is a major factor
ensuring consistent fermentation performance.
In order to restrict the access of undesirable contaminating micro-
organisms (which would be propagated along with the culture yeast)
propagation systems are designed to a high standard. Under ideal circum-
stances, the culture plant is placed in a room that is well separated from the
rest of the brewing plant. The room should be equipped with its own
filtered air supply and maintained at a slight positive pressure with respect
to the surrounding areas. Access to the plant should be restricted to a few
personnel and self-closing doors with disinfectant mats in the doorways
are advantageous. The vessel(s) used for propagation are closed and
 Yeast propagation 33

Steam or
- Inoculum

Steam in
(oOlan..;.t::.o-ut,....-..r,----------t,

,
(ulture out or
Wort in

Fig.2.5 A modem yeast propagation vessel of 49 hI (30 barrels) working volume.

C, in-place cleaning spray-ball; F, secondary air filter; C, sight glass; Cu, pressure
gauge; M, manway and dip-point; P, sample point; R, pressure and vacuum relief
valve, fitted with sterilizing filter; T, temperature probe; V, vent with shut-off
valve. (Courtesy of Shobwood Engineering Ltd, Burton-on-Trent.)

capable of being effectively cleaned either with a combined detergent

sanitizer or preferably with separate caustic detergent cleaner followed by
a sterilant or live steam. The wort used is either 'pasteurized' using a plate
heat exchanger and passed into the sterile vessel or boiled in the vessel for
30 min or so to effect 'sterilization'; the necessary heating is obtained using
pressurized steam in jackets around the vessel or by injecting live steam
into the vessel. After venting air and steam, the vessel may be sealed so that
sterilization may occur under pressure. After the sterilizing period, the
vessel and contents are cooled and during this process air is admitted
through a sterilizing filter. Most propagation vessels are fitted with cooling
jackets and a system for admitting sterile air or oxygen. Agitators are not
usually used, the aeration process and evolution of CO2 during active
fermentation being sufficient to maintain the yeast in suspension. A dia-
gram of a typical propagation vessel is shown in Fig. 2.5: in this vessel the
34 The biochemistry and physiology of yeast growth
jacket serves the dual purpose of being filled with steam during sterilization
and coolant during the propagation phase.
In a propagation system more than one vessel is often employed. Two
vessels of different sizes, e.g. 1 hl and 20 hl, would be found and the product
from the larger vessel used to recharge the smaller and pitch the first
fermenter, containing (in this case) 200-400 hl wort. By retaining a propor-
tion of propagated yeast in the smaller vessel the brewer may propagate
fresh yeast from there without recourse to the laboratory culture every time.
The coolant circulation around the small vessel would be maintained at
such a rate and temperature as to keep the yeast below ambient tempera-
ture. In practice, recourse to the laboratory culture may be made only at
intervals of 6-12 months. The sizes of the propagation vessels may be
determined by the size of the first fermentor and the desire to keep the
inoculum size to 5-10% of the volume of the fermentor.
In propagation the brewer wishes to obtain the maximum amount of
yeast possible consistent with the flavour of the beer produced being not
too far removed from the flavour of the product of a normal fermentation.
In most breweries the beer from the first fermentation following propaga-
tion would in any event be blended with greater volumes of 'normal' beer.
Accordingly, propagation is carried out using worts of greater specific
gravity than those used in the typical fermentation, at higher temperatures
and with the use of intermittent aeration throughout the process. The
concentration of yeast produced would be of the order of 5 x 108 cells mr!
at a viability of > 98%. The time of propagation would be of the order of
24-48 h. Yeast growth in a propagator may be described by the standard
equations for microbial growth in a batch culture (Briggs et al., 1982). A
typical growth curve for yeast growing under conditions similar to those
obtaining in a yeast propagator is shown in Fig. 2.6. In practice the lag phase
would be almost non-existent (since the cells are fully adapted before
inoculation) and the culture would be transferred prior to the onset of
stationary phase.

2.5 BREWERY FERMENTA nON

Systems of fermentation have gradually evolved around the type of brew-

ing yeast used. It is usual to distinguish between two types of brewing yeast
(and hence fermentation system): thus brewers are said to employ top
yeasts (top fermentation) or bottom yeasts (bottom fermentation). The
distinction arises from the practice of removing the yeast from the top of
the fermentor by a skimming process or from the bottom of the vessel. It is
current practice to classify all brewer's yeast as S. cerevisiae (Barnett, Payne
and Yarrow, 1983), although in brewing practice ale yeasts are referred to
as S. cerevisiae whereas lager strains are called S. carlsbergensis (or S. uvarum)
(Lodder, 1970). It would seem useful from a practical standpoint to retain
 Brewery fermentation 35

I A
8
144---- ------.,.~I...- - D ----+I E I

8·2
A lClg phClse of growth
8·0 8 PhClse of ClculerClting growth
( PhClse of exponentiClI growth
D PhCl5e of decelermting growth
E StationClr, phClse of growth

-...
e
I II e

-....,
"ii 7-0 1.107
....,
-
III

a
a
.D
e ....,
~ I'IIClX • 0·26 hr-' .D
CI'
e
::2
to .0·62hr 1.106
.3 6·0 Z

5·0
5 10 15 20 25 30 35 40
Time (hI

Fig. 2.6 Batch growth curve for brewing yeast culture in shake flasks at 20°C.

a distinction between the two types of brewing yeast, since they are used
to produce beers under quite different conditions of fermentation. In addi-
tion the two types of organism differ in characteristics not generally in-
cluded as taxonomic criteria, e.g. serotype (Campbell and Brudzynski,
1966) and maximum temperature for growth (Walsh and Martin, 1977).
More recent attempts to classify brewing yeasts are based on relevant
brewing characteristics (Bryant and Cowan, 1979; Cowan and Bryant,
1981), and consequently are of more value to the brewer.
Top yeasts necessarily tend to rise to the top of the fermenter and form
a head, whereas bottom yeasts do not form a substantial head but tend to
be sedimentary. Use of the different systems and yeasts is also typical of
brewing in different geographical areas. Bottom yeasts and bottom fermen-
tation systems were traditionally used to produce so-called lager beers in
the brewing centres of mainland Europe whilst the top fermentation system
is typical of the UK and was traditionally used to make the so-called ales.
The production of traditional ale employed open square fermenters or one
of several quite distinct types of fermentation system, e.g. Yorkshire square,
Burton union, and premium quality beers are still produced in these
systems. The modem trend in fermentation, however, has been to replace
36 The biochemistry and physiology of yeast growth
the open fermentation vessels (usually used where yeast is removed from
the top of the vessel) with closed vessels. Closed (or covered) vessels are
obviously more hygienic in operation since the fermentation is not exposed
to the atmosphere and furthermore it is a relatively simple matter to arrange
for them to be equipped with automatic cleaning-in-place (CIP) systems.
Modern practice has also led to an increase in size of fermentation vessel
and the siting of well-lagged fermenters in the open, thereby saving much
on expensive building costs involved in fabricating a fermentation room
(Briggs et al., 1982).
Many different designs of modern batch fermenter are used; Fig. 2.7 is a
diagram of a typical cylindroconical vessel used by many UK breweries for
the production of both lagers and ales. These vessels were originally
intended to be dual purpose and used for fermentation and conditioning
(Nathan, 1930; Shardlow, 1972). However, in practice they are mainly
employed for fermentation only and additional tanks are used for the
conditioning process (Maule, 1977). Cylindroconical fermenters range in
size from 100 to 4800 hi. Each vessel has an angled cone at its base and the
most common internal angle encountered is 70°. Very large vessels tend to
have much shallower slopes and angles of 105°. The cone allows for the
yeast to compact at the end of fermentation and beer may be separated
relatively cleanly from the sediment. Experience has shown that even top
yeasts may be induced to sediment in this type of vessel although, where
difficulty is experienced, antifoam is used to collapse the yeast head.
Alternatively a sedimentary variant may be isolated from the yeast which
routinely settles in the cone (Shardlow and Thompson, 1971). In order to
allow for foam or head formation during fermentation, vessels are usually
sized at 25-30% in excess of their working volume; therefore antifoam may
be employed to increase the effective working volume of the fermenter
(Button and Wren, 1972).
In spite of their success, cylindroconical vessels have not been used by
the traditional ale brewer to produce his premium quality beer. This results
from the fact that the CO2 content of beer produced in deep fermenters is
in excess of that required in a traditional ale and there is no doubt that beer
produced in these vessels differs in flavour from that made in traditional
fermentation systems. Nevertheless, the traditional brewer has used
cylindroconical fermenters to produce lager beers most successfully.
Attemperation of fermentation is achieved by circulating coolant
through cooling jackets (Maule, 1976). In the vessel shown in Fig. 2.7 three
jackets are used, although many vessels do not have a cone jacket and
may only have a single wall jacket two-thirds of the way up the cylindrical
part of the fermentor. Cone cooling is generally accepted as advantageous
when yeast for pitching is to be stored in the vessel; if a separate yeast-
collecting vessel is used then cone cooling is not required. The presence
of sedimented yeast effectively insulates the cone and restricts transfer of
heat to and from the fermenter contents. If the vessel is to be used for
 Brewery fermentation 37

1""\
t.,l VE
-S P
~
~
( --------
/Fl
i)
,
F

coolANT
oUT
J

302m 11 '43m

~ -SP /TP
",-S(

COOL ANT
J t
IN F
70·

'H
-S
YEAST IN- WORT IN
- YEAST OUT
£lEER OU1

Fig.2.7 Modem cylindroconical brewery fermentation vessel. C, in-place cleaning

spray-ball; F, floors; FL, flange; J, cooling coils ('limpets'); L, inspection lamp; M,
manway (detachable cone); P, pressure relief valve; S, sight glasses; SC, supporting
collar; SP, sample point; TP, temperature probe; V; vacuum relief valve; VE, vent
and detergent entry. (Courtesy of Gordon Smith of Bristol Ltd.)
38 The biochemistry and physiology of yeast growth
conditioning then cone-cooling is also necessary. The positioning of cool-
ing jackets and of temperature probes within the vessel are subjects of
much research (Maule, 1976).
Cylindroconical fermentors are fitted with pressure and vacuum release
valves. These relieve pressure during filling and vacuum during emptying,
or cleaning with hot alkali which absorbs CO2, thereby lowering the pres-
sure which is further reduced as the vessel cools. The outlet from the vessel
is from the apex of the cone but some fermenters have an additional outlet
for beer at the junction of the cone and vertical side. Temperature probes
and CIP heads are the only intrusions to the inside of the vessel, which is
fabricated throughout in stainless steel and is therefore easy to clean. In
some vessels, lines for admitting COz for rousing and purging and air for
rousing and aeration are fitted at the base of the cone. These features make
the vessels suitable for use as yeast propagators.
Large fermenters have the capacity to contain more than one production
volume (brew length) of wort. Such vessels are filled sequentially and the
brewer may choose to aerate and pitch any or all lengths. When the first
length only is pitched, the pitching rate must be increased to account for
the subsequent lengths. However, as the yeast grows in the first length the
amount needed is less than a pro rata increase based on the number of
lengths in the vessel. The choice of aeration and pitching regimes influences
to some extent the flavour of the final product.
A vessel taking a single brew length and used to produce ale would be
filled with aerated wort, and yeast slurry would be pumped in-line with
the wort. The temperature of the wort would typically be 15-18°C and the
yeast concentration 2 x 107 cells mrl (corresponding to a pitching rate of
3 kg pressed yeast of 25% solids hl-I wort). As fermentation proceeds the
temperature will rise and at a given value (top temperature) cooling will
be applied to keep the temperature steady. A typical top temperature
would lie in the range 18-25°C. The modern trend is to use higher fermen-
tation temperatures, thereby shortening the fermentation time, although
this practice does affect beer flavour. Some brewers claim that flavour
changes may be minimized to some extent by fermenting under pressure
(5-15 psi).
Fermentation in cylindroconical vessels is faster than in traditional
fermenters mainly because the rapid evolution of CO2 from the base of the
cone, coupled with attemperation of the vessel walls, results in a very
thorough mixing of wort and yeast (Ladenburg, 1968; Maule, 1976). A
typical fermentation would take 40-48 h and as the fermentation rate
declines, heat output falls and the cooling is increased to lower the temper-
ature of the beer and encourage the yeast to settle out. The settling period
may take 2-3 days and in some operations freshly fermented beer (green
beer) is centrifuged to remove yeast and passed to a separate vessel, thus
releasing the fermenter for cleaning and reuse. The fermentation is
monitored by taking samples for measuring the specific gravity and is
 Brewery fermentation 39

controlled by varying the rate of cooling (reducing the rate allows the
temperature to rise and accelerates the fermentation, and vice versa). A
typical fermentation profile is shown in Fig. 2.8.
Fermentations to produce lager beers are conducted in a similar manner
to those for ales except the temperatures used are lower and consequently
fermentation times are longer. Typical values would be pitch at 12°C, top
temperature 15°C and fermentation time 7 days.
One of the greatest operational advantages of a cylindroconical
fermenter over traditional vessels is that precise control of fermentation
profile may be achieved by controlling temperature (Hoggan, 1977). This
flexibility has stimulated research workers to present computer simula-
tions of the fermentation process and develop programs to enable auto-
mated fermentation to be achieved (Luckiewicz, 1978; Ruocco, Coe and
Hahn,1980).
Continuous systems of fermentation have also been used (Coutts, 1967;
Ault et ai., 1969; Bishop, 1970; Seddon, 1975) but in the UK have been
abandoned in favour of modern batch methods.
In contrast to the modern fermentation, traditional ale systems would
employ a pitching temperature of 14°C, a top temperature of 17°C and take
5-7 days, and traditional German lager systems would pitch at 7°C, use a
top temperature of 9°C and take 14-21 days.
In most ale fermentation systems, the yeast grows to produce a 5-8-fold
increase. In lager fermentations the overall increase in yeast is generally

I' (ooling .,
1-040 21
Cells/ml
20
6.10 7
1-030
5.107
S.G.
1-020 4.101

3.101
·1·010
2.10 1

1.10 7

TIME (HOURS)

Fig.2.8 Typical fermentation profile for modern batch fermentation to produce

ale. SG, specific gravity; T, temperature.
40 The biochemistry and physiology of yeast growth
nearer 5-fold. The amount of growth is dependent on the pitching rate,
temperature regime, oxygen content and specific gravity of the wort. Exces-
sive growth is undesirable in that wort sugars would be converted to undue
amounts of yeast rather than to alcohol. Yeast is cropped from the fermentor
maintained at 4°C, sometimes after washing with cold water to 'clean' it or
with acidic solutions (pH 2.5) to kill any contaminating bacteria, and used
to pitch subsequent fermentations. The surplus is sold to manufacturers of
yeast extract and to distilleries for use in their fermentation systems.

REFERENCES

Ault, RG., Hampton, AN., Newton, R and Roberts, RH. (1969) Journal of the
Institute of Brewing, 75, 260.
Bajomo, M.F. and Young, T.w. (1993)Journal of theInstitute of Brewing, 99, 153.
Barnett, J.A, Payne, RW. and Yarrow, D. (1983) Yeasts: Characteristics and
Identification, Cambridge University Press, Cambridge.
Barney, M.C and Helbert, J.R (1976) Journal of the American Society of Brewing
Chemists, 34, 61.
Bishop, L.R (1970) Journal of the Institute of Brewing, 76, 173.
Bloomfield, D. and Bloch, K. (1960) Journal of Biological Chemistry, 235, 337.
Boulton, CA, Jones, AR and Hinchliffe, E. (1991) Proceedings of the 23rd Congress
of the European Brewery Convention, Lisbon, IRL Press, Oxford, p. 385.
Briggs, D.E., Hough, J.S., Stevens, R and Young, T.w. (1982) Malting and Brewing
Science, Vol. 2, Chapman & Hall, London.
Bryant, TN. and Cowan, W.D. (1979) Journal of the Institute of Brewing, 85, 89.
Button, AH. and Wren, J.J. (1972) Journal of the Institute of Brewing, 78,443.
Campbell, I. and Brudzynski, A. (1966) Journal of the Institute of Brewing, 72, 556.
Coutts, M.W. (1967) International Brewer and Distiller, 1, 33.
Cowan, W.D. and Bryant, T.N. (1981) Journal of the Institute of Brewing, 87,45.
David, M.H. (1974) Journal of the Institute of Brewing, 80, 80.
Devuyst, R, Dyon, D., Ramos-Jeunehomme, C and Masschelein, C.A (1991)
Proceedings of the 23rd Congress of the European Brewery Convention, Lisbon, IRL
Press, Oxford, p. 377.
Fiechter, A, Fuhrmann, G.F. and Kappeli, O. (1981) Advances in Microbial Physiology,
22,123.
Gilliland, RB. (1959) Journal of the Institute of Brewing, 65,424.
Hagele, E., Neff, J. and Mecke, D. (1978) European Journal of Biochemistry, 83, 67.
Hansen, E.C (1896) Practical Studies in Fermentation. Cited in Laufer, S. and Schwarz,
R (1936) Yeast Fermentation and Pure Culture Systems, Schwarz Laboratories
Inc., New York.
Harris, G. and Thompson, C (1960) Journal of the Institute of Brewing, 66, 293.
Harrison, J. and Webb, T.J.B. (1979) Journal of the Institute of Brewing, 85, 231.
Haynes, W.C, Wickerham, L.J. and Hesseltine, CW. (1955) Applied Microbiology,
3,361.
Heredia, CF., Sols, A and De la Fuente, G. (1968) European Journal of Biochemistry,
5,321.
Hoggan, J. (1977) Journal of the Institute of Brewing, 83, 133.
Holzer, H. (1976) Trends in Biochemical Science, 1, 178.
Horecker, H. (1968) In Aspects of Yeast Metabolism (eds A.K. Mills and H. Krebs),
Blackwell, Oxford, p. 71.
 References 41
Hsu, N.P., Vogt, A and Bernstein, L. (1980) MBAA Technical Quarterly, 17, 85.
Institute of Brewing Analysis Committee (1971) Journal of the Institute of Brewing,
77,18I.
Jones, M. and Pierce, J.s. (1964) Journal of the Institute of Brewing, 70, 307.
Jones, M., Pragnell, M.J. and Pierce, J.s. (1969) Journal of the Institute of Brewing,
75,520.
Kara, B.Y., Daoud, I. and Searle, B. (1987) Proceedings of the 21st Congress of the
European Brewery Convention, Madrid, IRL Press, Oxford, p. 409.
Kara, B.V., Simpson, W.J. and Hammond. J.RM. (1988) Journal of the Institute of
Brewing, 94, 153.
Kirsop, B. (1955) Journal of the Institute of Brewing, 61, 466.
Kirsop, B. (1974) Journal of the Institute of Brewing, 80, 565.
Kirsop, B.E. (1984) In Maintenance ofMicroorganisms (eds B.E. Kirsop andJ.J.s. Snell),
Academic Press, London, p. 109.
Ladenburg, K. (1968) MBAA Technical Quarterly,S, 81.
Lodder, J. (ed.), (1970) The Yeasts, a Taxonomic Study, North-Holland, Amsterdam.
Luckiewicz, E.T. (1978) MBAA Technical Quarterly, 15,190.
Lynen, F. (1963) Control Mechanisms in Respiration and Fermentation (ed. B. Wright).
Proceedings of Annual Symposium 8, Society of General Physiologists, 1961,
Woods Hole, MA, p. 290.
Macy, J.M. and Miller, M.w. (1983) Archives of Microbiology, 134,64.
Maule, D.R (1976) The Brewer, 62, 140.
Maule, D.R (1977) The Brewer, 63,204.
Maynell, G.G. and Meynell E. (1965) Theory and Practice in Experimental Bacteriology,
Cambridge University Press, Cambridge.
Miller, L.F., Mabee, M.s., Gress, H.s. and Jangaard, N.O. (1978) Journal of the
American Society of Brewing Chemists, 36, 59.
Mills, D.R (1941) Food Research, 6, 36I.
Nathan, L. (1930) Journal of the Institute of Brewing, 36, 538.
Nykanen, L. and Suomalainen, H. (1983) Aroma of Beer, Wine and Distilled Beverages,
Reidel, Dordrecht.
Oura, E. (1977) Process Biochemistry, 12, 19.
Parkinnen, E., Oura, E. and Suomalainen, H. (1976) Journal of the Institute of Brewing,
82,283.
Pasteur, L. (1867) Etudes sur la Biere, Imprimerie de Gauthier-Villers, Paris.
Peddie, F.L.,Simpson, W.J., Kara, B.V., Robertson,S.C and Hammond,J.RM.(1991)
Journal of the Institute of Brewing, 97, 21.
Pringle, J.R and Hartwell, L.H. (1981) In Molecular Biology of the Yeast Saccharomyces:
Life Cycle and Inheritance, Cold Spring Harbor Laboratory, Cold Spring Harbor,
NYp.97.
Quain, D.E. and Tubb, RS. (1982) MBAA Technical Quarterly, 19,29.
Richards, M. (1967) Journal of the Institute of Brewing, 73, 162.
Rose, AH. and Harrison, J.s. (eds) (1969) The Yeasts, Vol. 1, Biology of Yeasts,
Academic Press, London.
Rose, AH. and Harrison, J.s. (eds) (1970) The Yeasts, Vol. 3, Yeast Technology,
Academic Press, London.
Rose, AH. and Harrison, J.s. (eds) (1971) The Yeasts, Vol. 2, Physiology and Biochem-
istry of Yeasts, Academic Press, London.
Ruocco, J.J., Coe, RW. and Hahn, CW. (1980) MBAA Technical Quarterly, 17, 69.
Russell, I. and Stewart,G.G. (1981) Journal ofthe American Society of Brewing Chemists,
39,19.
Seddon, AW. (1975) MBAA Technical Quarterly, 12, 130.
Shard low, P.J. (1972) MBAA Technical Quarterly, 9, 1.
Shardlow, P.J. and Thompson, CC (1971) Brewers' Digest, 46 (August), 76.
42 The biochemistry and physiology of yeast growth
Sols, A. and De la Fuente, G. (1961) In Membrane Transport and Metabolism (eds
A. Kleinzeller and A. Kotyk), Academic Press, London, p. 36l.
Sols, A., Gancedo, G. and De la Fuente, G. (1971) In The Yeasts, Vol. 2, Physiology and
Biochemistry of Yeasts (eds A.H. Rose and J.5. Harrison), Academic Press,
London.
Stryer, L. (1995) Biochemistry, 4th edn, W.H. Freeman & Co., New York.
Thomas, K.c. and Ingledew, W.M. (1990) Applied and Environmental Microbiology,
56,2046.
Thome, RS.W. (1949) Journal of the Institute of Brewing, 55, 20l.
Van den Berg, R (1978) European Brewery Convention, Fermentation and Storage
Symposium, Monograph V, Zoeterwoude, p. 66.
Walsh, RM. and Martin, P.A. (1977) Journal of the Institute of Brewing, 83, 169.
Wang, P.Y., Johnson, B.F. and Schneider, H. (1980) Biotechnology Letters, 3, 273.
Wynants, J. (1962) Journal of the Institute of Brewing, 68,350.
 CHAPTER 3

Yeast genetics
J.R.M. Hammond

3.1 INTRODUCTION

Yeasts have been employed in the production of alcoholic beverages for

thousands of years. Originally adventitious organisms present on raw
materials and equipment were responsible for the often unpredictable
fermentation which took place. More recently, strains have been selected
on the basis of their fermentation behaviour and the availability of pure
cultures has led to marked improvements in the reproducibility of
fermentations and to beer quality.
Genetic studies with Saccharomyces cerevisiae were pioneered by Winge
and colleagues at the Carlsberg Laboratories in Copenhagen and subse-
quently developed by Lindegren and co-workers. Yeasts, mainly of indus-
trial origin, were extensively interbred to produce strains which could be
mated to give healthy diploid cells capable of sporulating to produce four
viable ascospores. Strain S288C and its diploid derivative X2180 provide
the source of most genetically marked strains used in laboratory studies
throughout the world. However, as described below, these strains differ
markedly from those used for brewing and, as a consequence, studies of
the genetics of brewing yeast strains have lagged behind investigations of
laboratory organisms.
Against a background of fundamental knowledge of the genetics of
S. cerevisiae, this chapter will highlight the genetic peculiarities of brewing
yeasts. The limitations in their performance will be described as will the
difficulties encountered in trying to breed new yeast strains using so-called
'traditional' breeding methods. Finally, the tremendous developments
which the advent of recombinant DNA methods has brought about will be
explained together with the new insight gained into the genetic make-up
of brewing yeasts.

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and 1. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412 59150 2
44 Yeast genetics
3.2 GENETIC FEATURES OF SACCHAROMYCES CEREVISIAE

3.2.1 Life cycle and sporulation

The life cycle of a typicallaboratory strain of S. cerevisiae is shown in Fig. 3.l.
Such a yeast has both haploid (one set of chromosomes) and diploid (two
sets of chromosomes) modes of existence. Strains in which the haploid form
is stable and can be maintained for many generations are termed
heterothallic. The haploids from such strains exist as one of two mating

mi7
t."";,,/ DIPLOID

4- SPORIE~o::CUS
germination

O()GG
DIPLOID

mitosis

I
ating-type TETRAD
conversion
and mating
(homothallic strains)
mitOSiS
(heterothallic
strains)

ZYGOTE

HAPLOID

Fig. 3.1 Life cycle of Saccharomyces cerevisiae.

 Genetic features of Saccharomyces cerevisiae 45
types, a or a, and mate to form diploids when a cell of one mating type
comes into contact with a cell of the other mating type. Strains in which cell
fusion and diploid formation occur among cells derived from a single spore
are termed homothallic. The presence of the HO gene in such strains brings
about a high frequency of switching between mating types during vegeta-
tive growth. Under the influence of this gene, the mating type locus, MAT,
of such strains readily changes fromMATa to MA Ta or vice versa. The MAT
gene is found on chromosome III of the yeast genome together with two
silent genes, HMLa and HMRa, which provide the information to allow the
switch of mating type at the MAT locus. Cells of homothallic yeasts have
to bud at least once before they are able to switch mating type, but thereafter
a high frequency of switching occurs at each budding for many generations
(for further details see Herskowitz and Oshima, 1981).
In both homothallic and heterothallic strains, mating takes place when
cells of opposite mating type come into close proximity. Cells of a mating
type produce an oligopeptide called a-factor which stops the growth of a
cells and causes a and a cells to adhere to each other. Cells of mating type
a produce a-factor which has similar effects on a cells. In the presence of
these factors, the cells adhere and cytoplasmic fusion takes place to form
a heterokaryon (Thorner, 1981). Nuclear fusion follows rapidly to give a
zygote (Lindegren and Lindegren, 1943). By subsequent cell division this
forms the diploid phase of the yeast life cycle which can be stably
maintained for many generations. Meiosis and sporulation of diploid cells
is triggered by nitrogen deprivation in the presence of a non-fermentable
carbon source but will only occur if both MATa and MATa genes are
present. Following entry into meiosis the chromosomes in the yeast
nucleus undergo DNA synthesis, pairing, recombination and segregation.
Spore walls grow and envelop the four haploid genomes (two each of a
and a mating types) forming the characteristic four-spored ascus. The
spores, when placed in suitable nutrient media, germinate to form hap-
loids and begin the whole cycle once more. Meiotic recombination is
important for evolutionary change and provides the means whereby
chromosomes can be mapped (for further details see Esposito and
Klapholz, 1981).
Brewing strains behave very differently from laboratory-bred yeasts.
They sporulate poorly, rarely form spores in tetrads and many of the spores
formed are non-viable even when great efforts are made to produce ideal
conditions for sporulation, such as growing cells under complete catabolite
derepression and at lower temperatures than is usual for laboratory strains
(Gjermansen and Sigsgaard, 1981; Bilinski, Russell and Stewart, 1986,
1987a). In one study with brewing yeasts (Anderson and Martin, 1975) ale
yeasts were found to sporulate better than lager yeasts but only two- or
three-spored asci were formed and many of the resulting strains were
aberrant, being capable of mating with both a and a haploid yeasts, or were
sterile.
46 Yeast genetics
3.2.2 Chromosomes, ploidy and genetic stability
The chromosomes of S. cerevisiae are located in the nucleus and make up
between 80% and 85% of the total cellular DNA (Petes, 1980). In haploid cells
there is sufficient DNA to account for about 15 000 genes, although to date
only 769 genes have been mapped to the 16 chromosomes (Mortimer et al.,
1989). Each chromosome consists of a single DNA molecule of molecular
size between 150 and 2500 kilobase pairs (kbp) together with basic histone
protein molecules. Most genes of the haploid genome are present as single
copies, the major exceptions being the ribosomal RNA genes present as
about 100 copies and the approximately 15 copies of each transfer RNA gene
(Fangman and Zakian, 1981). Mobile genetic elements called Ty also occur
in Saccharomyces yeasts. Up to 35 copies of Ty can be present in haploid yeast
cells and their presence, together with their ability to move from one
chromosomal location to another, can cause rearrangements of the yeast
genome (Philippsenet al., 1983).
Whereas laboratory-bred strains of S. cerevisiae are either haploid or
diploid most yeasts used for brewing are polyploid or aneuploid. This
conclusion has been reached by measuring the DNA content of individual
cells and comparing the results with those obtained from haploid strains,
a notoriously unreliable procedure. Such studies have identified the
presence of cells containing between two (diploid) and seven (heptaploid)
times the normal haploid content of DNA as well as many aneuploids,
which have a DNA content intermediate between triploid and tetraploid
values (Johnston and Oberman, 1979). More recent studies involving direct
genetic analysis have confirmed that brewing strains are indeed polyploid,
particularly triploid, tetraploid or aneuploid.
Industrial yeasts may gain a number of benefits from being polyploid.
For instance, extra copies of important genes such as those responsible for
maltose utilization (MAL) could improve their fermentation performance.
Indeed, production of a-glucosidase and hence the rate of maltose fermen-
tation is increased with the dosage of MAL genes (Mowshowitz, 1979;
Stewart et al., 1981). It has also been argued that polyploid yeasts are more
stable than haploid yeasts since multiple mutational events are required in
order to change them. However, because of their very nature, polyploid
yeasts can harbour non-functional recessive mutations as demonstrated by
Delgado and Conde (1983). Therefore genetic stability of polyploid yeasts
is likely to be a function of the frequency of segregational events leading to
the expression of mutant genes, rather than the frequency of mutation itself.

3.2.3 Extrachromosomal elements

A number of extrachromosomal genetic elements are also known in yeasts,
the most important being 2 f.1m DNA, mitochondrial DNA and
double-stranded RNA (dsRNA).
 Genetic features of Saccharomyces cerevisiae 47

Two micrometre DNA is an extrachromosomal element found in most

strains of S. cerevisiae (Broach, 1981) including brewing strains (Tubb, 1980;
Stewart, Russell and Panchal, 1981; Aigle, Erbs and Moll, 1984). When
present, there are between 50 and 100 copies of 2 Ilm DNA per cell. Two
micrometre DNA occurs as circular plasmids (Hartley and Donelson, 1980)
but the only function which can be ascribed to these extrachromosomal
elements is their own maintenance; there appears to be no obvious advan-
tage for cells to possess 2 Ilm DNA. Although early work indicated that
these plasmids were cytoplasmic, a more detailed examination now sug-
gests they have a nuclear extrachromosomal location. Much genetic mod-
ification work carried out in recent years with brewing yeasts has involved
the use of 21lm DNA-based plasmids and for this reason these chromo-
somal elements have taken on a much greater importance than their lack
of function in vivo would suggest.
Mitochondrial DNA consists of a 75 kbp circular molecule and is present
as 10-40 molecules per cell in laboratory haploid yeast strains. It has a lower
buoyant density than chromosomal DNA, thereby facilitating separation
of the two types of molecule in density gradients (Fangman and Zakian,
1981). Mitochondrial DNA shows typical cytoplasmic inheritance and its
replication is highly independent of nuclear control, taking place through-
out the cell cycle (Newlon and Fangman, 1975). The mitochondrial genome
carries the genetic information for only a few essential components, more
than 90% of mitochondrial proteins being synthesized from nuclear-
encoded genes (Dujon, 1981). Mutations in mitochondrial DNA produce
petite strains which are unable to metabolize non-fermentable substrates.
Such respiratory-deficient mutations can range from point mutations (miq
through deletion mutations (rho-) to complete elimination of the mitochon-
drial DNA (rhoO). Petites of S. cerevisiae are often deficient in uptake of
sugars such as maltose and galactose (Evans and Wilkie, 1976; Mahler and
Wilkie, 1978) and petite mutants of brewing yeasts have been reported
which are deficient in maltotriose utilization and which therefore inade-
quately attenuate wort (Gyllang and Martinson, 1971). Brewing yeasts with
different mitochondrial genomes produce beers of dissimilar flavour
(Hammond and Eckersley, 1984) and petite mutants are very different from
their parents in flocculation behaviour, lipid metabolism and higher alco-
hol production (Lewis, Johnston and Martin, 1976). The mitochondrial
genome also has effects on diacetyl formation (Conde and Mascort, 1981;
Morrison and Suggett, 1983; Debourg et al., 1991) and on the production of
4-vinylguaiacol (Tubb et al., 1981).
Many strains of Saccharomyces contain cytoplasmic linear dsRNA mole-
cules enclosed in virus-like particles. There are two main varieties,
L-dsRNA, which is present in most yeast strains including some brewing
yeasts (Kreil, Kleber and Teuber, 1975; Young, 1981), and M-dsRNA, which
is present only in killer strains of Saccharomyces (Wickner, 1983). L-dsRNA
provides the code for the capsid protein of the virus particles and M-dsRNA
48 Yeast genetics
encodes both killer toxin and the immunity factor which prevents self-
killing. Cells of killer strains normally contain about 12 copies of M-dsRNA
and about 100 copies of L-dsRNA. They can be cured of M-dsRNA by
growth at elevated temperature or by treatment with cycloheximide (Fink
and Styles, 1973). Only one brewing killer yeast has ever been found (Young
and Yagiu, 1978), brewing yeasts generally being sensitive to the action of
killer toxin.

3.3 THE NEED FOR NEW BREWING YEASTS

3.3.1 The nature of a 'good' brewing yeast

In general the aim of brewing, like any other industrial process, is to obtain
the highest efficiency in the use of raw materials and production plant
without distorting the quality of the end-product. One way of achieving
this in the case of beer prod uction is by breeding more efficient yeast strains.
Breeding programmes also provide scope for the production of completely
new products from brewing yeasts.
The properties which are of most importance for a 'good' brewing yeast
are:
1. a rapid fermentation rate without excessive yeast growth;
2. an efficient utilization of maltose and maltotriose with good conversion
to ethanol;
3. an ability to withstand the stresses imposed by the alcohol concentra-
tions and osmotic pressures encountered in breweries;
4. a reproducible production of the correct levels of flavour and aroma
compounds;
5. an ideal flocculation character for the process employed;
6. good 'handling' characteristics (e.g. retention of viability during
storage, genetic stability).

3.3.2 Brewing yeast improvement

There is considerable scope for improving many of the properties of brew-
ing yeasts listed in the preceding section. In general, such improvements
will either involve extending the biochemistry of the organisms, modifying
their physiology or generating completely novel strains for employment in
new technologies.

(a) Extension of yeast biochemistry

There are many ways in which the biochemistry of brewing yeasts could
be extended to the advantage of the brewer.
The ability of brewing yeasts to secrete amylolytic enzymes
 The need for new brewing yeasts 49

(glucoamylases, amylases, pullulanases etc.) would considerably improve

the efficiency of utilization of wort carbohydrates. Amylolytic yeasts could
be used to produce well-attenuated or low-carbohydrate beers and to
reduce or replace the amounts of malt and enzymes needed to convert
starch to fermentable sugars. Since S. cerevisiae strains secrete few extra-
cellular proteins (Lampen, 1968), amylolytic yeasts would also provide the
opportunity to obtain high-purity enzyme preparations without the need
for elaborate processing.
The introduction of genes for proteolytic enzymes into brewing yeasts
could provide opportunities for degrading haze-forming proteins and so
naturally chill-proof beer. If the proteolysis was sufficiently specific to
produce 'head-positive' polypeptides this would lead to an improvement
in beer foam quality.
The secretion by yeast of a suitable endo-~-glucanase would enable
barley ~-glucans normally present in beer to be degraded during fermen-
tation, and so prevent the development of hazes and gels. Removal of
~-glucans would also improve the filterability of the beer at the end of the
process.
By engineering into yeast cells the ability to utilize carbohydrates other
than those normally metabolized, more complete degradation of conven-
tional raw materials becomes possible together with the use of new raw
materials. Enzymes for the degradation of pentoses and cellulose would
increase the degree of utilization of usual brewery raw materials whilst
enzymes for the degradation of lactose would permit the use of whey.
Most of these modifications are technically feasible and all have already
been worked on in some detail (sections 3.4.5(e) and 3.4.S(f)).

(b) Modification of yeast physiology

Many possibilities are available for introducing new characteristics into
yeast cells in order to improve their fermentation performance. The ability
to predictably change the flocculation ability of a strain would enable it to
be used in a number of very different fermentation systems. Similarly,
changes in cell adhesion could be useful if yeasts are to be used in im-
mobilized cell systems. Fermentations at high specific gravities require
strains with improved tolerance to ethanol and osmotic pressure. Yeasts
capable of fermenting at higher temperatures, while still producing accept-
able beers, would lead to reductions in cooling costs, higher productivities
and a reduced risk of microbial contamination. An increased tolerance to
pressure and carbon dioxide would permit fermentations to be carried out
in very large, deep fermentors or in pressurized vessels. Other possibilities
for the improvement of brewing yeasts include increasing fermentation
rates (possibly by increasing the dosage of the genes involved in maltose
utilization), improving fermentation efficiency by reducing their specific
growth yields, improving control of both the type and amount of flavour
50 Yeast genetics
compounds produced and reducing microbial contamination problems by
producing yeasts with anti-contaminant properties.
Many of these potential improvements are not well defined either bio-
chemically or genetically and so are not amenable to a direct genetic
approach. Those that are better understood such as flocculation, flavour
production and microbial contamination have again been worked on in
some detail (sections 3.4.3(b), 3.4.5(g), 3.4.5(h) and 3.4.5(i)).

(c) Novel strains for new technologies

This area represents the most difficult one for brewing geneticists but is
potentially the most exciting. Yeasts with a restricted pattern of fermenta-
tion could be engineered in order to produce a low-alcohol beer. Another
worthwhile target would be to introduce into yeasts genes for the produc-
tion of new flavour compounds in order to produce beers with unusual
flavours. Finally there is the concept of using spent brewer's yeast to
produce valuable by-products such as vitamins, enzymes, lipids etc., their
formation being enhanced by suitable genetic modification. Alternatively
brewing yeasts could be genetically developed to produce completely new
products after they have served their role in the brewery. The possibilities
here are limitless. The concept has already been tried for the production of
human serum albumen (Hinchliffe, Kenny and Leaker, 1987) but because of
both supply and legislative difficulties the use of spent brewing yeast has
been abandoned in favour of specially grown yeast for this application. With
this one exception, very little effort has so far been put into exploring the
possibilities offered by the production of novel yeasts for new technologies.

3.4 GENETIC TECHNIQUES AND THEIR APPLICATION TO THE

ANALYSIS AND DEVELOPMENT OF BREWING YEAST STRAINS

3.4.1 Mutation and selection

A simple and direct means for strain development is to isolate mutants,
with or without prior mutagenesis. Variation is always present within a
yeast population and successful isolation of mutants depends on the fre-
quency with which they occur and the ability of the experimenter to devise
appropriate selection procedures. One spontaneous mutation frequently
encountered in breweries is that of changing yeast flocculation character-
istics. These changes have enabled bottom-fermenting ale yeasts, for use in
cylindroconical fermenting vessels, to be selected from their top-cropping
predecessors. Continuous culture can be readily used to isolate variants
within a yeast population without prior mutagenesis. This was demon-
strated by Thome (1970) when he grew an apparently stable lager yeast in
this way for 9 months. After this time more than half of the yeasts isolated
 Genetic techniques in brewing yeast strains 51
showed a difference in at least one property of brewing significance.
Selection in continuous culture has also been used to obtain yeasts with
improved ethanol tolerance (Brown and Oliver, 1982; Korhola, 1983).
The use of mutagens increases the proportion of mutants within a given
population and has proved surprisingly successful with polyploid brewing
yeasts. Molzahn (1977) mutated both ale and lager yeasts using N-methyl-
N-nitro-N-nitrosoguanidine and ultraviolet light. By selecting for valine
and methionine auxotrophs he successfully isolated a number of strains
which respectively produced less diacetyl and hydrogen sulphide, both
important negative flavour attributes of beer. Flocculation variants were
also obtained in the same series of experiments. In other experiments,
strains with a reduced ability to form esters and higher alcohols have been
isolated as spontaneous mutants showing resistance to glucosamine
(Hockney and Freeman, 1980). Mutants resistant to thia-isoleucine have
been obtained which produce beer containing higher levels of amyl alcohol
(Kielland-Brandt, Peterson and Mikkelsen, 1979). Although successful, the
isolation of mutants in most of these experiments was too frequent to be
explained by simultaneous inactivation of several copies of wild-type genes
in polyploid brewing yeasts. Other factors such as mitotic recombination
and chromosome loss must also be involved (see also section 3.2.2).
The selection of mutants resistant to the glucose analogue 2-
deoxyglucose (DOG) has proved extremely valuable when applied to
brewing yeasts (Jones, Russell and Stewart, 1986). In normal fermentations,
due to catabolite repression, maltose is only taken up by yeast when about
half of the glucose present in the wort has been metabolized. Mutants
resistant to the action of DOG are derepressed, utilizing maltose and
glucose simultaneously, and have improved fermentation characteristics.
Mutation methods can also be used to produce genetically marked yeast
strains for further analysis. Gjermansen (1983) used ultraviolet light and
ethylmethane sulphonate mutagenesis to induce auxotrophic mutations in
meiotic segregants of a production lager yeast. These mutants were then
used as part of a breeding programme to produce new production strains
(section 3.4.2) and for detailed genetic analysis of the lager yeast (section
3.4.3).

3.4.2 Hybridization
Classically, hybridization of strains of S. cerevisiae involves mating haploids
of opposite mating type to give a heterozygous diploid. Recombinant
meiotic progeny are recovered by sporulating the diploid and recovering
the individual haploid spores. This is the basis for the classical approach to
genetic mapping, called tetrad analysis because of the four (tetrad) spores
found in each ascus formed by a sporulating diploid cell. Detailed analysis
of the results of crosses between haploid yeasts enables individual genes to
be mapped to particular chromosomes and to particular parts of these
52 Yeast genetics
chromosomes. A more detailed description of these and other mapping
procedures can be found in Sherman and Wakem (1991).
As already mentioned in sections 3.2.2 and 3.2.3, brewing yeasts are
polyploid and sporulate very poorly. Consequently, the use of hybridiza-
tion techniques for strain development has proved difficult. Several work-
ers have persevered and managed to isolate meiotic segregants from
brewing yeasts which can then be used in breeding programmes.
Gjermansen and Sigsgaard (1981) carried out pairwise crosses between
segregants obtained from one production yeast. Although most hybrids
were inferior to the parent yeast, one with good brewing properties was
obtained. This strain has been used at full production scale where it
fermented faster, flocculated better and produced a better flavoured beer
than did the normal production yeast (von Wettstein, 1983). The segregants
were originally thought to be haploid (Gjermansen, 1983) but more detailed
genetic analysis (section 3.4.3) has revealed that, despite their lower ploidy,
they are almost as genetically complex as their polyploid parent.
An alternative approach was taken by Bilinski, Russell and Stewart
(1987b) who obtained meiotic segregants from both ale and lager yeasts. By
mating these, a species hybrid was produced which had both ale and lager
yeast characteristics since it could both grow at 37°C and utilize the sugar
melibiose. The fermentation performance of the hybrid was as good as the
parent lager yeast and the ethanol yield was better. The beer produced
lacked the sulphury flavour characteristic of the lager yeast parent but had
an estery note more typical of beers produced by the ale yeast.
Although such hybridization methods are less specific than recombinant
DNA techniques (section 3.4.5), they may provide the best approach for
improving yeast characteristics which are poorly defined genetically or
which result from the complex interaction of many genes.
Because of their poor sporulating ability, it is hopeless to attempt to carry
out tetrad analysis with brewing yeasts or their progeny. Consequently
genetic analysis of the genes of importance in brewing yeasts has been
carried out in model systems based on laboratory strains of S. cerevisiae.
Genes controlling flocculation have been defined and mapped in this way
(Lewis, Johnston and Martin, 1976; Russell et al., 1980; Stewart and Russell,
1986) and gene dosage effects of the genes controlling maltose metabolism
have been described in constructed polyploid strains (Mowshowitz, 1979;
Stewart, Goring and Russell, 1977). Further, hybridization and tetrad anal-
ysis have been used to assign the ability of yeasts to decarboxylate cinnamic
acid to a single dominant gene (POFl). This is important because cells
containing this gene impart a phenolic flavour to beer if used for fermen-
tation (Goodey and Tubb, 1982). The POFl gene has been cloned and
partially characterized (Meaden and Taylor, 1991). The introduction of this
cloned gene into brewing yeasts leads to the production of an aroma
characteristic of phenolic off-flavour when the transformed strain is used
to ferment wort. This strongly suggests that brewing yeasts do not normally
 Genetic techniques in brewing yeast strains 53
produce a phenolic off-flavour when used to produce beer because of their
lack of a functional POFl allele.

3.4.3 Rare mating

(a) Rare mating techniques

Brewing yeasts which fail to show a mating type can be mated with haploid
strains by a procedure known as rare mating. A large number of cells of the
parental strains are mixed together and a strong positive selection proce-
dure is applied to obtain the rare hybrids formed (Gunge and Nakatomi,
1972). These hybrids are usually selected as respiratory-sufficient
prototrophs from crosses between a respiratory-deficient mutant of the
industrial strain and an auxotrophic haploid strain (Fig. 3.2).
The technique has been used to construct brewing strains with the ability
to ferment dextrins (Tubb et al., 1981). The beers produced by these hybrids
had an unacceptable phenolic off-flavour because of the presence of the
POFl gene in the diastatic parent yeast. Only after the gene had been
eliminated from this latter yeast by conventional hybridization techniques
were satisfactory diastatic hybrids produced (Goodey and Tubb, 1982).
Using this approach, followed by classical genetic improvement proce-
dures, mitotically stable brewing yeasts capable of superattenuation have
been constructed (Guinova-Stoyanova and Lahtchev, 1991).

(b) Cytoduction
Rare mating can also be used to generate progeny (cytoductants or
heteroplasmons; Fig. 3.2) which receive cytoplasm from both parents but
retain the nucleus of only one. By using haploid maters which carry the
karl mutation (karyogamy-deficient; i.e. deficient in nuclear fusion)
cytoduction can be favoured over hybrid formation during rare mating
experiments. This technique has been used extensively to transfer the
dsRNA determinants for yeast killer factors from laboratory strains into
suitable recipient strains in order to produce brewing yeasts with anti-
contaminant properties (Young, 1981, 1983; Hammond and Eckersley,
1984). In some cases, construction of killer yeasts by cytoduction impaired
the brewing characteristics of the recipient strain probably as a conse-
quence of introducing a 'foreign' mitochondrial genome (Hammond and
Eckersley, 1984). It is clear that the source of mitochondria in a modified
brewing yeast can have marked effects on fermentation rate, flocculation
behaviour and beer flavour (Conde and Mascort, 1981; Russell, Jones and
Stewart, 1985). Although in principle the manipulation of the mito-
chondrial genome by cytoduction provides a means for strain improve-
ment, an attempt to correct a sugar fermentation problem using this
approach was not successful (Hammond and Wenn, 1985).
54 Yeast genetics

(f)
o 0
®
l.
0
INDUSTRIAL LABORATORY
STRAIN STRAIN

"·ti~ .,
r.spiratory
d.ficiency

a
PETITE
MUTANT
rare
mating

HYBRID CYlODUCTANTS

Fig. 3.2 Rare mating between industrial and laboratory strains of yeast.

(c) Single chromosome transfer

During matings between strains, one of which carries the karl allele,
occasional rare progeny are formed which contain the nuclear genome of
one parent together with an additional chromosome from the other
(Nilsson-Tillgren et al., 1980; Dutcher, 1981; Goodey et al., 1981). This
phenomenon has been used to transfer single chromosomes from meiotic
segregants of the lager brewing yeast strain M244 into laboratory yeasts,
thereby allowing detailed genetic analysis of the brewing yeast
 Genetic techniques in brewing yeast strains 55
chromosomes to be carried out. Chromosomes III, V, X, XII and XIII have
been studied in this way.
Two forms of chromosome III have been identified. One is functionally
and structurally identical to the chromosome III found in laboratory strains.
The other chromosome type appears to be functionally homologous to the
laboratory strain chromosome but is structurally different, recombination
between the two chromosome types only occurring in certain regions
(Nilsson-Tillgren et ai., 1981; Holmberg, 1982). The presence of two forms
of chromosome III has been confirmed by an examination of the structure
of the LEU2 gene, which is found on this chromosome. Two different LEU2
genes are present in the lager strain (Pedersen, 1985b) and one LEU2 gene
cloned from the brewing yeast has a restriction endonuclease map consid-
erably different from that of classical S. cerevisiae strains (Casey and
Pedersen, 1988.)
In a similar manner two forms of chromosome V have been identified,
one of which is both structurally and functionally homologous to the
chromosome found in laboratory strains of S. cerevisiae. The other fails to
recombine at all despite being functionally homologous (a so-called
homeologous chromosome; Nilsson-Tillgren et ai., 1986). Chromosome X
has been examined using pulsed-field electrophoresis (section 3.4.5(j)) and
three independently migrating structures identified, two of which are not
homologous to chromosome X from S. cerevisiae, recombination occurring
along only part of the chromosome (Casey, 1986a,b). The third chromo-
some, although the same length as thatfrom laboratory strains, has not been
examined in any detail. Petersen et ai. (1987) analysed both chromosomes
XII and XIII. They were able to isolate two different types of chromosome
XII, one homologous and the other homeologous to the equivalent chromo-
some from laboratory strains. In contrast three different chromosomes XIII
were isolated, one showing homology to that of S. cerevisiae, a non-
recombining homeologous chromosome and a mosaic-type chromosome
exhibiting partial recombination.
Thus there appear to be three types of chromosomes present in lager
brewing yeast M244: chromosomes which are functionally and structurally
homologous to the equivalent S. cerevisiae chromosome, represented by one
form of chromosomes III, V, XII and XIII; homeologous chromosomes
which are functionally but not structurally homologous to the equivalent
S. cerevisiae chromosomes, represented by one form of chromosomes V, XII
and XIII; mosaic chromosomes where recombination with S. cerevisiae
chromosomes is limited to only parts of the chromosomes, represented by
one form of chromosomes III and XII and both identified forms of chromo-
some X. Clearly lager yeast M244 has two recombination-incompatible
genomes. The question of the copy number of each version of the various
chromosomes has been examined by means of the IL V2 locus found on
chromosome XIII. IL V2 gene copies were replaced by ilv2 deletion mutants.
It was found that the brewing yeast contained two copies of each of the two
56 Yeast genetics
versions of the ILV2 region (Gjermansen et al., 1988). If this can be applied
generally, then the brewing yeast would appear to be allotetraploid,
although the alloploidy is clearly irregular with many mosaic chromo-
somes occurring. In addition, some chromosomes apparently lack certain
genes (Nilsson-Tillgren et al., 1986; Casey, 1986b). If all brewing yeasts are
allotetraploids with irregular chromosome structures and missing or de-
fective genes this may help to explain their low sporulation frequencies and
low spore viabilities.

3.4.4 Spheroplast fusion

Spheroplast fusion provides another way of overcoming the mating/
sporulation barrier found in industrial yeasts but, unlike rare mating, it is
a direct asexual technique. It can be used to produce both hybrids and
cytoductants. The procedure, first described by van Solingen and van der
Plaat (1977), is outlined in Fig. 3.3. Spheroplasts are formed by removing
the cell wall with a suitable lytic enzyme, in a medium containing an
osmotic stabilizer (often 1.0 M sorbitol) to prevent cell lysis. Spheroplasts
from different strains are mixed together in the presence of polyethylene
glycol and calcium ions and then, after fusion has taken place, allowed to
regenerate their cell walls in an osmotically stabilized agar medium. Cells
can also be induced to fuse by the application of a high-intensity electric
field as described by Halfmann et al. (1982).
The major problem with spheroplast fusion is that a wide range of cell
types is produced which are often mitotically unstable, especially in the
case of inter-species and inter-generic fusants. Nevertheless the technique
has been applied to the development of brewing yeasts. Dextrin-fermenting
strains have been constructed by fusing together S. cerevisiae var. diastaticus
and brewing yeasts (Barney, Jansen and Helbert, 1980; Freeman, 1981;
Russell, Hancock and Stewart, 1983; de Figueroa and de van Broock,1985;
Hansen, Rocken and Emeis, 1990). As with rare mating studies, the produc-
tion of phenolic off-flavours was a problem until the POFl gene was
eliminated by preliminary hybridization of the diastatic strain. Spheroplast
fusion has also been used to create yeast strains with enhanced tolerances
to ethanol, temperature and osmotic pressure (Panchal, Peacock and
Stewart, 1982; Crumplen et al., 1990). The fusion of a brewing yeast with a
FL05 flocculent laboratory strain has been reported (Vrano, Nishikawa and
Kamimura, 1990). The fusants were intermediate between the two parents
with respect to carbohydrate assimilation and flocculation. A further
application of the technique has been to produce lager brewing yeasts with
anti-contaminant properties; either producing yeast killer factor (Rocken,
1984) or possessing both anti-yeast and anti-bacterial properties (Sasaki et
al., 1984). The latter was achieved by mating a laboratory killer strain of
yeast with a strain having anti-bacterial activity and then fusing the hybrid
with a brewing yeast. The constructed yeast produced poor quality beer,
 Genetic techniques in brewing yeast strains 57

~ ~ YEA~ CElLS

1 1
® ®
sp/1eroplasting
enzyme

S.... ROPLASTS

];
@® tusing agent (PEG/Ca--)

~
® FUsm SPliEROPLASTS

!cell wall regeneration medium

FUSED CELLS

Fig. 3.3 Spheroplast fusion applied to yeasts.

probably because of the large contribution to its genome from the non-
brewing strains, always a problem with this type of experimental system.
Of course, the production of inter-generic hybrids can only be carried
out using spheroplast fusion. Stable fusion hybrids between S. cerevisiae and
Yarrowia lipolytica (de van Broock, Sierra and de Figueroa, 1981) or Candida
utilis (Perez, Val1in and Benitez, 1984) have been reported whereas fusion
products obtained between S. cerevisiae and Kluyveromyces Iactis degenerate
rapidly into strains showing parental characteristics (Stewart, Russell and
Panchal,1981).
 58 Yeast genetics
3.4.5 Transformation and recombinant DNA methods

(a) Transformation methods

Transformation involves the use of DNA molecules to bring about a
change in a recipient organism. Because the mating system is bypassed
the technique can readily be applied to industrial yeasts. Following early
unconfirmed reports of transformation using native DNA (Oppenoorth,
1959; Khan and Sen, 1974), an unequivocal demonstration of yeast trans-
formation awaited the availability of recombinant DNA techniques. Using
this approach (Fig. 3.4), DNA is fragmented using an endonuclease (re-
striction enzyme). The fragments are then annealed to plasmid DNA
which has been linearized, usually with the same restriction enzyme, after
which recombinant plasmids are generated by 'stitching up' the joins
using a DNA ligase. In this way a pool of recombinant plasmids is
produced, each containing a small piece of the original donor DNA. A
sufficiently large random pool in which there is a high probability of
finding any single gene from the donor strain is called a gene 'bank' or
'library' (Clarke and Carbon, 1979). A number of approaches can be used
to isolate or 'clone' individual genes from such a bank. Details of the

YEAST CELL

spheroplastlng
enzyme

o
TRANSFORMED@
YEAST CELL

·...:1 (J
SPHEROPLAST
cuf DNA CUT PLASMID
PIECES

1
DNA
re.trlctlon i
o
enzyme
~
DONOR PLASMID DNA
DNA

Fig. 3.4 The methods used for yeast transformation.

 Genetic techniques in brewing yeast strains 59
procedures used in recombinant DNA research can be found in Guthrie
and Fink (1991).
There are various ways in which yeast cells can be induced to take up
DNA from the medium in which they are suspended. One technique
involves removing the cell walls using a spheroplasting enzyme as de-
scribed in section 3.4.4 (Hinnen, Hicks and Fink, 1978), whilst another
involves treatment of the cells with lithium or caesium salts (Ito et al., 1983).
In both cases the yeasts and DNA are mixed under conditions similar to
those used for spheroplast fusion. Alternatively DNA can be induced to
enter yeast cells by the application of an electric current, a technique known
as electroporation (Becker and Guarente, 1991).
The single chromosome transfer procedure (section 3.4.3(c) has also been
used to transfer plasmids from laboratory strains of yeast carrying the karl
mutation into brewing yeast strains. This provides an alternative trans-
formation method to those involving isolated DNA molecules (Navas,
Esteban and Delgado, 1991).

(b) Dominant selection markers

Since the first demonstration of transformation of yeast, a wide range of
plasmid vector systems have become available and many have been applied
to the production of new industrial yeast strains. Since brewing yeasts are
polyploid, it is extremely difficult to generate the auxotrophic mutants
normally used for selection purposes in transformation experiments involv-
ing laboratory strains. Consequently the plasmids used for transformation
experiments with brewing yeast strains have to carry a dominant marker
which is selectable against the wild-type polyploid background. The most
widely used dominant selection markers are described below. Other select-
able markers with potential for use with industrial yeasts have been
discussed elsewhere (Knowles and Tubb, 1986; Esser and Kamper, 1988).
Copper resistance, coded by the CUPl gene, is probably the most pop-
ular selection method used in transformation experiments with industrial
yeast strains (Henderson, Cox and Tubb, 1985; Meaden and Tubb, 1985;
Hinchliffe and Daubney, 1986). The CUPl gene is derived from S. cerevisiae
and is found in many wild and laboratory strains. Selection is possible since
almost all brewing yeast strains are sensitive to copper. Brewing yeasts
transformed with plasmids containing the CUPl gene ferment at the same
rate as untransformed yeasts and produce beer which is indistinguishable
from that normally produced (Meaden and Tubb, 1985).
Sacchararomyces yeasts are sensitive to geneticin (G418), an antibiotic
which is structurally related to gentamycin and kanamycin. A bacterial
gene which encodes an inactivating enzyme (aminoglycoside phospho-
transferase) is expressed in yeast, thus resistance to G418 can be used as a
selectable marker Gimenez and Davies, 1980). A disadvantage of this
system is that the resistance factor is a bacterial gene which must be
60 Yeast genetics
eliminated before the transformed strains can be used for beer production.
Various strategies enable this to be done (section 3.4.5(e».
A more recently described dominant selection marker is that for resis-
tance to the herbicide sulphometuron-methyl (SMR1; Casey, Xiao and
Rank, 1988a). Like CUP1, it is derived from yeast and has no significant
effect on fermentation behaviour. The gene is in fact a mutant allele of the
IL V2 gene of S. cerevisiae which encodes the enzyme acetolactate synthase.
It is particularly useful where integration of a new gene into the yeast
chromosome is required since it is not only a dominant selectable marker
but also provides a site of integration in the IL V2 gene (Casey, Xiao and
Rank, 1988c; Xiao and Rank, 1989).
Resistance to chloramphenicol has also been used for selecting trans-
formants of industrial yeasts (Hadfield, Cashmore and Meacock, 1986).
Since this resistance is a mitochondrial function it can only be used when
yeast cells are grown on non-fermentable substrates. However, with this
proviso, it is very effective and has been successfully used for transforma-
tion of brewing yeasts. Like G418 resistance, chloramphenicol resistance is
coded by a bacterial gene and so must be eliminated prior to commercial
use of the transformed yeast.
Since ale yeast strains cannot use melibiose as a carbon source it is
possible to use the acquisition of this ability as a selection system in such
yeasts (Tubb et al., 1986). However, the melibiose utilization gene, MELl,
encodes an extracellular enzyme and so there is a problem of Mer strains
cross-feeding adjacent Mer strains, making selection more difficult. Direct
selection of dextrin-utilizing yeasts on starch plates suffers from the same
problem (Meaden et al., 1985).

(c) Gene stability and expression

The various recombinant DNA methodologies, such as the use of multi-
copy plasmids based upon 2 ).lm DNA, single copy centromere-containing
plasmids or integrative transformation, have been reviewed recently
(Knowles and Tubb, 1986; Esser and Kamper, 1988). Most initial transfor-
mation experiments with brewing yeasts used multi-copy plasmids such
as pJDB207 (Beggs, 1981) containing yeast chromosomal DNA, part of the
yeast 2 ).lm plasmid and DNA from a bacterial plasmid. The presence of
bacterial DNA provides the opportunity to replicate the plasmid in
Escherichia coli, without which transformation experiments would be very
tedious. These multi-copy plasmids have the advantage that, due to the
number of copies present in the cell, there is usually good expression of the
gene. This is, however, sometimes disadvantageous since overproduction
of gene product can have deleterious effects on fermentation performance
(Perry and Meaden, 1988). A further shortcoming is that such plasmid-
based systems are relatively unstable, the desired trait gradually being lost
with successive fermentations (Meaden and Tubb, 1985; Cantwell et al.,
 Genetic techniques in brewing yeast strains 61
1985). In contrast, when the required gene is integrated into the yeast
chromosome resulting transformants are extremely stable, although the
level of expression may be much reduced because of lower gene copy
number (Enari et al., 1987). An alternative, intermediate system has been
developed which has the expression advantages of a multi-copy plasmid
but is less unstable than the typical bacterial-yeast hybrid plasmid
(Hinchliffe, Fleming and Vakeria, 1987). All brewing yeasts contain 2 /lm
DNA which is stably maintained by a highly developed maintenance
system. This stability is lost when bacterial sequences are present in the
plasmids. By inserting the gene of interest directly into the endogenous
2 /lm plasmid, a high copy number transformant can be obtained which is
considerably more stable than conventional chimaeric plasmids. Similar
effects can be achieved using so-called 2 /lm 'disintegration vectors' which
contain both yeast and bacterial DNA but which are constructed in such a
way that the bacterial sequences are eliminated once they have entered and
transformed a yeast cell (Chinery and Hinchliffe, 1989).
Although transformation of brewing strains with native DNA has been
reported (Jansen et al., 1979; Barney, Jansen and Helbert, 1980), the use of
recombinant methods offers a much higher degree of specificity together
with the possibility of amplifying the amount of a gene product made by a
yeast strain. Such amplification can be achieved by inserting the gene of
interest into a multi-copy plasmid, by integrating the gene at several sites
within the yeast chromosomes or by attaching a highly efficient promoter
sequence to the gene in order to increase its expression. This last approach
is the one that has now been almost universally adopted. Typical highly
efficient promoters are those that are involved in controlling the genes for
enzymes of the glycolytic pathway, such as alcohol dehydrogenase (ADH1;
Hitzeman et al., 1981), glyceraldehyde-3-phosphate dehydrogenase
(GAPDH;Montgomeryetal., 1980) and phosphoglycerokinase (PGK1;Tuite
et al., 1982). As well as using efficient promoters, it is also possible to attach
regulated promoters to genes, so that they can be switched on and off in
response to external factors. This has potential for controlling accurately
the levels of gene expression.

(d) Application of recombinant DNA methods to brewing yeasts

Although a relatively recent technology, recombinant DNA methods have
already been used to introduce several new properties into brewing yeasts.
These have included the ability to degrade various carbohydrates and
proteins as well as modifications to the processes responsible for beer
flavour production and flocculation during fermentation. A number of
criteria must be met before such strains can be considered for commercial
production of beers for sale. The yeasts must be stable under the conditions
of fermentation normally employed and must produce sufficient new
product in order that the desired phenotypic change is realized. On the
62 Yeast genetics
other hand, the fermentation behaviour of the yeast and the quality of the
product must remain unaffected except, of course, for the desired new
attribute. Finally it is desirable that the new strain should contain no DNA
other than that necessary for generation of the new property. This will
minimize any objections from the regulatory authorities (the topic of
regulation is covered in more detail in section 3.5).

(e) Brewing yeasts able to use a wider range of carbohydrates

Glucoamylases
Brewing yeast strains are normally only able to utilize mono-, di- and
trihexoses. Larger oligomers (dextrins), which represent about 25% of wort
sugars, are not metabolized. Yeasts able to ferment these additional sugars
convert raw materials to ethanol much more efficiently than do brewing
yeasts. They can be used to produce low carbohydrate diabetic beers or, by
dilution to normal ethanol contents, low calorie 'light' beers are produced.
Currently such beers are often produced by adding commercial enzymes,
either prior to wort boiling or during fermentation. The commercial
enzymes used are typically fungal glucoamylases from Aspergillus niger.
A yeast capable of producing its own glucoamylase has several advan-
tages. Enzyme addition is much more controllable, mistakes in addition
rates are eliminated and deleterious enzyme activities which are often
present in commercial enzyme preparations are avoided. Consequently
there has been much interest in the generation of brewing yeasts capable
of fermenting dextrins. The limited success achieved using rare mating and
spheroplast fusion has already been described (sections 3.4.3(a) and 3.4.4).
Much greater success has been achieved utilizing recombinant DNA meth-
ods by transforming yeast strains with a gene for glucoamylase (a-1,4-
glucan glucohydrolase, EC3.2.1.3), an enzyme capable of degrading the
normally unfermentable wort dextrins to glucose. A gene for this enzyme
(DEXl or STA2) was cloned from S. cerevisiae var. diastaticus (Meaden et al.,
1985) and inserted into a multi-copy plasmid containing both bacterial and
yeast DNA sequences including the CUPl selectable marker. The plasmid
was used to transform both ale and lager brewing yeast strains (Meaden
and Tubb, 1985). The transformants produced extracellular glucoamylase
but, because of the lack of a-1,6 debranching activity, superattenuation was
limited. As expected for a chimaeric multi-copy plasmid, some degree of
plasmid instability was experienced. Additionally, growth and fermenta-
tion rates were adversely affected by the production of excess gluco-
amylase (Perry and Meaden, 1988). The plasmid instability problem
was largely overcome by the development of an all-yeast multi-copy
plasmid (Vakeria and Hinchliffe, 1989) but the fermentation rate of the
transformant was still somewhat slower than that of the parent strain.
More recently another brewing yeast has been transformed using this
same plasmid construct. The modified yeast fermented at about the same
 Genetic techniques in brewing yeast strains 63
rate as the parent strain but superattenuated the wort to produce approxi-
mately 1% (v Iv) more ethanol (Hammond, 1995). The new yeast was
judged to be commercially stable in that, although it slowly lost the STA2
gene, this had no noticeable effect on superattenuation or beer flavour over
ten successive fermentations. Under normal circumstances a newly prop-
agated yeast would be introduced before any significant effect on yeast
performance would be noticed. Beers have been successfully produced
using this yeast at the small commercial (100 hI) scale.
The STAl gene, which is closely related to STA2, has also been cloned
into multi-copy plasmids and used for transformation of brewing yeasts,
using both G418 resistance (Sakai et ai., 1989) and copper resistance (Park
et ai., 1990) as the selection systems. Plasmid stability and fermentation
performance were comparable to those achieved with STA2 trans-
formants. Only when the gene was integrated into the yeast chromosome
was a stable transformant obtained (Park et ai., 1990). A further develop-
ment has been the transformation of brewing yeasts with a centromeric
plasmid containing both STA2 and the a-amylase gene AMY from Bacillus
amyioliquefaciens (Steyn and Pretorius, 1991). The transformed yeasts were
capable of secreting both enzymes and carried out an efficient one-step
utilization of starch.
The major problem with all systems based on glucoamylases derived
from S. cerevisiae var. diastaticus is the lack of a-1,6 deb ranching activity
which leaves considerable amounts of wort dextrins undigested. This has
been overcome by using genes for glucoamylases from other microbial
sources. The GA gene from Aspergillus niger, which encodes a glucoamylase
with both a-1,4 and a-1,6 hydrolytic activities, has been cloned and inserted
into an integrating vector (Yocum, 1986). The vector was designed to
achieve integration of the GA gene into the yeast genome whilst ensuring
that any bacterial DNA sequences were eliminated from the transformants
(Fig. 3.5). In order to ensure genetic stability three copies of the GA gene
were integrated into three different copies of the yeast HO gene. Good
expression was achieved by using a yeast glycolytic enzyme promoter and
good secretion was obtained by using a yeast secretion signal sequence.
Fermentation trials at scales up to 100 hI have been extremely successful;
high levels of secreted extracellular glucoamylase have been achieved and
superattenuated beer of good quality has been produced on a regular basis
(Gopal and Hammond, 1992).
One disadvantage of the glucoamylase from A. niger is that it is relatively
heat stable and retains activity in beer after pasteurization. Consequently
there has been some interest in the glucoamylase from Schwanniomyces
occidentaiis which, although not currently used commercially, has the ad-
vantage of being much less heat stable than the fungal enzyme, as well as
possessing debranching activity. The GAMI gene for this enzyme has been
cloned (Lancashire et al., 1989; Dohmen et al., 1990) and, for transforming
brewing yeast strains, placed under the control of the Saccharomyces ADHl
64 Yeast genetics

1 Cut atA
transform

~[Q~"""""~i:~lac~zC:~ ..

1
Yeast
chromosome
-------~. . . . . . . . . . . . . .t_-------

Select G418 r

-I"IC~GACI""""~C:~~==~~~--~""""""~-

Recombination, white colonies

GA

Fig. 3.5 Transformation protocol for achieving stable integration of Aspergillus

niger glucoamylase gene into brewing yeasts. The GA gene is inserted into a copy
of the redundant yeast HO gene contained on a plasmid which also possesses
copies of the bacterial LacZ and G418 resistance genes. The lack of a yeast origin
of replication ensures transformation by integration. The plasmid is cut at A and
used to transform brewing yeasts. Transformants are selected by their ability to
grow in the presence of G418, Recombination between the two copies of the HO
gene leads to either complete loss of the insert or loss of all extraneous bacterial
DNA with retention of the GA gene. These recombination events can be detected
by growing the cells on a medium containing X-gal (5-bromo-4-chloro-3-indolyl-
~-D-galactopyranoside). Initial transformants produce blue colonies because of the
presence of the LacZ gene. Once this has been lost the colonies become white. By
screening white colonies for glucoamylase activity the desired transformants can
be isolated.

promoter (Lancashire et ai., 1989). Vectors for both multi-copy and

integrative transformation have been constructed using SMRI as the select-
able marker. Both types of transformant produced glucoamylase, the fer-
mentations proceeding normally to a lower gravity than usual to produce
acceptable beers.
 Genetic techniques in brewing yeast strains 65
~-Glucanases
Barley ~-glucanase is heat labile and easily destroyed during malting. As a
result, barley ~-glucans are often found in beer, where they present filtra-
tion difficulties and can give rise to hazes, sediments and gels. To overcome
this problem, microbial ~-glucanases are often added to mashes or worts.
Consequently there has been considerable interest in transferring genes
coding for ~-glucanase into brewing yeasts. Genes have been cloned from
Bacillus subtilis (Cantwell and McConnell, 1983; Hinchliffe and Box, 1984;
Lancashire and Wilde, 1987), from Trichoderma reesei (Knowles et al., 1985)
and from barley (Jackson, Ballance and Thomsen, 1986). Initial attempts to
express the B. subtilis gene in yeast were disappointing, yielding only low
levels of intracellular activity (Hinchliffe and Box, 1985). Expression was
increased dramatically by placing the ~-glucanase gene under the control
of the yeast ADHI promoter (Cantwell et al., 1985). Transformed brewing
yeasts produced high levels of ~-glucanase intracellularly but only very low
levels were detected in the beer, probably due to cell lysis. Only when a
yeast secretion signal sequence was also introduced (Lancashire and Wilde,
1987) were significant levels of extracellular ~-glucanase obtained. For this
last work the bacterial gene was placed under the control of the yeast MFa.
promoter and the a. mating factor signal sequence. Selection of trans-
formants was based on chloramphenicol resistance. The fermentation per-
formance of the transformed yeasts was identical to that of the parent
brewing yeast. However, the ~-glucan contents and viscosities of the
experimental beers were considerably reduced and their filterabilities
much improved.
All reported work involving transformation of brewing yeasts with
genes for bacterial glucanases has involved the use of multi-copy plasmids
with the inevitable instability problems. However it did seem that instabil-
ity was much greater in transformed ale yeasts than in lager strains
(Cantwell et al., 1985). This was confirmed by Perry and Meaden (1988)
using different plasmids. If this is generally true it supports the hypothesis
that the two yeast types are fundamentally different (see section 3.4.5(j».
In order to ensure adequate expression in transformed yeasts, the
~-glucanase gene from T. reesei was placed in the yeast PGK expression
cassette which was in turn inserted into both multi-copy and integrating
plasmids (Enari et al., 1987). Specific yeast secretion signal sequences were
not required since fungal extracellular enzymes, unlike their bacterial
equivalents, are efficiently excreted by yeasts. Selection of transformants
was based on copper resistance. With both types of transformant, wort
~-glucans were effectively degraded and beer filterability was much im-
proved whilst fermentation and beer characteristics were largely unaltered
(Enari et al., 1987; Penttila et al., 1987). Although the multi-copy plasmid
constructs were more unstable than the integrated transformants, they
produced considerably higher f3-glucanase activities.
The ~-glucanase from barley, which is normally responsible for ~-glucan
66 Yeast genetics
degradation during brewing, shows highest activity at a pH of 4.7, in
contrast to the bacterial enzyme which is most active at pH 6.7. It is thus
better suited to the conditions found during beer production and is more
appropriate for introduction into brewing strains. In order to clone the gene
for this enzyme, two separate DNA sequences isolated from barley aleu-
rone, which together encompassed the barley ~-glucanase gene, were
joined together. The gene was then fused to a mouse a-amylase secretion
signal sequence and inserted into an expression vector between yeast
ADHl promoter and terminator sequences (Jackson, Ballance and
Thomsen, 1986). Subsequently an integrating vector was constructed along
the same lines as that used for expression of A. niger glucoamylase (see
Fig. 3.5). Using resistance to G418 as the selection marker, a number of yeast
transformants were produced and all produced extracellular ~-glucanase
(Thomsen, Jackson and Brenner, 1988). A similar integrative transformant
of a brewing yeast has been used for pilot-scale beer production. The
~-glucan content of the beer was much lower than normal, flavour and
foam stability were unaffected but filtration rates were markedly increased
(Berghof and Stahl, 1991).
The cloning of the barley gene opens up a host of possibilities outside
yeast transformation. For instance, the gene could be changed in vitro to
produce a more heat-stable enzyme which could then be reintroduced into
barley. The new enzyme would be more able to survive the temperatures
found during malting and mashing and would probably eliminate the need
for f3-glucanase-producing yeasts. The techniques for the manipulation of
barley in this way have only recently been developed (Mannonen et al.,
1993) and, as yet, no barley plants producing recombinant ~-glucanases
have been reported.

Other carbohydrases
Other carbohydrase genes which have been expressed in S. cerevisiae in-
clude cellobiohydrolases from T. reesei (Penttila et al., 1988), a-amylase from
A. niger (Knowles and Tubb, 1986), from Schwanniomyces occidentalis
(Strasser et al., 1989), from wheat (Lancashire, 1986), from rice (Kumagai et
al., 1990) and from barley (Sogard and Svensson, 1990), a-glucosidase from
Candida tsukubaensis (Kinsella and Cantwell, 1991) and melibiase from
S. uvarum (Post-Beittenmiller et al., 1984; Liljestrom, 1985). The last enzyme,
normally found only in lager yeasts, has been expressed in ale yeasts (Tubb
et al., 1986), thus enabling the melibiase pasteurization test (Enevoldsen,
1981,1985) to be applied to ales as well as lagers.

(j) Brewing yeasts with proteolytic activity

So far the least successful application of recombinant DNA methods to
yeast has been in the quest for proteolytic yeast strains. There is a need to
reduce the level of polypeptides in beer to prevent the formation of hazes
 Genetic techniques in brewing yeast strains 67
on chilling and during storage. Such haze stabilization is conventionally
carried out using papain. A gene for an extracellular protease has been
cloned and inserted into a centromeric plasmid (Young and Hosford, 1987).
Transformation of a lager yeast strain was successfully carried out using
G418 resistance as the selection procedure. However, although the fermen-
tation characteristics of the new yeast were acceptable, the flavour of the
beer produced was not satisfactory (Sturley and Young, 1988). The intro-
duction of proteolytic activity into yeast would also raise concerns about
the hydrolysis of foam-promoting polypeptides in wort and beer.

(g) Modification offlocculation properties of brewing yeasts

The flocculation of brewing yeasts is a complex phenomenon and the
genetics involved are far from clear (Calleja, 1987). A number of genes
(FL01, FL05, FLOB and tupl) have been identified as being associated with
flocculation but their presence in brewing yeasts is still the subject of
research interest. Stratford and Assinder (1991) examined 42 flocculent
yeast strains and concluded that brewing yeasts showed two types of
flocculation behaviour. The flocculation behaviour of bottom-fermenting
lager strains and laboratory strains, containing the known flocculation
genes, was clearly different from that of top-fermenting ale strains, suggest-
ing that flocculation of ale strains may not be caused by the so-far identified
FLO genes. Hybridization studies support the probable involvement of
FLOl in the flocculation of bottom-fermenting yeasts (Watari et al., 1991a).
The FLOl gene product has been tentatively identified as a hydrophobic
cell wall protein (van der Aar, Straver and Teunissen, 1993).
The FLOl gene has been cloned (Watari et al., 1989), inserted into a
multi-copy plasmid and used to transform both top- and bottom-
fermenting non-flocculent yeasts. The transformed yeasts were more floc-
culent than their parents (Watari et al., 1991b) but since the character was
encoded on a multi-copy plasmid the new property was easily lost. Never-
theless the transformed yeasts fermented satisfactorily (Wa tari et al., 1991a).
The same gene has, more recently, been integrated into the genome of a
non-flocculent bottom-brewing yeast (Watari et al., 1994) and the modified
yeast stably expressed flocculence much more strongly than did its parent
strain.

(h) Yeasts with anti-contaminant properties

The breeding of killer yeasts by cytoduction (section 3.4.3(b» has been
superseded by another technique in which dsRNA is isolated from killer
yeasts and inserted directly into protoplasts of S. cerevisiae by means of
electroinjection (Salek, Schnettler and Zimmermann, 1990). The trans-
formants stably maintain their newly obtained killer activity through many
generations.
68 Yeast genetics
Killer wine yeasts have also been produced by more normal transforma-
tion procedures. A DNA copy of the K1 killer gene was cloned and placed
into an ADHl expression cassette which was then integrated into the yeast
genome. The parent wine yeasts used for these experiments already con-
tained K2 dsRNA and so the resulting transformants expressed both K1
and K2 killer factors (Boone et al., 1990). Such double killer strains cannot
normally exist since the K1 and K2 dsRNA molecules appear to be mutually
incompatible (Wickner, 1981). The transformants were stable and they
fermented grape must in exactly the same manner as the parent yeast. Such
transformants have not, as yet, been made from brewing yeasts but the
same protocol should be equally applicable.

(i) Elimination of the necessity for flavour maturation of beer

Vicinal diketones, especially diacetyl, are a major off-flavour in finished
beers. Diacetyl is formed during fermentation by the chemical oxidation of
a-acetolactate, which diffuses from yeast cells into the fermenting wort. Its
presence in green beer is the major reason for an expensive, time-
consuming flavour maturation period at the end of fermentation. In order
to accelerate beer production, various genetic approaches have been used
to reduce or eliminate diacetyl formation. Since the conversion of
a-acetolactate to diacetyl is the rate limiting step, the best approaches to
this problem have involved either preventing a-acetolactate formation or
removing it rapidly before it is converted to diacetyl.
The enzyme acetolactate decarboxylase (ALDC), which occurs in a
number of bacteria, catalyses the conversion of a-acetolactate to acetoin,
thus preventing its oxidation to diacetyl. The gene for this enzyme has been
cloned from Enterobacter aerogenes (Sone et al., 1987, 1988), from Klebsiella
terrigena (Blomqvist et al., 1991), from Lactococcus lactis (Goelling and Stahl,
1988), fromAcetobacterpasteurianus (U. Stahl, personal communication) and
from Acetobacter aceti var. xylinum (Yamano, 1994). In initial experiments
the E. aerogenes gene was expressed off the yeast ADHl promoter on a
multi-copy plasmid and this was used to transform brewing strains. In
fermentation trials a considerable reduction in the level of diacetyl was
observed when the modified yeast was used in both batch and continuous
fermentations (Shimizu, Sone and Inoue, 1989). There was no effect on other
aspects of yeast metabolism since the ALDe made by the yeast was located
in the cytoplasm. Production of a-acetolactate occurs in yeast mitochondria
and so only that material which escaped into the cytoplasm was converted
to acetoin. Any a-acetolactate required for valine synthesis remained in the
mitochondria where it was unaffected by the decarboxylase enzyme.
Subsequently the ALDe gene was integrated into the ribosomal DNA
(rDNA) of yeast. Because of the tandem repeat nature of rDNA more than
20 copies of the ALDe gene were integrated and all transformants showed
a high enzyme activity (Fujii et al., 1990). Because the new genes were
 Genetic techniques in brewing yeast strains 69
integrated, the transformants were extremely stable even under non-
selective conditions. Again beers from laboratory-scale fermentation trials
exhibited very low final diacetyl concentrations. In an extension of this
work, cloned ALDC genes from two strains of E. aerogenes and from one
strain of K. terrigena were placed under the control of both ADHl and PGK
promoters and terminators and were used to transform brewing yeasts by
integration into the rDNA, ADHl and PGK loci. All the transformants
fermented well and in the case of two of them (both integrated at the PGK
site and under the control of the PGK promoter and terminator) no matu-
ration was necessary since the diacetyl concentration in the beer at the end
of primary fermentation was below the taste threshold. Even the beers
produced by the other strains had markedly reduced diacetyllevels. The
quality of the finished beer was uniformly good and the transformed yeasts
were completely stable (Suihko et al., 1989, 1990; Blomqvist et al., 1991). For
food approval purposes the ALDC genes from L. lactis or Acetobacter sp.
would be preferable to those from other sources since these organisms are
already used for food production (Goelling and Stahl, 1988). Since it occurs
as part of the natural flora of Lambic beer, similar status has been claimed
for E. aerogenes (Blomqvist et al., 1991).
Because of the importance of diacetyl for beer flavour, much research
effort has been concentrated on the amino acid biosynthetic pathway
leading to the formation of ex-acetolactate (Fig. 3.6). The enzymes of the
pathway are regulated by quite complex control mechanisms, in particular
acetolactate synthase is subject to very strong feedback inhibition by valine
(de Robichon-Szulmajster and Magee, 1968). Since the pathway has been
studied in considerable detail and the IL V genes coding for the various
enzymes have been cloned (Petersen et al., 1983), there have been a number
of attempts to control the level of diacetyl in beer by manipulating these
genes in brewing yeasts. Diacetyl production can be completely eliminated
in mutants lacking threonine deaminase and acetolactate synthase but,
because of their inability to synthesize a number of amino acids including
valine, such yeasts ferment poorly (Ryder and Masschelein, 1983). Partial
reduction in acetolactate synthase activity can be achieved by the use of
dominant mutation. Sulphometuron methyl inhibits acetolactate synthase
activity and dominant mutations of the IL V2 gene are known which bring
about resistance to the herbicide by reducing the binding affinity of the
enzyme for the herbicide (Falco and Dumas, 1985). Some of the resistant
forms of the enzyme are less active. Spontaneous dominant mutants of this
type have been isolated but reductions in diacetyllevels were disappoint-
ingly small (Galvan et al., 1987; Gjermansen et al., 1988). More recently, a
procedure has been devised whereby recessive mutations in the IL V2locus
of brewing yeasts can be produced (Kielland-Brandt et al., 1989). This
involves mutagenic treatment of spontaneous herbicide-resistant meiotic
segregants from brewing yeasts followed by screening for slow growth on
media lacking isoleucine and valine. In this way strains have been
70 Yeast genetics

1
Pyruvate

ILV2 Acetolactate synthase

!
a-Acetolactate ., Diacetyl

ILV5 Aeductolsomecase

Dihydroxy isovalerate

ILV3 1 Olhydroxyacld dehydcatase

a-Ketoisovalerate

1 Tcansamlnase

Valine

Fig. 3.6 Enzymes and genes of the valine synthetic pathway in Saccharomyces
cerevisiae.

produced which have very low acetolactate synthase activities. Mating of

these strains has yielded hybrids that ferment well, produce little diacetyl
and make acceptable beer (P. Sigsgaard, personal communication).
Another way of reducing the activity of acetolactate synthase has been
described by Vakeria, Box and Hinchliffe (1991). A fragment of the ILV2
gene was used to transform a brewing yeast in such a way that antisense
mRNA was made. This had the effect of reducing the level of normal
acetolactate synthase mRNA and hence the amount of enzyme synthesized.
Although the amount of diacetyl produced during fermentation was less,
the transformed yeast fermented wort much more slowly.
An alternative approach to reducing diacetyllevels is to increase the
flux through the pathway leading to valine production. This has been
achieved by transforming yeasts with multi-copy plasmids containing
copies of IL V3 and IL VS genes. Yeasts transformed with IL VS showed an
increase in acetolactate reductoisomerase activity and a considerable de-
crease in diacetyllevels (Dillemans et al., 1987; Villanueba, Goossens and
Masschelein, 1990; Goossens et al., 1991, 1993) whereas those transformed
with IL V3 displayed no effect on diacetyl level despite a considerable
increase in the corresponding enzyme activity (Goossens et al., 1987).
Another way of increasing the copy number of the IL VS gene has been
 Genetic techniques in brewing yeast strains 71

suggested by Gjermansen et al. (1988). Using a gene replacement tech-

nique, functional IL V5 genes can be stably integrated at the IL V2 locus
with the simultaneous loss of ILV2. In this way the level of acetolactate
synthase would be reduced at the same time as the level of
reductoisomerase is increased.

(j) Molecular biological approaches to yeast differentiation

Traditionally, yeasts strains have been differentiated by a mixture of
biochemical, morphological and physiological parameters. With the ad-
vent of recombinant DNA techniques, it is now possible to analyse yeast
DNA in detail and to distinguish strains on the basis of some form of DNA
fingerprint.
Probably the most widely applied technique is that which involves
digestion of DNA using specific restriction endonucleases, separation of
the fragments by agarose gel electrophoresis and then either direct exami-
nation of the banding patterns or examination of the patterns produced by
hybridizing labelled 'probes' to the DNA fragments. The probes are them-
selves short lengths of DNA labelled with either a radioactive marker
(typically 32p) or some other marker which can be detected easily. Non-
radioactive methods are becoming more popular especially as the tech-
nique is now being applied in routine quality control applications (for
further details see Meaden (1990». Direct examination of mitochondrial
DNA, without the use of a labelled probe, can be used to differentiate ale
yeasts from lager yeasts (Martens, van den Berg and Harteveld, 1985; Lee,
Knudsen and Poyton, 1985; Casey, Pringle and Erdmann, 1990) but is
unable to distinguish between strains of lager yeasts (Aigle, Erbs and Moll,
1984; Martens, van den Berg and Harteveld, 1985; Casey, Pringle and
Erdmann, 1990) except for flocculation variants (Donhauser, Vogeser and
Springer, 1989). Direct examination of total DNA after cutting with the
restriction enzyme HpaI has also been used to distinguish ale from lager
yeasts (Panchal et al., 1987). Normally, however, the probing approach has
to be used.
A detailed analysis of a wide range of industrial yeasts has been carried
out using labelled probes. With an RDNl probe three basic patterns were
obtained but almost all of the brewing yeasts examined produced the same
pattern (Pedersen, 1983a,b). In contrast, with a HIS4 probe, ale yeasts could
be distinguished from lager strains (Pedersen, 1983a,b). Similar results were
obtained using a LEU2 probe although in this case great heterogeneity was
observed amongst the ale strains (Pedersen, 1985a,b, 1986a). In general the
results obtained with lager yeasts were so homogeneous as to suggest that
they are all closely related, whereas the ale strains showed considerable
variability. This general conclusion is further supported by work involving
the use of the Tyl transposable element as a probe. The Tyl element is found
distributed about the yeast chromosomal DNA at about 30-35 copies per
72 Yeast genetics
haploid genome, and would seem to be a good candidate for DNA
fingerprinting since changes in the number and location of these elements
should lead to alteration in the restriction patterns observed. Great hetero-
geneity was observed amongst ale strains (Pedersen, 1986a), whilst differ-
entiation of lager yeasts was possible with some Tyl probes (Decock and
Iserentant, 1985; Martens, van den Berg and Harteveld, 1985; Pedersen,
1985a,b, 1986a) but not with another such probe (Aigle, Erbs and Moll, 1984;
Decock and Iserentant, 1985). The heterogeneity of ale strains has been
confirmed using a CUPl probe and moreover this has been used to distin-
guish two very closely related lager yeasts (Taylor, Hammond and Meaden,
1990). The DEXl (STA2) gene has also proved extremely useful for analys-
ing brewing yeasts since it cross reacts with other related sequences found
in brewing yeasts to produce quite complex patterns which can provide a
DNA fingerprint (Taylor, Hammond and Meaden, 1990). Even more com-
plex fingerprints are produced using synthetic polyGT probes, and these
have been used to differentiate a number of industrial yeast strains
(Walmsley, Wilkinson and Kong, 1989).
Another technique which has been used to investigate yeast genome
structure is that of pulsed field electrophoresis of whole chromosomes. This
method (karyotyping) has not been widely applied to brewing yeasts but
the work carried out so far suggests that it will prove to be a powerful tool,
especially for distinguishing lager strains. Initial work in this field
Oohnston and Mortimer, 1986; Pedersen, 1986b, 1987; Takata et al., 1989)
used orthogonal field-alternation gel electrophoresis (OFACE; Carle and
Olsen, 1984, 1985) which yields curved tracks and so makes interpretation
of the results difficult. Nevertheless, distinct karyotypes were distin-
guished among the brewing yeasts examined. More recently, contour-
clamped homogeneous electric field electrophoresis (CHEF; Chu, Vollrath
and Davis, 1986) and transverse alternating pulsed field electrophoresis
(TAFE; Gardiner, Laces and Patterson, 1986) which yield straight tracks
have found widespread acceptance. By careful examination of patterns
obtained using these methods, either by densitometry (Casey, Xiao and
Rank, 1988b) or by computer image processing (Pedersen, 1989), it has
proved possible to distinguish different strains of brewing yeasts. Subse-
quent work has refined the techniques used, particularly the speed at which
the chromosome samples can be prepared (Sheehan et al., 1991) and the
resolution of the bands obtained in the electrophoresis gel (Casey, Pringle
and Erdmann, 1990). It is clear from a detailed examination of pulsed field
electrophoresis gels of brewing yeasts that they contain many more bands
than do gels from laboratory strains and that the intensity of the bands
varies. This is presumably a reflection of the polyploid nature of the yeasts
and their possession of a mixture of different but similar chromosomes (see
section 3.4.3(c)). This great number and variety of bands enables karyo-
typing to be used for distinguishing many closely related industrial yeast
strains (Casey, Pringle and Erdmann, 1990; Sheehan et al., 1991).
 Commercial use of genetically modified brewing yeasts 73
The techniques of genetic manipulation may also provide the means
whereby unique selectable markers can be introduced into individual yeast
strains so that the strains can be distinguished by simple laboratory tests.
Alternatively small changes can be made to the genome of a brewing yeast
which have no effect whatever on its behaviour but which can be detected
by means of the DNA fingerprinting techniques described above.

3.5 THE COMMERCIAL USE OF GENETICALLY MODIFIED

BREWING YEASTS

3.5.1 Safety considerations

When genetic modification methods were first developed the scientists
involved raised a number of concerns about the technology (Berg et al.,
1974). These included the colonization of the environment by genetically
modified organisms, the unwanted transfer of genes to other organisms in
the wild and the possible inadvertent production of new pathogenic organ-
isms. With the passage of time it has generally been agreed that these
potential problems were much exaggerated. However, as a result of the
expression of these concerns, a diversity of legislation has developed
worldwide to address such issues. These regulations are mainly concerned
with laboratory-scale work and, although differing in detail, are broadly
similar in most countries. They recognize the need for different laboratory
facilities for work with different genetically modified organisms, the degree
of containment required varying with the perceived risk. All genetic mod-
ification work carried out with brewing yeasts has always fallen into the
lowest containment category, merely requiring the facilities which are
found in any good microbiological laboratory.
Until 1993, regulations applying to genetically modified organisms were
concerned solely with human health and safety. Now environmental pro-
tection also has to be considered. In the United Kingdom, all work with
genetically modified organisms is controlled by two sets of regulations: the
Genetically Modified Microorganisms (Contained Use) Regulations, 1992
and the Genetically Modified Organisms (Deliberate Release) Regulations,
1992. These apply to the construction, use and release of genetically modi-
fied organisms and involve prior notification of an intention to carry out
genetic modification work, together with risk assessments (further details
can be found in Hammond and Bamforth, 1994).

3.5.2 Novel foods approval

In the United Kingdom, before beers made with genetically modified yeasts
can be produced on a large scale and sold commercially, the approval of
several Government committees is required. The primary approval must
74 Yeast genetics
come from the Advisory Committee on Novel Foods and Processes
(ACNFP) which reports to the Ministry of Agriculture, Fisheries and Food
(MAFF). Guidelines have been issued to enable manufacturers to decide
whether any new product comes within the category of a 'novel food' and,
if so, what information will be required by ACNFP for their consideration
(Department of Health, 1991). It is clear from this document that beers made
from genetically modified yeasts do fall into the category of a 'novel food'.
The information requirements include data on the genetic procedures used
and the properties of the modified strain including its stability and
toxicology.
As part of their assessment, ACNFP may consult with two other com-
mittees, the Advisory Committee on Genetic Modification (ACGM) of the
Health and Safety Executive who provide advice on all aspects of human
health and safety, and the Advisory Committee on Release to the Environ-
ment (ACRE) of the Department of Environment who provide advice on
waste disposal and incidental release of organisms in fermenter off-gases.
Following these consultations, ACNFP makes the decision as to whether
the new yeast can be used commercially to make beer.
Novel food approval has recently been obtained for the commercial use
of a brewer's yeast genetically modified with the STA2 gene to produce
extracellular glucoamylase. This is the first such permission that has been
granted anywhere in the world. Batches of beer have been made with the
new yeast and distributed to interested parties worldwide (Hammond, 1994).
At the time of writing (November 1994) the novel food approval system
is voluntary although the Food Safety Act (Ministry of Agriculture, Fisher-
ies and Food, 1990) does contain provisions that will enable the scheme to
be put on a statutory basis. However, this is unlikely to happen until
parallel EC novel food regulations are approved. The current state of affairs
in Europe has been described recently (Hammond and Bamforth, 1994), as
have the likely effects of European legislation when compared with the
situation in the USA (Hammond, 1991).

3.6 CONCLUSIONS

Since the previous edition of this book was written there have been huge
advances in the application of modem genetic methods to brewing yeasts.
The structure of the genome of brewing yeasts is being slowly unravelled
and details of the control of metabolic processes are beginning to be
understood. Additionally, the techniques of modem yeast genetics have
been applied to the construction of new yeast strains for the more efficient
production of beer. Many of these novel yeasts have been assessed at the
pilot brewery scale and have been technically successful. One such strain
has now been approved for the production of beer for consumption by
the general public but has not, as yet, been used commercially. This need
 References 75
for approval is a major barrier to commercial exploitation. The rigorous
'novel food' requirements are clearly a response to perceived consumer
concerns about the risk of genetically modified organisms. These same
perceptions have also led to proposals that foods produced with geneti-
cally modified organisms should be labelled 'Products of Gene
Technology' (Food Advisory Committee of MAFF, quoted as an appendix
in Department of Health, 1991). Pressure groups and administrators be-
tween them are making the application of recombinant DNA technology
in the food and drink sector very difficult. From numerous surveys it is
clear that the average consumer is singularly ill-informed about the ben-
efits and risks of biotechnology. It is essential that the food and drink
sector addresses this problem and educates the public about the factors
involved in the use of genetically modified organisms for food and drink
production. There is a clear need to convince public opinion that the risks
are small, before beer manufactured with genetically modified yeasts will
become generally acceptable. All food and drink biotechnologists must
devote time and effort to ensuring that any risks are seen in the correct
perspective and are not wildly exaggerated by pressure groups. Otherwise
the application of novel biotechnology in the food and drink sector will
be severely curtailed for many years and the significant available benefits
will be lost.

ACKNOWLEDGEMENTS

The author thanks Dr C.W. Bamforth for critical reading of the manuscript
and the Director General of BRF International for permission to publish.

REFERENCES

Aigle, M., Erbs, D. and Moll, M. (1984) Journal of the American Society of Brewing
Chemists, 42, l.
Anderson, E. and Martin, P .A. (1975) Journal of the Institute of Brewing, 81, 242.
Barney, M.C, Jansen, G.P. and Helbert, J.R. (1980) Journal of the American Society of
Brewing Chemists, 38, l.
Becker, D.M. and Guarente, L. (1991) In Methods in Enzymology, Vol. 194 (eds
C Guthrie and G. Fink), Academic Press, London, p. 182.
Beggs, J.D. (1981) In Molecular Genetics in Yeast, Alfred Benzon Symposium, Vol. 16
(eds D. von Wettstein, A. Stenderup, M. Kielland-Brandt and J. Friis),
Munksgaard, Copenhagen, p. 383.
Berg, P., Baltimore, D., Boyer, H.W. et al. (1974) Proceedings of the National Academy
of Sciences, USA 71,2593.
Berghof, K. and Stahl, U. (1991) BioEngineering, 7, 27.
Bilinski, CA., Russell, 1. and Stewart, G.G. (1986) Journal of the Institute of Brewing,
92,594.
Bilinski, CA., Russell, 1. and Stewart, G.G. (1987a) Journal of the Institute of Brewing,
93,216.
76 Yeast genetics
Bilinski, CA, Russell, I. and Stewart, G.G. (1987b) Proceedings of the 21st Congress of
the European Brewery Convention, Madrid, IRL Press, Oxford, p. 497.
Blomqvist, K., Suihko, M.-L., Knowles, J. and Penttila, M. (1991) Applied and
Environmental Microbiology, 57, 2796.
Boone, C, Sdicu, A-M., Wagner, J. et al. (1990) American Journal of Enology and
Viticulture, 41,37.
Broach, J.R (1981) In The Molecular Biology of the Yeast Saccharomyces, Life Cycle and
Inheritance (eds J.N. Strathem, E.W. Jones andJ.R Broach), Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY, p. 445.
Brown, S.W. and Oliver, S.G. (1982) European Journal of Applied Microbiology and
Biotechnology, 16, 119.
Calleja,G.B. (1987) In The Yeasts (edsAH. RoseandJ.5. Harrison), Vol. II,Academic
Press, London, p. 165.
Cantwell, B.A. and McConnell, D.J. (1983) Gene, 23,211.
Cantwell, B., Brazil, G., Hurley, J. and McConnell, D. (1985) Proceedings of the 20th
Congress of the European Brewery Convention, Helsinki, IRL Press, Oxford p. 259.
Carle, G.F. and Olsen, M.V. (1984) Nucleic Acids Research, 12, 5647.
Carle, G.F. and Olsen, M.v. (1985) Proceedings of the National Academy of Sciences,
USA, 82, 3756.
Casey, G.P. (1986a) Carlsberg Research Communications, 51,327.
Casey, G.P. (1986b) Carlsberg Research Communications, 51, 343.
Casey, G.P. and Pedersen, M.B. (1988) Carlsberg Research Communications, 53,209.
Casey, G.P., Pringle, AT. Erdmann, P.A. (1990) Journal of the American Society of
Brewing Chemists, 48, 100.
Casey, G.P., Xiao, W. and Rank, G.H. (1988a) Journal of the Institute of Brewing, 94, 93.
Casey, G.P., Xiao, W. and Rank, G.H. (1988b) Journal of the Institute of Brewing,
94,239.
Casey, G.P., Xiao, W. and Rank, G.H. (1988c) Journal ofthe American Society ofBrewing
Chemists, 46, 67.
Chu, G., Vollrath, D. and Davis, RW. (1986) Science, 234,1582.
Chinery, S.H. and Hinchliffe, E. (1989) Current Genetics, 16, 21.
Clarke, L. and Carbon, J. (1979) In Methods in Enzymology, Vol. 68 (ed. R Wu),
Academic Press, New York, p. 396.
Conde, J. and Mascort, J.L. (1981) Proceedings of the 18th Congress of the European
Brewery Convention, Copenhagen, IRL Press, Oxford, p. 177.
Crump len, RM., D'Amore, T., Russell, I. and Stewart, G.G. (1990) Journal of the
American Society of Brewing Chemists, 48,58.
Debourg, A, Goossens, E., Villanueba, K.D. et al. (1991) Proceedings of the 23rd
Congress of the European Brewery Convention, Lisbon, IRL Press, Oxford, p. 265.
Decock, J.P. and Iserentant, D. (1985) Proceedings of the 20th Congress of the European
Brewery Convention, Helsinki, IRL Press, Oxford, p. 195.
de Figueroa, L.I.C and de van Broock, M.RG. (1985) Applied Microbiology and
Biotechnology, 21, 206.
Delgado, M.A and Conde, J. (1983) Proceedings of the 19th Congress of the European
Brewery Convention, London, p. 465.
Department of Health (1991) Guidelines on the Assessment o/Novel Foods and Processes,
Report on Health and Social Subjects, No. 38, HMSO, London.
de Robichon-Szulmajster, H. and Magee, P.T. (1968) European Journal ofBiochemistry,
3,497.
de van Broock, M.R, Sierra, M. and de Figueroa, L. (1981) In Current Developments
in Yeast Research (eds G.G. Stewart and I. Russell), Pergamon Press, Toronto,
p.171.
Dillemans, M., Goossens, E., Goffin, O. and Masschelein, CA (1987) Journal of the
American Society of Brewing Chemists, 45,81.
 References 77
Dohmen, RJ., Strasser, A.W.M., Dahlems, U.M. and Hollenberg, CP. (1990) Gene,
95,111.
Donhauser, S., Vogeser, G. and Springer, R (1989) Monatsschrift fur Brauwissenschaft,
42,4.
Dujon, B. (1981) In The Molecular Biology of the Yeast Saccharomyces, Life Cycle and
Inheritance (eds J.N. Strathem, E. W. Jones and J.R Broach), Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY, p. 505.
Dutcher, S.K. (1981) Molecular and Cellular Biology, 1, 245.
Enari, T.M., Knowles, J., Lehtinen, U. et al. (1987) Proceedings of the 21st Congress of
the European Brewery Convention, Madrid, IRL Press, Oxford, p. 529.
Enevoldsen, B.s. (1981) Carlsberg Research Communications, 46, 37.
Enevoldsen, B.s. (1985) Journal of the American Society of Brewing Chemists, 43, 183.
Esposito, RE. and Klapholz, S. (1981) In The Molecular Biology of the Yeast
Saccharomyces, Life Cycle and Inheritance (eds IN. Strathem, E.W. Jones and
J.R Broach), Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, p. 211.
Esser, K and Kamper, J. (1988) Process Biochemistry, April, 36.
Evans, LH. and Wilkie, D. (1976) Genetic Research, 27, 89.
Falco, S.C and Dumas, KS. (1985) Genetics, 109,21.
Fangman, W.L. and Zakian, V.A. (1981) In The Molecular Biology of the Yeast
Saccharomyces, Life Cycle and Inheritance (eds IN. Strathem, E.W. Jones and
J.R Broach), Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, p. 27.
Fink, G.R and Styles, CA. (1973) Proceedings ofthe National Academy ofSciences, USA,
69,2846.
Freeman, RF. (1981) Proceedings of the 18th Congress of the European Brewery
Convention, Copenhagen, IRL Press, Oxford, p. 497.
Fujii, T., Kondo, K, Shimizu, F. et al. (1990) Applied and Environmental Microbiology,
56,997.
Galvan, L., Perez, A., Delgado, M. and Conde, J. (1987) Proceedings ofthe 21st Congress
of the European Brewery Convention, Madrid, IRL Press, Oxford, p. 385.
Gardiner, K, Laces, W. and Patterson, D. (1986) Somatic Cell and Molecular Genetics,
12,185.
Gjermansen, C (1983) Carlsberg Research Communications, 48, 557.
Gjermansen, C and Sigsgaard, P. (1981) Carlsberg Research Communications, 46, 1.
Gjermansen, C, Nilsson-Tillgren, T., Petersen, J.G.L. et al. (1988) Journal of Basic
Microbiology, 28, 175.
Goelling, D. and Stahl, U. (1988) Applied and Environmental Microbiology, 54,1889.
Goodey, A.R and Tubb, RS. (1982) Journal of General Microbiology, 128, 2615.
Goodey, A.R, Brown, A.J.P. and Tubb, RS. (1981) Journal of the Institute of Brewing,
87,239.
Goossens, E., Dillemans, M., Debourg, A. and Masschelein, CA. (1987) Proceedings
of the 21st Congress of the European Brewery Convention, Madrid, IRL Press,
Oxford, p. 553.
Goossens, E., Debourg, A., Villanueba, KD. and Masschelein, CA. (1991)
Proceedings of the 23rd Congress of the European Brewery Convention, Lisbon, IRL
Press, Oxford, p. 289.
Goossens, E., Debourg, A., Villanueba, KD. and Masschelein, CA. (1993)
Proceedings of the 24th Congress of the European Brewery Convention, Oslo, IRL
Press, Oxford, p. 251.
Gopal, CV. and Hammond, J.RM. (1992) Proceedings of the 5th International Brewing
Technology Conference, Harrogate, Brewing Technology Services Ltd, London,
p.297.
Guinova-Stoyanova, T.A. and Lahtchev, KL. (1991) Proceedings of the 23rd Congress
of the European Brewery Convention, Lisbon, IRL Press, Oxford, p. 313.
Gunge, N. and Nakatomi, Y. (1972) Genetics, 70, 41.
78 Yeast genetics
Guthrie, C. and Fink, G.R (eds) (1991) Methods in Enzymology, Vol. 194, Academic
Press, London.
Gyllang, H. and Martinson, E. (1971), Proceedings of the 13th Congress of the European
Brewery Convention, Estoril, IRL Press, Oxford, p. 265.
Hadfield, c., Cashmore, A.M. and Meacock, P.A. (1986) Gene, 45,149.
Halfmann, H.J., Rocken, W., Emeis, c.c. and Zimmermann, U. (1982) Current
Genetics, 6,25.
Hammond, J.RM. (1991) Proceedings of the 23rd Congress of the European Brewery
Convention, Lisbon, IRL Press, Oxford, p. 393.
Hammond, J.RM. (1995) In Proceedings of the 4th Aviemore Conference on Malting,
Brewing and Distilling (eds EG. Priest and I. Campbell), Institute of Brewing,
London, in press.
Hammond, J. and Bamforth, C. (1994) The Brewer, Feb., 65.
Hammond, J.RM. and Eckersley, KW. (1984) Journal of the Institute of Brewing,
90,167.
Hammond, J.RM. and Wenn, RV. (1985) Proceedings of the 20th Congress of the
European Brewery Convention, Helsinki, IRL Press, Oxford, p. 315.
Hansen, M., Rocken, W. and Emeis, c.c. (1990) Journal of the Institute of Brewing,
96,125.
Hartley, J.L., and Donelson, J.E. (1980) Nature, 286, 860.
Henderson, RC.A., Cox, B.s. and Tubb, RS. (1985) Current Genetics, 9, 133.
Herskowitz, I. and Oshima, Y. (1981) In The Molecular Biology of the Yeast
Saccharomyces, Life Cycle and Inheritance (eds J.N. Strathem, E.W. Jones and
J.R Broach), Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, p. 181.
Hinchliffe, E. and Box, W.G. (1984) Current Genetics, 8, 471.
Hinchliffe, E. and Box, W.G. (1985) Proceedings of the 20th Congress of the European
Brewery Convention, Helsinki, IRL Press, Oxford, p. 267.
Hinchliffe, E. and Daubney, c.J, (1986) Journal of the American Society of Brewing
Chemists, 44, 98.
Hinchliffe, E., Fleming c.J. and Vakeria, D. (1987) Proceedings of the 21st Congress of
the European Brewery Convention, Madrid, IRL Press, Oxford, p. 505.
Hinchliffe, E., Kenny, E. and Leaker, A. (1987) European Brewery Convention
Monograph, XIII, 139.
Hinnen, A., Hicks, J.B. and Fink, G.R (1978) Proceedings of the National Academy of
Sciences, USA, 75, 1929.
Hitzeman, RA., Hagie, F.E., Levine, H.L. et al. (1981) Nature, 293, 717.
Hockney, RC. and Freeman, RF. (1980) Journal of General Microbiology, 121, 479.
Holmberg, S. (1982) Carlsberg Research Communications, 47, 233.
Ito, H., Fukuda, Y., Murata, K and Kimura, A. (1983) Journal of Bacteriology,
153,163.
Jackson, E.A., Ballance, G.M. and Thomsen, KK. (1986) Carlsberg Research
Communications, 51, 445.
Jansen, G.P., Barney, M.C., Helbert, J.R and Esposito, M.s. (1979) Journal of Applied
Biochemistry, 1, 369.
Jimenez, A. and Davies, J. (1980) Nature, 287,869.
Johnston, J.R. and Mortimer, R.K (1986) International Journal of Systematic
Bacteriology, 36, 569.
Johnston, J.R. and Oberman, H. (1979) Progress in Industrial Microbiology, 15, 151.
Jones, RM., Russell, I. and Stewart, G.G. (1986) Journal of the American Society of
Brewing Chemists, 44, 161.
Khan, N.C. and Sen, S.P. (1974) Journal of General Microbiology, 83, 237.
Kielland-Brandt,M.C.,Peterson,J.G.L. and Mikkelsen, J.D. (1979) Carlsberg Research
Communications, 44, 27.
Kielland-Brandt, M.C., Gjermansen, c., Nilsson-Tillgren, T. and Holmberg, S. (1989)
 References 79
Proceedings of the 22nd Congress of the European Brewery Convention, Zurich, IRL
Press, Oxford, p. 37.
Kinsella, B.T. and Cantwell, B.A. (1991) Yeast, 7, 445.
Knowles, J.KC and Tubb, RS. (1986) European Brewery Convention Monograph,
XII, 169.
Knowles, J.K.C, Penttila, M., Teeri, T.T. et al. (1985) Proceedings of the 20th Congress
of the European Brewery Convention, Helsinki, IRL Press, Oxford, p. 25I.
Korhola, M. (1983) In Gene Expression in Yeast, Proceedings ofthe Alko Yeast Symposium
(eds M. Korhola and E. Vaisanen), Foundation for Biotechnical and Industrial
Fermentation Research, Helsinki, p. 23I.
Kreil, H., Kleber, W. and Teuber, M. (1975) Proceedings of the 15th Congress of the
European Brewery Convention, Nice, IRL Press, Oxford, p. 323.
Kumagai, M.H., Shah, M., Terashima, M. et al. (1990) Gene, 94, 209.
Lampen, J.O. (1968) Antonie van Leeuwenhoek, 34, I.
Lancashire, W.E. (1986) The Brewer, 72,345.
Lancashire, W.E. and Wilde, RJ. (1987) Proceedings of the 21st Congress ofthe European
Brewery Convention, Madrid, IRL Press, Oxford, p. 513.
Lancashire, W.E., Carter, A.T., Howard, J.J. and Wilde, RJ. (1989) Proceedings of the
22nd Congress of the European Brewery Convention, Zurich, IRL Press, Oxford, p.
49I.
Lee, S.Y., Knudsen, F.B. and Poyton, RO. (1985) Journal of the Institute of Brewing,
91,169.
Lewis, CW., Johnston, J.R and Martin, P.A. (1976) Journal of the Institute of Brewing,
82,158.
Liljestrom, P.L. (1985) Nucleic Acids Research, 13, 7257.
Lindegren, CC and Lindegren, G. (1943) Proceedings of the National Academy of
Sciences, USA, 29, 306.
Mahler, H.R and Wilkie, D. (1978) Plasmid, I, 125.
Mannonen, L., Kurten, V., Ritala, A. et al. (1993) Proceedings of the 24th Congress of
the European Brewery Convention, Oslo, IRL Press, Oxford, p. 85.
Martens, P.B., van den Berg, Rand Harteveld, P.A. (1985) Proceedings of the 20th
Congress of the European Brewery Convention, Helsinki, IRL Press, Oxford, p. 21I.
Meaden, P. (1990) Journal af the Institute afBrewing, 96, 195.
Meaden P.G, and Taylor N.R (1991) Journal of the Institute of Brewing, 97, 353.
Meaden, P.G. and Tubb, RS. (1985) Proceedings of the 20th Congress of the European
Brewery Convention, Helsinki, IRL Press, Oxford, p. 219
Meaden, P., Ogden, K, Bussey, H. and Tubb, RS. (1985) Gene, 34, 325.
Ministry of Agriculture, Fisheries and Food (1990) Food Safety Act, HMSO, London.
Molzahn, S.W. (1977) Journal of the American Society of Brewing Chemists, 35, 54.
Montgomery, D.L., Leung, D.W., Smith, M., et al., (1980) Proceedings of the National
Academy of Sciences, USA, 77, 54I.
Morrison, KB. and Suggett, A. (1983) Proceedings of the 19th Congress of the European
Brewery Convention, London, IRL Press, Oxford, p. 489.
Mortimer, RK, Schild, D., Contopoulou, CR and Kans, J.A. (1989) Yeast,S, 32I.
Mowshowitz, D.B. (1979) Journal of Bacteriology, 137,1200.
Navas, L., Esteban, M. and Delgado, M.A. (1991) Journal of the Institute of Brewing,
97,115.
Newlon, CS. and Fangman, W.L. (1975) Cell,S, 423.
Nilsson-Tillgren, T., Petersen,J.G.L., Holmberg, S. and Kielland-Brandt, M.C (1980)
Carlsberg Research Communications, 45, 113.
Nilsson-Tillgren, T., Gjermansen, C, Kielland-Brandt, M.C, et al. (1981) Carlsberg
Research Communications, 46, 65.
Nilsson-Tillgren, T., Gjermansen, C, Holmberg, S., et al. (1986) Carlsberg Research
Communications, 51, 309.
80 Yeast genetics
Oppenoorth, W.F.F. (1959) Proceedings of the 7th Congress of the European Brewery
Convention, Rome, IRL Press, Oxford, p. 29.
Panchal, CJ., Peacock, L. and Stewart, G.G. (1982) Biotechnology Letters, 4, 639.
Panchal, CJ., Bast, L., Dowhanick, T. and Stewart, G.G. (1987) Journal of the Institute
of Brewing, 93, 325.
Park, CS., Park, Y.J., Lee, Y.H., et al. (1990) Technical Quarterly of the Master Brewers
Association of the Americas, 27, 112.
Pedersen, M.B. (1983a) Proceedings of the 19th Congress of the European Brewery
Convention, London, IRL Press, Oxford, p. 457.
Pedersen, M.B. (1983b) Carlsberg Research Communications, 48, 485.
Pedersen, M.B. (1985a) Proceedings of the 20th Congress of the European Brewery
Convention, Helsinki, IRL Press, Oxford, p. 203.
Pedersen, M.B. (1985b) Carlsberg Research Communications, 50,263.
Pedersen, M.B. (1986a) Carlsberg Research Communications, 51, 163.
Pedersen, M.B. (1986b) Carlsberg Research Communications, 51, 185.
Pedersen, M.B. (1987) Proceedings of the 21st Congress of the European Brewery
Convention, Madrid, IRL Press, Oxford, p. 489.
Pedersen, M.B. (1989) Proceedings of the 22nd Congress of the European Brewery
Convention, Zurich, IRL Press, Oxford, p. 521.
Penttila, M.E., Suihko, M.L., Lehtinen, V., et al. (1987) Current Genetics, 12, 413.
Penttila, M.E., Andre, L., Lehtovaara, P., et al. (1988) Gene, 63, 103.
Perez, C, Vallin, C and Benitez, J. (1984) Current Genetics, 8, 575.
Perry, C and Meaden, P. (1988) Journal of the Institute of Brewing, 94, 64.
Petersen, J.G.L., Holmberg S., Nilsson-Tillgren, T., et al. (1983) European Brewery
Convention Monograph, IX, 30.
Petersen, J.G.L., Nilsson-Tillgren, T., Kielland-Brandt, M.C, et al. (1987) Current
Genetics, 12, 167.
Petes, T.D. (1980) Annual Review of Biochemistry, 49, 845.
Philipp sen, P., Eibel, H., Gafner, J. and Stotz, A. (1983) In Gene Expression in Yeast,
Proceedings of the Alko Yeast Symposium (eds M. Korhola and E. Vaisanen),
Foundationfor Biotechnical and IndustrialFermentation Research, Helsinki, p.189.
Post-Beittenmiller, M.A., Hamilton, RW. and Hopper, J.E. (1984) Molecular and
Cellular Biology, 4, 1238.
Rocken, W. (1984) Monatschrift fur Brauwissenschajt, 37, 384.
Russell, I., Hancock, I.F. and Stewart, G.G. (1983) Journal of the American Society of
Brewing Chemists, 41, 45.
Russell, I., Jones, R and Stewart, G.G. (1985) Proceedings of the 20th Congress of the
European Brewery Convention, Helsinki, IRL Press, Oxford, p. 235.
Russell, I., Stewart, G.G., Reader, H.P., et al. (1980) Journal of the Institute of Brewing,
86,120.
Ryder, D.S. and Masschelein, CA (1983) European Brewery Convention, Monograph,
IX,2.
Sakai, K., Fukui, S., Yabuuchi, S., et al. (1989) Journal of the American Society of Brewing
Chemists, 47, 87.
Salek, A, Schnettler, R and Zimmermann, V. (1990) FEMS Microbiological Letters,
70,67.
Sasaki, T., Watari, J., Kohgo, M., et al. (1984) Journal of the American Society of Brewing
Chemists, 42, 164.
Sheehan, CA, Weiss, AS., Newsom, LA, et al. (1991) Journal of the Institute of
Brewing, 97, 163.
Sherman,F. and Wakem,P. (1991) In Methods in Enzymology, Vol. 194 (edsC Guthrie
and G. Fink), Academic Press, London, p. 38.
Shimizu, F., Sone, H. and Inoue, T. (1989) Technical Quarterly of the Master Brewers
Association of the Americas, 26, 47.
 References 81
Sogaard, M. and Svensson, B. (1990) Gene, 94, 173.
Sone, H., Kondo, K., Fujii, T., et al. (1987) Proceedings of the 21st Congress of the
European Brewery Convention, Madrid, IRL Press. Oxford, p. 545.
Sone, H., Fujii, T., Kondo, K., Shimizu, F., Tanaka, J. and Inoue, T. (1988) Applied and
Environmental Microbiology, 54, 38.
Stewart, G.G. and Russell, 1. (1986) European Brewery Convention Mongraph, XII, 53.
Stewart, G.G., Goring, T.E. and Russell, 1. (1977) Journal of the American Society of
Brewing Chemists, 35, 168.
Stewart, G.G., Russell, 1. and Panchal, C (1981) In Current Developments in Yeast
Research (eds G.G. Stewart and 1. Russell), Pergamon Press, Toronto, p. 17.
Steyn, A.J.C and Pretorius, 1.S. (1991) Gene, 100, 85.
Strasser, W.M., Selk, K, Dohmen, KJ., et al. (1989) European Journal of Biochemistry,
184,699.
Stratford, M. and Assinder, S. (1991) Yeast, 7, 559.
Sturley, S.L. and Young, T.w. (1988) Journal of the Institute of Brewing, 94, 133.
Suihko, M.-L., Penttila, M., Sone, H., et al. (1989) Proceedings of the 22nd Congress of
the European Brewery Convention, Zurich, IRL Press, Oxford, p. 483.
Suihko, M.-L., Blomqvist, K., Penttila, M., et al. (1990) Journal ofBiotechnology, 14, 285.
Takata, Y., Watari, J., Nishikawa, N. and Kamada, K. (1989) Journal of the American
Society of Brewing Chemists, 47, 109.
Taylor, N.K, Hammond, J.KM. and Meaden, P.G. (1990) In Proceedings of the 3rd
Aviemore Conference on Malting, Brewing and Distilling (ed. 1. Campbell),
Institute of Brewing, London, p. 403.
Thomsen, K.K., Jackson, E.A. and Brenner, K. (1988) Journal of the American Society
of Brewing Chemists, 46, 31.
Thome, KS.W. (1970) Process Biochemistry, 4, 15.
Thorner, J. (1981) In The Molecular Biology of the Yeast Saccharomyces, Life Cycle and
Inheritance (eds J.N. Stra them, E. W. Jones and J.K Broach), Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY, p. 143.
Tubb, KS. (1980) Journal of the Institute of Brewing, 86, 76.
Tubb, KS., Searle, B.A., Goodey, A.K and Brown A.J.P. (1981) Proceedings of the 18th
Congress of the European Brewery Convention, Copenhagen, IRL Press, Oxford,
p.487.
Tubb, KS., Liljestrom, P.L., Torkkeli, T. and Korhola, M. (1986) In Proceedings of the
2nd Aviemore Conference on Malting, Brewing and Distilling (eds F.G. Priest and
1. Campbell), Institute of Brewing, London, p. 298.
Tuite, M.F., Dobson, M.J., Roberts, N.A., et al. (1982) EMBO Journal, 1, 603.
Urano, N., Nishikawa, N. and Kamimura, M. (1990) Proceedings of the Institute of
Brewing (Australia and New Zealand Section) Convention, Auckland, Australian
Industrial Publishers, Adelaide, p. 154.
Vakeria, D. and Hinchliffe, E. (1989) Proceedings of the 22nd Congress of the European
Brewery Convention, Zurich, IRL Press, Oxford, p. 475.
Vakeria, D., Box, W.G. and Hinchliffe, E. (1991) Proceedings of the 23rd Congress of the
European Brewery Convention, Lisbon, IRL Press, Oxford, p. 305.
van der Aar, P.C, Straver, M.H. and Teunissen, W.KH. (1993) Proceedings of the 24th
Congress of the European Brewery Convention, Oslo, IRL Press, Oxford, p. 259.
van Solingen, P. and van der Plaat, J.B. (1977) Journal of Bacteriology, 130, 946.
Villanueba, KD., Goossens, E. and Masschelein, CA. (1990) Journal of the American
Society of Brewing Chemists, 48, 111.
von Wettstein, D. (1983) Proceedings of the 19th Congress of the European Brewery
Convention, London, IRL Press, Oxford, p. 97.
Walmsley, KM., Wilkinson, B.M. and Kong, T.H. (1989) Biotechnology, 7,1168.
Watari, J., Takata, Y., Ogawa, M., et al. (1989) Agricultural and Biological Chemistry,
53,901.
82 Yeast genetics
Watari, J., Takata, Y., Murakami, J. and Koshino, s. (1991a) Proceedings of the 23rd
Congress of the European Brewery Convention, Lisbon, IRL Press, Oxford, p. 297.
Watari, J., Takata, Y., Ogawa, M., et al. (1991b) Agricultural and Biological Chemistry,
55,1547.
Watari, J., Nomura, M., Sahara, H., et al. (1994) Journal of the Institute of Brewing,
100,73.
Wiekner, RB. (1981) In The Molecular Biology of the Yeast Saccharomyces, Life Cycle and
Inheritance (eds J.N. Strathem, E.w. Jones andJ.R Broach), Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY, p. 415.
Wiekner, RB. (1983) Archives of Biochemistry and Biophysics, 222, 1.
Xiao, W. and Rank, C.H. (1989) Gene, 76,99.
Yamano, 5., Tanaka, J. and Inoue, T. (1994) Journal of Biotechnology, 32,165.
Yocum, RR (1986) Proceedings ofBioExp086, 171.
Young, T.w. (1981) Journal of the Institute of Brewing, 87, 292.
Young, T.w. (1983)Journal of the American Society of Brewing Chemists, 41, 1.
Young, T.W. and Hosford, E.A. (1987) Proceedings of the 21st Congress of the European
Brewery Convention, Madrid, IRL Press, Oxford, p. 521.
Young, T.W. and Yagiu, M. (1978) Antonie van Leeuwenhoek, 44, 59.
 CHAPTER 4

The microflora of barley

and malt
B. Flannigan

4.1 THE MICROFLORA OF BARLEY

4.1.1 Introduction
The nature and the magnitude of the microflora of barley depend on both
the field conditions under which the crop was grown and the post-harvest
history of the grain. The microflora includes bacteria, actinomycetes, yeasts
and filamentous fungi which contaminate and colonize the grain in the field,
as well as others, particularly filamentous fungi, which are associated with
storage. The filamentous fungi, or moulds, in the first category are usually
referred to as field fungi, and include species of Alternaria, Cladosporium,
Epicoccum, Fusarium, Cochliobolus, Drechslera, Pyrenophora (the latter three
formerly known as Helminthosporium) and several other genera. The moulds
in the second category are known as storage fungi, and are mainly species
of Aspergillus, Eurotium and Penicillium. Although Pepper and Kiesling
(1963) listed a wide range of microorganisms which had been isolated from
barley kernels (and subsequent publications have added to this range), the
microfloras of different barleys are remarkably similar to each other, and to
other cereals, generally being dominated by the same limited number of
species. It should be said at this juncture, that since Pepper and Kiesling
(1963) compiled their list, and even since more recent work cited below has
been published, there have been considerable changes in the nomenclature
offungi. Where nomenclatural changes have taken place, the organisms will
be from here on referred to by their current names, with the older names
used by the authors of cited papers following in parentheses. Notice is also
taken that the International Commission on Botanical Nomenclature
(ICBN) has decreed that, where fungi have both asexual (anamorph) and

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and J. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0412591502
84 The microflora of barley and malt
sexual (teleomorph) phases, the teleomorphic name has priority. In this
chapter, therefore, teleomorphic names will be used for those members of
the anamorphic genera Aspergillus, Fusarium, Helminthosporium and Penicil-
lium which have sexual stages, with the anamorphic name appearing in
parentheses after the abbreviation 'ana.', e.g. Cochliobolus sativus (ana.
Helminthosporium sativum) or Pyrenophora teres (ana. H. teres).

4.1.2 The field microfIora

(a) Contamination and colonization

As in wheat (Flannigan and Campbell, 1977), barley kernels are first con-
taminated by airborne bacteria, yeast and moulds soon after the ears
emerge from the enveloping leaf sheaths (Hill and Lacey, 1983a). At first,
the developing caryopsis in the opened floret is fully exposed to contami-
nation from the air, as is the inside surface of the lemma and pale a, which
later mature to form the husk. However, as the caryopsis swells, filling the
space enclosed by these bracteoles, further direct aerial contamination is
prevented, although the outer surface of the lemma and palea remains
exposed. Climatic conditions appear to play an important part in the
deposition of microorganisms on the surface of the barley kernel. Warnock
(1973a) has found that the number of Cladosporium spores deposited on the
outside of the lemma and palea is related to the numbers in the atmosphere,
and both increase markedly after rain. The number of spores germinating
on the outside of the bracteoles also increases after rain, but there appears
to be little if any invasion of underlying tissues from the outside surface of
these organs.
The environment between the caryopsis and the bracteoles, however, is
favourable for microbial growth, which may then extend into adjacent
tissues. For example, whilst some spores deposited when the inside surface
of the bracteoles was exposed remain dormant, the majority of Cladosporium
spores appear to germinate. Initially, the dark-coloured hyphae developing
from these spores spread to form a superficial mycelium, which later
invades the parenchymal layer of the lemma and palea (Warnock, 1973a).
Mycelium of other field fungi, such as Alternaria and Cochliobolus sativus
(ana. Helminthosporium sativum), has also been noted on the inner surface
of, and within, the bracteoles (Mead, 1943; Warnock, 1973b). The pericarp
layer of the caryopsis itself may also be extensively invaded by mycelium
(Mead, 1943; Warnock and Preece, 1971), but perhaps less so than the
lemma and palea (Tuite and Christensen, 1955). Although most reports
appear to suggest that colonization arises from spores germinating on the
inner surface of the bracteoles and caryopsis, there is some evidence that
mycelium of Cladosporium invading the bracteoles and the outer layer of
the caryopsis may originate in trapped anthers which have previously been
contaminated and colonized (Warnock, 1973a).
 The microflora of barley 85

In addition to affecting spore deposition and germination on the outer

surface of the kernel, climatic conditions have a profound effect on the
colonization of the kernel by microorganisms. The prevalence of
seedborne phytopathogens, including various species of Cochliobolus and
Pyrenophora (ana. Heiminthosporium), in barley in eastern Canada has been
attributed to a moister climate than in other parts of the country (Machacek
et al., 1951). The presence of larger numbers of microorganisms, including
ciliated protozoa and the slime-mould Physarum poIycephaIum, in stained
and weathered barley (Kotheimer and Christensen, 1961) has been
ascribed to high relative humidity (RH) or rainfall after heading (Follstad
and Christensen, 1962). Lodging of crops frequently leads to profuse
mould growth; a high RH around the fallen ears may exist long enough
to allow development of the dark-coloured sporing structures of, partic-
ularly, Cladosporium on the surface of the kernel. However, it has been the
experience in this laboratory that the viable microbial counts for stained
or weathered barleys are not invariably higher than those for bright
barleys.
The determination of the degree to which kernels have been invaded by
microorganisms requires painstaking microscopical examination of serial
sections and excised organs. It is therefore not surprising there has been
only one investigation reported in which attempts have been made to
obtain quantitative estimates of the amount of fungal mycelium present
(Warnock, 1971; Warnock and Preece, 1971).
Warnock (1971) examined 20 kernels in detail and estimated that the
total length of hyphae present in individual kernels ranged from 19 to
177 cm, the mean values for two ten-kernel samples being 60 and 74 em.
It was suggested that these means represented a mycelial dry mass of
1.5-1.9Ilg. Whilst the spores of a number of field fungi, including
Cladosporium (Warnock, 1973a), are distinctive enough to be identified and
enumerated, the mycelium of one species can seldom be readily distin-
guished from that of another. The location of the mycelium of a particular
species requires the use of fluorescent antibody technique in conjunction
with exacting microscopical examination (Warnock, 1971, 1973b). In ad-
dition to the time-consuming nature of such work, another disadvantage
is that these methods cannot distinguish between living and dead myce-
lium or spores. Furthermore, although clusters of yeast cells have been
observed on the inner surface of the lemma and palea (Tuite and
Christensen, 1955), it is not possible to identify them. Consequently, since
many workers have been concerned with the presence of viable micro-
organisms which are phytopathogens or spoilage agents, most of what is
known about the components of the microflora derives from studies using
methods to enumerate such viable microorganisms, i.e. direct and dilution
plating (section 4.6).
86 The microflora of barley and malt
(b) Bacteria
Viable counts obtained by dilution plating show that even during the early
stages of development of the kernel in the field the microflora is numerically
dominated by bacteria (Kotheimer and Christensen, 1961; Follstad and
Christensen, 1962; Hill and Lacey, 1983a), and as many as 108 bacteria per
gram of barley have been isolated by Kotheimer and Christensen (1961) at
the late dough stage, or growth stage 87-89 (Zadoks, Chang and Konzak,
1974). No detailed investigation of the bacterial flora of barley has been
published, but Erwinia herbicola and Xanthomonas campestris are often men-
tioned as being prevalent in pre-harvest barley (Clarke and Hill, 1981;
Flannigan et al., 1982). It is likely, however, that other bacteria, such as
pseudomonads, micrococci and Bacillus spp., which have been noted in
dry-stored barley for malting (Haikara, Makinen and Hakulinen, 1977), will
also be present before harvest. Generally, the counts of actinomycetes in
barley in the field are low (Clarke and Hill, 1981; Hill and Lacey, 1983a). Most
appear to be Streptomyces spp. (Hill and Lacey, 1983a), Str. griseus probably
being the commonest, but Str. albus and Str. thermoviolaceus also being
notable. Thermoactinomyces vulgaris, one of the causative agents of the respi-
ratory disease farmer's lung, may also be evident in ripening barley (Hill
and Lacey, 1983a) and barley at harvest (Clarke and Hill, 1981), as well as in
recently harvested grain (Flannigan, 1970).

(c) Yeasts
Yeasts are usually the next most abundant components after bacteria in
viable counts prepared from pre-harvest barley, although their numbers
may be exceeded by filamentous fungi during the later stage of ripening.
By harvest, 50-85% of kernels may be colonized by yeasts (Hill and Lacey,
1983a). The pink yeasts, Sporobolomyces and Rhodotorula are frequently
predominant (Lund, 1956; Clarke, Hill and Niles, 1966; Flannigan, 1974;
Clarke and Hill, 1981; Hill and Lacey, 1983a), but Hansenula, Torulopsis and
Candida have been isolated from Danish barley before harvest (Lund, 1956).
Although Kotheimer and Christensen (1961) reported the presence of
Saccharomyces in pre-harvest barley in Minnesota, it is generally regarded
that members of this genus are rare in barley (Pepper and Kiesling, 1963).
Considering the strong resemblances between other components of the
microfloras of the two cereals, it is probable, however, that species of
Cryptococcus and Trichosporon noted in pre-harvest wheat (Flannigan and
Campbell, 1977) are also present in barley.

(d) Filamentous fungi

As might be expected from the microscopical studies mentioned pre-
viously, viable field fungi appear to be associated mainly with the lemma
 The microfLora of barley 87
and palea, judged by their distribution in naked barley before harvest
(Flannigan, 1974). Here, approximately two-thirds of the inoculum is asso-
ciated with the bracteoles and one-third with the caryopsis. After harvest,
the high percentages of surface-disinfected kernels which give rise to
colonies on plating on agar media confirm that field fungi are present
within the outer layers of the kernel as viable mycelium or spores
(Machacek et al., 1951; Follstad and Christensen, 1962; Flannigan, 1969,
1970; Haikara, Makinen and Hakulinen, 1977). The field fungus most
frequently present in kernels is Alternaria alternata (Alt. tenuis), both in
Europe (Clarke, Hill and Niles, 1966; Jorgensen, 1969; Flannigan, 1969, 1970;
Haikara, Makinen and Hakulinen, 1977; Clarke and Hill, 1981) and North
America (Machacek et al., 1951; Follstad and Christensen, 1962; Christensen
and Kaufmann, 1969), although Cladosporium spp. (mainly Clad.
cladosporioides and Clad. herbarum) may also be present in a large number of
kernels. For example, Hill and Lacey (1983a) have found that 90% of kernels
are contaminated by Cladosporium in some pre-harvest crops, and 100% by
Alt. alternata. Other saprophytic fungi which are common, but usually less
abundant than Alt. alternata and Cladosporium spp., are Epicoccum nigrum
(E. purpurascens) and Aureobasidium (Pullularia) pullulans (Machacek et al.,
1951; Flannigan, 1969, 1970). Acremonium (Cephalosporium) spp. (Flannigan,
1970; Haikara, Makinen and Hakulinen, 1977; Ylimaki et al., 1979) may be
prominent, and Verticillium lecanii may be present on every kernel of some
crops in the field (Hill and Lacey, 1983a).
Among seedborne phytopathogens, Cochliobolus sativus (ana.
Helminthosporium sativum) is rarer in the British Isles than in other parts of
Europe and North America, but Pyrenophora graminea (ana. H. gramineum)
and P. teres (ana. H. teres) are found (Flannigan, 1970). Fusarium spp. may
be frequent in wetter regions of North America (Machacek et al., 1951;
Follstad and Christensen, 1962), and in northern Europe, where the harvest
is late (Uoti and Ylimaki, 1974; Haikara, Makinen and Hakulinen, 1977;
Ylimaki et al., 1979; Flannigan and Healy, 1983). The most frequent fusaria
in European barleys appear to be the root-rot species, F. avenaceum (Ylimaki
et al., 1979) and F. culmorum (Haikara, Makinen and Hakulinen, 1977), but
in some studies the snow mould, Monographella nivalis (ana. F. nivaIe) has
been noteworthy, (e.g. Flannigan, 1969, 1970; Jorgensen, 1969).
F. sporotrichioides (F. tricinctum) has also been reported as prominent in
Finnish barley (Ylimakiet al., 1979). Ina Canadian survey, F. poae, F. equiseti
and F. acuminatum (F. scirpi var. acuminatum) were the commonest species
(Machacek et al., 1951), and in Japan, F. graminearum (lchinoe, Hagiwara
and Kurata, 1984).
If the total amount of inoculum is assessed by viable count methods, it
usually appears that the mould flora immediately before harvest is domi-
nated in Europe by Cladosporium spp. (Flannigan, 1974; Hill and Lacey,
1983a), and in the United States by Alt. alternata (Follstad and Christensen,
1962). Studies on the development of the microflora of wheat (Flannigan and
88 The microflora of barley and malt
Campbell, 1977) suggest that in Europe Aureobasidium pullulans and
Cladosporium spp. are likely to be the earliest contaminants of barley and to
increase rapidly in numbers after anthesis, followed by Alt. alternata and
often Verticillium lecanii, and finally by Epicoccum nigrum. At this stage, a
single kernel may carry approximately 1.6 x 104 colony forming units (CFU)
of bacteria, 8 x 103 yeast CFU and 2.5 x 103 field fungi (Flannigan, 1974).

4.1.3 The storage microflora

(a) Bacteria in stored barley

Unless there has been extensive contamination by the excretory products
of rodents, insects or mites, the bacteria found in stored barley are likely to
be the hardier remnants of the field bacterial flora. Petters, Flannigan and
Austin (1988) reported that the flora of stored English malting barley
comprised representatives of 13 taxa, mostly Gram-positive pigmented
organisms. The Gram-positive bacteria included Aureobacterium flavescens,
Bacillus spp., Brevibacterium linens, Corynebacterium spp., Clavibacterium
iranicum, Microbacterium imperiale and Oerskovia xanthineolytica, whilst
among the Gram-negative bacteria were Erwinia herbicola, Pseudomonas
fluorescens and Chromobacterium sp. The highest populations of individual
types were of Corynebacterium sp. (18% of total isolates) and Ps.fluorescens
(16%).

(b) The source of storage fungi

Storage fungi are not found exclusively in stored grain; they are to be found
as contaminants of grain in the field. In studies where kernels have been
surface disinfected, after the manner of seed pathologists, it has been
concluded that these fungi are virtually absent, e.g. by Clarke, Hill and
Niles (1966) and Clarke and Hill (1981). When kernels have been plated
directly onto agar media without surface disinfection, however, small
numbers have yielded penicillia and aspergilli, with the latter being
reported to be most frequently in the Aspergillus glaucus group (Tuite and
Christensen, 1957; Flannigan, 1978; Hill and Lacey, 1983a). This very
important Asp. glaucus group of species possess sexual stages and are now
referred to by the teleomorphic name Eurotium. Very occasionally, as many
as 20% of kernels may be contaminated by Eurotium spp. in the field
(Flannigan, 1978), but seldom more than 5% bear these storage fungi
(Christensen and Kaufmann, 1969). During harvesting there is apparently
little, if any, increase in the level of contamination by Eurotium, but it
appears there can be massive contamination by Penicillium spp. The num-
ber of kernels yielding penicillia may increase from < 1% on the ear to
approximately 60% in the combine harvester (Flannigan, 1978). The source
of this inoculum seems to be soil thrown up by the combine harvester,
 The microflora of barley 89
culms and leaves which are threshed in the drum along with the grain and
residues of previous crops in the auger and storage tank of the combine. If
newly harvested grain is surface disinfected, it is likely that < 0.5% of
kernels will yield storage fungi (Christensen and Kaufmann, 1969),
confirming that at this stage contamination is almost entirely superficial,
and that few kernels will have been actively colonized by Eurotium spp. and
Penicillium spp.
Although they have received most attention, the Eurotium species
E. amstelodami, E. chevalieri, E. repens and E. rubrum (ana. Asp. amstelodami,
Asp. chevalieri, Asp. repens and Asp. ruber) are not the only storage aspergilli
which contaminate barley in the field. Asp. candidus, Asp. jlavus (Flannigan,
1978), Asp. fumigatus, Asp. nidulans, Asp. niger, Asp. ochraceus and Asp.
versicolor (Hill and Lacey, 1983a) may also be present. Asp. fumigatus, a
thermotolerant species associated with the spoilage of moist-stored feed
barley (Lacey, 1971; Clarke and Hill, 1981; Hill and Lacey, 1983b), has been
noted as being surprisingly widespread in pre-harvest barley. Other such
spoilage species which may also be present as contaminants in the field (Hill
and Lacey, 1983a) include mesophilic Penicillium piceum and P. roqueforti
and thermophilic Rhizomucor (Mucor) pusillus, Talaromyces thermophilus
(ana. P. dupontii) and Thermomyces lanuginosus (Humicola lanuginosa).

(c) Post-harvest colonization

The moisture content (MC) of barley at harvest may range from <10%
(Lutey and Christensen, 1963) in drier areas of North America and Australia
to 25-30% in northern Europe (Hellberg and Kolk, 1972). Grain with a high
MC must be dried to <14% if it is to be stored for any period of time, and
to <12.5% to exclude the possibility of any mould growth in storage. Whilst
the need to dry grain is clearly understood, it is often the case that the rate
at which it can be dried cannot keep pace with the rate at which it is
harvested. The resulting delay in drying the harvested grain can allow
microorganisms to develop and the grain to heat. One method of reducing
microbial growth and cooling the grain in such an event is to blow air at
ambient temperature through the bulk, i.e. low-volume ventilation (LV).
LV is frequently employed in pre-drying silos until the grain can be dried
to a safe level using a hot-air drier, but it can only be regarded as a
temporary holding operation. Since many farmers do not have the capacity
to satisfactorily dry barley for malting, grain merchants and maltsters often
prefer to dry the grain themselves. Although, by having specifications
which state that, for example, 'Grain must be sound, sweet and free from
heat, (visible) mould and infestation' (Dolan, 1979), the maltster appears to
safeguard his interests, it would be possible for a farmer to present for
purchase barley which had heated and then been cooled using LV. The
maltster is well advised 'to be wary of barley held on the farm for any length
of time before drying' (Flannigan and Healy, 1983).
90 The microflora of barley and malt
Depending on the moisture content of the harvested grain, there may be
some growth of field fungi, e.g. Fusarium spp., if drying is delayed. How-
ever, such delays are usually characterized by the development of storage
fungi. It has been found that, compared with barley delivered for drying to
a maltster directly after harvesting, the viable mould count of barley held
on the farm for a few days can rise from 3 x 103 to 1.3 X 104 CPU per kernel,
with species of Eurotium, Penicillium and especially Absidia, Mucor and
Rhizopus being prominent in the farm-held grain. Numbers of yeasts can
increase from 5 x 103 to 2.2 X 104 per kernel (B. Murphy, unpublished re-
sults, cited in Flannigan and Healy (1983)). The development of these
storage fungi usually appears to be at the expense of the field species. Mills
and Wallace (1979) noted that the numbers of kernels bearing the predom-
inant field fungi Alt. alternata and Cladosporium spp. apparently declined as
those contaminated by Eurotium spp. (Asp. glaucus) and P. aurantiogriseum
(P. verrucosum var. cyclopium) increased in piles of undried grain after a wet
harvest in Manitoba. In the UK, Flannigan and Healy (1983) have observed
that undried barleys received in a 'hot' or 'hot and smelly' state are
characterized by lower levels of field fungi, particularly Cladosporium spp.,
and higher levels of storage fungi, especially Eurotium (Asp. glaucus), than
are present in barleys acceptable to the maltster for purchase and drying.
Even when there is no delay in drying, the number of kernels contami-
nated by storage fungi increases by the time the grain enters storage. Much
of the increase results from cross-contamination as the grain is moved
during the post-harvest operations leading up to, and including, drying
and cleaning, but part may be the result of proliferation during the early
part of drying. Mills and Wallace (1979) have found that the drying process
can reduce the numbers of kernels bearing viable myxomycetes, and lead
to increases in the levels of Absidia, Eurotium and Penicillium. Where,
because of a low MC at harvest, the grain does not require drying, it has
been demonstrated (Tuite and Christensen, 1957) that there will still be a
large increase between the combine harvester and elevator in the number
of kernels contaminated by Eurotium spp. Christensen and Kaufmann
(1969) have drawn attention to the large numbers of spores of storage fungi
which are habitually present in dust and the air of elevators and act as the
inoculum for incoming grain. There is no practical method of reducing the
number of these spores, far less eliminating them, so that control of storage
fungi, which can initiate spoilage of dry-stored grain, lies in ensuring that
storage conditions prevent their growth.

(d) Storage fungi as agents of spoilage

Whilst barley entering storage is contaminated or colonized by a wide
range of microorganisms, the storage conditions determine which, if any,
of these organisms can grow on the grain. The moisture content of the
grain is clearly important, but other interacting factors, including
 The microflora of barley 91
temperature, aeration, inclusion of chaff and foreign material such as weed
seeds and infestation by insects and mites, playa part in selecting which
microorganisms proliferate.
Although maltsters and grain merchants think in terms of the Me of
grain as far as safe storage is concerned, it is perhaps more appropriate from
the microbiological standpoint to consider the water activity (a w ) of the
grain. In this context, it is defined as the ratio of the vapour pressure of
water in the grain to that of pure water at the same temperature, and is
equivalent to equilibrium relative humidity (ERH). If seeds are stored, for
example, in an atmosphere at 80% RH they will eventually come to equi-
librium with that atmosphere, their aw being 0.80, but their Me depending
on the chemical nature of the seed. Oily seeds will have a lower Me than
starchy grain, e.g. in groundnuts the Me will be approximately 9.2%, and
in barley 16.6%. The relationship between Me, aw and ERH for barley over
a range relevant to mould growth is shown in Table 4.1, and between Me
and aw for barley and finished malt in Fig. 4.1. At aw 1.0, the grain is fully
hydrated; at aw 0.6, even the moulds least demanding in their water require-
ments cannot grow.
The interacting parameters of aw and temperature define whether the
spores of different microorganisms germinate or not, and the rate at which
the organisms grow. In Table 4.2, the temperature range and minimum aw
of some of the commoner species of actinomycetes and moulds which have
been isolated from stored barley are listed, the physiological classification
of Lacey, Hill and Edwards (1980) being adopted. Of the moulds, Asp.
restrictus is the most extreme xerophile: Hill and Lacey (1983b) have re-
corded that it grows slowly in barley at aw 0.65-0.70. Eurotium spp. are less
extreme than Asp. restrictus, although nevertheless strongly xerophilic.
Most other storage aspergilli are regarded as being moderately xerophilic,
but the thermotolerant species Asp.fumigatus is only slightly xerophilic,like
the storage penicillia. As stated previously, drying barley to an Me of 12%
(a w 0.6) prevents any microbial growth. Drying to 14% (a w 0.7) should be
sufficient to prevent growth of all but the most xerophilic of moulds, and
clearly precludes growth of field fungi and yeasts, generally with a

Table 4.1 The relationship between

equilibrium relative humidity (ERH), water
activity (aw) and moisture content (MC) of
barley
ERH aw Approx. Me
(%) (%)

60 0.6 12.3
70 0.7 14.0
80 0.8 16.6
90 0.9 20.3
95 0.95 24.8
92 The microflora of barley and malt

30-

20

10

[
M

I I I I I
0·2 04 06 o·s 1·0
Ow

Fig. 4.1 The relationship at 25°C between moisture content (MC) and water
activity (a w ) of barley (B) (after Pix ton and Warburton, 1971) and finished malt (M)
(after Pixton and Henderson, 1981).

minimum aw > 0.90, and bacteria, with a minimum aw usually> 0.93. Some
field fungi, including Fusarium (Welling, 1969) and Cladosporium, are able
to grow in grain with high aw which has been stored at low temperatures.
In such situations, psychrotolerant penicillia (with a temperature range
-lOoC to 35°C) seem also to be prominent, e.g. P. brevicompactum,
P. chrysogenum, P. aurantiogriseum (P. verrucosum var. cyclopium) and
P. verrucosum (P. verrucosum var. verrucosum).
By their metabolic activity, microorganisms generate water and heat, so
they have the capacity to raise both the aw and temperature of stored grain,
a poor conductor of heat. As has been mentioned above, growth of extreme
xerophiles at aw 0.65-0.70 is slow and the small amount of heat that is
generated is dissipated, so that the temperature of the grain does not rise.
However, by generating water these organisms gradually raise the mois-
ture level in the grain, and in so doing make possible the growth of other
less xerophilic species, as well as allowing accelera tion of their own growth.
As further species grow, the build-up of heat is such that thermophilic
organisms are able to develop rapidly and the temperature of the grain may
reach 65°C or more.
 The microflora of barley 93
Table 4.2 Physiological classification of storage moulds and actinomycetes
according to their temperature limits and minimum aw for growth (after Lacey, Hill
and Edwards, 1980)
Lower Upper Thermotolerantl Extremely
mesophilic mesophilic thermophilic thermophilic
Category (2-37°C) (5-50°C) (1O-57°C) (25-70°C)
Extremely Aspergillus
xerophilic restrictus
(min. aw <0.75) Eurotium spp.*
Moderately Asp. versicolor Asp. candidus
Asp·fIavus
Asp. niger
Asp. ochraceus
Asp. terreus
Slightly Penicillium P. capsula tum Asp. fumigatus
xerophilic frequentans P. citrinum
(min. aw 0.80 P. rugulosum P. funiculosum
--0.89 P. piceum
Hydrophilic Fusarium spp. Absidia Rhizomucor Thermomyces
(min. aw >0.90) Mucor corymbifera pusillus lanuginosus
hiemalis Rhizopus Thermoascus Talaromyces
M. racemosus arrhizus crustaceus thermophilus
R. stolonifer Streptomyces S. albus Faenia
S. aureofaciens griseus rectivirgula
S. olivaceus Thermoactino-
myces vulgaris
• 2-50°C.

In storage experiments, Hill and Lacey (1983b) observed that after 9-12
months the slow growth of Asp. restrictus in barley at aw 0.65-0.70 gave way
to Eurotium spp. (Asp. glaucus group) at aw 0.75 (MC > 14.4%), and deteriora-
tion of the grain (judged by percentage germination) accelerated. This
group continued to increase in numbers as the aw rose and reached a
maximum at aw 0.90-0.93 (MC 20.0-22.3%), although they were then no
longer the predominant organisms. Above aw 0.85 (MC 17.2%), spontan-
eous heating occurred and P. aurantiogriseum (P. verrucosum var. cyclopium)
and Asp. candidus became dominant. At aw 0.90-0.95 (MC 20.0-25.3%) the
temperature reached a maximum of 50°C, and P.funiculosum, P. capsulatum,
Asp. flavus, Asp. nidulans, streptomycetes and bacteria predominated. To-
gether with the respiratory pathogen Asp. fumigatus, Absidia corymbifera
increased in numbers, but only reached a maximum above aw 0.95, when
the temperature rose to a maximum of 65°C and thermophilic organisms
were present in large numbers. These included the actinomycetes
Thermoactinomyces vulgaris and Micropolyspora faeni (causative agents of
farmer's lung) and fungi such as Rhizomucor pusillus and Thermomyces
lanuginosus.
94 The microflora of barley and malt
Although the development of a microbial succession culminating in
massive proliferation of thermophilic organisms is usually associated with
the faulty storage of undried feed barley (Clarke, Niles and Hill, 1967;
Clarke et al., 1969; Lacey, 1971; Clarke and Hill, 1981) high temperatures
may be reached in dry-stored grain when 'hot spots' develop. Tempera-
tures as high as 53°C have been recorded in hot spots within grain bulks
during the Canadian winter (Wallace and Sinha, 1962), and in one hot spot
a maximum of 64°C in May (Sinha and Wallace, 1965). The initiation of
heating in this last case was attributed to the growth of P. aurantiogriseum
(P. verrucosum var. cyclopium) and P. funiculosum at -5°C to 8°C during the
winter after harvest. Asp.flavus, Asp. versicolor, Absidia spp. and Streptomyces
spp. succeeded the penicillia as the temperature rose. In addition to Asp.
flavus and Asp. versicolor, Asp. fumigatus was prominent in other hot spots
(Wallace and Sinha, 1962). Since Gilman and Barron (1930) observed in
laboratory experiments that in less than 1 week Asp. flavus and Asp.
fumigatus can raise the temperature of barley, oats or wheat by as much as
26°C, it appears that after the initial stages these aspergilli play an
important part in the accelerating development of such hot spots.
Sinha and Wallace (1965) found that, after the maximum had been
reached, the temperature of the hot spot fell to that of the surrounding grain
in about 3 weeks, and there was no later resurgence of growth. Although
the time scale may vary, this can be regarded as being typical of the course
of events in a hot spot. The fall in temperature presumably results from the
cessation of microbial growth due to the combined effects of the elevated
temperature and associated drying of the grain. Sinha and Wallace (1965)
observed that the MC at the focus of the hot spot fell in 4 weeks from
20.0-23.5% at the peak temperature to 16.0-18.9%, and that the size of the
central area with an MC > 16.0% was halved between the second and the
fourth weeks. Growth may cease within the heated pocket of grain, but
many viable spores remain and can present a health hazard (Lacey, 1975).
Whilst storage fungi survive as spores (Sinha and Wallace, 1965), field fungi
do not, despite being able to exist (in decreasing numbers) for a period of
several years in dry-store grain (Machacek and Wallace, 1952).
Development of a large hot spot can start from mould growth in a very
small patch of moister grain within a bulk. Moisture migrates from this
small focus of activity, as it does throughout the development of the hot
spot, so that the zone of fungal growth extends and more grain is affected.
There are various reasons for some grain initially being moister than the
remainder of the bulk, and consequently capable of supporting mould
growth. In the first instance, it may not have been dried to a safe level. MC
measurements are average values, but grain lots are not homogeneous and
there can be marked differences between the true moisture content of one
kernel and the next. It may be that in a batch dried on the basis of the MC
of grain harvested in the middle of the day, there is barley which was
harvested at either end of the day and consequently had a higher initial
 The microflora of barley 95
MC It can be that some grain is not cooled sufficiently after drying, so that
when it cools within a bin or silo, condensation forms on the surface of the
kernels. Sufficient water for mould growth can also become available at the
surface of the kernel as a result of the diurnal evaporation/ condensation
cycle on the insolated side of a storage bin. There is also the possibility that
structural faults may allow water to leak into a grain bulk.
There is often a close association between spoilage fungi and the pres-
ence of mites and insects (Howe, 1973; Sinha, 1973). If insects are present
in large numbers they can increase the moisture of grain and initiate the
development of hot spots. Mites and insects, including the grain weevil
Sitophilus granarius, may feed on moulds, and they act as vectors for the
spread of moulds by carrying spores on the surface of their bodies as they
migrate. In a typical hot spot developing as a result of insect infestation
Gacobson and Thomas, 1981) the temperature at the origin may reach 39°C
and the adult insects migrate upwards. As moisture also migrates upwards,
barley at the top of the bulk may become moist enough to sprout and
support profuse mould growth.
Whichever is the case, there is a rapid drop in the germinative capacity
of kernels as storage fungi develop and grain heats (Sinha and Wallace,
1965; Lund, Pedersen and Sigsgaard, 1971; Hill and Lacey, 1983b). Where
the MC of barley is 15-18% or greater, it has been found (Tuite and
Christensen, 1955; Armolik, Dickson and Dickson, 1956; Follstad and
Christensen, 1962) that even when there is no self-heating in storage there
is reduced germination and an associated build-up of the more xerophilic
moulds, particularly the Eurotium spp. (Asp. glaucus group). From numer-
ous experiments (reviewed in Christensen and Kaufmann (1969, 1974) and
Christensen (1973)) in which grain was heavily inoculated with Eurotium
(Asp. gZaucus), Christensen and his colleagues concluded that the storage
fungi are directly responsible for this reduction in germination. It cannot
be disputed that in such experiments, and in many other cases where there
is a low germinative capacity, Eurotium spp. will grow out from the embryo
and outer layers if the kernels are incubated in a moist chamber. However,
Harrison and Perry (1976) noted various deteriorative changes of
endogenous origin, developing in advance of any substantial invasion by
fungi, and concluded that, although Eurotium rubrum (Asp. repens),
P. aurantiogriseum (P. cyclopium) and, at higher moisture levels, Fusarium
culmorum increase in numbers, they are not the primary causes of
deteriora tion.
Flannigan and Bana (1980) have drawn attention to characteristics of the
embryo which may make it a favoured site for colonization by fungi. Lipid-
and ethanol-soluble sugars and oligo saccharides, including sucrose, raffi-
nose and fructosans, are concentrated in the embryo and aleurone. As
experimental work showed, Eurotium spp. (Asp. glaucus group) show lipo-
lytic activity and are able to utilize sucrose and raffinose as sole sources of
carbon, although they show little or no capacity to degrade starch or
96 The microflora of barley and malt
structural polysaccharides in grain. In addition to providing a readily
assimilable source of carbon for these moulds, the concentration of sugars
in the embryo means that the embryo will more readily absorb moisture
than the starchy endosperm, e.g. from a humid atmosphere. In correctly
dried grain, the MC of the embryo should be similar to that for the whole
kernel; in badly stored grain, the MC of the embryo will be higher than that
of the whole. This raising of the MC of the embryo will in theory make it
more susceptible to invasion by storage fungi, but it may be that it must be
debilitated to some degree before it is invaded.

4.2 THE MICROFLORA OF MALT

4.2.1 The microflora during malt production

During steeping of grain, dormant microorganisms are activated: mould
spores germinate, mycelium grows and yeasts and bacteria multiply. In
laboratory studies, increases in bacterial numbers range from x5 (Follstad
and Christensen, 1962; Flannigan et al., 1982) to x36 (Kotheimer and
Christensen, 1961), and in yeasts from xl. 1 in stained and weathered barley
and x5 in bright barley (Kotheimer and Christensen, 1961) to x200 in
Finnish barley (Haikara, Makinen and Hakulinen, 1977). In commercial
malting, Douglas (1984) found that where steeping is interrupted by an
air-rest, there may be increases of x4 to x25 in bacterial numbers in the grain
by the end of the first steep and x35 to x150 by the time it is couched.
However, during a different year in the same maltings, the number of viable
heterotrophic bacteria on the grain actually fell during the first steep (Table
4.3), but had increased to nearly x4 by the end of the second (petters,
Flannigan and Austin, 1988). The range of bacteria on the barley at the first
steep was narrower than in dried barley, consisting of Aureobacterium
flavescens, Alcaligenes sp., Clavibacterium iranicum and Clav. michiganense,
Flavobacterium esteroaromaticum, Microbacterium imperiale, Erwinia herbicola,
Pseudomonas fluorescens and Ps. putida. By the end of the second steep, the

Table 4.3 Viable counts of microorganisms associated with kernels during com-
mercial production of a sulphured malt (after Petters, Flannigan and Austin, 1988)
Aerobic
heterotrophic
Stage bacteria Lactobacilli Moulds Yeasts
6
Stored barley 1.8 x 105 2.0 x 10; 2.0 x 10; 4.7 x 10;
First steep 6.7 x 106 4.2 x 104 8.0 x 103 4.6 x 106
Second steep 6.6 x 107 7.8 x 106 1.7 x 102 1.1 x 106
Green malt (5 days) 5.7 x 106 8.7 x 105 1.5 x 102 3.9 x 104
Kilned malt 5.6 x 106 1.6 x 104 2.0 x 102 3.2 x 104
Screened malt 5.5 x 10 5.7 x 10 8.3 x 10 1.8 x 10
 The microflora of malt 97

range had widened, with Gram-positive rods being particularly prominent.

Lactobacilli appeared in substantial numbers (Table 4.3), the increase being
proportionally much greater than for total viable bacteria (Petters,
Flannigan and Austin, 1988). This is in line with the earlier observation of
Sheneman and Hollenbeck (1960) that mesophilic lactic acid bacteria rose
during steeping from 0-10 2 to >5 X 105 CPU g-l, while total bacterial numbers
increased from around 106 g-l to 7 X 106 g-l.
Just as bacterial counts generally increase with steeping, so do yeast
numbers. For example, Douglas (1984) reported that viable yeast numbers
at the end of the first and second steeps could be respectively x3 to x20 and
x5 to x1300 those of dry barley. Petters, Flannigan and Austin (1988)
observed increases of nearly x100 and more than x230 from a level of
4.7 x 103 CPU per dry barley kernel at the first and second steep, respec-
tively (Table 4.3). In comparison, for filamentous fungi the corresponding
increases from the original 2 x 102 CFU per dry kernel were only 4- and
8-fold.
Examination of steep liquor shows that, in addition to activating dormant
organisms, steeping washes superficial contaminants off kernels (Healy,
1985; Petters, Flannigan and Austin, 1988), e.g. Aureobasidium, Cladosporium,
Mucor and Penicillium (Healy, 1985). A proportion of these contaminants is
deposited on other kernels. In micro-malting, Haikara, Makinen and
Hakulinen (1977) reported an increase from 15% to 90% in kernels bearing
Fusarium spp., although Follstad and Christensen (1962) earlier noted a fall
with steeping from 47% to 3% in kernels contaminated by Cladosporium spp.
Increases may, of course, result in part from the growth of mycelium from
one kernel to another, but it seems unlikely that this would account for all
the cross-contamination observed by Haikara, Makinen and Hakulinen
(1977). Douglas (1984) observed that, as well as increases in the numbers of
kernels contaminated by Fusarium spp. by the end of the first steep, in some
commercial malting runs there could be increases in viable counts, indicat-
ing proliferation during steeping. Healy (1985) noted increased contamina-
tion by F. avenaceum and F. culmorum during steeping, and detected F. poae
and other fusaria which had not been evident in the original dry barley,
although presumably they had been present in a small number of kernels.
Increases have also been noted in both viable counts of, and percentage
frequency of kernels bearing, Aureobasidium pullulans and Cladosporium spp.
(Douglas and Flannigan, 1988). Together with Trichosporon beigelii
(T. cutaneum), Geotrichum candidum, a yeast-like species rarely found in dry
barley (Pepper and Kiesling, 1963), was reported by Healy (1985) to be
present on appreciable numbers of steep-out kernels. Although Haikara,
Makinen and Hakulinen (1977) observed an increase in the percentage
frequency for what is usually the commonest field fungus, Alt. alternata,
during laboratory steeping, Gyllang and Martinson (1976b) and Healy
(1985) recorded falls during steeping in the course of production of a
commercial malt. It was found by Douglas (1984) that in some cases there
98 The microflora of barley and malt
was little change and in others a fall in the percentage of kernels yielding
Alt. aiternata, even when there was an apparent increase in the total amount
of inoculum (viable count) of the organism. Haikara, Makinen and
Hakulinen (1977) recorded increases in the percentage frequency of kernels
yielding the storage aspergilli and penicillia, and Gyllang and Martinson
(1976b) reported an increase for the latter, but Douglas (1984) found that
these two types can increase or decrease independently of each other.
Whether in laboratory or commercial malting, viable counts of bacteria
and yeasts reach a maximum during germination of the barley. Although
Follstad and Christensen (1962) recorded smaller increases, Haikara,
Makinen and Hakulinen (1977) reported bacterial counts for germinated
barley x400 those for the original dry barley, and yeasts x100, and
Kotheimer and Christensen (1961) reported corresponding increases of
x4300 and x79, respectively, in laboratory trials. Sheneman and
Hollenbeck (1960) noted a peak in bacterial numbers in green malt after
5 days, when the mean total count was 7 x 108 g-l (x700 the count for
barley). Numbers of mesophilic lactic acid bacteria increased in parallel
with the total numbers, but accounted for < 2% of the total counts. Douglas
(1984) noted x85 to x440 more bacteria on green malt in box maltings than
on the dry barley, and x40 to x1800 more yeasts. Corresponding rises
(Table 4.3) were observed by Petters, Flannigan and Austin (1988), with
the range of Gram-negative bacteria isolated being greater than earlier.
The dominant species was Flav. esteroaromaticum (26% of isolates), and
Erw. herbicola, Ps. fluorescens, Ps. putida and Serratia plymuthica were also
prominent. Lactobacilli increased nO-fold from the end of the second
steep, L. alimentarius and L. plantarum being among those noted. It should
be noted here that there can be large differences between the relative
increases in different malting systems (Flannigan et al., 1982). The increases
in bacterial counts between casting and the end of the germination period
are proportionally greater in floor malting than in saladin box malting,
and in germination vessels there is apparently little or no change. The
corresponding increases in yeast numbers are, however, apparently
greater in box malting than in floor malting, but in germination vessels
the increases are much smaller.
During laboratory malting, Kotheimer and Christensen (1961) found
that there was no change in the viable mould count as bright barley
germinated, although, compared with the dry grain, there was a 4-fold
increase in the case of stained and weathered barley. Follstad and
Christensen (1962) reported that in their laboratory studies counts in some
cases increased by roughly x6 relative to the dry barley, but in others there
were either no or considerably smaller changes. Flannigan et al. (1982) also
recorded a x6 increase during germination, but noted in floor malting a
corresponding 13-fold increase, and in some saladin boxes and germination
vessels apparent falls of 20-70% and 60-90%, respectively. On some occa-
sions, Douglas (1984) recorded lower counts in green malt produced in
 The microflora of malt 99

saladin boxes than in the original barley, but also found an increase of xll
during one malting run, largely due to Fusarium spp.
When the literature is examined, it is difficult to discern any consistent
pattern of development of moulds from the evidence on viable counts, and
this is compounded by the results for direct plating. Haikara, Makinen and
Hakulinen (1977) found that during germination, Fusarium spp. already
being present in 90% of kernels after steeping, the largest increase in the
number of kernels bearing any particular category of mould during germi-
nation was for those bearing Penicillium spp. (from 30% to 77%). Smaller
proportional increases were observed in those contaminated by, in
descending order, Mucor, Eurotium (Asp. glaucus), Alt. alternata and
Rhizopus. In commercial malting, Gyllang and Martinson (1976b) again
found that the greatest increase, although smaller, was in the percentage
frequency of Penicillium spp. in kernels, followed by Pyrenophora teres (H.
teres) and Alt. alternata. However, Douglas (1984) reported that during
germination the greatest increases in frequency were for Geotrichum can-
didum, Mucor spp. and, once, Alt. alternata. Direct and dilution platings
(Douglas, 1984; Douglas and Flannigan, 1988) indicated that the level of
contamination by xerophilic storage aspergilli (Eurotium spp.) and peni-
cillia may fall during the production of green malt in Saladin boxes. This
decrease in xerophilic fungi was also observed by Healy (1985) in other
maltings. Although the levels of field fungi such as Cladosporium spp.,
Epicoccum nigrum (E. purpurascens) and sometimes Alt. alternata may de-
crease during production of green malt in Saladin boxes (Flannigan et al.,
1984), increases in Cladosporium and Alternaria, as well as Fusarium, have
also been observed in commercial maltings (Healy, 1985). However, as
Haikara, Makinen and Hakulinen (1977) have indicated, the initial level of
contamination in the dry barley is important in determining how successful
a mould is in colonizing other kernels during the production of green malt.
Kilning has a marked effect on bacterial numbers. Follstad and
Christensen (1962) found that the very high numbers on green malt were
reduced to between 63% of the count for the barley before steeping and x8
the barley count. Haikara, Makinen and Hakulinen (1977) recorded a drop
of 95% in numbers, but they were still x20 those in the barley. In contrast,
Haikara, Makinen and Hakulinen (1977) did not observe any reduction in
yeast numbers, which remained xlOO those on barley, but Follstad and
Christensen (1962) found that there was a fall to x6-27 the original count
for barley.
Having noted 15 yeast species in eight genera during malting in Saladin
boxes, the commonest being species of Candida, Cryptococcus, Debaryomyces,
Pichia and Rhodotorula, Healy (1985) observed a reduction to < 10% of
former levels as a result of kilning. In another system, in which the same
vessel was used for grain drying, steeping, germination and kilning, she
recorded 22 yeast species, adding Kluyveromyces to those genera observed
in the Saladin boxes. The total viable yeast count in this case was x2.5 that
100 The microflora of barley and malt
of the original barley, and nearly one-half of the species survived kilning,
to varying extents. Petters, Flannigan and Austin (1988) only recorded eight
species, with the predominant species on green malt being Candida
catenulata and Debaryomyces hansen ii, and on screened malt the former and
Rhodotorula mucilaginosa.
The combination of sulphuring (to produce a pale malt or reduce nitro-
samine formation) and kilning results in greater reductions in numbers of
bacteria (Graff, 1972; Flannigan, 1983). Sheneman and Hollenbeck (1960)
found that bacterial counts for finished malt were similar to those for the
original barley, and Douglas (1984) observed thatthey were xO.5 to x2 those
in barley. Later, a fall greater than one order of magnitude in both total
viable aerobic heterotrophic bacteria and lactobacilli was observed by
Petters, Flannigan and Austin (1988). The dominant types on the kilned
malt were Gram-positive, viz. Clav. iranicum, Flav. esteroaromaticum,
Arthrobacter globiformis and, among the lactobacilli, L. alimentarius.
Flannigan (1983) found in the laboratory that yeast numbers were also
affected by sulphuring, and Douglas (1984) observed in a maltings that
yeasts were reduced to < 1-20% of the number in barley, although Petters,
Flannigan and Austin (1988) later reported a 6-fold increase relative to the
barley (Table 4.3).
Follstad and Christensen (1962) found that, in general, kilning brought
about reductions in the numbers of moulds on germinated grain with the
final counts being 63-170% of those for the original barley. However, in a
malt prepared from weathered barley the count rose from x5 the original
to x7 with kilning. As far as the percentage frequency of kernels contami-
nated by different moulds is concerned, Haikara, Makinen and Hakulinen
(1977) noted large increases with kilning for Mucor spp. and Rhizopus spp.,
and smaller increases for Penicillium spp., Eurotium spp. (Asp. glaucus
group) and, from a very low level, Cladosporium spp. However, there was
no change for Alt. alternata and Fusarium spp. Gyllang and Martinson
(1976b) also noted increases in the number of kernels bearing Rhizopus spp.
and Cladosporium spp. after kilning in a maltings for 24 h, together with an
increase in the frequency of the respiratory pathogen Asp. jumigatus. Kiln-
ing with sulphur appears to have a profound effect in reducing both viable
counts and the percentage of kernels bearing moulds. Douglas (1984) found
that most field fungi and storage penicillia were frequently reduced to a
level below the limits of detection in dilution plating, although they might
still be detected in small numbers of kernels by direct plating. However,
Eurotium spp. appear to be much less affected, and the number of kernels
yielding Mucor spp. may increase. An increase in Mucor spp. with kilning
can be responsible for the total mould count being greater than that for the
original barley, although the counts usually appear to be lower (Douglas,
1984). Despite the proliferation of Geotrichum candidum resulting in contam-
ination of nearly all kernels of green malt, it has been reported that viable
inoculum of this species may be virtually eliminated by kilning with
 The microflora of malt 101
sulphur (Douglas, 1984; Douglas and Flannigan, 1988), although other
work (Healy, 1985) indicates that this species may survive on malt kernels.
It is worth noting that the design of the kiln may influence whether
moulds survive. Heaton, Callow and Butler (1992) reported that a Phoma
and two species of Fusarium with a maximum temperature for growth of
40-45°C were able to survive kilning at an air temperature of 70°C and
caused disfigurement of the painted wall in a concrete malting tower. It
appeared that the temperature of the poorly insulated kiln wall was below
40°C for most of the kiln cycle during the winter months and only reached
the kiln air temperature of 70°C for short periods during the summer.
Although agar plate cultures of the organism totally lost viability when
placed in the centre of the kiln, it is not known whether malt kernels
adjacent to the wall failed to attain high kilning temperatures or whether
the fungi survived in these kernels.

4.2.2 The microflora of finished malt

Reports indicate that the total numbers of bacteria in distiller's malts
(Boruff, Claasen and Sotier, 1938) are likely to be in the range 0.7-7.7 x
106 g-I for barley malt and 1.6-4.6 x 106 g-I for rye malt, although Flannigan
et al. (1982) have recorded a maximum of 9.8 x 107 g-I for barlel malt.
Sheneman and Hollenbeck (1960) obtained totals of 1.4-7.0 x 10 g-l for
brewer's malts, with the maximum count for lactic acid bacteria being
1.4 x 104 g-I. They reported that Lactobacillus leishmanii was the predomi-
nant mesophilic lactic acid species, and that Pediococcus acidilactici and
L. delbrueckii were the main thermophilic species, although they were pres-
ent in very small numbers. The laboratory studies of Haikara, Makinen and
Hakulinen (1977) indicated that the principal Gram-negative bacteria pres-
ent are likely to be Erwinia herbicola and Pseudomonas spp., with micrococci
and Bacillus spp. being the main Gram-positive types.
Yeast numbers in finished malt have been reported to lie in the range
7.2 x 102 to 1.1 X 104 g-I (Flannigan et al., 1982), with the pink yeasts in the
genera Sporobolomyces and Rhodotorula often being abundant. Other yeasts
usually present include those found in grain (section 4.1.2(c».
Counts for fungi, including yeast-like species such as Aureobasidium
pullulans and Geotrichum candidum, and filamentous yeasts in the genus
Trichosporon, may be 103-104 g-l (Flannigan et al., 1982). In a survey of
finished malts, the moulds isolated most frequently from kernels by Gyllang
and Martinson (1976b) were Eurotium (Aspergillus) amstelodami, Asp.
fumigatus and Rhizopus spp. These authors also found that the percentages
of kernels bearing the three categories quoted were much greater in malt
from one Swedish malting plant than from another, where the most fre-
quently isolated types were Penicillium spp. and Alt. alternata. Relatively
high frequencies have been noted for Alt. alternata and Fusarium spp. in some
Scandinavian malts (Gyllang et al., 1981), and Cochliobolus sativus
102 The microflora of barley and malt
(H. sativum) in a distiller's malt (Flannigan et al., 1982), but in sulphured
malts the moulds most often on kernels appear to be mucoraceous types and
Eurotium spp., with Penicillium spp. usually being less common and field
fungi present on small numbers of kernels (Douglas, 1984). It is to be
expected that kilning temperature and procedure will strongly influence the
final microflora: more microorganisms are likely to survive the lower kiln-
ing temperatures used in the production of distiller' sor high-diastaticmalts.
The microbiological status of malt reaching a brewery or distillery will
depend very much on its handling after production. Handling operations
result in cross-contamination between kernels, and also fresh aerial con-
tamination. In addition, malt in transit usually picks up moisture. It was
observed that malt exported to Nigeria in sacks which became damaged
showed greater levels of storage fungi than in the reference sample retained
in the UK (Flannigan et al., 1984). Badly stored malt is also likely to show
increasing colonization by storage fungi. At RH > 79%, malt becomes
progressively more colonized by storage aspergilli, including Asp. candidus
and Asp. versicolor (Flannigan et al., 1982), and can support growth of the
toxigenic species Asp.flavus (Flannigan, Healy and Apta, 1986).

4.3 EFFECTS OF MICROORGANISMS ON MALTING

4.3.1 Water sensitivity and dormancy

One explanation put forward to account for water sensitivity in barley is
the presence in the lemma, palea and pericarp testa of large populations of
field microorganisms which compete for available oxygen during and
immediately after harvest, when the numbers of viable field fungi and
bacteria on and in the kernel are also greatest, and water sensitivity disap-
pears with time, just as with time the viability of the field microflora
declines (Lutey and Christensen, 1963). When water-sensitive barley is
placed in an excess of water, the limitation of entry of oxygen to the embryo
results in the kernel either failing to germinate or germinating only slowly.
Under these circumstances, some microorganisms are able to invade the
embryo, and the kernel loses its viability.
In a paper where they reviewed earlier work and discussed various
methods of dealing with water sensitivity, Gaber and Roberts (1969) re-
ported that combinations of antibacterial and antifungal antibiotics were
effective in overcoming the inhibition of germination. They concluded that
it was necessary to suppress the activity of both bacteria and fungi. How-
ever, Matthews and Collins (1975) found that antimicrobial compounds did
not eliminate differences in oxygen uptake between water sensitive and
control barleys, although they did reduce water sensitivity to some extent,
and considered that it was characteristics of the kernel, rather than the
presence of microorganisms, which determined whether it was water
 Effects of microorganisms on malting 103

sensitive or not. Nevertheless, in experiments on field germination with

Gliocladium roseum, a saprotrophic mould which has been isolated from
barley (Pepper and Kiesling, 1963), Lynch and Prynn (1977) noted that the
effects of low oxygen availability on germination and viability were ex-
acerbated by the presence of fungi. Harper and Lynch (1981) have since
observed that this organism grows at the embryo end of the kernel, i.e. at
the point of entry of oxygen. From measurements of oxygen uptake of
kernels and isolated moulds they have concluded that, if in imbibed barley
100 Ilg of mycelium of G. roseum are present below the husk, the require-
ment of the fungus for oxygen will be nearly twice that of the embryo; with
the rapidly growing Mucor hiemalis it will be approximately x 4; and with
a slower growing Penicillium sp. about two-thirds of the requirement of the
embryo. Earlier, Harper and Lynch (1979) had demonstrated that the
nitrogen-fixing bacterium Azotobacter chroococcum could inhibit germina-
tion by competing for oxygen, its effects being greater at low oxygen
concentrations. It therefore appears that both bacteria and fungi could play
a part in causing water sensitivity, but the evidence that they are the sole
cause is far from conclusive.
In a recent series of papers, Briggs and his colleagues (Kelly and Briggs,
1992a, b, 1993; Doran and Briggs, 1993; Briggs and McGuiness, 1993) have
re-examined the role of microorganisms, noting that during malting the
germination of undried dormant barley appears to be the result of excessive
microbial growth occasioned by aeration and mixing. They have demon-
strated that controlling microbial growth by antimicrobials leads to
increased germinability and vigour after germination, as well as improving
a-amylase levels. Experiments with de-embryonated grain indicate that the
surface microflora is one factor influencing responsiveness of the aleurone
to gibberellic acid. The work has confirmed that steeping with antibiotics
results in improved germination, especially in the case of dormant barley,
and, as microorganisms at the surface of the grain have a substantial uptake
of oxygen, it has been concluded that they are a major cause of dormancy.

4.3.2 Seedling blight fungi

In the field, seedling blight and root rot caused by Fusarium spp. and
Cochliobolus sativus are well known. Jorgensen (1983) found that in
Denmark, over a IS-year period, C. sativus (H. sativum), F. avenaceum,
F. culmorum, F. graminearum and Monographella nivalis (F. nivale) were,
together with Pyrenophora graminea (H. gramineum), the most important
seedborne pathogens of barley. He suggested that if >15% of kernels were
contaminated by C. sativus seed barley required treatment with a seed
dressing. Much earlier, Christensen and Stakman (1935) found that as many
as 91 % of kernels in some barleys from Iowa and southern Minnesota were
contaminated by C. sativus (H. sativum), and up to 54% by Fusarium spp.
They observed that there was a negative correlation between percentage
104 The microflora of barley and malt
contamination and percentage germination. In special germination tests to
assess possible effects on malting, badly infected kernels were set to
germinate on moist blotting paper at 16°C. Only a little over two-thirds
germinated, and of these more than 25% showed a stunted, rotted or
aborted primary rootlet and a contorted or very short acrospire. C. sativus
was the more destructive of these two types of field fungi. Fusarium spp.
were nevertheless recognized as lowering germinative capacity and giving
uneven growth on malting floors (Mason, 1923). When fusaria were applied
in steep by Sloey and Prentice (1962), although most isolates did not, single
isolates of F. oxysporum f.sp. lycopersici (F. bulbigenum var. lycopersici),
F. graminearum, Monographella nivalis (F. nivale) and an unidentified species
caused significant reductions in rootlet growth. The presence of Fusarium
spp. and C. sativus (H. sativum), and of Alt. alternata and Cladosporium spp.,
in greater numbers of kernels than in bright barley may also have accounted
for the lower percentage germination and rootlet dry weight in stained and
weathered barley observed by Kotheimer and Christensen (1961).
By inoculating barley in the field with several strains of F. avenaceum and
F. culmorum, Haikara (1983) has been able to demonstrate that subsequent
decreases in the germinative capacity and germinative energy of the con-
taminated grain are more dependent on the weather between anthesis and
harvest than on the strain. For example, with one strain of F. culmorum the
germinative capacity was 99.5% after a drier growing season and 48% after
the following wetter season. The effects of this species on germination were,
in general, greater than those of F. avenaceum. In some instances, there was
a marked disparity between the germinative capacity and the germinative
energy, indicating that the fungi had weakened the kernels, so that they
were able to germinate only under optimum conditions (Haikara, 1983).
A possible explanation for these effects on germination might be the
production of toxic metabolites, or mycotoxins, by the fungi. C. sativus
produces toxic substances to which barley is more susceptible than wheat
or oats (Ludwig, 1957). Nummi, Niku-Paavola and Enari (1975) and
Flannigan, Morton and Naylor (1985) have found that the trichothecene,
T-2 toxin, which is produced by strains of F. sporotrichioides (F. tricinctum),
F. avenaceum, F. culmorum and other fusaria (Smith et al., 1984), retards
acrospire and rootlet development, but Haikara (1983) did not detect
trichothecenes in her barleys or malt. However, when barley was stored
undried before malting, there was a lO-fold increase in the level of
zearalenone (F-2) and a reduction in germinative capacity. Low levels of
this toxin inhibit germination in maize (Brodnik, 1975), so that it may have
had a direct effect in this case.
As indicated in section 4.2.2, and by the fact that Andersen, Gjertsen and
Trolle (1967) noted an increase in the number of superficial spores on
kernels, conditions during production of green malt may be ideal for the
growth of Fusarium spp. They form the 'red mould' which can grow on
malting floors, and in some cases in the past grew to the extent that
 Effects of microorganisms on malting 105
production of satisfactory malt was impossible and maltings were tempo-
rarily closed for disinfection. Although growth of fusaria can be accompa-
nied by the discoloration of rootlets, noted in germination tests on blotters
by Jorgensen (1983), there is no published evidence that they produce
mycotoxins during malting.

4.3.3 Stimulation of germination

Although some of the Fusarium strains examined by Sloey and Prentice
(1962) had a deleterious effect on rootlet growth during malting, the only
isolate of F. culmorum tested significantly enhanced the production of
rootlets. In her experiments with barley inoculated in the field with
F. avenaceum and F. culmorum, Haikara (1983) found that relative to unin-
oculated controls there was enhanced rootlet growth during malting of
grain from a wetter growing season, although there had been a reduction
in the malt produced after the previous drier season. This discrepancy was
probably related to the level of inoculum on the grain: wetter conditions
favour the growth of fusaria. Another species, F. moniliforme (teleomorph
=Gibberella fujikuroi) produces a number of growth-regulating compounds,
the gibberellins (Bu'lock, 1984), which are used to accelerate malting. It may
be that if this species is present in abundance there can be some stimulation
of germination, since Gjertsen, Trolle and Andersen (1965) found that one
F. moniliforme isolate added at steeping apparently stimulated rootlet
growth. However, isolates of other species not recognized as being gibber-
ellin producers gave the same effect. In any event, the effects of gibberellins
could be counteracted by other factors. For example, F. moniliforme is
known to produce a number of toxins (Drysdale, 1984), including fusaric
acid, which is injurious to cell membranes.
In contrast to its phytotoxic properties (Ludwig, 1957), C. sativus
(H. sativum) was shown to produce two metabolites which stimulate sugar
release in de-embryonated barley (Briggs, 1966). It has also been reported
that both living mycelium and extracts of Alt. alternata stimulate elongation
of wheat coleoptiles (Ponchet, 1966). Clad. herbarum produces in culture the
auxins indole-3-acetic acid and indole-3-acetonitrile (Valadon and Lodge,
1970), and apparently has a stimulatory effect on the growth of maize
seedlings (Pidoplichko, Moskovels and Zholanova, 1960). It appears
possible that these last two species, which are often the predominant field
fungi on barley, could have similar effects in malting, although the presence
of other, antagonistic, species could have a modifying effect.

4.3.4 Malt analysis

Weathered barley not only shows poorer germination (section 4.3.1), it also
produces malts with undesirable characteristics, the most widely reported
being high protein modification, and consequently high wort nitrogen
106 The microflora of barley and malt
(Malt Research Institute, 1955; Kneen, 1963). In reviewing the effects of the
microflora of barley on malt and beer properties, Etchevers, Banasik and
Watson (1977) reported that, compared with malts prepared directly from
high-moisture barley, they had found pronounced changes in the charac-
teristics of those prepared after storage of the barley for 1 month. The extract
and wort nitrogen levels were higher, but a-amylase and diastatic power
(DP) were much lower.
When they have been used to inoculate barley in the field (Haikara, 1983)
or in the steep (Prentice and Sloey, 1960; Sloey and Prentice, 1962; Kneen,
1963; Gjertsen, Trolle and Andersen, 1965), Fusarium spp. have been found
to affect malt characteristics. Sloey and Prentice (1962) noted a general
tendency towards increased steep and respiration losses, with a concomi-
tant reduction in malt recovery. The greatest reduction was seen with an
isolate of F. culmorum, where the steep and respiration losses were accom-
panied by greater rootlet production. Earlier, Prentice and Sloey (1960) had
found that isolates of F. graminearum and the gibberellin producer,
F. moniliforme increased both a-amylase and DP. In their second paper they
reported that malt prepared from barley inoculated with further isolates of
these species, and F. culmorum, F. equiseti, F. poae and unidentified species,
also showed higher a-amylase and extract values. Using some of these
strains and others from Danish barley, Gjertsen, Trolle and Andersen (1965)
confirmed that there were increases in a-amylase in most cases, and in DP
in a few cases. Kneen (1963) noted that an unidentified Fusarium sp. gave a
distiller's malt with a greater DP. Moreover, Haikara (1983) found that
malts produced from barley inoculated with F. avenaceum and F. culmorum
showed a higher degree of modification, although in some cases the
a-amylase activity and DP were lower than in controls. This was attributed
to fungal enzyme activity undetected by the analytical methods employed.
In the investigations mentioned above, increased protein modification and
wort nitrogen have also been observed, and F. graminearum and
F. moniliforme have been noted in particular (Prentice and Sloey, 1960; Sloey
and Prentice, 1962).
Few of the wide range of other microorganisms have as yet been
implicated in alterations in malt quality. Prentice and Sloey (1960) exam-
ined nearly 100 bacteria, yeasts and moulds, and found that, apart from
fusaria, only a relatively small number affected wort nitrogen and none
raised a-amylase or DP. Among those which increased wort nitrogen were
several bacteria, including Pseudomonas spp., and unidentified species of
Aspergillus (two) and Mucor. Kneen (1963) reported that Asp. niger,
Rhizopus arrhizus and unidentified species of Absidia, Trichothecium
(Cephalothecium), Coniothyrium, Helminthosporium and Rhizopus also had
this effect. Eurotium (Aspergillus) amstelodami and Asp. fumigatus, which
were reported to grow during malting (Gyllang and Martinson, 1976a),
were additionally found to raise nitrogen levels and extract yields
(Gyllang, Satmark and Martinson, 1977).
 Effects of the microflora on beer and distilled spirit 107
The extent to which mycotoxins are involved in the effects discussed
above is uncertain. Both Nummi, Niku-Paavola and Enari (1975) and
Flannigan, Morton and Naylor (1985) found in laboratory experiments that
the trichothecene toxin T-2, well known as a potent inhibitor of protein
synthesis, depressed synthesis of a-amylase during malting. Flannigan,
Morton and Naylor (1985) also noted reductions in a-amino nitrogen in the
wort, although not apparently sufficient to alter the gravity of the wort. In
other laboratory experiments, diacetoxyscirpenol (DAS) (Ferguson,
Flannigan and Schapira, 1984) and, to a lesser extent, deoxynivalenol
(DON) (Schapira, Whitehead and Flannigan, 1989) have also been shown
to affect germination and enzyme development during malting of barley.
As with T-2, growth of the coleoptile and rootlets growth is inhibited and
a-amylase and a-amino nitrogen levels reduced as a result of the presence
of these toxins in barley. It can be argued that, although individual myco-
toxins apparently affect malting, the levels of toxin used to spike the grain
in order to demonstrate this are higher than would be encountered in
cereals for malting. However, grain carries mixtures of mycotoxins, not
only because a variety of different moulds are present but also because
particular species produce different ranges of toxins. For example, Haikara
(1983) found that the strain of F. culmorum which had the greatest effect on
the germinative capacity of barley and the characteristics of malt also
produced the greatest amounts of the oestrogenic mycotoxin zearalenone
during malting. However, she did not test for production of deoxynivalenol
(DON), also known to be produced by this species. These two toxins, and
other toxins in natural assemblages, have entirely different modes of action
and target different metabolic functions. It is therefore not surprising that
experimental animal studies have presented evidence that natural mixtures
of mycotoxins are much more potent than the sum of the effects of the
individual toxins might suggest (Flannigan, 1991). Such synergy, resulting
from combinations of low levels of different toxins, might therefore be of
significance to malting, but would make tracing the real effects of particular
species or mycotoxins extremely complex.

4.4 EFFECTS OF THE MICROFLORA ON BEER AND DISTILLED

SPIRIT

4.4.1 Beer
The best known effect of the micro flora of barley and malt is that of reduced
gas stability, or gushing. It is particularly associated with late-harvest areas,
e.g. Scandinavia, and wetter growing seasons. The earlier literature on the
phenomenon was discussed by Gjertsen, Trolle and Andersen (1963), who
investigated its causes in Danish beer and concluded that the problem of
primary gushing arose from the use of weathered or badly stored barley for
108 The microflora of barley and malt
malt production. Prentice and Sloey (1960) investigated the problem by
preparing malts inoculated at steep with a range of organisms, and found
that it was mainly fusaria which created the problem. Further work using
30 Fusarium isolates established that inoculation with many of these
resulted in increased wort nitrogen and formol nitrogen in wort and beer,
and that F. graminearum and F. moniliforme were particularly notable in
causing gushing (Sloey and Prentice, 1962). Using some strains from these
workers, Gjertsen, Trolle and Andersen (1965) established that gushing was
not dependent on when the inoculations were made during steeping, and
also that it was the result of interaction between the mould and the germi-
nating barley. It has since been demonstrated that malt made with field-
inoculated barley can give rise to gushing (Haikara, 1983). Haikara found
that gushing caused by F. avenaceum and F. culmorum was both weather and
strain dependent and that there appeared to be a relationship between
gushing and the ability of the strains to produce the mycotoxin zearalenone.
It was noted that gushing induced by F. graminearum-contaminated malt
was unpredictable; it could occur shortly after bottling or several weeks later
(Donhauser et al., 1989). Niessen et al. (1992) confirmed that strains of
F. graminearum and F. culmorum in German grain could initiate gushing. It
was subsequently noted by Niessen et al. (1993) that gushing was associated
with DON in malted barley and wheat. Although some gushing beers did
not contain DON and some DON-positive beers were non-gushing, taken
overall, the concentration of the toxin in gushing beers was found to be
significantly higher than in non-gushing beers.
Other field fungi have also been found to be capable of causing gushing.
Culture filtrates of Alternaria sp., Nigrospora sp. and Stemphylium sp. caused
gushing when added to normal beer (Amaha et al., 1973; Yoshida et al.,
1975). Various storage fungi have also been found to have the same poten-
tial, including Eurotium spp. (A. glaucus group), P. chrysogenum (including
strains previously called P. notatum), P. griseoroseum (P. cyaneofulvum) and
Rhizopus sp. (Amaha et al., 1973; Yoshida et al., 1975; Fukushima, Kitabatake
and Amaha, 1976). E. amstelodami and Asp. fumigatus added in the steep
were also found to cause gushing (Gyllang and Martinson, 1976a), and
were considered to be responsible for periodic episodes of this in a Swedish
brewery, where they appeared to be the predominant species on malt.
In investigating the cause of gushing, Amaha et al. (1973) found that a
Fusarium isolate produced at least two gushing inducers, one of which was
a peptide-containing substance of low molecular weight. A polypeptide
(MW about 15 000) which induced gushing was isolated from culture
filtrates and grain contaminated with Nigrospora sp., which was originally
identified wrongly as being Rhizopus sp. (Amaha et al., 1973, 1974). The
purified substance induced vigorous gushing in normal beer at concentra-
tions as low as 0.05 ppm. The isolate was morphologically very similar to
type strains of N. oryzae, one of which was also found to produce gushing
inducers (Kitabatake and Amaha, 1974). A gushing-inducing peptido-
 Effects of the microflora on beer and distilled spirit 109
glycan causing gushing at a level of 4 ppm was isolated from Stemphylium
sp. (Amaha et al., 1973; Kitabatake and Amaha, 1976).
Elevated nitrogen levels have been mentioned above in relation to
Fusarium spp., but Kneen (1963) found that Asp. niger and Rhizopus arrhizus
also increased nitrogen in beer. This last species and, particularly, Eurotium
rubrum (Asp. ruber) and Asp. ochraceus caused reduced haze stability, al-
though unidentified species of Alternaria, Penicillium and Geotrichum in-
creased the stability. In addition, Asp. ochraceus inhibited the action of the
beer stabilizer papain. Asp. fumigatus and R. oryzae have also been found to
increase a-amino nitrogen and soluble nitrogen in beer (Gyllang, Satmark
and Martinson, 1977).
The effect of moulds on beer flavour was studied by Kneen (1963), who
observed that Asp. niger, Asp. ochraceus, R. arrhizus and unidentified species
of Cladosporium, Coniothyrium and Fusarium were responsible for strong
off-flavours in beer. These ranged from 'molasses, burned' in the first case
to 'unclean, winey, harsh' in the last. Isolates of Absidia, Trichothecium
(Cephalothecium), Cladosporium (Hormodendron) and Rhizopus gave slight
off-flavours. Beer brewed with malt contaminated with R. oryzae was
reported by Gyllang, Satmark and Martinson (1977) to be distinctive,
without having any special off-flavour, butA.fumigatus gave a pronounced
roughness and a stale flavour.
Colour is also affected by the presence of moulds. Kneen (1963) noted
that malt prepared from weathered barley produced beer with a high
colour, as was the case when malt prepared after inoculation with Asp.
niger, Trichothecium (Cephalothecium), Rhizopus and, especially, Fusarium sp.
or R. arrhizus was used. Increased colour has subsequently been observed
when Asp. fumigatus, R. oryzae (Gyllang, Satmark and Martinson, 1977),
F. avenaceum and F. culmorum (Haikara, 1983) have been present.
Having mentioned earlier the apparent association between two
Fusarium mycotoxins and gushing, it is appropriate to consider here the
possibility that, as various mycotoxins are known to be toxic to a wide range
of organisms, including yeasts, there may be some effect on fermentation.
The fact that maltsters in Western countries use high quality cereals might
be considered to preclude the possibility of appreciable levels of myco-
toxins passing into commercial fermentations. However, grain that is used
as a cheap starchy adjunct, e.g. maize, may be of much lower quality and
may therefore contribute mycotoxins to the wort. It has been shown that
mycotoxins such as zearalenone and trichothecenes can have a
concentration-dependent effect on yeast growth (Schappert and
Khachatourians, 1983, 1984; Flannigan, Morton and Naylor, 1985). The
effects of different trichothecene toxins on log phase growth of Saccharo-
myces cerevisiae (Flannigan, Morton and Naylor, 1985) reflect their potency
against animal systems. At comparable concentrations, T-2 strongly inhib-
ited growth, DAS (Flannigan et al., 1986) also retarded growth, but to a
lesser extent, and DON had very little effect (Whitehead and Flannigan,
110 The microflora of barley and malt
1989). With another yeast, KIuyveromyces marxianus, the inhibition pro-
duced by T-2, i.e. the reduction in dry mass relative to toxin-free controls
over a period of 6 h at 25°C, was approximately 13 times that by DAS
(Flannigan and Barnes, 1990). The difference in potency between members
of the trichothecene family is also seen when yeast growth is allowed to
continue into the stationary phase. For example, at the extremely high DON
concentration of 50 Ilg DON mr1 culture, the dry mass of the yeast was
reduced by only 10% compared with toxin-free controls, i.e. not much more
than one-half of the reduction caused by 2.5llg T-2 ml-1. The degree to
which yeast viability is affected depends on both the toxin and its concen-
tration, with DON having only slight effects (Whitehead and Flannigan,
1989; Flannigan, 1989).
Although Lafont, Romand and Lafont (1981) reported thatT-2 and DAS
had greater effects than aflatoxin Bl and patulin on carbon dioxide evolu-
tion by S. carisbergensis, their results for T-2 and DAS actually show that the
velocity of gas production began to recover towards the end of the 4 h
experiments, even when toxin was present at 50 Ilg mrl. In laboratory
experiments with S. cerevisiae and K. marxianus, T-2 toxin was found to
inhibit fermentation initially, causing a concentration-dependent lag in the
attenuation of glucose and production of ethanol in Wickerham medium
(Flannigan et al., 1986; Flannigan and Barnes, 1990); at the high level of
10 f.lg T-2 mrl medium a 42 h lag in controls lengthened by around 120 h
(Schapira, 1985). Despite this extended lag, however, the rate of attenuation
recovered to roughly the same rate as controls (Flannigan et al., 1986), this
recovery being associated with a time-dependent increase in viable cell
number (Schapira, 1985). Koshinsky, Cosby and Khachatourians (1992)
have confirmed that ethanol production per se is not affected by T-2, and
that any apparent inhibition of production results from inhibition of yeast
growth.
As in laboratory malting investigations, DAS had a similar but less
potent effect on fermentation than T-2 (Schapira, 1985; Flannigan, 1989),
but concentrations of DON as high as 20 Ilg ml-1 had little effect on attenu-
ation and ethanol production (Whitehead and Flannigan, 1989). Also as in
malting investigations, the concentrations of these toxins producing the
effects mentioned above are considerably higher than those encountered
in even very heavily contaminated grain. However, as Flannigan (1989) has
suggested, combinations of lower concentrations of different toxins acting
in concert might contribute to slow starting or 'sticking' of fermentations.
Other experiments (Schapira, 1985) indicate that increasing the pitching
rate would probably overcome the effects of these combinations.

4.4.2 Distilled spirit

Using a ratio of 90 parts cooked maize to 10 parts lightly kilned distiller's
malt in the mash, Kneen (1963) found that moulds appeared to have little
 Health hazards 111
effect on alcohol yield. He did note, however, that even 10% of heavily
mould-contaminated malt could give off-flavours in the distillate. Isolates
of Cladosporium, Fusarium, Coniothyrium and the yeast Rhodotorula pro-
duced marked off-flavours, with Asp. niger and an unidentified species of
Alternaria giving less pronounced effects. There was not necessarily a
correlation between these results and those for beer, e.g. Rhizopus arrhizus
had an effect on the flavour of beer, but not of distillate.
Although Harrison and Graham (1970) stated that it requires a very high
level of contamination in wort - about 108 bacteria mr! - before there is
substantial competition with the yeast during fermentation, Dolan (1979)
has pointed out the need to have the millroom and grist cases separate from
the mash house and tun room in order to reduce contamination by
microorganisms present in airborne malt dust.

4.5 HEALTH HAZARDS

4.5.1 Respiratory hazards

The hazards to the health of people in the grain industry arising from
inhalation of high concentrations of grain dust over long periods have been
reviewed by Dennis (1973). Some idea of the degree of hazard can be
gauged from the fact that, in Saskatchewan, Williams, Skoulas and
Merriman (1964) found that 54% of some 500 elevator agents had a history
of a persistent cough, wheezing, breathlessness, dermatitis or 'grain fever'.
In a follow-up investigation (Skoulas, Williams and Merriman, 1964), it was
found that there was less impairment of lung function in those without
these symptoms than in agents exhibiting them. Chronic bronchitis, pul-
monary fibrosis and emphysema regularly develop over a period of years
in such elevator agents.
Grain dust consists of more than particles of grain and chaff: soil frag-
ments, bacteria, yeasts, spores and mycelium of moulds and actinomycetes,
mites and insects, and their fragments and products are all present. It is
now well understood that workers handling grain can become sensitized
to microorganisms (Lacey, 1975) and mites (Dennis, 1973; Wraith,
Cunnington and Seymour, 1979) in grain and develop allergic conditions.
Although skin allergies may develop, allergic reactions most frequently
occur in the respiratory system. Where the allergic reaction is provoked in
the lungs, repeated exposure to the antigen, or allergen, which causes the
allergy leads to degenerative changes in the lung tissue and serious
respiratory impairment.
The spores of certain fungi and actinomycetes are potent sources of
allergens, which are frequently associated with the spore wall. After inha-
lation, spores which are larger than 10 11m in diameter are largely deposited
on the mucous membranes of the nasopharynx. Any allergic reaction they
112 The microflora of barley and malt
cause will be of a rhinitis or hay fever type. Spores in the range 4-10 Ilm are
mainly deposited in the bronchi and larger bronchioles and provoke an
asthmatic type of response, in which there is difficulty in breathing (dys-
pnoea). In both these cases the allergic responses are Type I reactions, i.e. the
response is evident more or less immediately after exposure to the allergen
and the effects last only a few hours. The Type I reaction is characteristic of
the 10% of the population who are atopic. Such individuals are usually
allergic to a wide range of allergens and react to spore concentrations which
do not normally affectthe remainder of the population, <10 4-10 6 m -3 air.
The greatest deposition of spores which are less than 4 Ilm in diameter
occurs on the respiratory surface of the terminal bronchioles and air sacs,
the alveoli; if the spores provoke an allergic response the condition is
described as extrinsic allergic alveolitis. This is a Type III reaction and
usually occurs when the non-atopic individual is exposed to greater con-
centrations (>106 m-3) of a particular type of spore containing a single
specific allergen. Unlike the Type I reaction, it does not develop for about
4 h and may continue for 48 h or more. The Type III reaction is also unlike
Type I because it is mediated by precipitating antibodies, and there is tissue'
damage. In the lung, there is a perivascular inflammatory response and
granulomata are formed. The overt symptoms are fever, malaise, dry cough
and breathlessness. With repeated attacks, progressive fibrosis of the lung
occurs: it is permanently damaged, its efficiency is considerably reduced
and there is increasing dyspnoea. Extreme respiratory distress is experi-
enced if both Type I and Type III responses are provoked in the same
individual, although this is a relatively rare event.
The types of spore associated with barley which are most likely to
provoke Type I reactions include the common field fungi Alt. alternata and
Clad. herbarum, with Epicoccum nigrum (E. purpurascens) causing rhinitis
only. Type III reactions can be caused by the spores of Penicillium spp.
and the thermophilic actinomycetes Micropolyspora faeni and Thermo-
actinomyces vulgaris, both of which are associated with self-heated barley
and fodders (Lacey, 1975) and are the causal agents of farmer's lung.
Although its spores are in the size range associated with Type III reactions,
the storage fungus Asp. jumigatus may provoke both types of reaction. The
clumping together of spores, with consequent deposition in the upper part
of the respiratory system, may account for its involvement in a Type I
reaction.

4.5.2 Maltworker's lung

Although Bridge (1932) noted the incidence and severity of chronic respi-
ratory diseases, such as bronchitis and emphysema, in maltworkers, it was
not until 1968 that the condition known as maltworker's lung was de-
scribed (Riddle et al., 1968). This extrinsic allergic alveolitis is now
recognized as an occupational disease, like farmer's lung. It is caused by
 Health hazards 113
exposure to high concentrations of the spores of Asp. clavatus. This organ-
ism appears to be relatively rare in dry grain. It has, however, been found
in barley in Canada (Wallace, 1973), Egypt (Abdel-Kader et al., 1979) and
Romania (Stankushev, 1969).
After first reports of the occurrence of this type of allergic alveolitis
among workers in floor maltings (Riddle et al., 1968; Channell et al., 1969),
56 Scottish maltings were surveyed by Grant et al. (1976). Grant and his
colleagues only detected Asp. clavatus in 12 of these. Of 711 maltworkers in
the 56 maltings, nearly as many in box maltings (6.2%) as in floor maltings
(6.8%) were affected by the disease. Overall, 5.2% of the workers surveyed
had symptoms of maltworker's lung, but the incidence was lowest (1.1 %)
among those where modern mechanical systems were employed, e.g. drum
maltings.
Where Asp. clavatus has become established in floor maltings, it can be
seen sporing profusely on the surface of kernels. In consequence, clouds of
spores are released into the atmosphere during turning of green malt
(Riddle et al., 1968). Other operations which further increase the spore load
in the atmosphere are loading and stripping the kiln, and screening off the
culms and rootlets. It is to be expected that enclosure of the malting barley,
e.g. in drums, will result in a reduction of the number of spores in the air.
The spore walls are particularly rich in allergenic substances which induce
alveolitis (Blyth, 1978). Although it appears that maltworker's lung results
from a Type III reaction, it has been suggested that the allergens in the
spores also provoke a Type I reaction, which may be a component of the
disease (Grant et al., 1976).
As Asp. clavatus is rarely found in dry barley, the primary source of
contamination of malting premises remains uncertain. Riddle et al. (1968)
considered that pigeons which had free access to the grain store might have
been implicated in one maltings, and it was suggested that the fungus
might have been introduced with a particular consignment of grain in
another (Channell et al., 1969). Whatever the case, it is certain that once Asp.
clavatus is introduced into a maltings it meets with favourable conditions
for growth and sporulation. Although it only grows during germination of
the grain, and perhaps also in the early stages of kilning, its spores are easily
spread throughout the premises. In addition to isolating the organism from
all parts of a distillery maltings, Channell et al. (1969) detected the organism
in the mash house. They also isolated it from the sputum of employees who
did not work as maltmen.
The growth of the mould becomes a particular problem when grain is
germinated at elevated temperatures, e.g. when the temperature is deliber-
ately raised to speed up malting (Riddle et al., 1968), or when, during warm
weather, there is no means of preventing it from rising above the normal
16°C (Flannigan et al., 1984). It has been found in experimental malts that
there may be a 1Q4-fold increase in the viable counts of this organism if the
malting temperature is raised from 16°C to 25°C (Flanniganet al., 1984), the
114 The microflora of barley and malt
optimum temperature for growth of the fungus. Consequently, it is not
surprising that the organism is one of the most frequently isolated toxigenic
fungi (see section 4.5.3) in sorghum malt in South Africa (Rabie, Thiel and
Marasas, 1983; Rabie and Lubben, 1984), where the temperature of malting
may be 28°C. The presence of cracked corns (Riddle et al., 1968), pre-
germinated grain or accidentally introduced finished malt (Flannigan et al.,
1984), which provide easily colonized loci from which contamination can
spread, may also lead to greater growth on the germinating barley. When
the malting industry is taken as a whole, it would appear that the incidence
of this allergenic fungus is low. However, when it does appear, the fact that
its spores are spread so readily makes control of this organism extremely
difficult. Even dry grain stored on the premises becomes contaminated by
spores prior to malting, and addition of hypochlorite to the steep water in
the concentrations often employed is not effective in killing these spores
(Flannigan et al., 1984). Riddle et al. (1968) have suggested that dilute
hypochlorite solutions may directly stimulate growth of Asp. clavatus, but
it appears more likely that hypochlorite promotes growth by reducing the
numbers of other microorganisms competing with it. The higher concen-
trations used in cleaning floors or plant generally give only temporary relief
from the problem, since the spores are still present on inaccessible surfaces.
Caustic disinfectants are probably more effective than hypochlorite, but
scrupulous attention to dust control and hygiene is necessary if the organ-
ism is to be controlled. Measures taken to control Asp. clavatus would also
reduce the prevalence of other airborne fungi which were commonly
isolated from the sputum of Scottish maltworkers and acted as sensitizing
agents (Blyth et al., 1977).

4.5.3 Mycotoxins
In the preceding sections, the effects of recognized mycotoxins and other
mould metabolites on germination and malt characteristics were men-
tioned, and it is therefore appropriate to consider mycotoxins associated
with brewing materials as they relate to health. Viewed on a worldwide
scale, the most important and best researched of mycotoxins found in
cereals (Flannigan, 1991) are the aflatoxins produced by Asp. flavus and Asp.
parasiticus. They are not only toxic but also potent carcinogens. Other
important mycotoxins are produced by various species of Fusarium associ-
ated with grain, including the oestrogenic toxin zearalenone, various
trichothecenes, which are potent inhibitors of protein synthesis and are also
immunosuppressive, and a group of toxins first reported in 1988, the
fumonisins, which if not directly carcinogenic are cancer promoters. Two
nephrotoxins, citrinin and ochratoxin A, produced by P. verrucosum and
some other penicillia are not uncommonly found in European barley.
In section 4.3 it was noted that Asp. clavatus is toxigenic as well as
allergenic, and it is therefore important to recognize that inhalation of the
 Health hazards 115
spores which cause the allergic disease maltworker's lung may also be a
means by which its toxins enter the human system. Indeed, the author has
been told that some victims of maltworker's lung also exhibited neuro-
logical symptoms, including tremor. Among the known mycotoxins which
the fungus can synthesize are patulin, cytochalasin E and two tremorgenic
compounds, tryptoquivaline and tryptoquivalone (Glinsukon et al., 1974;
Clardy et al., 1975). Recently, it has been shown in a laboratory investigation
of patulin and cytochalasin E production that, although there are strain
differences, the fungus can produce these toxins during malting of both
barley and wheat (T.M. Lopez, unpublished results). It is not known what
the effects of natural mixtures of these various toxins are on humans, but
mycotoxicosis and tumour development have been observed in experi-
mental mice inoculated nasally with spores (Blyth and Hardy, 1982). This
finding gives further reason for the use of efficient facemasks, respirators
or air-flow helmets in maltings.
There may also be hazards to stock fed on malting by-products contam-
inated with this organism. Blyth et al. (1977) were able to isolate the mould
from screened-off culms and rootlets, and in a recent investigation of a
mycotoxicosis in the UK (Gilmour et al., 1989) culms were found to be
contaminated by the organisms at a level of 15 x 106 CFU g-1. In this case, a
supplementary feed containing culms from a distillery maltings as the
principal ingredient was the cause of death among cattle, but not among
sheep consuming the feed. However, in another very recent case there was
96% mortality in sheep fed on sprouted barley grains from an Israeli factory
producing malt extract (Shlosberg et al., 1991). In addition to Asp. clavatus
growing and contaminating coleoptiles and rootlets during malting, it is
also possible that it could adventitiously contaminate culms subsequent to
kilning and later grow under damp storage conditions. It has been reported
that the organism is able to grow on culms stored at RH 92.5% or higher,
i.e. culms with an aw of 0.925 or above will support growth and toxin
production (Flannigan and Pearce, 1994).
It is not just culms or sprouted grains which have been involved in
mycotoxicoses, however. Wet residues from sorghum beer production
which had been spread out to dry and were found to be contaminated with
Asp. clavatus were the cause of fatal tremorgenic disease among cattle
(Kellerman et al., 1976). In this case, the residues had been spread out to
dry, but it is not known whether the mould was present in the residues
beforehand or whether it contaminated the material during drying. Which-
ever, the practice of spreading the residues out to dry was considered to
favour growth of the mould (Kellerman et al., 1976). In laboratory experi-
ments, Flannigan and Pearce (1994) found that Asp. clavatus could also grow
on brewer's grains, but treatment of this material with propionic acid,
which is applied to moist-stored barley for animal feed, could prevent
growth of the fungus.
As far as beer is concerned, Flannigan (1989) has pointed out that the
116 The microflora of barley and malt
effects of some natural toxins in beer were recognized a long time ago,
although their existence was not. For example, the use of the grass Lolium
temulentum (temulentia =drunkenness) to 'fortify' or enhance the effects of
beer was prohibited in France in the thirteenth century during the reign of
Louis IX. In 1669, it was written that the presence of this 'weed or grayne'
in barley crops in the west of Scotland had 'such a mischievous effect that
one gill of ale or bere wherin such grayn hath been will fuddle a man more
than a gallon of other drink' (Stones, 1984). Although it has not yet been
demonstrated, there is a distinct possibility that the toxins in L. temulentum
are alkaloid mycotoxins produced by endophytic fungi similar to those
responsible for outbreaks of mycotoxicosis in stock feeding on infected
pasture grasses. Somewhat more recently, circumstantial evidence indi-
cated that a case of acute human illness in the USA was attributable to
mycotoxins in beer (Cole et al., 1983). An individual who drank a single can
of beer developed neurological and other symptoms of toxicosis, which
persisted for more than a day. A pellicle-like growth of P. crustosum was
subsequently found in the empty can. When grown in a medium containing
beer, the fungus produced four alkaloid mycotoxins which are active
against the central nervous system, viz. festuclavine and three
roquefortines. Since the can was reportedly normal at opening, being fully
filled and carbonated, the circumstances under which the organisms
entered, grew and produced toxin remain a mystery.
It is clear from various surveys which will be mentioned later that a
variety of mycotoxins do reach the final product, beer, but generally only
in limited concentrations. In experiments in which toxin was added at
various stages during the malting and brewing processes, Gjertsen et al.
(1973), Krogh et al. (1974), Chu et al. (1975) and Nip et al. (1975) found that
little aflatoxin B1, citrinin or ochratoxin A could be detected in beer. It
appeared that about 40% of ochratoxin A in malt remained in the spent
grains, with a further third being degraded and up to 20% being recoverable
from the yeast after fermentation (Nip et al., 1975). It was also noted
(Gjertsenet al., 1973; Kroghet al., 1974) that, where high levels of ochratoxin
A or citrinin were naturally present, the grain would have been unaccept-
able for malting because of its appearance or failure to germinate.
Flannigan, Morton and Naylor (1985) deduced from work with toxin-
spiked barley that most zearalenone and trichothecene T-2 toxin would be
lost or degraded during malting, and El-Banna (1987) reported that 77% of
the trichothecene DON added to barley was destroyed during germination.
Mannio and Enari (1973) observed that no zearalenone or T-2 appeared to
pass into beer from malt prepared from barley heavily contaminated with
F. culmorum.
A group of mycotoxins which are currently the object of considerable
research effort are the fumonisins produced by F. moniliforme, which were
first discovered in South Africa in 1988. These toxins, which can cause fatal
mycotoxicoses in pigs and horses, are of particular concern because they
118 The microfLora of barley and malt
possible reason for this is that zearalenone in wort is apparently rapidly
reduced to zearalenol during fermentation (Scott et al., 1992). In their study
of 49 European light lagers, Payen et al .. (1983) did not find other Fusarium
toxins in more than a few samples either, i.e. the trichothecenes DON, DAS
and T-2 were present in only one, two and three samples, respectively.
When Cerrutti et al. (1987) carried out a thin layer chromatography exam-
ination of 24 beers imported into Italy, as in an earlier investigation of
Italian beers, they failed to detect any of these Fusarium toxins or aflatoxins,
sterigmatocystin, citrinin and ochratoxin A. However, the last-named
toxin, most probably originating from grainborne P. verrucosum or other
penicillia, but possibly from Asp. ochraceus, was found in five lagers tested
by Payen et al. (1983). Tressl, Hommel and Helak (1989) were also able to
detect ochratoxin in beer by high performance liquid chromatography, as
well as in barley and malt. The detection limit in all three was 51lg kg-I.
The European evidence on zearalenone is backed up by a recent
Canadian survey of 50 samples of beer brewed in or imported into Canada
(Scott, Kanhere and Weber, 1993). No zearalenone was detected, but the
investigation did reveal DON in 29 samples, at levels ranging from 0.33
to 50.3 ng ml- I, and nivalenol in three, at 0.10--0.84 ng mrl. The incidence
of DON in Canadian cereals has been of concern in recent years, and Scott
et al. (1992) showed that in Germany gushing in beer was associated with
the abundance of this toxin in cereals in 1987 (Lepschy-von Gleissenthal
et al., 1989; Muller and Schwadorf, 1993) and 1991 (Niessen et al., 1993).
Other Fusarium toxins found in the German crops included
3-acetyldeoxynivalenol (3-AcDON), DAS, nivalenol, T-2 and zearalenone
(Lepschy-von Gleissenthal et al., 1989; Muller and Schwadorf, 1993). In a
survey of 196 samples of beers from several breweries, Niessen et al. (1993)
found that three-quarters of the beers brewed from wheat malt (WeiB Bier)
and around one-quarter brewed from barley malt contained DON. As the
breweries were mostly in southern Germany and around 20% of the beers
had been selected on suspicion of gushing, the amounts of toxin detected
give rather a skewed picture. Nevertheless, the levels in the wheat beers
(mean 0.25 ng ml- I; maximum 0.57 ng mrl) were significantly higher than
in beers derived from barley malt (0.15 ng mrl; 0.48 ng mr l).
It is impossible to assess the real importance of mycotoxins in beer to
human health, but it is worth noting again that yeasts may bring about
conversion of mycotoxins to both less and more toxic compounds, which
may not be assayed. The oestrogenic and tumour-promoting toxin
zearalenone can be taken as an example. In suggesting that there was a risk
from mycotoxins in beer Schoental (1984) commented on the fact that the
toxin is relatively common in barley and may be produced during malting
(Haikara, 1983), but yet is apparently uncommon in beer. She suggested
that a reason for this is that during fermentation it is reduced to
a-zearalenol, a compound which apparently is ten times more oestrogenic
than the zearalenone from which it is derived and therefore presents a
 Health hazards 117
are cancer promoters and are widespread in maize in South Africa, Europe
and the USA. Studies carried out at the USA Department of Agriculture
fermentation laboratory in Peoria by Bothast et al. (1992) on a-amylase-
treated maize with heavy natural fumonisin Bl contamination showed that
little of the toxin was degraded during fermentation to produce industrial
ethanol. Some 85% of the toxin was recoverable, with most of this being in
the distiller's dried grains, and much smaller quantities in the thin stillage
and distiller's solubles. These results were in line with earlier investigations
at that laboratory on utilization of aflatoxin- and zearalenone-contami-
nated maize, in that there was accumulation of toxin in the spent grains and
none in the distilled ethanol. Taking into account this and other evidence,
it is clear that mycotoxins are most unlikely ever to appear in distilled spirit.
However, zearalenone, DON and another trichothecene, nivalenol, have
been found together in Korean malt (Lee et al., 1985, 1986), and Lovelace
and Nyathi (1977) detected up to 4.6 mg zearalenone rl in some home-
brewed opaque maize beers in Zambia. In a survey of pito beer prepared
from millet and/ or red guinea com in Nigeria, zearalenone was found in
the products of 28 out of 46 breweries, although at 12.5-200 llg rl the
amounts were much lower than the extreme values in Zambia (Okoye,
1986). Subsequently, it was discovered that 51% of zearalenone in malt
carries over into the final product (Okoye, 1987).
The survey of Okoye (1986) also confirmed earlier work which showed
that contamination by aflatoxins was extensive (Okoye and Ekpenjong,
1984). In this first study, all 23 beers examined contained aflatoxin BI. The
concentrations ranged from 1.7 to 137 llg rl and compared with the
92-262 llg aflatoxin rl reported by Alozie, Rotimi and Oyibo (1980) for
traditional Nigerian beers. These concentrations contrast with those in a 5
year survey of commercial sorghum beer brewing in South Africa (Trinder,
1988). On average, the aflatoxin Bl content of sorghum malt in this much
larger survey was 2.18 llg kg-I, with 68% of malts containing ::;1 llg kg-I,
20% having 1-3 llg kg-1 and only 4.5% yielding more than 10 llg kg-I. While
39% of samples of strainings, normally separated from the wort before
fermentation and bottling and then dried, contained the toxin (0.3-6.0 llg
kg-I), it was only detected in 5% of sorghum beers (0.05-0.13 llg kg-I). On
the basis of these results, it was concluded that consumption of sorghum
beer was unlikely to constitute a health hazard in South Africa. In Europe,
no aflatoxins were detected in 86 beers from various countries and 88
samples of grain, malt, hop pellets and other brewing materials (Woller and
Majerus, 1982). However, using an immunochemical method with a detec-
tion limit of 0.2-1.0 llg kg-I, Fukal, Prosek and Rakosova (1990) did detect
aflatoxins in one-third of 37 malting barleys and one-fifth of 42 malts,
although none was found in any of 34 beers examined.
In contrast to the apparent widespread incidence of zearalenone in
regional African beers, only one of four surveys of beers in Europe has
detected the toxin, and even then only in one sample (Payen et al., 1983). A
 Assessment of mould contamination 119

greater health risk. As mentioned earlier in this chapter, zearalenone does

appear to be reduced rapidly to zearalenol during fermentation. However,
the principle derivative was ~-zearalenol, not its isomer, a-zearalenol (Scott
et al., 1992), which is much less oestogenic.
At present, the evidence suggests that there is a greater risk to animal
health than to human health from mycotoxins associated with beer produc-
tion. Aflatoxins, fumonisins and zearalenone (Bothast et al., 1992) have been
shown to accumulate, and trichothecenes (Mannio and Enari, 1973) and
ochratoxin A (Nip et al., 1975) to be present in significant amounts in spent
grains. As Bothast et al. (1992) has pointed out, it might be necessary to
detoxify spent grains containing large amounts of mycotoxins, e.g. by
formaldehyde or ammoniation, before they are used to feed farm animals.

4.6 ASSESSMENT OF MOULD CONTAMINATION

Since it is the evaluation of the fungal contamination of barley and malt

which is likely to present the greatest problem to the brewing micro-
biologist, it is worth noting that methods for assessment of the percentage
contamination of kernels by different moulds have been set out in the
European Brewery Convention Analytica Microbiologica, Part II (Moll, 1981).
In addition to direct plating without surface disinfection, Flannigan (1977)
has described dilution plating methods for viable mould counts. A general
text of use for the identification of moulds is Smith's Introduction to Industrial
Mycology (Onions, Allsopp and Eggins, 1981), which includes references to
monographs for the identification of species in large or difficult genera such
as Aspergillus, Fusarium and Penicillium. Another extremely useful aid to
identification is Introduction to Food-Borne Fungi by Samson et al. (1995). In
addition to having extremely well-illustrated keys, this book has chapters
on detection and quantification of moulds and on mycotoxins and their
detection, as well as appended mycological media.
Although for some time there will continue to be reliance on traditional
methods for detection, quantification and identification of moulds, rapid
methods are being developed and adopted. For example, an enzyme-linked
immunosorbent assay (ELISA) has been developed and used at Carlsberg
Breweries for detection of Fusarium in barley and malt (Vaag, 1991; Vaag
and Pedersen, 1993), with very high levels of Fusarium antigen indicating
a high risk of gushing. Another immunological method has been employed
by Schwabe et al. (1993, 1994), who have used a latex agglutination test
employing an antibody to extracellular polysaccharide to assess the prev-
alence of Fusarium in barley, and have detected a quantitative relationship
between test results and DON / AcDON levels. Whilst these immunological
methods are genus specific, there is scope for developing species-specific
methods. Highly specific monoclonal antibody-based ELISA and dip-stick
methods for a single grainborne storage fungus, Penicillium islandicum, have
120 The microflora of barley and malt
already been developed (Dewey et al., 1990). In addition to such immuno-
logical methods, research in various laboratories in Europe and North
America is being directed to using molecular methods in plant pathology
and food microbiology. One type of approach is to develop a rapid ampli-
fied polymorphic DNA method (RAPD) for typing particular organisms.
For example, Ouellet and Seifert (1993) have employed RAPD to distin-
guish the head blight fungus F. graminearum from other species of Fusarium
and to track the organism in the field.

REFERENCES

Abdel-Kader, M.I.A., Moubasher, A.H. and Abdel-Hafez, S.1.1. (1979)

Mycopathologia, 68, 143.
Alozie, T.C, Rotimi, CN. and Oyibo, B.B. (1980) Mycopathologia, 70, 125.
Amaha, M., Kitabatake, K., Nakagawa, A et al. (1973) Proceedings ofthe 14th Congress
of the European Brewery Convention, Salzburg, IRL Press, Oxford p. 38l.
Amaha,M., Kitabatake, K.,Nakagawa, A, Yoshido,J. and Harada, T. (1974) Bulletin
of Brewing Science, 20,35.
Andersen, K., Gjertsen, P. and Trolle, B. (1967) Brewers' Digest, 42,76.
Armolik, N., Dickson, J.G. and Dickson, AD. (1956) Phytopathology, 46, 457.
Blyth, w. (1978) Clinical and Experimental Immunology, 32,272.
Blyth, W. and Hardy, J.C (1982) British Journal of Cancer, 45,105.
Blyth, W., Grant, I. W.B., Blackadder, E.S. and Greenberg, M. (1977) Clinical Allergy,
7,549.
Boruff, CS., Claassen, R.I. and Sotier, AL. (1938) Cereal Chemistry, 15, 45l.
Bothast, R.J., Bennett, G.A., Vancauwenberge, J.E. and Richard, J.L. (1992) Applied
and Environmental Microbiology, 58,233.
Bridge, J.C (1932) Annual Report of the Chief Inspector of Factories, HMSO, London.
Briggs, D.E. (1966) Nature, 210, 418.
Briggs, D.E. and McGuiness, G. (1993) Journal of the Institute of Brewing, 99, 249.
Brodnik, T. (1975) Seed Science and Technology, 3,69l.
Bu'lock, J.D. (1984) In The Applied Mycology of Fusarium (eds M.O. Moss and
J.E. Smith), Cambridge University Press, Cambridge, p. 215.
Cerrutti, G., Vecchio, A., Finoli, C and Trezzi, A (1987) Monatsschrift for
Brauwissenschaft, 40, 455.
Channell S., Blyth, W., Lloyd, M. et al. (1969) Quarterly Journal of Medicine, 38, 35l.
Christensen, CM. (1973) Seed Science and Technology, 1, 547.
Christensen, CM. and Kaufmann, H.H. (1969) Grain Storage, The Role of Fungi in
Quality Loss, University of Minnesota Press, Minneapolis.
Christensen, CM. and Kaufmann, H.H. (1974) In Storage of Cereal Grain and its
Products (ed. CM. Christensen), American Association of Cereal Chemists, St
Paul, p. 158.
Christensen, J.J. and Stakman, E.C (1935) Phytopathology, 25,309.
Chu, F.S., Chang CC, Ashoor, S.H. and Prentice, N. (1975) Applied Microbiology,
29,313.
Clardy, J., Springer, J.P., Buechi, G. et al. (1975) Journal of the American Chemical
Society, 97, 663.
Clarke, J.H. and Hill, S.T. (1981) Transactions of the British Mycological Society,
77,557.
Clarke, J.H., Hill, S.T. and Niles, E.V. (1966) Pest Infestation Research, 1965, 13.
 References 121
Clarke, J.H., Niles, E.V. and Hill, S.T. (1967) Pest Infestation Research, 1966, 14.
Clarke, J.H., Hill, S.T., Niles, E.V. and Howard, M.A.R (1969) Pest Infestation
Research, 1968, 17.
Cole, RJ., Domer, J.W., Cox, RH. and Raymond, L.W. (1983) Journal of Agricultural
and Food Chemistry, 31, 655.
Dennis, C.A.R (1973) In Grain Storage: Part of a System (eds RN. Sinha and
W.E. Muir), Avi, Westport, CT, p. 367.
Dewey, F.M., MacDonald, M.M., Phillips, S.1. and Priestley, RA (1990) Journal of
General Microbiology, 136, 753.
Dolan, T.CS. (1979) The Brewer, 65, 60.
Donhauser, S., Weideneder, A, Winnewisser, W. and Geiger, E. (1989) Brauwelt,
129,1658.
Doran, P.J. and Briggs, D.E. (1993) Journal of the Institute of Brewing, 99, 165.
Douglas, P.E. (1984) A microbiological investigation of industrial malting of barley,
MSc Thesis, Heriot-Watt University.
Douglas, P.E. and Flannigan, B. (1988) Journal of the Institute of Brewing, 94,85-88.
Drysdale, RB. (1984) In The Applied Mycology of Fusarium (eds M.O. Moss and
J.E. Smith), Cambridge University Press, Cambridge, p. 95.
EI-Banna, AA (1987) Mycotoxin Research,3,37.
Etchevers, G.C, Banasik, O.J. and Watson, CA (1977) Brewers' Digest, 52, 46.
Ferguson, P.E., Flannigan, B. and Schapira, S.F.D. (1984) SIM News 34 (4), 45-46.
Flannigan, B. (1969) Transactions of the British Mycological Society, 53, 371.
Flannigan, B. (1970) Transactions of the British Mycological Society, 55, 267.
Flannigan, B. (1974) Transactions of the British Mycological Society, 62, 51.
Flannigan, B. (1977) In Biodeterioration Investigation Techniques (ed. AH. Walters),
Applied Science, London, p. 185.
Flannigan, B. (1978) Transactions of the British Mycological Society, 71, 37.
Flannigan, B. (1983) Journal of the Institute of Brewing, 89, 364.
Flannigan, B. (1989) In Biodeterioration Research II (eds G.C Llewellyn and
CE. O'Rear), Plenum, New York, p. 191.
Flannigan, B. (1991) In Toxic Substances in Crop Plants (eds J.P.F. D'Mello,
CM. Duffus and J.H. Duffus), Royal Society of Chemistry, London, p. 226.
Flannigan, B. and Bana, M.5.0. (1980) In Biodeterioration, Proceedings of the Fourth
International Symposium, Berlin (eds T.A Oxley, D. Allsopp and G. Becker),
Pitman, London, p. 229.
Flannigan, B. and Barnes, S.D. (1990) Proceedings of the Japanese Association of
Mycotoxicology, 31, 25.
Flannigan, B. and Campbell, 1. (1977) Transactions of the British Mycological Society,
69,485.
Flannigan, B. and Healy, RE. (1983) Journal of the Institute of Brewing, 89, 341.
Flannigan, B. and Pearce, A.R (1994) In Biology of Aspergillus (eds K.A Powell,
J. Peberdy and E. Renwick), Plenum, New York, p. 115.
Flannigan, B., Healy, RE. and Apta, R (1986) In Biodeterioration 6 (eds S. Barry,
D.R. Houghton, G.c. Llewellyn and c.E. O'Rear), Commonwealth
Agricultural Bureaux International, Farnham Royal, p. 300.
Flannigan, B., Morton, J.G. and Naylor, RJ. (1985) In Trichothecenes and Other
Mycotoxins (ed. J. Lacey), Wiley, New York, p. 171.
Flannigan, B., Okagbue, RN., Khalid, Rand Teoh, C.K. (1982) Brewing and Distilling
International, 12, 31.
Flannigan, B., Day, S.W., Douglas, P.E. and McFarlane, G.B. (1984) In Toxigenic Fungi
- Their Toxins and Health Hazard (eds K. Kurata and Y. Ueno), Kodanshal
Elsevier, Tokyo, p. 52.
Flannigan, B., Naylor, R.J., Prescott, G.M. and Schapira, S.F.D. (1986) In
Biodeterioration 6 (eds S. Barry, D.R. Houghton, G.c. Llewellyn and
122 The microflora of barley and malt
e.E. O'Rear), Commonwealth Agricultural Bureaux International, Farnham
Royal, p. 238.
Follstad, M.N. and Christensen, e.M. (1962) Applied Microbiology, 10, 33l.
Fukal, L., Prosek, J. and Rakosova, A (1990) Monatsschrift for Brauwissenschaft,
43,212.
Fukushima, S., Kitabatake, K and Amaha, M. (1976) Bulletin of Brewing Science,
22,37.
Gaber, S.D. and Roberts, E.H. (1969) Journal of the Institute of Brewing, 75, 303.
Gilman, J.e. and Barron, D.H. (1930) Plant Physiology, 5,565.
Gilmour,J.S.,Inglis, D.M.,Robb,I. and Madean,M. (1989) Veterinary Record, 124,133.
Gjertsen, P., Trolle, B. and Andersen, K (1963) Proceedings of the 9th Congress of the
European Brewery Convention, Stockholm, IRL Press, Oxford, p. 320.
Gjertsen, P., Trolle, B. and Andersen, K (1965) Proceedings of the 10th Congress of the
European Brewery Convention, Brussels, IRL Press, Oxford, p. 428.
Gjertsen, P., Myken, F., Krogh, P. and Hald, B. (1973) Proceedings of the 14th Congress
of the European Brewery Convention, Salzburg, IRL Press, Oxford, p. 373.
Glinsukon, T., Yuan,S.5., Wightman, R, Kitaura, Y., Buechi, G., Shank, Re., Wogan,
G.N. and Christensen, e.M. (1974) Plant Foods for Man, 1,113
Graff, AR (1972) MBAA Technical Quarterly, 9, 18.
Grant, I.W.B., Blackadder, E.S., Greenberg, M. and Blyth, W. (1976) British Medical
Journal 1, 490.
Gyllang, H. and Martinson, E. (1976a) Journal of the Institute of Brewing, 82, 182.
Gyllang, H. and Martinson, E. (1976b) Journal of the Institute of Brewing, 82, 350.
Gyllang, H., Satmark, L. and Martinson, E. (1977) Proceedings of the 16th Congress of
the European Brewery Convention, Amsterdam, IRL Press, Oxford, p. 245.
Gyllang, H., Kjellen, K, Haikara, A and Sigsgaard, P. (1981) Journal of the Institute
of Brewing, 87, 248.
Haikara, A (1983) Proceedings of the 20th Congress ofthe European Brewery Convention,
London, IRL Press, Oxford, p. 40l.
Haikara, A, Makinen, V. and Hakulinen, R (1977) Proceedings of the 16th Congress
of the European Brewery Convention, Amsterdam, IRL Press, Oxford, p. 35.
Harper, S.H.T. and Lynch, J.M. (1979) Journal of General Microbiology, 112, 45.
Harper, S.H.T. and Lynch, J.M. (1981) Journal of General Microbiology, 122, 55.
Harrison, J.S. and Graham, J.e.J. (1970) In The Yeasts, Vol. 3, Yeast Technology (eds
AH. Rose and I.S. Harrison), Academic Press, London, p. 283.
Harrison, I.G. and Perry, D.A. (1976) Annals of Applied Biology, 84, 57.
Healy, RE. (1985) Effects of the microflora of barley on malt characteristics, PhD
Thesis, Heriot-Watt University.
Heaton, P.E., Callow, M.E. and Butler, G.M. (1992) International Biodeterioration and
Biodegradation, 29, 135.
Hellberg, A and Kolk, H. (1972) Acta Agriculturae Scandinavica, 22, 137.
Hill, RA and Lacey, J. (1983a) Annals of Applied Biology, 102,455.
Hill, RA and Lacey, J. (1983b) Annals of Applied Biology, 103,467.
Howe, RV. (1973) Seed Science and Technology, I, 563.
Ichinoe, M., Hagiwara, H. and Kurata, H. (1984) In Toxigenic Fungi - Their Toxins and
Health Hazard (edsH.Kurataand Y. Ueno),Kodansha/Elsevier, Tokyo, p.190.
Jacobson, RJ. and Thomas, KP. (1981) Plant Pathology, 30, 54.
Jorgensen, J. (1969) Friesia, 9, 97.
Jorgensen, I. (1983) Seed Science and Technology, 11, 615.
Kellerman, T.S., Pienaar, J.G., van der Westhuizen, G.e.A, Anderson, L.AP. and
Naude, T.W. (1976) Onderstepoort Journal of Veterinary Research, 43,147.
Kelly, L. and Briggs, D.E. (1992a) Journal of the Institute of Brewing, 98,329.
Kelly, L. and Briggs, D.E. (1992b) Journal of the Institute of Brewing, 98, 395.
Kelly, L. and Briggs, D.E. (1993) Journal of the Institute of Brewing, 99,57.
 References 123
Kitabatake, K and Amaha, M. (1974) Bulletin of Brewing Science, 20, l.
Kitabatake, K and Amaha, M. (1976) Bulletin of Brewing Science, 22, 37.
Kneen, E. (1963) Proceedings of the Irish Maltsters Technical Conference, 1963, Irish
Maltsters Association, Dublin, p. 5l.
Koshinsky, H.A, Cosby, RH. and Khachatourians, G.G. (1992) Biotechnology and
Applied Biochemistry, 16,275.
Kotheimer, rB. and Christensen, CM. (1961) Wallerstein Laboratory Communications,
24,2l.
Krogh, P., Hald, B., Gjertsen, P. and Myken, F. (1974) Applied Microbiology, 28, 3l.
Lacey, J. (1971) Annals of Applied Biology, 69,187.
Lacey, J. (1975) Transactions of the British Mycological Society, 65, 17l.
Lacey, J., Hill, S.T. and Edwards, M.A (1980) Tropical Stored Products Information,
39,19.
Lafont, J., Romand, A and Lafont, P. (1981) Mycopathologia, 74, 119.
Lee, U.-S., Jang, H.-S., Tanaka, T. et al. (1985) Food Additives and Contaminants, 2, 185.
Lee, U.-S., Jang, H.-S., Tanaka, T. et al. (1985) Food Additives and Contaminants, 3, 253.
Lepschy-von Gleissenthal, J., Dietrich, R, Martlbauer, E. et al.(1989) Zeitschrift flir
Lebensmittel Untersuchung und Forschung, 188,521.
Lovelace, CE.A and Nyathi, CB. (1977) Journal of the Science of Food and Agriculture,
28,288.
Ludwig, RA (1957) Canadian Journal of Botany, 35, 291.
Lund, A (1956) Friesia, 5, 297.
Lund, A, Pedersen, H. and Sigsgaard, P. (1971) Journal of the Science of Food and
Agriculture, 22, 458.
Lutey, RW. and Christensen, CM. (1963) Phytopathology, 53, 713.
Lynch, J.M. and Prynn, S.J. (1977) Journal of General Microbiology, 103,193.
Machacek, J.E. and Wallace, H.AH. (1952) Canadian Journal of Botany, 30, 164.
Machacek, J.E., Cherewick, W.J., Mead, H.w. and Broadfoot, W.C (1951) Scientific
Agriculture, 31, 193.
Malt Research Institute (1955) Publication of the Malt Research Institute, 12,2.
Mannio, M. and Enari, T. (1973) Brauwissenschaft, 26, 134.
Mason, F.A. (1923) Bulletin of the Bureau of Biotechnology, 11/12, 78.
Matthews, S. and Collins, M.T. (1975) Seed Science and Technology, 3, 863.
Mead, H.W. (1943) Wallerstein Laboratory Communications, 6, 26.
Mills, J.T. and Wallace, H.AH. (1979) Canadian Journal of Plant Science, 59,645.
Moll, M. (1981) Journal of the Institute of Brewing, 87,303.
Miiller, H-M. and Schwadorf, K (1993) Mycopathologia, 121, 115.
Niessen, L., Donhauser, S., Weideneder, A., et al. (1992) Brauwelt 132, 702.
Niessen, L., Bohm-Schraml, M., Vogel, H., and Donhauser, S. (1993) Mycotoxin
Research, 9, 99.
Nip, W.K., Chang, CC, Chu, F.5. and Prentice, N. (1975) Applied Microbiology,
30,1048.
Nummi, M., Niku-Paavola, M.-L. and Enari, T.-M. (1975) Brauwissenschaft, 28,130.
Okoye, Z.5.C (1986) Journal of Food Safety, 7,233.
Okoye, Z.5.C (1987) Food Additives and Contaminants, 4, 57.
Okoye, Z.5.C and Ekpenjong (1984) Transactions of the Royal Society of Tropical
Medicine and Hygiene, 78, 417.
Onions, AH.5., Allsopp, D. and Eggins, H.O.w. (1981) Smith's Introduction to
Industrial Mycology, Arnold, London.
Ouellet, T. and Seifert, KA (1993) Phytopathology, 83,1003.
Payen, J., Girard, T., Gaillardin, M. and Lafont, P. (1983) Microbiologie-Aliments-
Nutrition, 1, 143.
Pepper, E.H. and Kiesling, RL. (1963) Proceedings of the Association of Official
Analytical Chemists, 53, 199.
124 The microjlora of barley and malt
Petters, H.L, Flannigan, B. and Austin, B. (1988) Journal ofApplied Bacteriology, 65, 279.
Pidoplichko, M.N., Moskovels, V.S. and Zholanova, N.M. (1960) Mikrobiolohichnyi
Zhurnal,22,15.
Pixton, S.W. and Henderson, S. (1981) Journal of Stored Products Research, 17, 191.
Pixton, S.W. and Warburton, S. (1971) Journal of Stored Products Research, 7,261.
Ponchet, J. (1966) Etude des communautes mycopericarpiques du caryopse de ble,
DSc Thesis, University of Paris.
Prentice, N. and Sloey, W. (1960) Proceedings of the American Society of Brewing
Chemists, p. 28.
Rabie, CJ. and Lubben, A. (1984) South African Journal ofBotany,3,251.
Rabie, CJ., Thiel, P.G. and Marasas, W.F.O. (1983) Third International Mycological
Congress, Tokyo, Abstracts, Mycological Society of Japan, Tokyo, p. 599.
Riddle, H.F.V., Channell, S., Blyth, W., et al. (1968) Thorax, 23, 271.
Samson, RA., Hoekstra, E.S., Frisrad, J.C and Filtenborg, O. (1995) Introduction to
Food-Borne Fungi, 4th edn, Centraalbureau voor Schimmelcultures, Baam,
The Netherlands, 299 pp.
Schapira, S.F.D. (1985) The effect of trichothecene mycotoxins on yeast and malting
barley, MSc Thesis, Heriot-Watt University.
Schapira, S.F.D., Whitehead, M.P. and Flannigan, B. (1989) Journal of the Institute of
Brewing, 95, 415.
Schappert, KT. and Khachatourians, G.G. (1983) Applied and Environmental
Microbiology, 45, 862.
Schappert, KT. and Khachatourians, G.G. (1984) Applied and Environmental
Microbiology, 47,681.
Schoental, R (1984) Proceedings of the Fifth Meeting on Mycotoxins in Animal and
Human Health, Edinburgh, August 1984, (eds M.O. Moss and M. Frank),
University of Surrey, Guildford, pp. 62-67.
Schwabe, M., Rath, F., Golomb, A., et al. (1993) Monatsschrift /iir Brauwissenschaft,
46,408.
Schwabe, M., Fenz, R, Engels, R, et al. (1994) Monatsschrift /iir Brauwissenschaft,
47,160.
Scott, P.M., Kanhere, S.R and Weber, D. (1993) Food Additives and Contaminants,
10,381.
Scott, P.M., Kanhere, S.R, Daley, E.F. and Farber, J.M. (1992) Mycotoxin Research,
8,58.
Sheneman, J.M. and Hollenbeck, CM. (1960) Proceedings of the American Society of
Brewing Chemists, p. 22.
Shlosberg, A., Zadikov, I., Perl, S. et al. (1991) Mycopathologia, 114, 35-39.
Sinha, RN. (1973) In Grain Storage: Part of a System (eds RN. Sinha and W.E. Muir),
Avi, Westport, Connecticut, p. 15.
Sinha, RN. and Wallace, H.A.H. (1965) Canadian Journal of Plant Science, 45,48.
Skoulas, A., Williams, N. and Merriman, J.E. (1964) Journal of Occupational Medicine,
6,359.
Sloey, W. and Prentice, N. (1962) Proceedings of the American Society of Brewing
Chemists, p. 24.
Smith, J.E., Mitchell, I. and Chiu, M.L.C (1984) In The Applied Mycology of Fusarium
(eds M.O. Moss and J.E. Smith), Cambridge University Press, Cambridge,
p.157.
Stankushev, K (1969) Mikologia i Fitopatologia, 3, 268.
Stones, E.L.G. (1984) In Miscellany II (ed. D. Sellar), The Stair Society, Edinburgh,
p.148.
Tressl, R, Hommel, E., Helak, B. (1989) Monatsschrift /iir Brauwissenschaft, 42,331.
Trinder, D.W. (1988) Journal of the Institute of Brewing, 95, 307.
Tuite, J.F. and Christensen, CM. (1955) Cereal Chemistry, 32, 1.
 References 125
Tuite, J.F. and Christensen, C.M. (1957) Phytopathology, 47,265.
Uoti, J. and Ylimaki, A (1974) Annales Agriculturae Fenniae, 13,5.
Vaag, P. (1991) Proceedings of the 23rd Congress of the European Brewery Convention,
Lisbon, IRL Press, Oxford, p. 553.
Vaag, P. and Pedersen, S. (1993) Proceedings of the 24th Congress of the European
Brewery Convention, Oslo, IRL Press, Oxford, p. 111.
Valadon, L.R.G. and Lodge, E. (1970) Transactions of the British Mycological Society,
55,9.
Wallace, H.A.H. (1973) In Grain Storage: Part of a System (eds R.N. Sinha and
W.E. Muir), Avi, Westport, Connecticut, p. 71.
Wallace, H.AH. and Sinha, R.N. (1962) Canadian Journal of Plant Science, 42,130.
Warnock, D.W. (1971) Journal of General Microbiology, 67,197.
Warnock, D.W. (1973a) Transactions of the British Mycological Society, 61,49.
Warnock, D.W. (1973b) Transactions of the British Mycological Society, 61, 547.
Warnock, D.W. and Preece, T.F. (1971) Transactions of the British Mycological Society,
56,267.
Welling, B. (1969) Tidsskrift for Planteavl, 73,291.
Whitehead, M.P. and Flannigan, B. (1989) Journal of the Institute of Brewing, 95, 411.
Williams, N., Skoulas, A and Merriman, I.E. (1964) Journal of Occupational Medicine,
6,319.
Woller, R. and Majerus, P. (1992) Monatsschrift Jilr Brauwissenschaft, 35,88.
Wraith, D.G., Cunnington, AM. and Seymour, W.M. (1979) Clinical Allergy, 9, 545.
Ylimaki, A, Koponen, H., Hintikka, E., et al. (1979) Technical Research Centre of
Finland, Materials and Processing Technology, Publication 21.
Yoshida, J., Nakagawa, A, Eto, M., Kitabatake, K. and Amaha, M. (1975) Journal of
Fermentation Technology, 53,184.
Zadoks, J.c., Chang, T.T. and Konzak, C.F. (1974) Weed Research, 14, 415.
 CHAPTERS

Gram-positive brewery
bacteria
F.e. Priest

5.1 INTRODUCTION

Bacteria have been recognized as important spoilage agents of beer since

the end of the nineteenth century. Comparison of the descriptions of
bacteria prevalent in breweries and beers at that time with those of today
reveals that the range of bacteria encountered is remarkably constant
although the names have been changed frequently over the years. Recently,
as new processes have been introduced, in particular the packaging of beers
with very low oxygen contents, and microbiological techniques have been
improved, so 'new' contaminants have been isolated from spoiled beers.
These include the strictly anaerobic strains of Pectinatus cerevisiiphilus and
Megasphaera (see Chapter 6) but in general brewers are facing today the
same bacterial contaminants as they did last century.
Bacteria are traditionally divided into two categories according to their
reaction to a differential staining procedure, the Gram stain. Cells that are
bounded by a relatively thick, single layer of peptidoglycan, a net-like
polymeric molecule that provides structural strength to the cell, generally
retain the crystal violet/iodine complex of the Gram stain when washed
with ethanol or acetone and are referred to as Gram positive when sub-
jected to the Gram stain. These cells appear purple under the microscope.
Gram-negative bacteria, on the other hand, possess a multilayered envel-
ope comprising the cytoplasmic membrane, the periplasmic space, a thin
layer of peptidoglycan and a hydrophobic outer membrane. Such cells do
not retain the crystal violet/iodine complex when washed with ethanol or
acetone and appear pink under the microscope when stained by Gram's
procedure. Shimwell (1945) was first to report the particular relevance of
the Gram reaction to brewery microbiology, when he noted that the growth

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and I. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412 59150 2
128 Gram-positive brewery bacteria
of most Gram-positive bacteria was inhibited by hop bittering substances,
but Gram-negative bacteria were unaffected. This antiseptic action of hops,
together with the poor nutrient status of beer, its low pH and the presence
of ethanol, restricts the range of bacteria that can grow in finished beers.
With regard to the Gram-positive organisms, lactic acid bacteria of the
genera Lactobacillus and Pediococcus are the most common and dangerous
spoilage agents and strains of some species have developed marked hop
tolerance (Fernandez and Simpson, 1993). Some members of the genera
Micrococcus and Staphylococcus can survive in beer and under certain con-
ditions Micrococcus kristinae will grow and cause spoilage. On rare occa-
sions aerobic, spore-forming bacteria belonging to the genus Bacillus have
caused problems, particularly in unhopped worts (Healy and Armitt, 1980).
However, like the micrococci, these bacteria are inhibited by hop com-
ponents (Smith and Smith, 1993), do not appear to develop hop tolerance
and are therefore not generally a serious threat. This chapter will review
the physiology, taxonomy, isolation and enumeration of these various
bacteria and will cover their effect on beer quality.

5.2 LACTIC ACID BACTERIA

The lactic acid bacteria are closely related to the endospore-forming bacteria
of the genera Bacillus and Clostridium and, with numerous other genera,
form a major division of the bacteria known as the low G + C group because
they all possess a relatively low « 55%) content of guanine plus cytosine in
their chromosomal DNA. The lactic acid bacteria share considerable genetic
and molecular homology and have long been recognized as a natural group
(Ingram, 1975). However, a definitive description of the lactic acid bacteria
cannot be agreed upon and with current developments in the taxonomy of
these bacteria is probably more distant now than at any previous time.
However, the typical lactic acid bacterium is a Gram- positive, non-
sporulating rod or coccus. It lacks the enzyme catalase and is strictly ferment-
ative producing either a mixture of lactic acid, CO2, acetic acid and/or
ethanol (heterofermentation) or almost entirely lactic acid (homofermenta-
tion) as the major metabolic end-product from sugar catabolism. Conse-
quently, it will be acid tolerant and have complex nutritional requirements.
Lactic acid bacteria are generally associated with habitats rich in nutrients
such as milk and dairy products, vegetation and some are also members of
the normal flora of the mouth, intestine and vagina of mammals.
The pioneering work of Orla-Jensen (1919), who divided the group into
four genera, Lactobacillus for the rod-shaped organisms, Streptococcus for
the homofermentative, facultatively anaerobic cocci, 'Betacoccus' and
'Tetracoccus', remains influential. The classification and identification of
the lactic acid bacteria are being revolutionized however, by the applica-
tion of modem molecular approaches to systematics, in particular the use
 Lactic acid bacteria 129
of nucleic acid comparisons as indicators of evolutionary or phylogenetic
relationships (Woese, 1987). These molecular taxonomies are based on
sequence comparisons of the small subunit ribosomal RNA genes and are
generating more detailed and reliable indications of the genera and species
of all bacteria (see Priest and Austin, 1993, for details). Of the original
OrIa-Jensen taxa, Streptococcus has been divided into several genera:
Enterococcus for 'Streptococcus faecalis' and relatives, Lactococcus for the
dairy streptococci, 'S. lactis' and relatives and Vagococcus for some motile
organisms otherwise resembling lactococci, as well as Streptococcus sensu
stricto.
'Betacoccus' has been renamed Leuconostoc and includes hetero-
fermentative cocci that occur in pairs or short chains. This genus has not
escaped the attentions of the 'new systematicists' and rRNA gene sequence
studies have shown that three phylogenetically diverse generic-ranked
taxa have been contained in Leuconostoc sensu lata:
1. Leuconostoc sensu stricto;
2. Leuc. oenos;
3. the new genus Weissella which includes Leuc. paramesenteroides and
some closely related heterofermentative lactobacilli (such as L. confusus
and L. kandleri) now reclassified as W. confusa and W. kandleri (Martinez-
Murcia and Collins, 1990; Collins et al., 1993).
'Tetracoccus' has become Pediococcus and contains those homo-
fermentative cocci that divide in two planes to produce pairs and tetrads.
One member of this genus, P. halophil us, has been elevated to generic status
by rRNA sequence studies as Tetragenococcus (Collins et al., 1990).
Aerococcus is a monospecific genus which is related to Pediococcus and
Tetragenococcus, and yet another genus, Globicatella, has been introduced
recently for cocci from human clinical sources which were related to, but
phylogenetically distinct from, the streptococci (Collins et al., 1992). Fortu-
nately, few of these lactic acid bacteria are commonly encountered in
breweries, but the features of the principal genera are shown in Table 5.1
for reference purposes.
OrIa-Jensen in his far-reaching studies, also divided Lactobacillus into
three subgenera, mainly on the basis of temperature range for growth and
mode of fermentation. Heterofermentative lactobacilli were placed in
'Betabacterium'. Homofermentative strains that grew at 45°C but not at 15°C
were placed in 'Thermobacterium' and those with the opposite temperature
relationships constituted 'Streptobacterium'. Although these subgeneric al-
locations are still used, the names were not included in the Approved Lists
of Bacterial Names (Skerman, McGowan and Sneath, 1980) and therefore
have no standing in nomenclature.
Evidence for genetic heterogeneity in the genus can be gained from gross
analysis of chromosomal DNA in the form of mol% guanine plus cytosine.
Bacteria with widely different G + C ratios (> 12%) should be genetically
Table 5.1 Principal genera of lactic acid bacteria and their characteristics (data from Axelsson (1993) and Collins et al. (1992,1993»
Characteristics
CO2 Growth Growth Growth in Lactic
from at at 6.5% acid
Genus Shape Tetrads glucose 1Q°C 45°C NaCl formed
Carnobacterium Rods + • L
Lactobacillus Rods d d d d D, L,
DLt
Aerococcus Cocci + + + L
Enterococcus Cocci + + + L
Globicatella Cocci + •
Lactococcus Cocci + L
Leuconostoc Cocci + + d D
Pediococcus Cocci + d d d L, DLt
Streptococcus Cocci d L
Tetragenococcus Cocci + + L
Vagococcus Cocci + L
Weissella Small rods/ + d d D,DL
Cocci
+, Positive; - negative; d, response varies; ., not determined.
* Configuration of lactic acid produced from sugar
t Production of 0-, L-, or DL-lactic acid varies between species.
 Lactic acid bacteria 131
distinct and are unlikely to be members of the same genus (De Ley, 1969).
That strains of Streptococcus (34-46% G + C), Leuconostoc (37-46% G + C)
and Pediococcus (34--44% G + C) show narrow and overlapping ranges of
G + C content indicates that these genera are not genetically heterogeneous
and could be closely related (Bradley, 1980). Lactobacilli, on the other hand,
display 32-53% G + C, suggesting that genetic diversity exists and some
revision of the genus is required since this range exceeds the 10-12%
maximum variation expected within a genus (Bradley, 1980; Priest, 1981a).
This process has already been initiated with the establishment of the genus
Carnobacterium (Collins et al., 1987) for some species such as L. divergens and
L. carnis which are commonly associated with meats and are able to grow
at low temperature. A simple key for the identification of the camobacteria
has been published (Montel et al., 1991).

5.2.1 Carbohydrate metabolism

Lactic acid bacteria seldom hydrolyse polysaccharides although
L. amylophilus, which hydrolyses starch, is an exception. They do however,
catabolize a range of disaccharides often including maltose which, at least
in the case of lactococci, enters through a permease system and is then
cleaved by a maltose phosphorylase into glucose and glucose-1-phosphate
(Konings, Poolmanand Driessen, 1989). Indeed, most sugars are assimilated
by specific permeases driven by the proton motive force and phosphory-
lated in the cytoplasm by an ATP-dependent kinase. Some species use the
phosphoenolpyruvate (PEP)-dependent phosphotransferase system (PTS)
which phosphorylates sugars using PEP as they are transported into the cell.
The two major pathways of hexose fermentation within the lactic acid
bacteria differ in the way in which the C6 skeleton is split and this yields
different sets of end-products (Fig. 5.1). Homofermentation, characteristic
of the streptococci, pediococci and some lactobacilli, involves the splitting
of fructose-1,6-bisphosphate by the enzyme aldolase into two
triosephosphates which are converted to lactate via pyruvate.
Heterofermentation is practised by leuconostocs and some lactobacilli. In
this process, glucose-6-phosphate is oxidized to form gluconate-6-
phosphate and decarboxylated. The resultant xylulose-5-phosphate is split
by phosphoketolase to yield a triosephosphate that is converted to lactate,
and acetyl phosphate that is a precursor of either ethanol or acetate depend-
ing on the oxidation-reduction potential of the system. Thus equimolar
amounts of CO2 lactate and acetate/ethanol are formed from glucose.
Should an additional hydrogen acceptor in the form of fructose or oxygen
be available, it would be reduced to mannitol or H 20 2 respectively and
ethanol would not be formed.
There can be some confusion about the use of the terms homo-
fermentative and heterofermentative. It is prudent to reserve the term
homofermentativefor lactic acid bacteria which use the glycolytic pathway
132 Gram-positive brewery bacteria

_ _ _ _ _ _ glucose _______

fructose-!,6P glucose-6P
~
6P-gluconate

aldolase
[ ' - - C0 2
xylulose-SP

triose-3P
---j triose-3P + acetyl-P

l(~~~)
j
CADP
ATP

pyruvate pyruvate

1
lactate
1
lactate acetate (ethanol)

Glycolysis (homofermentation) 6P-Gluconate (phosphoketolase)

pathway (heterofermentation)

Fig. 5.1 Main pathways of hexose fennentation in lactic acid bacteria (modified
from Kandler, 1983).

for glucose catabolism and therefore produce predominately lactate as

end-products, and heterofermentation for lactic acid bacteria which use
the gluconate-6-phosphate/phosphoketolase pathway. However, it
should be emphasized that some lactobacilli produce various end-products
from glycolysis of glucose, and others which use glycolysis for glucose
catabolism switch to the gluconate-6-phosphate/phosphoketolase
pathway for certain substrates, notably pentose sugars.
Heterofermentative lactic acid bacteria generally ferment pentoses, al-
though exceptions do occur (Kandler, 1983). The various pentoses are taken
up by specific permeases and induce synthesis of the relevant enzymes for
conversion of the substrate into the key intermediate D-xylulose-5-
phosphate, which enters the phosphoketolase pathway. Some lactic acid
bacteria also ferment pentitols via o-xylulose-5-phosphate, although these
sugar alcohols are taken up via the PEP-dependent PTS (London and
Chace, 1977, 1979). Since a functional phosphoketolase pathway is required
for pentose metabolism by heterofermentative lactic acid bacteria it might
 Lactic acid bacteria 133
be expected that homofermentative strains would be unable to utilize these
sugars. Indeed no authentic strains of the 'Thermobacterium' group have
been found that utilize pentoses (Kandler, 1983) but the streptobacteria and
pediococci are often able to use such sugars. These bacteria possess a
phosphoketolase that is induced by pentose sugars and the organisms thus
function heterofermentatively on such substrates, although with glucose
alone they would be homofermentative. For this reason these bacteria are
referred to as facultatively heterofermentative.
Lactic acid bacteria produce either D(-)- or L(+)-lactate and sometimes
both isomers. Lactate racemase was originally thought to be responsible for
the production of racemic mixtures from L( +)-lactate but it has become
evident that this enzyme is not widely distributed. Instead two NAD-
dependent lactate dehydrogenases of different stereospecificity are gener-
ally responsible for mixtures of L- and D-lactate. There are alternative minor
pathways for pyruvate dissimilation in lactic acid bacteria which are of
particular importance to the brewer since they result in the formation of
volatile flavour compounds such as diacetyl. These will be covered in
section 5.3.3(b).

5.2.2 Nitrogen metabolism

Lactic acid bacteria are typically fastidious and require a variety of amino
acids and vitamins for growth. It seems that in many environments they
obtain these amino acids through the action of proteases although, com-
pared with bacilli or pseudomonads, lactic acid bacteria are weakly pro-
teolytic. Proteolysis has been particularly well documented in relation to
the growth of lactococci in milk, where they are largely responsible for
flavour development during cheese production. In these circumstances a
complex system is involved comprising:
1. cell wall-associated caseinolytic protease;
2. extracellular peptidases;
3. amino acid transport systems;
4. peptide transport systems;
5. intracellular peptidases (Smid, Poolman and Konings, 1991).
Although the proteolytic activity of lactococci is low, the overall system
is efficient at providing all the amino acids for growth in milk and protease-
deficient mutants soon become growth-limited. Although little is known
of the proteolytic systems of brewery lactic acid bacteria, El Soda and
Desmazeaud (1982) and El Soda et al. (1982) have demonstrated proteolytic
activity in both heterofermentative and homo fermentative lactobacilli.
Surface-bound peptidases have also been demonstrated in lactobacilli. This
is particularly relevant to the brewer since beer is deficient in various amino
acids and it is often assumed that this is important in restricting the growth
of spoilage lactobacilli.
134 Gram-positive brewery bacteria
Some amino acids such as glutamate and histidine are degraded by
lactobacilli rather than incorporated into proteins; arginine in particular
can be used as an energy source by many heterofermentative strains
including L. brevis (Rainbow, 1981; Manca de Nadre, Pesce de Ruiz Hogado
and Oliver, 1988). Arginine is hydrolysed to citrulline and ammonia and
the citrulline is converted to ornithine with the production of ATP.

5.2.3 Oxygen requirements

An interesting feature of the lactic acid bacteria is their relationship to
oxygen. These bacteria are described as aerotolerant anaerobes in that they
cannot use oxygen as a terminal electron acceptor in respiration but they
contain NADH oxidases which are often induced by oxygen (Condon,
1987). The products of the reaction are either NAD+ and H20z or NAD+ and
H 20 depending on whether two or four electrons are transferred. Unlike
strict anaerobes, lactic acid bacteria are not killed by oxygen. It is thought
that toxic oxygen species such as peroxide (0z2-) and particularly superox-
ide (Oz-) are produced by cells in the presence of air. Aerobic bacteria have
developed the enzymes catalase and superoxide dismutase respectively to
remove these radicals, whereas strict anaerobes have not. Lactic acid bac-
teria are unable to synthesize haem and therefore cannot produce catalase
unless haematin is provided. Nevertheless, they synthesize a
pseudocatalase that requires a reducing substance, usually NADH, to
remove peroxides. Pseudocatalase was first reported in pediococci
(Delwiche, 1961) and soon after in lactobacilli, leuconostocs and strepto-
cocci Oohnston and Delwiche, 1965). The enzyme from L. plantarum has
been purified and is a Mn2+-containing molecule. Lactobacilli, leuconostocs
and pediococci do not synthesize superoxide dismutase, however
(Archibald and Fridovich, 1981). Instead, these bacteria contain high levels
of Mn2+, which is believed to be responsible for scavenging intracellular
toxic oxygen species, in particular the superoxide radical. Indeed there is a
good correlation between the intracellular Mn2+ concentration and
tolerance to oxygen amongst these bacteria. Nevertheless, it is advisable to
grow these bacteria under anaerobic or partially anaerobic conditions,
particularly upon initial isolation.

5.3 LACTOBACILLUS

5.3.1 Classification
Following the initial work of Grla-Jensen (1919) the taxonomy of the
lactobacilli progressed slowly despite the economic and environmental
importance of the group (London, 1976). The three subgenera became
firmly entrenched and formed the basis of a most useful article on the
 Lactobacillus 135
identification of these bacteria published by Sharpe, Fryer and Smith (1966).
It then became apparent that the heterofermentative 'betabacteria' could be
divided into two groups, one comprising metabolically inert strains that
were highly ethanol tolerant and common in wines, and a group of more
widely distributed, physiologically versatile species. This classification was
incorporated in Sharpe's revision of her identification scheme (Sharpe,
1979) but was slightly modified in Bergey's Manual of Systematic Bacteriology
(Kandler and Weiss, 1986). The subgeneric names were no longer used and
the 'thermobacteria' were labelled Group I (obligate homofermentative
strains), the 'streptobacteria' (facultative heterofermentative strains)
formed Group II and the 'betabacteria', both ethanol tolerant and other
types, constituted the obligately heterofermentative Group III. Hammes,
Weiss and Holzapfel (1993) similarly consider these three groups for iden-
tification purposes in their comprehensive review of these bacteria in the
second edition of The Prokaryotes.
This grouping is helpful for identification purposes but has been chal-
lenged as an appropriate approach to the classification of these bacteria.
Perhaps it is useful to stress at this point that classification and identifica-
tion are different disciplines with different aims. Classification is the
arrangement of organisms into groups on the basis of their relationships
which may be either evolutionary (phylogenetic) and/or phenetic (based
on similarities between the properties of the organisms and with no evolu-
tionary considerations). For various reasons, classifications based on these
two different approaches do not always concur (see Priest and Austin,
1993). Identification, on the other hand, is the assignment of an unknown
organism to its correct place in a pre-existing classification. Identification
makes no assumptions about the classification processes.
With the current interest in phylogenetic classifications based on 16S
rRNA gene sequences, the traditional (phenetic) classification of the lacto-
bacilli has been challenged. Initially, certain 'atypical' lactobacilli isolated
from refrigerated vacuum-packed meats including L. divergens, L. piscicola,
L. maltaromicus and L. carnis were placed in a new genus Carnobacterium.
These bacteria are not encountered in breweries but a simple scheme to aid
identification of these bacteria has been published (Montel et al., 1991). The
remaining lactobacilli have been assigned to three groups which are similar
to, but not identical with the physiological groups (Table 5.1). Group 1
(designated the L. delbrueckii group) contains L. delbrueckii, the type species
of the genus, and 11 other obligately homofermentative species most of
which are uncommon in breweries. Group 2 is based on the the facultatively
heterofermentative L. casei and comprises 32 Lactobacillus species and five
Pediococcus species including P. damnosus (see below). Some obligately
homofermentative species and most of the obligately heterofermentative
species including those found in breweries, such as L. brevis, L. buchneri and
L.fermentum, were recovered in RNA Group 2. Finally, Group 3 comprises
some atypical heterofermentative lactobacilli, including L. confusus and
Table 5.2 Some Lactobacillus species found in fermented beverages and related habitats and
their allocation to taxonomic groups
Fermentation Subgeneric Common
Species RNA group group* allocation habitats
L. delbrueckii 1 I Thermobacterium Grain, mashes
L. yamanashiensis+ 2 II Streptobacterium Wine, cider
L. casei II Streptobacterium Milk and milk
products
L. coryneformis 2 II Streptobacterium Fermented
subsp. coryneformis vegetables
L. coryneformis 2 II Streptobacterium Beer
subsp. torquens
L. homohiochii 2 II Streptobacterium Wine
L. paracasei subsp. 2 II Streptobacterium Beer
paracasel+
L. sake 2 II Streptobacterium Wine
L. brevis 2 III Betabacterium Beer, wine
L. buchneri 2 III Betabacterium Beer, wine
L. fructivorans§ 2 III Betabacterium Wines, vinegar
preserves
L. fermentum 2 III Betabacterium Fermented
vegetables
L. hilgardii 2 III Betabacterium Wine
L. lindneri ? III Betabacterium Beer
Weissella confusa'll 3 III Betabacterium Fermented
vegetables
* Group I, obligately homofermentative; Group II, facultatively heterofermentative (homofermentative with
hexose sugars); Group III, obligately heterofermentative.
t Includes L. mali (Carr et al., 1977).
+ Previously L. casei subsp. pseudoplantarum.
§ Includes L. trichodes and L. heterohiochii.
'Il Previously L. confusus.
 Lactobacillus 137
Leuconostoc paramesenteroides; this group was designated the
'paramesenteroides' group (Collins et al., 1991) initially and more recently
has been raised to generic status as Weissella (Collins et al., 1993). Although
this classification describes the molecular evolutionary pathways which
gave rise to the lactobacilli and relatives, it is unclear at present how it will
fit into the established physiological classification.
Numerical or phenetic taxonomy involves the estimation of similarity
between strains based on a large number of attributes, usually data from
phenotypic tests. Strains are then allocated to clusters (generally species) of
high mutual relatedness derived from these estimates (Sneath and Sokal,
1973). If each of the fermentative groups of lactobacilli contains pheno-
typically similar bacteria this should be revealed by numerical analysis.
Unfortunately, the few studies published to date provide conflicting con-
clusions. Barre (1969) examined 65 strains of lactobacilli from wine together
with some reference strains. The bacteria were recovered in two large
clusters representing RNA Group 2 and based on L. casei / L. plantarum and
L. buchneri/ L. fermentum, but most of the wine isolates were identified as
L. buchneri. Support for the three physiological groups was provided by the
numerical analysis of 30 lactobacilli by Seyfried (1968). However, other
studies (Wilkinson and Jones, 1977; Barbour and Priest, 1983; Shaw and
Harding, 1984) have not resulted in the precise allocation of species to
fermentation groups. Indeed numerical taxonomic analysis of Lactobacillus
has been neither popular nor particularly successful, for reasons that have
been discussed at length elsewhere (Priest and Barbour, 1985).
One of the principal problems in Lactobacillus taxonomy that has been
highlighted by the numerical approach is the high phenetic similarity
amongst the species. Most numerical taxonomic studies generally show
80-85% similarity between strains from different bacterial species.
Amongst the lactobacilli, however, strains from separate species sometimes
possess as much as 95% similarity despite being unrelated by criteria such
as rRNA sequence or DNA homology. It would seem that evolutionary
convergence has resulted in species that are genetically distinct but have
similar phenotypes. Thus species are distinguished by few characters and
such monothetic classifications lead to problems in identification since a
strain need only be aberrant in one or two key features to be incorrectly
identified. Since such schemes involving phenotypic attributes have
proved unreliable for the identification of lactobacilli, other approaches
such as serology, protein profiling and DNA probes have been explored
and will be discussed in connection with brewery lactobacilli in section 5.3.5
and in Chapter 9.
In conclusion, the classification of Lactobacillus is emerging in a stable
fashion. Classification at the genus level and above will be based on rRNA
sequences supported where appropriate by chemotaxonomic considera-
tions such as cell wall structure and physiological traits. Species will be
based largely on DNA sequence homology and related techniques such as
138 Gram-positive brewery bacteria
whole cell protein profiling which provide sound criteria for species dis-
tinction. However, this leaves a difficult situation for the microbiologist
wishing to identify an unknown strain and for the present the traditional
tables presented in Bergey's Manual of Systematic Bacteriology (Kandler and
Weiss, 1986) and the second edition of The Prokaryotes (Hammes, Weiss and
Holzapfel, 1992) remain the best option for the non-specialist laboratory.
Fortunately, for the brewery microbiologist, the range of lactobacilli
encountered is fairly restricted thus simplifying identification (see
section 5.3.5).

5.3.2 Distribution of lactobacilli in beer and breweries

The frustrations associated with identifying Lactobacillus species have hin-
dered ecological studies aimed at determining the range of species found
in the brewing environment. Most such studies have focused on spoiled
beer, in which one of the most common contaminants appears to be L. brevis.
This obligate heterofermentative bacterium is generally tolerant towards
hops and grows optimally at 30°C and pH 4-5. Various other species have
been described in the past, e.g. 'L. diastaticus' (amylolytic, responsible for
superattenuation in beers) and 'L. pastorianus' (produces extracellular poly-
saccharide or rope). These atypical properties are not considered sufficient
to merit species status and both names are considered synonymous with
L. brevis. It should be remembered therefore that L. brevis is physiologically
versatile and can cause various problems in beer. L. lindneri was originally
used to describe strains from lager that grow optimally at 19°C; it was later
deemed to be synonymous with L. brevis (Rogosa, 1974). However, a more
recent study indicates that strains of 'L. lindneri' are sufficiently different
from L. brevis in the spectrum of sugars fermented and DNA base compo-
sition (L. brevis, 44-47%; L.lindneri 42-46%) to warrant species status (Back,
1981, 1982). 'L. frigidus' and 'L. parvus' are considered to be synonyms of
L. buchneri, a close relative of L. brevis that differs only in its ability to
ferment melezitose and some strains have a requirement for riboflavin
(Sharpe, 1979; Back, 1981).
Recently, a new lactic acid bacterium was isolated from spoiled beer
which differs from L. brevis and L. lindneri in the range of sugars fermented
and in DNA base composition to a sufficient extent to warrent species
status, and which has been given the name L. brevisimilis (Back, 1987). This
bacterium is difficult to culture and does not cause serious problems in
bottled beer since it dies rapidly in this environment. Other lactobacilli that
appear to occur less frequently in beer and breweries include L. casei
(Eschenbecher, 1968; Dolezil and Kirsop, 1975), L. delbrueckii (McMurrough
and Palmer, 1979), L. fructivorans, L. fermentum (Lawrence and Leedham,
1979), L. coryneformis, L. coryneformis subsp. torquens, L. curvatus
(Eschenbecher, 1969; Back, 1981; Makinen, Tanner and Haikara, 1981) and
L. plantarum (Wackerbauer and Emeis, 1968; Back, 1981).
 Lactobacillus 139
One of the most comprehensive studies of brewery lactobacilli was that
of Eschenbecher (1968, 1969), who examined 660 Lactobacillus cultures
capable of causing beer spoilage from 81 breweries. These strains were
identified to five species using sugar fermentations and other conventional
tests. Of the 660 strains, 185 were identified as L. casei, 97 as 'L. casei subsp.
fusiformis', six as 'L. plantarum var. arabinosus', 56 as L. coryneformis, four as
L. coryneformis subsp. torquens, 250 as L. brevis, 14 as 'L. brevis var.lindneri'
and 48 as L. buchneri. Eschenbecher noted that Group II ('streptobacteria')
occurred more frequently than heterofermentative (Group III) bacteria in
heavily hopped beers: this is inconsistent with the general notion that only
L. brevis is hop tolerant and suggests that more work is needed in this area.
Back (1981, 1982) has provided excellent reviews of the taxonomy of
brewery lactobacilli and assigned brewery isolates to species broadly sim-
ilar to those of Eschenbecher with comparable frequencies. In addition to
those described by Eschenbecher, Back included L. curvatus and two un-
named groups designated species group 8 and species group 9 (Back, 1981).
The variety of lactobacilli in beer and breweries is therefore considerably
more diverse than is usually believed and it becomes important to know if
these species vary in their potential for beer spoilage.

5.3.3 Spoilage of beer by lactobacilli

(a) Growth of lactobacilli in beer

Since strains of at least nine species of Lactobacillus can be detected in beer
and breweries, the traditional view that all lactobacilli are spoilage organ-
isms may be erroneous if some of these organisms are unable to grow in
beer. Unfortunately information in this area is scant. Back (1982) consid-
ered lactobacilli almost invariably responsible for beer spoilage and sug-
gested that trace evidence of their presence in beer should be rectified
immediately. Nevertheless different species vary in their ability to grow in
beer and in their tolerance to hop bittering compounds. Using five different
beers, Back (1981) showed that strains of L. curvatus, L. brevis, L.lindneri and
species groups 8 and 9 could grow in virtually every beer examined
whereas strains of L. casei grew in only three beers and L. plantarum,
L. buchneri and L. coryneformis grew in only one. Obviously, the presence in
beer of an organism that is unable to grow does not pose a serious threat to
the brewer and, if these findings could be extended to other beers, then it
may be that some lactobacilli could be ignored as spoilage organisms.
The growth of lactobacilli in beer appears to depend as much on the type
of beer as it does on the offending organism. This has been noted above,
and in an examination of 13 lactobacilli incubated in 31 different beers,
Dolezil and Kirsop (1980) found that only in three beers were all the
organisms able to grow. In five beers none developed, and diverse results
were obtained with the remaining 23 beers. The reasons for the resistance
140 Gram-positive brewery bacteria

2 x pyruvate

1 CO2 -r-
"
CH 3-C-COOH
TPP
pyruvate

1
F
hydroxyethyl-TPP CHl-CHO-TPP acetaldehyde TPP
r:pyruvate
TPP (3) P
CO
''"''''
2
0 acetyl
,coenzyme
a-acetolactate
"
CH1-C-COH-CH, a-acetolactate (3)
I
I
I
.~ ~OOH .
I(1) CO 2 TPP HS-CoA (2)
I
I
I
I
0 0
t
diacetyl
" "
CH 3-C-C-CH 3 diacetyl

j
acetoin
FNADH'
NAD

CH, -CHOH -C-CH3

FNADH2
0

"
I
acetoin

1
2.3-butanediol
NAD

CH1-CHOH-CHOH-CH,
1
2.3-butanediol

(1) Oxidative decarboxylation in fermenting wort.

(2) Enzymic decarboxylation of a-acetolactate.
(3) Alternative route for diacetyl synthesis.

Fig. 5.2 Pathways of diacetyl and acetoin formation in yeast fermentations and
Lactobacillus. TPP, thiamine pyrophosphate.

of some beers to infection are unclear. There was no correlation between

any of the usual analytical parameters such as pH, specific gravity, nitrogen
and amino nitrogen content, the quantity of fermentable carbohydrate or
of sulphur dioxide and the ability of lactobacilli to proliferate. Nor was
there any particular feature of the bacteria responsible for growth in a wide
range of beers, although metabolic diversity, as judged by the spectrum of
sugars fermented, appeared to correlate with the spoilage potential. This
 Lactobacillus 141
contradicts Back's (1981) findings, however, since Lactobacillus species
groups 8 and 9 ferment only a few sugars and yet are potent beer spoilage
organisms.
Bacterial growth in beer depends on the ability to scavenge residual
nutrients left following yeast growth, and tolerance to the relatively hostile
environment posed by a fermented beverage. Those nutrients that support
the growth of lactic acid bacteria in beer have been examined by inoculating
beers with lactic acid bacteria and following the changes in sugar and
amino acid content with growth (Pfenninger et al., 1979). In most cases small
amounts of sucrose were metabolized along with varying amounts of
maltose, maltotriose and maltotetraose. In some cases dextrins up to 14 or
15 glucose units were hydrolysed. Amongst the amino acids absorbed by
lactobacilli, lysine, tyrosine and arginine predominated.
The major growth-inhibiting substances in beer are derived from hops
and include trans-humulone and the related (-)-humulone and colupulone
(Verzele, 1986). These compounds act as ionophores and dissipate the
trans-membrane pH gradient of sensitive bacteria (Fernandez and
Simpson, 1993). Lactobacilli vary more than lO-fold in their sensitivity to
trans-humulone with beer spoilage organisms invariably showing high
tolerance to this compound. Resistance is not associated with plasmids but
seems to be chromosomally encoded (Fernandez and Simpson, 1993).
Resistance of specific beers to spoilage by lactobacilli has been tenta-
tively attributed to the presence of a heat-labile yeast metabolite (Dolezil
and Kirsop, 1980), but in a small screening exercise, Young and Talbot
(1979) were unable to detect any extracellular substances from four yeasts
that inhibited bacterial growth. An interesting approach was that of
Munekata (1973), who examined various dipeptides for their inhibitory
effect on the growth of lactic acid bacteria. Some alanyl dipeptides were
found to be potent growth inhibitors. Resistance to infection has also been
attributed to the cosedimentation of certain bacteria with brewing yeasts
(White and Kidney, 1979). It is likely that a combination of these factors will
be responsible for the susceptibility of beers to spoilage by various bacteria.

(b) Effect of spoilage on beer quality

Lactobacilli are most dangerous during conditioning of beer and after
packaging. Spoilage is characterized by a 'silky' turbidity but sometimes
before this is apparent the 'buttery' flavour of diacetyl may be noticed.
Although lactic acid is the major metabolic end-product of these bacteria,
its high flavour threshold in beer, which contains up to 300 ppm lactate
(Hough et al., 1982), compared with the low threshold for diacetyl, about
0.15 ppm (Hough et al., 1982), makes the latter a potent flavour-modifying
compound.
Diacetyl (and 2,3-pentanedione) are produced during normal fermenta-
tions. During growth, yeast synthesizes a-acetolactate as an intermediate
142 Gram-positive brewery bacteria
in valine biosynthesis. Some a-acetolactate is excreted by the yeast and is
non-enzymically converted to diacetyl in the fermenting wort. The diacetyl
is then enzymically reduced by the yeast to form acetoin and finally the
relatively innocuous 2,3-butanediol (Fig. 5.2, route 1). Premature removal
of the yeast from the beer leaves a-acetolactate which subsequently oxida-
tively decarboxylates to form diacetyl which accumulates and causes an
unacceptable flavour and aroma (Haukeli and Lie, 1978; Inoue, 1981).
Similarly, contamination by lactobacilli results in accumulation of diacetyl
but by a slightly different route (Fig. 5.2, route 3). Acetolactate synthesis
followed by enzymic oxidation to diacetyl was originally thought to be the
only pathway operating in lactobacilli, but direct synthesis from acetyl-
CoA and acetaldehyde-TPP was demonstrated later (Speckman and
Collins, 1968, 1973) and it appears that both pathways can operate in the
same microorganism Oonsson and Pettersson, 1977). Thus trace infections
of beer, in the absence of yeast, can under certain conditions give rise to
high levels of diacetyl accumulation. Some lactobacilli have been im-
plicated in the production of extracellular polysaccharide or 'rope', but this
is not a serious problem with lactobacilli and is more often attributable to
pediococci or acetic acid bacteria (Rainbow, 1981).

5.3.4 Media for the isolation of lactobacilli

Perhaps nothing in brewery microbiology has generated as much interest
and controversy as the development of culture media for the selective
isolation and enumeration of lactobacilli and pediococci. In their excellent
review of the topic, Casey and Ingledew (1981) list some 23 media designed
for the growth of lactic acid bacteria; some of the more successful formula-
tions are listed in Table 5.3. A more recent review by Holzapfel (1992) lists
31 media for lactobacilli and related bacteria but is not restricted to brewery
varieties. Nevertheless, this is a comprehensive listing of media for lactic
acid bacteria which may give some ideas for new approaches specific for
brewery lactobacilli.
There have been almost as many studies of the effectiveness of
Lactobacillus media as there have been media designed. Unfortunately they
reach almost as many conclusions! It should be remembered that these
media are often being tested for two criteria although this is not always
made clear: recovery and selectivity. Generally, media that are effective at
recovering the majority of lactobacilli from a sample have low selectivity
against other brewery bacteria such as enterobacteria, and those that are
highly selective for lactobacilli demonstrate lower recovery rates.
Nevertheless, some general comments can be made.
For recovery of lactobacilli, maltose is a preferred carbon source. Many
lactobacilli cannot use glucose as a carbon source on initial isolation and,
by supplementing MRS with maltose, Lawrence and Leedham (1979) found
this medium to be superior to Sucrose Agar, Universal Beer Agar, MRS
 Lactobacillus 143
Table 5.3 Some common media suitable for the cultivation of lactic acid bacteria
from breweries
Medium Comments Reference
Cinnamond Medium 2 Fructose as carbon Harrison (1979)
(CM2) source. Bromocresol
green indicator
KOTmedium Enhanced growth of Taguchi et al. (1990)
pediococci
de Man, Rogosa and Glucose as carbon de Man, Rogosa and
Sharpe (MRS)* source Sharpe (1960)
Modified MRSt Supplemented with Lawrence and Leedham
maltose (1979)
MRS and beer Improvement of MRS for Back (1978)
brewery microbiology
Modified Homohiochi General medium used Kleynmans, Heinzl and
Medium for sourdough, Hammes (1989)
vegetables and wine
Homoferm-Heteroferm- For differential recovery McDonald et al. (1987)
Differential medium of homoferm. /heteroferm.
lactic acid bacteria
Improved Nakagawa Semi-solid or solidified Eto and Nakagawa (1975)
Medium agar, with glucose as
carbon source
Lactobacillus Growth Glucose-based complex Van Keer et al. (1983)
Medium (LGM) medium supplemented
with vitamins
NBB Rapid growth of Back (1980)
Lactobacillus and
Pediococcus
Modified NBB Improved version of NBB Nishikawa and Kohgo
(1985)
Raka-Ray No.3 t Complex medium with Saha, Sonda and
maltose and fructose as Middlekauf (1974)
carbon source
Sucrose aga/t: Sucrose-based medium Boatwright and Kirsop
similar to MRS (1976)
VLLB-L41 Beer-based medium Wackerbauer and Emeis
supplemented with (1967)
nutrients, maltose and
xylose
VLB-S7 Designed for isolation Emeis (1969)
and identification of
pediococci

* Recommended by International Committee on Systematic Bacteriology; Subcommittee on

the Taxonomy of the Genus Lactobacillus, for the growth of Lactobacillus.
t Recommended by European Brewery Convention, Analytica Microbiologia Subcommittee.
t Recommended by Institute of Brewing, Analysis Committee.
144 Gram-positive brewery bacteria
Agar and Wallerstein Laboratories Differential Agar. Hsu and Taparowsky
(1977) favoured MRS and Improved Nakagawa Agars for recovery of
lactobacilli and Hug, Schlienger and Pfenniger (1978), in a comparison of
six media, concluded that MRS and Raka-Ray-3 (RR-3) were the most
effective for the detection of lactobacilli. A similar conclusion was reached
by Matsuzawa, Kirchner and Wackerbauer (1979), but they showed that
these media had low selectivity (wort bacteria and Escherichia coli could
grow), which would limit their application in the enumeration of lactic acid
bacteria in yeast and beer samples. To improve the selectivity,
phenylethanol and cycloheximide were added but growth of lactobacilli
was then restricted and when compared with VLB-S7, the latter was found
to be superior. Van Keer et al. (1983) concluded that RR-3 was a valuable
general-purpose medium for lactic acid bacteria and described Lactobacillus
Growth Medium (LGM) which was as efficient as RR-3 in recovery but
promoted better growth. A recent general-purpose medium that promoted
good growth of lactobacilli and pediococci is NBB Agar (Back, 1980; Dachs,
1981) which has been modified by Nishikawa and Kohgo (1985). When
compared with MRS and VLB-S7, development of lactic acid bacteria
occurred very much more rapidly on NBB Agar (Wackerbauer and Rinck,
1983; Back, Durr and Anthes, 1984). However, a recent inter-laboratory
evaluation exercise concluded that no significant differences in colony
counts for Lactobacillus or Pediococcus strains were detected for three media,
NBB, modified NBB or Universal Beer Medium, and that no preference for
one medium was reported (Crumplen, 1988).
Thus, depending on the purpose of the medium and desirability for
selectivity, it would seem that MRS, modified MRS, RR-3, VLB-S7 and NBB
are probably the most favoured media for the isolation and enumeration
of lactobacilli. With regard to the incubation conditions, 30°C in an
anaerobic or semi-anaerobic environment is most suitable.

5.3.5 Identification of lactobacilli

Virtually any non-sporulating, rod-shaped bacterium that is Gram positive
and catalase negative and has been isolated from yeast, beer or brewery
plant will be a Lactobacillus, and since these bacteria are usually responsible
for beer spoilage it is generally unnecessary to identify it further. Neverthe-
less, for some purposes it may be desired to identify an isolate to species
level, particularly since different species have been attributed with different
spoilage potential (Back, 1981). Information for the identification of brew-
ery lactobacilli based on the results of Eschenbecher (1968, 1969), Sharpe
(1979), Back (1981), Kandler and Weiss (1986) and Hammes, Weiss and
Holzapfel (1993) is given in Table 5.4. Because of the phenotypic similarity
of many of these bacteria, accurate identification is often a problem. General
methods for conducting these tests are given by Harrigan and McCance
(1976) and most practical microbiology manuals.
Table 5.4 Phenotypic features of Lactobacillus species of brewery origin

'§ til
I:::
<li
~ ;:::
0"'
& ....
0 .8
u
>i.
]- ~
'"
;:;
~
til '"
.~ til
til
] '6.... E .;::
6 a....
u til
6;::: .;::
<li .... 0
;>
..8<li
til <li
..8<li .a ~'<il ~<li
e ~ .;; .;; .§u
'OtilJ ::9 ~
.... §, '~" <li <li ;:::
6.... ":e;::: -B
<li 0 0 ;::: .... ....
u "0 u u u 0..
'" ~ ~ ~ ~ J::: :3
Test '"
.J .J .J .J .J .J .J .J .J .J .J .J
Acidfrom:
arabinose d + + + d
cellobiose + d + + d
dextrin + • d d d d • • •
glucose + + + + + + + + + + + +
lactose d d + d + d d +
maltose + d + + + + + + + + d +
mannitol + + + + •
mannose + + + d + + • w
melezitose + d +
melibiose d + + + +
rhamnose + •
ribose + + + + + + + w
sorbitol + d + •
sucrose + + + + + d d + d •
trehalose + d + d
xylose d d + d d
Gas from glucose + + + + + +
Growth at 15°C + + + + + + + + + •
Arginine dihydrolase d + + + +
Lactate configuration L D D(L) D DL DL DL DL DL DL DL DL
+,90% or more strains positive; -, 90% or more strains negative; d, 11--89% strains positive; w, positive to weak
reaction; • not determined.
146 Gram-positive brewery bacteria
(a) Serological identification of lactobacilli
Serological detection and identification of lactobacilli is attractive due to
the specificity and rapidity of the antibody / antigen reaction but this ap-
proach has not been exploited to any extent. This probably stems from the
taxonomic complexity of the group and problems due to cross-reactions of
multivalent sera (Wackerbauer and Emeis, 1968). Oolezil and Kirsop (1975)
demonstrated at least two antibody components in sera raised to lacto-
bacilli from beers: one that reacted with a species-specific antigen and a
second non-specific antibody that reacted with pediococci and yeast. By
adsorbing these preparations, rapid identification of spoilage lactobacilli
including L. casei and L. brevis was possible, but difficulties were still
encountered with L. plantarum. Similarly, Nishikawa, Kohgo and
Karakawa (1979) used antisera to detect L. brevis in beer and Rinck and
Wackerbauer (1987) detected L. lindneri amongst other lactobacilli serolog-
ically. It is unfortunate that this promising approach has not been exploited
and it is likely that the introduction of monoclonal antibodies will stimulate
interest in this area (Whiting et al., 1992).

(b) Identification kits for lactobacilli

One of the most significant recent developments in diagnostic micro-
biology has been the introduction of commercial miniaturized systems for
the identification of bacteria. In general, these comprise disposable trays of
media for the testing of about 20 phenotypic attributes (see Chapter 9). The
pattern of results is then compared with patterns for reference organisms
and an identification obtained (0'Amato, Holmes and Bottone, 1981).
Concomitant with these kits, computerized data bases and probabilistic
identification schemes have been developed (Sneath, 1978; Priest and
Williams, 1993). Such systems were developed for medically important
bacteria such as Enterobacteriaceae and have been used to identify enteric
bacteria from breweries and related environments (Van Vuuren et al., 1978,
1981; Ingledew, Sivaswamy and Burton, 1980). The phenotypic homo-
geneity of the lactobacilli necessitates the use of at least 50 tests (e.g. the
API 50 tray) to obtain an identification and still the results may not be
entirely reliable. This is often due to strain variability on storage of the
organism. In one example, upon initial isolation a strain was identified as
L. brevis but reidentified after 15 weeks as L. plantarum. This transition
required the change of five test results which occurred independently of
experimental error (O.R. Lawrence, personal communication). In this con-
nection it should be remembered that lactobacilli harbour numerous
plasmids, many of which code for carbohydrate utilization pathways
(McKay, 1983), and it may be that plasmid loss is responsible for the altered
phenotype.
A second reason for the failure of commercial identification kits is simply
 Pediococcus 147
that they have been used for brewery microbiology so infrequently that
results for brewery bacteria such as L. lindneri or L. brevisimilis are not
included in the data bases. Nevertheless, the Minitek system has been used
successfully for the identification of lactobacilli (Gilliland and Speck, 1977)
and a simple modification of this system was claimed to produce results
for lactobacilli that were consistent with those obtained by traditional
procedures (Benno and Mitsuoka, 1983). As the ecology of brewery bacteria
is understood in more detail and the benefits of the identification of
dangerous spoilage bacteria are acknowledged, identification kits will
probably be used more extensively in brewery quality control and research
laboratories.

5.4 PEDIOCOCCUS

5.4.1 Classification
Pediococci are homofermentative cocci that occur in pairs and tetrads
through division in two planes. They have a long association with brewing
microbiology and were originally known as 'sarcinae' because their cell
morphology was confused with the cubical packets of eight cells of true
sarcinae. Shimwell and Kirkpatrick (1939) first showed that the brewery
cocci were lactic acid bacteria but assigned them to the genus Streptococcus
as 'Str. damnosus' (Shimwell, 1941). This classification was not accepted,
although the close relationships of pediococci and streptococci are often
stressed (Whittenbury, 1978), and these bacteria were placed in the genus
Pediococcus which had been used by Claussen for strains he had earlier
isolated from European beers. The classification and nomenclature of the
pediococci continued to cause confusion however, largely because
Nakagawa and Kitahara (1959) used the name 'P. cerevisiae' to describe
the common cocci from beer and breweries rather than P. damnosus as
used by Gunther and White (1961) and Coster and White (1964). Solberg
and Claussen (1973a) noted that these two names were being used for the
same bacterium and in response to Garvie's (1974) request, the Judicial
Commission of the International Committee on Systematic Bacteriology
ruled that 'P. cerevisiae' had not been validly published and the name
P. damnosus was conserved. Thus P. damnosus and 'P. cerevisiae' are syn-
onyms and the former is used for the common brewery cocci (Sharpe,
1979; Priest, 1981b).
The close relationship of the pediococci with Lactobacillus has been
emphasized by the rRNA studies in which most pediococci form a small
cluster with the L. casei group of obligately homofermentative and
heterofermentative bacilli. The closest phylogenetic relatives include L. kefir
and L. buchneri (Collins et al., 1991). The rRNA studies have also clarified
the boundaries of the genus Pediococcus and reveal it as a homogeneous
148 Gram-positive brewery bacteria
Table 5.5 Phenotypic features of Pediococcus species of brewery origin

til
;::l
:a<11
S
....
....<11
.S
Ii.
V}
..c
;:!
V}

til til
;::l ;::l til
til til
;::l ·0
;::l <11
u
<11
u .8!\l U til ..0
til
0
!\l !\l ·2 ;::l u
til til
....c0 .S ·c '3 ...'S
§
!\l
....c0
<11
0...
0
....X
<11
i::
!\l
:a·0
<11
'"0 0... 0... .S '"0 0... !\l
Test 0..: 0..: 0..: 0..: 0..: 0..: 0..:

Acid from:
arabinose + + +
cellobiose + + + + + + +
dextrin +
lactose + + + d +
maltose + + + + + +
maltotriose + + +
melibiose d
raffinose d
ribose + + + +
sorbitol
starch +
sucrose d d d + +
trehalose + + + + d +
xylose d +
Growth at: + + + + + +
35°C
50°C +
6.5%NaCl + + + + +
15% NaCl +
pH 4.5 + + + + + +
pH 7.5 + + + + + +
Arginine dihydrolase + + +
Pseudocatalase + • • +
Lactate configuration DL DL DL DL L(+) DL L(+)

All species grow at lOoC and at pH 5.5, produce acid from glucose, fructose and mannose and
split aesculin.
For key see footnote to Table 5.4.

taxon with P. dextrinicus as the most distantly related true Pediococcus

forming an independent lineage. Within the Pediococcus taxon, P. damnosus
and P. parvulus form one group while P. acidilactici and P. pentosaceus form
a second (Collins et al., 1991) and this distinction is also evident from
 Pediococcus 149
physiological tests such as the arginine dihydrolase reaction (Table 5.5). The
non-aciduric, microaerophilic species P. urinae-equi is synonymous with
'Aerococcus viridans' and the non-aciduric, salt-requiring species,
P. halophilus has been removed to a separate genus as Tetragenococcus
(Collins, Williams and Wallbanks, 1990).
A comprehensive study of 840 cocci from various sources including
wine, beer and brewery equipment was based on phenotypic tests and
DNA sequence homology and supported the integrity of the six established
species, P. acidilactici, P. damnosus, P. dextrinicus, P. parvulus, P. pentosaceus
as well as T. halophilus (Back, 1978; Back and Stackebrandt, 1978). Some
strains, largely derived from plant material, which had previously been
placed in P. damnosus because of their inability to ferment pentoses, showed
high DNA homology with typical strains of P. pentosaceus and were as-
signed to the subspecies 'P. pentosaceus subsp. intermedius'. Furthermore,
some brewery strains showed only 40% reassociation with DNA from
P. damnosus or P. parvulus, their closest relatives, and were placed in the
new species, P. inopinatus (Back, 1978). A numerical taxonomic analysis of
96 Gram-positive cocci from beer and breweries and reference strains
confirmed the species P. damnosus, P. dextrinicus, P. parvulus and
T. halophil us. No strains of P. inopinatus were recovered in this study,
however; virtually all the beer isolates were identified as P. damnosus
(Lawrence and Priest, 1981).

5.4.2 Distribution of pediococci in beer and breweries

Pediococcus damnosus is undoubtedly the most common, and feared,
Pediococcus found in beer (Solberg and Claussen, 1973a; Back, 1978;
Lawrence and Priest, 1981). It is particularly interesting that this organism
is apparently only found in beer, brewing yeast and wines and not in
brewing raw materials or plant materials. This suggests considerable
adaptation to this particular habitat. P. inopinatus has been isolated from
wine, beer and beer yeast but is also present in fermented vegetables and
milk (Back, 1978). P. dextrinicus and P. pentosaceus can be isolated from beer
and brewery plant but more rarely than P. damnosus.
The distribution of pediococci within the brewery has been examined in
a study of some 200 isolates (McCaig and Weaver, 1983). P. damnosus was
common in beer and late fermentations but seldom encountered in pitching
yeast. P. damnosus var. I (probably P. inopinatus) and P. pentosaceus were
frequently detected in pitching yeasts (44% and 35% respectively of those
cocci detected) but rarely found in late fermentations or the finished prod-
uct. With regard to those pediococci found in fermented foods such as
silage, cucumbers or olives, P. parvulus, P. inopinatus and P. dextrinicus are
commonly isolated. The biotechnological applications of the pediococci in
the processing and preservation of meat and plant foods have been
reviewed recently (Raccach, 1987).
150 Gram-positive brewery bacteria
5.4.3 Spoilage of beer by pediococci

(a) Growth ofpediococci in beer

Only P. damnosus and, to a lesser extent, P. inopinatus can grow in beer
although other species will survive in beer for long periods (Back, 1978;
Lawrence and Priest, 1981). Not only does the low pH inhibit these other
species but they are markedly less hop tolerant than P. damnosus, which is
particularly resistant to the antiseptic action of hop constituents (Fernandez
and Simpson, 1993). As with the lactobacilli, there is evidence that some
beers are more resistant to spoilage by P. damnosus than others. This may
result from availability of growth factors. There is some controversy con-
cerning the vitamins required for the growth of these bacteria, but biotin
and riboflavin promote the growth of most P. damnosus strains and calcium
pantothenate is essential for the growth of this organism (Solberg and
Claussen, 1973b). Calcium pantothenate can be replaced by a ~-glucoside
of pantothenate, a compound that occurs in tomato juice and is essential
for the growth of Leuconostoc oenos (Eto and Nakagawa, 1974, 1975).

(b) Effect of pediococci on beer quality

Pediococci are traditionally associated with' sarcina sickness', characterized
by acid formation and the 'buttery' aroma of diacetyl. The pathways for
diacetyl synthesis conform to those shown in Fig. 5.2 for lactobacilli but the
amount synthesized varies between species. P. pentosaceus does not prod uce
diacetyl and P. inopinatus does not make it as prolifically as P. damnosus
(McCaig and Weaver, 1983). In trial fermentations, contamination by either
P. damnosus or P. inopinatus extended the fermentation time, depressed the
amount of yeast in suspension by some 30% and produced high diacetyl
contents in the beer. The diacetyl concentration was much greater in the case
of P. damnosus. Ropiness in beers has been attributed to pediococci
(Shimwell, 1945) but is dependent on the presence of some fermentable
sugar. The extracellular slime produced by these bacteria is thixotropic and
comprises a complex heteropolymer containing glucose, mannose and
nucleic acid (Hough et al., 1982).

5.4.4 Media for the isolation of pediococci

Those media designed for the cultivation of lactobacilli can generally be
used for pediococci. However, the importance of pediococci in relation to
beer spoilage and their poor growth rates has led to the development of
specific media for their isolation. Nakagawa's medium based on unhopped
beer is a semi-solid agar in which pediococci appear as granular colonies
after 5 days at 25°C. This has since been modified for use for all lactic acid
bacteria (Table 5.3). Medium VLB-S7 (Emeis, 1969) was introduced for the
 Leuconostoc 151
isolation of beer pediococci and found to be superior to several general-
purpose media. However, NBB medium and modified NBB had no advan-
tages for the growth and recovery of pediococci when tested against
Universal Beer Agar (Crumplen, 1988).

5.4.5 Identification of pediococci

Pediococci are typically catalase-negative organisms which can be distin-
guished from leuconostocs by virtue of their homofermentative mode of
sugar metabolism and characteristic formation of tetrads. However, the
catalase test is not always reliable since many strains of P. pentosaceus, when
grown on a low sugar-containing medium, produce a 'pseudocatalase'
which decomposes H 20 2 but differs from catalase in being sensitive to azide
and does not contain a haem group (Whittenbury, 1978). Thus there is the
possibility of confusing these strains with the catalase-positive micrococci,
but only the latter will grow at pH 7.0 on nutrient agar. A scheme for the
identification of the genus Pediococcus is given at the end of this chapter;
this section is devoted to identification at the species level.
Since pediococci vary in their potential for beer spoilage, accurate
identification may sometimes be important. The information in Table 5.5
has been derived from the results of Back (1978, 1982), Sharpe (1979),
Lawrence and Priest (1981) and Weiss (1992). Only P. damnosus,
P. inopinatus, P. pentosaceus, P. pentosaceus subsp. intermedius and
P. dextrinicus are likely to be encountered in the brewery but data for
P. acidilactici, P. parvulus and T. halophilus have been included for com-
pleteness. Methods for these tests are the same as those for lactobacilli and
are outlined at the end of this chapter. Other approaches to the identifi-
cation of pediococci have successfully used the API identification kit for
lactobacilli (Dolezil and Kirsop, 1977).
Serological methods for the identification of pediococci were first intro-
duced by Dolezil and Kirsop (1976) and have received added interest from
the introduction of monoclonal antibodies. A membrane filter-based
immunofluorescent antibody test has been developed which can detect
contaminating pediococci at 0.001% of pitching yeasts in less than 4 h
(Whiting et al., 1992). This is, of course, a major improvement on traditional
plating tests.

5.5 LEUCONOSTOC

Heterofermentative cocci that are sometimes oval or even short rods and
occur in pairs or short chains are classified in the genus Leuconostoc. These
organisms are found on vegetables and fruit and in fermenting vegetable
matter but apparently occur rarely in breweries. Leuconostocs and
obligately heterofermentative lactobacilli have similar nutritional
152 Gram-positive brewery bacteria
requirements, they grow on the same media, and can often be confused since
some lactobacilli can occur as coccobacilli. Nevertheless, rRNA studies have
clearly demonstrated the distinction between the leuconostocs and Lacto-
bacillus (Martinez-Murcia, Harland and Collins, 1993). Leuconostocs can be
distinguished from most lactobacilli by their inability to produce ammonia
from arginine and by forming D(-)- rather than DL-lactate from glucose
(Sharpe, 1979). Leuconostocs form three major phylogenetic lineages:
• Leuconostoc sensu stricto, which includes Leuc. mesenteroides and five
other Leuconostoc species;
• the Leuc. paramesenteroides group now reclassified as Weissella and
which includes the single Leuconostoc species (as W. paramesenteroides)
and W. confusa (previously Lactobacillus confusus) and some other
heterofermentative lactobacilli;
• Leuc. oenos, which forms a single membered taxon (Collins et al., 1991,
1993; Martinez-Murcia, Harland and Collins, 1993).
Leuconostocs occur in the early stages of Scotch whisky fermentations
(Bryan-Jones, 1975) and a recent numerical taxonomic study of these organ-
isms classified the majority of isolates into two groups: Leuc. mesenteroides
and a possible new species (Priest and Barbour, 1985; Priest and Pleasants,
1988). There was no indication that Leuc. oenos, the bacterium responsible
for the malolactic fermentation in wines, is present in distilleries. The only
species to be found in breweries appears to be Leuc. mesenteroides (Back,
1982) but it is not responsible for beer spoilage. Exceptionally acid-tolerant
strains of Leuc. mesenteroides have been isolated from fruit mashes and they
are also associated with traditional African beverages (Sanni and Oso,
1988). Data for the identification of Leuconostoc species are given in
Table 5.6.

5.6 HOMOFERMENTATIVE COCCI

Gram-positive cocci that are homofermentative and occur in pairs or chains

are classified in the genera Enterococcus, Lactococcus, Streptococcus and
Vagococcus. These bacteria are widely distributed in raw milk and dairy
products, on plant material and in the mouth and intestines of humans and
animals (Sharpe, 1979). The only member of this group likely to be encoun-
tered in the brewery is Lactococcus lactis, a common organism on plants, but
best known for the production of diacetyl from citrate and its role in butter
manufacture. Although Lactococcus lactis occurs rarely in breweries (Back,
1982) it could be confused with other catalase-negative, Gram-positive
cocci since it would grow on media designed for the cultivation of lactic
acid bacteria. It can be distinguished from the leuconostocs by its
homofermentative metabolism of sugars and from the brewery pediococci
by its cell morphology (chains rather than tetrads) and growth at pH 9.2
 Micrococcus and Staphylococcus 153
(pediococci are unable to grow at high pH). A more definite identification
could be obtained using commercially available Group N antiserum for the
typing of lactococci but it should be remembered that cross-reactions
between this and pediococci have been noted (Dolezil and Kirsop, 1976).
There have been no reports of Lactococcus lactis growing in beer.

5.7 MICROCOCCUS AND STAPHYLOCOCCUS

The genera Micrococcus and Staphylococcus contain Gram-positive cocci

which form catalase and do not have a lactic fermentative mode of metab-
olism. Microscopically these bacteria form clusters of cells but these are not
always apparent and then the bacteria are virtually indistinguishable from
lactic cocci. Nevertheless, they can be readily differentiated by their ability
to grow at pH 7.0 on nutrient agar and the catalase test (see section 5.9).
Although Micrococcus and Staphylococcus differ widely in DNA base com-
position (30-40 and 66-75 mol%, respectively) and have very different
rRNA sequences, there are few reliable tests that distinguish them. The
most useful criterion is that staphylococci are facultative anaerobes al-
though Staph. saprophyticus and related species grow poorly or not at all
under anaerobic conditions (Kloos and Schleifer, 1975a,b) whilst micro-
cocci, with few exceptions, are strict aerobes. Moreover, staphylococci are
erythromycin resistant and not lysed by lysozyme, but most micrococci are
erythromycin and lysozyme sensitive. Finally, staphylococci are sensitive
to lysostaphin whereas most micrococci are resistant.
These bacteria are not generally considered to be important brewery
bacteria but recent studies have shown that they are widely distributed in
beer and breweries and in some instances have been responsible for beer
spoilage. A numerical taxonomic study of 96 cocci from the brewing
environment revealed that the skin commensals Staph. saprophyticus, Staph.
epidermidis and the widely distributed M. varians were common in beer and
breweries, with the latter sometimes being isolated from pitching yeast
(Lawrence and Priest, 1981). However, these bacteria grew poorly, if at all,
at pH 4.5 and below, were susceptible to hop constituents, and in the case
of M. varians were obligately aerobic. They are therefore unable to spoil beer
and are probably adventitious contaminants arising from humans or from
raw materials. Many can survive in beer for long periods, however, and
could be confused with spoilage organisms in quality control procedures,
particularly since some strains of M. kristinae display tetrad-like packets of
cells in Gram stains.
Micrococcus kristinae is an atypical Micrococcus in that it is facultatively
anaerobic (Kloos, Tornabene and Schleifer, 1974). Strains of this organism
are relatively acid tolerant and hop resistant and have been responsible for
beer spoilage particularly in relatively high-pH, low-bitterness products
(Back, 1981; Lawrence and Priest, 1981). The intensity of growth is related
154 Gram-positive brewery bacteria
to the oxygen content of the beer, but even traces of growth under anaerobic
conditions can yield a fruity aroma and an atypical taste (Back, 1981). It is
therefore important to be able to distinguish M. kristinae from the harmless
M. varians and this is most readily achieved by its ability to grow anaero-
bically. Phenotypic descriptions of M. kristinae have been published by Back
(1981) and Kocur, Kloos and Schleifer (1992).

5.8 ENDOSPORE-FORMING BACTERIA

Aerobic endospore-forming bacteria belonging to the genus Bacillus have

on occasion caused problems in breweries. Spores from these bacteria are
present in malt and cereal adjuncts and will survive wort boiling but are
subsequently unable to germinate because of the low pH of fermenting
wort and beer. Moreover, all bacilli are susceptible to hops so there is no
problem from these bacteria in beers (Smith and Smith, 1993). It is a
sobering thought, however, that a spore-forming variant of L. brevis would
create havoc in the modem brewing industry!
Thermophilic, endospore-forming bacilli have been isolated from brew-
ery plant, malt and sweet worts (McMurrough and Palmer, 1979; Healy and
Armitt, 1980; Smith and Smith, 1992). Identified as B. coagulans and
B. coagulans-B. stearothermophilus intermediates, these bacteria produce co-
pious quantities of lactic acid in sweet wort held at 55-70°C for more than
2 h. Although usually considered detrimental, controlled acidification of
the wort can have benefits both in beer preparation and for the beer in
general, including improvements in flavour balance and microbiological
stability (Back, 1988). However, one of the more disturbing features of the
B. coagulans strains is their ability to contribute to nitrosamine formation
through the reduction of nitrate to nitrite (Smith and Smith, 1992).

5.9 IDENTIFICATION OF GENERA OF GRAM-POSITIVE BACTERIA

OF BREWERY ORIGIN

This scheme (Table 5.6) is designed to enable rapid identification, to the

generic level, of a Gram-positive bacterium isolated from the brewing
environment. This will generally be sufficient to predict the spoilage po-
tential of the organism but in some instances it may be desirable to pursue
the identification to the species. Tables and keys for this purpose have been
included in the relevant sections of this chapter.

5.9.1 Test procedures

The tests involved in the identification key use common microbiological
procedures but details of the methods are given below for those readers
 Identification of Gram-positive bacteria of brewery origin 155
Table 5.6 A key for the identification of genera of Gram-positive bacteria of
brewery origin
(1) Catalase* Positive -7 (5)
Negative -7 (2)
(2) Cell morphology Rods: Lactobacillus
Coccit -7 (3)
(3) Gas production from glucose Positive: Leuconostoc/ Weissella+
Negative -7 (4)
(4) Cells in tetrads Pediococcus
Cells in pairs / chains Lactococcus
(5) Endospore-forming rods Positive: Bacillus
Negative -7 (6)
(6) Anaerobic growth Positive -7 (7)
Negative: Micrococcus
(7) Lysis by lysostaphin Positive: Staphylococcus
Negative: M. kristinae

* Some pediococci may be catalase positive on media of low sugar content. If doubtful, repeat
test using culture from medium with at least 1% glucose or test reaction for sensitivity to 0.01
M azide; catalase is inhibited, pseudocatalase is not (Whittenbury, 1978).
t Some leuconostocs are elongate and can be confused with heterofermentative lactobacilli
and vice versa: see section 5.5.
+ Differentiation of Leuconostoc from Weissella is problematic and requires combinations of
characters for the particular species (see Collins et ai., 1993).

unfamiliar with bacterial identification. For further information the reader

should consult Cowan and Steele's Manual for the Identification of Medical
Bacteria (Barrow and Feltham, 1993) for general tests, Sharpe (1979) and
Hammes, Weiss and Holzapfel (1993) for tests involving lactic acid bac-
teria and Kloos, Tornabene and Schleifer (1974), Baird Parker (1979) and
Kocur, Kloos and Schleifer (1992) for those involving micrococci and
staphylococci.

(a) Catalase test

Method 1
Grow the organism on a plate or slope of a suitable medium. Add 1 ml of
3% (v Iv) H 20 2 and examine immediately for evolution of gas which
indicates catalase activity. Anaerobically incubated cultures should be
exposed to air for at least 30 min before adding the H 20 2•

Method 2
Grow the organism in a suitable broth; after incubation add 1 ml of 3%
(v Iv) H 20 2 and examine for gas evolution.

Controls
Positive: Staphylococcus epidermidis or Micrococcus varians; negative:
Lactobacillus brevis.
156 Gram-positive brewery bacteria
(b) Gas production from glucose
Since the amount of gas produced is often slight, the method of Gibson and
Abd-el-Malek (1945) is used. A suitable semi-solid medium comprises:
meat extract (Lab-Lemco), 0.5 g; tryptone, 0.5 g; yeast extract, 0.5 g; Tween
80, 0.05 ml; agar, 0.2 g; water to 100 ml. Glucose, sterilized separately, is
added to 5% (w Iv) and the medium adjusted to pH 6.0. Cultures are stab
inoculated into the medium in tubes using a vigorous heavy inoculum. The
cultures are overlayered with a seal of sterile molten agar and, after
incubation, gas bubbles collecting beneath the agar seal denote
heterofermentative growth.

Controls
Positive: Lactobacillus brevis; negative: Pediococcus damnosus.

(c) Anaerobic growth

A modification of the Hugh and Leifson (1953) oxidation/fermentation
(O/F) test is used. Duplicate tubes containing tryptone, 1 g; yeast extract,
0.1 g; glucose, 1 g; bromocresol purple, 0.004 g; agar, 0.2 g; in 100 ml of
water, pH 7.0 are stab inoculated with the organism and one tube is over-
layered with paraffin oil to provide anaerobic conditions. Growth and acid
(yellow) production in both tubes denote fermentative metabolism; growth
and acid in only the open tube indicate strictly oxidative metabolism.

Controls
Fermentative: Staphylococcus epidermidis; oxidative: Micrococcus varians.

(d) Lysis by lysostaphin and lysozyme

Sterile filtered lysostaphin (final concentration 200 J.lg mrl) or lysozyme
(final concentration 25 J.lg mrl) is added to sterile, molten peptone (1%),
yeast extract (0.1 %), glucose (1.0%), sodium chloride (0.5%), agar medium
at pH 7.0. Plates are streaked with the organism and incubated for 2 days.

Controls
Lysostaphin-sensitive, lysozyme-resistant, Staphylococcus epidermidis.
Lysostaphin-resistant, lysozyme-sensitive, Micrococcus varians.

5.10 CONCLUDING REMARKS

In this article I have emphasized the diversity of Gram-positive bacteria

that can be recovered from beer. It is evident that rather than simply 'lactic
rods', at least ten species of Lactobacillus can be the cause of beer spoilage.
 References 157
Similarly 'lactic cocci' encompasses several Pediococcus species and prob-
ably M. kristinae. The time may be ripe to consider some of the rapid
detection and identification methods for bacteria described in later chap-
ters in a quality control context and on a routine basis. Only then will we
become aware of the full spectrum of brewery bacteria and their role in
beer spoilage. The cost might be high, but savings would accrue for
example if it could be demonstrated unequivocally that a beer contami-
nated with Lactobacillus X could be sent to trade rather than destroyed or
reprocessed because strains of that species were unable to grow in and
consequently spoil that particular product. Beyond such immediate inter-
ests a more detailed picture of brewery microbiology would emerge that
would be of interest and value to all those involved in the technical side
of the industry.

REFERENCES

Archibald, F.S. and Fridovich, L. (1981) Journal of Bacteriology, 146, 928.

Axelsson, L.T. (1993) In Lactic Acid Bacteria (eds S. Salminen and A van Wright),
Marcel Dekker, New York, p. 1.
Back, W. (1978) Brauwissenschaft, 31, 237, 312, 336.
Back, W. (1980) Brauwelt, 120, 1562.
Back, W. (1981) Monatsschrift fUr Brauerei, 34, 267.
Back, W. (1982) Brauwelt, 45, 1562, 2090.
Back, W. (1987) Monatsschrift fUr Brauwissenschaft, 40, 484.
Back, W. (1988) Monatsschrift fUr Brauwissenschaft, 41, 152.
Back, W. and Stackebrandt, E. (1978) Archives of Microbiology, 118, 79.
Back, W., Durr, P. and Anthes, S. (1984) Monatsschrift fUr Brauwissenschaft, 37,126.
Baird-Parker, AC. (1979) In Identification Methods for Microbiologists, 2nd edn (eds
F.A. Skinner and D.W. Lovelock), Academic Press, London, p. 201.
Barbour, E.A. and Priest, F.G. (1983) In Current Developments in Malting, Brewing and
Distilling (eds F.G. Priest and I. Campbell), Institute of Brewing, London,
p.289.
Barre, P. (1969) Archiv fUr Mikrobiologie, 68, 74.
Barrow, G.I. and Feltham, RK.A. (eds) (1993) Cowan and Steel's Manual for the
Identification of Medical Bacteria, 3rd edn. Cambridge University Press,
Cambridge.
Benno, Y. and Mitsuoka, T. (1983) Systematic and Applied Microbiology, 4, 123.
Boatwright, J. and Kirsop, B.H. (1976) Journal of the Institute of Brewing, 82, 343.
Bradley, S.G. (1980) In Microbiological Classification and Identification (eds
M. Goodfellow and RG. Board), Academic Press, London, p.11.
Bryan-Jones, G. (1975) In Lactic Acid Bacteria in Beverages and Food (eds J.G. Carr,
C.V. Cutting and G.c. Whiting), Academic Press, London, p. 165.
Carr,J.G., Davies, P.A., Dellaglio, F. et al. (1977) Journal ofApplied Bacteriology,42,219.
Casey, G.P. and Ingledew, W.M. (1981) Brewers' Digest, 55, (March), 38.
Collins, M.D., Williams, A.M. and Wallbanks, S. (1990) FEMS Microbiology Letters,
70,255.
Collins, M.D., Farrow, J.AE., Philips, B.A. et al. (1987) International Journal of
Systematic Bacteriology, 97, 310.
Collins, M.D., Rodrigues, U., Ash,C. etal. (1991) FEMSMicrobiology Letters, 77, 5-12.
158 Gram-positive brewery bacteria
Collins, M.D., Aguirre, M., Facklam, RR et al. (1992) Journal of Applied Bacteriology,
73,433.
Collins, M.D., Samelis, J., Metaxopoulos, I. and Wallbanks, S. (1993) Journal ofApplied
Bacteriology, 75, 593.
Condon, S. (1987) FEMS Microbiology Reviews, 46, 269.
Coster, E. and White, H. (1964) Journal of General Microbiology, 37,15.
Crump len, R (1988) Journal of American Society of Brewing Chemists, 46, 129.
Dachs, E. (1981) Brauwelt, 121,1778.
D'Amato, RF., Holmes, B. and Bottone, E.J. (1981) Critical Reviews in Microbiology,
9, l.
De Ley, J. (1969) Journal of Theoretical Biology, 22, 89.
Delwiche, E.A. (1961) Journal of Bacteriology, 81, 416.
de Man, I.e., Rogosa, M. and Sharpe, M.E. (1960) Journal of Applied Bacteriology,
23,130.
Dolezil, L. and Kirsop, B.H. (1975) Journal of the Institute of Brewing, 81, 28l.
Dolezil, L. and Kirsop, B.H. (1976) Journal of the Institute of Brewing, 82, 93.
Dolezil, L. and Kirsop, B.H. (1977) Journal of Applied Bacteriology, 42, 213.
Dolezil, L. and Kirsop, B.H. (1980) Journal of the Institute of Brewing, 86, 122.
EI Soda, M. and Desmazeaud, M.J. (1982) Canadian Journal of Microbiology, 28, 118l.
EI Soda, M., Zeyada, N., Desmazeaud, M.J., Mashaly, R and Ismail, A (1982) Science
Alimentaire, 2, 26l.
Emeis, e.e. (1969) Monatsschrift for Brauerei, 22, 8.
Eschenbecher, F. (1968) Brauwissenschaft, 21,424, 464.
Eschenbecher, F. (1969) Brauwissenschaft, 22, 14.
Eto, M. and Nakagawa, A (1974) Bulletin of Brewing Science, 20,37.
Eto, M. and Nakagawa, A (1975) Journal of the Institute of Brewing, 81,232.
Fernandez, J.L. and Simpson, W.J. (1993) Journal of Applied Bacteriology, 75, 369.
Garvie, E.I. (1974) International Journal of Systematic Bacteriology, 24, 30l.
Gibson, T. and Abd-el-Malek, Y. (1945) Journal of Dairy Research, 14, 35.
Gilliland, S.E. and Speck, M.L. (1977) Applied and Environmental Microbiology,
30,1289.
Gunther, H.L. and White, H. (1961) Journal of General Microbiology, 26,199.
Hammes, W.P., Weiss, N. and Holzapfel, W. (1993) In The Prokaryotes, 2nd edn (eds
A Balows, H.G. Triiper, M. Dworkin et al.), Springer-Verlag, New York,
p.1535.
Harrigan, W.F. and McCance, M.E. (1976) Laboratory Methods in Food and Dairy
Microbiology, Academic Press, London.
Haukeli, AD. and Lie, S. (1978) Journal of the Institute of Brewing, 84, 85.
Healy, P. and Armitt, J. (1980) Journal of the Institute of Brewing, 86,169.
Holzapfel, W.H. (1992) International Journal of Food Microbiology, 17,113.
Hough, J.5., Briggs, D.E., Stevens, R and Young, T.W. (1982) Malting and Brewing
Science, Vol. 2, 2nd edn, Chapman & Hall, London.
Hsu, W.P. and Taparowsky, J.A. (1977) Brewers' Digest, 52, (February), 48.
Hug, H., Schlienger, E. and pfenniger, H. (1978) Schweizer Brauerei-Rundschau,
89,145.
Hugh, Rand Leifson, E. (1953) Journal of Bacteriology, 66, 24.
Ingledew, W.M., Sivaswamy, G. and Burton, J.D. (1980) Journal of the Institute of
Brewing, 86, 165.
Ingram, M. (1975) In Lactic Acid Bacteria in Beverages and Food (eds I.G. Carr,
C.V. Cutting and G.e. Whiting), Academic Press, London, p. l.
Inoue, T. (1981) MBAA Technical Quarterly, 18, 62.
Johnston, M.A. and Delwiche, E.A. (1965) Journal of Bacteriology, 90, 347.
Jonsson, H. and Pettersson, H.-E. (1977) Milchwissenschaft, 32,587.
Kandler,o. (1983) Antonie van Leeuwenhoek, 49, 209.
 References 159
Kandler, O. and Weiss, N. (1986) In Bergey's Manual of Systematic Bacteriology, Vol.
2 (eds P.H.A. Sneath, N.S. Mair, M.E. Sharpe and J.G. Holt), Williams and
Wilkins, Baltimore, p. 1208.
Kleynmans, V., Heinzl, H. and Hammes, W.P. (1989) Systematic and Applied
Microbiology, 11, 267.
Kloos, W.E. and Schleifer, K-H. (1975a) Journal of Clinical Microbiology, 1, 82.
Kloos, W.E. and Schleifer, K.-H. (1975b) International Journal of Systematic
Bacteriology, 25, 62.
Kloos, W.E., Tornabene, I.G. and Schleifer, K-H. (1974) International Journal of
Systematic Bacteriology, 24, 79.
Kocur, M., Kloos, W.E. and Schleifer, KH. (1992) In The Prokaryotes, 2nd edn (eds
A. Balows, H.G. Triiper, M. Dworkin et al.), Springer-Verlag, New York,
p.1301.
Konings, W.N., Poolman, B. and Driessen, A.J.M. (1989) CRC Critical Reviews in
Microbiology, 16, 419.
Lawrence, D.R and Leedham, P.A. (1979) Journal of the Institute of Brewing, 85, 119.
Lawrence, D.R and Priest, F.G. (1981) Proceedings of the 18th Congress of the European
Brewery Convention, Copenhagen, IRL Press, Oxford, p. 217.
Lawrence, R.e. and Thomas, I.D. (1979) Symposium of the Society for General
Microbiology, 29, 187.
London, J. (1976) Annual Review of Microbiology, 30,279.
London, J. and Chace, N.M. (1977) Proceedings of the National Academy of Sciences of
the United States of America, 74, 4296.
London, J. and Chace, N.M. (1979) Journal of Bacteriology, 140,949.
Makinen, V., Tanner, Rand Haikara, A. (1981) Brauwissenschaft, 34,173.
Manca de Nadre, M.e., Pesce de Ruiz Holgado, A.A. and Oliver, G. (1988) Biochemie,
70,367.
Martinez Murcia, A.J. and Collins, M.D. (1990) FEMS Microbiology Letters, 70, 73.
Martinez-Murcia, A.J., Harland, N.M. and Collins, M.D. (1993) Journal of Applied
Bacteriology, 74,532.
Matsuzawa, K, Kirchner, G. and Wackerbauer, K (1979) Monatsschrift flir Brauerei,
32,312.
McCaig, R and Weaver, RL. (1983) MBAA Technical Quarterly, 20, 31.
McDonald, L.e., McFeeters, RF., Daeschiel, M.A. and Fleming, H.P. (1987) Applied
and Environmental Microbiology, 53, 1382.
McKay, L.L. (1983) Antonie van Leeuwenhoek, 49, 259.
McMurrough,1. and Palmer, V. (1979) Journal of the Institute of Brewing, 85, 11.
Montel, M.e., Talon, R, Fournoud, J. and Champomier, M.e. (1991) Journal of
Applied Bacteriology, 70, 469.
Munekata, M. (1973) Bulletin of Brewing Science, 19, 27.
Nakagawa, A. and Kitahara, K (1959) Journal of General and Applied Microbiology,
5,95.
Nishikawa, N. and Kohgo, M. (1985) Technical Quarterly of the Master Brewers
Association of the Americas, 22,61.
Nishikawa, N., Kohgo, M. and Karakawa, T. (1979) Bulletin of Brewing Science, 25,
13,17,23.
Orla-Jensen, S. (1919) Lactic Acid Bacteria, Andre Fred Host, Copenhagen.
Pfenninger, H.B., Schur, F., Anderegg, P. and Schlienger, E. (1979) Proceedings of the
17th Congress of the European Brewery Convention, Berlin, IRL Press, London,
p.491.
Priest, F.G. (1981a) In The Aerobic Endospore10rming Bacteria: Classification and
Identification (eds Re.W. Berkeley and M. Goodfellow), Academic Press, Lon-
don,p.33.
Priest, F.G. (1981b) Journal of the Institute of Brewing, 87, 279.
160 Gram-positive brewery bacteria
Priest, F.G. and Austin, B. (1993) Modern Bacterial Taxonomy, 2nd edn, Chapman &
Hall, London.
Priest, F.G., and Barbour, E.A (1985) In Computer-Assisted Bacterial Systematics (eds
M. Goodfellow, D. Jones and F.G. Priest), Academic Press, London, p. 137.
Priest, F.G. and Pleasants, J.G. (1988) Journal of Applied Bacteriology, 64, 379.
Priest, F.G. and Williams, S.T. (1993) In Handbook of New Bacterial Systematics (eds
M. Goodfellow and AG. O'Donnell), Academic Press, London, p. 361.
Raccach, M. (1987) CRC Critical Review of Microbiology, 14, 291.
Rainbow, C (1981) In Brewing Science, Vol. 2 (ed. J.RA Pollock), Academic Press,
London, p. 491.
Rinck, M. and Wackerbauer, K (1987) Monatsschrift flir Brauwissenschaft, 40, 324.
Rogosa, M. (1974) In Bergey's Manual of Determinative Bacteriology, 8th edn (eds
RE. Buchanan and N.E. Gibbons), Williams and Wilkins, Baltimore, p. 576.
Saha, RB., Sonda, RJ. and Middlekauf, J.E. (1974) Proceedings of the American Society
of Brewing Chemists, 32, 9.
Sanni, AI. and Oso, B.A. (1988) Nahrung, 32, 319.
Seyfried, P.L. (1968) Canadian Journal of Microbiology, 14, 313.
Sharpe, M.E. (1979) In Identification Methods for Microbiologists, 2nd edn (eds
F.A Skinner and D.W. Lovelock), Academic Press, London, p. 233.
Sharpe, M.E., Fryer, T.F. and Smith, D.G. (1966) In Identification Methods for
Microbiologists Part A (eds B.M. Gibbs and F.A Skinner), Academic Press,
London, p. 65.
Shaw, B.G. and Harding, CD. (1984) Journal of Applied Bacteriology, 56,25.
Shimwell, J.L. (1941) Wallerstein Laboratories Communications, 4, 41.
Shimwell, J.L. (1945) Wallerstein Laboratories Communications, 8,23.
Shimwell, J.L. and Kirkpatrick, W.F. (1939) Journal of the Institute of Brewing, 45, 137.
Skerman, V.B.D., McGowan, V. and Sneath, P.H.A (1980) International Journal of
Systematic Bacteriology, 30, 225.
Smith, N.A. and Smith, P. (1992) Journal of the Institute of Brewing, 98, 409.
Smith, N.A. and Smith, P. (1993) Journal of the Institute of Brewing, 99,43.
Smid, E.J., Poolman, B. and Konings, W.N. (1991) Applied and Environmental
Microbiology, 57,2447.
Sneath, P.H.A (1978) In Essays in Microbiology (eds J.R Norris and M.H. Richmond),
John Wiley, London, p. 10/1.
Sneath, P.H.A and Sokal, RR (1973) Numerical Taxonomy, W.H. Freeman, San
Francisco.
Solberg, O. and Claussen, O.G. (1973a) Journal of the Institute of Brewing, 79, 227.
Solberg, O. and Claussen, O.G. (1973b) Journal of the Institute of Brewing, 79, 231.
Speckman, RA and Collins, E.B. (1968) Journal of Bacteriology, 95,174.
Speckman, RA and Collins, E.B. (1973) Applied Microbiology, 26, 744.
Taguchi, H., Ohkochi, M., Vehara, H. et al. (1990) Journal of the American Society of
Brewing Chemists, 48, 72.
Van Keer, C, Van Melkebeke, L., Vertriest, W. et al. (1983) Journal of the Institute of
Brewing, 89, 361.
Van Vuuren, H.J.J., Kersters, K, De Ley, J. et al. (1978) Journal of the Institute of
Brewing, 84, 315.
Van Vuuren, H.J.J., Kersters, K, De Ley, J. and Toerien, D.F. (1981) Journal ofApplied
Bacteriology, 51, 51.
Verzele, M. (1986) Journal of the Institute of Brewing, 92,32.
Wackerbauer, K and Emeis, CC (1967) Monatsschrift flir Brauerei, 20, 160.
Wackerbauer, K and Emeis, CC (1968) Monatsschrift flir Brauerei, 22,3.
Wackerbauer, K and Rinck, M. (1983) Monatsschrift flir Brauwissenschaft, 36, 392.
Weiss, N. (1992) In The Prokaryotes, 2nd edn (eds A. Balows, H.G. Triiper,
M. Dworkin et al.), Springer-Verlag, New York, p. 1502.
 References 161
White, F.H. and Kidney, E. (1979) Proceedings of the 17th Congress of the European
Brewery Convention, Berlin, IRL Press, Oxford, p. 80l.
Whiting, M., Crichlow, M., Ingledew, W.M. and Ziola, B. (1992) Applied and
Environmental Microbiology, 58, 713.
Whittenbury, R. (1978) In Streptococci (eds F.A. Skinner and L.B. Quesnel), Academic
Press, London, p. 5l.
Wilkinson, B.J. and Jones, D. (1977) Journal of General Microbiology, 98,399.
Woese, c.R. (1987) Microbiological Reviews, 51, 22l.
Young, T.W. and Talbot, N.P. (1979) Proceedings of the 17th Congress of the European
Brewery Convention, Berlin, IRL Press, Oxford, p. 817.
 CHAPTER 6

Gram-negative spoilage
bacteria*
H.I.J. Van Vuuren

6.1 INTRODUCTION

Brewing bacteriology was born when microorganisms responsible for the

spoilage of beer were investigated by Louis Pasteur in his classic nineteenth
century study. He was called upon to determine why French beer was
inferior to German beer. He isolated a number of bacterial contaminants
from French beer and malt wort and in 1876 published his famous book
Etudes sur la Biere, ses Maladies, Causes qui les Provoquent. Procedes pour la
Rendre Inalterable, avec une Theorie Nouvelle de la Fermentation (Studies of
Beer, its Diseases and the Causes That Provoke Them. Procedures for
Making it Unalterable, with a New Theory of Fermentation). Today, it is
generally believed that the presence of bacteria in the beer brewing process
is undesirable.
The Gram stain was developed in 1884 by the Danish physician
Christian Gram. Many modifications and improvements have subse-
quently been made (Society of American Bacteriologists, 1957). The Gram
reaction is governed by certain structural properties of the bacterial cell
wall. The distinction between Gram-positive and Gram-negative organ-
isms is thus of prime importance in the taxonomy of eubacteria. Both
Gram-positive and Gram-negative bacteria are involved in beer spoilage,
but this chapter is confined to the latter.
A diagnostic key for the identification of beer spoilage bacteria was
devised by the eminent brewery microbiologist Shimwell (1947). The key
was simplified by Ault (1965), who restricted it to bacteria likely to be
encountered in the beer brewing process. Van Vuuren (1976) modified
Ault's key further by only considering beer spoilers. A number of

'Manuscript submitted December 1991

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and I. Campbell.
Published in 1996 by Chapman & Hall, London. ISBN 0412591502
164 Gram-negative spoilage bacteria

Gram-negative bacteria

Growth on solid media Growth on solid media

(Aerobic conditions) (Only anaerobic conditions)

~
No acetic acid Acetic acid
~
Cocci Rods
from ethanol from ethanol Megasphaera Pectinatus

AC:~:=%r Selenomonas
Zymophilus

No growth in Growth in
Zymomonas Zymomonas
selective medium selective medium
Enterobacteriaceae Zymomonas

Fig. 6.1 Diagnostic scheme for the identification of Gram-negative beer spoilage
bacteria.

diagnostic keys have subsequently been published (Back, 1980; Ingledew,

Sivaswamy and Burton, 1980; Casey and Ingledew, 1981; Harper, 1981;
Campbell, 1983). The simple key shown in Fig. 6.1 can be used to identify
Gram-negative contaminants in beer, including the recently isolated genera
Selenomonas and Zymophilus.
Enterobacteriaceae isolated from beer breweries can be identified further
by the API 20E system (Ingledew, Sivaswamy and Burton, 1980; Van
Vuuren et ai., 1981).
The occurrence of Gram-negative bacteria in the beer brewing process
is generally regarded as being undesirable. Important Gram-negative spoil-
age bacteria considered in this chapter include acetic acid bacteria, certain
members of the family Enterobacteriaceae, Zymomonas, Pectinatus
cerevisiiphilus, Pectinatus jrisingensis, Selenomonas lacticifex, Zymophilus
raffinosivorans, Zymophilus paucivorans and Megasphaera. The occurrence of
a few non-fermentative bacteria occasionally found in beer is also
discussed.
This brings me to the term infection, often used incorrectly by brewery
microbiologists. Bacteria infect humans, animals and plants but they con-
taminate substrates such as beer, other beverages and food. They are
therefore contaminants of beer and the term contamination should be used
instead of infection.
 Acetic acid bacteria 165
6.2 ACETIC ACID BACTERIA

The acetic acid bacteria comprise an industrially important group of Gram-

negative, rod-shaped bacteria capable of converting ethanol to acetic acid.
They spoil food, soft drinks and alcoholic beverages such as beer, wine and
cider (for a review see Swings and De Ley, 1981). The acetic acid bacteria
are subdivided into the genera Acetobacter and Gluconobacter (Acetomonas).
In the eighth edition of Bergey's Manual of Determinative Bacteriology
(Buchanan and Gibbons, 1974), Gluconobacter was included in the family
Pseudomonadaceae, whereas Acetobacter was regarded as a genus of uncer-
tain affiliation among the aerobic Gram-negative rods. However, recent
taxonomic studies have shown that Acetobacter and Gluconobacter are
closely related and share little affinity with the pseudomonads. They are
therefore placed in the new family Acetobacteriaceae (Gillis and De Ley,
1980).

6.2.1 Acetobacter

(a) General characteristics

Acetobacter cells are Gram negative or Gram variable, ellipsoidal to rod-
shaped, straight or slightly curved, 0.6-0.8 11m x 1.0--4.0 11m, occurring sin-
gly, in pairs or chains. Pleomorphic forms occur which may be spherical,
elongated, swollen, club shaped, curved or filamentous. They are non-
motile or motile; if motile, peritrichous or lateral flagella are present.
Acetobacter spp. are obligately aerobic with a respiratory metabolism (02 is
the terminal electron acceptor), catalase positive, oxidase negative, and
oxidize ethanol to acetic acid and acetate to CO2 and H 20. They are often
referred to as the' overoxidizers'. In liquid media, Acetobacter forms a ring,
film or pellicle; uniform turbidity of the medium and a cell deposit is
sometimes observed (De Ley, Swings and Gossele, 1984).

(b) Taxonomy
There have been extensive taxonomic studies on the genus Acetobacter. The
eighth edition of Bergey's Manual of Determinative Bacteriology (Buchanan
and Gibbons, 1974) contained three species with nine subspecies. At pres-
ent, only Acetobacter aceti, Acetobacter liquefaciens, Acetobacter pasteurianus
and Acetobacter hansenii are recognized (De Ley, Swings and Gossele,
1984b). Features differentiating these species are given in Table 6.1.

(c) Metabolism
Acetobacter spp. possess a respiratory metabolism (De Ley, 1961; Leisinger,
1965) and, like Gluconobacter, directly oxidize sugars, alcohols and steroids
166 Gram-negative spoilage bacteria
Table 6.1 Characteristics differentiating species of the genus Acetobacter (data from
Bergey's Manual of Systematic Bacteriology (Krieg and Holt, 1984»
Characteristics A. aceti A. liquefaciens A. pasteurianus A. hansenii
Formation of:
water-soluble brown
pigments on GYC* +
y-pyrones from
D-glucose d
y-pyrones from
D-fructose +
5-ketogluconic acid
from D-glucose + d d
2,5-diketogluconic
acid from D-
glucose +
Ketogenesis from
glycerol + + +
Growth on carbon
sources:
ethanol + + d
dulcitol d
Na acetate + d d
Growth on L-amino
acids in the presence
of D-mannitol as
carbon source:
L-glycine, L-threonine,
L-tryptophan d
L-asparagine,
L-glutamine d + +
Growth in the
presence of 10%
ethanol d
G +C mol% (Tm) 55.9-59.5 62.3-64.6 52.8-62.5 58.1-62.6
* GYC: 5% o-glucose + 1% yeast extract + 3% CaC03 + 2·5% agar.
-,0-15% strains positive; d, 16--84% strains positive; +, 85-100% strains positive.

(for reviews see De Ley and Kersters, 1964; Asai, 1968). For example,
glucose is oxidized to gluconic and 2- and 5-oxogluconic acids, glycerol to
dihydroxyacetone and sorbitol to sorbose (for a review see Rainbow, 1966).
They also metabolize sugars via the hexose monophosphate pathway and
the tricarboxylic acid cycle (Asai, 1968; Rainbow, 1981). In certain
A. pasteurianus strains, the Entner-Doudoroff pathway appears to be more
active than the hexose monophosphate shunt (White and Wang, 1964a,b).
Strains which utilize ethanol as sole carbon source, and ammonium sul-
phate as the only source of nitrogen, utilize enzymes of the glyoxylate
by-pass (Stouthamer, Van Boom and Bastiaanse, 1963; Leisinger, 1965).
Acetobacter grows well on media containing ethanol, glycerol and
 Acetic acid bacteria 167
sodium DL-lactate as sole carbon source. Several strains are able to grow
using NH/ and ethanol as sole nitrogen and carbon sources respectively.
Only a few strains are able to use single amino acids as sole nitrogen source.
Growth factor requirements depend on the carbon source supplied. Some
A. pasteurianus and A. hansenii strains produce extracellular cellulose.

6.2.2 Gluconobacter

(a) General characteristics

The morphology of Gluconobacter is very similar to that of Acetobacter.
However, motile strains have between three and eight polar flagella; a
single flagellum is rarely observed. They are obligately aerobic, catalase
positive, oxidase negative, do not reduce nitrate to nitrite but reduce
ethanol to acetic acid. Furthermore, Gluconobacter does not oxidize acetate
or lactate to CO 2 and H 20. A strong ketogenesis occurs and acid formation
from D-glucose and D-xylose is pronounced (pH::; 4.5). Some strains pro-
duce a pink, non-diffusible pigment whereas others may produce a soluble,
dark brown pigment and y-pyrone. The pathway for y-pyrone formation
has been elucidated (Asai, 1968) and it is believed that the production of
brown pigments is related to y-pyrone synthesis (Rainbow, 1981). Acetic
acid formed in Frateur's (1950) ethanol medium dissolves the calcium
carbonate, which is not re-deposited afterwards. This method, as well as
that of Carr (1968), can be used to distinguish between Acetobacter and
Gluconobacter. In beer, wort or glucose-containing liquid media, a pellicle
or film forms at the surface. Some strains produce viscous growth in beer,
due to formation of dextrans and levans (De Ley and Swings, 1984).
Gluconobacter oxydans is the most important species of this genus for the
brewer.

(b) Metabolism
Gluconobacter possesses a respiratory metabolism with oxygen as the ter-
minal electron acceptor. The hexose monophosphate shunt constitutes the
most important route for the metabolism of sugars to CO2 and H 20. The
complete glycolytic and tricarboxylic acid cycles do not function in
Gluconobacter (De Ley, 1961; Williams and Rainbow, 1964; Greenfield and
Claus, 1972). Enzymes of the Entner-Doudoroff pathway are present
(Leisinger, 1965; Kersters and De Ley, 1968). GI uconobacter is able to oxidize
sugars, steroids and aliphatic and cyclic alcohols. Gluconobacter utilizes a
number of sugars and alcohols as sole source of carbon and good growth
is obtained on D-mannitol, sorbitol, glycerol, D-fructose and D-glucose.
Ammonium ions as well as single amino acids are used as sole nitrogen
source by many Gluconobacter strains. A detailed description of Acetobacter
and Gluconobacter spp. is given in Krieg and Holt (1984).
168 Gram-negative spoilage bacteria
6.2.3 Beer spoilage
Acetification of beer was studied by pioneers of microbiology such as
Persoon (see Rainbow, 1981), Pasteur, Hansen, Henneberg and Beijerinck
(see Swings and De Ley, 1981). Acetobacter and Gluconobacter spp. have been
isolated from breweries (Van Vuuren, 1976; Ploss, Erber and Eschenbecher,
1979; Hough et al., 1982). Acetomonads are resistant to the bacteriostatic
activity of hops, acid and ethanol (Masschelein, 1973; Hough et al., 1982)
and are therefore capable of growing and spoiling beer. Being strict aerobes,
they should not grow in wort or beer once anaerobic conditions develop.
However, beer strains probably grow under microaerophilic conditions
and have been isolated from beer with a low oxygen content (Harper, 1980).
Laboratory strains of acetic acid bacteria fail to grow on media incubated
under carbon dioxide but comparable strains isolated from public houses
grow successfully (Harper, 1980).
Acetic acid bacteria are particularly prevalent in public houses and
draught beer is rapidly acetified if air is allowed to enter casks. These
bacteria spoil beer by causing acid, off-flavours, turbidity and ropiness
(Shimwell, 1936; Cosbie, Tosic and Walker, 1942, 1943; Tosic and Walker,
1944; Walker and Tosic, 1945; Shimwell, 1948a; Masschelein, 1973). The
production of off-flavours without acetification is probably caused by the
oxidation of polyalcohols, such as glycerol, to dihydroxyacetone. Ropiness
is often observed as a pellicle or greasy-looking covering on the surface of
beer. The extent of beer spoilage by acetic acid bacteria is strain dependent
and the alcohol content of beer also affects spoilage. Comrie (1939) and
Shimwell (1936) found that the bacteria were completely inhibited by
alcohol concentrations higher than 6%. However, some strains grow in
media containing 10% ethanol (De Ley, Swings and Gossele, 1984).
Acetobacter sp. BS05 and Gluconobacter oxydans subsp. oxydans strain NCIB
9013 are able to survive 12-13% v Iv ethanol produced in high gravity
brews (Magnus, Ingledew and Casey, 1986). Gilliland and Lacey (1966)
showed that an Acetobacter sp. produced a compound toxic to yeasts.
Ploss, Erber and Eschenbecher (1979) failed to find acetic acid bacteria
on barley and hops but a few malt, wort, yeast and air samples were
contaminated. Storage vessels, pressure equilibration, stability and filtra-
tion samples were highly contaminated. However, the low incidence of
acetomonads in air samples is rather surprising since they are ubiquitous
in breweries and beer exposed to air can be rapidly acetified. Acetomonads
occur on many plant species and contamination from plant materials and
air should not be ignored (lngledew, 1979; Hough et al., 1982). Acetobacter
thrives in alcohol-enriched niches whereas Gluconobacter prefers sugars as
carbon source. Acetomonads are resistant to hop antiseptics and acid
conditions and are ethanol tolerant. They can therefore grow and spoil beer
in the presence of oxygen.
 Enterobacteriaceae 169
6.3 ENTEROBACTERIACEAE

6.3.1 General characteristics

The family Enterobacteriaceae comprises a large number of bacterial
species isolated from different ecological niches. Some species are free
living whereas others are pathogenic to humans, animals, insects and
plants. The following genera are included in the family: Escherichia, Shigella,
Enterobacter, Klebsiella, Citrobacter, Salmonella, Erwinia, Kluyvera, Serratia,
Cedecea, Morganella, Hafnia, Edwardsiella, Providencia, Proteus, Yersinia,
Obesumbacterium, Xenorhabdus, Rahnella and Tatumella (Brenner, 1984).
Enterobacteriaceae are Gram-negative, non-acid-fast, straight rods,
0.3-1.0 11m x 1.0-6.0 11m. Motile as well as non-motile species occur. Motile
strains have peritrichous flagella. They do not form endospores or
microcysts, are facultatively anaerobic, resistant to bile salts and capable of
growing in a mineral salts medium containing glucose as sole carbon and
energy source. However, some require vitamins or amino acids or both.
Acid and often detectable gas are produced during fermentation of
o-glucose. All species important in brewing are catalase positive and
oxidase negative. Nitrate is reduced to nitrite except by some Erwinia
strains and Yersinia, and this has implications in the synthesis of
N-nitrosamines in beer (Smith, 1994). The G + C content of the DNA of
members of the family is 38-60 mol%.

6.3.2 Metabolism
Enterobacteria ferment o-glucose via the Embden-Meyerhof-Parnas and
hexose monophosphate pathways to yield varying amounts of formic acid
(or CO2 and H 2), acetic acid, lactic acid, succinic acid, ethanol, acetoin and
2,3-butanediol (Fig. 6.2). The concentration of these metabolites depends to
a large extent on the bacterial species and is also affected by growth
conditions (Van Vuuren, 1978; Tracey, 1984). Two main fermentation types
occur. In the mixed acid fermentation typified by Escherichia coli, ethanol
and acids are formed with little or no acetoin and 2,3-butanediol. In
contrast, relatively large amounts of acetoin and 2,3-butanediol but smaller
amounts of acids are produced by Klebsiella spp. via the 2,3-butanediol
pathway. The methyl red and Voges-Proskauer tests are often used as an
indication of the mixed acid and 2,3-butanediol fermentations respectively.
The results obtained are affected by the ratio of end-products produced and
care should be taken when the tests are used for identification purposes.
Gas chromatographic analyses of metabolic end-products of 60 entero-
bacterial contaminants from beer showed that Citrobacter freundii, Proteus
mirabilis and Klebsiella oxytoca ferment o-glucose via the mixed acid fermen-
tation pathway. Rahnella aquatilis (previously Enterobacter agglomerans),
Enterobacter cloacae, Klebsiella pneumoniae, Serratia marcescens and Hafnia alvei
170 Gram-negative spoilage bacteria
glucose

1
oxaloacetate _--=+COa",,-_ PEP

7- 1
M
•
+ pyruvate

~
succinate lactate oo-acetolactate

acetyl-S-CoA formate
I
~ acetoIn

. . . . 1~~. .
I
I
) .......... +2"

1~" ,
;'
;' .....
.....
/
;'

acetate ethanol Co, H, 2.3-butanedlOl

Fig. 6.2 Metabolism of glucose by Enterobacteriaceae.

produced large but varying amounts of 2,3-butanediol (Van Vuuren, 1978).

Although H. alvei is regarded to ferment D-glucose via the 2,3-butanediol
pathway, Tracey (1984) found that the organism is metabolically diverse;
some H. alvei strains ferment D-glucose via the mixed acid fermentation
pathway but others produce large amounts of 2,3-butanediol.
Obesumbacterium proteus has the mixed acid fermentation pathway.
In the eighth edition of Bergey's Manual of Determinative Bacteriology
(Buchanan and Gibbons, 1974), the family Enterobacteriaceae was divided
into tribes mainly on the basis of D-glucose fermentation via the mixed acid
or 2,3-butanediol pathways. The tribe concept has been abandoned in
Bergey's Manual of Systematic Bacteriology (Brenner, 1984) as it was felt that
'the use of tribes is of no diagnostic significance and of questionable
taxonomic significance'.

6.3.3 Beer spoilage

Enterobacterial contaminants in beer breweries have been referred to as
termobacteria (Lindner, 1895), wort bacteria (Ault, 1965; Cowbourne, Priest
and Hough, 1972; Makinen, Tanner and Haikara, 1981) and as coli-aerogenes
or coliform bacteria (Ingledew, 1979; Keevil, Hough and Cole, 1979; McCaig
and Morrison, 1984). The term termobacteria has no scientific standing;
wort bacteria include non-enterobacterial contaminants whereas coli-
aerogenes or coliform bacteria exclude important beer spoilers such as
 Enterobacteriaceae 171
O. proteus. Enterobacteriaceae, enterobacterial contaminants or
enterobacteria are the correct terms to describe these contaminants.
Lindner in 1895 was probably the first to report contamination and
spoilage of beer by Enterobacteriaceae. Several enterobacterial species such
as Enterobacter aerogenes, E. cloacae, Rahnella aquatilis, H. alvei, O. proteus,
C. freundii, K. pneumoniae, K. oxytoca, Serratia strains and P. mirabilis have
subsequently been isolated from beer breweries (Thorwest, 1965; Bernstein,
Blenkinship and Brenner, 1968; Von Riemann and Scheible, 1969; Prucha
and Scheible, 1970; Anderson, Howard and Hough, 1971; Niefind and
Spilth, 1971; Cowbourne, Priest and Hough, 1972; Priest and Hough, 1974;
Eschenbecher and Ellenrieder, 1975; Van Vuuren and Toerien, 1981; Van
Vuuren et al., 1981; McCaig and Morrison, 1984; Hamze et al., 1991).
Pathogenic species occurring in beer have not been described. Ault
(1965), Shimwell (1948b) and Seidel (1980) considered members of the
family Enterobacteriaceae of little importance to the brewing industry. It is
now well known that enterobacterial species retard or accelerate the fer-
mentation process and significantly influence the flavour and aroma of the
final product. It is generally believed that apart from O. proteus and
R. aquatilis, other enterobacterial contaminants do not survive the adverse
conditions during fermentation (Priest, Cowbourne and Hough, 1974;
Hamze et al., 1991). However, bacteria may exist in a 'non-recoverable'
stage (Xu et al., 1982), i.e. where no growth is observed on standard
laboratory media. They are thus viable and exist in a 'dormant' stage.
'Non-recoverable' bacterial contaminants could have serious implications
for the brewing industry and question the often-held view that many
enterobacteria do not survive the brewing process. When present in re-
cycled pitching yeast, they will not be detected by standard culture media
and methods; contamination and spoilage of beer might therefore occur
when they develop under favourable conditions in a new fermentation.
Investigations should be undertaken to determine the efficacy of media and
methods to detect possible 'non-recoverable' bacterial contaminants. En-
riched or more selective culture media and prolonged incubation should
be considered.

(a) Obesumbacterium proteus

O. proteus is the best known enterobacterial contaminant. This bacterium
was probably first described by Lindner as 'Termobacterium lutescens'
(Priest, 1974). Shimwell (1936) named it 'Bacterium Y', Shimwell and
Grimes (1936) renamed it 'Flavobacterium proteum' (later 'F. proteus') and
finally Shimwell (1963, 1964) created a new genus Obesumbacterium, with a
single species O. proteus. The extensive studies of Priest et al. (1973) and
Priest, Cowbourne and Hough (1974) indicated that O. proteus is closely
related to Hafnia and thus a member of the family Enterobacteriaceae. They
proposed that the genus Obesumbacterium be discontinued and that
172 Gram-negative spoilage bacteria
O. proteus be transferred to Hafnia as 'Hafnia protea'. Furthermore, they
recognized two genetic groups of 'H. protea'. This has been confirmed by
Brenner (1981) and Prest, Hammond and Stewart (1994). However,
O. proteus rather than 'R. protea' was included in the Approved Lists of
Bacterial Names (Skerman, McGowan and Sneath, 1980) and subsequently
in Bergey's Manual of Systematic Bacteriology (Brenner, 1984).
O. proteus has thus far only been isolated from the brewery environ-
ment. It is often found in pitching yeast and referred to as the 'short fat
rod of pitching yeast'. O. proteus is highly pleomorphic, grows well in
hopped or unhopped wort and tolerates ethanol concentrations of up to
6% v Iv. However, the inhibitory concentration of ethanol appears to be
pH dependent. O. proteus can grow in unhopped wort with pH ranging
from 4.4 to 9.0. Brown (1916) considered it a harmless yeast contaminant
whereas Strandskov (1965) concluded that it could spoil beer only when
present in pitching yeast at a concentration exceeding 1.5% of the yeast
count. However, O. proteus is an important beer spoilage organism; it is
able to grow alongside yeast in a brewery fermentation, retard the fer-
mentation process and result in beer with a high final specific gravity and
pH (Case, 1965; Priest, Cowboume and Hough, 1974). O. proteus is also
responsible for increased levels of dimethyl sulphide, dimethyl disulph-
ide, n-propanol, isobutanol, isopentanol, 2,3-butanediol and diacetyl
(Cowboume, Priest and Hough, 1972; Priest, Cowboume and Hough,
1974). The final concentration of these compounds is often strain
dependent. Beer produced with yeast contaminated with O. proteus has a
parsnip-like or fruity odour and flavour. Pitching yeast is the main source
of contamination in breweries.

(b) Rahnella aquatilis

Enterobacter agglomerans was first recognized as an important contaminant
in beer breweries by Van Vuuren et al. (1978). This finding was confirmed
by Makinen, Tanner and Haikara (1981) and McCaig and Morrison (1984).
However, at that stage E. agglomerans comprised a large and heterogeneous
group of bacteria and its nomenclature and classification was in a state of
flux (Van Vuuren, 1987). The 18 aerogenic strains of E. agglomerans isolated
from breweries (Van Vuuren et al., 1981) have been reclassified as strains
of Rahnella aquatilis (Hamze et al., 1991). The taxonomic position
E. agglomerans strains studied by Makinen, Tanner and Haikara (1981) and
McCaig and Morrison (1984) have not been clarified. For the purpose of this
review all strains of E. agglomerans previously isolated from breweries will
be regarded as strains of R. aquatilis.
R. aquatilis cells are small rod-shaped cells, 0.5-D.7 ~m x 2.0-3.0 ~m.
They are motile when grown at 25°C but motility is lost at 37°C. D-Glucose
is fermented with the production of acid and most strains form gas. Lactose,
maltose, L-rhamnose, raffinose and salicin are fermented. R. aquatilis is
 Enterobacteriaceae 173
negative for lysine and ornithine decarboxylases and for arginine
dihydrolase. All strains are Voges-Proskauer positive and most strains are
methyl red positive. Strains are weakly positive for phenylalanine deami-
nase. The mol% G + C of the DNA is 51-56 (determined by melting tem-
perature, T m) (Izard et ai., 1979, 1981; Brenner, 1984). R. aquatilis occurs
mainly in fresh water but has occasionally been isolated from human
clinical specimens.
R. aquatilis grows well in hopped or unhopped wort, in the presence or
absence of yeast. It survives the brewing process when normal gravity wort
is used and, like o. proteus, accumulates in recycled pitching yeast (Van
Vuuren, Cosser and Prior, 1980; Magnus, Ingledew and Casey, 1986). In
high-gravity brews, however, R. aquatilis is killed by ethanol concentrations
of 11-12% v Iv (Magnus, Ingledew and Casey, 1986).
In ale fermentations, O. proteus rises to the surface of the wort with the
top yeasts and remains viable for 60-70 h; other enterobacteria remain
suspended in the wort and are killed by adverse conditions, i.e. a low pH
and a high ethanol concentration (Priest, Cowboume and Hough, 1974).
During lager beer fermentations, however, bacterial contaminants such as
R. aquatilis settle or flocculate with the yeast. Bacteria accumulated in
harvested pitching yeast are a serious source of contamination. It also
becomes impossible to obtain accurate counts of bacteria during
fermentation.
R. aquatilis greatly affects the brewing process as well as beer flavour
and aroma when present in fermenting wort (Van Vuuren, Cosser and
Prior, 1980; McCaig and Morrison, 1984). According to Cowboume, Priest
and Hough (1972), enterobacterial contaminants retard the fermentation
process; this effect might, however, be yeast strain dependent. The pres-
ence of R. aquatilis during lager beer fermentations increases the rate of
fermentation during the initial stages but the final specific gravity of
contaminated beer is higher than that of control beer (Van Vuuren, Cosser
and Prior, 1980). In ale brewing, however, the specific gravity of the final
beer is lower when fermentations are contaminated by R. aquatilis (McCaig
and Morrison, 1984). The final pH of fermenting wort contaminated with
R. aquatilis is higher than usual, probably due to neutral metabolites
produced during fermentation. R. aquatilis ferments D-glucose via the
2,3-butanediol fermentation pathway (Van Vuuren, Cosser and Prior,
1980). However, in ale fermentations the final pH of the contaminated
beer is reportedly similar to that of control beer (McCaig and Morrison,
1984).
Abnormally high levels of diacetyl (0.7 mg r!) and dimethyl sulphide
(142.8 Ilg r!) were detected in beer produced from wort contaminated with
106 R. aquatilis cells mr!. The concentration of these compounds in the
uncontaminated control beer was considerably lower (0.04 mg r! and
54.6 Ilg r! respectively). The production of these compounds was affected
by the level of contamination. Significant increases were only found with
174 Gram-negative spoilage bacteria
a contaminant population of 106 cells ml-!; a slight increase in the concen-
tration of both compounds was observed at 103 cells mr! (Van Vuuren,
Cosser and Prior, 1980). The amount of dimethyl sulphide produced by
R. aquatilis is strain dependent (McCaig and Morrison, 1984). The produc-
tion of excess vicinal dike tones by R. aquatilis in beer was confirmed by
McCaig and Morrison (1984). It has been suggested that other entero-
bacteria are indirectly responsible for synthesis of diacetyl by producing
higher levels of <F128Ma-acetolactate (Priest and Hough, 1974), a heat-la-
bile precursor of diacetyl (Inoue et al., 1968; Suomalainen and Ronkainen,
1968). The mechanism by which R. aquatilis produces diacetyl during
fermentation has not yet been elucidated. R. aquatilis contamination also
leads to increased levels of acetaldehyde and methyl acetate; however, the
concentration of ethyl acetate, n-propanol, isobutanol, isoamyl acetate and
isoamyl alcohol is lower (Van Vuuren, Cosser and Prior, 1980). Beer
contaminated with R. aquatilis is unpalatable, with a fruity, milky and
sulphury taste and aroma. Barley, malt and hops as well as recycled
pitching yeast could well be the source of contamination (Van Vuuren,
Cosser and Prior, 1980).

(c) Citrobacter freundii

C. freundii has been reported as an occasional contaminant of pitched wort
(Cowboume, Priest and Hough, 1972; Priest, Cowboume and Hough, 1974;
Eschenbecher and Ellenrieder, 1975; Van Vuuren et al., 1981). This bac-
terium is closely related to E. coli. The cells are straight rods, -1.0 11m in
diameter and 2.<.H,.0 11m long. They occur singly and in pairs and are
usually motile by peritrichous flagella. C. freundii is facultatively anaerobic,
oxidase negative and catalase positive. Citrate can be used as sole carbon
source; however, citrate negative strains have been found in beer breweries
(Van Vuuren et al., 1981). D-Glucose is fermented via the mixed acid
fermentation pathway and gas is produced. The G + C content of the DNA
is 50-51 mol%. C. freundii has been found in humans, animals, soil, water,
sewage, clinical specimens and wound swabs. It is regarded as an
opportunistic or secondary pathogen (Krieg and Holt, 1984).
Keevil, Hough and Cole (1979) studied the effect of C. freundii on beer
fermentation and flavour. The organism reduces yeast viability and affects
the redox potential of fermenting wort. Contamination of fermenting wort
with C.freundii leads to faster fermentations. Similar results were obtained
with R. aquatilis (Van Vuuren, Cosser and Prior, 1980). The presence of
C.freundii during fermentation greatly increases concentrations of lactate,
pyruvate, succinate, isocitrate and dimethyl sulphide in the final beer. Even
a low number of C. freundii cells, viable for only a restricted period during
fermentation, may spoil beer (Keevil, Hough and Cole, 1979). C. freundii is,
however, sensitive to high concentrations of ethanol and does not survive
the brewing process (Magnus, Ingledew and Casey, 1986).
 Enterobacteriaceae 175
(d) Klebsiella
Klebsiella spp. have been isolated from different stages of the brewing
process and are associated with beer spoilage (Anderson, Howard and
Hough, 1971; Masschelein, 1973). Klebsiella cells are straight rods,
0.3-1.0 11m x 0.6-6.0 ~m, occurring singly, in pairs or short chains. They are
non-motile, capsulated, facultatively anaerobic and oxidase negative.
D-Glucose is fermented by brewery isolates via the mixed acid or
2,3-butanediol fermentation pathways (Van Vuuren, 1978). Acid and gas
are produced but anaerogenic strains occur. The G + C content of the DNA
is 53-58 mol%. Klebsiella spp. have also been isolated from intestinal
contents, clinical specimens, soil, water and plants (Krieg and Holt, 1984).
K. pneumoniae is a pathogen of humans and causes inflammation of the
lungs. It is most unlikely that this pathogen occurs in samples from beer
breweries. A new species, Klebsiella terrigena, was recently described (Izard
et al., 1981). It occurs mainly in aquatic and soil environments and can be
distinguished from K. pneumoniae by its ability to grow at lODC and inability
to produce gas from lactose at 44.soC. Its ability to ferment melezitose
distinguishes K. terrigena from Klebsiella planticola (Brenner, 1984).
K. pneumoniae strains isolated from beer breweries conform to criteria laid
down for the new species (Van Vuuren et al., 1981). Therefore, K. terrigena,
rather than K. pneumoniae, occurs as a contaminant in beer breweries. This
information will certainly be received with relief by brewers.
Beer produced from wort contaminated with an indole-negative
Klebsiella strain (possibly K. terrigena) had a typical phenolic off-flavour.
Only indole-negative enterobacteria produce phenolic off-flavours in beer
(West, Lautenbach and Brumsted, 1963; Von Riemann and Scheible, 1969;
Prucha and Scheible, 1970). Cowbourne, Priest and Hough (1972) suggested
that bacteria produce these off-flavours by deaminating tyrosine (Fig. 6.3).
Abnormally high concentrations of 4-ethylguaiacol have been detected
in beer having a phenolic off-flavour (Halcrow et al., 1966). Klebsiella spp.
decarboxylate ferulic acid, normally present in beer, to 4-vinylguaiacol
(Fig. 6.4). According to Finkle et al. (see Halcrow et al., 1966) hydrogenation
of the vinyl group occurs as a side reaction. Both 4-vinylguaiacol (Dadic,
van Gheluwe and Valyi, 1971) and 4-ethylguaiacol (Halcrow et al., 1966)
impart a phenolic taste to beer.
Indole-positive K. pneumoniae strains have been isolated from beer brew-
eries (Van Vuuren, 1978). Indole-positive K. pneumoniae strains have been
reclassified as K. oxytoca (Brenner, 1984) and strains of this species have
been isolated from fermenting wort in two different breweries (Van Vuuren
and Toerien, 1981). The organism is now considered a contaminant of beer
but it is not known how it affects the fermentation process, flavour or
aroma. K. oxytoca is sensitive to ethanol and does not survive the brewing
process (Magnus, Ingledew and Casey, 1986).
Klebsiella spp. have also been associated with production of excessive
176 Gram-negative spoilage bacteria

o o o
OH OH OH

C~-CH-COOH C~-CH-COOH C~-COOH

OH N~
p-hydroxyphenylacllc acid Tyrosine p-hydroxyphenylacetlc aCId

I
o
OH

CH=CH-COOH

p-hydroxyphenylacryloc aCId

1
Volatile phenols

Fig. 6.3 Production of volatile phenols from tyrosine.

volatile organo-sulphur compounds such as dimethyl sulphide (Anderson,

Howard and Hough, 1971) which affect beer quality adversely. Use of
35S-labelled sulphur showed that less than 2% of sulphur produced by
Klebsiella is derived from methionine, cysteine, cystine or sulphate
(Wainwright, 1972). It has been shown that R. aquatilis produces dimethyl
sulphide mainly by reduction of dimethyl sulphoxide (McCaig and
Morrison, 1984).

(e) Other enterobacterial contaminants

Other enterobacterial contaminants isolated from brewery worts include
E. cloacae, E. aerogenes, H. alvei, Serratia strains and P. mirabilis. Two genetic

OH OH OH

Cr~ -C~
Cr~
I~
+~

Cr~
I~
CH=CH-COOH CH=C~ CH,-C~

lerulic aCId 4-vinylguaiacol 4-ethylguaiacol

Fig. 6.4 Production of 4-vinylguaiacol and 4-ethylguaiacol from ferulic acid.

 Zymomonas 177
groups of E. cloacae have been isolated from beer breweries (Van Vuuren,
1978). E. cloacae and E. aerogenes produce phenolic compounds and dimethyl
sulphide respectively in beer. H. alvei contamination leads to increased
concentrations of n-propanol and dimethyl sulphide (for a review see Priest,
Cowbourne and Hough, 1974). Spoilage of beer by Serratia and P. mirabilis
has not been reported.
Enterobacterial contaminants occur most commonly in pre-fermentation
wort and are associated with early stages of fermentation. However, a
much wider distribution of most types occurs (Van Vuuren et al., 1978) than
was recognized previously. They continue to grow in fermenting wort and
produce metabolites which have a detrimental effect on beer quality.
Sulphury or phenolic off-flavours in beer might indicate enterobacterial
contamination. Another feature of enterobacteria is their ability to reduce
nitrate to nitrite. Unacceptable levels of nitrite may be formed in beer if
nitrate-containing water is used for brewing and contamination with en-
terobacteria occurs (Weiner, Rolph and Taylor, 1975; Hough et al., 1982). It
has recently been shown that O. proteus contains the enzyme nitrate reduc-
tase which converts nitrate to nitrite (Calderbank and Hammond, 1989). In
experimental fermentations, levels of apparent total N-nitroso compounds
(ATNC) at the end of fermentation were dependent on the initial wort
nitrate levels as well as the initial number of bacteria present in the pitching
yeast (Smith, 1994).
Enterobacteriaceae occurring in beer breweries are regarded as free-
living forms normally encountered in water, soil and plant material. Water,
barley, malt and hops are probably the main sources of contamination. It
is obvious that members of the Enterobacteriaceae are actively involved in
beer spoilage. Many brewery microbiologists have so far underestimated
the significance of these contaminants.

6.4 ZYMOMONAS

6.4.1 General characteristics

Zymomonas cells are short, plump, Gram-negative rods, 1.0-1.4 !lm
x 2.0-6.0 !lm. They occur mostly in pairs but single cells are usually present.
Some strains form chains, rosettes, and in many strains filamentous cells
are present. Endospores are not formed. Most strains are non-motile. Motile
strains have one to four polar flagella. Glucose and fructose are fermented
to nearly equimolar amounts of ethanol. They do not ferment maltose, are
oxidase negative and anaerobic but tolerate some oxygen. No growth is
observed on nutrient agar. Zymomonas has a G + C content of 47.5-
49.5 mol% (Swings and De Ley, 1984).
178 Gram-negative spoilage bacteria
6.4.2 Taxonomy
Swings and De Ley (1977) studied a large collection of strains from different
ecological niches. They suggested that only a single species, Zymomonas
mobilis, be recognized with two subspecies: Z. mobilis subsp. mobilis and
Z. mobilis subsp. pomacii. The latter was created specifically for the cider
sickness zymomonads. This classification system was followed in Bergey's
Manual of Systematic Bacteriology (Swings and De Ley, 1984). Synonyms
assigned to Zymomonas are listed in Table 6.2.

6.4.3 Metabolism
The almost quantitative fermentation of glucose and fructose to ethanol
and CO2 via a modified Entner-Doudoroff pathway (Fig. 6.5) is a distinctive
characteristic of the genus Zymomonas. Small amounts of acetaldehyde,
acetylmethylcarbinol, acetic acid, lactic acid and glycerol are formed
(Schreder, Brunner and Hampe, 1934).
The ability of Zymomonas to grow and metabolize in the presence of high
ethanol concentrations is unusual for a bacterium and emphasizes its
importance as a contaminant in the alcoholic beverage industry. Comrie

Table 6.2 Synonyms of Zymomonas (data from Swings and De Ley, 1977)
Zymomonas mobilis subsp. mobilis (Lindner) De Ley & Swings (1976)
Synonyms
Termobacterium mobile Lindner (1928)
Pseudomonas lindneri Kluyver & Hoppenbrouwers (1931)
Sacc111lromonas lindneri (Kluyver & Hoppenbrouwers) Shimwell (1950)
Zymomonas mobile (sic) (Lindner) Kluyver & Van Niel (1936)
Zymomonas mobilis (Lindner) Kluyver & Van Niel (1936)
Zymomonas mobilis var. anaerobia Richards & Corbey (1974)
Zymomonas congolensis Van Pee & Swings (1971)
Achromobacter anaerobium Shimwell (1937)
Sacc111lrobacter Shimwell (1937)
Sacc111lromonas anaerobia Shimwell (Shimwell) (1950)
Zymomonas anaerobia Shimwell (Kluyver) (1957)
Zymomonas anaerobia var. anaerobia (Shimwell) Carr (1974)
Sacc111lromonas anaerobia var. immobilis Shimwell (1950)
Zymomonas anaerobia var. immobilis (Shimwell) Carr (1974)
Zymomonas mobilis var. recifensis Goncalves de Lima, De Araujo,
Schumacher & Cavalcanti Da Silva (1970)

Zymomonas mobilis subsp. pomacii (Millis) De Ley & Swings (1976)

Synonyms
Cider sickness organism Barker & Hillier (1912)
Zymomonas anaerobill var. pomaceae Millis (1956)
 levan

, "">"'" "D' g,",a.:', /H 0 2

? glucose- NADP+ [9Iuconate- ] gluconate- 2-keto,3-deoxy-

t---- !I 9IUCose':iiiE6_ hos hate ~ 6-phosphate - 6-phosphate ~ gluconate-6-
ATP p. p a-lactone I phosphate

sucrose Ilactate I
I l"""~ §]
ATP
fructose-6-
fructose ~ phosphate
I
I
I
I
glyceraldehyde-
3-phosphate
I
pyruvate ~ CH J
t CHO
NADH
=:;;::j CH 3
~~~~~I CH 2 OH
I
I
I
I
Pi t / \
? fructose-1,6-
• - NADH+
ATP
, ,
acetoin CH 3
aldolase ~ diphosphate I
COOH
-
glyceric acid- phospho-
phosphatase 1,3-diphosphate enolpyruvate

ATP
" NADH H2 0

"D") -
NADPH . '
glyceric acid- • glyceric acid-
(NAD'
particulate 3-phosphate 2-phosphate

Fig. 6.S Mechanism of carbohydrate catabolism in two Zyrnornonas strains. Strain NCIB 8938: - , active, - -., weakly active; strain
NeIB 8777: - - , active, - - -, weakly active. (From Swings and De Ley, 1977, reproduced with permission of Bacteriological Reviews.)
180 Gram-negative spoilage bacteria
(1939) reported that growth of Zymomonas is inhibited by 8% v Iv ethanol.
However, Swings and De Ley (1977) found that almost half of their strains
grew at 10% v Iv ethanol. Z. mobilis strain BS057 survived ethanol concen-
trations of 12-13% v Iv produced during high gravity brewing (Magnus,
Ingledew and Casey, 1986). Although growth of some Zymomonas strains
may be arrested by 8-10% v Iv ethanol, they remain metabolically active
and can produce up to 15% v Iv ethanol. Zymomonas is an attractive
organism for the industrial production of ethanol (Swings and De Ley,
1977).

6.4.4 Beer spoilage

Shimwell (1937) first isolated Zymomonas from beer and named it
'Achromobacter anaerobium'. Zymomonas strains studied by Shimwell pro-
duced hydrogen sulphide and acetaldehyde in primed beer. This was
confirmed by Anderson, Howard and Hough (1971), Dadds, MacPherson
and Sinclair (1971), Dadds and Martin (1973) and Wainwright (1972). Trace
amounts of dimethyl sulphide and dimethyl disulphide are also produced
(Dadds, Martin and Carr, 1973). Shimwell considered Zymomonas the most
malignant beer spoilage bacterium recorded at that time. Beer contam-
inated withZymomonas has a rotten apple (Shim well, 1950) or fruity (Dadds
and Martin, 1973) odour. Zymomonas is an important contaminant in ale
breweries, mainly in the UK (for a review see Dadds and Martin, 1973).
Zymomonas has not been reported in lager breweries. Its relatively
stringent carbohydrate requirements probably restrict the organism to ale
breweries (Dadds, MacPherson and Sinclair, 1971). However, according
to Swings and De Ley (1977) its possible presence in lager breweries
should not be ignored. They suggested that the low temperatures (8-12°C)
of lager fermentations are unfavourable for the development of
Zymomonas.
Zymomonas has also been isolated from soil near breweries and from
brushes of cask-washing machines (Shimwell, 1937, 1950), well water
(Stevens, according to Dadds and Martin, 1973), fermented apple juice
(cider; Barker and Hillier, 1912; Millis, 1951, 1956), pear juice (perry; Millis,
1951) and ripening honey (Ruiz-Argueso and Rodriquez-Navarro, 1975).
Zymomonas strains are also responsible for the spontaneous alcoholic fer-
mentation of agave sap (pulque) (Lindner, 1928), palm wine (Swings and
De Ley, 1977) and sugar-cane juice (Goncalves de Lima et al., 1970).

6.5 ANAEROBIC GRAM-NEGATIVE RODS

Brewery microbiologists have generally accepted that beer spoilage bac-

teria are restricted to certain aerobic and facultatively anaerobic species.
However, spoilage of packaged beer by strictly anaerobic rods has been
 Anaerobic Gram-negative rods 181
reported (Lee, Mabee and Jangaard, 1978; Lee et al., 1980). Lee, Mabee and
Jangaard (1978) named a new genus and species for these obligately
anaerobic rods, viz. Pectinatus cerevisiiphilus. Bacterial contaminants sim-
ilar to P. cerevisiiphilus have since been isolated from spoilt beer in
Germany (Back, Weis and Seidel, 1979; Seidel-Rufer, 1990), Scandinavia
(Haikara, 1985; Haikara et al., 1980) and Japan (Takahashi, 1983). In an
extensive taxonomic study of anaerobic rods isolated from breweries,
Schleifer et ai. (1990) identified a new species of the genus Pectinatus,
P. frisingensis, as well as a new species of the genus Selenomonas, S.Iacticifex
and a new genus, Zymophilus, comprising two species Z. raffinosivorans
and Z. paucivorans.

6.5.1 Pectinatus cerevisiiphilus

The description of P. cerevisiiphilus below is based on the amended descrip-
tion by Schleifer et al. (1990). P. cerevisiiphilus cells are slightly curved rods,
0.7-0.9 /lm x 2.0-30 /lm; they occur singly, in pairs, and only rarely in short
chains. The occurrence of longer, helical filaments is characteristic for older
cells. They are motile by flagella emanating from only one side of the cell
(comb-like). This organism is an obligately anaerobic, non-spore-forming
mesophile. No growth is obtained on agar plates under aerobic conditions
or under a CO2 atmosphere. They can, however, be cultured on thiogly-
collate agar by the roll-tube technique of Hungate (1969), or in a modified
MRS medium under anaerobic conditions (Schleifer et ai., 1990). Colonies
are circular, entire, beige to white, glistening and opaque. Glucose is
fermented to acetic and propionic acids. The G + C content of the DNA is
38-41 mol%.

6.5.2 Pectinatus frisingensis

Morphologically P. frisingensis is similar to P. cerevisiiphilus. Colonies on
modified MRS agar medium are shiny, opaque, circular, slightly yellow,
and 1-2 mm in diameter after 3 days at 30°C. In addition to acetic and
propionic acids, acetoin and sometimes minor amounts of succinic acid are
produced from glucose under fermentative conditions. The G + C content
of the DNA is 38-41 mol%. P. frisingensis can be distinguished from
P. cerevisiiphilus by its ability to ferment cellobiose, inositol and
N-acetyl-glucosamine and its inability to utilize xylose and melibiose
(Schleifer et al., 1990).

6.5.3 Selenomonas lacticifex

Four strains of S. Iacticifex have been isolated from pitching yeast. These
strains were obligately anaerobic, usually motile and non-spore-forming
rods, 0.6-0.9 /lm x 5.0-15 /lm. The cells are curved and crescent shaped and
182 Gram-negative spoilage bacteria
occur predominantly as single cells or (rarely) in pairs. The optimum
temperature for growth is 30°e. Colonies on modified MRS agar are
smooth, opaque, circular, yellowish and 2-3 mm in diameter after 3 days
at 30°e. D-Glucose is fermented to acetic, lactic and propionic acids. The
G + C content of the DNA is 51-52 mol% (Schleifer et al., 1990).

6.5.4 Zymophilus raffinosivorans

Z. raffinosivorans was isolated from pitching yeast and brewery waste. Cells
are straight to slightly curved rods, 0.7--O.9/lm x 3.0-15/lm, and occur
predominantly as single cells; pairs or short chains are sometimes observed
(Schleifer et al., 1990). They are non-spore-forming motile rods but motility
can be lost after several subcultivations. Colonies on modified MRS agar
are circular, smooth, opaque, slightly yellow and 1-2 mm in diameter after
3 days at 30°e. D-Glucose is fermented to acetic and propionic acids. The
G + C content of the DNA is 38-41 mol%.

6.5.5 Zymophilus paucivorans

Z. paucivorans was isolated from pitching yeast. In contrast to
Z. raffinosivorans, Z. paucivorans cells are curved or helically shaped. The
rod-shaped cells are 0.8-1.0 /lm x 5.0-30 /lm and occur singly, in pairs or in
short chains. They are motile but motility may disappear after subculturing.
Colonies on modified MRS agar are circular, smooth, slightly yellow, and
1-2 mm in diameter after 3 days at 30°e. D-Glucose is fermented to acetic
and propionic acids and trace amounts of lactic acid. Z. raffinosivorans and
Z. paucivorans differ in their ability to utilize xylose, rhamnose, xylitol,
raffinose, melibiose and inositol (Schleifer et al., 1990). The G + C content
of Z. paucivorans DNA is 39-41 mol%.

6.5.6 Beer spoilage

It has become clear that strictly anaerobic rod-shaped bacteria constitute an
important group of spoilage bacteria of packaged beer. Bacterial isolates
considered to be Pectinatus strains produced a considerable amount of
acetic and propionic acids as well as acetoin in wort and packaged beers
(Seidel, Back and Weis, 1979; Lee et al., 1980). High concentrations of H 2S
were also detected in beer contaminated by Pectinatus strains. The beer
became turbid with an odour of rotten eggs. However, strains previously
thought to be similar to P. cerevisiiphilus have now been reclassified; they
belong to at least three different genera (Schleifer et al., 1990). Selenomonas
lacticifex is considered to be a spoilage organism and Zymophilus
raffinosivorans a potential spoilage organism of beer (Seidel-Rufer, 1990).
Z. paucivorans, however, is not regarded as a beer spoilage organism by
Seidel-Rufer (1990).
 Miscellaneous non-fermentative bacteria 183
Pectinatus strains have been isolated from lubrication oil mixed with beer
and water, and from a drainage system (Lee et al., 1980). However, the
source of contamination by anaerobic rods in breweries is not clear. Trans-
mission of anaerobic rod-shaped bacteria may well be by contaminated
pitching yeast.

6.6 MEGASPHAERA

Weiss, Seidel and Back (1979) isolated strictly anaerobic Gram-negative

cocci from finished beer spoilt by cloudiness and unpleasant odours. The
isolates were phenotypically similar to Megasphaera elsdenii. However, the
G + C values differed from that of the M. elsdenii type species and they were
regarded as Megasphaera species I and II. Engelmann and Weiss (1985)
studied the taxonomic position of 12 Megasphaera strains isolated from beer.
These strains were genetically closely related and they suggested a new
species, M. cerevisiae, in the family Veilonellaceae.
M. cerevisiae cells are Gram-negative, slightly elongated cocci, 1.3-1.6 11m
in diameter, in pairs and occasionally in short chains. They are non-motile
and non-spore forming. They are strictly anaerobic and colonies on a basal
medium containing fructose or lactate are whitish, smooth and flat, 2-3 mm
in diameter. Growth occurs between 15°C and 37°C with an optimum of
28°C. M. cerevisiae is negative for catalase and benzidine; it produces H 2S.
Nitrate is not reduced to nitrite and it is indole negative. Fructose is
fermented by all strains but only some strains ferment arabinose. The mol%
G + C of the DNA is 42.4--44.8 (Tm). Their habitat is unknown (Engelmann
and Weiss, 1985).
Megasphaera strains produce considerable amounts of butyric acid in
beer. Acetic, isovaleric, valeric and caproic acids as well as acetoin are also
produced in smaller quantities (Seidel, Back and Weiss, 1979). In wort, large
quantities of butyric and caproic acids and smaller amounts of acetic,
isovaleric and valeric acids are produced. No acetoin is produced in wort
(Seidel, Back and Weiss, 1979). Megasphaera is considered a true beer spoiler
by Seidel, Back and Weiss, (1979). M. cerevisiae, however, is sensitive to
alcohol and low pH (Haikara and Lounatmaa, 1987). Growth of this
contaminant in beer is also restricted at an alcohol content of 2.8% w Iv.

6.7 MISCELLANEOUS NON-FERMENTATIVE BACTERIA

Alcaligenes, Acinetobacter, Flavobacterium and Pseudomonas spp. have been

isolated from worts and brewing liquors (Masschelein, 1973; Eschenbecher
and Ellenrieder, 1975; Priest, 1981a). These bacteria have a strict respiratory
type of metabolism. Spoilage of beer by these occasional contaminants has
not been reported.
184 Gram-negative spoilage bacteria
6.8 DETECTION, ENUMERATION AND ISOLATION

The ability of certain Gram-negative bacteria to spoil beer is now widely

accepted. To prevent or minimize spoilage, effective microbiological
controls are essential. A number of methods exist for the detection of
bacterial contaminants in beer breweries. Traditionally, bacteria are de-
tected by direct microscopic observation and by cultivation on suitable
nutrient media. The need for more rapid detection has been recognized
and several other methods have been employed. These include the ATP
assay, pH change assay, measurement of the uptake of radioactive 14C,
and use of the Coulter counter, fluorescence microscopy, serology, optical
brighteners and staining techniques. These alternative methods are dis-
cussed in Chapter 8. Therefore, only microscopic techniques and some
methods of culturing Gram-negative spoilage bacteria are considered in
this chapter.

6.8.1 Microscopic examination

Direct microscopic examination of a suspension or stained smears can be
used to detect bacterial contaminants. These rapid methods yield valuable
information concerning the presence, morphology and motility of bacteria.
When used in conjunction with the Gram stain, a distinction can be made
between Gram-negative and Gram-positive isolates. However, these meth-
ods can only be used when the numbers of bacteria exceed 30 000 mrl.
Furthermore a distinction between live and dead cells cannot be made and
it is sometimes difficult to distinguish non-motile bacteria from insoluble
matter.
The Petroff-Hausser counting chamber can be used to determine bacte-
rial cell numbers in suspensions. The disadvantages of the method are
threefold: distinguishing dead cells from living cells; low bacterial popula-
tions cannot be calculated unless the sample is centrifuged beforehand; and
the actual counting procedure is tiresome. Despite these shortcomings, it is
a good practice to examine samples from breweries regularly under the
microscope.

6.8.2 Culture media

Many Gram-negative bacterial species grow well on simple laboratory
media. Others require specific nutrients such as vitamins and various
growth-promoting substances. Furthermore, special-purpose media are
available to facilitate recognition, enumeration and isolation of certain
types of bacteria.
The proper application of media in brewing bacteriology has been
thoroughly reviewed by Casey and Ingledew (1981). No attempt has been
made to repeat that review. I will rather concentrate on media that I have
 Detection, enumeration and isolation 185
found suitable for rapid culture of acetic acid bacteria, Enterobacteriaceae
and Zymomonas. In addition, the detection of the strictly anaerobic contam-
inants, Pectinatus, Selenomonas, Zymophilus and Megasphaera, will also be
discussed.
For the selective inhibition of yeasts, actidione (cycloheximide) can be
added to culture media at 20-100 Ilg ml-1• The lower concentration will
inhibit brewing yeasts and the higher concentration will prevent the
growth of most 'wild' yeasts.
Membrane filtration is most useful for enumerating low levels of organ-
isms in suspensions. A relatively large volume of fluid can be passed
through the filter which is then placed on a suitable agar medium. Colonies
developing on the surface of the filter are counted after a suitable incubation
period.

(a) Acetic acid bacteria

The nutrient requirements of acetic acid bacteria depend on the carbon
source supplied (De Ley, Gillis and Swings, 1984; De Ley, Swings and
Gossele, 1984). Williamson (1959) developed a medium for the detection
and isolation of acetic acid bacteria in breweries. However, Acetobacter
grows well on a standard medium containing 0.5% Bacto yeast extract
(Difco), 1.5% ethanol and 2.5% agar (Suomalainen, Keranen and
Kangasperko, 1965). De Ley and Swings (1984) described enrichment and
isolation procedures for Gluconobacter.

(b) Enterobacteriaceae
Enterobacteriaceae comprise a large and heterogeneous family of bacteria
with differing nutritional requirements. Numerous media, some selective,
have been developed to sustain growth of these organisms (for a review see
Casey and Ingledew, 1981). Lactose broth can be used to detect Citrobacter,
Enterobacter, Rahnella and Klebsiella strains whereas Universal Liquid Me-
dium supports good growth of most bacterial contaminants, including
enterobacteria, encountered in beer breweries (Van Vuuren et al., 1977).
Furthermore, enterobacterial contaminants can readily be cultured at 30°C
on selective media such as MacConkey agar. The lactose-positive Citrobac-
ter, Enterobacter and Klebsiella strains form red colonies whereas those of O.
proteus and H. alvei are pink to colourless. A prolonged incubation period
(36-48 h) is needed for growth of O. proteus (priest, 1981b).

(c) Zymomonas

Dadds (1971) described a suitable medium for the detection of Zymomonas.

The presence of Zymomonas is indicated by abundant gas production at
30°C after 2-6 days. False positive results may be due to the growth of
186 Gram-negative spoilage bacteria
lactobacilli or wild yeasts. Swings and De Ley (1977, 1984) list suitable
media for the isolation and maintenance of Zymomonas.

(d) Pectinatus, Selenomonas and Zymophilus

A selective differential medium (LL Agar; lactate-lead acetate) was devel-
oped for isolation and differentiation of Pectinatus (Lee, Moore and Mabee,
1981). Lee tubes (Ogg, Lee and Ogg, 1979) containing LL agar can be used
for the detection and isolation of Pectinatus, a strict anaerobe. Pectinatus,
selenomonas and Zymophilus spp. can be grown under anaerobic conditions
at 30°C in modified MRS medium (Schleifer et al., 1990).

(e) Megasphaera
Megasphaera strains can be cultured under anaerobic conditions in a GasPak
system (Seidel and Back, 1979) on beer agar enriched with 1% glucose and
1% peptone, Micro-assay culture agar and nutrient agar (Merck AG, 7881).

6.8.3 Summary
Many breweries rely on their own media and methods. The nature of this
review precludes the possibility of considering these applications, of which
many undoubtedly have merit.
The nutritional requirements of Gram-negative bacterial contaminants
vary considerably and I doubt if any single medium meets all their require-
ments. The use of more than one medium type is therefore advisable.
Furthermore, it is good practice occasionally to include other medium types
to ensure that all bacterial contaminants are detected.

6.9 CONCLUSIONS

A wide range of Gram-negative bacterial contaminants including acetic

acid bacteria, certain enterobacteria, Zymomonas, Pectinatus, Selena monas
and Megasphaera are able to spoil beer. The presence of these contaminants
during different stages of the brewing process or in the final beer affects the
concentration of certain volatile compounds to some degree depending on
the level of contamination. A complex and delicate balance of flavour
compounds determines flavour, aroma and quality of beer. Increased
concentrations of certain volatiles affect beer quality adversely; a reduction
thereof could also lead to off-flavours by allowing other compounds pres-
ent at normal concentrations to dominate flavour and aroma. To ensure
flavour consistency, bacterial contaminants in the brewing process either
should be eliminated or their numbers should be kept as low as possible.
It is not possible to operate a commercial brewery under aseptic conditions,
 References 187
but special attention should be given to ensure that cleaning procedures
are adequate.
Since the previous edition of this book appeared in 1987, knowledge
concerning Gram-negative spoilage bacteria, particularly the anaerobic
rods and cocci, has increased considerably. P. jrisingensis, S. lacticifex,
Z. raffinosivorans and Z. paucivorans have been described as contaminants
in beer. P. jrisingensis and s. lacticifex are now regarded as spoilage organ-
isms and Z. raffinosivorans as a potential spoilage bacterium. The taxonomic
position of Megasphaera species I and II and that of E. agglomerans has been
clarified. However, it seems highly likely that hitherto undescribed
Gram-negative spoilage bacteria exist in breweries. Brewing micro-
biologists should accept the challenge to isolate and classify such strains
and determine their effect on beer fermentation.

ACKNOWLEDGEMENTS

I thank Professor M.J. Hattingh, Department of Plant Pathology, University

of Stellenbosch, for a critical review of this manuscript and The South
African Breweries (Pty) Ltd for cooperation, financial assistance and access
to their breweries.

REFERENCES

Aho, P.E., Seidler, RJ., Evans, H.J. and Raju, P.N. (1974) Phytopathology, 64, 1413.
Anderson, RJ., Howard, G.A. and Hough, J.5. (1971) Proceedings of the 13th Congress
of the European Brewery Convention, Estoril, IRL Press, Oxford, p. 253.
Asai, T. (1968) Acetic Acid Bacteria. Classification and Biochemical Activities, University
of Tokyo Press. Tokyo and University Park Press, Baltimore, p. 103.
Ault, RG. (1965) Journal of the Institute of Brewing, 71, 376.
Back, W. (1980) Brauwelt, 43,1562.
Back, W., We is, N. and Seidel, H. (1979) Brauwissenschaft, 32, 233.
Barker, B.T.P. and Hillier, V.F. (1912) Journal of Agricultural Science,S, 67.
Bernstein, L., Bienkinship, B.K. and Brenner, M.W. (1968) Proceedings of the Annual
Meeting of the American Society of Brewing Chemists. St Paul, MN, p. 150.
Brenner, D.J. (1981) In The Prokaryotes. Vol. 2 (eds M.P. Starr, H. Stolp, H.G. Trilper
et al.). Springer, Berlin. p. 1105.
Brenner, D.J. (1984) In Bergey's Manual of Systematic Bacteriology, 9th edn, Vol. 1 (eds
N.R Krieg and J.G. Holt), Williams and Wilkins, Baltimore, p. 408.
Brown, H.T. (1916) Journal of the Institute of Brewing, 22, 265.
Buchanan, RE. and Gibbons, N.E. (eds) (1974) Bergey's Manual of Determinative
Bacteriology, 8th edn, Williams and Wilkins, Baltimore, p. 29l.
Calderbank, J. and Hammond, J.RM. (1989) Journal of the Institute of Brewing, 95, 277.
Campbell, 1. (1983) The Brewer (October) 414.
Carr, J.G. (1968) In Identification Methods for Microbiologists, Part B (eds B.M. Gibbs
and D.A. Shapton), Academic Press, London, p. l.
Carr, J.G. (1974) In Bergey's Manual of Determinative Bacteriology, 8th edn (eds
RE. Buchanan and N.E. Gibbons), Williams and Wilkins, Baltimore, p. 352.
188 Gram-negative spoilage bacteria
Case, AC. (1965) Journal of the Institute of Brewing, 71,250.
Casey, G.P. and Ingledew, W.M. (1981) Brewers' Digest (February), 26.
Comrie, AA.D. (1939) Journal of the Institute of Brewing, 45, 342.
Cosbie, AJ.C., Tosie, J. and Walker, T.K. (1942) Journal of the Institute of Brewing,
48,82.
Cosbie, AJ.C., Tosie, J. and Walker, T.K. (1943) Journal of the Institute of Brewing,
49,88.
Cowboume, M.A, Priest, F.G. and Hough, J.5. (1972) Brewers' Digest, 47, 76.
Dadds, M.J.5. (1971) In Isolation ofAnaerobes, Technical Series No.5 (eds D.A. Shapton
and RG. Board), Academic Press, London, p. 219.
Dadds, M.J.5. and Martin, P.A. (1973) Journal of the Institute of Brewing, 79, 386.
Dadds, M.J.S., MacPherson, AL. and Sinclair, A (1971) Journal of the Institute of
Brewing, 77, 453.
Dadds, M.J.5.,Martin,P.A. and Carr,I.G. (1973) Journal ofApplied Bacteriology, 36,531.
Dadic, M., van Gheluwe, J.E.A. and Valyi, Z. (1971) Wallerstein Laboratories
Communications, 34, 5.
De Ley, J. (1961) Journal of General Microbiology, 24,31.
De Ley, J. and Kersters, K. (1964) Bacteriological Reviews, 28, 164.
De Ley, J. and Swings, I. (1976) International Journal ofSystematic Bacteriology, 26, 146.
De Ley, I. and Swings, J. (1984) In Bergey's Manual of Systematic Bacteriology, 9th
edn, Vol. 1 (eds N.R Krieg and J.G. Holt), Williams and Wilkins, Baltimore,
p.275.
De Ley,J., Gillis, M. and Swings,J. (1984) In Bergey's Manual ofSystematic Bacteriology,
9th edn, Vol. 1 (eds N.R Krieg and I.G. Holt), Williams and Wilkins, Baltimore,
p.267.
De Ley, J., Swings, I. and Gossele, F. (1984) In Bergey's Manual of Systematic
Bacteriology, 9th edn, Vol. 1 (eds N.R Krieg and I.G. Holt), Williams and
Wilkins, Baltimore, p. 268.
Engelmann, U. and Weiss, N. (1985) Systematic and Applied Microbiology, 6, 287.
Eschenbecher, F. and Ellenrieder, M. (1975) Proceedings of the 15th Congress of the
European Brewery Convention, Nice, IRL Press, Oxford, p. 497.
Ewing, W.H. and Fife, M.A (1972) International Journal of Systematic Bacteriology,
22,10.
Frateur, J. (1950) La Cellule, 53,287.
Gilliland, RB. and Lacey, J.P. (1966) Journal of the Institute of Brewing, 72,29lo
Gillis, M. and De Ley, I. (1980) International Journal of Systematic Bacteriology, 30, 7.
Goncalves de Lima, 0., De Araujo, J.M., Schumacher, E. and Cavalcanti Da Silva,
E. (1970) Revista do Instituto de Antibioticos Universidade do Recife, 10, 3.
Graham, D.C. and Hodgkiss, W. (1967) Journal of Applied Bacteriology, 30,175.
Greenfield, S. and Claus, G.W. (1972) Journal of Bacteriology, 112, 1295.
Haikara, A (1985) Brauwissenschaft, 38, 239.
Haikara, A and Lounatmaa, K. (1987) Proceedings of the 21st Congress of the European
Brewery Convention, Madrid, IRL Press, Oxford, p. 473.
Haikara, A., Enari, T.M. and Lounatmaa, K. (1981) Proceedings of the 18th Congress
of the European Brewery Convention, Copenhagen, IRL Press, Oxford, p. 229.
Haikara, A, Penttila, L., Enari, T.M. and Lounatmaa, K. (1980) Applied and
Environmental Microbiology, 41, 51lo
Halcrow, RM., Glenister, P.R, Brumsted, D.D. and Lautenbach, AF. (1966)
Proceedings of the 9th Convention of the Institute of Brewing (Australian Section),
Institute of Brewing, London, p. 273.
Harnze, M., Mergaert, J., Van Vuuren, H.J.I. et al. (1991) International Journal of Food
Microbiology, 13, 63.
Harper, D.R (1980) Process Biochemistry, (December /January), 2.
Harper, D.R (1981) Brewers' Guardian (August), 23.
 References 189
Hough, J.S., Briggs, D.E., Stevens, R and Young, T. W. (1982) In Malting and Brewing
Science, Vol. 2, Chapman & Hall, London, p. 74l.
Hungate, RE. (1969) In Methods in Microbiology, Vol. 3B (eds J.R Norris and
D.W. Ribbons), Academic Press, London, p. 117.
Ingledew, W.M. (1979) Journal of the American Society of Brewing Chemists, 37, 145.
Ingledew, W.M., Sivaswamy, G. and Burton, J.D. (1980) Journal of the Institute of
Brewing, 86, 165.
Inoue, T., Masuyama, K, Yamamoto, Y. et al. (1968) Proceedings of the Annual Meeting
of the American Society of Brewing Chemists, St Paul, MN, p. 158.
Izard, D., Gavini, F., Trinel, P.A. and Leclerc, H. (1979) Annales de Microbiologie
(Paris), 130, 163.
Izard, D., Ferragut, C, Gavini, F. et al. (1981) International Journal of Systematic
Bacteriology, 31, 116.
Joyner, Jr, A.E. and Baldwin, RL. (1966) Journal of Bacteriology, 92, 132l.
Keevil, W.J., Hough, J.s. and Cole, J.A. (1979) Journal of the Institute of Brewing, 85,99.
Kersters, K and De Ley, J. (1968) Antonie van Leeuwenhoek, 34, 393.
Kirchner, G., Lurz, R and Matsuzawa, K (1980) Brauwissenschaft, 33, 46l.
Kluyver, A.J. (1957) In Bergey's Manual of Determinative Bacteriology, 7th edn (eds RS.
Breed, E.G.D. Murray and N.R Smith), Williams and Wilkins, Baltimore, p.
199.
Kluyver, A.J. and Hoppenbrouwers, W.J. (1931) Archiv fUr Mikrobiologie, 2, 245.
Kluyver, A.J. and Van Niel, CB. (1936) Zentralblatt fUr Bakteriologie, Parasitenkunde,
Infektionskrankheiten und Hygiene, Abt. II, 94, 369.
Krieg, N.R and Holt, J.G. (eds) (1984) Bergey's Manual of Systematic Bacteriology, 9th
edn, Vol. I, Williams and Wilkins, Baltimore.
Lee, S.Y., Mabee, M.s. and Jangaard, N.D. (1978) International Journal of Systematic
Bacteriology, 28, 582.
Lee, S. Y., Moore, S.E. and Mabee, M.s. (1981) Applied and Environmental Microbiology,
41,386.
Lee, S.Y., Mabee, M.s., Jangaard, N.O. and Horiuchi, E.K. (1980) Journal of the
Institute of Brewing, 86, 28.
Leisinger, T. (1965) Zentralblatt fur Bakteriologie, Parasitenkunde, Infektionskrankheiten
und Hygiene, Abt. II, 119, 329.
Lindner, P. (1895) Mikroskopische Betriebskontrolle in den Giirungsgewerben mit einer
EinfUhrung in die technische Biologie, Hefenreinkultur und Infektionslehre, 1st Aufl.,
Parey, Berlin, p. l.
Lindner, P. (1928) Bericht des Westpreussischen botanisch-zoologischen Vereins, 50,253.
Magnus, CA., Ingledew, W.M. and Casey, G.P. (1986) Journal of the American Society
of Brewing Chemists, 44,158.
Makinen, V., Tanner, Rand Haikara, A. (1981) Brauwissenschaft, 34,173.
Masschelein, CA. (1973) Brewers' Digest, 48,54.
McCaig, R and Morrison, M. (1984) Journal of the American Society of Brewing
Chemists, 42, 23.
Millis, N.F. (1951) PhD Thesis, University of Bristol, Bristol, UK
Millis, N.F. (1956) Journal of General Microbiology, 15, 52l.
Niefind, H.J. and Spath, G. (1971) Proceedings of the 13th Congress of the European
Brewery Convention, Estoril, IRL Press, Oxford, p. 459.
Ogg, J.E., Lee, S.Y. and Ogg, B.J. (1979) Canadian Journal of Microbiology, 25, 987.
Olsen, E. (1939) Svenska Bryggareforeningens Monadsblad, 54, 8.
Ploss, M., Erber, J. and Eschenbecher, F. (1979) Proceedings of the 17th Congress of the
European Brewery Convention, West Berlin, IRL Press, Oxford, p. 39.
Prest, A.G., Hammond, J.R.M. and Stewart, G.S.A.B. (1994) Applied and
Environmental Microbiology, 60, 1635.
Priest, F.G. (1981a) Journal of the Institute of Brewing, 87, 299.
190 Gram-negative spoilage bacteria
Priest, F.G. (1981b) In An Introduction to Brewing Science and Technology, Part II,
Contamination, Institute of Brewing, London, p. 23.
Priest, F.G. and Hough, J.S. (1974) Journal of the Institute of Brewing, 80, 370.
Priest, F.G., Cowbourne, M.A. and Hough, J.S. (1974) Journal of the Institute of
Brewing, 80, 342.
Priest, F.G., Somerville, H.J., Cole, J.A. and Hough, J.S. (1973) Journal of General
Microbiology, 75,295.
Priicha, J. and Scheible, E. (1970) Brauwelt, 110, 1233.
Rainbow, C. (1961) In Progress in Industrial Microbiology, Vol. 3 (ed. D.J.D.
Hockenhull), Heywood, London, p. 45.
Rainbow, C. (1966) Wallerstein Laboratories Communications, 29,5.·
Rainbow, C. (1981) In Brewing Science, Vol. 2 (ed. J.RA Pollock), Academic Press,
London, p. 49l.
Richards, M. and Corbey, D.A (1974) Journal of the Institute of Brewing, 80,24I.
Ruiz-Argueso, T. and Rodriques-Navarro, A (1975) Applied Microbiology, 30, 893.
Schleifer, K.H., Leuteritz, M., Weiss, N. et al. (1990) International Journal of Systematic
Bacteriology,40,19.
Schreder, K., Brunner, R and Hampe, R (1934) Biochemische Zeitschrift, 273, 223.
Seidel, H. (1980) Brauwelt, 43, 1216.
Seidel, H., Back, W. and Weiss, N. (1979) BrauwissenschaJt, 32, 262.
Seidel-Rufer, H. (1990) BrauwissenschaJt, 3, lOI.
Shimwell, J.L. (1936) Journal of the Institute of Brewing, 42, 585.
Shimwell, J.L. (1937) Journal of the Institute of Brewing, 43, 507.
Shimwell, J.L. (1947) Wallerstein Laboratories Communications, 10, 195.
Shimwell, J.L. (1948a) Wallerstein Laboratories Communications, 11, 27.
Shimwell, J.L. (1948b) Wallerstein Laboratories Communications, 11,135.
Shimwell, J.L. (1950) Journal of the Institute of Brewing, 56, 179.
Shimwell, J.L. (1963) Brewers' Journal, 99, 759.
Shimwell, J.L. (1964) Journal of the Institute of Brewing, 70, 247.
Shimwell, J.L. and Grimes, M. (1936) Journal of the Institute of Brewing, 42, 348.
Skerman, V.D.B., McGowan, V. and Sneath, P.H.A (1980) International Journal of
Systematic Bacteriology, 30,225.
Smith, N.A. (1994) Journal of the Institute of Brewing, 100, 347.
Society of American Bacteriologists (1957) Manual of Microbiological Methods,
McGraw-Hill, New York, p. 15.
Stouthamer, A.H., Van Boom, J.H. and Bastiaanse, AJ. (1963) Antonie van
Leeuwenhoek, 29,393.
Strandskov, F.B. (1965) Wallerstein Laboratories Communications, 28, 29.
Soumalainen, H. and Ronkainen, P. (1968) Nature, 220, 792.
Soumalainen, H., Keranen, E.J.A. and Kangasperko, J. (1965) Journal of the Institute
of Brewing, 71, 4l.
Swings, J. and De Ley, J. (1977) Bacteriological Reviews, 41, l.
Swings, J. and De Ley, J. (1981) In The Prokaryotes, Vol. 1 (eds M.P. Starr, H. Stolp,
H.G., Triiper, A Balows and H.G. Schlegel), Springer, Berlin, p. 771.
Swings, J. and De Ley, J. (1984) In Bergey's Manual of Systematic Bacteriology, 9th
edn, Vol. 1 (eds N.R Krieg and J.G. Holt), Williams and Wilkins, Baltimore,
p.576.
Takahashi, N. (1983) Bulletin of Brewing Science, 28, II.
Thorwest, A (1965) Brauwelt, 105, 845.
Tosic, J. and Walker, T.K. (1944) Journal of the Institute of Brewing, 50, 296.
Tracey, RP. (1984) MSc Thesis, University of the Orange Free State, Bloemfontein,
South Africa.
Turner, J.G., Walker, S.K. and Blanko, M. (1976) Proceedings of the American
Phytopathological Society, 3, 268.
 References 191
Van Pee, W. and Swings, J. (1971) East African Agricultural and Forestry Journal,
36,311.
Van Vuuren, H.J.J. (1976) MSc Thesis, University of Stellenbosch, Stellenbosch,
South Africa.
Van Vuuren, H.J.J. (1978) PhD Thesis, Rijksuniversiteit-Gent, Gent, Belgium.
Van Vuuren, H.J.J. and Toerien, D.F. (1981) Journal of the Institute of Brewing, 87,229.
Van Vuuren, H.J.J., Cosser, K. and Prior, B.A. (1980) Journal of the Institute of Brewing,
86,31.
Van Vuuren, H.J.J., Louw, H.A., Loos, M.A. and Meisel, R (1977) Applied and
Environmental Microbiology, 33,246.
Van Vuuren, H.J.J., Kersters, K., De Ley, J. et al. (1978) Journal of the Institute of
Brewing, 84, 315.
Van Vuuren,H.J.J., Kersters, K., De Ley,J. and Toerien, D.F. (1981) Journal ofApplied
Bacteriology, 51,51.
Von Riemann, J. and Scheible, E. (1969) Brauwelt, 109, 1074.
Wainwright, T. (1972) Brewers' Digest, 47, 78.
Walker, T.K. and Tosic, J. (1945) Journal of the Institute of Brewing, 51, 245.
Wallnofer, P. and Baldwin, RL. (1967) Journal of Bacteriology, 93, 504.
Weiner, J.P., Rolph, D.J. and Taylor, R (1975) Proceedings of the 15th Congress of the
European Brewery Convention, Nice, IRL Press, Oxford, p. 565.
Weiss, N., Seidel, H. and Back, W. (1979) Brauwissenschaft, 32, 189.
West, D.B., Lautenbach, A.F. and Brumsted, D.D. (1963) Proceedings of the Annual
Meeting of the American Society of Brewing Chemists, St Paul, MN, p. 194.
White, G.A. and Wang, CH. (1964a) Biochemical Journal, 90,408.
White, G.A. and Wang, CH. (1964b) Biochemical Journal, 90, 424.
Williams, P.J.le B and Rainbow, C (1964) Journal of General Microbiology, 35, 237.
Williamson, D.H. (1959) Journal of the Institute of Brewing, 65, 154.
Xu, H.S., Roberts, N., Singleton, F.L. et al. (1982) Microbial Ecology, 8, 313.
 CHAPTER 7

Wild yeasts in brewing and

distilling
1. Campbell

7.1 INTRODUCTION

Wild yeasts must have been a problem in beverage fermentations through-

out history, to brewers and winemakers in particular, and have been the
subject of numerous reviews since their microbiological nature became
understood. Wiles (1953), Gilliland (1967, 1971) and Rainbow (1981) have
given useful and detailed accounts of the adverse effects of wild yeasts,
which have changed little over the years. Ingledew and Casey (1982)
provided a comprehensive survey of culture media for detection and
isolation of wild yeasts, which they defined as 'yeasts not playing a signi-
ficant part in a normal fermentation'. In the review of wild yeasts in the
earlier edition of this book (Campbell, 1987), it was considered more
appropriate to use Gilliland's (1967) definition 'yeasts not deliberately used
and under full control', which will be used again to indicate the accidental
and haphazard nature of wild yeast infection.

7.2 DETECTION OF WILD YEASTS

It is good microbiological practice to test for the absence of undesirable

yeasts at each sensitive stage of the brewing process. This was strongly
emphasized by Wiles (1953), Gilliland (1967, 1971), Rainbow (1981) and the
present author (Campbell, 1987) in their reviews. The Analysis Committees
of the European Brewery Convention (Moll, 1981) and the Institute of
Brewing (Baker, 1991) have published standard methods for detection of
wild yeast contaminants, although with only minimal background
information on the microbiological principles.

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and 1. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0412 59150 2
194 Wild yeasts in brewing and distilling
In breweries, the various steps prior to hop boiling are of little import-
ance as a source of wild yeasts in the wort, since none is sufficiently heat
resistant to survive even brief boiling. Wiles (1953) reported small numbers
of yeasts in malt, hops and priming sugars, but even when beers were dry
hopped and primed for cask conditioning there was no evidence of any
harmful effect.
The unboiled worts of distillery practice might seem to be more at risk
from yeast contaminants of the malt (Chapter 4). Although a proportion of
bacterial contaminants may persist through mashing to the fermentation,
yeasts from malt normally have no adverse effect on the fermentation. With
few exceptions, yeast contaminants of malt are strictly aerobic species
(Chapter 4) and therefore, even if they survive mashing temperatures, they
are unable to grow during the subsequent fermentation, swamped by the
great excess of fermentative culture yeast and rapid generation of CO2 and
anaerobic conditions.
In both breweries and distilleries, samples of cooled aerated wort, and
swab samples of the surface of the cleaned empty fermentation vessel,
should be taken routinely before every fermentation. Since no yeast should
be present in either of these samples, any which are discovered should be
regarded with concern: even if by chance they are not dangerous contam-
inants, certainly their presence indicates imperfect sterilization procedures.
As indicated in Table 7.1, non-selective media have to be used for these
samples. Any wild yeasts, culture yeasts or bacteria detected indicate poor
hygiene and potential microbiological problems. Commercially available
malt extract agar or Wallerstein Laboratories nutrient (WLN) agar are
useful general non-selective media for the brewing and distilling indus-
tries, and are suitable for the detection of contaminant bacteria as well as
yeasts.
Samples of pitching yeast must be examined before use to ensure satis-
factory viability or vitality (Pfenninger, 1977; Baker,1991) but it is also good
practice to check each inoculum for a suitably low number, and preferably
the absence, of contaminant bacteria and wild yeasts. For this purpose, a
selective medium is necessary. Unfortunately there is no one medium
which suppresses brewery or distillery culture yeasts and supports the
growth of any possible contaminant yeast which may be present.
The most useful and versatile medium for detection of wild yeasts is
lysine agar (Campbell, 1987), a synthetic medium with lysine as the sole
source of nitrogen. Almost all yeast genera other than Saccharomyces have
the ability to utilize the amino groups of lysine for biosynthesis of other
amino acids, and hence proteins. Therefore on plating out contaminated
pitching yeast, or a sample taken during fermentation or from early stages
post-fermentation before removal of yeast, only those non-Saccharomyces
yeasts which may be present are able to grow to colonies of normal size.
Two important disadvantages of lysine agar are:
 Detection of wild yeasts 195
Table 7.1 Media for isolation of wild yeasts
1. Examination of wort, washings or swab samples from clean fermentors,
pipework, kegs, etc., filtered or pasteurized beer and other samples which
should not contain live yeasts.
Non-selective media:
(a) wort agar;
(b) malt extract agar;
(c) Wallerstein Laboratories nutrient (WLN) agar.
2. Examination of pitching yeast or samples during fermentation (Ingledew
and Casey, 1982).
(a) Inhibitory media (culture yeasts are suppressed; only certain contaminant
yeasts are able to grow):
(i) 'actidione' agar (WLN agar + actidione), suppresses culture yeasts
and many sensitive wild yeasts;
(ii) copper sulphate_?r crystal violet agar (wort agar or malt extrac!..Tgar
+ either 61lg ml CUS04 (Taylor and Marsh, 1984) or 20 Ilg ml
crystal violet), suppress culture strains of S. cerevisiae but wild strains
of Saccharomyces mat grow
(iii) fuchsin/sulphite agar (various formulations): similar principle to
CuS04 agar.
(b) Selective media (nutrients available only for wild yeast):
(i) lysine agar, selective for wild yeasts capable of growing on lysine as
sole source of N;
(ii) dextrin agar, or starch agar, selective for yeasts capable of utilizing
dextrin or starch as sole source of C;
(iii) citrate agar, xylose agar, etc., selective for yeasts capable of using sole
C source supplied.

1. its selective effect is by starvation, and carried-over extracellular or

intracellular nutrient will allow some slight growth of Saccharomyces
yeasts to visible colonies;
2. Saccharomyces spp., biochemically adapted to grow well in brewery and
distillery fermentations, are potentially the commonest contaminants
of these situations but cannot be detected.
Therefore, although S. bayanus, S. cerevisiae, S. exiguus, S. pastorianus and
S. unisporus are the only brewery contaminants listed in Table 7.2 which are
unable to grow on lysine medium, inability to detect wild strains of
S. cerevisiae in particular is a serious problem.
A different approach can be used to detect diastatic wild yeasts. The most
important, particularly for its role as a post-fermentation contaminant of
beers, was first isolated by Andrews and Gilliland (1952) and named
S. diastaticus, but is now regarded simply as a diastatic strain of S. cerevisiae
(Kreger-van Rij, 1984). A useful selective medium for this organism relies
on its ability to utilize dextrin or starch as a sole carbon source for growth
(Ingledew and Casey, 1982): synthetic medium of Wickerham' s (1951) yeast
196 Wild yeasts in brewing and distilling
Table 7.2 Yeast species reported as brewery contaminants (Barnett, Payne and
Yarrow, 1990)
Brettanomyces anomalus Hanseniaspora uvarum
B. claussenii H. valbyensis
B. custersianus H. vineae
B. custersii
B.lambicus Kluyveromyces marxianus

Candida beechii Pichia anomala

C. ernobii P·fabianii
C. humilis P·farinosa
C. intermedia P. fermentans
C. norvegica P. guilliermondii
C. oleophila P. membranaefaciens
C. parapsilosis P.ohmeri
C. sake P.onychis
C. solani P. orientalis
C. stellata P. subpelliculosa
C. tenuis
C. tropicalis Saccharomyces bayanus
c. vartiovaarai S. cerevisiae
C. versa til is S. exiguus
S. pastorianus
Debaryomyces hansenii S. unisporus
D. marama Schizosaccharomyces pombe
Dekkera bruxellensis
D. intermedia Torulaspora delbrueckii

Filobasidium capsuligenum Zygosaccharomyces bailii

Z. rouxii
When both perfect and imperfect forms of the yeast occur, only the perfect (sporing) form is
named (see Table 1.2).

nitrogen base + 1% starch or dextrin allows growth of'S. diastaticus' and

other amylolytic yeasts, while normal brewing strains of S. cerevisiae and
other non-amylolytic yeasts of any genus are unable to grow.
The general principle of providing a selective single or mixed carbon
source, which can be utilized by contaminant Saccharomyces strains but not
brewing yeast, has been examined with a range of compounds, but with
only limited success (see Ingledew and Casey, 1982). In general, contami-
nants recovered on media with specific carbon nutrients other than dextrin
or starch are of genera other than Saccharomyces and would have been
detected anyway on lysine agar.
Another possible medium which is occasionally effective is 'actidione
agar'. Although used primarily for the detection of contaminant bacteria
(Chapters 5 and 6), it is also suitable for detection of those wild yeasts which
happen to be actidione resistant. Actidione (cycloheximide) is an anti-
fungal antibiotic, but various wild yeasts are resistant to the 100 Ilg mrl
added to the commercially available medium, and can be detected, or
 Detection of wild yeasts 197
counted if quantitative procedures are used. Most culture yeast strains are
inhibited by 5 Ilg mr1 actidione, so WLN, malt extract or any other suitable
non-selective medium can be supplemented with the lowest concentration
of antibiotic which will reliably inhibit culture yeast, thereby producing a
medium with greater potential for isolation of wild yeast than the commer-
cially available 100 Ilg ml- 1 medium. Wild yeasts of many genera, including
some strains of Saccharomyces spp., may be sufficiently resistant to actidione
to be recovered in this way. Unfortunately, the resistant strains which can
be detected may represent only a part of the wild yeast population.
Numerous ingenious attempts to provide selective media that suppress
brewing strains of S. cerevisiae but allow growth of all wild Saccharomyces
strains, particularly of the species S. cerevisiae itself, have been generally
unsuccessful as routine media, despite their presumed success in their
inventors'laboratories with the yeast strains at their disposal. Ingledew and
Casey (1982) have provided a useful account of these media, many of
which, despite their limited selective value, work well with some
breweries' strains of culture and wild yeasts and may be no worse than
lysine, actidione or dextrin media. Subsequent to that review, Taylor and
Marsh (1984) described a medium containing CUS04 as selective agent, but
with the increasing success of molecular biological methods, as indicated
below, little further attention seems to have been devoted to development
of new selective media.
Lysine and dextrin media exert their selective effects by starvation.
Ideally, several washings by centrifugation are required to limit carry-over
of nutrient from the previous medium. However, it is my impression that
most microbiologists plate unwashed samples on dextrin or lysine media,
and tolerate slight growth of culture yeast on the traces of nutrient trans-
ferred with the inoculum; the extra effort of washing cultures may be
justified in research on yeast nutrition but probably not in routine isolation
of contaminants. Actidione and other inhibitor media suppress the growth
of culture yeast but it is important to remember that the cells are not killed,
and colonies of wild yeasts could be contaminated with surviving culture
yeast. Therefore colonies must never be inoculated directly to identification
media, but must first be plated on a non-selective medium, e.g. malt extract
broth, to ensure a pure culture.
Descriptions above apply to traditional culture methods on agar plates,
but the same principles can be applied to rapid culture methods, e.g.
growth analysers which measure change in impedance of a growing cul-
ture (Chapter 8). For example, with duplicate broth cultures in lysine and
complete media, growth in the former only indicates the presence of
non-Saccharomyces wild yeasts.
Immunological methods were described in detail in the earlier edition
of this book (Campbell, 1987). Immunofluorescence (IF) is a standard
recommended method (Pfenninger, 1977; Baker, 1991), useful for rapid
detection and counting of microbial, including yeast contaminants,
198 Wild yeasts in brewing and distilling
provided the appropriate specific antiserum is available. A particularly
important application is the detection of contaminants of the genus
Saccharomyces in pitching yeast and of fermentation samples, when no
selective medium is possible (Campbell, 1987). Although not yet published
as recommended methods for the brewing industry, applications of
enzyme-linked immunosorbent assay (ELISA) and the direct epifluor-
escence technique (DEFT) are common in the food and drink industries
(Pettipher et al., 1980).
Recent developments in molecular biology have produced new methods
for detection of specific contaminants. The polymerase chain reaction
(PCR) for amplifying unique fragments of DNA or RNA has been sug-
gested for detection of many different pathogens and spoilage bacteria in
the food and drink industries (Chapters 3 and 8), but with equivocal results
in beer (DiMichele and Lewis, 1993; Thompson et al., 1995). Also, like IF
which requires specific sera, PCR detection of a particular contaminant is
possible only when the specific primer, reacting with the contaminant but
not culture yeast, is available.
A commonly used version of DEFT based on enzymic hydrolysis of dark
fluorescein di-acetate to fluorescent fluorescein (Baker, 1991) ignores dead
cells. A serious disadvantage of PCR and standard IF methods is the
inability to distinguish living from dead cells: both have the nucleic acid
and cell wall structures which are detected in these tests. Furthermore,
these methods are limited to detection and counting of specific
contaminants; no culture is available for further analysis.

7.3 IDENTIFICATION OF WILD YEASTS

The principles of yeast classification and identification were discussed in

Chapter 1. Many contaminant yeasts of the brewing and related industries
are themselves strains of S. cerevisiae, but other species of Saccharomyces, and
other fermentative and non-fermentative genera are also liable to occur.
Table 7.2 shows the list of yeast species recorded by Barnett, Payne and
Yarrow (1990) as contaminants of brewery fermentations, but is almost
certainly incomplete as many instances of contamination are not followed
through to full identification, and even then the information would not
necessarily be published in the scientific literature. In distillery fermenta-
tions, few of these listed contaminants could compete with distiller's yeast,
and then only if introduced as a substantial proportion of the culture.
Therefore, provided the distilling yeast is of acceptable quality, the likely
causes of wild yeast contamination would be either improperly cleaned
fermentation vessels from a previous contaminated fermentation, or mas-
sively contaminated brewer's yeast used as supplementary pitching yeast.
How important is it to identify fully all yeast contaminants? It has been
suggested (e.g. Gilliland, 1971) that it is helpful to determine the genus and
 Identification of wild yeasts 199

Vegetative growth by
I
I
Multilateral budding
1
Polar budding

I
Strong fermentation of glucose Hanseniaspora

+ I I w/- (weak or no
fermentation)
Saccharomyces
Kluyveromyces Acetic acid production
Torulaspora (obvious by smell)
Zygosaccharomyces 1
+ 1-
r-I- - - ' - - - -....

Dekkera Debaryomyces, Pichia

Fig. 7.1 Simple identification scheme for common perfect yeast genera of the
brewing and related industries.
The genera above are 'perfect' yeasts which form spores. Equivalent non-sporing
'imperfect' genera are given below:

Spore-forming: Dekkera Hanseniaspora Other genera above

Non-sporing: Brettanomyces Kloeckera Candida

Yeast spoilage effects

Hanseniaspora/Kloeckera, Saccharomyces/Kluyveromyces/Torulaspora/Zygosaccharomyces
and equivalent Candida spp:
Unintended fermentation, producing turbidity and off-flavours
Dekkera / Brettanomyces:
Acetic acid spoilage, turbidity.
Debaryomyces/Pichia* and equivalent Candida spp:
Turbidity, yeasty or estery off-flavour, often with production of surface film/
pellicle fragmenting to flaky particles or deposit.
*Hansenula is no longer recognized as a separate genus: former Hansenula spp. are
now regarded as nitrate-utilizing spp. (or, in some cases, strains) of Pichia.

species of contaminants for full access to the literature on their adverse

effects and their elimination. On the other hand, a simple scheme such as
Fig. 7.1 for determination of the principal types of yeasts should provide
sufficient information to suggest possible sources of the contamination and
appropriate remedial action, as suggested below (section 7.5). Colony
morphology and colour (Hall, 1971) or resistance to antibiotics and other
inhibitors with a multi-agent 'yeastar' (Simpson, Fernandez and
Hammond, 1992) are simple labour-saving methods which are sufficient to
recognize similarity or difference of separate isolates, but give no
200 Wild yeasts in brewing and distilling
information on identity. Colony morphology and colour are difficult to
describe well enough for reliable exchange of information between labora-
tories, but the extent of inhibition by the agents on the 'yeastar' is more
easily expressed quantitatively. However, either exchange of cultures or
identification to genus and preferably to species level is advisable for
operations involving more than one laboratory.
In recent years the API identification kits (supplied by bioMerieux UK,
Basingstoke or API, Montalieu-Vercieu, France) have simplified identifica-
tion of bacteria and yeasts. The API 20C kit for yeasts, with 20 test media,
is primarily intended for the identification of pathogenic Candida spp. and
distinction from non-pathogenic yeasts which colonize body surfaces. The
system is equally applicable to identification of industrial yeasts, and the
manual available from API (Anon., 1994) will identify culture yeasts and
the commoner contaminants of the brewing and distilling industries. The
suppliers maintain a database of properties of all known yeasts, and
provide a consultancy service for identification of unusual isolates.
Where identification of substantial numbers of isolates is planned, it may
be more economical to use traditional identification media used according
to the methods of Kreger-van Rij (1984) or Barnett, Payne and Yarrow
(1990). A computer program is available as a supplement or alternative to
the latter publication to assist with identification.
If a yeast isolate from pitching yeast or a fermentation is identified as
S. cerevisiae, it is almost certain to be the pitching yeast itself. If isolated from
beer post-fermentation, further testing is necessary to determine whether
it is the culture yeast or a contaminant. Small-scale fermentations in tall-
tube fermentors (Pfenninger, 1977) will definitively distinguish culture
yeasts from each other and from contaminants, by demonstrating the
characteristic and constant brewing properties of the culture yeasts
(Campbell, 1987). As an alternative to the full range of tests in Table 7.3, gas
chromatographic analysis of the characteristic 'flavour profile' of indi-
vidual strains will suffice, but also requires several days' incubation. In
cases of doubtful identity of a batch of pitching yeast, it is impracticable to
wait for results of such a test, but even for less urgent situations, a lengthy
and labour-intensive small-scale fermentation is an unsatisfactory method
of identification.
Although all strains of S. cerevisiae, by definition, ferment glucose, su-
crose and maltose but fail to ferment lactose, strains vary with regard to
fermentation of other sugars, e.g. galactose, melibiose, melezitose, a-
methyl glucoside and trehalose (Kreger-van Rij, 1984). Even after only 6 h
incubation in these sugars, fermentative ability is obvious and difference
from, or identity with, known culture yeast can be established. Alterna-
tively, the 'yeastar' test can be used, as above. It is important to have a small
number of rapid tests which can reliably distinguish the culture strains of
the various brands of a brewery.
Molecular biology provides an alternative, but not necessarily more
 Effects of wild yeasts in the brewery 201
Table 7.3 Tests for distinguishing strains of Saccharomyces cerevisiae (after Gilliland,
1971, amended in accordance with current taxonomy)
1. Laboratory fermentation tests:
(a) all strains of S. cerevisiae ferment glucose, sucrose and maltose, mayor
may not ferment galactose, but do not ferment lactose (Kreger-van Rij,
1984);
(b) further differentiation can be achieved by testing fermentation of
maltotriose, melibiose, raffinose, trehalose, dextrin (and galactose, used in
initial identification tests).
2. Small-scale fermentation tests, e.g. EBC tall-tube fermentors:
(a) flocculation, fining;
(b) head formation;
(c) final pH of beer;
(d) uptake of nitrogen (especially amino N), isohumulones;
(e) rate of fermentation, attenuation limit;
(f) total yeast crop;
(g) liability to autolysis;
(h) volatiles produced, taste of final beer.
3. Additional laboratory tests: for distinguishing strains:
(a) giant colony morphology (and colour, if on WLN agar);
(b) sensitivity to antibiotics (e.g. different concentrations of cycloheximide,
nystatin, etc.);
(c) vitamin requirements (very labour intensive, and largely abandoned).

rapid possibility. The concept of 'DNA fingerprinting' pioneered by

Pedersen (1983) has not yet become an established analytical technique but
is expected to supplant cultural methods for rapid identification of different
culture yeasts, even in the most difficult situation of recognizing two
different lager strains (Meaden, 1990). The 'fingerprint' can be matched
with those of known cultures, but is unsuitable for identification of an
unknown culture. Recent restoration of the former specific names
S. bayanus and S. pastorianus has been possible only by DNA analysis:
traditional, i.e. cultural, identification tests can no longer distinguish them
from S. cerevisiae (Barnett, 1992; Martini, Martini and Cardinali, 1993):
electrophoretic analysis or 'DNA fingerprinting' is necessary.

7.4 EFFECTS OF WILD YEASTS IN THE BREWERY

Gilliland's (1967) definition 'yeasts not deliberately used and under full
control' can now be seen to cover four main types of wild yeast infection:
1. fermentative yeasts, which could also occur as
2. 'killer' yeasts and
202 Wild yeasts in brewing and distilling
3. the wrong type of culture yeast, and
4. 'aerobic' (non-fermentative) yeasts.
In the following pages we examine these types in more detail.

7.4.1 Fermentative yeasts

Yeasts of the genera Saccharomyces, Kluyveromyces, Torulaspora and Zygo-
saccharomyces are biochemically similar and, therefore, potentially able to
compete with culture strains of S. cerevisiae. With the exception of 'killer'
yeasts (see below), contaminant yeasts have little or no direct adverse effect
on culture yeasts, but if capable of growth at even a slightly faster rate will
increase relative to the culture yeast over successive fermentations. Wild
Saccharomyces contaminants with a more efficient and therefore faster rate
of nutrient uptake, with simpler vitamin requirements or with simpler
metabolic activity (e.g. inability to utilize maltotriose) are all potentially
capable of slightly faster growth than the culture yeast and of gradually
increasing in proportion.
Contamination of pitching yeast, even if initially pure, is probably
inevitable after a number of successive fermentations and repitchings in
traditional open rectangular vessels. The risk is reduced, but certainly not
eliminated, by the use of closed vessels. Whether chance contamination
increases or disappears over successive fermentations depends on various
process factors and the characteristics of the culture and wild strains. In ale
fermentations, the traditional head-forming yeast is skimmed at intervals
from the surface of the fermenting beer. Contaminants which rise with the
head will therefore be incorporated in the skimmed yeast and continue as
contaminants of the pitching yeast. If only the middle of three skimmings
is used as pitching yeast, it should be possible to control the proportion of
contaminants which happen to rise to the head early or late in the
fermentation.
Flocculation, the ability of yeast to adhere in clumps and sediment
rapidly from the medium in which they are suspended (Stratford, 1992),
is an important property of brewing yeast. It is important that yeasts
remain in suspension during the fermentation, but to avoid expensive
clarification by filtration or centrifugation it is important that brewing
yeasts spontaneously clump and settle out of the beer late in the fermen-
tation. A good brewing yeast does this, but many wild yeasts are non-
flocculent and so, in addition to any flavour faults they cause, the beer
remains turbid.
Detection of contaminants or mutants of unsuitably strong flocculence
or non-flocculence is difficult or impossible by the use of selective media
alone, but the changed flocculation properties are themselves selective and
assist recovery of the contaminant. The correct degree of flocculation is an
important property of brewing yeast (Stewart, 1975; Speers et al., 1992):
 Effects of wild yeasts in the brewery 203
small-scale fermentations will confirm the changed properties, and
comparison with the authentic original culture will show whether the
change is due to a contaminant, in which cultural properties will be
different, or a mutant, whose properties other than flocculence should be
unchanged. As above, simple tests for fermentation of sugars or resistance
to inhibitors give a rapid indication of a different yeast strain.
Changes in flocculence, although genetically controlled Gohnston and
Reader, 1983; Stratford, 1992), are expressed as changes in cell wall struc-
ture. Other changes of the yeast cell wall also affecting brewing behaviour
are fining and head-forming ability; the effect of mutations in these prop-
erties was discussed by Campbell (1987). For beers treated with isinglass
or other fining agents, the presence of non-fining contaminants creates haze
problems. Haze may also result from the fact that many of these contami-
nants are smaller than the culture yeasts which filtration or centrifugation
procedures have been developed to remove.
Although fermentative diastatic yeasts are possible contaminants of the
fermentation, they are also possible post-fermentation contaminants, hav-
ing available a rich nutrient source by their ability to utilize dextrin
(Andrews and Gilliland, 1952). Low levels, possibly only one viable cell, of
post-fermentation contamination can grow to spoilage levels as turbidity,
phenolic and other off-flavours and a reduction of the viscosity of the beer
by hydrolysis of dextrins (Gilliland, 1971; Campbell, 1987).
'5. diastaticus' strains are often flocculent, and it is possible that cells in
the centre of flocs could be insulated from pasteurization conditions. Al-
though such a situation has been postulated to explain failure of pasteuriza-
tion (Brightwell, 1972) it is unlikely that flocs would pass the preliminary
filtration stage. More likely as a cause of failure of pasteurization is the heat
resistance of yeast spores. Although brewing yeasts seldom sporulate, and
certainly are unlikely to do so in wort or beer (Chapter 3), many of the
potential contaminants of the brewery are spore formers (Table 7.2). Yeast
spores are slightly more heat resistant than vegetative cells (Put and de
Jong, 1980) but since pasteurization programmes are designed primarily to
destroy those non-sporing culture yeasts which pass the filtration stage,
spore-forming contaminants may survive.

7.4.2 Killer yeast

Normally in fermentations, contaminant yeasts or bacteria are in competi-
tion with the culture yeast and, at worst, increase in proportion over
successive fermentations with each reuse of the pitching yeast. Killer yeast
(Chapter 3) is an extreme form of contaminant, killing sensitive culture
yeast, as almost all are, and leaving the killer strain itself as the dominant
yeast of the fermentation. Maule and Thomas (1973) gave a detailed case
history of one of the earliest confirmed examples of an outbreak of killer
yeast infection replacing the culture yeast and producing an objectionable
204 Wild yeasts in brewing and distilling
off-flavour in addition to poor attenuation and other defects. Although
various zymocins, the preferred name for killer factors, are known and
yeasts of many genera produce them, under the conditions of brewery or
distillery fermentations, Saccharomyces spp. are the most likely killer strains.
Therefore with little chance of a suitable selective medium for early detec-
tion, the contamination is unlikely to be recognized until the killer strain
has become the dominant organism and detectable on non-selective me-
dium, e.g. by different colony colour or morphology. It is possible to
engineer resistant production yeast strains by transferring specific resist-
ance factors, or even killer factors, into cultures (Chapter 4), but so many
different killer factors exist that anticipating necessary resistance factors is
impracticable. Furthermore, such manipulation risks changing the valu-
able properties of the culture yeast, and also with current popular aversion
to 'genetic engineering', customers would react adversely.

7.4.3 Different culture yeasts

For most beers the specific yeast strain, or mixture of known strains, is an
essential component of the recipe. A different culture yeast, however well
it may perform in one of the other products of the brewery, is likely to give
an unsatisfactory beer when used in error. Breweries where different types
or brands are produced are permanently at risk of such accidents. Rate of
fermentation, final attenuation, nature and amount of the flavour by-
products of fermentation, flocculation and head formation are all important
properties of individual strains (Nykanen and Suomalainen, 1983;
Molzahn, 1993). Obviously the 'correct' yeast is completely absent after the
accidental substitution of the wrong culture, so some difference from
normal is possible for all of the above properties of the beer. Careful
documentation is the most effective safeguard, but either rapid cultural
methods or rapid tests with genetic probes, as described above (section 7.2),
will confirm the identity of the strain.

7.4.4 Aerobic yeasts

Yeasts of the genus Pichia, and particularly P. membranaefaciens, are the
commonest of the non-fermentative spoilage yeasts. The literature of both
brewing (e.g. Wiles, 1953; Rainbow, 1980) and wine microbiology (e.g.
Sponholz, 1993) treats these non- or weakly fermenting yeasts as aerobic
organisms, but this is not entirely true. In pure culture in wort, spoilage
Pichia spp. are capable of growth under completely anaerobic conditions
(Campbell and Msongo, 1991). In the acid and ethanol concentrations of
beer, however, these yeasts are no longer able to grow anaerobically and
show their generally recognized behaviour as strict aerobes.
Acetic acid-forming Brettanomyces and Dekkera spp., although ferment-
ative, require oxygen for growth and fermentation. They form an important
 Elimination of wild yeasts 205

component of the yeast flora of fermenting Belgian Iambic beer (van

Oevelen et al., 1977) but also, like the acetic acid bacteria Acetobacter and
Gluconobacter, are liable to grow in beer post-fermentation, if air
accidentally gains access.
The nature of the spoilage flora of beer is greatly affected by access of
air. Such air may itself be the source of the yeast or bacterial contaminants,
but even if not, access of air to packaged beer allows growth of any aerobic
yeasts which have escaped filtration and pasteurization, and encourages
growth of surviving facultative anaerobes. Of the possible contaminants
listed in Table 7.2, all are stimulated by oxygen, but yeasts of the genera
Debaryomyces, Filobasidium and Pichia, the last including the former genus
Hansenula, grow only under aerobic conditions in beer (Campbell and
Msongo, 1991), as do many spoilage species of Candida. The effects of
'aerobic' yeast growth are turbidity, often but not always associated with
a surface film or thick pellicle, which if present, fragments to a flaky
suspension or deposit. Production of acetic acid by Brettanomyces and
Dekkera spp. often leads to excessive production of ethyl and other acetate
esters, but unpleasantly high levels of ester production are particularly
associated with Hansenula spp., now recognized as nitrate-utilizing species
of Pichia, e.g. P. subpelliculosa.

7.5 ELIMINATION OF WILD YEASTS

It is obvious that to eliminate wild yeasts from brewery fermentations or

beer, the brewer must know the source of the contamination. While the
original source must obviously be from outside the brewery, and reinfec-
tion from that external source must always be a possibility, subsequent
instances of contamination are likely to be one or more of the internal
sources:
1. pitching yeast;
2. equipment surfaces;
3. air;
4. water.
Raw materials added directly (priming sugar, dry hopping) are a possi-
ble, but rare, cause of infection (Wiles, 1953). The nature of the wild yeast,
fermentative or oxidative, can be a clue to its source, e.g. aerobic yeasts are
often associated with airborne infection from grain dust or growth on the
dilute medium created by wort or beer rinsings.
It is unlikely that boiled and cooled wort is the initial source of
contamination, except perhaps in the unfortunate combination of a faulty
heat-exchanger plate and a contaminated water supply. The most likely
result would be bacterial contamination, although introduction of a wild
yeast is possible. Pipework leading to the fermentor, or the walls of the
206 Wild yeasts in brewing and distilling
fermentation vessel itself, are possible sources of infection from the previ-
ous fermentation. Plant must be of perfectly smooth internal construction,
and both mains and vessels must be cleaned and sterilized immediately
after every fermentation (Chapter 10), subsequently confirmed by the
standard microbiological tests (Baker, 1991).
With routine microbiological examination of each batch of pitching
yeast, any development of microbial contaminants is obvious long before
reaching unacceptable levels. The concept of 'unacceptable level' of wild
yeast contamination is interpreted differently between breweries, and must
also take into account the potential of the contaminant to outgrow the
culture yeast. Normally the yeast is pitched before microbiological results
are available, but with the exception of killer yeast, contaminants develop
only slowly from one fermentation to the next. Most breweries routinely
replace yeast cultures after a fixed number of successive fermentations, to
prevent any substantial development of contaminants. If no new culture is
ready to replace unacceptably contaminated yeast, washing with phos-
phoric or other acid to reduce the pH to 2.8-3.0 for 2 h is often a satisfactory
remedy. Bacterial contaminants are less resistant than S. cerevisiae to low
pH, but unfortunately many wild yeasts are at least as tolerant of low pH
as brewing yeasts. Indeed there is a risk that their proportion may increase
as a result of acid washing, which is therefore not necessarily an effective
treatment for removal of wild yeast contamination from pitching yeast. It
is often preferable to tolerate a low but stable level of contamination rather
than to undertake potentially troublesome elimination of wild yeast. Yet
even with that attitude, it is essential that contamination is kept to an
acceptably low level by good hygiene.
Post-fermentation contamination is possible from various sources; rou-
tine microbiological sampling at each stage of post-fermentation process-
ing is essential to ensure satisfactory quality (Hjortshoj, 1984).
Pasteurization is primarily to eliminate the few surviving culture yeasts
rather than spoilage organisms, but excessive numbers of either in the
filtered beer presented to the pasteurizer will overcome the system and a
proportion of these organisms will survive as contaminants. Micro-
biological quality of beer leaving filters and pasteurizers must be checked
by membrane filtration; kegs, bottles and cans, and lids, crown caps and
jet water must be free of spoilage organisms, and the quality of the
packaged product must be monitored by membrane filtration and forcing
tests. Standard microbiological tests required have been described by Moll
(1981) and Baker (1991). At the final stage, the dispense system is also a
possible source of contamination requiring careful surveillance (Harper,
1981).
Elimination of detected spoilage often requires extensive investigation
for possible sources. Ultimately the elimination of wild yeast con-
tamination depends on the experience and good sense of the brewer and
microbiologist.
 References 207

REFERENCES

Andrews, J. and Gilliland, RB. (1952) Journal of the Institute of Brewing, 58, 189.
Anonymous (1994) API 20C Analytical Profile Index, BioMerieux UK, Basingstoke.
Baker, CD. (1991) Recommended Methods of Analysis, Institute of Brewing, London.
Barnett, J.A. (1992) Yeast, S. 1.
Barnett, J.A., Payne, R.W. and Yarrow, D. (1990) Yeasts, Characteristics and
Identification, 2nd edn, Cambridge University Press, Cambridge.
Brightwell, R (1972) Proceedings of the 12th Convention of the Australian and New
Zealand Section of the Institute of Brewing, Institute of Brewing, London, p. 209.
Campbell, I (1987) In Brewing Microbiology (eds F.G. Priest and 1. Campbell), Elsevier,
London.
Campbell, 1. and Msongo, H.5. (1991) Journal of the Institute of Brewing, 97,279.
DiMichele, L.J. and Lewis, M.J. (1993) Journal of the American Society of Brewing
Chemists, 51,63.
Gilliland, RB. (1967) Brewers' Guardian, 96 (December), 37.
Gilliland, RB. (1971) Journal of the Institute of Brewing, 77, 276.
Hall, J.F. (1971) Journal of the Institute of Brewing, 77,513.
Harper, D.R (1981) Brewers' Guardian, 110 Guly), 23.
Hjortshoj, B. (1984) European Brewery Convention Symposium on Quality Assurance,
Zoeterwoude, IRL Press, Oxford, p. 65.
Ingledew, W.M. and Casey, G.P. (1982) Brewers' Digest, 57 (March), 18.
Johnston, J.R and Reader, H.P. (1983) In Yeast Genetics: Fundamental and Applied
Aspects (eds J.F.T. Spencer, D.M. Spencer and A.RW. Smith), Springer, New
York, p. 205.
Kreger-van Rij, N.J.W. (1984) The Yeasts, a Taxonomic Study, 3rd edn, Elsevier/
North-Holland, Amsterdam.
Martini, A.V., Martini, A. and Cardinali, G. (1993) Antonie van Leeuwenhoek, 63, 145.
Maule, A.P. and Thomas, P.D. (1973) Journal of the Institute of Brewing, 79, 137.
Meaden, P.G. (1990) Journal of the Institute of Brewing, 96, 195.
Moll, M. (1981) Journal of the Institute of Brewing, 87,303.
Molzahn, S.W. (1993) Proceedings of the 25th Congress of the European Brewery
Convention, Oslo, IRL Press, Oxford, p. 203.
Nykanen, L. and Suomalainen, H. (1983) Aroma of Beer, Wine and Distilled Beverages,
Riedel, Dordrecht, New York and London.
Pedersen, M.B. (1983) Proceedings of the 19th Congress of the European Brewery
Convention, London, IRL Press, Oxford, p. 457.
Pettipher, G.L., Mansel, R, McKinnon, CH. and Cousins, CM. (1980) Applied and
Environmental Microbiology, 39, 423.
Pfenninger, H.B. (1977) Journal of the Institute of Brewing, 83, 109.
Put, H.M.C and de Jong, J. (1980) In Biology and Activities of Yeasts (eds F.A. Skinner,
S.M. Passmore and RR Davenport), Academic Press, London, p. 18l.
Rainbow, C (1981) In Brewing Science, Vol. 2 (ed. J.RA. Pollock), Academic Press,
London, p. 491.
Simpson, W.J., Fernandez, J.L. and Hammond, J.RM. (1992) Journal of the Institute
of Brewing, 98, 33.
Speers, RA., Tung, M.A., Durance, T.D. and Stewart, G.G. (1992) Journal of the
Institute of Brewing, 98, 525.
Sponholz, W.-R (1993) In Wine Microbiology and Biotechnology (ed. G.H. Fleet),
Harwood, Chur, p. 395.
Stewart, G.G. (1975) Brewers' Digest, 50 (March), 42.
Stratford, M. (1992) Advances in Microbial Physiology, 33, 1.
Taylor, G.T. and Marsh, A.S. (1984) Journal of the Institute of Brewing, 90,134.
208 Wild yeasts in brewing and distilling
Thompson, A.N., Wright, D.M., Pawson, E.C. and Meaden, P.G. (1995) In Proceed-
ings of the Fourth Aviemore Conference on Malting, Brewing and Distilling 1994
(eds 1. Campbell and F.G. Priest), Institute of Brewing, London, p. 213.
Van Oevelen, D., Spaepen, M., Timmermans, P. and Verachtert, H. (1977) Journal of
the Institute of Brewing, 83, 356.
Wickerham, L.J. (1951) Technical Bulletin no. 1029, US Department of Agriculture,
Washington, DC.
Wiles, A.E. (1953) Journal of the Institute of Brewing, 59, 265.
 CHAPTER 8

Rapid detection of
microbial spoilage
1. Russell and T.M. Dowhanick

8.1 INTRODUCTION

Minimization of the length of time required to detect the presence of

spoilage micro flora is a paramount concern in the food and beverage
industry, for failure to keep such microorganisms to a minimum practicable
level can lead to significant economic losses as a result of recalls, reduced
shelf life and inconsistency in product quality. In the brewing industry, the
last decade or more has witnessed the development of a plethora of novel
or improved methods for the detection of microorganisms. The reasons for
this include:
1. increased consumer awareness on product quality;
2. tightened government regulations;
3. increased competition among brewers due to declining consumption;
4. the increasing trend to avoid pasteurization of packaged beer;
5. technological advancements.
It has been fortuitous for the brewer that most microorganisms, partic-
ularly those that are pathogenic to humans, do not survive in a typical beer
due to factors such as the presence of alcohol (usually in the range of 4-5.5%
ethanol), subphysiological pH, anti-microbial hop components and an
anaerobic environment. As a result, the range of microorganisms generally
found in a brewery is relatively few in comparison to the diverse array of
microbia found in the medical, environmental, cosmetic or food and dairy
industries. Table 8.1 lists the genera of Gram-positive and Gram-negative
bacteria, along with wild yeasts which have been detected in various
breweries (lngledew, 1979; Rainbow, 1981; Back, 1987; Dowhanick, 1990;
Stenius et al., 1991; see also Chapters 5, 6 and 7).

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and I. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412 59150 2
210 Rapid detection of microbial spoilage
Table 8.1 Bacteria and yeast genera found in breweries
Gram-positive Gram-negative Wild yeasts
Bacillus Acetobacter Brettanomyces
Lactobacillus Acinetobacter Candida
Leuconostoc Alcaligenes Cryptococcus
Micrococcus Citrobacter Debaryomyces
Pediococcus Enterobacter Endomyces
Streptococcus Flavobacterium Hansenula
Gluconobacter Kloeckera
Hafnia Pichia
Klebsiella Rhodotorula
Megasphaera Saccharomyces
Obesumbacterium Torulopsis
Pectinatus Zygosaccharomyces
Proteus
Pseudomonas
Zymomonas

Detection techniques mayor may not require microbial growth. For those
requiring growth, selective conditions may be necessary to suppress the
proliferation of desirable microorganisms such as brewer's yeast, in order
to detect the presence of undesirable ones such as potential beer spoilage
microorganisms. As described by Kyriakides and Thurston (1989), there are
two broad groups of critical control points for microbiological sampling
found in the brewing process. The first group encompasses locations where
brewing yeasts are present, and where selective conditions (i.e. excluding
the brewing yeast strain being employed) are required to detect contam-
inants. Examples of such critical control points include the yeast slurry,
fermentation and ageing. The second group consists of locations where any
microorganism should either be completely absent or present only in very
low numbers. Such critical control points include raw materials, bright beer,
finished product and strategic surfaces of process machinery such as filler
heads. In the latter group, relatively non-selective conditions are employed
to detect as broad a range of contaminating organisms as possible.
Detection implies establishing the presence or absence of particular
microbes. Traditionally, microbial detection methods in a brewery most
often begin with streaking or inoculating a sample to a specific medium (of
a selective or non-selective nature), followed by incubation of the medium
under defined conditions of temperature and environment (aerobic, anaer-
obic or CO2 enriched) for a lengthy but specified period of time. While
classic detection techniques usually result in accumulation of the required
information, they are too slow to benefit the brewer in time to implement
corrective measures. It is with this in mind that improvements in the speed
of microbial detection and identification procedures would be very useful
to the brewer. Yet, considerable caution must be exercised when the term
 Impedimetric techniques (conductance, capacitance) 211
'rapid' is prefixed to any microbial detection method. In any detection
scenario where a growth period is required in order for microbes to reach
a particular titre, it must be remembered that microorganisms have their
own specific growth rates. These rates may be very slow compared to
commonly encountered clinical species such as Escherichia coli. While a
rapid detection test for E. coli, with a mean generation time of 20-30 min,
might require only a few hours incubation, the same rapid test for fastidious
growers such as Pediococcus damnosus or Lactobacillus lindneri, with doub-
ling times in excess of 5 h, could require a minimum period of several days
(Barney and Kot 1992).
With the relatively recent advances in molecular biology and genetic
biotechnology, many techniques have been developed to characterize or-
ganisms at the molecular level and to study their genetic expression under
a multitude of conditions. These techniques have been applied to the
clinical, food and beverage areas, and to the environmental fields in order
to achieve one or both of the following goals:
1. the detection of problematic organisms;
2. the identification and/or differentiation of organisms at the genus,
species or subspecies level.
This chapter will offer an overview of several detection techniques. It
will become apparent that some of the detection techniques are also iden-
tification techniques. For example, specific probes such as monoclonal
antibodies or oligonucleotides not only detect the presence of a particular
microorganism in, for example, a yeast slurry, but due to the high specificity
of detection, these probes can also identify the particular microbe. Simi-
larly, it can be argued that identification of that same contaminant in the
yeast slurry signifies the detection of the contaminant. This type of meth-
odology is in stark contrast to the use of identification kits such as API and
others which require a pure isolate in order to perform the identification.
Therefore, although this chapter is dedicated to detection techniques, some
of the same techniques are also discussed in Chapter 9 but in the latter case
the emphasis is on identification.

8.2 IMPEDIMETRIC TECHNIQUES (CONDUCTANCE,

CAPACITANCE)

Impedimetric techniques are based on the observation that growth-related

microbial activity can be measured in culture media over time by moni-
toring changes in the chemical and ionic composition in the medium
(Harrison and Webb, 1979; Kilgour and Day, 1983; Firstenberg-Eden and
Eden, 1984; Evans, 1985; Schaertel, Tsang and Firstenberg-Eden, 1987;
Adams, 1988; Adams, Bryan and Thurston, 1989; Kyriakides and Thur-
ston, 1989; Unkel, 1990; Barney and Kot, 1992; Fleet, 1992; Owens,
212 Rapid detection of microbial spoilage
Konirova and Thomas, 1992; Pishawikar, Sinhal and Kulkarni, 1992; Foster,
1994). When a substance, such as a growth medium is subjected to an
alternating electrical current, the impedance (resistance to the flow of the
current through the medium) is affected by the conductance (ability of the
medium to allow electricity to pass through it) and the capacitance (the
ability to store an electrical charge). A pair of electrodes is submerged in
a conducting culture medium to determine total impedance (the vectoral
sum of conductance) which is associated with changes in the bulk ionic
medium, and capacitance, which is associated with changes in close
proximity to the electrodes (Schaertel, Tsang and Firstenberg-Eden, 1987).
In the growth medium, large molecules such as carbohydrates lack elec-
trical charge and consequently this increases the impedance of the medium
while decreasing both the conductance and capacitance. As these un-
charged macromolecules are metabolically broken down by microorga-
nisms into smaller subunits such as bicarbonate, which carries a charge,
the capacitance and conductance begin to increase while the impedance
decreases. Thus, microbial growth can be monitored in the culture me-
dium by measuring either decreases in total impedance, or increases in
either conductance or capacitance.
When monitoring a culture using the appropriate medium, measure-
ments of either impedance or conductance, when plotted against time,
produce a curve similar, but not superimposable, to the respective growth
curve. Impedimetric measurements require growth of microorganisms to
a particular threshold value, which when reached can be accurately
detected and measured by computerized accessories. For these instru-
ments, threshold levels, or detection times (i.e. the time required to reach
the threshold level in which a change in the electrical property occurs),
usually correspond to microbial levels of about 105-106 colony forming
units (CFU) mrl. Since detection times are calculated when threshold
levels are reached, two factors significantly affect the detection times of
healthy c¢lls inoculated into a suitable medium. These are the quantity of
cells initially inoculated and the natural growth rates or generation times
of the respective cultures. The greater the quantity of viable micro-
organisms initially placed into the specialized sample cells, or the shorter
their doubling times, the shorter the detection time. Figure 8.1 shows a
hypothetical growth and conductance response of two microorganisms
inoculated to the same initial concentration, but possessing different
growth rates.
There are a number of advantages to using impedimetric technology.
They include:
1. a time reduction of days in detecting even the most fastidious of
microorganisms when compared to standard plating;
2. some selectivity for microbial growth depending on the medium
employed (see Chapter 9);
 Impedimetric techniques (conductance, capacitance) 213

8 100
f:l.N. tii
f:l.G •
a.
..- Cl
~
'ii
7 <J
CII
CI
.!:!.. t::
II
Z 6 50 ~
<J u
CII
t::
0 II
C 1)
0 5 ::J
ol 'C
.9 0
t::
0

2 4 6 8 10 12 14 16 18
lime (hours)

Fig.8.1 Hypothetical growth and conductance response of two organisms (des-

ignated 1 and 2) inoculated to the same initial concentration but possessing
different growth rates.

3. user-friendly automation for assessing multiple samples for threshold

detection.
The disadvantages include:
1. the instruments are expensive to purchase and operate;
2. due to ionic interferences, not all types of media are amenable to
electrometric analysis.
There are at least two good examples of how impedimetric methods
may be used in the brewery situation. First, they have been used to replace
the standard forcing test for bacterial spoilage by filtering a beer sample
and incubating the filter in brewery wort in a Bactometer (Evans, 1982).
Standard and impedimetric forcing tests were in complete agreement for
'spiked' beers but the latter were obtained in 2-4 days rather than up to
3 weeks for traditional forcing tests. A second brewery application is the
detection of bacterial contamination in pitching yeasts (Kilgour and Day,
1983). Selectivity for bacterial growth was obtained by incubating pitching
yeast samples in WLN broth containing cycloheximide to prevent yeast
growth. Bacterial growth was detected in a Malthus instrument in less
than 18 h depending on the inoculum size (10 6_107 cells ml-! were detected
immediately, but 10 cells mr! required the full 18 h). In practice, this
allowed detection of 100 bacteria per million yeasts in 18 h, a sensitivity
which compares favourably with traditional plating, but in about half the
time.
214 Rapid detection of microbial spoilage
8.3 MICROCALORIMETRY

Calorimetric or heat measurements can be performed by either measuring

changes in quantities of heat by use of a thermometer, or by measuring
thermoelectric effects produced in thermoelements as a result of heat
passage. As microorganisms grow in a culture medium, they generate small
but detectable amounts of heat, primarily as a result of catabolic bio-
chemical reactions. These microfluctuations in heat (typically in the micro-
watt range) can be measured and recorded over time by instruments called
microcalorimeters. Microcalorimeters designed for microbial measure-
ments generally function by passing heat through thermoelectric couples,
producing an electrical current which is proportional to the heat flow. The
thermocouple produces a potential (in the millivolt range) which is ampli-
fied and read off a recorder, producing a thermogram. The beginning of a
thermogram usually occurs during exponential growth of a microbial
culture and increases as the culture continues to grow. As exponential
growth is taken over by the onset of stationary phase, metabolic activity in
the culture diminishes, as does the heat level in the culture (Pishawikar,
Sinhal and Kulkarni, 1992).
Microcalorimeters are capable of detecting as few as 103 actively growing
bacteria per ml and have been used for microbial detection in clinical
applications (Russell et aI., 1975; Beezer et aI., 1978; Herman et aI., 1980) as
well as in the food and dairy industries (Berridge, Cousins and Cliffe, 1974;
Lampa et aI., 1974; Gram and Soggaro, 1985). While microcalorimetry has
been mentioned in brewing literature (Hope and Tubb, 1985; Simpson,
1990), it does not appear that this technique has undergone much testing
or implementation among brewers. However, with the recent increase in
popularity of filter-sterilized beer, microcalorimetry may find application
as an early indicator of microbial contamination in a manner similar to
methods used with ATP bioluminescence.

8.4 TURBIDOMETRY

When microorganisms proliferate in a clear liquid broth, both the turbidity

and absorbance of the broth increase. These changes to the broth can be
detected and measured by a spectrophotometer or turbidometer. If the
initial inoculum size is low and the broth is quite clear, the growth curve
obtained will appear very similar in shape to that produced by impedio-
metry (Firstenberg-Eden and Sharpe, 1991). With the use of an automated
turbidometer (Bioscreen Analysis System, Labsystems, Helsinki, Finland),
microbial growth has been assessed from a variety of food and dairy ma-
trices Gorgensen and Schultz, 1985; Mattila and Alivehmas, 1987; Mattila,
1987) and from beer (Haikara, Mattila-Sandholm and Manninen, 1990).
Haikara, Mattila-Sandholm and Manninen (1990) employed automated
 Flow cytometry 215
turbidometry in order to detect microbial contaminants in pitching yeast.
These contaminants included Lactobacillus casei, Enterobacter agglomerans,
Pediococcus damnosus, Pectinatuscerevisiiphilus and Saccharomyces diastaticus.
As would be expected for microorganisms with naturally diverse growth
rates, detection times varied significantly between the different species. In
general, detection times using the automated turbidometer were
approximately 1 day shorter than the time required to visually discern
turbidity, and the detection limit for these studies was approximately 106
cells mrl. The time saved in using turbidometry, compared to standard
plating, varied from species to species, but ranged anywhere from 2- to
8-fold, or 2-4 days.

8.5 FLOW CYTOMETRY

Flow cytometry combines both microscopic and biochemical analyses in a

single technique to measure microorganisms in liquids. The cells are intro-
duced into the centre of a rapidly moving fluid stream and are forced to
flow single file out of a 50-100 !-lm orifice at a uniform speed (typically
1-10 m S-I). As they do so, the cells are carried past a station where they are
illuminated by a light source (laser) and measurements are made at rates
of 104_106 cells min-I. As the microorganisms in the flow stream pass
through the light beam, the illuminating light is scattered by the cells and
the intensity of light scattered at different angles yields information about
cell size, shape, viability, density and surface morphology (Kruth, 1982;
Salzman, 1982; Muirhead and Horan, 1984; Traganos, 1984).
There have been a number of applications of flow cytometry relevant to
brewing described in the literature (Hutter, 1991; Donhauser et ai., 1992).
Hutter (1992a) demonstrated that immunofluorescent staining could be
used to mark the cells of a particular species of microorganism in order to
facilitate identification and counting. Early work only allowed one species
at a time to be counted, thus not proving very useful to the brewing
industry. Newer developments in fluorescent dyes and dye-antibody con-
jugates, as well as the detection and spectral separation of fluorescent light
in flow cytometers, now allow several species to be stained and counted in
a single operation. Using species-specific antibodies marked with four
different fluorochrome conjugates, Hutter (1992b) demonstrated that four
species in a mixed culture (Saccharomyces cerevisiae, Schizosaccharomyces
pombe, Lactobacillus brevis and Pediococcus damnosus) could be clearly
distinguished and counted.
The major advantage to the method is sensitivity. The major disadvan-
tages include slow processing of large volumes and small particulate debris
interfering with the assay. There is also a high capital cost for multi-purpose
cytometers and a high level of operator training is required.
216 Rapid detection of microbial spoilage
8.6 ATP BIOLUMINESCENCE

An important cellular metabolite which can be used in biochemical assays

in order to detect the presence of microorganisms is adenosine 5'-tri-
phosphate (ATP). ATP is a high energy molecule that is found in all living
organisms. ATP bioluminescence offers the opportunity to significantly
reduce, from days to hours, the time required to identify the presence of
living organisms. This technology is based on detection of the presence of
ATP in samples using enzyme-driven light production, or biolumines-
cence. The ATP molecule can be assayed efficiently by a two-step reaction
employing the luciferin-Iuciferase enzyme reaction, which is the basis of
bioluminescence in fireflies:
Step 1: luciferin + luciferase + ATP + Mg2+ :} (luciferin-Iuciferase-AMP) +
pyrophosphate
Step 2: (luciferin-Iuciferase-AMP) + O2 :} oxyluciferin + luciferase + CO2
+ AMP + light
The amount of light that is produced correlates with the amount of ATP in
the sample. When properly assayed, most of the ATP, but not necessarily
all, is directly indicative of the presence of living organisms in the sample.
There are commercially available reagent kits which can detect as low as
100 yeast cells per sample without enrichment. The required instruments
(luminometers) have limits of detection as low as 1-2 yeast cells per sample.
Thus, one yeast cell in a given sample, with a doubling time of 2-3 h, can
be detected after approximately 20 h incubation. This results in detection
of living microbes in considerably less time than that required for standard
plating techniques. This method is also comparable in cost to standard
plating when factors such as labour and use of disposables such as anaer-
obic system envelopes (when screening for microaerophilic or anaerobic
microorganisms) are included.
The use of ATP-driven bioluminescence for the detection of yeast and
bacteria in non-pasteurized beer or at certain locations along the brew path
has become a useful and practical procedure in the brewing industry
(Hysert, Kovecses and Morrison, 1976; Kilgour and Day, 1983; Dick et al.,
1986; Harrison, Theaker and Archibald, 1987; Krause and Barney, 1987;
Avis and Smith, 1989; Miller and Galston, 1989; Simpson, 1989; Simpson et
al., 1989; Takamura et al., 1989; Barney and Kot, 1992; Ogden, 1993;
Dowhanick and Sobczak, 1994). In the last few years, improvements have
been made both in development of more efficient reagents for the extraction
of ATP from brewery microorganisms (Jago et al., 1989; Simpson and
Hammond, 1989) and in the development of more sensitive luminometers.
A recent survey by Stanley (1992) reported that more than 90 luminometers
from more than 60 companies are currently commercially available, and
some of the more widely used reagents have undergone comparative
evaluations (Griffiths and Phillips, 1989).
 nor>-mlcroblal cell
microbial cell

L.umlnometer
A l
R.adout
(Totol non-microbial RLU)

Mlcroblol & S.lectl ... rei ....

non-microbial 01 ATP from
c.lI. non-microbial call •.
Mlcroblol CIllo
rem.ln Intact

L.umlnomatar
8 Readout
.e'@ -t (Total microbial RLU)

~'0
Mlcrobl.1 & S.lectl"" rei .... Hydrolysl. 01
non-microbial 01 ATP from non-microbial ATP
c.lI. non-microbial c.n.. with ATPese.
Microbial celli Microbial cell.
..,maln Intact remain Intact

L.umlnometar
Readout
j-t I (Total microbial RLU)
•'AA.T? J
:no "AT?J -th!£~•
'~~ ~
bJ"
c •• '11 .. "'- o--'hv"

Selectl.... relea.e Addition 01

~
Microbial lucllerin-
01 microbial ATP
c.lI. luclfe.. se

Fig. 8.2 Measurement of ATP-driven bioluminescence_ (A) Total non-microbial bioluminescence from a mixture of microbial and
non-microbial cells_ (B) Total microbial bioluminescence from a mixture of microbial and non-microbial cells. (C) Total bioluminescence
from microbial cells only_ RLU, relative light units.
218 Rapid detection of microbial spoilage
ATP bioluminescence as a quality control device for hygienic surface
swabbing of process machinery, water analysis or beer analysis is currently
in use in many breweries. As outlined in Fig. 8.2, the assay can be used to
detect the presence of either microbial or non-microbial ATP. The result is
qualitative, but not necessarily quantitative, as unknown samples may
contain numerous species of microorganisms possessing vastly different
ATP levels (e.g. yeasts versus bacteria). Quantitative levels of detection
compared to numbers of organisms observed by plating techniques can be
determined when pure cultures of organisms are assayed under reproduc-
ible conditions. This is because different organisms contain vastly different
levels of ATP, at different phases of growth, on a per cell basis. For example,
a typical yeast cell contains approximately 100 times the amount of ATP
found in a bacterial cell (Hysert, Kovecses and Morrison, 1976). In addition,
bacteria which vary significantly in size (such as an isolate of Lactobacillus
plantarum compared to one of Pediococcus damnosus) will also vary in
relative amounts of ATP per cell. Lastly, especially when conducting sur-
face hygiene or water analyses, not all organisms extracted for ATP will
necessarily develop into colonies when plated on the comparative solid
media. Therefore, employment of bioluminescence for routine detection of
either microbial or non-microbial ATP is useful as a qualitative, or
semi-quantitative assay along the brew path.

8.7 MICROCOLONY METHOD

The microcolony method employs microscopy to detect growing cells

which have not yet reached colony forming units visibly discernible to the
naked eye. This method is essentially a rapid modification to membrane
filtration of a sample followed by incubation of the membrane on a
nutrient medium. In lieu of waiting several days for the growth of
discernible colony forming units, microcolonies are selectively stained and
then visualized microscopically within, at most, 24 h incubation. In the
early stages of microcolony method development, stains such as ponceau
S, janus green, methylene blue and safranin were used for the rapid
detection of yeasts (Richards, 1970; Winter, York and EI-Nakha, 1971;
Barney and Helbert, 1975; Anon., 1975; Portno and Molzahn, 1975). After
the microcolonies were stained, the membrane was dried, heated and
cleared (usually with immersion oil), and then examined by transmitted
light microscopy. An alternative microcolony method using stains was
developed whereby the membrane filters and the cells were selectively
stained rather than the cells alone (Tse and Lewis, 1984). Using this
method, the filter membranes containing microcolonies were stained with
bromocresol green and methyl red. This technique was Originally devel-
oped to distinguish easily between microcolonies of bacteria from water
samples. The bacterial colonies stained dark blue while the membrane
 Microcolony method 219
filter stained pink. As with the direct staining of yeast cells, a notable
disadvantage was that post-stain treatment of the membrane was lethal
to the microcolonies, thereby eliminating the possibility of reincubating
the filter membrane for further study.
Significant improvements to the microcolony technique were made by
use of incident-light fluorescence microscopy, which allowed fluorescent
dyes to be added directly to the nutrient medium. Such dyes, when
incorporated by yeasts and other microorganisms, allowed direct fluores-
cence detection of the living cells without requiring further treatment of
the membrane. While initial difficulties were encountered, as most fluor-
escent dyes were either not incorporated or growth inhibitory to many
microorganisms, certain optical brighteners (i.e. fluorescent whitening
agents found in most washing powders and soaps) were found to be
readily taken up from nutrient media (Paton and Jones, 1971). Harrison
and Webb (1979) utilized optical brighteners derived from 4,4'-diamino-
stilbene-disulphonic acid to develop a microcolony method for yeasts and
moulds. Using the optical brightener Uvitex 2B (Ciba-Geigy), Harrison
and Webb (1979) were able to not only detect fluorescing microcolonies
against a black filter membrane, but were also able to differentiate between
living and dead yeast cells, since the dead cells fluoresced uniformly while
the living cells fluoresced primarily at sites of budding where chitin was
most concentrated. In addition, these optical brighteners were non-inhib-
itory to cellular growth of both yeasts and moulds, and the filter mem-
branes could be reincubated, allowing for continued examination of a
sample over several days and/ or further identification of the microorga-
nisms if desired. The optical brighteners tested, however, were not suitable
for staining bacteria.
The microcolony method was further refined, allowing for detection of
both yeasts (brewing and wild) and brewery bacteria (including members
of the genera Lactobacillus, Pediococcus, Bacillus, Pectinatus, Hafnia and
Escherichia) using the optical brightener Tinopal CBS-X (Ciba-Geigy), a
derivative of distryl-~-phenyl (Haikara and Boije-Backman, 1982). While
some difficulties were observed in correlating aerobic bacteria microcolony
counts with those of standard colonies, the correlation of yeasts and anaer-
obic bacteria using both methods was very good. While the micro colony
technique is useful for the rapid detection of very low numbers of micro-
organisms, examination of the microbes is both laborious and tedious,
resulting in operator fatigue (Haikara and Boije-Backman, 1982; Simpson,
1991; Barney and Kot, 1992). It is also noteworthy that microcolony
methods, as with standard plating, do not give results which necessarily
reflect the ability of proliferating microorganisms on a nutrient medium to
actually grow in and spoil beer (Hope and Tubb, 1985).
220 Rapid detection of microbial spoilage
8.8 DIRECT EPIFLUORESCENCE FILTER TECHNIQUE

The direct epifluorescent filter technique (DEFT) is a method which em-

ploys polycarbonate membrane filtration of samples (for concentration of
microflora) combined with epifluorescent microscopy in order to visualize
individual, fluorochrome-stained cells (Pettipher et al., 1980; Kilgour and
Day, 1983; Pettipher, 1985; Navarro et al., 1987; Simpson, 1990; Barney and
Kot, 1992; Fleet, 1992). This method, which is considerably more sensitive
and rapid than the microcolony method, was initially developed as a
quality control technique for counting bacteria in raw milk (Pettipher et al.,
1980). DEFT can be completed in approximately 30 min as no pre-growth
stage is required, and it has the potential to detect single cells on a mem-
brane filter. Polycarbonate membranes are used for filtration instead of
standard cellulose acetate membranes as they have a much flatter,
smoother surface which is better suited for microscopy (Pettipher, 1985).
Acridine orange has been the most commonly used fluorescent dye for
DEFT because of its ability to bind to nucleic acids and theoretically
differentiate metabolically active or living cells (which stain orange) from
inactive or dead cells (which stain green). The reason for the observed
difference in colour between living and dead cells is that acridine orange
stains orange when it binds to single stranded DNA or to the relatively
abundant levels of RNA which are found in metabolically active cells.
When cells die, these single stranded nucleic acids are rapidly degraded by
nucleases, while double stranded DNA molecules, which naturally fluor-
esce green, are not rapidly degraded in dead cells. However, in studies
conducted on filtered brewery samples, differentiation of viable from
non-viable cells was not successfully achieved, and nondescript stained
debris was a particular problem (Kilgour and Day, 1983; Navarro et al.,
1987; Barney and Kot, 1992). Peladanand Leitz (1991) were able to improve
results using acridine orange by adding a second stain, berberine sulphate,
to the epifluorescent procedure, as the latter dye selectively attaches to the
double stranded DNA of dead cells, giving them a green fluorescence while
remaining impermeable to living cells. Others have reported that when
using fluorescent dyes such as aniline blue, viablue or certain tetrazolium
derivatives, viable and non-viable yeasts can easily and reliably be differ-
entiated from each other (Koch, Bandler and Gibson, 1986; Hutcheson et al.,
1988; Betts, Bankes and Banks, 1989). In addition, some of the nondescript
stained debris can be removed by prefiltering the sample through 12 !lm
pore size polycarbonate membranes (Navarro et al., 1987).
When manually performed, DEFT (as well as the microcolony tech-
nique) is tedious, particularly when screening for low numbers of cells on
large membrane filters. Such analyses can quickly lead to operator fatigue
and, in effect, become an impractical procedure. To offset this difficulty,
an automated version of DEFT can be performed by use of computer-
 Protein fingerprinting by PAGE 221
assisted microscopic scanning of the filter membrane coupled to
computer-enhanced image analysis (Kilgour and Day, 1983; Pettipher
et al., 1989; Simpson, 1990; Barney and Kot, 1992; Fleet, 1992; Niwa, 1993).
Such automation is currently expensive and takes significant periods of
time for computer processing, due to the large number of tasks performed
by the computer (e.g. filter scanning, size separation and debris exclusion,
colour recognition etc.). However, an automated counting system is mar-
keted by Micro-Measurements Ltd (Saffron Walden, Essex, UK) which
removes the problems associated with eye fatigue and reportedly gives
good agreement with visual counts. Moreover, a completely automated
DEFT system has been announced (Cobra 2024, Biocom, France) which
features three computers controlling sample preparation, filtration, drying
and image analysis. A throughput rate of 150 samples per hour is claimed
with improved reproducibility and lower counting limits than manual
systems.

8.9 PROTEIN FINGERPRINTING BY POLYACRYLAMIDE GEL

ELECTROPHORESIS

Microorganisms generally possess two groups of proteins: the constitu-

tively synthesized structural or regulatory 'housekeeping' proteins, and the
differentially regulated polypeptides that are either induced or repressed
as a result of environmental stimuli. The former category of proteins tends
to be found under most conditions conducive to cell viability and growth,
whereas the latter category is regulated by conditions that include the
growth medium, growth temperature, nature of the gaseous environment
and growth phase at the time of harvesting. Based on these observations,
polyacrylamide gel electrophoresis (PAGE) of cellular proteins has been
used in recent years as a means of classifying and identifying yeasts and
other microorganisms (Kersters and De Ley, 1975; Van Vuuren et al., 1981;
Drawert and Bednar, 1983; Van Vuuren and Van Der Meer, 1987, 1988;
Dowhanick et al., 1990; Vancanneyt et al., 1991; Newsom, O'Donnell and
McIntosh, 1992; Chapter 9). In a series of steps, soluble proteins are ex-
tracted from pre-grown cells (quantities as low as single colony isolates)
and subjected to electrophoretic separation. This separation can be made
in one dimension based on differences among proteins in size using sodium
dodecyl sulphate (SDS PAGE) or native PAGE, or based on differences in
ionic charge using isoelectric focusing. Separations can also be performed
in two dimensions by combining size and ionic charge separation in one
gel. The resulting separation pattern or 'protein fingerprint' is specific to
the gene expression of the isolate in question. As such, it can be analysed
for relative differences or similarities to other strains, and based on this
information, the microorganism can be categorized. Visual analysis can be
used to compare the patterns, or densitometer/computer analysis can be
222 Rapid detection of microbial spoilage
used to numerically cluster closely related strains that display only minor
but reproducible differences.
When determining whether or not two or more isolates are essentially
the same, it is very important that the respective isolates are treated in the
same manner, since the cellular inventory (and respective quantities) of
gene products can be significantly altered by differences in growth condi-
tions. Attempts to compare the protein profiles of one organism to another
should include common pre-growth steps immediately prior to harvesting
and preparation of cells. Once pre-growth of cells has been completed,
extraction, electrophoretic separation and staining of proteins can be
accomplished within the span of a few hours.

8.10 IMMUNOANALYSIS

Immunoanalysis has been considered a useful means of identifying con-

taminating microorganisms along the brew path because of its potential to
detect or identify microbes in either a semi-quantitative or directly quanti-
tative way (Campbell and Allan, 1964; Tsuchiya, Fukazawa and Kawakita,
1965; Dolezil and Kirsop, 1975; Hutter, 1978). Using polyclonal or mono-
clonal antibodies, different assays have been designed to differentiate or
even speciate microorganisms from each other.
Rapid, semi-quantitative 'sandwich' immunoassays can be used to
colorimetric ally differentiate or identify relatively abundant (> 104) num-
bers of intact cells. Such assays employ secondary antibodies to which
enzymes have been attached (e.g. alkaline phosphatase, horseradish per-
oxidase). Figure 8.3 is an example of a sandwich enzyme-linked immuno-
sorbent assay (ELISA). Typical ELISAs employ a matrix such as a filter
membrane, dipstick, test tube or microtitre plate coated with antibody
specific to the desired antigen of interest. The sample is placed in contact
with the immobilized antibody, and the antigen is allowed to bind to the
matrix-bound antibody. The matrix is washed (leaving the bound antigen
attached), and a second antigen-specific antibody-enzyme conjugate is
allowed to bind to the matrix-bound antigen. The unbound secondary
antibody is removed and a substrate is added to the matrix. This substrate
is colorimetric ally altered by the enzyme attached to the secondary anti-
body. This results in a semi-quantitative colorimetric change to the matrix
if the specific antigen, such as a beer spoilage microorganism, is present.
ELISA kits have been designed and are being used extensively for envi-
ronmental testing (for pesticides and other compounds), in at-home preg-
nancy testing kits, workplace drug screening and for AIDS testing. In the
food and beverage industry, the main application for these ELISA kits is
in screening for organisms such as Salmonella and Listeria. To date such
kits are not yet commercially available for beer spoilage microorganisms,
but research has been undertaken (Umesh-Kumar and Nagarajan, 1991).
 Immunoanalysis 223

y Step 1 -
Primary antibody is adsorbed to
a solid matrix (eg. test tube,
dipstick, etc.) prior to use.

Step 2 -
Bacteria in the sample bind to
the primary antibody

Step 3 -
Sample is washed off and
enzyme-linked secondary antibody
is bound to a different site on
the bacteria.

Step 4 -
Excess secondary antibody is washed
away and color substrate is added.
Enzyme-substrate interaction results
in color development proportional to
number of bacteria bound.

Fig. 8.3 Enzyme linked immunosorbent assay.

The major advantages these kits will offer is their low cost, speed and
portability.
Antibodies have been designed and employed to identify brewery
microorganisms (Legrand and Ramette, 1986; Phillips and Martin, 1988;
Yasui, Taguchi and Kamiya, 1989, 1992; Hutter, 1992a,b; Vidgrenet ai., 1992;
Whiting et ai., 1992; Gares et ai., 1993; Laurent et ai., 1993). In these cases,
single or very low numbers of cells (including microcolonies) have been
visualized by immunofluorescent microscopy. Typically, specific anti-
bodies are allowed to bind to contaminant microbes. After removal of
unbound specific or primary antibodies, secondary 'indicator' antibodies
(such as antibodies specific to the type of immunoglobulin from which the
primary antibody is derived) to which fluorochromes (e.g. fluoroscein
thiocyanate) have been attached are added. Unbound secondary antibodies
are removed, and microscopic observation of samples under ultraviolet
illumination is used to identify the presence of target microorganisms
which are seen as brightly fluorescing cells.
As with many techniques requiring microscopy, eye fatigue is often a
224 Rapid detection of microbial spoilage
problem. It is also necessary to select antibodies that are specific to easily
accessible (usually surface) cellular components which are abundant
regardless of the growth conditions of the cells (e.g. antibodies raised
against cytosolic glucose-repressed or stress-induced gene products would
be of limited use). However, as with the use of any specific probe, once the
potential problems of cross-reactivity of antibodies to other sources have
been resolved, the technique offers fast, qualitative and quantitative anal-
ysis of brewery samples without requiring incubation periods for microbial
growth to obtain threshold detection values.

8.11 HYBRIDIZATION USING DNA PROBES

Deoxyribonucleic acid (DNA) is the genetic material found in almost all

living organisms, with the exception of certain viruses where the genetic
material is ribonucleic acid (RNA). The biological properties which essen-
tially define all organisms are manifest in sequences of nucleotides, which
consist of four bases: adenine, cytosine, guanine and either thymine for
DNA or uracil for RNA. The nucleotide sequences found in strains from
different species are highly specific and remain, for the most part, constant
from generation to generation. Given the magnitude of roughly millions of
bases of DNA found in a typical yeast or bacterium, this detection/
identification technique is capable of readily distinguishing specific se-
quences of anywhere from less than 100 nucleotides to several thousand
nucleotides within total genomes.
Hybridization (reviewed extensively by Wetmur (1991» is described as
the formation of a double-stranded nucleic acid (either DNA to DNA or
DNA to RNA) by base pairing between single-stranded nucleic acids
derived (usually) from different sources. DNA-DNA hybridization and
DNA-RNA hybridization techniques have been developed and utilized
primarily by molecular biologists to isolate, characterize and study the
expression of genes (DNA) and gene transcripts (RNA). Of particular
interest when performing hybridizations is the ability to control the level
of specificity of base pairing by altering conditions of the hybridization,
such as temperature or salt concentrations. When conditions are used to
maximize the stringency of the reaction, hybridization becomes a highly
selective tool to either differentiate minute differences in DNA sequences
that can be associated within and between species, or identify target
sequences found specifically within the genetic information of certain
organisms. The DNA probes employed are sensitive and can detect 104_107
organisms per test sample. Few problems with cross-reactivity are
encountered once the system has been optimized. The advantage of the
DNA-DNA hybridization assay, as compared to PAGE or immuno-
analysis, is that detection of the target organism is not dependent on the
products of gene expression, which as already discussed can vary with
 Hybridization using DNA probes 225

Step 1 - Step 2 - Step 3 -

Collection of Cell lysis and Binding of DNA
organisms on a DNA strand to filter matrix
filter matrix separation

Step 4 - Step 5 -
Addition of labeled Hybridization of labeled
DNA Probes probes to complementary
DNA from organisms

Fig. 8.4 DNA-DNA hybridization test.

growth conditions, physiological state of the organisms in the test sample

etc.
Hybridization assays can be performed using single colony isolates, cells
collected on membrane filters or purified nucleic acid digested with restric-
tion endonucleases and size separated by gel electrophoresis (i.e. restriction
fragment length polymorphisms (RFLPs)). Diagnostic assays based on
DNA hybridization include the following:
1. propagation of the organism from the test sample to sufficient titre;
2. DNA release from the target organism;
3. a DNA probe specific for the organism;
4. a hybridization format;
5. labelling of DNA probes and detection of the resultant hybrids.
Figure 8.4 illustrates a typical assay which is described in more detail in
Chapter 9.
A variety of DNA probes have been used to identify, differentiate or
characterize yeasts and bacteria relevant to the brewing industry. These
probes include genes derived from specific microorganisms (e.g. the HIS4
or LE U2 genes from Saccharomyces cerevisiae or the S-layer protein gene from
Lactobacillus brevis), endogenous plasmids (such as those found in
Pediococcus damnosus), transposable or ty elements, arbitrary repeated se-
quences such as poly-GT, or even viral DNA such as the single-stranded
phage M13 (Decock and Iserentant, 1985; Martens, van den Berg and
226 Rapid detection of microbial spoilage
Harteveld, 1985; Pedersen, 1983, 1985, 1986; Steffan et al., 1989; Walmsley,
Wilkinson and Kong, 1989; Delley, Mollet and Hottinger, 1990; Meaden,
1990;Colminetal., 1991; Lonvaud-Funeletal., 1991; Lonvaud-Funel,Joyeux
and Ledoux, 1991; Miteva, Abadjieva and Stefanova, 1992; Reyes-Gavilan,
et al., 1992; Vidgren et al., 1992; Lonvaud-Funel, Guilloux and Joyeux, 1993).
While this technology is not yet user friendly enough to be employed as a
routine quality control test in the brewery quality control laboratory, the
potential to produce hybridization kits does now exist. Using hybridization
techniques and DNA probes, several commercial diagnostic kits are al-
ready available to the food industry. These kits are designed to detect
microorganisms such as Salmonella or Listeria with high specificity.

8.12 KARYOTYPING (CHROMOSOME FINGERPRINTING)

Karyotyping is the determination of chromosomal size and number.

Karyotyping has been used for many years to either characterize, differen-
tiate or identify eukaryotic organisms. In higher eukaryotes such as animal
and plant species, karyotyping can easily be performed by selectively
staining DNA in situ (e.g. by use of the Feulgen reaction) and then viewing
the clearly discernible chromosomes under a light microscope. The indi-
vidual chromosomes can then be sorted by micromanipulation and identi-
fied. In lower eukaryotes such as yeast, the chromosomes, while still being
considered as very large molecules, are too small in size to be readily
karyotyped by selective staining and light microscopy. In yeast, karyo-
typing is achieved by the electrophoretic separation of whole chromosomes
through an agarose gel. Conventional one-dimensional agarose gel electro-
phoresis, in which DNA fragments approximately 1/100 to 1/1000 the size
of a yeast chromosome move through a uniform electric field, cannot
resolve large DNA molecules such as yeast chromosomes, since such large
molecules tend to migrate independently of their size. By applying two
orientations of an electric field it is possible to karyotype yeast chromo-
somes, since smaller chromosomes have the ability to respond more
quickly to changes in the electric field than larger ones (Bustamante,
Gurrieri and Smith, 1993). Such changes in electric field to size separate
large DNA molecules form the basis of pulsed-field gel electrophoresis
(PFGE).
When chromosomes separated by PFGE are stained with ethidium
bromide, they appear as discrete bands under ultraviolet light. Various
methods for chromosome fingerprinting are available, with the primary
difference being the orientation of the alternating fields during electro-
phoresis (Fig. 8.5). As with protein fingerprinting, computer-designed
cluster analyses can be used to differentiate and/ or speciate sample yeast
isolates.
Karyotyping of Saccharomyces chromosomes has been well documented
 Karyotyping (chromosome fingerprinting) 227

A-
: '~";::..';:~': PFGE - (Pulsed Reid Gel Electrophoresis)
s- ~ "-.
:
i: s+ Employs 2 altemating fields,
one homogeneous,
:. : one non-homogeneous.

OFAGE - (Orthogonal Reid Altematlng

Gel Electrophoresis)
Employs 2 non-homogeneous altemating
fields

e- • • A- TAFE (Transverse Altematlng Reid

At ....
... (1j
~:n~. . -..s+ Electrophoresis)
The electric field is oriented tmnsversely
to the gel
Gel

RGE (Reid Inversion Gel Electrophoresis)

A uniform electric field is periodically inverted
in one dimension using a 180 reorientation
0

angle.

CHEF (Clamped Homogeneous

Electr1c Reid)
The 24 electrodes are arranged in a
hexagonal contour which offers
reorientation angles of 60 or 1200

Crossed Reid and Rotating Reid

G
Electrophoresis
Employs a single homogeneous field and
changes the orientation in relation to the
gel by discontinuously and periodically
rotating the gel
+

·•••••
.
_\+: ..=l : +\_
PHOGE (Pulsed Homogeneous
Orthogonal Gel Electrophoresis)
·• I - •. Employs a field orientation angle of 90 0

•••••
+

Fig. 8.5 Examples of common gel electrophoresis systems for karyotyping yeasts.
228 Rapid detection of microbial spoilage
(De Jong et al., 1986; Johnston and Mortimer, 1986; Casey, Xiao and Rank,
1988; Degre et al., 1989; Casey, Pringle and Erdmann, 1990; Vezinhet,
Blondin and Hallet, 1990; Querol, Barrio and Ramon, 1992), and has been
utilized in the brewing industry not only as a research and development
tool, but from a quality control standpoint as a means of differentiating or
fingerprinting ale and lager production strains. This technique has been
useful in differentiating closely related lager yeast strains through observed
chromosomal polymorphisms, which by classical means of characteriza-
tion (e.g. giant colony morphology, sporulation, growth characteristics etc.)
have often been difficult or even impossible to differentiate (Casey, Pringle
and Erdmann, 1990; Good et al., 1993). Such chromosomal polymorphisms
are the result of insertions, deletions and translocations of DNA fragments
large enough (typically 10 kilobase pairs or greater) to be electrophoretic-
ally discerned. Pre-grown cells can be prepared and electrophoretically
separated by PFGE within 48 h, although more than one run may be
required to optimally separate chromosomes of similar size.

8.13 POLYMERASE CHAIN REACTION

The polymerase chain reaction (PCR) (Mullis et al., 1986) is one of the newest
and potentially most promising techniques to allow rapid detection of the
presence of brewery microbial contaminants. PCR is a widely used in vitro
method for amplifying very small amounts of selected nucleic acids (DNA
or RNA) by several orders of magnitude over a short period of time (hours).
PCR has revolutionized DNA technology by allowing any nucleic acid
sequence to be simply, quickly and inexpensively generated in vitro in both
great abundance and at a discrete length. Several publications describe the
use of PCR as a rapid detection method for identification of pure cultures of
foodborne pathogens (Bej et al., 1990a,b; Bessesenet al., 1990; Boddinghaus
etal., 1990). Problems have however been encountered when PCR is applied
directly to various foods due to losses in sensitivity (Grant et al., 1993).
At the molecular level, the PCR procedure consists of repetitive cycles
of DNA denaturation, primer annealing, and extension by a highly thermo-
stable DNA polymerase. This process is illustrated in Fig. 8.6. Two short
DNA probes (also called primers) of typically 15-25 nucleotides flank the
DNA segment to be amplified and are repeatedly heat denatured,
hybridized to their complementary sequences, and extended with DNA
polymerase. The two primers hybridize to opposite strands of the target
sequence, such that synthesis proceeds across the region between the
primers, replicating that DNA segment. The product of each PCR cycle is
complementary to, and capable of binding, primers so the amount of DNA
synthesized is doubled in each successive cycle. The original template DNA
can be in a pure form and as a discrete molecule or it can be a very small
part of a complex mixture of biological substances. After 20 replication
 Polymerase chain reaction 229

A F

I
Target DNA
5' 3'
3' 5' Denaturation
94
p
B Cycle 1
! 72
Denaturation :::I
5' 940C
3' i 55
I
E
3' 5' {!
30

C Primer Annealing
SSOC 1 min 1 min 2 min 1 min
5' 3'
+-- +-- Cycle 1 --+
-.+ 5'
3'

D Primer
extension
5'
3'
720C
5'
3' G
5' 3' c
3' 5' .g
Sc
E Cycle 2
•uc
0
5' 3' U
+-- c(
z
3'-'+ 5' 0

5' 3'
+-- Number of cycles

3'
-.+
5'

~
25-40 cycles

Fig.8.6 Polymerase chain reaction. Target DNA (A) is heat denatured (B) at 94°e.
Primers are annealed (C) at 55°C and then primer extension (D) proceeds at 72°e.
The cycle (A-D) is then repeated (E) unti125-40 cycles have been completed. (F)
Time-temperature representation of a typical PCR cycle. (G) Quantitation of
amplified DNA product. Copies of amplified DNA increase exponentially as
number of cycles increases.

cycles, the target DNA will (theoretically) have been amplified over a
million-fold. Amplified samples can be size separated by one-dimensional
agarose or polyacrylamide gel electrophoresis, stained with ethidium
bromide and visualized under ultraviolet light.
230 Rapid detection of microbial spoilage
To differentiate microorganisms such as yeasts or bacteria at the genus,
species or subspecies level, primers are designed from comparative se-
quence analysis of variable or highly conserved regions of DNA, most
notably ribosomal DNA (rONA). Different probes can be designed to
produce peR products of various sizes, and the presence or absence of
electrophoretically size-separated bands can be indicative of the presence
or absence of specific genera or species in the test sample. DNA probes
may also be chromogenically 'tagged' to produce peR products of differ-
ent colours for differentiation of subsets of microorganisms (Emburyet al.,
1990; Kropp, Fucharoen and Embury, 1991). It should be possible to
manipulate the specificity of peR so that, if desired, nucleic acid differen-
tiation could be accomplished for Gram-positive or Gram-negative bac-
teria in one scenario, or for Pediococcus damnosus or Obesumbacterium
proteus in another. Such identification could be elucidated at the single cell
level within a yeast slurry, a fermenter sample or virtually any sample
vulnerable to microbial contamination. To date, although in its relative
infancy in the brewing industry, PCR using specific DNA probes has been
reported as effective in identifying different isolates of brewery micro-
organisms (Klijn, Weerkamp and de Vos, 1991; Tsuchiya et al., 1992;
Tsuchiya, Kano and Koshino, 1993a,b; DiMichele and Lewis, 1992; Prest,
Hammond and Stewart, 1994).

8.14 RANDOM AMPLIFIED POLYMORPHIC DNA peR

As already noted, highly specific probes which are carefully designed to

amplify a specific target sequence typically range in size from 15 to 25
nucleotides. An alternative to the use of specific probes is the use of
non-specific or randomly designed probes that range in size from 9 to 12
nucleotides. The use of these smaller, randomly designed or non-specific
probes for peR diagnostics is called random amplified polymorphic DNA
or RAPD. RAPD peR will typically produce several PCR products due to
binding at a considerably greater number of different locations in a genome
than would be observed with a specific probe. These products can then be
electrophoretically size separated to give DNA fingerprints which can be
genus or even species specific (Williams et al., 1991; Waugh and Powell,
1992). RAPD fingerprints are dependent on the sequence of the primers,
the reaction conditions for each cycle of amplification, and the source of the
DNA. This technology has recently been employed to characterize and
differentiate brewery microorganisms including members of Pediococcus,
Lactobacillus, Obesumbacterium spp. as well as ale and lager yeasts (Savard
and Dowhanick, 1993; Savard, Hutchinson and Dowhanick, 1993). In place
of a mixture of oligonucleotide probes, RAPD peR can be performed using
highly repetitive, specific sequences such as poly-GT oligonucleotides. In
the yeast genome, poly-GT sequences are located at the ends of
 References 231
chromosomes and are considered to be 'hot spots' for genetic recombina-
tion (Walmsley, Wilkinson and Kong, 1989).

8.15 SUMMARY

Significant advances have been and are being made in the field of rapid
microbiological methods. Some of the rapid detection techniques discussed
in this chapter still rely on microbial growth until a particular threshold
value is attained. When using such techniques, it must be remembered that
the concept of 'rapid' remains limited by the growth rate, growth condi-
tions and initial inoculum size of the microbes being analysed. The primary
benefit of using these techniques is that elucidation of the presence or
absence of microbes can now be performed in only a fraction of the time
previously required with classic plating techniques. In addition to detection
alone, certain techniques now offer both rapid verification and identifica-
tion of major beer spoilage microorganisms such as Lactobacillus,
Pediococcus, Obesumbacterium and other potentially problematic microbes,
to both the genus and species level. These methods truly constitute a major
breakthrough for the brewing industry.
Rapid methods and automation in the brewing microbiology laboratory
is an area of great potential and many exciting new developments can be
expected in the coming years. Currently some of the techniques reviewed
in this chapter remain better suited for a research and development labor-
atory rather than a quality control laboratory. However, as design of
equipment and reagent kits coupled to advancements in methodology
catch up to the needs of new and old customers alike, these advances in
detection and identification will be able to quickly, efficiently and cost
effectively assist the brewer in maintaining beer of the highest quality and
consistency, through minimization of spoilage brought on by microbial
contamination.

REFERENCES

Adams, M.R. (1988) Analytical Proceedings, 25, 325.

Adams, M.R., Bryan, J.J. and Thurston, P.J. (1989) Letters in Applied Microbiology,
8,55.
Anonymous (1975) Millipore Technical Bulletin ABB07.
Avis, J.W. and Smith, P. (1989) In Rapid Microbiological Methods for Foods, Beverages
and Pharmaceuticals (eds CJ. Stannard, S.B. Petitt and F.A. Skinner), Blackwell
Scientific Publications, Oxford, p. 1.
Back, W. (1987) Brauwelt International, 2,174.
Barney, M.C and Helbert, J.R. (1975) Master Brewers Association ofAmerica Technical
Quarterly, 12, 23.
Barney, M. and Kot, E. (1992) Master Brewers Association of America Technical
Quarterly, 29, 91.
232 Rapid detection of microbial spoilage
Beezer, A.E., Bettelheim, KA., AI-Salihi, S. and Shaw, E.J. (1978) Science Tools, 25, 6.
Bej, A.K, Mahbubani, M.H., Miller, R et al. (1990a) Molecular Cellular Probes, 4, 353.
Bej, A.K., Steffan, RJ., Dicesare, J.L., Haff, L. and Atlas, RM. (1990b) Applied and
Environmental Microbiology, 56, 307.
Berridge, N.J., Cousins, C.M. and Cliffe, A.J. (1974) Journal of Dairy Research, 41, 203.
Bessesen, M.T., Luo, Q., Rotbart, H.A. et al. (1990) Applied and Environmental
Microbiology, 56,2930.
Betts, RP., Bankes, P. and Banks, J.G. (1989) Letters in Applied Microbiology, 9, 199.
Boddinghaus, B., Rogall, T., Flohr, T. et al. (1990) Journal of Clinical Microbiology,
28,175l.
Bustamante, c., Gurrieri, S. and Smith, S.V. (1993) Trends in Biotechnology, 11, 23.
Campbell, 1. and Allan, A.M. (1964) Journal of the Institute of Brewing, 70, 316.
Casey, G.P., Pringle, A.T. and Erdmann, P.A. (1990) Journal of the American Society
of Brewing Chemists, 48, 100.
Casey, G.P., Xiao, W. and Rank, G.H. (1988) Journal of the Institute of Brewing, 94, 239.
Colmin, c., Pebay, M., Simonet, J.M. and Decaris, B. (1991) FEMS Microbiology
Letters, 81, 123.
Decock, J.P. and Iserentant, D. (1985) Proceedings of the 20th Congress of the European
Brewery Convention, Helsinki, IRL Press, Oxford, p. 195.
Degre, R, Thomas, D.Y., Ash, J. et al. (1989) American Journal of Enology and
Viticulture, 40, 309.
De Jong, P., De Jong, C.M., Meijers, R et al. (1986) Yeast, 2, 193.
Delley, M., Mollet, B. and Hottinger, H. (1990) Applied and Environmental
Microbiology, 56, 1967.
Dick, E., Wiedman, R, Lempart, K and Hammes, W.P. (1986) Chemie Mikrobiologie
Technologie der Lebensmittel, 10, 37.
DiMichele, L.J. and Lewis, M.J. (1992) Presented at the Second Brewing Congress of
the Americas, St Louis, 20-24 September.
Dolezil, L. and Kirsop, B.H. (1975) Journal of the Institute of Brewing, 81, 281.
Donhauser, S., Eger, c., Hubl, T. et al. (1992) Brauwelt, 132, 130l.
Dowhanick, T.M. (1990) Brewers' Digest, 65, 34.
Dowhanick, T.M. and Sobczak, J. (1994) Journal of the American Society of Brewing
Chemists, 52, 19.
Dowhanick, T., Sobczak,J., Russell,1. and Stewart,G.G. (1990) Journal of the American
Society of Brewing Chemists, 48, 75.
Drawert, F. and Bednar, J. (1983) Journal of Agriculture and Food Chemistry, 31, 848.
Embury, S.H., Kropp, G.L., Stanton, T.c. et al. (1990) Blood, 76,619.
Evans, H.A.V. (1982) Journal of Applied Bacteriology, 53,423.
Evans, H.A.V. (1985) Food Microbiology, 2,19.
Firstenberg-Eden, R and Eden, G. (1984) Impedance Microbiology, Research Studies
Press, John Wiley and Sons, Toronto.
Firstenberg-Eden, R and Sharpe, A.N. (1991) In Biotechnology and Food Ingredients
(eds 1. Golgberg and R Williams), Van Nostrand Reinhold, New York, p. 507.
Fleet, G. (1992) Critical Reviews in Biotechnology, 12, l.
Foster, A. (1994) Presented at the 60th Annual Meeting of the American Society of
Brewing Chemists, Toronto, Canada.
Gares, S.L., Whiting, M.s., Ingledew, W.M. and Ziola, B. (1993) Journal of the
American Society of Brewing Chemists, 51, 158.
Good, L, Dowhanick, T.M., Emandes, J.E. et al. (1993) Journal of the American Society
of Brewing Chemists, 51, 35.
Gram, L. and Soggaro, H. (1985) Journal of Food Protection, 48, 431.
Grant, KA., Dickinson, J.H., Payne, M.J. et al. (1993) Journal of Applied Bacteriology,
74,260.
Griffiths, M.w. and Phillips, J.D. (1989) In Rapid Microbiological Methods for Foods,
 References 233
Beverages and Pharmaceuticals (eds CJ. Stannard, S.B. Petitt and F.A. Skinner),
Blackwell Scientific Publications, Oxford, p. 13.
Haikara, A. and Boije-Backman, S. (1982) Brauwissenschaft, 35,113.
Haikara, A., Mattila-Sandholm, T. and Manninen, M. (1990) Journal of the American
Society of Brewing Chemists, 48, 92.
Harrison, J. and Webb, T.J.B. (1979) Journal of the Institute of Brewing, 85, 231.
Harrison, M., Theaker, P.D. and Archibald, H.W. (1987) Proceedings of the 21st
Congress of the European Brewery Convention, Madrid, IRL Press, Oxford, p. 168.
Herman, J.P.M., Jakubczak, E., Izard, D. and Leclerc, H. (1980) Canadian Journal of
Microbiology, 26, 413.
Hope, CF.A. and Tubb, RS. (1985) Journal of the Institute of Brewing, 91, 12.
Hutcheson, T.C, McKay, T., Farr, L. and Seddon, B. (1988) Letters in Applied
Microbiology, 6, 85.
Hutter, K-J. (1978) Brauwissenschaft, 31,278.
Hutter, K-J. (1991) Brauwissenschaft, 44, 216.
Hutter, K-J. (1992a) Brauwelt International, 2, 183.
Hutter, K-J. (1992b) Monatsschrift fUr Brauerei, 45, 280.
Hysert, D.W., Kovecses, F. and Morrison, N.M. (1976) Journal of the American Society
of Brewing Chemists, 34,145.
Ingledew, W.M. (1979) Journal of the American Society of Brewing Chemists, 37, 145.
Jago, P.H., Stanfield, G., Simpson, W.J. and Hammond, J.RM. (1989) In ATP
Luminescence - Rapid Methods in Microbiology (eds P.E. Stanley, B.J. McCarthy
and R Smither), Blackwell Scientific Publications, Oxford, p. 53.
Johnston, J.R and Mortimer, RK. (1986) International Journal of Systematic
Bacteriology, 36,569.
Jorgensen, H.L. and Schultz, E. (1985) International Journal ofFood Microbiology, 2,177.
Kersters, K and De Ley, J. (1975) Journal of General Microbiology, 87, 333.
Kilgour, W.J. and Day, A. (1983) Proceedings of the 19th Congress of the European
Brewery Convention, London, IRL Press, Oxford, p. 177.
Klijn, N., Weerkamp, A.H. and de Vos, W.M. (1991) Applied and Environmental
Microbiology, 57, 3390.
Koch, H.A., Bandler, R and Gibson, RR (1986) Applied and Environmental
Microbiology, 52,599.
Krause, J.A. and Barney, M.C (1987) Paper presented at the 53rd Annual Meeting
of the American Society of Brewing Chemists, Cincinnati, Ohio.
Kropp, G.L., Fucharoen, S. and Embury, S.H. (1991) Blood, 78,26.
Kruth, H.S. (1982) Analytical Biochemistry, 125, 225.
Kyriakides, A.L. and Thurston, P.A. (1989) In Rapid Microbiological Methods for Foods,
Beverages and Pharmaceuticals (eds CJ. Stannard, S.B. Pettit and F.A. Skinner),
Blackwell Scientific Publications, Oxford, p. 101.
Lampa, RA., Mikelson, D.A., Rowley, D.B. et al. (1974) Food Technology, 28, 52.
Laurent, F., Drocourt, J.L., Jakobsen, M. et al. (1993) Proceedings of the 24th Congress
of the European Brewery Convention, Oslo, IRL Press, Oxford, p. 487.
Legrand, M. and Ramette, P. (1986) Cerevisia, 11, 173.
Lonvaud-Funel, A., Joyeux, A. and Ledoux, o. (1991) Journal ofApplied Bacteriology,
71,501.
Lonvaud-Funel, A., Guilloux, Y. and Joyeux, A. (1993) Journal ofApplied Bacteriology,
74,41.
Lonvaud-Funel, A., Fremaux, C, Biteau, N. and Joyeux, A. (1991) Food Microbiology,
8,215.
Martens, F.B., van den Berg, Rand Harteveld, P.A. (1985) Proceedings of the 20th
Congress of the European Brewery Convention, Helsinki, IRL Press, Oxford, p. 211.
Mattila, T. (1987) Journal of Food Protection, 50, 640.
Mattila, T. and Alivehmas, T. (1987) International Journal of Food Microbiology, 4,157.
234 Rapid detection of microbial spoilage
Meaden, P. (1990) Journal of the Institute of Brewing, 96, 195.
Miller, Rand Galston, G. (1989) Journal of the Institute of Brewing, 95, 317.
Miteva, V.I., Abadjieva, A.N. and Stefanova, Tz.T. (1992) Journal of Applied Bacteri-
ology, 73, 349.
Muirhead, K.A. and Horan, P.K. (1984) In Advances in Cell Culture, Vol. 3 (ed.
K. Maramorosch), Academic Press, New York.
Mullis, K.B., Faloona, S.J. Scharf, S.J. et al. (1986) Cold Spring Harbor Symposium on
Quantitative Biology, 51,263.
Navarro, A. Moreno, P. Carbonell, J.V. and Pinaga, F. (1987) Proceedings of the 21st
Congress of the European Brewery Convention, Madrid, IRL Press, Oxford, p. 465.
Newsom, LA., O'Donnell, D.C and McIntosh, J. (1992) Proceedings of the 22nd
Convention of the Institute of Brewing (Australian and New Zealand Section),
Melbourne, Institute of Brewing, London, p. 189.
Niwa, M. (1993) Brauwelt International, 11, 149.
Ogden, K. (1993) Journal of the Institute of Brewing, 99, 389.
Owens, J.D., Konirova, L. and Thomas, D.S. (1992) Journal of Applied Bacteriology,
72,32.
Paton,A.M. andJones,S.M. (1971) In Methods in Microbiology, Vol. 5A (edsJ.R Norris
and D.W. Ribbons), Academic Press, London, p. 135.
Pedersen, M.B. (1983) Carlsberg Research Communications, 48, 485.
Pedersen, M.B. (1985) Carlsberg Research Communications, 50, 263.
Pedersen, M.B. (1986) Carlsberg Research Communications, 51, 163.
Peladan, F. and Leitz, R (1991) Proceedings ofthe 23rd Congress ofthe European Brewery
Convention, Lisbon, IRL Press, Oxford, p. 48I.
Pettipher, G.L. (1985) In Rapid Methods and Automation in Microbiology and
Immunology (ed. K.-O. Habermehl), Springer-Verlag, New York, p. 629.
Pettipher, G.L, Mansell, R, McKinnon, CH. and Cousins, CM. (1980) Applied and
Environmental Microbiology, 39, 423.
Pettipher, G.L., Kroll, RG., Farr, L.J. and Betts, RP. (1989) In Rapid Microbiological
Methods for Foods, Beverages and Pharmaceuticals (eds CJ. Stannard, S.B. Pettit
and F.A. Skinner), Blackwell Scientific Publications, Oxford, p. 33.
Phillips, A.P. and Martin, K.L. (1985) In Rapid Methods and Automation in Microbiology
and Immunology (ed. K.-O. Habermehl), Springer-Verlag, Berlin, p. 408.
Phillips, A.P. and Martin, K.L. (1988) Journal of Applied Bacteriology, 64, 47.
Pishawikar, M.S., Sinhal, RS. and Kulkarni, P.R (1992) Trends in Food Science and
Technology, 3, 165.
Portno, A.D. and Molzahn, S.W. (1975) Brewers' Digest, 3,44.
Prest, A.G., Hammond, J.R.M. and Stewart, G.S.A.B. (1994) Applied and
Environmental Microbiology, 60, 1635.
Querol, A., Barrio, E. and Ramon, D. (1992) Systematic and Applied Microbiology, 15, 439.
Rainbow, C (1981) In Brewing Science, Vol. 2 (ed. J.RA. Pollock), Academic Press,
London, p. 49I.
Reyes-Gavilan, CG. de los, Limsowtin, K.Y., Tailliez, P. et al. (1992) Applied and
Environmental Microbiology, 58, 3429.
Richards, M. (1970) Wallerstein Laboratory Communications, 33,97.
Russell, W.J., Farling, S.R, Blanchard, G.C and Boling, E.A. (1975) In Microbiology
(ed. D. Schlessinger), American Society for Microbiology, Washington, DC,
p.22.
Salzman, G.c. (1982) In Cell Analysis, Vol. 1, (ed. N. Catsimpoolas), Plenum, New
York, p. 11I.
Savard, L. and Dowhanick, T.M. (1993) Presented at the 59th Annual Meeting of the
American Society of Brewing Chemists, Tucson, 12-16 June.
Savard, L., Hutchinson, J. and Dowhanick, T.M. (1993) Journal ofthe American Society
of Brewing Chemists, 52, 62.
 References 235
Schaertel, B.J., Tsang, N. and Firstenberg-Eden, R (1987) Food Microbiology, 4, 155.
Simpson, W.J. (1989) Brewers' Guardian, 118,20.
Simpson, W.J. (1990) In Proceedings of the Third Aviemore Conference on Malting,
Brewing and Distilling (ed.1. Campbell), Institute of Brewing, London, p. 16l.
Simpson, W.J. (1991) Brewers' Guardian, 120, 30.
Simpson, W.J. and Hammond, J.RM. (1989) In ATP Luminescence - Rapid Methods
in Microbiology (eds P.E. Stanley, B.J. McCarthy and R Smither), Blackwell
Scientific Publications, Oxford, p. 45.
Simpson, W.J., Hammond, J.RM., Thurston, P.A. and Kyriakides, A.L. (1989)
Proceedings of the 22nd Congress of the European Brewery Convention, Zurich, IRL
Press, Oxford, p. 663.
Stanley, P.E. (1992) Journal of Bioluminescence and Chemiluminescence, 7, 77.
Steffan, RJ., Breen, A., Atlas, RM. and Sayler, G.s. (1989) Canadian Journal of
Microbiology, 35, 68l.
Stenius, V., Majamaa, E., Haikara, A. et al. (1991) Proceedings of the 23rd Congress of
the European Brewery Convention, Lisbon, IRL Press, Oxford, p. 529.
Takamura, 0., Nakatani, K, Taniguchu, T. and Murakami M. (1989) Proceedings of
the 22nd Congress of the European Brewery Convention, Zurich, IRL Press, Oxford,
p.148.
Traganos, F. (1984) Cancer Investigation, 2, 149.
Tse, K-M. and Lewis, CM. (1984) Applied and Environmental Microbiology, 48, 433.
Tsuchiya, T., Fukazawa, Y. and Kawakita, S. (1965) Mycopathologia et Mycologia
Applicata, 26, l.
Tsuchiya, Y., Kano, Y. and Koshino, S. (1993a) Journal of the American Society of
Brewing Chemists, 51, 40.
Tsuchiya, Y., Kano, Y. and Koshino, S. (1993b) Journal of the American Society of
Brewing Chemists, 52, 95.
Tsuchiya, Y., Kaneda, H., Kano, Y. and Koshino, S. (1992) Journal of the American
Society of Brewing Chemists, 50,64.
Umesh-Kumar, S. and Nagarajan, L. (1991) Folia Microbiologica, 36, 305.
Unkel, M. (1990) Brauwelt, 4, 100.
Vancanneyt, M., Pot, B., Hennebert, G. and Kersters, K (1991) Systematic and Applied
Microbiology, 14, 23.
Van Vuuren, H.J.J. and Van Der Meer, L.J. (1987) American Journal of Enology and
Viticulture, 38,49.
Van Vuuren, H.J.J. and Van Der Meer, L.J. (1988) Journal of the Institute of Brewing,
94,245.
Van Vuuren, H.J.J., Kersters, K, De Ley, J. and Toerien, D.F. (1981) Journal of Applied
Bacteriology, 51, 5l.
Vezinhet, F., Blondin, B. and Hallet, J.N. (1990) Applied Microbiology and
Biotechnology, 32, 568.
Vidgren G., PaIva, I., Pakkanen, R et al. (1992) Journal of Bacteriology, 174,7419.
Walmsley, RM., Wilkinson, B.M. and Kong, T.H. (1989) Bio(fechnology, 7, 1168.
Waugh, R and Powell, W. (1992) Trends in Biotechnology, 10,186.
Wetmur, J.G. (1991) Critical Reviews in Biochemistry and Molecular Biology, 26, 227.
Whiting, M., Crichlow, M., Ingledew, W.M. and Ziola, B. (1992) Applied and
Environmental Microbiology, 58, 713.
Williams,J.G.K, Kubelik,A.R, Livak,KJ. etal. (1991) Nucleic Acid Research, 18,653l.
Winter, F.H., York, G.K and EI-Nakha, H. (1971) Applied Microbiology, 22, 89.
Yasui, T., Taguchi, H. and Kamiya, T. (1989) Proceedings of the 22nd Congress of the
European Brewery Convention, Zurich, IRL Press, Oxford, p. 190.
Yasui, T., Taguchi, H. and Kamiya, T. (1992) Proceedings of the 22nd Convention of the
Institute of Brewing (Australian and New Zealand Section), Melbourne, Institute of
Brewing, London, p. 190.
 CHAPTER 9

Methods for the rapid

identification of
• •
mIcroorganIsms
C.S. Gutteridge and F.G. Priest

9.1 WHAT IS IDENTIFICATION?

Identification is correctly defined as an activity leading to the assignment

of unidentified microorganisms to a particular class in a previously made
classification (Norris, 1980). This definition notes the link between identi-
fication and classification, which is the ordering of microorganisms into
groups. It is often emphasized that satisfactory schemes for identification
can only be based on good classifications (Sneath, 1972), although in
identification, speed is imperative and second only in importance to accu-
racy (Steel, 1962). Nevertheless, identification is understood by most micro-
biologists to mean the naming of an unknown culture against a developed
taxonomic scheme. This activity provides a vital framework that allows
microbiologists to recognize when they are working with the same
microorganism and is thus fundamental to the pursuit of the science of
microbiology.
Identification does need to be clearly distinguished from the process of
constructing taxonomic groups. Once groups are known, individual char-
acters can be weighted and used predictively. The production of keys and
diagnostic tables and methods for discriminating between organisms that
are easily confused is a traditional task for the diagnostic microbiologist.
Microbial identification is at its most intense in the clinical diagnostic
laboratory where the daily workload of a medium-sized hospitallabora-
tory may be between 170 and 350 samples (Bascomb, 1980). Only 10-30%
of the specimens tested actually contain a pathogen and these are subse-
quently processed further. Testing for antibiotic sensitivity is the priority,

Brewing Microbiology, 2nd edn. Edited by F. G. Priest and I. Campbell.

Published in 1996 by Chapman & Hall, London. ISBN 0 412591502
238 Methods for the rapid identification of microorganisms
but pathogens are identified because this sometimes aids diagnosis and
treatment, but more often provides epidemiological information to monitor
the spread of disease. It is not surprising that the main developments in
rapid microbiological identification in the past two decades have been
targeted at the clinical laboratory and clinical problems.
Outside of the clinical laboratory, few microbiologists face the same
intensity of identification problems. In industrial laboratories, unless they
are specifically involved in screening for the isolation of organisms with
novel properties, identification tasks are kept to a minimum. The food and
beverage industry spends millions of pounds annually on microbiological
testing, and although this is primarily enumeration, it does include the
detection of specific organisms that are either potential pathogens or are
involved in spoilage or biodeterioration problems.
Jarvis (1985) has commented that the food microbiologist would be
prepared to carry out more identification if time could be saved on enumer-
ation. There is much to be gained from a detailed understanding of the
distribution of microorganisms in the food processing or brewing environ-
ment but surveys are rarely undertaken because of the resources they
divert.
A further aspect of the identification problem in industrial laboratories
is that it is often aimed at key properties and may not be designed along
purely taxonomic lines. For example, in the study of the spoilage of chem-
ically preserved beverages, the recognition that a yeast is preservative
resistant is important. Preservative resistance is not necessarily exclusive
to one genus or species. Thus, recognizing preservative resistance is a type
of pragmatic identification that does not relate to a formally established
taxonomy. For such tasks it is perhaps better to use the term characteriza-
tion to describe the search for key properties in microorganisms. In this
chapter methods for identifying and characterizing microorganisms are
reviewed and techniques that offer rapid identification are described in
more detail. The methods discussed are drawn from all areas of micro-
biology and examples are not restricted to the brewing industry. Future
trends in the identification of microorganisms are considered.

9.2 LEVELS OF EXPRESSION OF THE MICROBIAL GENOME

A suitable scheme for relating the various techniques used to identify

microorganisms was proposed by Norris (1980) and is shown in Table 9.l.
Four levels of expression of the genetic information in microorganisms
can be recognized. At Level 1, genetic information is expressed in terms
of the DNA molecule and the genome of the cell. One of the earliest and
simplest methods to characterize DNA is measurement of the base com-
position of the chromosome in terms of mol% guanine and cytosine (mol%
G + C). This is not particularly useful since two organisms with the same
 Levels of expression of the microbial genome 239
Table 9.1 Techniques for studying the genetic information in microbial cells at
different levels of expression (adapted from Norris (1980»
Level Techniques
1 The genome DNA/DNA, DNA/RNA hybridization
Gene probes
%G+C
Plasmids
Phage
2 Proteins Amino acid sequencing
Electrophoresis
Immunology
3 Cell components Amino acid pools
Light-scattering techniques
Cell wall composition
Chromatography
Spectroscopy
Bacteriocins
Phage
4 Morphology and behaviour Microscopy
Motility
Enzyme tests
Physiology
Nutrition

G + C content need not necessarily be related. For example cows and rats
have about the same chromosomal G + C contents of around 44%. Never-
theless, if two organisms have very different base compositions, differing
by more than 5%, they cannot be related. This negative implication of base
composition can often be useful to exclude certain taxa when identifying
difficult organisms. This is particularly the case if base composition can
be detected rapidly and simply with minimum sample preparation, for
example using differential scanning calorimetry (Miles, Mackey and
Parsons, 1986) or high performance liquid chromatography (HPLC)
(Tamaoka, 1994).
Complete nucleotide sequencing of the genetic material in a cell would
provide an absolute identification but, although it might be feasible with
modem genetic techniques, it is certainly not practical. Separation of the
DNA molecule into single strands followed by hybridization with single
strands from another organism can be used to produce a measure of the
similarity of base sequences between DNAs of different microorganisms
and hence offers a direct measure of relatedness. Indeed, DNA hybridiza-
tion forms the basis of the generally accepted species definition in bacterial
systematics; strains belonging to the same species should possess more than
70% DNA sequence homology (Grimont, 1988). Similar techniques can be
used to compare DNA and RNA. These hybridization techniques have
made an important contribution to microbial taxonomy over the last two
240 Methods for the rapid identification of microorganisms
decades. They have also spawned the development of gene probes (DNA
probes) which are unique nucleotide sequences that will only hybridize
with a specific gene(s) of interest. If the sequence is chosen judiciously its
detection within the genome of a microorganism can be used as an accurate
means of identification (Falkow, 1985; Schleifer, 1990).
At Level 2, the genetic information is expressed as proteins. Amino acid
sequencing, like nucleotide sequencing, is expensive and time consuming,
and is not practical for rapid identification. In contrast, electrophoresis, i.e.
determining the ability of protein molecules to migrate under the influence
of an electric field through a gel, the pores of which are of molecular
dimensions, is a technique with wide applicability to taxonomic and iden-
tification problems. The electrophoresis of whole-protein extracts of micro-
organisms can be used as an identification method which, while it is not
necessarily rapid, can be efficient in terms of sample throughput (Kersters
et al., 1994). Studies on the molecular structure of particular proteins are
valuable for taxonomic studies but, again, are not practical for identi-
fication. Immunology is often directed at proteins and is an important
diagnostic tool.
Level 3 is the cell composition level, which can be addressed by a host
of analytical techniques applied to whole cells or particular cell extracts.
For example, amino acid pools, cell wall structure and cellular and mem-
brane lipids can be profiled using techniques such as gas chromatography
(GC), HPLC and thin layer chromatography (TLC). Whole cells can be
volatilized by pyrolysis and the products detected by pyrolysis gas chro-
matography (Py-GC) or pyrolysis mass spectrometry (Py-MS). Mass spec-
trometry per se can be used to identify precise structures for diagnostic
purposes. In addition to MS, a range of physical techniques has been
examined with a view to microbial characterization, including circular
intensity differential scattering (CIDS), flow cytometry, impedance mea-
surements, microcalorimetry, nuclear magnetic resonance and infrared (IR)
and Raman spectroscopy (Nelson, 1991). Several of these require expensive
and highly specialized equipment and we shall discuss later only the
pyrolysis and IR techniques since these have most promise for strain
recognition and typing. Immunology can also be targeted at non-protein
cell components and sensitivity to bacteriophage and bacteriocins may also
be related to cell structure.
The final level of expression, Level 4, has been called morphology and
behaviour and is the level at which most conventional identification meth-
ods operate. Studies on the presence or absence of particular enzymes,
either directly or by the detection of end-products, come into this category.
It includes all the multi-test systems that are marketed to standardize the
examination of microbial behaviour and to extend the range of tests
available. Morphological studies, usually by microscopy, are also tests
performed within Level 4.
The remainder of the chapter will be devoted to a more detailed
 Identification at the genomic level 241
examination of developments in identification techniques at each of these
four levels. It is, however, important to recognize that the longer term aim
of diagnostic microbiology must be to devise rapid methods that can be
applied generally (i.e. to a wide range of microorganisms).

9.3 IDENTIFICATION AT THE GENOMIC LEVEL

We have noted above the ability of single-stranded DNA to hybridize with

complementary DNA and how this can lead to the concept of the gene
probe (DNA probe) which is surely one of the most powerful diagnostic
tools to be introduced in the past decade. The attraction to the problem of
microbial identification is that a probe consisting of> 20 nucleotides can be
shown to be statistically unique and if it hybridizes with part of the
microbial genome then this can be used to give an identification of
unparalleled accuracy. The basic steps in the nucleic acid probe test are:
1. design of probe;
2. labelling of probe;
3. preparation of target DNA;
4. hybridization;
5. detection of hybrids.
Some of these steps will be addressed briefly here to give the reader an
idea of the processes involved and to stress that they are becoming
increasingly simple. Such hybridizations are carried out by undergraduate
students in the International Centre for Brewing and Distilling and could
be similarly performed in the brewery laboratory with commercial kits
and other proprietary reagents. For a fuller discussion of the techniques,
the reader should consult recent reviews such as those written by Schleifer
(1990), Schleifer, Ludwig and Amann (1993) and Amann and Ludwig
(1994).

9.3.1 The probe

There are two broad classes of probe, those directed to unknown parts of
the chromosome and those targeted to specific genes. The first class is
seldom used today, but the most typical example is the use of chromosomal
DNA itself. For microorganisms conforming to the genomic definition of a
species (i.e. strains showing more than 70% DNA sequence relatedness)
chromosomal DNA can be used as the probe since it should possess less
than 70% (and in practice usually only 30% or less) sequence homology
with DNA from other species of microorganism. Thus the DNA should
hybridize strongly with DNA from members of the same species and
weakly with all other DNA samples. This is generally found to be the case
and by inclusion of suitable control DNA samples, accurate identification
242 Methods for the rapid identification of microorganisms
by hybridization can be obtained (e.g. for the food poisoning bacterium
Campylobacter; Chevrier et al., 1989).
The specificity of chromosomal DNA is insufficient for accurate identi-
fication of closely related bacteria. Identification of such organisms requires
specific gene fragments. One possibility is to clone random pieces of DNA
and test them for specificity. There are many examples of this approach,
and it has been used successfully for Lactobacillus (Delley, Mollet and
Holtinger, 1990; Pilloud and Mollet, 1990) and Streptococcus (Schmidhuber,
Ludwig and Schleifer, 1988).
Probes are most often directed at a particular locus in the target DNA.
This could be the gene for a particular metabolic trait, a virulence factor
(often used in clinical diagnostic microbiology) or a gene such as a ribo-
somal RNA (rRNA) gene. Probes aimed at protein coding genes may be
derived from the cloned gene or, if this is not available but the sequence is
published, an oligonucleotide can be synthesized which is complementary
to part of that gene. One simple way to make a probe to any published gene
sequence is to amplify the gene from chromosomal DNA using the poly-
merase chain reaction (PCR) and to include a labelled nucleotide in the
reaction. In this way large amounts of labelled probe can be made in a single
PCR. Typical probes to virulence factors include those for the haemolysin
of Listeria monocytogenes (Datta et al., 1988) and the enteropathogenic toxins
of Escherichia coli (Sommerfelt et al., 1988).
The 16S and 23S rRNA molecules are particularly suitable as targets for
probes because they are present in high copy number (104_10 5 molecules
per bacterial cell) thus greatly improving the sensitivity of the hybridiza-
tion. Within these genes there are regions which are highly variable and
differ significantly between species. These regions are valuable for species
identification. Other areas in the genes are more conserved and sequences
are identical within all species in a genus. These parts are therefore useful
for generic probes. This combination of factors has led to identification
through rRNA gene probes becoming very popular, and Schleifer, Ludwig
and Amann (1993) list numerous applications of this technology.

9.3.2 Labelling of the probe and detection

One obstacle to the widespread introduction of DNA hybridization meth-
ods into the food industry was that the probes were initially labelled with
radioisotopes for detection purposes. Although for some purposes radio-
isotopes are still preferred, giving greater sensitivity of detection, for most
routine applications simple and robust non-radioactive labelling and de-
tection systems are generally used. Two popular systems available as
commercially produced kits are based on biotin and digoxigenin (a plant
steroid) incorporation into the probe. After the hybridization, labelled
probe is detected by enzyme (e.g. alkaline phosphatase) conjugated avidin
or antibody reactions respectively. The bound antibody is then detected
 Identification at the genomic level 243

colorimetrically. These systems have improved markedly over the past

decade and now offer extreme sensitivity, simplicity of operation and none
of the hazards associated with radioactivity.

9.3.3 The target

The target DNA can be purified directly from colonies on an isolation or
quality control Petri dish of microorganisms. In these 'colony
hybridizations', the microorganisms are removed to a nitrocellulose or
nylon membrane, lysed and the DNA partially purified by treatment of the
membrane with protease and detergent. The DNA is then bound to the
membrane by heat or UV treatment and hybridized to the labelled probe.
After development, hybridizing colonies are revealed.
Colony hybridizations often suffer from high background signals due to
cellular material binding the probe in a non-specific way. In these circum-
stances, it is better to purify DNA from microbial cultures using a routine
procedure. The DNA is then denatured and spotted onto a membrane using
a 'dot-blot' or 'slot-blot' manifold. These pieces of apparatus allow the
immobilization of large numbers of DNA samples on a membrane in a
specific array of dots (or slots). The membrane can then be hybridized with
a specific probe DNA and positive hybridization reactions are revealed as
shown in Fig. 9.l.
An innovative detection/ identification procedure is the reverse dot-blot
hybridization which has particular application in quality control proce-
dures. A sample of beer could be collected and 'forced' to provide microbial
biomass. Total DNA is then prepared and labelled with a non-radioactive
label using a standard kit. The membrane is impregnated with dots of
denatured DNA from reference organisms such as representative lacto-
bacilli or spoilage yeasts. The labelled DNA from the sample is hybridized
with the filter and any bound DNA is subsequently detected. Any hybrids
such as the Lactobacillus brevis DNA showing up indicate that L. brevis was
present in the original sample.
The biggest barrier to the adoption of these hybridization techniques in
the quality control or research laboratory is the perceived difficulty of
working with molecular techniques. We stress that much of the complica-
tion of the techniques has been removed with the introduction of labelling
and detection kits by most of the major biochemical manufacturers. Once a
potential specific target has been identified, if the DNA sequence is available
a probe can be readily prepared by PCR and a highly specific, economically
realistic detection/ identification procedure can be developed.
At present only one line of products for use in food microbiology is
marketed as a complete probe kit: the Gene-Trak Systems (Integrated
Genetics, Framingham, MA, USA). These are based on an ingenious
'dipstick' technology targeted at regions of the 165 rRNA. The probe has
attached to it a 'tail' carrying a fluorescein label. Dipsticks specific for E. coli,
244 Methods for the rapid identification of microorganisms

Fig. 9.1 Example of a slot-blot hybridization reaction. Chromosomal DNA

samples were loaded onto a membrane using a slot-blot manifold and hybridized
with a non-radioactively (digoxigenin) labelled DNA probe. Positive reactions are
evident as dark bands, negative reactions as light (background) bands.

Salmonella and Listeria are currently marketed together with some clinically
relevant bacteria such as Yersinia and Staphylococcus.

9.4 TECHNIQUES FOR EXAMINING PROTEINS

9.4.1 Immunology
Immunological techniques are not applied exclusively to proteins but this
is a convenient time to consider them as many of the most successful
 Techniques for examining proteins 245

immunological identification systems detect specific proteins. To chronicle

fully all the exciting developments in immunology over the past decade
and forecast their impact on diagnostic microbiology is beyond an article
of this sort. The principal developments have been well publicized. Mono-
clonal antibody technology (Kohler and Milstein, 1975) provides the means
of mass producing antibodies of defined specificity, and enzyme-linked
immunoassays (ELISA) provide a convenient, non-radioisotopic means of
detecting antigen-antibody reactions (for an introduction to monoclonal
antibodies see Macario and Conway de Macario (1985), for ELISA tech-
niques see Voller, Bidwell and Bartlett (1980) and Claussen (1988) and for
general applications of serology in bacterial identification see Bowden
(1993)).
As an example of the developing impact of immunology on microbial
identification we can consider the problem of detecting the food poisoning
organism, Salmonella. The fluorescent antibody (FA) technique was one of
the first immunological techniques applied to the detection of salmonellae
(Thomason, Cherry and Moody, 1957; Thomason, 1980). The principle of
the FA is the application of conjugates of Salmonella antiserum with fluor-
escein isothiocyanate to smears of bacteria on microscope slides. When
viewed against a dark background, Salmonella cells fluoresce orange-green
(Schrade, 1984). Methodology based on FA was approved for use in the
USA in 1976 following improvements in the quality of the reagents over a
number of years. However, even with improved antisera, the FA technique
has not become widely used. The main problem has been the high level of
false positives (5-9%), exacerbated by the obvious difficulties in using a
microscopical technique to examine an antigen-antibody reaction on
hundreds of samples per day.
A different approach was reported by Sperber and Diebel (1969), who
called their technique enrichment serology. Again, problems with the
specificity of antisera gave an unacceptable level of false positives (13-15%)
and the technique failed to detect some serotypes of Salmonella, including
S. agona (Mohr, Trenk and Yeterian, 1974).
Krysinski and Heimsch (1977) developed the first enzyme immunoassay
for salmonellae, but despite improvements by Minnich (1978) and
Swaminathan and Ayres (1980) the antibody preparation still suffered from
cross-reactions with other enteric bacteria.
The first enzyme immunoassays for salmonellae to employ a mono-
clonal antibody were reported by Robison, Pretzman and Mattingly (1983).
The MOPC 467 protein was originally isolated by Potter (1970) and was
later found to be specific for a flagella determinant found on several
Salmonella strains (Smith, Miller and Whitehead, 1979). The MOPC 467
monoclonal could detect, specifically, salmonellae in mixed cultures with
no cross-reactions with other enteric bacteria (Mattingly and Gehle, 1984).
Unfortunately it could not detect all the known Salmonella serotypes.
Mattingly (1984) developed a second monoclonal (6H4) by immunizing
246 Methods for the rapid identification of microorganisms
mice with strains of salmonellae that could not be detected by the MOPC
467 system. Mattingly et al. (1985) describe the ELISA system resulting from
these developments. The monoclonal antibodies are attached to a solid-
phase matrix (a ferrous metal bead) which can be transferred between the
wells of a microtitre plate for exposure to the antigen, the enzyme-
conjugated antibody and the substrate. The system uses a horseradish
peroxidase enzyme producing a green colour if salmonellae are present.
The ELISA is used by first raising the Salmonella population to 105_106
cells mr! by the normal pre-enrichment/ selective enrichment procedures,
and then the assay can be applied directly, taking 2-4 h to perform.
Conventional procedures take 24-48 h to isolate salmonellae and confirm
suspect colonies by biochemical and serological techniques.
The development of a usable ELISA system for salmonellae is typical of
the enormous potential for modern immunological methods. Monoclonal
techniques simplify the search for specific antisera which are required for
accurate typing. ELISA techniques simplify the detection of antigen-
antibody reactions, replacing more difficult techniques such as radio-
immunoassays. The main limitation on the use of immunological
techniques for microbial identification is the effort, in terms of manpower,
required for screening clones and developing assays. This will probably
limit applications to those pathogens or organisms of industrial importance
that are identified with a frequency which will support the commercializa-
tion of an ELISA kit. Indeed the plethora of commercial kits available today
based on monoclonal antibodies and ELISA methodology testifies to the
effectiveness of the technology (Sharpe, 1994).

9.4.2 Electrophoresis
Electrophoresis of whole cell protein extracts is another technique that has
found its major application in taxonomic and characterization studies, but
is becoming increasingly important in rapid identification. The expression
of the microbial genome produces more than 2000 protein molecules within
a microbial cell. Techniques that can address the characterization of these
proteins obviously have enormous potential in diagnostic microbiology.
Jackman (1985) has proposed a standard polyacrylamide gel electro-
phoresis (PAGE) technique which has been adopted by the National Col-
lection of Type Cultures as the foundation of a system for the computer
identification of medically important bacteria. A key feature of the tech-
nique is the standardization (as far as is practicable) of the conditions under
which the organisms are first grown and then harvested and disrupted,
releasing their proteins. Jackman carried out cell disruption in sodium
dodecyl sulphate (SDS) buffer at 100°C, which has the added benefit of
killing the cells, allowing pathogens to be handled safely. This is followed
by electrophoresis in a homogeneous, discontinuous, SDS-containing alka-
line buffer. A similar system has been developed in the laboratory of
 Methods that examine aspects of cell composition 247
K. Kersters (Vauterin, Swings and Kersters, 1993; Pot, Vandamme and
Kersters, 1994). Both systems result in gels which are scanned by a densi-
tometer following standardization by reference to marker proteins. The
results from both approaches are complex banding patterns in which each
band represents several protein species since the resolution of one-
dimensional electrophoresis is insufficient to separate all proteins in the
cell. The quantitative data from the densitometer are amenable to analysis
by computer taxonomic techniques to provide cluster or species finger-
prints which can be stored in a computer. For identification, unknown
organisms can be compared with a library of fingerprints.
The taxonomic studies to which electrophoresis has been applied have
been reviewed by Kersters et al. (1994). The most common findings are that
electrophoretic patterns are most discriminatory at the species, subspecies
or biotype levels. The main advantages of the electrophoretic approach are
the efficiency with which large numbers of strains can be compared and,
provided a cell disruption method can be found, its applicability to differ-
ent types of microorganism. The main disadvantage is that it requires a
large number of cells for the preparation of the initial extract.
Within the brewery, protein electrophoresis is particularly suited to
identification of lactic acid bacteria. It has proven to be useful for identifi-
cation of the L. acidophilus (Pot et al., 1993) and L. casei groups (HertI et al.,
1993) as well as leuconostocs (Dicks, Van Vuuren and Dellaglio, 1990).
A novel electrophoretic technique, introduced by Robert Silman of St
Bartholomew's Hospital, London, was commercialized as the AMB-ID
system (V.A. Howe, London). In this technique, single colonies of the
organism to be identified are incubated from 30 min to 2 h (depending on
the type of organism) in a medium containing 3sS-methionine. The methio-
nine is taken up during protein synthesis and radiolabels the proteins.
Growth and protein synthesis is halted by the addition of a buffer and the
cells are disrupted by heating at 100°C for 2 min. The extract is then
electrophoresed for 1 h in a gradient polyacrylamide gel system. The
protein bands are detected and recorded by a specially developed beta
scanner producing a fingerprint which can be matched against a computer
library.

9.5 METHODS THAT EXAMINE ASPECTS OF CELL COMPOSITION

A host of analytical techniques can be applied to the study of the cellular

composition of microorganisms. Although originally these techniques
were applied largely to problems of taxonomy and characterization they
are increasingly being used for rapid identification. Many of them have
great potential to become more important in diagnostic microbiology and
are perhaps held back by the lack of analytical expertise and equipment
available to most microbiologists.
248 Methods for the rapid identification of microorganisms
9.5.1 Physical techniques
One of the most intriguing techniques to be suggested as a rapid identifica-
tion method is circular intensity differential scattering (CIDS) (Gregg et al.,
1985). CIDS involves the differential scattering of left and right circularly
polarized light, generated by a laser and allowed to impinge on a microbial
suspension in a cuvette. A detector measures the intensity of light at differ-
ent scattering angles producing a spectrum. Spectra can be accumulated at
different wavelengths. The physical basis of the ems spectra is still poorly
understood but has been proposed to be the three-dimensional 'packaging'
of helical molecules, principally those in the microbial genome.
Gregg et al. (1985) envisage future instruments, linked to a flow cyto-
meter, that will separate cells so that organisms can be examined one at a
time and a CIDS spectrum obtained. There is little published work to justify
the optimism for CIDS as a general identification technique. Salzman,
Griffith and Gregg (1982) have shown that different CIDS spectra were
obtained from five different crude influenza virus preparations, two strains
of Salmonella typhimurium and three strains of Escherichia coli.
Flow cytometry (for reviews see Muirhead, Horan and Poste (1985) and
Nelson (1991)) is another physical technique with promise for microbial
identification. A flow cytometer is a device in which cells are introduced
into the centre of a fast-moving fluid stream and are forced to flow in a
single file out of a 50-100 !lm orifice at a uniform speed within a layer of
fluid. The cells pass a measurement station where they are illuminated by
a light source (usually a tunable laser). Measurements are made at rates of
104-106 cells min-I. When a particle in the flow stream passes through the
light beam, the illuminating light is elastically scattered by the cell and the
intensity of scatter at different angles can yield information about cell size,
shape, viability, density and surface morphology. In most applications of
flow cytometry, fluorochromes are used to label the cellular component(s)
of interest. The fluorescence emitted by these molecules, when excited by
the illuminating laser beam, can yield information on the expression of
target molecules within single cells. Flow sorters are then able to separate
phYSically the cells of interest from the heterogeneous population.
Phillips and Martin (1985) provide an example of the potential use of
flow cytometry in diagnostic microbiology. Using a fluorescein-conjugated
antibacterial IgG, they were able to separate Bacillus anthracis spores above
a background of Gram-negative bacteria.
CIDS and flow cytometry are expensive techniques in terms of equip-
ment costs and have had little or no impact on diagnostic microbiology.
However, they are techniques that can be applied to a heterogeneous
population of cells with the potential to separate and type each individual
member of the population. Thus they are rapid techniques that might be
applied to mixed cultures, whereas the vast majority of identification
techniques will only work with pure cultures.
 Methods that examine aspects of cell composition 249
9.5.2 Analytical techniques
Goodfellow and Minnikin (1985) have pointed out that the nature of the
layer binding the bacterial cell has held a special place in microbial system-
atics since the invention of the Gram stain over 100 years ago. Salton (1952)
was the first to introduce methods for isolating and purifying cell walls.
Since then the walls of many different bacteria have been analysed and
several procedures developed to detect specific components. Examples of
these are shown in Table 9.2. O'Donnell, Minnikin and Goodfellow (1985)
have shown how many of these analytical procedures can be integrated to
generate a chemo-systematic profile of actinomycetes. However, most of
the analytical procedures that have been applied to microbial characteriza-
tion are slow because they require the production of a small quantity of
biomass for extraction and the methods are generally not automated.
Consequently, they have been largely applied to classical taxonomic
studies.
Gas chromatographic techniques, however, have been used extensively
for microbiological characterization. Good reviews have been published by
Odham, Larsson and Mardh (1984), Jantzen and Bryn (1994) and Embley
and Wait (1994). Gas chromatography is a general-purpose analytical
technique for separating covalent compounds of moderate molecular
weight. In simple terms, a sample is introduced into a stream of heated
carrier gas (the mobile phase) where its components are volatilized and
swept through a chromatography column containing a stationary phase.
In the column the components are selectively retarded according to their
interactions with the stationary phase. The separated components enter a
detector and are recorded and quantitated as chromatographic peaks. Gas
chromatographs can be readily connected to mass spectrometers (GC-MS)
to effect the identification of the separated components.
As discussed, GC techniques can be used to generate information about
many constituents of microbial cells. For the identification of micro-
organisms they have, until recently, found particular application in anaer-
obic bacteriology where strict anaerobes can be identified by the
determination of their fermentation end-products. The VPI Anaerobe
Laboratory Manual (Holdeman, Cato and Moore, 1977) describes many of
the essential procedures. In some situations GC can be applied to extracts
of infected materials to make a direct diagnosis of the contaminating
organism. Examples include the presence of short-chain fatty acids in pus
specimens containing anaerobes (Phillips, Taylor and Eykyn, 1980) and in
amniotic fluid (Gravett et al., 1982).
The major development in GC for identification purposes, however, is
the introduction of complete microbial identification systems based on this
technology. Originally introduced by Hewlett Packard and now marketed
as the MIDI system (Newark, Delaware, US), the basis of the method is the
extraction of whole cell fatty acids and GC profiling of their methyl esters
Table 9.2 Chemotaxonomic information from bacterial cell envelopes (adapted from Goodfellow and Minnikin (1985»
Site Component Techniques Reference
Plasma membrane Isoprenoid quinones HPLC, MS, RPTLC Collins (1985)
Lipid-soluble pigments
Lipoteichoic acids
Polar lipids
Fatty acids GC Moss (1985)
Isoprenoid ethers HPLC, MS, RPTLC Jantzen and Bryn (1985)
Other long-chain components
Proteins
Walls Peptidoglycans and analogues GC O'Donnell, Minnikin and
Goodfellow (1985)
Polysaccharide sugar patterns
Teichoic acids and analogues
Outer membrane Gram-negative bacteria Lipopolysaccharides
(K and a antigens)
Polar lipids TLC Dobson et al. (1985)
Outer membrane Gram-positive bacteria Bound lipids-mycolic acids TLC Dobson et al. (1985)
Free lipids
Glycolipids
Sulphoglycolipids
Waxes
HPLC, high-performance liquid chromatography; MS, mass spectrometry; RPTLC, reverse-phase thin layer chromatography; GC, gas chroma-
tography; TLC, thin layer chromatography.
 Methods that examine aspects of cell composition 251
(FAME profiles). The methodology is based on the work of C.W. Moss at
the Centers for Disease Control, Atlanta, USA (Moss, 1981; Moss and Dees,
1976,1978; Moss, Dees and Guerrant, 1980) and uses standardized cultiva-
tion of the microorganisms and preparation of the FAME samples. Bacteria
are grown for 24 h as spread plates and the cells are saponified in a strong
base. After saponification the solution is acidified and the liberated fatty
acids are methylated to increase their volatility for Gc. The methyl esters
are extracted and 1 III of a washed extract injected on to a 50 m
methylsilicone fused silica capillary column. The system is completed by a
computer for recording data and matching FAME profiles against a
reference library.
The library covers a number of Gram-negative and Gram-positive gen-
era but the technique can be applied to any microorganism and individual
reference libraries constructed. The production of FAME profiles takes
approximately 60-90 min (60 min sample preparation time) although sam-
ples are prepared and analysed in batches and the actual preparation time
per sample is quoted at less than 4 min. Examples of FAME profiles for two
Gram-negative bacteria are shown in Fig. 9.2 (Miller, 1984). The technique
of FAME profiling has wide-ranging potential as an identification tool as
evidenced by the introduction and success of the MIDI system. Possible
disadvantages are, firstly, that running any GC system requires consider-
able dedication and analytical expertise, and secondly, it is still not clear
whether FAME profiles vary sufficiently to allow typing of strains as well
as genera and species.
Some analytical techniques can be applied directly to whole cells with
no requirement to make an extract. One example is infrared (IR) spectro-
metry which has received increasing attention as a simple procedure to
obtain fingerprints of microorganisms (for reviews see Giesbrecht et al.,
1985; Nichols et al., 1985; Magee, 1993; Naumann, Helm and Schultz, 1994).
Attempts to use IR spectroscopy to characterize microorganisms were
made in the 1950s and 1960s (e.g. Norris, 1959), but no successful routine
technique could be established because of the time-consuming preparation
procedures, the low scanning speeds of the conventional spectrometers and
the difficulties in discriminating the complex and relatively similar spectra.
Many of these problems have been overcome with the modem Fourier
transform (FT) IR instruments.
The IR spectrum of any compound is known to express a unique
'fingerprint' and it is this characteristic that allows IR spectroscopy to be
used for identification using spectral data libraries. The recent micro-
biological applications rely on FT-IR spectra of whole cells which comprise
the vibrational features of all cell components, that is DNA/RNA, proteins,
membrane and cell wall constituents (Helm et al., 1991). FT-IR therefore
probes the total composition of a given organism in a single experiment
and allows selectivity to a very high level, for example at the strain or
serotype level. Moreover, since whole cells are being tested complicated
252 Methods for the rapid identification of microorganisms

..•
Ii
0

3}3} ji~
.I
~ .%
%
... uu
0

.!
0
eu
.
.I
9-
U
U %0 ...
9.
..,
1&1 0'"
NS U
z 19.
..,2 ..
9. 0
~:: :t
1&1 ::u .!
a: U ~
U

• •
I I I I I I I I
0 2 4 10 12 14
MINUTES

0
:!!
~
....
cae;
...
~

0
U uri U
.!!
...
9-
u
1&1
I/)
z
1J
u %
%
..,
0
~
U
..u .....
'?
(;
2 ...
0
.!!
0
13
...'?
t
a: '?

- L1- 11 f ; 1 ~
------
j
o 5 15
MINUTES

Fig.9.2 Fatty acid methyl ester profiles for Xanthomonas maltophilia (top trace) and
Pseudomonas putrefaciens (bottom trace). (From Miller (1984), reproduced by kind
permission of Hewlett Packard.)

and time-consuming preparation of cell constituents is avoided. The pro-

cedure has been used to distinguish intra generic classification and provide
identification of staphylococci (Helm, Labischinski and Naumann, 1991)
and in principle the method would be of value for lactic acid and yeasts.
Some typical examples of FT-IR spectra are shown in Fig. 9.3.
Ultraviolet resonance Raman spectrometry (UVRRS) applied to bacterial
identification has been reviewed by Nelson and Sperry (1991) and Magee
(1993). This method depends on spectra produced by UV excitation of
bacteria at various wavelengths yielding information on nucleic acids,
 Methods that examine aspects of cell composition 253

r-
ID
o

ILIm
u ...
<:111
a: •
!I?
o
0
III
m
a:.
ID
N
o

9200 -2000 1100 200

\.IAVENUMBERS

o
III
r-
o

o
III
N
o 2

2000 11UO 600 200

\.IAVENUMBERS

Fig.9.3 Fourier transfonn infrared spectra of: (top trace) 1, Streptococcus pyogenes;
2, Klebsiella pneumoniae; (bottom trace) 1, Streptococcus pyogenes Group A; 2, Strep-
tococcus agalactiae Group B. (From Giesbrecht et al. (1985), reproduced by kind
permission of Springer-Verlag.)
254 Methods for the rapid identification of microorganisms
proteins and cell wall constituents. This powerful method is capable of
distinguishing bacteria at the species and strain level but it is in its infancy
and routine methods for its application have yet to be described.
Pyrolysis, which is the controlled thermal degradation of a compound,
in this case microbial biomass, in an inert atmosphere produces a complex
mixture of low molecular weight volatile compounds. This mixture can be
separated using GC and the constituents detected to produce a 'fingerprint'
known as a pyrogram. This technique of pyrolysis gas chromatography
(Py-GC) has been applied by several groups to the characterization of
microorganisms (Gutteridge and Norris, 1979; Gutteridge, Mackey and
Norris, 1980; Magee, Hindmarch and Meecham, 1983; Magee, 1993;
Goodfellow et al., 1994). The main difficulties with Py-GC are the slow
speed per sample (modern capillary columns will separate over 100 com-
ponents and an analysis can take up to 2 h) but most important of all is the
lack of quantitative reproducibility as columns age and need to be replaced.
To overcome these problems, Meuzelaar and Kistemaker (1973) combined
pyrolysis with mass spectrometry (Py-MS) producing an instrument with
fast analysis times (3 min/sample) and no column to degrade. Various
instrumental developments have taken place (Meuzelaar et al., 1976; Shute
et al., 1984; Aries, Gutteridge and Ottley, 1986) culminating in the design of
dedicated, low-cost, commercial systems with automated sample input
such as the PYMS 200X and RAPyD 400 machines marketed by Horizon
Instruments (Heathfield, Sussex, UK).
The attraction of Py-MS as a characterization technique, apart from the
fast analysis times, is that it can be applied to a single microbial colony taken
directly from the surface of a culture plate and transferred to the pyrolysis
filament. Thus, preparation steps are kept to an absolute minimum. In
addition, it is capable of discriminating organisms at taxonomic levels
below species and providing fingerprints of individual strains. An excellent
example of this is provided by Wieten, Meuzelaar and Haverkamp (1984)
who showed the differentiation of E. coli strains with and without the K1
polysaccharide antigen. A spectrum of the purified antigen (colominic acid)
accounted for the key differences between the spectra of the whole bacteria.
Other similar examples, particularly for epidemiological typing of patho-
gens, have been described by Freeman et al. (1990) and reviewed by Magee
(1993) and Goodfellow et al. (1994).
One of the earlier disadvantages of Py-MS was the requirement for
complex statistical techniques to handle the large amounts of quantitative
data produced. A typical pyrolysis mass spectrum is shown in Fig. 9.4. The
intensities of all masses above m/z 12 and below m/z 300 are recorded,
although, in practice, microorganisms only give signals in the range
m/z 30-160 depending on the type of mass spectrometry employed (Tas
et al., 1985). Observations on a wide variety of microorganisms suggest that
spectra of different genera or species are qualitatively similar and discrim-
ination has to be based on complex combinations of small, but reproducible
 o
o
(r)

o
o
N

-
N
E

.~
<rJ
:::
'"
:::
...
~

...
0 <::!
0 <::!
:::
~
u

c
~
:::
.E
t...

--
!:Q

§
.t:u
IJ.)
0.-
rJJ
rJJ
rJJ
ce
S
.~
0 rJJ
..b
0
~
...
0 0 0 Cl..
0 It)

"'a\"
oil
li:
256 Methods for the rapid identification of microorganisms
variations in mass intensities. This requires the sorting of matrices contain-
ing tens of thousands of data points which can only be realistically
accomplished by using multivariate statistical techniques such as principal
components analysis, canonical variates analysis and cluster analysis.
These methods and their application to Py-MS data have been discussed
by Gutteridge, Vallis and MacFie (1985) and Magee (1993). Fortunately
these computational problems have now been resolved with the introduc-
tion of dedicated Py-MS instruments as described above. Furthermore,
developments in the use of artificial neural networks offer a powerful
system for the discrimination of samples based on Py-MS data (Goodacre,
Kell and Bianchi, 1992).
Further concerns about the use of Py-MS for identification of micro-
organisms relate to the long-term stability of the mass spectrometers used
and the problem this may pose for the construction of reference libraries of
key spectra. These problems have yet to be investigated fully but can be
overcome by always including suitable reference strains.
Most microbiological applications of Py-MS involve the fingerprinting
of batches of isolates to examine short-term problems. The rapid screening
of batches of new isolates to gain an appreciation of the number of types of
microorganism present is an important application of Py-MS. Figure 9.5

GROUP I
(·wllel· SacohaN.YC.. I

---- --------
GROUP 2 {
'pltcldng SacchaN.Yc•• I

~~======~--r-------T-

I I , I , I
10 10 70 10 50 40 30

Fig. 9.5 Dendrogram of the relationships between 51 yeast isolates based on the
analysis of pyrolysis mass spectrometry data.
 Developments in studying morphology and behaviour 257
shows a dendrogram obtained after the computer analysis of spectra for 51
yeast strains isolated from a home-brew product. Data were collected and
analysed in 48 h producing a functional classification that divided the
yeasts into four groups. Subsequently, conventional tests were applied to
the strains and it was found that Py-MS had successfully discriminated
wild Saccharomyces strains (Group 1) from the pitching yeasts (Group 2) and
other wild yeasts (Group 3). The strains in Group 3 could be subdivided by
Py-MS but the taxonomic basis of this was not determined. This example
indicates the potential for Py-MS in screening and epidemiological invest-
igations. We use Py-MS in our laboratory to get a 'first look' at the relation-
ships between isolates when we have sampled the micro flora of an
environment for the first time. This allows us to make a selection of strains
for further study and typing, saving a great deal of time in the process. This
application of Py-MS is convenient because it avoids the thorny problems
of data bases and long-term stability. There is no doubt that Py-MS can be
used for the identification of microorganisms but whether it can be made
sufficiently routine to appeal to the ordinary diagnostic laboratory remains
in doubt.
One final technique that can be applied to whole cells, even single cells,
is the laser microprobe mass analyser (LAMMA). LAMMA (Leybold-
Heraeus, Koln, Germany) is a specialized laser pyrolysis MS apparatus. A
specimen on a grid is located by light microscopy and a laser beam is
focused along the same path. When the laser is activated a small section of
the grid is pyrolysed and the pyrolysis products are detected and quantified
by a time-of-flight mass spectrometer. LAMMA has been used by Bohm,
Kapr and Schmitt (1985) to differentiate Bacillus anthracis, B. cereus and
B. thuringiensis based on the analysis of single vegetative cells. One inter-
esting feature of LAMMA is that it can also detect some inorganic ions and
can be used to study, for example, Na, K and Ca metabolism. Nevertheless,
because of the extreme capital costs, LAMMA remains a curiosity in terms
of microbial identification despite its interesting ability to analyse single
cells.

9.6 DEVELOPMENTS IN TECHNIQUES FOR STUDYING

MORPHOLOGY AND BEHAVIOUR

9.6.1 Selective media, DEFT and impedance

Studies on the morphology and behaviour of microorganisms are several
stages removed from the microbial genome. However, this is the level at
which most conventional identification systems operate and important
developments continue to be made which, by and large, can be im-
plemented by most microbiologists as they do not require specialized
analytical equipment. For example, the ability of microbiologists to
258 Methods for the rapid identification of microorganisms
construct selective media for the isolation of specific groups or types of
microorganism continues unabated. The copper sulphate medium for the
detection of Saccharomyces and non-Saccharomyces wild yeasts in the pres-
ence of brewing yeasts (Taylor and Marsh, 1984) is a fine example of
characterization for pragmatic purposes. Raka-Ray agar (Saha, Sondag and
Middlekauff, 1974), which is used for the selective isolation of lactobacilli
from brewing processes, is another (Chapter 6).
The use of microscopy will continue to be the first line of attack for a
microbiologist faced with an identification problem. The direct epifluor-
escent filter technique (DEFT), which is described in detail by Pettipher
(1983,1989), was developed as a rapid counting procedure for dairy sam-
ples. DEFT involves the recovery of organisms from a sample by filtration,
staining in situ with a fluorescent dye, and examination under an
epifluorescent microscope. The use of fluorescence staining under the
carefully controlled conditions of the DEFT can improve the objectivity of
microscopical examination and permit tentative identification of types of
organisms in situations where there is previous experience, such as the
examination of bovine materials for mastitis-causing organisms. The
application to brewing microbiology has been fully described in Chapter 8.
Another technique devised for rapid enumeration but with potential for
recognition of specific groups of microorganisms is electrical detection,
otherwise known as conductance or impedance detection (Firstenberg-
Eden and Eden, 1984; Easter and Gibson, 1989). In general terms, the growth
of microorganisms in a medium will alter the electrical resistance and the
changes can be monitored continuously by passing a small current through
the medium between electrodes, and comparing it with an uninoculated
control or an electronic reference. Three electrical signals can be
distinguished: conductance, capacitance and impedance.
The major application of electrical detection is as a replacement for plate
counts as a means of detecting growth (see Chapter 8). The development
of selective media that produce appropriate electrical responses can allow
this technology to be used for the detection of specific classes of micro-
organism. For example, Easter and Gibson (1985) developed an electrical
detection system for salmonellae. Following pre-enrichment in buffered
peptone water modified by the addition of dulcitol and trimethylamine
oxide and selective enrichment in selenite-cystine broth with similar
modifications, salmonellae can be detected by their ability to give a fast
(100 f.lS h-1) and large (> 600 f.lS) conductance change. Other enteric bacteria
produce a small change or no change at all (Fig. 9.6). The principle of the
test is the reduction of trimethylamine N-oxide to trimethylamine, a reac-
tion which occurs during post-mortem bacterial spoilage of marine fish
muscle and is accompanied by a significant change in conductance. This
test can reduce the time taken for the detection of salmonellae by 24 h
although its specificity can be improved by running it in parallel with other
media and by using immunology as a confirmatory tool. Nevertheless,
 170
Salmonella 18angl
1501

130+
/
f... 110+
<II
...c
ell
90·
/
ti
~
"D 70
c
0
...
50 Enterobacter cloacae

30 ~ ---------
-------------------------
Enterobacter aero genes
10
--+- I I! '=-=t=-:.. I; ·····I·······j .. ·····j·· .. ·· I I

o 2 4 6 8 10 12 14 16 18 20 22 24 26 28
hours

Fig. 9.6 Change in the conductance of a medium containing trimethylamine oxide

inoculated with single strains of Salmonella isangi, Enterobacter cloacae and Enterobacter
aerogenes.
260 Methods for the rapid identification of microorganisms
electrical detection systems have the potential to be useful for the recogni-
tion of many broad classes of microorganism provided some key aspect of
physiology can be adapted to produce an electric response. This limitation
will perhaps not permit the use of electrical systems for the delineation of
specific genera or species.

9.6.2 Miniaturized methods and commercial systems

The major development in rapid identification over the past 10-15 years
has been the miniaturization of conventional biochemical and nutritional
tests into commercially manufactured kits. A vast range of commercial
systems is now available mainly, but not exclusively targeted at organisms
of clinical significance. For the identification of members of the
Enterobacteriaceae and other Gram-negative bacteria there are several
systems available, including the API20E (bioMerieux Basingstoke, UK),
Micro-ID (General Diagnostics, Morris Plains, USA) and Enterotube
(Roche, Geneva, Switzerland), with results available in 4-24 h depending
on the system used. The main advantage comes from standardization of
reagent manufacture which produces reproducible identifications. Sneath
(1974) estimated the probability of errors occurring with conventional
identification tests within a laboratory to be 2-4% and between laboratories
6-10%. Within-laboratory studies of commercial identification systems
yielded average probabilities of errors comparable with conventional tests.
Definitive data on between-laboratory reproducibility for commercial
identification systems are not readily available but Holmes, Dowling and
Lapage (1979) expect it to be superior to conventional methods. Commer-
cial identification systems are backed up with computerized collections of
reference profiles generated from thousands of strains which can be used
to produce an assignment of an unknown based on a probability score.
Recent microbiological developments of miniaturized systems have
been in two opposing directions. Firstly, kits have been produced for more
specialized groups of organisms, including several for clinical yeasts, e.g.
Mycotube (Roche), Auxodisk (Sobioda, New York, USA), Uni-Yeast-Tek
System (Corning Medical, Halstead, UK), Micro-drop Assimilation Test
System (Clinical Science, Washington, USA) and API20C (bioMerieux Ltd).
In the other direction, systems such as the API50CH allow the profiling of
the carbohydrate assimilation pattern of any organism for which a suitable
inoculation and incubation protocol can be established. Logan and
Berkeley (1981) used API20E, API50E and APIZYM test kits to produce a
taxonomic scheme for the Gram-positive, aerobic, spore-forming genus
Bacillus which was in accord with schemes produced by conventional
biochemical and morphological tests. As a result, identification of Bacillus
spp. can be achieved by the current API50CH system backed by a
computer-based pattern-matching program. In my own laboratory we
have adapted the API50CH kit for the identification of yeasts involved in
 Developments in studying morphology and behaviour 261
Table 9.3 Examples of API50CH profiles for three Brettanomyces strains
B. anomalus B. intermedius B. naardenensis
Carbohydrate NCYC 615 CBS 73 NCYC 813
Glycerol + +
Erythritol
D-Arabinose
L-Arabinose
Ribose
D-Xylose +
L-Xylose
Adonitol
p-Methylxyloside
Galactose + + +
D-Glucose + + +
D-Fructose + + +
D-Mannose + + +
L-Sorbose
Rhamnose
Dulcitol
Inositol
Mannitol +
Sorbitol +
a-Methyl-D-mannoside
a-Methyl-D-glucoside + +
N-Acetylglucosamine + + +
Amygdalin + + +
Arbutin + +
Aesculin + + +
Salicin + +
Cellobiose + + +
Maltose + + +
Lactose +
Melibiose
Sucrose + +
Trehalose + + +
Inulin
Melezitose + + +
D-Raffinose +
Starch +
Glycogen
Xylitol +
p-Gentiobiose + + +
D-Turanose + +
D-Lyxose +
D-Tagatose
D-Fucose
L-Fucose
D-Arabitol + +
L-Arabitol
Gluconate
2-Ketogluconate
3-Ketogluconate
+, growth; -, no growth.
262 Methods for the rapid identification of microorganisms
the spoilage of soft drinks. Used in conjunction with a few fermentation
tests and some morphological observations, it can produce a profile that
can be compared with an existing yeast data base, such as that published
by Barnett, Payne and Yarrow (1983) which is now also available on
computer. The adaptation of this system for yeast profiling is not a simple
matter. Care has to be taken to avoid carry-over of substrates. Long incu-
bation periods, compared with bacteria, may be necessary, increasing the
risk of contamination. Nevertheless, perseverence can produce a profiling
system geared to the identification needs of your own laboratory. Examples
of API50CH profiles for three type cultures of Brettanomyces are shown in
Table 9.3.

9.6.3 Automated systems

Technically, miniaturized systems are also proceeding in two directions:
automation to provide rapid and reproducible sample handling, and a
switch to enzyme- or metabolism-based tests to avoid the delays involved
in traditional growth-based tests. Miniaturization of biochemical tests,
accompanied by photo-optical detection of growth, are the principles be-
hind the VITEK System marketed in the UK by bioMerieux. The VITEK is
an approach to automation of screening, identification and antimicrobial
assay in the laboratory. Plastic transparent test cards are factory-filled with
chemical substrates. The cards are rehydrated with a suspension of the
unknown organism and are incubated in stacks. Detection of growth by
monitoring changes in optical density is automatic and the system uses a
computer to produce a probabilistic identification score against a library of
standard profiles. Three standard test cards have proved popular, those for
Gram-negative identification, Gram-positive identification and yeast
biochemical profiles. Emphasis was on organisms of clinical significance
and the yeast card covers mainly Candida and Cryptococcus species. How-
ever, bioMerieux have been working to extend the application of VITEK
into the industrial laboratory. Sister instrumentation for automated blood
cultures (VITAL), automated immunoanalysis (VIDAS) and automated
fluorescence microscopy (COBRA) has also been introduced.
Biolog (California, USA) have introduced methodology for identifica-
tion based on the indication of carbon substrate utilization by the inclusion
of tetrazolium-based indicator of respiration with the substrate. Carbon
sources specific for Gram-negative or Gram-positive bacteria or yeasts and
tetrazolium salt are dispensed in microtitre dishes which are purchased
complete. Standard inoculation of the 96 wells with a microbial suspension
and incubation for 4 and 16 h provides a biochemical fingerprint of the
microorganism for the 95 substrates (one control well). These patterns can
be read by eye or with an automated reader and compared with a database
using standard computer algorithms. Although Biolog did not provide
particularly high accuracy for identification of enterobacteria in a recent
 Developments in studying morphology and behaviour 263
comparative survey (Stager and Davis, 1992), Biolog have one of the most
extensive databases for microbial identification and recent introduction of
trays for lactobacilli and yeasts make this system attractive to the brewery
microbiologist.
Other automated systems such as those from Sensititre and Baxter
Diagnostics are described below since they are based on enzyme tests.

9.6.4 Enzyme-based systems

A different and more rapid approach to the identification of clinical isolates
is based on enzyme tests for characterizing the microorganisms (Bascomb,
1980, 1985; Manafi, Kneifel and Bascomb, 1991). Conventional methods rely
on microbial growth and, with the dilution of the bacterial population that
occurs during inoculation, at least 10-17 generation times are required to
increase numbers to levels where enough metabolites are produced to
trigger a detection system, such as a drop in pH indicated by a colour
change. In addition, many conventional tests detect products or effects of
enzymes that have to be induced or are the result of multienzyme
pathways. Tests targeted at the detection of constitutive enzymes or their
direct products are faster and more specific. Further advantages of enzyme
activity tests are:
1. amplification of the signal is more easily achieved by optimizing
conditions for enzyme activity than by increasing the concentration of
a product;
2. the shorter incubation times required for enzyme activity testing
minimize concerns over chance contamination;
3. enzyme activity tests are more sensitive and can be completed with
lower numbers of bacteria than conventional tests;
4. equipment for monitoring enzyme activities is available.
Three categories of enzyme test can be recognized:
1. qualitative tests, where the presence of an enzyme is detected
subjectively by observation of a change in the appearance, colour or
fluorescence of an enzyme/substrate mixture;
2. studies of isoenzymes using electrophoresis;
3. quantitative tests in which activities are measured and expressed in
terms of the quantity or rate of reaction product formation.
Enzymes whose activities have been used for identification have been
mainly catabolic, including esterases, glycosidases, lipases, phosphatases,
peptidases and proteases (Manafi, Kneifel and Bascomb, 1991). Detection
has been made easier by the introduction of a variety of synthetic chromo-
genic and fluorogenic substrates where the chromogen or fluorogen is
attached to a simple moiety such as an amino acid, monosaccharide, fatty
acid, phosphate or sulphate ion. Quantitative measurements of activity
264 Methods for the rapid identification of microorganisms
require some degree of automation if large numbers of determinations are
needed. Bascomb (1985) has chronicled developments in the use of contin-
uous flow analysers, discrete analysers, fast centrifugal analysers,
specialized analysers and multi-well plates.
Bascomb (1984) used a fluorimetric system in multi-well plates. Pre-
prepared plates using dried substrates can be stored at 5-8°C. Plates were
inoculated manually and incubated for 2 h after which reagents were
added and the plates were incubated for a further 15 min to allow the
reaction to occur. The fluorescence of each well was recorded automatically
and fed into a computer running a discriminant analysis procedure for
identification. Unknowns could be typed within 2-5 h of receiving the
primary isolation plate. A comparison between identifications based on
rapid fluorescence enzyme tests and conventional methods for 1543 cul-
tures, representing 24 genera, showed an overall agreement of 88%
(Bascomb, 1985). In general, the enzyme activity approach to identification
is a good one that can give genuinely rapid results and it could be
specifically adapted to the identification of brewery microorganisms.
Complete microbial identification systems based on enzyme activities
are produced by Sensititre using fluorescently labelled substrates in 96-well
micro titre plates for the identification of Gram-negative bacteria. Test
plates are incubated off line and may be read after 5 or 18 h using a
dedicated plate reader. Walkaway and autoScan are both manufactured by
Baxter Diagnostics. These computer-controlled machines enable incuba-
tion of multi-well test plates and automatic interpretation of test results.
Fluorogenic panels of enzyme substrates specific for Gram-negative rods,
Gram-positive cocci and yeasts are available. In comparative studies, the
VITEK, Sensititre and Baxter Diagnostics products all gave high and
accurate identifications (Stager and Davis, 1992).

9.7 FUTURE TRENDS IN RAPID IDENTIFICATION

In order to make some sensible predictions about the future of rapid

identification some comments on the costs, speed and applicability of the
techniques discussed have been collated in Tables 9.4 and 9.5. Table 9.4
shows the capital cost of investing in the necessary equipment to use a
technique and gives a comparative estimate of the cost of developing an
identification system and of the running costs per sample. Obviously
development costs depend on the application. The use of VITEK to identify
Enterobacteriaceae from a novel source can be achieved with the existing
test card(s) and is comparatively cheap. In contrast, the use of VITEK to
identify non-clinical yeasts probably requires the development of a new
test card and would consequently be more expensive. Development costs
must also be judged alongside the benefits that might accrue from having
a better identification method. ELISA and gene probe assays may be
 Future trends in rapid identification 265
Table 9.4 Comparison of costs for various identification techniques
Capital cost
Technique (£thousands) Development cost Cost per test*
Levell: genetics
Gene probes None High High

Level 2: proteins
ELISA <5 High High
Electrophoresis 10-15 Moderate Low
AMB-ID system 25--45 Moderate Moderate

Level 3: cell
composition
CIDS >100 High Low
Flow cytometry 50-100 High Low
HP FAME GC system 20-30 Moderate Moderate
FT-IR 30-40 High Low
Py-MS 50--60 High Low
LAMMA >100 High Moderate

Level 4: morphology
and behaviour
Selective media None Moderate Low
DEFT 10-15 Low Low
Impedance 20-40 Moderate Moderate
Commercial kits None Low Modera te /high
VITEK >30 Moderate Moderate/high
Enzyme tests <5 High Low
* Costs per test: low <£1, moderate £1-2, high >£2.

expensive to develop (£5000 to £50 000) but if successful they might provide
accurate diagnosis with minimum sample preparation and save on running
costs. The costs per test in Table 9.4 are rated low « £1), moderate (£1-£2)
or high (> £2). ELISAs and gene probes are costly on reagents for individual
tests but these costs are continually declining in real terms with improved
technology and the cost savings of high volume and the benefits of produc-
ing accurate identifications will often justify the higher cost. With respect
to capital costs, it is clear from Table 9.4 that the analytical techniques that
address cell composition all have high equipment costs. Microbiology is
geared to the use of high volumes of disposables, in contrast to chemistry,
and the need to invest in expensive analytical equipment is beyond the
experience and influence of most microbiologists. Such entrenched
attitudes are a real obstacle to the enhanced use of analytical techniques for
microbial identification.
In Table 9.5, speed per sample is a judgement based on the amount of
preparation time and the time taken by the analysis. Applicability is taken
to mean whether the technique can be applied to different types of micro-
organism (assuming that the user will construct his/her own reference
266 Methods for the rapid identification of microorganisms
Table 9.5 Comparison of speed and applicability for various identification tech-
niques
Volume
Technique Speed per sample throughput Applicability
Levell: genetics
Gene probes Moderate High Wide

Level 2: proteins
ELISA Moderate High Restricted
Electrophoresis Moderate High Wide
AMB-ID system Moderate High Wide

Level 3: cell
composition
cms Fast Low Not known
Flow cytometry Fast High Not known
HP FAME GC system Moderate High Wide
FT-IR Fast Moderate Not known
Py-MS Fast High Wide
LAMMA Fast Moderate Not known

Level 4: morphology
and behaviour
Selective media Slow Low Restricted
DEFT Fast Low Restricted
Impedance Moderate High Restricted
Commercial kits Moderate/slow Low Wide
VITEK Fast/moderate High Wide
Enzyme tests Fast High Wide

libraries). Thus, an individual ELISA or gene probe assay is restricted in the

sense that it only has one 'target' per development budget. In contrast,
investment in FAME profiling by GC gives the user a versatile tool that can
be turned to a wide range of identification problems.
In conclusion, it seems obvious that, in the short term, the demand for
accurate and rapid detection of organisms of public health significance will
fuel the development of an increasing range of gene probes and ELISA
assays. The cost of these developments, however, will restrict the commer-
cial opportunities to those where large numbers of identifications per year
can be guaranteed. In laboratories where identification and characteriza-
tion of microorganisms is carried out less intensively and covers more
unusual groups, the technologies of enzyme tests, FAME profiling, Py-MS
and electrophoresis (including the AMB-ID system) will prosper because
the users can establish, rapidly, their own database. Interestingly, all four
of these approaches require the use of sophisticated computing techniques
to handle the data generated. The computer power to allow this to happen
on-line is now available at a reasonable cost. In the longer term, techniques
 References 267

with the ability to identify, directly, members of a mixed population must

gain in prominence if diagnostic microbiology is to advance.

ACKNOWLEDGEMENTS
We would like to thank Mike Amott and Renny Ison for permission to refer
to some of their unpublished data.

REFERENCES
Amann, R. and Ludwig, W. (1994) In Bacterial Diversity and Systematics (eds
F.G. Priest., A. Ramos-Cormenzana and B. Tindall), Plenum Press, New York,
p.115.
Aries, RE., Gutteridge, CS. and Ottley, T.W. (1986) Journal of Analytical and Applied
Pyrolysis, 9, 8l.
Barnett, J.A., Payne, RW. and Yarrow, D. (1983) Yeasts: Characteristics and
Identification, Cambridge University Press, Cambridge.
Bascomb, S. (1980) In Microbiological Classification and Identification (eds
M. Goodfellow and RG. Board), Academic Press, London, p. 359.
Bascomb, S. (1984) In New Horizons in Microbiology (eds A. Sanna and G. Morace),
Elsevier, Amsterdam, p. 24l.
Bascomb, S. (1985) In Rapid Methods and Automation in Microbiology and Immunology
(ed. K.-O. Habermehl), Springer, Berlin, p. 367.
Bohm, R, Kapr, T. and Schmitt, H. U. (1985) Journal ofAnalytical and Applied Pyrolysis,
8,449.
Bowden, G.H.W. (1993) Handbook of New Bacterial Systematics (eds M. Goodfellow
and A.G. O'Donnell), Academic Press, London, p. 429.
Chevrier, D., Larzul, D., Megrand, F. and Guesden, J.1. (1989) Journal of Clinical
Microbiology, 27, 32l.
Claussen, J. (1988) In Laboratory Techniques in Biochemistry and Molecular Biology,
Vol. 1 (eds RH. Burdon and P.H. van Knippenberg), Elsevier, New York.
Collins, M.D. (1985) In Chemical Methods in Bacterial Systematics (eds M. Goodfellow
and D.E. Minnikin), Academic Press, London, p. 237.
Datta, A.R, Wentz, B.A., Shook, D. and Trucksess, M.W. (1988) Applied and
Environmental Microbiology, 54, 2933.
Delley, M., Mollet, B. and Holtinger, H. (1990) Applied and Environmental
Microbiology, 56, 169).
Dicks, 1.M.T., Van Vuuren, H.J.J. and Dellaglio, F. (1990) International Journal of
Systematic Bacteriology, 40, 83.
Dobson, G., Minnikin, D.E., Minnikin, S.M. et al. (1985) In Chemical Methods in
Bacterial Systematics (eds M. Goodfellow and D.E. Minnikin), Academic Press,
London, p. 237.
Easter, M.C and Gibson, D.M. (1985) Journal of Hygiene, Cambridge, 94,245.
Easter, M.C and Gibson, D.M. (1989) Progress in Industrial Microbiology, 26,57-100.
Embley, T.M. and Wait, R (1994) In Chemical Methods in Prokaryotic Systematics (eds
M. Goodfellow and A.G. O'Donnell), John Wiley, Chichester.
Falkow, S. (1985) In Rapid Methods and Automation in Microbiology and Immunology
(ed. K.-O. Habermehl), Springer, Berlin, p. 30.
Firstenberg-Eden, R and Eden, G. (1984) Impedance Microbiology, Research Studies
Press, Letchworth.
268 Methods for the rapid identification of microorganisms
Freeman, R, Goodfellow, M., Gould, F.K. et al. (1990) Journal ofMedical Microbiology
32,283.
Giesbrecht, P., Naumann, D., Labischinski, H. and Barnickel, G. (1985) In Rapid
Methods and Automation in Microbiology and Immunology (ed. K-O. Habermehl),
Springer, Berlin, p. 198.
Goodacre, R, Kell, D.B. and Bianchi, G. (1992) Nature, 359,594.
Goodfellow, M. and Minnikin, D.E. (1985) In Chemical Methods in Bacterial
Systematics (eds M. Goodfellow and D.E. Minnikin), Academic Press, London,
p.l.
Goodfellow, M., Chun, J., Atalan, E. and Sanglier, J.-J. (1994) In Bacterial Diversity
and Systematics (eds F.G. Priest., A Ramos-Cormenzana and B. Tindall), Ple-
num Press, New York, p. 87.
Gravett, M.G., Eschenbach, D.A., Spiegel-Brown, CA and Holmes, KK (1982) New
England Journal of Medicine, 306, 725.
Gregg, CT.,McGregor, D.M., Grace, W.K. and Salzman, G.C (1985) In RapidMethods
and Automation in Microbiology and Immunology (ed. K-O. Habermehl),
Springer, Berlin, p. 184.
Grimont, P.A.D. (1988) Canadian Journal of Microbiology, 34, 54l.
Gutteridge, CS. and Norris, J.R (1979) Journal of Applied Bacteriology, 47, 5.
Gutteridge, CS., Mackey, B.M. and Norris, J.R (1980) Journal ofApplied Bacteriology,
49,165.
Gutteridge, CS., Vallis, L. and MacFie, H.J.H. (1985) In Computer-assisted Bacterial
Systematics (eds M. Goodfellow, D. Jones and F.G. Priest), Academic Press,
London, p. 369.
Helm, D., Labischinski, H. and Naumann, D. (1991) Journal ofMicrobiological Methods
14,127.
Helm, D., Labischinski, H., Schallehn, G. and Naumann, D. (1991) Journal of General
Microbiology, 137,69.
Hertel, c., Ludwig, W.,Pot, B.etal. (1993) Systematic and AppliedMicrobiology, 16,463.
Holdeman, L.V., Cato, E.P. and Moore, W.E.C (1977) Anaerobe Laboratory Manual,
4th edn, VPI, Blacksburg.
Holmes, B., Dowling, J. and Lapage, S.P. (1979) Journal of Clinical Pathology, 32, 78.
Jackman, P.J.H. (1985) In Chemical Methods in Bacterial Systematics (eds
M. Goodfellow and D.E. Minnikin), Academic Press, London, p. 115.
Jantzen, E. and Bryn, K (1985) In Chemical Methods in Bacterial Systematics (eds
M. Goodfellow and D.E. Minnikin), Academic Press, London, p. 145.
Jantzen, E. and Bryn, K (1994) In Chemical Methods in Prokaryotic Systematics (eds
M. Goodfellow and AG. O'Donnell), John Wiley and Sons Ltd., Chichester, p.
2l.
Jarvis, B. (1985) In Rapid Methods and Automation in Microbiology and Immunology (ed.
K-O. Habermehl), Springer, Berlin, p. 593.
Kersters, K, Pot, B., Dewettinek, D., Torek, V., Vancanneyt, M., Vauterin, L. and
Vandamme, P. (1994) In Bacterial Diversity and Systematics (eds F.G. Priest., A.
Ramos-Cormenzana and B. Tindall), Plenum Press, New York, p. 5l.
Kohler, G. and Milstein C (1975) Nature, 256,495.
Krysinski, E.P. and Heimseh, R.C (1977) Applied and Environmental Microbiology,
33,947.
Logan, N.A and Berkeley, RCW. (1981) In The Aerobic Endosporeforming Bacteria:
Classification and Identification (eds R.CW. Berkeley and M. Goodfellow), Aca-
demic Press, London, p. 105.
Macario, AJ.L. and Conway de Maeario, E. (1985) Monoclonal Antibodies Against
Bacteria, Vols. 1 and 2, Academic Press, New York.
Magee, J. (1993) In Handbook of New Bacterial Systematics (eds M. Goodfellow and
AG. O'Donnell), Academic Press, London, p. 383.
 References 269
Magee, J.T., Hindmarch, J.M. and Meecham, D.F. (1983) Journal of Medical
Microbiology, 16, 483.
Manafi, M., Kneifel, W. and Bascomb, S. (1991) Microbiological Reviews, 55, 335-348.
Mattingly, J.A. (1984) Journal of Immunological Methods, 73,147.
Mattingly, J.A and Gehle, W.D. (1984) Journal of Food Science, 49, 807.
Mattingly, J.A., Robison, B.J., Boehm, A and Gehle, W.A. (1985) Food Technology,
39,90.
Meuzelaar, H.L.C and Kistemaker, P.G. (1973) Analytical Chemistry, 45, 587.
Meuzelaar, H.L.C, Kistemaker, P.G., Eshuis, W. and Engel, H.W.B. (1976) In Rapid
Methods and Automation in Microbiology (eds S.W.B. Newsom and
H.H. Johnston), Learned Information, Oxford, p. 225.
Miles, CA, Mackey, B.M. and Parsons, S.E. (1986) Journal of General Microbiology,
132,939.
Miller, L. (1984) Hewlett Packard Gas Chromatography Application Note 228-37.
Minnich, S.A. (1978) MS Thesis, University of Idaho.
Mohr, H.K., Trenk, H.L. and Yeterian, M. (1974) Applied and Environmental
Microbiology, 27, 324.
Moss, CW. (1981) Journal of Chromatography, 203, 337.
Moss, CW. (1985) In Rapid Methods and Automation in Microbiology and Immunology
(ed. K.-O. Habermehl), Springer, Berlin, p. 232.
Moss, CW. and Dees, S.B. (1976) Journal of Clinical Microbiology, 4, 492.
Moss, CW. and Dees, S.B. (1978) Journal of Clinical Microbiology, 8, 772.
Moss, CW., Dees, S.B. and Guerrant, G.O. (1980) Journal of Clinical Microbiology,
12,127.
Muirhead, K.A., Horan, P.K. and Poste, G. (1985) Bio/fechnology, 3,337.
Naumann, D., Helm, D. and Schultz, C (1994) In Bacterial Diversity and Systematics
(eds F.G. Priest., A Ramos-Cormenzana and B. Tindall), Plenum Press, New
York, p. 67.
Nelson, W.H. (1991) Modern Techniques for Rapid Microbiological Analysis, VCH
Publishers, New York.
Nelson, W.H. and Sperry, J.F. (1991) In Modern Techniques for Rapid Microbiological
Analysis (ed. W.H. Nelson), VCH Publishers, New York, p. 7.
Nichols, P.D., Henson, J.M., Guckert, J.B. et al. (1985) Journal of Microbiological
Methods,4,79.
Norris, J.R. (1980) In Microbiological Classification and Identification (eds
M. Goodfellow and R.G. Board), Academic Press, London, p. l.
Norris, K.P. (1959) Journal of Hygiene, 57,326.
Odham, G., Larsson, L. and Mardh, P.-A. (1984) Gas Chromatography/Mass
Spectrometry Applications in Microbiology, Plenum Press, London.
O'Donnell, AG., Minnikin, D.E. and Goodfellow, M. (1985) In Chemical Methods in
Bacterial Systematics (eds M. Goodfellow and D.E. Minnikin), Academic Press,
London, p. 13l.
Pettipher, G.L. (1983) The Direct Epifluorescent Filter Technique for the Rapid
Enumeration of Microorganisms, Research Studies Press, Letchworth.
Pettipher, G.L. (1989) Progress in Industrial Microbiology, 26, 19-56.
Phillips, AP. and Martin, K.L. (1985) In Rapid Methods and Automation in Microbiology
and Immunology (ed. K.-O. Habermehl), Springer, Berlin, p. 408.
Phillips, T., Taylor, E. and Eykyn, S. (1980) Infection, 8, S155.
Pilloud, N. and Mollet, B. (1990) Systematic and Applied Microbiology 13, 345.
Pot, B., Vandamme, P. and Kersters, K. (1994) In Chemical Methods in Prokaryotic
Systematics (eds M. Goodfellow and AG. O'Donnell), John Wiley & Sons Ltd,
Chichester, p. 493.
Pot, B., Hartel, C, Ludwig, W. et al. (1993) Journal of General Microbiology, 139, 513.
Potter, M. (1970) Federation Proceedings, 29,85.
270 Methods for the rapid identification of microorganisms
Robison, B.J., Pretzman, C.L and Mattingly, J.A. (1983) Applied and Environmental
Microbiology, 45, 1816.
Saha, RB., Sondag, RJ. and Middlekauff, J.E. (1974) Proceedings of the 9th Congress
of the American Society of Brewing Chemists, American Society of Brewing
Chemists, St Paul, MN, 5, I.
Salton, M.RJ. (1952) Biochimica et Biophysica Acta, 8, 510.
Salzman, G.c., Griffith, J.K. and Gregg, C.T. (1982) Applied and Environmental
Microbiology, 44, 108I.
Schleifer, KH. (1990) Food Biotechnology, 4,585.
Schleifer, KH., Ludwig, W. and Amann, R (1993) In Handbook of New Bacterial
Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic Press,
London, p. 464.
Schmidhuber, S., Ludwig, W. and Schleifer, K-H. (1988) Journal of Clinical
Microbiology, 26, 1042.
Schrade, J.P. (1984) In FDA Bacteriological Analytical Manual, 6th edn, FDA,
Washington, DC, Chapter 8.
Sharpe, A.N. (1994) Food Research International, 27, 237.
Shute, L.A., Gutteridge, C.S., Norris, J.R and Berkeley, RC.W. (1984) Journal of
General MicrobiologJj, 130, 343.
Smith, A.M., Miller, J.5. and Whitehead, D.5. (1979) Journal ofImmunology, 123,1715.
Sneath, P.H.A. (1972) In Methods in Microbiology, Vol. 7A (eds J.R Norris and
D.W. Ribbons), Academic Press, London, p. 29.
Sneath, P.H.A. (1974) International Journal of Systematic Bacteriology, 24, 508.
Sommerfelt, H., Kalland, KH., Raj, P. et al. (1988) Journal of Clinical Microbiology,
26,2275.
Sperber, W.H. and Diebel, RH. (1969) Applied and Environmental Microbiology,
17,533.
Stager, C.E. and Davis, J.R (1992) Clinical Microbiology Reviews, 5, 302-327.
Steel, KJ. (1962) In Microbial Classification (eds G.c. Ainsworth and P.H.A. Sneath),
Cambridge University Press, Cambridge, p. 20.
Swaminathan, B. and Ayres, J.c. (1980) Journal of Food Science, 45, 352.
Tamaoka, J. (1994) In Chemical Methods in Prokaryote Systematics (eds M. Goodfellow
and A.G. O'Donnell), John Wiley, Chichester, p. 463.
Tas, A.c., van der Greef, J., de Waart, J. et al. (1985) Journal of Analytical and Applied
Pyrolysis, 7, 249.
Taylor, G.T. and Marsh, AS (1984) Journal of the Institute of Brewing, 90,134.
Thomason, B.M. (1980) Journal of Food Science, 44, 38I.
Thomason, B.M., Cherry, W.B. and Moody, M.D. (1957) Journal of Bacteriology,
74,525.
Vauterin, L., Swings, J. and Kersters, K (1993) In Handbook of New Bacterial
Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic Press,
London, p. 25l.
Voller, A.D., Bidwell, D. and Bartlett, A. (1980) In Manual ofClinical Immunology, 2nd
edn (eds N.R Rose and H. Friedman), American Society for Microbiology,
Washingon, p. 54.
Wieten, G., Meuzelaar, H.L.C. and Haverkamp, J. (1984) In Gas Chromatography/
Mass Spectrometry Applications in Microbiology (eds G. Odham, L. Larsson and
P.-A. Mardh), Plenum Press, London, p. 335.
 CHAPTER 10

Cleaning and disinfection

in the brewing industry
M. Singh and J. Fisher

10.1 INTRODUCTION

'Cleaning and disinfection in the brewing industry' encompasses such a

broad area that it would be impracticable to concentrate on any particular
aspect in detail. Detergents based on caustic, acids and enzymes have all
been used; active ingredients of disinfectant compositions from most of the
oxidizing and non-oxidizing types have been considered at some stage. All
systems have some disadvantages, and technology has yet to reach a stage
where a universally acceptable product is available. The choice of detergent
and disinfectant must be a carefully considered compromise.
The objectives of this chapter are to provide a broad view of cleaning
and disinfection in terms of methods of application and the types of
chemicals used, and it is not intended to be a detailed academic portrayal
of detergent and disinfectant theory.

10.2 DEFINITIONS

It is important to define the terminology used in this section. Some are

standard British Standard definitions (Anon., 1976), whereas other terms
have evolved within the industry and are occasionally misused. The rea-
sons why something is cleaned and the standard of cleaning required are
obviously very closely interlinked. The standard of cleaning can be broadly
divided into three classifications, the first three of the definitions given
below.
1. Physical cleanliness: visually clean to a satisfactory standard.
Brewing Microbiology, 2nd edn. Edited by F. C. Priest and 1. Campbell.
Published in 1996 by Chapman & Hall, London. ISBN 0412 59150 2
272 Cleaning and disinfection in the brewing industry
2. Chemical cleanliness: clean to a standard where anything in contact
(e.g. product) with the cleaned surface suffers no contamination. This
may also be defined as 'water break free': clean water will completely
wet the surface and drain as a continuous film without forming
rivulets or droplets.
3. Microbiological cleanliness: cleaned to a level such that no physical or
microbiological contamination is present.
4. Detergent: a cleaning agent. By a combination of physical and chemical
processes a detergent removes soil from a substrate.
5. Soil: any substance which is in the wrong place. For example, a glass
of stout is a fortifying beverage. However, spilt on one's clothes, it
becomes soil, quite likely to leave a dark, difficult-to-remove stain.
6. Disinfection: the destruction of microorganisms, but not usually bac-
terial spores; it does not necessarily imply killing all microorganisms,
but reducing them to a level which is harmful neither to health nor to
the quality of perishable goods. The term is applicable to the treatment
of inanimate objects and materials and may also be applied to the
treatment of the skin and other body membranes and cavities (Anon.,
1976). Chemicals used for disinfection are termed disinfectants.
7. Chemical sterilizing agent (sterilant): misnomers used as synonyms for
disinfectant.
8. Sanitization: a term used mainly in the food and catering industry: a
process of both cleaning and disinfecting utensils and equipment
(sanitizer: a chemical agent used for sanitization).
9. CIP: cleaning in place without the need to dismantle.

10.3 STANDARDS REQUIRED WITHIN A BREWERY

The concept of why anything should be cleaned, with the results required,
can be applied specifically to the brewing process. Even simplification of
the discrete stages within the overall process leads to over a dozen stages,
each with its own plant, soil type and the level of cleanliness desired. The
main processes in a brewery are summarized in Table 10.1.
As Table 10.1 indicates, cleaning within the brewery is indeed a complex
operation involving many types of plant and soil. The actual result required
varies from one area to another, but in general, CIP methods are used to
achieve microbiologically clean plant. Cleaning water of good and consis-
tent standard greatly facilitates the operations and provides a basis for
high-quality results.
In order to make recommendations for the type of detergent, apart from
knowledge of the process, soil types and method of application, it is
necessary to have information on water quality. The following is suggested
as the minimum information necessary:
 Standards required within a brewery 273
Table 10.1 Levels of cleanliness required and cleaning methods for various
brewing operations
Process Plant Soil type Result required How obtained
Malting Floors Particulate Physically clean Manually
Screens
Milling Mills Particulate Physically clean Manually
Rollers
Mashing Mash tun Particulate Chemically clean Manually or
Starch crr
Sugar
Protein
Scale
Tannin
Straining Mash tun Particulate Chemically clean Manually or
Lauter tun Starch crr
Mash filter Sugar
Strainmaster Protein
Boiling Copper Particulate Chemically clean Manually or
(Hop kettle) Starch crr
Sugar
Protein
Tannin
Scale
Separation Hop strainer Hops Microbiologically CIP with
Sedimentation Trub clean detergent/
vessel Tannin disinfectant
Whirlpool or sanitizer
Cooling Heat Protein Microbiologically CIP with
exchanger Scale clean detergent/
Rust disinfectant
Particulate or sanitizer
Fermentation Open or Yeast Microbiologically CIP with
closed Protein clean detergent/
fermentors Oxidation disinfectant
products or sanitizer or
Tannin Manually,
Sugar followed by
Scale disinfecting
rinse
Separation Yeast press Yeast Microbiologically CIP with
Centrifuge Protein clean detergent/
disinfectant
or sanitizer or
Manually,
followed by
disinfecting
rinse
274 Cleaning and disinfection in the brewing industry
Table 10.1 continued
Process Plant Soil type Result required How obtained
Conditioning Closed Yeast Microbiologically CIP with
vessels Protein clean detergent/
Scale disinfectant or
Manually,
followed by
disinfecting
rinse
Clarification Filters Yeast Microbiologically CIP with
Centrifuge Protein clean detergent/
disinfectant
or sanitizer
Storage Bright-beer Yeast Microbiologically CIP with
Transport tanks Protein clean detergent/
Road tankers Scale disinfectan t
or sanitizer
Pasteurization Heat Protein Microbiologically CIP with
exchanger Scale clean detergent/
disinfectant
or sanitizer
Packaging Kegs Protein Microbiologically CIP with
(Kegs) Casks Scale clean detergent/
Fillers Particulate disinfectant
deposits or sanitizer
Packaging Bottles Paper Microbiologically In machine by
(Bottles) Filler Glue clean temperature
Scale and causticity
Mould CIP detergent/
disinfectant
or sanitizer
In-bottle Algae Microbiologically Dosing of
pasteuriza tion Scale clean biocide
Can-warming Particulate Scale control
deposits

1. water hardness (temporary and total);

2. total dissolved solids;
3. pH;
4. microbiological quality.
Table 10.2, showing data provided by the South Staffordshire Water
Authority, gives the outline water analysis for a bore hole and the town
supply in the Burton-on-Trent (UK) area. Untreated bore-hole water would
give rise to scale formation if used with alkaline or caustic detergents, or if
heated. The high level of dissolved solids could give rinsing problems.
 Cleaning methods available 275
Table 10.2 Analysis of bore-hole water from a Burton-on-Trent
brewery and the town water supply
Bore-hole supply Town supply
Total hardness
(as:Rpm CaC03) 791 232
Ca + (as ppm CaC03) 473 160
Mg2+ (as ppm CaC03) 318 72
Total dissolved solids
(ppm) 1060 375
Total alkalinity
(as ppm CaC03) 516 103
pH 8.3 7.2

Obviously bore-hole water would not be the liquor of choice for cleaning,
even though it may be relatively inexpensive and readily available. The
town supply is relatively hard and although it may be satisfactory for
cleaning, a detergent of lower specification could be used if this liquor was
softened. It is usually more cost effective to soften water by ion exchange
than chemically by sequestering agents.
Many of the problems attributable to water hardness are eliminated by
the use of acid detergents and sanitizers. However, these types of product
have certain other limitations compared with caustic detergents, and also
their use is restricted mainly to light- and medium-duty cleaning.

10.4 CLEANING METHODS AVAILABLE

A variety of cleaning techniques are used in the brewery. Examples of the

following would be found in most operations:
1. manual cleaning (bucket and brush);
2. immersion cleaning (soak baths, tackle baths);
3. high-pressure spray cleaning (general environmental cleaning, e.g.
walls and floors);
4. foam cleaning, gel cleaning (fillers, small open fermenters);
5. CIP (of vessels and pipework) by recirculation using either high or low
pressure devices.
The brewing process involves the movement of a liquid from one area
to another. At each stage the composition of this liquid changes as it
undergoes complex reactions and is heated, cooled and filtered.
The composition of the soils deposited by these processes also alters as
previously described (Table 10.1). The greater part of the plant is large,
enclosed, inaccessible and not amenable to manual cleaning. It is not
surprising therefore that CIP techniques have evolved to cope with the
276 Cleaning and disinfection in the brewing industry

Fig. 10.1 Sprayball configuration in brewhouse vessels.

demands of towering cylindro-conical fermenters and the kilometres of

pipework associated with the modem brewery.
It is beyond the scope of this chapter to detail the engineering aspects of
modem cleaning systems. However, a basic understanding of the princi-
ples of CIP and the types of system available is necessary, as this chapter is
largely concerned with cleaning and disinfection by CIP.
For cleaning vessels, there are two principal methods of applying
detergents:

Fig. 10.2 Sprayball configuration in conditioning tanks.

 Cleaning methods available 277

Fig.10.3 Rotating spray head in fermenting vessels.

1. through low-pressure devices such as spray balls placed strategically

to give the desired coverage (Figs 10.1 and 10.2) and
2. through rotating high-pressure sprays (Fig. 10.3).
The cleaning of pipework is carried out by recirculating detergent solution
through the pipes. Cleaning velocities and associated flow rates are pre-
sented in Table 10.3, which shows that there is an optimum flow range to
achieve the best results.
Numerous types of cleaning systems (CIP sets) are available to match
closely the equipment to cleaning requirements. There are three basic types:
1. total loss system;
2. partial recovery system;
3. full recovery system.
Diagrams of these three systems are shown as Figs 10.4, 10.5 and 10.6. The
major difference between these systems is that the total loss system neither
278 Cleaning and disinfection in the brewing industry
Table 10.3 Cleaning velocities and associated flow rates (1 min-I)
Pipe outside Cleaning velocity (m s-1)
diameter
(inches) 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3.0
1.0 27.8 37.1 46.3 55.6 64.8 74.1 83.4 92.6
1.5 62.5 83.4 104.2 125.1 145.9 166.8 187.6 208.4
2.0 111.2 148.2 185.3 222.3 259.4 296.4 333.5 370.6
2.5 173.7 231.6 289.5 347.4 405.3 463.2 521.7 579.0
3.0 250.2 333.5 416.9 500.3 583.6 667.0 750.4 833.8
4.0 444.7 592.9 741.1 889.3 1038 1185 1334 1482
6.0 1001 1334 1668 2001 2335 2668 3002 3335
8.0 1779 2372 2964 3557 4150 4731 5336 5929
12.0 4002 5336 ,6670 8004 9338 .1 0672 12006 13340.
A B C
Range A: poor cleaning; B: efficient cleaning; C: no improvement in cleaning, increasing risk
of water hammer and damage to pipework and fittings.

recovers dilute detergent for reuse nor recovers water of pre- or post-rinses.
Recovery systems not only save dilute detergent to be topped up and
reused, they also save post-rinses for subsequent pre-rinse application.
In all cases, the cleaning cycles involve a pre-rinse to remove gross loose
soil, and detergent recirculation to clean the vessel and pipework. After
discarding or recovering the dilute detergent, the vessel is rinsed with clean
water. A disinfectant may be added to this rinse water. Sometimes a pulsed
rinse cycle ('burst rinsing') is used to ensure more efficient use of rinse liquor.
Total loss systems are more likely to be found in applications where

Rinse
water . -_ _ _ _ 4 Detergent
supply
... supply
"'CIP ~~
retum bd
CIP CIP
supply scavenge
~ pump

CIP
supply
pump

Fig.10.4 Single use (total loss) system.

 Cleaning methods available 279

Rinse
Detergent water
supply supply

..._-!-___... ~_ .. CIP
return
.. -ooo()iic)-....c)lQoo-...
CIP
supply

Dilute Rinse
detergent water
tank tank

CIP
supply
pump

Fig. 10.5 Partial recovery system.

gross soiling is present and a recovered detergent may suffer heavy con-
tamination, e.g. yeast recovery plants. In areas of light soiling, e.g. bright
beer tanks, buffer tanks and road tankers, a recovery system may well be
the best choice.

Rinse
Detergent water
supply
, ,
supply

r----T----~I;..........- . CIP
return
.-~-,",-­
CIP
supply

CIP
supply
pump

Fig. 10.6 Full recovery system.

280 Cleaning and disinfection in the brewing industry
The detergent process is a complex mixture of chemical and physical
effects. There is a basic energy requirement to achieve satisfactory
cleaning:
total cleaning energy required = chemical energy + mechanical energy +
thermal energy
As a 'rule of thumb', plant used for hot processes (e.g. mashing,
pasteurization) must be cleaned hot, and plant used for cold processes may
be cleaned cold. In all cases, when a satisfactory cleaning regime has been
established, it is vital to take into account the energy balance equation and
to ensure that the total cleaning energy is available.

10.5 SOIL COMPOSITION

It is important to recognize how soil changes as we progress from the hot

brewhouse through fermentation to cold conditioning: denatured proteins
are gradually replaced by tannins and minerals. The physical volume of
soil changes from one process to another and the effects of time and
temperature further modify soils. The detergent chosen for the cleaning
task must be reactive against at least a proportion of the soil if an acceptable
cleaning standard is to be achieved.

10.6 PROCESS OF DETERGENCY

At least four distinct stages are involved.

1. The wetting of the surface to allow intimate contact between detergent
and soil.
2. Chemical action on the soil. Here, at least three separate interactions
are possible:
(a) action of acids on mineral scales to dissolve the inorganic scale
deposits;
(b) hydrolysis reactions to solubilize proteins;
(c) the saponification by caustic reaction on 'oily' components of soils.
3. The dispersion of large to finely divided soil particles.
4. The suspension in solution of any removed soil.
Cleaning of brewery plant by various soak, spray and recirculation
techniques involves to a greater or lesser extent all of the above physical
and chemical processes. The type of detergent and method of application
determine how effectively the soil is removed. The following sections deal
with the chemistry of detergents with reference to the removal of brewing
soils.
 Chemistry of detergents 281
10.7 CHEMISTRY OF DETERGENTS

Nearly all cleaning processes in a brewery employ either strongly acidic or

strongly alkaline conditions (Fig. 10.7). The reason is simple: soil is most
readily removed by chemical reaction and the two principal reactions,
solution of inorganic soil and organic residues, occur in conditions of excess
acidity (H+) or alkalinity (OH}
The reactions are solubilization of mineral scale by acids:
CaC03 + 2HCl -t CaCh + H 20 + CO2i
solubilization of proteins:
(H+) -t
[complex protein] --_- - NH CH2 CO NH CH2 CO NH-
(OH ) -t 'peptide chain'
and solubilization of oils and fatty acids:
R-COOH + NaOH -t R-COONa + H 20
(R = linear saturated or unsaturated alkyl group ~ Cs).
The use of near-neutral detergents is highly desirable because of han-
dling, storage and effluent considerations. Currently the use of this type of
product is limited mainly to manual operations where chemical energy is
supplemented by physical action. For most CIP operations the use of
detergents based on caustic soda prevails. However, sophisticated acidic
detergents and sanitizers are becoming more widely adopted.

Neat
caustic
detergents
I:'l
'NEAR' NEUTRAL
~~~tdetergents DETERGENTS ~~~:t~~~tions
Use solutions I Use ~
(acids) Beer I solutions
t>§$,WS'Sl ~ • (alkaline)
6S§$S§SSSSSSSSSS$§$SSS'S9 ~

IIIIIIIIIIIIIII
o 2 3 4 5 6 7 8 9 10 11 12 13 14

... Increase H+ pH Scale Increase -OH

Fig. 10.7 pH scale showing relative positions of caustic, alkaline, 'neutral' and
acid detergents.
282 Cleaning and disinfection in the brewing industry
10.8 CAUSTIC AND ALKALINE DETERGENTS

Caustic soda is a very reactive compound. It is the basis of most heavy-duty

detergents used in the brewery. Undoubtedly, caustic soda has a number
of advantages over other chemicals which may be considered in detergent
compositions. The anti-microbial effect of NaOH solutions was demon-
strated by Hobbs and Wilson (1942): Table 10.4 shows the lethal effect on
Bacillus subtilis spores.
In washing beer bottles, temperatures around 70°C and causticities of
2-4% w /w are frequently employed. Under such conditions, a very pro-
nounced anti-microbial effect is achieved and, provided the rinse water is
of potable quality, microbiologically clean bottles may be obtained.
The principal disadvantages of NaOH as a detergent component for
brewery applications are its reactions with water hardness salts and C~
atmospheres. Precipitation of water hardness salts could be one of:
1. Ca(HC03h + 2NaOH ~ CaC03 J, + Na2C03 + H20
2. MgS04 + 2NaOH ~ Mg(OH)2 J, + Na2S04
3. CaS04 + Na2C03 ~ CaC03 J, + Na2S04
The reaction with CO2is
2NaOH + CO2 ~ Na2C03 + H20
It is chemically impossible to prevent these reactions. However, the
deleterious effects of water hardness precipitation can be largely obviated
by the formulation of carefully chosen additives into the caustic soda.
Sequestrants are the principal additives used to prevent the precipitation
of water hardness salts. Sequestrant technology has advanced significantly
over recent years and the next section deals with the role of this group of
compounds in brewery formulations.

Table 10.4 Percentage concentration of NaOH (w Iv) required to

destroy 25% of a population of Bacillus subtilis spores at various
temperatures
Temperature Time (min)
OF °C 1 2 4 8 16
120 49.0 2.44 1.64 1.10 0.74 0.50
130 54.5 2.15 1.44 0.97 0.65 0.44
140 60.0 1.90 1.28 0.86 0.58 0.39
150 65.5 1.66 1.12 0.75 0.51 0.34
160 71.0 1.46 0.89 0.66 0.45 0.30
 Sequestrants 283
10.9 SEQUESTRANTS

The sequestrants currently in use in detergent formulations can be sub-

divided into two main categories, based principally on their mode of action.

10.9.1 Stoichiometric sequestrants

This type functions by chelation, a process in which metal ions are held in
a stable structure by the sequestrant, usually in a fixed ratio of metal ions
to sequestrant. Important members of the group for brewery cleaning
operations are:
1. amino carboxylic acids, nitrilotriacetic acid (NT A) and ethylene-
diamine tetra-acetic acid (EOTA);
2. hydroxycarboxylic acids (gluconic acid and its derivatives);
3. polyphosphates (sodium tripolyphosphate, sodium hexameta-
phosphate).

(a) Amino carboxylic acids

As their tri- and tetra-sodium salts, NTA and EOTA are widely used in
detergent compositions. Both are useful additives if soil contains a high
proportion of inorganic scale, since these sequestrants are able to form
stable chelates with metal ions. This chelating action helps to break down
the mineral content of soils, facilitating removal. The higher stability con-
stants (Table 10.5) of NTA and EOTA chelates with metal ions indicate
species of greater thermodynamic stability (Anon., 1972). The stability
constant of the metal ion/ sequestrant complex MZ is given by the
expression:

Table 10.5 Logarithms of the stability

constan ts of nitrilotriacetic acid and
ethylenediamine tetra-acetic acid
Metal ion NTA EDTA
M~2+ 5.4 8.7
Ca + 6.4 10.6
Mn2+ 7.4 13.8
Ai3+ 16.1
Fe2+ 8.8 14.3
Zn2+ 10.7 16.5
Pb2+ 11.4 18.0
Ni 2+ 11.5 18.6
Cu2+ 13.0 18.8
Fe3+ 15.9 25.1
284 Cleaning and disinfection in the brewing industry
[M+ 1[Z -]
KMZ = [MZ]

As Table 10.5 shows, the log KMZ values of Ca with NTA and EDTA are
6.4 and 10.6, respectively. Therefore the ratios of the concentration of
undissociated complex [MZ] to the dissociated complex [M+] [Z-] are 106.4:1
and 1010.6 : 1 for NTA and EDTA, respectively. Clearly EDTA forms the more
stable complex, and on this evidence would be the preferred choice for
removing mineral scale containing a high proportion of Ca salts.
Table 10.6 shows the complexing efficiency of two commercial grades
of Na3NTA and Na4EDTA within the pH ranges specified. It is evident
that Na3NTA is the more efficient agent in terms of actual amount of metal
ion complexed per gram of active sequestrant (Anon., 1972). Brewery
detergents containing these types of sequestrant are normally optimized
mixtures, their composition depending on their final applications. For
example, simply to soften water by removing Ca2+ and Mg2+ in the form
of soluble chelates, NTA would be more efficient than EDTA. However,
for removing tenacious scale as found in heat exchangers, EDTA would
be more efficient.

(b) Hydroxycarboxylic acids

Gluconic acid is the most commonly used of this type of sequestrant. It is
very soluble in caustic soda and has good long-term and high-temperature
stability. To be fully effective as a chelating agent, gluconic acid must be
used in the presence of free caustic alkali. At pH II, 1 g sodium glue onate

Table 10.6 Complexing efficiency of 1 g of sodium

salts of nitrilotriacetic acid and ethylenediamine
tetra-acetic acid at 25°C
Metal ion Na3NTA NMEDTA
140-156 mg 110mg
ci+
(pH 9.0-12.0) (pH 8.5-11.5)

78mg 64mg
Mg2+
(pH ?-10.0) (pH 8.0-10.5)

247mg 167mg
Cu2+
(pH 3.0-12.0) (pH 1.5-13.0)

217mg 147mg
Fe3+
(pH 1.5-3.0) (pH 1.0-6.0)
 Sequestrants 285

will chelate only about 25 mg CaC03, whereas in 3% NaOH solutions

(pH 14), 1 g will chelate 325 mg CaC03. Sodium gluconate may be formu-
lated into high levels (> 40% w /w) NaOH, principally as liquid mixtures.
This type of product finds widespread use in brewhouse cleaning opera-
tions and may be used at up to 5% w /w NaOH under 'boil out' conditions.
Sodium gluconate is also a very effective sequestrant for Fe3+and AI3+.
This feature is extremely valuable in formulating detergents for washing
beer bottles, since removal and suspension of aluminium is necessary for
dealing with neck foils and labels. Additionally, removal of rust rings from
bottle necks is facilitated by sodium gluconate.

(c) Polyphosphates
Polyphosphates are a very versatile sequestrant group but because of
limited solubility and stability in caustic solutions, are more commonly
incorporated in powdered detergents. Polyphosphates not only behave as
stoichiometric sequestrants at high concentrations, they also have a
threshold effect. Furthermore, they help to build up the detergency of
formulations by improving dispersion and rinsing properties.

10.9.2 Threshold sequestrants

This type of sequestrant functions at a sub-stoichiometric level. A relatively
small amount of a threshold sequestrant can prevent water hardness salts
from producing a scale deposit. Unlike stoichiometric sequestrants which
form soluble chelates, threshold sequestrants act by modifying the physical
characteristics of a precipitate. In this respect they are said to behave as
crystal growth modifiers in that they can distort crystal lattices, thereby
preventing scale formation. They do not actually prevent the precipitation
of insoluble salts. Threshold sequestrants find application in CIP and
bottle-washing detergents, usually in combination with stoichiometric
sequestrants.
Polyphosphates can exhibit threshold effects at sub-stoichiometric levels
but the most notable threshold sequestrants are probably phosphonic acid
derivatives such as amino-tris-(methylenephosphonic acid) (ATMP) and
1-hydroxyethane diphosphonic acid (HEDP). Polyacrylic acid derivatives
can behave both as threshold agents and as dispersants and are often
included in caustic and alkaline CIP detergents.
It is clear that sequestrants play an important role in alkaline and caustic
detergent formulations. Amongst other factors, water conditions and soil
composition in breweries determine the choice of product. A well-
formulated detergent will allow some margin of variation in these para-
meters by combining the properties of both stoichiometric and threshold
sequestrants.
286 Cleaning and disinfection in the brewing industry
10.10 ACIDS

Acids have been on the brewery detergent scene for many years. Originally
used on an 'ad hoc' basis mainly for descaling, their use has spread more
and more to CIP cleaning of vessels and pipework from fermentation
through to racking. One of the major reasons for the introduction of acids
is compatibility with C02. The use of C~-compatible detergents offers
savings in time, detergent and CO2. Elmore (1980) discussed experimental
results of CO2losses with a caustic detergent routine (Table 10.7): the effect
of CO2 on caustic solutions is substantial, even where the level of CO2 is
only 15%. The removal of CO2 from a vessel is highly desirable before
cleaning with a caustic-based detergent. However, where time is at a
premium, this may not be practicable. It takes approximately 5 h to evacu-
ate CO2from a vessel through the only exit available, a standard manway
door, to achieve a level of CO2 which is not detrimental to the caustic
solution (Table 10.8).
The most commonly used acid for brewery applications is phosphoric
acid. Blended with suitable surfactants it can form the basis of a good CIP
detergent; also, it is often used in keg-washing products. When blended
with nitric acid, the mixture is more aggressive but offers bacteristatic
properties if enough nitric acid is present. Also, HN03 is cheaper than
apo4 and has a passivating influence on stainless steel, but is totally
destructive to any copper, brass or phosphor-bronze in the system.
In its raw form, sulphuric acid is rather corrosive to stainless steel, and
should be formulated with a corrosion inhibitor if regular use is envisaged.
H2S04 is the cheapest source of acidity but its inherent detergency is not as
good as H 3P04• It is used in some sanitizer formulations where very low
pH is a prerequisite for most effective use of the biocide.
HCl can be very effectively inhibited from attacking mild steel with the
use of cationic corrosion inhibitors. When formulated in this way, the

Table 10.7 Losses of C02 and NaOH during a caustic detergent routine with 2%
NaOH
CO2
After After Loss in
Initial After detergent final complete Fall in Increase
CO2 pre-rinse cycle rinse cycle NaOH in Na2C03
(%) (%) (%) (%) (%) (%) (%)
97 85 66 57 40 1.00 0.90
50 34 15 13 37 1.15 0.95
40 32 18 16 24 0.75 0.60
32 25 9 7 25 0.75 0.55
23 13 4 3 20 0.45 0.45
15 12 5 5 10 0.70 0.60
0 0.15 0.09
 Surface active agents 287
Table 10.8 Rate of evacuation of C02
from a brewery vessel
Time (h) C02 content (%)
o 98
0.5 45
1 25
1.5 20
2 13
2.5 9
3 7
3.5 6
4 5
4.5 4
5 3
5.5 3

products are used mainly for removing water hardness scale from
bottle-washing machines.
Sulphamic acid is available as a powder and has good descaling prop-
erties. It can be formulated into liquids, but tends to hydrolyse. Since it is
quite expensive and does not offer any significant advantages over H 3P04,
it has limited application.
For any acid, corrosion characteristics need careful consideration. The
mineral acids commonly used in brewery detergents are highly corrosive
to mild steel and must not be used without effective corrosion inhibitors in
the formulation. Figures 10.8 and 10.9 show the corrosion profiles of HCl,
HN03, H 3P04 and H 2S04 against stainless steel types 304 and 316. Clearly,
nitric and phosphoric acids demonstrate far better compatibility with these
metals and may be used without problems (Elmore, 1980). However, the
level of cr in dilution water for detergents should always be checked when
considering acid detergents, because of the increased risk of pitting
corrosion.
Organic acids such as citric, gluconic or tartaric are less powerful
descalers than the mineral acids and much more expensive. They are not
widely used in brewery detergents.

10.11 SURFACE ACTIVE AGENTS

Surface active agents, 'surfactants', are used in brewery detergents only in

certain highly specialized areas, e.g. a type of surfactant is often used in
bottle washing as a defoaming agent and in acidic CIP detergent formula-
tions to improve detergency. Surfactants have four main functions in
brewery detergent formulations.
288 Cleaning and disinfection in the brewing industry
Phosphoric acid Type 304 Type 316
120
.......
0 100
t"
! 80
~CD 60
Q. 40
E
{!2 20
0
0 20 40 60 80 100 0 20 40 60 80 100
Acid by weight ('Yo) AcId by weight (%)
Nitric Acid Type 304 Type 316
120
6t" 100

! 80
~CD 60
Q.
E 40
{!2 20
0
0 20 40 60 80 100 0 20 40 60 80 100
Acid by weight ('Yo) Acid by weight (%)

S <O.311111Vyear D 0.03-0.13 nvn/year

D 0.13-0.75 mmlyear ~ >0.75 mmlyear

Fig. 10.8 Corrosion profile of phosphoric and nitric acids.

1. To reduce the surface tension of the liquid. This will help the wetting
process.
2. A surfactant may produce foam, as a consequence of a reduction in
surface tension and entrainment of air by physical action. The stability
of this foam is then influenced by other surface effects. Foam gen-
eration is undesirable in CIP but vital for 'foam-cleaning' operations.
3. Some surfactants which are water insoluble can be used to defoam
detergent solutions and application is found for these compounds in
bottle-washing products.
4. Apart from helping in the physical processes of detergency by reducing
surface tension, surfactants may also prevent readhesion of soil which
has previously been removed by dispersion and micelle-forming action.
All surfactant molecules have a hydrophilic and hydrophobic structure.
When added to water, the molecules concentrate at the surface in an
attempt to have as much as possible of the hydrophobic portion out of the
 Surface active agents 289

Hydrochloric acid Type 304 Type 316

120

6~ 100
80
!!
~
Gl
60
Co 40
E
~ 20
0
0 5 10 15 20 0 5 10 15 20
Acid by weight (%) Acid by weight (%)
Sulphuric acid Type 304 Type 316
120
6 100
~
!! 80
~ 60
~ 40
E
~ 20

20 40 60 80 100 0 20 40 60 80 100
Acid by weight (%) Acid by weight (%)

~ <0.03 mmlyear D 0.03-0.13mrnlyear

ITIIIIll 0.13-0.75 mmlyear ~ >0.75 mrnlyear

Fig. 10.9 Corrosion profile of hydrochloric and sulphuric acids.

water. Above a certain concentration, the molecules which cannot find

space at the surface form micelles in the bulk solution, capable of holding
soils in suspension and even of solubilizing them.
Surfactants are classified according to the charge on the dissociated
species, the anionic being negatively charged and the cationic positively
charged molecules. There is no net charge on non-ionic surfactants, and the
amphoteric group may be of either charge, depending on pH.
In the anionic group are soaps which are good detergents and lubricants,
and alkyl sulphonates and sulphates which are used in high-foaming
'neutral' detergents or 'washing-up liquids'. Soaps are often used in brew-
ery conveyor lubricants.
Cationic surfactants include fatty amine salts which are used as corro-
sion inhibitors, especially in water treatment applications. Quaternary
ammonium compounds (QACs) are used mainly for their biocidal proper-
ties. Most but not all QACs have an undesirable effect on beer head
retention and have limited brewery applications.
290 Cleaning and disinfection in the brewing industry
For brewery cleaning, non-ionic surfactants play an important role. They
have good hard-surface detergency and are used as both track lubricants
and CIP products. Certain non-ionic surfactants also find application as
defoamers.
The amphoteric compounds are a versatile, if somewhat expensive,
group of compounds. Certain amphoterics are biocidal, whilst others are
used for their 'mildness' in shampoos and bubble bath, but are not of great
significance as brewery detergents.
It is envisaged that the use of specialized surfactants in brewery deter-
gents will increase in the future. This may well be driven by a requirement
to reduce caustic or acid strength of heavy-duty products for environmen-
tal reasons, or to tackle the cleaning problems associated with pH-sensitive
parts of the plant where high or low pH would not be acceptable.

10.12 DISINFECTANTS AND SANITIZERS USED IN BREWERIES

Detergents are used in CIP, soak, spray and manual applications. Disinfect-
ants can also be used in these areas, but there are special types used in the
treatment of recirculating water such as cooling towers and pasteurizers.
Water treatment biocides will be considered in section 10.15.
Table 10.1 showed the required levels of cleanliness at various stages of
the brewing process. The next sections of this chapter will be concerned
with the stages, from the brewhouse to the final packaged product, where
disinfectants are required.
The disinfection stage is used as an extra guarantee of cleanliness, e.g. in
CIP, and in some cases to disinfect and to prevent recontamination, e.g.
soak baths. It does not compensate for a bad detergent stage or badly
designed cleaning processes or equipment. Disinfection may be defined as
'the reduction of microbial numbers to an acceptable level', a level usually
determined by the brewery for a particular area by sampling the final rinse
of CIP or by swabbing surfaces of cleaned equipment. Usually in breweries
the product is examined at a number of stages for the presence of bacteria
and wild yeasts.
A disinfectant used in breweries should comply with the following
factors:
1. compatible with other chemicals;
2. compatible with plant materials;
3. tolerant of hard water;
4. tolerant of soil;
5. low or non-foaming;
6. safe in use, non-irritant to skin;
7. non-tainting;
8. no effect on head retention;
 Oxidizing disinfectants 291
9. broad anti-microbial activity at low concentration and low
temperature;
10. economical in use;
11. of low environmental impact.
Unfortunately, no disinfectant meets all of these requirements. The
chemistry and application advantages and disadvantages of the disinfec-
tants most commonly used in the brewing industry will be discussed.
Disinfectants can be considered as oxidizing and non-oxidizing types.

10.13 OXIDIZING DISINFECTANTS

10.13.1 Halogens
Usually chlorine or iodine disinfectants are used, but recently bromine has
found an application.

(a) Chlorine
Chlorine gas was first used for fumigation in hospitals in 1791 (Sykes, 1972)
but this usage has an important disadvantage: the gas is highly toxic to
humans.
Active chlorine is supplied in two forms:
1. inorganic compounds containing hypochlorite ions either as a liquid,
e~g. NaOCI, or as a powder, e.g. chlorinated trisodium phosphate
(Na3P04.11H20kNaOCl.NaCl;
2. powdered organic chlorine-release agents, e.g. trichloroisocyanurate:

CI
010
II/N,II
r r
N N
I'C/I
CI II CI
o

In solution, all of these compounds hydrolyse to produce hypochlorous

acid and/ or hypochlorite ions, depending on pH.
Acid Alkaline
Ch HOCI ocr
chlorine gas hypochlorous acid hypochlorite ion
In breweries, NaOCl is normally used. It is most stable in a slightly
292 Cleaning and disinfection in the brewing industry
alkaline solution, and for this reason the concentrate is supplied stabilized
with NaOH at a pH of up to 12. In an in-use solution of between 50 and 300
ppm the pH will be between 7 and 9. At the optimum pH for disinfection,
pH 5.0, the solution is less stable. Below pH 5.0 chlorine gas will be pro-
duced. NaOCl hydrolyses proteins making them more soluble and more
easily removed.
NaOCI has many advantages: it is non-foaming, has some hard-water
tolerance, does not leave an active residue and has a wide anti-microbial
spectrum which includes activity against viruses and bacterial spores. It is
fast acting and economical to use. The disadvantages are that it can be
corrosive at high concentrations, irritating to skin, the in-use solution is
unstable, it is affected by organic soil and may give taint problems due to
formation of chlorophenol and chloramines.
A recent application of a chlorine compound is the use of chlorine
dioxide for disinfection of process waters in breweries. The disinfecting
power of Cl02, an unstable, potentially explosive and toxic gas, has been
known since the 1920s. It is soluble in water but unstable in solution. Cl02
can be manufactured thus:

Sodium chlorite itself is explosive: it is supplied in solution and mixed

with HCl when the chlorine dioxide is required. Because of the hazardous
nature of the reactants and product, expensive equipment is required to
manufacture chlorine dioxide at the point of use. In most applications
0.1-0.5 ppm is sufficient to disinfect liquor but higher levels are required
for a positive disinfection action on surfaces.
Cl02 at use dilution overcomes some of the disadvantages of NaOCI in
that it is non-tainting, non-corrosive and non-toxic. The reaction products
of Cl02 with organic substances are far less toxic than those of NaOCI.

(b) Iodine
Iodine itself is not very soluble in water and the vapour is irritating to the
eyes which makes it difficult to handle. Iodine is a very reactive element,
but it is this reactivity which makes it a good disinfectant. Compounds used
in the brewing industry contain iodine complexed with surface-active
agents which act as carriers or solubilizers. These complexes, known as
iodophors, were first introduced in 1949 (Kelley, 1961). The complexes are
water soluble and overcome the handling difficulties of h but retain the
disinfecting power. It is the free h which is the disinfecting agent. h is
released by the iodophor on dilution. The optimum pH for anti-microbial
activity is around 5.
 Oxidizing disinfectants 293

Acid Alkaline
Iz HOI or 103-
greatest some inactive inactive
activity activity
The surface-active agents provide better wetting and organic soil
penetration, so iodophors are less affected by soil than hypochlorite.
lodophors have a broad anti-microbial spectrum which is similar to
hypochlorite, although they tend to be less active against bacterial spores.
Like hypochlorite they are fast acting but tend to be more expensive. They
are used in soak baths and spray applications at up to 10 ppm available Iz
but only low-foam iodophors can be used in CIP. In solution, iodophors
are yellow-brown in colour. The colour can be an advantage, e.g. in a soak
bath the colour indicates the presence of active iodine. In-use solutions are
unstable; as the iodine dissipates the solution will go colourless. The colour
can be a disadvantage by causing staining, particularly with some plastics.
This may also result in taint problems and possible effects on beer head
retention. Unlike hypochlorite, iodine does not react with amines.
lodophors can be corrosive, therefore it is necessary to ensure that the
correct concentration is used. In hand use, skin irritation can be a problem.

(c) Bromine

Bromine itself is not used as a brewery disinfectant (except for the treatment
of recirculating water, described in section 10.15.1) because of handling
difficulties and because it offers no advantages over chlorine. Bromine
hydrolyses in water to form a weak acid, hypobromous acid (HOBr), which
is the primary active anti-microbial agent.

10.13.2 Oxygen-releasing compounds

(a) Peracetic acid

Peracetic acid was first introduced as a disinfectant in 1955 (Hugo, 1991).

The product supplied is an equilibrium mixture:

hydrogen
peracetic acid acetic acid peroxide

As the equation above shows it is soluble in water; the breakdown products

are harmless:
294 Cleaning and disinfection in the brewing industry

There is sometimes concern about the production of oxygen from the

breakdown of peracetic acid, but at the dilution used for disinfection this
should not be a problem. Peracetic acid has a very unpleasant smell and in
the concentrate is corrosive, highly reactive and strongly oxidizing. Because of
these properties it is very unpleasant to handle and is therefore not recom-
mended for manual use. Its main use is as a terminal disinfectant in CIP.
In use, solutions are not very stable and will react with organic materials.
Peracetic acid may attack plant material such as rubber gaskets, and at
higher concentrations pitting corrosion may be a problem in water with
high levels of cr.
Peracetic acid has a wide anti-microbial spectrum which includes viruses
and bacterial spores, but resistant yeasts have been reported. Its activity is
fast and is maintained at temperatures lower than ambient, which is partic-
ularly important in disinfecting bright beer tanks and conditioning vessels.

(b) Hydrogen peroxide

Hydrogen peroxide was introduced as a disinfectant in 1887 (Hugo, 1991).
It is supplied in solution which has a tendency to decompose:
H20 2 ~ H20 + 112D2
Manual use of H 20 2 is not recommended; its main use is in spray
applications such as aseptic packaging. Hydrogen peroxide is bactericidal
and fungicidal, but some bacteria and fungi are less sensitive because they
produce the enzyme catalase which can destroy H 20 2• Since it is slow
acting, a long contact time and/ or elevated temperatures are required for
good anti-microbial activity.

10.14 NON-OXIDIZING DISINFECTANTS

10.14.1 Quaternary ammonium compounds

QACs were first introduced in 1916 (Hugo, 1991), and are probably the
best-known cationic surface-active agents. Their general formula is
 Non-oxidizing disinfectants 295
where X is usually a halide but is sometimes 5042-. R1, R2, RJ and & may be
a variety of alkyl or aryl groups.
QACs are generally poor detergents but good wetting agents allowing
penetration of soil. In solution they ionize producing a cation, the substi-
tuted nitrogen part of the molecule, which provides the surface-active
property. The length of the carbon chain in the group affects the disinfectant
ability; usually Cs-Cs are the most effective. QACs which are effective
disinfectants are generally too high-foaming for CIP applications. Low-
foaming products generally have less anti-microbial activity and may be
difficult to dissolve in water. Their main uses are in soak and manual
cleaning at concentrations above 200 ppm active QAC. Optimum anti-
microbial effect is at about neutral pH, but generally they show activity
between pH 3.0 and 10.0. The use dilution is stable even at higher temper-
atures. They are less affected by organic soil than the halogens, but may be
precipitated by tannins present in beer.
The cationic nature of these products means that they can absorb to
surfaces and may be difficult to rinse off. QACs used in a brewery must be
free rinsing or taint and head retention problems may arise.
The anti-microbial range of QACs is less than for the oxidizing disinfect-
ants. They are generally less effective against Gram-negative bacteria than
against Gram positive. There is also a possibility of resistant bacteria
developing with prolonged use. QACs have limited activity against bacte-
rial spores and almost no effect against viruses. To be effective against
yeasts and moulds, a higher concentration is required. QACs tend to be
expensive in use compared to the halogens.

10.14.2 Biguanides
Biguanides with anti-microbial activity were first reported in the 1950s
(Hugo, 1971). The biguanides are derived from guanidine, a naturally
occurring substance found in cereals and vegetables such as turnips:

guanidine

Biguanides are usually supplied as polymers and in the form of salts,

mostly as the hydrochloride. Optimum activity lies between pH 3.0 and pH
9.0. Below pH 3.0 they are inactive against microorganisms. Above pH 9.0
they are precipitated, therefore care should be taken to ensure thorough
rinsing after caustic detergent treatment. Biguanides are cationic in nature
but are not regarded as surface active. If properly rinsed they will not affect
beer head retention. They are also non-foaming and so are suitable for CIP.
296 Cleaning and disinfection in the brewing industry
Biguanides are also used in soak, manual cleaning and as tunnel pasteurizer
biocides. Most biguanides have equal antibacterial activity against Gram-
negative and Gram-positive bacteria. They are less effective against moulds
and yeasts, especially at acid pH, and are ineffective against bacterial spores
and viruses.

10.14.3 Amphoterics
Amphoterics have been in use as disinfectants since the early 1950s (Block,
1977). They are based on an amino acid, usually glycine, with substituents.
Sometimes the term ampholyte is used because in solution they ionize to
produce cations, anions or zwitterions depending on pH.

H H H 0
I I I II
R--N-C-C-C-OH acid cation
I I I
H H H
H H H 0

R+ -b-b-b-~-o- zwitterion
I I I
H H H
H H H 0
I I I 1/
R - N-C-C-C-O alkaline anion
I I I
H H H

Only certain amphoterics have both disinfecting and surface activity. The
disinfecting ability appears to increase with the number of basic groups.
Amphoterics have good wetting properties but cannot dissolve protein or
beer stone. Although viscous liquids, they are freely soluble in water. Because
of their surface activity, they tend to be too high-foaming for use in CIP.
Amphoteric disinfectants can be expensive to buy and use, but generally
have low mammalian toxicity and low skin irritation, which can justify the
cost. Their main applications are soak, spray and manual use. Optimum
activity lies between pH 3.0 and 9.0, but depends on the formulation. They
are equally effective against Gram-negative and Gram-positive bacteria,
but are less effective against yeasts and moulds, and have little effect against
viruses and bacterial spores. Properties such as soil tolerance and corrosion
vary with the ampholyte, but corrosion is not normally a problem. The
in-use solution, usually 1000 ppm, is stable.

10.14.4 Acid anionics

The active molecule in acid anionics varies considerably. There are two
main types:
 Water treatment 297

1. based on anionic surfactants combined with mineral acid

o
II
R-S-O-H·
II
o

2. based on carboxylic acids, including fatty acids and derivatives

o
R-C~ carboxylic acid
..........
OH

Acid anionics tend to be formulated products with additions to assist

activity and/or solubility. These properties will vary with the product.
Acid anionics tend to have some detergency and wetting ability. The higher
foaming products are unsuitable for CIP. Their general use is in spray
applications. They are not suitable for hand use because of the low pH, 2,
which is required for optimum anti-microbial activity.
In general, acid anionics show good anti-microbial activity against bac-
teria but are less effective against bacterial spores and viruses. Only some
of the carboxylic acid types are active against yeasts and moulds.

10.15 WATER TREATMENT

Treatment of recirculating water in cooling towers and tunnel pasteurizers

is a specialized area. In such situations, biocides are used to prevent slime
developing through microbial growth. In tunnel pasteurizers, slime devel-
opment blocks filters and spray bars, affecting the efficiency of pasteuriza-
tion, and could result in problems with the product. Slime development in
recirculating water is associated with bad odours, possible corrosion and
potential health problems, e.g. legionellosis, caused by Legionella spp.
Any biocides used in tunnel pasteurizers must not affect can ink pig-
ments or crown corks and must be compatible with other chemicals used,
such as scale inhibitors and corrosion inhibitors. Some disinfectants already
mentioned, such as QACs, biguanides and bromine, are used in recirculat-
ing water system. The properties of QACs and biguanides have already
been discussed. The use of bromine is relatively new.

10.15.1 Bromine
Hypobromous and hypochlorous acids can be produced by dissolving
bromo-chloro-dimethylhydantoin, supplied as a solid or powder, in water
in a controlled manner through a brominator:
298 Cleaning and disinfection in the brewing industry
CsaBrClN2Ch + 2H20 ~ CsHsN20 2 + HOBr + HOCI
The water flow through the brominator is regulated to give a residual level
of bromine of 0.5 ppm available bromine. Bromine in this form is less
corrosive than hypochlorite. Bromo-chloro-dimethylhydantoin provides a
relatively safe way of handling bromine.

10.15.2 Aldehydes
This group of biocides contains highly reactive aldehyde groups, respon-
sible for their anti-microbial properties. The main aldehydes for disinfec-
tion are formaldehyde and glutaraldehyde:

R = H formaldehyde
R =CHO-(CH~3 glutaraldehyde

Aldehydes have an optimum pH range of alkaline to slightly acidic.

For good anti-microbial activity, depending on the concentration, 15 min
minimum will be needed, but frequently several hours' treatment is
required.
Formaldehyde was first used as a disinfectant in 1886 (Hugo, 1991), but
is rarely used as a disinfectant in breweries now because of its toxicity, by
inhalation and skin contact, and its reported carcinogenicity. HCHO com-
plexed in formaldehyde-release agents has provided a safe method for use
as preservative or biocide.
Glutaraldehyde, currently used as a tunnel pasteurizer biocide, was
introduced as a disinfectant in 1957 (Hugo, 1991). Glutaraldehyde is
sometimes used in mixtures with QACs to increase their anti-microbial
activity.

10.15.3 Isothiazolinones
Several isothiazolinones are used as biocides and preservatives. The most
common is a mixture of 2-methyl-4-isothiazolin-3-one and 5-chloro-2-
methyl-4-isothiazolin-3-one:

2-methyl-4-isothiazolin-3-one 5-chloro-2-methyl-4-isothiazolin-3-one
These products are used in tunnel pasteurizers and cooling towers as
anti-microbial agents and are added to dilute detergent tanks in CIP to
prolong the useful life of the detergent.
Isothiazolinones tend to be bacteriostatic at very low concentrations, but
 Summary 299

are bactericidal after long contact of several hours. Their anti-microbial

activity includes activity against sulphate-reducing bacteria, which if al-
lowed to grow in tunnel pasteurizers can contribute to corrosion and
produce bad odours and black slime.

10.15.4 BNPD
BNPD (2-bromo-2-nitropropane-1,3-diol) was originally developed as a
preservative for cosmetic, pharmaceutical and toiletry products. It has now
found an application in recirculation water treatment, although BNPD is
only a slow-acting biocide.
Br
I
HO H2C-C-CH2 OH
I
N02
BNPD
Laboratory and field studies have shown BNPD to be effective against a
wide range of microorganisms including Legionella spp. BNPD is readily
soluble in water, but is more stable in acid conditions. At use dilutions it is
non-toxic and the molecule is readily biodegradable.

10.16 STEAM

Steam is very effective for sterilizing equipment. It is effective in killing

bacteria, moulds, yeast, bacterial spores and viruses. It has limited use in
breweries, where in many areas chemical disinfection has replaced steam.
Steam must be wet and air free. In breweries, steam sterilization can take
up to 1.5 h to be effective.
Steam is expensive to produce, is dangerous to handle and bakes some
types of soil onto surfaces, making it more difficult to remove. Heat-
sensitive materials obviously cannot be steam sterilized.

10.17 SUMMARY

In the brewhouse, no terminal disinfectants are used. Cleaning is either

manual or CIP, relying on heat and causticity to obtain a satisfactory level
of cleanliness. From fermentation through to yeast recovery, cleaning and
disinfection are achieved by CIP or manually. The products used here are
detergents followed by disinfectants or combined detergent-sanitizer.
The main criteria for a CIP disinfectant are that it should not foam, be
chemical and plant compatible, have broad anti-microbial activity at low
temperature and concentration, and have no detrimental effect on the
300 Cleaning and disinfection in the brewing industry
product. The disinfectants used in CIP are usually sodium hypochlorite,
peracetic acid and iodophors.
The combined detergent-sanitizers are usually of the acid anionic type
which are particularly effective at low temperatures. For manual cleaning,
safety has to be a major consideration. For this reason, amphoterics are
often used as the disinfectants. The other commonly used disinfectants are
iodophors, chosen because of their broad anti-microbial spectrum and for
their colour, which indicates whether there is any free h left in solution. For
cleaning conditioning vessels and bright beer tanks the disinfectant needs
to be effective at low temperatures and not to have a detrimental effect on
the finished product. Peracetic acid is often used in these areas.
The anti-microbial actives used in water treatment tend to be more
aggressive chemicals. Here there is no cleaning stage, so the active com-
pounds have to be able to cope with soil. Obviously chemical, plant and
packaging compatibilities are major considerations, along with a broad
anti-microbial activity.

REFERENCES

Anonymous (1972) BASF handbook - the 'Irilon' Range of Sequestrants, p. 10.

Anonymous (1976) BS 5283: Glossary of Terms Relating to Disinfection, British
Standards Institute, London, p. 4.
Block, S5. (1977) Disinfection, Sterilisation and Preservation, Lea & Febiger.
Elmore, D.G. (1980) Proceedings of the Brewing Technology Conference, Harrogate, 8.
Hobbs, G. and Wilson, G. S. (1942) Journal of Hygiene, 42,436.
Hugo, W.B. (1971) Inhibitors and Destruction of the Microbial Cell, Academic Press,
London.
Hugo, W.B. (1991) Journal of Applied Bacteriology, 71, 9.
Kelley, F.e. (1961) Proceedings of the Royal Society of Medicine, 54,83l.
Sykes, G. (1972) Disinfection and Sterilisation, Spon, London.
 Index

a-factor 44-5 Alternaria alternata 87, 90, 99, 105

Absidia 90, 106 on malt 99, 101
Absidia corymbifera 93 AMB-ID system 247
Acetic acid bacteria 165-8, 185 Amino acids 14
beer spoilage 168 uptake of isotopically labelled 26
Acetobacter 165-7, 205, 210 yeast nutrition 26
characteristics differentiating species Aminocarboxylic acids 283
of 166 Amino-tris-(methylene phosphoric
general characteristics 165-6 acid) (ATMP) 285
metabolism 165-6 Ammonium ions 14
taxonomy 165 Amphibolic pathways 15
Acetolactate decarboxylase 68-71,140, Amphoterics 296
142 Anabolic pathways 15
Acetyl coenzyme A (acetyl CoA) 20 Anaerobic
fatty acid synthesis 27 bacteria 180-83
Lactobacillus 140 growth test 156
Acids Anaplerotic pathways 15
cleaning/ disinfecting 286-7 Antigen-antibody reaction 222-4
hydrochloric 289 API test kits 200, 260, 261
nitric 286, 288 Arthrobacter globiformis 100
phosphoric 286 Ascomycetes 1-3
sulphanamic 286, 289 Aspergillus 84-115
see also individual acids as storage fungi 89-96
Acinetobacter 183 Aspergillus candidus 102
Acridine orange staining 30 Aspergillus flavus 94, 102
Actidione agar 196-7 Aspergillus fumigatus 89, 91, 93, 94
Adenosine triphosphate (ATP) Aspergillus glaucus 88, 90
bioluminescence 216-8 effect on beer 108
glycolysis 17-19 Aspergillus restrictus 91
method for rapid viable cell counts Aspergillus versicolor 94, 102
216 Aureobasidium pullulans 101
Aerococcus 129, 130 Azotobacter chroococcum 103
Alcaligenes 96, 183, 210
Alcohol dehydrogenase (ADH) 19 Bacillus 210
Aldehydes 298 identification 155
Ales 33, 36, 39 on malt 101
Alternaria 84, 87 nitrosamines 154
302 Index
Bacillus (contd.) electrophoresis (CHEF) 72,
in wort 154 226-8
Bacillus coagulans 154 Cleaning-in-place (CIP) systems 36, 275
Bacillus stearothermophilus 154 detergents 286
Barley, microflora of 83-96 methods 275-80
Basidiomycetes 1, 6 techniques 275-8
Betabacterium 129 sequestrants 285
Betacoccus 128 velocities 278
Biguanides 295-6, 297 Colupulone 141
Binary fission 4 Conductance detection 21-4, 258-60
Biocides 297-9 Continuous fermentation 39
Blastomycetes 2 Copper sulphate medium 195, 197
Brettanomyces 6 Corrosion profiles 288,289
as contaminants 196,204-5 Crabtree effect 25
identification 199 Cryptococcus 210
Brewers' wort composition 14-15 Cylindroconical fermentors 36, 38
Bromine 293, 297
Bromonitropropane diol (BNPD) 299 Debaryomyces 2, 5,199
Budding Detergent
multilateral 4 acid 275
polar 4 alkali 275, 282
chemistry 281
Candida 2, 6, 196, 205 C02-compatible 286
on barley 86 definition 272
in malting 99 near neutral 281
Carbohydrate metabolism process 280
acetic acid bacteria 165-7 recircularization 278-80
during fermentation 17-26 recovery systems 278-89
enterobacteria 169-70 Deuteromycetes 1
Gluconobacter 167 DiacetyI68-71
lactic acid bacteria 131-3 Lactobacillus 241-2
synthesis 24-26 Dimethl sulphide 172-4
Carnobacterium 130, 131 Direct epifluorescent filter technique
Catabolic pathways 15 (DEFT) 30, 198,220-21,257-60
Catabolite repression 25 Disinfectant
Catalase test 155 definition 272
Caustic soda 281 use 290-91
Cell Disinfection 272, 290
composition Dispersants 285
analytical techniques 249-52 DNA 46
physical techniques 248 random amplified polymorphic
counts 28-9 (RAPD) 230-31
DEFT 220-21 recombinant 58,61-71
Chlorine 291-2 transformation 58-9, 61
Chlorine dioxide 292 211m 46-8,60,61
Chromosomes 46
karyotyping 72, 226-8 Electrical detection, see Conductance
transfer 54-6 detection
CIDS spectrum 248, 265 Electrophoresis
Citrobacter freundii 174 whole cell proteins 221-2, 246-7
Cladosporium Embden-Meyerhof-Pamas route
on barley 87, 90 (EMP) 17-23
on malt 99,100 in bacteria 169
Clamped homogeneous electric field Endospore-forming bacteria 154
 Index 303
Enterobacter aerogenes 68, 176 during malting 97
Enterobacter agglomerans, see Rhanella germination stimulation 105
aquatalis gushing 107-8
Enterobacter cloacae 171, 176 malt 106-7
Enterobacteriaceae 169-77, 185 mycotoxins 116, 118
beer spoilage by 170-71
metabolism 169-70 Gas
nitrosamines 177 chromatographic techniques 249-51
Enterococcus 152 production from glucose 156
Enzyme-linked immunosorbent assay Gene probe 224-6, 241-4
(ELISA) 198,222-4,244-6,265 Gene-Trak system 243
Enzyme tests for identification 263-4 Genetic information levels 238-41
Epicoccun 112 Genetic techniques 50-73
Epicoccun nigrum 112 Geotrichum candidum 97,99-100
Equilibrium relative humidity (ERH) 91 Germination stimulation 105
Erwinia herbicola 96 Gliocladium roseum 103
in malting 98, 101 ~-Glucanase 65-66
Ethylenediamine tetra-acetic acid Gluconic acid 283-5
(EDTA) 283-4 Gluconobacter 205
4-Ethyl guaicol 175 general characteristics 167
Eurotium metabolism 167
on barley 86, 90, 93, 95 Glutaraldehyde 298
and gushing 108 Glycolytic sequence 17-23
in malting 99-102 Gram-negative bacteria 163-91
Gram-positive bacteria 127-62
Fatty acid methyl ester (FAME) profiles identification of genera 154
249-52,265 test procedures 154-6
Fermentation 34-40 Gram stain 127-8, 163
attemperation of 36, 38 Guanine + cytosine content (G + C%)
carbohydrate metabolism during 17-24 238-9
computer simulations 39 generic description 238-9
continuous systems 39 Gram positive bacteria 128
improved performance 48
monitoring 39 Hafnia 219
profile 39 Hafnia alvei 169, 176
temperatures 39, 40 Hansenula 5
time for 38-9 Health hazards 111-19
Ferulic acid 175-6 Helminthosporium 84--5
Filamentous fungi 86-8 Heterofermentation 131-2
Flavobacterium 96, 98, 171 test for 156
see also Obesumbacterium Hexose monophosphate pathway
Flocculation (FLO) genes 67 bacteria 166
Flocculation of yeasts 67, 202-3 yeast 23
Flow cytometry 215, 248 Homofermentation 131-2
Fluorescence microscopy 29-30, 219 Hops 141
Fourier transform infrared Humulone 141
spectroscopy (FT-IR) 251-2 Hydrogen peroxide 293-4
Freeze-drying 31 Hydroxycarboxylic acid
Fungi Hydroxyethanol diphosphonic acid
classification of 1 (HEDP) 285
filamentous 86-88 Hypochlorite 291
seedling blight 103-5
storage 88-96 Identification 237-8
Fusarium yeasts 9-10
304 Index
Immunological techniques 193-4, Malting
222-4,244-6 effects of microorganisms on 102-9
Impedimetric techniques 211-13,258-60 Maltworkers' lung 112-14
Iodine 292-3 Mass spectrometers 254-7
Iodophores 292-3 Mating type locus (MAn 44-5
Isothiazolinones 298-9 Megasphaera 164, 183
culture process 186
Killer yeasts 7, 67-8, 203-4 Microbacterium imperiale 96
Klebsiella 169-71, 175-6 Microcalorimetry 214
beer spoilage by 171 Micrococcus 153-4
Klebsiella pneumoniae 175 Micrococcus kristanae 153-4
Klebsiella terrigena 69, 175 Microcolony methods 218-19
Kluyveromyces 4,199 Microflora
barley 83-96
Lactic acid bacteria 128-53 effects on beer 107-10
Lactobacillus 134-47 effects on distilled spirit 110-11
classification 134-8 malt 96-102
distribution in beer and breweries Micropolyspora faeni 112
138-9 Miniaturized methods for
growth in beer 141 identification 260-62
identification of 144-7 Mitochondrial DNA 47
kits for 146-7 Morphological studies 257
isolation media 142-4 Morphology and behaviour 257
phenotypic features of 145 techniques for studying 257-64
serological detection and identifi- Mould contamination assessment
cation 146 119-20
species found in fermented beverages Mucor 100
and related habitats 136 Mucor hiemalis 103
spoilage of beer by 141-2 Mucor pusillus 89
Lactobacillus acidophilus 247 Mycotoxins 114-19
Lactobacillus brevis 138 in beer 115-19
Lactobacillus casei 247
Lactobacillus coryneformis 138-9 NADH2 16, 19, 20
Lactobacillus delbrueckii 138 Nicotinamide adenine dinucleotide
Lactobacillus fermentum 138-9 (NAD+) 19,20,23,25
Lactobacillus lindneri 138-9, 211 Nicotinamide adenine dinucleotide-
Lactobacillus plantarum 98 phosphate (NADP+) 23, 28
Lactococcus 152 Nitrate reductase 177
Laser microprobe mass analyser Nitrite 177
(LAMMA) 257 Nitrogen
Legionella 297, 299 metabolism (yeast) 26
Legionellosis 297 metabolism (lactic acid bacteria) 133
Leuconostoc 129-31, 151-2,247 Nitrosamines 177
Leuconostoc oenos 152 N-nitroso compounds 177
Lipid metabolism 26-8 Nomenclature of yeasts 7-9
Luciferase reaction for ATP Non-fermentative bacteria 183
determination 216
Lysine agar 194-5 Obesumbacterium proteus 171-2
Lysis culture media 185
by lysostaphin and lysozyme 156 detection 230, 231

Malt Pectinatus 180-81

analysis 105-7 beer spoilage by 182-3
microflora 96-102 detection 215
 Index 305
identification of 186 Raka-Ray agar 144
isolation of 186 Rapid detection of microbial spoilage 209-31
taxonomy 181 Rapid identification of
Pectinatus cerevisiiphilus 181, 215 microorganisms 237-67
Pectinatus frisingensis 181 costs 265
Pediococcus 147-52 Rare mating 53-6
classification 147-9 Respiratory hazards 111-14
distribution in beer and breweries 149 Rhanella aquatalis 172-4
effect on beer quality 150 beer spoilage 171, 173-4
growth in beer 150 culture 185
identification of 151 Rhizopus
isolation media 150-51 effect on beer 109
spoilage of beer 150 on malt 99
Pediococcus acidilactici 147 Rhodotorula
on malt 101 on barley 86
Pediococcus damnosus 147-51 on malt 99
detection 211, 215, 218, 230 RNA
Pediococcus inopinatus 149-51 double-stranded 47-8
Pediococcus pentosaceus 149-51 ribosomal 135
Penicillium
and gushing 108 Saccharomyces 4
in malting 100, 103 Saccharomyces bayanus 195
mycotoxins 114, 116 Saccharomyces carlsbergensis 7, 34
psychrotolerant 92-4 Saccharomyces cerevisiae
as storage fungi 88-9 acid washing 206
Pentane dione 141 biochemistry and physiology 13-40,
Peracetic acid 293 49-50
Phenolic off-flavour 52,175-6 flow cytometry 215
Phosphoenolpyruvate (PEP-dependent genetic features 44-8
phosphotransferase) system hybridization 51-3
(PTS) 131 identification 9-10, 200
Physarurn polycephalum 85 life cycle 44-6
Pichia 3, 5, 204-5 rare mating 53-6
Ploidy 46 recombinant DNA methods 58-71
Polar budding 4 spheroplast fusion 56-8
Polymerase chain reaction (PCR) 198, strain typing 71-3, 201, 226-8
228-31,242-3 systematics 6-8
Polyphosphate 283, 285 wild strains 197
Post-harvest colonization 89-90 Saccharomyces cerevisiae var. ellipsoideus 7
Proteins 244 Saccharomyces diastaticus 62-3, 195
examining techniques 244-7 Saccharomyces ellipsoideus 7
Pseudomonas spp. 183 Saccharomyces exiquus 195
in malting 96 Saccharomyces pastorianus 195
Pulsed-field gel electrophoresis (PFGE) Saccharomyces uvarum 7, 34
226-8 Saccharomyces willianus 7
Pyrolysis 254-7 Sanitization 272
Pyrolysis gas chromatography (Py-GC) Sanitizers 281
254 Schizosaccharomyces 215
Seedling blight fungi 103-5
Pyrolysis mass spectrometry (Py-MS)
254-7 Selective media 195-7
Selenomonas lacticifex 181-2, 186
Qua ternary ammonium compounds Sequestrants 282, 283-6
(QACS) 289, 294-5 stoichiometric 283-5
threshold 285
306 Index
Single chromosome transfer 54-6 Yeasts
Soaps 289 aerobic 204--5
Spheroplast fusion 71-2 on barley 86
Spray biochemistry and physiology 13-42,
high pressure 277 48-50
low pressure 277 bottom 34
Sprayball 276 classification 1--6
Staphylococcus 153-4 culture plant operation 32-4
Steam 299 definition 1
Sterilant, see Disinfectant flocculation of 67
Storage genetic features 44-8
bacteria 88 genetic targets for brewing and
fungi 88-96 distilling 48
microflora 88-96 genetic techniques 50--61
Streptobacterium 129 handling 30-32
Streptococcus 129, 152-3 identification of 9-10
Streptomyces spp. 86 improvement 48
Sucrose agar 142 killer 67, 203-4
Surface active agents, see Surfactants life cycle 44--5
Surfactants 287-90 mating types 45
classification 289 metabolism 15-28
new products 73-4
Tatumella 169 nomenclature 7-8
Tetracoccus 129 nutrition 14--15
Thermoactinomyces vulgaris 86, 93 propagation 28-34
and respiratory hazards 112 proteolytic activity 66-7
Thermobacterium 129 pure cultures 30-32
Torulaspora 4, 5 spores 3
Torulopsis 86 sporulation 2--6, 44--5
Transformation 58-61 systematics 1-11
Transposable element, Ty I, 71 top 34
viability of 30
Universal beer agar 142-4 wild 193-200
Uridine diphosphate glucose (UDPG) 24 see also Saccharomyces cerevisiae
Vagococcus 152 Zygomycetes 1
VITEK identification system 262 Zygosaccharomyces 4, 5
Zymomonas 177-80
Water-sensitivity in barley 102-3 beer spoilage by 180
Weissella 152, 155 detection of 185--6
Wild yeasts 193-208 general characteristics 177
detection of 193-8 metabolism 178-80
effects in brewery 201-5 taxonomy 178
elimination of 205-6 Zymophilus 182
identification of 198-201 beer spoilage by 182-3
isolation media 194--7 detection 186