Biotechnology Applications in Beverage Production

BIOTECHNOLOGY APPLICATIONS IN

BEVERAGE PRODUCTION

ELSEVIER APPLIED FOOD SCIENCE SERIES

BIOTECHNOLOGY APPLICATIONS

IN BEVERAGE PRODUCTION

Edited by

C. CANTARELLI Department of Food Science and Technology and Microbiology,

University of Milan, Italy

and

G. LANZARINI Department of Industrial Chemistry and Materials,

University of Bologna, Italy

ELSEVIER APPLIED SCIENCE LONDON and NEW YORK

ELSEVIER SCIENCE PUBLISHERS LTD Crown House, Linton Road, Barking, Essex IGll 8.JU, England

Sole Distributor in the USA and Canada ELSEVIER SCIENCE PUBLISHING CO., INC.

655 Avenue of the Americas, New York, NY 10010, USA

WITH 53 TABLES AND 36 ILLUSTRATIONS

© 1989 ELSEVIER SCIENCE PUBLISHERS LTD Softcover reprint of the hardcover 1 st edition 1989

British Libnry CatalogaiDg in Publcation Data

Biotechnology applications in beverage production. 1. Drinks. Manufacture. Applications of biotechnology I. Cantarelli, C. (Corrado) II. Lanzarini, G. 663

ISBN-13: 978-94-010-6992-2 e-ISBN-13: 978-94-009-1113-0

DOl: 10.1007/978-94-009-1113-0

Libnry of Congress Cataloging.in.PublcatioD Data

Biotechnology applications in beverage production.

Bibliography: p. Includes index. 1. Beverages--Congresses. 2. Biotechnology-Industrial

applications--Congresses. I. Cantarelli, C. (Corrado) II. Lanzarini, G. TP501.B56 1989 663 88-33620

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material

herein.

Special regulations for readers in the USA

This publication has been registered with the Copyright Clearance Center Inc. (Ccq, Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside the USA, should be referred to the

publisher.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission of the

publisher.

PREFACE

Beverage production is among the oldest, though quantitatively most significant, applications of biotechnology methods, based on the use of microorganisms and enzymes.

Manufacturing processes employed in beverage production, originally typically empirical, have become a sector of growing economic importance in the food industry.

Pasteur's work represented the starting point for technological evolution in this field, and over the last hundred years progress in scientifically based research has been intense. This scientific and technological evolution is the direct result of the encounter between various disciplines (chemistry, biology, engineering, etc.).

Beverage production now exploits all the various features of first and second-generation biotechnology: screening and selective improvement of microorganisms; their mutations; their use in genetic engineering methods; fermentation control; control of enzymatic processes, including industrial plants; use of soluble enzymes and immobilized enzyme reactors; development of waste treatment processes and so on.

Research developments involving the use of biotechnology for the purpose of improving yields, solving quality-related problems and stimulating innovation are of particular and growing interest as far as production is concerned. Indeed, quality is the final result of the regulation of microbiological and enzymatic processes, and innovation is a consequence of improved knowledge of useful fermentations and the availability of new ingredients.

The Council of Europe's sponsorship of the work which led to the contributions to this volume is clear evidence of the growing need for adequate information about scientific and technological progress.

The objective pursued in preparing this volume was to bring v

vi PREFACE

together knowledge from various sources, thus providing an up-to-date and hopefully stimulating framework for a unitary approach to a number of problems common to different beverages. This kind of approach, now firmly established and widely adopted in food industry technology, should stimulate cross-fertilization between different production sectors, which suffer from an excessive compartmentalization of technologies and knowledge.

Biotechnology, which includes both basic and applied sciences, is a multidisciplinary field and can thus act as a common denominator, providing a single, integrated picture of the state of the art in beverage production.

C. CANTARELLI

G. LANZARINI

CONTENTS

Preface. v

1. A Proposal for Correct Nomenclature of the Domesticated Species of the Genus Saccharomyces 1

A. VAUGHAN MARTINI and A. MARTINI

2. Microorganisms of Wine 17 CARLO ZAMBONELL!, PATRIZIA ROMANO and GIOVANNA SUZZI

3. Genetic Manipulation of Brewing and Wine Yeast 31 C. FALCONE and L. FRONTAL!

4. Killer Yeasts: Notes on Properties and Technical Use of the Character 41

GIANFRANCO ROSINI

5. The Effects of Carbon Dioxide on Yeasts 49 J. C. SLAUGHTER

6. Microbial Spoilage of Canned Fruit Juices 65 A. CASOLARI

7. Recent and Future Developments of Fermentation Technology and Fermenter Design in Brewing 77

C. A. MASSCHELEIN

VII

viii CONTENTS

8. Fermenter Design for Alcoholic Beverage Production 93 MAURO MORESI

9. Optimal Fermenter Design for White Wine Production 107 MAURO MORESI

10. Factors Affecting the Behaviour of Yeast in Wine Fermentation 127 CoRRADO CANTARELLI

11. On the Utilisation of Entrapped Microorganisms in the Industry of Fermented Beverages. . 153

C. DIVIES

12. Preparation of Yeast for Industrial Use in Production of Beverages 169 KNUT ROSEN

13. Enzymes in the Fruit Juice Industry 189 G. LANZARINI and P. G. PIFFERI

14. Enzymatic Processing of Musts and Wines . 223 ARTURO ZAMORANI

Index . 247

LIST OF CONTRIBUTORS

CORRADO CANTARELLI

Department of Food Science and Technology and Microbiology, University of Milan, Via Celoria 2, 1-20133 Milan, Italy

A. CASOLARI

Adv. Research Laboratory, Plasmon Dietetici Alimentari SpA, 123 Via Nazionale, I-430460zzano Taro, Parma, Italy

C. DIVIES

Microbiology Laboratory, ENSBANA, University of Bourgogne, 21100 Dijon, France

C. FALCONE

Department of Cellular and Developmental Biology, University of Rome, 'La Sapienza', 1-00100 Rome, Italy

L. FRONTALI

Department of Cellular and Developmental Biology, University of Rome, 'La Sapienza', 1-00100 Rome, Italy

G. LANZARINI

Department of Industrial Chemistry and Materials, University of Bologna, Viale Risorgimento 4, 1-40136 Bologna, Italy

A. MARTINI

Department of Plant Biology, University of Perugia, Borgo 20 giugno 74, 1-06100 Perugia, Italy

C. A. MASSCHELEIN

Department of Biochemical Industries, Institute of Fermentation Industries, C.E.R.I.A., 1070 Brussels, Belgium

ix

x LIST OF CONTRIBUTORS

MAURO MORESI

I.M. T.A.F., University of Basilicata, Via Nazario Sauro 85, 1-85100 Potenza, Italy

P. G. PIFFERI

Department of Industrial Chemistry and Materials, University of Bologna, Viale Risorgimento 4, 1-40136 Bologna, Italy

PA TRIZIA ROMANO

Department for the Protection and Improvement of Food and Agricultural Products, Microbiology Division, University of Bologna, Coviolo, 1-42100 Reggio Emilia, Italy

KNUT RosEN

The Danish Distillers Ltd, 10, Raffinaderivej, PO Box 1738, DK-2300 Copenhagen-s, Denmark

GIANFRANCO ROSINI


J. C. SLAUGHTER

Department of Brewing and Biological Sciences, Heriot-Watt University, Chambers Street, Edinburgh, EH1 1HX, UK

GIOV ANNA SUZZI


A. VAUGHAN MARTINI


CARLO ZAMBONELLI


ARTURO ZAMORANI

Department of Agricultural Biotechnology, University of Padova, Via Gradenigo 6, 1-35131 Padova, Italy

Chapter 1

A PROPOSAL FOR CORRECT NOMENCLATURE OF THE DOMESTICATED SPECIES OF

THE GENUS SACCHAROMYCES

A. VAUGHAN MARTINI and A. MARTINI

Department of Plant Biology, University of Perugia, Perugia, Italy

INTRODUCTION

The ability to produce ethanol by fermentation of simple sugars is almost completely restricted to yeasts. Although many species are known to carry out this transformation, only a few are able to yield significant amounts of ethyl alcohol during the natural fermentation of the juices of various sugary fruits. Only a handful of these, are commercially exploitable as actual or potential selected starters.

Zambonelli et al. (see Chapter 2) list Schizosaccharomyces pombe, Torulaspora delbrueckii and Saccharomyces cerevisiae with all its synonyms sensu Yarrow1 as yeast species strictly related to the production of alcoholic beverages.

Even though Schizosaccharomyces pombe and Torulaspora delbrueckii have been proposed respectively for the biological deacidification of grape musts containing high quantities of malic acid and as a possible alternative to S. cerevisiae for highly 'pure' fermentation processes, these applications have rarely gone beyond the laboratory or pilot plant levels.

Information on the natural flora of fermenting grape musts may be found in the review by Kunkee and Goswell,2 while the monograph of Phaff et al. 3 may be consulted for a summarized review on yeast species associated with the surface of other sugary fruits or with the production of less conventional alcoholic beverages.

This review will consider only those yeasts that are traditionally considered the main protagonists of alcoholic fermentation, regardless of the name of the beverage, of the carbon source used and of the technological or natural approach to the process. In fact, it is

C. Cantarelli et al. (eds.), Biotechnology Applications in Beverage Production© Elsevier Science Publishers LTD 1989

2 A. VAUGHAN MARTINI AND A. MARTINI

Saccharomyces cerevisiae, together with all its many relatives (S. bayanus, S. chevalieri, S. oviformis, S. pastorianus, etc.), that predominate in the majority of cases?

For the above reasons, this discussion will be centred on the yeasts belonging to the sensu strictu group of the genus Saccharomyces (sensu Yarrow),! with emphasis on: (i) their ecology according to the latest investigations; (ii) the modern taxonomic procedures utilized for discriminating between species; (iii) the present classification based on genetic analysis as well as nDNA-nDNA reassociation data.

Before entering into the actual matter of this report, which is the taxonomy of the species of the sensu stricto complex of the genus Saccharomyces, it would seem appropriate to briefly review some recent findings on the ecology of wine yeasts in nature that denote a peculiar and unsuspected situation.

A SHORT SUMMARY ON mE ORIGIN OF YEASTS ASSOCIATED WIm GRAPE MUST AND WINE

After the first demonstration that yeasts were normal inhabitants of the epidermis of ripe grapes,s Hansen6 proposed the idea that yeast cells, washed off by rain or along with ripe fruits, fall to the ground, where they somehow survive the winter. In summertime they return to the fruit surface carried by various vectors such as wind, air currents or insects. It was later found that the apiculate yeast Kloeckera apiculata is normally present on the surface of numerous fruits such as cherries, gooseberries, grapes, plums and strawberries.7

More recent evidence3 demonstrated that the elliptical yeast S. cerevisiae var. ellipsoideus (at the time considered the yeast of wine as opposed to the brewer's yeast S. cerevisiae) is only rarely present on fruit surfaces, while appearing on the scene only at the end of an initial occupation of the must by K. apiculata (double domination effect).

As a matter of fact, S. cerevisiae and related species were consistently isolated in the past 90 years only in those ecological surveys performed by using an enrichment culture in liquid media (sterile grape must or malt). As a result, a strong selective pressure was imposed by the high sugar concentration (up to 18-19% w/v) and by anaerobic conditions in favour of those species capable of fermenting sucrose, in particular S. cerevisiae and related species. On the contrary, in the few investigations carried out without enrichment,

CORRECf NOMENCLATURE OF THE GENUS SACCHAROMYCES 3

K. apiculata always predominated (>75%), followed by Metschinikowia pulcherrima, a group of film-forming or pigmented species and the yeast-like organism Aureobasidium pullulans.

A series of ecological surveys carried out by using more vigorous preisolation treatments of samples (fast shaking, jet-streaming of surfaces, ultrasonication) confirmed the above conclusions and showed that members of the collective species S. cerevisiae are practically absent from natural surfaces.8 As a result, if high ethanol tolerant yeasts are not natural residents of grape surfaces, their origin must be found elsewhere.

Following the logical supposition that wine yeasts may more easily colonize the winery environment exposed each vintage to billions and billions of cells, Peynaud and DomercqlO demonstrated that various surfaces (floors, walls, ceilings, vats, equipment, etc.) host a yeast flora belonging in a large majority to the collective species S. cerevisiae. In order to verify those conclusions, Rosinill studied the colonization of the surfaces of a newly established winery by using a labelled yeast starter. After two consecutive years of wine-making, all surfaces of the winery were colonized by the labelled S. cerevisiae strain. When during the third year fermentation was allowed to proceed naturally, without the addition of a starter, the grape must was immediately taken over by the winery-resident labelled yeast.

Additional indirect evidence in favour of this peculiar ecological situation is provided by the findings of recent comparisons of electrophoretically separated yeast chromosomes. Johnston and Mortimer12 demonstrated that S. cerevisiae and related strains possess a significantly higher number of medium and small-sized chromosomes than do most other yeast species. This is considered by the authors to be the result of thousands of years of continuous selection for stronger fermenting capabilities, with polymeric genes from a common progenitor continuously duplicating and rearranging in larger numbers of chromosomal units.

At this point, in order to better understand their taxonomic position, we must keep in mind the fact that the wine-associated collective yeast species S. cerevisiae may be considered a 'domesticated organism' living in the wineries rather than circulating in nature.

YEAST CLASSIFICATION: A BRIEF mSTORY

Since the beginning of this century, the procedures used for yeast classification have undergone profound modifications. Initially, class-


ification was mostly carried out by studying morphological characters such as the shape of the cell and/or the macroscopic appearance of the colony. 6 Other formal characters, sexual reproduction and the capability to form mycelium or pseudomycelium, were later introduced by Guilliermond. 13 In the following decades, nutritional tests based on the ability to aerobically utilize different carbohydrates as sole carbon sources (assimilation) or in the absence of oxygen (fermentation) were introduced by the taxonomic school of the Centraalbureau voor Schimmelcultures (CBS) of Delft in Holland. I 4--18 Accordingly, the number of taxonomic tests kept increasing up to the 40 required today for the determination of an unknown yeast. 19

Additional criteria proposed, such as ascospore morphology,20 serological characteristics of the cell wall,21 proton magnetic resonance spectra of cell wall mannans,22 the type of coenzyme Q of the electron transport system,23 the ability to assimilate n-alkanes24 and results of numerical taxonomy25,26 were found to be effective only for discrimination to the genus level.

In spite of the large number of tests required, the separation of two taxa was often established on a simple difference of a single character.27,28 Classical examples of this are the results of taxonomic studies of the past 80 years in which the positive fermentation of galactose or maltose or sucrose was the sole discriminating criteria between the traditionally wine-related yeasts such as S. cerevisiae, S. bayanus, S. chevalieri, S. italicus and S. oviformis, while all their remaining phenotypic properties are essentially identical. Studies by Scheda and Yarrow29,30 demonstrated that fermentation patterns can vary significantly when repeated after a period of time. More recently Rosini et al., 31 in the course of a taxonomic revision of over 1000 wine-associated Saccharomyces strains of the Industrial Yeasts Collection of the Department of Biologia Vegetale of the University of Perugia, Italy, reported that changes in the ability to ferment various sugars appeared randomly among all the old epithets, with a relatively high frequency. Minor genetic modifications may occur so rapidly in fermenting yeast populations that the decision of separating species on the basis of differences in single phenotypic characters, often governed by a single or at the most very few genes, can no longer be accepted. 27 ,32,33

Accordingly, in the latest monograph 1 the only species recognized in the sensu strictu group was S. cerevisiae. This decision was the direct consequence of the numerous observations accumulated on the

CORRECT NOMENCLATURE OF THE GENUS SACCHAROMYCES 5

variabilty of fermentation characters but also of the pressure of the introduction of molecular taxonomy which pointed out the evident inconsistencies of the conventional classification procedure.

STUDY OF MACROMOLECULAR RELATIONSmpS BETWEEN MICROORGANISMS

It is commonly accepted by taxonomists that relationships between organisms are based upon two postulates: common ancestral origin and differentiation due to progressive substitution in nucleotide sequences. In other words, two organisms may be considered conspecific, in spite of their phenotypic expression, only when they have conserved a major portion of their genomes directly descending from a common ancestor. This approach to the classification of microorganisms, known as 'molecular taxonomy', is based on the evaluation of affinities between two organisms at the level of their macromolecules, particularly nuclear DNA. During the past two decades several methods of genome comparison have been proposed as an aid in the classification of yeasts such as DNA base composition expressed as mole percent of guanine plus cytosine (mol %G + C) and more precisely nDNA/nDNA homology. Since it is not the scope of the present work to expand on these methods, the reader is referred to the review by Kurtzman et al. 28

It must be remembered, however, that the taxonomic value of G + C percentages is mainly exclusionary. In fact, while different mol %G + C values between two strains automatically excludes conspecificity, identical values do not necessarily mean that the two taxa belong to the same species. DNA/DNA homology, on the other hand, is much more indicative on the species level since it is an in-vitro reassociation reaction of whole nuclear DNA. On the basis of extensive comparisons between species it was proposed for yeasts that strains exhibiting 80% or higher DNA/DNA relatedness be considered conspecific.27,34 Base sequence divergence, pointing to species separation, is not yet precisely established even though it is commonly accepted that reassociation values below 20% indicate absence of complementarity. 28

The above procedures of molecular analysis have put into dramatic evidence some of the shortcomings of conventional classification procedures and led to the unification of the numerous epithets of


Saccharomyces sensu stricto under a single species due to similar phenotypic characters as well as identical %G + C values.1,35

Conversely, when the same species were subjected to nuclear DNA/DNA reassociation, the situation was somewhat different. In fact, Rosini et al.,31 Vaughan Martini and Kurtzman,36 and Vaughan Martini and Martine7 demonstrated that Saccharomyces sensu stricto is composed of at least three separate species: S. cerevisiae, S. bayanus and S. pastorianus. In addition, recent unpublished data from this laboratory introduced into the scene a fourth relative of S. cerevisiae, S. paradoxus, ecologically separated and characterized by a complete absence of relationships with the alcoholic fermentation environment. These results will be discussed later.

A PRACTICAL APPROACH TO THE CLASSIFICATION OF SPECIES OF SACCHAROMYCES ASSOCIATED WITH

THE ALCOHOLIC FERMENTATION INDUSTRY

The old epithets to which the fermentation industry throughout the world is accustomed are essentially: S. cerevisiae which indicates brewer's top yeast; S. carlsbergensis, the agent responsible for "low" fermentation in brewing; S. ellipsoideus, later called S. cerevisiae var. ellipsoideus: the wine yeast 'par excellence'; S. oviformis, later denominated S. bayanus, believed to be especially endowed for refermentation processes; and S. pastorianus, the agent of fermentation in cold climates.

The history of yeast classification largely coincides with that of the group of species of the genus Saccharomyces defined as sensu stricto by van der Walt,4 which includes the most important strains mentioned above for the alcoholic beverages industry. This group, in fact, is an excellent example of the continuous changes encountered by yeast taxonomy during the last century.

Since the publication in 1912 by Guilliermond of the first taxonomic monograph, Les Levures,38 throughout those of the Dutch School of Delft,1,4,17 the genus Saccharomyces underwent innumerable modifications with the initial tendency of describing numerous new species followed by a period dominated by the practice of grouping more species under a single epithet. This was due to the findings of innumerable workers that the practice of separating species on the basis of the fermentation of a single or a few sugars is unsound because of the extreme variability of these properties, previously


TABLE 1 VARIATIONS IN THE FERMENTATION PROFILES OF

1014 STRAINS OF Saccharomyces sensu stricto CONSERVED IN THE INDUSTRIAL YEAST COLLECfION

OF THE DEPARTMENT OF PLANT BIOLOGY, PERUGIA,

ITALY

1. Acquisition of the ability to ferment galactose: S, bayanus becomes S. cerevisiae: (9 cases) S, oviformis becomes S, cerevisiae: (29 cases)

2. Loss of the ability to ferment galactose: S, cerevisiae becomes S, bayanus: (25 cases)

3, Acquisition of the ability to ferment maltose: S, chevalieri becomes S, cerevisiae: (50 cases) S. fructuum becomes S. cerevisiae: (21 cases)

4. Acquisition of the ability to ferment raffinose: S. italicus becomes S. cerevisiae: (10 cases)

considered cardinal criteria for speciation in Saccharomyces sensu stricto. 29,30 In fact, as already mentioned, the results of a study31 on the variation of some physiological properties of 1014 S. cerevisiae strains conserved for up to 40 years in the Industrial Yeasts Collection of the Dipartimento di Biologia Vegetale of the University of Perugia showed that changes in fermentative characters appear randomly, though consistently, with high frequency (c. 16%) (Table 1). Galactose fermentation, for example, was acquired in 38 cases and lost in 25; maltose fermentation was acquired in 71 cases and raffinose fermentation in 10.

A recent genetic study39 offers an explanation to the fact that taxonomically indistinguishable strains may exhibit rather different fermentative patterns. In the genus Saccharomyces the fermentation of sugars is under the control of families of genes, dispersed and repeated in the genome. Each family includes multiple copies of these genes, generally unlinked though functionally equivalent. An example of the above situation can be the SUC family that contains genes capable of coding for the synthesis of the enzyme invertase, but which can be repeated six times separately in different chromosomes of the genome. As a consequence, different copies of these SUC genes may be either present or completely lacking in different strains of Saccharomyces cerevisiae. In other words, genetically identical strains may appear phenotypically quite different in relation to the fermentation of sugars.

All the above fermentative versatility and variability is very useful for the yeast that can adapt to various environmental conditions, and


also for the technologist who utilizes them as starters; but makes much more miserable the life of those who must classify them and give them a name. It is not surprising, therefore, that, while on one side taxonomists found that the yeast strains selected for alcoholic fermentation belong to a single species (S. cerevisiae), many direct utilizers of these starters still insist on using traditional names, that no longer exist, sometimes even for thirty years.

In fact, if we observe the evolution of the group sensu stricto of the genus Saccharomyces throughout this century (Fig. 1), an extremely mutable situation can be seen. Twenty species were described in Les Levures, the first treatise on yeast classification.38 In the monograph The Yeasts, a taxonomic study of Lodder and Kreger van Rij17 those 20 original species were grouped under eight epithets (S. ellipsoideus was then already reduced to synonymy with S. cerevisiae), while eight more new species had been described after 1912. In the second edition of The Yeasts, a taxonomic study,4 the 16 species of 1952 had been reduced to 8; but 13 newly described species had been introduced, bringing to 21 the number of taxa in Saccharomyces sensu stricto. At that time some of the traditionally considered 'wine yeasts' were still classified separately as: S. cerevisiae, S. uvarum, S. bayanus, S. chevalieri, and S. italicus.

In the following years, additional studies exhaustively showed that the species included in Saccharomyces sensu stricto are ecologically, physiologically and technologically identical. As a consequence, in the third edition of The Yeasts, a taxonomic study 1 all 21 species included in 1970 were reduced under the single name of S. cerevisiae. This conclusion represented the logical consequence of a long series of investigations based also on the novel procedures introduced in yeast taxonomy during the 1960s mentioned earlier. In fact such procedures as electrophoretical analysis of cellular enzymes,40 PMR spectrum22

and antigenic activitfl of cell wall mannans, and the percentages of guanine and cytosine (mol %G + C) of nuclear DNA/5 have all unequivocally demonstrated that the starters of the alcoholic beverages industry, in spite of different technological properties, belong consistently to the same species, S. cerevisiae.

At this point the story could seem finished if, in the meantime, new techniques for the investigation of genetic affinity such as molecular taxonomy had not been introduced. Almost immediately, in fact, the classification of Saccharomyces sensu stricto based on conventional methods began showing many discrepancies. Already in the first


application to yeasts of the nONA/nONA reassociation in vitro,41 S. cerevisiae exhibited a low relatedness with S. uvarum and S. globosus. As described in the section on molecular taxonomy, this method is based on the principle that two organisms can be considered synonymous only when their informational macromolecules (i.e. ONA) possess the same nucleotide sequences.

Following those preliminary results, further nONA/nONA reassociation experiments were carried out using the type strains of some of the traditionally utilized wine species such as S. cerevisiae, S. chevalieri, S. italicus, S. bayanus and S. uvarum. 31 The results showed that these taxa may be separated into two distinct groups in spite of being essentially indistinguishable on a phenotypic level.

This research was continued by Vaughan Martini and Kurtzman,36 considering the type strains of many synonyms of Saccharomyces cerevisiae and later37 analysing several strains more specifically related to wine and beer fermentation. The results, while confirming the division of Saccharomyces sensu stricto into two distantly related groups, also revealed the presence of an intermediate species, S. pastorianus. These results were later largely confirmed by genetic studies. Holmberg42 and Nilsson-Tillgren et al. 43 found significant differences as well as similarities between chromosomes III and V of S. cerevisiae and S. carlsbergensis (S. pastorianus) therefore explaining in part the intermediate relationship (about 52%) between the two taxa. Later, three separate studies44-46 confirmed the distant relationship between S. cerevisiae and S. bayanus by demonstrating a low percentage of mating between strains of the two species.

More recent studies of this group (Vaughan Martini, unpublished data) have brought to light still another member of the complex: Saccharomyces paradoxus47 a yeast species isolated exclusively from natural sources such as tree exudates, insects and soil. When this species was compared by the nONA/nONA optical reassociation technique to the species of the Saccharomyces sensu stricto complex, low as well as intermediate homology values were found (Fig. 2). The fact that this species is homologous to S. cerevisiae for about 50% of its genome, leads to the possible conclusion that it may represent the natural progenitor of the family of yeasts associated with alcoholic fermentation. The three species strictly exclusive of the alcoholic fermentation environment: S. cerevisiae, S. bayanus and S. pastorianus could have been preceeded by S. paradoxus before formal fermentative processes were initiated several thousand years ago. In view of

1912

Le

s L

evur

es

s. cere

vis

iae--

----

. S.

elli

psoi

deus

S.

tur

bida

ns

S. i

licis

S.

vor

derm

anni

S.

sak

e S.

car

tilag

inos

us

S. b

atat

ae

S. t

okyo

1952

Th

e Ye

asts

a

taxo

nom

ic s

tudy

-I

* S.

cer

evis

iae

S. y

eddo

------'

S. w

illia

n us

S. i

nter

med

ius

1_

~ S.

w

illia

n us

S. v

ali

du

s--

----

J·

S. c

orea

nus

S. carlsberge,!s~ ~

S. m

onac

ensl

S S.

uva

rum

S.

log

os

S. c

orea

nus

S. c

arls

berg

ensi

s =r

S.

uva

rum

S.

log

os

1970

Th

e Ye

asts

a

taxo

nom

ic s

tudy

-II

* S.

cer

evis

iae

S. c

orea

nus

* S.

uva

rum

1984

Th

e Ye

asts

a

taxo

nom

ic s

tudy

-III

-o ?> ~ c:: Cl ~ ~ ~ > z o ?> ~ ~ ~

S. b

ayan

us

S. p

asto

rian

us

S. b

aya

nU

SE

}

S. p

asto

rian

us

* S.

bay

anus

S.

ovi

form

is

S. b

etic

us

S. h

eter

ogen

icus

S.

het

erog

enic

us

S. c

heva

lier

i -----,

S. f

ruct

uum

f-

--*

S. c

hev

alie

ri

S. i

talic

us

S. s

tein

eri

r----

* S.

ita

licus

S.

glo

bosu

s S.

glo

bosu

s

S. a

ceti

S. p

rost

oser

dovi

i S.

ole

agin

osus

S.

ole

aceu

s S.

cap

ensi

s S.

dia

stat

icus

S.

hie

nipi

ensi

s S.

inu

sita

tus

S. n

orbe

nsis

S.

abu

lien

sis

S. c

ordu

ben

sis

S. g

adite

nsis

S.

his

pale

nsis

__

_ --

l

FIG

. 1.

E

volu

tion

of

th

e no

men

clat

ure

of

Sac

char

omyc

es

sen

su

stri

cto:

19

12-1

984.

* S.

cer

evis

iae

8 ~ Z ~ ~ ~ gj o '!l 5J trl

Cl

trl Z

C '" ~ R s: ::.:

, ~ 8 ......

......


/ S. paradoxus

1% I "53\ S. bayanus ( 20% ' J. cerevisiae

\% 1 i ~s. pastorianus ~

FIG. 2. Interspecific relationships within Saccharomyces sensu stricto as determined by DNA/DNA reassociation.

the results of molecular taxonomy and traditional genetic analysis within this group, it can be confirmed that the Saccharomyces sensu stricto complex comprises four species variably related between themselves.

CONCLUDING REMARKS

What is the significance of all this new information in the context of alcoholic beverages? For the time being, it can be deduced that the strains associated with wine-making belong to the species S. cerevisiae while there are strong indications that those utilized by the brewing industry are part of the taxon S. pastorianus (which replaces the old name S. carlsbergensis). Any technological variability present can then be referred to the strains within these taxa, rather than maintaining old, confusing generalizations.

This routine may avoid a great deal of confusion since by relating to the name of the species the most important technological properties, the message is transmitted that all strains belonging for instance to S. ellipsoideus must be good fermentation starters or that those of S. oviformis must be the only ones able to carry out a decent refermentation. On the contrary, S. cerevisiae must be considered as the only species related to wine-making which is characterized by a large spectrum of technological properties distributed randomly among its strains.

In spite of the above taxonomic evidence, however, it will be hard to update the jargon of the sector and it will probably be for some

CORRECT NOMENCLATURE OF THE GENUS SACCHAROMYCES 13

time that oenologists and brewers will keep calling their selected starters with the traditional, often out-dated names such as S. ellipsoideus, S. oviformis, S. bayanus, S. carlsbergensis or S. pastorianus.

At this point one may ask: 'How important is the correct name of a yeast in the industry of alcoholic beverages, when it is the technological properties of the strain that count?' This was probably valid when these beverages were produced in small plants; but not today when, for example, an expert oenologist, working in a cooperative winery producing 20,000 metric tons of wine, makes use of a selected starter classified and patented with a wrong name. In this case the name, together with a specific and defined identity, becomes not only important but economically essential. As a consequence, in this world of microbial strains often identified to the smallest sequence of nucleotides, perhaps we should try to begin calling our friendly domesticated yeasts with their actual names, leaving the good old traditions to the history books.

REFERENCES

1. Yarrow, D. (1984). Saccharomyces Meyen ex Reess. In: The Yeasts: A taxonomic study, N. J. W. Kreger-van Rij. (Ed.), Amsterdam; Elsevier Science Publishers, B.V., pp. 379-95.

2. Kunkee, R. E. and Goswell, R. W. (1977). Table wines. In: Alcoholic Beverages, A. H. Rose (Ed.), Academic Press, London, pp. 315-86.

3. Phaff, H. J., Miller, M. W. and Mrak, E. M. (1978). The Life of Yeasts, Harvard University Press, Cambridge, MA.

4. Walt, J. P. van der. (1970). The genus Saccharomyces (Meyen) Reess. In: The Yeasts, a taxonomic study, J. Lodder (Ed.), North-Holland Publisher, Amsterdam, pp. 555-718.

5. Pasteur, L. (1872). Nouvelles experiences pour demontrer que Ie germe de la levure qui fait Ie vin provient de l'exterieur des grains de raisin. c.R. Ac. Sci. Paris, 75,781.

6. Hansen, E. C. (1888). Recherches sur la physiologie et la morphologie des ferments alcooliques. VII. Action des ferments alcooliques sur les diverses especes de sucre. C.R. Trav. Lab. Carlsberg, 2, 143-67.

7. Kloecker, A. (1915). Chronologische Zusammenstellung der Arbeiten fiber Saccharomyces apiculatus van 1870 bis 1912. Centro Bakt., 43, 369-419.

8. Martini, A., Federici, F. and Rosini, G. (1980). A new approach to the study of yeast ecology of natural substrates. Can. 1. Microbiol., 26, 856-9.

9. Rosini, G., Federici, F. and Martini, A. (1982). Yeast flora of grape berries during ripening. Microb. Ecol., 8,83-9.


10. Peynaud, E. and Domercq, S. (1959). A review on microbiological problems in wine-making in France. Amer. J. Vitic. Enol., 10,69-77.

11. Rosini, G. (1984). Assessment of dominance of added yeast in wine fermentation and origin of Saccharomyces cerevisiae in wine-making. J. Gen. Appl. Microbiol., 30, 249-56.

12. Johnston, J. R. and Mortimer, R. K. (1986). Electrophoretic karyotyping of laboratory and commercial strains of Saccharomyces and other yeasts. Int. J. Syst. Bacteriol., 36,569-72.

13. Guilliermond, A. (1913). Nouvelles observations sur la sexualite des levures. Arch. Protistenk., 28, 52-77.

14. Stelling-Dekker, N. M. (1931). Die sporogenen Hefen. Verh. Kon. Ned. Akad. Wet., Afd. Natuurk., Sect. n, 28, 1-547.

15. Lodder, J. (1934). Die anaskosporogenen Hefen. I. Hiilfte. Verh. Kon. Ned. Akad. Wet., Afd. Natuurk., Sect. n, 32: 1-256.

16. Diddens, H. A. and Lodder, J. (1942). Die anaskosporogenen He/en. II. Halfte, North-Holland Publ. Co., Amsterdam.

17. Lodder, J. and Kreger-van Rij, N. J. W. (1952). The Yeasty, a taxonomic study, North-Holland Publ. Co., Amsterdam.

18. Lodder, J. (Ed.) (1970). The Yeasty, a taxonomic study, North-Holland Publ. Co., Amsterdam.

19. Kreger van-Rij, N. J. W. (Ed.) (1984). The Yeasty: a taxonomic study, Elsevier Science Publishers, B.V., Amsterdam.

20. Kurtzman, C. P., Smiley, M. J. and Barker, F. L. (1975). Scanning electron microscopy of ascospores of Debaryomyces and Saccharomyces. Mycopathol. Mycol. Appl., 55,29-34.

21. Tsuchiya, Y., Fukuzawa, Y., Taguchi, M., Nakase, T. and Shinoda, T. (1974). Serological aspects of yeast classification. Mycopathol. Mycol. Appl., 53,77-91.

22. Gorin, P. A. J. and Spencer, J. F. T. (1970). Proton magnetic resonance spectroscopy-An aid in identification and chemotaxonomy of yeasts. Adv. Appl. Microbiol., 13, 25-87.

23. Yamada, Y. and Kondo, K. (1972). Taxonomic significance of the coenzyme Q system in yeasts and yeast-like fungi (2). In: Fermentation Technology Today Pro. Wth Int. Ferment. Symp., Soc. Ferment. Technol., G. Terui (Ed.), Osaka, Japan, pp. 781-4.

24. Bos, P. and De Bruyn, J. C. (1973). The significance of hydrocarbon assimilation in yeast identification. Antonie van Leeuwenhoek, 39, 99-107.

25. Campbell, J. (1973). Computer identification of yeasts of the genus Saccharomyces. J. Gen. Microbiol., 77, 127-35.

26. Barnett, J. A., Payne, R. W. and Yarrow, D. (1983). Saccharomyces cerevisiae Meyen ex Hansen. In: Yeasty: Characteristics and Identification, Cambridge University Press, p. 467.

27. Price, C. W., Fuson, G. B. and PhatI, H. J. (1978). Genome comparison in yeast systematics: delimitation of species within the genera Schwanniomyces, Saccharomyces, Debaryomyces and Pichia. Microbiol. Rev., 42, 161-93.

28. Kurtzman, C. P., PhatI, H. J. and Meyer, S. A. (1983). Nucleic acid


relatedness among yeasts. In: Yeast Genetics-Fundamental and Applied Aspects, J. F. T. Spencer, D. M. Spencer and A. R. W. Smith (Eds.), Springer-Verlag, New York, pp. 139-66.

29. Scheda, R. and Yarrow, D. (1966). The instability of physiological properties used as criteria in the taxonomy of yeasts. Arch. Mikrobiol., 55,209-25.

30. Scheda, R. and Yarrow, D. (1968). Variations in the fermentative pattern of some Saccharomyces species. Arch. Microbiol., 61, 310-16.

31. Rosini, G., Federici, F., Vaughan, A. E. and Martini, A. (1982). Systematics of species of the yeast genus Saccharomyces associated with the fermentation industry. European 1. Appl. Microbiol. Biotechnol., 15, 188-93.

32. Kurtzman, C. P., Smiley, M. J. and Johnson, C. J. (1980). Emendation of the genus Issatchenkia Kudriavzev and comparison of species by deoxyribonucleic acid reassociation, mating reaction, and ascospore ultrastructure. Int. 1. Syst. Bacteriol., 30,503-13.

33. Starmer, W. T., PhatI, H. J., Miranda, M. and Miller, M. W. (1978). Pichia amethionina, a new heterothallic yeast associated with the decaying stems of cereoid cacti. Int. 1. Yst. Bacteriol. 28,433-41.

34. Martini, A. and PhatI, H. J. (1973). The optical determination of DNA-DNA homologies in yeasts. Ann. Microbiol., 23,59-68.

35. Yarrow, D. and Nakase, T. (1975). DNA base composition of species of the genus Saccharomyces. Antonie van Leeuwenkoek, 41,81-8.

36. Vaughan Martini, A. and Kurtzman, C. P. (1985). Deoxyribonucleic acid relatedness among species of the genus Saccharomyces sensu stricto. Int. 1. Syst. Bacteriol., 35, 508-11.

37. Vaughan Martini, A. E. and Martini, A. (1987). Three newly delimited species of Saccharomyces sensu strictu. Antonie van Leeuwenhoek, 52, 77-84.

38. Guilliermond, A. (1912). Les Levures, Octave Doin et Fils, Paris. 39. Carlson, M. (1987). Regulation of sugar utilization in Saccharomyces

species. 1. Bacteriol., 169,4873-7. 40. Yamazaki, M., Goto, S. and Komagata, K. (1983). An electrophoretic

comparison of the enzymes of Saccharomyces yeasts. 1. Gen. Appl. Microbiol., 29, 305-18.

41. Bicknell, J. N. and Douglas, H. C. (1970). Nucleic acid homologies among species of Saccharomyces. 1. Bacteriol. 101, 505-12.

42. Holmberg, S. (1982). Genetic ditIerences between Saccharomyces carlsbergensis and S. cerevisiae. II. Restriction endonuclease analysis of genes of chromosome III. Carlsberg Res. Commun., 47,233-44.

43. Nilsson-Tillgren, T., Gjermansen, T. c., Holmberg, S. and Petersen, J. G. L. (1986). Analysis of chromosome V and the ILV 1 gene from Saccharomyces carlsbergensis. Carlsberg Res. Commun., 51,309-26.

44. Naumov, G. I. (1987). Genetic basis for classification and identification of the ascomycetous yeast. In: The Expanding Realm of Yeast-like Fungi. Proc. Int. Symp. Perspectives of Taxonomy, Ecology ad Phylogeny of Yeast and Yeast-like Fungi, G. S. de Hoog, M. Th. Smith, A. C. M. Weijman. (Eds), Amsterdam, Elsevier Science Publishers, pp. 469-75.


45. Banno, I. and Kaneko, J. (1988). Genetic analysis of taxonomic relation between S. cerevisiae and S. bayanus. 7th Int. Symp. Yeasts, 1-5 August, Perugia, Italy.

46. Hawthorne, D. C. (1988). Recombination and speciation within the genus Saccharomyces. 7th Int. Symp. Yeasts, 1-5 August, Perugia, Italy.

47. Batschinskaya, A. A. (1914). Entwicklungsgeschichte und Kultur des neue HefepiIzes Saccharomyces paradoxus. J. Microbiol. Epidemiol. Immunobiol., 1,231-47.

Chapter 2

MICROORGANISMS OF WINE

CARLO ZAMBONELLI, PATRIZIA ROMANO and GIOVANNA SUZZI

Department for the Protection and Improvement of Food and Agricultural Products, University of Bologna, Reggio Emilia, Italy.

GRAPE MUST AS A NATURAL SOURCE OF NUTRIENTS FOR MICROORGANISMS

A number of microorganisms can grow in grape must, depending on its composition:

-fermentable sugars with the bulk consisting of glucose and fructose;

-nitrogenous compounds, i.e. ammonia, amino acids and polypeptides;

-minerals with the bulk consisting of potassium, calcium, magnesium, phosphates and sulphates;

-growth factors, i.e. pantothenic acid, biotin, thiamine, pyridoxine and others.

The high acidity of must inhibits the growth of many microorganisms; the pH value of grape must ordinarily lies between 3·0 and 3·5. In this range the pH value has a definite selective action allowing the following microorganisms to grow: yeasts, lactic acid bacteria, acetic acid bacteria and moulds.

Another factor influencing the growth of microorganisms in wine-making is the establishment of anaerobic conditions, which inhibit the growth of acid-tolerant aerobic microbes, such as acetic acid bacteria and moulds. Consequently, yeasts and lactic acid bacteria are the only microorganisms able to develop during must fermentation. Acetic acid bacteria (and sometimes moulds) may grow as a consequence of technological errors.

17


18 CARLO ZAMBONELLI, PATRIZIA ROMANO AND GIOVANNA SUZI

FERMENTATION WITH NATURAL OR SELECTED YEASTS

The advantages of performing must fermentation with pure yeast cultures have long been recognized; the use of starters of selected yeast strains (sometimes lactic acid bacteria) initially presented some difficulties. Nowadays, however, wine-makers from traditional wineproducing countries have successfully adopted this technique. In Italy centres for the production and distribution of selected yeasts have been set up to supply large wine-producing areas. 1

The wine microorganisms suitable for use as starter cultures are the yeast Saccharomyces cerevisiae for alcoholic fermentation and the bacteria Leuconostoc oenos for malolactic fermentation.

The natural microflora, which is always present even in musts inoculated with selected strains, can influence wine-making. This problem is overcome by adding sulphur dioxide, which restricts the growth of undesirable microorganisms and favours natural or selected strains of Saccharomyces cerevisiae. The compound, however, is recognized as having a negative effect on man and the desirability of minimizing S02 addition is often emphasized in the literature. On the other hand, its reduction poses various problems which can only partially be overcome through yeast selection. 2

As a consequence, knowledge of microorganisms growing in must assumes great interest. Yeasts only will be considered here, the reader is referred to recent papers by Wibowo et al. 3 and by Davis et al., 4 for information regarding lactic acid bacteria.

YEASTS

A yeast may be defined as an unicellular fungus reproducing by budding or fission.

Three editions of The Yeasts, a Taxonomic Study have been published.5-7 The present scheme of yeast classification is reported in Table 1, in which the genera are asterisked according to their importance.

Very important modifications concern the taxonomy of the genera of relevant oenologicaI interest, i.e. Saccharomyces, Torulaspora and Zygosaccharomyces, which differ as follows:

-Saccharomyces: the vegetative phase is predominantly diploid, conjugation occurring soon after germination of the ascospores.

Sutx

livis

ion:

As

co

my

co

tin

a

Cla

ss:

Hem

iosc

omyc

etes

O

rder

: E

ndom

ycet

ales

F

amil

y: S

perm

opht

hora

ceae

Fam

ily:

Sac

char

omyc

etac

eae

Sub

fam

.: S

chiz

osac

char

om

ycet

oide

ae

Sub

fam

.: N

adso

nioi

deae

Sub

fam

.: L

ipom

ycet

oide

ae

Sub

fam

.: S

acch

arom

ycet

oide

ae

( }-

-Spe

cies

acc

epte

d in

the

gen

us.

Coc

cidi

ascw

..

Met

sch

nik

owia

N

emat

ospa

ra

*. '" S

chiz

osac

char

omyc

es

'" '" H

anse

nias

porQ

"'

Nad

sort

ia

'" '" S

acch

arom

ycod

es

Wic

kerh

amia

Lip

omyc

es

Am

bros

iozy

ma

Art

hro

ascu

s C

itero

myc

es

Cla

visp

ora

Cyn

iclo

myc

es

• Deb

aryo

myc

es

• "'D

ekke

ra

Gui

lJie

rmon

della

··

Han

sen

uIa

·ls

.atc

henk

iD

'" Klu

yver

omyc

es

Lod

tJ.e

rom

yces

P

achy

.ole

n P

achy

ticho

spor

a ·*

Pic

hia

'" '"

'" '" S

accJ

uuom

yces

'" S

acch

arom

ycop

sis

Schw

anni

omyc

es

Spor

opac

hyde

rmia

St

epha

noas

cus

'" '" '"

Tor

ula

spor

a W

icke

rham

ieU

a W

inge

a ..

. Zyg

osac

char

omyc

es

(1)

(6)

(1)

(4)

(6)

(3)

(1)

(1)

(5)

(4)

(1)

(1)

(1)

(1)

(9)

(2)

(1)

(30)

(4

) (l

l)

(1)

(1)

(1)

(56)

(7

) (7

) (1

) (2

) (1

) (3

) (1

) (1

) (8

)

TA

BL

E 1

C

LA

SS

IFIC

AT

ION

OF

YE

AS

TS

Div

isio

n: E

um

yco

lll

Subd

ivis

ion:

Bas

idio

myc

otin

a

Ord

er:

Ust

ihzg

inlli

es

Fam

ily:

Filo

bosi

diac

eae

Chi

onos

phae

ra

FilD

bosi

diel

la

FU

obas

idiu

m

Tel

iosp

ore-

form

ing

Leuc

o.sp

arid

ium

Y

east

s R

hodo

spar

idiu

m

Spor

idio

bolu

s

Ord

er:

Trem

ella

les

Fib

uJob

osid

ium

F

amil

y: S

irob

asid

iace

ae

Siro

bosi

dium

Fam

ily:

T,e

mel

lace

ae

Hol

term

anni

a Tr

emel

la

(1)

(1)

(3)

(6)

(9)

(4)

(1)

(1)

(1)

(9)

Subd

ivis

ion:

Deu

tero

myc

otin

a C

lass

: B

last

omyc

etes

Fam

ily:

Cry

ptoc

occa

ceae

Fam

ily:

Spo

robo

lo

myc

etac

eae

Aci

culo

coni

dium

••

Bre

tttm

omyc

es

···C

andi

da

• Cry

ptoc

occu

s '" '

" '" K

loec

kera

M

alas

sezi

a O

ospa

ridi

um

Pha

ffia

• Rho

doto

rula

S

arci

nos

poro

n

Schi

zob/

asto

spar

ion

Ster

igm

atom

yces

Sy

mpa

diom

yces

'" T

rich

ospo

ron

'" T

,igo

nop

sis

Bul

/era

S

poro

bolo

myc

es

(1)

(9)

(196

) (1

9)

(6)

(2)

(1)

(1)

(8)

(1)

(1)

(6)

(1)

(15)

(1

) (6

) (7

)


-Zygosaccharomyces: the vegetative phase is predominantly haploid, conjugation is usually between individual cells, just before the formation of asci.

-Torulaspora: the vegetative phase is predominantly haploid, conjugation occurring usually between a cell and its bud, just before the ascus is formed.

These three genera were recognized in Lodder and Kreger-van Rij's classification of 1952,6 whereas in Lodder's classification of 19707 they were all included in the genus Saccharomyces. Indeed the different sporification types were not considered significant because sporification can occur after conjugation between haploid cells also in Saccharomyces. Later the genus Saccharomyces was recognized as having predominantly diploid cells and thus the three genera were reinstated. In particular:

-In 1952 Zygosaccharomyces bailii represented a single taxon and continues to do so today, whereas in 1970 it was classified as Saccharomyces bailii. The other species of this genus followed suit.

-In 1952 the genus Torulaspora included the species Torulaspora rosei and Torulaspora delbrueckii. In 1970 these two species were classified as Saccharomyces rosei and Saccharomyces delbrueckii and in 1984 they were recognized as a single species, Torulaspora delbrueckii.

-Therefore in 1970 the genus Saccharomyces also included the genera Torulaspora and Zygosaccharomyces. Nowadays, the genus Saccharomyces consists of only seven species, i.e. Saccharomyces cerevisiae, Saccharomyces exiguus and five other species of no oenological interest.

Compared to the former classifications, the genus Saccharomyces has undergone many relevant modifications, besides those related to the genera Torulaspora and Zygosaccharomyces. The most relevant change is the suppression of all the technologically important species (which are recognized as physiological races),with the exception of Saccharomyces cerevisiae. This elimination was based on the instability of the characters used to differentiate the species, and confirmed by DNA-DNA homology and DNA base composition (guanine (G) + cytosine (C) content).

On the other hand, Winge and Laustsen8 had already shown the interfertility of some Saccharomyces species, recognized at that time.

MICROORGANISMS OF WINE 21

The present classification of the genus Saccharomyces is reported in Table 2.

Yeasts from Spontaneous Fermentation Of the 500 yeast species included in the present classification, only few play an important role in wine-making.

Castelli9 found which yeasts most frequently perform spontaneous must fermentation and pointed out the environmental factors affecting their presence. The initial stages of must fermentation are begun by yeasts of the genera Kloeckera and sometimes Hanseniaspora, which may be accompanied by Saccharomycodes ludwigii and species of the genus Candida. At the end of fermentation, species of the genus Saccharomyces always predominate and species of the genera Torulaspora and Zygosaccharomyces may be present. At that time, these dominant species of the genus Saccharomyces were recognized in the classification as follows: Sacch. ellipsoideus, Sacch. bayanus, Sacch. oviformis, Sacch. mangini, Sacch. uvarum, Sacch. italicus, Sacch. veronae, Sacch. eiegans, Sacch. exiguus. CastellilO considered these species significant for their different characteristics from both ecological and technological points of view. Other authors11 found that these species differed in oenological behaviour and suggested their use in various technological applications. In fact, Sacch. ellipsoideus has always been considered the most suitable for must fermentation, Sacch. bayanus for sparkling wine and Sacch. uvarum for fermentation at low temperatures.

At present, it is difficult to define what importance to give to the different oenological properties of these species, because the number of strains studied was often not significant, even though certain characteristics are more frequent in some physiological races. For example, flocculent strains, the most suitable for refermentation, are frequently found to belong to the physiological race bayanus.

In conclusion, it can be affirmed that studies carried out with the Saccharomyces species recognized in 1952 contribute important information and that it is convenient to take into consideration the physiological races.

Yeast Ecology Recently scanning electron microscopy studies of the distribution of microbial cells on natural surfaces, such as soil particles, 12,13,

grapes,14,15 leaves16 and rhizosphere17 showed that the presence of

22 CARLO ZAMBONELLI, PATRIZIA ROMANO AND GIOVANNA SUZZI

TABLE 2 THE GENERA Saccharomyces, Torulaspora AND Zygosaccharomyces ACCORDING TO THE CLASSIFICATIONS OF LODDER (1970)' AND KREGER-VAN

RIJ (1984)5

1970 1984

Saccharomyces cerevisiae Synonyms: Sacch. cerevesiae var. el/ipsoideus Sacch. vini

, .-" .. ,

_______ J

-----j Saccharomyces bayanus : Synonyms: ~ ......... ___ .. ___ .... Sacch. pastorianus : Sacch. oviformis ______ :

------1

Saccharomyces chevalieri : Synonyms: : ____________ _

Sacch. mangini Sacch. fructuum

, , ______ 1

------, Saccharomyces uvarum : Synonyms: ~ ____________ _ Sacch. carlsbergensis Sacch. logos .......... _:

Saccharomyces cerevisiae

-------------f.r." cerevisiae

----- -------- f.r. bayanus

-------------- f.r. chevalieri

--- ---------- f.r. uvarum

Saccharomyces aceti ------------------- --------------f.r. aceti Saccharomyces capensis -------------- -- ---- ----------f.r. capensis Saccharomyces coreanus ------ -- ------- --------------f.r. coreanus Saccharomyces diastaticus ------- ------- --- ---- -------f.r. diastaticus Saccharomyces globosus- ----- ---------- ------------ --f.r. globosus Saccharomyces heterogenicus- ---------- --------------f. r. heterogenicus Saccharomyces hienipiensis- ----.. ------ --------------f. r. hienipiensis Saccharomyces inusitatus- ----- --------- --------------f.r. inusitatus Saccharomyces italicus ----------------- ------ --------f.r. italicus Saccharomyces norbensis- -- ---- -- ------ -- -- -------- -of. r. norbensis Saccharomyces oleaceus- --------------- ------------- -f. r. oleaceus Saccharomyces oleaginosus-- ---------- - ----------. --of. r. oleaginosus Saccharomyces prostoserdovii- - -------- --------------f. r. prostoserdovii

MICROORGANISMS OF WINE

TABLE 2-contd.

Saccharomyces dairensis-------Saccharomyces dairensis Saccharomyces exiguus Saccharomyces exiguus Saccharomyces kluyveri Saccharomyces kluyveri Saccharomyces telluris Saccharomyces telluris Saccharomyces unisporus Saccharomyces unisporus

Saccharomyces servazzii

1------Torulaspora delbrueckii

Saccharomyces delbrueckii Saccharomyces fermentati Saccharomyces incospicuus Saccharomyces rosei Saccharomyces saitoanus Saccharomyces vafer __ ---J

Saccharomyces pretoriensis ------Torulaspora pretoriensis Saccharomyces bailii Zygosaccharomyces bailii

Saccharomyces bisporus-------Zygosaccharomyces bisporus

Saccharomyces amur:Jae 1 ______ _ Zygosaccharomyces cidri Saccharomyces cidri

Saccharomyces montanus-------Zygosaccharomyces montanus

Saccharomyces eupa~I--____ _ Saccharomyces flore~ Zygosaccharomyces florentinus

23

Saccharomyces microellipsoides'----Zygosaccharomyces microellipsoides

Saccharomyces mrakii'--------Zygosaccharomyces mrakii

Saccharomyces rouxil:·--------- Zygosaccharomyces rouxii

a f.r.: Formerly registered.


microbial cells is not casual. In particular, the sudaces of leaves and fruits can be natural habitats for various microbial species; their cells grow to form microcolonies which firmly adhere to the sudaces.

The major problem of studies on yeast ecology results from the methods used for isolating the microflora. Therefore, comparisonsl 8-20

of the frequency of yeast occurrence were made between past and recent ecological surveys. As a result, the number of recoverable species obtained by enrichment in grape must and consequent isolation9 was much lower than that achieved by other methods.21-24

Knowledge of yeast ecology is also technologically important for providing an exhaustive picture of yeast flora present at crushing. Many of these yeasts, overwhelmed by others during fermentation, were found to develop subsequently, giving unexpected results. Of the natural microflora occurring on grapes and in musts, yeasts of the genus Schizosaccharomyces are generally considered rare microorganisms, nevertheless studies18 carried out with new criteria and methods, showed that these could be more frequent than expected. It is well known that these yeasts pedorm a complete conversion of malic acid to carbon dioxide and ethanol. Therefore, it is possible that the low acidity, frequent in wines of some Italian regions, is due to the malo-alcoholic fermentation by Schizosaccharomyces.

Kunkee and Amerine25 listed the most frequent species of yeast found in grapes, musts and wines from various countries. Table 3 reports this list modified according to the present classification.

Selected Cultures After the early studies by Winge and Laustsen8,26 and by Lindegren and Lindegren,27 yeast genetics and selection for oenology did not progress. This can be ascribed to the difficulty in defining what is required of a yeast selected for oenology, i.e. the difficulty in establishing what are the desirable characteristics of wine yeasts. Numerous characters have been found and studied to date. These can be divided into two categories:

(1) technological characters affecting the wine-making process: -fermentation vigour -alcohol tolerance -sulphur-dioxide resistance -types of cell growth -foaming ability

TA

BL

E 3

Y

EA

ST

S

FO

UN

D

IN

GR

AP

ES

, M

US

TS

A

ND

W

INE

S,

AC

CO

RD

ING

T

O T

HE

C

LA

SS

IFIC

AT

ION

O

F K

RE

GE

R-V

AN

R

IJ

Bre

ttan

omyc

es b

ruxe

llens

is

Bre

ttan

omyc

es c

laus

seni

i B

rett

anom

yces

cus

ters

ii B

rett

anom

yces

int

erm

ediu

s B

rett

anom

yces

lam

bicu

s C

andi

da a

lbic

ans

Can

dida

boi

dini

i C

andi

da c

anta

relli

i C

andi

da c

aten

ulat

a C

andi

da e

dax

Can

dida

gla

brat

a C

andi

da h

elle

nica

C

andi

da h

umic

ola

Can

dida

ing

ens

Can

dida

int

erm

edia

C

andi

da n

orve

gica

C

andi

da p

arap

silo

sis

Can

dida

rug

osa

Can

dida

sak

e C

andi

da s

olan

i C

andi

da s

tella

ta

Can

dida

tro

pica

lis

Can

dida

util

is

Can

dida

ver

onae

C

andi

da v

ersa

tilis

C

andi

da v

ini

Can

dida

zey

lano

ides

C

itero

myc

es m

atri

tens

is

Cry

ptoc

occu

s al

bidu

s C

rypt

ococ

cus

laur

entii

C

rypt

ococ

cus

lute

olus

D

ebar

yom

yces

han

seni

i D

ebar

yom

yces

han

seni

i (C

andi

da f

amat

a)

(198

4)5

Filo

basi

dium

cap

suli

genu

m

Han

seni

aspo

ra g

uill

ierm

ondi

i H

anse

nias

pora

occ

iden

talis

H

anse

nias

pora

osm

ophi

la

Han

seni

aspo

ra u

varu

m

Han

seni

aspo

ra v

alby

ensi

s H

anse

nias

pora

vin

eae

Han

senu

la a

nom

ala

Han

senu

la s

atur

nus

Han

senu

la s

ubpe

llic

ulos

a Is

satc

henk

ia o

ccid

enta

lis

(Can

dida

sor

bosa

) Is

satc

henk

ia o

rien

talis

K

loec

kera

afr

ican

a K

loec

kera

api

cula

ta

Klo

ecke

ra c

ortic

is

Klo

ecke

ra j

avan

ica

Klu

yver

omyc

es m

arxi

anus

va

r. l

actis

va

r. m

arxi

anus

va

r. v

anud

enii

K

luyv

erom

yces

the

rmot

oler

ans

Lodd

erom

yces

elo

ngis

poru

s M

etsc

hnik

owia

pul

cher

rim

a (C

andi

da p

ulch

erri

ma)

M

etsc

hnik

owia

reu

kauf

ii

Pac

hytic

hosp

ora

tras

vaal

ensi

s P

ichi

a be

ssey

i P

ichi

a et

chel

lsii

Pic

hia

fari

nosa

P

ichi

a fe

rmen

tans

P

ichi

a m

embr

anae

faci

ens

(Can

dida

val

ida)

P

ichi

a oh

mer

i

Rho

doto

rula

aur

antia

ca

Rho

duto

rula

glu

tinis

R

hodo

toru

la m

inut

a R

hodo

toru

la r

ubra

Sa

ccha

rom

yces

cer

evis

iae

(and

its

phy

siol

ogic

al

race

s)

Sacc

haro

myc

es e

xigu

us

Sacc

haro

myc

odes

lud

wig

ii

Sacc

haro

myc

opsi

s fib

ulig

era

Sacc

haro

myc

opsi

s lip

olyt

ica

Sacc

haro

myc

opsi

s vi

ni

Schi

zosa

ccha

rom

yces

jap

onic

us

var.

jap

onic

us

var.

ver

satil

is

Schi

zosa

ccha

rom

yces

mal

idev

oran

s Sc

hizo

sacc

haro

myc

es o

ctos

poru

s Sc

hizo

sacc

haro

myc

es p

ombe

To

rula

spor

a de

lbru

ecki

i To

rula

spor

a gl

obos

a Tr

icho

spor

on c

utan

eum

Tr

icho

spor

on f

erm

enta

ns

Tric

hosp

oron

pul

lula

ns

Wic

kerh

amie

lla

dom

ercq

ii

Zygo

sacc

haro

myc

es b

ailii

Zy

gosa

ccha

rom

yces

bis

poru

s Zy

gosa

ccha

rom

yces

flor

enti

nus

Zygo

sacc

haro

myc

es m

icro

elli

psoi

des

Zygo

sacc

haro

myc

es r

ouxi

i

~ ~ ~ ;I>

Z ~ '" o 'r1

~ Z

tTl ~


-film-forming capacity -ability to settle rapidly -ability to ferment at low temperatures -'Killer' factor

(2) qualitative characters affecting wine quality: -production of low volatile acidity (mostly accounted for as acetic

acid) -sulphite production -hydrogen sulphide production -break down of malic acid -production of secondary compounds (acetaldehyde, ethyl acetate,

high-alcohols)

Selected cultures possessing the desirable characteristics are chosen according to their application. For example, yeasts selected for must fermentation are required to develop with disperse cells, whereas yeasts for sparkling wine must be selected for their ability to flocculate. Some characters are always desirable, such as fermentation vigour and alcohol tolerance: others are always undesirable, such as foaming ability and hydrogen sulphide production.

METHODS FOR YEAST SELECTION

Clonal Selection of Natural Variants Clonal strain selection consists of collecting a large number of pure cultures, determining their most suitable oenological characteristics and choosing the best strains.

In nature, strains possessing the more common characteristics, such as high fermentation vigour and alcohol tolerance, are easily recoverable, whereas strains possessing characteristics referred to as rare are recoverable only in a few instances. For example, only 1% of yeasts show flocculation at the highest levels28 and 1% does not produce hydrogen sulphide;29 for this reason it is highly improbable that natural yeast strains possessing both these characteristics can be found.

Even if clonal strain selection does not achieve immediate results, it is a fundamental step in collecting particular yeast strains which can be improved by various techniques.


Classical Methods for Yeast Improvement Saccharomyces cerevisiae has predominantly diploid cells, it can be genetically improved by spore conjugation, as proposed by Winge and Laustsen.26 Good results can be achieved by crossing parental strains which are homozygous for the characters on trial, sporulate well and yield highly viable spores. In some cases, hybridization of spores from diploids with haploid cells gave successful results.

New and genetically improved strains have been programmed and constructed by classical methods: by selective hybridization, Eschenbruch et al. 30 eliminated undesirable properties, such as foaming ability, in wine yeasts; in 1982 Thornton31 increased fermentation vigour and sulphur dioxide resistance in wine yeasts, and in 198532 introduced flocculation ability character into other strains; by spore conjugation Romano et al. 33 obtained a new wine yeast which is highly flocculent, non-hydrogen sulphide forming and suitable for sparkling wine.

Hybridization cannot always be used as a means to manipulate wine yeasts because they are generally homothallic,34 sporulate with low viability or do not sporulate at all,35 due to a possible polyploidy or aneuploidy. 36

Spheroplast (Protoplast) Fusion A technique that shows greatest promise as an aid in the genetic manipulation of wine yeast strains is spheroplast fusion. This technique has great applicability to wine yeast strains, which sporulate poorly or do not sporulate at all. Svoboda,37 van Solingen and van der Plaat,38 and Spencer and Spencer39 developed this technique on yeasts.

Protoplasts are the forms resulting from the removal of the yeast cell wall with lytic enzymes and can be induced to fuse if they are mixed in polyethylene glycol solution. After fusion the product must be induced to regenerate its cell wall in suitable media and to begin cell division.

Genetic improvements of industrial yeasts have already been achieved by spheroplast fusion. Russell and Stewart40 have successfully fused a number of brewer's yeast strains and Hara et al. 41 introduced the 'Killer' factor into cryophilic wine yeast strains.

Unfortunately, the fusion product is often very different from both original partners because the genome of both donors becomes integrated. Consequently, this technique is not specific enough to selectively introduce a single character into a yeast strain. 36

28 CARLO ZAMBONELLI, PATRIZIA ROMANO AND GIOVANNA SUZZI

Induction of Mutants Nutritional mutants of Saccharomyces cerevisiae can be obtained with high frequency on haploid cells, or on spores of diploid strains. Suitable procedures for inducing mutants have been described by Dawes and Hardie42 and by Romano et al. 43 These nutritional mutants are then employed for genetic improvement.

Giudici and Zinnato44 carried out must fermentation with different auxotrophic mutants and found significant differences in the content of high-alcohols.

By using mutants requiring leucine, Rous et al. 45 considerably decreased isoamyl alcohol content.

DISCUSSION

Species of natural yeasts, present in must and in wine, are numerous, but only few are interesting from a biotechnological point of view. Saccharomyces cerevisiae and related physiological races are employed in the main fermentative processes and Schizosaccharomyces species are used to reduce acidity in wine.

In ancient times, wine-makers left must fermentation to spontaneous microftora; nowadays this procedure is considered out of date and all the phases of vinification can be conducted and controlled by using selected and specific cultures.

In addition, the methods of genetic improvement offer a means of programming and constructing new strains of Saccharomyces cerevisiae.

REFERENCES

1. Zambonelli, C. and Tini, V. (1983). Atti Accad. Ita/. Vite Vino, 35,203. 2. Suzzi, G., Romano, P. and Zambonelli, C. (1985). Am. 1. Enol. Vitic.,

36,199. 3. Wibowo, D., Eschenbruch, R., Davis, C. R., Fleet, G. H. and Lee, T. H.

(1985). Am. 1. Enol. Vitic., 36,302. 4. Davis, C. R., Wibowo, D., Eschenbruch, R., Lee, T. H. and Fleet, G. H.

(1985). Am. 1. Enol. Vitic., 36,290. 5. Kreger-van Rij, N. J. W. The yeasts, a taxonomic study. Elsevier Science

Publishers, Amsterdam, (1984). 6. Lodder, J. and Kreger-van Rij, N. J. W., The yeasts, a taxonomic study.

North-Holland Publishing Company, Amsterdam, (1952).


7. Lodder, J., The yeasts, a taxonomic study. North-Holland Publishing Company, Amsterdam, (1970).

8. Winge, O. and Laustsen, O. (1939). C.R. Trav. Lab. Carlsberg Ser. Physiol., 22,337.

9. Castelli, T. (1954). Arch. Mikrobiol., 20,323. 10. Castelli, T. (1960). Lieviti e fermentazioni in enologia L. Scialpi (Ed.),

Roma, pp. 31-40. 11. Ribereau-Gayon, J. and Peynaud, E. (1960). Traite d'Onologie, Vol. 1

Maturation du Raisin, Fermentation Alcoolique, Vinijication, Libraire Poly technique Ch. Berenger, Paris.

12. Gray, T. R. G. (1967). Science, 155, 1668. 13. Kilbertus, G. and Proth, J. (1979). Can. 1. Microbiol., 25, 943. 14. Belin, J. M. (1972). Vitis, 11, 135. 15. Davenport, R. R. (1974). Vitis, 13, 123. 16. Beech, F. W. and Davenport, R. R. (1970). A survey of methods for the

quantitative examination of the yeast flora of apple and grape leaves. In: Ecology of Leaf Surface Microorganisms, T. F. Preece and C. H. Dickinson (Eds) Academic Press, London, New York, p. 139.

17. Locci, R., Petrolini Baldan, E. B., Quaroni, S. and Sardi, P. (1977). Riv. Patol. Veg., 13,49.

18. Florenzano, G., Balloni, W. and Materassi, R. (1977). Vitis, 16, 38. 19. Martini, A., Federici, F. and Rosini, G. (1980). Can. 1. Microbiol., 26,

856. 20. Rosini, G., Federici, F. and Martini, A. (1982). Microb. Ecol., 8,83. 21. Federici, F., Martini, A. and Rosini, G. (1976). Ann. Fac. Agrar. Univ.

Studi di Perugia, 33,483. 22. Federici, F., Rosini, G. and Martini, A. (1977). Ann. Microbiol. (Milan),

27,95. 23. Federici, F., Rosini, G. and Martini, A. (1977). Ann. Fac. Agrar. Univ.

Studi di Perugia, 32, 101. 24. Martini, A. and Federici, F. (1976). Bot. Ital., 110,297. 25. Kunkee, R. E. and Amerine, M. A. (1970). In: The Yeasts, Vol. 3, A. H.

Rose and J. S. Harrison (Eds) Academic Press, New York, p. 5. 26. Winge, O. and Laustsen, O. (1938). C. R. Trav. Lab. Carlsberg ser.

Physiol. 22,235. 27. Lindegren, C. C. and Lindegren, G. (1943). Proc. Natl. Acad. Sci.

USA, 29, 306. 28. Suzzi, G., Romano, P. andZambonelli, C. (1984). Can. 1. Microbiol., 30,36. 29. Zambonelli, C., Soli, M. G. and Guerra, D. (1984). Ann. Microbiol.

(Milan), 34,7. 30. Eschenbruch, R., Cresswell, K. J., Fischer, B. M. and Thornton, R. J.

(1982). Eur. 1. Appl. Microbiol. Biotechnol., 14, 155. 31. Thornton, R. J. (1982). Eur. 1. Appl. Microbiol. Biotechnol., 14, 159. 32. Thornton, R. J. (1985). Am. 1. Enol. Vitic., 36,47. 33. Romano, P., Soli, M. G., Grazia, L., Suzzi, G. and Zambonelli, C.

(1985). Appl. Environ. Microbiol., 50, 1064. 34. Thronton, R. J. and Eschenbruch, R. (1976). Antonie van Leeuwenhoek

1. Microbiol. Serol., 42,503.


35. Gjermansen, C. and Sigsgaard, P. (1981). Carlsberg Res. Commun., 46, 1. 36. Stewart, G. G. (1981). Can. J. Microbiol., 27,973. 37. Svoboda, A. (1976). Folia Microbiol., 21, 193. 38. Van Soiingen, P. and van der Piaat, J. B. (1977). J. Bacteriol., 130,946. 39. Spencer, J. F. T. and Spencer, D. M. (1977). J. [nst. Brew., 83,287. 40. Russell, I. and Stewart, G. G. (1979). J. [nst. Brew., 85,95. 41. Hara, S., Iimura, Y., Oyama, H., Kozeki, T., Kitano, K. and Otsuka, K.

(1981). Agric. Bioi. Chem., 45, 1327. 42. Dawes, I. W. and Hardie, I. D. (1974). Mol. Gen. Genetics, 131, 281. 43. Romano, P., Soli, M. G. and Suzzi, G. (1983). J. Bacteriol., 156,907. 44. Giudici, P. and Zinnato, A. (1983). Vignevini, 10, 63. 45. Rous, C. V., Snow, R. and Kunkee, R. E. (1983). J. [nst. Brew., 89,274.

Chapter 3

GENETIC MANIPULATION OF BREWING AND WINE YEAST

C. FALCONE and L. FRONTAL!

Department of Cellular and Developmental Biology, University of Rome, Rome, Italy

INTRODUCTION

Genetic improvement of beer or wine yeast strains has taken only limited advantage of classical techniques, including the formation of hybrids from strains with different desirable qualities and mutagenesis followed by screening for selection of the desired phenotype. This is mainly due to the polyploid or aneuploid nature of these strains. This characteristic ensures the high genetic stability of industrial strains that is necessary for the maintenance of product quality, but makes both sporulation and mutagenesis very difficult, and so has hence severely limited the use of classical genetic improvement techniques.

In past years, several attempts were made to generate recombination of desirable qualities by protoplast fusion, and some good results were obtained. More recently, several important results have been obtained by transformation of yeast strains with recombinant plasm ids bearing desirable genes. In this review, we will illustrate the general techniques and describe some of the results obtained with these two different genetic manipulations of beer and wine strains.

GENETIC IMPROVEMENT BY PROTOPLAST FUSION

Protoplast fusion allows the exchange of genetic material between two different yeast strains regardless of their chromosome number, mating type or species.

While interspecific fusions have usually yielded very unstable strains, intraspecific protoplast fusion techniques have in some cases

31


32 C. FALCONE AND L. FRONTALI

been successfully used to overcome the poor sporulation and mating abilities of industrial brewing strains of S. cerevisiae.

The first step in the protoplast fusion technique is the removal of the yeast cell wall with snail or microbial lytic enzymes. The resulting protoplasts (or spheroplasts) are osmotically fragile and must be suspended in a medium of high osmotic pressure (e.g. O·8-1·2M sorbitol). Spheroplast fusion is brought about by the use of polyethylene glycol and Ca2+ ions. The fusion products are then induced to regenerate in solid sorbitol-containing media. Selection of fused cells obviously requires the presence in each strain of at least one selectable marker, so that fusion is followed by selection of markers from both strains.

Fusion of several different ale and lager brewing strains has been achieved by Russell and Stewart, l who overcame the difficulty of obtaining auxotrophic industrial strains by fusing ale and lager strains bearing respiratory deficient phenotypes ('petite') due to nuclear or mitochondrial mutations. Fused cells were then selected for respiratory capacity on medium containing glycerol as the sole carbon source. In this experiment, a nonflocculent lager strain was fused with a flocculent haploid strain bearing the FLOI gene and several nutritional requirements. The recombinant was flocculent and had no nutritional deficiencies. A similar fusion between a nonflocculent ale strain and the FLOI haploid yielded giant recombinant colonies with greatly increased sporulation capacities, and tetrad analysis was possible. This analysis showed that the fusion product was a diploid, although it is probable that the original fusion product had a higher ploidy. The fusion of ale and lager strains bearing different flocculation characteristics and different capabilities of fermenting maltotriose was also successful, but the results were of limited practical use.

YEAST TRANSFORMATION

With the advent of recombinant DNA technology, it is now possible to transform yeast cells with DNA fragments or purified genes from the same or other organisms. Yeast transformation was first achieved in 1978 by Hinnen et al., 2 who transformed a Leu 2- strain of S. cerevisiae to Leu 2+ with a segment of yeast DNA containing the yeast Leu 2+ gene, ligated into the E. coli plasmid pBR322. After that, many groups obtained similar results with different genes, vectors, and methodologies.

GENETIC MANIPULATION OF BREWING AND WINE YEAST 33

Efficient transformation essentially requires:

(1) cell permeabil.ization to donor DNA; (2) an appropriate DNA vector; (3) an easily-selectable genetic marker; (4) a suitable recipient strain.

Since exogenous DNA normally does not enter yeast cells, it is necessary to alter the cell wall, which acts as a barrier to DNA uptake. The best results in terms of transformation efficiency have been obtained by partial removal of the cell wall by lytic enzymes, which are commercially available as mixtures of fJ-glucanases (Helicase, Glusulase, Zymolyase, Mutanase). The resulting spheroplasts are suspended in hypertonic medium in the presence of calcium chloride, the fusing agent polyethylene glycol (PEG) and the donor DNA. Spheroplasts are then embedded in 3% agar to allow the regeneration of the cell wall and plated on selective medium that supports only the growth of the transformed cells.

Many auxotrophic strains of yeast are now available in various laboratories as recipient strains for transformation. The plasmids used for yeast transformation are usually yeast-E. coli shuttle vectors, in that they can replicate in E. coli at high copy number. This is important when amplification of the vector is required before yeast transformation. Vectors should carry easily selectable genetic markers for the selection of transformant clones both in E. coli and yeast.

Antibiotic resistance markers are generally used for E. coli. The selection of transformants in yeast is usually obtained by the utilization of genes coding for enzymes of biosynthetic pathways (Leu, His, Ura, Trp) as markers. Bacterial genes that code for Kanamycin and Neomycin resistance and confer on yeast the resistance to the aminoglycoside G418 are useful as selectable markers for the transformation of those yeasts (polyploid strains, uncharacterized species) for which auxotrophic mutations are not available. Finally, vectors should contain unique sites for restriction endonucleases for simplified insertion of the desired gene and for other molecular manipulations.

Integrating Vectors Yeast integrating plasmids (YIp; Fig. l(A» are usually composed of a bacterial DNA plasmid, such as pBR322 (carrying sequences for replication and genetic selection in E. coli) and a yeast selectable


• • • FIG. 1. Schematic representation of vectors used in yeast transformation. Black and white (including shaded and slashed) areas represent bacterial and yeast DNA sequences respectively. Genes for the selection of transformants and the sequences carrying the origin of replication are also shown. See text

for further explanations.

marker (i.e. URA3). These components are common to almost all vectors used for yeast transformation.

YIp vectors do not replicate in yeast but can integrate into the nucleus of the recipient cells (ura-) essentially by recombination between the yeast sequences present on the plasmid and the homologous sequences on the chromosomes. A YIp vector carrying the Leu 2 gene was utilized by Hinnen et al. in the transformation of S. cerevisiae discussed above.

The efficiency of transformation with these vectors is 1-10 transformants/I-'g DNA/l07 cells, and the stability of the acquired phenotype is very high since the excision of the transforming DNA from the chromosome is a rare event (10-9-10-10). One can introduce a foreign gene in such a vector and integrate it stably in one or a few copies in the chromosome.

Episomal Vectors Yeast episomal vectors (YEp; Fig. 1(B)) carry sequences of the 2-l-'m plasmid that is normally present in most strains of S. cerevisiae, where it seems to have no function. This 6000 base pair long plasmid is present at 50-100 copies per cell and exists in two molecular forms that differ in the reciprocal orientation of two unique sequences which are separated by two 600 base pair inverted repeats (Fig. 2).

YEp vectors replicate autonomously at a high copy number and are useful when high expression of the cloned gene is required. Moreover, they are easily recovered from the transformed strains. These vectors


fir ••

Eclll flrllB

FIG. 2. Representation of the A and B forms of the 2/-lm plasmid from S. cerevisiae. IR indicate the inverted repeat regions. Restriction sites for some

endonucleases are also shown.

transform yeast cells at high efficiencies (103_105 transformants/,ug DNA) and do not integrate into the chromosome.

Recombination has been observed however, between this kind of vector and the endogenous 2 ,urn plasmid. The presence of the latter in the same cells strongly increase the stability of the vectors, which are lost at a rate of 1 % per generation on non-selective medium. Plasmids (2,um like) have recently been described in Kluyveromyces drosophilarum3 ,4 and in a group of osmophilic and osmotolerant yeast strains belonging to the genus Zygosaccharomyces. 5,6 Using these plasmids, vectors that successfully transform the respective yeast species have been developed. The search for plasmids in other yeast strains, especially those of industrial interest, could open the possibility for genetic manipulation of these strains.

Replicating Vectors Yeast replicating vectors (YRp; Fig. 1 (C» contain sequences (derived from yeast chromosomal DNA or from other organisms) that can function as origins of DNA replication (autonomously replicating sequences or ARS) in S. cerevisiae. These plasmids transform yeast at high efficiencies (102_103 transformants/,ug DNA) and are present in high copy numbers per cell. The transform ants are very unstable and the loss of the transforming phenotype is much greater than 1 % per generation in non-selective medium.


Expression and Secretion Vectors Vectors have also been constructed for the expression and/or secretion of cloned gene products in yeast. They contain strong yeast promoter and terminator sequences that direct high and precise expression of the cloned genes. Promoters can be constitutive (alcohol dehydrogenase, phosphoglycerate kinase) or regulated (UOP galactose epimerase, acid phosphatase). In the latter, the expression of the gene is under the control of galactose or inorganic phosphate, respectively.

A further requirement has been the secretion of the products of the cloned genes. To this end, vectors have been constructed that encode the signal sequence for secretion of the a--factor or invertase from S. cerevisiae or the killer factor secretion signal from K. lactis. The cloned gene can be fused to one of these signal sequences and the protein can be secreted into the medium.

To date, many genes have been cloned and expressed in S. cerevisiae and the products account for from 1-5% of the total cell protein. Secretion to the medium has been achieved for many human gene products (epidermal growth factor, interieukin-2, a--interferon, p-endorphin, etc.) and also for genes encoded by other organisms (including E. coli galactosidase, calf prochymosin, wheat amylase, etc.). The expression and secretion vectors have greatly enhanced the utilization of yeast in biotechnology.

USE OF RECOMBINANT DNA TECHNOLOGY FOR THE IMPROVEMENT OF WINE AND BREWING YEAST

STRAINS

The use of recombinant ONA (rONA) technology has some obvious advantages for the improvement of industrial yeast strains. Protoplast fusion can overcome the difficulties encountered in mating two different industrial strains, but shares with classical cell fusion techniques the difficulties arising by the recombination of two entire genomes and hence the possibility of losing some of the desirable qualities of the parental strains. On the contrary, the transformation of industrial yeast by chimaeric plasmids can introduce a single desirable gene and hence a specific and well-defined genetic improvement, without altering any of the other properties of the strains. However, in the field of alcoholic beverages the use of rONA technologies has been somewhat retarded by the difficulties in obtaining auxotrophic mutants


TABLE 1 POSSIBLE GOALS OF GENETIC MANIPULATION OF BREWING

YEAST STRAINS

(1) Utilization of new substrates: amilolysis proteolysis hydrolysis of tJ-glucans

(2) Increase of productivities: low volume systems diminished biomass production

(3) Quality of product: control of aroma compounds (4) New alcoholic beverages

and also to a certain degree, by the lack of confidence in the possibility of really obtaining a genetic improvement of the complex set of qualities required for this kind of production. Only in recent years have important programmes of improvement of industrial yeast strains been attempted, mainly due to activity of centres such as Carlsberg and Guinness Laboratories, and several promising results have been obtained.

A list of possible goals of genetic manipulation of industrial brewing strains is reported in Table 1. Several attempts to introduce the genes for amylase, glucoamylase or even cellulase into brewing strains have been successfully performed. However the most important results concerning the capability of utilizing new substrates are probably the ones regarding the introduction in brewing yeast of bacterial or fungal tJ-glucanases.

Barley grains can contain as much as 10% of their total carbohydrate as tJ-glucan, and incomplete hydrolysis of this polysaccharide can cause serious problems in the brewing process, especially due to difficulties in beer filtration and beer clarity. Cantwell et al., 7

from Guinness laboratories, have introduced endo-p-glucanase from B. subtilis into brewing strains, using a 2,um based vector bearing copper resistance as the selectable marker. The gene was found to be expressed and the product partially secreted in an active form. A similar result was obtained by Knowles et al. 8 who introduced three genes from Trichoderma reesei coding for different fJ-1-4-glucanases into a laboratory strain of S. cerevisiae.

The control of aromatic compounds is another rapidly developing field. Special interest has been devoted to the study of diacetyl formation. Diacetyl and pentanedione are formed in beer as


side-products of the isoleucine-valine pathway from acetolactate and a-aceto-a-hydroxybutyrate, respectively. A long secondary fermentation is usually required to reduce the levels of the two vicinal diketones and a detailed knowledge of the molecular genetics of their formation might eventually lead to the breeding of strains which produce only low amounts of diacetyl and hence require lower lagering times.

For this reason, the ILVI and ILV2 genes and their flanking regions have been identified and sequenced in a lager beer strain. 9

Auxotrophic mutants from this strain have been obtained and transformed with the cloned ILVI and ILV2 genes.

A further example of genetic manipulation of yeast which might result in the improvement of wine fermentation has been reported by Williams et al. 10 A genomic bank from Lactobacillus delbrueckii was cloned in E. coli and transformants were selected by their ability to grow on minimal medium containing malate and to produce L-Iactate. A 5 kb insert was identified and shown to confer malolactic activity on transformants. The 5 kb fragment was then inserted into a shuttle vector that can replicate both in E. coli and yeast in order to obtain a wine yeast strain which might perform malolactic fermentation in the absence of malolactic bacteria. Malolactic activity of the transformed yeast strain was however, very low and problems involving plasmid stability and gene expression must still be solved before the practical use of this strain is possible.

In conclusion, in the last few years the application of recombinant DNA technologies to the improvement of beer and wine strains has undergone rapid development, and although practical applications have not yet been achieved, the basis for future advances has been firmly established.

REFERENCES

1. Russell, I. and Stewart, G. C. (1979). J. Inst. Brew. London, 85,95-8. 2. Hinnen, A., Hinks, J. B. and Fink, G. R. (1978). Proc. Natl. Acad. Sci.

USA, 75, 1929-33. 3. Falcone, C., Saliola, M., Chen, X. J., Frontali, L. and Fukuhana, H.

(1986). Plasmid, 15, 248-51. 4. Chen, X. J., Saliola, M., Falcone, C., Bianchi, M. and Fukuhara, H.

(1986). Nucl. Acids Res., 14,4471-81.


5. Araki, H., Jeampipatkul, A., Tatsumi, H., Sakurai, T., Ushio, K., Muta, I. and Oshima, Y. (1983).1. Mol. Bioi., 182, 191-203.

6. Toh-e, A. and Utatsu, I. (1985). Nucl. Acids Res., 13,4267-83. 7. Cantwell, B. A., Ryan, T., Hurley, J. C. and McCormell, D. J. (1986).

Yeast, 2, 52. 8. Knowles, J. K. c., PenUila, M., Teeri, T. T., Andre, L., Salovuori, I. and

Lehtovaara, P. (1985). EBC Congress, pp. 251-58. 9. Petersen, J. G. L. (1985). EBC Congress, pp. 275-82.

10. Williams, S. A., Hodges, R. A., Strike, T. L., Snow, R. and Kunkee, R. E. (1984). Appl. Environ. Microbiol., 47,288.

Chapter 4

KILLER YEASTS: NOTES ON PROPERTIES AND TECHNICAL USE OF THE CHARACTER

GIANFRANCO ROSINI

Department of Plant Biology, University of Perugia, Perugia, Italy

INTRODUCTION

In 19631 it was observed that some laboratory yeast strains belonging to the species Saccharomyces cerevisiae were able to kill other yeasts. Three different phenotypes of this property have been determined: 'Killer', 'Sensitive' and 'Neutral'.

Cells of the killer phenotype are able to kill 'Sensitive' strains, while cells of the 'Neutral' type are unable to kill and are insensitive to the action of killer cultures.

The study of the killer characteristic involves the use of growth media containing methylene blue, buffered between pH4·2 and 4·7, inoculated with a 'Sensitive' strain. After solidification the probable killer cultures are inoculated onto the surface of the medium and incubated at 20°C for 72 h. Those yeasts which killed 'Sensitive' strains, as shown by zones of inhibition, were designated as killer.

In nature more killer activities are present and they are classified by cross-reaction according to the method of Young and Yagiu. 2 Since the discovery of this property, various researchers have studied the character in order to determine such factors as: diffusion in nature; inheritance of the character; nature and forms of action of the toxic principle; optimum conditions for production and activity of the killer toxin.

The results of these investigations are briefly reviewed in the following pages.

41


42 GIANFRANCO ROSINI

PRESENCE OF KILLER CULTURES IN NATURE

The presence of these particular yeasts was evaluated among collection strains as well as strains isolated from natural fermenting substrates. In 1975 Philliskirk and Young3 examined the cultures present in the National Collection of Yeast Cultures (NCYC) and observed that only 7 out of the 28 examined yeast genera included species having killer activity. Furthermore, the highest frequency of these yeasts was found among those belonging to the genus Hansenula (12 out of the 29 examined strains). In the Saccharomyces genus only 5·7% of the cultures demonstrated killer properties, and even lower percentages were observed in the genera Kluyveromyces, Debaryomyces, Pichia, Candida, Torulopsis.

After carrying out studies on cultures of Saccharomyces present in the Collocation of the Institute of Magarateh (Yalta, Crimea), Tyurina et al. 4 concluded that killer yeasts are not widespread and a prolonged storage of yeast strains in a laboratory (periods longer than 25 years) causes a loss in killer activity.

In 1973 Naumov et al. 5 observed that killer yeasts of the Kl type can easily be found among collection cultures, while those of the K2 type characterize the microftora of fermenting grape musts.

In Perugia (Italy) cultures of the Industrial Yeast Collection of the Department of Vegetable Biology at the University of Perugia, were analyzed for killer character. The results obtained showed a low occurrence (3·9%) of killer strains in the genus Saccharomyces. On the other hand high frequency of killer phenotype (71,4%) was found in Hansenula strains. A decrease in virulence of the killer character was observed after keeping the cultures in the collection for a long period of. time. 6-9

Upon analysis, yeasts isolated from natural sources such as fruit, edible mushrooms, agricultural soil and decomposing vegetables, Stumm et al.1O observed that only 26 (17%) out of the 157 yeasts belonging to 9 different genera had killer activity, and 17 (11 % ) appeared to be 'Sensitive'. The same study also found an interesting relationship between genera. Killer strains belonging to the genera Pichia and Hansenula are able to kill yeasts of Saccharomyces and Candida but not Hanseniaspora. On the other hand yeasts of this last genus and Saccharomyces are killed by Kluyveromyces strains, while strains belonging to the genus Candida appear to be immune.

The low occurrence of killer yeast in nature is refuted by the results

KILLER YEASTS: PROPERTIES AND TECHNICAL USE 43

of other researches. In fact, analyzing 907 strains isolated from grape musts, Barrell found 514 (56·7%) 'Killer', 94 (10·4%) 'Neutral' and 299 (33%) 'Sensitive' strains. Upon analysis of 103 French wines, Cuinier and Gros12 found killer cultures in 29 products, the most frequently in those produced in the Beaujolais area and southern regions.

Other studies on the presence of killer strains in oenological environments and products, was carried out by Kitano et al.13 and Thornton. 14 They examined the frequency of killer strains in Japanese wineries as well as in Australian and New Zealand wines, and characterized them. The presence of killer yeasts as contaminants was also found in industrial fermentative processes for the production of beer/5 sake16 and bakers yeast.3

Upon examining the diffusion Qf killer strains in nature, Tipper and Bostian17 advanced a tentative but realistic hypothesis: 'The killer property may affect the ecology of yeasts significantly, but low prevalence may indicate that yeast killers are inefficient assassins'.

INHERITANCE OF KILLER CHARACTER AND ACTIVITY OF THE KILLER FACTOR

The inheritance of the killer character18 is supported by the presence of a dominant gene and the simultaneous presence of cytoplasmatic determinants found in the M fraction of dsRNA in Saccharomyces cerevisiae. In fact, by carrying out electrophoresis on cell extracts of killer yeast, Young and Yagiu2 demonstrated that killer strains lost both the M type dsRNA and the killer property, when subjected to treatment with such agents as cycloheximide or high temperature. From the above studies it was found that the M type dsRNA is present only in Saccharomyces cerevisiae killer strains, but not in other genera. The authors advanced three hypotheses:

(1) the genetic determination of killer activity is not cytoplasmatic in non Saccharomyces yeasts.

(2) dsRNA is present but not detectable by usual electrophoretic techniques

(3) the examined yeast species produce nucleases able to degrade M type dsRNA quickly, thus preventing its electrophoretic detection.


Killer activity can also be present in the absence of killer cells. This demonstrates that cultures secrete some 'killer factors' into the medium which are lethal to sensitive strains. Studies carried out on the toxin demonstrated that it is a glyco-protein19 consisting of 90% carbohydrates (mostly mannose) and 10% protein. Separation of the protein and carbohydrate showed that toxic action is ascribed to the protein fraction.

Examining the toxin of K12-1 Saccharomyces cerevisiae strain, in 1984 Bostian et al. 20 found that it consists of two polypeptides linked and arranged in one dimer, probably of 'ab' type. However a high fraction of this dimer shows no killer property and only 10-30% of it has any toxic activity. 17 The activity of the killer factor would imply its preliminary binding to a cell wall receptor. The chemical nature of this receptor was identified in 1,6-D-glucan21 found on cell walls of the sensitive phenotype. This killer factor is then transferred from the cell wall to the membrane, which appears to be the specific point of action of the toxin; thus changing cell permeability.22 All anchorage and transfer operations require energy consumption. 22-24 This active mechanism would explain why sensitive cells do not die immediately. Their death only occurs after a variable amount of time (the latency period)-between 40 and 90 min-which is determined by growth medium composition and microbe development. After this interval there is a rapid fall in intracellular pH, loss of K+ ions and ATP which are found in the growth medium in high amounts (even 5 times higher than usual amounts). 25,26

OPTIMUM CONDmONS FOR KILLER TOXIN ACTIVITY

The production of killer toxin and its stability in culture broth depend on the pH and temperature of the medium. Optimum pH values, for Saccharoymces cervisiae were found by various authors to range between 4.6-4.810 and 4.2-4.8.27 For killer cultures of other genera, a higher resistance was found at lower pH.2,28 Temperatures higher than 25°C generally inactivate the killer factor in liquid culture, while agarized substrates give higher stability to the toxic principle which remains active up to 42°C. Toxins produced by using natural substrates are more temperature sensitive than those obtained on synthetic substrates. The optimum temperature range of toxin production as well as stability appears to be between 20 and 25°C. Because of the


essentially protein nature of the killer factor, killer activity loss occurs in the presence of proteolytic enzymes such as papain and pronase ,2,29

or after treatments with deproteinizing substances such as bentonite. l1

Conversely, the addition of protein compounds such as gelatin, lactalbumin or other organic compounds to a medium containing toxin tend to stabilize its toxic activity.

KILLER CULTURES AND POTENTIAL UTILIZATION

After determining and studying the killer phenomenon a particular interest in possible uses of this character arose. In industrial fermentation, microbiological purity is essentiaeo since any contamination could radically change the economy of the transformation, especially when continuous fermentations are involved.

The use of cultures which can prevail over wild sensitive yeasts by merit of their killer property can prevent or greatly reduce the possibility of pollution. This interesting prospect is especially relevant to industries such as breweries and bakers yeast manufacturers, but appears to be even more interesting for enological industries. In fact, with regard to the first industries mentioned, even under strict sterility conditions, it is important to improve safety by using killer strains. For this reason it is useful to produce industrial strains having the killer property by means of traditional and not genetic engineering techniques. 31-34

For the enological industry the use of competitive cultures would be even more interesting, since grape musts undergo only a treatment with sulphur dioxide before fermentation. This treatment is able to kill only some of the undesirable yeasts (apiculate yeasts of the genera Kloeckera and Hanseniaspora) while prolonging the developmental latency of strains belonging to other genera, especially Saccharomyces. Using cultures able to prevail over wild microftora by killing rivals, would therefore be desirable. At present however we believe the presence of killer character in cultures selected for wine-making should be considered only as an additional but not indispensable characteristic for a good starter of wine fermentation. 8

This conclusion derives from various considerations, only some of which can be considered as valid. In fact, it is worth noting that among some Saccharomyces cerevisiae Killer strains observed in our laboratory collection of wine yeasts, it was possible to find some strains suitable for wine-making since they are positive in accordance with


traditional selection tests. Furthermore it was possible to verify that at around pH 3, (similar to that of grape must), killer action is still present. But these valid remarks are opposed to other negative, as well as interesting observations:

-Saccharomyces cerevisiae killer cultures are unable to exert activity on apiculate yeasts considered as undesirable in vinification.6

-as various authors clearly pointed out, there are different types of killer character, and each being able to act only on specific killer or non-killer cultures.2 ,8

-temperatures of 35-40°C, which can often be reached during industrial vinifications, are able to inhibit killer activity in a short time.

-killer cultures are able to dominate sensitive populations only if their presence in the medium is between 1 and 5% of the total population. 7 ,11

Our scepticism is due to these negative observations. For maximum utilization of killer characters in the oenological

industry, the following were considered as effective:

-genetic improvement of strains, which however is already carried out ,35--37

-in wine-making, use of strains of the Neutral phenotype able to resist the action of natural killer yeasts, which are often present in grape musts,

-research about cultures able to exert killer action on apiculate, sulphur dioxide-resistant, film forming, osmotolerant yeasts, all undesirable in the oenological industry.

Once these particular killer cultures have been determined, the possibility could be considered of transferring this killer property to Saccharomyces strains selected for wine-making, or producing the toxin to be added to musts or wines as a natural antiseptic.

REFERENCES

1. Bevan, E. A. and Makower, M. (1963). In: Genetics Today. XI Int. Congr. on Genetics, S. J. Geerts (Ed.) 1, 202-3.


2. Young, T. W. and Yagiu, M. (1978). Antonie van Leeuwenhoek, 44, 59-77.

3. Philliskirk, G. and Young, T. W. (1975). Antonie van Leeuwenhoek, 41, 147-5l.

4. Tyurina, L. X., Bur'yan, N. 1. and Skarikowa, T. K. (1980). Bull. O/V, 53,573-6.

5. Naumov, G. 1., Tyurina, L. X., Bur'yan, N. I. and Naumova, T. I. (1973). Bioi. Nauki, 16, 103-7.

6. Rosini, G. (1983). Can. J. Microbiol., 29, 1462-72. 7. Rosini, G. (1983). XVIII Congr. Int. de la Vigne et du Vin, Cape Town. 8. Rosini, G. (1974). J. Atti Ac. Italiana della Vite e del Vino, 35,261-71. 9. Rosini, G. (1985). J. Microbiol., 31,300-2.

10. Stumm, c., Hermans, J. M., MiddeJbeek, E. J., Croes, A. F. and De Vries, G. J. M. L. (1977). Antonie van Leeuwenhoek, 43, 125-8.

11. Barre, P. (1980). Bull. OIV, 53,560-7. 12. Cuinier, M. C. and Gros, C. (1983). Vignes et Vin, 318,25-36. 13. Kitano, K., Sato, M., Shimazaki, T. and Hara, S. (1984). J. Ferment.

Technol., 62, 1-6. 14. Thornton, R. J. (1986). Antonie van Leeuwenhoek, 52,97-103. 15. Maule, A. P. and Thomas, P. D. (1973). J. Inst. Brew., 79, 137-41. 16. Imamura, T., Kawamoto, M. and Takaoka, Y. (1974). J. Ferment.

Technol., 52,293-9. 17. Tipper, D. J. and Bostian, K. A. (1984). Microbiological Reviews, 48,

125-56. 18. Somers, J. M. and Bevan, E. A. (1969). Genet. Res., 13,71-83. 19. Palfree, R. and Bussey, H. (1979). Eur. J. Biochem., 93,487-93. 20. Bostian, K., Bussey, H., Elliot, Q., Burny, V., Smith, A. and Tipper, D.

J. (1984). Cell., 36, 741-51. 21. Hutchins, K. and Bussey, H. (1983). J. Bacteriol., 154, 161-9. 22. Skipper, N. and Bussey, H. (1977). J. Bacteriol., 129,668-77. 23. AI-Aidross, K. and Bussey, H. (1978). Can. J. Microbiol., 24,228-37. 24. Bussey, H., Saville, D., Hutchins, K. and Palfree, R. G. E. (1979). J.

Bacteriol., 140, 888-92. 25. Bussey, H., Sherman, D. (1973). Biochim. Biophys. Acta, 298,868-75. 26. De la Pena, P., Barros, F., Gascon, S., Lazo, P. S. and Ramos, S. (1981).

J. BioI. Chem., 256, 10420-5. 27. Young, T. W. and Philliskirk, G. (1977). J. Appl. Bacteriol., 43,425-36. 28. Middelbeek, E. J., Hermans, J. M. H. and Stumm, C. (1979). Antonie

van Leeuwenhoek, 45, 437-50. 29. Woods, D. R. and Bevan, E. A. (1968). J. Gen. Microbiol., 51, 115-26. 30. Burrow, S. (1979). In: Economic Microbiology 4, A. H. Rose (Ed.),

Academic Press, London pp. 31-64. 31. BOTtol, A., Nudel, c., FraiJe, E., Torres, R., GiuJietti, A., Spencer, J. F.

T. and Spencer, D. M. (1986). Appl. Microbiol. Biotechnol., 24,414-6. 32. Ouchi, K. and Akiyama, H. (1976). J. Ferment. Technol., 54,615-23. 33. Ouchi, K., Wickner, R. B., Toh-e, A. and Akiyama, H. (1979). J.

Ferment. Technol., 57, 483-7. 34. Young, T. W. (1981). J. Inst. Brew., 87,292-5.


35. Hara, S., Iimura, Y. and Otsuka, K. (1980). Am. J. Enol. Vitic., 31, 28-33.

36. Hara, S., Yuzura, I., Oyama, H., Kozeki, T., Kitano, K. and Otsuka, K. (1981). Agric. Bioi. Chern., 45, 1327-34.

37. Seki, T., Choi, E. and Ryu, D. (1985). Appl. Envir. Microbiol., 49, 1211-15.

Chapter 5

THE EFFECTS OF CARBON DIOXIDE ON YEASTS

J. C. SLAUGHTER

Department of Brewing and Biological Sciences, Heriot-Watt University, Edinburgh, UK

INTRODUCTION

The great importance of the medium composition in determining the course of yeast fermentation is widely recognized. Not only are the major compounds such as sugars and amino acids taken into account but there is an awareness of the significance of many minor compounds such as minerals, vitamins and oxygen. The compounds mentioned are very important in determining the initial rate and extent of yeast growth but it is clear that after some time in batch fermentation, the efficiency of fermentation begins to fall. The most obvious reasons for this would be the accumulation of toxic products and several researchers are looking into the phenomenon of ethanol tolerance together with the toxicity of higher alcohols and esters. In contrast, little is heard about carbon dioxide as an influence in fermentation performance even though it is a major product and has been known for over 50 years to exert profound effects on yeast. Many papers have been published on the effects of CO2 on microorganisms in general, including yeasts, but the work is scattered amongst several journals, and has often been carried out spasmodically and from a fairly empirical point of view. 1

However, it seems that CO2 has the same main effects on all microorganisms although, of course, detailed responses vary. To generalize, low concentrations tend to stimulate growth whilst higher concentrations inhibit growth and metabolism. The effects of CO2

concentrations which can easily be encountered in industrial practice are so marked with yeasts that the gas is potentially a very important factor in determining the fermentation performance of yeast and in the maintenance of the viability of pitching yeast.

49


so J. C. SLAUGlITER

PHYSICO-CHEMICAL ASPECTS OF CARBON DIOXIDE

Simple though CO2 is in chemical terms, its behaviour is often anomalous and there are still aspects of its hydration which are not well understood. Carbon dioxide is a symmetrical linear molecule with zero dipole moment (O=C=O). It dissolves unexpectedly well in water, but is even more soluble in hydrophobic solvents. It is most soluble in solvents such as ethanol and acetone which contain both hydrophobic and strongly polar groups. At 2SoC the solubility of CO2

in ethanol is about 4·S times greater than in water; the corresponding figure for acetone is 9·9 times.

Solution of Carbon Dioxide in Water The basic relationship between gaseous and aqueous forms is determined by temperature and pressure according to Henry's law. However, the relationship between CO2 and the water is not clearly understood. Although the molecule appears non-dipolar it has been suggested that each part of the molecule could act as a small dipole, C-06-, and so orientate the surrounding water molecules. Whatever the mechanism, CO2 (aq) is a major species as the solubility coefficient for CO2 in water is approximately one. At lSoC and lOS Pa pressure, 1 ml of water dissolves about 1 ml of CO2 ; at O°C the coefficient is about 1·68 and at 2SoC it is about 0·7S.

After solution, the gas reacts with water to form carbonic acid:

CO2 (aq) + H20 ~ H2C03

Both forward and back reactions are relatively slow and in many organisms this reaction is catalysed by the enzyme carbonic anhydrase. At equilibrium carbonic acid is a very minor component as it is present at 0·001 % of the concentration of CO2 (aq). Once carbonic acid is formed it instantly ionizes:

H2C03 ~ H+ + HC03'

HC03' ~ H+ + CO~-

Carbonic acid is a weak acid and CO~- is effectively absent below pH 8 so the second step can be ignored in a study of fermentation where medium pH is normally in the mildly acidic range. Formation of bicarbonate is strongly affected by pH as shown by data in Table 1 calculated from known equilibrium constants and the HendersonHasselbalch equation.

THE EFFECfS OF CARBON DIOXIDE ON YEASTS

TABLE 1 EFFECf OF pH ON THE RELATIVE

IONIZATION OF CARBONIC ACID

pH CO2 (aq)

7 1 6 1 5 1 4 1

0·001 0·001 0·001 0·001

HCO;

2·5 0·25 0·025 0·0025

Solution of Carbon Dioxide in Complex Media

51

The solution of CO2 in complex media is likely to be affected by a wide range of compounds. Solutes in general will tend to reduce the solubility of the gas. An example of particular significance is the effect of the addition of ethanol to water (Table 2). This effect is definitely significant in the range of ethanol concentrations which occur in commercial yeast fermentations. 1 Compounds which react with CO2

(aq) or one of the derived species will increase the amount of CO2

which can dissolve in the medium although they will not affect the CO2

(gas)/C02 (aq) relationship. Ions which react with H+ or HCO; will produce a buffering effect and allow greater ionization of H2C03 to occur. Carbon dioxide (aq) reacts rapidly with un-ionized amino groups to form carbamates. This applies to free amines, peptides and proteins.

The reaction is more rapid than that with water so that on initial exposure to CO2, carbamate formation occurs faster than carbonic acid formation. Once the amino groups are saturated or the addition

TABLE 2 EFFECf OF ETHANOL ON THE

SOLUBILITY OF CO2 IN WATER

Ethanol concentration

(% w/v)

o 5

10 15 20

Relative solubility of CO2

1·00 0·82 0·69 0·58 0·53

52 J. C. SLAUGHTER

of CO2 ceases then carbonic acid production is favoured because of the high concentration of water and the carbamates begin to dissociate. This phenomenon is thought to be the mechanism of transient CO2

effects in some organisms. Another, ill-defined, area of 'C02 reaction' is the association of H2C03 with the positively charged groups of proteins through a dipole interaction.

Solution of Carbon Dioxide within the Cell The presence of yeast cells in a medium provide an extra layer of complexity to the solution of CO2 , Whether CO2 is being produced within the cell or is being added to the medium from an external source, CO2 (aq) is the only species which can effectively pass through the cell membrane. Carbon dioxide (aq) has a high solubility in water and lipids whereas the other species (H2C03 and the ions) are only very sparingly soluble. In practice, CO2 (aq) within the cell is very nearly at equilibrium with CO2 (aq) outside the cell. The rest of the picture depends on the local environment inside and outside the cell. The situation in the medium has already been dealt with. Within the cell the pH value is probably near 7 and the protein concentration is higher. This means that at any given value of CO2 (gas) the bicarbonate concentration will be much higher within the cell than in the medium. The higher pH will also make carbamate formation possible. Taken along with the higher protein concentration this means that more CO2 will dissolve within the cell than outside on a volume basis as more derived species will form within the cell.

PHYSIOLOGICAL EFFECTS OF CARBON DIOXIDE

Stimulation of Yeast Growth In gas flow fermenters where it is possible to maintain a partial pressure of CO2 of less than 105 Pa it has been shown that pC02 up to about 3 x 104 Pa (equivalent to about 10-15 mM CO2 in solution) tends to improve yeast growth. Carbon dioxide can account for 6-7% of cell carbon but this is greatly reduced by the presence of adequate amounts of aspartate in the medium. Examination of the map of biochemical pathways suggests an obvious explanation in that CO2 is a substrate in several essential central reactions. Oxaloacetate is formed from pyruvate or phosphoenol pyruvate and CO2 as part of the biosynthetic route for aspartate and glutamate. Carbamoyl phosphate,

THE EFFECTS OF CARBON DIOXIDE ON YEASTS 53

essential for the formation of arginine and the pyrimidine nucleotides, is synthesized from ammonia and CO2• Other CO2 fixation reactions occur during synthesis of the purine nucleotides, and it seems obvious from these basic metabolic relationships that the availability of CO2

will be a potential factor in determination of growth rate.

Potential Mechanisms of Growth Inhibition In contrast to the stimulating effect of CO2 at low concentrations, the inhibition caused by higher concentrations is harder to explain. Currently, no mechanism can be reliably invoked but several of the ways in which CO2 is known to act could be responsible. The following sections describe a range of possible mechanisms.

Mass Action Effects Throughout metabolism many decarboxylation steps occur and it is possible that CO2 could exert its inhibitory effect through product inhibition of these reactions. Whilst this idea is clearly theoretically correct, there does not seem to be any evidence that it actually is important under physiological conditions. In fact the opposite may be true; the major CO2 producing step, i.e. glycolysis, is known not to be inhibited by pressures up to an excess of 4 x 105 Pa whereas growth begins to fall away at pC02 of 0·5 x 105 Pa.

Another generalized mass action effect of CO2 is its potential influence on pH. Clearly even quite small changes in pH could affect all enzymically catalysed reactions to some extent. However, it is not known how effective the buffering mechanism of the cell is and just how high the CO2 concentration must be before there is any appreciable effect.

Specific Effects on Enzymes Conventional enzyme studies in vitro have often shown that HC03 can affect catalytic activity. It may be an activator or inhibitor or both depending on the concentration, e.g. ATP: citrate lyase is activated above 20 mM; isocitrate dehydrogenase is inhibited in the range 40-100 mM; cytochrome oxidase is activated up to 200 mM and inhibited above. The known examples seem to represent an arbitrary group of enzymes and there is no particular connection to CO2-

metabolizing enzymes. Carbamate formation could clearly have a pronounced effect on

enzyme activity as the process converts an uncharged group (the

54 J. C. SLAUGHTER

amino group) into a negatively charged one (the carboxyl group). There is no specific evidence for this mechanism with regard to a known enzyme, however. Most work has been done with model peptides or haemoglobin. Interaction of H2C03 with the positively charged groups on an enzyme clearly has the potential to change the surface chemistry of the molecule and hence its catalytic activity. As with carbamate formation, carbonic acid interaction has not been implicated in the inhibition of any known enzyme.

Although in certain situations, changes in enzyme activity may have major effects on metabolism, recent work on control of metabolic flux has shown that in situations approaching the steady state, individual enzymes are unlikely to have a great influence.2

At the moment, both theory and the limited number of experimental results indicate that under physiological conditions the controllability coefficient which indicates how sensitive metabolic flux is to small changes in the activity of a particular enzyme, is likely to be low. In practical terms it seems that the activity of most enzymes can be halved, and possibly quartered, before there are any significant effects on metabolism. The point in the present context is that if CO2 acts through inhibition of individual enzymes the effect must either be very marked with a single enzyme or distributed over a wide range of enzymes.

Membrane Effects As with enzymes, CO2 can influence membrane properties in a number of ways.

Bicarbonate ions can affect the surface charge and membrane potential and thus influence the uptake of solutes, such as sugars and amino acids, which depend on the electrical properties of the membrane to provide the energy for their absorption. Carbon dioxide (aq) can dissolve in the lipid core of the membrane and alter its properties towards a more 'fluid' state. This can be expected to have a generalized effect on membrane properties. Changes in membrane properties would also result from reaction of proteins with CO2 (aq) to form carbamates and interaction of proteins with H2C03 •

As well as the instantaneous chemical and physical effects mentioned above, exposure of cells to CO2 can also have biochemical consequences for the membrane composition. This has been documented for the fatty acid composition which has been shown to change markedly towards a more 'fluid' composition after 6 h exposure to CO2 •3

THE EFFECfS OF CARBON DIOXIDE ON YEASTS

Known Inhibitory Eft'ects of Carbon Dioxide on Yeast

Cell Properties

55

Cell division. There have been occasional reports, certainly since the 1930s, that CO2 reduces and eventually prevents cell division. Work in the 1960s and 70s has identified the critical range as being from 0·5 X lOS-3·90 X 105 Pa CO2 • l The figures must be taken as approximate as they are derived from the work of several authors who were using different yeast strains, different media and substantially different fermentation vessels and techniques.

In our laboratory, using our ale strain of Saccharomyces cerevisiae (NCYC 1108) fermenting a malt extract medium at 25°C in a vessel to which an excess pressure of CO2 can be applied, we have confirmed that pC02 of 3·90 X 105 Pa effectively prevents cell division.4 Reducing the excess pressure to 2·07 x 105 Pa allows a certain amount of growth.5 Experiments at 20°C, but with otherwise identical conditions, have shown reduction of growth at excess pressures of 1·01 x 105 and 0·51 x 105 Pa.6

Lipid composition. In 1969, Castelli et al. 3 found that whilst changes in the pH of the medium between 5·5 and 6·0 and in the bicarbonate concentration between 0·73 and 28·35 mM at constant pC02 had no effect on the lipid composition, changes in pC02 had profound effects. An increase in pC02 from 0·18 x 105 to 0·56 X 105 Pa at 28°C in a gas flow fermenter caused an increase in the total fatty acid content from 9·4% to 14·7% after 6 h exposure to the new conditions. The proportion of unsaturated fatty acids increased from 47·4% to 67·3%.

Thus the lipid composition of the yeast cell appears to depend on pC02 in both a quantitative and qualitative fashion.

DNA content. The only other major item of biochemical information concerned with cell composition was reported by Norton and Krauss in 1972.7 They were studying the effect of CO2 on cell division and used sealed fermenters which could either be pressurized or allow pressure to build up inside them as fermentation proceeded. This work first established that an excess pressure of 2·90 x 105 Pa at 25°C prevented cell division and also showed that the effect was due to the gas itself rather than the physical pressure by carrying out experiments with nitrogen in place of CO2 , The new biochemical observation from their work was that although the cells did not bud at 2·90 x lOS Pa, DNA replication occurred with the result that the cells exposed to the

56 J. C. SLAUGHTER

CO2 had double the DNA content of the control cells. We have investigated this phenomenon and have obtained essen

tially the same results. Throughout a 24 h fermentation period the DNA content of cells exposed to 2·90 x lOS Pa excess CO2 pressure at 25°C steadily increased so that at the end, the experimental cells had approximately twice the DNA content of the contro1.4 Most compounds which block cell growth do so in G 1 phase of the cell cycle so CO2 appears to be a relatively rare compound in that it blocks after the S phase in G2. This means that CO2 could be a useful tool for workers investigating the mechanisms which control the cell cycle.

RNA and protein content. Examination of cells at 2·90 X 105 Pa excess showed RNA contents rather lower than in the control. 4

However, the pattern of change was much the same. In contrast, CO2

had little effect initially on the protein content but after about 12 h the cells rapidly lost about 40% of their protein. The viability, as measured by the methylene blue technique, showed parallel changes to the protein content but the loss began slightly earlier.

Cell size. Exposure to 2·90 X 105 Pa led to a substantial increase in volume which was maintained throughout the experiment. Closer investigation of the early period showed that the increase took about H-2 h for completion. 4 Very similar results were obtained at 2·02 X 105 Pa.8 At 1·01 X 105 Pa a much smaller increase in volume was found;6 the experimental cells were 30-40 Ilm3 larger than the control throughout the fermentation. At 0·5 x 105 Pa no significant difference from the control could be found.

At the moment we have no explanation for the increase in cell size. Some of it could be attributed to an increase in cellular material as the cell grows but fails to form a bud. This cannot, however, be a complete explanation as the growth rate in the control is much too slow. Possibly the increase in size is a result of increased fluidity of the membrane caused by solution of CO2, However, this suggestion runs counter to the current idea that the shape and size of yeast cells is maintained by the cell wall. The idea that CO2 could dissociate the proteins of the cytoskeleton and so result in an increase in cell size runs into the same problem.

Regardless of the mechanism, the increase in cell volume is the earliest response known to CO2 and it could be that the changes in

THE EFFECTS OF CARBON DIOXIDE ON YEASTS 57

composition of the cell follow from this primary change. Two types of hypothesis suggest themselves. In the first approach several workers have suggested that the increase in cell volume which occurs during growth and hence a reduction in the concentration of certain compounds and ions triggers the various stages of the cell cycle.9

Something of this type of mechanism could underlie the doubling of DNA under CO2• The normal mechanisms of the cell could be responding to the change in volume as if it were occurring in the normal course of events.

Secondly, the increase in cell volume could result in some disruption of the internal organization of the cell, particularly at the membrane level. This idea may be particularly useful in explaining CO2-induced alterations in metabolism and it is interesting in this context that glycolysis, which does not depend on structural organization, is relatively independent of CO2 concentration whereas the TCA cycle, which depends entirely on the physical integrity of the mitochondria, is affected by quite low CO2 concentrations.

Fermentation Pressure fermentation systems. The use of CO2 pressure in

industrial fermentations was suggested in the early 1960s as part of a system to accelerate the production of lager beer. At a time of rising consumption there was a lot of interest in techniques which could speed up the long slow traditional lager production methods. The simplest way to accelerate fermentation, and particularly the lager fermentation which was conducted at temperatures as low as 8°C, was to raise the temperature a few degrees by reducing the cooling rate. Unfortunately, this resulted in production of undesirable volatile compounds which affected the beer flavour, in particular the concentrations of higher alcohols and ester was increased. It was found that application of an excess CO2 pressure up to 2 X 105 Pa at a suitable point in the fermentation could inhibit production of the undesirable compounds whilst allowing the main fermentation reactions to proceed more quickly at a slightly elevated temperature. During the 1960s and 70s much empirical work was carried out in breweries throughout the world to establish fermentation conditions which reduced the time of fermentation but maintained the quality of the product by a combination of increased temperature and CO2

pressure. In many cases this was successful and used commercially,

58 J. C. SLAUGlITER

although some breweries gave up the process later as it resulted in a decreased viability of the yeast.

Flavour volatiles. In 1974, Miedaner et al. 1O reported the results of a series of 10-litre laboratory fermentations carried out at 8·5°C, 12°C, 16°C and 20°C. They measured acetaldehyde, ethyl acetate, isoamyl acetate, n-propanol, isobutanol, isoamyl alcohol and vicinal diketones. In all cases, the effect of increasing temperature was to stimulate fermentation speed, accelerate production of the volatiles and increase the maximum concentration of the volatiles. Application of an excess CO2 pressure of approximately 0·8 x lOS, 1·2 X 105 or 2·0 X 105 Pa at a fixed point between one and two days of fermentation more or less stopped production of all the volatiles except vicinal diketones and their precursors (VDKs). The concentration pattern of VDKs seemed not to be significantly affected by the CO2 pressure.

In 1975, Norstedt et al. 11 reported some work which they had carried out in full-sized brewery fermentation tanks. They found that the production of esters fell as the height of the fermenting wort increased. Production of decanoic and dodecanoic acids was also reduced but the concentrations of octanoic and hexanoic acids were unaffected. Further experiments carried out by purging the wort with N2/C02 mixtures and use of CO2 counter pressure established that the effects of wort height were due to changes in the CO2 concentration and not hydrostatic pressure.

Also in 1975, Kumada et al. 12 presented work on large-pilot and full-commercial scale fermentation. They carried out fermentations at 9°C, 14·5°C and 18°C with and without the application of CO2 pressure at 1·01 x lOS and 1·81 X 105 Pa throughout. They found that CO2

reduced yeast growth, speed of fermentation and amino acid uptake. The production of acetaldehyde increased whilst that of isoamyl alcohol and isoamyl acetate decreased. Vicinal diketones showed a more complex pattern. A pressure of 1·01 x 105 Pa applied to the fermentation at 9°C led to increased VDK production and after eight days the VDK level was still higher in the experimental fermenter as opposed to the control. However, at both 14·5°C and 18°C, CO2

repressed the maximum concentration of VDK achieved although by the end of fermentation it seems likely that there would be no difference between test and control.

Up to this stage all the reports under brewing conditions referred to lager strains (then known as S. carlsbergensis but since merged into S.

THE EFFECfS OF CARBON DIOXIDE ON YEASTS 59

cerevisiae) and in the early 1980s we began experiments with the ale yeast, S. cerevisiae NCYC 1108. Fermentations were carried out using 5-litre batches of malt wort in l1-litre steel vessels at 12°C, 16°C and 20°C with and without the application of 2·02 x 105 Pa of CO2 excess pressure throughout the fermentation period. We found inhibition of yeast growth and a slowing of fermentation as reported by the earlier workers. 5 Likewise, production of the fusel oils was reduced. The size of the effect varied from a 12% reduction of isoamyl alcohol concentration at 16°C to complete elimination of n-propanol at 12°e. In general, the effect of CO2 was greatest at 12°e.

All reports on the effect of CO2 during fermentation have shown that it reduces the production of fusel oils and esters. Reports have come from Japan, Europe, America and South Africa using different yeasts, worts and fermentation systems so it appears that this is a fairly fundamental aspect of S. cerevisiae. The mechanism of the effect is, however, unknown. An attractive possibility is that the CO2 causes product inhibition of the decarboxylation step in fusel oil production or specifically inhibits one or both of the two enzymes involved. Further investigation suggests that this is unlikely as it is considered that fusel oils are produced from a-keto acids by pyruvate decarboxylase and alcohol dehydrogenase-the same enzymes that catalyse the final steps of glycolysis. If this is true, enzyme inhibition seems an unlikely mechanism as, as was mentioned above, glycolysis is unaffected by the kind of CO2 pressure investigated here.

When the concentration of total VDK was examined we obtained results at all temperatures which indicated an increase, in contrast to all the published results except those of Kumada et al. at 9°e. 12 In our experiments the effect of CO2 was not so much on the initial rate of formation but rather on the length of time of production and the rate of disappearance. We do not know whether our results represent a genuinely greater production of a-acetohydroxy acids, a failure in the ability of the cell to absorb and reduce the diketones or whether the CO2 inhibits the spontaneous decarboxylation of a-acetohydroxy acids to diketones. The last explanation seems very unlikely as if it were true, we should expect that slower removal of VDKs would occur in all CO2-pressurized systems and, as mentioned above, this is clearly not the case.

Amino acid absorption. In 1975, Kumada et alY reported that the overall absorption of amino acids was slowed in rate and reduced in

60 J. C. SLAUGHTER

total quantity when 1·01 x 105 Pa CO2 was applied to fermentations. The effect was definite but not very large and seemed consistent with the reduced cell growth which also occurred. We have examined this phenomenon and seen a similar effect on free a-amino nitrogen assayed using ninhydrin at both 0·5 x 105 and 1·01 x 105 Pa. A more detailed examination, however, revealed that the effect is not a simple overall reduction in the rate of absorption of all amino acids. There appeared to be little, if any, effect on the Group A amino acids, i.e. those normally absorbed rapidly, whilst absorption of Group B amino acids was very much slowed down. 6 More experimental work will be needed to tell if the changed amino acid absorption pattern is a cause or a result of reduced growth. It is possible that CO2 could inhibit the permeases responsible for uptake of Group B amino acids and so bring growth to a halt through nitrogen deficiency. Alternatively, of course, some other event could be slowing metabolism so reducing the cellular demand for amino acids after the first phase of the fermentation.

We have carried out further work on amino acid uptake at 2·03 x 105 Pa and have found a completely different picture.s At this pressure, cell division was severely retarded, there was a near doubling of cell volume, a 50-60% increase in DNA per cell, a loss in protein and a loss in viability as compared to the control. The pattern of absorption of amino acids also changed completely. When changes in total free a-amino nitrogen were measured using ninhydrin the concentration in the medium declined for the first 4 h but thereafter began to increase and, after 24 h at 25°C, the medium content was higher than at the outset. Detailed investigation using HPLC amino acid analysis showed that in the first 4 h the absorption rate of all the amino acids except lysine was altered. Some compounds were not absorbed at all, the rate of absorption of others was slower than the control and, in a few cases, faster than in the control. By 8 h it was clear in all cases that the concentration in the medium was increasing, presumably due to leakage from the cell. A possible scenario is that the CO2-induced increase in cell volume referred to earlier causes disruption of the vacuolar membrane which, as the vacuole is the location of a range of general proteases, allows degradation of the cell protein to begin. This would result in a build up in the internal amino acid pools which would inhibit uptake and then amino acids would begin to leak through the partially permeabilized cell membrane. This scheme is supported by the pattern of loss of cell protein and viability. However, as in several areas mentioned earlier, more research is


needed to clarify exactly what is happening on the cellular and molecular level.

Carbon Dioxide as an Incidental Control Factor All the work reported so far on CO2 effects on yeast and fermentation has been from the point of view of cellular properties or with a view to controlling the products of fermentation by deliberate use of CO2

pressure. However, it is well known that during fermentation the wort can become super-saturated with CO2, particularly in large fermenters and lager fermentations. Values as high as 0·5% w/v have been reported under conditions where no attempt to retain the CO2 in the liquid was made. The CO2 concentration reached will depend on the vigour of the fermentation, the temperature, the surface area to volume ratio of the wort and the capacity of the escape pipes. Given the range of commercial fermentation practice in terms of wort strength, pitching rate, fermentation temperature and the variety of vessels used it can be calculated that in the fermentation industries the yeast will be exposed to CO2 concentrations ranging from c. 40 mM (small open vessel at c. 20°C) to over 100 mM (large vessels at c. 12°C). The pattern of change will depend on all the fermentation parameters mentioned above and it seems likely that CO2 could be a contributor to difficulties experienced in scale-up of processes and to variation from one fermentation to another. In order to define the potential of CO2 as an important incidental control parameter we have carried out some laboratory experiments on the 5-litre scale at 20°C using counter pressures of 0·50 x HP and 1·01 x 105 Pa (theoretically equivalent to 57 and 76 mM CO2 respectively). These concentrations are within the range which might be expected in commercial practice where no attempts to control the CO2 concentration were being taken.

Several of the results of these experiments have already been mentioned but to recap: fermentation under a constant excess CO2

pressure of 0·50 x 105 Pa caused reduced growth rate, viability, and reduced absorption of amino acids. Use of 1·01 x 105 Pa had a greater effect on all these parameters and, in addition, caused a distinct increase in the mean volume of the cells.

Several other aspects of fermentation performance have been shown to be sensitive to 0·5 x lOS Pa CO2 •6 The results obtained at 1·01 x 105 Pa tend simply to be greater numerically but were always of the same tendency.

62 J. C. SLAUGHTER

Fall in specific gravity and ethanol production. A pressure of O·SO x 105 Pa led to a decline in specific gravity which was just detectably slower than the control but the effect on the rate of ethanol production was clear. However, the final ethanol concentration achieved was the same in both fermentations. The effect of 1·01 x 105 Pa was slightly greater but again there was no detectable difference in the yield. These CO2 concentrations clearly have no important effect on the major reactions of fermentation and this is consistent with earlier reports that glycolysis is not affected by CO2

until much higher levels are reached. The slight reductions we found can be explained in terms of the lower cell numbers in the CO2

fermentations.

Fusel oils. As reported by all other workers CO2 reduced fusel oil production. With n-propanol and isobutanol the effect was on both rate and total production, whereas with isoamyl alcohol at O·SO x 105 Pa the effect was more on rate and the final concentration under CO2 was the same as in the control. At 1·01 x lOS Pa the effect was very similar but the final concentration of isoamyl alcohol was slightly less in the CO2 fermentation than in the control.

As mentioned earlier, the mechanism of the effect is unknown. The only comment which can be made at this stage is that CO2 also slows down the uptake of the branched chain amino acids which are potential precursors of isobutanol and isoamyl alcohol. However, CO2

has no clear effect on the absorption of threonine, the potential precursor of n-propanol so the situation is obviously more complex than simply an influence on amino acid absorption. What matters, of course, is not the uptake of amino acids but the pool sizes of the a-keto acids and currently we have no information on how these are affected by CO2,

Esters. The only ester for which we have information is ethyl acetate and the production and final concentration of this compound are reduced by both O·SO x 105 and 1·01 x lOS Pa. This agrees with other reports on esters.

Acetaldehyde. Acetaldehyde is a normal product of fermentation which is commonly present in freshly fermented beverages at concentrations above the threshold value. This compound is thought to make an important contribution to the 'green' flavour found after primary fermentation.


Low CO2 pressures clearly have an influence on acetaldehyde concentration. At most points during fermentation, acetaldehyde concentration was higher in the CO2 vessel except for a brief period at the end of fermentation when the concentration in the control rose rapidly. However, acetaldehyde continued to be produced in the CO2

vessel and after a short time the concentration once again became greater than in the control.

There is no clear explanation for these observations. However, given the slight reduction in the rate of ethanol production under CO2 ,

the effect of CO2 on acetaldehyde could suggest that alcohol dehydrogenase is slightly inhibited by CO2 , although, at the moment, this suggestion seems to run against other observations of the effect of CO2 on glycolysis. An alternative explanation is that CO2 renders the cell membrane slightly more permeable to acetaldehyde so the concentration in the medium increases coincidentally.

Vicinal diketones. The two vicinal diketones found in fermented beverages are diacetyl and 2,3-pentanedione. Reference will be made only to diacetyl as this is the organoleptically important compound and the response of 2,3-pentanedione to CO2 is essentially the same as that of diacetyl.

As reported for experiments with 2·02 x 105 Pa CO2 , our strain of S. cerevisiae responded to CO2 at O· 50 x 105 and 1·01 x 105 Pa by the apparent production of greater amounts of diacetyl and its precursor. The results appear to support the role of valine as feed-back inhibitor in control of diacetyl production; production occurs whilst the yeast is fermenting but not absorbing valine. The fall in diacetyl occurs earlier and is greatest in the fermentations where the cells absorb valine more rapidly and to a greater extent.

REFERENCES

1. Jones, R. P. and Greenfield, P. F. (1982). Enzyme and Microbiol. Technology, 4,210-23.

2. Kell, D. B. and Westerhoff, H. V. (1986). FEMS Microbiology Reviews, 39,305-20.

3. Castelli, A., Littarru, G. P. and Barbaresi, G. (1969). Archiv. Mikrobiol., 66,34-9.

4. Lumsden, W. B., Duffus, J. H. and Slaughter, J. C. (1987).1. of General Microbiology, 133, 877-81.

64 J. C. SLAUGHTER

5. Arcay-Ledezema, G. J. and Slaughter, J. G. (1984). J. of Institute of Brewing, 90, 81-4.

6. Knatchbull, F. B. and Slaughter, J. C. (1987). J. of Institute of Brewing, 93,420-4.

7. Norton, J. S. and Krauss, R. W. (1972). Plant and Cell Physiology, 13, 139-49.

8. Slaughter, J. C., Flint, P. W. N. and Kular, K. (1987). FEMS Microbiology Letters, 40,239-43.

9. Walker, G. M. and Duffus, J. H. (1983). Magnesium, 2, 1-16. 10. Miedaner, H., Narziss, L. and Womerg, O. (1974). Brauwissenschaft, 27,

208-14. 11. Norstedt, G., Bengtsson, A., Bennet, P., Lindstrom, I. and Ayrapaa, T.

(1975). Proceedings of the 15th European Brewery Convention, Nice, pp. 581-99.

12. Kumada, J., Nakajjma, S., Takahashi, T. and Narziss, L. (1975). Proceedings of the 15th European Brewery Convention, Nice, pp. 615-23.

Chapter 6

MICROBIAL SPOILAGE OF CANNED FRUIT JUICES

A. CASOLARI

Advanced Research Laboratory, Plasmon Dietetici Alimentari SpA, Parma, Italy

INTRODUCTION

Acidity is the single most important factor affecting microbial spoilage of fruit juices. More generally, hydrogen ion concentration is among physico-chemical factors of major concern affecting microbial growth. Microorganisms have an optimum pH for growth and a characteristic range of pH over which they can grow. Most bacteria have an optimum pH near 6·8 and may grow at pH values ranging from 4-8; a narrow spectrum of bacteria can mUltiply at pH < 4 or pH> 8. Yeasts and moulds can grow at pH < 2.

Usually, the growth rate decreases as pH drops below the optimum value. Approaching the lower limiting pH for growth, microorganisms are firstly inhibited and eventually killed. Figure 1 shows growth/inhibition/death curves usually obtained in media of different hydrogen ion concentrations. As can be seen, the curve describing microbial growth at optimum pH becomes increasingly flattened as pH decreases (the lag phase lengthens, the slope of the exponential growth phase decreases, the maximum cell count drops, the length of the stationary phase shortens, the slope of the death phase increases), eventually leading to a true death curve at very low pH. However in suitable environmental conditions, the end of the exponential growth phase, the stationary and the death phases are regarded in fact as being caused by the increased concentration of acids (mainly lactic, acetic and formic) produced during the metabolism.

The exact mechanisms of microbial inhibition by acids is not known. However, it is well known that most macromolecules are active in a narrow range of pH and that the intracellular acidity is maintained in a

65


66

o Z -

9

A. CASOLARI

6

zO~~~~~------------~~ 01 .2

-9

FIG. 1. Microbial response to environmental pH (6-1). The number identifying each curve also identifies the pH value of environment at which the

curve is obtained.

condition of relative independence from the environmental onel - 3

through a regulated flow of protons (H+) across the microbial membrane.4 Several membrane-bound enzymes regulate the size of this flow,5 protons being unable to flow freely across membranes. Nevertheless, as environmental pH decreases it can be expected that the acidity of the cell interior will increase by at least two mechanisms: (1) directly, through the lessening of the proton-impermeability of the membrane due to the high external proton pressure, and (2) indirectly, through the penetration of the cell by molecules undissociated in the acidic environment but which dissociate at the higher interior pH. In both cases, the intracellular pH will be higher than the environmental one. However, as the interior pH decreases an increasing fraction of macromolecules becomes inactive and the growth rate likewise decreases. As the exterior acidity is low enough, the pH of the cell interior drops to levels incompatible with the physiological activity. On the basis of available experimental data it is not possible to distinguish conclusively between the effect of pH and that of the undissociated molecules of acids.

Recently the activity of seven acids (hydrochloric, sulphuric, phosphoric, citric, lactic, malic and acetic) was assayed against 49

MICROBIAL SPOILAGE OF CANNED FRUIT JUICES 67

bacterial strains (belonging to 20 species and 10 genera) (Casolari and Contini, unpublished results). The main results obtained through the statistical analysis performed over 308 curves of inhibition were for all tested bacteria: (1) the degree of inhibition increased as pH decreased; (2) the relationship between the degree of inhibition (RBE, relative biological effectiveness) and pH or the logarithm of the molarity was linear; (3) the inhibition curves can be regarded as parallel, irrespective of the acid used; (4) given the same pH the degree of inhibition increased linearly with pK of the acid used; (5) the rate of change of the RBE as a function of pK or of Log M was equal and about 0·8 times the rate of change of RBE against pH. The more or less pronounced inhibiting and/or lethal effects of acids having different pK values is well known.

Citric (pK 3·08) is the principal acid in citrus fruits, strawberries, cranberries, blueberries and tomatoes; malic acid (pK 3·4) predominates in apples, pears, apricots, and peaches; grapes contain malic and tartaric acids (pK 2·98). Oxalic, succinic, fumaric and isocitric are

TABLE 1 TYPICAL P H VALUES FOR VARIOUS

FRUIT JUICES

Fruit

Lemon Lime Raspberry Blueberry Grapefruit Grape Strawberry Blackberry Cherry Apple Orange Apricot Pineapple Peach Cocktail Plum Pear Mango Tomato

pH

2·2-2·4 2·2-2·6 2·7-3·3 2·8-2·9 2·9-3·4 3·0-4·0 3·1-3·9 3·2-3·4 3·2-3·9 3·3-3·5 3·3-4·0 3·4-3'6 3·4-3·7 3·4-4·2 3·6-4·0 3·7-4·3 3·7-4·7 3·8-4·7 3·9-4·6

68 A. CASOLARI

some of the other organic acids that can be found in fruits. Acetic acid (pK 4·75) is usually present only in trace amounts.

Table 1 shows the range of pH usually found in fruit juices. As can be seen, the higher pH values are lower than 4·5 in nearly all cases. In canned foods a pH of 4·5 is used as borderline between acid and low-acid foods, that is foods not requiring-and those requiring, respectively-the minimum botulinum cook (12D), according to the view that only a limited spectrum of microorganisms can grow at pH < 4·5 and these are mostly non-pathogenic. However, this assumption disregards the ability of Staphylococous aureus and Clostridium botulinum of growing at pH levels near 4 (in defined environmental conditions) and of several Salmonella strains also growing at pH < 4.

The typical spoilage flora of fruit juices is represented by some Clostridium, Bacillus, Enterobacteriaceae, lactic acid bacteria, Acetobacteraceae, yeasts and moulds.

CLOSTRIDIUM

The anaerobic sporeformers belonging to the genus Clostridium and collectively named butyric anaerobes are characterized by having the ability to produce large quantities of butyric acid through carbohydrate metabolism. Strains most frequently involved in fruit juice spoilage belong to the species Cl. butyricum and Cl. pasteurianum. Butyric anaerobes are ubiquitous and commonly found in soil and in decaying vegetation. The surfaces of healthy fruits can harbour both spores and vegetative cells of butyric anaerobes, although in low numbers. More usually they can be found at very high concentrations in rotten areas, where local higher pH and lower p02 is more suitable for growth of clostridia. Vegetative growth of Cl. pasteurianum and Cl. butyricum may occur at pH ~ 3·6;6 germination and outgrowth of Cl. pasteurianum spores may occur at pH ~ 3·9 (pear juice) (Casolari, unpublished results) and at water activity (Aw) > 0·97.7

Vegetative cells of butyric anaerobes are unable to grow at pH 4· 7 if Aw is lower than 0·97 (RI = 1·1168), although Cl. acetobutylicum may grow at Aw = 0·95 and pH 4· 7. Fruit juice spoilage by butyric anaerobes is characterized by lowering of pH (0·2-0·4 units), development of very high volumes of hydrogen and carbon dioxide (about 1·2: 1), and strong cheesy (butyric) odour.


The spores of Cl. pasteurianum are more heat resistant than those of the other butyric anaerobes tested. S The decimal reduction time at 100°C and pH 4·5 is 1 min and z = 10°C. The failure of the relatively mild heat process usually applied to fruit products in order to achieve commercial sterility, is caused by high contamination levels coupled with pH values higher than expected, both circumstances occurring in rotten fruits. As a matter of fact, heat processes usually applied to fruit juices are unable to destroy significant numbers of Ci. pasteurianum spores. Nevertheless, germination and outgrowth of spores occurs with the probability decreasing as the pH is lowered. The probability of spoilage (SP) of a fruit juice held at 30°C is given for Cl. pasteurianum spores inoculated into tomato juice:9

SP = 1O(-7s.12+17XpH) (1)

therefore the expected number of growing spores will be: GSN = No x 1O(-so.12+17XpH) (2)

where No is the residual living spores level of the heat-treated juice. The maximum level of contamination found in raw tomatoes was

104 spores/g. 10 It follows that, according to eqn (2), at pH 4·3 about one spore per kg of juice will be able to develop, while at pH 4·2 only two spores per 100 kg or 190 spores per 10 000 kg of juice will be able to grow. Similarly, the development of not more than one spore in 100 tanks of 10 000 kg juice is expected to occur if pH is lower than 3·9. Unfortunately, there is neither information about the maximum level of contamination by butyric anaerobes that can be expected in juice of different fruits, nor is there available from the literature a relationship analogous to eqn (1) valid for fruits other than tomatoes. However, the above reported relationship (eqn (1)) can provide good results when also applied to different fruit juices. Thus, taking into account that the maximum contamination level cannot exceed about lOS butyric anaerobes spores per g, the fermentation of a tank containing 10 000 kg juice (and therefore 1015 spores) can be expected to occur only one out of three times if pH = 3·8 and one in every 100 times if pH=3·7.

BACILLUS

Three types of Bacillus are usually involved in spoilage of fruit juices, i.e. B. coaguians, B. macerans and B. polymyxa.

70 A. CASOLARI

Usually, the spoilage of tomato products by B. coagulans ('flatsour') is characterized by lowering of pH (0·2-1 pH units) due to lactic acid produced during metabolism. B. coagulans is a thermophilic bacterium able to grow with increasing probability at temperatures higher than 30°C (and lower than 65°C) in substrates having pH > 4 and water activity higher than 0·97. Factors which have been determined to influence the minimum pH for growth include type of acid, substrate, inoculum size and strain examined.u-14 However, a relationship analogous to the above reported for butyric anaerobes (eqn (1)) is not available from the literature. The decimal reduction time D of the spores at pH 4·5 is 6-8 min at 100°C (z = 8°C).

Both B. macerans and B. polymyxa may spoil fruit juices producing hydrogen and carbon dioxide (1: 1), lactic, formic and acetic acids, etc. by storage at temperatures lower than 46°C. The minimum pH values for growth of both sporeformers are identical to those of butyric anaerobes.

ENTEROBACTERIACEAE

E. coli and E. aerogenes may grow at pH < 4· 5 and several strains of Salmonella at pH.::;; 3·9. The minimum pH value for growth decreases when increasing the inoculum size, according to a relationship of the type:

Nm = expA x (B - pH) (3)

where Nm is the minimum contamination level required for growth at the measured pH, A is a constant specific of single species and B is the pH value at which there is no inhibition at all of the growth.

Enterobacteriaceae may grow from chilling temperatures to 46°C at Aw > 0·92. Fruit juice spoilage by bacteria of this family is characterized by the production of lactic, acetic and formic acids, large quantities of hydrogen and carbon dioxide, together with some enhancement of fruit flavour. Since the heat resistance of Enterobacteriaceae is quite low (the D at 60°C ranges between 1 and 15 min, 4 < z < 12) their presence in heat-treated juices is a result either of a gross underprocessing or of a post-pasteurization contamination.


LACTIC ACID BACTERIA

Bacteria belonging to this group are very often associated with fruit juice spoilage. These organisms convert many sugars present in fruit juice to a mixture of lactic, acetic and formic acids (homofermentative), together with carbon dioxide, mannitol, ethanol, etc. (heterofermentative species); some species produce diacetyl and acetoin. Fresh fruit juices can be spoiled readily by members of this group; products of the secondary metabolism can be traced to estimate the quality of raw materials.

Several Lactobacillus (L. plantarum, L. fermenti, L. brevis and L. buchneri) are able to convert malic acid to lactic acid and carbon dioxide (the so-called malo-lactic fermentation), and/or citric acid to succinic acid; the fall in acidity gives fruit juices a rather flat taste, coupled with a loss of astringency. A number of species (L. plantarum, Leuconostoc mesenteroides, Streptococcus viscosum) may produce dextran-type polysaccharides giving slimy juice. Most bacteria belonging to the genus Lactobacillus grow at pH values higher than 3 (while some species--e.g. L. plantarum-only at pH> 3·6), at temperatures ranging from about 5°C-53°C and at Aw > 0·90. Decimal reduction times at 61°C range between 2 and 15 min (10 < z < 13).15

Cocci belonging to the genus Leuconostoc grow at pH values ranging from 2-7·5, at temperatures ranging from 10°C-40°C and at Aw> 0·94. The D at 61°C is less than 1·5 min. 16 They usually ferment dextrose and fructose producing lactate, ethanol, and carbon dioxide; decarboxylate malic acid to acetic acid and carbon dioxide; convert citric acid to lactic and acetic acids and carbon dioxide.

Streptococci may spoil fruit juices producing lactic acid and sometimes dextran. They can grow at pH values 3·9 up to pH 10 and at Aw> 0·91. To this type of cocci belong most heat-resistant vegetative bacterial cells: D at 60°C may be greater than 10 min and z may be as high as about 20°C.17,18

Acetobacter may grow in juices having pH> 3·6, at temperatures ranging from 50-43°C. They oxidize the dextrose to gluconic acid and the ethanol to acetic acid; some strains are able to oxidize acetic and lactic acids to carbon dioxide and water. Like Acetobacter, Gluconobacter oxidize ethanol to acetic acid while acetate and lactate are not further oxidized. These Gram-negative aerobic rods grow at pH> 3·6 and at temperatures ranging from 7°C-41°C.

Pediococci are homofermentative lactic acid bacteria growing at

72 A. CASOLARI

2·9 < pH < 8·S, Aw> 0·90 (18% NaCl) and up to SO°C. They sometimes produce dextrans.

YEASTS

Together with lactic acid bacteria, yeasts are among the agents most often involved in fermentation of fruit products. Yeasts can grow in a wide range of pH (from 1·S_9).19 Most of them grow in substrates of 0·80 Aw; some yeasts, called osmophiles, are able to grow just to Aw 0·61. The temperature suitable for growth was found to range between -12°C and SOOC.2O The decimal reduction time at Aw near 0·98 equals about 1 min at 61°C.

Species most often involved in juices fermentation belong to Saccharomyces, Torulopsis, Candida, Pichia, Hansenula and Hanseniaspora genera. Fruit juice concentrates are fermented almost exclusively by Saccharomyces rouxii, S. mellis, Torulopsis and Hansenula. Typical fermentation products are ethanol and carbon dioxide.

MOULDS

The surfaces of fresh fruit usually harbour a large number of moulds most often belonging to the genera Mucor, Rhizopus, Penicillium, Aspergillus, Alternaria, Cladosporium, Acremonium and Botrytis. The contamination level may increase drastically as a result of infection and subsequent rotting of deep tissues. Mould counts of healthy fruits may range from 1W -lOS, those of rotten fruits between lOS and 107 propagules/ g, and may reach 109/ g as a function of the type of fruit, degree of ripeness/rotting, the time elapsed between harvesting and processing and thermo-hygrometric conditions during that time. The contamination level may increase during processing as a result of contact with machinery surfaces often colonized by Geotrichum. Living moulds usually do not occur in processed fruit products since fungal elements are very heat-sensitive. Nevertheless, hyphal filaments can be revealed by microscopic examination of the juice; the relative abundance of hyphal elements is correlated with quality of both raw materials and general hygienic conditions of processing lines.

Moulds can develop at pH values ranging from l·S-11 and higher, at temperatures higher than -10°C and lower than SSOC. Nearly all


species can grow in substrates with water activity higher than 0·80; some Aspergillus and Wallemia grow at Aw = 0·75 and species belonging to Eurotium and Monascus are able to grow at Aw near 0·61.

Hyphae and conidia have a decimal reduction time at 55°C usually ranging from 3-6 min at Aw near 0·98. In pear juice (pH 3·7) and in apple juice (pH 3·6) asexual spores of Mucor spinescens show a decimal reduction time at 60°C of 1·7-1·4min and a z of 6·3-4·6, respectively.21 The heat resistance of conidia may increase dramatically as Aw decreases: at Aw = 0·90 the decimal reduction time at 55°C may rise to about 100-240 min (A. parasiticus).22 Usually, sexual spores (ascospores) are more heat resistant than conidia. The ascospores of A. chevalieri'show a decimal reduction time at 80°C (pH 3·8 and Aw 0·98) of 3·3 min and z = 13°C.23 Those of Byssochlamys have a D at 90°C ranging between 1·3 and 17·5 min as a function of the strain (given the same pH the heat resistance decreases in the presence of fumaric, lactic and acetic acids, and increases with malic and tartaric acids) and z = 6·8°C. However, it should be remembered that usually heat inactivation curves of ascospores are not exponential and the D values reported in the literature very often represent average values obtained from first 4-5 decimal reduction of counts.

Moulds require oxygen for development and then they usually grow at the surface of processed fruit products where the head space of the container may hold limited amounts of air for some time after packaging. Bottled and canned fruit juices are seldom spoiled by moulds. Byssochlamys fulva and B. nivea are more usually involved. Mycelial development is coupled with more or less pronounced pectolytic activity; some gas is produced.

Until a few years ago, Byssochlamys was the single most relevant mould of fruit products spoilage. Only recently it was found that several fruit juices (tomato, orange, mixed tropical fruit juices) packed in brik containers were spoiled by Mucor spinescens and Acremonium roseogriseum.21 The juices showed a varying degree of turbidity, a whitish surface mycelial mass and a pronounced alcoholic smell. Briks were blown. Mould strains isolated from spoiled juices were able to ferment purposely inoculated juices (pear, apple, apricot, orange, peach, mixed fruit, tomato, grapefruit juices) producing carbon dioxide and ethanol. 24 The heat resistance of mould strains isolated from spoiled juices was very low, while resistance to hydrogen

74 A. CASOLARI

peroxide (usually employed to sterilize the brik-forming film in aseptic filling technology) was very high.21 By testing 52 mould strains belonging to 25 species and 6 genera (Mucor, Rhizopus, Acremonium, Aspergillus, Fusarium and Penicillium) it was found that most (94%) strains produce carbon dioxide and ethanol during JUIce fermentation. 24,25 Since several species produce highly toxic substances and some species (A. flavus, A. parasiticus, A. versicolor, etc.) produce the most potent carcinogens known (aflatoxin and sterigmatocystin), there is a twofold reason to regard fruit juice spoilage by moulds as a very important occurrence.

REFERENCES

1. Conway, E. J. and Downey, M. (1950). Biochem. J., 47,347. 2. Neal, A. L., Weinstock, J. O. and Lampen, J. O. (1965). J. Bact., 90,

126. 3. Harold, F. M. and Baarda, J. R (1968). J. Bact., 96,2025. 4. Hamilton, W. A. (1975). Adv. Microb. Physiol., 12, 1. 5. Garland, P. B. (1977). Symp. of the Soc. for Gen. Microbiol., 27, 1. 6. Blocher, J. C. and Busta, F. F. (1983). Food Technol., 37, (11) 87. 7. Jakobsen, M. and Jensen, H. C. (1975). Lebensm.-Wiss. U. Technol., 8,

158. 8. Casolari, A. and Giannone, L. (1966). Ind. Cons. (Parma), 41,95. 9. Campanini, M., Casolari, A., Lancillotti, F. and Lucisano, A. (1971). Ind.

Cons. (Parma), 46, 182. 10. Casolari, A. and Ercolani, G. L. (1965). Ind. Cons. (Parma), 40,306. 11. Erickson, F. J. and Fabian, F. W. (1942). Food Res., 7,68. 12. Murdock, D. I. (1950). Food Res., 15, 107. 13. Rice, A. C. and Pederson, C. S. (1954). Food Res., 19, 115, 124. 14. York, G. K., Heil, J. R, Marsh, G. L., Ansar, A., Merson, R. L.,

Wolcott, M. T. and Leonard, S. (1975). J. Food Sci., 40, 764. 15. Tomlins, R. I. and Ordal, Z. J. (1976). In: Inhibition and Inactivation of

Vegetative Microbes, F. A. Skinner and W. B. Hugh (Eds) , Academic Press, London, NY, San Francisco, p. 153.

16. Casolari, A. and Campanini, M. (1973). Ind. Cons. (Parma), 48, 140. 17. Zakula, R. (1969). Fifteenth Europ. Meet. of Meat Res. Workers, 67,

Institute of Meat Technology, University of Helsinki, pp. 157-63. 18. Zivanovic, R, Oluski, A. and Tadic, Z. (1965). Technol. Mesa., 6, 198. 19. Dawes, J. W. (1976). In: Inhibition and Inactivation of Vegetative

Microbes, F. A. Skinner and W. B. Hugo (Eds) , Academic Press, London, NY, San Francisco, p. 279.

20. Van Uden, N. (1984). In: CRC Handbook of Microbiology, 2nd Edn, Vol. VI, A. I. Laskin and H. A. Lechevalier (Eds), CRC Press Inc., Boca Raton, FL, pp. 1-8.


21. Vicini, E., Barbuti, S., Spotti, E., Campanini, M., Castelvetri, F., Gola, S., Manganelli, E., Cassara, A. and Casolari, A. (1984). Microb. Alim. Nutrit., 2, 21.

22. Doyle, M. P. and Marth, E. H. (1975). J. Milk Food Technol., 38,750. 23. Pitt, J. I. and Christian, J. H. B. (1970). Appl. Microbiol., 20,682. 24. Casolari, A. and Gherardi, S. (1985). Symp. on Adv. in Fruit and Veget.

Juice Ind., Tel Aviv 1984, Juris Druck and Verlag, Zurich. 25. Spotti, E., Casolari, A. and Cassara, A. (1984). Microb. Alim. Nutrit., 2,

37.

Chapter 7

RECENT AND FUTURE DEVELOPMENTS OF FERMENTATION TECHNOLOGY AND

FERMENTER DESIGN IN BREWING

c. A. MASSCHELEIN

Department of Biochemical Industries, Institute of Fermentation Industries, C.E.R.I.A. Brussels, Belgium

INTRODUCTION

During traditional brewery fermentations, metabolic regulatory control systems determine such cyclic phenomena as sequential uptake of amino acids and fermentable sugars by the brewing yeast cell which, together with synthesis of essential lipids, allows regulated cell growth and division and maintains cellular integrity.

The principle that the rate of fermentation is a function of the rate and extent of cellular growth is well established. It has indeed been shown that each growth period is characterised by a maximal fermentation capacity for glucose, maltose and maltotriose per unit yeast, the extent of which is specific for each yeast strain, and that this diminishes rapidly during the stationary phase to a maintenance level. 1-3

As a result of these studies and others demonstrating the importance of free amino nitrogen,4 lipid and oxygen concentrations5 for improved yeast growth and performance, it may be concluded at the practical brewing level that inadequate growth not only reflects on the economics of inconsistent fermentation times, but, at the limit also results in poor attenuations and altered beer flavour notes. However, cellular growth per se is of little interest to the brewer when he is primarily interested in producing ethanol and a quality beer with the desired sensory and analytical profile. The production of cell mass and yeast storage material (glycogen) by growth is of limited interest only, to enable the brewer to have sufficient yeast for repitching purposes and to ensure that the yeast is in a reasonable condition for the next fermentation.

77


78 C. A. MASSCHELEIN

Over-production of biomass and yeast storage material ultimately means that less substrate is channelled to the product.

The brewer, therefore, is faced with a compromise, viz. to achieve sufficient yeast growth to gain an optimal rate and extent of attenuation and desired flavour development whilst balancing but not over-expending nutrients and energy for accumulation of reserve material. Moreover, modern day brewing is complicated by stress factors that impact negatively on yeast growth and consequently on specific and volumetric productivities viz.:

-wort density (high gravity brewing) -ethanol concentration -pressure fermentations -deep fermentation -adjunct rate -agitation -continuous fermentation concepts with internal and external

recycling -mass transfer limitations in immobilised systems

It is, therefore, clear that the full potential of brewing strains is not presently achieved and needs further study at the metabolic level to secure more efficient and controlled fermentations.

ENVIRONMENTAL FACTORS AFFECTING PRODUCTIVITY AND BEER FLAVOUR

While temperature is a dominant rate-limiting factor in accelerated batch fermentation, a parallel effect on yeast metabolism and beer flavour must be considered.

Fermentation efficiency, being highly dependent on the initial lag phase temperature adjustments, should logically be made from inoculation onwards. Chances for a more intensive biosynthesis and overflow of metabolic intermediates will, however, be greater at higher temperatures since amino acids involved in their formation are only gradually taken up in the early stages of fermentation. The main problem to be considered in addition to this overall stimulating effect, is the influence of fermentation temperature on the distribution between various volatile by-products. According to the results of Ayarapaa6 and Nordstedt et aC different beer aroma components

FERMENTATION TECHNOLOGY AND FERMENTER DESIGN 79

HIGHER ALCOHOLS ESTERS

mg 1-1 mg 1-1 .(1)

75 -11(1)30 / / •

/ *

/ • 50 • 20

/ ""-•• ,6'(2)

• '<2) .. 101(3)

*/ ./ 25

/ 10 / • *'

*_*(3) :;::::::* -*( 4)

•

./ '" .x1O-1( 4)

10 15 20 10 15 20 TEMPERATURE °c

FIG. 1. Effect of temperature on the formation of higher alcohols. (1), Pentanols; (2), 2-phenylethanol; (3), isobutanol; (4), n-propanol, and esters (1), ethyl acetate; (2), isoamyl acetate; (3), phenylethyl acetate;

(4) ethylcaprylate.

show different temperature dependencies. As can be seen from Fig. 1 production of higher alcohols increases between lOoC and 20°C.

High rates are observed for pentanols and 2-phenylethanol. In contrast rates are rather slow for isobutanol and propanol is nearly independent of temperature. Only the formation of ethyl- and phenylacetates increases within the range lO-20°C, isoamylacetate and ethylcaprylate having a maximum near 15°C. Similar results are obtained for caprylic acid, the other free fatty acids (C6 , ClO, C12)

decreasing with increasing temperatures. A more intensive overflow of acetohydroxy acids and vicinal

diketones is also observed when higher starting temperatures are used. In contrast to higher alcohols, vicinal diketones are only produced through anabolic processes from the carbon source. Diacetyl and 2,3 pentanedione are formed by spontaneous oxidative decarboxylation of


the corresponding acetohydroxy acids, metabolic intermediates of the common biosynthetic pathway leading to isoleucine and valine.8,9

The unavoidable accumulation of acetohydroxy acids during the initial stages of fermentation on wort suggest that, in brewery yeasts, the regulatory mechanisms involved in gene expression are inadequate to balance demand and utilisation of valine and isoleucine. Consequently, selection of strains with mutationally impaired catalytic function of acetohydroxy acid synthetase would appear to be a promising approach to prevent overflow of these highly undesirable fermentation by-products. Direct isolation of specific mutants from brewing yeasts is severely hampered by the fact that such strains are often polyploid or aneuploid in nature. Moreover, the low degree of sporulation and poor spore viability make such an approach extremely difficult. Another strategy is to increase the anabolic flow of intermediates leading to isoleucine and valine by inserting additional copies of ILV5 and ILV3 genes.lO

Indeed, the reduced ability to produce vicinal diketones as a result of IL V5 amplification is a good indication that reductoisomerase is one of the rate-limiting steps for the conversion of vicinal diketone precursors to isoleucine and valine. 11

The alternative approach is the use of acetolactate decarboxylase which catalyses the conversion of a-acetohydroxy acids into acetoin and acetylethylcabinol. 12

a-Acetolactate decarboxylases are known to occur in several species of Enterobacter, Lactobacillus and Bacillus. Addition of enzymes produced by these microorganisms results in a rapid removal of a-acetohydroxy acids to a level below the taste threshold value of the corresponding vicinal diketones. Accordingly, application of enzymatic procedures to overcome diacetyl problems, could play an important role for the commercial success of fermentation systems geared to achieve higher productivities.

FERMENTER DESIGN, FERMENTATION EFFICIENCY AND BEER FLAVOUR

If temperature has a stimulating effect on fermentation efficiency, this effect is considerably reduced at the conclusion of the growth phase. At this stage, drop of gravity is mainly controlled by the amount of yeast in suspension and consequently by the carbon dioxide generation


and the natural agitation taking place in the fermenter. Experiments carried out by Delente et al.13 and Borkmann and Manger14 have shown that all gas bubbles are formed at the bottom of the fermenter. The bubbles move upwards by buoyancy force, grow in size by reduction in hydrostatic pressure and by carbon dioxide diffusion from the liquid, and slow down by drag force. As a result the higher the liquid in the fermenter, the more vigorous the natural agitation. The bubble power of agitation is proportional to the rate of carbon dioxide production and correlated to the logarithm of the height of the tank. This is the main reason why fermentation efficiency is much greater in tall thin cylindroconical vessels with higher height to diameter ratios than in conventional horizontal brewery tanks. This would be an ideal situation provided that environmental conditions created by deep fermentation are not adversely affecting beer flavour. The influence of bubble agitation on ester formation in tank fermenters has been studied by Vrieling. 15 Fermentations carried out under the same technological circumstances with respect to wort aeration, wort composition, pitching rate and fermentation temperature indicate that ester content of the final beer is decreasing when the liquid level in the fermenter is increasing.

Ester content is directly correlated to the logarithm of height of fermenter which indicates a correlation with the amount of agitation. Striking similarities are observed between mechanical agitation and bubble agitation. 16

As a result the overall effect of agitation is that beers contain less esters but higher amounts of fusel alcohols and have a lower pH. The negative effects of excessive turbulences are generally intensified by the practice of filling tanks over a long period with aerated charges of wort. 17 According to the volume of a single brew, the number of brews per day and the size of fermenting vessels, considerable variations may occur between different breweries. If all pitching yeast is introduced with the first charge of wort and all charges of wort are equally aerated, important changes in yeast behaviour and beer flavour may be expected compared to a conventional batch process where dissolved oxygen is only available for a short period of time.

Chances for more intensive biosynthesis and overflow of metabolic intermediates are considerably increased with longer filling times and higher temperatures during the infeed period.

Rapid yeast growth will lead to enhanced anabolic formation of amino acid precursors with concomitant overflow of higher alcohols, oxo-acids, organic acids and vicinal diketones.


In contrast reduced ester formation will result from more vigorous turbulence during the early stages of fermentation. These findings provide evidence that the move to greater volume processing cannot be achieved successfully without evaluating the significance of geometry, size and dimensions of fermenting vessels to fermentation kinetics, yeast behaviour and beer flavour.

IMMOBILISED YEAST SYSTEMS IN BREWING

There are many reasons for changing from batch to continuous fermentation processes. Most important are economics of invested capital, reduced running costs, potential for automation and saving of manpower. Therefore, high rate continuous systems offer an interesting challenge to the economics of large volume batch processing. The removal of flavour-matching constraints makes continuous cell recycle reactors on ideal choice for industrial ethanol fermentation. The APV tower system has been operated successfully on a large scale. IS The reactor consists of a vertical cylindrical tower with a conical bottom. The tower is topped by a large diameter settling zone fitted with baffles. The overall geometry is from 7: 1-10: 1 with tower diameters from 0·9-2 m.

High cell densities of 50-80 g/litre are achieved with productivities of 60 g/litre/h ethanol. In brewing, internal recycling using flocculent strains has been utilised for continuous beer production. 19 However, efforts were very much constrained by the need to replicate exactly the flavours of existing batch fermented products. Other major drawbacks of tower fermentation systems in brewing are complexity, lacking of flexibility, contamination risks, heterogeneity as regards cellular distribution in the reactor, restriction in the choice of yeasts and finally a starting time of two to three weeks to build up the desired high cell density and achieve stable operation.

The utilisation of cells immobilised by entrapment within natural polymers eliminates most of these inconveniences. An immobilising matrix may be made into bead form for use in continuous and batch packed-bed or fluidised-bed reactors. Such gel occlusion systems are attractive not only due to improved volumetric productivity but also because more substrate is channelled to the product and less to the production of unnecessary biomass.

Additional advantages for using an immobilised system when


compared to a free cell system are that:

(1) A higher cell mass per unit fermenter can be achieved, resulting in potentially faster process times and, thus, emphasising greater volumetric efficiency.

(2) For batch systems there is no need to remove the immobilised cells at the end of the fermentation; the process is reduced to simply racking the product and refilling the reactor/fermenter with substrate; thus, complicated microbial handling systems are avoided.

(3) For continuous systems ease of operation control facilitates kinetic optimisation for yield improvement.

(4) Smaller reactor volumes may be envisaged, hence, lower capital costs.

(5) For brewing yeast strains flocculation properties need not be taken into consideration.

Gel occlusion systems have been described by various workers for application in the brewing process.20-23 However, for such strategies to be successful there is an increasing need to find a solution which satisfies both productivity and the quality of the final product.

Experience acquired over the past five years in our laboratory has shown that the stability of entrapped systems is related to the renewal of cells, and that productivity gains are proportional to the concentration of cells immobilised in the reactor. 22.24

Diffusional limitations have been observed in entrapped cell matrices and are therefore considered to be a major drawback for optimal cellular metabolism in such systems.25 In this regard an understanding of the physiology of brewing yeast cells operating without or with very limited growth is a prerequisite for commercial application of immobilised cell systems. The activity of entrapped yeast, however, is not only a direct function of growth-related aspects. Yeast cells entrapped in an immobilising matrix are subject to a completely different environment to cells in free suspension, where mass transfer and immobilisation rank equal in defining and explaining the activity of cells in the carrier polymer.

With this background we have been exploring the association of growth-related aspects with mass transfer limitations between batch and continuous immobilised cell systems for amino acid uptake and production of the primary and secondary products of yeast metabolism which define the flavour-active quality of beer.


STABILITY AND PRODUCTIVITY OF A CONTINUOUS ALGINATE ENTRAPPED CELL SYSTEM USING

DIFFERENT DILUTION RATES

Closer study of diffusional limitations in immobilised type continuous systems have revealed that growth rate is a function of the leakage of cells from the alginate matrix so that the time for yeast to double in amount can rise to as much as 2000 h. As metabolic flux and fermentation performance are regulated by cellular growth, it is evident that the full potential of brewing yeast strains could not be achieved in such systems. With this background we have examined the behaviour of lager yeasts with respect to their ability to utilise wort sugars in free and immobilised systems.

Continuous fermentation experiments were carried out at 15°C in the upward flow packed-bed reactor as previously described.25 The alginate entrapped yeast was at an initial concentration of 1· 5 x 106 cells/mg gel. After one week's operation in the reactor, the cell concentration in the gel beads reached 2·47 x 106 cells/mg gel.

Flow rates were adjusted in order to obtain steady state conditions at 30, 50 and 80% attenuation. Beads were removed from the

TABLE 1 CONTINUOUS FERMENTATION PARAMETERS OF WORT WITH

ALGINATE ENTRAPPED Saccharomyces uvarum IN A PACKED

BED REACTOR WITH FLOW RATES ADJUSTED AT 30, 50 AND

80% ATTENUATION

Parameters Attenuation (% )

Total volume (ml) Total bead volume (ml) No. of beads Bead volume (mm3)

Loading capacity (% dry yeast/beads) Yeast (g dry weight/litre) Free cells (%) Wort volume (ml) Flow rate (ml/h) Dilution rate (per h) Fermentation time (h)

30

500 370

8600 43·2 17·1

166

130 620

4·76 0·2

50 80

750 800 392 395

8570 8500 45·8 46·5 22·2 24

145 140 0·1 2·45

358 405 115 50

0·3 0·12 3·0 8·0


fermenter after 5 and 30 days operation and dissolved in a 10% sodium gluconate solution. Relevant experimental parameters are shown in Table 1. Cells were harvested and washed by centrifugation and finally resuspended in a 0·66 M KH2P04 solution.

Manometric determination of the fermentation rates on glucose, maltose and maltotriose at non-limiting concentrations and at 30°C were undertaken and compared to free and immobilised cell fermentations conducted batchwise in EBC tubes. Both growing and non-growing cells were used in the immobilised tall tube experiments. Changes in 'fermentative capacity are shown in Table 2.

Results obtained in free cell experiments are in agreement with previous studiesl- 3 showing that each sugar reaches a fermentation capacity maximum during the growth phase which diminishes as the yeast enters the stationary phase.

Fermentation rates for each of the measured wort sugars are substantially higher at 30% attenuation than those obtained towards the end of fermentation. These results also highlight the relatively severe effects of growth limitations on the specific fermentation rates for maltose and especially for maltotriose compared to the more mild effects on glucose. Similar batch fermentation experiments carried out with alginate entrapped cells indicate a further and substantial loss of fermentation capacity per unit yeast. Compared to free cell experi-

TABLE 2 CHANGES IN SPECIFIC FERMENTATION RATES (~l CO2 h- 1 mg- 1 DRY

YEAST) ON GLUCOSE (1) MALTOSE (2) AND MALTOTRIOSE (3) IN FREE

AND IMMOBILISED CELL SYSTEMS

Fermentation Attenuation(% ) conditions ... _---

30 50 80 ~-------

(1) (2) (3) (1) (2) (3) (1) (2) (3)

Free cells/batch 400 438 182 254 224 86 222 182 86 Immobilised growing 272 152 55 160 125 48 76 89 12

cells/batch Immobilised non- 148 130 48 159 142 36 72 75 44

growing cells/batch Immobilised cells 204 215 64 251 187 41 177 123 33

continuous/5 days Immobilised cells 100 66 19 82 74 20 85 58 17

continuous/30 days


ments, the minimum levels obtained by adjusting the flow rate at a preselected attenuation of 80% decreased from 250 to 70 JlI CO2/h/mg yeast (dry). Interestingly, cyclic variations in the rates of fermentation of the individual sugars observed at various stages of the free cell batch fermentation tend to disappear when yeast growth is more or less completely restricted by the immobilising matrix. This is typically the case of a packed-bed reactor operating continuously for a long period of time. Prolonged non-growth conditions imposed by cell immobilisation led invariably to extremely low but stable maintenance levels even when flow rates were adjusted at attenuations where a maximum rate was measured in the free cell system. The viability of the yeast did not show a similar trend. Indeed, after using immobilised cells in the packed-bed reactor for 30 days, a viability level of 90% was demonstrated. Glycolysis is an energy producing pathway, and we are in fact carrying out the fermentation using energy-producing cells under conditions of low or no energy demand. While this situation is satisfactory to some extent with respect to the stability of the system, it raises numerous questions if full control over yeast behaviour is to be secured.

EFFECT OF IMMOBILISATION AND FERMENTATION TECHNOLOGY ON SPECIFIC AND VOLUMETRIC RATES

OF SUBSTRATE UTILISATION

Previous manometric determinations of the fermentation rates on individual wort sugars were made at 30°C and at non-limiting concentrations. Such conditions are giving maximum rates which are not necessarily reflecting those prevailing during a brewery fermentation. In order to have an overall picture of the fermentative behaviour of yeast during free and immobilised cell experiments, specific (qs:g/g cell mass/h) and volumetric (Q: g/litre/h) rates of substrate utilisation were calculated in terms of the course of change in gravity and yeast concentration (X/g/litre). Results are summarised in Table 3.

Specific rates are independent of cell mass concentration and describe the effectiveness of the cells in substrate utilisation. In this way useful information can be gained to assess fermentation conditions which are the most effective. It can be seen from results shown in Table 3, that it is almost essentially the continuity of the system which is responsible for the dramatic decrease in the specific rate of


TABLE 3 CHANGES IN SPECIFIC AND VOLUMETRIC SUGAR UTILISATION RATES DURING

FREE AND IMMOBILISED CELL FERMENTATIONS

Fermentation Attenuation(% ) conditions Qs: gig yeast/h 30 50 80 Q: g/litre/h X: g/litre qs X Q qs X Q qs X Q

Batch, free cells 0·62 4 2·5 0·48 4·3 2·1 0·47 4·3 2·0 Batch, immobilised 0·83 1·3 1·1 0·80 1·3 1-1 0·46 1·3 0·6 Continuous, 5 days 0·22 166 35·9 0·05 145 7·5 0·04 126 5·0 Continuous, 30 days 0·12 108 12·5 0·03 108 3·3 0·03 108 3·2

fermentation. Indeed, even after five days continuous operation in a packed-bed reactor the specific rate dropped to approximately one-tenth of the maximum rate measured near the end of both the free and immobilised cell batch fermentation. Although the above limitations weigh heavily against the application of continuous methodology for beer production, a higher cell mass per unit fermenter can be achieved through immobilisation, resulting in potentially faster process time and, thus emphasising greater volumetric efficiency.

The effect of cell mass concentration on the volumetric rate of sugar utilisation is shown in Table 3. Even at a specific rate ten times lower than that obtained at the end of the batch fermentation, the volumetric rate is still more than double compared to a free system.

In order to explain anomalies encountered in immobilised brewing yeast systems concerning retention of fermentation capacity over extended operating periods, extensive research has been carried out to understand the metabolic interrelationship between yeast growth and regulation of glycolytic flux. 25

MASS TRANSFER LIMITATIONS AND FATTY ACID SYNTHESIS IN IMMOBILISED YEAST SYSTEMS

It has already been pointed out that cell concentration in an immobilised cell reactor may be several times higher than in an equivalent sized free cell system. An adequate cellular oxygen supply is critical for conventional brewery fermentations so it should be even


more important for the high cell densities as encountered when cells are immobilised. As a result, it is reasonable to suppose that cells in the outer layer of the beads will consume most of the available oxygen, creating an oxygen gradient within the bead.

In order to investigate whether the above suggestions are metabolically correct, we have examined the relationship between oxygen uptake and synthesis of unsaturated fatty acids in free and immobilised cell systems. Yeast samples were harvested after three successive fermentations and fatty acid determinations were carried out before and after all oxygen had been removed from air-saturated wort. The calcium alginate pellet was fractionated in three layers, each representing one-third of the total cell mass. Fatty acids contents of outer and inner layers are shown in Table 4.

It can be seen that the synthesis of unsaturated fatty acids is essentially dependent upon the supply of oxygen to the interior of the gel bead thus confirming a less favourable oxygen balance in the centre of the calcium alginate pellet.

It is clear from these results that the bottleneck for optimal oxygen

TABLE 4 FATTY ACID COMPOSITION OF FREE AND IMMOBILISED CELLS BEFORE

AND AFTER OXYGEN REMOVAL FROM AIR-SATURATED WORT

Fatty acid Oxygen absorption (mg/g dry wt) Before After

--~

Free Free Immobilised layer Outer Inner

CIO 2·25 1·9 1·4 1·7 CI2 1·3 1·2 1·05 1-15 CI4 1·0 0·9 0·95 0·9 CI6 7·8 8·3 8·2 8·6 C16.1 10·7 21·6 26·7 18·7 CIS 2·4 5·25 5·3 4·2 C18.1 3·95 14·8 17·0 7·9 C18.2 ND ND 0·3 0·15 Total unsaturated fatty acids 14·65 36·4 44·0 26·75 Total fatty acids 29·4 53·85 61·8 43·3 Unsaturated x 100

50·0 67·5 72·5 61·7 Total fatty acids

ND-Not detected.

FERMENTATION TECHNOLOGY AND PERMENTER DESIGN 89

transfer and fatty acid synthesis is the physical-mechanical characteristics of the immobilising matrix. In this regard it is obvious that size, texture and porosity are of great importance for the performance of immobilised brewing yeast cell systems.

EFFECT OF IMMOBILISATION ON THE FORMATION OF FLAVOUR-ACTIVE COMPOUNDS

According to the extremely low metabolic activity of alginate entrapped yeast with respect to growth rate, fermentation activity and amino acid uptake25 the rate of formation of higher alcohols and esters would most likely also be reduced in immobilised systems. It can indeed be seen from results shown in Table 5 that there is a considerable decrease in the amount of flavour-active compounds produced per unit yeast between the free and immobilised cell systems. It was further shown that in immobilised systems ester formation occurs as an independent function of yeast growth and that alcohol acetyl transferase activity is entirely controlled by changes in the fatty acyl composition of the plasma membrane depending on the oxygen balance within the alginate bead.26

Free vicinal diketones and their precursors a-acetolactate and a-acetohydrobutyrate, were found to be excreted at significantly higher concentrations during the continuous packed-bed reactor fermentation than encountered under batch fermentation conditions using free and immobilised cells. Nevertheless, as for esters and higher alcohols amounts produced per unit yeast were relatively low when compared to a free cell system.

TABLE 5 HIGHER ALCOHOL AND ESTER CONTENT OF BEERS FERMENTED IN

FREE AND IMMOBILlSED(PACKED-BED) CELL SYSTEMS

Higher alcoholl ester

n-Propanol Isobutanol Isoamyl alcohol Isoamyl acetate Ethyl acetate

Free (mg/litre)

9·g 16·0 62·0 2·0

19·0

Immobilised (mg /litre)

g·o 7·5

31·0 0·06

11·0

Free I immobilised (mgll00g dry wt

yeast I litre)

19x 22x 32x

600 x 25x


CONCLUSIONS

The development of high gravity brewing procedures, the reduction of fermentation and lagering time, the design and construction of large capacity unconventional vessels and the combination of unit processes for fermentation and maturation are obviously the result of both economic pressures and technological progress.

To maximize volumetric productivities, fermentation parameters and fermenter design must be optimised with respect to agitation and yeast concentration. The bubble power of agitation is proportional to the rate of carbon dioxide production and increases logarithmically with the liquid depth in the fermenter. Fermentation efficiency is therefore considerably increased in tall thin cylindroconical vessels. Correct pressure and temperature regimes should be determined carefully when such vessels are used. The negative effects of excessive turbulences on beer flavour are generally intensified by the practice of filling over a long period with aerated charges of wort.

Immobilised cell technology offers an interesting challenge to the economics of large volume processing. Immobilised brewing yeast systems are attractive not only due to improved volumetric productivity but also because more substrate is channelled to the product and less to the production of unnecessary biomass. In our opinion further optimisation of immobilised systems with respect to mass transfer, reactor design and yeast physiology is the most logical strategy in providing the means of increasing productivity without changing the flavour characteristics of the fermented products.

REFERENCES

1. Harris, G. and Millin, D. J. (1963). Proceedings of the European Brewery Convention Congress, Brussels, 400.

2. Masschelein, C. A., Dupont, J., Ramos-Jeunehomme, C. and Devreux, A. (1960). Journal of the Institute of Brewing, 66, 502.

3. Masschelein, C. A., Jenard, H., Ramos-Jeunehomme, C. and Devreux, A. (1965). Proceedings of the European Brewery Convention Congress, Stockholm, 209.

4. Kirsop, B. H. (1978). European Brewery Convention on Fermentation and Storage Symposium, Monograph V, pp. 3-16.

5. Aries, V., Kirsop, B. H. and Taylor, G. T. (1977). Proceedings of the European Brewery Convention Congress Amsterdam, 255.

6. Aylirlipaa, T. (1970). Brauwissenschaft, 23,48.


7. Nordstedt, C., Bengtsson, A., Bennet, P., Lindstrom, I. and Ayrapaa, T. (1975). Proceedings of the European Brewery Convention Congress, Nice, 581.

8. Inoue, T., Masuyana, K., Yamamoto, Y., Okada, K. and Kuroiwa, Y. (1968). Proceedings of the American Society of Brewing Chemists, 158.

9. Masschelein, C. A., Ramos-Jeunehomme, c., Jenard, H. and Devreux, A. (1971). Proceedings of the European Brewery Convention Congress, Estoril, 211.

10. Dillemans, M., Goossens, E., Goffin, O. and Masschelein, C. A. (1987). Journal of the American Society of Brewing Chemists, 45, 8l.

11. Goossens, E., Dillemans, M., Debourg, A. and Masschelein, C. A. (1987). Proceedings of the European Brewery Convention Congress, Madrid, 563.

12. Godtfredsen, S. E. and Ottesen, M. (1982). Carlsberg Research Communications, 47,93.

13. Delente, J., Akin, C., Kralobe, E. and Ladenburg, K. (1968). Technical Quarterly of the Master Brewers' Association of America, 5,228.

14. Borkmann, K. and Manger, H. J. (1975). Die Lebensmittelindustrie, 22, 501.

15. Vrieling, A. M. (1978). Proceedings of the European Brewery Convention, Symposium on Fermentation and Storage, 135.

16. Masschelein, C. A. (1981). Brewing and Distilling International, 37. 17. Masschelein, C. A. (1975). Brauwelt, 115, 608. 18. Greenshield, R. and Smith, E. (1971). Chemical Engineer, (May) 249. 19. Portno, A. D. (1978). Proceedings of the European Brewery Convention,

Symposium on Fermentation and Storage, 145. 20. Godtfredsen, S. E., Ottesen, M. and Svenson, B. (1981). Proceedings of

the European Brewery Convention Congress, Copenhagen, 505. 21. Linko, M. (1981). Proceedings of the European Brewery Convention

Congress, Helsinki, 37. 22. Masschelein, C. A. (1987). Proceedings of the European Brewery

Convention Congress, Madrid, 208. 23. White, F. H. and Portno, A. D. (1978). Journal of the Institute of

Brewing, 84,228. 24. Masschelein, C. A. (1986). Proceedings of the European Brewery

Convention, Symposium on Brewers' Yeast, 2. 25. Ryder, D. S. and Masschelein, C. A. (1985). Journal of the American

Society of Brewing Chemists, 43,66. 26. Masschelein, C. A., Carlier, A., Ramos-Jeunehomme, C. and Abe, I.

(1985). Proceedings of the European Brewery Convention Congress, Helsinki, 339.

rj rH S t t lag

T TK TR Uo VM VF

W X,X' Y;

!iT

Chapter 8

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCTION

MAURO MORESI

l.M. T.A.F., University of Basilicata, Potenza, Italy

NOTATION

Acetic acid concentration (mol/litre) Diffusion coefficient (m2/s) Ethanol concentration (mol/litre) Glycerol concentration (mol/litre) Cell-death rate constant (mol h/litre) Non-competitive ethanol inhibition constant (mol/litre) Saturation constant (mol/litre) Substrate inhibition constant (mol/litre) Ammoniacal nitrogen concentration (mol/litre) Consumption or formation rate of component i (mol/litre h) Fermentation heat rate (kJ /mol substrate consumed) Substrate concentration (mol/litre) Fermentation time (h) Duration of the induction phase (h) Fermentation temperature (0C) Absolute temperature (K) Reference temperature = 289 K Design overall heat transfer coefficient (W /m2 K) Maximum specific cell growth rate (l/h) Working volume of each fermentation vessel (m3)

Concentration of water (mol/litre) Biomass concentration (mol/litre or g/litre) Yield coefficient (mol ith component/mol substrate consumed)

Mean temperature difference (0C) 93


94 MAURO MORESI

Subscripts A Acetic acid C Carbon dioxide E Ethanol G Glycerol N Ammoniacal nitrogen o Initial S a-,8-D-glucose X Cell biomass W Water

INTRODUCTION

Up to almost fifty years ago the design of the equipment used for ethanol production (be it for beer, wine, cider, perry, or potable spirit) was more an art than a science, being the result of a large number of practical observations well rooted in traditional beliefs and techniques.

Overviewing a century of beer fermenter design, Maules1 makes clear the difficulty of defining a satisfactory fermenter for two main reasons:

(1) Inaccurate knowledge of yeast metabolism, which does not allow the overall pattern of ethanol fermentation to be modelled reliably.

(2) Multipurposity of the equipment so as to be suitable for fermentation, maturation and storage.

At the same time, physical phenomena, such as heat and mass transfer among must (or wort), yeast and the external environment, may limit the biological reactions associated with ethanol fermentation. All this makes beer or wine fermenters unconventional. Thus, their design is still, by no means, standard. Generally speaking, a fermenter for alcoholic beverage production should:

(1) guarantee the residence time necessary for the right evolution of the biochemical reactions associated with beer or wine fermentation;

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCTION 95

(2) keep the fermentation temperature within the optimal range of values;

(3) allow adequate mixing of its content (must and yeast); (4) produce a drinkable beverage.

Since the alcoholic fermentation occurs in the liquid phase, the Fundamentals of Chemical Reaction Engineering2 suggest using open or closed tanks, eventually of the stirred type. It is worth noting that the early equipment used for beer- or wine- making was a wooden cask, subsequently replaced by epoxy resin- or enamel-lined cement vats either open or closed. So, to assemble a complete picture of the operation of ethanol fermenters, it is necessary to integrate biological reactions kinetics and mass and heat transfer with gas-liquid mixing and contacting patterns in the equipment. In fact, different design and scaling-up procedures are required for reactors with different flow and mixing characteristics. The major task in modern fermenter design will be to blend these various ingredients to obtain a coherent overall strategy for design and analysis of biological fermenters for ethanol production, as a preliminary step for the development of their automatic control.

In this communication the basic features of the ethanol fermentation kinetics, as well as the main fermenter designs, will be reviewed with special emphasis on wine fermentation.

WHITE WINE FERMENTATION MICRO KINETICS

Such a fermentation is controlled by a number of parameters, such as:

-substrate concentration -ethanol concentration -dissolved oxygen level -pH -temperature -additional nutrients requirements -secondary metabolite inhibition -concentration of polyphenolic substances, especially in wme

fermentation

There is an extensive literature on this topic and in other chapters of this volume the effects of the above variables on ethanol fermentation kinetics are reviewed.

96 MAURO MORES!

To better specify how the type of information available can be used to construct a reliable model of the fermentation examined, the case of wine fermentation will be more specifically examined. To this end, the anaerobic fermentation of grape must by a strain of Saccharomyces cerevisiae with ammonia as the nitrogen source was described by the following stoichiometric equation:3--7

C6H 120 6 + YNNH3~ YXCHl.7S00.4SNo.2 + YECzH60 + YG C3Hg0 3 substrate biomass ethanol

+ YA C2H40 2 + YCC02 + YwH20 acetic acid

glycerol (1)

where 1'; is the stoichiometric coefficient of any reagents or products in eqn (1).

Second, having defined the above eight relevant compounds, four elementary balances for the elements carbon, hydrogen, oxygen and nitrogen and an enthalpy balance were formulated, thus establishing five relationships between the conversion rates of substrate (rs), ammoniacal nitrogen (rN) , biomass (rx) , ethanol (rd, glycerol (rG) , acetic acid (rA), carbon dioxide (rd, water (rw), and heat evolved (rH)' In particular, the heats of formation of the compounds examined were extracted from Ref. 4.

Third, the mathematical description of wine fermentation process was completed by introducing four kinetic equations so as to solve the aforementioned equations. By selecting rs, rx, rE and rc as key parameters of this fermentation, it was possible to produce the following:

rA = -3 rs - 0·3126 rx + 2 rE - 3·5 rc (2)

rG = -0·125 fx - 2 fE + 2 rc (3)

rw = +0·55 rx + fE - rc (4)

fH = -194·76 rs + 2·48 fx + 193·37 fE - 240·85 fc (5)

Moreover, it was found that fE and fc are functions of fs, that is also a function of fx:4

fC= -Ycfs

fs = -rx/Yx

(6)

(7)

(8)

Finally, yeast growth rate was described by applying the well-known

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCTION 97

TABLE 1 REGRESSION EQUATIONS OF THE STOICHIOMETRIC COEFFICIENTS (YX ' YE , Yd AND KINETIC CONSTANTS (VM' K E, Ks, KSI> AND Ko) OF WHITE WINE

FERMENTATION, AS EXTRACTED FROM REFS 3, 4 AND 7

Parameter Regression Unit

VM = 0·812 exp( -7161 (l/TK -l/TR» -0·187 exp( -16551 (l/TK -l/TR » l/h

KE = 0·875 exp(5938 (l/TK - l/TR» mol/litre Ks = 0,988 exp(5938 (l/TK - l/TR» mol/litre KSI =0·15gexp(5938 (l/TK -l/TR» mol/litre Ko =913 mol h/litre Yx = 0·052 + 2·9 x 1O- 3T mol biomass

mol substrate consumed YE = 1·897 - 1·46 x 1O-3T mol ethanol

mol substrate consumed Yc = 1·961- 3·14 x 1O- 3 T mol CO2

mol substrate consumed -----"----- - ---.-~

Monod equation modified to take account of substrate and ethanol inhibition, as well as cell decay resulting from ethanol accumulation in the fermenting mUSe-iO

(9)

where VM is the maximum specific cell growth rate, Ks the saturation constant, KE the ethanol non-competitive inhibition constant, KSI the substrate inhition constant, KD cell death rate constant, X, E and S the active yeast, ethanol and reducing sugars concentrations. Table 1 shows the regression equations of both the stoichiometric coefficients (Yx , YE , Yd and kinetic constants (VM' K e , K s , KSh and K D ), as determined by fitting a series of experimental data obtained from white wine fermentation carried out by Cantarelli et al. 5,6 and Castor and Archer, 11 respectively,

The lag phase of yeast growth was also modelled by assuming no cell growth for a period decreasing with the inoculum size: 12

rx = 0 for t ~ tlag (10)

with

tlag = 3·12(X~)-O'585 (11)

98 MAURO MORESI

15 300r-----------------------~150 1·5

~10 200 E 100 1·0 011 L. ~ 0, GI GI t; ~ L. L. L. « ~ ~ ~

1 ~ 0. -~ ~

Vl lIJ Z

E> 5 100 G 50 0.5 I!)

x· A

X

o L...oIi~t:.-.-_ _O____O___+_4__....:===+==!==I 0 o 40 80 120

Time (hl

FIG. 1. Time course of a typical isothermal white wine fermentation as simulated by using the mathematical model here described: Concentrations of substrate (S), yeasts (X), ethanol (E), glycerol (G), acetic acid (A) and

ammoniacal nitrogen (N) against time (t).

where X~ is the initial yeast concentration (in g/litre) and t lag is expressed in hours. Figure 1 shows the evolution of a typical traditional white wine fermentation at constant temperature.'

DESIGN CONSIDERATIONS

Bioreadors Among the many different types of bioreactors that can be used in ethanol fermentation, Fig. 2 illustrates their basic structures, thus leading to the following classification:

(a) barrels; (b) vats; (c) cylindroconical fermenters; (d) cylindrical fermenters with gently sloping bases and low

height-to-diameter (i.e. aspect) ratios;

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCfION 99

- - - - -

0 a b c

d

e

FIG. 2. Alternative designs of the main bioreactors used in alcoholic beverage production: (a) barrel; (b) vat; (c) cylindroconical fermenter; (d) cylindrical tanks with gently sloping bases; (e) tower fermenter; (f) spheroconical

fermenter.

100 MAURO MORESI

(e) cylindrical fermenters with high aspect ratios (Le. tower fermenters) ;

(f) spheroconical fermenters.

Chronologically speaking, the wide variety of wooden casks (Fig. 2(a» traditionally used to ferment grape must or wort has been progressively replaced by cement vats lined with glazed tiles or epoxy resins (Fig. 2(b». These were then substituted with epoxy-resin lined mild steel tanks or, more recently, with stainless steel ones resembling the so-called Nathan vessel (Fig. 2(c» patented as early as 1908 and 1927,tJ Such equipment is nowadays largely used for beer and wine fermentation in the range 1-340 m3 •1 Type (d) is also widely used not only in white wine fermentation, but also in red fermentation since its sloping base allows an easy sedimentation of grape seeds and their ready removal from the vessel while it is full.

The tower fermenters, originally developed for the production of lager and British mild ale,14 consists of stainless steel cylindrical vessels with aspect ratios of 6-12 equipped on the top with foam separators to promote yeast sedimentation and attain yeast concentrations up to 350 g/litre. 15 Successful operation of the system depends, to a large extent, on the nature of the yeast employed. In particular, the yeast must be extremely flocculent and sedimentary in nature in order to counteract both the upward movement of beer and the large quantities of carbon dioxide evolved.

The spheroconical fermenters have the maximum surface area-tovolume ratio as compared to the other vessels in Fig. 2, but suffer from the disadvantage of high construction costs owing to difficult metal sheet rolling and large material waste. In the circumstances, their construction becomes competitive with that of cylindrical vessels for capacities greater than 300 m3• 1 A large number of such tanks were installed in Spain in 1972 and were successfully used for lagering in the same vessel. 16

As said before, almost all the ethanol fermentation processes are still conducted batchwise. So, once the vessel has been filled, the wort or must is not removed until a certain reaction time has elapsed, after which all the fermenter contents are harvested. Then, the vessel is cleaned and newly filled with fresh raw must to start another fermentation cycle.

Although continuous fermentation for beer production has been attempted since 1906,17 nowadays no company, with the exception of

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUcnON 101

two breweries in New Zealand,18 is operating continuous brewing for such a fermentation technique generally tends to create more problems than it solves. In fact, it requires constant laboratory monitoring and complex automatic control of process parameters. At the same time, it reduces the flexibility of a brewery to satisfy consumer requirements. l ,18

On the contrary, one of the main disadvantages of batch fermentation is low ethanol productivity. In tower fermenters use of flocculent yeasts helps in maintaining a high cell concentration within the system, thus sustaining higher ethanol formation rates.

When non-flocculent yeasts are employed, the only way to accomplish such requirements is to retain the cell population inside the equipment by means of the following techniques:

-recycling after centrifugation or hydrocycloning, sedimentation and cooling, or membrane separation;

-cell immobilisation by adsorption to several solid supports (i.e. wood chips, sand, activated charcoal, exchange resins, etc.), entrapment within natural polymers (e.g. sodium alginate, K-carrageenan, polyacrylamide, etc.), or chemical reactions, such as copolymerisation with, or covalent binding to, chemically activated supports.

Although ethanol fermentation by free or immobilised cells can be performed in each type of the bioreactors shown in Fig. 2, the tower fermenters are best suited for batch or continuous immobilised-cell fermentation as packed-bed or fluidised-bed reactors. Table 2 lists the main characteristics of a few continuous immobilised-cell bioreactors. According to the type of mixing between the feed and fermenting liquor, the aforementioned fermenters may also be classified as:

(1) perfectly mixed, when the concentration of reagents and products, though variable with time, is independent of size. Under continuous operation, the feed is instantaneously mixed with the vessel contents, the composition of which coincides with that of the stream leaving the fermenter. In barrels and vats or cylindrical vessels with low aspect ratios perfect mixing is generally achieved by virtue of the agitation induced by carbon dioxide bubbles rising, unless grape lees are present (as in the red wine fermentation). In the circumstances, mechanicallyinduced mixing has to be applied, at least from time to time, to

102 MAURO MORESI

TABLE 2 MAIN CHARACfERISTICS OF A FEW CONTINUOUS IMMOBILISED-CELL

BIOREACfORS

Characteristics CSTW Packed-bed Fluidised-reactor bed reactor

Ease of operation Easy Easy Difficult Ease of reaction control Easy Difficult Easy Ease of catalyst replacement Easy Difficult Easy Pressure drop Low High Low Ease of cell removal Easy Difficult Easy Mixing degree Good Poor Good Distribution of viable cells Uniform Uneven Uniform Catalyst attrition High Low Medium Ease of scale-up Easy Easy Difficult Suitability for:

-Product inhibited kinetics Poor Good Fair -Substrate inhibited kinetics Good Poor Fair

Cost Low Low High

a Continuously-stirred tank reactor.

allow a more uniform distribution of substrate and yeasts and a more effective extraction of aromatic and polyphenolic substances from grape skin and seeds.

(2) Completely segregated (ideal plug-flow ferrnenter), when at constant fluid velocity across each empty section of the reactor longitudinal gradients of concentration and temperature occur. Such a situation generally applies to packed-bed reactors, in which backmixing is relatively small. 19

(3) Partially segregated (longitudinal heat and mass diffusivity), when backmixing is more or less pronounced as observed in liquid fluidised- and loosely packed-bed reactors, respectively. As the liquid recirculation ratio is increased, the degree of backmixing increases and the fluidised-bed reactors tend to behave as perfectly mixed ones.2O

Mass Transfer In wine or beer production the ethanol fermentation rate does not appear to be limited by mass transfer resistances, unless immobilisedcell reactors are used. In this case, the limited mass transfer steps can be either interparticle mass transfer or intraparticle diffusion.

PERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCfION 103

The former refers to the nutrient mass transfer from the fermenting must to the external surface of the immobilised-cell support particles. Owing to the low ion and substrate diffusivity in liquids and the low values of the Reynolds number generally encountered with liquid flow through packed beds, large concentration gradients usually occur, unless the rate of agitation in a stirred reactor or the flow rate through a tubular reactor is increased.

The intraparticle diffusion concerns the transfer of nutrients within the tortuous pores of each immobilised-cell particle and makes the reaction rate generally less at the centre than at the outside surface of the pellet. Such a phenomenon can be taken into account by introducing the effectiveness factor, that is the ratio of the observed reaction rate to that which would be obtained if all the cells entrapped within the pellet were available to the reagents at the same concentrations they have at the external surface. 21

Internal diffusional restrictions may be detected by a lowering of the apparent activation energy for the reaction and have been associated with values of the Thiele modulus greater than unity. 21 In fact, such a parameter is proportional to the pellet size and to the square root of the ratio of the reaction rate constant per unit area to the effective molecular diffusion coefficient of substrate within the tortuous intrapellet pores (De).

The latter reflects the hydrogen bond mediated interactions of substrate with the support material and steric hindrance between substrate and the support and can be estimated by correcting the bulk diffusion coefficient of substrate for the porosity of the support and tortuosity of the pores, since the pores in the matrix are sufficiently large to allow reasonably free diffusion of substrate into, and product out of, the support, but small enough to completely retain the entrapped or linked cells.

However, division of the entrapped cells causes complications. Often, the mechanical strength of the immobilised-cell preparation is weakened and, in extreme cases, the granules may be broken. Although the concentration of immobilised cells has been increased, the new cells are not evenly dispersed within the support material, which leads to increased diffusional restrictions. Furthermore, the presence of these additional cells reduces the effective diffusion coefficient of the substrate inside the support in proportion to the percentage of the gel cross-section occupied by cells.22

Moreover, the problems of getting product out of the immobilised-

104 MAURO MORESI

cell preparation are very similar to those concerned with substrate movement into the cell.

In general, a balance has to be achieved between immobilised-cell preparations of a high activity per unit reactor volume and with a consequently low effectiveness factor, and preparations of low activity and high effectiveness factor. Within certain limits, the interparticle mass transfer and internal diffusional restrictions on reaction rate can be minimised by decreasing the size of the immobilised-cell particles and promoting as vigorous mixing throughout the bed as that induced by the co-current flow of a liquid and a gas, thus leading to fluidised-bed reactors. Further considerations about the design and operation of immobilised enzyme and/or microbial-cell reactors are given in Refs 23 and 24.

TABLE 3 RELATIVE FEATURES OF THE MAIN OUT- OR IN-VESSEL HEAT TRANSFER

EQUIPMENT USED IN WINE AND BEER PRODUCTlON7,25

Heat transfer UD AT Relative equipment (kWm 2K) (0C)

FFA LRT LOR MC Cost

Shell-and-tube 0·50-1-50 >10 M E D C D Double-pipe 0·80-1·50 >10 M D C B C Trombone cooler 0·09-0·11 >10 M D C A A Scraped-surface 0·70-0·90 <10 VH A A E G Spiral-tube 0·10-0·30 >10 M D C A B Plate-type 1·80-2·10 <10 C B A D E Spiral-type 0·90-1·20 <10 H C B D F

Pipe-tank Coils 0·20-0·35 >10 M B Vertical falling- 0·01-0·03 >10 M A

film cooler Jacket 0-06-0·15 >10 M D External-tank 0·08-0·20 >10 M C

coils

FFA = Feed fouling attitude: C-clean; M-mild; H-high; VH-very high. LRT = liquid residence time. LOR = Lay-out requirements. MC = Maintenance cost. Each relative parameter increases from A (least figure) to G (highest figure).

FERMENTER DESIGN FOR ALCOHOLIC BEVERAGE PRODUCfION 105

Heat Transfer The exothermic nature of ethanol fermentation requires temperature control, unless very small size fermenters are used. For fermenter capacities smaller than 150 m3 , both in- and out-cooling systems may be applied. However, whatever their size cement vats and mild steel tanks both lined with epoxy resin need external heat exchanges to control the reaction temperature owing to the poor thermal conductivity of their wall materials. The relative features of the main out- or in-vessel heat transfer equipment used in winery and brewery are compared in Table 3.7,25

FUTURE PROSPECTS AND CONCLUSIONS

As a result of this review, it is possible to emphasise the following prospects for the industrial alcoholic beverage fermenters:

(1) radical changes in design are not anticipated in the near future; (2) automatic control of fermentation is the ultimate goal of present

development effort; (3) continuous operation for the production of beer and white table

wine, as a drawback of point (2).

From a research viewpoint, the following may be expected in the near future:

(1) development of more reliable packed- or fluidised-bed reactors, (2) fermentation of genetically-engineered microorganisms in highly

concentrated slurry systems, especially for unmalted beer production.

REFERENCES

1. Maules, D. R. (1986).1. Inst. Brew., 92, 137. 2. Brotz, W. (1965). Fundamentals of Chemical Reaction Engineering,

Academic Press, New York. 3. Battley, E. H. (1960). Physiol. Plantarum, 13,628. 4. Williams, L. A. (1982). Am. 1. Enol. Vitic., 33, 149. 5. Cantarelli, C. and Moresi, M. (1983). Atti Accademia Ital. Vite e Vino, 35,

225. .

106 MAURO MORESI

6. Cantarelli, c., Moresi, M. and Manfredini, M. (1984). In: Proceedings of the 7th Int. Oenological Symp., Rome (Italy), 7-9 May 1984, E. Lemperle and H. Rasenberger (Eds), Edition of the Int. Association for Modem Winery Technology and Management, Breisach, West Germany, p. 317.

7. Moresi, M. (1984). Industrie delle Bevande, 13, 273. 8. Ghose, T. K. and Tyagi, R. D. (1979). Biotechnol. Bioeng., 21, 1401. 9. Boulton, R. (1980). Am. J. Enol. Vitic., 31,40.

10. Jones, R. P., Pammet, N. and Greenfield, P. F. (1981). Process Biochemistry (April/May), 42.

11. Castor, J. G. B. and Archer, T. E. (1956). Am. J. Enol. Vitic., 7, 19. 12. Bailey, J. E. and Ollis, D. F. (1977). Biochemical Engineering

Fundamentals, McGraw-Hill Kogakusha Ltd., Tokyo, p. 338. 13. Nathan, L. (1930). J. Inst. Brew., 36,549. 14. Bishop, L. R. (1970). J. Inst. Brew., 76, 172. 15. Ricketts, R. W. (1971). In: Modem Brewing Technology, W. P. K.

Findlay (Ed.), Macmillian Press, London, p. 83. 16. Martin, S. (1975). In: Eur. Brew. Convention Proceedings of the 15th

Congress, Nice, p. 301. 17. Schalk, H. A. (1906). New York Eng. Pat. No. 13,915. 18. Stewart, G. G. and Russell, I. (1985). In: Comprehensive Biotechnology.

Vol. 3. The Practice of Biotechnology: Current Commodity Products, M. Moo-Young, H. W. Blanch, S. Drew and D. I. C. Wang (Eds), Pergamon Press, Oxford, p. 335.

19. Gencer, M. A. and Mutharasan, R. (1981). In: Advances in Biotechnology. Vol. 1. Scientific and Engineering Principles, M. Moo-Young, C. W. Robinson and C. Vezina (Eds), Pergamon Press, Toronto, p. 627.

20 .. Shieh, W. K. (1981). Biotechnol. Bioeng., 23,2147. 21. Satterfield, C. N. (1970). Mass Transfer in Heterogeneous Catalysis, MIT

Press, Cambridge, Massachusetts, USA. 22. Mavituna, F. and Sinclair, C. G. (1975). In: Proc. lst Eur. Congr.

Biotechnol., Dechema, Frankfurt, FRG, Part 1, 2/182-5. 23. Pitcher, W. H. Jr (1978). In: Advances in Biochemical Engineering. Vol.

10, T. K. Ghose, A. Fiechter and N. Blakebrough (Eds), SpringerVerlag, Berlin, p. 1.

24. Cheetham, P. S. J. (1980). In: Topics in Enzyme and Fermentation Biotechnology. Vol. 4, A. Wiseman (Ed.), Ellis Horwood Ltd, Chichester, p. 189.

25. Cantarelli, C. and Baccioni, L. (1984). Industrie Bevande, 7, 7.

Chapter 9

OPTIMAL FERMENTER DESIGN FOR WHITE WINE PRODUCTION

MAURO MORESI

I.M. T.A.F., University of Basilicata, Potenza, Italy

NOTATION

A Acetic acid concentration (mol/litre) AH Heat transfer surface of the fermenter jacket (m2)

Ap Heat transfer surface of each plate (m2)

Aij Activity constant of a two-suffix van Laar binary equation representing the limiting value of log Yi as its composition in the binary mixture approaches zero (Xi = 0) (dimensionless)

Cc Specific heat of carbon dioxide (J kg-1 K- 1)

CEv Specific heat of ethanol in the vapour phase (J kg- 1 K-1)

CM Specific heat of raw must (J kg-1 K-1)

Cp Specific heat of fermenting liquor (J kg- 1 K- 1)

Cwv Specific heat of steam (J kg- 1 K-1)

C Dissolved CO2 concentration in the fermenting liquor (mol/litre)

C* Saturation CO2 concentration in the fermenting liquor (mol/litre)

CEF f.o.b. cost of each plate heat-exchanger for fermentation temperature control (MLit)

CEM f.o.b. cost of the plate-heat exchanger used to pre-cool the raw must (MLit)

CF f.o.b. cost of each VF-m3 fermentation vessel (MLit) C, Overall investment costs of the fermentation unit (MLit) Co Operating costs of the fermentation unit (Lit/h) Cp f.o.b. cost of a generic pump (MLit) CR Investment costs of the refrigeration system (MLit) Cv Specific wine production costs (Lit/litre)

107


108 MAURO MORES}

Dr< Diameter of the fermentation vessel (m) E Ethanol concentration (mol/litre) Fr Temperature-difference correction factor (dimensionless) G Glycerol concentration (mol/litre) HF Fermenter height (m) HH Height of the fermenter jacket (m) H; Molar ratio between the ith component and carbon dioxide

(dimensionless) K Solubility parameter for non-electrolytes, as defined by eqn

(33) (litre/g) Ko Cell-death rate constant (mol h/litre) KE Non-competitive ethanol inhibition constant (mol/litre) Ks Saturation constant (mol/litre) KSI Substrate inhibition constant (mol/litre) N Ammoniacal nitrogen concentration (mol/litre) NE Overall number of external centrifugal pumps and heat

exchangers ( dimensionless) NF Overall number of VF-m3 fermenters (dimensionless) NR Electric power of the refrigeration system (kW) Pc Partial pressure of CO2 at half fermenter height (atm) Psi Vapour pressure of the generic ith component (atm) P Overall pressure on the top fermentation vessel, (kPa) qR Instantaneous refrigeration flow rate (kW) QM Maximum flow rate of raw must treated (m3/h or m3/day) Qp Winery production per vintage (hI) QR Rate of heat flow required to keep the fermentation

temperature constant (kW) QRF Overall rate of heat flow required to keep the fermentation

temperature of all fermenters within given limits (kW) QRM Rate of heat flow required to pre-cool the raw must (kW) QRT Overall refrigeration load (kW) 'i Consumption or formation rate of the ith component

(mol/litre h) 'H Fermentation heat rate (kJ /mol substrate consumed) S Substrate concentration (mol/litre) I Fermentation time (h) Ilag Duration of the induction phase (h) T Fermentation temperature eC) TH Temperature of liquor leaving the heat exchanger eC)

OPTIMAL FERMENTER DESIGN FOR WHITE WINE PRODUCTION 109

TMo TR Ts Twh Tw2 UD

VM

Vp

Vo W Xi

X Yi

z, Zo

~p ~p

~R

P T

Subscripts

Initial temperature of raw grape must eC) Reference temperature eC) Threshold temperature eC) Inlet and outlet temperatures of cooler fluid eC) Design overall heat transfer coefficient (W m-2 K- 1)

Maximum specific cell growth rate (h -1) Working volume of each fermentation vessel (m3)

CO2 molar volume at O°C and 1 atm (dm3/mol) Concentration of water (mol/litre) Molar fraction of the generic i component in the liquid phase (dimensionless) Biomass concentration (mol/litre) Molar fraction of the generic i component in the gas phase (dimensionless) Yield coefficient (mol ith component/mol substrate consumed) Bunsen coefficients of CO2 and water (dimensionless)

Activity coefficient of the ith component (dimensionless) Coefficient of reaction heat (kJ/mol biomass formed) Overall duration of the fermentation cycle (h) Duration of the ith phase of the fermentation process (h) Total module factor of a generic plate heat exchanger (dimensionless) Total module factor of a generic fermenter (dimensionless) Total module factor of a generic centrifugal pump (dimensionless) Total module factor of the refrigeration system (dimensionless) Latent heat of vapourisation of the generic i component at the reference temperature TR (J/mol) Liquor density (kg/m3)

Time required to saturate the liquor with CO2 (h)

A Acetic acid c Refers to the closed system ca Refers to the filling operation C Carbon dioxide

110 MAURO MORESI

E Ethanol F Refers to the fermentation phase G Glycerol M Grape must or referred to the maximum value N Ammoniacal nitrogen o Initial or refers to the open system R Refers to the out-vessel cooling phase sa Refers to the cleaning operation sv Refers to the emptying operation S a-tJ-D-glucosio X Cell biomass

INTRODUCTION

The overall pattern of grape must fermentation is the combined result of a large number of factors, such as must components (i.e. sugar, nutrient and growth factor concentrations, acidity and phenolic content), microftora and operating conditions (viz. temperature, oxygen level and agitation).

When the initial concentrations of both reducing sugars and essential nutrients in the grape must and the inoculum size of viable, selected yeast strains are kept constant, the evolution of the fermentation process may be somewhat standardised, provided that the fermentation temperature is maintained within more (± 1°C) or less (±5-1O°C) limited margins by using continuous or batch cooling systems.

Nevertheless, during white wine fermentation the mode of temperature control affects in a rather complex manner not only the wine composition and organoleptic properties, but also the overall operating costs of the fermentation unit. In particular, the lower the fermentation temperature the longer the fermentation process becomes, thus asking for a greater number of fermenters to be installed so as to ensure the expected wine production per vintage.

According to Cantarelli and Baccioni, l the fermentation temperature during wine fermentation can be controlled by using outside- or inside-vessel cooling, thus resulting in the so-called open or closed fermentation systems (see Fig. 1).

In previous works the stoichiometrr,3 and kinetics4 of white wine fermentation were assessed and used to construct a mathematical

OPTIMAL PERMENTER DESIGN POR WHITE WINE PRODUCTION 111

(al

VR-I

(bl

~--'-----------------T----------------~

L VI_1&B F-I TCIH

I L

r I

_J V1 - 1A

FIG. 1. P & I diagram for white wine fermentation units operating as a closed (A) or an open (B) system. Equipment and instrument identification symbols: E - plate-heat exchanger; F - fermenter; G - centrifugal pump; M - electric motor; QC - central board; TCI - fermentation temperature controller and indicator with alarm for minimum (L) and maximum (H) values; VR - servomotor operated regulation valve; VI - two-way solenoid valve;

-+- pneumatic transmission; -fIf-- electrical transmission.

model of white wine fermenters equipped with in- or out-cooling systems and operating batchwise. Then, a strategy for designing such fermentation units was developed and used to optimise the working volume of each fermenter and cooling system, the overall number of fermenters and so on as a function of winery production. 5

The main aim of this work is to review the principal aspects of the above mathematical model and its application to fermentation control and operation.

112 MAURO MORESI

MATHEMATICAL MODEL OF BATCH WlDTE WINE FERMENTERS

To simulate the evolution of white wine fermentation in cylindrical or cylindroconical vessels, a knowledge of the biological reaction and liquid mixing is needed. In another chapter of this volume,6 white wine fermentation stoichiometry and kinetics are fully described and will be used in the following modelling exercise.

As far as the reactor mixing is concerned, it may be observed that mechanical mixing is usually more efficacious to avoid layering and disperse yeast than the agitation due to the gas-lift action of carbon dioxide bubbles. However, the degree of micromixing in the bioreactor increases with the fermenter aspect ratio (that is, the fermenter height-to-diameter ratio). Moreover, similar results are due to either the convection currents thermally induced during in-vessel cooling or the mechanical recirculation of the fermenting must during out-vessel cooling.

In the circumstances, the liquid phase in the above vessels may, for the sake of simplicity, be represented by the completely mixed model. Thus, the unsteady-state material and energy balances for the fermentation reaction (eqn (1» of Ref. 6 can be written as follows:

dS/dt=rs (1) dN/dt= rN (2) dX/dt=rx (3) dE/dt=rE-qcHE/Vp (4) dG/dt= rG (5) M~=~ 00 dW/dt=rw-qcHw/Vp (7) dC/dt=rc forC~C* (8)

act at = (dC* /dT)(dT /dt) = rc - qc/Vp, for C(T, t) = C*[T(t)] (9)

pCpdT/dt = rH - qR/Vp- qc/Vp{cc(T- TR) + HE[AER + cEv(T- TR)] + Hw[AWR + cwv(T- TR)]} (10)

to be integrated with the following initial conditions:

S = So; N = No; X = Xo; E = Eo; G = Go; A = Ao; W= Wo; C=O; T = To fort =0

(11)

OPTIMAL FERMENTER DESIGN FOR WHITE WINE PRODucrION 113

Since the above rates rs, rN, rE, rG, rA, re and rw were found to be functions of rx only, 6 the above system of differential equations can be simplified as follows:

for t:::;; tlag

for t > tlag

(12)

dE/dt= YErx/Yx - qeHE/VF (13)

dW /dt= Ywrx/Yx - qeHw/VF (14)

C = YdX - Xo)/Yx for C:::;; C* (15)

ac/at= (dC*/dt)(dT/dt)

= Yerx/Yx - qe/VF for C(T, t) = C*[T(t)] (16)

dT /dt = AHRrX/(pcp) - qR/(VFPCp) - qc/(VFPcp){ cdT - TR)

+ HE[)'ER + cEiT - TR)] + Hw[AwR + cwiT - TR)]} (17)

S = So - (X - Xo)/Yx (18)

N = No - 0·2(X - Xo) (19)

G = Go + YG(X - Xo)/Yx (20)

A = Ao + YA(X - Xo)/Yx (21)

with YG = 2(Yc - YE) - 0·125Yx (22)

YA = 3 - 0·3126Yx - 3·5Ye + 2YE (23)

Yw = 0·55Yx - Ye + YE (24)

AHR = (194·76 + 2·48Yx + 193·37YE - 240·85Yc)/Yx (25)

to be integrated with the following initial conditions:

X = Xo; E = Eo; W = Wo; T = To for t = 0 (26)

where the stoichiometric and kinetic parameters (Yx , YE , Ye , VM, K E ,

Ks, KSI and Kn) are given in Table 1 of Ref. 6. Then, the above system of differential equations can finally be

numerically integrated once assigned qR(t) and qdt).

114 MAURO MORESI

TABLE 1 INSTANTANEOUS REFRIGERATION FLOW RATE (qR) AS A FUNCTION OF THE

COOLING SYSTEM USED TO CONTROL THE FERMENTATION TEMPERATURE

(ALL THE SYMBOLS ARE GIVEN IN THE NOTATION)

Type of fermenter

Isolated vessel: -With external heat

exchanger

Unisolated vessel using: -External-wall air-cooling -External-wall falling-film

cooling -Jackets or coils

Instantaneous Refrigeration Flow Rate This parameter depends upon the type of cooling system used, as shown in Table 1. In particular, the heat flow to be instantaneously removed to perform an isothermal process represents the contribution of each single fermenter to the refrigeration load of the cooling system:

QR = VF !J.HRrx - qd cdT - TR)

+ HE[AER + cEv(T - TR )] + Hw[AwR + cwv(T - TR )]} (27)

So, in a closed vessel isothermal ethanol fermentation may be carried out by varying the flow rate or input temperature (Tw1) of the colder fluid flowing inside the tank jacket or coils or falling along the fermenter walls in order to obtain:

(28)

In isolated vessels, adiabatic and anergic ethanol fermentation may be allowed to occur as long as the liquor temperature is lower than a threshold value (1s). Then, the fermenting must is circulated through an external heat exchanger to be cooled. When the mean temperature is newly To, the external circulation of liquor is stopped and the fermentation goes on adiabatically until T ~ 18; otherwise, another refrigeration phase is performed in the same manner mentioned before.


Instantaneous Flow Rate of Carbon Dioxide This flow is evolved during the fermentation and depends upon the physico-chemical conditions of the fermenting must. In this work, it was simply assumed that:

qc = 0 for t ~ l' (29)

qc = VFYcrx/Yx - (dC* /dT)(dT /dt) for t > l' (30)

where l' is the time required to saturate the fermenting must with carbon dioxide. The equilibrium concentration of COz(C*) in grape must, which is a complex function of temperature and electrolyte and organic compound (excluding short-chain alcohols) concentrations, decreases with respect to that in pure water as the above parameters increase. By neglecting either the salting-out effect of sugars, organic acids and salts or the salting-in effect of ethanol and methanol on the assumption of their self-compensation effects in grape must, C* was estimated by using the prediction method developed by Schumpe et al. :7

C* = zPc/Vo

z = zo10-KS

K = 0·1345 - 0·001 T

Zo = 1·72 - 6·69 x 1O-2T + 1·62 x 1O-3T 2

+ -2·28 X 1O-5T 3 + 1·39 X 1O-7T4

(31)

(32)

(33)

(34)

where Pc is the partial pressure of CO2 , Vo ( = 22·258 dm3/mol) is the CO2 molar volume at OQC and atmospheric pressure, Zo and z the Bunsen coefficients for pure water and fermenting must.

The evolution of CO2 strips the fermenting must of a certain amount of volatile metabolites according to their molar fractions and activity coefficients in the liquor. In this case, ethanol seems to be the only volatile component to be significantly transferred from the liquid into the gas evolved. By assimilating the fermenting must to a water-ethanol mixture, the molar ratios between the molar fractions of ethanol or water and CO2 in the gaseous effluent from the fermenter can be estimated as follows:

HE = YE/YC

Hw= Yw/Yc

(35)

(36)

116 MAURO MORESI

with Yc = 1 - YE - Yw (37) Yi = YiXiPsJ P (38)

log YE = AEw/[l + AEWXE/(AWEXW)]2 (39) log Yw =AWE/[l + AWEXW/(AEWXE>Y (40)

PsE = exp(19·729 - 4243·0S/(T + 24S·S4» (41)

Psw = exp(18·613 - 4004·0/(T + 234·24» (42)

where Yi, Psi' Yi and Xi are respectively the activity coefficient, vapour pressure and molar fractions in the gas and liquid phases for the generic ith component, while AEW (= 0·67) and AWE (= 0·42) are the activity constants of the Van Laar equation at 2SoC, as extracted from Perry et al. 8

OPTIMISATION PROCEDURE

The design of a winery of working capacity Qp requires several environmental (viz. period of grape ripening and harvesting, mean temperatures of ambient, Ta, and raw grape must, TMo, as well as raw must composition) and design parameters (that is, configuration and geometric dimensions of fermenters, type of coolers and their overall heat transfer surface) and operating variables (initial composition of raw must, fermentation temperature, 1'0, and the mode of temperature control) to be defined.

More specifically, the design of the .fermentation unit for a given production of white wine may be optimised as follows:

(1) Simulate the evolution of the fermentation process by solving numerically the aforementioned system of differential equations.

(2) Determine the fermentation time (AtF) corresponding to 8S% utilisation of the initial concentration of reducing sugars.

(3) Estimate the overall duration of each fermentation cycle (Ate) by taking account of filling (Atea = l·S h), emptying (Atsv = l·S h) and cleaning (Atsa = 1 h) times:

Ate = AtF + Atea + Atsv + Atsa (43)

(4) Evaluate the number of fermenters of a given capacity VF

required to fulfill winery production:

NF = QM I1tjVF (44)

OPTIMAL FERMENTER DESIGN FOR WHITE WINE PRODUCfION 117

where QM is the maximum flow rate of grape must entering the fermentation unit during vintage.

(5) Calculate the amount of heat flow (QRM) required to pre-cool the raw must from TMo to To:

(45)

(6) Estimate the maximum refrigeration load (QRF) of the fermentation unit by combining linearly the heat flow rate of each fermenter undergoing fermentation. When out-vessel cooling is applied, the refrigeration system has to lower the temperature of the liquors in all the fermentation vessels at T ~ Ts to 10. So, to avoid mixing the contents of such fermenters, it is necessary to install a number of plate-heat exchangers at least equal to the maximum number of fermenters requiring out-vessel cooling.

(7) Rate the refrigeration system with a 10% overdesign factor as follows:

(46)

(8) Design each heat transfer equipment in order to cool the liquor in each vessel requiring out-cooling from Ts to To in a given period of time (~tR)' while the liquor flowing through such an exchanger is cooled no more than 5°C per passage to avoid thermal shock to yeast.

(9) Evaluate the investmettt costs of the fermentation unit (which consists of NF vessels, (NE + 1) plate-heat exchangers, NE centrifugal pumps and a mechanical vapour-compression refrigeration system), by taking into account both the bare equipment costs and auxiliary costs (instruments, piping and valves, insulation, civil work, electrical, installation, etc.). By using Guthrie's concept9 of total module factor (;i), which represents the contribution of all direct and indirect costs in the bare process module plus the contingencies necessary to adjust for unlisted items or insufficient design definition, as well as for contractor fees, the total cost of each item of the fermentation unit can be roughly evaluated on the basis of free on board (f.o.b.) equipment costs, thus yielding the following overall cost (C)) of the fermentation unit.

C) = ~FNFCF + ~pNECp + ~E(NECE + CEM) + ~RCR (47)

where CF, Cp, CE, CEM and CR are, respectively, the f.o.b.

118 MAURO MORESI

costs of any fermenter, centrifugal pump, plate-heat exchangers for fermenting liquor and raw grape must and refrigeration system. All the aforementioned costs were calculated via the correlation reported in Table 2 of Ref. 5.

(10) Evaluate the operating costs (Co) of the fermentation unit by taking into account the following items: investment-related, utility and selected yeast costs. The first item includes depreciation and maintenance. Whereas the latter includes ordinary maintenance (estimable as a percentage of about 3% of CI ), the former can be estimated over a 5- or lO-year period for the process machinery or fabricated equipment at an interest rate of 15%.

TABLE 2 INPUT DATA REQUIRED FOR WHITE WINE FERMENTATION UNIT DESIGN

Parameter Value Unit

Raw must -Flow Rate (QM) (1/60-1/20)Qp m3/day -Temperature (TMO) 25 °C -Sugar Concentration (~) 220 g/litre -Nitrogen Concentration (No) 1·2 g/litre

Fermentation conditions -Inoculum (X~) 0·2 g/litre -To 15 °C -1; 22 °C -TR 25 °C -L1tR 4 h -P 103 kPa

Fermentation units

-VF 10-110 m3

-lJp 2·3-3·8 m -HF 2·9-10·4 m -HH 2-7·3 m -Ap 0·18 m2

Cost factors

-~F,c 2·00 -~F,O 2·13 -~E 1·78 -~p 1·78 -~R 2·13


The utility costs refer to the electric power consumed to pilot the refrigeration system and all the centrifugal pumps and were evaluated on the basis of Litt 120/kWh. The inoculation with selected yeast strains costed c. Lit 5000 per kg dried yeast. Finally, the labour costs w~re evaluated according to the mode of temperature control. When the fermentation unit was operated as a closed system, the manpower employed consisted of two seasonal workers for the day-shifts at Lit 30000 per hour and of a single worker for the night-shift at Lit 50 000 per hour. In the other system, the manpower employed was doubled. Therefore, the specific wine production costs (Cv )

are: (48)

(11) Determine the operating conditions associated with the minimum specific wine production costs (Cv ) as a function of the vessel capacity (VF ) and the mode of temperature control, by repeating the procedure from step 1.

OPTIMAL DESIGN

The optimisation procedure outlined in the previous section has been applied to design three typical wineries of small, medium and large size with working capacities of 30000, 60 000 and 120000 hI of wine during a typical grape harvest period of 28 days with incoming flow rates of raw must varying day by day as plotted in Fig. 2.

Qp/20

:u Qp/30 ",-

E

:I Qp/60 o

o 4

I 8 12 16 20 24 28 Working period (days)

FIG. 2. Typical diagram of daily flow rates of raw must submitted to wine fermentation as a function of the grape harvest period and winery production

(Qp) per vintage. t Lit = Italian Lira.

120 MAURO MORESI

All the input data required for this study are summarised in Table 2_ More specifically, the composition of raw must submitted to fermentation was chosen as suggested by Cantarelli et al_, 3 so as to maximise grape sugars conversion into ethanol with minimum formation of glycerol and acetic acid_ The geometric dimensions of cylindrical vessels with or without jackets were reported in Table 3 of

TABLE 3 MAIN DESIGN PARAMETERS, INVESTMENT AND OPERATING COSTS OF THE

OPTIMAL WHITE WINE FERMENTATION UNITS OPERATING AS CLOSED OR

OPEN SYSTEMS FOR Qp SMALLER OR GREATER THAN 30000 hi PER VINTAGE

Parameter Winery capacity per vintage, Qp(hl)

30000 60000 120000

Cooling system Closed Open Open VF (m3) 110 40 30 AtF (h) 81 79 78 Atc (h) 85 83 82 NF (-) 5 26 69 QRT (kW) 288 517 1000 NR (kW) 96 172 334

Equipment costs -NFCF (MLit) 160-4 221-6 478-3 -NECp (MLit) 1-5 3-2 -CEM (MLit) 3-1 4-6 7-7 -NECEF (MLit) 15-4 33-2 -NE (-) 4 10 -CR (MLit) 117-8 196-1 364 Cr (MLit) 577-3 927-8 1872-4

Operating costs -Depreciation

Fabricated Equipment (MLit) 65-0 101-1 217-4 Machinery (MLit) 74-9 125-4 233

-Maintenance (MLit) 17-3 27-8 56-2 -Electric Energy (MLit) 7-7 18-4 26-9 -Inoculum (MLit) 3 6 12 -Manpower (MLit) 38-1 76-2 76-2

Co (MLit) 206-0 354-9 621-7

Cv (Lit/litre) 68-7 59-1 51-8


Ref. ~, whereas the plate-heat exchangers used to control the fermentation temperature were equipped with plates of 0·18 m2

effective surface area. 10

Finally, the cost correlations used to estimate the f.o.b. costs of process machinery and fabricated equipment, as reported in Table 2 of Ref. 5, were updated by introducing the most recent values actually available from the Chemical Engineering Plant Cost Indexes (July 87).11

As an example, Figs 3 and 4 show the evolution of white wine fermentation when in- or out-vessel cooling is applied. Under the hypotheses assumed, the maximum substrate consumption and ethanol and heat production rates in the closed fermentation system are observed about 60 h after the inoculation (see Fig. 3(a», while CO2

bubble rising strips the maximum amount of ethanol from the liquor a little later, that is at t = 68 h (see Fig. 3(b».

When the open fermentation system is used, the variation of the fermenting liquor temperature within 15°C and 22°C, especially during the refrigeration phase, has a strong effect not only upon the sugar consumption rate, but also upon the molar fraction of ethanol (YE) in the gaseous stream leaving the fermenter (see Fig. 4(b».

Finally, the dashed areas in Figs 3( a) and 4( a) are proportional to the heat evolved during the fermentation process and show the duration of each refrigeration phase. When in-cooling is used, such a phase lasts as long as the fermentation process itself. In the open fermentation system, each refrigeration phase coincides with the external Circulation of the fermenting liquor through a plate-heat exchanger.

In order to minimise the operating costs of each fermentation unit to be realised, several calculations based on the aforementioned design strategy were carried out by varying either the working voiume of each fermenter (Vp) or the refrigeration system.

Figure 5 shows the specific wine production costs (Cv ) as a function of the working volume of each fermenter (Vp ) and winery production when in- (continuous lines) or out-vessel (broken lines) cooling was used. It is worth noting that the operating costs of the out-cooling system are lower than those of the in-cooling one for medium and large size wineries only.

As far as the latter system is concerned, the specific wine production costs (Cv) monotonically decrease as Vp increases for all winery size and fermenter capacities examined. However, it is unlikely that such a

122 MAURO MORESI

(a) T = 15°C

ok--L ____ ~~--~--+_--~--~~--~O

o 20 40 60 80 Time (h)

FIG. 3. Typical evolution of white wine fermentation at 15°C in a closed system consisting of a 50-m3 jacketed vessel, as predicted by eqns (12-42). (a) Concentrations of reducing sugars (S), ammoniacal nitrogen (N), cell biomass (X), ethanol (E), glycerol (G) and acetic acid (A) and fermentation heat flow rate (QR) against fermentation time (1). (b) Molar flow rates of carbon dioxide (qc) and ethanol (qCHE) and molar fraction of ethanol CYd against

fermentation time.

trend can be maintained for VF much larger than 110 m3. In fact, as VF

increases the heat transfer surface of the vessel may become insufficient for an effective control of wine fermentation temperature, unless the temperature of the colder fluid is appropriately reduced, thus increasing further the refrigeration costs.

In the open fermentation system, the optimal operating conditions are affected by NF and QRF, both parameters increasing as VF


(a)

o 10 20 30 40 50 60 70 80 Time (h)

FIG. 4. Typical evolution of white wine fermentation in an open system consisting of a 50-m3 unisolated vessel and a plate heat-exchanger to keep the fermentation temperature within 1YC and 22°C, as predicted by eqns (12-42). (a) Concentrations of reducing sugars (S), ammoniacal nitrogen (N), cell biomass (X), ethanol (E) and fermentation heat flow rate (QR) against fermentation time (t). (b) Molar flow rates of carbon dioxide (qc) and ethanol

(qCHE) and molar fraction of ethanol (YE) against fermentation time.

decreases. In this specific case, the optimal working volume of each fermenter varies in the range 30-40 m3 (Fig. 5).

Table 3 shows the main design parameters and the investment and operating costs for three optimal white wine fermentation units operating as closed or open systems for Q p smaller or greater than 30000 hi per vintage.

124

a; ~ .... ::i

> u

150

100

50

MAURO MORESI

,

, , , ,

, ,

, ,

Qp(hl)

,///~30000

" ",""

"~--::--,; ,--'~'--~30000 ~' ,-- 60000

--- 120000

O~~~~~~~~~ ______ --. o 50 100

vF (m3)

FIG. 5. Wine production costs (Cv ) against working volume (VF ) of each fermenter at different winery capacity (Qp) when in-(continuous lines) or

out-vessel (broken lines) cooling is used.

To verify the stability of Cv with respect to several parameters, such as inoculum size (X~), threshold value of temperature (Ts) and duration of the refrigeration phase (~tR)' a sensitivity analysis was also carried out. For the medium size winery, an increase in the inoculum size from 0·2 to 0·5 g/litre would involve a 4·8% reduction in the production costs, whereas the standard values of Ts and ~tR were found to be optimal.

CONCLUSIONS

Following a series of previous works2- 5 on the stoichiometry, kinetics, and reactor modelling of white wine fermentation, a design strategy for white wine fermentation units was established so as to choose the structure and temperature control system associated with minimum wine production costs.

In particular, the above costs can be minimised by installing jacketed fermentation vessels of about 110 m3 or isolated vessels of appropriate working volumes equipped with external plate-heat


exchangers for winery outputs respectively smaller or greater than 60,000 hI per vintage, once the initial composition of raw must, the amount of selected yeast strains inoculated, the range of variation of the fermentation temperature and the duration of the refrigeration phase are best chosen.

In conclusion, it is worthwhile pointing out that the mathematical model here described can be used not only to design and optimise a new fermentation unit, but also to calculate the suitability of an existing winery for new process conditions.

REFERENCES

1. Cantarelli, C. and Bacc.ioni, L. (1984). Recipienti per la fermentazione termocondizionata: i criteri di scelta fra i diversi sistemi di refrigerazione e di riscaldamento. Industrie delle Bevande, 7,7.

2. Cantarelli, C. and Moresi, M. (1983). Atti Accademia Ital. Vite e Vino, 35, 225.

3. Cantarelli, c., Moresi, M. and Manfredini, M. (1984). In: Proceedings of the 7th Int. Oenological Symp., Rome (Italy), 7-9 May 1984, E. Lemperle and H. Rasenberger (Eds) , Edition of the Int. Association for Modern Winery Technology and Management, Breisach, West Germany, p. 317.

4. Moresi, M. (1984). Industrie delle Bevande, 13, 273. 5. Moresi, M. (1985). Industrie delle Bevande, 14,7. 6. Moresi, M. (1989). Fermenter design for alcoholic beverage production.

In: Biotechnology Applications in Beverage Production, C. Cantarelli and G. Lanzarini (Eds), Elsevier Applied Science, London, pp. 93-106.

7. Schumpe, A., Quicker, G. and Deckwer, W. D. (1982). In: Advances in Biochemical Engineering, Vol. 24, A. Fiechter (Ed.), Springer-Verlag, Berlin, p. 1.

8. Perry, R. H., Chilton, C. H. and Kirkpatrick, S. D. (1963). Chemical Engineers' Handbook, 4th Edn, McGraw-Hill Book Co., New York.

9. Guthrie, K. M. (1969). Chem. Eng., 76, 114. 10. Anon. (1969). Thermal Handbook, Alfa-Laval, Vasteras, Sweden. 11. Anon. (1987). Chem. Eng., 94,7.

Chapter 10

FACTORS AFFECTING THE BEHAVIOUR OF YEAST IN WINE FERMENTATION

CORRADO CANTARELLI

Department of Food Science and Technology and Microbiology, University of Milan, Milan, Italy

NOTATION

M Yeast mortality, as the percentage of dead cells, estimated at intervals of 24-48 h by direct microscopic count (methylene blue staining). The average of all the values estimated at various time until the exhaustion of fermentation

Mmax Highest yeast mortality value and the time (h) at which this value appeared

1iag Length (h) of the lag phase of the fermentation 4sp Length (h) of the exponential phase of the fermentation V Mean fermentation velocity expressed as rate of CO2 forma

tion (gcoJlitre h) estimated gravimetrically until exhaustion of fermentation (weight loss less than 0-01 g/litre day)

V. Specific rate of CO2 formation, referred to the yeast X (dry weight) recovered at the end of fermentation (V /gx)

X Yeast concentration (mg (dry weight)/ml) X Rate of yeast formation, as the average of the yeast content

(dry weight) estimated at various times during the fermentation (mg/litre h)

YA Acetic acid (A) yield, in relation to ethanol (E) formation (gA/gE x 10-3)

YE Ethanol (E) yield, in relation to sugar (S) uptake (gE/gS) YG Glycerol (G) yield, in relation to ethanol (E) formation

(gG/ gE X 10-3)

YH Acetaldehyde (H) yield, in relation to ethanol (E) formation (gH/ gE x 10-4)

Yx Yeast biomass (X) yield, in relation to sugar (S) uptake (gx/ gs x 10-3)

127


128 CORRADO CANTARELLI

THE EVOLUTION OF WINE FERMENTATION TECHNOLOGY

Unlike other fermentation processes for the production of ethanol for beverages or industrial uses, enology presents considerable and precise constraints on the applicative transfer of increased knowledge about the metabolism of yeast.

The seasonality, size and set-up of wine production are the main limitations on the profitability of investments in fermentation equipment.

In fact, the turnover in fermentation vats is a critical factor in transformation costs because of the length of the winemaking process. Furthermore, this cost must be added to that of the energy consumed for thermal regulation. However, in recent years significant developments have been made in new fermentation plants with fermenters equipped with devices for thermoregulation and mechanisation of pomace leaching. 1

On the other hand, viewing the fermentation process as a biological activity, the only important innovation in recent years has been the spread of dry yeast starters as a result of their ease of use. Unlike in the case of beer and industrial ethanol, continuous and semicontinuous fermentation processes (listed in Table 1) have not yet been applied to wine fermentation. These processes have obtained important results, despite the well-known limitations connected with their cost, problems

TABLE 1 PROCESS INNOVATIONS FOR BEVERAGES FERMENTATION

Discontinuous, accelerated fermentation -Recycling high concentrations of flocculent yeast (by settling from conical

vats or from centrifugal separation) -Gradual feeding of the must -Mechanical stirring -Addition of dispersed inert solids

Continuous fermentation -Must feeding without mixing (heterogeneous)

Open: yeast outflowing Closed: yeast recycle, or immobilised yeast

-Must feeding with mixing Open: cascade, multistage Closed: single stage/multistage biostat, yeast kept within a membrane

fluidised bed of immobilised yeasts

THE BEHAVIOUR OF YEAST IN WINE FERMENTATION 129

in must feeding and the instability of the microftora, and the need for skilled personnel for their operation.

The so-called 'semicontinuous' wine fermenters used at present are actually an application of the debated 'superquatre' system proposed by Semichon during the 1920s.2 In fact, the benefits provided by recent 'fast' and continuous alcohol fermentation processes are actually derived from other operating conditions, such as the high concentration of resting yeast, the regulation of the level of oxygen and the stirring of the liquid.

In addition to the structural characteristics mentioned above, the transfer to wine fermentation of findings related to beer and industrial ethanol production is hampered by the features of the raw material itself. In fact, grape must presents: (a) high sugar content which cannot be diluted; (b) low pH; (c) a low concentration of nitrogen nutrients; (d) considerable phenolic content, and (e) spontaneous microftora. In addition, wine fermentation is a process which takes place once a year without yeast recycling, and the process itself may be quite long and encounter problems related to sticking. Finally, there is the fact that 'white' must usually has been treated in such a way that the final oxygen and nutrient content is quite low.

Therefore, the objectives of the development of wine fermentation technology are varied. This chapter mainly concerns 'white' fermentation without pomace and the refermentation of sparkling wine. The objectives to be reached may be summarised as follows:

1. To obtain a linear kinetics in the fermentation process, excluding the lag and exponential yeast growing phases, in order to: (a) reduce the final content of glycolytic fermentation byproducts which accumulate during the yeast reproduction phase; (b) prevent the negative effects of oxygen on the stability of the future wine and improve its freshness; (c) reduce the amount by which the temperature increases and hence the required cooling operations;

2. shorten the length of the fermentation process; 3. prevent fermentation stops and avoid the presence of residual

sugars.

This chapter examines the aspects of technological interest connected with yeast reproduction and fermentation in terms of transformation yield, byproduct accumulation and fermentation speed.


It is useful to briefly summarise current knowledge about the role of the three groups of variables in wine fermentation: (a) the characteristics of the yeast; (b) the conditions of the substrate, and (c) the operating conditions.

mE CHARACTERISTICS OF mE YEAST

Difterences Among Yeast Strains The search for an 'ideal yeast' began even before there was a definition of the nature of the microorganism. Indeed, the 'hunt' for good yeast has continued from just after Pasteur's discoveries until the present day. In terms of yeast ecology and control of physiological

TABLE 2 CRITERIA FOR WINE YEAST SELECTION: TIlE PROPERTIES REQUIRED FOR A

CHIMERICAL YEAST STRAIN

Growth and fermentation -High growth rate, high fermentative power, high tolerance to intracellular

ethanol and to carbon dioxide, low oxygen demand, ability to ferment at low temperature

Other physiological characteristics: -Flocculence, high/low settling rate, low foam production, 'good' solubles

from autolysis

Resistance to chemicals: -Sulphites, pesticides

Resistance to physical factors: -High temperatures, mechanical stress, drying

Productivity : -High alcohol yield at high/low temperatures and pH -Low production of acetic acid, acetaldehyde and other sulphite-binding

carbonyls, higher alcohols -Production of 'good' esters -Production of malate -Low production of other fixed acids

Adaptability to the substrate components: -High sugar content -Low pH -Sulphur metabolism (volatile compounds, sulphite production) -Nitrogen metabolism; requirement of essential amino acids, ammonia

utilisation -Exogenous vitamins requirements (autoauxotrophic) -Malate fermentation -Minerals tolerance


TABLE 3 COMPOSITION OF WINES ISSUED FROM THE SAME WHITE MUST FERMENTED

BY VARIOUS YEASTS AT DIFFERENT TEMPERATURES (Adapted from Cantarelli6 )

Strain

Sacch. cerevisiae Sacch. bayanus Sacch. rosei No. 20 No. 838 No. 63

Fermentation temperature ("C) 18 35 18 35 18 35

Cell count (max. x 105) 115 67 133 83 57 43 Final biomass

g/litre (dry wt) 8·40 4·95 7·35 3·83 7·95 4·07 Alcohol (ml/litre) 11·64 11·39 11·39 11·48 11·39 10·96 Acetic acid (g/litre) 0·30 0·32 0·35 0·41 0·21 0·44 Glycerol (g/Iitre) 6-42 7·68 6·80 7·60 6·60 11·50 Butanediol (g/Iitre) 0·28 0·41 0·30 0·54 0·24 0·30 Acetaldehyde (mg/litre) 52 24 51 27 16 15 Higher alcohols (mg/litre) 200 228 140 180 200 202 Changes of the fixed

acidity (as tartaric) -0·22 +1·10 -1·11 -0·36 -0·94 +0·05

Must composition: reducing sugars, 191·5 g/Iitre; total acidity, 7·65 g/Iitre. Starter: 80 g of compressed yeasts per litre (c. 70% humidity). Species names following the classification are valid at the time of the trials; numbers refer to IMAT (University of Perugia) yeast collection.

characteristics, this search intensified between the 1930s and 1960s and was affected by several revisions of the yeast classification. Development of the use of dry starters has recently focused interest on the features of the selected yeast, which may be improved by genetic engineering. A list of the desired features is presented in Table 2.3

It should be noted that the search is presently directed at identification of specific yeast strains for a given wine-growing area; surprisingly, this was also a subject of research at the turn of the century.4 In any case, the sum of the positive characteristics cannot be the heritage of a single strain because of the diversity of wine typology.

The difference in results obtained with the fermentation of various strains has been demonstrated by numerous authors.5 Table 3 presents a few examples referring to some of the more common strains and significant components of the wines obtained under conditions which assured the unquestionable predominance of each strain.6

With all conditions being the same, fermentation power and glycolysis yield present significant differences, such as alcohol tolerance, speed of sugar fermentation and byproduct accumulation


(especially acetic acid). Other differences can be detected by analysis of volatile components and sensory evaluation.7 The reason for these differences is not completely understood, except in the case of alcohol tolerance, which is considered to be the reciprocal of ethanol productivity. The threshold of alcohol tolerance, which reaches limits of 10% for cell reproduction and 20% for ethanol production, is actually influenced by a series of factors:

(a) the intracellular level of ethanol linked to the permeability of the cell membrane, which in yeast, unlike in higher plants is not free;8

(b) permeability, and thus intracellular ethanol content, is affected by the availability of membrane lipids in addition to osmotic effects;9

(c) the age of the cell, whose internal volume decreases over time;lO (d) immobilisation (in polysaccharides), which increases the alcohol

tolerance. 11

Proliferation and Fermentation Because of the high sugar and acid concentration of the must, wine yeast does not demonstrate the 'Pasteur effect'. Even when exposed to high oxygen levels, diauxia does not take place when ethanol is utilised, except in the case of Sherry 'ftor' yeast. Fermentation prevails over respiratory glycolysis given the condition of the substrate and the physiological features of the yeast.

On the other hand, the yeast's ability to reproduce conditions fermentation activity: the biomass concentration determines the overall speed of sugar transformation, but the yields are negatively correlated with the yeast's reproductive capacity.

During the lag and exponential growth phases, sugar transformation yields are radically different from those obtained during the stationary phase, in which yields approaching theoretical levels are obtained.

Acetaldehyde, acetic acid, glycerol, non-volatile acids and other catabolic products of amino acid metabolism are produced during the first phase of the fermentation process in which the biomass is formed. 12

Starters: .Fermenting and Dried Yeast In addition to their ease of application and the certain predominance of the selected yeast, the increased use of dry starters appears to be justified by their increased fermentation capacity when compared to liquid fermenting starters. Some data (presented in Table 4) obtained


TABLE 4 ACfIVITY OF DRIED AND LIQUID FERMENTING STARTERS. MEAN OF VALUES

FROM COMPARATIVE ASSAYS ON 12 STRAINS OF Sacch. cerevisiae FOR

OENOLOGICAL USE ON WHITE AND RED MUST FERMENTATION (Adapted from Cantarelli and Aresani34)

Must White Red"

Starter Dry Fermenting Dry Fermenting

Mean velocity (if) 3·09 2·25 6·65 5·08 Specific velocity (Y,.) 1·91 1·54 2·01 1·89 Biomass produced

(g/litre) 1·68 1·53 3·25 2·72 Fermentation length

(days) 24 33 11 13 ------

a Phenolics from red grapes added (8 g/litre). Dry yeasts and liquid fermenting cultures of 24 h on the same musts inoculated at the same cell number (about 7 x 106 ml- 1).

Must: sugars, 153·7 g/litre; total nitrogen, 0·58 g/litre.

from experiments conducted on a series of wine strains clearly demonstrate the increased activity of the inoculum with dry yeast in comparison to a pied-de-cuve of the same strain at the same concentration. The dry yeast deriving from strongly aerated cultures is much richer in sterols, unsaturated fatty acids and trealose, thereby improving the yeast's ability to start fermentation.

The membrane lipid content explains higher fermentation capacity in terms of alcohol tolerance (because of intracellular ethanol excretion) and survival under anaerobiosis.

The possibility of increasing fermentation speed and yield by raising the yeast cell concentration was advanced at the end of the last century by Delbruck.13 This concept was implicit in numerous instruments proposed for accelerated and continuous fermentation, such as stirred and tower fermenters. In terms of the latter system, Hough et at. 14

maintained that the advantage would be suppression of the respiratory phase. In practice, this condition occurs when the biomass deriving from a propagator is transferred to fermenters in which carbon dioxide creates a condition of anaerobiosis. 15

Another original attempt to exclude the respiratory phase of the yeast has been proposed based on the utilisation of a respiratorydeficient mutant, known as 'petite-colonie', as the starter. This mutant


lacks cytochrome oxidase, and hence this yeast may be responsible for the excessive concentration of diketones. 16

Immobilised Yeast Immobilisation tends to split cell growth from the formation of products, thereby maintaining a high capacity for substrate conversion with low proliferation. This split may justify the additional costs of immobilisation.

In the case of wine fermentation, there are two basic objectives:

(a) to facilitate yeast settling in the production of bottled sparkling wine (see Chapter 11);

(b) to produce fermentation with high concentrations of yeast in the steady-state, under anaerobiosis and without addition of yeast activators.

The latter application is interesting for both the production of so-called 'base' wine (white) with low byproduct content and the production of sparkling wines processed in tanks. Immobilised yeast may be used as a fluidised bed or a 'reactor' in order to prevent fermentation stops, aeration, the addition of nutrients and fining agents, and biomass separation at the end of the process.

Experiments conducted on yeast immobilised in alginate, as opposed to free cells, demonstrate limitations on cell growth, survival with lower energy expenditure, higher tolerance for ethanol, high sugar content and heavy metals. 17

Trials on pilot and semi-industrial plants utilising yeasts immobilised in alginate (other than in-bottle, fermented Champagne) suspended in cylindrical fermenters (as fluidised bed or held in a 'reactor') provided positive results in terms of linear kinetics, byproduct content, organoleptic characteristics and retention of fermentation capacity during subsequent recycling. 18

SUBSTRATE CONDmONS

The substrate variables which condition fermentation in winemaking may briefly be summarised as the concentration in the substrate of fermentable sugars, the availability of nutrients and other solutes such as phenolics and minerals, the pH and the level of oxygen.

Sugar Content The osmotic effects produced by must solutes are related to alcohol tolerance. Many experimental observations on this subject demon-


strate that a low threshold exists beyond which the so-called 'Crabtree effect' occurs, and that another, higher threshold also exists beyond which fermentation yield, in terms of ethanol production, worsens. 19

The effect of a critical concentration of sugar was demonstrated by Muller Thurgau's early experiments. Gray20 observed plasmolysis and loss of fermentative capacity in yeast at sugar levels over 14%. High osmotic pressure and intracellular ethanol concentration act as synergical inhibitors on the yeast's enzymatic activity, according to Strehaiano and Goma and coworkers. 21

The transport of fermentable sugars changes during fermentation; while active and mediated by transferase in the exponential phase, it becomes purely diffusive during the subsequent stationary phase. Alcohol tolerance, and hence the final production of ethanol, increases as the sugar concentration gradually increases. Immobilised yeast demonstrates a higher resistance to elevated osmotic gradients. 21

Growth Factors and Activators The so-called 'activation' of fermentation is a traditional practice in enology and is obtained with the use of various additives which cause, or should cause, the increased proliferation of yeast and/or more active glycolysis. As we know, different forms of assimilable nitrogen (ammonia, amides, a-amino acids), thiamine (niacin and pantothenate to a lesser degree) and phosphates are added. These additives are currently employed in 'difficult fermentation', with or without yeast starters. They are used as a single chemical or yeast extract, which corresponds to a natural pool of these factors, including purines. Comparative experiments demonstrate that the acceleration of fermentation is essentially promoted by an increase in yeast reproduction obtainable with the addition of assimilable nitrogen. It has also been demonstrated22 that yeast with a large source of assimilable nitrogen has greater specific activity.

The effectiveness of vitamins and phosphates is less evident. The addition of thiamine is mainly justified by its ability to reduce the quantity of sulphite-reacting substances and pantothenate to prevent the formation of hydrogen sulphide and volatile mercaptans.

There is no technical evidence of the effectiveness of survival factors under anaerobiosis, although they are naturally present in fruit as prune bloom and intracellular lipids. 23

The obvious conclusion is that the availability of assimilable nitrogen is always critical, not only in terms of yeast growth but also to improve its fermentative capacity. In addition, to obtain fermentation


TABLE 5 AN EXAMPLE OF THE EFFECTS OF THE AVAILABILITY OF ASSIMILABLE

NITROGEN AND OF 'FACTORS OF SURVIVAL' ON AEROBIC AND ANAEROBIC

MUST FERMENTATION (Adapted from Ingledew and Kunkee24)

Atmosphere Additions Nitrogen

Air Nitrogen Air Nitrogen Nitrogen

None None YE YE YE+SF

YE: yeast extract.

utilisation (mg/litre)

87 65

314 107 293

SF: survival factors (Tween 80, ergosterol).

Cell count Fermentation rate (x107 /litre) (gsugar per 100mi per 24h)

35·0 1·3

36·0 3·0

32·0

1·80 0·50 4·00 1·22 3060

with resting yeast in the steady-state, it is necessary to have a sufficient level of oxygen or, alternatively, the so-called 'survival factors'. Some data taken from Ingledew and Kunkee's experiments24 clearly show the relationship between the availability of assimilable nitrogen, oxygen and survival factors (Table 5).

Another indirect attempt to activate or reactivate fermentation consists of the use of absorption materials which are physically able to remove chemicals inhibiting fermentation, such as fermentation byproducts (e.g. long-chain acids) or pesticides. In addition to activated charcoal and bentonite, the use of cell walls deriving from yeast autolysis has been proposed. A characteristic of this material is assumed to be its specific absorption of long-chain acids. 25

The pH Values for Yeast Growth Within the range of optimal pH values for yeast growth (from 4-6), the activity level has little influence on metabolic activities because intracellular pH is relatively independent of the medium's acidity. However, even if a deviation in fermentation cannot occur within the range of must pH values, it has been found that ethanol and glycerol yields, in addition to the concentration of acetic acid, depend on the pH. Furthermore, it is impossible to determine the significance of the individual effects of pH and sugar concentration, since they are closely interconnected in fruit ripening. 26


Polyphenols and Yeast The presence of polyphenols is relevant for the organoleptic features and colour stability of wines. Therefore, the effects of these components on yeast metabolism are fairly indefinite, despite the importance empirically assigned to so-called 'tannins' in terms of fermentation behaviour26. Singleton and Esau27 in their exhaustive monograph on grape phenolics assert that 'the potential effect of yeast on phenols of wine would appear to be greater than has yet been defined'.

The main groups of phenolics in grapes27 are hydroxycinnamic acid and their esters, flavans (catechins and their oligomers and proanthocyanins), pigments (flavonols and anthocyanidins).

The available data28 on the effects of vegetable phenolics on the microflora of wine dating back to the last century are rather heterogeneous and even contradictory. The ability of yeast to metabolise catechins and other phenolics has been demonstrated,29 but the metabolic pathways of some effects of these compounds on yeast have yet to be explained.

Many authors have demonstrated colour and phenolic content changes during the fermentation of musts into wine, but relatively little is known at present about the release of these compounds from solids and about enzymatic activities other than phenolase. The loss of phenolics during the fermentation of grape musts with different parts of the fruit have been determined.30 Based on the results of these trials, the following points should be stressed:

-there is a considerable loss in terms of total phenolics, but a relative increase in flavan oligomers (reacting with vanillin);31

-at the end of the fermentation process, a high concentration of flavans is associated with the settled biomass;32 the quantity of phenolics removed in this way increases considerably after treatment with laccase;33

-a loss of anthocyanins takes place with a corresponding colour change. This fact is generally ascribed to an adsorption of these compounds on the yeast wall; in fact, significant adsorption has been demonstrated in model systems containing resting yeast cells and yeast cell walls;34

-the presence of grape phenolics causes a sharp lowering in oxidoreduction levels in the medium.3!

A set of yeast strains selected for wine fermentation was tested to


TABLE 6 INFLUENCE OF RED GRAPE PHENOLICS ON MUST FERMENT A nON IN TERMS

OF PRODUCT YIELD (Adapted from Cantarelli and Avesani34)

Fermentation products

Fermentation velocity (V) Fermented sugar (g/litre) Biomass (g/litre) CO2 (g/litre) Ethanol (g/litre) Glycerol (g/litre) Acetic acid (mg/litre) Pyruvic acid (mg/litre) Volatile esters

(as ethyl acetate) (mg/litre)

Total

Composition

Rawo Phenolics addedb

1·79 4·82 194·8 177·24

1-44 3·59 94·93 96·44 89·95 87-42

7-42 7·87 168 637 132 13

63 18

Carbon balance

Raw Phenolics added

9·3 25·6 332 371 602 643 37·3 43-4 0·9 3·6

0·4 0·1

982 1087

o Filtered white grapes must. b Same must, addition of 8 g/litre of phenolics extracted from red grape.

demonstrate the effects of phenolics extracted from red and white grapes on must fermentation. 35 The most important result of these trials was the demonstration of the activity of red grape phenolics in the stimulation of both yeast growth and fermentative activity in terms of biomass production, fermentation speed and product yield. Phenolics from white grapes did not produce these effects. Proanthocyanins had an inhibitory effect on yeast growth and turned out to be absorbed by yeast in stirred cultures (see Tables 4 and 6).

The effects of phenolics from red grapes was thoroughly examined. 34

It was observed that the removal of phenolics from the medium (by PVPP) sharply lowered the production of biomass and the fermentation speed. On the other hand, the addition of purified anthocyanins from blackberries confirmed their clear stimulative effect on yeast growth; the increase in biomass production was proportional to the anthocyanin content of the medium up to 200 mg/lOO ml. A Coulter Counter test demonstrated a reduction in cell size at higher anthocyanin levels.

Availability of Molecular Oxygen Lack of oxygen compromises yeast growth by preventing respiratory glycolysis and the synthesis of membrane lipids. Under total


anaerobiosis, cell reproduction stops if the substrate lacks those 'survival factors'; the ability to metabolise carbon dioxide is not sufficient to sustain growth, but cell viability remains. If oxygen is administered, growth will be restored. 36

It has also been observed37 that when small quantities of oxygen are administered to the substrate, the growth of 'young' cells-but not 'old' ones--is stimulated. On the other hand, the vicarious effect of 'survival factors', in comparison with that of oxygen, demonstrates that respiration is not involved in fermentation.

These observations lead to the assumption that must fermentation may result in high ethanol yields (and reduced accumulation of byproducts) if the transformation occurs under anaerobiosis with resting yeast in the steady-state. This supports Amerine and Ough's assertion that 'it is to the winemakers advantage to maintain anaerobiosis'.38

OPERATING CONDmONS

This heading could include several different subjects, such as temperature, gaseous environment (carbon dioxide, oxygen), stirring and suspended solids, free vs immobilised yeast and fermenter size and geometry. Most of these subjects have been treated by the other authors (see Chapter S on Carbon Dioxide by Dr Slaughter, Chapters 8 and 9 on Fermenters by Dr Moresi and Chapter 11 on Immobilised Yeast by Dr Divies), and consequently only the effects of temperature will be considered here.

The growth of yeast, like that of all other living organisms, requires heat, which is supplied by the environment and by metabolic reactions. The speed of the latter reactions increases as the temperature increases within a range which is limited by denaturation. The thermal coefficient of fermentation, QlQ, decreases as the temperature increases: within the range between 30°C and 20°C (optimal for growth), the coefficient is about 4-S. However, between SoC and lSoC, the coefficient reaches values of about 10-12. These values are much higher than those indicated for in-vitro enzymatic reactions. 39 The reason for this behaviour is the intake and output of solutes by the yeast cells, which are extremely sensitive to temperature. Therefore, in the case of wine yeast, the optimum temperature for growth is different from that for fermentation.

There are considerable data available on the effect of temperature on


yeast growth and fermentation. A well-known experiment conducted by Castelli40 on 20 wine yeast strains demonstrated that fermentation starts with 24 h when the temperature is between 30°C and 37°C, after 24 h at 25°C, and after two days at 15°C. At 5°C only four strains were able to ferment after 15 days. On the other hand, Castelli was able to show that the best yields are obtained at 15-17°C, and that the yield worsens at higher temperatures. In particular, the quantity of acetic acid increases as the temperature increases. He also observed that optimal fermentation of Saccharomyces cerevisiae and Sacch. bayanus required higher temperatures than those required by Sacch. uvarum.

These results from over 40 years ago, which have been confirmed and reconfirmed by numerous tests, demonstrate that: (a) the optimal temperature for fermentation (15-20°C) is different from that for growth (25-30°C); (b) wild yeasts have lower optimum temperatures, and (c) the balance of glycolysis changes with temperature.

It should also be noted that temperature has an effect on the leaching of solids (pomace), the development of Schizomycetes, the figures of combined fractions of sulphur dioxide, the vapour pressure of volatiles and their stripping by the stream of carbon dioxide during fermentation.

Significant differences noted during sensorial analysis may be related to these observations. 41 The production of acetaldehyde, volatile esters, long-chain alcohols and volatile components of sulphur are of particular interest. It is a well-known fact that white wine fermented at low temperatures retains its primary aromas, but it may have a higher concentration of acetaldehyde and hydrogen sulphide (see Table 3). High-temperature fermentation has more complex nutritional requirements, which are partially the result of damage in the synthesis of membrane lipids. 42 High temperatures can inactivate critical oxidative enzymes resulting in an accumulation of pyruvate and ethanol; the optimum temperature for alcohol dehydrogenase under anaerobiosis is around 40°C.

The recently revised thermal balance of fermentation provides values of thermal content Il.H equal to -24·96 kcal/mole for gaseous carbon dioxide and -34·78 kcallmole when this gas is dissolved.43 The maximum value during the starting phase of fermentation is ten times higher than the theoretical value. Only 30-50% of the heat is dissipated by heat exchange through the walls of the vat and through the stream of carbon dioxide. Cooling operations are necessary in the


m~jority of cases. Temperature regulation may be achieved with 'open' or 'closed' systems, with heat exchangers located outside or on the wall of the vats, respectively. A description of the various systems appears in Chapter 9 written by Dr Moresi. Devices for temperature regulation provide results which are quite varied in terms of final costS.44

If we relate this information to yeast activity, which is the subject of this chapter, we reach the following conclusions: (a) starter must be added to precooled must in order to maintain a lower temperature with lower cooling costs; (b) the must's dispersed materials should be removed by fining, centrifugation and filtration, while paying particular attention to the admission of oxygen, and (c) the addition of sulphur dioxide to slow fermentation once the process has begun may be risky.

The possibility of obtaining a linear kinetics of fermentation has been mentioned with reference to what has been said earlier about the use of high concentrations of free or immobilised yeast under conditions of anaerobiosis in order to repress growth. One of the major objectives of this technique is to limit the rise in temperature during the starting phase, which is part of the conventional process.

Attempts at Optimisation of Fermentation Conditions Several papers discuss approaches for the improvement of fermentation conditions based on comparative trials on homogeneous lots of must or the use of fermenters provided with devices regulating operating conditions (such as temperature, stirring and gaseous atmosphere). Amerine and Ough.45 developed a significant series of assays on variables concerned with red wine fermentation. Nevertheless, it should be stressed that the experimental conditions for all of these assays do not allow valid scaling-up to the actual conditions of winemaking.

More recently, a study was developed46 to ascertain the possibility of obtaining a planning map for optimising 'white' wine fermentation. Five variables (sugar content, pH, available nitrogen in the form of ammonia, phenolics from red grapes and temperature) were selected in order to assess their individual and combined effects on fermentation length, the rate of carbon dioxide evolution and the production of ethanol, glycerol, acetic acid and volatile esters.

In conformity with the 'composite' design technique, five different levels were chosen for each variable within the range of practical


350

2

III

liO ~ oJ

3

-1

100 -2

50

0 2 4 6 8 10 Polyphenols (gllitre)

i i i i

-2 -1 0 2 Levels

FIG. 1. Fermentation velocity as a function of polyphenols and sugar concentrations at constant pH (=3·5), Nitrogen content (=0·62 g N/litre) and

temperature (=25°C).

interest. Once the statistical consistency of the experimental results of two replicate series of 47 fermentation assays was checked on the basis of the carbon balance, these data were submitted to canonical analysis. The results may be summarised as follows:

-Fermentation rate: this parameter was found to be mainly dependent on temperature, sugar concentration and, to a lesser extent, polyphenols. By way of example, Fig. 1 shows the loci of the constant rate of carbon dioxide evolution as a function of temperature and polyphenols. This demonstrates the existence of an optimal concentration of polyphenols, which in tum depends on the fermentation temperature. Furthermore, the higher the temperature, the higher the fermentation rate will be. At a constant temperature and sugar concentration, the effect of the polyphenols on the fermentation rate is positive at concentrations lower than 4 g/litre and negative at higher values up to 8 g/litre. This demonstrates their inhibitory effect on the fermentation rate.


2 40

1 2

-1

-2

41 30 I... :::J .... 111 I... 41 D.

~ 20 I-

10

------------ -Z1=Sso/0

- r---- I--r--r-- '---z1=S7·so/0

- r---- -I----I--- _ Z, =600/0

50 100 150 200 250 300 350

I

-2

Sugar (g/litre) I

-1 I

o Levels

I

2

FIG. 2. Ethanol yield (Zt) as a function of temperature and sugar concentration at constant pH (=3·5), Nitrogen content (=O·62g N/litre) and

polyphenols content (=2 g/litre as gallic acid).

Similar effects are revealed by increasing either sugar concentration at constant polyphenol content or both polyphenol and sugar content.

-Cell growth: yeast growth appears to be controlled by sugar concentration, temperature and polyphenols, whereas the effects of pH and nitrogen content (over 120 mg/litre) are statistically irrelevant. Cell yield appears to be inhibited by increasing either sugar concentration or temperature; however, it is increased by adding polyphenolic compounds to the must, no matter what the fermentation temperature.

-Ethanol production: the most important parameter controlling ethanol production is temperature, followed by sugar concentration. Figure 2 reports the loci of constant ethanol yield at various levels of these variables. As these variables increase, either individually or together, ethanol productivity decreases, and this fact has often been reported in the literature.

-Fermentation byproducts: glycerol yield changes as a function of


temperature and sugar levels; the higher these values, the greater the glycerol production.

Available nitrogen, polyphenols, pH and their interaction appear to be statistically insignificant as far as the final glycerol content of the wine is concerned. Acetic acid yield (referring to the ethanol production) increases as temperature and sugar concentration increase; polyphenols lower the content of this byproduct.

Volatile esters increase as the temperature increases up to 22°C. At higher temperatures they are quite easily stripped by the evolving carbon dioxide, and this can result in lower final ester content in the wine. On the other hand, increasing the sugar content favours ester production.

Optimal operating conditions: by combining the patterns of each variable described above, it is possible to obtain an overall picture indicating the optimal conditions for 'white' fermentation of white and rose (pressed, 'vin gris' type) wines.

Figure 3 and Table 7 show the optimal range of values for the

I

-t. 2 40 t

ell

1 U ~

.... 30 I..

:J -

~o .... ~ ~ ~ ....

...J

-1

a. "M E (') ~ 20 N

-2 10

/

50 100 150 200 250 300 350

I -2

Sugar (g / litre) I

-1 I o

Levels

I 2

FIG. 3. Contour map of ethanol (Zt), glycerol (Zz) and acetic acid (~) yields against the main operating variables: temperature and sugar concentration.

lHE BEHAVIOUR OF YEAST IN WINE FERMENTATION 145

TABLE 7 A TENTATIVE PLANNING MAP FOR OPTIMISING WHITE WINE FERMENTATION

(Adapted from Cantarelli and Moresi46)

According to the 'composite design' technique, the following levels for the main variables were:

Levels: -2 -1 0 +1 +2

Factors: pH 2·5 3·0 3·5 4·0 4·5 Nitrogen

(g/litre) 0·12 0·62 1-12 1·62 2·12 Sugar (g/litre) 80·0 147·3 217·8 268·0 320·0 Polyphenols

(g/litre) 0·0 2·0 4·0 6·0 8·0 Temperature eC) 10·0 17·5 25·0 32·5 40·0

An overall picture to optimise the main productions of white musts fermentation can be given showing the optimal values for temperature, sugar and polyphenols content (pH and nitrogen content have a negligible statistical significance) :

Temperature Sugar Polyphenols eC) (g/litre) (g !litre) Maximum fermentation rate 35 200 5 Maximum ethanol production 15 160-270 Minimal acetic acid production 23 200 Minimal glycerol production 5 150-280 Minimal yeast growth 10 300 7

temperature, sugar and polyphenol content of must in order to obtain maximum ethanol yield and minimum accumulation of undesirable byproducts such as acetic acid. Data on pH and nitrogen are not included, given their negligible statistical significance in these trials.

Fennentation with High Yeast Concentrations Various experimental approaches which generally concern industrial ethanol production (but are also applicable to beer and wine fermentation) are oriented towards maximisation of the substrate conversion rate, since yeast productivity is extremely low with, the use of conventional technology. The latter fact should be attributed to the inhibitory effect of the substrate (sugar concentration), the alcohol produced and other factors, such as the yeast's auto-inhibition and the


crossed inhibition between different strains (which differs from the killer effect).

There are two kinds of methods used to overcome this inhibition: (a) separation of the must's sludge and addition of adsorbent agents active on inhibitors, or (b) addition of significant quantities of yeast biomass in order to obtain substrate fermentation in the steady-state. At high concentrations, yeast growth virtually stops, and elevated productivity occurs; the effects of the inhibitors and substrate are cancelled. For instance, it has been observed that the biosynthesis of short-chain fatty acids takes place when yeast growth ceases.47

Different systems may be used for the processing of high concentrations of yeast, including: (a) cell retention through flocculence; (b) retention through microfiltration membranes; (c) biomass recovery by centrifugation and recycling, and (d) yeast immobilisation.

With conventional fermentation, yeast concentrations are in the order of 0·2-0·5 g/litre of yeast (dry matter). However, these values may be increased by .two orders of magnitude, thereby producing extremely high fermentative activity on the order of 2 g of ethanol/hour/g of dry yeast.

Experiments of this sort, conducted on grape must and wine to be refermented, have provided interesting results. Their main objective was to set up a one-tank winemaking process. Results obtained over the course of the past five years are summarised below.

The experiments48--50 were conducted on pilot plants (200-litre cylindrical fermenters and 30-litre jar fermenters) with yeast obtained under various conditions in the form of liquid, malt agar or commercial dried cultures or recycled from previous fermentation by centrifugation or immobilised alginate. Yeast was also obtained by recirculation of fermenting must using an external pump, addition of inert solids (diatomaceous earth), maintaining the immobilised yeast as the ftuidised bed with a nitrogen stream, and mechanical stirring.

In all cases, comparison was made between experimental and conventional conditions; comparison was also made between free yeast cells and immobilised yeast and various yeast concentrations from 2-8 g/litre. The following aspects were observed (see Tables 8 and 9):

-High yeast concentrations produce must fermentation with higher yields under anaerobiosis;

-immobilised yeast ferments under anaerobiosis with a linear kinetics and without log or exponential phases;

TA

BL

E 8

S

TE

AD

Y

ST

AT

E

WIN

E

FE

RM

EN

TA

TIO

N.

YE

AS

T

GR

OW

TH

, F

ER

ME

NT

AT

ION

A

ND

S

UR

VIV

AL

IN

G

RA

PE

M

US

T

UN

DE

R

AE

RO

BIC

AN

D

AN

AE

RO

BIC

C

ON

DIT

ION

S

Fer

men

tati

on

of

a m

ust

inoc

ulat

ed

wit

h ye

ast

from

ag

ar

slan

t (A

G)

and

as

biom

ass

recy

cled

(R

C)

from

a

grev

ious

fe

rmen

tati

on,

as

drie

d ye

ast

(DY

) an

d as

im

mob

ilis

ed y

east

(1M

).

(Ad

apte

d f

rom

Bal

sano

48 a

nd C

anta

rell

i 0)

Ref

. St

arte

r A

tmos

pher

e B

iom

ass

Gro

wth

7;

ag

T.,xp

C

ell

mor

tali

ty

yiel

d ra

te

Typ

e X

(Y

x)

(X)

M

Mm

ax

h

Ben

ch s

cale

B

.1

AG

0·

5 A

ir

140

0·93

24

96

2

B.2

l.

RC

1·

3 N

itro

gen

85

0·20

15

5 96

13

35

1

B.3

2

.RC

4·

5 N

itro

gen

129

0·57

1

290

6 20

2

B.4

A

G

1·8

Air

16

5 1·

23

12

120

3 B

.5

l.R

C

1·8

Nit

roge

n 54

0·

16

70

190

14

33

4 B

.6

2.R

C

3·6

Nit

roge

n 12

0 0·

94

1 17

0 4

10

3 -------.-~-. ---" --

.. ---~--

Ref

. St

arte

r A

tmos

pher

e St

irri

ng

Bio

mas

s A

lcoh

ol

Fer

men

tati

on

Byp

rodu

cts

yiel

d yi

eld

vel~

.cit

y yi

eld

Typ

e X

(Y

,J

(~)

(V)

(YA

) (Y

o)

Pilo

t sca

le

A.l

D

Y

5·7

Air

N

one

21

0·57

0·

78

6·5

47

A.2

D

Y

2·0

Air

Y

esa

48

0·53

0·

77

4·3

64

A.3

D

Y

7·4

Nit

roge

n Y

es·

23

0·56

0·

90

8·9

77

A.6

1M

20

·0

Nit

roge

n Y

esb

0·57

0·

87

6·2

59

P.7

D

Y

1·0

Air

Y

esc

96

0·57

1·

32

4·8

P.1

4 D

Y

2·0

Nit

roge

n Y

esb

6 0·

63

0·61

3·

5 -

----------

a M

echa

nica

l (i

mpe

ller

, N

:30)

. b

Gas

bub

blin

g.

C R

ecyc

led

by e

xter

nal

Mon

o pu

mp.


-mechanical stirring speeds up fermentation; -the presence of dispersed solids produces a similar effect.

The composition of the wines produced by these experiments differed in terms of the content of volatile components. Results were statistically evaluated for 22 analytical variables (multivariate analysis), but this did not demonstrate the existence of differences higher than those due to the original characteristics of the must (Tables 8 and 9). Sensory evaluation confirmed this basic similarity.7.18 Other experiments conducted on tank refermentation of sparkling wine have made use of an immersed 'reactor' of immobilised yeast under static

TABLE 9 PILOT SCALE FERMENTATIONS (50-2oo-litre) WITH FREE AND IMMOBIUSED

YEAST CELLS AT CONVENTIONAL AND HIGH CONCENTRATION (Adapted from Balsano48 and Villa49)

Conventional High concentration

Free cells Free cells Immobilised 0·10 2 2 Air Air Nitrogen

Musts (white) ref. Y YA Y YA Y YA

A.1 0·54 5·7 0·65 10·0 0·36 3·3 B.1 0·37 5·6 0·51 7·1 0·41 2·2 C.1 1·05 2·9 1·31 2·5 0·57 2·4 D.1 1·36 2·1 1·72 3·4 0·97 3·0

Mean values 0·83 4·0 1·05 5·7 0·57 2·7

Increasing yeast concentration

Starter Lag time Y YA YH

(h) Concentration Conditions

(X)

Musts (white) ref. A.2 0·15 Free 80 0·45 2·7 2·7 A.2 0·15 Immobilised 90 0·32 2·8 4·6 B.2 2·5 Free 35 1·20 2·3 4·2 B.2 2·5 Immobilised 25 0·77 2·4 4·6 A.2 6·0 Immobilised 6 0·86 3·2 6·5 D 8·0 Immobilised 0 8·70 3·2 13·0


conditions and complete anaerobiosis. The kinetics of fermentation was linear (without slow-downs), and the quality of the final product was judged to be good as a result of the lower content of acetaldehyde and higher alcohols.

The technique using high yeast concentrations appear interesting for the fermentation of light white or rose wine with a neutral, fresh taste. However, its limitations include:

-The cost of yeast, which can be amortised through recycling that is easily attainable with immobilised yeast;

-the loss of activity of the biomass, or so-called 'intoxication', which is due to intracellular ethanol content and other metabolites (long-chain aliphatic acids, higher alcohols, esters); an intermediate detoxication cycle must be provided for based on the addition of oxygen and available nitrogen. In our experience, the immobilised yeast can regain its activity through the addition of co-factors (a vitamin B pool) and minerals;

-the need for de-aerated or possibly even pasteurised must without dispersed materials and with low nitrogen content;

-the fact that all of the experiments conducted under anaerobiosis provided satisfactory results in terms of fermentation yield, but the formation of acetic acid was observed in every case. Consequently, the latter product is not suppressed even when oxygen is excluded.

The improvement of wine fermentation technology, in order to optimise the process, is a wide-ranging subject for applied microbiological research. It is important to establish the role to be played by the different variables, the most important of which are those conditioning yeast activity: biomass concentration, the growth phase, the oxygen level and the temperature.

REFERENCES

1. Cantarelli, C. and Baccioni, L. (1984). Ind. Bevande, 7,7-23. 2. Fages-Bonnery, A. (1968). In: Fermentations et Vinijications, INRA,

Paris, pp 553-83. 3. Minarik, E. (1978). Bull. DIV, 51, 352-67; Rankine, B. C. (1978). Ann.

Techno/. Agric., 27, 189-200; Zambonelli, C. (1977). Atti Ace. Ital. Vite Vino, 29,51-60.

4. De'Rossi, G. (1927). Microbiologia Agraria e Technica, UTET, Torino.


5. Dittrich, H. H. (1977). Mikrobiologie des Weines, E. Ulmer, Stuttgart, p. 84.

6. Cantarelli, C. (1966). Bull. OW, 39,428, 1191-1203. 7. Bertuccioli, M. (1987). Personal communication. 8. Nagodawithana, T. V. and Steinkraus, K. H. (1978). Appl. Envir.

Microb., 31, 158-62. 9. Leao, C. and Van Uden, N. (1982). Biotechn. Bioeng., 24,2601-4.

10. Guijarro, I. M. and Lagunas, R. (1984). J. Bact., 160, 874-80. 11. Cantarelli, C. (1985). Atti Acc. Ital. Vite Vino, 37,431-48. 12. Castor, J. B. C. and Guymon, J. F. (1952). Science, 115, 147-9; Millan,

C. and Ortega, J. M. (1988). Am. J. Enol. Vitic., 39, 107-12. 13. Kleber, W. (1987). Systems of fermentation. In: Brewing Science, Vol. 3,

J. R. H. Pollock (Ed.), Academic Press, New York. pp. 329-75. 14. Hough, J. S., Briggs, D. E., Stevens, R. and Young, T. W. (1982).

Malting and Brewing Science, Vol. 2, Chapman & Hall, London, p. 677. 15. Cahill, J. T., Carroad, P. A. and Kunkee, R. E. (1980). Am. J. Enol.

Vitic., 31, 46-52; Monk, P. R. and Storer, R. J. (1988). Am. J. Enol. Vitic., 39, 62-76.

16. Slominski, P. P., Perrodin, G. and Croft, J. A. (1968). Bioch. Bioph. Res. Comm., 30, 232-43.

17. Linko, M. (1988). Proceed. Eur. Brew. Conv. Helsinki, Elsevier, Amsterdam, pp. 39-50.

18. Bongianino, P. F. Fermentazione vinaria con lieviti immobilizzati: studio delle condizioni operative e delle. Thesis S.P.A., University of Milan, 1987.

19. Wiken, T. and Richard, O. (1953). Experientia, 9, 417-21; Dedeken, R. H. (1966). J. Gen. Microb., 44, 149-53.

20. Gray, W. D. (1945). J. Bact., 49/A, 445-54. 21. Goma, G., Uribelarrea, J. L., Mota, M. and Strehaiano, P. (1985).

Proceed. Eur. Brew. Conv., Elsevier, Amsterdam, pp. 13-87. 22. Cantarelli, C. (1957). Am. J. Enol. Vitic., 8, 113-20, 167-75. 23. Macy, J. M. and Miller, M. W. (1983). Arch. Mikrob., 134,64-7. 24. Ingledew, W. M. and Kunkee, R. E. (1985). Am. J. Enol. Vitic., 36,

65-71. 25. Geneix, c., Lafon Lafourcade, S. and Ribereau Gayon, P. (1983). Conn.

Vigne Yin, 17,205-17. 26. Eroshin, V. K. (1976). Biotech. Bioeng., 18,2893. 27. Singleton, V. L. and Esau, P. (1969). Phenolic substances in grapes and

wine and their significance, Adv. Food Res. Academic Press, New York. 28. Rosenstiehl, A. (1902). C.R. Acad. Sciences, 134, 119-22. 29. Harris, G. and Ricketts, R. W. (1962). Nature, 195, 473-4. 30. Cantarelli, C. and Peri, C. (1964). Am. J. Enol. Vitic., 15, 146-53. 31. Peri, C. (1968). C.R. Symp. IntI. Oenol., Bordeaux, INRA, Paris, pp.

325-9. 32. Singleton, V. L. (1967). Techn. Quart. Master Brewers Ass. Amer., 4,

215-53. 33. Cantarelli, C. and Mantovani, L. (1988). Ind. Bevande, 17,337-44. 34. Cantarelli, C. and Avesani, P. (1986). Atti Acc. Ital. Vite Vino, 37,


285-303; Saraiva, R. (1983). Incidence des composes phenoliques a l'egard de l'activite des levures et des bacteries. Thesis, University of Bordeaux II.

35. Cantarelli, C. (1988). Proc. 7th Inti. Symp. Yeasts, Perugia, p. 33. 36. Cantarelli, C. (1956). Ann. Inst. Pasteur, 90, 645-9. 37. Viken, T. O. (1968). In: Fermentations et vinijications, INRA, Paris, pp.

155-71. Tyagi, R. D. (1984). Process Biochem., 136-141. 38. Amerine, M. A. and Ough, C. S. (1957). Am. 1. Enol. Vitic., 8, 18-30. 39. Rose, A. H. (1976). Chemical Microbiology, Plenum Press, New York.

pp.129-69. 40. Castelli, T. (1941). Ann. Microb. Enzim., 2, 1-14. 41. Wilson, K. and McLeod, B. J. (1976). Ant. v. Leeuwenhoek, 42,397-401. 42. Cottrell, T. H. E. and McLellan, M. R. (1986). Am. 1. Enol. Vitic., 37,

190-4. 43. Williams, L. A. (1982). Am. 1. Enol. Vitic., 33, 149-56. 44. Cantarelli, C. (1984). L'Enotecnico, 20,383-96. 45. Amerine, M. A. and Ough, C. S. (1951). Am. 1. Enol. Vitic., 8, 18-30;

(1960). Am. 1. Enol. Vitic., 11,5-14; (1961). Am. 1. Enol. Vitic 12, 9-19 and 117-28.

46. Cantarelli, C. and Moresi, M. (1984). Atti Acc. Ital. Vite Vino, 35, 225-42.

47. Liu, L., Youzhi, Q. and Yi, Z. (1982). Food & Ferment. Ind. (China), 4, 21-5.

48. Balsano, A. (1983). Fermentazione vinaria con lievito non proliferante, Thesis, S.P.A., University of Milan.

49. Villa, M. (1987). Ricerche sperimentali sull'impiego di lieviti immobilizzati nella fermentazione dei mosti, Thesis, S.P.A., University of Milan.

50. Cantarelli, C. (1988). Proc. Congo Biotecnologie in Enologia, Acc. Ital. Vite Vino, Milan, 29-30 June 1988 (in press).

Chapter 11

ON THE UTILISATION OF ENTRAPPED MICROORGANISMS IN THE INDUSTRY

OF FERMENTED BEVERAGES

C. DIVIES

Microbiology Laboratory, ENSBANA, University of Bourgogne, Dijon, France

INTRODUCTION

Most fermented drinks have a vegetable origin. They result mainly from enzymic activities which can come either from the raw material or from microorganisms.

The major part of the biochemical reactions is known but still not completely controllable which in many cases limits the possibilities of industrialisation.

The reactions of the enzymes from the raw material are usually specific: for example, sucrose is hydrolysed to glucose and fructose by invertase, proteases and amylases are hydrolytic enzymes which hydrolyse proteins and starch.

Microorganisms can have more complex reactions, generally involving more than one enzyme: for example, yeasts ferment C6 or C12 sugars to give ethanol by means of more than ten interlinked enzymic stages.

The progress made in the last 25 years in the use of these two kinds of 'biocatalysers' will have important repercussions on the transformation processes of fermented drinks in the near future.

Since 1960, the process of enzyme immobilisation has shown a great expansion and many practical applications have been suggested and used in the analytical area (enzyme electrodes) and in the area of the industrial production of biomolecules. 12 The main advantage which led to the research is that the price of the treated products can be brought down since the enzymic preparation is used continuously and it is possible, using concentrated enzymes, to increase the productivity of the reactors.

153


154 C. DIYlES

The purification of enzymes is onerous. Therefore it was of interest to use whole viable microorganisms with high enzymatic activities in those cases where enzymatic reactions could overwhelm the selected transformations.

This conception is remarkably illustrated by enzymes without soluble coenzyme as is the case with aspartase3 and glucose isomerase.4

Moreover, the stability of the biocatalyser obtained by that process is greater than the stability of the immobilised enzymes. 2 These observations concern unienzymic reactions without soluble coenzyme.

Divies5,6 demonstrated that it was possible to immobilise microorganisms in alginate gels with a high percentage of the cells remaining viable for a long period of time and as a result it was possible to carry out multienzymic stage reactions. Alcoholic, lactic and malolactic fermentations, denitrification and anaerobic purification of water were taken as examples.

Later, many authors confirmed this possibility and a real explosion of research has taken place in the last 10 years, the applications of this technology are really promising.7- 10

Microorganism cell immobilisation or fixation can be defined as the localisation of whole cells in the reacting space with remaining catalytic properties so that it permits a repeated and continuous utilisation.

TECHNIQUES OF IMMOBILISATION

A number of physical and chemical methods developed for enzyme fixation can be used for cell immobilisation.9

Five principal techniques can be used (Fig. 1):

(1) Immobilisation by means of adsorption, for microorganisms as well as enzymes, is very simple. The cells and the carrier are left in contact in defined physical and chemical conditions. The carriers can be of different types, e.g. brick, glass, wood, PVC. However, the biofilm is likely to undergo a quick desorption caused either by an increase of the liquid velocity or by the underlying cells autolysing when running.

In the past few years the adhesion force of microorganisms on the carriers has been widely studied. 12 In most cases, several stages beside the purely physical and chemical phenomena are involved: 13

ENTRAPPED MICROORGANISMS IN FERMENTED BEVERAGES 155

Adsorption Polymertsation

fG®~* Physical Ionic bond bond

Covalent bond

I Cellular retention

Entrapment by membranes Flocculation

I I ... 1.... I

• , ;~~' ®®r.;,: .; .... ; -II :® ® I ® \!!I.

M f~ ... _ \ .. ~.-# L~ __ ~j Beads ;'® Encapsulation

Fibres Membrane reactor

FIG. 1. Immobilisation techniques. S = carrier; M = microbial cell.

-previous adsorption of a polymer on the carrier (e.g. protein, polyose)

-attraction of the microorganism -reversible adsorption -unreversible adsorption

Yeasts are naturally important biofilm producers/4 in contrast to lactic bacteria.

(2) Fixation methods using covalent bonds require very reactive molecules (carboiimide, glutaraldehyde) so that only a few cells remain capable of division after immobilisation.

(3) Flocculation of microorganisms occurs naturally in a number of strains and can be artificially induced as well. This is an attractive method of retention because it uses the simplest method of liquid-solid separation, i.e. decanting. 15 A quantity of 60-110 g/litre of yeasts can be held in activity without physical constraint on the microorganism.

The character is often taken into account and used in yeast selection. 16

(4) The entrapment technique is now a well used method. It consists of entrapping cells in a rigid network which prevents the surrounding liquid from cell invasion but still allows the diffusion of substrates and products. A number of polymers have been proposedpolyacrylamide, polyvinyl chloride and polyurethane are among the synthetic polymers in use. In these cases, the polymerisation conditions can be harsh. Lactic bacteria and yeasts tolerate, however, relatively well the polymerisation of polyacrylamide5 since 10-80% of the cells remain viable after the treatment.

The natural polymers in use are often polysaccharides such as

156 C. DIYlES

alginate, carrageenan or agar. Gelatin, collagen, chitosin and cellulose can also be used. They allow gentler conditions of polymerisation. The entrapment technique in alginate is a good illustration of this. It consists of thoroughly mixing a suspension of cells with a 2% sodium alginate solution (soluble form of D-manntlronic and L-guluronic acid polymer) and then adding the homogeneous system thus obtained to a concentrated solution of divalent cations (calcium chloride).

The cation leads to a breakdown of the gel because of ionic interactions with the carboxylic groups of the polysaccharide. Thus, standardised beads (0·5-5 mm) are obtained. Co-immobilisations involving microorganisms and enzymes are also possible. 17

(5) Membrane processes are attractive-they allow a complete recycling of the cells and the retention of an important biomass in the reactor. Microfiltration and ultrafiltration have already been successfully used for ethanol or carbonated beverages production. 18

Each technique has its own advantages and disadvantages. Only the utilisation of immobilised microorganisms through the entrapment method will be considered in the following sections.

KINETIC CONSTRAINTS OF ENTRAPPED IMMOBIUSED MICROBES AND unUSABLE REACfORS

Preparation methods of entrapped microorganisms lead to insoluble by-products such as membranes, and particles with different shapes and sizes. The choice of the reactor will depend on the nature of the immobilisation carrier, the microbial concentration, the concentration of the substrate and the toxicity of the accumulated products. When a microorganism is immobilisedwithin a polymer, its behaviour can no longer be described by means of kinetics usually applied in homogeneous media.

The microorganism fixation is liable to change its intrinsic kinetics (e.g. permeability, metabolic pools, macromolecule turnover, DNA replication). In many cases, the consequences are yet to be discovered.

In a heterogeneous medium the concentration of substrates, products or effectors in the very near vicinity of the immobilised microbe is often different from the one encountered in the solution away from the carrier, i.e. in the macroenvironment. These differences of concentration come both from the fact that the substances are


microkinetic and microkinetic and

physiology physiology interactions

free cell (microbe - carrier) immobi I ized cell

parti tion ~ E diffusion

macrokinetic and

physiology

immobil ized cell

FIG. 2. Relationship between the microkinetics of the free microorganism and the macrokinetics of the immobilised microorganism.

shared between the aqueous phase and the solid phase (interactions with the carrier), and from the diffusional resistances. Figure 2 shows the said phenomena.

The external transport rates can be evaluated with precision for most of the profiles by using chemical engineering correlations. Very incomplete information is available about the diffusivities of solutes within the gel. The coupling of diffusion with the reaction and the modelling of the system, when immobilised cells are studied, have given rise to many publications and reviews. 19,20

When using growing cells, inhibition of the growth due to a lack of a nutritional factor within the carrier is noted. Tosa and Sato21 have given theoretical calculations for the growth of Candida /ipolytica in strictly aerobic conditions, using agar beads. The growth is not limited for a diameter of the beads smaller than 0·2 mm. Gosmann and Rehm22 have observed using Pseudomonas putida, Saccharomyces cerevisiae and Aspergillus niger a limited growth in the peripheral part of the alginate beads. The same phenomenon occurs in continuous alcoholic fermentation with surface concentration up to 200 g/litre dry wt gel. Figure 3 shows the intraparticle distribution. This difference in cell distribution between the surface and the centre can be related to

158

Qi ~5.109

'" '" cu E o iii

;

o

C. DIYIES

-r--

0'5 Radius (mm)

FIG. 3. Yeast distribution in alginate beads after 200h running: beads diameter, 2·8 mm; population 3·6 x 109 cells/ml; temperature, 25°C; volume of reactor, 300 ml; volume of particles, 50 ml; residence time, 2 h; synthetic

fermentation medium, 100 g/litre.

imperfect diffusion of essential nutrient as the dissolved oxygen and internal ethanol concentrations are temporarily higher. 23,24

At moderate temperature (20°C), it is possible to get satisfactory results with 100 g/litre of yeasts immobilised in 2-mm alginate beads with an external sugar concentration of 20-50 g/litre. When the temperature reaches 35°C, the diffusional constraints appear (personal results).

In spite of the internal diffusion, reactors with immobilised cells remain productive in the fermentation process.

In the case of alcoholic fermentation which involves high substrate concentration, a heterogeneous fixed bed reactor allowing a reduced toxicity of ethanol is advantageous. The main drawback is the large release of carbon dioxide.

Fluidised bed reactors have many advantages because of a better external transfer. 25.26

BElL\. VIOUR OF ENTRAPPED YEASTS AND LACTIC BACTERIA

If the cells remain viable during the entrapment and if their nutritional requirements are satisfied during their later utilisation, the multi-

~30 1: E

1J ·u nI u ~ 20 E '0 c o

:;::; nI !:; 10 c &I u C o U

o


_____________ c~ ___________________ _

2

10 20 Days

FIG. 4. Evolution of malic acid concentration in the reactor effluent according to the composition of the feed. Volume of reactor, 100 ml; feed rate 38 ml/h. 1) Complete growth medium; 2) growth medium +50 Ilg chloramphenicol/ml; 3) growth medium without yeast extract and tryptone, with 10 ml tomato juice, the other elements remaining the same; 4) growth medium without yeast

extract, tryptone and tomato juice.

enzymic stages could remain active for several months. We demonstrated it in two cases related to enology-malolactic and alcoholic fermentations, and in brewing5 and cider-making??

The transformation of malic acid is shown in Fig. 4. The analysis showed that when there are no diffusional limitations, cell division occurs at the same rate whether the cells are entrapped or free (Refs 28 and personal results).

When equilibrium is reached, two critical growth phenomena occur in the gel-the ageing of the cells and the appearance of new cells in sufficient quantity to limit the activity.

This is clearly demonstrated in Table 1 which shows the thermal stability of lactic bacteria malolactic activity isolated at different times. After a 28-day run the thermal stability of the enzyme has decreased but has been restored after 65 days. In order to obtain a

160 C. DIYlES

TABLE I DECARBOXYLATION ACTIVITIES OF Lactobacillus casei

ISOLATED FROM A REAcrOR WHILE RUNNING

Cell activities after Samples (number of days) a heat treatment

at 0 1 9 28 45 65

30°C 491 103 65 60 61 53 40°C 511 105 68 39 27 42 % activity between 102 102 104 65 44·3 80

300C and 40°C % malic acid

transformed in the 95 95 93·3 80 92 86 reactor

The cells, isolated, have been heat-treated for 30 min. The malic acid decarboxylation has been measured by the Warburg method. Results are expressed in JlI of CO2 released per ml of protein and per hour at 300C and at pH 4·5. The feed is a 10 g/litre glucose, pH 4·5 medium. The feed rate is 38 ml/h, volume of reactor 100 ml, initial concentration of malic acid 33 mM.

successful microbial enzymic activity, it is better to cultivate the microorganism in gel.

It is also possible to get better fermentation activities with entrapped cells. Using one strain of Saccharomyces cerevisiae entrapped in 2-mm diameter alginate beads with 2 x 109 cells/ml of gel, we measured an activity of 0·6 g ethanol/g cells per hour against 0·3 g ethanol/g cells per hour using free cells (smaller beads give even better results). This increase in the fermentation rate has also been measured by Navarro and Durand29 with adsorbed yeasts.

TABLE 2 INFLUENCE OF IMMOBILISATION BY ENTRAPMENT ON

THE EVOLUTION OF SECONDARY PRODUcrs DURING

THE ALCOHOLIC FERMENTATION OF GRAPE JUICE

Producta

Glycerola

Acetaldehydea

Propanola

Isobutanola

Isoamyl alcohola

a mg/101o cells.

Free yeasts

88 7·3 0·62 2·5 4

Entrapped yeasts

100 2·9 1 2·5 5·2


Although the conditions of oxygen diffusion are limiting, we were able to measure an activation of the glucose respiration using immobilised Rhodorotula glutinis. The entrapment ensures the yeast is protected from the ethanol toxicity.

In the case of continuous stirred reactors (CSTR), we demonstrated that the ethanol inhibits its own production following the equation: V = Ve -kVP instead of V = Voe -kVP as proposed by Aiba et al. 30 for free cells, where Vo and V are the velocity of the reaction; g, the sugar consumed per hour and per unit biomass; P, ethanol concentration (gil); a and K, experimental constants. Lee et al. 31 and Ryu et aC2 gave V = Vo (1- aP) for a heterogeneous reactor. Recently, Casey33

made a complete review of these inhibition aspects. We followed the main secondary products of the fermentation which

are produced in unequal amounts by the immobilised cells. Table 2 shows the changes measured during the growth of immobilised and free cells in grape juice. We found that the production of glycerol, propanol and isoamyl alcohol was higher for immobilised cells than for free cells but less acetaldehyde was accumulated. Moreover, we found a better assimilation of amino acids for immobilized cells, which explains the production of primary, secondary and tertiary alcohols.

The reduction in acetaldehyde production can be explained by a better utilisation of the reduced coenzymes due to the cellular confinement in the gel. When the growth is limited and when the amino acids are less used, there is large accumulation of glycerol and a lesser accumulation of propanol. 23

This aspect of the changes in the fermentation profiles seems to be general. Hahn-Hagerdal34 proposed a more marked effect of water activity within the gel. Another explanation based on the fermentation profiles of acetaldehyde and glycerol could be a better utilisation of the reduced nucleotides pool. The composition of the immobilised yeast is also different (see Table 3). The glycogen content, the glucan, mannan, as well as the DNA contents are higher. Recently, Doran and Baile~5 have found the same differences. Substrate diffusion could explain the increase in reserve substances but the increase in DNA content needs further experiments to be undertaken before it can be explained.

Using lactic bacteria, we have demonstrated that the cellular confinement in gels leads to changes in fermentation profiles and in extrachromosomic factors replication. Entrapped cells have a high number of plasmid copies.36

Studies on citrate (-) variants in free or entrapped populations

162 C. DIYlES

TABLE 3 EVOLUTION OF MACROMOLECULAR CONSTITUENTS OF FREE CELLS AND

ENTRAPPED CELLS

Ratio: Proteins/ deoxyribose Proteins/ mannane Proteins/ glucane Proteins/trehalose Proteins/total glycogen Proteins/ alkaline glycogen Ribose/ deoxyribose

Free cells during growth phase

1551-1026 21·4-14 12·1-9·3

7-14·8 9-6

33-47 46-76

Entrapped cells

48h 648 h

734 12·8 7·2

30 4·55

23·46 53

500 11·5 4·25

85 2·5 9·6

25

show a very good stability of the strains when the cellular growth occurs near the surface. Likewise, the activities of the malolactic fermentation remain stable (unpublished results). De Taxis du Poet et al. 37 measured the same effect using E. coli with the plasmid PTG 201.

INDUSTRIAL APPLICATIONS

The Malolactic Fermentation (MLF) The malolactic fermentation of wine is generally carried out by a small bacterial population-5 x 107_108 CFU /ml (10 mg/litre dry wt assuming that the weight of a cell is 10-13 g).38

The method of cell immobilisation leading to an artificial increase of biomass could be, in that case, very useful.

As early as 1975 Divies5 emphasised the potential of this technique. Since then, many authors have tested the method and two relatively efficient pilot plants have been described.39,40 The half-life of the reactors is not more than ten days. The performance is still linked to the ability of bacteria to remain viable in the wine. The knowledge of Leuconostoc oenos has improved and better selected strains are now available.

In a recent work41 , we have found that it was possible to keep lO lD CFU / ml gel in activity using 2-mm diameter particles and to obtain a continuous inoculation of the wine. A residence time as short


as 10 min in the reactor is sufficient to realise an inoculation of 106 CFU/ml of wine. A residence time of the wine in the reactor from 1-6 h is necessary to achieve the entire degradation of malic acid.

The Bottle Fermentation 'Methode Cbampenoise' Among the utilisations of entrapped yeasts, the development of champagne-making is of interest. In this case a second fermentation of the wine occurs in the bottle itself. The use of entrapped yeasts in alginate beads allows freedom from the 'remuage' step42 which is a delicate phase to handle and which imposes important constraints on the manufacturer.

Figure 5 shows a bottle fermentation carried out using an optimised entrapped yeast preparation at 13°C. The performance must be stressed because it was complete after only 20 days for a normal cell

25

20

~6 15 ~

8 ~

,g Q) '- '-" <11 1/1 Cl 1/1 " Q) III 6:4 10

2 5

15 20 30 Time (days)

FIG. 5. Bottle fermentation kinetics using different methods of particle preparation. (0), (O}---sugar, pressure, optimised medium with a wine base; (.), (e)--sugar, pressure, synthetic medium. Entrapment: alginate 40 min in 0·5 M CaCI2, pH 7. Reactivation: 20°C, aerobic conditions. Bottle fermenta-tion: wine 25 g/litre glucose. Inoculum: 5 x 106 cells/ml. Temperature: 13°C.

164 c. mVIES

concentration. It must also be stressed that the entrapment allows a normal ageing of the wine.

The same steps of amino acid absorption by the yeasts and release when the mechanism slows down are found. After a five-months ageing period, cells have returned all their nitrogen to the wine. The bottle fermentation corresponds to the utilisation of 25 g/litre of sucrose which cannot entail deep changes in the secondary products of the fermentation. Sensory tests carried out with the final product did not give significant differences between a wine fermented by entrapped yeasts and a wine fermented by free yeasts.

The technology developed for champagne-making can be transposed to all the bottle-fermented sparkling beverages, particularly, the fermentation of wine with an admixture of a fruit infusion coming from the hydroalcoholic maceration of fruits43 or the fermentation of already fermented beverages such as cider or pineapple wine.27

Alcoholic Fermentation and Elaboration of Sparkling Drinks As early as 1975, Divies5 stated that the method was very efficient for carrying out the alcoholic fermentation of fermented drinks (wine, beer) and for ethanol production.

In brewing, White and Portno44 and Ryter and Masschelein45

demonstrated its feasibility. It is possible to obtain beer with a good organoleptic quality.

Onaka et at. 46 found that it was possible to reduce the diacetyl concentration if the fermentation was carried out in a very anaerobic microenvironment (about 0·04 mg/litre O2). The conditions of the systems are still to be optimised. In enology, the studies undertaken in Italy by Cantarelli confirm this possibility. Divies and Deschamps have just proposed27 this technology for cider-making.

For these drinks it is possible to keep 2 x 109_1010 cells/ml of gel in activity, using 1-3-mm alginate beads (20-100 g/litre of yeasts, assuming the weight of a yeast cell to be 10-11 g).

This important biomass can be kept in activity in a reproducible and controlled way using a relatively simple process, i.e. decanting. Continuous reactors, preferentially heterogeneous reactors or successive batches can be used.

In these conditions, the yeast flocculation is no longer needed. Furthermore, a fermentation under CO2 pressure is even easier to carry out. The aromatic characteristics of the products are even better controlled, especially the decrease in higher alcohols and the increase in esters. Divies and Deschamps27 demonstrated that continuous


fermentations under relatively high pressure (up to 6 bars) were possible. This new process could be used in sparkling wine manufacture with good results.

The cellular entrapment makes easier all the fermentation processes involving mixed cultures of microorganisms. Mixed cultures of lactic bacteria are obtained by this process.47

A multistep process is also possible:

-in a first reactor, the sugar content of the fruit juice is brought down with an immobilised yeast;

-in a second reactor, a low alcoholic drink is obtained using a fermentative yeast, thus we get rid of the usual undesirable sweetness of low alcoholic drinks. In that case immobilised cell reactors lead to good products with satisfactory organoleptic qualities.

CONCLUSION

The utilisation of immobilised yeasts or lactic bacteria will certainly show an industrial expansion in the next few years.

Among the immobilisation methods, the technique involving entrapped cells are of interest in the following situations:

- When high productive fermentations are wanted, cellular confinements up to 1010 yeasts/ml or 1011 bacteria can be reached. These cellular preparations allow a reproducible and controlled utilisation during long periods of time and require simple technology. Centrifugation (recycling) or filtration (clogging) steps are no longer needed.

-Microorganisms obtained from genetic manipulations can be used and the number of variants reduced.

-The entrapment method is well adapted to keep in activity complex microbial populations, e.g. different types of yeasts, and lactic bacteria.

New types of drinks are likely to appear (low alcohol flavoured drinks). The technology of entrapped systems allows fermentations without cellular loss when this loss is not sought (champagne-making).

Research on the physiology of immobilised yeasts as a function of their spatial distribution is still to be carried out. The physiology of entrapped cells is still too imperfectly known to be able to make the best use of them.

166 C. DIYlES

REFERENCES 1. Lemuel, B. and Wingard, J. R (1974). In: Advances in Biochemical

Engineering, Vol. 2, A. Fiechter (Ed.) Springer-Verlag, Berlin, pp. 1-54. 2. Chibata, I., ed. (1978). In: Immobilized Enzymes Research and

Development, Kodansha, Tokyo, p. 284. 3. Sato, K., Mori, T., Tosa, T., Chibata, I., Fumi, M., Yamashita, K. and

Sumi, A. (1975). Biotechnol. Bioeng., 17,1787-1804. 4. Vieth, W. R, Wang, S. S. and Saini, R (1973). Biotechnol. Bioeng., 15,

565-9. 5. Divies, C. (1975). Brevet ANV AR, no 7524509. 6. Divies, C. (1977). Brevet ANV AR, no 770502. 7. Larson, P.O., Ohlson, S. and Mosbach, K. (1976). Nature, 263,796-7. 8. Klein, J. and Wagner, F. (1978). First European Congress on

Biotechnology, pp. 142-64. 9. Mattiasson, B. (1983). Immobilisation Methods. In: Immobilised Cells and

Organelles, Vol. 1, Vol. 2, CRC Press, pp. 3-27. 10. Webb, C., Black, G. M. and Atkinson, B. (1986). In: Process

Engineering Aspects of Immobilised Cell Systems, Publ. The Inst. Chern. Eng.

11. Chibata, I., Tosa, T. and Fujimura, M. (1984). Immobilized living microbial cells. In: Annual Reports of Fermentation Processes, Vol. 6, T. Tasao (Ed.) Academic Press, pp. 1-22.

12. Marshall, K. C. (1985). In: Bacterial Adhesion, D. C. Savage and M. Fletcher (Eds), Plenum Press, New York and London, pp. 133-56.

13. Mitchell, R. (1976). In: Microbial Aspects of Dental Caries, Vo!' 1, H. M. Stiles, W. J. Loesche and T. C. O'Brien, Inf. Retrivial Inc., Londres, pp. 47-53.

14. Navarro, J. M. (1978). Production d'ethanol par les cellules adsorbees. In: Utilisation Industrielle du Carbone d'Origine Vegetale par Voie Microbienne, L. Bobichon and G. Durand, (Eds) , S. F. M., Paris, pp. 211-217.

15. Ghommidh, C. and Navarro, J. M. (1986). Flocculation-fermentation. In: Bioreacteurs, G. Goma (Ed.), S. F. M., Paris, pp. 89-112.

16. Stewart, G. G. and Russel, I. (1986).1. Inst. Brew., 92,537-58. 17. Hartmeier, W. and Mucke, I. (1982). Basic trials to co-immobilize living

yeast cells and glucoamylase for beer wort fermentation. In: Use of Enzymes in Food Technology. P. Dupuy (Ed.), Lavoisier, Paris pp. 519-24.

18. Goma, G. and Durand, G. (1986). Nouvelles conceptions de mise en oeuvre des microorganisms et de bioreacteurs basees sur les cultures it forte densite cellulaire. In: Bioreacteurs, G. Goma (Ed.), S. F. M., Paris, pp. 127-43.

19. Radovitch, J. M. (1985). Enzyme Microb. Techno!., 7,2-10. 20. Karel, S. F., Libichi, S. B. and Robertson, C. R. (1985). Chemical

Engineering Science, 40, 1321-54. 21. Tosa, K. and Sato, K. (1985).1. Ferment. Technol., 63,251-8. 22. Gosmann, B. and Rehm, H. J. (1986). App. Microbiol. Biotechnol., 23,

163-7.


23. Siess, M. H. and Divies, C. (1981). European J. Appl. Microbiol. Biotechnol., n, 10-15.

24. Seki, M. and Furusaki, S. (1985). J. Chem. Eng., Japan, 18,461-3. 25. Karkare, S. B., Dean, R. C. and Venkatasubramanian, K. (1985).

Biotechnology, 3, 247-51. 26. Da Fonseca, M. M., Black, G. M. and Webb, C. (1986). In: Process

Engineering Aspects of Immobilised Cell Systems, C. Webb, G. M. Black and B. Atkinson (Eds), The Inst. Chern. Eng., Warwickshire, pp. 63-74.

27. Divies, C. and Deschamps, P. (1986). ANV AR, Patent/France no 8610472. 28. Calligari, J. P., Francotte, C., de Wanuemaeker, B., Debonne, I. and

Simon, J. P. (1986). In: Process Engineering Aspects of Immobilised Cell Systems, C. Webb, G. M. Black and B. Atkinson (Eds), The Inst. Chern. Eng., Warwickshire, pp. 264-71.

29. Navarro, J. M. and Durand, G. (1977). European J. Appl. Microbiol., 4, 243-54.

30. Aiba, S., Shoda, M. and Nagatani, M,. (1968). Biotechnol. Bioeng., 10, 845-64.

31. Lee, T. H., Ahn, J. C. and Ryu, D. D. Y. (1983). Enzyme Microb. Technol., S, 41-5.

32. Ryu, D. D. Y., Kim, H. S. and Taguchi, H. (1984). J. Ferment. Technol., 62,255-61.

33. Casey, G. P. (1986). Crit. Rev. In Microbiol., 13,219-80. 34. Hahn-Hagerdal, B. (1986). Enzyme Microb. Technol., 8,322-7. 35. Doran, P. and Bailey, J. E. (1986). Biotechnol. Bioeng., 28,73-87. 36. Cavin, J. F., Prevost, H., Charlet, B. and Divies, C. (1986). Fonctionn

ement et equipment plasmidique de Streptococcus lactis subsp. diacetylactis en fermentation continue sous forme libre et incluse dans un gel d'alginate. Abstracts: ler Congres Societe Fran~ise de Microbiologie, Toulouse, p. 171.

37. De Taxis Du Poet, P., Dhulster, P., Barbotin, J. N., and Thomas, D. (1986). J. Bacteriol., 165,871-7.

38. Divies, C. an~ Cavin, J. F. (1986). In: Microorganismes et Aliments. Evolution, Maitrise, Destruction des Flores, 6, VI, pp. 43-55.

39. Gestrelius, S. (1982). Enzyme Eng., 6,245-50. 40. Cuenat, P. and Villetaz, J. C. (1984). Rev. Suisse Vitic. Arboric. Hortic.,

16,145-51. 41. Cavin, J. F. Gauvrit, I. and Divies, C. (1987). Fermentation malolactique

et production de levains en continu par les cellules de Leuconostoc oenos immobilisees. Abstracts: 12 erne colloque Societe Fran~aise de Microbiologie, Lille, p. 288.

42. Bidan, P., Divies, C. and Dupuy, P. (1978). ANVAR, Patent/France no 7822131.

43. Lenzi, P. and Cavin, J. F. (1985). (Ed) ANVAR, Patent/France no 8506147.

44. White, F. H. and Portno, A. D. (1978). J. Inst. Brewing, 84, 228-30. 45. Ryter, D. S. and Masschelein, C. A. (1985). ASBC Journal, 43,66-75. 46. Onaka, T., Inoue, T. and Kubo, S. (1985): Biotechnology, 3,467-70. 47. Prevost, H., Divies, C. and Rousseau, E. (1985). Biotechnol. Letters, 7,

247.

Chapter 12

PREPARATION OF YEAST FOR INDUSTRIAL USE IN PRODUcnON OF BEVERAGES

KNUT ROSEN

The Danish Distillers Ltd, Copenhagen, Denmark

INTRODUCTION

Yeast is an one-celled microscopic organism, in Latin it is called Saccharomyces, which means sugar fungus.

The organism has the ability to transform sugar into alcohol and carbon dioxide and utilize the liberated energy for production of new cells and for its own metabolism. This ability is utilized in wine and beer production, in alcohol production and in bread-making where the production of carbon dioxide is used for dough-rising.

Another interesting quality of yeast which was discovered by Pasteur in 1876 is the fact that the cells develop much better aerobically than anaerobically, which is nowadays utilized in all commercial yeast production.

In wine production a great number of various natural sorts of yeast are active at different stages as the alcohol content in the fermenting must rises. At least 18 genera and 147 species are involved. 1

At the beginning of fermentation, wild yeasts of the types Kloeckera apiculata, Metschnikowia pulcherrima and Torulopsis stellata slowly produce alcohol and various by-products, then the fast ferlQenting Saccharomyces species prevail and dominate the rest of the fermentation.

Pure culture wine yeasts are selected strains of yeast flora belonging to Saccharomyces cerevisiae, S. uvarum and S. bayanus. (According to the latest taxonomy they are all Saccharomyces species.)2 The use of pure culture wine yeasts began in the early 1950s when it was demonstrated that yeast, which was cultivated aerobically, could be used in anaerobic fermentations with good results. In the early 1960s

169


170 KNUT ROSEN

wine yeast in the form of moist press cakes was marketed in the USA however, the keeping quality was limited. The American company continued working with the product and after a few years a stable dry yeast was put on the market. This was an active dry wine yeast, WADY. Nowadays there are various producers of WADY, both in America and Europe.

COMPosmON OF THE YEAST

When analysing wine dry yeast the following average composition is found (%):

Dry matter Water

92-95 8-5

The composition of dry matter is as follows (%):

Carbon Hydrogen Oxygen Nitrogen Phosphorus Magnesium Calcium Potassium Sulphur Iron

48·2 6·5

33·8 6·0 1·0 0·1 0·04 2·1 0·01 0·005

From these figures a gross formula for yeast can be worked out:

C4.02H6oS02011Noo43POo03

If we assume that the yeast is produced from sugar in molasses and that 1 kg molasses contains 480 g saccharose and that this quantity yields 208 g yeast dry matter, the following gross equation for yeast production can be calculated:

0·674 x C12H220 11 + 0·43 NH3 + 3·82 O2 + 0·015 P20 S

= C4002~.502'llNo'43PO.03 + 4·81 H20 + 4·07 CO2

or

1 kg molasses + 15·2 g NH3 + 254 g 02( = 0·9 NM3 air)

PREPARATION OF YEAST FOR BEVERAGE PRODUCTION 171

giving

208 g yeast dry matter + 180 g H20

+ 372 g CO2( = 0·19 NM3) + 950 kcal.

This gross equation supplies a lot of important data. Yeast production uses molasses, NH3, air, phosphorus or phosphate (and other chemicals which have not been taken into account in the equation, first and foremost magnesium and a few trace elements and vitamins from the B-group---biotin, pantothenic acid and inositol).

Apart from yeast and carbon dioxide a certain amount of heat is produced. The carbon dioxide produced results from a complete decomposition of sugar. This shows that approximately 50% of the sugar is used as an energy source by the propagation and approximately 25% is found in the yeast as carbohydrates.

In this gross equation we have assumed that no alcohol is produced by yeast propagation. It is only 70 years since it was discovered how to carry out the cultivation to avoid formation of alcohol: the nutrients must be added slowly and in stages throughout the entire cultivation period, keeping pace with the consumption so that the aerobic conditions exist during the entire fermentation. This principle was developed and patented by S0ren Sak, who was technical manager of De Danske Spritfabrikker in Denmark, and since then it has formed the basis of yeast production all over the world.

Cultivation of yeast is carried out as a batch process and as the nutrients are added continuously, this type of process is called Fed-Batch.

The production of wine yeast on a large scale follows the principles that are used in the production of bakers' yeast.3 A more detailed description will follow.

SURVEY OF YEAST PROOUCTION

On the basis of the mass balance shown earlier a diagram for yeast production can be derived (see page 172). The basis, molasses, is diluted by water and the mixture is sterilized to destroy foreign micro-organisms. Various· impurities are precipitated during the heating and are removed by separation, and then the molasses is suitable for yeast manufacture.

Saccharomyces cerevisiae is a micro-organism proliferating by mitotic division. To ensure a pure yeast culture, the process is initiated

172 KNUT ROSEN

Molasses-Treatment of molasses -----~

Pure yeast culture-Propagation -----~

Auxiliary materials-Preparation of auxiliary materials

Air -------------~) Fermentation

water---------------~i Separation

1 Yeast cream

1 Yeast drying and packing

at the laboratory by producing a starter culture containing an adequate amount of yeast grown from one yeast cell. This quantity of yeast is transferred to the propagating plant of the factory where the cultivation is continued.

Apart from molasses, nitrogen and phosphorus must be added to the yeast. These materials are added in the form of aqueous solutions of ammonium sulphate, ammoniacal water or urea as the nitrogen source and diammonium phosphate or phosphoric acid as the phosphorus source.

The cultivation of the yeast is continued in the big fermenter in the yeast plant, adding molasses, nitrogen and phosphate compounds according to a fixed scheme. At the same time large quantities of filtered air are blown through the substrate for the respiration of the yeast. The air flow is controlled by the oxygen demand.

When the fermentation has ended, the yeast suspension is led to the separators where the yeast is concentrated into yeast cream, being simultaneously washed with water. Subsequently, the yeast cream is cooled and stored in special reservoirs. From here the yeast cream is pumped to filters where excess water is removed. After filtration, the yeast can be dried in a special yeast-drying apparatus for processing into active dry yeast which is then packed into bags and hermetically sealed.

PREPARATION OF YEAST FOR BEVERAGE PRODUcrION 173

RAW MATERIALS

As is shown in the gross equation, the main raw material is saccharose which is normally supplied in the form of molasses, being a by-product from sugar production. There are two types: beet molasses and cane molasses and they differ as shown in Table l.

Beet molasses has a low invert sugar content, a high nitrogen content, lower calcium and low biotin, whereas cane molasses is an excellent biotin source. A high calcium content may cause problems in the plate apparatus, pipes and valves and if the content is higher than

TABLE 1 COMPARISON OF BEET AND CANE MOLASSES

Beet Cane

Dry matter (%) 74-78 75 Sugar--Total (%) 48-52 48-56

Invert (%) 0·2-1·2 15-20 Raffinose (%) 0·5-2·0 Fermentable (%) 45-47 46-52 Unfermentable (%) 2-4

Organic non-sugars (%) 12-17 9-12 N-containing compounds (%) 6-8 2-3 Betain (%) 3-4 Glutamic acid (%) 2-3 Organic acids (%) 6-8 3

Gums, etc. (%) 4

Ash 10-12 10-15 (%) Na (%) 0·3-0·7 0·1-0·4

K (%) 2-7 1·5-5 Ca (%) 0·1-0·5 0·4-0'8 CI (%) 0·5-1·5 0·7-3·0 P (%) 0·02-0·07 0·6-2·0

pH 7-9 5-6

Vitamins (ppm) Biotin 0·04-0·13 1·2-3·2 Inositol 6000-8000 6000 Pantothenic acid 50-100 54-64 Thiamin 1·3 1·8 Nicotinic acid 20-45 30-80

174 KNUT ROSEN

0·3-0·5%, special precautions must be taken to remove the insoluble calcium sulphate. Beet molasses sometimes contains so much sulphur dioxide that this substance acts as toxin for the yeast and reduces the yield. With an S02 content higher than 0·05-0·1 % it will be necessary to reduce the amount, for instance by oxidation with hydrogen peroxide or potassium chlorate.

Beet molasses always contains small amounts of acetic acid but occasionally butyric acid is found, caused by infection during the sugar production. This acid is toxic to the yeast in quantities in excess of 0·1 %. There exists no easy way to remove it and one may be compelled to mix such an 'infected' molasses with molasses from a wellfunctioning sugar factory.

Nitrite is also toxic to yeast and the maximum content allowed is 0·005%. Of the nitrogenous compounds in beet molasses the amino acids, making up approximately one-third, can be utilized by the yeast. The rest (and for cane molasses the whole amount) of the necessary nitrogen must be supplied by addition of easily assimilable compounds such as ammonia, ammonium sulphate or urea.

In addition phosphate must be added and this is ,usually done in the form of diammonium hydrogen phosphate or phosphoric acid. Previously superphosphate was often used, this chemical being cheaper, and it is still used in some countries. However, superphosphate often contains considerable amounts of fluorine which is toxic to the yeast, so the use of this material should be avoided, if possible.

is: For all chemicals the maximum content of toxic impurities allowed

Fluorine Arsenic Lead

400mg/kg 20mg/kg 2mg/kg

The very important question of organic growth factors should be mentioned in regard to raw materials. The usual yeast strains require addition of various vitamins of the B-group, a fact which has been used in the so-called biosnumbering for grouping the yeast strains into a number of classes. The strains most often used for producing wine and bakers' yeast require addition of biotin, pantothenic acid and inositol to give optimum yields. Molasses contains some of these substances as mentioned already and the following figures show the


contents compared with the demand:

Beet Cane Yeast dry ppm molasses molasses matter

Biotin activity 0·04-0·13 1·2-3·2 0·9 Pantothenic acid activity 50-100 54-64 165 I~ositol activity 6000-8000 6000 3700

For determination of these activities a microbiological method is used but the test organism is the yeast strain used for the production of wine yeast or bakers' yeast. Thus the figures include precursors and compounds having the same growth-promoting effect.

If the molasses does not contain a sufficient content of these growth factors it may become necessary to supplement by artificially-produced vitamins or to use mixtures of beet and cane molasses. In this case it must be remembered that similar substances or precursors can often supply the demand. In many cases the cheaper D.L. desthiobiotin can substitute for D-biotin and tJ-alanine can be used in lieu of pantothenic acid.

Magnesium is of great importance for the enzymes involved in glycolysis. A lack of this means the pyruvate is converted into acetaldehyde which is blown away with the exhaust air and thus the yield and quality of the yeast is essentially reduced.

The processing water must be of potable water quality according to the demands of WHO (see Table 2).

Process air is filtered so that the air is sterile on admission to the compressor. Normally a filter built up from three components is used:

1. Coarse filter separating approximately 80-85% of the particles in the air.

2. Fine filter removing up to approximately 93-97% of the particles.

3. Micro-filter separating 99·97% of particles of 0·3/-lm.

The filter material used is fibre glass and control of the efficiency can be carried out by the so-called sodium flame test where salt particles are atomized in front of the filter and the effect is registered by a flame behind the filter.

176 KNUT ROSEN

TABLE 2 SPECIFICATION FOR WATER USED IN PROCESSING

Total haldness

Iron (maximum) Manganese (maximum) KMn04 used by organic

substances Chloride Nitrate Sulphate

15°dH = 150 mg/litre CaO = 26·8FH

0·1 mg/litre 0·05 mg/litre

lOmg/litre 250 mg/litre 50mg/litre 500 mg/litre

The water must be free of organic substances

Microbiological factors: Total count-21°C

tluorescents Total count-37°C Total coliforms Total E. coli Total Clostridium perfringens

100 bacteria per ml 5 bacteria per ml 10 bacteria per ml o per 100 ml o per 100 ml o per 100 ml

EQUIPMENT IN A YEAST PLANT

In the following sections the equipment in a yeast factory is described.

Molasses Storage Plant The molasses is pumped into a raw-molasses storage tank which should be provided with a heating coil so that the molasses can be kept at a suitable temperature, but not over 40°C, guaranteeing a reasonable viscosity for pumping. All molasses pipes in the molasses tank farm and to the production plant should be heated. The molasses is pumped from the storage tank to a buffer tank.

Molasses Treatment Plant This part comprises buffer tanks, a sterilizing unit and a sludge centrifuge. (Fig. 1). A metering pump with three pistons transports the molasses, water and sulphuric acid to a mixing tank so that quantity, dilution ratio and pH may be controlled by varying the speed and the cylinder volume of each piston.

The resulting wort is then led to a sterilization plant consisting of: a plate heat exchanger and a so-called 'palarisator' with steam injection in a pressure tank, sterilization cell, vacuum flash cooling tank and the

178 KNUT ROSEN

necessary pumps, etc. Here the wort is sterilized by heating to 140°C for a few seconds and is then cooled, first under vacuum and then by passing through the plate heat exchanger in countercurrent with the inlet of molasses wort. The sterile wort is finally pumped to a sludge centrifuge and the clarified wort is then led to a wort tank which is insulated as the wort enters at a temperature of about 75°C.

If the molasses has a very high content of calcium it may be necessary to remove the majority of this as calcium sulphate.

Propagation Plant In order to obtain the best yeast quality and bacteriological purity it is essential that the fermentation plant, in which the sterile laboratory pure culture of approximately 200 g yeast is propagated to approximately 300 tons yeast, is built as suitable as possible.

The method of single yeast cell production was initially systematized by Mr Emil Chr. Hansen, a microbiologist, about 100 years ago at the Danish Carlsberg brewery.

The propagation plant comprises a vat of approximately 550 litre and a vat of approximately 12 000 litres. All vats are made of stainless steel and equipped with cooling jacket and all necessary apparatus (Fig. 2).

The propagation of the yeast is started by the separation of a pure culture in the laboratory. The culture is used as inoculum in a sterile culture medium in the small propagation vat and when the fermentation has finished, the liquid containing the yeast is conducted to the large propagation vat in which a sterile culture medium has been prepared in advance. While the fermentation in the small vat takes about 20 h, wort in the large vat is fermented after about 12 h, after which the contents are pumped to one of the large fermenting vats.

The two vats are placed in a special room where all necessary precautions have been taken to avoid microbial contamination from bacteria, wild yeast and fungi.

Yeast Vats The main features in yeast production are the large fermenting vats. Each has a total volume of up to 200 m3 and is equipped with a sectional cooling jacket so that cooling can take place by means of a closed cooling circuit and a refrigerating compressor plant.

Each vat is equipped with a stationary aeration system with distributing pipes over the whole bottom, two manhole covers, sight


CW

I I 1 ":rw <l

I I I I I --- h_h ---rh------~-l~II--L

A

r--' I I I

B

FIG. 2. Propagation plant.

glass, light fitting, vent pipe to the open air as well as all necessary pipe connections for instruments and inlets. The air, which is blown to the fermenting vat through the aeration systems, is compressed in centrifugal compressors with continuously variable flow regulation without loss of efficiency. The air is sucked through filters so that it becomes completely pure and sterile as described previously.

All necessary internal piping for the plant including necessary pumps, vents, etc. from the tanks to the separation plant, is supplied in stainless steel.

FERMENTATION OR YEAST CULTIVATION

The initial steps of yeast cultivation take place in the laboratory in a laminar flow bench. Colonies with a diameter of 1-2 mm have been

180 KNUT ROSEN

grown in a petri dish from isolated cells of the required strain. From such a colony the cultivation is started, first in test tubes, later in flasks and after one week a 200 g pure starter culture in a Carlsberg flask is produced. This is used as inoculum into the first vat (the 550-litre vat), of the propagation plant. In this vat a substrate has been prepared consisting of 40 kg molasses, nutrient salts, and up to approximately 300 litres of water. The fermentation takes about 20 h under light aeration and results in about 3 kg yeast dry matter and 9-10 litres of alcohol.

All of this quantity is used as seed yeasts in the second vat of the propagation plant, with approximately 12000-litre capacity. In this vat is about 1000 kg molasses in 9000 litres of liquid and after about 12 h fermentation under light aeration this liquid contains about 75 kg yeast dry matter and about 250 litres of alcohol.

The next step is that this liquid is transferred under sterile conditions to one of the big yeast vats where an aerobic, fed-batch fermentation takes place with the addition of molasses wort, nitrogen and phosphorus. After about 26 h this fermentation has ended and, after separation, the result is 6 tons of seed yeast dry matter in the form of about 27 m3 yeast cream. The cream is cooled and can now be used as seed yeast for 6-7 final fermentations.

A final fermentation is carried out as follows. The selected yeast vat is cleaned and sterilized with steam before the fermentation is started.

At the beginning process water, diammonium phosphate, magnesium sulphate, and a small amount of molasses are added automatically when activated from the control system. A portion of seed yeast is added after treatment with sulphuric acid at pH 2, air is blown into the wort, and molasses wort is added from the wort tanks concurrently with the yeast growth in the fermenter. At the same time ammoniacal water is added as a nitrogen source, while the pH is controlled by means of a sodium carbonate solution and sulphuric acid.

The heat generated by the yeast growth is removed through the cooling jacket. When the calculated amount of molasses wort has been added and assimilated, the fermentation is terminated. The fermenters are equipped with instruments for measuring the process parameters mentioned in Fig. 3.

Molasses: The flow of wort is measured and controlled. The set point is changed according to the actual fermentation scheme which has been keyed into the control system.


L

.!! I1J ~ 01 .S o o u

Fermentation control

pH

Foam

FIG. 3. Fermenter.

Foam sensor

pH

Nitrogen: A solution of ammoniacal water (25%) is added as a nitrogen source to the fermentation. The admission is controlled in the control system as a constant percentage of the molasses added.

pH: The pH value in the fermenting wort is measured and controlled by means of a 20% sodium carbonate solution and sulphuric acid.

Air: The quantity of air is controlled according to a programme in the control system at the beginning of the fermentation cycles whereafter the addition of air is controlled by means of an oxygen potential meter.

Temperature: The temperature in the fermenting wort in all phases

182 KNUT ROSEN

is measured and controlled to a pre-set value by admission of cooling water to the cooling jacket.

Anti-foam (foam level control); Anti-foam agent is added to the fermenter when the foam has covered a foam electrode for a certain time. An extra foam electrode has been built in for emergency supply. On the control panel a signal is given when anti-foam agent is added and if the addition becomes continuous, an alarm sounds.

YEAST SEPARATION

The fermented wort containing the produced yeast is pumped to the separation plant. This plant consists of a strainer followed by three separators in series which are fed with 90 m3/h of wort. After each of the first two separators follows a washing injector to which wash water is conducted.

The separated wort and wash water are collected and pumped to the sewage system. The yeast appears as a cream which is pumped through a plate heat exchanger and cooled to 2-3°C in counterflow with ice-water at + 1°C. After cooling, the yeast cream is conducted to the yeast cream plant which consists of stainless steel tanks of 30 m3 ,

insulated and with a stainless steel covering jacket on the lower cylindrical part for forced circulation of cooling water of + 1°C maintaining a temperature of 3-4°C for the yeast cream. As the yeast cream reaches the tank at a temperature of 3-4°C the cooling should only remove the heat developed during standing.

The vats are equipped with agitator, manhole cover, pipe connections and flanges for instruments. Moreover, all necessary internal piping and pumps for pumping of yeast cream are supplied in stainless steel. Each yeast cream vat can contain the yeast produced from one fermentation.

The excess water is removed by filtration in filter presses or vacuum filter drums, in both cases resulting in a filter cake having about 30% dry matter.

ACTIVE DRY WINE YEAST

The wine yeast produced at the beginning of the process is recovered as filtered yeast cake, but the shelf life of this yeast is relatively limited,


3-4 weeks. Therefore, it has always been desirable to be able to dry the yeast in a way that would extend the keeping quality without injuring the fermenting power significantly. The first experiments were carried out as tray drying; the yeast was distributed in small lumps or granules on a tray with a perforated bottom and hot air was blown through.

The method was further developed by using belt dryers where the yeast was extruded as strands about 2-4 mm in diameter on a belt of perforated steel plate. This belt was slowly passed through drying zones with different drying conditions, temperatures and air humidity and after 10-15 h the discharge end was reached where the product, which had now been dried to about 8% water, was collected, led to a silo, and was then ready to be packed.

Another way of drying consists of filling the strands into a rotolouvre dryer in which the yeast is dried in a rotating drum. The resulting yeast appears in the form of small smooth pellets.

When the yeast is to be used, it has to be rehydrated. This takes place by pouring the yeast into water or a mixture of grape must and water which has been heated to about 4O-43°C and it takes 10-15 min.

Yeast cream

EmUlsifier tank system

, I \ , (-) ~ \ j I I ,,! I /

" V [ __ L.L-,

Vacuum filter

Extruder Instant active - - - - - - - - - - - --- ---..

Ir----'---:F:::-IU--:-id:-:b-e--:-d --------,I dry yeast

FIG. 4. Yeast drying plant.

184 KNUT ROSEN

At lower temperatures the re-establishment of the cell membranes takes longer and this means that contents from the cells will more easily leak into the liquid. At higher temperatures the yeast will be killed.

About 15 years ago a new type of dry yeast was developed, the instant active dry wine yeast. By using fluid-bed drying and the addition of about 1 % of emulsifier a product is produced that can be used directly. To obtain the best result (i.e. the highest number of living yeast cells), it is recommended that it is rehydrated as described above.

The yeast is extruded in a specially designed machine which extrudes the yeast in thin strings with a diameter of about 0·5 mm. The extrusion takes place directly into the 'fluid-bed' drying plant. The yeast stays in the drying plant for 30-60 min, and the drying is performed very carefully by blowing warm, dry air through, so that the yeast particles are not exposed to temperatures which will damage the yeast (Fig. 4).

WASTE WATER

By cultivation of yeast, carbohydrates and assimilable nitrogen compounds are removed from the molasses but there is still about 20% organic matter left in the waste water, the separated wort and the wash water.

A yeast factory having a production of 6000 tons of yeast dry matter per year pollutes as much as a town with about 220000 inhabitants measured as BOD. However, the nitrogen and phosphorus discharged are significantly lower, corresponding to 50 000 and 12 000 inhabitants respectively.

Accordingly, it will be necessary in most cases to undertake a purification, either by irrigation or by biological treatment in the purification plant of the town (and this plant would benefit from the mixing of yeast waste water and household waste water as there is normally a deficit of BOD compared with nitrogen and phosphorus in household waste water) or in a special plant belonging to the yeast plant. In this the effluent may be purified by an anaerobic fermentation in which the organic substances are decomposed to methane and carbon dioxide followed by an ordinary aerobic (active sludge) treatment.

PREPARATION OF YEAST FOR BEVERAGE PRODUCfION 185

SPECIFICATIONS AND ANALYSES

Demands on 'Selected Yeasts' In some wine-producing countries specifications are defined for yeast used in wine production. As an example, the following demands from France and Italy can be mentioned:

France: Living cells by cultivation on malt-Wickerham Other yeast strains Moulds Total bacteria, aerobic plating Water content

Italy:

109/g 0·01 % of living cells l/mg 105/g Max. 8%

WADY must not contain pathogenic organisms or their toxins and it must not be possible to demonstrate the presence of coliform bacteria, faecal Streptococci, or sulphite-reducing Clostridia in 0·1 g.

Analyses The following items are checked in the normal control of dry yeast for wine production:

(1) Dry matter: The dry matter content of dry yeast is normally above 92%. The method of fluid-bed drying produces about 95% dry matter.

(2) Number of cells: The number of living cells is defined as number of cells per gram. The number will of course vary according to the size of the cells, i.e. according to the content of reserve carbohydrates.

If the yeast has a high content of trehalose, etc. and a correspondingly low content of protein, the cells are comparatively big and there are fewer cells per gram than in cases where the content of reserve carbohydrates of the cells is lower. For this reason the total number of cells varies from 10-40 x 109 cells per gram.

(3) Number of living cells: Two different methods are used: (a) plating on a malt-extract agar or (b) dyeing with methylene blue and counting of cells which have not absorbed the dye.

There are differences between the two methods. If the cells agglomerate, several cells will-according to (a)-make only one colony, and the result is therefore too low. When using (b) it may be

186 KNUT ROSEN

difficult to distinguish the cells, which cannot propagate under the existing circumstances, from the cells which are dead.

In practice, however, it appears that (b) normally yields quite good results which can be of practical use.

(4) Fermentation activities: The fermentation activity is often used as a measure of the activity of a wine dry yeast. It is determined by measuring the rate of fermentation in a grape juice with an admixture of S02. Suitable conditions for the mixture:

Grape juice, pH 3·6 S02,68ppm Yeast, 0·11 g WADY/litre = 11 g/hl Incubation at 28°C Loss of weight determined after 24, 48 and 72 h. Mter one week the residual sugar content is determined.

The rate of fermentation is calculated as gC02/24 h/litre from the analysis value.

TABLE 3 RESULTS OF ANALYSIS

Saccharomyces S. cerevisiae bayanus

Montrachet Pasteur

Chemical composition Dry matter (%) 91·9 91·6 Nin DM (%) 6·32 5·76 P20 S in DM (%) 1·93 1·67 Gassing power Without S02 (%) 80 36 With S02 (%) 40 26 Fermentation power Weight loss 24 h (g) 1·8 0·6

48 h (g) 5·6 5·7 72h (g) 7·0 7·0

Residual sugar 0·13 0·14 Activity 23·3 23·3 Living cells Plating x 109 40 31 Counting x 109 38 21 % dead cells 18 33 Bacteriology Total count 9xl~ 2x 103

Wild yeast 20 10


Primarily, this method demonstrates the qualities of the yeast strain and is a suitabl~ method for checking that there has not been any change during the yeast production.

(5) Gassing power: In this case the COz-development is determined in a mixture of grape juice and water with and without S02 (43 ppm S02) and with 1·7 g wine yeast per litre at 30°C.

The CO2-development is measured after 2! h and expressed in per cent of a standard dry yeast. Compared with (1) this method indicates to a higher degree the activity of the yeast sample and is therefore important for the yeast producer as a control of the quality of the yeast produced.

(6) Infecting micro-organisms: The wine dry yeast produced is of course analysed for the content of infecting micro-organisms, as 'total count' of bacteria, wild yeast, mould and pathogenic bacteria such as coliform bacteria, especially E. coli, faecal Streptococci, sulphitereducing Clostridia, Salmonella, etc.

Table 3 shows the results of analysis of two different wine yeasts.

CONCLUSION

It can be concluded from the literature that the use of pure culture gives a quicker start for fermentation, a better control and uniformity and a better stability of the wine produced.

Some authors have noted that you will not get the quite special taste sensation which can be produced by the wild yeast flora at the beginning of the fermentation, but which however, may give the wine a totally undesirable aroma when the fermentation is abnormal.

Experience from many years of production proves that the use of the pure culture of a wine yeast yields a first-class wine with a stable quality and a good taste, which ensures the wine producers obtain a satisfactory product.

REFERENCES

1. Kraus, J. K., Reed, G. and Villetaz, J. C. (1983). Connaissance de La Vigne et du Yin, 17,93 and (1984) Connaissance de La Vigne et du Yin, 18, 1.

2. Barnett, J. A., Payne, R. W. and Yarrov, D. (1983). Yeasts, Characteristics and Identification. Cambridge University Press, p. 467.

3. Rosen, K. (1987). In: Yeast BiotechnoLogy, ed. D. R. Berry, I. Russell, and G. G. Stewart (Eds), Allen and Unwin, London, p. 471.

Chapter 13

ENZYMES IN THE FRUIT JUICE INDUSTRY

G. LANZARINI and P. G. PIFFERI

Department of Industrial Chemistry and Materials, University of Bologna, Bologna, Italy

INTRODUCTION

Amongst all biotechnological applications the percentage relative to the food sector and agricultural chemistry will, in the very near future, be more than 25 %.

At present enzymes are normal processing aids and are increasingly employed in the production of fruit and vegetable juices; their function as biological catalysers of specific reactions, under mild pH and temperature conditions and at low concentrations, is strictly related to the search for new processes and products. 1,2

The production of fruit and vegetable juices involves the transformation of organized, whole, solid tissue into a semifluid system of cells and fragments of plant cell walls suspended in cellular liquid.

In the production of fruit and vegetable juices, enzymes are used with two aims:

(a) in the treatment of crushed pulp: to increase the yield in juice, to reduce the processing time, to improve the extraction of some components (aroma and colour) and to obtain the partial or total liquefaction of the plant tissue (homogenized products, nectars, concentrates of pulpy juices and nectars);

(b) in the treatment of juice to: -lower viscosity and facilitate more or less forced

concentration; -clarify to obtain clear juices for concentration; -increase filtering speed during finishing treatments of clear

juices; 189


TA

BL

E 1

E

NZ

YM

ES

AN

D M

ICR

OO

RG

AN

ISM

S U

SE

D I

N T

HE

FR

UIT

JU

ICE

AN

D C

ITR

US

FR

UIT

JU

ICE

IN

DU

ST

RY

Enz

yme

Cla

ssifi

catio

n

Pec

tine

ster

ase(

PE

) E

C3.

1.1.

11

End

opol

ymet

hylg

alac

turo

nase

(en

do-P

MG

) E

C3.

2.1.

41

End

opol

ygal

actu

rona

se( e

ndo-

PG

) E

C3.

2.1.

15

Exo

poly

met

hylg

alac

tura

se

EC

(unc

lass

ifie

d)

Exo

poly

gala

ctur

onas

e (e

xo-P

G)

EC

3.2.

1.40

End

opec

tinly

ase

(end

o-P

L)

EC

4.2.

2.10

End

opol

ygal

actu

rona

tely

ase

(end

o-P

AL

) E

C4.

2.2.

2

Exo

pect

inly

ase

(exo

-PL

) E

C (u

ncla

ssif

ied)

E

xopo

lyga

lact

uron

atel

yase

( exo

-P A

L)

EC

4.2.

2.9

End

o-,8

-(ce

llula

se C

X)

gluc

anas

e E

C3.

2.1.

4

Exo

-,8(c

ellu

lase

C1)

glu

cana

se

EC

3.2.

1.91

End

ogal

acta

nase

E

C3.

2.l.

S9

Cat

alys

ed r

eact

ion

Hyd

roly

sis

of

este

r bo

nd

Hyd

roly

sis

of a

-1,4

bo

nds:

of

pect

in

of p

ecti

c ac

id

of p

ecti

n of

pec

tic

acid

Tran

s-el

imin

atio

n o

f a-

1,4

bond

s:

of p

ecti

n

of p

ecti

c ac

id

of p

ecti

n of

pec

tic

acid

H

ydro

lysi

s o

f ,8-

1,4

bond

s:

of c

ellu

lpse

of c

ellu

lose

of g

alac

tane

s

Mec

hani

sm

Exo

-mec

hani

sm

Ran

dom

mec

hani

sm

Ran

dom

mec

hani

sm

Ran

dom

mec

hani

sm

Exo

-mec

hani

sm

Exo

-mec

hani

sm

Ran

dom

mec

hani

sm

Ran

dom

mec

hani

sm

Exo

-mec

hani

sm

Exo

-mec

hani

sm

Ran

dom

mec

hani

sm

Exo

-mec

hani

sm

Ran

dom

mec

hani

sm

Fun

ctio

n

Self

-cla

rifi

cati

on o

f ju

ices

or

wit

h en

do-P

G

Pul

p tr

eatm

ent b

efor

e ju

ice

pres

sing

Ju

ice

clar

ific

atio

n (i

.e.

appl

e) f

ruit

/veg

etab

le

mac

erat

ion,

red

ucti

on in

vi

scos

ity

of

juic

es to

be

conc

entr

ated

Pre

-pre

ssin

g pu

lp-t

reat

m

ent

Self

-cla

rifi

cati

on o

f ju

ices

or

wit

h en

do

PG

/PE

mix

ture

M

acer

atio

n o

f ve

geta

bles

Pul

p tr

eatm

ent

befo

re

pres

sing

and

pre

di

ffus

ion

Tot

al ft

uidi

fica

tion

of

frui

ts a

nd v

eget

able

s;

prod

ucti

on o

f nec

tar

conc

entr

ates

E

lim

inat

ion

of

sedi

m

ents

in c

lear

con

cent

rat

ed ju

ices

Hyd

roly

sis

of a

-l,5

bo

nds:

E

ndo

arab

inas

e E

C (

uncl

assi

fied

) o

f ar

abin

ans

Ran

do

m m

echa

nism

E

lim

inat

ion

of

sedi

men

t in

cle

ar c

once

ntra

ted

juic

es

Hyd

roly

sis

of a

-l,4

bo

nds;

a-

amyl

ase

EC

3.2.

1.1

of

star

ch

Ran

do

m m

echa

nism

P

reve

ntio

n o

f st

arch

se

dim

enta

tion

~-amylase

EC

3.2

.1.2

o

f st

arch

E

xo-m

echa

nism

in

cle

ar a

pple

jui

ce

Pro

tein

ase

Hyd

roly

sis

of

prot

eins

P

re-p

ress

ing

frui

t pu

lp

to p

epti

des

and

trea

tmen

t, p

rodu

ctio

n o

f am

inoa

cids

ho

mog

eniz

ed t

ropi

cal

frui

ts

Glu

cose

-oxi

dase

E

C 1

.1.3

.4

Oxi

dati

on o

f ~-glucose

Red

ox w

ith

O2

Ant

i-ox

idan

t ac

tion

in

to g

luco

nola

cton

e co

nsum

ptio

n ju

ices

a-

L-r

ham

noox

idas

e E

C3

.2.1

.40

H

ydro

lysi

s o

f na

ring

in

Red

ucti

on o

f bi

tter

tas

te

to p

runi

n o

f gr

apef

ruit

~-D-glucosidase

EC

3.2.

1.21

H

ydro

lysi

s o

f pr

unin

to

Exo

-mec

hani

sm

nari

ngen

in

Imm

obil

ized

mic

roor

gani

sms

Bac

teri

a A

rthr

obac

ter

Tra

nsfo

rmat

ion

of:

li-H

ydro

lysi

s an

d re

dox

Red

ucti

on o

f bi

tter

m

onin

17-

tast

e in

ora

nge

and

glob

ifor

mis

I de

hydr

olim

onoa

te A

-le

mon

jui

ce

lact

one

nom

ilin

to

17-

dehy

dron

omil

inoa

te A

-la

cton

e B

acte

ria

Art

hrob

acte

r gl

o-T

rans

form

atio

n o

f lim

-R

edox

bi

form

is II

onin

to

lim

onol

, an

d no

mil

in t

o no

mil

ol

Bac

teri

a C

oryn

ebac

teri

um

Tra

nsfo

rmat

ion

of

lim-

Red

ox

fasc

ians

on

in t

o li

mon

ol,

nom

i-li

n to

oba

cuno

ne

Ace

tyl-

lyas

e

192 G. LANZARINI AND P. G. PIFFERI

-obtain chemico-physical stabilization of the JUlce, to prevent formation of precipitates, improve organoleptic features or prevent their alteration from present quality standards.

The large number and variety of commercial preparations is a disadvantage, as it often results in lack of precise knowledge about the enzymatic activity of these products, their specific application and the side-effects produced by the presence of other enzymatic activity. Table 1 shows the main enzymes used in the production of fruit juices at an industrial or research level.

CELLULAR STRUCTURE AND JUICE EXTRACTION

The principles of the basis of extraction of juice from the cell3,4 and the partial or total ftuidification of the tissue and the plant cell,5,6 are given in Table 2. The cellular structure consists of (a) the cell wall, with a composition which varies in the same fruit during ripening. Fifty percent of the weight, or more, is made up of cellulose and hemicellulose. There is also a remarkable amount of pectin with a high degree of esterification, a decisive factor for fruit firmness; (b) the cytoplasm, stratified on the cell wall, containing the nucleus, the plastids, the enzymatic proteins and growth factors; and (c) the vacuole, in which sugars, acids, salts, polyphenols and pigments are dissolved.

The cell wall consists of individual segments which are thus interconnected; the cellulose fibrils are connected by means of xyloglucan chains to the pectin, which in its turn is connected by means of arabi nogal act an to the protein serine residues of the cell wall.

FRUIT AND VEGETABLE ENZYMES

Three large groups of plant enzymes may be distinguished:

(1) Hydrolases: saccharases, proteases, phosphatases and pectolytic enzymes;

(2) oxidoreductases: peroxidases, catalases, glucosidases and polyphenoloxidases;

(3) cellulases.

TA

BL

E 2

E

LE

ME

NT

S R

EL

AT

IVE

TO

RE

LE

AS

E A

ND

EX

TR

AC

TIO

N O

F J

UIC

E F

RO

M T

HE

CE

LL

S A

ND

IT

S S

EP

AR

AT

ION

Pri

ncip

le o

f ju

ice

ex

trac

tion

fro

m t

he c

ell

Pri

ncip

le o

f ju

ice

sepa

rati

on

Cru

shin

g an

d pr

essi

ng

Mec

hani

cal

rupt

urin

g o

f ce

ll w

alls

an

d m

embr

anes

(re

sult

s o

f cu

ttin

g,

pres

sing

, st

retc

hing

, ab

rasi

on).

N

eces

sary

tre

atm

ent w

ith

pect

olyt

ic

and

cell

ulol

ytic

enz

ymes

Pre

ssin

g an

d fi

lter

ing

soli

d re

sidu

es

usin

g fi

lter

-pre

ss

Pro

cess

Ext

ract

ion

Mod

ific

atio

n o

f al

l m

embr

ane

perm

eabi

lity

(he

at d

enat

ura

tioJ1

); in

sign

ific

ant

mec

hani

cal

rupt

ure.

Oth

er p

heno

men

a:

acti

vati

on o

f pec

toly

tic

enzy

mes

, de

tach

men

t o

f m

iddl

e la

mel

la,

gela

tini

zing

of

star

ch,

and

'fus

ion'

of

pect

ins,

exp

ansi

on o

f ai

r in

inte

rcel

lula

r sp

aces

, ce

ll de

tach

men

t E

xtra

ctio

n by

: di

ffus

ion

wit

hin

the

phas

es,

pass

age

of

solu

tes

to

the

phas

e li

mit

sur

face

s by

liqu

id

exch

ange

in

alm

ost

osm

otic

pro

ce

sses

an

d b

y hy

dros

tati

c pr

es

sure

; us

ing

infl

ow o

f m

echa

nica

l ap

para

tus

Flu

idifi

catio

n

Sol

ubil

izat

ion

of c

ell

wal

ls

and

mem

bran

es b

y en

zy

mat

ic a

ctio

n

Sie

ving

of

non-

degr

aded

m

ater

ial

(sto

nes,

see

ds,

etc)

af

ter

flui

difi

cati

on o

f ti

ssue

TA

BL

E 3

CL

ASS

IFIC

ATI

ON

OF

DEP

OLY

MER

IZIN

G P

EC

fIC

EN

ZYM

ESa

Rea

ctio

n EL

ectiv

e su

bstr

ate

Mod

e o

f act

ion

Nam

e E

CN

o.

Syst

emat

ic n

ame

Pol

ygal

actu

rona

te

Ran

dom

E

ndo-

poly

gala

ctur

onas

e 3.

2.1.

15

Pol

y(1,

4-a-

D-

gala

ctur

onid

e)

glic

anoh

ydro

lase

H

ydro

lysi

s P

olyg

alac

turo

nate

E

xo-m

echa

nism

E

xo-p

olyg

alac

turo

nase

3.

2.1.

67

Pol

y(1

,4-a

-D-g

alac

turo

nide

) ga

lact

uron

ideh

ydro

lase

P

olyg

alac

turo

nate

E

xo-m

echa

nism

E

xo-p

oly-

a-D

-3.

2.1.

82

Pol

y(1,

4-a-

D-

gala

ctur

onox

idas

e ga

lact

uron

ide)

di

gala

ctur

onhy

drol

ase

Pol

ygal

actu

rona

tes

Ran

dom

P

ecta

te ly

ase

4.2.

2.2

Pol

y(1,

4-a-

D-

Oll

goga

lact

uron

ate

lyas

e ga

lact

uron

ide)

lya

se

Oli

gola

ctur

onat

es

Exo

-mec

hani

sm

4.2.

2.6

Oli

goga

lact

uron

ide

lyas

e fj

-c1e

avag

e P

olyg

alac

turo

nate

s E

xo-m

echa

nism

E

xo-p

ecta

te ly

ase

4.2.

2.9

Pol

y(1,

4-a-

D-

gala

ctur

onid

e) e

xo-I

yase

P

ecti

nate

s R

ando

m

Pec

tin

lyas

e 4.

2.2.

10

Pol

y(m

etho

xy-

gala

ctur

onid

e) l

yase

a A

s pe

r th

e N

omen

clat

ure

Com

mit

tee

of

the

Inte

rnat

iona

l B

ioch

emis

try

Com

mis

sion

, A

cade

mic

Pre

ss.

(197

9).

ENZYMES IN THE FRUIT JUICE INDUSTRY 195

In the hydrolase group, the pectic enzymes have a prominent role due to their function in the modification of pectic substances, which are partly responsible for the organoleptic features of the product and are used in numerous technological processes.

PECTIC SUBSTANCES AND PECTIC ENZYMES

The pectic substances are structural heteropolysaccharides, mainly found in the middle lamella and in the primary parenchymatic cell wall of the pulpy fruit mesocarp, with a molecular weight of 30000-400 000 daltons. They consist of 1,4-a-D-galacturonan chains partially methoxylated. A certain number of L-rhamnose units are bound by C1 and C2 carbon atoms to the main chain; side chains of neutral sugars, mainly galactose, arabinose and xylose are bound in a covalent manner to the C2 or C3 carbon of the galacturonidic unit or to the C4 carbon of rhamnose. The degree of esterification with methanol varies according to the fruit and the degree of ripeness (e.g. apple 85-90%, grape 44-65%, orange 47-60%). There are a wide variety of pectic enzymes in nature which can be classified into two large groups (Table 3). Figure 1 shows the attack of the pectic enzyme on the pectic chain.

ROLE OF ENDOGENIC PECTINASES IN THE PRODUCTION OF BEVERAGES

The presence of PE in citrus fruits is responsible for the 'break' of turbidity in the respective fruit juices.7,s This strong pectin esterasic activity, if it is not inhibited by heating or freezing, de-esterifies the

FIG. 1. Attack of the pectolytic enzymes on the pectic chain.


native pectin making it coagulate with the calcium ions present in the fruit juice and forming flocks which settle out. In the case of concentrated fruit juices, calcium pectate jellies are formed from which the juice cannot be reconstituted. It has been shown9 that there are at least two forms of PE in orange juices: one of these is more easily inhibited than the other by the presence of oligomers. This explains the partial success obtained in the attempt to clarify citrus fruit-juice by adding pectate10 or oligomers with a low degree of polymerization. s-u

The enzymatic autoclarification of citrus fruit-juice is a practice adopted by small family-run farms in SicilyY In tomato juice, obtained either by the cold or the hot pressing process, there are large quantities of PG and PE.13 During the cold process there is an interval between the disintegration and pasteurization to allow the PE to convert the highly esterified pectins into material with a low methoxyl content so that the PG can act. In this way a product suitable for subsequent concentration into a sauce is obtained. Whereas during the hot process, the juice is treated rapidly to maintain a high viscosity, a feature that the consumer expects of the product.

Both PE and PG are present in grapes. The presence of PE, which brings about the release of methanol, varies a great deal in quantityl4-16 and in form. 9 As regards the constituent parts of the grape berry, the PE activity is almost all concentrated in the skin. The combined effect of ethyl alcohol and polyphenols drastically reduce its activity17 in wines.

SOLUBLE ENZYMES IN THE PRODUCTION OF FRUIT JUICES

Treatment of Pulp with Pectic Enzymes Due to the raw material price trends, the increase in the cost of labour, investment and maintenance, constant attention has been given to the search for better methods of transforming fruit. The attempts to obtain a higher juice yield and fruit-juices with a higher soluble solid content, have witnessed the improvement in or conception of new machinery for the mechanical recovery of juice, the development of different extraction methods, the combination of pressing-extraction or the repeated pressing of crushed pUlp,1S-24 or the enzymatic treatment of the latter before extraction of the fruit juice6,25-27 and/or residual solids.


The fruit-juices are obtained by mechanical processes which separate the liquid phase from the residual solids by pressing, centrifuging, filtering, and then, if necessary, clarifying and concentrating, pasteurizing or freezing. The pectic substances with the most significant technological role account for 0·5-4% of the weight of fresh material. When the tissue is ground, the pectin divides itself between the liquid phase (soluble pectin) causing an increase in viscosity and the pulp particles where other pectin molecules remain bound to the cellulose fibrils by means of side chains of hemicellulose and thus facilitate water retention.

In pectin-rich fruit mechanical crushing gives a fruit juice with a high viscosity which remains bound to the pulp under the form of a jellified mass. It is difficult to extract this juice by pressing or with other mechanical methods; with added pectinases the viscosity of the fruit juice drops, the pressability of the pulp improves, the jelly structure disintegrates, and the fruit juice is more easily obtained and with a higher yield; moreover, due to the breakdown of the skin tissue, a diffusion of pigments in the fruit juice is obtained. This treatment with pectinase, essential for berry fruits and others with soft pulp, already used for the extraction of blueberry, raspberry and strawberry juice,28,29 of cherry30,31 and other stone fruits32,33 was introduced in the 1970s to apple juice,25,26.34 extraction technology. The most widely used pectinases are those recommended for the clarification of fruits and must be capable of rapidly degrading highly esterified pectins; in other words PE and PG and/or PL which solubilize pectin from the cell wall and facilitate extraction of the cell contents. The improvement in the pulp pressing and the increase in fruit juices yield are due to the degraded pectins' lost capacity to bind water. 35 The treatment of apple pulp makes it possible to use continuous presses, or rotary vacuum filters. Treatment improves the yield from types of apples which are difficult to press (e.g. Golden Delicious) and in the case of apples stored for long periods and over-ripe apples. In the latter case the juice remains trapped in a semidegraded pectin network of the fruit's endogenic pectolytic system and hence the interest in an enzyme which can lower viscosity by specifically degrading only soluble pectin. Since the polyphenols in the pulp inhibit pectic enzymes, the phenomenon may be reduced by mixing the pulp for 15-30 min before adding pectic enzymes to provide oxygen for the fruit's polyphenoloxidase. As an alternative it may also be useful to add PVP (polyvinylpyrrolidone) (0·04% in weight).36


Treatment time varies from 10-60 min at 40-S0°C with enzyme concentrations of 0·004-0·04%, and the increase in juice yield is remarkable. 25 The juice obtained contains from 200-300 ppm of methyl alcohol against 30-80 ppm in the juice obtained from untreated pulp.37

Recently, it was proposed to optimize treatment of the crushed apple before pressing5,38,39 by using a new pectic enzyme preparation, based on Pilnik's6 theoretical observations.

The objective was to improve apple juice yield, press capacity, soluble solid content, the quality of the final product and reduce finishing treatments. The process runs for 2 h at 20-2SoC. It completes pulp enzymatic liquefaction before screening and permits different juice separation methods thus greatly reducing the screening time.

Unlike the method which involves times of up to 2 h at 40-S0°C with cellulase and pectinase preparations and unfavourable effects on the aroma, there are no modifications of the juice's organoleptic features. Finally, it permits a reduction of 30-S0% in the residual solids, an increase in pressing capacity from 1 to 1·S-1·8, with the aroma perfectly preserved in the concentrates obtained.

Pectic Enzymes in the Maceration of Fruit Pulp (Homogenized Products, Nectars and Concentrates) The thermo-mechanical process is usually ideal for preparing pulpy fruit juices from fruit which, during maturation, undergoes a natural softening and a maceration of the tissues. In the case of fruit with firm flesh such as apples, the use of pectinase is recommended to determine the hydrolysis of pectins from the intercellular membranes and produce, in the end, a suspension of whole cells. The process, described by Grampp,40 known as maceration, is effected by enzymatic preparations known as'macerases'. Their active components are depolymerases, in particular endo-PG, pectinlyases together with cellulases and sometimes proteases. The theory is that such enzymatic preparations degrade the pectin of the middle-lamella without attacking the pectin of the cell walls which, it is presumed, have a different (higher) degree of esterification, and thus remain intact. The finished products are characterized by high viscosity and good turbidity stability. Care must be taken to prevent the reaction continuing too long, to avoid affecting the turbidity stability;41 it is also necessary to know the endogenic pectolytic enzymes. In fact the combination of


additional endo-PG with the PE eventually contained in the fruit may affect product stability.

Therefore, for the production of nectars, obtained from fruit and berries by grinding, homogenization and the addition of water, sugar and citric acid (citric and/or ascorbic acid), to maintain a creamy consistency and a stable turbidity of the finished product, it would be opportune to subject the homogenized pulp to a rapid heating to inhibit endogenous PE and other enzymes.

Macerase is useful in homogenized apple, apricot, carrot and tomato production. Macerases are also widely employed in the technology of nectar base production (concentrates of pulpy juices). The viscosity drop in the macerate permits the separation of the pulp from the serum, in a decanter, the latter is then concentrated without heat damage and mixed again with the solid part. 42 The nectars obtained, by dilution of the concentrate thus obtained, demonstrate a stable turbidity.

Pectic Enzymes in Total Fruit Pulp Fluidification Pectolytic enzymes mixed with cellulase may permit an almost total 'liquefaction' of the fruit and vegetable pulp.43 Depending on the original material, juices can be obtained which are almost clear (cucumber, papaya), turbid (peach) or pulpy (apple, apricot, carrot). As the plant cells separate from one another, the cell walls are also mainly digested.

The structural viscosity of the puree6 drops drastically if pectinase is associated with cellulase and therefore smaller quantities of pectinase are used; on the other hand, soluble pectins only degrade slowly permitting the turbidity of the juice to remain more or less unaltered, but without the property of a long sedimentation stability, whereas the macerated pulp of the whole fruit has a low viscosity and may be concentrated.

Total ftuidification may be aimed at the production of nectars or nectar bases as above or the production of clear juices. Thus with simple technology a juice yield of 95% and more is obtained; the separation of the juice from the residual solids may be obtained by using a rotary vacuum filter without the need for expensive equipment such as the press. The process may be ideal for products for which there is no specific juice extraction machinery (tropical fruit) and therefore very important for producer countries of these fruits.44

By using enzymes which attack the middle lamella and the primary

200

Washing

Sorting

Pitting

G. LANZARINI AND P. G. PIFFERI

1-----1--Crushing -----II ••. ~ZYmi ng ________

Screening ----.... Processing: Peeling

Blanching

Pasteurization_Filling

" ~mogeniZing

Concentration

" Freezing (II)

;reSSing

t'oUdy juIce

~zyme

Filtration

" Clear juice

" Concentration

" Clear juice

" Concentrated JUIce

(III)

FIG. 2. Use of enzymes in the preparation of juices (I), homogenates (II) and clear juices (III).

cell wall without destroying the membranes around the protoplasts, it is possible to avoid free access to the air and uncontrolled reactions between enzymes and their substrates. By flash-pasteurization at a high temperature in plate-type or tube heat exchangers, oxygen is eliminated and the enzymes are destroyed before the cell contents are extracted. Products can be obtained with a higher content not just of soluble solids, but also of pigments, vitamins and aromas. The liquefaction process involves times of 1-4 h at 35-50°C with continuous shaking to achieve close contact between the enzymatic preparation and the substrate.45 Figure 2 shows a block diagram for the production of juices with and without enzymatic fluidification.

Pectolytic Enzymes and Use of Citrus Fruit-Juice By-products Pectic enzymes are also used in the citrus fruit industry especially for the exploitation of by-products. The coarse pulp which remains in the juice strainers contains remarkable quantities of juice and soluble solids, which may be recovered by counter-current washing and consist of the so-called pulp-wash. The use of pectinase leads to a considerable increase in the soluble solids extracted and reduces the viscosity of the pulp-wash, facilitating its concentration46-48 at more than 70° Brix. The pectinases used must not contain pectinesterase.


Another important application is the use of the peels remaining after juice extraction. This material is interesting since it provides a product which increases the desirable cloudiness and colour of beverages, for which a great demand exists in the soft-drink sector.

After heat treatment to deactivate the enzymes and soften the tissue, the peel is ground, pectinase is added and then it is pressed. A dense turbid liquid is obtained which, after centrifuging, is concentrated and pasteurized. 49

Pectinases, associated with cellulase and xylanase, were used to increase the extraction yield of essential oil from the peels.50,5l Again from the peels, it is possible to obtain a pectin with a low methoxyl content, by activating the endogenic PE, which may be of interest in the food industry and which under different conditions forms jellies with bivalent cations. 52

Use of Cellulase in Fruit Pulp Treatment The commercial availability of cellulases has provided new possibilities for the processing of fruit and vegetables.

Cellulose molecules are linear chains of l,4-a-D-glucopyranose in Cl conformation. The diequatorial glucosidic bond determines the stiffness of the chain, so that these line themselves up, side by side, in microfibrils which are held together by hydrogen bonds, which give a part of the fibrils a crystalline character; other parts are amorphous and the latter are attacked by the so-called Cx-cellulase which is an endo-p-glucanase (EC 3.2.1.4). For an effective cellulose solution an exo-p-glucanase (Cl-cellulase) is necessary to separate the cellobiose from the non-reducing tip of the molecule and requires that the Cx

activity first creates these points of attack. The use of cellulase in treating the pulp to obtain dense or clear juices by pressing is rather limited; if the juice yield is better and the total solids increase in the case of plums,53 pears,54 apples, 54-56 and apricots,53,57 the results regarding the organoleptic quality of the clear juices obtained are contradictory. It would appear that treatments at 50°C or for more than 90 min at 25°C cause a decline in the organoleptic quality of apple juice, due to the reduction in aldehyde C6 and the C6 esters, whereas the effect on the colour and the aroma of apricot and plum would appear good57 or improved. 53 Therefore the cellulases are used in the treatment of stone-fruit pulp to obtain clear fruit juices with a rich colour, for the production of turbid fruit juices, in homogenized fruit to be concentrated and also for the production of fruit or apple juice


after the total fluidification of fruit. 38.39,45,58 The combination of cellulase with pectinase drastically reduces the viscosity of the pulp during fluidification, which may become complete.6,58

In fact it has been shown59,60 that mixtures of cellulase C1, endo-PG and PE solubilize 100% of the apple cell wall; mixtures of PG and PE with or without cellulase C1 completely solubilize the cell wall pectin. Without PE, endo-PG alone or associated with cellulase C1 solubilizes only 48% of the pectin.

On the other hand, pectinlyase alone degrades 83% of pectin and this yield is increased by the addition of cellulase C1. The viscosity drop with the pectinases increases with the temperature, whereas the cellulase action is little affected by the temperature. 55 Cellulase alone and in a mixture with pectinase, contributes to increase markedly the diffusion speed of sliced apple juice,61 in the production of juice by pulp extraction.

Use of HemiceUulase in Fruit Pulp Treatment Endogalactanase (EC 3.2.1.89: galactan 4-f3-D galactan hydrolase) and endoarabinase (araban 5-a-L-arabinohydrolase) are able to release approximately one-third of the arabinose and the galactose contained in the apple cell wall and they probably have the same type of polysaccharide substrate (arabinogalactans). In particular the endoarabinase may permit the degradation of hazes which form in clarified and concentrated apple juice62-M and which consist of 1,5 arabinase.39 These enzymes already available on a commercial level, are added to the crushed apple pulp39 together with pectolytic preparations to destroy these polymers as soon as they are solubilized.

They are also used for the enzyming of crushed pulp in the preparation of clear fruit juices from blackcurrants, raspberries, blueberries, cherries, grapes, and apples, and to obtain a greater juice yield than with pectinase alone, with a greater colour extraction65 due to the drastic drop in viscosity. With these preparations45 it is also possible to obtain the fluidification of soft and tropical fruit, homogenized fruit which flows easily, or pulp concentrates.

Use of Proteases in Fruit Pulp Treatment The use of proteases in the treatment of fruit pulp is limited. Acid proteases are used in mixtures with endo-PG and PE in the treatment of fruit pulp to obtain juice from berries,65 fruit juices for canning,66 from strawberries,67 to increase the juice yield and improve the


extraction of anthocyanin pigments; and in homogenized tropical fruits to obtain fluid products with good organoleptic characteristics. 68

TREATMENT OF JUICES WITH SOLUBLE ENZYMES

Pectolytic Enzymes The production of turbid concentrated juices and clear juices, whether concentrated or not, raises the problem of the viscosity drop. In the first case depectinization must be driven to the point where viscosity is close to its lowest value and requires the cleavage of not more than 10% of the glucosidic bonds; in the second case, beginning with pressed juices, with a permanent viscosity and turbidity, the pectin must be sufficiently degraded to avoid hindering or slowing down either the setting phase or sedimentation and to facilitate the subsequent filtering of the sediment. Depectinization is forced to the point of the negative alcohol test. 69 The apple juice clarification mechanism has been studied by End070,71 and Yamasaki et al. 72 ,73

The combined action of PE and endo-PG appears to be necessary and sufficiene0-73 to degrade highly esterified pectin (degree of esterification = 82-95%) in solution, which acts as a protective colloid, and which coats the micelle of the pulp, which partially hydrolysed, determines the coagulation of the micelle. Clarification would take place in three stages: solubilizing of insoluble pectin (protopectin); breaking down of soluble pectin and precipitation of the micelles.

The cloud particles are formed of a protein-carbohydrate complex with approximately 36% protein. With a pH of 3·5 the external surface of the complex, consisting of pectin and other carbohydrates, has a negative charge whereas the nucleus has a positive charge. Partial hydrolysis of the external negatively charged shell causes the exposure of part of the positively charged protein: the destabilized cloud particles then coagulate and precipitate through electrostatic interaction.

With a pH of 4·7 when the protein has a negative charge, there is a drop in viscosity but not coagulation. The parameters influencing the process are pH, temperature, contact time, enzyme concentration. The cultivar and ripeness of the fruit influence the non-enzymatic stage of the setting process; with a lower pH clarification is more rapid.


Clarification of apple juice is also possible using pectinlyase (PL) alone74-77 for which apple pectin, with its high degree of esterification, is an ideal substrate. According to the most recent methods, clarification is carried out at 45-50°C. The juice, after dearomatization at 9O"C, is cooled at 50°C, a pectolytic preparation is added and subsequently a heat-stable tJ-amylase; after 90 min, with the addition of the finishing agents gelatine and bentonite, it is then filtered and concentrated with the obvious cyclic advantages when compared to the conventional one which involves an amylase which is not heat-stable and thus more time-consuming.

A different method is traditionally used in the clarification of apple juice for cider production; in this case pectin, de-esterificated with PE, reacts with added calcium ions, leaving a clear juice.

Pectinesterase may play a useful role in the production of citrus fruit juices, in particular lemon, for the production of clear juices and those to be concentrated; endogenic PE acts even at the low pH values of these fruit juices,9,78 however it is a slow process, and recently PE-based preparations have been devised to clarify fruit juices in industrial situations;79 the use of an exo-PG has also been proposed. 80

The main problem in the technology of citrus fruit juices, besides the maintenance of a stable turbidity in orange juice, given the PE activity even at low temperatures,81 is to avoid the formation of calcium pectate jelly in concentrated juices, which prevents the juice being reconstituted78,82-85 by addition of water; the heat stability of PE must be noted, with a rather low pH this increases significantly with the increase in the degree of concentration. 86

Amongst the various stabilization methods studied, some employ enzymes. The depolymerization of soluble pectin in the juice (DE 65%) has been proposed, using pectinlyase (pL)87 before demethylation with endogenic pectinesterase. The oligomers obtained, even if demethylated can no longer precipitate due to the pH or due to reaction with the calcium ions of the juice.

It is possible, by adding PG to the juice, to obtain depolymerization to oligomers of the low DE pectin produced by PE;lO,83 the oligomers obtained containing less than 18 units of galacturonic acid cannot form insoluble calcium saltsll and may contribute to PE inhibition.

Even the concentration of natural orange juice, at 60% (w/w) (60° Brix) is facilitated by the lowering of viscosity obtained by adding pectinase (500 ppm per Brix degree, for 1 h at 25°C), without causing

ENZYMES IN THE FRUIT JUICE INOUSTRY 205

problems for the turbidity of juice rediluted at 12·5° Brix, for the organoleptic characteristics and with a significant improvement in the settlement stability. 88

Amylase A problem, defined starch haze, may occur in concentrated apple juice prepared from unripe fruits, when the starch gelatinized during the aroma stripping stage dissolves in the juice and subsequently undergoes retrogradation,89 this gives rise to starch granules which are difficult to eliminate.32 This may be avoided with amylase treatment,90-92 which, added at the same time or after pectinase, if heat stable, permits a shorter cycle than the traditional one.93

Proteases Proteases are little used in the treatment of fruit juices; acid proteases are recommended94 in the clarification of fruit juices to be canned; treatment of citrus fruit concentrates with pepsin at 50° Brix, permits the preparation of soft drinks with 8·5% fruit juice without losing turbidity or the formation of precipitate for approximately three months. 95

Debittering Enzymes Citrus fruit products contain two classes of bitter compounds: the flavonoids and limonoids. Naringin, the main bitter flavonoid, is found in large quantities in grapefruit peel, and at a production technology level various factors affect the bitter taste of the juice obtained.96-98 An excessive naringin content, more than 700-800 ppm, is held to be the cause of an unpleasantly bitter taste. Besides the suggestion of using relatively specific adsorbents, the use of enzymatic preparations (naringinase) has been proposed, which by means of a-L-rhamnosidase transforms naringin into prunin «(:1-0-glucosil-1, 7 naringenin) and then with (:1-o-glucosidase, prunin to naringenin, which is not bitter.99-101 The scheme is shown in Fig. 3.102 Cyclic terpene limonin with two lactone-rings, which is mainly responsible for the bitter taste of orange, lemon and mandarin juice, is already perceptible at 6 ppm, and originates in the fruit juice after pressing from a non-bitter precursor, limonoic-monolactone-A -acid, present in the albedo, in the rags and in the seeds. Besides different methods which reduce the limon in content mainly with adsorbents, biologicalmethods have been proposed which instead of enzymes (0-


vf~,O HO~ OH 0

OH

H~ ~ a - L - Rhamnosidase

HO OH

OH

OH 0 OH

I ~ - D - Glucosidase

HO ~ OH

Naringenin

HO 0

FIG. 3. Enzymic hydrolysis of naringin.

lactonhydrolase limonin, limonoato-dehydrogenase, limoninepoxydase, deoxylimoninhydrolase), use microorganisms: Pseudomonas, 103

Arthrobacter globiformis,104 Acinetobacter Sp.,105 Corynebacterium fascians. 107 The first two transform limonin along two routes: to 17-dehydrolimonoate-A-lactone acid through limonin D-lactonhydrolase and limonoatodehydrogenase; and to deoxylimonoate through limoninepoxydase and deoxylimoninhydrolase (Fig. 4); the third microorganism also follows this second route to degrade limonin, whereas Corynebacterium fascians transforms limonin into limonol and nomilin, another bitter limonoid perceptible at concentrations greater than 3 ppm, to obacunone (which is not bitter) with nomilinacetyltranseliminase, and subsequently to obacunoate with obacunon-A lacton-hydrolase (Fig. 5).

The use of enzymes extracted from these microorganisms does not


Deoxylimonin

~

o

Deoxylimonate

+

E, ow -

Limonol

~

Limonoate A-ring lactone

JE'

17 - dehydrolimonoate A - ring lactone

207

FIG. 4. Conversion of limonin via limonoate A-ring lactone to C17-

dehydrolimonoate A-ring lactone and debittering of limonin.

provide an effective degradation of limonoids at the juice pH, due to their rather high optimum pH (approximately 8·0).

Oxidoreductases Glucose oxidase, which oxidizes glucose to gluconic acid with oxygen consumption, was proposed for the stabilization of the colour and odour of juices107,l08 due to oxygen elimination and the reduction of glucose susceptible to the Maillard reaction (reaction between an aldehyde and protein or amino acid) and to prevent the development of off-flavours in orange juice due to oxidation of limonene. 109 The ascorbic/ ascorbicoxidase acid system may be useful as an oxygen scavenger in beverages such as apple juice. lio Alternatively ascorbic acid and other fruit juice constituents may be protected against oxidation by adding superoxidedismutase. lll Diacetylreductase has


Nomilin

Obacunon Obacunoate

FIG. 5. Proposed pathway for debittering nomilin.

been proposed to eliminate diacetyl from orange juice;1l2 aldehydoxidase or aldehydedehydrogenase might be useful in reducing acetaldehyde and propionaldehyde which may form in orange juicel13

and subsequently give rise to condensation products which contribute to modifying the aroma (smell of carton).

Enzymes Acting on Polyphenols Oxygenases capable of breaking aromatic rings of phenols114 or enzymes capable of phenol methoxylation,115 e.g. methyltransferase, thus reducing the reactivity of orthodiphenols with oxygen and polyphenoloxidase and, hence, the alterating of some organoleptic characteristics, have been proposed for processing beverages and fruit juices; however the heat inactivation of the polyphenoloxidases in the fruit pulp, which is possible with the present heat exchangers, has led to a fall-off in interest in these enzymes.

Also the hydrolysis of orange juice anthocyanins, rich in pigments ('Sanguinello'), by an anthocyanase (f:J-glucosidase) of fungal origin, might improve the taste and prevent unpleasant colour variations during juice pasteurizing and represent an interesting application for the Italian citrus fruit industry.


USE OF ENZYMES IN TREATMENT OF VEGETABLES

Pectinases are the most widely used enzymes in the preparation of vegetable JUIces and for their liquefaction, together with cellulase.38,39,58 As with fruit, enzyme treatment may be carried out on the crushed vegetables before pressing and/or on the solid residues to increase the overall juice yield. However, partial liquefaction by maceration116 or total liquefaction, appears to be the best method of obtaining both juices and homogenized vegetables. 58 For this purpose pectolytic and cellulase enzyme mixtures are the most commonly employed. An interesting application is that which produces juices rich in carotene from carrot,117,118 courgettes, lettuce, spinach, and green peppers.117 The disintegration of the vegetable is related to the enzyme concentration (0·06-0·6%), time, pH, type of vegetable. The most frequent diameter of the particles, after treatment for 1-3 h at 50°C with 0·06-0·6% endo-PG, is between 0·01 and 0·06 mm.ll9 After treatment with endo-PG most of the 18 vegetables treated120,121 are made up of single cells and the product is suitable for further processing to produce juices, baby foods, dehydrated soups and sauces. The concentration of tomato juice collected by machine can be improved to obtain concentrates with 38-45% dry substance. The yield in juice increases by an average of 20% and the dry substance content by 7%. A great deal of research has been done on the production of carrot juice. Pulp treatment with enzymes increases the juice yield122 and in the liquefaction of carrot the best results are obtained with enzymatic preparations richer in PE,123 whereas endo-PG may be partly replaced by cellulase.6

Munsch prepares carrot juice by treatment with pectolytic enzymes and subsequent pressing124 to obtain a significant increase in juice, dry substance, °Brix, viscosity and colour. Blanching increases yield but reduces the nutritive value. Simultaneous treatment with endo-PG and cellulase125 is more effective than treatment with single enzymes in succession and, at 50°C, pH 4·0 for 30 min, permits the degradation of 80% of the sliced carrot cooked with microwaves, against 60% with single treatments.

TREATMENT OF JUICES WITH IMMOBILIZED ENZYMES

Besides the general advantages (reuse, process continuity, constant product quality, finished product free from enzymes, multi-enzymatic


complexes) and disadvantages (loss of activity during immobilization, cost of support and immobilization, higher investment costs, more technology, microbiological problems, etc.) of immobilization, in the case of juices there are also legislative scientific-technological and economic problems.

As regards legislative problems it can be said that the employment, permitted at the moment, of crude enzymatic preparations makes immobilization less interesting.

The main scientific-technological problems are:

(1) The difficulty of obtaining pure enzymatic activity to be able to prepare tailored active mixtures, e.g. PE and endo-PG or PL and endo-PG, ideal for obtaining the pre-established results;

(2) the difficulty of immobilizing enzymes with expensive and labile co-enzymes (e.g. debittering enzymes);

(3) problems linked to the kind of fruit or citrus fruit juice; e.g. pH significantly different from the optimum pH of the soluble enzyme, high molecular weight of substrate to be treated, presence of adsorbed polyphenols on the immobilized enzyme with possible inhibitions, viscosity of the solution, inhibition by the product, etc.;

(4) coexistence of opposing factors, such as for example the convenience of using porous supports to have a high surface area for enzyme fixation and therefore smaller reactors and a short residence-time, and conversely the juice turbidity which occludes the pores and would therefore require non-porous supports, etc.

All this makes it difficult to choose the most suitable design for the biocatalytic reactor, which in its turn, may condition the choice of the support and/or the immobilization method.

The scarce amount of information in the literature, which is, furthermore, limited to laboratory experiences, and sometimes replaces enzyme immobilization with that of microorganisms, is a result of these factors.

The economic aspect remains the most significant, and the obvious question is whether it is worthwhile to use immobilized enzymes limited to the juice-treating sector. Several factors must be considered:

(1) The cost and the quantity of soluble enzymes used, which if low reduce the economic advantage of immobilization;


(2) the cost of immobilization (support, materials, labour); (3) the degree of immobilization optimization;

211

(4) the operational stability and the stability of the enzymatic matrix;

(S) the kind of catalysed reaction; if the quality standard of the treated JUice is distinctly improved, a new technological treatment, even if expensive, may prove worthwhile if an increased consumer preference for the product is obtained;

(6) the assessment of alternative technological treatments to the new one proposed, in relation to the quality of the product obtained.

Generally the third and fourth factors mentioned are very positive, it can be said that, bearing in mind the average cost, enzyme immobilization may be worthwhile on an economic basis.

Use of Immobilizing Enzymes in Fmit Juice Production Today If, up until now, there have been numerous papers on enzyme immobilization as applied to fruit juice technology, the technological application of enzyme immobilization has rarely been experimented with, and, to our knowledge, only at a laboratory level. Pectic enzymes, immobilized on different supports,126,127 have been experimented on in the clarification of apple juicel 2B--133 and citrus fruit juicey4

On polyacrylic acid129 , on porous glass135 and on sisal fibre,134 the working stability of endo-PG is quite good with a half-life of up to 13 days, but the specific catalytic activity is low, except for the second support (up to 1S00 enzymatic units (EU)/g) which is, however, very expensive. PE immobilized on nylon has a very high stability, 132,136 but the immobilization yield is low, around 10%.

Good yields, at low cost and with reasonable preservation stability at pH 3·0 of around 45-50 days were obtained on oxycellulose,137 on y-alumina135 and modified chitin138 which makes their use potentially interesting for the clarification of juices, in particular apple.

The debittering of grapefruit and citrus fruit juice with immobilized enzymes has witnessed the covalent immobilization of naringinase on DEA-Sephadex,139 on hydrophylic polymeric materials127 and on chitin. 140

The best results are obtained by physical immobilization on hollow fibres,141-142 there is a reduction of 60% in the initial naringin, without


negative effects on' the part of the suspended solids in the juice,142 with a good correlation between the organoleptic perception of bitterness and the quantity of naringin remaining after treatment.

To reduce limonin and nomilin, responsible for the bitter taste of orange juice, microorganisms immobilized in polyacrylamide gel were successfully used, such as Arthrobacter globiformis 1,144 Arthrobacter globiformis 11,145 Corynebacterium fascians. 146 The first reduces limonin and nomilin to the corresponding 17-dehydrolimonoate-Alactone and 17-dehydronomilinoate-A-Iactone, the second to limonol and nomilinol and the third nomilinol to limonol and obacunone, all non-bitter compounds. Corynebacterium fascians has the advantages of having constitutive debittering enzymes without the need for an inducing limonoid in the medium, and of using inexpensive carbon compounds.

EXPERIMENTAL CONTRIBUTIONS

Immobilization of Polygalacturonase and Pectinesterase Two studies are mentioned which provide interesting practical results on the immobilization of endo-polygalacturonases (en-PG) and pectinesterase.

The immobilization of en-PG has been tested on inorganic supports such as glass, silica, talc, bentonite, graphite,147 celite, pumice, alumina, zeolite, calcium oxide, and on organic supports such as cellulose and its derivatives, silk, chitin and its derivatives, 6-6 and 11 nylon, PVP, maleic styrene-anhydride copolymers, phenolic resins prepared 'ad hoc', using as support activators volatile metallic salts (TiCh), bifunctional agents (glutaraldehyde, diisocyanates) and reactive dyes. The stability of the immobilized enzyme at pH 3·0 and at 25°C was assumed as the criterion of efficiency. The best results were obtained by immobilizing the en-PG on oxycellulose (11-13% of COOH) , on modified chitin (mono- and trimaleil chitosan) and on r_alumina134,137,138,148 (Table 4).

The PE was immobilized on r-alumina (spherules or powder) using a methodology involving the activation of the support by treatment with dodecadiamine (1% in chloroform) and subsequently with glutaraldehyde (1% w/v). The PE was then bound in a buffer at pH 3·0 and at 4°C (Table 5).149

TA

BL

E 4

E

ND

O-P

OL

YG

AL

AC

TU

RO

NA

SE

IM

MO

BIL

IZA

TIO

N.

PH

YS

ICA

L

AN

D

CH

EM

ICA

L

CO

NS

TA

NT

S

OF

T

HE

F

RE

E

AN

D

IMM

OB

ILIZ

ED

E

NZ

YM

E

tIl

Supp

ort

AU

U

I A

Y(%

) IY

(%)

Act

ivit

y ra

tio

Max

. KM

H

alf-

life

(t/

2) h

ours

z N

3·

0 p

H/m

ax.

pH

p

H

(mol

/lit

re)

-< 3:

T

pH

N

at

Imm

. tIl

N

at.

PC

I m

m.

PC

("

C)

PC

P

C

'" Z

Chi

tin

2400

0 6

70

0

98

27·9

0·

08

0·13

5·

1 0-

0000

2 25

3·

0 67

2 13

72

..., :::: 4·

0 84

5 15

23

tIl

.."

40

3·0

46

55

::t'

4·0

90

110

c:: =l A

lum

ina

1200

0 2

20

0

90

18·3

0·

08

0·20

5·

0 0·

0004

3 25

3·

0 67

2 20

00

'""' 4·

0 84

5 10

60

c:: n 40

3·

0 46

80

tI

l

4·0

90

110

Z

Oxy

cell

ulos

e 3

60

0

400

97

11

0·08

0·

10

5·0

0·00

008

25

3·0

672

1406

0 c::

4·0

846

1103

~

40

3·0

46

338

-< 4·

0 90

21

8 --------.--,~-------------

----

----

----

-, ---

----------_

._,-

-----------" ,-

AU

= A

dsor

bed

unit

/g.

UI

= A

ctiv

e un

it/g

. A

Y =

Ads

orpt

ion

yiel

d. I

Y =

Imm

obil

izat

ion

yiel

d.

tv -VJ


TABLE 5 IMMOBILIZATION OF PECTINESTERASE ON Y-ALUMINA (1·0-1·5 MM DIAMETEI

SPHERES OR 0·075-0·25 MM POWDER)

Enzyme lmmobiliza- AY(%) IY(%) FY(%) AU/g U//g t/2 day Support preparation tion atpH3 (mg)

pH and 25°C

Crude PE 3a 84-8 15-6 13·2 10600 1650 30 10 preultra- 5a 12·0 n.d. n.d. n.d. n.d. n.d. 10 filtered on 3b 87 9·7 8·5 10900 1060 26 25 UM10 5b 33 n.d. n.d. 4200 n.d. n.d. 25 membrane 7b 90 n.d. n.d. 11275 n.d. n.d. 25

3c 100 7 7 12528 877 20 25 3c 100 5 5 12528 626 22 25

aWithout glutaraldehyde. b With glutaraldehyde and dodecadiamine. c Glutaraldehyde alone (0·10% and 0,25%). AU/g = Adsorbed unit/g. UI/g = Active unit/g. FY = Formal yield. AY = Adsorption yield. IY = Immobilization yield. n.d. = not detected.

Biocatalyzer

cage

~,J:2'5cm ~-/O'6cm ~m

6·8cm

FIG. 6. CFST reactor.

TA

BL

E 6

P

EC

TO

LY

TIC

A

CT

IVIT

Y

OF

CO

MM

ER

CIA

L

PR

EP

AR

AT

ION

S

US

ED

IN

CL

AR

IFY

ING

IN

DU

ST

RIA

L A

PP

LE

Jm

CE

AN

D

IMM

OB

ILIZ

AT

ION

YIE

LD

O

N

Y-A

LU

MIN

A S

PH

ER

ES

(D

IAM

ET

ER

1·0

-1·5

MM

)

Enz

yme

End

o-P

G

PL

P

E

Cel

lulo

se

Hem

i-A

ctiv

ity

ratio

E

ndo-

PG

imm

obil

izat

ion

prep

arat

ion

(EU

/g)

(EU

/g)

(EU

/g)

(EU

/g)

cellu

lose

(E

U/g

) P

G/P

L

PG

/PE

P

G/P

L

Yiel

d"

EU

E

U

t/2b

(%)

(Ads

./g)

(A

ct./

g)

(day

s)

1 5

65

0

5·4

45

1046

12

5·5

8·35

10

·5

1916

0 20

21

31

2 10

550

3·5

40

3014

26

5 11

·4

2·1

1024

0 45

7 13

3

1571

0·

18

4 87

27

392·

7 22

·2

10·1

89

00

1812

17

4

1415

0·

09

19

1572

2 74

·5

211-

1 7·

4 24

400

1795

22

5

884

0·36

13

·5

2455

·5

65·5

37

·5

4·1

1774

0 72

0 18

6

2243

0 3·

60

25·0

25

000

6230

89

7·2

6·95

14

·4

9600

13

82

31

7 34

850

5·40

64

53·7

20

·5

1065

0 21

83

45

8 8

70

0

5·50

12

0 22

000

1581

·8

72·5

21

·8

9 49

2000

0·

45

270

1·09

x 1

06

1822

60

0 16

11

2400

18

82

33

10

2610

14

78

0 50

000

186·

4 3·

35

55·7

1·

4 20

000

278

36

11

66

40

5·

6 11

85·7

12

22

850

3·2

7140

17

·0

9100

15

47

37

" ....

Act

ive

EU

Im

mob

ilIz

atio

n Y

ield

= A

dsor

bed

EU

x 1

00

bA

t p

H3

·0


Clarification of Apple and Lemon Juice with Immobilized EnzymeslSO,1S1

There are many preparations on the market for the clarification of fruit juices; some of these have been analysed to study their different pectolytic activity and the immobilization of the endo-PG on y-A1203 (Table 6). It may be noted that with constantly low PL values the corresponding endo-PG and PE values vary a great deal; this means that an accurate examination must be made of their use in relation to the technological effect required and the value of the clarifying cost/effect ratio obtained.

For clarification of apple juice, preparation 7 was used which does, however, show no pectinesterase activity.

Industrial apple-juice, pretreated with soluble PE (0·5 EU/mL) for 15 min at 25°C, was treated at 25°C in the continuous flow stirred tank (CFST) reactor of Fig. 6 with 42 EU of immobilized endo-PG continuously (320 ml/h) or in batch (36 cycles of 7 min each). The relative turbidity was read from the absorbance at 660 nm. The lemon juice, fed at 25°C in a continuous flow (90 ml/h) into the same reactor, was treated with PE (80 EU), securing the biocatalysing mesh to the reactor shaft without the container. The turbidity of the lemon juice, stored for 12 h at 4°C after treatment, was determined by measuring

100

o 2345678 I I I I I I Til11e fh) I I I I I

o 3 6 9 ~ ~ ~ ~ ~ v ~ ~ ~ Cycles

FIG. 7. Clarification of apple juice in continuous (0) and batch (e) processes with immobilized endo-PG (7' -cycles).

E c: Ii '" ~100

> ..... g 80 cc ~ 60 w > 40 ~ u:l 20 a::


o 2 3 4 5 6 7 8 TIME (hI

FIG. 8. Clarification of lemon juice in continuous process (CFST reactor) with immobilized pectinesterase.

the absorbance at 660 nm. The residence time in the reactor was 6 min for the apple juice and 24 min for the lemon juice.

The results indicate that apple juice can be clarified at 25°C in a continuous or batch process by treatment with endo-PG immobilized on y-alumina. In the continuous process there is no evidence of enzyme inhibition and the turbidity is reduced by 70% compared to 80% for batch treatment (Fig. 7). Lemon juice is clarified effectively up to an 80% reduction of its initial turbidity by treatment at 25°C with PE immobilized on y-alumina after an initial induction phase (Fig. 8).

REFERENCES

1. Pifferi, P. G. and Lanzarini, G. (1988). In: Use of Enzymes in Food Industry, G. F. Montedoro (Ed.), Monography IPRA-CNR, Vol. 22, Rome, p. 93.

2. Lanzarini, G. and Pifferi, P. G. (1988). In: Use of Enzymes in Food Industry G. F. Montedoro (Ed.), Monography IPRA-CNR, Vol. 22, Rome, p. 129.

3. Cantarelli, C. (1977). Innovazioni tecnologiche nella produzione di vini e succhi di frutta, Fast SIPIA, Milano, p. 9.

4. Gierschner, K., Haug, M. and Wimer, H. (1978). Deutsche Lebensm. Rundschau, 74, 438.

5. Doerreich, K. (1983). Fluess. Obst., 50(7),304.


6. Pilnik, W., Voragen, A. G. V. and De Vos, L. (1975). Fluess. Obst., 42, 448.

7. Joslin, M. A. and Pilnik, W. (1961). In: The Orange, Its Biochemistry and Physiology, W. B. Sinclair (Ed.), University of California Press, Berkeley, CA, pp. 373-435.

8. Krop, J. J. P. (1974). The mechanism of cloud loss phenomena in orange juice. Doctoral Thesis, Agricultural University, Wageningen, The Netherlands.

9. Rombouts, F. M., Versteeg, c., Karman, A. H. and Pilnik, W. (1982). In: Use of Enzymes in Food Technology, P. Dupuy (Ed.), Symp. Intern., Versailles, p. 483.

10. Baker, R. A. and Bruemmer, J. H. (1972). J. Agric. Food Chem., 20(6), 1169.

11. Termote, F., Rombouts, F. M. and Pilnik, W. (1977). J. Food Biochem., 1,15.

12. Uhlig, H. (1978). Rohm Spektrum, 20, 44. 13. McColloch, R. J., Nielsen, B. W. and Beavans, E. A. (1950). Food

Technol., 4(9),339. 14. Montedoro, G. (1968). Ann. Fac. Agraria Univ. Perugia, XXIII, 585. 15. Montedoro, G. and Angelini, L. (1973). Sci. Tecnol. Alim. (STA), 3,

159. 16. Montedoro, G. and Angelini, L. (1973). Sci. Tecnol. Alim. (STA), 3,

305. 17. Usseglio Tomasset, L. and Tarantola, C. (1963). Vitis, 3, 190. 18. Bielig, H. J. and Rouwen, F. M. (1976). Fluess. Obst., 43,42. 19. Bielig, H. J. and Rouwen, F. M. (1976). Fluess. Obst., 43,426. 20. Dousse, R. and Ugstad, E. (1975). Lebensm. Wiss. u. Technol., 8,255. 21. Neubert, K. P. (1972). Fluess. Obst., 39,238. 22. Schobinger, V. and Durr, P. (1974). Fluess. Obst. 41, 129. 23. Luethi, H. R. and Glunk, V. (1974). Fluess. Obst., 41,498. 24. Wucherpfennig, K., Possmann, P. and Hamacher, A. (1973). Fluess.

Obst., 40, 266. 25. Bielig, H. J., Wolff, I. and Baiche, K. J. (1971). Fluess. Obst., 38,408. 26. Pilnik, W. and De Vos, L. (1970). Fluess. Obst., 37,430. 27. Voragen, A. G. J. and Pilnik, W. (1981). Fluess. Obst., 48,261. 28. Roumbouts, F. M. and Pilnik, W. (1971). Fluess. Obst., 38,93. 29. Cejkova, R. and Vykoukova, I. (1977). Kvasny Prum., 23(3),63. 30. Suic, D. and Krstic, V. (1965). Fruchsaftind., 10, 100. 31. Vykoukova, I. (1977). Kvasny Prum., 23(8), 176. 32. Delecourt, R. (1972). BIF Bruxelles, 15, 82. 33. Suic, D. and Vujicic, B. (1973). Fleuss. Obst., 40,79. 34. Verspuy, A., Pilnik, W. and De Vos, L. (1970). Fluess. Obst., 37,518. 35. Endo, A. (1965). Agric. BioI. Chem., 29,229. 36. De Vos, L. and Pilnik, W. (1970). Fluess. Obst., 37,430. 37. Wucherpfennig, K., Dietrich, H. and Bechtel, I. (1983). Fluess. Obst.,

50,348. 38. landa, W. (1983). Fluess. Obst., S0, 308. 39. landa, W. and Doerreich, K. (1985). Transformazione e Valorizzazione


della Frutta. AlTA AIDIC, Bari, p. 49. 40. Grampp, E. (1982). Dechema Monographien, 70, 175. 41. Roumbouts, F. M. and Pilnik, W. (1971). Fluess. Obst., 38,93. 42. Sulc, P. and Ciric, D. (1968). Fluess. Obst., 35,230. 43. Grampp, E. (1969). Deutsche Lebensm. Rundschau, 65,343. 44. Bielig, H. J. and Wolff, J. (1973). Fluess. Obst., 40,413. 45. Urlaub, R. (1985). Trasformazione e Valorizzaiione della Frutta, AlTA

AIDIC, Bari, p. 49. 46. Braddock, R. J. and Kesterton, J. W. (1975). Proc. Fla. State Hort. Soc.,

87,310. 47. Braddock, R. J. and Kesterton, J. W. (1975). Proc. Fla. State Hort. Soc.,

87,292. 48. Braddock, R. J. and Kesterton, J. W. (1976). l. Food Sci., 41,82. 49. Larsen, S. (1969). [FlU. Techn. Rep., 9, 109. 50. Anil, S. and Richard, J. (1983). Biotechnol. Lett., 5(7), 481. 51. Platt, W. C. and Poston, A. L. (1962). US Patent 3,058,887. 52. Buckley, K. and Mitchell, J. R. (1976). US Patent 3,973,051. 53. Dandamaev, G. S. H., Vlasova, O. K., Mungieva, N. A. and Tiunova,

N. A. (1981). Tsellyulazy Mikroorg., U, 191. 54. Schobinger, V., Duer, P. and Akesson, A. (1981). Alimenta, 20(2),37. 55. Rak, M. J. and Vukov, K. (1978). Konzerv. Paprikaip, 1, 17. 56. Schreier, P., Drawert, F., Steiger, G. and Mick, W. J. (1978). Food Sci.,

43,1797. 57. Samsonova, A. N., Brumstein, V. D., Bomogolova, J. N., Nova

sel'tseva, N. Yu. (1979). Konservn. Ovoschchesush. Prom.-st., 1,22. 58. Schmitt, R. (1983). Fluess. Obst., 50(1),23. 59. Voragen, A. G. J., Heutnik, R. and Pilnik, W. J. (1980). App.

Biochem., 2, 452. 60. Voragen, A. G. J. and Pilnik, W. (1981). Fluess. Obst., 48(8),261. 61. Binkley, C. R. and Wiley, R. C. (1981). l. Food Sci., 46,729. 62. Churms, S. C., Merrifield, E. H., Stephen, M. A., Walwyn, D. R.,

Polson, A., Van der Merwe, K. J., Spies, H. S. C. and Costa, N. (1983). Carbohydrate Research, 113,339.

63. Whitaker, J. R. (1984). Enzyme Microb. Technol., 8,341. 64. Pilnik, W. and Voragen, A. G. J. (1984). Gordian, 9, 166. 65. Gierschner, K. (1981). Gordian, 6,205. 66. Poluyanova, M., Turkina, M. and Mikhailoiva, V. V. (1977). Konservn.

Ovoshchesush. Prom. -st, 9, 21. 67. Flores, J. and Heatherbell, D. (1984). Fluess. Obst., 51(7), 320. 68. Satoro, K. (1978). Japan Kokai 7896347 App. 23/8/1978. C.A. 89

213876. 69. Baumann, J. W. (1981). In: Enzymes and Food Processing, G. G. Birch,

N. Blakebrough and K. J. Parker (Eds), Applied Science Publishers, London, p. 129.

70. Endo, A. (1965). Agric. Bioi. Chem., 29, 129. 71. Endo, A. (1965). Agric. BioI. Chem., 29,229. 72. Yamasaki, M., Yasni, T. and Arima, K. (1967). Agric. BioI. Chem., 31,

552.


73. Yamasaki, M., Yasmi, T. and Arima, K. (1964). Agric. Bioi. Chem., 28, 779.

74. Ishii, S. and Yokotsuda, T. J. (1971). Agric. Food Chem., 19,958. 75. Ishii, S. and Yokotsuda, T. J. (1972). Agric. Food Chem., 20, 787. 76. Ishii, S. and Yokotsuda, T. J. (1973). Agric. Food Chem., 21,269. 77. Szaier, I. and Szaier, Cz. (1982). Biotechnol. Letters, 4(9),553. 78. Versteeg, C. (1979). Pectinesterase from the orange fruit. Their

purification, general characteristics and juice cloud destabilizing properties. PhD Thesis, Dept of Food Science, Wageningen, The Netherlands.

79. Grampp, E., Schmitt, R. and Uhlig, H. (1980). US Patent 4,211,799. 80. Omran, H. (1982). In: Reinigung und Characterizierung von exo

polygalacturonase aus Aspergillus Niger Dissertation Institut fuer Lebensmitteltechnologie, Universitaet Hohenheim, Stuttgart, p. 30.

81. Versteeg, C., Roumbouts, F. M., Spaansen, C. H. and Pilnik, W. (1970). J. Food Sci., 45,969.

82. Krop, J. J. P. and Pilnik, W. (1974). Lebensm. Wiss. u. Technol., 7,62. 83. Krop, J. J. P. and Pilnik, W. (1974). Lebensm. Wiss u. Technol. 7, 121. 84. Rothschild, G. and Karsenty, A. (1974). J. Food Sci., 39, 1037. 85. Baker, R. A. (1977). Proc. Ind. Soc. Citriculture, 3,751. 86. Marshall, M. R., Marcy, J. E. and Braddoch, R. J. (1985). J. Food Sci.,

50,220. 87. Krop, J. J. P. (1974). Agric. Res. Rep. N. 830, Wageningen, The

Netherlands. 88. Braddock, R. J. (1981). Proc. Fla. State Hort. Soc., 94,270. 89. HeatherbeU, D. A. (1976). Con!ructa, 21, 16. 90. Krebs, J. (1971). Fluess. Obst., 38, 137. 91. Durr, P. and Schobinger, H. U. (1976). Alimenta, 15, 143. 92. Grampp, E. (1977). Food Technol., 11, 38. 93. Grampp, E. and Urlaub, R. (1982). Swiss CH 626, 513 30/11/1981: C.A.

9684398d. 94. Poluyanova, M., Turkina, M. and Mikhailova, V. V. (1977). Konservn.

Ovoschchesush. Prom.-st, 9,31. 95. Wobben, J. H. and Tan, H. B. (1981). Ger. Offen. 3.045.817 19/6/1981;

C.A. 95, 787849. 96. Maier, V. P., Bennet, R. D. and Hasegawa, S. (1977). In: Citrus Science

and Technology, Vol. 1, S. Nagy, E. Shaw and M. K. Veldhuis (Eds) , A VI, Westport, CT., p. 355.

97. Rouseff, R. L. (1980). Flavonoids and Citrus Quality, ACS Symp. Ser. No. 143,83.

98. Albach, R., Redman, G. H. and Cruse, R. R. (1981). J. Agric. Food Chem., 29, 808.

99. Versteeg, C., Martens, L. J. H., Roumbouts, F. M., Voragen, A. O. J. and Pilnik, W. (1977). Lebensm. Wiss. u. Technol., 10(5),268.

100. Roe, B. and Bruemmer, J. H. (1977). Proc. Fla. State Hort. Soc., 90, 180.

101. Sodial, Fr. Demande 2.442.598 27/6/1980. 102. Roumbouts, F. M. and Pilnik, W. (1979). Microb. Ind. Alim. Ann.

Congr. Int., APRIA, Paris, p. 158.


103. Hasegawa, S., Kim, S. K., Border, S. N., Brewster, L. C. and Maier, V. P. (1973). J. Food Sci., 41,706.

104. Hasegawa, S., Brewster, L. C. and Maier, V. P. (1973).1. Food Sci., 38, 1153.

105. Yaks, B. and Lifshitz, A. (1981).1. Agric. Food Chem. 29(6), 1258. 106. Herman, Z., Hasegawa, S. and Ou, P. (1985).1. Food Sci., SO, 118. 107. Scott, D. and Reed, G. (1975). In: Enzymes in Food Processing, G.

Reed (Ed.), Academic Press, NY, pp. 519-47. 108. Chagorazde, K. S. and Bakuradze, N. S. (1972). Lebensm. Ind., 19(7),

284. 109. Hammer, F. E. and Scott, D. (1982). Bios. (Nancy), 13(8-9),5. 110. 1st. Spero V. P. A. French Pat. n. 2241260 (1973). 111. Agence Nationale de la Valorisation de la Recherche DOS 2417508

(1973). 112. Brewster, L. C. (1976).1. Agric. Food Chem., 24,21. 113. Raymond, W. R., Hostetter, I. B., Assar, K. and Varsel, C. (1979).1.

Food Sci., 44, 777. 114. Kelby, H. S. and Finkel, B. I. (1969).1. Sci. Food Agric., 28,629. 115. Finkle, B. I. and Nelson, R. F. (1963). Biochim. Biophys. Acta, 78,

747. 116. Musulin, K. and Sos, I. (1978). Konzero. Paprikaip, 6,217. 117. Gatai, K., Zetelaki-Horvath, K. (1978). Proc. Hung. Ann. Meet.

Biochem., 18, 207. 118. Bielig, H. I. and Woll, I. (1973). Fluess. Obst., 40,413. 119. Zeletaki-Horvath, K. and Gatai, K. (1980). Acta Aliment. Acad. Sci.

Hung., 9(4), 367. 120. Zeletaki-Horvath, K. and Gatai, K. (1977). Acta Aliment. Acad. Sci.

Hung., 6(4),227. 121. Zeletaki-Horvath, K. and Gatai, K. (1977). Acta Aliment. Acad. Sci.

Hung., 6(5), 355. 122. Deelen, W. Van (1984). Voedingsmiddelentechnologie, 17,62. 123. Dongowski, G. and Bock, W. (1978). Lebensm. Ind., 44,347. 124. Munsch, M. H. (1982). Influence des procedes de fabrication sur la

valeur nutritionelle du jus de carottes. Thesis, Universite de Dijon. 125. Szeenath, H. K., Frey, M. D., Radola, B. I. and Scherz, H. (1984).

Biotech. Bioeng. 26, 788. 126. Wood, L. L., Hartdegen, F. I. and Hohan, P. A. (1972). US Patent

3,929,574. 127. Chibata, I., Tosa, T., Ono, M. and Watanabe, T. (1978). lpn. Kokai

Tokkyo Koho, 78113052. 128. Sheiman, M. I. (1974). Properties and applications of free enzymes and

immobilized pectic enzymes. PhD Thesis, Rutgers University NY. 129. Young, L. S. (1976). Preparation, Characterization and Performance of

an Immobilized Multipectic Enzyme System. PhD Thesis, Cornell University, NY.

130. Bock, W., Kranze, M., Goebel, H., Anger, H., Schwaller, H. I., Flemming, C. and Gabert, A. (1978). Niihrung, 22, 185.

131. Rexova-Benkova, L., Mrackova, M. and Babor, K. (1980). Coli. Czech. Chem. Comm., 45, 163.


132. Vijayalakshmi, M., Baron, A. and Drilleau, J. F. (1982). In: Use of Enzymes in Food Technology, P. Dupuy (Eds), Lavoisier, Paris, p. 537.

133. Baraniak, A. (1982). Acta Alimentaria Polonica, 8(1), 11. 134. Anaya, M. c., Lopez, M. C. A. and Arijona, J. L. (1982). In: Use of

Enzymes in Food Technology, P. Dupuy (Ed.) Lavoisier, Paris, p. 503. 135. Baraniak, A. (1981). Acta Alimentaria Polonica, 7(3), 219. 136. Vijayalakshmi, M. A., Picque, D., Jamouille, R. and Segard, E. (1980).

In: Food Process Engineering, Vol. 2, P. Linko and J. Larinkari (Eds) Applied Science Publishers, London, p. 152.

137. Lanzarini, G. and Pifferi, P. G. (1985). Metodo di Immobilizzazione di Pectinasi su Ossicellulosa. It. Pat. n.C. 2040.12.IT.9-N297 Patent C.N.R.-IPRA n. 297.

138. Pifferi, P. G. (1985). Immobilizzazione di pectinasi su chitina modificata per la rottura di sostanze pectiche. It. Pat. n.c. 2040 12.1T.11 Patent C.N.R.-IPRA n. 298.

139. Ono, M., Tosa, T. and Chibata, I. (1977). J. Ferment. Techno!., 55(5), 493.

140. Tsen, H. Y. (1981). K'O Hsueh Fa Chan Yueh K'an, 9(10), 871. 141. Soderquist, J. S. (1978). Immobilization of a Naringinase in a hollow

Fiber Reactor. NS Thesis, University of California, Berkeley. 142. Olson, A. C., Gray, G. M. and Guadagni, D. G. (1979). J. Food Sci.,

44,1358. 143. Gray, G. M. and Olson, A. C. (1981). J. Agric. Food Chem., 29, 1298. 144. Hasegawa, S. and Pelton, A. V. (1983). J. Agric. Food Chem., 31, 180. 145. Hasegawa, S., Pelton, A. V. and Bennet, D. R. (1983). J. Agric. Food

Chem., 31, 1002. 146. Hasegawa, S., Vandercook, C. E., Choi, Y. G., Herman, Z. and Ou, P.

(1985). J. Food Sci., 50,330. 147. Lanzarini, G. and Pifferi, P. G. (1983). Industrie Alimentari, 22(3), 179. 148. Pifferi, P. G., Lanzarini, G. and Varela, A. (1985). Immobilizzazione di

poligalatturonasi su allumina modificata. It. Pat. n.C. 2040 12.IT.1O Patent C.N.R.-IPRA n. 298.

149. Pifferi, P. G. and Lanzarini, G. (1986). Proceedings of the Symposium On Mild Technologies, Rome, November 26-27, p. 216.

150. Pifferi, P. G. and Lanzarini, G. (1986). Proceedings of XIX IFFJP Symposium, Den Haag, May 12-15, pp. 493-501.

151. Lanzarini, G., Pifferi, P. G. and Manenti, I. (1987). Proceedings of II World Congress of Food Technology, Barcelona, March 3-6 (1987). pp. 268-9.

Chapter 14

ENZYMATIC PROCESSING OF MUSTS AND WINES

ARTURO ZAMORANI

Department of Agricultural Biotechnology, University of Padova, Padova, Italy

INTRODUCTION

Whereas the use of enzyme preparations in the food industry is well-established and expanding rapidly, enzyme processing in enology is less common.

In spite of Cruess' and Besone's earliest proposal of enzyme clarification (1941), it is indicative that the only routine use of enzymes occurs in the production of RCM (rectified concentrated musts), whose production technology concerns fruit juice rather than wines.

It seems that a high specificity would make enzymes suitable for accomplishing the change of some compounds to give a product of reliable quality. When biological or physical chemical processes cause these changes to occur they often involve undesired modifications of compounds other than those directly involved in the enzymatic treatment.

The reasons why enzymatic techniques are not commonly used include:

-Some traditional techniques obtain fairly good results by creating the conditions that control the activity of grape enzymes;

-the 'classic' wine industry is still based on traditional methods; -the narrow profits in the field cannot bear new technologies which

would cause costs to rise; -difficult conditions for enzymatic reactions in musts and wines

(low pH, ethanol concentration and presence of tannin compounds) ;

-low-grade purity of enzyme preparations; -possible enzyme persistence in wine; -legal restrictions.

223


224 ARTURO ZAMORANI

It is therefore more advisable to mention here research works and proposals aimed at replacing traditional techniques with 'mild technologies' based on enzyme preparations in order to reach the following targets:

(a) for musts: to increase yield to improve colour in red wines to improve flavour to avoid madeirization

(b) for wines: to improve clarity of white wines to decrease malic acidity

MUST AND PRESSED PREPARATIONS

The enological importance of must pressed preparations lies in their being tissues rich in pectocellulose. Their cultigen makes them a definite hindrance to liquid-fluid separation, to must fluidification, as well as to skin-located aromatic and colouring substances.

For all types of wines, the final goals to be reached are an increase in yield and an improvement in aroma; and for white wines an increase in colour is also required.

There are techniques and special methods by which we try to reach these objectives in traditional ways: intense grape and skin pressing, static or centrifugal preliminary clarification of musts, fermentation through maceration, thermovinification, carbonique or cold maceration. The latter shows some disadvantage~.g. an over-extraction of polyphenolic substances, an undesired activity of 'parasite' enzymes in long-term vinification, poor quality and high cost.

Proposals about the use of enzymes during vinification, (some attempts with cellulases and glucosidases preparations have already been made), mainly concern pectolytic and Proteolytic enzymes. l -4

Pectolytic enzymes can be classified as follows (International Enzyme Committee (lEe) classification):

-De-esterifiers, acting mainly on pectins; 3.2.1.11-pectinmethylesterase or pectin esterase (PME or PE)

-depolymerases acting mainly on pectins; 3.2.1.41-endo- and exopolymethylgalacturonase (PMG) 4.2.2.3-endo- and exopectin transeliminase (PTE)

ENZYMATIC PROCESSING OF MUSTS AND WINES 225

-depolymerases acting mainly on pectic acid; 3.2. 1. 15---endopolygalacturonase (PO) 3.2.1.4O----exopolygalacturonase (PO) 4.2.2.1-endopectic acid transeliminase (APTE) 4.2.2.2-exopectic acid transeliminase (APTE)

For enzyme preparations which are to be suitable for wine treatment, the possibility of reaching pectin macromolecule degradation with fairly limited methanol formation becomes very important. In this area we have noted that PE endogenous activity involves methyl alcohol liberation; this occurs in variable concentrations in different cultigens although the concentrations are quite high when compared to the ones derived by addition of commercially available enzyme preparations.

However, because of the effect of ethanol and polyhydric phenols, the PE activity of wines has been shown to decrease to one-third of the original value.1 The endoPO-like activity of grapes (which can lower the viscosity of a pectin solution), unlike that of PE, is less than that to be found by addition of commercial preparations in normal amounts. As far as the degradation process of pectin substances is concerned, the activity itself might also be a limiting factor. In fact, by liberating pectic acid, the demethoxylated compounds could inhibit PE activity through a feed-back mechanism.

Therefore it seems that PO activity is involved in the degradation process of pectic acid, representing a limiting factor for PE activity pectin demethoxylation, hence for the continuation of the hydrolytic process. This is not the only reliable hypothesis, since the presence of other (POM and PTE) pectolytic enzymes, of polyphenols and several side mechanisms, may be involved in the intermediate phase of the process. 1 Therefore we can conclude that PE activities are interdependent, hence it is necessary to use enzyme preparations whose composition would be complementary to the one of must and wine, and whose concentration should be defined by preliminary tests.

Proteolytic enzymes can be classified as follows (1EC):

-Hydrolytic enzymes acting on terminal amino acids of protein chains; 3.4. 11-aminopeptidases 3.4. 12--carboxylpeptidases

-hydrolytic enzymes acting on peptides; 3.4. 13-dipeptidases

226 ARTURO ZAMORANI

-hydrolytic enzymes acting on the inside peptide bonds of the proteinaceous chain; 3.4.23-acid proteases 3.4.24-neutral proteases.

According to some authors, grapes show a proteolytic activity which should persist at low pH values and at high temperatures (70-80°C; 158-176°P); so that hot maceration wines show a lower protein content but a higher content of peptide compounds. l

Cellulolytic activity enzymes can be classified as follows (IEC):

-3.2.1.4--endoglucanase, which cleaves the internal glucosidic bonds in native cellulose;

-3.2. 1. 91-cellobio-hydrolase , which releases cellobiose from the non-reducing end of the cellulose chain;

-3.2.1.21.-beta-glucosidase, which cleaves cello-oligosaccharides to produce glucose.

Using hydrolytic depolymerizing enzyme preparations, as reported in the scientific literature, it was concluded that the following improvements were possible, either separately or in combination:

-An increase in yield of free-run must and wine, when the addition is made before fermentation;

-greater compactness and easier separation of pressed cake; -increase in colour and colour stability during the aging of white

wines; -improvement of aromatic features.

It must be emphasized here that in order to prove the effective validity of the results, these should be subjected to marketing trials. There is a great deal of evidence as far as the increase in yields and improvement in filtrability are concerned.

The results obtained by Montedoro l are reported in Table I. It should also be noted that the must processed wines yield a bigger quantity of dry lees; this confirms therefore, the enzyme intervention on the colloid complex compounds bound to cell structures. No agreement, however, can be found on the two related topics, i.e. colour gain and effects on the aroma. Opinions range from recording a more intense or at least similar colour, to finding that the use of pectolytic enzymes yields weaker, less intensely coloured, wines.

The same could be said about organoleptic properties, which are

TA

BL

E 1

IN

FL

UE

NC

E O

F C

OM

BIN

AT

ION

O

F S

OM

E E

NZ

YM

AT

IC P

RE

PA

RA

TIO

NS

O

N T

HE

Y

IEL

D

AN

D F

lLT

RA

BIL

ITY

OF

FE

RM

EN

TE

D M

US

TS

A

ND

T

HE

IR C

OR

RE

SP

ON

DIN

G

WIN

ES

Yie

ld(%

) D

ry p

ress

F

iltra

bilit

y ca

ke w

ine

(ml/

min

) M

ust

Win

e (g

/litr

e)"

Mus

t W

ine

Fre

e ru

n P

ress

ed

Tota

l

Con

trol

65

·0

23·7

88

·0

92·8

5·

19

2·30

68

·0

Cel

lula

se

57·5

28

·8

86·2

92

·4

5·31

2·

32

81·2

C

ellu

lase

+ h

emic

ellu

lase

59

·4

23·7

89

·0

88·8

6·

67

2·30

75

·0

Cel

lula

se +

hem

icel

lula

se +

pro

teas

e 68

·7

22·5

90

·8

83·4

6·

62

2·25

84

·9

Cel

lula

se +

pec

tina

se

56·2

23

·5

89·6

88

·8

6·40

2·

45

86·5

C

ellu

lase

+ p

ecti

nase

+ p

rote

ase

61·3

25

·6

90·5

90

·2

6·59

2·

13

88·0

C

ellu

lase

+ x

ylan

ase

63·8

23

·1

89·8

89

·8

6·38

2·

23

78·8

C

ellu

lase

+ x

ylan

ase

+ p

rote

ase

67·5

21

·2

90·8

90

·0

6·71

2·

13

77·5

aCon

stan

t w

eigh

t at

105

°C (

221°

F).

ttl Z

N ~ ~ n "" :oc 0 n ttl '" ~ Cl 0 'r1

3: c:: '" til > z \::

) ~ ttl '" tv

tv

-..J

TA

BL

E 2

(j

IN

FL

UE

NC

E

OF

S

OM

E

DIF

FE

RE

NT

C

ON

CE

NT

RA

TIO

NS

O

F

EN

ZY

MA

TIC

P

RE

PA

RA

TIO

NS

O

N

AN

AL

YT

ICA

L

FE

AT

UR

ES

O

F

00

MU

ST

S A

FT

ER

24

h M

AC

ER

AT

ION

AN

D I

NC

UB

AT

ION

OF

'IH

EIR

CO

RR

ES

PO

ND

ING

WIN

ES

Enz

yme

Opt

ical

D

yeb

Ant

hocy

anin

sc

Pol

yphe

nols

(m

g/li

tre)

V

/U

Pro

tein

sg

amou

nt

dens

ity a

(mg/

litr

eY

Win

e M

ust

(mg/

litr

e)

Mus

t W

ine

Tota

ld

Tann

ic·

Mus

t W

ine

Mus

t W

ine

Mus

t W

ine

Mus

t W

ine

Con

trol

2·

100

1·80

0 1·

02

1·09

67

31

53

2 51

2 13

6 14

4 0·

71

164

Cel

lula

se

10

2·33

0 2·

080

0·95

1·

48

88

32

526

514

130

144

134

> 50

2·

175

2·11

6 1·

03

1·14

80

34

51

2 50

0 12

6 13

0 0·

78

100

~ 10

0 2·

115

2·22

6 1·

04

1·25

77

37

50

4 46

4 11

5 12

9 96

::c

Pec

tina

se

10

2·27

5 1·

700

1·14

1·

83

66

24

540

528

185

107

0 17

4 ~

25

2·65

0 1·

990

1-13

1·

65

77

26

556

540

197

128

0·77

18

5 a::

50

2·96

0 1·

920

1-11

1·

37

100

25

594

572

246

140

193

0

Pro

teas

e 10

2·

340

2·30

8 0·

95

1·16

86

39

54

5 52

2 16

5 14

9 10

2 ~

25

2·27

5 2·

260

0·92

1·

17

83

37

558

535

170

157

0·74

84

~

50

2·11

0 2·

210

0·90

1·

21

80

36

579

560

174

171

70

aOD

450

nm

and

520

nm

, cu

vett

e 1

cm.

b R

atio

OD

450

: 52

0 n

m,

cuve

tte

1 cm

. cD

iffe

renc

e O

D 5

20 n

m m

easu

red

in c

uvet

te 1

cm

wit

h an

d w

itho

ut K

2S20

5;

data

are

exp

ress

ed i

n te

rms

of

mal

vidi

ne

3,5

digl

ucos

ide.

d

Dat

a ar

e ex

pres

sed

in t

erm

s o

f ga

llic

aci

d m

easu

red

wit

h F

olin

-Cio

calt

eau.

e

Dat

a ar

e ex

pres

sed

in t

erm

s o

f ga

llic

aci

d m

easu

red

wit

h F

olin

-Cio

calt

eau

befo

re a

nd a

fter

pre

cipi

tati

on w

ith

met

hyl

cell

ulos

e.

f R

atio

van

illi

ne t

o le

ucoa

ntho

cyan

ins.

g

Dat

a ar

e ex

pres

sed

in t

erm

s o

f al

bum

in u

sing

biu

ret

met

hod.

TA

BL

E 3

E

FF

EC

f O

F

INC

UB

AT

ION

T

EM

PE

RA

TU

RE

O

N

PE

CT

INA

SE

A

CT

Ivrr

Y

AN

D

ITS

IN

FL

UE

NC

E

ON

T

HE

C

OL

OU

R

AN

D

PO

LY

PH

EN

OL

CO

MP

OS

ITIO

N O

F M

US

TS

AN

D W

INE

S

Tem

pera

ture

O

ptic

al

Nua

nceb

A

ntho

cyan

insc

P

olyp

heno

ls (

mg/

litr

e)

Pro

tein

sll

dens

ity a

(mg/

litr

e)

(mg/

litr

e)

Mus

t W

ine

Tann

icd

Tota

le M

ust

Mus

t W

ine

Mus

t W

ine

Mus

t W

ine

Mus

t W

ine

Con

trol

10

°C (5

0°F

) 1·

416

1·14

48

10

0 11

9 30

6 40

4 15

°C (5

9°F

) 1·

010

1·43

6 1·

45

2·03

16

6 56

11

7 12

0 29

2 25

°C (

77°F

) 2·

100

1·80

0 1·

02

1·49

34

6 72

13

6 14

4 53

2 54

2 16

4 35

°C (9

5°F

) 3·

550

2·46

8 0·

92

1·33

61

6 10

0 16

4 18

0 77

0 48

0 24

0 +

Pec

tina

se

10°C

(50°

F)

1·75

0 1·

37

80

150

107

303

504

15°C

(59°

F)

1·05

0 2·

068

1·33

1·

91

193

92

170

112

307

25°C

(77

°F)

2·27

5 1·

990

1·44

1·

83

413

86

197

128

556

410

78

35°C

(95°

F)

3·80

0 1·

990

1·02

1·

38

633

80

240

167

754

602

152

Foo

tnot

es a

s T

able

2.

230 ARTURO ZAMORANI

thought to be better than their controls in some experiments, while in others they are considered to be worse.

The results obtained by Montedoro and Bertuccioli of the effect of some hydro lases on must colour and its corresponding wine are reported in Tables 2 and 3.

Data on the effect of enzyme concentrations at macerating conditions of 25°C (77°F)/24h emphasize that all enzyme preparations cause an increase in colour intensity of must and wine. The pectinase effect occurs in a peculiar way. In fact while the must shows the highest colour intensity in all experiments, the wines, on the other hand, show the least intensity, one of them is of even lower intensity than the control.

The nuance of the musts deriving from various preparations are rather similar, one exception being the protease-processed ones, which are the best in all assays. Also, in this negative pectinase process, an increase in enzyme concentrations corresponds to an increase in colour intensities of must and to a proportional increase in the corresponding intensities of wine. These results are in agreement with the corresponding values of anthocyanin and tannin concentrations of wine and must. The loss of anthocyanin and tannin occurred during fermentation and are the highest in all tests. An increase in the colour of musts and wines is caused by cellulase, though the effect is less than the one caused by pectinase.

The loss in colour found in this processing can be explained by suggesting that other enzyme systems are activated under the influence of pectolytic preparations. These systems could consist of phenoloxidase and glucosidase; the former oxidizes orthodihydrophenols which then oxidize (by autocatalysis), anthocyanins which are available through depolymerization of colloid complexes; the latter hydrolyse the 3-glycosidic linkages and give unstable aglycon.

Similar results, which confirm the above mentioned ones, were obtained by enzyme hydrolytic associations on traditionally processed Rosso Conero grapes (Table 4) and carbonique macerated Rosso Piceno wines (Table 5). In both cases the protease cellulase association produces an increase in colour intensity. 1

It has been found that enzyme addition to must involves a change in the gas chromatographic profile of the wine headspace, which can be considered from the qualitative or quantitative point of view (Table 6).

The degustation of pectinase- and cellulase-treated wines after two months fermentation shows a more intense bouquet than the control; but the pectinase-treated is definitely oxidized after one year. 1

TA

BL

E 4

E

FF

Ecr

OF

ENZY

ME

PRO

CESS

ING

IN

RO

SSO

C

ON

ERO

PR

ESSE

D

GR

APE

S O

N

CO

LOU

R

AN

D

POLY

PHEN

OLI

C C

OM

Posr

nON

OF

ITS

CO

RR

ESPO

ND

ING

WIN

E

Opt

ical

N

uanc

eb

Ant

hocy

anin

sc

Pol

yphe

nols

(m

g / l

itre)

de

nsity

a

(mg/

litr

e)

Tann

ice

Tota

ld

Con

trol

5·

280

1·46

52

21

0 72

0 P

rote

ase

+ p

ecti

nase

' 4·

600

1·42

49

43

2 87

2 P

rote

ase

+ c

ellu

lase

' 6·

320

1·35

66

81

0 13

20

a-e

As

Tab

le 2

. f

Pro

teas

e-C

GA

20

238;

p

ecti

nas

e-C

GA

20

408;

ce

llu

lase

-CG

A

2038

5 (C

iba-

Gei

gy,

Bas

ilea

-Sw

iss)

.

TA

BL

E 5

E

FF

Ecr

O

F EN

ZYM

E-PR

OC

ESSE

D,

CA

RB

ON

IQU

E M

AC

ERA

TED

RO

SSO

PI

CEN

O

GR

APE

S O

N

CO

LOU

R A

ND

PO

L YPH

ENO

L C

OM

POSr

nON

OF

ITS

CO

RR

ESPO

ND

ING

WIN

E

Opt

ical

N

uanc

eb

Ant

hocy

anin

sc

Pol

yphe

nols

(m

g/li

tre)

de

nist

ya

(mg/

litr

e)

Tann

ice

Tota

ld

Con

vent

iona

l m

acer

atio

n 5·

260

0·75

30

50

246

1·44

0 C

arbo

niqu

e m

acer

atio

n,

7 da

ys a

t 30

"C (

86°P

):

Pre

ssed

mus

t 5·

075

0·73

50

75

362

1·44

0 +

skin

s fo

r 24

h

5-47

5 0·

7520

98

53

6 1·

480

+ en

zym

es'

4·77

5 0·

7207

11

1 33

7 1·

464

a-e

As

Tab

le 2

_ f

Pec

tina

se +

pro

teas

e (p

ecti

nas

e-C

GA

204

08;

pro

teas

e-C

GA

203

88,

Cib

a-G

eigy

, B

asil

ea

Swis

s).

~ ~ ~ i ~ ~ ~ ~ i tj -

232 ARTURO ZAMORANI

TABLE 6 EFFEcr OF DIFFERENT ENZYME PREPARATIONS ON SOME VOLATILE

COMPOUNDS OF RED WINE HEADSPACE

Number Peak Compound Control Pectinase Protease Cellulase

1 12 n.i. tr. 0·04 0·03 0·03 2 13 Ethyl isobutyrate tr. 0·02 0·01 0·03 3 15 n.i. 0·62 tr. tr. 0·01 4 16 n.i. 0·33 0·68 0·20 0·63 5 17 Ethyl butyrate 0·38 0·12 0·06 0·40 6 18 n.i. 0·04 0·00 0·02 tr. 7 21 n.i. 20·70 0·25 0·00 3·03 8 22 Isoamyl acetate 1·47 2·97 2·75 1·94 9 25 n.i. 1-46 0·00 0·00 0·02

10 26 Ethyl capronate 1·57 1·96 1·77 3·78 11 28 Octanal 0·12 0·04 0·04 0·07 12 31 Ethyl lactate 0·96 1-49 1·31 1·34 13 32 I-Esanol 1-68 2-19 1·99 1·99 14 34 Nonanal 0·06 0·50 0·04 0·04 15 36 Ethyl caprilate 0·67 0·45 0·57 0·74 16 41 Ethyl pelargonate 0·03 0·16 0·05 0·05 17 48 Ethyl caprinate 0·07 tr. 0·03 0·05 18 49 Ethyl succinate 0·11 0·10 0·10 0·13 19 57 2-Phenyl ethanol 0·10 0·08 0·08 0·14

Data are expressed as ratio peak area of volatile compound/peak area of internal standard. n.i. = not identified.

An increase in the colour intensity of differing Piemonte wines such as Barbera and Freisa was caused by the use of pectolytic preparations (Figs 1, 2 and 3); this intensity remains for the time equivalent to one year's marketing, along with an unfavourable evolution of aromatic features. 3 In any case, the increase in colour intensity is slight, and therefore of little practical value.

The cause of these conflicting results, in the use of pectolytic preparations, has often been attributed to the presence of impurities in various enzyme preparations (anthocyanase, esterase, galacturonase, glycosidase, etc.). Yet the action of pectolytic enzymes affects the conditions for the occurrence of reactions caused by the endogenous enzyme of grapes in different ways because of the high degree of purity reached nowadays by commercial preparations. In different situations these enzymes could not have carried out their catalytic action.

ENZYMATIC PROCESSING OF MUSTS AND WINES

27

21

~ ~ 15 " "C

ii Z 9 a. o

3

3 5 7 Days after must prepar;otlon

233

FIG. 1. Optical density trend in vinification of Freisa and Barbera musts, with or without pectic enzymes.

From the above mentioned considerations we can conclude that hydrolytic (especially pectolytic) enzyme preparations should be used on must and pressed grapes only when the advantage deriving from must ftuidification is greater than probable disadvantages in the colour and aromatic complex. These preparations should be used only when we can prevent undesired changes happening by suitable operations such as ready centrifugation, use of low temperatures and sulphur dioxide, etc.

The use of pectolytic enzymes (which are the only ones whose use can be considered as codified) is legally permitted in the processing of

3-6

Freisa

/~ Barbera

~t:;::"""-""""'1

0 ·4

o 2 4 6 8 Days after must prepar;rtlon

FIG. 2. Total polyphenols trend in vinification of Freisa and Barbera musts, with or without pectic enzymes.

234 ARTURO ZAMORANI

15

-+.,

12 \\~.. Fre.,sa

!:\~:==: g 3 ~: . ........ _. __ ..... :::

Barbera

70 140 210 280 350 Days after must preparation

FIG. 3. Optical density trend after vinification of Freisa and Barbera, with or without pectic enzymes.

pressed grapes and must for white wine-making without maceration of wines made through thermovinification along with wines after separation from vinasses. This use, however, is legally prohibited for red-vinified must.

Pectolytic enzymes for wine-making:

(a) must come from producing strains not belonging to pathogenic species which can form toxic metabolites;

(b) must not contain pathogenic organisms, and their toxins, particularly aflatoxins;

(c) must contain aerobic mesophilic saprophytes lower than 100 000 CFU (colony forming units) per gram;

(d) must not contain organisms related to faecal pollution per gram of product;

(e) must not contain residual fermentation products with antimicrobial and antimycotic activity;

(f) must be diluted so as not to alter the chemical-physical composition of the processed product.

The pectolytic enzymes which are put on the market must be kept in closed containers clearly labelled in Italian as follows:

(a) details as requested for other enological compounds (see 'agents substances for wine-making');

(b) instructions for use; (c) advised doses;


(d) a notice 'The processed product must be deprived of its solid parts by filtration, centrifugation or by other legally permitted systems';

(e) a boldface, clearly readable, notice 'methyl alcohol is likely to occur in higher quantity than permitted whenever the suggested dose is added in excess, or in case of non-filtration, noncentrifugation or non-separation of the liquid from its solid parts within five hours from the introduction of enzyme preparation' .5

In the various uses suitable conditions should be established to produce the desired results.

The various operating conditions are as follows:

(1) In white wine-making, to facilitate must or wine clarification and to reduce time contact in cold maceration: enzyme treatment (1-2 g/hl) within a range of 15-25°C (59°-77°F), must sulphitage, operating time 5-10 h, clarifier processing, then separation, fermentation.

(2) In thermovinification in order to obtain an increase in free run must, with a highly reduced viscosity producing a remarkable increase in liquid efficiency and its quicker and easier filtration: pressing sulphitation (15-20 g/hl sulphur dioxide), hot processing of pressed grapes enzyme treatment a 30-60 min break at 46-50°C (114-122°F), free run must, pressing, must cooling, fermentation.

(3) In clarification of must: enzyme treatment, about 60-90 min break, addition of clarifiers, filtration. In order to prevent fermentation and oxidation, other operations could also be carried out.

(4) In clarification of young and pressing-derived wines from sound and Botrytis-free grapes: enzyme treatment, 3-4 days break, addition of clarifiers, filtration and bottling. For pressingderived wines, the use of a higher quantity of enzyme preparation and an extension of break at about 30°C (86°F) might be advisable.

WHITE WINE PROCESSING TO IMPROVE CLARITY

For low protein content wines, the enological importance of proteolytic enzyme processing lies in the destabilization of proteins in

236 ARTURO ZAMORANI

colloidal dispersion in white wines. This causes turbidity and deposits to occur during presentation and marketing. This alteration commonly called 'casse' does not take place in red wines because of their tannin polyphenol content. These react at colloidal level with proteinaceous substances. The result is the almost complete removal of proteins during fermentation and subsequent aging.

Suitable enzyme preparations for this purpose are proteolytic enzymes (see classification in the previous section) which should guarantee hydrolysis of proteins to amino acids or short peptides no longer flocculent at the wine conditions. Partially satisfactory results have been reached by treating the must with acid proteases.

Concerning the repeated suggestion that wine is processed using proteolytic preparations it has to be recorded that it proves scarcely active. In any case, it creates the problem of the further removal of the added enzyme which can be carried out only with traditional clarification and pasteurization. For the reasons previously mentioned we think that the use of immobilized proteases6•7 might solve the problem.

Table 7 lists the results obtained by Bakalinsky and Boulton7 by processing wines in a bioreactor filled with too-nm diameter agarose spheres, to which an acid protease (from Aspergillus niger) has been covalently bonded.

TABLE 7 EFFECTS OF IMMOBILIZED ACID PROTEASE ON PROTEINACEOUS STABILITY

OF WINES

pH Etlumol Treatment conditions Stobility (%)

Time (h) Temp. Loadinga

Riesling 3·00 12·0 1 30aC (86"F) 189 + 2 30aC (86"F) 189 + 3 30aC (86"F) 189 +

Sauvignon blanc 3·65 14·2 0·5 3WC (86"F) 226 1 3WC (86"F) 226 1·5 3WC (86"F) 226

Chardonnay 3·57 12·9 1 3'rC(98"F) 238 2 37·C (98"F) 155

Gewurz-traminer 3·59 13-6 1 37·C (98"F) 238 2 37·C (98"F) 155

ami wine/g dry weight conjugate.


It should be noted that immobilized protease has reached proteinaceous stability in one wine only, though active in all of them. This fact could be accounted for by the characteristics of the enzyme activity: hydrolysis of some proteinaceous fractions in the wine is either absent or occurs in part up to ftocculable peptone derivatives.

At present, the deproteinization of wines by immobilized protease depends on two factors: (a) a new knowledge of wine protein structure and behaviour, and (b) a chance to spot a pool of acid protease which have their own specificity and can globally hydrolyse non-ftocculable sub-unit proteinaceous molecules.

For this purpose the research carried out in the field of isolation, characterization and immobilization of grape endogenous protease is particularly relevant: these proteases could be particularly active on the proteins derived from the same grapes.

PROCESSING TO AVOID MADEIRIZATION

A basic requisite of wines is that their whole optimal organoleptic properties should remain unchanged until consumption. To obtain such a result special care should be given to common white wines. Consumers' taste demands freshness and fruitiness within a limit of a sufficient stability of their first year.

Consumers in general like white wines limpid, slightly or weakly yellow coloured and with an average acid taste. Madeirization in white wines occurs most frequently along with turbidity, colour intensification, changes in aroma and taste which become rounder, more delicate and Marsala-like. It is therefore of the utmost importance to obtain and stabilize the above mentioned characteristics.

Their change is due to a complex sequence of events, where polyphenols play an important role: these substances are changed by oxidation reactions which occur in the presence of oxidizers and pro-oxidizers, and which are catalysed by many substances (iron, enzymes, calcium, amino acids, proteins, etc.). This oxidation involves all grape polyphenolic compounds, though with different kinetics.

In must preparation conditions there occurs both interaction between enzyme oxidation (catalysed by catecholase and laccase) and physical-chemical oxidation, catalysed by quinone accumulation. The oxidizability8 of some polyphenols as a regression percentage of the starting concentration is reported in Table 8.

238 ARTURO ZAMORANI

TABLE 8 GRAPE PHENOLS STABILITY: REGRESSION PERCENTAGE

WITH RESPECT TO INITIAL CONCENTRATION

( + ) Catechin ( - ) Epicatechin Procyanidin Bl Procyanidin B2 Procyanidin B3 Procyanidin B4 Gallic acid

Enzymatic (after 100 min)

15 25 25 85 65 55 65

Autocatalytic (after 50h)

32 27 45 76 48 65 18

The reactivity of oxidation products with amino acids, hexose and alcohol causes the formation of chromophores and volatile substances to occur; these substances in their turn, cause the already mentioned loss in limpidity, yellowing up to darkening, and turbidity.

In traditional wine technology the prevention of madeirization can avail itself of stabilizing procedures which either act on catalytic factors, or block oxidizers, or remove polyphenols. The latter can be considered as the most suitable for that purpose: they are defined therefore, as anti-madeirizers.

Proteinaceous, clarification, use of polyamides, and high doses of sulphur dioxide can be considered as very important nowadays. These methods, however, are neither effective, nor legally permitted, nor toxicologically safe. We have also found that an indiscriminate removal of polyphenols can lead to an undesired taste flattening, without reaching a better stabilization than using the catechin and anthocyanidin removal processes.

The decolourizing process using high doses of active carbons may give undesired solubles; yet it is not more effective than in selective processing.

An alternative to chemical-physical adsorbents can be the use of enzymes acting on the polyphenols which started madeirization; these polyphenolic substances are hydrolysed and oxidized by such enzymes so as to be previously destabilized by polymerization and flocculation, or to be easily and completely removable by clarification. 8

The enzyme preparations which are available for this purpose contain enzymes active on polyphenolic substances. The most interesting of them, according to Cantarelli8 are listed in Table 9.


Oxidases

EC 1.11.1.7. Peroxidase

TABLE 9 PHENOLIC ENZYMES

Hydro/ases

EC 3.1.1.20. Tannin-acilhydrolase (tannase)

EC 1.10.3.1. o-Diphenoloxidase EC 3.1.1. Anthocyanase (catecholoxidase, polyphenoloxi-dase, phenolase, tyrosinase) EC 1.10.3.2. p-Diphenoloxidase (Iaccase) EC 1.13.1.1. Catechol 1,2 oxygenase EC 1.13.1.3. Protocatechuic-acidoxygenase

Transferases

EC 2.1.1. 6. Methyltransferase (catechol-omethyltransferase )

The possibility of obtaining the elimination or reactivity loss of the polyphenols responsible for instability was investigated using the following enzyme preparations:

-tannase, probably from Aspergillus; -phenolase, prepared as an extract or acetoin powder from

Agaricus campestris; -laccase, from Polyporus versicolor in liquid, partly purified

culture; -industrially-made anthocyanase.

The enzyme treatments were compared later on; the most successful

TABLE 10 POLYPHENOL CONTENT AND COLOUR INCREASE FOUND IN MADEIRIZATION

TEST OF PINOT WINES OBTAINED BY VARIOUS ENZYME TREATMENTS

Without Polyporus Agaricus Tannase Anthocyanase treatment laccase phenolase

Polyphenol: Total 339 310 319 333 330 Non-tannic 267 261 284 284 283

Colour: OD 420 nm-l·OOO cm 92 82 78 64 72 increase due to made- 121 95 160 156 174 irization test

TA

BL

E 1

1 P

OL

YP

HE

NO

L.

CO

LO

UR

A

ND

S

TA

BIL

ITY

V

AL

UE

S

OF

T

RE

BB

IAN

O

WIN

ES

T

RE

AT

ED

W

ITH

L

AC

CA

SE

A

ND

S

ILIC

A

SO

LU

TIO

N

AS

CO

MP

AR

ED

W

ITH

CA

SE

INA

TE

AN

D A

CT

IVE

CA

RB

ON

AN

D B

EN

TO

NIT

E-T

RE

AT

ED

W

INE

S

Vin

tage

198

4 V

inta

ge 1

985

Enz

yme

C+

C+

B

C+

C+

B

Enz

yme

C+

C+

B

C+

C+

B

S02-

oniy

tr

eatm

ent

wit

hout

w

ith

trea

tmen

t w

itho

ut

with

tr

eatm

ent

S02

S02

S02

S02

Pol

yphe

nols

: T

otal

26

5 24

2 24

0 20

1 20

0 21

3 21

2 N

on-t

anni

c 18

8 18

6 18

0 17

9 18

1 19

6 19

3 C

olou

r:

OD

420

nm

-1·0

00 c

m

161

126

111

77

108

87

111

incr

ease

du

e to

m

adei

riza

tion

tes

t:

In b

ottl

e 64

73

61

28

59

13

24

In

via

l 83

21

0 17

6 14

3

C +

C +

B =

trea

tmen

t w

ith

case

inat

e (1

6 g/

hl),

act

ive

carb

on (

8·5

g/hl

) an

d be

nton

ite

(80

g/hl

).


one was compared with clarification carried out with casein and active carbon and bentonite, with or without sulphur dioxide.

The polyphenol concentration in wines derived from the same, differently processed, Pinot mixed grapes (Table 10) showed that laccase was more effective than other enzymes in decreasing the titre of said compounds. In madeirization assays colour increase values confirmed that result.

If we want processing to be successful we have to carry out sulphitation and clarification along with filtration, in order to remove oxidation products. A fairly comprehensive work cycle is as follows:

Grapes-pressing, decIustering, separating, free run must, enzyme processing, fermentation in the presence of silica solution, decanting with addition of sulphur dioxide, refrigeration, filtration, inert conditioning, bottling.

In Table 11 the results obtained with Trebbiano wine processed

TABLE 12 ORGANOLEPTIC ANALYSES OF LACCASE· AND SILICA SOLUTION· TREATED

WINES AS COMPARED TO CASEINATE· AND ACTIVE CARBON- AND SILICA

SOLUTION-TREATED WINES: 0-5 SCORE; AVERAGE VALUES AND VARIATION

COEFFICIENTS

Vintage 1984 Vintage 1985

Enzyme" C+C+B S020nlyb Enzyme" C+C+B S020nlyC treatment without treatment treatment without treatment

S02 S02

Colour 2·25 (53) 2·62 (42) 2·50 (42) 4·55 (11) 3·11(24) 4·11(21)

Odour: Fruit 1·37 (95) 2·00 (56) 2·00 (61) 2·22 (59) 2·00 (41) 2·55 (49) S02 2·50 (56) 1·87 (62) 1·62 (52) 2·00 (41) 2·33 (35) 1·55 (53)

Flavour: Body 1·62 (52) 2·37 (36) 1·87 (31) 2·77 (37) 2·00 (40) 2·77 (28) Acidity 2·87 (53) 2·75 (30) 3·25 (25) 2·22 (19) 2·88 (25) 2·55 (28) Astringency 1·50 (94) 1·75 (37) 1·62 (52) 2·22 (46) 2·22 (28) 2·22 (41) Off-tlavour 2·37 (59) 1·25 (87) 0·12 (264) 0·55 (149) 0·33 (141) 0·44 (54)

Persistency 1·87 (67) 2-62 (26) 2·37 (20) 2·55 (27) 2·11 (27) 2·77 (28)

C + C + B = treatment with caseinate (16 g/hl), active carbon (8·5 g/hl) and bentonite (88 g/hl). "Polyphenollaccase, 6000 U/hl. b S02 addition in must 5 g/hl. c S02 addition in must-wine 3·5 g/hl.

242 ARTURO ZAMORANI

within the above mentioned sequence (laccase + silica solution) are compared with the results obtained by clarification (caseinate + active carbon + bentonite) in the presence or absence of sulphur dioxide. The use of the enzyme was shown to be highly effective, preferable or practically identical to the results obtained by traditional processing (caseinate + active carbon + bentonite) (Table 11).

Even organoleptic analysis confirmed that wines treated with laccase and silica solution or sulphur dioxide show different, better organoleptic madeirization-resistant features (Table 12).

mE USE OF ENZYMES TO DECREASE MALIC ACIDITY OF WINES

Malo-lactic fermentation is essentially a decomposition of malic acid into lactic acid and carbon dioxide, which occurs in wine through anaerobic lactic bacteria. The result is a decrease in total acidity. We can outline the whole reaction of malo-lactic fermentation as follows:

I-malic acid-I-lactic acid + CO2

A decrease in malic activity is necessary for the acquisition of effective organoleptic features in a number of red wines. In order to gain the soft full taste of a ripe product, these wines must lose that slightly bitter sourness which is due to malic acid.

Malo-lactic fermentation occurs by the activity of Lactobacillus, Leuconostoc and Pediococcus, which find suitable conditions in the wine-making process (usually in the spring after alcoholic fermentation). In modern technology it is difficult to induce spontaneous malo-lactic fermentations in low-pH wines containing sulphur dioxide, and also for the formation of secondary compounds, particularly acetic acid, which are unsuitable for the quality features of wine.9

In order to circumvent these difficulties, selected bacteria have been introduced into wine; but this attempt failed. If we were able to transform malic acid into lactic acid by the use of enzymes, the drawbacks of these microbial fermentations could possibly be overcome.

The chemical pathway of malo-lactic fermeqtation is fairly simple and involves the following enzymes:


-EC 1.1.1.4~malic NADP-dependent enzyme (EM); it causes the conversion of malic into pyruvic acid;

-EC 1.1.1.37-malic NAD-dependent dehydrogenase (MDH); -EC 1.1.1.27-lactic, NAD-dependent, dehydrogenase (LDH); it

causes the conversion of pyruvic into lactic acid; -EC 1.6. 1. I-nicotinamide nucleotide transhydrogenase, it trans

fers hydrogen from NADP to NAD; -EC (unclassified}-malo-Iactic NAD-dependent activity (MLA) ,

it transforms malic acid into lactic acid.

The first four enzymes were extracted from Raboso grapes, which are very rich in malic acid;1(}"'12 they are also commercially available. The last one was extracted from Leuconostoc oenos ML34.13--15

For malo-lactic enzyme transformation, two (immobilized) systems were proposed: in the former system the NAD-dependent EM was immobilized on Sepharose CL-4B and chitosan-glutaraldehyde in order to carry out the conversion from malic to pyruvic acid.16 For the next step from pyruvic to lactic acid, the NADH-dependent LDH was immobilized on the same matrix.

In order to obtain a steady malic-to-Iactic acid enzyme degradation, the use of an enzyme (TH) which catalyses hydrogen transfer from NADPH to NAD is necessaryY TH has been immobilized on the above mentioned matrixes.

Finally, the following multistage immobilized enzyme system was obtained

I-malic acid + 2NADP+ ~ 2NADPH + pyruvic acid + CO2

1 1 NADP+

TH NADPH --+ +

NADH NAD+

1 i LDH

2NAD+ + I-lactic acid pyruvic acid + 2NADH --+

This system was effective on aqueous malic acid solutions but it failed

244 ARTURO ZAMORANI

in wine processing. In the latter system MLA was immobilized on alginate calcium gels and used in wine processing. Probably owing to low wine pH, immobilized MLA which averages an excellent ca. pH 6, gave low results.

The most promising methods of controlling wine biological acidification seems to be the industrial application of immobilized cells of selected bacterial strains.l~22 Leuconostoc oenos ML34 were entrapped in alginate calcium gels and their behaviour was studied in a batch system. A maximum activity at pH 4, in the presence of 15% ethanol was found. After 36 days that activity was still equivalent to 60% of the maximum. In wine, immobilized microorganisms transformed I-malic into I-lactic acid with yields usually recorded in free-cell malo-lactic fermentations. It was found that the reaction had taken place fairly quickly and that transformation could be stopped at the required deacidification level. After repeated tests, however, we noted that Leuconostoc activity gradually decreased. 22

Continuous flow experiments were also carried out in which both malo-lactic activity and likely variations in wine features were evaluated. Leuconostoc oenos ML34 were immobilized in 2-mm alginate calcium gel and a 20°C (68°F) thermostatted glass column was filled with them. Nitrogen at 0·17 atm. and a peristaltic pump controlled the wine flow. The nitrogen pressure was aimed at avoiding air contact and at preventing carbon dioxide bubbles inside the column. This system proved effective but after only 4 days it was found that the activity was 60% of the initial value.

A pH increase, a decrease in total acidity, the lactic acid yield compared to the metabolized malic acid, were all in agreement with the values we usually find in spontaneous malo-lactic fermentation. 23

In another test Lactobacillus sp. 48 were immobilized in K

carrageenan in the presence or absence of Pentagel (purified bentonite), which increases gel porosity, thus facilitating mass transfers to the immobilized system. The subject of this microbiological deacidification was a Durello white wine. The results are listed in Table 13. The immobilized Lactobacilli proved to be very effective in malic acid degradation; pH variation, acidity decrease and lactic acid yield were the same as in normal malo-lactic acid fermentation.

Furthermore, while in previous experiments no Lactobacillus developed in wines with a pH of less than 3·5, malic acid decarboxylation was not delayed by the low pH value (3·15) of the starting wine. As to the reactive stability of the bioreactor containing


TABLE 13 CHEMICALS IN A DURELLO WINE BEFORE AND AFfER MALO-LACTIC

FERMENTATION BY 1·5 g OF Lactobacillus sp. 48 IMMOBILIZED IN 2% K

CARRAGEENAN IN THE PRESENCE OF 5% PENTAGEL (BENTONITE SILICA).

FLOW WAS 0·51 vol/h

Chemicals Before ML fermentation

Alcohol (% vol/vol) 11·00 Free S02 (mg/litre) 3·80 Total S02 (mg/litre) 57·60 Molecular S02 (mg/litre) 0·24 pH 3·15 Titratable acidity 15·67 (g/litre tartaric acid) I-Lactic acid (g/litre) 0·04 I-Malic acid (g/litre) 8·30 Acetic acid (g/litre) 0·12 Citric acid (g/litre) O· 36 Glycerol (g/litre) 4·36 Diacetyl (mg/litre) 0·90 Acetoin (mg/litre) 6·20 Fructose (g/litre) 0·26 Glucose (g/litre) 0·15 Conversion ratio"

" I-lactic acid produced 00 I-malic acid metabolized x 1

After ML fermentation

3·35 11·73

3·36 2·83 0·15 0·33 4·15 1·50 8·30 0·04 0·00

60·7

Differences

+0·20 -3·94

+3·32 -5·47 +0·03 -0·03 -0·21 +0·60 +2·10 -0·22 -0·15

Lactobacillus sp. 48, it lasted 46 days, much longer than previously recorded.

The possibility of inducing and controlling malo-lactic fermentation with immobilized microorganisms for fairly long periods has been confirmed by these experiments. It is necessary, however, to keep on researching in order to adapt bacteria to lower pH and to higher sulphur dioxide concentrations. Fermentation, therefore, can proceed under the same conditions as those wines which request malo-lactic transformation.

REFERENCES

1. Montedoro, G. (1976). Riv. Sci. Teen. Alim. Nutr. Um., 6, 133. 2. Montedoro, G. and Bertuccioli, M. (1976). Lebensm. Wiss. u. Technol.,

9,225.

246 ARTURO ZAMORANI

3. Castino, M. and Ubigli, M. (1979). Riv. Vitic. Enol., 32, 65. 4. Cantarelli, C. (1984). Ind. Alim., 23,845. 5. Garoglio, P. G. (Ed.) (1980). Nuova Enologia, Enciclopedia Vitivinicola

Mondiale, AEB, Brescia. 6. Gaina, B. S., Pavlenko, N. M., Datunashvili, E. N., Krylova, Yu. I.,

Kozlov, L. V. and Antonov, V. K. (1976). Appl. Biochem. Microb., 12, 117.

7. Bakalinsky, A. T. and Boulton, R. (1985). Am. I. Enol. Vitic., 36,23. 8. Cantarelli, C. (1986). Vini d'Italia, 28(3),87. 9. Kunkee, R. L. (1974). In: Chemistry of Winemaking, A. D. Webb (Ed.),

Adv. Chern. Ser. no. 137, Washington, DC, p. 151. 10. Zamorani, A., Spettoli, P., Bottacin, A. and Varanini, Z. (l97~. Riv.

Vitic. Enol., 32,354. 11. Spettoli, P., Bottacin, A. and Zamorani, A. (1980). Vilis, 9,4. 12. Spettoli, P., Bottacin, A. and Zamorani, A. (1981). Phytochemistry, 20,

29. 13. Spettoli, P., Nuti, M. P., Bottacin, A. and Zamorani, A. (1982). In: Use

of Enzymes in Food Technology D. Dupuy (Ed.), Technique et Documentation Lavoisier, Paris, p. 545.

14. Spettoli, P. and Zamoraili, A. (1983). In: Progress in Food Engineering, C. Cantarelli and C. Peri (Eds) Forster-Verlag-AG, Federal Republic of Germany, p. 377.

15. Spettoli, P., Nuti, M. P. and Zamorani, A. (1984). Appl. Environ. Microbiol., 48, 900.

16. Spettoli, P., Bottacin, A. and Zamorani, A. (1980). Tecnl. Aliment., 3, 31.

17. Spettoli, P. and Bottacin, A. (1981). Am. I. Enol. Vitic., 32,87. 18. Monsan, P. and Durand, G. (1976). Ind. Aliment. Agric., 93,543. 19. Divies, G. H., Siess, M. H. and Jeanblanc, M. F. (1979). In: Cellules

Immobilisees, J. M. Labault and G. Durand (Eds), S. F. M., p. 151. 20. Totsuka, A. and Hara, S. (1981). Hakkokogaku, 59, 231. 21. Gestrelius, S. (1982). In: Enzyme Engineering, I. Chibata, S. Fukui and

L. B. Wingard (Eds), Plenum Press, New York, p. 245. 22. Spettoli, P., Bottacin, A., Nuti, M. P. and Zamorani, A. (1982). Am. I.

Enol. Vitic., 33, 1. 23. Spettoli, P., Nuti, M. P., Dal Belin Peruffo, A. and Zamorani, A. (1984).

In: Enzyme Engineering, I. A. Laskin, G. T. Tsao, Jr and L. B. Wingard (Eds) Annals of the New York Academy of Science, New York, p. 461.

INDEX

Acetaldehyde carbon dioxide effects on formation

of, 58, 62-3 production by immobilised cells,

160, 161 Acetic acid

factors affecting formation, 138, 144

formation in fermentation, 149 in fruit juices, 68

Acetobacer spp., fruit juices spoilt by, 71

Acetohydroxy acids, formation in brewing, 79,80

Acetolactate decarboxylases, 80 Acinetobacter spp., citrus fruit juice

debittered by, 206 Acremonium spp., fruit juices spoilt

by, 73 Activators, 135-6 Active carbon, wines treated with,

238,240,241,242 Active dry wine yeast, 171

drying of, 183-4 rehydration of, 182 specification for, 185 see also W ADY

Alcohol tolerance threshold, 131-2 factors affecting, 132

Alcoholic fermentation, microorganisms for, 18

Aldehydoxidase, 208 Alginate-entrapped systems

advantages of, 134 cell distribution in, 157-8

247

Alginate-entrapped systems--contd. preparation of, 156 productivity of, 84-5

Alternaria spp., fruit juices spoilt by, 72

Amino acids, absorption affected by carbon dioxide, 59-60, 61

Amylase, fruit juices treated by, 205 Anaerobiosis, fermentation affected

by, 139 Anthocyanase, 239 Anthocyanins, 137

effect of enzymic processing on, 228,229,230,231

Antibiotic resistance markers, 33 Apple juice, clarification of, 204,

216-17 Apple pulp, enzymic treatment of,

197, 198 APV tower system, 82 Arabinase, 191, 202 Arthrobacter globiformis citrus fruit

debittered by, 191, 206, 212 Aspergillus spp.

fruit juices spoilt by, 72, 73, 74 immobilised reactors using, 157

Aureobasidium pullulans, 3 Australian wines, killer yeasts in, 43

Bacillus spp., fruit juices spoilt by, 69-70

Bacteria, pH range for growth, 65 Barbera wines, 233, 234 Barley grains, polysaccharides in, 37

248 INDEX

Batch fermentation, disadvantages of,101

Beer continuous fermentation process,

100-1 fermenter design, difficulties

encountered, 94 flavour

environmental factors affecting, 78-80

equipment design affecting, 81-2 Beet molasses, composition of, 173-4 Bentonite, wines treated with, 240,

241,242 Bioreactors, 98-102 Bitter compounds, citrus fruit, 205 Botrytis spp., fruit juices spoilt by, 72 Bottle-fermented sparkling

beverages, immobilised yeast used,l64

Brettanomyces spp., 19,25 Brewing, immobilised yeast used, 164 Butyric anaerobes, 68 Byssochlamys spp., fruit juices spoilt

by, 73

Candida spp., 19,21,25, 72 immobilised reactors using, 157 killer strains, 42

Cane molasses, composition of, 173 Canned foods, pH criteria for, 68 Canned fruit juices, microbial

spoilage of, 68-74 Carbamates, formation of, 51-2,

53-4 Carbamoyl phosphate, formation of,

52 Carbon dioxide

growth inhibition by, 53-63 incidental control by, 61-3 instantaneous flow rate in white

wine batch fermenters, 115-16 mass action effects of, 53-4 physico-chemical properties of,

50-2 physiological effects of, 52-63 solution in complex media, 51-2

Carbon dioxide--contd. solution in water, 50 solution within the cell, 52 yeast growth affected by, 55-63

Carbonic acid, 50 Carrot juice, enzymic treatment of,

209 Caseinate, wines treated with, 240,

241,242 'Casse', 236 Cell division, carbon dioxide effects

on,55 Cell growth, factors affecting, 143 Cell size, carbon dioxide effects on,

56-7 Cellobio-hydrolase, 226 Cellulases, 190

fruit pulp treated with, 201-2 grape musts processed using, 227-

8 red wine headspace volatiles

affected by, 232 Cellulolytic activity enzymes, 205,

226 Centraalbureau voor

Schimmelculture (CBS-Delft), 4

Champagne-making, entrapped yeasts used, 163-4

Cider production apple juice clarified for, 204 immobilised yeast used, 164

Citeromyces matritensis, 19,25 Citric acid, in fruit juices, 67 Citrus fruit juice by-products,

enzymic treatment by, 200-1 Citrus fruit juices

debittering of, 205-7 enzymatic autoclarification of, 196 enzymic treatment of, 204

Citrus fruit peel, enzymic treatment of,201

Cladosporium spp., fruit juices spoilt by, 72

Clonal strain selection, yeasts, 26 Clostridium spp.

fruit juices spoilt by, 68-9 heat resistance of, 68

Continuous fermentation process, beer, 100-1

Continuous fermentation processes, wine, 128, 129

INDEX

Continuous flow stirred tank (CFST) reactors, fruit juice enzymic clarification, 214, 216

Continuously-stirred tank reactor (CSTR), characteristics of, 102

Corynebacterium /ascians, citrus fruit debittered by, 191,206,212

Crabtree effect, 134-5 Cryptococcus spp., 19,25

Debaryomyces spp., 19,25 Debittering enzymes, 205-7 Depolymerases, 198, 224-5 Diacetyl

formation of, 37-8, 79-80 carbon dioxide affecting, 63

Diacetylreductase, 207-8 DNA, cell content, carbon dioxide

effects on, 55-6 Dried yeast, 132-3 Durello wine, malo-lactic

fermentation in, 245

Ecological surveys, 3 Enterobacteriaceae

fruit juices spoilt by, 70 heat resistance of, 70

Entrapment methods, 155-6 polymers used, 155-6

Entrapped cell systems, industrial applications of, 162-5

Entrapped microorganisms, kinetics of,156-8

Entrapped yeasts, behaviour of, 160-62

Enzyme immobilisation advantage of, 153 applications of, 153

Enzymes carbon dioxide effects on, 53-4 reactions of, 153 use in juice production, 189

249

Enzymic processing applications in winemaking, 235 musts, 224-35 wines, 235-45

Episomal vectors, 34-5 Escherichia coli, use in recombinant

DNA technology, 33, 38 Esters

carbon dioxide affecting, 59, 62 equipment design affecting, 81-2 immobilised cell systems affecting,

89 temperature affecting, 79

Ethanol carbon dioxide solubility affected

by, 51 production of

carbon dioxide affecting, 62 factors affecting, 143

Expression vectors, 36

Fatty acids, immobilised yeast system, 88-9

Fed-batch process, yeast produced by, 171

Fermentation, carbon dioxide affecting, 57-8, 61

Fermentation activity, dry yeast, 186 Fermentation byproducts, factors

affecting, 143-4 Fermentation efficiency

alginate-entrapped yeast system, 84-6

environmental factors affecting, 78, 80

equipment design affecting, 80-1 Fermentation rate, factors affecting,

77,142 Fermenters

completely segregated, 102 design, effects of, 80-2 design of, 98-105 design requirements, 94-5 future prospects for, 105 heat transfer in, 104, 105 mass transfer in, 102-4 partially segregated, 102

250 INDEX

Fermenters-contd. perfectly mixed, 101-2 yeast production, 181

Filobasidium spp., 19,25 Flat-sour spoilage, 70 F1avans, 137 Flavour-active compounds

carbon dioxide effects on production, 58-9, 62-3

immobilisation affecting formation of,89

operating conditions affecting, 140 temperature effects on formation

of,78-80 F1uidification processes 199-200 F1uidised-bed reactor, ~haracteristics

of, 102 Freisa wines, 233, 234 French wines, killer yeasts in 43 Fruit juices '

acids in, 67-8 clarification using immobilised

enzymes, 216-17 effect of enzymes on, 189 enzymes used, 190-91 immobilised enzymes used in

production, 211-12 microbial spoilage of, 68-74 pH values for, 67, 68 production of, 197 soluble enzymes used in production

of,196-203 treatment with immobilised

enzymes, 209-12, 216-17 treatment with soluble enzymes

203-8 ' Fruit pulp

cellulasesused, 201-2 enzyme treatment of, 196-203 ftuidification of, 199-200 maceration of, 198-9 pectic enzymes used, 196-201 proteases used, 202-3

Fruit and vegetables cellular structure of, 192 enzymes in, 192

Fusarium spp., fruit juices spoilt by 74 '

Fusel oils, carbon dioxide effects on formation of, 59, 62

Galactanase, 190, 202 Gassing power, dry yeast, 186, 187 Gel occlusion systems, 82 83 Genetic manipulation, 31~8

possible goals of, 37 Geotrichum spp., fruit juices spoilt

by, 72 Glucanases, 190, 226

fruit pulp treated with, 201 Il-Glucanases, 33, 37 Glucose oxidase, 207 Il-D-Glucosidase, 191,205 226 Glucoxidase, 191 ' Glycerol, production by immobilised

cells, 160, 161 Grape musts

composition of, 17, 129 enzymic processing of, 224-35 pH range of, 17 yeasts on, 2-3, 25

Growth inhibition by carbon dioxide 55-63

mechanisms for, 53-4 ' pH effects on, 65-7

Growth factors, yeasts affected by 135-6 '

Guanine-cytosine percentage relationship, 5, 6

Guinness Laboratories, 37

Hansen, Emil Chr., 178 Hanseniaspora spp., 19,21,25, 72 Hansenula spp., 19,25 72

killer strains, 42 ' Heat transfer, 140-1

fermenters, 104, 105 Hemicellulases

fruit pulp treated by, 202 grape musts processed using, 227

Henry's law, 50 Higher alcohols

immobilised cell systems affecting, 89, 160, 161

temperature affecting formation, 79

High-yeast fermentation, 145-9 immobilised system used, 146, 165 limitations of, 149 pilot-plant results, 147, 148 processing techniques used, 146

Hydroxycinnamic acid, 137

Immobilisation adsorption used, 154-5 entrapment methods used, 155-6 fixation used, 155 flavour-active compounds affected

by,89 flocculation used, 155 membrane retention used, 156 substrate utilisation rates affected

by, 86-7 techniques for, 101, 154-6

Immobilised cell systems, industrial applications of, 162-5

Immobilised enzymes cost considerations, 210-11 juices treated with, 209-12, 216-

17 legislative problems, 210 scientific-technological problems,

210 Immobilised microorganisms

fruit juices treated with, 191 malo:-Iactic fermentation using,

244-5 Immobilised proteases, white wine

clarified using, 236-7 Immobilised yeast systems

advantages of, 83 brewing use of, 82-3 equipment used, 134 fatty acid synthesis in, 88-9 mass transfer limitations of, 87-8 objectives of, 134 productivity of, 84-5

Immobilised-cell bioreactors characteristics of, 102 problems with, 103-4

In-bottle fermentation, 163-4 Integrating vectors, 33-4 Issatchenkia spp., 19, 25

INDEX 251

Japanese wines, killer yeasts in, 43 Juice extraction processes, 192, 193

Killer toxin identification of, 44 optimum conditions for

production, 44-5 temperature effects, 44, 46

Killer yeasts characteristics of, 41 first discovered, 41 genetic introduction of, 27 inheritance of character, 43 occurrence in nature, 42-3 potential utilisation of, 45-6

Kloeckera apiculata, 2,3,25,169 Kloeckera spp., 19,21,25 Kluyveromyces drosophilarum, 35 Kluyveromyces spp., 19,25

killer strains, 42

Laccase, 239, 240, 241, 242 Lactic acid bacteria

entrapped, behaviour of, 158-60 fruit juices spoilt by, 71-2

Lactobacillus delbrueckii, 38 Lactobacillus spp.

fruit juices spoilt by, 71 immobilised system for malo-lactic

ferm~ntation, 244-5 Lager production, 57, 100 Lemon juice, clarification by

immobilised enzyme, 216-17 Leuconostoc oenos

enzyme extracted from, 243 immobilised system for malo-lactic

fermentation, 244 malo-lactic fermentation started

by, 18 Leuconostoc spp., fruit juices spoilt

by, 71 Limonene, 205 Limonin, conversion of, 205-6, 207,

212 Lipid composition, carbon dioxide

affecting, 55

252 INDEX

Lodderomyces elongisporus, 19,25 Low-alcohol drinks, immobilised cell

reactors used, 165

Macerases, 198-9 Maceration process, 198 Macromolecular constituents,

entrapped cells affecting, 162 Macromolecular relationships,

taxonomy use of, 5-6 Madeirization

effects of, 237 processing to avoid, 237-42

Magareteh Institute, killer yeasts in, 42

Malic acid, in fruit juices, 67 Malo-lactic fermentation, 242

enzymes involved, 243 immobilised bacteria used, 159-60,

162-3 immobilised enzyme system used,

243-5 microorganisms for, 18,24

Mass action effects, carbon dioxide, 53

Mass transfer fermenters, 102-4 immobilised yeast system, 87-8

Mathematical models, white wine batch fermenters, 112-16

Membrane properties, carbon dioxide affecting, 54

'Methode Champenoise', 163-4 Metschinikowia pulcherrima, 3, 25,

169 Metschinikowia reukaufi, 25 Microorganisms, reactions of, 153 Molasses

composition of, 173-5 requirement for yeast production,

170, 171 storage of, 176 treatment of, 176-8 vitamin content of, 173, 174-5

Molecular taxonomy, 5 Monod equation, 97

Moulds fruit juices spoilt by, 72-4 pH range for growth, 65, 72

Mucor spp., fruit juices spoilt by, 72, 73,74

Must pressed preparations, enzymic processing of, 224-35

Naringin, 205 enzymic hydrolysis of, 205, 206,

211-12 Nathan vessel, 100 National Collection of Yeast Cultures

(NCYC), killer yeasts in, 42 Nectars, production of, 199 New Zealand wines, killer yeasts in,

43 Nitrogen

fermentation affected by, 135, 144 requirement for yeast production,

170 supply for yeast production, 174,

181 Nomilin, debittering of, 206, 208, 212

Oligogalacturonate lyase, 194 Operating conditions, fermentation

affected by, 139-49 Optimisation, fermentation

conditions, 141-5 Optimisation procedure, white wine

batch fermenters, 116-19 Orange juice, concentration of,

204-5 Oxaloacetate, formation of, 52 Oxidoreductases, 207-8 Oxygen, requirement for yeast

production, 170 Oxygen availability, yeast affected

by, 138-9

Pachytichospora transvaalensis, 19, 25

Packed-bed reactor, characteristics of, 102

INDEX 253

Pasteur effect, 132, 169 Pectatelyase, 194 Pectic enzymes

citrus fruit-juice by-products treated with, 200-1

fluidised fruit pulp affected by, 199-200

fruit pulp treated with, 196-8 macerated fruit pulp affected by,

198-9 Pectin chain, attack of pectolytic

enzymes on, 195 Pectin enzymes, 195 Pectin substances, 195, 196 Pectinases

endogenic, role in beverage production, 195-6

grape musts processed using, 227-9 red wine headspace volatiles

affected by, 232 Pectinesterase, 190,204

immobilisation of, 212, 214 Pectinlyase, 190, 194, 202

fruit juice treated with, 204 fruit pulp treated with, 202

Pectolytic enzymes fruit juices treated with, 203-5 fruit pulp treated with, 196-201 labelling of, 234-5 pectic chain attacked by, 195 vegetable juices treated with, 209

Pediococci spp., fruit juices spoilt by, 71-2

Penicillium spp., fruit juices spoilt by, 72,74

Pentaumedione, formation of, 37 Perugia University

Industrial Yeast Collection, taxonomic study, 4, 7

Wine Yeast Collection, killer yeasts in, 42

'Petite-colonie', 133 PH

fruit juices, 67, 68 microbial growth affected by, 65-7 yeast growth range, 65, 136

Phenolase, 239 Phenolic enzymes, 239

Phenolics, fermentation affected by, 137-8

Phenols, grape, stability of, 238 Phosphate

requirement for yeast production, 170

supply for yeast production, 174 Pichia spp., 19,25,72

killer strains, 42 Piemonte wines, enzymatically

processed, 232-4 Pigments

fruit juice, 197 enzymic treatment of, 208

wine, 137 Pinot wines, enzymically treated, 239,

241 Polyethylene glycol, spheroplast

fusion by, 32, 33 Polygalacturonases, 190, 194

immobilisation of, 212-13, 215 Polygalacturonatelyase, 190 Polygalacturonoxidase, 194 Polymethygalacturonase, 190 Polyphenols

effect of enzymic processing on, 228,229,230,231,239

enzymes acting on, 208 enzymes affected by, 197 organoleptic effect of removal, 238 yeast affected by, 137-8, 142

Pressure fermentation systems, carbon dioxide used in, 57-8

Proteases fruit juices treated by, 205 fruit pulp treated with, 202-3 grape musts processed using, 227-8 red wine headspace volatiles

affected by, 232 white wine clarified using, 235-7

Proteinase, 191 Proteins, cell content, carbon dioxide

affecting, 55 Proteolytic enzymes

classification of, 224-5 requirements for, 234

Protoplast fusion, genetic improvement by, 27, 31-2

254

Pseudomonas putida, immobilised reactors using, 157

Raboso grapes, 243 Recombinant DNA technology,

yeasts improved using, 32, 36-8

Red grape phenolics, fermentation affected by, 138

Red wine effect of enzymic processing on

headspace volatiles, 232 starters used, 133

Refrigeration flow rate, white wine batch fermenters, 114

Replicating vectors, 35 a--L-Rhamnoxidase, 191, 205 Rhizopus spp., fruit juices spoilt by,

72,74 Rhodorotula glutinis, 25

entrapped use, 161 Rhodotorula spp., 19,25 RNA, cell content, carbon dioxide

affecting, 56 Rose wines, production of, 144 Rosso Conero grapes, 231 Rosso Piceno wines, 232

Saccharomyces spp. characteristics of, 18 classification of, 19,20-1,22-3 fruit juices spoilt by, 72 interspecies relationships, 9, 12 nomenclature for, 1-13 practical approach to classification

of,6-12 Sacch. bayanus, 6, 7, 8, 9, 11, 12,

21,131 analytical data, 186

Sacch. carlsbergensis, 6,9, 10, 12, 58

Sacch. cerevisiae alginate-entrapped system, 160 analytical data, 186 carbon dioxide effects on, 59, 63 chromosomal aspects, 3, 9

INDEX

Saccharomyces spp.--contd. Sacch. cerevisiae--contd. composition of wines made

using, 131 immobilised reactors using, 157 killer strains, 41, 43, 44, 45-6 mutants induced in, 28 'neutral' strains, 41, 46 'sensitive' strains, 41 starter activity of, 133 taxonomy, 6-12 technologically diverse species,

3,6,8,21 traditional use of name, 6

Sacch. chevalieri, 7, 8 Sacch. ellipsoideus, 2, 6, 10, 12, 21 Sacch. exiguus, 20, 21 Sacch. fructuum, 7, 11 Sacch. italicus, 7,8,9,11 Sacch. oviformis, 6, 7, 11, 12 Sacch. paradoxus, 6,9, 12 Sacch. pastorianus, 6,9, 11, 12 Sacch. rosei, 131 Sacch. uvarum, 8,9, 10, 21

taxonomy of, 10-11, 18, 20-1, 22-3

traditional names, 6, 13 see also Yeast

Saccharomycodes ludwigii, 19,21,25 Saccharomycopsis spp., 19,25 Sak, S0ren, 171 Schizosaccharomyces pombe, 1,25 Schizosaccharomyces spp., 24, 25, 28

classification of, 19 Secretion vectors, 36 Silica, wines treated with, 241-2 Sodium flame test, 175 Sparkling drinks, immobilised yeast

used,l64 Specific gravity, carbon dioxide

effects on, 62 Spheroconical fermenters, 99, 100 Spheroplast fusion, genetic

improvement by, 27, 32 Spontaneous fermentation, yeasts

from, 21 Starch haze, enzymic treatment for,

205

INDEX 255

Starter cultures fermenting activity of, 132-3 microorganisms used, 18 selected microorganisms for, 18

Stone-fruit pulp, enzymic treatment of,201

Streptococci spp., fruit juices spoilt by, 71

SUCgenes, 7 Sugar content, fermentation affected

by, 134-5, 142, 143 Sulphur dioxide

effects of, 18 in molasses, 174

Superquatre system, 128 Survival factors, 136, 139

Tannase, 239 Taxonomy, criteria used, 3-4, 5-6,

18,20 Temperature effects, fermentation

processes, 140 Thermal balance

fermentation, 140 yeast production, 171

Tomato juice concentration of, 209 microbial contamination of, 69, 70 pH of, 67

Torulaspora delbrueckii, 1,20,25 Torulaspora rosei, 20 Torulaspora spp., 19,21,25

characteristics of, 20 classification of, 19,20,22-3

Torulopsis stellata, 167 Total module factor, 117 Tower fermenters, 82, 99, 100, 101 Trebbiano wines, enzymically

treated, 240,241-42 Trichosporon spp., 19,25

Vegetables, enzymic treatment of, 209

Vicinal diketones carbon dioxide affecting, 58, 59, 63

Vicinal diketones-contd. formation in brewing, 79, 80

Vitamins demand by yeast, 171, 175 supply by molasses, 173, 174-5

WADY drying of, 183-4 rehydration of, 183 specification for, 185 see also Yeast

Waste-water treatment, yeast factory, 184

White wine fermentation closed system, 111, 122 kinetics of, 95-8 mathematical modelling of batch

fermenters, 112-16 instantaneous carbon dioxide

flow rate for, 115-16 instantaneous refrigeration flow

rate for, 114 objectives of development work,

129 open system, 111, 121, 122 optimal design of batch fermenters,

119-24 optimisation procedure for batch

fermenter design, 116-19 input data required, 118

optimisation procedures, 141-5 planning maps for optimisation of,

142-5 starters used, 133

White wines, enzymic clarification of, 235-7

Wickerhamiella domercqii, 19,25 Wine, see also Red ... ; White

wine ... Wine fermentation, 95-8

factors affecting, 95 kinetics, 96-8

Wine production enzymatic techniques

improvement made by, 224 reasons for non-use, 223

Wine yeasts, 6, 8, 9, 20, 21

256 INDEX

Wine yeasts--contd. active dry, 182-4

see also WADY selected strains, 9, 131, 169

Winery design of, 119-24

closed fermentation system used, 122

cooling system data, 120 daily flowrate data, 119 open fermentation system used,

121, 123 equipment costs of, 120 operating costs of, 120, 124 yeast colonization of, 3

Wines, enzymic processing of, 235-45 Wooden casks, 95

Xylanase, grape musts processed using, 227

Yeast (Sacch. spp.) analyses required, 185-7 basis of production, 171-2 composition of, 170 cultivation of, 179-82 drying of, 183-4 fermentation activity of, 186 fluid-bed drying of, 184 gassing power of, 186, 187 infecting micro-organisms in, 186,

187 production of

air filtration used, 175 I)ir injection for, 175, 181 anti-foam agents used, 182 equipment used, 176-9 fermentation processes, 179-82 molasses used, 173-5, 180 outline of process, 171-2 propagation plant used, 178, 179 raw materials used, 172, 173-6 separation process, 182 waste water produced, 184 water quality required, 175, 176 yeast vats used, 178-9

Yeast (Sacch. spp.}-contd. propagation of, 178 separation of, 182 shelf life of, 182 specifications for, 185 see also WADY

Yeasts auto-inhibition of, 145-6

methods to overcome, 146 characteristics of, 24, 26, 130-4 classical methods for improvement

of,27 classification of, 19,20-1,22-3

historical background, 3-5 practical approach, 6-12

clonal strain selection of, 26 definition of, 18 differences between strains, 130-2 ecology of, 21, 24 entrapped/immobilised

behavior of, 158-60 distribution of, 157-8

episomal vectors, 34-5 expression vectors, 36 fermentation efficiency of, 132 fruit juices spoilt by, 72 genetic manipulation of, 31-8

brewing strains, 37-8 wine strains, 38

growth factors affecting, 135-6 growth stimulated by carbon

dioxide, 52-3 inhibitory effects of carbon dioxide

on,55-63 integrating plasmids, 33-4 lag phase of growth, 97 mutants induced in, 28 operating conditions affecting,

139-49 origins of, 2-3, 21, 24 oxygen availability affecting, 138-9 pH range for growth, 65, 136 replicating vectors, 35 secretion vectors, 36 selected cultures used, 24, 26 selection criteria, 130 selection methods for, 26-8 spontaneous fermentation, 21

Yeasts----contd. substrate effects on, 134-9 sugar content effects on, 134-5 temperature effects on, 140 transformation of, 32-6

INDEX

Zygosaccharomyces spp., 19,25 characteristics of, 20 classification of, 19, 20, 22-3 plasmids in, 35

257

Biotechnology Applications in Beverage Production

Documents

Transcript of Biotechnology Applications in Beverage Production