Molecular Wine Microbiology

MOLECULAR WINE MICROBIOLOGY

Edited by

ALFONSO V. CARRASCOSA

ROSARIO MUN˜ OZ

RAMO´ N GONZA´ LEZ

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Academic Press is an imprint of Elsevier Academic Press is an imprint of Elsevier 32 Jamestown Road, London, NW1 7BY, UK 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1800, San Diego, CA 92101-4495, USA First edition 2011 Copyright Ó 2011 Elsevier Inc. All rights reserved Translation by Anne Murray and lain Patten 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 the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively, visit the Science and Technology Books website at www.elsevierdirect.com/rights for further information

Notice 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. Because of rapid advances in the medical sciences, in particular, independent veriﬁcation of diagnoses and drug dosages should be made

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-12-375021-1

For information on all Academic Press publications visit our website at www.elsevierdirect.com

Typeset by TNQ Books and Journals Printed and bound in the United States of America 11 12 13 14 15 10 9 8 7 6 5 4 3 1 Preface

The publication of Louis Pasteur’s Me´moire It is hoped that Molecular Wine Microbiology sur la fermentation alcoolique in 1857 has come to will be a useful tool for researchers and educa- represent a milestone in the history of science tors working in both the private and public and its applications, as it marked the beginning sectors. Above all, however, it will be a valuable of a growing fascination with the biology of resource for those starting out on their fasci- wine microorganisms among researchers world- nating journey through the world of wine wide. Since then, unprecedented improvements microbiology. in winemaking processes have gone hand in Coordinated by Alfonso V. Carrascosa, hand with the development of modern microbi- Rosario Mun˜oz, and Ramo´n Gonza´lez from the ology, and it would now be impossible to under- Spanish National Research Council (CSIC), this stand the continuing progress made in the wine book brings together contributions from a range industry without taking into account the impact of experts on the microbiology of wine working of advances in microbiological research. in universities, research centers, and industry. A greater understanding of the microbiology of wine holds the key to critical issues affecting the industry, such as the management of safety and quality. For instance, by identifying and Translation by Anne Murray and lain Patten gaining a better understanding of the molecular mechanisms underlying the growth of microor- The editors would like to acknowledge ganisms that cause wine spoilage or pose a the excellent translation of the Spanish text. The threat to consumer health, winemakers will be translators have been able to capture all the better positioned to control and even eradicate nuances of the original, using accurate wine- them during the production process. making English terms.

vii CHAPTER 1

Saccharomyces Yeasts I: Primary Fermentation Agustıń Aranda 1, Emilia Matallana 1,2, Marcel$lı´ del Olmo 2 1 Departamento de Biotecnologıá, Instituto de Agroquı´mica y Tecnologıá de Alimentos, CSIC, Valencia, Spain and 2 Departament de Bioquı´mica i Biologia Molecular, Facultat de Cie`ncies Biolo`giques, Universitat de Vale`ncia, Valencia, Spain

OUTLINE

1. Yeasts of Interest in Wine Production 2 2.2.3. pH 9 1.1. Yeast Flora on the Grape, in the Winery, 2.2.4. Clarification 9 and in the Must 2 2.2.5. Carbon Dioxide 9 1.2. Morphology and Cellular Organization of 2.3. Yeast Growth and Fermentation Kinetics 9 Yeasts 3 2.4. Biochemistry of Fermentation 10 1.3. Genetic Characteristics of Wine Yeasts 4 2.4.1. Alcoholic Fermentation 11 2.4.2. Nitrogen Metabolism 12 2. Growth Characteristics of Saccharomyces 2.5. The Importance of Yeast Metabolism in Yeasts During Fermentation 4 Wine Aroma 14 2.1. Must Composition 4 2.1.1. Sugars 5 3. Gene Expression During Fermentation 17 2.1.2. Organic Acids 5 3.1. Glycolytic Genes 18 2.1.3. Nitrogenous Compounds 5 3.2. Osmotic Stress-response Genes 19 2.1.4. Polyphenols 7 3.3. Genes Induced During the Stationary 2.1.5. Mineral Salts 7 Phase 20 2.1.6. Lipids 7 3.4. Gene Expression in Wine Yeasts Exposed 2.1.7. Inhibitors 7 to Specific Stress Conditions 22 2.2. Physical Parameters of Fermentation 8 4. Genetic Improvement of Yeast Efficiency 2.2.1. Temperature 8 During Fermentation 23 2.2.2. Aeration 8

1. YEASTS OF INTEREST IN WINE numbers. This microflora can be affected by PRODUCTION a wide variety of factors, principally temperature, rainfall, altitude, ripeness of the crop, and 1.1. Yeast Flora on the Grape, in the use of fungicides (Boulton et al., 1996). The flora Winery, and in the Must associated with winery equipment is largely made up of S. cerevisiae (Fleet & Heard, 1993; The fermentation of grape must is a complex Fleet, 2007; Martini & Vaughan-Martini, 1990), microbiological process that involves interac- though species of the genera Brettanomyces, tions between yeasts, bacteria, and filamentous Candida, Hansenula, Kloeckera, Pichia, and Toru- fungi (Fleet, 2007; Fugelsang & Edwards, laspora have also been isolated. 2007). Yeasts, which play a central role in the The yeasts present in the must during the first winemaking process, are unicellular fungi that few hours after filling the tanks belong to the reproduce by budding. Most yeasts belong to same genera as those found on the grapes, the phylum Ascomycota on the basis of their predominantly Hanseniaspora/Kloeckera. In these sexual development. In these organisms, the spontaneous vinification conditions, Saccharo- zygote develops within a sac-like structure, the myces yeasts (mainly S. cerevisiae) begin to ascus, while the nucleus undergoes two meiotic develop after around 20 h and are present divisions, often followed by one or more mitotic alongside the grape-derived yeast flora. After 3 divisions. A wall forms around each daughter or 4 d of fermentation, Saccharomyces yeasts nucleus and its surrounding cytoplasm to predominate and are ultimately responsible for generate four ascospores within the ascus. The alcoholic fermentation (Ribe´reau-Gayon et al., ascus then ruptures and releases the ascospores, 2006). This change in the yeast population is which can germinate and produce new vegeta- linked to the increasing presence of ethanol, tive cells. Although thousands of yeast species the anaerobic conditions, the use of sulfites have been identified, only 15 correspond to during harvesting and in the must, the concen- wine yeasts (Ribe´reau-Gayon et al., 2006). tration of sugar, and the greater tolerance of Traditionally, wine has been produced using high temperatures shown by S. cerevisiae yeast strains found on the surface of grapes compared with other yeasts (Fleet & Heard, and in the winery environment. The yeasts 1993; Fleet, 2007). S. cerevisiae comprises reach the grapes by wind and insect dispersal numerous strains with varying biotechnological and are present on the wines from the onset of properties (Ribe´reau-Gayon et al., 2006). The fruit ripening (Lafon-Lafourcade, 1983). The importance of using genetic techniques to iden- predominant species on the grape is Kloeckera tify and characterize the different species and apiculata, which can account for more than 50% strains of yeast that participate in fermentation of the flora recovered from the fruit (Fugelsang should not be underestimated. This is consid- & Edwards, 2007). Other species of obligate ered further in Chapter 5, which addresses the aerobic or weakly fermentative yeasts with taxonomy of wine yeasts. very limited alcohol tolerance may also be Currently, the usual strategy employed in found in lesser proportions. These belong to winemaking involves inoculation of the must the genera Candida, Cryptococcus, Debaryomyces, with selected yeasts in the form of active dried Hansenula, Issatchenkia, Kluyveromyces, Metschni- yeast. This practice, which emerged in the kowia, Pichia, and Rhodotorula (Fleet & Heard, 1970s, shortens the lag phase, ensures rapid 1993; Ribe´reau-Gayon et al., 2006). The fermen- and complete fermentation of the must, and tative species Saccharomyces cerevisiae and helps to create a much more reproducible final Saccharomyces bayanus are present in limited product (Bauer & Pretorius, 2000; Fleet & Heard, YEASTS OF INTEREST IN WINE PRODUCTION 3

1993). The selection of wine yeasts with specific with specific wine-growing regions (Lafon- genetic markers provides a system for the Lafourcade, 1983; Snow, 1983). precise monitoring of the growth of particular strains during fermentation. Analyses of this 1.2. Morphology and Cellular type have shown that fermentation is driven Organization of Yeasts mainly by inoculated yeasts (Delteil & Aizac, 1988), although these sometimes become only Saccharomyces yeast cells have a rigid cell wall partially established (Esteve-Zarzoso et al., that allows them to resist the changes in osmotic 1999). Given that the growth of the natural flora pressure that can occur in the extracellular envi- is not completely suppressed during the initial ronment. Inside the cell wall, there is a periplas- days of vinification, these strains can make mic space and a plasma membrane surrounding substantial contributions to certain properties the cytoplasm. Various transport mechanisms of the wine (Querol et al., 1992; Schu¨ tz & Gafner, control the permeability of these structures 1993). Consequently, there is increasing interest and maintain their role as barriers. in the use of mixed starter cultures in which Yeasts have multiple subcellular organelles non-Saccharomyces yeasts contribute desirable characteristic of eukaryotic cells. These include characteristicsdparticularly in terms of the a nucleus surrounded by a nuclear envelope, organoleptic quality of the winedthat comple- a smooth and a rough endoplasmic reticulum, ment the fermentative capacity of Saccharomyces a Golgi apparatus, mitochondria, and vacuoles. yeasts (Fleet, 2008). The cytoplasm contains numerous enzymes The inoculated yeast strain must obviously involved in the metabolic events described be very carefully selected on the basis of certain below, such as the enzymes responsible for alco- necessary characteristics (Degre´, 1993; Fleet, holic fermentation. Although some Saccharomyces 2008). For instance, it must produce vigorous strains lack mitochondria (respiration-deficient fermentation with short lag phases and little or “petite” mutants), these organelles play residual sugar, have reproducible fermentation a fundamental role in metabolism. During characteristics, be tolerant of high pressure, fermentation, the high concentration of glucose ethanol, and suboptimal temperatures, and in the medium inhibits synthesis of enzymes produce glycerol and b-glucosidases in involved in the citric acid cycle and cytochromes adequate quantities to achieve a good aroma. from the respiratory chain through an effect Other valuable properties include fermentative known as glucose repression (Gancedo, 2008; capacity at low temperatures, low foaming, Santangelo, 2006 and references therein). As killer activity (Barre, 1980), certain levels of a result, mitochondrial oxidative metabolism is specific enzymatic activities (Darriet et al., limited under these conditions. However, aerobic 1988; Dubourdieu et al., 1988), and resistance metabolism, which is dependent upon mitochon- to the adverse growth conditions present during dria, does occur during the production of winemaking (Zuzuarregui & del Olmo, 2004a). commercial yeasts for must inoculation and It is particularly important in the secondary during some phases of the winemaking process. fermentation of some sparkling wines for the Vacuoles are important for homeostasis, since yeast to be flocculent or easily separated from enzymes that participate in the degradation and the medium (Degre´, 1993; Zaworski & Heimsch, recovery of cell constituents are exclusively or 1987). Autochthonous strains that meet these predominantly localized to these structures. criteria have been increasingly used in recent They also accumulate metabolites such as basic yearsinanefforttoobtainwinesthatmain- amino acids, S-adenosylmethionine, polyphos- tain the sensory characteristics associated phates, allantoin, and allantoate at much higher 4 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION concentrations than those found in the 1976). In contrast, in heterothallic (ho) strains, cytoplasm. the MAT locus is stable and cells remain in More in-depth reviews of the cellular organi- a haploid state until they encounter a cell of the zation of yeasts can be found in The yeasts opposite mating type with which to fuse (1991), edited by Rose and Harrison, and in (reviewed in Sprague, 1995). Wine strains also The molecular biology of the yeast Saccharomyces exhibit a high degree of heterozygosity (Barre cerevisiae (1982), edited by Strathern, Jones, et al., 1993; Codoń et al., 1995), including for and Broach. the HO locus (Guijo et al., 1997; Mortimer et al., 1994), and they can undergo mitotic recombina- 1.3. Genetic Characteristics of Wine tion (Longo & Ve´zinhet, 1993; Puig et al., 2000), Yeasts a characteristic that is not observed in haploid laboratory strains. This capacity for extensive Unlike their counterpart laboratory strains, genomic change means that wine yeasts do not wine strains of S. cerevisiae are prototrophs, display genetic stability (Pretorius, 2000; Snow, meaning that they do not require amino acids 1983). These factors and their relationship with or nucleotides for their growth. This has impor- evolutionary processes are discussed in detail tant consequences for genetic manipulation, by Pe´rez-Ortıń et al. (2002) and are also consid- since genes conferring resistance to antibiotics, ered in Chapter 6. such as cycloheximide (del Pozo et al., 1991)or As in all eukaryotes, the mitochondria of S. geneticin (Hadfield et al., 1990), must be used cerevisiae have a circular mitochondrial DNA or auxotrophies introduced prior to transforma- (mtDNA) (Christiansen & Christiansen, 1976; tion of these yeasts. Hollenberg et al., 1970). This is usually located Wine strains of yeast are usually diploid, in the mitochondrial matrix but may occasion- polyploid, or even aneuploid (Bakalinsky & ally be bound to the inner mitochondrial Snow, 1990; Codoń et al., 1995). Chromosome membrane. The mtDNA contains genes encod- length in these yeasts is highly polymorphic ing proteins essential for mitochondrial function (Bidenne et al., 1992; Rachidi et al., 1999), and and, in yeasts, exhibits a high degree of poly- this results in extensive variability in sporula- morphism due to variability in the presence of tion capacity and spore viability. This character- certain introns and differences in the size of istic also influences the options for gene intergenic regions (Clark-Walker et al., 1981). manipulation, since at least two copies of This variability has been used in taxonomic a gene need to be eliminated to obtain a deletion studies, as discussed in Chapter 5. mutant. The ploidy of wine yeasts may provide them with advantages in adapting to change- able environments or, perhaps, represent 2. GROWTH CHARACTERISTICS a way of increasing the dose of genes that are OF SACCHAROMYCES important for fermentation (Bakalinsky & YEASTS DURING FERMENTATION Snow, 1990; Salmon, 1997). Finally, wine yeasts are predominantly homo- 2.1. Must Composition thallic (HO), meaning that following sporulation the daughter cells can change mating type, Grape must is a complex medium containing conjugate with a cell of the opposite mating all of the nutrients necessary for the growth of type, and ultimately form a cell with 2n DNA S. cerevisiae. However, the varying composition content that is homozygous for all genes except of different musts, in addition to being crucial the MAT locus (Thornton & Eschenbruch, for the characteristics of the final product, GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 5 influences the growth dynamics of the yeast. acid are found at lower concentrations. Tartaric Vinification is a discontinuous, batch-type acid predominates in must from warmer fermentation process in which all of the nutrients climates, where it reaches concentrations of 2 are present in the culture medium from the outset to 8 g/L, whereas, in cooler climates, malic and the concentration of the nutrients declines acid concentrations may exceed those of tartaric as they are consumed by the yeast. As a result, acid, depending on the ripeness of the grapes. the availability of some nutrients may act as These acids have no direct effect on yeast a limiting factor for growth. Below we describe growth but do play a decisive role in the pH the main components of the must and their effect of the must (see Section 2.2.3). on the process of alcoholic fermentation. 2.1.3. Nitrogenous Compounds 2.1.1. Sugars Nitrogen content is important since it tends With the exception of water, monosaccha- to be limiting for the growth of S. cerevisiae rides are the most abundant component of (Ingledew & Kunkee, 1985) and the principal grape must. Glucose and fructose are the main cause of stuck fermentation (Bisson, 1999). The hexose sugars and are present in approximately concentration of soluble nitrogen varies equimolar concentrations. Other monosaccha- between 0.1 and 1 g/L (Henschke & Jiranek, rides present as minor components include 1993). The composition of nitrogen sources in arabinose (0.2e1.5 g/L) and xylose (0.03e the must depends on a large number of factors, 0.1 g/L); low concentrations of the disaccharide such as the grape variety, infection with Botrytis sucrose, which is generally hydrolyzed at the cinerea (which eliminates large quantities of the low pH found in must, are also present (Ough, nutrients that can be assimilated by Saccharo- 1992). Although polysaccharides such as myces yeasts), the timing of harvest, use of fertil- pectins, gums, and dextrin are present at izers, addition of supplements in the winery, concentrations of around 3 to 5 g/L, they are and the extent of clarification of the musts, not assimilable by wine yeasts. The total concen- particularly in white grape musts (Lagunas, tration of sugars is generally between 170 and 1986). Variations in the quantity and form of 220 g/L (Ribe´reau-Gayon et al., 2006). In musts the nitrogen sources in the must influence yeast with sugar concentrations of more than 200 g/ cell growth, fermentation rate, and ethanol L, there is a slowing of fermentation. Sugar tolerance. The main compounds are ammonia concentrations between 250 and 300 g/L can (3e10%), amino acids (25e30%), polypeptides inhibit yeast growth as a result of the high (25e40%), and proteins (5e10%). In addition, osmotic pressure and the elevated intracellular smaller quantities of nitrates, nucleotides, concentration of ethanol (Nishino et al., 1985). amines, and vitamins may be present. Nucleo- However, the low sugar concentrations typical tides are only present at very low concentrations of northerly wine-growing areas do not limit in the must (e.g., adenine and uracil nucleotides yeast growth and only affect the final alcohol are found at concentrations of 4e15 mg/L and concentration. 4e8 mg/L, respectively). These are taken up by the yeast and incorporated into their nucleic 2.1.2. Organic Acids acids, although yeast can also synthesize their The second most abundant compounds, own nucleotides (Monteiro & Bisson, 1992). organic acids, are present at concentrations of Saccharomyces yeasts cannot assimilate inor- between 9 and 27 g/L (Ough, 1992). Tartaric ganic nitrogen sources such as nitrates and and malic acid together account for 90% of the nitrites. They are also unable to assimilate fixed acidity (Jackson, 1994); citric and ascorbic proteins and polypeptides present in the 6 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION medium, since they do not have a system for exponential growth phase (Beltrań et al., 2005). extracellular digestion of these types of Because ethanol impedes the uptake of nitroge- compound. As a result, they are essentially nous compounds during later stages, it has been dependent on the concentrations of ammonia proposed that nitrogen should be added prior to and amino acids, their preferred nitrogen sour- or during the initial phases of fermentation, to ces (Ough & Amerine, 1988). The most abun- coincide with aeration of the must (Sablayrolles dant amino acids in the must tend to be et al., 1996). More recent data have shown proline and arginine, and their concentrations poorer recovery of fermentation activity with vary in different musts. Proline cannot be addition of ammonia alone than with addition metabolized by yeast under the low-oxygen of amino acids or a combination of the two conditions associated with alcoholic fermenta- (Jimeńez-Martı´ et al., 2007). tion and should therefore not be taken into Among the nitrogenous compounds, vita- account when considering nitrogen availability. mins deserve special mention. Wine yeasts are It has been reported that concentrations of able to synthesize all of their own vitamins assimilable nitrogen below 140 mg/L impair except for biotin (Ough, 1992) and nicotinic fermentation at normal sugar concentrations acid under anaerobic conditions (Panozzo (Bely et al., 1990), and a concentration of ammo- et al., 2002), meaning that they are not as depen- nium ions below 25 mg/L is generally consid- dent as more complex organisms on the avail- ered to be undesirable. However, outcomes ability of these cofactors. Nevertheless, the can vary according to the individual strain. presence of vitamins in the must stimulates Cases have been described in which normal the growth and metabolic activity of yeasts via fermentation occurred in the presence of the vitamins’ participation as coenzymes in 120 mg/L assimilable nitrogen (Carrasco et al., numerous biochemical reactions, and as a result 2003), while strains that require a minimum of they can be considered as growth factors 267 mg/L to complete the process have also (Ribe´reau-Gayon et al., 2006). Must is generally been reported (Mendes-Ferreira et al., 2004). rich in vitamins, but the concentrations of some Since Saccharomyces yeasts can synthesize their are suboptimal. As a result, addition of vitamins own amino acids, the simplest solution to the can stimulate growth, particularly when the problem of nitrogen deficiencies is to provide grapes have been subject to fungal infection, ammonium salt supplements, usually in the which always reduces the total concentration form of diammonium sulfate or phosphate of vitamins. Thiamine is also an important (Ribe´reau-Gayon et al., 2006). Addition of up component of the must. However, it is partially to 30 g/hl of diammonium phosphate (DAP) is degraded by the sulfite added to prevent the permitted in the European Union (EU), whereas appearance of spoilage organisms (Jackson, in the United States up to 96 g/hl is allowed 1994) and is also consumed by the yeasts over (Fugelsang & Edwards, 2007). Excessive the course of the fermentation. Consequently, nitrogen supplementation can alter the microbi- it is advisable to add it to the must. The amount ological stability of the wine (providing nutri- recommended by the EU is 50 mg/hl, whereas ents for spoilage organisms) and its aroma (in the maximum permitted level in the United many cases derived from deamination of amino States is 60 mg/hl (Fugelsang & Edwards, acids). The timing of nitrogen addition is also 2007). Deficiencies in other vitamins, such as important. Although reductions in fermentation pantothenic acid and pyroxidine, should also time have been reported to occur independently be avoided as they can lead to generation of of the timing of addition, better results are undesirable compounds such as acetic acid obtained when nitrogen is added during the and hydrogen sulfide (Wang et al., 2003). GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 7

2.1.4. Polyphenols of fermentation (Sablayrolles, 1996). The lack The many and varied phenolic compounds of these types of lipid (especially ergosterol, present in the must are essential elements in the principal sterol in the plasma membrane of determining the organoleptic character of the Saccharomyces yeasts) affects the structure and wine (Waterhouse, 2002). Although it has been function of the plasma membrane and leads to reported that the anthocyanins in red grape increased effects of ethanol and poor glucose musts and the procyanidins in white grape uptake (Jackson, 1994). These compounds are musts can stimulate and inhibit growth, respec- referred to as survival factors, since their pres- tively (Cantarelli, 1989), these compounds have ence is necessary for cell viability but their addi- no relevant influence on the growth of wine tion does not increase growth (Ribe´reau-Gayon yeasts. Their most noteworthy effect is as anti- et al., 2006). Generally, the presence of these oxidants, particularly in the case of quinones. types of lipid in the must is guaranteed given It has recently been described that resveratrol their abundance in grape skins. Problems are and other polyphenols with recognized preven- only encountered in excessively clarified white tive effects in cardiovascular disease (Fremont, wines, since up to 90% of unsaturated fatty acids 2000) can extend the replicative lifespan of may be lost under these conditions (Bertrand & Miele, 1984). In such situations, it is appropriate Saccharomyces yeasts (Howitz et al., 2003). to supplement the must with yeast extract or lysed yeast (Munõz & Ingledew, 1990). In other 2.1.5. Mineral Salts situations, the use of dried yeast grown under Inorganic elements are necessary for normal aerobic conditions usually guarantees the pres- metabolism and maintenance of pH and ion ence of sufficient lipids in the cell wall for balance in yeasts. Potassium, sodium, calcium, fermentation of the must to take place. and magnesium are the predominant cations in the must, and chlorates, phosphates, and 2.1.7. Inhibitors sulfates the main anions (Ough & Amerine, This section covers exogenous compounds 1988). Must generally provides the inorganic added to the grapes and must to prevent the elements required for yeast growth, but, if the appearance of undesirable microorganisms concentration of one of these elements is that can also influence the growth of wine limited, normal progression of fermentation yeasts. can sometimes be affected (Bisson, 1999). Phos- phate ions are particularly important given their 2.1.7.1. SULFITES vital metabolic role, as hexose sugars must be Sulfites are added to control the appearance phosphorylated in order to be metabolized. of spoilage organisms in the must. Industrial Deficiencies in this anion can be compensated yeasts have been selected to be resistant to the along with those of nitrogen by supplementa- quantities of sulfites used in wineries, and their tion with DAP (see Section 2.1.3). growth is not usually affected by the concentrations of between 0.8 and 1.5 mg/L that are nor- 2.1.6. Lipids mally used. Concentrations above 1.5 mg/L, Under the anaerobic conditions associated however, can inhibit growth (Sadraud & Chau- with wine fermentation, yeasts cannot synthe- vet, 1985). This inhibition is dependent upon size sterols or long-chain unsaturated fatty the pH of the must; SO2, the active molecular acids. Synthesis of these compounds will only species, is generated at lower pH, and as a result occur if oxygen is added during fermentation the toxicity of a given concentration of the to increase yeast cell viability and the quality compound increases under those conditions 8 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

(Farkas, 1988). Sulfite toxicity is also increased decline (Watson, 1987). Although there is a linear by the richness of methionine in the must, increase in the rate of fermentation between 10 whereas it is reduced by higher concentrations and 32C (doubling every 10C), this does not of adenine (Aranda et al., 2006). Nevertheless, mean that higher temperatures are the most sulfite normally only delays the onset of fermen- appropriate for fermentation of the must. tation and does not affect the rate or completion Ethanol toxicity increases with temperature, of the process. and higher temperatures lead to evaporation of ethanol and other volatile compounds that 2.1.7.2. PESTICIDES are essential to the organoleptic properties of Chemical compounds applied to the vines to the wine (Torija et al., 2003), particularly in the prevent parasite infection can sometimes affect case of white wines. Excessively low tempera- the growth of Saccharomyces yeasts during vini- tures are also not recommended, since they fication. Folpet and captan, traditionally the can cause stuck fermentation when yeast most commonly used fungicides, have membrane fluidity begins to be affected (Bisson, a substantial antiseptic effect on yeasts (Cabras 1999). It is also not economically viable to main- & Angioni, 2000). New-generation fungicides tain fermentations under these conditions for are only marketed if they have been shown to extended periods. Consequently, controlling have no effect on yeasts. For instance, metalaxyl, fermentation temperature is an essential cymoxanil, famoxadone, fenhexamid, fluquin- element of modern wine production. White conazole, kresoxim-methyl, quinoxyfen, and tri- wines are generally fermented at between 10 floxystrobin have no effect on yeast growth and 18C to improve the retention of aromas, (Jackson, 1994; Oliva et al., 2007). In addition, whereas red wines tend to be fermented at clarification of the must eliminates most of the higher temperatures (between 18 and 29C) to pesticides present on the surface of the grapes, achieve good extraction of phenolic compounds and many are degraded spontaneously under (Fugelsang & Edwards, 2007). Nevertheless, an the acidic conditions of the must. As with initial fermentation temperature of 20C is rec- sulfites, traces of fungicide in the must tend to ommended in both cases in order to stimulate inhibit the onset of fermentation rather than initiation of yeast growth (Jackson, 1994). Low interfere with fermentation rate or completion. temperatures may favor the growth of non- Saccharomyces yeasts during the initial stages of 2.2. Physical Parameters of fermentation. Fermentation 2.2.2. Aeration The main physicochemical factors that affect Saccharomyces yeasts are facultative anaer- the growth of Saccharomyces yeasts during alco- obes, able to consume sugars in the absence of holic fermentation are described below. oxygen more effectively than non-Saccharomyces yeasts (Visser et al., 1990). In fact, excess oxygen 2.2.1. Temperature can inhibit fermentation, a phenomenon known Temperature is the most important physical as the Pasteur effect. Nevertheless, a certain factor in the growth of yeasts and the progres- amount of oxygen is beneficial for the growth sion of fermentation (Fleet & Heard, 1993). of wine yeasts since it is required for the Although S. cerevisiae has an optimal growth synthesis of sterols (mainly ergosterol) and temperature of around 30C, it can adapt to unsaturated fatty acids. A more oxygenated a wide range of temperatures up to a maximum environment may be helpful in musts with of 40C, at which point viability begins to nitrogen deficiencies, as this will allow the GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 9 amino acid proline to be metabolized (Ingledew solid particles act as nuclei for the formation of & Kunkee, 1985). It is also advisable to add carbon dioxide bubbles and favor dissipation exogenous nitrogen sources during aeration of of the gas, which at high levels can inhibit the the must (Sablayrolles et al., 1996). The oxygen growth of Saccharomyces yeasts (Thomas et al., captured by the must during pressing is usually 1994). On the other hand, the final products sufficient to reach saturation, and is therefore obtained from clearer musts have better organo- generally adequate for normal progression of leptic characteristics. The extent of clarification fermentation. In red wines, oxygen consump- must therefore be optimized to produce better tion due to oxidation of phenols is compensated wines without affecting the fermentation by the aeration created during pump-over, process. resulting in oxygen concentrations of around 10 mg/L. This effect is most beneficial at the 2.2.5. Carbon Dioxide end of the exponential growth phase. Neverthe- Alcoholic fermentation of hexose sugars less, excessive aeration may lead to undesirable generates carbon dioxide, which can reach production of acetaldehyde and hydrogen volumes equivalent to 56 times that of the fer- sulfide, and reduced production of aromatic mented must (Boulton et al., 1996). The release esters (Nyka¨nen, 1986). of this gas contributes to the dissipation of some heat and produces convection currents 2.2.3. pH within the must that aid the diffusion of nutri- The typical pH of grape must is between 2.75 ents. However, its evaporation also favors loss and 4.2 (Heard & Fleet, 1988). These pH values of ethanol and volatile compounds (Jackson, do not have a negative effect on the growth of 1994). Furthermore, if produced in excess, Saccharomyces yeasts, and problems only begin carbon dioxide affects the viability of Saccharo- to present themselves at a pH below 2.8. The myces yeasts, mainly due to membrane damage. toxic effects of low pH are due to the increased effects of ethanol (Pampuhla & Loereiro-Dias, 2.3. Yeast Growth and Fermentation 1989) and sulfite (Farkas, 1988). Tolerance of Kinetics acidic pH depends on the abundance of potassium ions in the must (Kudo et al., 1998). Low Yeast growth during wine fermentation pH favors the hydrolysis of disaccharides and, differs from that occurring in other industrial therefore, fermentation. In addition, the acidic processes such as brewing, since the high concen- character of the must prevents the appearance tration of sugars leads to the production of of spoilage microorganisms. Consequently, ethanol at concentrations that inhibit growth. acids such as tartaric acid are sometimes added Fermentation begins rapidly with inoculums (addition of 1 g/L, for example, reduces the pH containing approximately 106 cells/mL. The by 0.1 units). However, addition of excess tarta- typical growth cycle of Saccharomyces yeasts ric acid can lead to undesirable precipitation. consists of three phases and begins following a short lag period (Lafon-Lafourcade, 1983). 2.2.4. Clarification The first phase is the limited growth phase and Elimination of solid particles from the must is lasts between 2 and 5 d, generating a population an important element in the production of white of up to 107 or 108 cells/mL. Fermentation during wines. However, elimination of the nutrients this phase occurs at a constant, maximal rate, and that are associated with them, particularly it tends to consume between a third and half of nitrogenous compounds, can impair yeast the initial sugar content (Castor & Archer, growth (Ayestaran et al., 1995). Furthermore, 1956). Next, growth enters a quasi-stationary 10 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION phase that lasts around 8 d. During this time, estimation of the number of viable cells, but there is no increase in the number of cells in the they are less reliable. population. However, the cells are metabolically Another parameter that is analyzed in wine active and the rate of fermentation remains yeasts is vitality; that is, the capacity of the cells maximal. Finally, the culture enters the death to achieve complete metabolic activity. There is phase, which is poorly characterized and highly a relationship between this metabolic activity variable. Whereas some authors claim that death and the time necessary to reach maximum does not occur until all of the sugars have been fermentation rate. This is usually measured by consumed (Boulton et al., 1996), others have indirect impedance; in other words, the reduc- assigned greater importance to this phase. tion in impedance due to a solution of potas- According to this view, the death phase is esti- sium hydroxide that reacts with the carbon mated to be three or four times longer than the dioxide produced by the metabolic activity of growth phase and still involves consumption of the yeast (Novo et al., 2007). a considerable quantity of sugar (Ribe´reau- Because all of the methods for monitoring Gayon et al., 2006). The loss of viability is accom- yeast growth are relatively difficult to imple- panied by a reduction in the rate of fermentation, ment in wineries, in practice, fermentation due not only to a reduction in the number of kinetics are analyzed using simpler techniques viable cells but also to inhibition of the metabolic such as monitoring the reduction in sugar activity of the nonproliferative cells. The loss of concentration, the increase in ethanol content, fermentative capacity of the cells in this final or the release of carbon dioxide (Ribe´reau- phase has been linked to the depletion of adeno- Gayon et al., 2006). However, the simplest sine triphosphate (ATP) and the accumulation of method to adapt to winery conditions is anal- ethanol, which have negative effects on ysis of the density of the must, since measure- membrane transport. It has been observed that, ment of the mass per unit volume provides an under these conditions, cellular enzyme systems approximate measure of sugar content. During are functional but the intracellular concentration the course of fermentation, the sugar concentra- of sugars decreases progressively. tion decreases while ethanol content increases, Yeast growth is monitored by microscopic and this leads to a reduction in density. The counts of the cells in diluted samples of ferment- initial density of the must and the final density ing must. The number of cells can also be esti- of the wine will depend on the initial sugar mated by measuring the optical density at 600 concentration, which will lead to a specific to 620 nm following the generation of standard percentage of ethanol (approximately 1% [vol/ curves for the inoculated strain. In both cases, vol] ethanol for every 17 g of sugars) (Ribe´r- estimations of the numbers of cells present in eau-Gayon et al., 2006). the fermenting must do not differentiate between viable and dead cells, a very important 2.4. Biochemistry of Fermentation distinction when monitoring the progression of wine fermentation. To differentiate between the The biochemistry of wine production is also two, plate counts can be performed with solid complex. The central metabolic process that nutrient media, on which only viable cells will takes place is alcoholic fermentation, a catabolic be able to produce colonies; however, this type pathway involving the transformation of the of analysis is slow, as the colonies take 3 to 4 d hexose sugars present in the must into ethanol to grow. Other more rapid techniques based and carbon dioxide. Compounds are also gener- on the use of fluorescent reagents or biolumines- ated that play a central role in yeast growth and cent quantification of ATP are available for in the organoleptic properties of the wine. GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 11

A more complete description of the biochem- cytoplasm and can be expressed in terms of ical processes that take place during wine the following simplified equation: production can be found elsewhere (Boulton et al., 1996; Rose & Harrison, 1991; Strathern C6H12O6 þ 2ADP þ 2HPO4 /2C2H5OH et al., 1982). The aim here is to introduce readers þ 2CO þ 2ATP þ 2H O to some of these pathways, in particular those 2 2 that are most relevant in terms of yeast growth Alcoholic fermentation involves the Embden- and the properties of the final product. Meyerhof-Parnas (EMP) pathway, which was described by Embden, Meyerhof, and Parnas 2.4.1. Alcoholic Fermentation around 1940 and is also known as glycolysis. Carbon sources, in particular the hexose The pathway involves 10 reactions. The first sugars glucose and fructose, allow cells to five reactions correspond to the energy invest- obtain energy by alcoholic fermentation. This ment phase, in which sugars are metabolically metabolic pathway (Figure 1.1) occurs in the activated by ATP-dependent phosphorylation to give rise to a six-carbon sugar, fructose-1,6- bisphosphate, which is cleaved to produce two moles of triose phosphate. During the energy ATP generation phase (reactions 6 to 10), the triose Glucose Glucose-6-phosphate phosphates are reactivated, generating two compounds with a high phosphate-transfer ATP potential: firstly 1,3-bisphosphoglycerate and Fructose-1,6-bisphosphate then phosphoenolpyruvate. Each of these compounds transfers a high-energy phosphate NADH DHAP Glyceraldehyde-3-phosphate group to adenosine diphosphate (ADP), thus NADH producing ATP in a process known as Glycerol-3-phosphate substrate-level phosphorylation. The chemical 1,3-bisphosphoglycerate energy of ATP can be subsequently transformed Glycerol ATP in the cell into other forms of energy necessary 3-phosphoglycerate for cell growth. The first reaction in this energy generation phase is an oxidation reaction catalyzed by the enzyme glyceraldehyde-3-phosphate dehydrogenase. This enzyme requires Ethanol Phosphoenolpyruvate nicotinamide-adenine dinucleotide (NADþ)as NADH ATP a coenzyme to accept the electrons from the Acetaldehyde Pyruvate substrate being oxidized. As a consequence, NADH this coenzyme is reduced to NADH. Acetate After glycolysis, alcoholic fermentation is completed with two additional reactions used FIGURE 1.1 Schematic diagram of the conversion of þ glucose into ethanol during alcoholic fermentation by the to reoxidize NADH to NAD to guarantee the yeast Saccharomyces cerevisiae. The figure also shows the continuation of glycolysis. In the first reaction, relationship between energy production in this pathway the resulting pyruvate is decarboxylated to acet- and the processes linked to the redox state of the coenzyme aldehyde and carbon dioxide by the enzyme þ NAD /NADH. The reactions in which consumption or pyruvate decarboxylase, which requires thia- synthesis of ATP and NADH occur are indicated. DHAP ¼ mine pyrophosphate as a coenzyme. Finally, dihydroxyacetone phosphate. Figure adapted from Norbeck and Blomberg (1997). the acetaldehyde is reduced to ethanol by the 12 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION enzyme alcohol dehydrogenase in a reaction are exposed at the beginning of vinification involving oxidation of NADH to NADþ. (Blomberg & Adler, 1992). The glycolysis pathway is not only involved The synthesis of glycerol also represents in energy production for the yeast. It also gener- a mechanism for the oxidation of molecules of ates metabolites that can be used as substrates NADH generated during glycolysis that have for the biosynthesis of molecules linked to not been reoxidized as a result of 1,3-bisphos- increased biomass (Figure 1.2). Glucose-6-phosphoglycerate or pyruvate being diverted phate can be directed towards the pentose phos- towards products other than ethanol. It is there- phate pathway, which allows the formation of fore essential to maintain the redox balance in NADPH and ribose phosphate, molecules that the cytoplasm. Higher alcohols are also are necessary for the biosynthesis of fatty acids produced during alcoholic fermentation. As dis- and nucleotides, respectively. Pyruvate is also cussed below, these compounds can also be an important substrate for the synthesis of mole- generated from certain amino acids and are cules such as oxaloacetate, succinate, organic important in determining the aroma of wine. acids, and amino acids. These molecules are produced at the beginning of vinification, 2.4.2. Nitrogen Metabolism when the activities of pyruvate decarboxylase S. cerevisiae is equally able to use amino acids, and alcohol dehydrogenase are low. 3-Phospho- ammonia, uracil, proline derivatives, and urea glycerate can also be diverted from glycolysis to as nitrogen sources (for a detailed review of participate in the synthesis of amino acids such the use of these compounds and their effect on as serine. Finally, dihydroxyacetone phosphate, yeast growth rate, see Cooper, 1982a). Among one of the end products of the energy invest- the nitrogenated components that can be found ment phase, is used to produce glycerol. This in the must, amino acids make the largest contri- molecule has a powerful effect on the quality bution to nitrogen provision for the synthesis of of the wine, participates in the biosynthesis of structural and functional proteins and the triacylglycerols, and is also the main compatible production of enzymes and transporters. osmolyte that is produced by yeasts in response Figure 1.3 shows the uptake and use of to the significant osmotic stress to which they nitrogen by yeasts when it is available in the

FIGURE 1.2 Biosynthetic precursors Glucose Glucose-6-phosphate Ribose-5-phosphate derived from alcoholic fermentation. DHAP ¼ Nucleic acids dihydroxyacetone phosphate. Figure adapted Fructose-1,6-bisphosphate from Henschke and Jiranek (1993). DHAP Glyceraldehyde-3-phosphate

Triacylglycerols Glycerol-3-phosphate 1,3-bisphosphoglycerate

Glycerol 3-phosphoglycerate Serine Proteins Proteins Aromatic amino acids Phosphoenolpyruvate Asparagine Aspartate Oxaloacetate Pyruvate Alanine Nucleic acids Pyrimidines

Acetaldehyde Acetyl-CoA Ketoacids

Fatty acyl-CoA Higher alcohols Ethanol Fatty acids Esters GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 13

+ FIGURE 1.3 Schematic diagram of the NH + 4 H H+ AMINO ACID uptake and metabolism of nitrogenous compounds. Reproduced from Llaurado´ i Rever- MUST chon (2002). CELL ADP + P ATP H+ AMINO ACID H+

+ NH -KETOGLUTARATE ATP H+ 4 ( -KETOACID)

VACUOLE NADP-GDH ADP + P H+ TRANSAMINATION (TRANSAMINASES) +

H AMINO ACID NAD-GDH

GLUTAMATE (AMINO ACID) HIGHER ALCOHOLS

H+ AMINO ACID POLYAMINES VITAMINS NUCLEIC ACIDS PROTEINS medium. Most nitrogenous compounds are releasing compounds that are toxic to the cell incorporated in the cell via active transport (Cooper, 1982a). systems, specifically symport with ions, usually During yeast growth, more than half of the protons (Cooper, 1982b). In S. cerevisiae, a general intracellular reserves of amino acids are in the amino acid permease (Gap1p) has been identi- vacuoles (Wiemken & Durr, 1974). This fied, and also specific permeases for different compartmentalization contributes to the regula- amino acids (Grenson et al., 1966; Horak, tion of the activity of various enzymes involved 1986). Ammonia undergoes active transport of in their degradation (Sumrada & Cooper, 1982). the protonated species that requires the pres- Ammonia and glutamate are central to all ence of glucose (Roon et al., 1977), and three nitrogen metabolism in yeasts. Ammoniacal systems have been identified involving the nitrogen is rapidly incorporated in the biosyn- proteins Mep1p, Mep2p, and Mep3p (Marini thetic pathways through the activity of et al., 1994; Marini et al., 1997). A detailed NADPþ-dependent glutamate dehydrogenase. description of the uptake mechanisms of these In addition, it represents the end product of the different nitrogenous compounds is provided catabolic pathways for nitrogenous compounds, by Cartwright et al. (1989), Cooper (1982b), in this case through the reaction catalyzed by and Henschke and Jiranek (1993). NADþ-dependent glutamate dehydrogenase. Nitrogenous compounds are assimilated In turn, amino acids undergo interconversion during the first few hours of fermentation (Mon- processes via the transaminase system, in which teiro & Bisson, 1991) and degraded in a specific glutamate plays an extremely important role as order that depends on factors such as the a donor and acceptor of amino groups. requirement for each compound in biosynthetic Metabolism of nitrogenous compounds by processes, efficiency of transport, and possible yeasts also contributes to the formation of conversion into ammonia or glutamate without products that play an important role in the final 14 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION quality of the wine by affecting its sensory origin of the main volatile compounds gener- properties. ated during yeast metabolism. Table 1.1 shows the compounds and their origins, their concen- 2.5. The Importance of Yeast trations in wine, and the characteristics of the Metabolism in Wine Aroma aroma produced. All of these elements are described in greater detail in the review by Wine aroma is generated by a series of Lambrechts and Pretorius (2000). aromatic and volatile compounds recognized The first group of compounds is the volatile by the senses of taste and smell. Some of these fatty acids. These include acetic acid, long-chain arise from the grapes and are responsible for fatty acids (C16 and C18), and short-chain fatty what is known as varietal aroma. However, acids (C8, C10, and C12). The levels of acetic most arise from the fermentation process and acid must be strictly controlled and should not their concentrations are essentially dependent exceed 1.0 to 1.5 g/L (Eglinton & Henschke, on the yeasts that predominate during fermen- 1999). In yeast, fatty acids are generated from tation and the conditions under which fermen- acetyl-coenzyme A (CoA) derived from oxida- tation takes place (Egli et al., 1998; Henick-Kling tive decarboxylation of pyruvate. Their et al., 1998; Steger & Lambrechts, 2000). Vinifi- synthesis requires two enzyme systems: acetyl- cation temperature plays a particularly impor- CoA carboxylase and fatty acid synthase. For tant role, since more aromatic wines are an extensive review of these pathways, see Pal- produced when the process is carried out at tauf et al. (1992) and Ratledge and Evans (1989). temperatures close to or below 15C(Bauer & Higher alcohols make the greatest contribu- Pretorius, 2000). tion to wine aroma. At concentrations below The main groups of aromatic compounds 300 mg/L they introduce a desirable complexity, derived from yeast metabolism are organic whereas at concentrations above 400 mg/L they acids, higher alcohols, esters, and, to a lesser have a negative effect on wine quality (Nyka¨nen, extent, aldehydes (Rapp & Versini, 1991). 1986). The higher alcohols produced in the Substances derived from fatty acids and from largest quantities are 1-propanol, 2-methyl-1- nitrogen- or sulfur-containing compounds also propanol, 2-methyl-1-butanol, 3-methyl-1- contribute (Boulton et al., 1996). Various studies butanol, hexanol, and 2-phenylethanol undertaken in recent years show that the (Henschke & Jiranek, 1993). One of the pathways composition of the must and, in particular, the through which these compounds are generated levels and nature of nitrogenous compounds is the conversion of branched-chain amino acids present in the must or subsequently added to (valine, leucine, isoleucine, and threonine); as limiting fermentations play an important role a result, their accumulation depends on the in determining the organoleptic properties of quantity and type of nitrogen sources in the the wine (see, for instance, Beltrań et al., 2005; must (Giudici et al., 1993). However, most higher Carrau et al., 2008; Jimeńez-Martı´ et al., 2007; alcohols are synthesized de novo from sugars Mendes-Ferreira et al., 2009; Torija et al., 2003; via the initial formation of the corresponding Vilanova et al., 2007). Some of the compounds ketoacids. arising from metabolism have a negative contri- Esters are largely responsible for the fruity bution, as is the case for acetaldehyde, acetic and floral character of a number of wines. acid, ethyl acetate, some higher alcohols Acetate esters of higher alcohols (such as ethyl when present at high concentrations, and, in acetate, 2-phenylethanol acetate, or isoamyl particular, reduced sulfur compounds, organic acetate) and ethyl esters of medium-chain satu- sulfates, and thiols. Below we describe the rated fatty acids (such as ethyl hexanoate) make GROWTH CHARACTERISTICS OF SACCHAROMYCES YEASTS DURING FERMENTATION 15

TABLE 1.1 Principal Compounds Responsible for Wine Aroma

Concentration Compound in wine (mg/L) Aroma Acetic acid 150e900 Vinegar Propionic acid Trace Rancid Butyric acid Trace Bitter

Volatile fatty acids Hexanoic acid Trace to 37 Rancid, vinegar, cheese Octanoic acid Trace to 41 Oily, rancid, sweet, buttery Decanoic acid Trace to 54 Unpleasant, rancid, bitter, phenolic Propanol 9e68 Powerful

Butanol 0.5e8.5 Petrol 2-methyl-1-butanol 15e150 Marzipan Higher alcohols Isobutylic acid 9e28 Alcoholic

Isoamyl alcohol 45e490 Marzipan Hexanol 0.3e12 Freshly mown grass 2-Phenylethanol 10e180 Floral, rose Isoamyl acetate 0.03e8.1 Banana, pear

2-Phenylethyl acetate 0.01e4.5 Rose, honey, fruity, ﬂoral Ethyl acetate 26e180 Varnish, nail polish, fruity Esters Isobutyl acetate 0.01e0.8 Banana

Ethyl butanoate 0.01e1.8 Floral, fruity Ethyl hexanoate Trace to 3.4 Apple, banana, violet Ethyl octanoate 0.05e3.8 Pineapple, pear Ethyl decanoate Trace to 2.1 Floral Acetaldehyde 10e300 Bitter, green pineapple

Carbonyl compounds Benzaldehyde 0.003e4.1 Bitter almond Diacetyl 0.05e5 Larder

4-Vinylphenol 0e1.15 Medicinal

4-Vinyl guaiacol 0e0.496 Smoky, vanilla

Volatile phenols 4-Ethylphenol 0e6.047 Horse sweat 4-Ethyl guaiacol 0e1.561 Smoky, vanilla

(Continued) 16 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

TABLE 1.1 Principal Compounds Responsible for Wine Aromadcont’d

Concentration Compound in wine (mg/L) Aroma Hydrogen sulfide Trace to >0.080 Rotten eggs Dimethyl disulfide Trace to 0.0016 Boiled cabbage Sulfur compounds Diethyl disulfide Trace Garlic, burnt rubber Methyl mercaptan Qualitative Rotten eggs, cabbage Ethyl mercaptan Qualitative Onion, rubber

Data taken from Cabanis et al. (1998) and Lambrechts and Pretorius (2000). the greatest contribution. According to the clas- are mainly formed during the second half of sification proposed by Baumes et al. (1986), all of fermentation, when the concentration of ethanol these are included in the apolar group. The in the medium is high (Herraiz & Ough, 1993). other group, the polar compounds, includes Volatile short-chain aldehydes are also compounds found in greater quantities but important, and contribute in particular to pine- that have more influence on the body than on apple and lemony aromas. Acetaldehyde and the aroma of the wine. These include diethyl diacetyl account for around 90% of these succinate, 2-ethyl-hydroxypropionate, diethyl compounds and have acceptable limits of malate, and ethyl-4-hydroxypropanoate. The 100 mg/L and 1 to 4 mg/L, respectively. These principal ester is ethyl acetate, although concen- compounds are generated from two ketoacids trations above 170 mg/L in white wines and derived from the synthesis or degradation of 160 mg/L in red wines are unacceptable (Corison amino acids or higher alcohols. et al., 1979). All these compounds are funda- Volatile phenols are important for flavor, mentally derived from sugar metabolism. color, and aroma of wines (Dubois, 1983). The Acetate esters are synthesized in reactions most important are vinylphenols in white wines between alcohol and acetyl-CoA catalyzed by and ethylphenols in red wines (Chatonnet et al., alcohol acetyltransferases (Peddie, 1990). Fatty 1997; Etievant, 1981; Singleton & Essau, 1969). acid esters, on the other hand, are generated These compounds are produced from the following activation of the corresponding fatty nonvolatile acids trans-ferulic and trans-p-cuo- acid by CoA, catalyzed by an acyl-CoA synthase maric acid, essentially through the activity in (Nordstro¨m, 1964a, 1964b, 1964c). Final concen- red wines of contaminating yeasts belonging trations of these compounds in the wine are to the genera Brettanomyces/Dekkera (Chatonnet affected by two factors: their hydrolysis during et al., 1997). early phases of wine maturation (Ramey & Finally, sulfur compounds make a significant Ough, 1980) and the extent to which they are contribution to the aroma of wine due to their transferred to the medium, which is reduced high reactivity, although in some cases they with increasing chain length and is influenced are responsible for undesirable aromas. The by the temperature at which fermentation takes main compound in this group is hydrogen place (Nyka¨nen et al., 1977). Ethyl esters of sulfide, which has an acceptable limit of amino acids have also been found in wines at between 10 and 100 m/L. This compound is concentrations of up to 58 mg/L (Herraiz & essentially derived from sulfate in the medium Ough, 1993; Heresztyn, 1984); these compounds and elemental sulfur introduced by fungicides, GENE EXPRESSION DURING FERMENTATION 17 and its formation by yeasts is linked to nitrogen TABLE 1.2 Saccharomyces cerevisiae Genes Mentioned and sulfur metabolism (Henschke & Jiranek, in This Chapter 1993; Rauhut, 1993). In fact, it has been observed that deficiencies in easily assimilable sources of Gene Molecular function or biological process nitrogen are a major cause of hydrogen sulfide ADE4 Phosphoribosyl pyrophosphate formation by yeasts (Stratford & Rose, 1986) aminotransferase and these levels can vary according to the initial ADH7 Alcohol dehydrogenase concentration of nitrogen in the must and the ALD2/3/4/6 Aldehyde dehydrogenases strain under consideration (Mendes-Ferreira et al., 2009). Other sulfur compounds that ATF1 Alcohol o-acetyltransferase contribute to wine aroma include methylmer- ATH1 Vacuolar acid trehalase captan, ethylmercaptan, dimethyl disulfide, CAR1 Arginase and diethyl disulfide (Rauhut & Ku¨ rbel, 1996). COX6 Cytochrome C oxidase CUP1 Metallothionein 3. GENE EXPRESSION DURING Catalase FERMENTATION CTT1 FBA1 Fructose bisphosphate aldolase

The capacity of yeasts to produce a wine with GLK1 Glucokinase desirable properties must be related to the Glyoxylase synthesis of speciﬁc molecules, proteins, and GLO1 products of enzymatic reactions, and, conse- GPD1 Glycerol-3-phosphate dehydrogenase quently, substantial efforts have been made GPH1 Glycogen phosphorylase in recent years to investigate the molecular Phosphoglycerate mutase processes occurring during winemaking. Most GPM1 studies of gene expression during winemaking GRE2 Lactaldehyde dehydrogenase have focused on the alcoholic fermentation GRX5 Mitochondrial glutaredoxin V phase, but reports have also been published on Glutathione synthase I studies undertaken during the phases of indus- GSH1 trial production and rehydration of commercial GSY1/2 Glycogen synthase biomass, and also during aging. Initially, anal- HOR7 Stress response yses focused on genes of interest, but more recent studies have analyzed global gene HSP12 Stress response expression using DNA microarrays. In this HSP26 Molecular chaperone chapter, we review the information available HSP30 Stress response on the expression of particular genes (summa- Molecular chaperone rized in Table 1.2, which indicates their molec- HSP78 ular function or the biological process in which HSP82 Molecular chaperone they are implicated). The application of DNA HSP104 Molecular chaperone microarrays to the understanding of gene Hexokinases expression in wine yeasts during the winemak- HXK1,2 ing process will be discussed in Chapter 6. HXT1-18 Glucose transporters Although in some cases gene expression has MET16 Phosphoadenylylsulfate reductase been analyzed in natural must fermentations, Cytosolic neutral trehalase most of the studies that have been published NTH1 (Continued) 18 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

TABLE 1.2 Saccharomyces cerevisiae Genes Mentioned in This Chapterdcont’d 3.1. Glycolytic Genes The high concentration of sugars in the must Gene Molecular function or biological process represses the expression of genes encoding PAU3 Stress response enzymes involved in mitochondrial respiration PGK1 Phosphoglycerate kinase and also inhibits the activity of the expressed enzymes. As a result, the growth of . PMA1/2 Proton transporters S cerevisiae during vinification largely involves fermenta- POT1 Acetyl-CoA C-acyltransferase tive metabolism. Gene expression analysis SNF3 Glucose sensor shows that this metabolic alternative involves increased flow through the glycolytic pathway. SPI1 Stress response Northern blot analysis of the levels and accumu- SSA3 Stress response lation of messenger RNA (mRNA) during SSA4 Molecular chaperone fermentation of natural musts by wine yeasts has shown that, despite being expressed STI1 Molecular chaperone throughout fermentation, genes linked to the TDH2/3 Glyceraldehyde-3-phosphate glycolytic pathway have specific, dynamic dehydrogenases expression profiles during the different phases THI4 Thiamine synthesis of the growth curve (Puig & Pe´rez-Ortıń, TPS1 Trehalose phosphate synthase 2000a). These studies analyzed the fermentation of both synthetic musts and natural musts Trehalose-6-phosphate phosphatase TPS2 derived from the grape varieties Bobal and TRR1 Thioredoxin reductase I Moscatel. The gene expression patterns TRX2 Thiol-disulfide thioredoxin exchanger observed in the different media led to certain general conclusions. Thus, the genes , Thioredoxin peroxidase I TDH2/3 TSA1 which code for isoenzymes II and III of glyceral- UBI4 Ubiquitin dehyde-3-phosphate dehydrogenase, have the YGP1 Stress response highest expression levels in all of the musts analyzed. In addition, all of the glycolytic genes The molecular function of the gene product is shown, or, when this is reach maximum expression levels during the not known, the biological process in which it is involved (in italics) according to the Saccharomyces Genome Database (http://www. first 24 to 48 h of fermentation, which coincides yeastgenome.org) is shown. with the phase of cell growth and maximum fermentation rate. The extent of mRNA accumu- are based on the use of synthetic musts with lation decreases progressively for all of the a defined chemical composition that mimics, genes analyzed during the stationary phase in among other characteristics, the composition of parallel with the slowing of fermentation. The sugars and total nitrogen in natural musts. results obtained by Puig and Pe´rez-Ortıń The advantage of synthetic media is that (2000a) also indicated a significant difference their defined composition allows reproducible between the musts analyzed. In the case of experiments to be carried out even though the Bobal musts, increases in the expression of all musts are prepared at different points in time. genes were detected between 4 and 6 d after Furthermore, it is possible to analyze the effect inoculation, a finding that did not occur in the that specific changes in must composition have other two musts. In addition, gene expression on the fermentation process and the organo- levels were generally higher in synthetic musts, leptic properties of the final product. and in some cases (FBA1, TDH2/3, and GPM1) GENE EXPRESSION DURING FERMENTATION 19 lower in Moscatel varieties. These differences indicated that the molecular response to these may be related to the way in which each of the conditions is complex and influenced by musts was prepared and the consequent varia- a number of factors. In response to hyperos- tions in their composition. In fact, the same motic stress, S. cerevisiae accumulates glycerol, group observed differences in the expression the main compatible osmolyte in yeasts. This of other genes according to the must used is explained by the observation that the (Puig & Pe´rez-Ortıń, 2000a, 2000b). Comparison main change in the gene expression pattern with the results obtained by Riou et al. (1997) in response to osmotic stress involves the using a synthetic must and a different strain induction of GPD1, which codes for glycerol- highlights differences in the time course of 3-phosphate dehydrogenase, an enzyme in- stationary-phase mRNA levels for various volved in the synthesis of glycerol. Induction genes, including the glycolytic gene PGK1. of the expression of this gene is a rapid phenom- Other data on the expression of glycolytic genes enon occurring in the initial phases of fermenta- in wine fermentations have been obtained from tions with high sugar concentrations (20%), and experiments designed to analyze global gene high levels of GPD1 mRNA are detected 15 min expression. The studies of Backhus et al. after inoculation and reach a maximum induc- (2001), Erasmus et al. (2003), Jimeńez-Marti tion approximately 1 h later. Analysis of the and del Olmo (2008), Marks et al. (2003), Marks expression of other genes that act as markers et al. (2008), and Rossignol et al. (2003) lead to for the general stress response and are depen- the conclusion that transcriptional regulation dent on Msn2/4p transcription factors, such as of glycolytic genes during wine fermentation is HSP12 and HSP104 (Martıńez-Pastor et al., affected under conditions that influence growth 1996), indicates that their expression is low rate and, especially, fermentation rate. following inoculation and their transcription begins to be induced some hours later, when 3.2. Osmotic Stress-response Genes the molecular response to osmotic stress has already finished. The gene-expression patterns The first few hours of fermentation constitute observed during the first few hours of vinifica- a critical period in which the capacity of the tion depend on a variety of factors (Pe´rez- inoculated cells to adapt to the extremely high Torrado et al., 2002a; Zuzuarregui et al., 2005), sugar concentration and initiate fermentation including the metabolic status of the inoculum following a short lag phase is crucial if the inoc- (differences are observed in the expression of ulated strain is to dominate the fermentation at HSP genes when rehydrated cells are used the expense of the natural flora of the must. compared with that seen in cells derived from Given that the inoculated yeast is derived from precultures), the nature of the osmolyte respon- dried and rehydrated or precultured cells, sible for osmotic stress (glucose or glycerol), the contact with the must is likely to involve pH, and the temperature (expression levels of substantial reprogramming of gene expression GPD1 and HSP104 increase with reducing in the yeast cells. temperature between 15 and 28C and with In a study designed to assess the effect of increasing pH between 3.0 and 3.6). hyperosmolarity due to high concentrations of In our laboratory, a study was undertaken glucose in must, the expression of stress- investigating the expression of 19 stress-response response genes was analyzed over the course genes over the course of fermentations using of the first day of fermentation in synthetic commercial and noncommercial strains with musts with differing sugar composition (Pe´rez- different fermentation behavior (Zuzuarregui & Torrado et al., 2002a). The results of that study del Olmo, 2004b). Aside from the differences 20 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION between strains, the data obtained indicated that understand the phenomenon of vinification, there was a significant reduction in mRNA levels since, as mentioned in Section 2.3, approxi- between 1 and 6 h following inoculation of mately the last two thirds of the wine fermenta- a synthetic must for the majority of the genes tion process occurs without cell division but in studied. This was the case for HSP26, SSA3/4, the presence of metabolically active yeasts that STI1, HOR7, GRE2, SPI1, COX6, CAR1,and produce many enologically desirable YGP1, which are implicated in the response compounds (Fuge et al., 1994). Termination of to different stress conditions and have specific cell division essentially occurs as a result of biological roles and regulatory mechanisms. the absence of one or more nutrients, leading These results highlight the importance of genes to a series of physiological, biochemical, and that participate in cellular processes other than morphological changes intended to ensure glycerol synthesis for the capacity of yeasts to survival during periods of shortage, which char- overcome osmotic stress and initiate growth in acterize entry of cultures into the stationary must. More recent studies have shown that phase (Herman, 2002; Werner-Washburne various signal transduction pathways (HOG, et al., 1996). Although under laboratory condi- PKA, and TOR) are involved to a greater or lesser tions the term “stationary phase” strictly applies extent in the transcriptional response to high to the termination of growth as a result of sugar concentrations, and that under these exhaustion of glucose (Werner-Washburne conditions the kinase Hog1p (essential for tran- et al., 1996), in the case of wine fermentation, scriptional activation under conditions of hyper- division can cease in the presence of high quan- osmolarity, reviewed in Hohmann & Mager, tities of sugars (100e150 g/L), since other nutri- 2003) is phosphorylated in a manner similar to ents, primarily nitrogenous compounds, are that seen in response to other types of osmotic usually the first to be consumed and cause stress (Jimeńez-Martı´ et al., unpublished). termination of division (Fleet & Heard, 1993). Recent studies of global gene expression At the end of wine fermentation, a high concen- during rehydration of dried yeast in different tration of ethanol also affects cell metabolism media and conditions and during subsequ- (Jones, 1989). Ethanol toxicity acts via an effect ent inoculation into must (Novo et al., 2007; on the fluidity and permeability of the plasma Rossignol et al., 2006) indicate that inoculation membrane (Alexandre et al., 1994). Both the of must does not lead to a typical stress res- absence of nutrients and the accumulation of ponse, despite the hyperosmotic conditions ethanol coincide with the end of fermentation that affect the yeast. The results show that and their effects are synergistic and difficult to the changes in gene expression, dependent separate. on the presence of fermentable carbon sources, The molecular events associated with entry affect genes coding for proteins involved in into the stationary phase have been described fermentative metabolism, in the nonoxidative under laboratory conditions (Herman, 2002). pentose phosphate pathway, and in ribosome Although winemaking conditions are not biogenesis. comparable to those used in the laboratory, much of the information that has been obtained 3.3. Genes Induced During the can be extrapolated to the conditions currently Stationary Phase used in wineries. For instance, Riou et al. (1997) studied the expression of 19 genes charac- Analysis of gene expression in Saccharomyces terized as being associated with the stationary yeasts during phases in which the yeast is phase in laboratory strains. When those genes not actively dividing is essential in order to were analyzed under vinification conditions in GENE EXPRESSION DURING FERMENTATION 21 an industrial strain of S. cerevisiae, the authors phase under winemaking conditions. However, found that 60% of the genes displayed expres- their role remains unclear. The gene ATF1 sion patterns similar to those obtained under encodes alcohol acetyltransferase, a key enzyme laboratory conditions. In addition to metabolic in the production of aromas (see Section 1.2.4). genes used to respond to external deficiencies, This gene also has a late expression profile. In many genes coding for heat shock proteins a study by Lilly et al. (2000), it became detect- (HSPs) are expressed during the stationary able after 7 d of fermentation and reached phase. Expression of HSPs is typically associ- maximal expression levels after 11 d. This ated with the response to stress and the gene expression pattern is of particular importance products tend to function as molecular chaper- to the final product, since it indicates that ones. These observations confirm the link aromas start to be produced at the end of between the absence of nutrients and other fermentation. Expression of the gene for forms of stress. For instance, in addition to being glycogen synthase (GSY2) is also increased strongly upregulated at the end of vinification, with time of fermentation (Pe´rez-Torrado et al., HSP26 and HSP30 expression is also induced 2002b), indicating that the accumulation of by ethanol (Piper et al., 1994), and this may be glycogen is a factor that defines entry into the responsible for the activation of their expression stationary phase. during the final phases. In studies of genes The expression of YGP1 and CAR1 has been expressed late during microvinification, Puig studied under winemaking conditions with et al. (1996) arrived at similar but not identical different quantities of nitrogen. Under these conclusions. The stress-response genes SSA3, conditions, the expression patterns did not HSP12, and HSP26 were activated at the begin- vary according to the quantity of nitrogen, ning of the stationary phase, but their levels whereas this was not the case under laboratory were reduced at the end of the fermentation growth conditions (Carrasco et al., 2003). These phase. HSP104 and POT1 (both identified as results indicate that information obtained under late-expressed genes under laboratory condi- controlled conditions in the laboratory cannot tions), in contrast, were not expressed under always be extrapolated to the conditions found winemaking conditions. These discrepancies in the winery. Subsequent studies have led to may be explained by differences in the strains the identification of genes that are highly or musts used. The same authors identified expressed under conditions in which nitrogen a new gene of unknown function, SPI1, that is is limiting or absent (Jimeńez-Martı´ et al., 2007; actively expressed during late phases under Mendes-Ferreira et al., 2007a, 2007b). Such vinification conditions (Puig & Pe´rez-Ortıń, genes could be used as markers to identify situ- 2000b). The peculiarities of the transcriptional ations in which fermentation is limited by control of this gene have allowed its promoter nitrogen deficiencies. to be used to manipulate the expression of In our laboratory, we have analyzed the certain genes for biotechnological purposes response to stress during the first half of vinifi- (Cardona et al., 2006; Jimeńez-Martı´ et al., 2009). cation in seven industrial strains previously In another study, the use of Northern blotting characterized in terms of their resistance to to address global gene expression limited to the specific stress conditions (Zuzuarregui & del right arm of chromosome 3, Rachidi et al. (2000) Olmo, 2004a). Coordinated expression of SPI1, identified two genes, PAU3 (a member of YGP1, CAR1, and COX6, which are activated a stress-response gene family of unknown func- in response to a lack of nutrients, was detected tion) and ADH7 (a putative alcohol dehydroge- in all strains during entry into the stationary nase), specifically expressed in the stationary phase after 149 h of culture (Zuzuarregui & del 22 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

Olmo, 2004b). Comparison of the expression of stress, and there was no clear correlation stress-response genes with fermentative between the two variables. Resistance to stress behavior indicates that tight regulation of the is a complex phenomenon and the genes response to various stress conditions may be involved are likely to differ in importance fundamental to the adaptation of the yeast to according to specific conditions. Consequently, the medium, although the absolute level of a more complete view of the process will require expression of the genes may sometimes be of identification of a larger number of genes. lesser importance. In our laboratory, we have also analyzed the effect of ethanol and, for the first time, acetalde- 3.4. Gene Expression in Wine Yeasts hyde as isolated stress conditions in different Exposed to Specific Stress Conditions wine and laboratory strains. Acetaldehyde accumulates within the cell during intense fermenta- Various stress conditions can affect the effi- tion and can halt cell division (Stanley et al., ciency of yeasts during the different phases of 1993), and it has been proposed to be largely wine production (Attfield, 1997; Bauer & Pretor- responsible for ethanol toxicity (Jones, 1989). ius, 2000). Although it is essential to study the Although both compounds generated transcrip- response of wine yeasts during real industrial tional responses in which the expression of processes, analysis of their behavior in response stress-response genes such as HSP12, HSP26, to specific individual stresses allows assessment HSP82, and HSP104 was activated, acetalde- of correlations with their adaptation to unfavor- hyde was a better transcriptional inducer than able environmental changes. ethanol at lower concentrations (Aranda & del Ivorra et al. (1999) characterized the expres- Olmo, 2004; Aranda et al., 2002). Those strains sion of typical stress-response genes (GPD1, displaying greater induction in response to HSP12, STI1, SSA3, and TRX2) to adverse condi- ethanol and acetaldehyde were more resistant tions such as thermal shock, oxidative stress, to these stress conditions. The expression of hyperosmolarity, ethanol, and the absence of genes encoding aldehyde dehydrogenases is glucose. Although there was variability among also transcriptionally regulated by acetaldehyde different strains of wine yeast, comparison and, to a lesser extent, by ethanol (Aranda & del with a laboratory strain showed that the stress Olmo, 2003). response mechanisms were essentially the The stress caused by addition of sulfite also same in all cases. HSP12, which encodes a small has transcriptional effects. Maximal induction HSP, was the best marker for differences of the genes MET16 and ADE4 (biosynthesis of between strains, since it was expressed in all methionine and adenine) is delayed during vini- conditions and its levels of transcription in fication in the presence of SO2 (Aranda et al., suboptimal conditions were lower in the strain 2006). These transcriptional changes probably displaying the greatest fermentative difficulty. represent the molecular mechanism that links This was the first indication of a link between tolerance of sulfite to the metabolism of adenine defects in the stress response and stuck fermen- and sulfur amino acids. tation. That study was extended in the first Other studies have revealed a correlation systematic analysis of stress responses in 14 between resistance to the stress conditions rele- commercial strains, which analyzed the expres- vant to wine production (oxidative stress and sion of the marker stress-response genes HSP12 ethanol) and appropriateness for winemaking and HSP104 in response to thermal shock (Zuzuarregui & del Olmo, 2004a), a correlation (Carrasco et al., 2001). These two genes were that can be extended to the biological aging of expressed in all strains under conditions of certain wines, since flor strains with greater GENETIC IMPROVEMENT OF YEAST EFFICIENCY DURING FERMENTATION 23 resistance to the substantial stress associated nutritional and organoleptic properties of the with this process (acetaldehyde, ethanol, product (reviewed in Dequin, 2001; Pretorius & moderate cold) are more abundant in soleras Bauer, 2002; Pretorius, 2003). Here we describe (Aranda et al., 2002). some strategies for the genetic improvement of It is increasingly apparent that post-tran- the fermentation efficiency of yeasts. We will scriptional events such as the processing and not discuss other types of genetic manipulation, transport of mRNA also influence gene expres- as these will be described in Chapter 7. sion. Thus, as the vinification process prog- S. cerevisiae is a highly efficient fermentative resses, mRNA is observed to accumulate in the organism. However, on occasion, stuck fermen- nucleus, and this is associated with the accumu- tations or extended lag phases can prevent the lation of ethanol at concentrations of more than yeasts from displacing the autochthonous flora. 6% (Izawa et al., 2005). Different strategies have been developed to Transcriptional analysis of the stress response augment the fermentative capacity of active has also been undertaken in laboratory-scale dried yeast; for instance, through accumulation simulations of the industrial processes of of certain carbohydrate reserves implicated in biomass propagation and dehydration to obtain stress resistance, such as glycogen and treha- active dried yeast (Pe´rez-Torrado et al., 2005, lose (Sillje´ et al., 1999). The accumulation of 2009). Those studies have addressed the expres- these metabolites can be modified by increasing sion of a panel of genes that act as markers for the expression levels of genes that participate different stresses (TRX2, STI1, HSP12, GPD1, in their biosynthesis (GSY1 and GSY2 for CUP1, GLO1, CTT1, GSH1, YGP1, GRE2, glycogen [Farkas et al., 1991] and TPS1 and GRX5, TSA1, TRR1). As a result, the response TPS2 for trehalose [Gonza´lez et al., 1992; Vuorio to oxidative stress has been identified as the et al., 1993]) or by eliminating those involved in most important in determining the fermentative their mobilization (GPH1 [Hwang et al., 1989] efficiency of commercial inoculums during both and NTH1 and ATH1 [Nwaka & Holzer, 1998; yeast growth and dehydration (Garre et al., Nwaka et al., 1995]). In our laboratory, wine 2010; Pe´rez-Torrado et al., 2005, 2009). strains have been developed that show greater accumulation of glycogen as a result of regulated overexpression of the glycogen synthase 4. GENETIC IMPROVEMENT OF encoded by the gene GSY2. The manipulated YEAST EFFICIENCY strain accumulates more glycogen during vini- DURING FERMENTATION fication in natural and synthetic musts, and under growth conditions similar to those used The use of pure strains for inoculation of in industry to obtain yeast biomass (Pe´rez-Tor- industrial fermentations, together with a good rado et al., 2002b). It also showed greater understanding of the biochemistry, genetics, viability when cells were recovered at the end and molecular biology of S. cerevisiae, make it of vinification following a period of 10 d in possible to develop strategies for the genetic the finished wine, and this was accompanied improvement of their efficiency, taking into by an increased fermentative capacity of the consideration the preservation of the genetic cells when reinoculated in fresh media. Other and genomic properties of natural strains in strategies that can increase the fermentative order to conserve their fermentative characteris- capacity are aimed at improving the efficiency tics (Gimeno-Alcan˜iz & Matallana, 2001). Many of uptake and phosphorylation of sugars and aspects are open to improvement, in terms have focused on the hexokinases HXK1 and of both the fermentation process and the HXK2, the glucokinase GLK1, and the hexose 24 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION transporters HXT1-HXT18 and SNF3 (Pretorius other microorganisms in the must. The accumu- & Bauer, 2002). It would also be of interest to lation of ethanol has negative consequences introduce heterologous genes coding for trans- during fermentation because of its multiple porters and kinases that allow improvement toxic effects. The main effect is on membrane in the utilization of fructose, a sugar that permeability, leading to loss of ions (principally usually accumulates as a result of the preferen- magnesium and calcium) (Dombeck & Ingram, tial use of glucose by the yeasts (Pretorius & 1986; Nabais et al., 1988), and the passive diffu- Bauer, 2002). sion of protons that alters the pH with Modulation of the stress response might also increasing concentrations of ethanol (Cart- contribute to improvements in the fermentative wright et al., 1986; Leao & van Uden, 1984). capacity of yeasts, given its relationship with the These effects appear to be explained by changes behavior of yeasts during winemaking (Aranda in the activity of ATPase proton pumps (Cart- et al., 2002; Ivorra et al., 1999; Zuzuarregui & del wright et al., 1989; Rosa & Sa´-Correia, 1991) Olmo, 2004a, 2004b). Recently, our group has and in membrane fluidity (Goldstein, 1987; developed strategies based on the modification Kunkee & Bisson, 1993; Sun & Sun, 1985). Strat- of stress-response genes in wine yeasts that egies designed to improve ethanol tolerance have led to an improvement in fermentative should be directed towards the stimulation of behavior (Cardona et al., 2006; Jimeńez-Martı´ sterol and long-chain unsaturated fatty acid et al., 2009; Pe´rez-Torrado et al., 2009). The metabolism to maintain membrane fluidity, following genes were used: TRX2 (encoding and should also focus on genes coding for a cytoplasmic thioredoxin), MSN2 (a transcrip- components of the membrane ATPase (PMA1 tion factor involved in the general stress and PMA2) to make it less sensitive to ethanol response), HSP26, and YHR087W (a gene of toxicity (Pretorius & Bauer, 2002). Wine strains unknown function that is induced under condi- have also been developed that produce more tions of osmotic stress, including those gener- ethanol due to an increased tolerance of osmotic ated by high concentrations of glucose; stress and ethanol (Hou et al., 2009). This was Jimeńez-Martı´, Zuzuarregui and del Olmo, made possible by simultaneous overexpression unpublished results). These examples indicate of the general transcription factor Spt15p and that it is possible to obtain an improvement in the Spt3p subunit of SAGA complexes, impli- the rate of consumption of sugars in different cated in the transcriptional activation of RNA- musts, under different conditions, and during polymerase II-dependent genes. specific stages of the vinification process by The examples discussed in this chapter high- modifying stress-response genes, and this could light how metabolic engineering can be help to facilitate the establishment of the inocu- exploited to improve the industrial applications lated strain and reduce fermentation time. of S. cerevisiae. Our increasing understanding of The low concentration of easily assimilable how this organism functions under wine nitrogen sources in musts is a significant cause fermentation conditions will lead to ever- of loss of fermentative capacity. This problem greater opportunities for genetic improvement may be alleviated through the use of modified of the yeast strains employed. wine yeasts able to use sources of nitrogen, such as proline, that are highly abundant in some musts but are not utilized during fermen- References tation (Henschke, 1997). Alexandre, H., Rosusseaux, I., & Charpenter, C. (1994). Ethanol tolerance is very important for the Ethanol adaptation mechanisms in Saccharomyces cer- capacity of inoculated yeasts to dominate the evisiae. Biotechnol. Appl. Biochem., 20, 173e183. REFERENCES 25

Aranda, A., & del Olmo, M. (2003). Response to acetalde- Bely, M., Sablayrolles, J. M., & Barre, P. (1990). Automatic hyde stress in the yeast Saccharomyces cerevisiae involves detection of assimilable nitrogen deficiencies during a strain-dependent regulation of several ALD genes and alcoholic fermentation in oenological conditions. is mediated by the general stress response pathway. J. Ferment. Bioeng., 70, 246e252. Yeast, 20, 749e759. Bertrand, A., & Miele, A. (1984). Influence de la clarification Aranda, A., & del Olmo, M. (2004). Exposure to acetal- du mout de raisin sur sa teneeur en acides gras. Connais dehyde in yeast causes an induction of sulfur amino Vigne Vin, 18, 293e297. acid metabolism and polyamine transporter genes, Bidenne, C., Blondin, B., Dequin, S., & Vezinhet, F. (1992). which depends on Met4p and Haa1p transcription Analysis of the chromosomal DNA polymorphism of factors respectively. Appl. Environ. Microbiol., 70, wine strains of Saccharomyces cerevisiae. Curr. Genet., 22, 1913e1922. 1e7. Aranda, A., Jimeńez-Martı´, E., Orozco, H., Matallana, M., & Bisson, L. (1999). Stuck and sluggish fermentations. Am. J. del Olmo, M. (2006). Sulfur and adenine metabolisms are Enol. Viticult., 50, 107e119. linked and both modulate sulfite resistance in wine Blomberg, A., & Adler, L. (1992). Physiology of osmotol- yeast. J. Agric. Food. Chem., 54, 5839e5846. erance in fungi. Adv. Microbiol. Physiol., 33, 145e212. Aranda, A., Querol, A., & del Olmo, M. (2002). Correlation Boulton, R. B., Singleton, V. L., Bisson, L. F., & Kunkee, R. E. between acetaldehyde and ethanol resistance and (Eds.), (1996). Principles and practices of winemaking. expression of HSP genes in yeast strains isolated during New York, NY: Chapman. the biological aging of sherry wines. Arch. Microbiol., 177, Cabanis, J. C., Cabanis, M. T., Cheynier, V., & Teissedre, P. L. 304e312. (1998). Tables de composition. In E. Flanzy (Ed.), Oenologie. Attfield, P. V. (1997). Stress tolerance: The key to effective Fondements scientifiques et technologiques (pp. 315e336). strains of industrial baker’s yeast. Nature Biotechnology, Paris, France: Technique & Documentation. 15, 1351e1357. Cabras, P., & Angioni, A. (2000). Pesticide residues in Ayestaran, B. M., Ancin, M. C., Garcıá, A. M., Gonza´lez, A., grapes, wines, and their processing productos. J. Agric. & Garrido, J. J. (1995). Influence of prefermentation Food Chem., 48, 967e973. clarification on nitrogenous contents of musts and Cantarelli, C. (1989). Phenolics and yeast: Remarks con- wines. J. Agric. Food Chem., 43, 476e482. cerning fermented beberages. Yeast, 5, S53e61. Backhus, L. E., DeRisi, J., Brown, P. O., & Bisson, L. F. (2001). Cardona, F., Carrasco, P., Pe´rez-Ortıń, J. E., del Olmo, M., & Functional genomic analysis of a commercial wine strain Aranda, A. (2006). Effect of a modification in the of Saccharomyces cerevisiae under differing nitrogen expression of MSN2 transcription factor in wine yeast on conditions. FEMS Yeast Res., 1, 111e125. stress resistance and fermentative behaviour. Int. J. Food Bakalinsky, A. T., & Snow, R. (1990). The chromosomal Microbiol., 114,83e91. constitution of wine strains of Saccharomyces cerevisiae. Carrasco, P., Pe´rez-Ortıń, J. E., & del Olmo, M. (2003). Yeast, 6, 367e382. Arginase activity is a useful marker of nitrogen limita- Barre, P. (1980). Role du facteur killer dans la concurrence tion during wine fermentations. System. Appl. Microbiol., entre souches de levures. Bulletin de l’Office Internat. de la 26, 471e479. Vigne et du Vin, 593e594, 560e572. Carrasco, P., Querol, A., & del Olmo, M. (2001). Analysis of Barre, P., Ve´zinhet, F., Dequin, S., & Blondin, B. (1993). the stress resistance of commercial wine yeast strains. Genetic improvement of wine yeast. In G. H. Fleet (Ed.), Arch. Microbiol., 175, 450e457. Wine Microbiology and Biotechnology (pp. 421e447). UK: Carrau, F. M., Medina, K., Farina, L., Boido, E., Harwood Academic, Reading. Henschke, P. A., & Dellacassa, E. (2008). Production of Bauer, F. F., & Pretorius, I. S. (2000). Yeast stress response fermentation aroma compounds by Saccharomyces cer- and fermentation efficiency: How to survive the making evisiae wine yeasts: Effects of yeast assimilable nitrogen of wine. A review. S. Afr. J. Enol. Vitic., 21,27e51. on two model strains. FEMS Yeast Res., 8, 1196e1207. Baumes, R., Cordonnier, R., Nitz, S., & Drawert, F. (1986). Cartwright, C. P., Juroszek, J.-R., Beavan, M. J., Ruby, F. M. S., Identification and determination of volatile constituents DeMorias, S. M. F.,& Rose, A. H. (1986). Ethanol dissipates in wines from different vine cultivars. J. Sci. Food Agric., the proton-motive force across the plasma membrane 37, 927e943. of Saccharomyces cerevisiae. J. Gen. Microbiol., 132, Beltrań, G., Esteve-Zarzoso, B., Roze´s, N., Mas, A., & 369e377. Guillamoń, J. M. (2005). Influence of timing of nitrogen Cartwright, C. P., Rose, A. H., Calderbank, J., & additions during synthetic grape must fermentations on Keenan, H. J. (1989). Solute transport. In A. H. Rose & fermentation kinetics and nitrogen consumption. J. S. Harrison (Eds.), The yeast. Metabolism and physiology J. Agric. Food Chem., 53, 996e1002. of yeast (pp. 5e49). London, UK: Academic Press. 26 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

Castor, J. G. B., & Archer, T. E. (1956). Amino acids in must Dubois, P. (1983). Volatile phenols in wine. In J. R. Piggot and wines, praline, serine and threonine. Am. J. Enol. (Ed.), Flavour of distilled beverages, origin and development Viticult., 7,19e25. (pp. 110e119). Chichester, UK: Ellis Horwood. Chatonnet, P., Viala, C., & Dubourdieu, D. (1997). Influence Dubourdieu, D., Darriet, O., Ollivier, J. N., & Ribereau- of polyphenolic components of red wines on the Gayon, P. (1988). Role de la levure Saccharomyces cer- microbial synthesis of volatile phenols. Am. J. Enol. Vit- evisiae dans l’hydrolyse enzymatique des he´terosides icult., 48, 443e448. terpeńiques du jus de raisin. C.R. Acad. Sci., 306, 489e493. Christiansen, G., & Christiansen, C. (1976). Comparison of Egli, C. M., Edinger, W. D., Mitrakul, C. M., & Henick- the fine structure in the mitochondrial DNA from Kling, T. (1998). Dynamics of indigenous and inoculated Saccharomyces cerevisiae and S. carlsbergensis: Electron yeast populations and their effect on the sensory char- microscopy of partially denatured molecules. Nucleic acter of Riesling and Chardonnay wines. J. Appl. Micro- Acids Res., 3, 465e476. biol., 85, 779e789. Clark-Walker, G. D., McArthur, C. R., & Daley, D. J. (1981). Eglinton, J. M., & Henschke, P. A. (1999). The occurrence of Evolution of mitochondrial genomes in fungi. Curr. volatile acidity in Australian wines. Australian Grape- Genet., 4,7e12. grower and Winemaker, 426a,7e12. Codoń, A. C., Gasent-Ramı´rez, J. M., & Benı´tez, T. (1995). Erasmus, D. J., van der Merwe, G. K., & van Vuuren, H. J. J. Factors which affect the frequency of sporulation and (2003). Genome-wide expression analyses: Metabolic tetrad formation in Saccharomyces cerevisiae bakers yeasts. adaptation of Saccharomyces cerevisiae to high sugar Appl. Environ. Microbiol., 59, 345e386. stress. FEMS Yeast Res., 3, 375e399. Cooper, T. G. (1982a). Nitrogen metabolism in Saccharomyces Esteve-Zarzoso, B., Belloch, C., Uruburu, F., & Querol, A. cerevisiae. In J. N. Strathern, E. W. Jones & J. B. Broach (1999). Identification of yeast by RFLP analysis of the (Eds.), The molecular biology of the yeast Saccharomyces: 5.8S rRNA gene and the two ribosomal internal tran- Metabolism and gene expression (pp. 39e99). Cold Spring scribed spacers. Int. J. Syst. Bacteriol., 49, 329e337. Harbor, NY: Cold Spring Harbor Laboratory. Etievant, P. X. (1981). Volatile phenol determination in wine. Cooper, T. G. (1982b). Transport in Saccharomyces cerevisiae. J. Agric. Food Chem., 29,65e67. In J. N. Strathern, E. W. Jones, & J. B. Broach (Eds.), The Farkas, I., Hardy, T. A., Goebl, M. G., & Roach, P. J. (1991). molecular biology of the yeast Saccharomyces: Metabolism and Two glycogen synthase isoforms in Saccharomyces cer- gene expression (pp. 399e461). Cold Spring Harbor, NY: evisiae are coded by distinct genes that are differentially Cold Spring Harbor Laboratory. controlled. J. Biol. Chem., 266, 15602e15607. Corison, C. A., Ough, C. S., Berg, H. W., & Nelson, K. E. Farkas, J. (1988). Technology and biochemistry of wine. New (1979). Must acetic acid and ethyl acetate as mold rot York, NY: Gordon and Breach. indicators in grapes. Am. J. Enol. Viticult., 30, 130e134. Fleet, G. H. (2007). Wine. In M. P. Doyle, & L. R. Beuchat Darriet, P., Boldron, J. N., & Dubourdieu, D. (1988). (Eds.), Food microbiology: Fundamentals and frontiers L’hydrolyse des hete´rosides terpeńiques du Muscat a`pe- (3rd ed.), (pp. 863e890) Washington, DC: ASM Press. tits grains par les enzymes pe´riplasmiques de Saccharo- Fleet, G. H. (2008). Wine yeasts for the future. FEMS Yeast myces cerevisiae. Connaiss. Vigne Vin, 22, 189e195. Res., 8, 979e995. Degre´, R. (1993). Selection and commercial cultivation of Fleet, G. H., & Heard, G. (1993). Yeast growth during wine yeast and bacteria. In G. H. Fleet (Ed.), Wine fermentation. In G. Fleet (Ed.), Wine microbiology and microbiology and biotechnology (pp. 421e447). Chur, biotechnology (pp. 27e57). Chur, Switzerland: Harwood Switzerland: Harwood Academic Publishers. Academic Publishers. del Pozo, L. D., Abarca, D., Claros, M. G., & Jimeńez, A. Fremont, L. (2000). Biological effects of resveratrol. Life Sci., (1991). Cycloheximide resistance as a yeast cloning 66, 663e673. marker. Curr. Genet., 19, 353e358. Fuge, E. K., Braun, E. L., & Werner-Washburne, M. (1994). Delteil, D., & Aizac, T. (1988). Comparaison de diffe´rences Protein synthesis in long-term stationary-phase cultures techniques de levurage par suivi de l’implantation d’une of Saccharomyces cerevisiae. J. Bacteriol., 176, 5802e5813. souche de levure oenologique marqueé. Connaiss. Vigne Fugelsang, K. C., & Edwards, C. G. (2007). Wine microbi- Vin, 21, 267e278. ology: Practical applications and procedures. New York, NY: Dequin, S. (2001). The potential of genetic engineering for Springer. improving brewing, wine-making and baking yeasts. Gancedo, J. M. (2008). The early steps of glucose signalling Appl. Microbiol. Biotechnol., 56, 577e588. in yeast. FEMS Microbiol. Rev., 32, 673e704. Dombeck, K. M., & Ingram, L. O. (1986). Magnesium limita- Garre, E., Raginel, F., Palacios, A., Julien, A., & Matallana, E. tion and its role in apparent toxicity of ethanol during (2010). Oxidative stress responses and lipid peroxidation yeast fermentation. Appl. Environ. Microbiol., 52, 975e981. damage are induced during dehydration in the REFERENCES 27

production of dry active wine yeasts. Int. J. Food Micro- Hollenberg, C. P., Borst, P., & van Bruggen, E. F. J. (1970). biol., 136(3), 295e303. Mitochondrial DNA. A 25 micron closed circular duplex Gimeno-Alcan˜iz, J. V., & Matallana, E. (2001). Performance of DNA molecule in wild type yeast mitochondria. Struc- industrial strains of Saccharomyces cerevisiae during wine ture and genetic complexity. Biochim. Biophys. Acta., 209, fermentation is affected by manipulation strategies based 1e15. on sporulation. System. Appl. Microbiol., 24,639e644. Horak, J. (1986). Amino acid transport in eukaryotic Giudici, P., Zambonelli, C., & Kunkee, R. E. (1993). microorganisms. Biochim. Biophys. Acta., 864, 223e256. Increased production of n-propanol in wine by yeast Hou, L., Cao, X., Wang, C., & Lu, M. (2009). Effect of strains having an impaired ability to form hydrogen overexpression of transcription factors on the fermen- sulphide. Am. J. Enol. Vitic., 44, 123e127. tation properties of Saccharomyces cerevisiae industrial Goldstein, D. (1987). Ethanol-induced adaptation in bio- strains. Lett. Appl. Microbiol., 49,14e19. logical membranes. Ann. N.Y. Acad. Sci., 492, 103e111. Howitz, K. T., Bitterman, K. J., Cohen, H. Y., Lamming, D. W., Gonza´lez, M. I., Stucka, R., Bla´zquez, M. A., Feldmann, H., Lavu, S., Wood, J. G., et al. (2003). Small molecule activa- & Gancedo, C. (1992). Molecular cloning of CIF1, a yeast tors of sirtuins extend Saccharomyces cerevisiae lifespan. gene necessary for growth on glucose. Yeast, 8, 183e192. Nature, 425, 191e196. Grenson, M., Mousset, M., Wiame, J. M., & Bechet, J. (1966). Hwang, P. K., Tugendreich, S., & Fletterick, R. J. (1989). Multiplicity of the amino acid permeases in Saccharo- Molecular analysis of GPH1, the gene encoding myces cerevisiae. I. Evidence for a specific arginine- glycogen phosphorylase in Saccharomyces cerevisiae. transporting system. Biochem. Biophys. Acta., 127, Mol. Cell. Biol., 9, 1659e1666. 325e338. Ingledew, M., & Kunkee, R. E. (1985). Factors influencing Guijo, S., Mauricio, J. C., Salmon, J. M., & Ortega, J. M. sluggish fermentations of grape juice. Am. J. Enol. Viti- (1997). Determination of the relative ploidy in different cult., 36,65e76. Saccharomyces cerevisiae strains used for fermentation Ivorra, C., Pe´rez-Ortıń, J. E., & del Olmo, M. (1999). An and “flor” film ageing of dry sherry-type wines. Yeast, inverse correlation between stress resistance and stuck 13, 101e117. fermentations in wine yeasts. A molecular study. Bio- Hadfield, C., Jordan, B. E., Mount, R. C., Pretorius, G. H., & technol. Bioeng., 64, 698e708. Burak, E. (1990). G418-resistance as a dominant marker Izawa, S., Takemura, R., Miki, T., & Inoue, Y. (2005). Char- and reporter for gene expression in Saccharomyces cer- acterization of the export of bulk poly(A)þ mRNA in evisiae. Curr. Genet., 18, 303e313. Saccharomyces cerevisiae during the wine making Heard, G. M., & Fleet, G. H. (1988). The effects of temper- process. Appl. Environ. Microbiol., 71, 2179e2182. ature and pH on the growth of yeast species during the Jackson, R. S. (1994). Wine science. Principles and applications. fermentation of grape juice. J. Appl. Bact., 65,23e28. San Diego, CA: Academic Press. Henick-Kling, T., Edinger, W., Daniel, P., & Monk, P. (1998). Jimeńez-Martı´, E., Aranda, A., Mendes-Ferreira, A., Selective effects of sulfur dioxide and yeast starter Mendes-Faia, A., & del Olmo, M. (2007). The nature of culture addition on indigenous yeast populations and the nitrogen source added to nitrogen depleted vinifi- sensory characteristics of wine. J. Appl. Microbiol., 84, cations conducted by a Saccharomyces cerevisiae strain 865e876. in synthetic must affects gene expression and the levels Henschke, P. A. (1997). Wine yeasts. In F. K. Zimmermann & of several volatile compounds. Anton. Leeuw., 92,61e75. K.-D. Entian (Eds.), Yeast sugar metabolism (pp. 527e560). Jimeńez-Martı´, E., & del Olmo, M. (2008). Addition of Lancaster, PA: Technomic Publishing Co. ammonia or amino acids to a nitrogen-depleted medium Henschke, P. A., & Jiranek, V. (1993). Metabolism of nitrogen affects gene expression patterns in yeast cells during compounds. In G. H. Fleet (Ed.), Wine microbiology and alcoholic fermentation. FEMS Yeast Res., 8, 245e256. biotechnology (pp. 77e164). Chur, Switzerland: Harwood Jimeńez-Martı´, E., Zuzuarregui, A., Ridaura, I., Lozano, N., Academic Publishers. & del Olmo, M. (2009). Genetic manipulation of HSP26 Heresztyn, N. (1984). Methyl and ethyl amino acid esters in and YHR087W stress genes may improve fermentative wine. J. Agric. Food Chem., 32, 916e918. behaviour in wine yeasts under vinification conditions. Herman, P. K. (2002). Stationary phase in yeast. Curr. Opin. Int. J. Food Microbiol., 130, 122e130. Microbiol., 5, 607e692. Jones, R. P. (1989). Biological principles for the effects of Herraiz, T., & Ough, C. S. (1993). Formation of ethyl esters ethanol. Enzyme Microbiol. Technol., 11, 130e153. of amino acids by yeasts during the alcoholic fermen- Kaeberlein, M., Andalis, A. A., Fink, G. R., & Guarente, L. tation of grape juice. Am. J. Enol. Viticult., 44,41e48. (2002). High osmolarity extends life span in Saccharo- Hohmann, S., & Mager, W. H. (2003). Yeast stress responses. myces cerevisiae by a mechanism related to calorie Berlin, Germany: Springer-Verlag. restriction. Mol. Cell. Biol., 22, 8056e8066. 28 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

Kudo, M., Vagnoli, P., & Bisson, L. F. (1998). Imbalance of D. M. Spencer (Eds.), Yeast technology (pp. 105e123). potassium and hydrogen ion concentrations as a cause Berlin, Germany: Springer-Verlag. of stuck enological fermentations. Am. J. Enol. Viticult., Mendes-Ferreira, A., Barbosa, C., Falco, V., Leaõ, C., & 49, 295e301. Mendes-Faia, A. (2009). The production of hydrogen Kunkee, R. E., & Bisson, L. F. (1993). Wine-making yeats. In sulphide and other aroma compounds by wine strains of A. H. Rose & J. S. Harrison (Eds.), (2nd ed.), The yeasts, Saccharomyces cerevisiae in synthetic media with Vol. 5 (pp. 69e127). London, UK: Academic Press. different nitrogen concentrations. J. Int. Microbiol. Bio- Lafon-Lafourcade, S. (1983). Wine and brandy. In G. Reed technol., 36, 571e583. (Ed.), Biotechnology (pp. 81e163). Heidelberg, Germany: Mendes-Ferreira, A., del Olmo, M., Garcıá-Martıńez, J., Verlag-Chemie. Jimeńez-Martı´, E., Leaõ, C., Mendes-Faia, A., et al. Lagunas, R. (1986). Misconceptions about the energy (2007a). Saccharomyces cerevisiae signatura genes for metabolism of Saccharomyces cerevisiae. Yeast, 2, 221e228. predicting nitrogen deficiency during alcoholic fermen- Lambrechts, M. G., & Pretorius, I. S. (2000). Yeast and its tation. Appl. Env. Microbiol., 73, 5363e5369. importance to wine aroma e A review. S. Afr. J. Enol. Mendes-Ferreira, A., del Olmo, M., Garcıá-Martıńez, J., Vitic., 21,97e129. Jimeńez-Martı´, E., Mendes-Faia, A., Pe´rez-Ortıń, J. E., & Leao, C., & van Uden, N. (1984). Effects of ethanol and other Leaõ, C. (2007b). Transcriptional response of Saccharo- alkanols on passive proton influx in the yeasts Saccha- myces cerevisiae to different nitrogen concentrations romyces cerevisiae. Biochim. Biophys. Acta., 774,43e48. during alcoholic fermentation. Appl. Env. Microbiol., 73, Lilly, M., Lambretsch, M. G., & Pretorius, I. S. (2000). Effect 3049e3060. of increased yeast alcohol acetyltransferase activity on Mendes-Ferreira, A., Mendes-Faia, A., & Leaõ, C. (2004). flavour profiles of wine and distillates. Appl. Environ. Growth and fermentation patterns of Saccharomyces Microbiol., 66, 744e753. cerevisiae under different ammonium concentrations Llaurado´ i Reverchon, J.M. (2002). Avaluacio´ dels con- and its implications in winemaking industry. J. Appl. dicionants del most en el desenvolupament de la fermentacio´ Microbiol., 97, 540e545. alcoho`lica a baixes temperaturas. Doctoral thesis. Tarragona, Monteiro, F., & Bisson, L. F. (1991). Amino acid utilization Spain: Universitat Rovira i Virgili. and urea formation during vinification fermentations. Longo, E., & Ve´zinhet, F. (1993). Chromosomal rearrange- Am. J. Enol. Viticult., 43,1e10. ments during vegetative growth of a wild strain of Monteiro, F. F., & Bisson, L. F. (1992). Utilization of adenine Saccharomyces cerevisiae. Appl. Environ. Microbiol., 59, by yeast during grape juice fermentation and investi- 322e326. gation of the possible role of adenine as a precursor of Marini, A. M., Soussi-Boudekou, S., Vissers, S., & Andre, B. urea. Am. J. Enol. Viticult., 43,18e22. (1997). A family of ammonium transporters in Saccha- Mortimer, R. K., Romano, P., Suzzi, G., & Polsinelli, P. (1994). romyces cerevisiae. Mol. Cell Biol., 17, 4282e4293. Genome renewal: A new phenomenon revealed from Marini, A. M., Vissers, S., Urrestarazu, A., & Andre, B. a genetic study of 43 strains of Saccharomyces cerevisiae (1994). Cloning and expression of the MEP1 gene derived from natural fermentation of grape musts. Yeast, encoding an ammonium transporter in Saccharomyces 10, 1543e1552. cerevisiae. EMBO J., 13, 3456e3463. Munõz, E., & Ingledew, W. M. (1990). Yeast hulls in wine Marks, V. D., Ho Sui, S. J., Erasmus, D., van der fermentation. A review. J. Wine Res., 1, 197e209. Merwe, G. K., Brumm, J., Wasserman, W. W., et al. Nabais, R. C., Sa´-Correia, I., Viegas, C. A., & Novais, J. M. (2008). Dynamics of the yeast transcriptome during wine (1988). Influence of calcium ion on ethanol fermentation fermentation reveals a novel fermentation stress by yeasts. Appl. Environ. Microbiol., 54, 2439e2446. response. FEMS Yeast Res., 8,35e52. Nishino, H., Miyazakim, S., & Tohjo, K. (1985). Effect of Marks, V., van der Merwe, G., & van Vuuren, H. J. J. (2003). osmotic pressure on the growth rate and the fermentation Transcriptional profiling of wine yeast in fermenting activity of wine yeast. Am. J. Enol.Viticult., 36, 170e174. grape juice: Regulatory effect of diammonium phos- Norbeck, J., & Blomberg, A. (1997). Metabolic and regula- phate. FEMS Yeast Research, 3, 269e287. tory changes associated with growth of Saccharomyces Martıńez-Pastor, M. T., Marchler, G., Schuller, C., Marchler- cerevisiae in 1.4 M NaCl. J. Biol. Chem., 272, 5544e5554. Bauer, A., Ruis, H., & Estruch, F. (1996). The Saccharo- Nordstro¨m, K. (1964a). Formation of esters from acids by myces cerevisiae zinc finger proteins Msn2p and Msn4p brewer’s yeast. II. Formation from lower fatty acids. are required for transcriptional induction through the J. Inst. Brew., 70,42e55. stress-response element (STRE). EMBO J., 15, 2227e2235. Nordstro¨m, K. (1964b). Formation of esters from acids by Martini, A., & Vaughan-Martini, A. (1990). Grape must brewer’s yeast. IV. Effect of higher fatty acids and fermentation: Past and present. In J. F. T. Spencer & toxicity of lower fatty acids. J. Inst. Brew., 70, 233e242. REFERENCES 29

Nordstro¨m, K. (1964c). Formation of esters from alcohols by Study of the first hours of microvinification by the use of brewer’s yeast. J. Inst. Brew., 70, 328e336. osmotic stress-response genes as probes. System. Appl. Novo, M., Beltran, G., Rozes, N., Guillamon, J. M., Sokol, S., Microbiol., 25, 153e161. Leberre, V., et al. (2007). Early transcriptional response of Pe´rez-Torrado, R., Gimeno-Alcan˜iz, J. V., & Matallana, E. wine yeast after rehydration: Osmotic shock and meta- (2002b). Wine yeast strains engineered for glycogen bolic activation. FEMS Yeast Res., 7, 304e316. overproduction display enhanced viability under Nwaka, S., & Holzer, H. (1998). Molecular biology of glucose deprivation conditions. Appl. Environ. Microbiol., trehalose and trehalases in the yeast Saccharomyces 68, 3339e3344. cerevisiae. Prog. Nucleic Acid Res. Mol. Biol., 58, 197e237. Pe´rez-Torrado, R., Go´mez-Pastor, R., Larsson, C., & Nwaka, S., Mechler, B., Destreulle, M., & Holzer, H. (1995). Matallana, E. (2009). Fermentative capability of dry Phenotypic features of trehalase mutants in Saccharo- active wine yeast requires a specific oxidative stress myces cerevisiae. FEBS Lett., 360, 286e290. response during industrial biomass growth. Appl. Nyka¨nen, L. (1986). Formation and occurrence of flavour Microbiol. Biotechnol., 81, 951e960. compounds in wine and distilled alcoholic beverages. Piper, P. W., Talreja, K., Panaretou, B., Moradas-Ferreira, P., Am. J. Enol. Viticult., 37,84e96. Byrne, K., Praekelt, U. M., et al. (1994). Induction of Nyka¨nen, L. L., Nyka¨nen, I., & Soumalainen, H. (1977). major heat-shock proteins of Saccharomyces cerevisiae, Distribution of esters produced during sugar fermenta- including plasma membrane Hsp30, by ethanol levels tion between the yeast cell and the medium. J. Inst. Brew., above critical threshold. Microbiology, 140, 3031e3038. 83,32e34. Pretorius, I. S. (2000). Tailoring wine yeast for the new Oliva, J., Cayuela, M., Paya, P., Martinez-Cacha, A., millennium: Novel approaches to the ancient art of Ca´mara, M. A., & Barba, A. (2007). Influence of fungicides winemaking. Yeast, 16, 675e729. on grape yeast content and its evolution in the fermenta- Pretorius, I. S. (2003). The genetic analysis and tailoring of tion. Commun. Agric. Appl. Biol. Sci., 72, 181e189. wine yeasts. In J. H. de Winde (Ed.), Functional genetics of Ough, C. S. (1992). Winemaking basics. Binghamton, NY: industrial yeasts (pp. 99e142). Berlin, Germany: Springer. Haworth Press Inc. Pretorius, I. S., & Bauer, F. F. (2002). Meeting the consumer Ough, C. S., & Amerine, M. A. (1988). Nitrogen compounds. challenge through genetically customized wine-yeast Methods for analysis of must and wines (2nd ed.), Univer- strains. Trends Biotechnol., 20, 426e432. sity of California, CA: John Wiley & Sons. Puig, S., & Pe´rez-Ortıń, J. E. (2000a). Expression levels and Paltauf, F., Kohlwein, S. P., & Henry, S. A. (1992). Regulation patterns of glycolytic yeast genes during wine fermen- and compartimentalization of lipid synthesis in yeast. In tation. System. Appl. Microbiol., 23, 300e303. E. W. Jones, J. R. Pringle, & J. R. Broach (Eds.), The Puig, S., & Pe´rez-Ortıń, J. E. (2000b). Stress response and molecular biology of the yeast Saccharomyces: Metabolism and expression patterns in wine fermentations of yeast genes gene expression (pp. 415e500). Cold Spring Harbor, NY: induced at the diauxic shift. Yeast, 16, 139e148. Cold Spring Harbor Laboratory. Puig, S., Querol, A., Barrio, E., & Pe´rez-Ortıń, J. E. (2000). Pampuhla, M. E., & Loereiro-Dias, C. (1989). Combined effect Mitotic recombination and genetic changes in Saccha- of acetic acid, pH, and ethanol on intracellular pH of fer- romyces cerevisiae during wine fermentation. Appl. menting yeast. Appl. Microbiol. Biotech., 31, 547e550. Environ. Microbiol., 66, 2057e2061. Panozzo, C., Nawara, M., Suski, C., Kucharczyka, R., Puig, S., Querol, A., Ramoń, D., & Pe´rez-Ortıń, J. E. (1996). Skoneczny, M., Becam, A. M., et al. (2002). Aerobic and Evaluation of the use of phase-specific gene promoters þ anaerobic NAD metabolism in. Saccharomyces cerevisiae. for the expression of enological enzymes in an industrial FEBS Lett., 517,97e102. wine yeast strain. Biotechnol. Lett., 18, 887e892. Peddie, H. A. B. (1990). Ester formation in brewery Querol, A., Barrio, E., Huerta, T., & Ramoń, D. (1992). fermentations. J. Inst. Brew., 96, 327e331. Molecular monitoring of wine fermentations conducted Pe´rez-Ortıń, J. E., Garcıá-Martıńez, J., & Alberola, T. M. by active dry yeast strains. Appl. Environ. Microbiol., 58, (2002). DNA chips for yeast biotechnology. The case of 2948e2953. wine yeasts. J. Biotechnol., 98, 227e241. Rachidi, N., Barre, P., & Blondin, B. (1999). Multiple Ty- Pe´rez-Torrado, R., Bruno-Barcena, J. M., & Matallana, E. mediated chromosomal translocations lead to karyotype (2005). Monitoring stress related genes during the changes in a wine strain of. Saccharomyces cerevisiae. Mol. process of biomass propagation of Saccharomyces cer- Gen. Genet., 261, 841e850. evisiae strains used for wine making. Appl. Environm. Rachidi, N., Barre, P., & Blondin, B. (2000). Examination of Microbiol., 71, 6831e6837. the transcriptional specificity of an enological yeast. A Pe´rez-Torrado, R., Carrasco, P., Aranda, A., Gimeno- pilot experiment on the chromosome-III right arm. Curr. Alcan˜iz, J., Pe´rez-Ortıń, J. E., Matallana, E., et al. (2002a). Genet., 37,1e11. 30 1. SACCHAROMYCES YEASTS I: PRIMARY FERMENTATION

Ramey, D. D., & Ough, C. S. (1980). Volatile ester hydrolysis Saccharomyces cerevisiae strain. J. Ferment. Bioeng., 83, on formation during storage of model solutions and 253e260. wines. J. Agric. Food Chem., 28, 928e934. Santangelo, G. M. (2006). Glucose signaling in Saccharo- Rapp, A., & Versini, G. (1991). Influence of nitrogen myces cerevisiae. Microbiol. Mol. Biol. Rev., 70, 253e282. compounds in grapes on aroma compounds of wine. In Schu¨ tz, M., & Gafner, J. (1993). Analysis of yeast diversity J. M. Rantz (Ed.), Proceedings of the International Symposium during spontaneous and induced alcoholic fermenta- on Nitrogen in Grapes and Wines (pp. 156e164). Davis, CA: tions. J. Appl. Bacteriol., 75, 551e558. American Society for Enology and Viticulture. Sillje´, H. H. W., Paalman, J. W. G., Schure, E. G., Ratledge, C., & Evans, C. T. (1989). In A. H. Rose, & Olsthoorn, S. Q. B., Verkleij, A. J., Boonstra, J., et al. J. S. Harrison (Eds.), The yeasts. Lipids and their metabo- (1999). Function of trehalose and glycogen in cell cycle lism, Vol. 3 (pp. 367e455). London, UK: Academic Press. progression and cell viability in Saccharomyces cer- Rauhut, D. (1993). Yeast-production of sulfur compounds. evisiae. J. Bacteriol., 181, 396e400. In G. H. Fleet (Ed.), Wine microbiology and biotechnology Singleton, V. L., & Esau, P. (1969). Phenolic substances in (pp. 183e223). Chur, Switzerland: Harwood Academic grapes and wine and their significance. Adv. Food Res., 1, Publishers. 1e282. Rauhut, D., & Ku¨ rbel, H. (1996). Identification of wine Snow, R. (1983). Genetic improvement of wine yeast. In aroma defects caused by sulfur-containing metabolites J. F. T. Spencer, D. M. Spencer, & A. R. W. Smith (Eds.), Yeast of yeast. In A. Lonvaud-Funel (Ed.), Enologie 95, 5e genetics. Fundamental and applied aspects (pp. 439e459). Symposium International d’Oenologie (pp. 515e519). Lon- New York, NY: Springer-Verlag. don: Technique & Documentation. Sprague, G. F. (1995). In A. H. Rose, A. E. Wheals, & Ribe´reau-Gayon, P., Dubourdieu, D., Done`che, B., & J. S. Harrison (Eds.), The yeasts. London, UK: Academic Lonvaud, A. (2006). Handbook of enology, Vol. 1. The Press. microbiology of wine and vinifications (2nd ed.), Chichester, Stanley, G. A., Douglas, N. G., Every, E. J., Tzanatos, T., & UK: John Wiley & Sons. Pamment, N. B. (1993). Inhibition and stimulation of Riou, C., Nicaud, J. M., Barre, P., & Gaillardin, C. (1997). yeast growth by acetaldehyde. Biotechnol. Lett., 15, Stationary-phase gene expression in Saccharomyces cer- 1199e1204. evisiae during wine fermentation. Yeast, 13, 903e915. Steger, C. L. C., & Lambrechts, M. G. (2000). The selection of Roon, R. J., Levy, J. S., & Larimore, F. (1977). Negative yeast strains for the production of premium quality interactions between amino acid and methylamine/ South African brandy base products. J. Ind. Microbiol. ammonia transport systems of Saccharomyces cer- Biotech., 24,1e11. evisiae. J. Biol. Chem., 252, 3599e3604. Stratford, M., & Rose, A. H. (1986). Transport of sulphur Rosa, M. F., & Sa´-Correia, I. (1991). In vivo activation by dioxide by Saccharomyces cerevisiae. J. Gen. Microbiol., ethanol of plasma membrane ATPase of Saccharomyces 132,1e6. cerevisiaie. Appl. Environ. Microbiol., 57, 830e835. Strathern, J. N., Jones, E. W., & Broach, J. R. (Eds.), (1982). Rose, A. H., & Harrison, J. S. (Eds.), (1991). The yeasts The molecular biology of the yeast Saccharomyces cerevisiae. (2nd ed.), London, UK: Academic Press. Cold Spring Harbor, NY: Cold Spring Harbor Rossignol, T., Dulau, L., Julien, A., & Blondin, B. (2003). Laboratory. Genome-wide monitoring of wine yeast gene expression Sumrada, R., & Cooper, T. G. (1982). Isolation of the CAR1 during alcoholic fermentation. Yeast, 20, 1369e1385. gene from Saccharomyces cerevisiae and analysis of its Rossignol, T., Postaire, O., Storaı¨, J., & Blondin, B. (2006). expression. Mol. Cell Biol., 2, 1514e1523. Analysis of the genomic response of a wine yeast to Sun, G. Y., & Sun, A. Y. (1985). Ethanol and membrane rehydration and inoculation. Appl. Microbiol. Biotechnol., lipids. Alcohol. Clin. Exp. Res., 9, 164e180. 71, 699e712. Thomas, K. C., Hynes, S. H., & Ingledew, W. M. (1994). Effects Sablayrolles, J. M., Dubois, C., Manginot, C., Roustan, J. L., of particulate materials and osmoprotectants on very high & Barre, P. (1996). Effectiveness of combined ammoni- gravity ethanolic fermentation by Saccharomyces cer- acal nitrogen and oxygen additions for completion of evisiae. Appl. Environ. Microbiol., 60, 1519e1524. sluggish and stuck fermentations. J. Ferm. Bioeng., 82, Thornton, R. J., & Eschenbruch, R. (1976). Homothalism in 377e381. wine yeasts. Anton. Leeuw., 42, 503e509. Sadraud, P., & Chauvet, S. (1985). Activite antilevure de Torija, M. J., Beltran, G., Novo, M., Poblet, M., l’anhydride sulfureux moleculaire. Connais. Vigne Vin, Guillamoń, J. M., Mas, A., et al. (2003). Effects of 19,31e40. fermentation temperature and Saccharomyces species on Salmon, J. M. (1997). Enological fermentation kinetics of an the cell fatty acid composition and presence of volatile isogenic ploidy series derived from an industrial compounds in wine. Int. J. Food Microbiol., 85, 127e136. REFERENCES 31

Vilanova, M., Ugliano, M., Varela, M., Siebert, T., & Werner-Washburne, M., Braun, E. L., Crawford, M. E., & Pretorius, I. S. (2007). Assimilable nitrogen utilisation and Peck, V. M. (1996). Stationary phase in Saccharomyces production of volatile and non-volatile compounds in cerevisiae. Mol. Microbiol., 19, 1159e1166. chemically defined medium by Saccharomyces cerevisiae Wiemken, A., & Durr, M. (1974). Characterization of amino wine yeasts. Appl. Microbiol. Biotechnol., 77, 145e157. acid pools in the vacuolar compartment of Saccharo- Visser, W., Scheffers, W. A., Batenburg-van der Vegte, W. H., myces cerevisiae. Arch. Microbiol., 101,45e57. & van Dijken, J. P. (1990). Oxygen requirements of Zaworski, P. G., & Heimsch, R. C. (1987). The isolation and yeasts. Appl. Environ. Microbiol., 56, 3785e3792. characterization of flocculent yeast. In G. G. Hiebsch (Ed.), Vuorio, O. E., Kalkkinen, N., & Londesborough, L. (1993). Biological research on industrial yeast III (pp. 185e195). Cloning of two related genes encoding the 56-kDa and Boca Ratoń, FL: CRC Press Inc. 123-kDa subunits of trehalose synthase from the yeast Zuzuarregui, A., Carrasco, P., Palacios, A., Julien, A., & del Saccharomyces cerevisiae. Eur. J. Biochem., 216, 849e861. Olmo, M. (2005). Analysis of the expression of some Wang, X. D., Bohlscheid, J. C., & Edwards, C. G. (2003). stress genes induced by stress in several commercial Fermentative activity and production of volatile wine yeast strains at the beginning of vinification. compounds by Saccharomyces grown in synthetic grape J. Appl. Microbiol, 98, 299e307. juice media deficient in assimilable nitrogen and/or Zuzuarregui, A., & del Olmo, M. (2004a). Analyses of stress pantothenic acid. J. Appl. Microbiol., 94, 349e359. resistance under laboratory conditions constitute a suit- Waterhouse, A. L. (2002). Wine phenolics. Ann. N.Y. Acad. able criterion for wine yeast selection. Anton. Leeuw., 85, Sci., 957,21e36. 271e280. Watson, K. (1987). Temperature relations. In A. H. Rose, & Zuzuarregui, A., & del Olmo, M. (2004b). Expression of J. S. Harrison (Eds.) (2nd ed.), The yeasts, Vol. 2 (pp. 41e47). stress-response genes in wine strains with different London, UK: Academic Press. fermentative behavior. FEMS Yeast Res., 4, 699e710. CHAPTER 2

Saccharomyces Yeasts II: Secondary Fermentation Alfonso V. Carrascosa 1, Adolfo Martinez-Rodriguez 1, Eduardo Cebollero 2, Ramoń Gonza´lez 3 1 Instituto de Investigacioń en Ciencias de la Alimentacioń (CIAL, CSIC-UAM), Madrid, Spain, 2 Instituto de Fermentaciones Industriales (CSIC), Madrid, Spain and 3 Instituto de Ciencias de la Vid y del Vino (CSIC-UR-CAR), Logronõ, Spain

OUTLINE

1. Sparkling Wines: Technology and 4.2. Morphological Changes in Yeast Cells Legislation 34 During Aging 42 1.1. Sparkling Wines: Description 4.3. The Genetics of Autolysis: Autophagy 43 and Classiﬁcation 34 5. Inﬂuence of Aging on the Quality of 1.2. Traditional-method Sparkling Wines 35 Traditional-Method Sparkling Wines 45 2. Production of Sparkling Wine Using the 6. Methods to Accelerate Yeast Autolysis in Traditional Method 35 Sparkling Wines and Implications for the 2.1. Primary Fermentation 36 Production Process 46 2.2. Prise de Mousse 37 6.1. Increased Temperature and Addition of 3. Secondary Fermentation 38 Autolysates 46 6.2. Genetic Improvements in Yeast 46 4. Aging 40 Acknowledgment 47 4.1. Biochemical Changes During Aging 40

1. SPARKLING WINES: and further analysis of the yeast strains consid- TECHNOLOGY AND LEGISLATION ered in the current chapter. Chapter 5 discusses the molecular characterization of wine yeasts As seen in the previous chapter, Saccharomyces and several of the points covered in Chapters yeasts are the main microorganisms responsible 1 and 6 will contribute to a better understanding for alcoholic fermentation in winemaking. Be- of the principles of proteomics and genomics. cause this fermentation takes place in open-top tanks, the carbon dioxide generated is spontane- 1.1. Sparkling Wines: Description ously released into the atmosphere. This release and Classification of carbon dioxide occurs during the production of most wines, generally referred to as still wines Traditional-method sparkling wines, which because of the small amounts of carbon dioxide are made using particular varieties of grape, they contain. Wines that contain more carbon contain carbon dioxide gas as a natural conse- dioxide are known as effervescent wines and quence of the process used in their production. include semi-sparkling and sparkling varieties. This gas is a byproduct of the secondary fermen- In natural semi-sparkling wines, the carbon tation of natural or added sugars in the base dioxide that forms during alcoholic fermentation wine. The fermentation takes place in closed becomes trapped in the wine after bottling. vessels and the resulting wine has a minimum When served, these wines produce bubbles that pressure of 4 atm at 20C. do not form a consistent or lasting mousse, Sparkling wines, which are served as an aper- unlike those of sparkling wines. Examples of itif or used to accompany meals or desserts, semi-sparkling wines are the French Blanquette have varying sugar content and are generally Me´thode Ancestral and Blanquette de Limoux acidic and white, although there are some rose´ (http://www.limoux-aoc.com) wines and cer- and a very small number of red varieties. tain Asti spumante wines from the Italian Pied- Depending on the winemaking method used, mont region (de Rosa, 1987). These wines are sparkling wines are classified as tank-fermented microbiologically stabilized prior to fermenta- wines (produced using large metallic tanks tion via cold processing, centrifugation, and/or using the Charmat or the continuous method) pasteurization. When served, sparkling wines or bottle-fermented wines (produced in the produce a persistent mousse and then gradually bottle using the transfer or the traditional release bubbles; the production costs of these [Champenoise] method) (Flanzy, 2000). wines are higher than those of semi-sparkling In the tank-fermented method, secondary wines. This chapter will look at the different fermentation takes place under isobaric condi- chemical and microbiological phenomena that tions in a sealed tank with a capacity of tens of take place during the production of sparkling hectoliters. These tanks are equipped with stir- wines and analyze their impact on sensory ring mechanisms that mix the yeast uniformly quality. It will also provide a detailed explana- into the base wine. The minimum time a wine tion of events such as secondary fermentation should remain in contact with the yeast before and aging, which is when yeast autophagy and it can be sold is 21 d (Pozo-Bayoń et al., 2003a). autolysis occur. Sparkling wine produced using this method is Most of the general microbiological aspects of bottled after clarification but cannot be labeled Saccharomyces yeasts have already been dis- either Champagne or Cava, for example. In cussed in Chapter 1 and many of the concepts certain cases, the wine is pasteurized at temper- are relevant, at least partly, to the understanding atures of between 33 and 70C for 2 to 5 d to PRODUCTION OF SPARKLING WINE USING THE TRADITIONAL METHOD 35 induce yeast autolysis and improve the sensory 1.2. Traditional-method Sparkling quality of the final product. The method is attrib- Wines uted to Euge`ne Charmat, who, in 1916, designed a system for producing large quantities of spar- In traditional-method sparkling wines, kling wine. It is both simpler and cheaper than secondary fermentation and the subsequent the traditional method and is used to produce aging process both take place in the bottle that low-cost sparkling wines. It is also suited to eventually reaches the consumer. It is an expen- making wines from certain aromatic varieties of sive, delicate procedure that is used to make grape such as Muscat in which aging with yeast high-quality, relatively expensive wines. In would mask the characteristic aromas of these France, the wines in this group are known as grapes and detract from the wine’s sensory Champagne and they are produced using base quality. The Charmat method is used to produce wines made with white Chardonnay grapes sparkling wines from Asti and Trento in Italy. and red Pinot Noir and Pinot Meunier grapes. In the continuous method, large tanks are The minimum aging period is 12 months. Italian used to reproduce the yeast autolysis that takes traditional-method sparkling wines are known place in bottles in the traditional method as Talento and they are made with white Char- (Flanzy, 2000). The process is conducted under donnay or Pinot Bianco grapes or with red Pinot isobaric conditions using a base wine to which Nero or Pinot Meunier grapes. The minimum 50 to 72 g/L of sugar is added. This is then aging period is 15 months and the wines must pasteurized at 70C to accelerate sucrose hydro- be produced in the regions of Trento, Piamonte, lysis. The juice is then cooled, filtered, and inoc- Lombardıá, el Trentino, el Alto Adige, Veneto, or ulated with yeast (w106 cells/mL). Secondary Friuli. fermentation now takes place and steps are In Spain, most sparkling wines produced us- taken to reduce the yeast population and induce ing the traditional method are known as Cava. cell death and autolysis. These wines must age on lees for at least 9 months. In the transfer method, the sparkling wine is The authorized grape varieties are Macabeo, Xar- produced in bottles, which are generally el$lo, Parellada, Subirat (Malvasıá riojana), and magnums measuring 1.5 or 2 L to minimize Chardonnay (white) and Garnacha tinta and storage space requirements. It is then left to age Monastrell (red). Rose´ Cava can be made using on lees for at least 2 months, after which it is Pinot Noir or Trepat. The vast majority of Cava transferred to a tank maintained under isobaric is produced in the Catalan region of El Penede`s, conditions with carbon dioxide or nitrogen to with only around 1% of the total production prevent loss of the gas. The wine is then cold pro- coming from outside this region. cessed at À5C, filtered, and sometimes transferred to a second tank, where the dosage, also known by the French term liqueur d’expedition, 2. PRODUCTION OF SPARKLING is added before rebottling. Another method WINE USING THE TRADITIONAL involves filtering the wine in the tank after add- METHOD ing the dosage and before bottling. With this technique, disgorging is not required and certain The production of traditional-method spar- advantage is taken of natural yeast autolysis, kling wines involves two main stages: primary thus helping to keep production costs down. fermentation and secondary fermentation. The The label of these wines must state that the former converts the must into base wine and wine has been naturally fermented in a bottle. the latter creates the final product. Secondary 36 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION fermentation is also known by the French term sometimes difficult to complete primary fermen- prise de mousse as it generates the carbon dioxide tation in these musts (due to extensive clarifica- that forms the frothy mousse in the glass as the tion, low pH, and the absence of grape skins), wine is served. selected yeasts, generally Saccharomyces cerevisiae, Toproduce the base wine, each variety of grape are often added in proportions of approximately is fermented separately. The grapes are hand- 106 cells/mL to ensure even fermentation and picked and any leaves or spoiled berries prevent the formation of byproducts that would removed. The grapes, still in bunches, are then adversely affect the organoleptic characteristics transported to the press in shallow containers to of the base wine (Bidan et al., 1986; Martıńez- protect the berries from bruising or damage and Rodrı´guez et al., 2001a). These yeasts are commer- thus prevent the onset of premature fermentation. cialized as active dry yeast. Furthermore, by To obtain the must, the grapes are pressed in using specially selected, pure inoculums, wine- several stages, with only the first press fractions makers can produce wines with distinctive char- used to make high-quality wines. Gentle crush- acteristics using similar fermentation processes ing of the grapes ensures that the juice is from one year to the next, thus circumventing extracted from the pulp of the berries without the random effects of spontaneous fermentation. breaking the seeds or extracting compounds The maintenance of a low, stable temperature from the skin or stems that would increase sedi- prevents the must from fermenting in an uncon- ment and colorant matter and add harsh vegetal trolled manner and thus protects against the aromas and flavors to the wine (Flanzy, 2000). loss of desirable aromas or the development of The light pressure applied means the extraction undesirable ones. is incomplete. The must should be transferred to The base wine is racked off the sediment the tanks as soon as possible after the grapes (solid particles that have settled at the bottom reach the press in order to minimize oxidation of the tank) and lees (yeast and adhered and thus safeguard against the development of bentonite particles) and the sulfite level cor- flat aromas and browning. rected. The level of free sulfites must be kept The next stage involves addition of sulfites to at under 15 mg/L to ensure normal yeast the must. Sulfur dioxide is an important addi- growth during secondary fermentation. tive in wine because of its antioxidant and anti- The next stage is the assembling or coupage microbial properties. Indeed, it plays a key role stage. This consists of blending base wines in determining which microorganisms partici- made from different varieties of grape or iden- pate in the initial phases of fermentation. The tical varieties from different years, or sometimes must is then clarified in large tanks at low even single varieties, in proportions that vary temperatures. During this process, known as depending on the quality of the harvest. static clarification, the forces of gravity cause The final step is tartrate stabilization, the aim the solid particles suspended in the must to of which is to prevent both the precipitation of settle at the bottom of the tank. The liquid is potassium bitartrate in the bottle (as a result of then racked off the sediment and thus cleared the low storage temperatures and an increase of impurities. in ethanol levels during secondary fermentation). One of the most common tartrate stabiliza- 2.1. Primary Fermentation tion methods is to induce the formation of potassium bitartrate crystals by reducing the Primary fermentation to produce the base temperature of the base wine to À4C. After wine takes place in tanks at a controlled tempera- several racking and filtration steps, all such ture of between 15 and 18C. Because it is crystals are removed from the base wine. In PRODUCTION OF SPARKLING WINE USING THE TRADITIONAL METHOD 37

TABLE 2.1 Optimal Characteristics of Base Wine Used levels that occurs during aging (when the wine to Make Cava is left in contact with the yeast) may be due not only to precipitation caused by the increase e Alcohol content 9.5 11.5 in alcohol levels but also to adsorption to the Total minimum acidity (tartaric acid) 5.5 g/L bentonite contained in the liqueur de tirage Nonreducing extract 13e22 g/L (Luguera et al., 1997; Martinez-Rodriguez & Polo, 2003). Maximum volatile acidity (acetic acid) <0.60 g/L The bottles are then stacked horizontally in Total sulfur dioxide <140 mg/L special aging rooms. Secondary fermentation, Ash 0.70e2 g/L prise de mousse, aging, and yeast autolysis all occur when the bottle is in this position. pH 2.8e3.3 When aging is complete, the wines are riddled. This consists of gently shaking the addition to appropriate organoleptic character- bottles to direct the sediment (lees) formed by istics, the final base wine must have analytical the yeast, bentonite, and any adhered sub- characteristics similar to those shown in Table stances towards the neck of the bottle. This 2.1. As the majority of sparkling wines are used to be done manually by rotating the bottles white, the rest of this chapter will refer to the one eighth of a turn every day for 15 d until they production of sparkling wines made from white were practically perpendicular to the floor, but base wines. nowadays it is performed by more or less automated systems that can rotate large numbers of 2.2. Prise de Mousse bottles simultaneously. The next stage, known as disgorging, consists Once the base wine has been produced, the of removing the lees that have settled at the neck next stage is the prise de mousse, during which of the bottle. The lees are frozen by placing the secondary fermentation, yeast autolysis, and neck of the bottle in a bath of freezing solution. possibly malolactic fermentation take place. The bottle is then placed in an upright position Sparkling wines must remain in the bottle for and the cork removed, and the ice plug contain- a minimum number of monthsdstipulated by ing the lees is expelled by the internal pressure national legislationdbefore they can be sold. in the bottle. Disgorging can be facilitated by The main operations that take place during using yeast immobilized on calcium alginate this period are tirage, stacking, riddling, dis- beads or enclosed in a special cartridge placed gorging, and dosage. in the neck of the bottle. Tirage consists of filling the bottle with the Some liquid may be lost during disgorging base wine and the liqueur de tirage. The base but this is compensated for with the addition of wine receives no further treatment once it has dosage. This liqueur may be pure sparkling been placed in the bottle. The liqueur de tirage wine, sparkling wine containing sucrose, grape is a suspension of yeast, sucrose (20e25 g/L), must, partially fermented grape must, grape and a small quantity of bentonite (3 g/100 L) must concentrate (which may or may not have to aid flocculation and the subsequent removal been rectified), base wine, or a combination of of yeast cells. The amount of bentonite used is all these. If necessary, wine distillates may also approximately 10 times less than that used in be added. The addition of dosage allows wine- treatments aimed at removing proteins from makers to give their sparkling wines a distinctive wine. Interestingly, several authors have shown finish. Finally, the bottle is sealed with its defini- that the reduction of protein and/or peptide tive cork, which is held in place with a muzzle. 38 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION

While in the bottle, the yeasts undergo or summarize in the following section using a prac- participate in a series of processes that are crit- tical example. ical to the quality of traditional-method spar- When dry active yeast is to be used, this kling wines. These are secondary fermentation, should be rehydrated for 20 min (500 g of yeast followed by autophagy and autolysis. in 5 L of water with 250 g of sucrose) at 35 to 40C. When an agar slant culture of a selected strain is to be used, the strain should be grown 3. SECONDARY FERMENTATION in sterilized must or complete medium with sucrose until a volume of 5 L and a population Secondary fermentation begins after tirage. It of 108e109 colony-forming units (CFU)/mL is starts with the inoculation of the base wine and achieved. The next stage is the conditioning ends when all fermentable sugars have been phase, in which the yeast adapts to the alcohol consumed. The yeasts used to make tradi- environment. In our example, 600 L of wine, tional-method sparkling wines must have 645 L of water, 120 kg of sucrose, 500 g of yeast certain, key, characteristics (Bidan et al., 1986). extract or 200 g of ammonium salts, 3 kg of tarta- Specifically, they should ric acid, and 5 L of active biomass are added to a 2000 L tank, which is kept at 20Cfor3to4d, Have high resistance to ethanol (10e12)as 1. until vigorous fermentation and a density of the base wines have an alcohol content of 1.000e1.002 kg/cm3 is achieved. The next stage over 9.5, which increases during secondary is the propagation phase. During this step, fermentation; 12.5 kg of sucrose and 500 L of base wine are Display fermentation activity at low 2. added to the mixture in the tank, which is kept temperatures as, on occasion, the at a temperature of 20C until a density of temperatures in cellars can be lower than 0.994e0.998 kg/cm3 is reached (approximately 12C; 24 h). This culture can now be added to the Be resistant to pressure caused by carbon 3. base wine with 20 to 25 g/L of sugar and a fining dioxide; agent in a proportion of between 8 and 10%. This Be able to flocculate as this facilitates the 4. produces a concentration of 8e12 Â 106 CFU/mL subsequent elimination of lees during and a sufficient amount with which to prepare disgorging and prevents yeast deposits from between 20 000 and 26 000 bottles; proportional adhering to the walls of the bottle (which is amounts can be used for other volumes. This the bottle in which the wine is sold); and operation is known as tirage. Not produce unpleasant aromas as the aroma 5. Following the inoculation of the base wine, of sparkling wines is influenced not only by there is a short lag period in which the yeast the grapes used but also by the metabolism of adapts to the new substrate conditions. The the yeast during primary and secondary growth pattern of in secondary fermentation and on-lees aging. Saccharomyces fermentation is similar to that in primary fermen- It was recently suggested that measurement of tation, although growth is generally slower as the autolytic and foaming capacity of yeast there are considerably fewer sources of carbon grown in synthetic media might be a valuable and nitrogen available. Other nutrients can tool in the selection of strains for secondary become limiting as fermentation progresses, fermentation (Martıńez-Rodrı´guez et al., 2001a). and increasing levels of ethanol and carbon Once a strain has been selected, the starter dioxide build-up can also restrict growth. Our culture used in industrial processes is prepared group found that a starter inoculum of approxi- following a general procedure that we will mately 106 cells/mL produced a population of SECONDARY FERMENTATION 39 close to 107 CFU/mL on termination of sec- Gonza´lez et al., 2008; Martıńez-Rodrı´guez ondary fermentation (Martıńez-Rodrı´guez et al., et al., 2001b; Martıńez-Rodrı´guez et al., 2004). 2002). Secondary fermentation tended to take In those studies, we compared viable cell place in the first 15 to 20 d after tirage. After counts and total microscopic cell count as we this, cell viability decreased slowly until it was had observed an increasing concentration of no longer detectable (between days 60 and 90). dead yeast cells, ranging from 105 cells/mL in Similar results have been reported by other the first week of incubation to 106 cells/mL groups (Feuillat & Charpentier, 1982). after day 20 of fermentation. Our analysis of Our group has analyzed morphological the morphological characteristics of these aspects of wine yeast viability and autolysis, cultures revealed the simultaneous presence of including the presence/absence of budding, dead and live cells, indicating an overlap vacuole size, cell size, and cytoplasm separa- between secondary fermentation and autolysis tion from the cell wall (Gonza´lez et al., 2003; (see Figure 2.1). The addition of bentonite

Exponential phase FIGURE 2.1 Saccharomyces cerevisiae EC1118 cells during exponential growth in synthetic medium and during secondary fermentation of base wine in the bottle at 11 and 16C. Note the smaller cell size and presence of a granular cytoplasm typical of dead cells in the images of yeast cells fermented at 16C. Images taken using Nomarski interference contrast microscopy; the bar corresponds to 10 mm.

Secondary fermentation in the bottle 11ºC 16 ºC

Day 32

Day 90 40 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION does not appear to have a significant effect on by various authors, and, in recent years, variations in yeast concentrations (Martıńez- numerous publications have contributed greatly Rodrı´guez et al., 2002). to the understanding of this process and high- Numerous variables have a considerable lighted its importance in various types of vinifi- influence on the entire secondary fermentation cation and related processes. In the next section, process and on the viability of the yeasts that we will look at the most relevant aspects of participate in this process. Temperature is one autolysis from two perspectives: a biochemical of the most critical variables, as was demon- perspective (dynamics and generation of strated in tests performed by our group, in different compounds released by yeast) and which viability was seen to decrease dramati- a microbiological perspective (addressing the cally after day 20 of fermentation in the bottle. mechanisms of autolysis based on changes At day 90 of fermentation, there was no occurring in the yeast cell). evidence of viability at 16C, but at 11C levels were similar to those seen at the start of fermen- 4.1. Biochemical Changes During Aging tation. Morphological and microscopic indicators of viability were visible in cultures that Sparkling wines contain a wide variety of remained viable (Figure 2.1). organic compounds, including proteins, peptides, polysaccharides, monosaccharides, lipids, fatty acids, nucleic acids, and volatile 4. AGING components. Many of these compounds, or their precursors, can originate in either grapes or Traditional-method sparkling wines must be yeast. Nitrogenous compounds are the most left to age once secondary fermentation is abundant and as such have been studied in complete. The length of this aging period is regu- greatest detail. Indeed, they are considered by lated by national legislation and may vary from many authors to be the most important organic one country to the next. Champagne, for compounds in sparkling wines. The most abun- example, must be left to age for at least 11 months dant nitrogenous compounds in these types of before it can be sold, whereas the minimum time wines are peptides and amino acids. Peptide stipulated for Cava by Spanish legislation is levels rise at the start of fermentation (possibly 9 months. The aim in all cases, however, is the due to the release of peptides as dead cells start same: to establish a minimum time during which to appear) and during autolysis (as mentioned the wine must remain in contact with the lees to in the previous section) and begin to fall guarantee a quality final product. towards the end of fermentation. This decrease The most important biological process that has been attributed to the consumption of takes place during aging is yeast autolysis, peptides by yeasts (Becker et al., 1973) and to a phenomenon that causes the release of intra- the presence of active acid proteases in the cellular compounds into the wine. In the wine (Lagace & Bisson, 1990). Because the previous section, we saw how a study of yeast protocols for isolating, analyzing, and character- population dynamics indicated an overlap izing peptides tend to be more complicated than between secondary fermentation and autolysis. those used for other nitrogenous compounds The term “autolysis” was first used in the sci- such as proteins and amino acids, it was only entific literature by Salkowsky at the end discovered in recent years, which is when of the nineteenth century to refer to the self- most of the studies of these molecules were per- degradation of cellular constituents that started formed, that peptides account for the dominant after cell death. Autolysis has since been studied fraction in autolysis. AGING 41

In 1998, Moreno-Arribas et al. (1998), using released by yeasts either at the end of fermenta- high-performance liquid chromatography tion or during autolysis. The analysis of amino (HPLC), observed that Cava wines made with acids released by yeasts in both model systems different grape varieties nevertheless had and sparkling wines aged for varying lengths a similar peptide profile, leading them to suggest of time has shown that autolysis increases that the peptide composition at the end of amino acid levels by just a few milligrams per secondary fermentation was more closely linked liter. This may be because more peptides than to yeast activity than to the initial composition amino acids are released during this process of the must. Their hypothesis was confirmed in or because the amino acids released (primarily later studies, which demonstrated that the yeast glutamic acid, arginine, and alanine) are con- strain used in secondary fermentation played verted through decarboxylation and deamina- a determining role in the final peptide fraction tion, resulting in a reduction of the final amino of wine (Martıńez-Rodrı´guez & Polo, 2000). It acid fraction. Nevertheless, although there are was also seen that the addition of bentonite, contrasting results regarding the specific a common operation in the making of sparkling behavior of certain amino acids, almost all wines, also influenced this fraction, even though authors agree that the concentrations of most only very small quantities of bentonite were used amino acids decrease during secondary fermen- (Martıńez-Rodrı´guez & Polo, 2003). Further- tation and increase again during aging, and that more, the substance was seen to have an insignif- these amino acids act as important precursors of icant influence on cell viability. Although aromatic compounds (Charpentier & Feuillat, peptides are the predominant compounds in 1993). secondary fermentation, they are the best indica- Proteins levels, like peptide levels, increase at tors of the dynamics of autolysis as they are the start of fermentation (possibly because of the released either directly by yeasts or indirectly presence of dead cells) and during autolysis, from proteins and are simultaneously converted and then decrease. This decrease is also influ- into free amino acids by enzymatic activity. enced by the presence of bentonite (adsorption Protease A is the best studied of all the proteases of proteins) and the increase in precipitation that participate in this conversion and it is also that occurs as a result of increasing alcohol considered the most active as its optimum pH content (Dizy & Polo, 1996). There is a progres- (3e3.5) is the same as that of wine. sive decrease in protein content during aging The most abundant amino acids in base wine and autolysis as proteins are hydrolyzed into are proline, glutamic acid, lysine, leucine, argi- compounds with a lower molecular mass, nine, and aspartic acid. Indeed, with the excep- which explains why traditional-method spar- tion of lysine, these amino acids show the kling wines tend to have a lower protein fraction greatest decline during fermentation (Martıńez- than the base wines used to make them. Rodrı´guez et al., 2002). The amino acid fraction The main sugar component of the polysaccha- of base wine is very important, as amino acids rides in base wine is arabinose (66%), but this are the main source of nitrogen during fermen- composition changes radically after secondary tation. They also serve as precursors of aromatic fermentation, with mannose and glucose becom- compounds that contribute to the special char- ing the dominant components (43 and 31%, acteristics of sparkling wines. The amino acids respectively) (Nuń˜ez et al., 2005). This indicates in sparkling wines are derived from various that the polysaccharides present in sparkling sources. Some come from the grapes used to wines after aging are primarily the result of the make the base wines and are not metabolized degradation of the yeast cell wall that takes place by yeasts during growth, and others are during autolysis. These polysaccharides are 42 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION essentially glycoproteins with a sugar content of formed during this stage are alcohols, esters, approximately 85 to 90% and a protein content of fatty acids, and aldehydes. The tertiary aroma, just 10 to 15% (Nuń˜ez et al., 2006). Glucan- and or bouquet, evolves as the compounds gener- mannose-containing polysaccharides increase ated during fermentation are transformed by during fermentation and aging in a similar aging. In a study conducted by our group, fashion to nitrogenous compounds. Their levels Pozo-Bayoń et al. (2003b) confirmed that there may therefore remain constant or decrease grad- was a close relationship between aging time ually if aging is prolonged. This decrease is and the final volatile compound fraction of primarily due to the activity of b-(1,3) gluca- sparkling wines, and explained that these nases, which are released by yeasts and remain compounds undergo both degradation and active in the wine. synthesis during aging. The lipids and fatty acids released by yeast Other compounds, such as nucleic acids, can during secondary fermentation and autolysis also be found in traditional-method sparkling are very difficult to quantify as their levels are wines, albeit at very low levels (similar to those very low (~2e4 mm). Pueyo et al. (2000) devel- of lipids). In experiments performed using oped an analytical method using HPLC and model systems, Hernawan and Fleet (1995) a light scattering detector to separate and quan- found that approximately 90% of RNA and tify lipids by classes. They found that sterol 40% of cellular DNA was degraded during yeast esters were the most common type of lipid autolysis and was soluble in wine. released during secondary fermentation (8.6%), followed by sterols (3.8%) and triglycer- 4.2. Morphological Changes in Yeast ides (2%). As several authors have reported, Cells During Aging lipid levels may decrease after an initial increase during secondary fermentation and at the Yeast autolysis, which is the dominant microbi- beginning of aging because these compounds ological process in aging, takes place after cell participate in the formation of esters, ketones, death. As mentioned above, important morpho- and aldehydes (Charpentier & Feuillat, 1993). logical changes, which are visible by optical The volatile compounds responsible for the microscopy, occur in the yeast cell during aroma of sparkling wines come from rather secondary fermentation. These changes continue heterogeneous groups, including alcohols, alde- to occur during yeast autolysis throughout aging. hydes, ketones, esters, volatile acids, terpenes, Most of the changes affect the cell wall, which and pyrazines, with levels varying considerably accounts for between 15 and 25% of the dry (from picograms to milligrams per liter). These weight of the cell and is formed mostly by poly- compounds have various origins, and the saccharides (80 to 90%), which are hydrolyzed aromas they generate can be classified into three during autolysis. These cell wall changes are groups according to whether they are derived visible only under an electron microscope. from the grape, from fermentation, or from In a study conducted by our group, Martıńez- processes occurring during aging. The primary Rodrı´guez et al. (2001b) observed the develop- aroma is derived from substances in the grape ment of folds on the yeast cell wall caused by and is also known as the varietal aroma as it is loss of volume during autolysis (see Figure 2.2), specific to the grape variety used (Codornnier & a finding that coincides with previous reports Bayonove, 1982). The secondary aroma, also (Charpentier & Feuillat, 1993; Gonza´lez et al., known as the fermentation aroma, is generated 2008). by the metabolic activity of yeasts during Not many studies have analyzed the changes fermentation. The most important compounds that take place inside the cell during autolysis in AGING 43

(a)

(b)

FIGURE 2.2 Saccharomyces cerevisiae IFI-473 cells after 12 months of aging in the bottle. Note the wrinkles and folds on the wall. Images taken using low-temperature scanning electron microscopy. winemaking conditions, but several groups have indicated that the disorganization of intracellular structures, together with the resulting release of hydrolytic enzymes, is the key step in the autolysis of yeast during the aging of traditional-method sparkling wines (Connew, 1998; Fornairon-Bonnefond et al., 2002). Our group found morphological differences, FIGURE 2.3 Saccharomyces cerevisiae IFI-473 cells after visible by optical microscopy, on comparing 24 h of accelerated autolysis in wine medium (a) and after 9 months of aging in the bottle (b). Note the absence of cells that had undergone accelerated autolysis autophagosomes after accelerated autolysis (a) and the for a few hours and cells from wines that had presence of autophagosomes after 9 months of aging (b). been aged for 9 months (see Figure 2.3). The Images taken using Nomarski contrast interference micro- main difference was the presence of structures scopy; the bar corresponds to 10 mm. similar to the autophagosomes described in the cytoplasm of cells from bottle-aged wines. release of vacuolar enzymes; autophagy, in These structures will be studied in greater detail contrast, is an exquisitely organized and regu- in Section 4.3. lated process that occurs in response to the absence of essential nutrients and involves the 4.3. The Genetics of Autolysis: trafficking of membranes and intracellular Autophagy components. It is a catabolic process that has been conserved in all eukaryotic cells to degrade It has traditionally been accepted that autol- cytoplasmic material in the vacuole. In S. cerevi- ysis in winemaking involves the uncontrolled siae, thanks to autophagy, cells can survive for 44 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION long periods of time in the absence of essential (a) nutrients by using products formed during the degradation of cell constituents. The cytoplasm is transported to the vacuole through double- membrane vesicles known as autophagosomes, which, while being formed, sequester the surrounding cytoplasmic region in a nonspecific manner (Baba et al., 1994). Autophagosome formation can be induced in wine yeasts through prolonged nitrogen starvation (see Figure 2.4). The outer membrane of the autophagosome fuses with the vacuole, releasing a vesicle surrounded by a single membrane (autophagic body) into the lumen of the vacuole. This autophagic body is then digested by vacuolar enzymes (Takeshige et al., 1992). This digestion process can be experimentally interrupted by (b) adding phenylmethylsulfonyl fluoride (PMSF), a protease inhibitor. The result is a vacuole full of autophagic bodies (see Figure 2.5) that is used as a marker of active autophagy. Considerable advances have been made in our understanding of the molecular mechanisms underlying autophagy since the genes responsible for this process, AUT and APG, were first identified (Thumm et al., 1994; Tsukada & Ohsumi, 1993). (For a detailed review of the molecular aspects of autophagy, see Klionsky, 2005 and Nakatogawa et al., 2009.) Cloning of the genes required for autophagy revealed that most were involved in the cytosol-to-vacuole targeting (Cvt) pathway, a constitutive pathway for the trans- FIGURE 2.4 Saccharomyces cerevisiae IFI-473 cells during port of pro-aminopeptidase I to the vacuole exponential growth (a) and after prolonged nitrogen star- that is morphologically and molecularly very vation (b). Note the presence of budding and the absence of similar to the autophagy pathway (Harding autophagosomes in the cytoplasm during the exponential et al., 1996; Scott et al., 1996). To unify nomencla- phase (a) and the opposite after prolonged nitrogen starvation (b). Images taken using Nomarski contrast interfer- ture, all genes, APG, AUT, and Cvt were ence microscopy; the bar corresponds to 10 mm. renamed ATG (autophagy) genes (Klionsky et al., 2003). Sixteen ATG genes have been identified to date as playing a role in the formation of development of other vesicle transport path- autophagosomes. Other genes involved in ways (reviewed in Levine & Klionsky, 2004). membrane fusion events and the degradation Most of the Atg proteins involved in auto- of vesicles in the vacuole have also been identi- phagosome formation localize to a single region fied as necessary for autophagy and the normal adjacent to the vacuole. This region is called the INFLUENCE OF AGING ON THE QUALITY OF TRADITIONAL-METHOD SPARKLING WINES 45

ATG gene products generally participate in the induction and control of autophagy via different mechanisms including the formation and degradation of protein complexes, enzymatic activities such as kinase activities (protein kinases, phosphatidylinositol kinases), and covalent modification of proteins through mechanisms similar to ubiquitination. To determine whether or not autophagy could also take place in winemaking conditions, our group used a series of mutants that affect the Cvt and autophagy pathways, either simultaneously or independently, and analyzed the transport of aminopeptidase 1 (which can be transported by either pathway) to the vacuoles. We demonstrated that autophagy does indeed FIGURE 2.5 Induction of autophagy in Saccharomyces occur in conditions similar to those of secondary cerevisiae IFI-473 strain by prolonged nitrogen starvation. fermentation (Cebollero et al., 2005a). We later Note the nondegraded autophagosomes in the vacuole (white arrow) caused by the interruption of the process developed another strategy to analyze indus- using the protease inhibitor, phenylmethylsulfonyl fluoride. trial strains (not necessarily mutant) in real Images taken using Nomarski contrast interference winemaking conditions and concluded that, in microscopy; the bar corresponds to 5 mm. sparkling wine production conditions, autolysis must be preceded by autophagy and the digestion of intracellular material in the vacuole has pre-autophagosomal structure (PAS) and is a clear influence on the nature and abundance thought to be a possible center for the formation of compounds released by yeasts into the wine of these vesicles (Kim et al., 2002; Suzuki et al., during autolysis (Cebollero & Gonza´lez, 2006). 2001). Analysis of the functional relationships between Atg proteins in the PAS has shed light on the cellular mechanism involved in 5. INFLUENCE OF AGING ON THE autophagosome synthesis (Reggiori et al., 2004; QUALITY OF TRADITIONAL- Suzuki et al., 2004). METHOD SPARKLING WINES Because the Cvt pathway is a constitutive pathway and, as mentioned, shares most of its There is no question that yeast compounds elements with the autophagy pathway, the released in wine during autolysis can consider- induction of autophagy in starvation condi- ably modify both the chemical composition and tions may imply reorientation of a functional the sensory properties of the resulting wine. transport system. The main differences bet- The interaction between these two aspects has ween the two pathways are vesicle size and been studied by various authors, who have cargo selectivity. Atg1 might play a key role in shown not only that these compounds have this transition by transducing the signal from a direct influence on the final quality of the other pathways such as the targets of rapamy- wine but also that they can act as intermediaries cin (TOR) pathway in response to starvation in the formation of other substances that conditions (Kamada et al., 2000; Scott et al., contribute to sensory quality. Mannoproteins, 2000). for example, which form part of the yeast cell 46 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION wall, are released during autolysis and play a key Consequently, increasing temperature was one role in the creation of a wine with small, lasting of the first strategies for accelerating autolysis bubbles, two of the most desirable characteristics to be explored. Although the activity of en- of sparkling wines (Brissonnet & Maujean, 1991; zymes involved in autolysis is known to in- Nu´ n˜ez et al., 2005). Fatty acids, although found in crease with temperature (Fornairon-Bonnefond low concentrations, are associated with the et al., 2002), increasing the storage temperature formation of esters, ketones, and aldehydes, all did not produce satisfactory results, as the fla- substances that have a very low sensory vor of the resulting wines was reported to be threshold, meaning that fatty acid concentrations excessively yeasty or toasty. Another technique can affect the flavor of wine (Charpentier & Feuil- that has been explored involves the addition of lat, 1993). The situation is similar for nitrogenous yeast extracts to the base wine together with compounds. Various aspects related to the the liquer de tirage (containing sucrose and yeasts sensory properties of sparkling wine and the to trigger secondary fermentation). The result, quality of the mousse, in particular, have been however, was not satisfactory either, as the associated with the nitrogen fraction. It is known, excessive proteolysis of the autolysates also for example, that amino acids are precursors of resulted in undesirable flavors (e.g., toasty) aromatic compounds (Feuillat & Charpentier, and aromas (Peppler, 1982). To resolve this 1982) and that the surfactant and sensory proper- problem, a specific yeast autolysate preparation ties of proteins and peptides can influence the procedure was developed. The result was organoleptic properties of sparkling wines (Mar- a lesser degradation of autolysates and, accord- tıńez-Rodrı´guez & Polo, 2003). In summary, in ing to some authors, accelerated aging and im- view of the importance of these aspects, it is clear proved aroma and bubble quality (Charpentier why traditional-method sparkling wines are & Feuillat, 1993). superior to those made using methods in which Another technique that has shown promising yeast autolysis has a lesser impact on the final results in the laboratory is the use of mixed product. Because autolysis in the bottle is cultures of killer and sensitive strains of S. a slow process requiring long aging times and cerevisiae. Tests performed in synthetic media considerable storage costs, being able to accel- have shown that these mixed cultures had a 20 erate this process or achieve similar effects with to 30% greater protein content than control the use of additives would represent a consider- cultures after 3 d of aging (Todd et al., 2000). able improvement for makers of sparkling wines. The next section summarizes a range of strategies 6.2. Genetic Improvements in Yeast that have been explored with this goal in mind. Because autolysis in wine is a lengthy process, which can last for years in some cases, it would 6. METHODS TO ACCELERATE be highly advantageous from a practical YEAST AUTOLYSIS IN SPARKLING perspective to find yeast strains capable of autol- WINES AND IMPLICATIONS FOR ysis in a shorter time. One way of achieving this THE PRODUCTION PROCESS would be through genetic improvements designed to create yeast strains with an acceler- 6.1. Increased Temperature and ated autolytic capacity. These yeasts would Addition of Autolysates help to accelerate autolysis without causing the problems associated with temperature increases The temperature at which wine is aged is one and the addition of yeast autolysates. Such of the main rate-limiting factors for autolysis. genetic improvements might indeed enhance REFERENCES 47 the final quality of the wine without the need to alter the production process. Genetic modification strategies include random mutagenesis and genetic engineering. While genetic engineering offers significant advantages, it also has several drawbacks. For example, it has limitations with respect to the alteration of complex genetic traits such as autolysis, it is subject to very strict legislative requirements, and it is currently negatively viewed by consumers (see Chapter 7). One of the advantages of random mutagenesis is that the system is relatively simple and an extensive knowledge of the targeted metabolic pathway is not required to create mutants. It is therefore likely to be much more widely accepted than FIGURE 2.6 Temperature-sensitive autolytic mutant of genetic engineering at present and would also Saccharomyces cerevisiae IFI-473 strain in which autolysis was induced by incubation at 37C for 24 h. Note the absence of offer greater commercial potential. In a study cytoplasmic content in a large number of cells. Images taken performed by our group involving the creation using Nomarski contrast interference microscopy; the bar of autolytic S. cerevisiae mutants by ultraviolet corresponds to 5 mm. mutagenesis, the most promising mutant rapid release of nitrogenous compounds into released greater amounts of nitrogenous the external environment; similar effects were compounds and amino acids in a model wine observed in both laboratory strains and indus- system at low temperatures, making it a poten- trial strains used in secondary fermentation tial candidate for use in secondary fermentation (Cebollero et al., 2005b; Cebollero et al., 2009). during sparkling wine production (Gonza´lez Strains in which autophagy was inhibited also et al., 2003). The mutants with this capacity showed an accelerated loss of viability, but this had a high level of cellular disorganization was associated with a more rapid release of when viewed under an optical microscope (see intracellular material only in cases of alterations, Figure 2.6). More recently, Nu´ n˜ez et al. (2005) with pleiotropic effects, in genes that both influ- showed that one of the mutants created in the ence autophagy induction and participate in above study exhibited accelerated autolysis many cellular processes related to starvation during the production of traditional-method response (Tabera et al., 2006). sparkling wines and gave rise to quality wines, despite an aging time of just 6 months. Following confirmation that autophagy also Acknowledgments occurs during secondary fermentation, as was Work at the author’s laboratories is funded by the Spanish strongly suggested by our microscopic findings Ministry for Science and Innovation (grants AGL2006- (see Figure 2.3), our group designed genetic 02558, AGL2009-07327, AGL2009-07894 and Consolider engineering strategies aimed at accelerating or INGENIO2010 CSO2007-00063) as well as Comunidad de Madrid (CAM) (grant S2009/AGR-1469). inhibiting the autophagy process. In the first case, using a gain-of-function allele of the CSC1 gene called CSC1-1, we showed that overexpres- References sion of this gene was associated with accelerated Baba, M., Takeshige, K., Baba, N., & Ohsumi, Y. (1994). autolysis, accelerated loss of viability, and more Ultrastructural analysis of the autophagic process in 48 2. SACCHAROMYCES YEASTS II: SECONDARY FERMENTATION

yeast: Detection of autophagosomes and their charac- wine aging on lees: A bibliographic review. J. Int. Sci. terization. J. Cell Biol., 124, 903e913. Vigne Vin, 36,49e69. Becker, J. M., Naider, F., & Katchalski, E. (1973). Peptide Gonza´lez, R., Martıńez-Rodrı´guez, A. J., & Carrascosa, A. V. utilization in yeast: Studies on methionine and lysine (2003). Yeast autolytic mutants potentially useful for auxotrophs of Saccharomyces cerevisiae. Biochim. Bio- sparkling wine production. Int. J. Food Microbiol., 84,21e26. phys. Acta, 291, 388e397. Gonza´lez, R., Vian, A., & Carrascosa, A. V. (2008). Bidan, P., Feuillat, M., & Moulin, J. P. (1986). Les vins Morphological changes in Saccharomyces cerevisiae during mousseux. Rapport de la France. 65e`me Asembleè the second fermentation of sparkling wines. Food Sci. Geńerale de L’OIV. Bull. OIV, 59, 563e626. Technol. Int., 14, 393e398. Brissonnet, F., & Maujean, A. (1991). Identification of some Harding, T. M., Hefner-Gravink, A., Thumm, M., & foam-active compounds in champagne base wines. Am. Klionsky, D. J. (1996). Genetic and phenotypic overlap J. Enol. Vitic., 42,97e102. between autophagy and the cytoplasm to vacuole protein Cebollero, E., Carrascosa, A. V., & Gonza´lez, R. (2005a). targeting pathway. J. Biol. Chem., 271, 17621e17624. Evidence for yeast autophagy during simulation of Hernawan, T., & Fleet, G. (1995). Chemical and cytological sparkling wine aging: A reappraisal of the mechanism of changes during the autolysis of yeast. J. Ind. Microbiol., yeast autolysis in wine. Biotechonol. Prog., 21, 614e616. 14, 440e450. Cebollero, E., & Gonza´lez, R. (2006). Induction of autophagy Kamada, Y., Funakoshi, T., Shintani, T., Pagano, K., by second fermentation yeasts during elaboration of Ohsumi, M., & Ohsumi, Y. (2000). Tor-mediated induc- sparkling wines. Appl. Environ. Microbiol., 72, 4121e4127. tion of autophagy via an Apg1 protein kinase complex. Cebollero, E., Gonzalez-Ramos, D., & Gonzalez, R. (2009). J. Cell Biol., 150, 1507e1513. Construction of a recombinant autolytic wine yeast Kim, J., Huang, W.-P., Stromhaug, P. E., & Klionsky, D. J. strain overexpressing the csc1e1 allele. Biotechnol. Prog., (2002). Convergence of multiple autophagy and cyto- 25(6), 1598e1604. plasm to vacuole targeting components to a perivacuolar Cebollero, E., Martıńez-Rodrı´guez, A., Carrascosa, A. V., & membrane compartment prior to de novo vesicle forma- Gonza´lez, R. (2005b). Overexpression of csc1e1.A tion. J. Biol. Chem., 277, 763e773. plausible strategy to obtain wine yeast strains under- Klionsky, D. J. (2005). The molecular machinery of autoph- going accelerated autolysis. FEMS Microbiol. Lett., 246, agy: Unanswered questions. J. Cell Sci., 118,7e18. 1e9. Klionsky, D. J., Cregg, J. M., Dunn, W. A., Emr, S. D., Charpentier, C., & Feuillat, M. (1993). Yeast autolysis. Sakai, Y., Sandoval, I. V., et al. (2003). A unified In G. H. Fleet (Ed.), Wine Microbiology and Biotechnology nomenclature for yeast autophagy-related genes. Dev. (pp. 225e242). Chur, Switzerland: Harwood Academic Cell, 5, 539e545. Publishers. Lagace, L. S., & Bisson, L. F. (1990). Survey of yeast prote- Codornnier, R., & Bayonove, C. (1982). Etude de la phase ases for effectiveness of wine haze reduction. Am. J. Enol. pre´fermentaire de la vinification: Extraction et formation Vitic., 41, 147e155. de certains composes de l’arome; cas des terpenols, des Levine, B., & Klionsky, D. J. (2004). Development by self- aldehydes et des alcohols en C6. Connais. Vigne Vin, 15, digestion: Molecular mechanism and biological func- 269e286. tions of autophagy. Dev. Cell, 6, 463e477. Connew, S. J. (1998). Yeast autolysis. A review of curret rese. Luguera, C., Moreno-Arribas, V., Pueyo, E., & Polo, M. C. Arch. Wine Ind. J., 13,61e64. (1997). Capillary electrophoretic analysis of wine de Rosa, T. (1987). La preparazione degli spumanti secondo proteins. Modifications during the manufacture of il sistema dell’ Asti spumante. In T. de Rosa (Ed.), sparkling wines. J. Agric. Food Chem., 45, 3766e3770. Tecnologia dei vini spumanti (pp. 247e249). Brescia, Italy: Martıńez-Rodrı´guez, A., Carrascosa, A. V., Barcenilla, J. M., AEB. Pozo-Bayoń, M. A., & Polo, M. C. (2001a). Autolytic Dizy, M., & Polo, M. C. (1996). Changes of the concentration capacity and foam analysis as additional criteria for the of nitrogenous compounds during fermentation of must selection of yeast strains for sparkling wine production. at pilot-plant scale. Food Sci. Technol. Int., 2,87e93. Food Microbiol., 18, 183e191. Feuillat, M., & Charpentier, C. (1982). Autolysis of yeasts in Martıńez-Rodrı´guez, A. J., Carrascosa, A. V., & Martin- champagne. Am. J. Enol. Vitic., 33,6e13. Alvarez, P. J. (2002). Influence of the yeast strain on the Flanzy, C. (2000). Los vinos espumosos. In A. Madrid changes of the amino acids, peptides and proteins Vicente (Ed.), Enologıá: Fundamentos cientı´ficos y tecnolo´- during sparkling wine production by the traditional gicos (pp. 497e515). Madrid, Spain: Mundi-Prensa. method. Int. J. Microbiol., 29, 314e322. Fornairon-Bonnefond, C., Camarasa, C., Moutounet, M., & Martıńez-Rodrı´guez, A. J., Gonza´lez, R., & Carrascosa, A. V. Salmon, J. M. (2002). New trends on yeast autolysis and (2004). Morphological changes in autolytic wine yeast REFERENCES 49

during aging in two model systems. J. Food Sci., 69, Pueyo, E., Martinez-Rodriguez, A., Polo, M. C., Santa- M233eM239. Marıá, G., & Bartolome´, B. (2000). Release of lipids Martıńez-Rodrı´guez, A., & Polo, M. C. (2000). Character- during yeast autolysis in a model wine system. J. Agric. ization of the nitrogen compounds released during yeast Food Chem., 48,116e122. autolysis in a model wine system. J. Agric. Food Chem., Reggiori, F., Tucker, K. A., Stromhaug, P. E., & Klionsky, D. J. 48, 1081e1085. (2004). The Atg1eAtg13 complex regulates Atg9 and Martıńez-Rodrı´guez, A., Polo, M. C., & Carrascosa, A. V. Atg23 retrieval transport from the pre-autophagosomal (2001b). Structural and ultrastructural changes in yeast structure. Dev. Cell., 6,79e90. cells during autolysis in a model wine system and Scott, S. V., Hefner-Gravink, A., Morano, K. A., Noda, T., sparkling wines. Int. J. Food Microbiol., 71,45e51. Ohsumi, Y., & Klionsky, D. J. (1996). Cytoplasm to Martıńez-Rodrı´guez, A., & Polo, M. C. (2003). Effect of the vacuole targeting and autophagy employ the same addition of bentonite to the tirage solution on the machinery to deliver proteins to the yeast vacuole. Proc. nitrogen composition and sensory quality of sparkling Natl. Acad. Sci. USA., 93, 12304e12308. wines. Food Chem., 81, 383e388. Scott, S. V., Nice, D. C., Nau, J. J., Weisman, L. S., Moreno-Arribas, V., Pueyo, E., Polo, M. C., & Martıń- Kamada, Y., Keizer-Gunnink, I., et al. (2000). Apg13p A´ lvarez, P. J. (1998). Changes in the amino acid and Vac8p are part of a complex of phosphoproteins that composition of the different nitrogenous fractions are required for cytoplasm to vacuole targeting. J. Biol. during the ageing of wine with yeasts. J. Agric. Food Chem., 275, 25840e25849. Chem., 46, 4042e4051. Suzuki, K., Kirisako, T., Kamada, Y., Mizushima, N., Nakatogawa, H., Suzuki, K., Kamada, Y., & Ohsumi, Y. Noda, T., & Ohsumi, Y. (2001). The pre-autophagosomal (2009). Dynamics and diversity in autophagy mecha- structure organized by concertad functions of APG nisms: Lessons from yeast. Nature Reviews Molecular Cell genes is essential for autophagosome formation. EMBO Biology, 10, 458e467. J, 20, 5971e5981. Nuń˜ez, Y. P., Carrascosa, A. V., Gonza´lez, R., Polo, M. C., & Suzuki, K., Noda, T., & Ohsumi, Y. (2004). Interrelationships Martıńez-Rodrı´guez, A. (2005). Effect of accelerated among Atg proteins during autophagy in. Saccharomyces autolysis of yeast on the composition and foaming cerevisiae. Yeast, 21, 1057e1065. properties of sparkling wines elaborated by a cham- Tabera, L., Munõz, R., & Gonza´lez, R. (2006). Deletion of penoise method. J. Agric. Food Chem., 53, 7232e7237. BCY1 from the Saccharomyces cerevisiae genome is Nuń˜ez, Y., Carrascosa, A. V., Gonza´lez, R., Polo, M. C., & semidominant and induces autolytic phenotypes suit- Martıńez-Rodrı´guez, A. J. (2006). Isolation and charac- able for improvement of sparkling wines. Appl. Environ. terization of a thermally extracted yeast cell wall fraction Microbiol., 72, 2351e2358. potentially useful for improving the foaming properties Takeshige, K., Baba, M., Tsuboi, S., Noda, T., & Ohsumi, Y. of sparkling wines. J. Agric. Food Chem., 54, 7898e7903. (1992). Autophagy in yeast demonstrated with Peppler, H. J. (1982). In A. H. Rose (Ed.), Economic microbi- proteinase-deficient mutants and conditions for its ology. Yeast extracts, Vol. 7 (pp. 293e312). London, UK: induction. J. Cell Biol., 119, 301e311. Academic Press. Thumm, M., Egner, R., Koch, B., Schlumpberger, M., Pozo-Bayoń, M. A., Martinez-Rodriguez, A. J., Moreno- Straub, M., Veenhuis, M., et al. (1994). Isolation of Arribas, M. V., Pueyo, E., Martin-Alvarez, P. J., & autophagocytosis mutants of. Saccharomyces cerevisiae. Polo, M. C. (2003b). La elaboracioń de vinos espumosos FEBS Lett., 349, 275e280. y su repercusioń en la composicioń quı´mica y la calidad. Todd, B. E. N., Fleet, G. H., & Henschke, P. A. (2000). Tecnologıá del Vino, 13,55e60. Promotion of autolysis through the interaction of killer Pozo-Bayoń, M. A., Pueyo, E., Martıń-Alvarez, P. J., and sensitive yeast: Potential application in sparkling Martıńez-Rodrı´guez, A., & Y Polo, M. C. (2003a). Influ- wine production. Am. J. Enol. Vitic., 51,65e72. ence of yeast strain, bentonite addition, and aging time Tsukada, M., & Ohsumi, Y. (1993). Isolation and character- on volatile compounds of sparkling wines. Am. J. Enol. ization of autophagy-defective mutants of Saccharo- Vitic., 54, 273e278. myces cerevisiae. FEBS Lett., 333, 169e174. CHAPTER 3

Yeasts Used in Biologically Aged Wines Tahıá Benı´tez, Ana M. Rincoń, Antonio C. Codoń Departamento de Gene´tica, Facultad de Biologıá, Universidad de Sevilla, Sevilla, Spain

OUTLINE

1. Introduction 51 3.3. Influence on the Membrane and Cell Wall 72 2. Characteristics of Yeasts Used in Biologically Aged Wines 56 4. Evolution of Cellular Genomes and Yeast 2.1. Physiological Characteristics of Populations in Biologically Aged Wines 73 Fermentation and Aging Yeasts 56 5. Genetic Improvement of Wine Yeasts 75 2.1.1. Flor Formation 61 5.1. Improving the Characteristics of 2.2. Genetic Characteristics of Fermentation Fermentation Yeasts 75 and Aging Yeasts 64 5.2. Improving the Characteristics of Aging 3. Influence of Environmental Factors on the Yeasts 77 Characteristics of Yeasts Involved in 6. Conclusions 78 Biological Aging 69 Acknowledgments 78 3.1. Influence on Mitochondria 69 3.2. Influence on Chromosomes 71

1. INTRODUCTION (literally, flower veil) or simply flor. In Spain, most biologically aged wines are produced in Most wines that undergo biological aging are the areas of Jerez, Montilla-Moriles, Condado known in Spanish as vinos generosos (literally, de Huelva, Aljarafe, and Rueda (Benı´tez et al., generous wines). They are fortified with grape in press). Outside Spain, the flor aging method spirit at the end of fermentation until an alcohol is used mostly in France (vins jaunes,or content of at least 16% is reached, and then aged yellow wines, from the Jura region), Sardinia, under a film of yeast known as velo de flor California, South Africa, Australia, and certain

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10003-7 51 Copyright Ó 2011 Elsevier Inc. All rights reserved. 52 3. YEASTS USED IN BIOLOGICALLY AGED WINES areas of Europe (mainly the Hungarian region used to produce different types of wines (Garcıá of Tokaj, which produces botrytized wines) Maiquez, 1995) (see Table 3.1). This extremely (Charpentier et al., 2009; Fleet, 2007; Kovacs delicate pressing system is a key step in a care- et al., 2008; Pirino et al., 2004; Sipiczki, 2008). fully controlled production process. Clusters of The best-known biologically aged wines are grapes from healthy vines are harvested at the from the Spanish region of Jerez, which boasts end of August or at the beginning of September a unique mix of soil, elevation, climate, and and vacuum-pressed to prevent contamination flora. Other characteristic features of this region of the must by skins, stems, pips, or similar include its planting and cultivation methods, material. This first press fraction is known by the architecture of the cellars in which the wines the Spanish term primera yema. The grapes are are aged, and its unique winemaking and aging then vacuum-pressed a second time, using techniques (Bravo-Abad, 1986). Thanks to their greater pressure, to produce the second yema. special climate, geography, and topography, The remaining grape products are generally the Guadalquivir valley and Jerez region were ground and fermented and then distilled to highly successful producers and traders of produce the grape spirit used to fortify the wine between the first century BC and the first wines at the end of fermentation. The grape century AD (Celestino-Pere´z, 1999). The exis- skins and seeds are then separated from the tence of vineyards and wines in the area has must by static sedimentation. The first yema is been documented as far back as Roman times, used to produce Fino wines (see Figure 3.1). and there are also records of the wineries of This superior must is normally inoculated Marco de Jerez in the urban land registers for with a Saccharomyces cerevisiae yeast strain from the city of Jerez de la Frontera undertaken at the winery. This locally occurring inoculum is the order of Alfonso X the Wise in 1264 (Garcıá known by the name of pie de cuba. Fermentation del Barrio, 1995). Since alcohol is added at the traditionally took place in oak barrels end of fermentation, most biologically aged measuring 500 to 600 L but nowadays it is care- wines are dry. The high alcohol content of these fully conducted in stainless steel tanks with wines is achieved with the addition of grape close monitoring of temperature and other spirit and the resulting wines are renowned parameters (Sua´rez-Lepe, 1997). Fermentation for their extraordinary finesse. The wines are ends at the beginning of December and the made from white grapes, mainly of the varieties wine is then stored until the solid particles Palomino (95%) and Pedro Ximeńez and Mosca- have settled. tel (5%), and produced under the appellation of Around January or February, the fermented Jerez-Xe´re`z-Sherry (see Table 3.1). The charac- musts may undergo a second selection process, teristic white soils of the area, known as albari- after which they are destined to produce Fino zas, are rich in limestone and retain water. wines (biological aging), Amontillado wines (bio- Given the nature of this soil, the local climate, logical aging followed by oxidative aging), or and the predilection for Palomino grapes, one Oloroso wines (oxidative aging only). The fer- would expect Jerez wines to be relatively insub- mented musts used for Finos, which have an stantial, but the unique winemaking process alcohol content of 10 to 12% vol/vol, are forti- that characterizes the area produces extraordi- fied with grape spirit to a strength of 15.5 to narily unique and flavorful wines (Martıńez- 16% and then transferred to oak butts (barrels), Llopis et al., 1992). which are filled to five sixths of their capacity The juice is extracted from the grapes in (see Figure 3.1). Shortly afterwards, a film of different press fractions. The resulting musts yeast (the flor), formed mostly (>95%) by strains are then separated according to quality and from different races of S. cerevisiae, starts to grow INTRODUCTION 53

TABLE 3.1 Grape Varieties, Types of Wine, and Aging Methods in Jerez-Sanlu´car and Montilla-Moriles Wines

Appellation Jerez-Xe´re`z-Sherry and Manzanilla-Sanlu´ car Montilla-Moriles

REGULATIONS Spanish Ministerial Order 2-V-77 Spanish Ministerial Order 12-XII-85

AUTHORIZED GRAPE VARIETIES Palomino de Jerez Aireń Palomino Fino Baladı´-Verdejo ¼ Jaeń blanco Pedro Ximeńez Moscatel Moscatel Pedro Ximeńez

TYPES OF WINE Fortiﬁed wines Alcohol strength Aged fortiﬁed wines Alcohol strength

Fino 15 Fino 14e17.5 Amontillado 16e18 Amontillado 16e22 Oloroso 18e20 Oloroso 16e20 Palo Cortado and Raya 18e20 Palo Cortado 16e18 Manzanilla 15 Raya 16e20 Natural sweet wine Natural sweet wine

Pedro Xime´nez Pedro Xime´nez White Without aging 10e12 With aging Min. 13

Without aging

Ruedos Min. 14 AGING METHOD In oak butts using the traditional In oak butts with a maximum criaderas and soleras method. capacity of 1000 L for at least All the wines sold under this 2 years using the traditional appellation must be at least criaderas and soleras method. 3 years old.

on the surface of the wine. These yeasts are sugars in the wine have already been metabo- responsible for the biological aging of the wine lized. The combined effect of the oxidative (Martı´nez et al., 1995). At this stage of the metabolism of the ﬂor yeasts and the physical process, alcohol is the only available carbon barrier they form on the surface of the wine source for the yeasts as all the fermentable creates reducing conditions that are responsible 54 3. YEASTS USED IN BIOLOGICALLY AGED WINES

FIGURE 3.1 Fermentation and biological aging of Jerez wines. The musts are fermented until a dry wine with an alcohol content of 10 to 12% is obtained (white vinification). Following an initial selection, the less delicate wines are fortified to 18% alcohol and left to undergo oxidative aging to produce Oloroso wines. The paler and more delicate wines are fortified to a strength of 15.5% and left to undergo biological aging. Following a second selection, some of these more delicate wines, used to produce Fino, are placed in oak butts in which they continue to age under a layer of yeasts. The other wines are fortified to a strength of 17.5% and used to produce Amontillado wines following a process of oxidative aging. Wines are aged for 1 to 3 years (anãdas) in a static system known as sobretablas before undergoing further aging in a dynamic system known as soleras and criaderas. In this system, a certain amount of wine is taken from the oldest butts (soleras)tobe bottled and is then replaced with a younger wine from the first criadera. This, in turn, is replaced by wine from the second criadera and so on until the youngest wine is replaced with wine from the anãdas (sobretablas) system (Benı´tez & Codoń, 2005).

for the pale color and many of the organoleptic The musts used to produce Oloroso wines are characteristics of the final product (Sua´rez-Lepe, fortified to a strength of 18 to 20%. At such 1997). The flor acts as an insulating layer high alcohol concentrations, yeasts are inca- between the wine and the surrounding oxygen, pable of growing and forming a flor. The absence which is continuously consumed by the of a biofilm during oak aging results in an strongly oxidative metabolism of the yeast. oxidized wine with a characteristic color and These reducing conditions, together with the aroma. Olorosos, for example, are very dark products of yeast metabolism, influence the and have an intense aroma and full body, and aroma, flavor, and color of the resulting Fino are either dry or tending towards medium wine, which is renowned for its pale, straw- sweet. In the combined aging system, wines gold color; intense yet delicate aroma that is are aged biologically and, following a second slightly reminiscent of hazelnuts; and rich, dry selection process (see Figure 3.1), fortified to 16 feel on the palate. to 22% alcohol, causing the flor to disappear. Biologically aged wines produced in Sanlu´ - They are then left to undergo oxidative aging, car de Barrameda are called Manzanillas. They producing Amontillados in the case of Finos, are made with the grape variety Listań, which and Manzanilla Pasada wines in the case of is a precursor of the Palomino Fino grape. The Manzanillas from Sanlu´ car. These amber-colored influence of the nearby Atlantic endows Manza- wines are smooth and dry and have a hazelnut nillas with iodine aromas and greater acidity. aroma (see Table 3.1)(Garcıá Maiquez, 1995). INTRODUCTION 55

Certain wines aged using the combined system Certain wines from nearby areas (Montilla- develop the distinctive aroma of Amontillados Moriles in the province of Co´rdoba, and and flavor of Olorosos (see Table 3.1). These are Huelva) also undergo biological aging under known as Palo Cortado wines (see Table 3.1). a layer of flor, but they have certain distinctive Fortified wines fermented on lees that undergo characteristics that set them apart from Jerez oxidative aging have less intense aromas than wines. The grapes used to produce Montilla- Olorosos and are called Raya wines (see Table Moriles wines are grown in limestone soils 3.1) (Benı´tez et al., 2009). and have a very high sugar content. Most Aging takes place in two phases: a static (90%) are of the variety Pedro Ximeńez but phase known as sobretablas, which takes place Aireń, Moscatel, and Baladı´ grapes are also in oak butts, and a dynamic phase using used (see Table 3.1). While Pedro Ximeńez a system known as criaderas and soleras. In the grapes are perfectly adapted to the local climate, second system, the butts are stacked on top of characterized by long, dry summers, Palomino each other to form what is known as a scale. grapes, which produce excellent wines in the The bottom row, nearest the floor, is called the neighboring Jerez, cannot withstand the heat. solera and contains the oldest wine (see Because the wines destined to produce Finos Figure 3.1). The row immediately above this is reach an alcohol content of 15% during fermen- called the primera criadera (literally, first nursery) tation, they do not need to be fortified. These and contains the second-oldest wine. Above this natural Finos, with their distinctive, unique come the second and the third criaderas, and so aroma and flavor, are considered among the on until a height of five or six rows is reached. best in the world and their production has The wines in the scales are mixed and standard- been documented as far back as 700 BC (Bravo- ized using a system known as saca and rocıó, Abad, 1986). White Finos that undergo biolog- where wine taken from the bottom row to be ical aging have complex aromas and almond bottled (twice a year) is replaced with an equal flavors and can reach an alcohol content of volume of wine from the row above. This occurs 17.5%. Amontillados, which are dry on the palate successively up through the rows until the and have hazelnut notes, can reach an alcohol youngest wine (the topmost criadera) is replaced content of 22%, while the amber-colored, by a sobretabla wine, which has been aged using velvety Olorosos can reach levels of 18 to 20%. the static system for 1 to 3 years (anãdas,or Montilla also has young wines with a low vintages). The wines taken from the soleras are alcohol content that are covered by the appella- standardized, stabilized, filtered, and bottled tion (see Table 3.1). This is not the case in Jerez, (Martıńez et al., 1995). They are also mixed however, where certain young wines do not with wines from other soleras from the same meet the requirements of the appellation. winery in suitable proportions to guarantee The wines from Huelva are made with grapes a final product with consistent properties. grown in well-drained, sandy soils. The grape of Jerez also produces sweet wines, made from choice is Zalema, although other varieties are grapes that have been exposed to a lot of sunlight used, including Listań, Moscatel, Palomino, (mainly Pedro Ximeńez and Moscatel grapes). and Garrido Fino. Wines from this area that After a partial fermentation phase, the wines undergo biological aging are straw colored are fortified to halt fermentation and retain the and have an alcohol content of 14 to 17%, sweetness of the wine. They then undergo oxida- although there is also a mahogany-colored tive aging, just like Olorosos, until a velvety, wine called Condado Viejo that is aged using mahogany-colored wine is obtained (Martıńez- the oxidative system and can reach alcohol Llopis et al., 1992; Martıńez et al., 1995). levels of 23% (Martıńez-Llopis et al., 1992). The 56 3. YEASTS USED IN BIOLOGICALLY AGED WINES sensory properties of these wines are similar to using natural fermentation, followed by aging those of Montilla and Jerez. Young wines with under a flor identical to the age-old method a low alcohol content are covered by the Huelva used by the wineries in Jerez (Zara et al., appellation. The first wines to reach the new 2008). None of these wines are fortified as they world following the discovery of America naturally reach an alcohol content of at least were from this region. 15% through fermentation. The area of Rueda, in the province of Valla- In addition to the controlled aging of white dolid, also produces biologically aged wines, wines already discussed, new methods are made with the grape variety Viura. While this being developed to age red wines using selected grape has a good set of primary aromas, it lacks flor yeasts (Sua´rez-Lepe, 1997). The resulting the complexity conferred by other varieties such wines have the characteristic organoleptic prop- as Verdejo. Until recently, the majority of wines erties of biologically aged Finos. produced in Rueda were dry, fortified wines, generally produced using oxidative aging, similar to the wines of Montilla. The Pa´lido 2. CHARACTERISTICS OF YEASTS Rueda wines from the same region are biologi- USED IN BIOLOGICALLY AGED cally aged and fortified with grape spirit to WINES a strength of 15%. They are fragrant, dry wines that are aged for at least 4 years. For the last 3 2.1. Physiological Characteristics of years, they are aged in oak barrels. The other Fermentation and Aging Yeasts fortified wine produced in the region, known as Dorado Rueda, undergoes oxidative aging The majority of yeasts responsible for and has an alcohol content of 15% (Martıńez- fermentation and aging correspond to strains Llopis et al., 1992). of S. cerevisiae. While yeasts that participate in Outside Spain, Australia produces flor fermentation are found on the vine and later sherry, a wine aged using the continuous in fermentation tanks for about just 4 weeks method through a column filled with wood a year, those that participate in aging are shavings. In South Africa and California, flor constantly present in aging barrels. Interest- yeast strains from Spain (S. cerevisiae beticus ingly, wine strains cannot be isolated in the vine- race) have been used to seed fermented musts yards in the weeks immediately before or after following a process similar to that used for harvesting (Mortimer, 2000). One possible Montilla wines (Benı´tez et al., 2009). In France, explanation is that these yeasts are carried to the vins jaunes of the Jura region are fermented the vineyard by insects when the grapes are until they reach an alcohol content of 14 to almost ripe. If this is the case, insects could be 15% and subsequently aged for at least 6 years considered the natural reservoir of these yeasts. under a yeast film that forms on the surface In the majority of today’s wineries, the main (Charpentier et al., 2009). The wine is character- fermentation yeasts are inoculated using ized by its yellow color and persistent walnut selected strains with desirable characteristics, flavor, derived from its high acetaldehyde although the indigenous microflora also makes content (Sua´rez-Lepe, 1997). In other areas of an important contribution to the organoleptic Europe, particularly in the Hungarian region properties of the final product. The yeast film of Tokaj, flor yeasts are used to age wines that forms on the surface of fortified wines after made from grapes infected with the noble rot fermentation is the result of colonization by (Botrytis cinerea)(Kovacs et al., 2008; Sipiczki, naturally occurring yeasts in the wineries. 2008). In Sardinia, fortified wines are produced Although these yeasts also belong to the species CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 57

S. cerevisiae, they do not participate in fermenta- the alcohol consumed by the yeasts (which can tion and are completely different to those that be as high as 7.5e9 L per 500 L barrel per year do in terms of metabolic, physiological, and of aging). genetic characteristics (Esteve-Zarzoso et al., A wide variety of fungi, yeasts, and bacteria 2001). They are autochthonous strains in exist alongside S. cerevisiae throughout the wineries that produce biologically aged wines. fermentation, production, and aging of wines They remain in the aging butts all year round produced using the biological aging method, and displace fermentation yeasts once the but this microflora is reduced to just a few yeasts wine has been fortified to high alcohol levels by increasing the alcohol content of the wine, (Infante et al., 2003). During aging, they release adjusting the pH level to between 3.0 and 3.5, acetaldehyde, consume glycerol and ethanol, and adding sulfur (to a level of between 100 reduce volatile acidity, and increase concentra- and 130 mg/L) (Campo et al., 2008). The musts tions of higher alcohols (see Table 3.2). The used to create Jerez wines contain fungi such cost of producing wines using this system, as Mucor, Rhizopus,andAspergillus species at however, is increased by the long aging period the beginning of fermentation. These are all required and the need to periodically replace natural components of the grape microflora

TABLE 3.2 Basic Characteristics of Jerez Wines

Wines

Parameters that determine the character of the wine Finos Olorosos Pedro Xime´nez

Soil Albariza Albariza Albariza Grape variety Palomino Palomino Pedro Ximeńez Vinification (pressure of must [atm]) <0.5 0.5e1.5 >5 Wine fortified to (degrees) 15.5 18 10.0e15.5

AGING Method Biological Biological/oxidative Oxidative Average age (years) 3e58e10 8e25

AVERAGE ANALYTICAL VALUES pH 2.9e3.3 3.1e3.5 3.6e4.1 Alcohol (degrees) 15.5e17.0 18e21 15.5 Total acidity (g/L of tartaric acid) 3.7e5.2 4.5e6.0 5.2e7.1 Volatile acidity (g/L of acetic acid) <0.3 <0.8e1.2 <0.8e1.3 Acetaldehyde (mg/L) 200e400 60e80 150e200 Glycerol (g/L) <1.0 5e83e5

Malic acid (mg/L) 134e268 335e603 2500 Lactic acid (mg/L) <900 <720 <400 Total polyphenols (mg/L) 250 275e350 500

Reproduced from Garcıá Maiquez (1995). 58 3. YEASTS USED IN BIOLOGICALLY AGED WINES but do not have a defined role in fermentation. above all due to the release of metabolites Other microorganisms present at this stage are through the activity of enzymes produced by bacteria from the genera Acetobacter, Pediococcus, these strains (proteases, lipases, esterases, and and Lactobacillusdwhich can alter the wine but pectinases) (Esteve-Zarzoso et al., 1998). Other are mostly eliminated when acidity levels are yeasts belonging to the genera Kluyveromyces, correcteddand yeasts from the genera Hanse- Torulaspora, and Saccharomyces make an enor- niaspora, Kloeckera, Candida, Pichia, Hansenula, mous contribution to the final aroma of wine Saccharomycodes, and Saccharomyces. The domi- thanks to their ability to convert monoterpene nant yeasts at the end of fermentation, however, alcohols during fermentation. are Saccharomyces yeasts. In Montilla-Moriles Wineries are currently working towards wines, new batches of must are added during controlling the fermentation process by inocu- fermentation to prevent excessive glycerol and lating musts with a selected yeast strain such acidity levels caused by the high sugar content as the pie de cuba to shorten fermentation time of the Pedro Ximeńez grapes used to make these and reduce but not completely eliminate the wines. Thanks to this process, the must retains number of other microorganisms present. a relatively high alcohol content throughout Although most of the glucose and fructose fermentation, explaining why the majority of present in must is converted to ethanol, small yeasts isolated during this period are S. cerevi- quantities undergo glycerol-pyruvic fermenta- siae (Sancho et al., 1986). Indeed, S. cerevisiae is tion, giving rise to the release of glycerol and the predominant species at the end of fermenta- pyruvate (Martıńez et al., 1998). Later, during tion in the majority of biologically aged wines biological aging, the flor yeasts convert ethanol (Charpentier et al., 2009; Martıńez et al., 1995). to acetaldehyde and acetate through oxidation; Exceptions are wines from Sardinia, in which they also consume glycerol, organic acids (ace- Saccharomyces prostoserdovii dominates (Fatichenti tic, lactic, citric, and succinic acid), and amino et al., 1983), and the botrytized wines of Tokaj acids (including proline), and produce higher (Sipiczki, 2003) and the Sauternes region in alcohols (isobutanol and isoamyl alcohol), acet- France (Naumov et al., 2000), where Saccharo- aldehyde, and acetoin (see Table 3.2)(Martıńez myces bayanus var. uvarum dominates. S. bayanus et al., 1998; Munõz et al., 2006). seems to develop in musts fermented at low The musts are clarified prior to fermentation temperatures (Naumov et al., 2000, 2002). to prevent the premature alteration of the proper- The value of non-Saccharomyces yeasts in the ties of the wine (Roldań et al., 2006). Interestingly, production of sweet wines is a topic of debate the fermentation of musts with a high solid (Urso et al., 2008). Generally speaking, they content gives rise to Finos with a lower volatile produce low levels of higher alcohols and ethyl acidity and greater concentrations of acetalde- esters compared to Saccharomyces yeasts. Pichia hyde, higher alcohols, and glycerol. Clarification yeast strains, for example, are undesirable may remove fatty acids and sterols, which would because they produce ethyl acetate (Garcıá Mai- then need to be produced by yeasts through quez, 1995), and Hanseniaspora and Kloeckera acetyl-coenzyme A (CoA) (Roldań et al., 2006). strains have been found to produce high, unde- Microaerobic conditions inhibit the synthesis of sirable levels of acetate, acetaldehyde, ethyl fatty acids, and the subsequent hydrolysis of acetate, and acetoin. Nonetheless, there have acetyl-CoA increases volatile acidity (Martıńez also been reports of strains from the genera Han- et al., 1998; Zara et al., 2009). Indeed, the senula, Kloeckera , Candida, and Pichia, and other inhibition of phospholipid and sterol biosyn- strains with low fermentation activity exerting thesis in microaerobic conditions often causes a seemingly favorable effect on wine aroma, stuck fermentation, even before high alcohol CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 59 concentrations are reached (Mauricio et al., during fermentation that are replaced by a single 1990). Oxygen is also necessary for the consump- dominant population belonging to the beticus tion of proline, which is the main source of race during aging (Esteve-Zarzoso et al., 2001). nitrogen in the must (Ingledew et al., 1987). If The flor that forms on the French vins jaunes proline levels become depleted or there is from the Jura region features S. cerevisiae of the a shortage of oxygen for proline consumption, beticus, montuliensis, and cheresiensis races (Char- other compounds that make an important contri- pentier et al., 2009). In Sardinia, the flor is bution to the aroma of the wine may be formed by S. prostoserdovii and S. bayanus in consumed (Berlanga et al., 2001; Go´mez et al., addition to S. cerevisiae (Zara et al., 2008). Differ- 2004). In an environment with limited oxygen ences in flor composition have been partly supply, yeasts can also release amino acids (thre- attributed to strain differences in sensitivity to onine, methionine, cysteine, tryptophan), which compounds such as acetaldehyde, leading to are synthesized de novo from ethanol to restore a natural evolution towards strains with greater redox potential; these amino acids serve as elec- tolerance of those compounds (Martıńez et al., tron acceptors to oxidate excess nicotinamide 1997b). Other authors have attributed these adenine dinucleotide (Berlanga et al., 2001; differences to the displacement of sensitive pop- Mauricio et al., 2001; Vriesekoop et al., 2009). ulations by killer strains (Mesa et al., 1999). In Analysis of flor films formed in Jerez and San- some wineries, all the strains analyzed have lu´ car wines during biological aging has been found to be sensitive to K1 but resistant revealed the presence of a relatively complex to K2, with the majority of strains not producing microflora. In some wineries, over 95% of the toxins (Ibeas & Jimeńez, 1996). Other studies, in flora has been found to be formed by S. cerevi- contrast, have found all beticus and cheresiensis siae, with two predominant racesdS. cerevisiae races to be resistant to K1 and K2 toxins and beticus and S. cerevisiae montuliensisdand two montuliensis and rouxii races to be sensitive to minority racesdS. cerevisiae cheresiensis and both. None of the strains were found to produce S. cerevisiae rouxii. These races can be distin- toxins and the authors reported a balance guished on the basis of their metabolic charac- between the four races in the static and dynamic teristics (Martıńez et al., 1997a). The remaining aging systems that they attributed to speed of population (approximately 4%) is composed of flor formation and resistance to acetaldehyde yeasts from the genera Debaryomyces, Pichia, rather than to the killer character of the strains Hansenula, and Candida. Just a single dominant (see Figure 3.2)(Martıńez et al., 1997a). The racedS. cerevisiae beticusdin coexistence with enormous variations that seem to exist in micro- minority Dekkera and Brettanomyces strains flora composition from one winery to the next (which might be responsible for sporadic would explain why wines from different spoilage of Fino wines due to acidification) has wineries in geographically close areas often been isolated in certain wineries, and molecular have very different organoleptic properties, analyses have even shown a single S. cerevisiae though made using similar processes (Budroni beticus population per butt (Ibeas et al., 1996). et al., 2005; Me´rida et al., 2005). In other wineries, the microflora has been found Finos, Amontillados, and Olorosos all extract to be mainly composed of beticus and cheresiensis tannins, phenols, and other compounds from races, with Pichia species existing as minority the wood of the butts in which they are aged. yeasts (Mesa et al., 1999). The literature also In the case of Finos and Amontillados, the meta- contains reports of wineries with minority pop- bolic activities of the flor yeasts also lead to ulations of Candida, Dekkera, Hanseniaspora, enrichment of 3-methylbutanal, phenylacetalde- Zygosaccharomyces, and Metschnikowia species hyde, methional, and sotolon, as well as methyl 60 3. YEASTS USED IN BIOLOGICALLY AGED WINES

(a) (b) 35 800

30 700 25

600 20

15 500 Time (days) Time

10 Acetaldehyde (mg/L) 400 5

0 300 B1 B2 B3 B16 CH15 M10 M12 M17 R13 B2 B3 B16 CH15 M10 M12 M17 M9 R13 Strains Strains

FIGURE 3.2 Flor formation speed (a) and acetaldehyde production (b) in different strains of Saccharomyces cerevisiae in young sherry wine (sobretabla). Strains from the same race can have different characteristics. S. cerevisiae beticus (B) and S. cerevisiae cheresiensis (CH), for example, are faster at forming a flor than S. cerevisiae montuliensis (M) or S. cerevisiae rouxii (R), but they are also characterized by less acetaldehyde production and tolerance. Reproduced from Martıńez et al. (1997a). esters derived from methylpentanoic acids. S. cerevisiae (Kawarai et al., 2007). These bacteria Finos, in particular, are especially rich in acetal- in the flor play a key role in the consumption of dehyde, diacetyl, ethyl esters of branched organic acids (gluconic, malic, and lactic acid). aliphatic acids, and 4-ethylguaiacol (Campo Lactic acid bacteria are responsible for malo- et al., 2008; Peinado et al., 2004). The character- lactic fermentation during the aging of younger istic flavor of Fino wines is principally derived wines, normally in the fourth or fifth criadera from acetaldehyde, but diacetylene and acetoin rows. These reactions have been detected in also have a role (Munõz et al., 2006). Concentra- wines from both Jerez and Montilla-Moriles tions of ethanol, glycerol, acetaldehyde, acetic (Bravo-Abad, 1986; Me´rida et al., 2005; Peinado acid, and nitrogenous and volatile compounds et al., 2004). While certain races of S. cerevisiae change continually throughout aging (Martıńez (beticus, cheresiensis) are more efficient than et al., 1998). While considerable amounts of others (montuliensis, rouxii) at reducing volatile ethanol, glycerol, organic acids, and amino acids acidity, they are less efficient when it comes to are consumed during biological aging (Martıńez consuming alcohol and producing acetaldehyde et al., 1998), this consumption is not continuous. (Martıńez de la Ossa et al., 1987b; Martıńez Ethanol, for instance, is consumed in the et al., 1993, 1998). Even within the same race, greatest quantities during the formation of the some strains are much more efficient than others flor, while glycerol, organic acids, and amino at reducing volatile acidity. Variations in volatile acids are consumed once the film has been concentrations depend not only on the yeast established. The flor also contains bacterial pop- strains that form the flor but also on aging condi- ulations, most of which are species of Lactoba- tions, number of criadera levels, butt replenish- cillus, associated with varying populations of ment methods, vineyard density, and climate CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 61

(Martıńez et al., 1993). In Fino wines, the dry causes the cells to aggregate. The resulting extract decreases to under 15 g/L, mainly aggregate adheres to the gas bubbles generated because of the consumption of glycerol by the during respiration and floats on the surface (see flor yeasts, while in Oloroso wines this extract Figure 3.3). It has been suggested that the flor is can exceed levels of 22 g/L due to the concentra- the result of ethanol-induced lipogenic activity tion of compounds in the wine produced by (Bravo-Abad, 1986). Later studies, however, evaporation (Martıńez de la Ossa et al., 1987a) reported that, while the addition of oleic acid (Table 3.2). or ergosterol did not affect flor formation, the addition of different proteases disintegrated Flor 2.1.1. Formation the flor and reduced hydrophobicity (Martıńez The flor is a structure of cells that forms et al., 1997c), indicating that flor formation a thick, white, rough film that floats on the depends on the presence of hydrophobic cell- surface of the wine. The yeasts that form this surface proteins. More recent studies have film are responsible for the biological aging of assigned a key role to mannoproteins containing Fino and Amontillado wines and can survive in over 90% mannose in hydrophobicity and flor alcohol concentrations of around 16%. Other formation (Caridi, 2006), and described how hostile conditions that these yeasts have to yeast cells secrete glucose and mannose poly- tolerate are high concentrations of acetalde- saccharides that surround cell aggregates (Beau- hyde, oxidative stress due to the metabolism vais et al., 2009). of nonfermentable carbon sources, water stress, While the presence of fermentable carbon and, often, high levels of metals (e.g., copper) sources or ammonium salts generally inhibits and nitrogen sources that are difficult to flor formation, both proline and ethanol appear assimilate (e.g., proline). The flor is thus consid- to activate the process (Fidalgo et al., 2006; Mar- ered to be an adaptive mechanism in which tıńez et al., 1997b). Indeed, hydrophobicity yeast cells change their size, shape, and hydro- increases and films become more compact as phobicity in response to different stresses. alcohol levels increase. Nonetheless, yeasts Hydrophobicity, which is imparted by the pres- have also been found to form flors in sweet ence of specific surface proteins (including Flo11 botrytized wines (Kovacs et al., 2008). Accord- [Muc1], which will be discussed in Section 3.3, ing to some authors, flor formation is very

FIGURE 3.3 Flor formation by wild- type strains and their Dflo11 disruptants. Different yeast strains were cultivated overnight and inoculated into flor SD medium with 3% ethanol (vol/vol) and a pH of 3.5. The tubes containing each culture were photographed after 3 d of static incubation at 30C. The data indicate that FLO11 is the primary factor in flor formation but that there are other genes involved. A9 is a diploid flor strain containing two functional copies of FLO11; A9- F1 is a strain derived from A9 with a disrupted copy of FLO11; and A9-F2, also derived from A9, has two disrupted copies of FLO11. Ar5-H12 is a flor strain with a single functional copy of FLO11; its derivative Ar5-H12-F1 has an undisrupted copy of FLO11. W3 is a wine strain incapable of forming a flor, as are its derivatives W3-F1 and W3-F2, which have one and two disrupted FLO11 alleles, respectively. Reproduced from Ishigami et al. (2006). 62 3. YEASTS USED IN BIOLOGICALLY AGED WINES positively influenced by the presence of poly- The genes responsible for flocculation (FLO phenol compounds (Budroni et al., 1995) and genes) form subtelomeric families that include biotin (Bravo-Abad, 1986). Others, in contrast, both functional genes and pseudogenes (Teunis- have found that the process requires panto- sen & Steensma, 1995; van Mulders et al., 2009). thenic acid in addition to the oxidative metabo- They encode cell-surface glycosylphosphatidy- lism (Martıńez et al., 1997b). These differences, linositol (GPI)-linked proteins that are cova- however, may simply be due to differences in lently bound to glucans in the cell wall the composition of the wine (Charpentier (Beauvais et al., 2009; Verstrepen & Klis, 2006). et al., 2009). These cell-surface proteins, known as adhesins Several authors have likened flor formation to (Huang et al., 2009), are composed of repeating a form of flocculation and to pseudohyphal motifs organized in a characteristic fashion development and invasive growth (Budroni from the plasma membrane, through the cell et al., 1995; Lambrechts et al., 1996). The simi- wall, to the cell surface (Douglas et al., 2007; larity between flor formation and flocculation van Mulders et al., 2009) (see Figure 3.4). and filamentation lies in the notable increase in Recombination of the internal repeats results cell hydrophobicity that occurs in all these cases in an increase or decrease in protein size, which, (Straver & Kijne, 1996) and the activation of these in turn, would alter phenotypes such as adher- processes in environments with limited nitrogen ence, flor formation, and flocculation (Rando & supply (Douglas et al., 2007; Ma et al., 2007), as Verstrepen, 2007; Verstrepen et al., 2005; Ver- will be discussed at the end of this section. strepen et al., 2004). One of the genes

YNL 190W* YOL155C* TIR1* EGT2* PIR3* DAN4* PIR1* MNN4* SED1* FLO5* HSP150* FLO10* CTR1* FLO9* TIR4* FLO1* FIT1* MSB2* AGA1* MUC1*

MFalpha SPA2*

NOP1 HKR1*

UBI5 SLA1

DDR48 YIL080W

NUM1

0 kb 1 kb 2 kb 3 kb 4 kb 5 kb 6 kb 7 kb 8 kb

FIGURE 3.4 Saccharomyces cerevisiae proteins containing conserved intragenic repeats. A screen of all open reading frames in the S. cerevisiae genome revealed 29 genes with large repeats (>40 nucleotides). Some of the repeats (vertical boxes) showed variations in size from one strain to the next. The majority of repeats occurred in cell-surface proteins. The names of the genes encoding proteins of this type are shown with asterisks. Reproduced from Verstrepen et al. (2005). CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 63

responsible for cell flocculation, FLO11, regu- (a) lates pseudomycelium formation, flor formation, and invasive growth in yeasts (Barrales et al., 2008; Lambrechts et al., 1996; Lo & Dranginis, 1996; Palecek et al., 2000; Tamaki et al., 2000). All of these phenomena are characterized by an increase in cell hydrophobicity (Purevdorj-Gage et al., 2007) (see Figure 3.5). WT WT FLO11 + Another of the genes responsible for floccula- (b) tion, FLO1, is involved in protecting cell aggre- 70 gates from a range of hostile conditions such as the presence of ethanol and certain antimicro- 60 bial compounds (Beauvais et al., 2009; Smukalla 50 et al., 2008), which is precisely the environment 40 the flor yeasts have to endure (see Figure 3.6). 30 Adverse conditions such as the presence of 20 ethanol and acetaldehyde, low pH, and a lack of nutrients activate the genes encoding adhe- 10 sins (Barrales et al., 2008; van Dyk et al., 2005; Aggregates of >5 cells (%) 0 Verstrepen & Klis, 2006) (see Figure 3.7). None- WT FLO11+ WT FLO11+ WT + cntrol WT FLO11 no induc post induc plasm. theless, synergistic effects between different 17.9 % 17.5 % 15.6 % 55.9 % 8.8 % hostile conditions (e.g., high temperature and ethanol) can lengthen the time required for (c) the flor to form and even cause it to disappear (Ibeas & Jimeńez, 1997) (see Figure 3.8). 50 The Flo11 adhesin is regulated via three main 40 signaling pathways: the mitogen-activated protein kinase (MAPK)-dependent pathway 30 (which is regulated by nitrogen depletion at 20 least) and two pathways regulated by glucose (Figure 3.7). Furthermore, other activators 10 involved in chromatin remodeling (e.g., Msn1) 0 and pH response pathways (e.g., Rim20) also WT FLO11 + WT FLO11 + WT + cntrol control the synthesis of Flo11 in these pathways

Cells separated in octane (%) WT FLO11 no induc post induc plasm. (see Figure 3.9). These observations, together 11.9 % 5.5 % 12.1 % 35.3 % 10.1 %

FIGURE 3.5 Flo11 overexpression confers greater ability to coaggregate and greater hydrophobicity. Micrograph of vortexing the tubes, and leaving the phases to separate. The wild-type (WT) strain (left) and strain overexpressing difference between the OD600nm before and after the addi- FLO11 under the control of the GAL1 promoter (WT tion of octane was used to determine the hydrophobicity of FLO11þ) (right), both during exponential growth (a). the culture. The greater the percentage of partitioned cells, Percentage of cells in multiple cell aggregates (b). Hydro- the greater the hydrophobicity. No induc ¼ GAL1 not phobicity (c). The percentage of cells partitioned in octane induced; post-induc ¼ GAL1 induced; DFLO11 ¼ strain was measured using an aqueous-hydrocarbon biphasic with disrupted FLO11 gene; WTþcntrol plasm ¼ wild-type assay. Optical density (OD)600 nm was measured after transformed with empty plasmid. Reproduced from placing a volume of octane on the surface of the aliquots, Purevdorj-Gage et al. (2007). 64 3. YEASTS USED IN BIOLOGICALLY AGED WINES

FIGURE 3.6 Presence of Flo1 protein 100 and flocculation capacity confer resistance to different stresses. Saccharomyces cer- 80 evisiae cells with (KV210) and without flo1- (KV22) (KV22) FLO1 expression were subjected FLO1+ (KV210) to various stress treatments, after which 60 the percentage of surviving cells was measured. Asterisks indicate statistically 40 significant differences between flocculent % survival and nonflocculent cultures (a ¼ 0.05); 20 error bars correspond to standard deviation. Reproduced from Smukalla et al. (2008). 0 15 mg/ml 5 mM peroxide 10% ethanol freezing/thaw 50 ºC 1 hr amphotericin B with the role attributed to FLO genes and the DNA (mtDNA) polymorphisms (restriction possible relationship between flocculation, fragment length polymorphism [RFLP] anal- flor formation, and pseudomycelium growth ysis) (Martıńez et al., 1995). Research performed (Ishigami et al., 2006), are further supported by since these techniques became available has the fact that high levels of these genes (mainly confirmed that the fermentation and flor strains of FLO11) have been detected during the flor of S. cerevisiae in biologically aged wines display formation phase (Infante et al., 2003). a high degree of genetic variability, not only in terms of DNA content (variations of 1.3 to 2.2. Genetic Characteristics of almost 4.0 n) but also in the number and size Fermentation and Aging Yeasts of nuclear chromosomes and mtDNA restriction fragments (Martıńez et al., 1995). Electropho- The emergence of molecular biology tech- retic karyotyping has also revealed important niques has permitted a more accurate classifica- differences between the size and number of tion of wine yeasts and revealed enormous chromosomes in both fermentation and flor variability among the different strains of S. cere- yeast strains in biologically aged wines (Valero visiae (Fernańdez-Espinar et al., 2003; Martorell et al., 2007). Furthermore, flor yeasts as a whole et al., 2005). While traditional metabolic seem to have a different chromosome pattern to methods were successfully used to distinguish other wine yeasts. between different races of S. cerevisiae in the Even greater variations, however, seem to fermentation and flor microflora of biologically exist in mtDNA (Martıńez et al., 1995). mtDNA aged wines, they were unable to unequivocally restriction analysis is, thus, a sufficiently simple, distinguish between different populations of rapid, and unequivocal method for studying the the same race (Martıńez et al., 1995). The emer- yeast populations involved in fermentation and gence of molecular techniques, however, has wine aging and monitoring the development of greatly improved the genetic characterization inoculated strains to determine whether or not of both fermentation and flor yeasts (Benı´tez they displace indigenous populations (Esteve- et al., 1996). The use of these techniques has Zarzoso et al., 2001). shed light on the DNA content of yeast strains The first genetic characterization studies of (flow cytometry), chromosome numbers and yeast strains in biologically aged wines were size (pulsed-field gel electrophoresis [PFGE]), conducted by Sancho et al. (1986) in Montilla- homology with genes from other yeasts (hybrid- Moriles wines. The authors isolated and ization with specific probes), and mitochondrial characterized strains during fermentation and CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 65

FIGURE 3.7 Signaling cascades that regulate Flo11. (a) The MAPK-dependent filamentous growth pathway. The core of this pathway in Saccharomyces cerevisiae is formed by the central kinases Ste11 (MAPKKK) and Ste7 (MAPKK). These kinases are shared by other MAPK signaling cascades, such as the mating response pathway and the high osmolarity glycerol (HOG) pathway. Msb2 is thought to function as a sensor at the top of the pathway, but the conditions that trigger Msb2 have not yet been characterized. Other known triggers of FLO11 that (at least partially) act through MAPK signaling include nitrogen starvation (which might be sensed through the ammonium permease Mep2) and elevated concentrations of certain fusel alcohols such as butanol. The specific downstream part of the MAPK-dependent filamentous growth pathway includes the MAPK Kss1 and the transcriptional regulators Dig1, Ste12, and Tec1. (b) The Ras/AMPc/PKA pathway. The Ras/cAMP/PKA pathway responds to the presence of glucose or sucrose in the medium. The pathway is activated by two independent triggers. First, the intracellular phosphorylation of glucose enhances the activity of adenylate cyclase Cyr1. Second, a G protein-coupled receptor system, consisting of the receptor Gpr1 and the Ga protein Gpa2, senses extracellular glucose and sucrose. Activation of the Gpr1/Gpa2 complex causes a further increase in Cyr1 activity, resulting in a transient cAMP peak. Subsequently, cAMP activates the protein kinase A complex (PKA), resulting in the dissociation of the Bcy1 subunits from the Tpk catalytic subunits of PKA. The three different Tpk subunits, Tpk1, Tpk2, and Tpk3, have been shown to have distinct roles in FLO11 regulation: Tpk2 mostly acts as an activator, while Tpk1 and Tpk3 function as inhibitors. Once released from the inhibitory Bcy1 subunits, the free Tpk2 kinase inactivates Sfl1 (suppressor of flocculation) and activates the positive regulator Flo8. (c) The main glucose repression pathway. The hexose transporters (Hxt) allow glucose uptake from the medium. Once inside the cell, glucose is phosphorylated to glucose-6-phosphate by one of the hexokinases (Hxk). This phosphorylation process and/or the depletion of AMP due to the increase in ATP production inactivate(s) the central Snf1 protein kinase. Inactivation of Snf1 allows the regulatory proteins Mig1 and Nrg1 to bind to the FLO11 promoter and recruit the general repressors Tup1 and Ssn6, resulting in repression of FLO11. Questions marks indicate unknown mechanism. Reproduced from Verstrepen and Klis (2006). biological aging and found that all the yeasts of flor strains in which they discovered hetero- belonged to different races of S. cerevisiae. zygotic lethal recessive alleles. This indicated More interestingly, however, they found that that these strains never sporulated or at least the races did not mix, which suggests strong that sporulation occurred less frequently than sexual isolation between the populations to mutations in lethal alleles. More recently, Puig prevent the random distribution of metabolic et al. (2000) attributed genetic variability in characteristics. Jimeńez and Benı´tez (1988) wine strains to mitotic recombination, repair, confirmed this sexual isolation in a later study and gene conversion during growth and not to 66 3. YEASTS USED IN BIOLOGICALLY AGED WINES

Temperature State of flor were able to form a flor film more quickly and % rho mutants Ethanol content that montuliensis races were more abundant in older wines because of their greater production 30 and tolerance of acetaldehyde (see Figure 3.2). 25 Other authors, on analyzing the molecular 20 profiles of the flor in other wineries, found 15 a single dominant strain (Esteve-Zarzoso et al., 10 2001; Ibeas et al., 1997). 5 Because the majority of flor strains never 0 sporulate (or sporulate only poorly) and meiotic Jul Jul Oct Oct Apr Apr Jan Jun Jan Jun Feb Feb Mar Mar Nov Dec Nov Dec Aug Sep Aug Sep May May products are often inviable (Martıńez et al., 1993 1994 1995), chromosome constitution has been estab- FIGURE 3.8 Alcohol content (% vol/vol), temperature lished by studying spores produced using mass ( C), rho (petite) mutants (%), and state of the flor during the mating methods and analyzing the segregation biological aging of Jerez wines. Synergistic effects of alcohol frequencies of markers in the chromosomes and temperature may give rise to a high rate of mutants after the sporulation of industrialelaboratory with nonfunctional mitochondria (petite mutants) and the hybrids (Bakalinsky & Snow, 1990). This deterioration or disappearance of the film. flor Reproduced method can be used to distinguish between from Ibeas et al. (1997). disomic, trisomic, and tetrasomic complements sporulation and meiotic recombination. There in parent strains. The study by Bakalinsky and have, however, been reports of sporadic intra- Snow showed numerous cases of aneuploidy and interspecific crosses between Saccharomyces in the strains analyzed but of particular interest strains from wine and other enological sources was the large number of extra copies of chromo- (Gonza´lez et al., 2008). somes V, VII, and XIII detected. Chromosome Mesa et al. (1999) found enormous genetic XIII contains the genes for alcohol and aldehyde variation between beticus and cheresiensis races dehydrogenases, which play a key role in the of flor yeasts isolated in a winery in terms of production and consumption of ethanol and both chromosome patterns and mtDNA acetaldehyde. Using similar methods, Guijo profiles. They also reported preferential associa- et al. (1997) analyzed the chromosome structure tions between chromosome and mitochondrial of wine strains involved in the fermentation and patterns. In a subsequent study, the same group aging of Montilla-Moriles wines. They detected found a correlation between strains with aneuploidy in all of the strains analyzed as well specific patterns and different aged wines, as a high rate of polysomy for chromosome XIII. with certain patterns found only in soleras and Considerable genomic plasticity has also been others found only at criadera levels (Mesa detected for chromosomes IV, VIII, and XII in et al., 2000). In an earlier study, Nadal et al. fermentation strains, including those that partic- (1996) had observed a strong association ipate in biologically aged wines (Infante et al., between specific mtDNA patterns and fermen- 2003; Puig et al., 2000). These variations have tation strains with high tolerance of ethanol been linked to chromosome translocations due and temperature. Shortly afterwards, Martıńez to homologous, asymmetric, or ectopic recombi- et al. (1997a) found that different aged wines nation between sequences of Ty, d,orY’ in the solera and criadera system were correlated elements that resulted in strains that were better with races rather than with specific mtDNA adapted to specific industrial conditions. Vari- profiles. They proposed that beticus races were ability in chromosome patterns has been attrib- more abundant in younger wines because they uted to recombination events that occur during CHARACTERISTICS OF YEASTS USED IN BIOLOGICALLY AGED WINES 67

FIGURE 3.9 related (a) 100 FLO11- 90 phenotypes in mutants with 80 70 different gene deletions. In addi- 60 tion to the regulatory function 50 40 described in Figure 3.6, proteins 30 involved in chromatin remodeling 20 10 and pH response can regulate the

Hydrophobicity (%) 0 synthesis of Flo11 protein. Hydrophobicity (a) was measured flo8Δ 133d flo11Δ ash1Δ rxt2Δ sds3Δ snf5Δ snf2Δ yta7Δ tub1Δ msn1Δmss11Δ gal11Δsap30Δpho23Δ rim20Δ using an aqueous-hydrocarbon biphasic assay (Purevdorj-Gage (b) et al., 2007). Flor ﬁlm on solid surface (b). Exponentially growing 5 cells were placed in microtiter 4,5 plate wells and incubated for 1 h 4 3,5 at 28C. The cells were then 3 stained with crystal violet and the

530 2,5

A 2 wells washed repeatedly with 1,5 1 water and photographed. For 0,5 bioﬁlm quantiﬁcation, the crystal 0 violet was solubilized using SDS

133d flo8Δ rxt2Δ snf5Δ snf2Δ yta7Δ tub1Δ and optical density530nm was flo11Δ msn1Δ ash1Δgal11Δ sds3Δ rim20Δ mss11Δ sap30Δpho23Δ measured. Invasive growth (c). Exponentially growing cells were (c) spotted on yeast extract peptone Unwashed dextrose (YPED) solid medium and photographed before Washed (unwashed) and after (washed) washing. 133d ¼ wild-type strain; ahs1 ¼ transcription factor; flo11 flo8Δ 133d flo11Δ ash1Δ rxt2Δ sds3Δ snf5Δ snf2Δ yta7Δ tub1Δ msn1Δmss11Δ gal11Δsap30Δpho23Δ rim20Δ ¼ adhesin; flo8 ¼ transcription factor; gal11 ¼ mediator complex component; msn1 ¼ transcriptional activator; mss11 ¼ transcription factor; pho23 ¼ component of Rpd3 histone deacetylase complex; rim20 ¼ protein involved in proteolytic activation of Rim101 in response to alkaline pH; rxt2 ¼ subunit of histone deacetylase complex Rpd3L; sap30 ¼ subunit of histone deacetylase complex; sds3 ¼ component of Rpd3p/Sin3p deacetylase complex; snf2 ¼ catalytic subunit of SWI/SNF chromatin remodeling complex; snf5 ¼ subunit of SWI/SNF chromatin remodeling complex; tup1 ¼ general transcriptional repressor; yta7 ¼ protein of unknown function. Reproduced from Barrales et al. (2008). vegetative growth due to the hostile environ- compare, gene by gene, the genomes of flor ment the yeasts have to endure (Infante et al., and laboratory yeasts have revealed the exis- 2003). Several chromosomal rearrangements tence of chromosomes, chromosomal regions, appear to be mediated by Ty1 elements, and and genes that affect events such as aneuploidy, yeasts that participate in biological aging have amplification, and deletion and are partly been found to possess Ty1 elements located responsible for the genetic variability detected exclusively on chromosome XII. Whether the and the enological properties of yeast strains severe restrictions affecting Ty1 mobility and (Hu et al., 2007; Infante et al., 2003). High expression in laboratory strains also apply to expression levels of genes involved in the wine strains is not known (Nyswaner et al., biosynthesis of amino acids and the metabolism 2008; Wu & Jiang, 2008). Hybridization experi- of nitrogen and sulfur, for example, have been ments based on DNA microarrays used to found in wine strains only, as has the 68 3. YEASTS USED IN BIOLOGICALLY AGED WINES

TABLE 3.3 Open Reading Frames (ORFs)1 Included in Genomic Regions Ampliﬁed in Saccharomyces cerevisiae Flor Yeast Strain 11.3 That Have Been Found Overexpressed in This Strain With Respect to S. cerevisiae X2180 Strain During Growth Under Enological-like Conditions

ORF Name Chromosome Characteristics of gene product

YBL092W RPL32 II (20e82 kb) 60S large subunit ribosomal protein YBR089C-A NHP6B II (427e436 kb) Regulation of transcription (chromatin architecture) YCL018W LEU2 III (76e105 kb) 3-Isopropylmalate dehydrogenase YCL050C APA1 III (3.5e70 kb) ATP adenyltransferase

YDL198C YHM1 IV (0e116 kb) Mitochondrial carrier protein (maintenance of mitochondrial genome)

YEL017C-A PMP2 V (30e128 kb) Plasma membrane Hþ-ATPase regulator a YER044C ERG28 V (196e313 kb) Involved in ergosterol biosynthesis YER163C V (488e554 kb) Biological process/function unknown

a YGR234W YHB1 VII (697e1095 kb) Flavohemoglobin (cell protection against nitrosylation) YHR053C CUP1-1 VIII (208e217 kb) Copper-binding (metallothionein) protein YHR055C CUP1-2 VIII (208e217 kb) Copper-binding (metallothionein) protein YHR096C HXT5 VIII (285e320 kb) Hexose transporter YHR162Wa VIII (320e481 kb) Biological process/function unknown

YIL065C FIS1 IX (232e243 kb) Involved in mitochondrial ﬁssion YIL155C GUT2 IX(18e57 kb) Glycerol 3-phosphate dehydrogenase (mitochondrial) YIR019C MUC1 IX (312e425 kb) Cell surface glycoprotein involved in bioﬁlm formation YIR037W HYR1 IX (312e425 kb) Glutathione peroxidase YMR009W XIII (196e427 kb) Biological process/function unknown

YPL092W SSU1 XVI (374e590 kb) Sulﬁte transport (sulﬁte resistance) YPR099C XVI (729e825 kb) Biological process/function unknown

1ORFs with significant log ratios, which indicate a higher copy in strain 11.3, but included within a chromosomal region with equal copy number in both 11.3 and 1.28 strains. The majority of ORFs correspond to proteins of enological interest. The limits of the genomic regions in each chromosome (from left telomere) are indicated. Reproduced from Infante et al. (2003). overexpression of genes linked to tolerance of Analysis of flor strains has revealed aneu- high levels of sulfur dioxide (Backhus et al., ploidy of chromosomes I, III, and IV and chro- 2001). Furthermore, in the study by Backhus mosomes X and XII in beticus and montuliensis et al., genes that regulate the change from races, respectively (Infante et al., 2003). In the fermentative to aerobic metabolism and are same study, variations were also found in the repressed by glucose in laboratory strains were copy number of 38% of the open reading frames induced in wine strains under low-nitrogen (ORFs) that make up the genome (see Table 3.3). conditions, despite high levels of glucose. The amplified regions appeared to be associated INFLUENCE OF ENVIRONMENTAL FACTORS ON THE CHARACTERISTICS OF YEASTS INVOLVED 69 with the presence of Ty transposons and long Although each race of S. cerevisiae confers terminal repeat elements. Moreover, it was specific characteristics to wine, the four majority found that, of the genes overexpressed during races of flor yeasts have the same restriction the flor phase that are of enological interest pattern in the intergenic region of 5.8S rDNA and involved in flor formation, sulfite tolerance, (Esteve-Zarzoso et al., 2004). This means that ergosterol synthesis, and the metabolism of metabolic differences are not detectable at the glutathione (GSH) and other substances (see molecular level. Furthermore, strains can vary Table 3.3), 37% were amplified genes, pointing even within the same race. This explains the to a possible association between gene rear- existence of phenotypes characterized by low, rangements and adaptation to specific condi- moderate, and high acetaldehyde production tions (Infante et al., 2003). There have also (550e800 mg/L for montuliensis strains and been reports of ribosomal DNA (rDNA) rear- 350e450 mg/L for beticus strains) and low, rangements involving chromosome XII in wine moderate, or high rate of flor formation yeasts not mediated by either homologous (15e20 d for beticus strains and 25e35 d for mon- recombination or Ty1 elements (Blake et al., tuliensis strains) (See Figure 3.2). These varia- 2006). Several authors have suggested that tions mean that a very large number of more than one gene may be involved in the individuals must be analyzed before a property formation and maintenance of the flor. Jimeńez can be assigned to a particular race. and Benı´tez (1988) and Castrejoń (2000),for example, using meiotic analysis of flor yeast strains found that flor formation was regulated 3. INFLUENCE OF by more than one gene. Those authors also ENVIRONMENTAL FACTORS ON found variations in aneuploidy. THE CHARACTERISTICS OF Restriction analysis of the intergenic region of YEASTS INVOLVED IN 5.8S rDNA has identified a 24-base-pair deletion BIOLOGICAL AGING in over 150 S. cerevisiae flor strains analyzed; this deletion was not found in fermentation strains, 3.1. Influence on Mitochondria providing further evidence that the two types of strain are different (Esteve-Zarzoso et al., Yeasts responsible for the biological aging of 2001). The 24-base-pair deletion was also found wines (basically flor yeasts) show enormous vari- in strains of Saccharomyces aceti and Saccharo- ability in terms of both chromosome and mtDNA myces gaditensis isolated from flor films in Jerez restriction fragment patterns (Martıńez et al., wines and in strains of S. prostoserdovii isolated 1995). mtDNA variability could be a result of from flor films in Vernaccia de Oristano wines mutations induced by the high mutagenic (Fernadez-Espinar et al., 2000). In more recent concentrations of alcohol the yeasts are exposed studies, the same deletion was detected in flor to. The frequency of spontaneous mutants with yeasts (all S. cerevisiae) isolated in different anã- nonfunctional mitochondria (petite mutants) das, soleras, and criaderas in Jerez and Montilla- has been reported to increase ten-fold at concen- Moriles wines (Naumova et al., 2005) and in trations of 24% ethanol (Bandas & Zakharov, the majority of strains isolated from flors in 1980; Castrejoń et al., 2002) (see Figure 3.8). The botrytized wines (Kovacs et al., 2008). The dele- mitochondrial genome is responsible for cell tion was not detected in S. cerevisiae flor strains viability in high-alcohol environments (Jimeńez from the French vins jaunes (Charpentier et al., & Benı´tez, 1988). In particular, the mitochondria 2009), indicating that flor yeasts possibly have of flor yeasts remain functional under these different phylogenic origins. conditions thanks to their exceptional resistance 70 3. YEASTS USED IN BIOLOGICALLY AGED WINES

FIGURE 3.10 Generation of reactive oxygen species (ROS) in the mitochondrion and role of superoxide dismutases. Mitochondrial DNA is the target of ROS. The increased resistance of flor yeasts to ROS and ethanol is largely due to a greater efficiency of the mitochondrial manganese superoxide dismutase (MnSOD). Reproduced from Raha and Robinson (2000). to ethanol. Because flor yeasts have an oxidative (Ristow et al., 1995). Ethanol toxicity has been metabolism, it has been suggested that ethanol correlated with the production of mitochondrial might induce a very high rate of mtDNA muta- reactive oxygen species (ROS) (see Figure 3.10) tion but that the requirement for functional mito- (Abbott et al., 2009; Costa et al., 1997; Du & chondria would lead to elimination of mutants Takagi, 2007; Gales et al., 2008; Landolfo et al., with nonfunctional mitochondrial genomes 2008; Longo et al., 1996; Piper, 1999; Raha & Rob- (Martıńez et al., 1995). inson, 2000), although data are unavailable on It has also been reported that ethanol causes the effect of ROS on mtDNA. The generation loss of mtDNA (Ibeas & Jimeńez, 1997). Because of superoxide radicals probably causes high ethanol is a membrane solvent, its mutagenic rates of mutagenesis in mtDNA as a result of effect has been attributed to alterations in the oxidative stress, and the tolerance shown by mitochondrial membrane that lead to the loss yeast mitochondria to these mutagenic effects of mtDNA. Nevertheless, both ethanol and acet- could be thanks to their ability to prevent aldehyde cause breaks in chromosomal DNA damage by these radicals (Costa et al., 1997). INFLUENCE OF ENVIRONMENTAL FACTORS ON THE CHARACTERISTICS OF YEASTS INVOLVED 71

Castrejoń et al. (2002) reported that both The preferential mutagenic action of acetalde- ethanol and acetaldehyde induced petite hyde and ethanol on mtDNA is due, on the mutants in populations of flor yeasts and demon- one hand, to the fact that mtDNA is very close strated that the induction mechanism initially to the production site of ROS, and, on the other, involved mtDNA damage. Indeed, mtDNA is to the fact that mtDNA polymerase lacks repair lost after prolonged incubations. Acetaldehyde ability. Flor yeasts possibly have more resistant and ethanol, thus, both cause irreversible mitochondria because of the greater efficiency changes in mtDNA and alter restriction patterns. of mitochondrial manganese superoxide dismu- While these mitochondrial alterations eventu- tase (MnSOD) (see Figure 3.10 and Table 3.4) ally lead to complete DNA loss, in Fino wines, and the enzymes associated with the metabo- mutant populations do not survive due to the lism of GSH and other compounds that protect lack of fermentable carbon sources (Castrejoń against oxidative damage (see Figure 3.11). et al., 2002). Petite mutants nevertheless remain Indeed, SOD genes are expressed much more detectable temporarily. Ibeas et al. (1997) found during the flor phase than during other growth that ethanol and temperature exerted a syner- phases (Castrejoń et al., 2002; Infante et al., gistic effect on the flor layer that led to the forma- 2003). In a study of 176 mutants with an tion of petite mutants (20e30%) (see Figure 3.8). ethanol-sensitivity phenotype (each with a dele- Esteve-Zarzoso et al. (2001), in turn, reported the tion in a different gene), almost all of the genes coexistence of large cells with functional mito- analyzed encoded proteins involved in respira- chondria and petite mutants (a minority) with tion and mitochondrial adenosine triphosphate different mtDNA restriction patterns. Finally, (ATP) synthesis (Kumar et al., 2008). the greatest mtDNA polymorphism has been Ethanol might also induce chromosome loss found in yeasts that generate wine with higher due to its role as a membrane solvent. This levels of acetaldehyde. would explain the high frequency of aneuploidy found in flor yeasts (Guijo et al., 1997; Martıńez

3.2. Influence on Chromosomes As mentioned, ethanol and acetaldehyde TABLE 3.4 Sensitivity to Ethanol in Respiration-deficient Mutants in Cytoplasmic Superoxide have been found to cause breaks in yeast chro- (CuZnSOD), Encoded by SOD1, and in mosomal DNA (Ristow et al., 1995). In Fino Mitochondrial SOD (MnSOD1), Encoded wines with high concentrations of ethanol and by SOD2 acetaldehyde, the chromosomal DNA sequences Viability (%) of flor yeasts in all probability undergo numerous changes due to errors generated Diauxic shift Post-diauxic during the repair of DNA breaks by recombina- 14% 20% 14% 20% tion. This would explain the high frequency of Saccharomyces cerevisiae EtOH EtOH EtOH EtOH chromosomal rearrangements seen in such cases and also the differences in chromosomal aBR10 90 (3) 56 (8) 79 (7) 82 (2) (wild-type) organization between laboratory strains and sod1 83 (10) 44 (15) 87 (11) 79 (8) wine and flor strains (Infante et al., 2003; Pirino et al., 2004; Puig et al., 2000; Zara et al., 2008). sod2 72 (3) 0 76 (3) 9 (3) Large differences in gene order due to trans- 1 locations have also been described in the Believed to play a key role in resistance to ethanol and oxidative stress. Values are means (SD) of five independent experiments. mtDNA of wine yeasts (Cardazzo et al., 1998). Reproduced from Costa et al. (1997). EtOH ¼ ethanol. 72 3. YEASTS USED IN BIOLOGICALLY AGED WINES

FIGURE 3.11 Resistance to oxidative stress in different Saccharomyces cerevisiae mutants with deletion of genes involved in glutathione (GSH) metabolism and response to oxidative stress. Reduced GSH plays an important role in resistance to oxidative stress, mainly in cells in the flor phase. In this assay, cells were taken during exponential or stationary growth phases or during flor formation (mats, or colonies of cells). Growth inhibition assays were performed in Petri dishes following exposure to 35% H2O2. cis2 ¼ g-glutamyl transpeptidase; glr1 ¼ glutathione reductase; gsh1 ¼ g-gluta- mylcysteine ligase; gsh2 ¼ glutathione synthethase; opt1/hgt1 ¼ cell-surface transporter of oligopeptides and GSH; yap1 ¼ stress-induced transcription factor; ycf1 ¼ vacuolar transporter of GSH and GSH-X. Reproduced from Gales et al. (2008). et al., 1995; Mesa et al., 2000; Naumova et al., phospholipids and a decrease in the sterol/ 2005). Finally, the maintenance of aneuploidy protein ratio (van Uden, 1989). Supplementation and chromosomal rearrangements is advanta- with ergosterol and unsaturated fatty acids geous in the conditions that flor yeasts have to increases ethanol tolerance in yeasts, strongly endure. It also plays a key role in sexual isola- indicating that the cell membrane is the main tion, which, in turn, prevents the random distri- target of ethanol-induced toxicity. The transfer bution of favorable characteristics, which is why of mitochondria from strains with high ethanol the majority of flor yeasts do not sporulate. Spor- tolerance to laboratory strains has been shown ulation does, however, increase cell resistance to to considerably increase ethanol tolerance hostile conditions, which explains why a small and thermotolerance in the receptor strains percentage of wine (not flor) yeasts are (Jimeńez & Benı´tez, 1988). The study by Jimeńez apomictic, meaning that they do not complete and Benı´tez also showed loss of ethanol- the first or second meiotic divisions and that induced mitochondrial functions following the two spores in the ascus have an identical transfer, indicating that the mitochondrial genetic structure to that of the parent cells (Cas- genome is partly responsible for tolerance. trejoń et al., 2004). Apomixis is less dependent This tolerance is partly due to the role of aerobic on environmental conditions and therefore metabolism in the biosynthesis of ergosterol and much more common than meiotic sporulation. unsaturated fatty acids, which are essential It always takes place in unfavorable conditions membrane components and key determinants and prevents recombination and the loss of of ethanol tolerance (Sulo et al., 2003; Zara optimal genotypes. et al., 2009). Indeed, there have been reports of considerable improvements in ethanol tolerance 3.3. Influence on the Membrane and during the fermentation of Montilla-Moriles Cell Wall wines following aeration, with direct correlations between oxygen concentrations, cell Yeast cell membranes undergo major changes membrane sterol, and phospholipid content when exposed to high concentrations of ethanol. (Mauricio et al., 1990). The yeasts synthesize lipids enriched in C18:1 to Ethanol also induces the expression of antiox- compensate for the decrease in palmitic acid. idant proteins (Piper, 1999) and proteins that There is also a general increase in the proportion protect against oxidative stress (e.g., SOD, of ergosterol, unsaturated fatty acids, and catalase, or enzymes associated with GSH EVOLUTION OF CELLULAR GENOMES AND YEAST POPULATIONS IN BIOLOGICALLY AGED WINES 73 metabolism) (Abbott et al., 2009; Du & Takagi, conditions. In addition to isolating proteins 2007; Gales et al., 2008; Infante et al., 2003; Land- encoded by the FLO1 and FLO11 genes (Barrales olfo et al., 2008), as indicated previously. Both et al., 2008; Smukalla et al., 2008), studies of flor ethanol and temperature induce the expression yeasts have also uncovered a strongly hydro- of proteins involved in the metabolism of treha- phobic 49 kDa cell surface mannoprotein of as- lose (a protective agent in cell membranes) and yet-undetermined function (Alexandre et al., stress-response proteins such as Hsp104 that 2000). Kovacs et al. (2008) isolated and contribute to thermotolerance and ethanol toler- compared another cell-surface Pir protein, ance. Resistance to osmotic shock, ethanol, cold, Hsp150 (see Figure 3.4), in numerous fermenta- oxidative stress, and acetaldehyde in flor strains tion and film-forming yeast strains from has been associated with the synthesis of heat different geographical areas. It is noteworthy shock proteins other than Hsp104, such as that all the flor strains analyzed were lacking Hsp12, Hsp82, and Hsp26 (Aranda et al., three of the 11 protein repeat regions found in 2002). Hsp12, which confers protection to other strains, but whether or not this deletion membrane liposomes, has also been associated is relevant to flor formation or maintenance is with flor strains (Zara et al., 2002). This protein, not known. Hydrophobin isolation protocols however, does not appear to play a determining have been used to isolate fungal proteins on role in ethanol tolerance, as deletion of the corre- the surface of flor yeasts that protect against sponding gene prevented film formation but desiccation and other stress conditions (Castre- did not increase cell sensitivity to ethanol. joń, 2000). The function of these proteins, As has already been indicated, high concen- however, has yet to be fully elucidated. trations of ethanol also induce the production of adhesins and other hydrophobic cell-surface proteins with internal repeats (Pir) (Huang 4. EVOLUTION OF CELLULAR et al., 2009) (see Figure 3.4), explaining why GENOMES AND YEAST cell hydrophobicity has been found to double POPULATIONS IN BIOLOGICALLY during flor formation (Barrales et al., 2008; AGED WINES Rando & Verstrepen, 2007; van Mulders et al., 2009) (see Figure 3.5). The synthesis of hydro- The chromosomal constitution of a popula- phobic proteins, particularly those encoded by tion of wine yeasts is determined by a combina- FLO genes, leads to increased resistance to tion of the variability introduced by mutation ethanol, temperature, pH, and mechanical stress and meiotic or mitotic recombination and the (Beauvais et al., 2009; Castrejoń, 2000; Smukalla adaptive selection that occurs in specific et al., 2008). Furthermore, it has been seen (in flor fermentation or aging conditions. There is yeasts only) that a deletion in the repression evidence that interspecific hybridization occurs domain of the FLO11 promoter considerably between different Saccharomyces wine strains. increased FLO11 gene expression levels and Examples include brewing strains that are that rearrangements in the repeat domains of hybrids of S. cerevisiae and Saccharomyces pastor- the coding region increased both the number ianus and cider strains with nuclear chromo- and hydrophobicity of the corresponding somes from three different species of proteins (Barrales et al., 2008; Fidalgo et al., Saccharomyces (Belloch et al., 2008; de Barros 2006). Lopes et al., 2002; Gonza´lez et al., 2007; Sipiczki, Many attempts have been made to identify 2008). No evidence of barriers to interspecies other proteins responsible for both flor forma- conjugation has been found. Flor yeasts, tion and increased tolerance of a range of hostile however, have less chromosome polymorphism 74 3. YEASTS USED IN BIOLOGICALLY AGED WINES than other wine yeasts, possibly because they et al., 1996) and the different rows in the have to endure very hostile conditions that dynamic solera and criadera aging system tend to select for a practically unique karyotype (Infante et al., 2003). (Martıńez et al., 1995; Naumova et al., 2005). Chromosomal constitution is thus influenced This relative uniformity is favored by sexual by variations that can occur in the genome of isolation during fermentation and aging (San- a yeast strain during fermentation or aging cho et al., 1986) (which would prevent meiotic and the selection of specific genotypes in certain recombination), a shortage of Ty1 elements environmental conditions. In the case of biolog- (Ibeas & Jimeńez, 1996; Nyswaner et al., 2008), ically aged wines, it has been suggested that poor mobility of Ty2 elements (Rachidi et al., beticus and cheresiensis races predominate in 1999; Wu & Jiang, 2008) (which would notably younger wines because of their ability to form reduce rearrangements and chromosome a flor much more quickly than other races (Mar- changes during mitosis), and frequent aneu- tıńez et al., 1997c). Likewise, it is thought that ploidy (Martıńez et al., 1995) (which would montuliensis and rouxii races predominate in render the majority of meiotic products invi- older wines because of their greater ability to able). The chromosome changes observed in tolerate and produce acetaldehyde (see wine strains during fermentation have been Figure 3.2). As explained previously, in the attributed to mitotic recombination and gene dynamic biological aging system, young wine conversion processes designed to eliminate from the upper rows is added to the older wines potentially deleterious alleles (Carro & Pina, below. Beticus and cheresiensis races would thus 2001; Puig et al., 2000). Cells might use recombi- gain access to these older wines, in which they nation to repair the continual damage caused by would initially proliferate because of their supe- ethanol and acetaldehydedthe main sources of rior ability to repair and re-form the flor. As the variabilitydduring both fermentation and concentrations of acetaldehyde in the wine aging (Benı´tez & Codoń, 2002; Castrejoń et al., increased, those yeasts would be gradually dis- 2002). There have also been reports of rearrange- placed by montuliensis and rouxii races until ments caused by reciprocal chromosomal trans- younger wine was added again (Martıńez location, which in all cases has been found to be et al., 1997c). The association detected in some the result of ectopic recombination between Ty wineries between specific genotypes and elements (Nyswaner et al., 2008; Rachidi et al., different levels in the aging scale has been attrib- 1999; Wu & Jiang, 2008). Several authors have uted to the speed with which the flor forms in identified flor strains with an identical karyo- younger wines and, in older wines, the tolerance type but different mtDNA RFLPs and vice to acetaldehyde (Infante et al., 2003). A compar- versa, and attributed this to conjugation and ison between strains capable of accelerated flor recombination processes between cells (Car- formation and strains with a greater acetalde- dazzo et al., 1998; Ibeas & Jimeńez, 1996). The hyde production capacity revealed copy conditions that develop during fermentation, number differences in 38% of their genes, as and above all during aging, determine whether well as differences in aneuploidy (detected in or not changes in the genome of the new chromosomes I, III, and VI in the former group recombinants will be maintained. This explains and in chromosomes X and XII in the latter the high copy number of certain chromosomes group) (Mesa et al., 1999, 2000) (see Table 3.3). or genes that have been observed in such cases Ethanol tolerance alleles have also been local- (Guijo et al., 1997) (see Table 3.3). It also explains ized to some of these chromosomes (VI, VII, the associations detected between specific geno- IX, and XII) (Hu et al., 2007). In the presence of types and both winemaking conditions (Nadal very dominant phenotypes, a population can GENETIC IMPROVEMENT OF WINE YEASTS 75 be reduced to a single race, or almost even a found in musts fermented with strains that do single genotype if the dominant strain succeeds not express this transporter. in completely displacing the other populations Modulation of the glycerol-to-ethanol ratio (Ibeas et al., 1997). during fermentation has been the focus of recent investigation. Low-alcohol wines can be obtained by increasing glycerol content and 5. GENETIC IMPROVEMENT OF consequently decreasing ethanol content. Glyc- WINE YEASTS erol is also important in wines lacking body as it enhances their sweetness and masks possible 5.1. Improving the Characteristics of acidity. Wine strains capable of producing Fermentation Yeasts more glycerol than ethanol have been created by overexpression of the gene encoding Fermentation yeasts must be able to favor glycerol-3-phosphate dehydrogenase (Michnick a quick onset of fermentation, convert all sugars et al., 1997). The wines produced using these to ethanol, and produce low levels of undesir- strains are characterized not only by greater able byproducts such as volatile acids. Rapid glycerol content but also by small variations in onset of fermentation is important as it elimi- other metabolites (Remize et al., 2003). Similar nates or reduces the contaminants in this phase, designs used with brewing strains have suc- while total consumption of sugars is important ceeded in increasing glycerol concentrations as it prevents contamination during aging, five- to six-fold. Considerable increases in glyc- particularly by strains of Brettanomyces, which erol production have also been seen in response have a major impact on the organoleptic quality to ethanol stress (Vriesekoop et al., 2009). None- of biologically aged wines (Barata et al., 2008; theless, in the case of biologically aged wines, Vigentini et al., 2008). The key elements in the fermentation of musts made with Palomino fermentation are sugar permeases and certain grapes containing high sugar concentrations glycolytic enzymes such as hexokinase 2 (Ber- (particularly when partially dehydrated grapes thels et al., 2008). Over 20 HXT genes have that have been exposed to many hours of sun been implicated in hexose transport in S. cerevi- are used) increases both glycerol levels and siae. HXT2 overexpression, in particular, has volatile acidity. Yeasts synthesize glycerol to been reported to considerably reduce the lag combat osmotic stress (Remize et al., 2003) and phase at the start of fermentation, although acetic acid (responsible for volatile acidity) to this effect has not been observed in the fermen- maintain the redox balance during glycerol tation of musts from Palomino grapes. Because production. Wine yeast strains expressing bacte- most wine strains of S. cerevisiae are more effi- rial genes linked to malolactic fermentation cient transporters of glucose than of fructose, have also been engineered to reduce the acidity the residual sugar is formed by fructose in of wines with a high concentration of malic acid musts with a glucose to fructose ratio of 1:1 at (Williams et al., 1984). the start of fermentation. The expression of Terpenes are characteristic substances found Fsy1, a specific fructose transporter found in in different grape varieties. They are present in S. pastorianus, complements an hxt null mutation sugar-bound forms, which are transported in S. cerevisiae (Rodrigues de Sousa et al., 2004) through the plant, and as free terpenes, which and increases the efficiency with which Palo- have a substantial impact on the organoleptic mino grapes are fermented. The resulting properties of wine. Wine yeasts expressing residual sugar level in such cases is under heterologous genes encoding hydrolases 0.5 g/L, contrasting with the level of 2 to 5 g/L capable of breaking the glycosidic bonds 76 3. YEASTS USED IN BIOLOGICALLY AGED WINES between sugars and terpenes have also been wine (Hernańdez-Orte et al., 2002). Strains that obtained, as have yeasts that hydrolyze plant overproduce leucine have been successfully cell materials that would otherwise retain used to increase levels of isobutyl and isoamyl aromas produced by the vine, resulting thus in alcohol (Watanabe et al., 1990); likewise, phenyl- more aromatic wines (Villanueva et al., 2000). ethyl alcohol levels have been increased using Constructs in which heterologous genes are strains that overproduce aromatic amino acids under the control of regulatable promoters (Fukuda et al., 1990, 1991a, 1991b). The alcohol have been used to control the release of aromas acetyltransferase gene family has been impli- at the end of fermentation or at any other cated in the synthesis of esters during alcoholic desired time (Puig & Perez-Ortıń, 2000). None- fermentation (Mason & Dufour, 2000), and, as theless, in one study, the addition of enzyme a consequence, attempts have been made to preparations during the fermentation of musts increase alcohol activity transferase and hence from Palomino grapes did not improve the increase ester levels while reducing those of organoleptic quality of the resulting wines (Rol- alcohols. It has been reported that acetate ester dań et al., 2006). This indicates the scarcity of concentration is increased by increasing alcohol primary aromas in this variety of grape and acetyltransferase activity in yeast strains used to suggests that the sensory properties of biologi- make sake (Fukuda et al., 1998), beer (Fujii et al., cally aged wines are mainly derived from 1994; Verstrepen et al., 2003a, 2003b), and wine secondary and tertiary aromas generated by and spirits (Dequin, 2001; Lilly et al., 2000). yeast metabolism (Garcıá Maiquez, 1995; Rol- The organoleptic properties of wine could also dań et al., 2006). be improved through the manipulation of In 1989, musts from the French Midi region Adh6 and Adh7 (alcohol dehydrogenase inoculated with Champagne or Burgundy enzymes involved in the synthesis of higher yeasts produced wines with aromas reminiscent alcohols) (de Smidt et al., 2008) or the Pdr12 of the wines from those regions (Sua´rez-Lepe, transporter (involved in the secretion of fusel 2002). These aromas are derived from the acids) (Hazelwood et al., 2006). production of higher alcohols, esters, fatty acids, Finally, efforts are also being made in the area aldehydes, sulfur compounds, phenols, and of “healthy” wines. The phytoalexin resveratrol, terpenes by yeasts. All of these compounds, a member of the stilbene family, has been iden- but particularly esters and alcohols, are formed tified in numerous plants, including vines. In as secondary metabolites during glycerol- Palomino grapes, for example, levels of this pyruvic fermentation (Cordente et al., 2009; phenolic compound vary from 2 to 7 mg/L in Linderholm et al., 2008; Sua´rez-Lepe, 2002). the juice, and from 14 to 64 mg/L in the skin Considering that yeasts use some amino acids (Roldań et al., 2003). These variations are related as precursors of higher alcohols and esters, it to numerous factors such as climate and the might be possible to enhance the fruity aroma health of the grapes. Resveratrol is also found of wine by increasing amino acid content (Car- in wine and has been associated with beneficial rau et al., 2008; Munõz et al., 2006; Thibon health effects in the areas of cancer and heart et al., 2008; Torrea-Gon˜i & Ancıń-Azpilicueta, disease (Gimeńez-Garcıá, 2000). Growing efforts 2001). The addition of amino acid solutions are thus being made to increase resveratrol (particularly threonine and serine) to must concentrations in wine by improving extraction made from 11 varieties of grape has been found of this compound from grapes, analyzing to increase the concentration of higher alcohols, factors that influence resveratrol levels, and some of their acetates, ethyl butyrate, and some inducing the expression of stilbene synthases acids, as well as the aromatic character of the in wine strains (Becker et al., 2003). GENETIC IMPROVEMENT OF WINE YEASTS 77

5.2. Improving the Characteristics of gradient, exert considerable oxidative stress, Aging Yeasts and are mutagens for mtDNA (Piper, 1999). Another of these proteins, Flo11 (described in The synergistic action of ethanol, acetalde- Section 3.3), is responsible for invasive growth hyde, low pH, water stress, oxidative stress, (van Dyk et al., 2005), filamentation (Ma et al., and temperature can contribute to the loss of 2007; Palecek et al., 2000; Tamaki et al., 2000), the flor film (Martıńez et al., 1997c) (see flocculation (Guo et al., 2000), and flor formation Figure 3.8). The use of strains capable of forming (Guo et al., 2000; Ishigami et al., 2006; Reynolds a stronger film in less time would thus enhance & Fink, 2001) (see Figure 3.3). It also appears to the quality of Fino wines and shorten the aging be responsible for the greater resistance shown time required. While hybrids of nonisogenic by cells to temperature variations, high alcohol strains capable of forming a flor faster than levels, low pH levels, and mechanical stress parental strains have been developed (Castre- during the flor formation phase (Barrales et al., joń, 2000), they do not confer the same organo- 2008; Castrejoń, 2000). Its overexpression might, leptic properties as their progenitors. thus, contribute to the formation of more stable Both flor formation and resistance to hostile flors in a shorter time. conditions are largely determined by the Another class of cell-surface protein involved proteins on the cell surface of the yeasts. in resistance to stress is highly glycosylated and Changes in external pH induce the synthesis attached to the cell wall via disulfide bridges of GPI-linked proteins anchored to the (Caridi, 2006). Thus, lower glycosylation is asso- membrane (Barrales et al., 2008) (see Figure 3.4). ciated with greater sensitivity to hostile condi- Many of these proteins are dependent on the tions. An example of such a protein is Hsp150, high-osmolarity glycerol response (osmotic which is strongly induced by osmotic stress or stress) (see Figure 3.7), which increases yeast low pH (Moukadiri & Zueco, 2001) and gives resistance to enzymatic lysis (Barrales et al., rise to increased stress resistance. Attempts to 2008). Sedi1, another GPI-linked membrane correlate differential expression of other HSP protein, is synthesized only in the stationary genes (mainly HSP12, HSP123, HSP82, HSP26, phase, and its absence increases cell sensitivity and HSP104) with resistance to cold, osmotic to enzymatic lysis. Reynolds and Fink (2001) stress, oxidative stress, and acetaldehyde- or described a family of GPI-linked membrane ethanol-induced stress in flor strains have not glycoproteins similar to adhesins that were succeeded in confirming that these genes are present in the yeast cell wall. The corresponding directly involved in either flor formation or genes are expressed in carbon- or nitrogen-star- resistance to hostile conditions (Aranda et al., vation conditions and the main function of the 2002; Zara et al., 2002). Products of the HSP70 proteins is to adhere to inert surfaces or other gene family, however, in association with other cells. An example of one of these proteins is factors but independently of Flo11, have been that encoded by the SPI1 gene, regulated by directly implicated in flor formation (Martineau Msn2 and Msn4 (Puig & Perez-Ortıń, 2000) et al., 2007). Hsp70 proteins are chaperones (see Figures 3.7 and 3.9). SPI1 confers resistance involved in protein folding and transport to 2,4 D and b-1,3-glucanase and controls the pH through the endoplasmic reticulum and above gradient across the membrane in adverse condi- all through the mitochondrial membrane. tions. Overexpression of this protein could High expression levels of other genes associ- improve a strain’s ability to form a film that is ated with ethanol metabolism, redox potential, resistant to hostile conditions because weak glycerol uptake, and oxidative stress have also organic acids alter the transmembrane proton been observed in flor yeasts (Infante et al., 78 3. YEASTS USED IN BIOLOGICALLY AGED WINES

2003; Longo et al., 1996). Ethanol and acetalde- aged wines with desirable characteristics such hyde toxicity have been associated with the as high levels of acetaldehyde and higher alco- production of mitochondrial ROS (Landolfo hols, and low volatile acidity. Furthermore, et al., 2008) (see Figure 3.10), leading to the accu- research efforts have uncovered genes that mulation of mutations and deletions in mtDNA appear to be involved in flor formation (FLO11 (Raha & Robinson, 2000). Enzymes such as gene) and tolerance of the hostile conditions N-acetyltransferase (Du & Takagi, 2007), cata- that characterize aging (SOD genes). The use lase (Abbott et al., 2009), and above all SOD of yeasts overexpressing these genes would (Costa et al., 1997), in contrast, protect against allow more stable films to be formed and this toxicity (see Table 3.4 and Figure 3.10). shorten the aging period required. Finally, the Resistance to oxidative stress has also been development of fermentation strains that over- found to be increased by proteins associated produce amino acids and produce higher levels with GSH metabolism (Gales et al., 2008) (see of alcohol transferase, or that express hydrolases Figure 3.11). Improved strains could also be or stilbene synthase, could give rise to more obtained by overexpressing the SOD enzymes aromatic, healthier young wines. Sod1 and Sod2, among others. The role of these two enzymes in ethanol tolerance, for example, Acknowledgments has been well established (Costa et al., 1997) (see Figure 3.10 and Table 3.4) and the SOD1 During the writing of this chapter, our work was supported gene is strongly expressed in the stationary by funds from the Spanish Ministry for Science and Innova- tion (project AGL2006-03947), the Autonomous Government phase and above all in the flor phase (Infante of Andalusia, Spain (Proyecto de Excelencia, PO6-CVI- et al., 2003). Finally, because the Sod1 protein 01646), the Interministerial Committee for Scientific and is involved in molecular crosslinking, it has Technology (CICYT) (Proyecto TRACE, PET 2008_0283), been possible to purify this protein in the flor and the company Bean Global, SA, in Jerez de la Frontera. using a purification protocol that favors aggregation at the liquideair interface (Castrejoń, References 2000). Abbott, D. A., Suir, E., Duong, G. H., de Hulster, E., Pronk, J. T., & van Maris, A. J. (2009). Catalase overexpression reduces lactic acid-induced oxidative stress 6. CONCLUSIONS in Saccharomyces cerevisiae. Appl. Environ. Microbiol., 75, 2320e2325. Alexandre, H., Blanchet, S., & Charpentier, C. (2000). Iden- S. cerevisiae strains that participate in the production of biologically aged wines have tification of a 49-kDa hydrophobic cell wall mannoprotein present in velum yeast which may be a highly heterogeneous genetic profile. They implicated in velum formation. FEMS Microbiol. Lett., are easily distinguishable from each other 185, 147e150. because of the enormous differences in their Aranda, A., Querol, A., & del Olmo, M. (2002). Correlation DNA content, mtDNA restriction profiles, and between acetaldehyde and ethanol resistance and chromosomal patterns. Strains isolated in expression of HSP genes in yeast strains isolated during the biological aging of sherry wines. Arch. Microbiol., 177, different wineries or even at different levels of 304e312. aging scales within the same winery also Backhus, L. E., DeRisi, J., & Bisson, L. F. (2001). Functional display genetic variability. Exploitation of genomic analysis of a commercial wine strain of under differing nitrogen condi- differences between S. cerevisiae races or strains Saccharomyces cerevisiae with varying genetic profiles associated with tions. FEMS Yeast Res., 1, 111e125. Bakalinsky, A. T., & Snow, R. (1990). The chromosomal specific metabolic characteristics will allow tar- constitution of wine strains of Saccharomyces cerevisiae. geted strategies aimed at producing biologically Yeast, 6, 367e382. REFERENCES 79

Bandas, E. L., & Zakharov, I. A. (1980). Induction of rho- Bravo-Abad, F. (1986). Crianza biolo´gica de vino: Procedi- mutations in yeast Saccharomyces cerevisiae by ethanol. miento tradicional de vinos finos de D.O. Jerez y D.O. Mutat. Res., 71, 193e199. Montilla-Moriles. Enologıá y Enotecnia, 1,15e19. Barata, A., Gonza´lez, S., Malfeito-Ferreira, M., Querol, A., & Budroni, M., Roggio, T., Pinna, G., Pretti, L., & Farris, G. A. Loureiro, V. (2008). Sour rot-damaged grapes are sources (1995). Which factors favour the formation of a biofilm of wine spoilage yeasts. FEMS Yeast Res., 8, 1008e1017. by Saccharomyces cerevisiae.InXII International Conference Barrales, R. R., Jimeńez, J., & Ibeas, J. I. (2008). Identification on Yeast Genetics and Molecular Biology. Lisbon, Portugal: of novel activation mechanisms for FLO11 regulation in J.W.S. Ltd. June 10e16. Saccharomyces cerevisiae. Genetics, 178, 145e156. Budroni, M., Zara, S., Zara, G., Pirino, G., & Mannazzu, I. Beauvais, A., Loussert, C., Prevost, M. C., Verstrepen, K., & (2005). Peculiarities of flor strains adapted to Sardinian Latge, J. P. (2009). Characterization of a biofilm-like sherry-like wine ageing conditions. FEMS Yeast Res., 5, extracellular matrix in FLO1-expressing Saccharomyces 951e958. cerevisiae cells. FEMS Yeast Res., 9,411e419. Campo, E., Cacho, J., & Ferreira, V. (2008). The chemical Becker, J. V., Armstrong, G. O., van der Merwe, M. J., characterization of the aroma of dessert and sparkling Lambrechts, M. G., Vivier, M. A., & Pretorius, I. S. (2003). white wines (Pedro Ximenez, Fino, Sauternes, and Cava) Metabolic engineering of Saccharomyces cerevisiae for the by gas chromatography-olfactometry and chemical synthesis of the wine-related antioxidant resveratrol. quantitative analysis. J. Agric. Food Chem., 56, 2477e2484. FEMS Yeast Res., 4,79e85. Cardazzo, B., Minuzzo, S., Sartori, G., Grapputo, A., & Belloch, C., Orlic, S., Barrio, E., & Querol, A. (2008). Carignani, G. (1998). Evolution of mitochondrial DNA in Fermentative stress adaptation of hybrids within the yeast: Gene order and structural organization of the Saccharomyces sensu stricto complex. Int. J. Food mitochondrial genome of Saccharomyces uvarum. Curr. Microbiol., 122, 188e195. Genet., 33,52e59. Benı´tez, T., & Codoń, A. C. (2002). Genetic diversity of Caridi, A. (2006). Enological functions of parietal yeast yeasts in wine production. In G. G. Khachatourians & mannoproteins. Anton. Leeuw., 89, 417e422. D. K. Arora (Eds.), Applied mycology and biotechnology, Carrau, F. M., Medina, K., Farina, L., Boido, E., Vol. 2 (pp. 19e44). Amsterdam, the Netherlands: Henschke, P. A., & Dellacassa, E. (2008). Production of Elsevier Science B.V. fermentation aroma compounds by Saccharomyces cer- Benı´tez, T., & Codoń, A. C. (2005). Levaduras. Saccharomyces evisiae wine yeasts: Effects of yeast assimilable nitrogen III. Levaduras de vinos de crianza biolo´gica. In on two model strains. FEMS Yeast Res., 8, 1196e1207. A. V. Carrascosa, R. Munõz, & R. Gońza´lez (Eds.), Carro, D., & Pina, B. (2001). Genetic analysis of the karyo- Microbiologıá del vino (pp. 78e113). Madrid, Spain: AMV type instability in natural wine yeast strains. Yeast, 18, Ediciones. 1457e1470. Benı´tez, T., Martıńez, P., & Codoń, A. C. (1996). Genetic Castrejoń, F. (2000). Mejora de cepas de Saccharomyces cer- constitution of industrial yeast. Microbiologia, 12, evisiae utilizables en la elaboracioń de vinos de Jerez. In 371e384. Gene´tica. Seville, Spain: University of Seville. Benı´tez, T., Rincoń, A.M., & Codoń, A.C. (in press). Yeasts Castrejoń, F., Codoń, A. C., Cubero, B., & Benı´tez, T. (2002). associated with production of fortified wines. In A. Acetaldehyde and ethanol are responsible for mito- Romano & G. H. Fleet (Eds.), Yeasts in the production of chondrial DNA (mtDNA) restriction fragment length wines. Springer. polymorphism (RFLP) in flor yeasts. Syst. Appl. Micro- Berlanga, T. M., Atanasio, C., Mauricio, J. C., & Ortega, J. M. biol., 25, 462e467. (2001). Influence of aeration on the physiological activity Castrejoń, F., Martıńez-Force, E., Benı´tez, T., & Codoń, A. C. of flor yeasts. J. Agric. Food Chem., 49, 3378e3384. (2004). Genetic analysis of apomictic wine yeasts. Curr. Berthels, N. J., Cordero Otero, R. R., Bauer, F. F., Genet., 45, 187e196. Pretorius, I. S., & Thevelein, J. M. (2008). Correlation Celestino-Pere´z, S. (Ed.) (1999). El vino en la antigu¨ edad between glucose/fructose discrepancy and hexokinase romana. In Simposio Arqueologıá del vino, Jerez. Consejo kinetic properties in different Saccharomyces cerevisiae Regulador de las Denominaciones de Origen Jerez-Xeres- wine yeast strains. Appl. Microbiol. Biotechnol., 77, Sherry y manzanilla-San-lućar de Barrameda y Universidad 1083e1091. Autońoma de Madrid, Madrid. October 2e4. Blake, D., Luke, B., Kanellis, P., Jorgensen, P., Goh, T., Cordente, A. G., Heinrich, A., Pretorius, I. S., & Penfold, S., et al. (2006). The F-box protein Dia2 over- Swiegers, J. H. (2009). Isolation of sulfite reductase comes replication impedance to promote genome variants of a commercial wine yeast with significantly stability in Saccharomyces cerevisiae. Genetics, 174, reduced hydrogen sulfide production. FEMS Yeast Res., 1709e1727. 9, 446e459. 80 3. YEASTS USED IN BIOLOGICALLY AGED WINES

Costa, V., Amorim, M. A., Reis, E., Quintanilha, A., & Fleet, G. (2007). Wine. In M. Doyle, & L. Beuchat (Eds.), Food Moradas-Ferreira, P. (1997). Mitochondrial superoxide microbiology: Fundamentals and frontiers (pp. 863e890). dismutase is essential for ethanol tolerance of Saccharo- Washington, DC: ASM Press. myces cerevisiae in the post-diauxic phase. Microbiology, Fujii, T., Nagasawa, N., Iwamatsu, A., Bogaki, T., Tamai, Y., & 143, 1649e1656. Hamachi, M. (1994). Molecular cloning, sequence anal- Charpentier, C., Colin, A., Alais, A., & Legras, J. L. (2009). ysis, and expression of the yeast alcohol acetyltransferase French Jura flor yeasts: Genotype and technological gene. Appl. Environ. Microbiol., 60, 2786e2792. diversity. Anton. Leeuw., 95, 263e273. Fukuda, K., Watanabe, M., Asano, K., Ouchi, K., & de Barros Lopes, M., Bellon, J. R., Shirley, N. J., & Takasawa, S. (1991a). Isolation and genetic study of Ganter, P. F. (2002). Evidence for multiple interspecific p-fluoro-DL-phenylalanine-resistant mutants over- hybridization in Saccharomyces sensu stricto species. producing beta-phenethyl-alcohol in Saccharomyces FEMS Yeast Res., 1, 323e331. cerevisiae. Curr. Genet., 20, 449e452. de Smidt, O., du Preez, J. C., & Albertyn, J. (2008). The Fukuda, K., Watanabe, M., Asano, K., Ouchi, K., & alcohol dehydrogenases of Saccharomyces cerevisiae:A Takasawa, S. (1991b). A mutated ARO4 gene for comprehensive review. FEMS Yeast Res., 8, 967e978. feedback-resistant DAHP synthase which causes both Dequin, S. (2001). The potential of genetic engineering for o-fluoro-DL-phenylalanine resistance and beta-phe- improving brewing, wine-making and baking yeasts. nethyl-alcohol overproduction in Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol., 56, 577e588. Curr. Genet., 20, 453e456. Douglas, L. M., Li, L., Yang, Y., & Dranginis, A. M. (2007). Fukuda, K., Watanabe, M., Asano, K., Ueda, H., & Ohta, S. Expression and characterization of the flocculin Flo11/ (1990). Breeding of brewing yeast producing a large Muc1, a Saccharomyces cerevisiae mannoprotein with homo- amount of b-phenylethyl alcohol and b-phenylethyl typic properties of adhesion. Eukaryot Cell, 6, 2214e2221. acetate. Agric. Biol. Chem., 54, 269e271. Du, X., & Takagi, H. (2007). N-Acetyltransferase Mpr1 Fukuda, K., Yamamoto, N., Kiyokawa, Y., Yanagiuchi, T., confers ethanol tolerance on Saccharomyces cerevisiae by Wakai, Y., Kitamoto, K., et al. (1998). Balance of activities reducing reactive oxygen species. Appl. Microbiol. Bio- of alcohol acetyltransferase and esterase in Saccharomyces technol., 75, 1343e1351. cerevisiae is important for production of isoamyl acetate. Esteve-Zarzoso, B., Fernańdez-Espinar, M. T., & Querol, A. Appl. Environ. Microbiol., 64, 4076e4078. (2004). Authentication and identification of Saccharo- Gales, G., Penninckx, M., Block, J. C., & Leroy, P. (2008). Role myces cerevisiae "flor" yeast races involved in sherry of glutathione metabolism status in the definition of ageing. Anton. Leeuw., 85, 151e158. some cellular parameters and oxidative stress tolerance Esteve-Zarzoso, B., Manzanares, P., Ramoń, D., & Querol, A. of Saccharomyces cerevisiae cells growing as biofilms. (1998). The role of non- Saccharomyces yeasts in industrial FEMS Yeast Res., 8, 667e675. winemaking. Int. Microbiol., 1, 143e148. Garcıá del Barrio, I. (1995). Las bodegas del vino de Jerez Esteve-Zarzoso, B., Peris-Torań, M. J., Garcıá Maiquez, E., (Historia, microclima y construccioń). In J. J. Iglesias- Uruburu, F., & Querol, A. (2001). Yeast population Rodrı´guez (Ed.), Historia y cultura del vino en Andalucıá dynamics during the fermentation and biological aging (pp. 141e178). Sevilla, Spain: Secretariado de Pub- of sherry wines. Appl. Environ. Microbiol., 67, 2056e2061. licaciones de la Universidad de Sevilla. Fatichenti, F., Farris, G. A., & Deiana, P. (1983). Improved Garcıá Maiquez, E. (1995). Sherry wine microorganisms. production of a Spanish-type sherry by using selected Microbiologia, 11,51e58. indigenous film-forming yeasts as starters. Am. J. Enol. Gimeńez-Garcıá, J. M. (2000). Contenido en resveratrol en la Vitic., 34, 216e220. uva de variedad Bobal. In Uva y Vino: Los valores salud- Fernańdez-Espinar, M. T., Barrio, E., & Querol, A. (2003). ables, (pp. 49e54). Valencia, Spain: Fundacioń Valenciana Analysis of the genetic variability in the species of the de Estudios Avanzados and Generalitat Valenciana, Saccharomyces sensu stricto complex. Yeast, 20, 1213e1226. Consellerıá de Agricultura, Pesca y Alimentacioń. Ferna´dez-Espinar, M. T., Esteve-Zarzoso, B., Querol, A., & Go´mez, M. E., Igartuburu, J. M., Pando, E., Luis, F. R., & Barrio, E. (2000). RFLP analysis of the ribosomal internal Mourente, G. (2004). Lipid composition of lees from transcribed spacers and the 5.8S rRNA gene region of Sherry wine. J. Agric. Food. Chem., 52, 4791e4794. the genus Saccharomyces: A fast method for species Gonza´lez, S. S., Barrio, E., & Querol, A. (2007). Molecular identification and the differentiation of flor yeasts. identification and characterization of wine yeasts iso- Anton. Leeuw., 78,87e97. lated from Tenerife (Canary Island, Spain). J. Appl. Fidalgo, M., Barrales, R. R., Ibeas, J. I., & Jimeńez, J. (2006). Microbiol., 102, 1018e1025. Adaptive evolution by mutations in the FLO11 gene. Gonza´lez, S. S., Barrio, E., & Querol, A. (2008). Molecular Proc. Natl. Acad. Sci. U.S.A., 103, 11228e11233. characterization of new natural hybrids of Saccharomyces REFERENCES 81

cerevisiae and S. kudriavzevii in brewing. Appl. Environ. caused by cell surface hydrophobicity in wild-type flor Microbiol., 74, 2314e2320. yeast. Biosci. Biotechnol. Biochem., 70, 660e666. Guijo, S., Mauricio, J. C., Salmon, J. M., & Ortega, J. M. Jimeńez, J., & Benı´tez, T. (1988). Yeast cell viability under (1997). Determination of the relative ploidy in different conditions of high temperature and ethanol concentra- Saccharomyces cerevisiae strains used for fermentation and tions depends on the mitochondrial genome. Curr. “flor” film ageing of dry sherry-type wines. Yeast, 13, Genet., 13, 461e469. 101e117. Kawarai, T., Furukawa, S., Ogihara, H., & Yamasaki, M. Guo, B., Styles, C. A., Feng, Q., & Fink, G. R. (2000). A (2007). Mixed-species biofilm formation by lactic acid Saccharomyces gene family involved in invasive growth, bacteria and rice wine yeasts. Appl. Environ. Microbiol., cellecell adhesion, and mating. Proc. Natl. Acad. Sci. 73, 4673e4676. U.S.A., 97, 12 158e12 163. Kovacs, M., Stuparevic, I., Mrsa, V., & Maraz, A. (2008). Hazelwood, L. A., Tai, S. L., Boer, V. M., de Winde, J. H., Characterization of Ccw7p cell wall proteins and the Pronk, J. T., & Daran, J. M. (2006). A new physiolog- encoding genes of Saccharomyces cerevisiae wine yeast ical role for Pdr12p in Saccharomyces cerevisiae: Export strains: Relevance for flor formation. FEMS Yeast Res., 8, of aromatic and branched-chain organic acids 1115e1126. produced in amino acid catabolism. FEMS Yeast Res., Kumar, G. R., Goyashiki, R., Tramakrishnan, V., Karpel, J. E., 6, 937e945. & Bisson, L. F. (2008). Genes required for ethanol toler- Hernańdez-Orte, P., Cacho, J. F., & Ferreira, V. (2002). ance and utilization in Saccharomyces cerevisiae. Am. J. Relationship between varietal amino acid profile of Enol. Vitic., 59, 401e411. grapes and wine aromatic composition. Experiments Lambrechts, M. G., Bauer, F. F., Marmur, J., & with model solutions and chemometric study. J. Agric. Pretorius, I. S. (1996). Muc1, a mucin-like protein that is Food Chem., 50, 2891e2899. regulated by Mss10, is critical for pseudohyphal Hu, X. H., Wang, M. H., Tan, T., Li, J. R., Yang, H., Leach, L., differentiation in yeast. Proc. Natl. Acad. Sci. U.S.A., 93, et al. (2007). Genetic dissection of ethanol tolerance in 8419e8424. the budding yeast Saccharomyces cerevisiae. Genetics, 175, Landolfo, S., Politi, H., Angelozzi, D., & Mannazzu, I. 1479e1487. (2008). ROS accumulation and oxidative damage to cell Huang, G., Dougherty, S. D., & Erdman, S. E. (2009). structures in Saccharomyces cerevisiae wine strains during Conserved WCPL and CX4C domains mediate several fermentation of high-sugar-containing medium. Biochim. mating adhesin interactions in Saccharomyces cerevisiae. Biophys. Acta, 1780, 892e898. Genetics, 182, 173e189. Lilly, M., Lambrechts, M. G., & Pretorius, I. S. (2000). Effect Ibeas, J. I., & Jimeńez, J. (1996). Genomic complexity and of increased yeast alcohol acetyltransferase activity on chromosomal rearrangements in wine-laboratory yeast flavor profiles of wine and distillates. Appl. Environ. hybrids. Curr. Genet., 30, 410e416. Microbiol., 66, 744e753. Ibeas, J. I., & Jimeńez, J. (1997). Mitochondrial DNA loss Linderholm, A. L., Findleton, C. L., Kumar, G., Hong, Y., & caused by ethanol in Saccharomyces flor yeasts. Appl. Bisson, L. F. (2008). Identification of genes affecting Environ. Microbiol., 63,7e12. hydrogen sulfide formation in Saccharomyces cerevisiae. Ibeas, J. I., Lozano, I., Perdigones, F., & Jimeńez, J. (1996). Appl. Environ. Microbiol., 74, 1418e1427. Detection of Dekkera- Brettanomyces strains in sherry by Lo, W. S., & Dranginis, A. M. (1996). FLO11, a yeast gene a nested PCR method. Appl. Environ. Microbiol., 62, related to the STA genes, encodes a novel cell surface 998e1003. flocculin. J. Bacteriol., 178, 7144e7151. Ibeas, J. I., Lozano, I., Perdigones, F., & Jimeńez, J. (1997). Longo, V. D., Gralla, E. B., & Valentine, J. S. (1996). Super- Effects of ethanol and temperature on the biological oxide dismutase activity is essential for stationary phase aging of Sherry wines. Am. J. Enol. Vitic., 48,71e74. survival in Saccharomyces cerevisiae. Mitochondrial Infante, J. J., Dombek, K. M., Rebordinos, L., Cantoral, J. M., production of toxic oxygen species in vivo. J. Biol. Chem., & Young, E. T. (2003). Genome-wide amplifications 271, 12 275e12 280. caused by chromosomal rearrangements play a major Ma, J., Jin, R., Jia, X., Dobry, C. J., Wang, L., Reggiori, F., et al. role in the adaptive evolution of natural yeast. Genetics, (2007). An interrelationship between autophagy and 165, 1745e1759. filamentous growth in budding yeast. Genetics, 177, Ingledew, W. M., Magnus, C. A., & Sosulski, F. W. (1987). 205e214. Influence of oxygen on proline utilization during the Martineau, C. N., Beckerich, J. M., & Kabani, M. (2007). wine fermentation. Am. J. Enol. Vitic., 38, 246e248. Flo11p-independent control of "mat" formation by hsp70 Ishigami, M., Nakagawa, Y., Hayakawa, M., & Iimura, Y. molecular chaperones and nucleotide exchange factors (2006). FLO11 is the primary factor in flor formation in yeast. Genetics, 177, 1679e1689. 82 3. YEASTS USED IN BIOLOGICALLY AGED WINES

Martıńez, P., Codoń, A. C., Pe´rez, L., & Benı´tez, T. (1995). Mesa, J. J., Infante, J. J., Rebordinos, L., & Cantoral, J. M. Physiological and molecular characterization of flor (1999). Characterization of yeasts involved in the bio- yeasts: Polymorphism of flor yeast populations. Yeast, 11, logical aging of sherry wines. Lebensm. Wiss. Technol., 32, 1399e1411. 114e120. Martıńez, P., Pe´rez, L., & Benı´tez, T. (1997a). Evolution of Mesa, J. J., Infante, J. J., Rebordinos, L., Sańchez, J. A., & flor yeast population during the biological aging of fino Cantoral, J. M. (2000). Influence of the yeast genotypes sherry wine. Am. J. Enol. Vitic., 48, 160e168. on enological characteristics of sherry wines. Am. J. Enol. Martıńez, P., Pe´rez, L., & Benı´tez, T. (1997b). Factors which Vitic., 51,15e21. affect velum formation by flor yeast isolated from sherry Michnick, S., Roustan, J. L., Remize, F., Barre, P., & wine. Syst. Appl. Microbiol., 20, 154e157. Dequin, S. (1997). Modulation of glycerol and ethanol Martıńez, P., Pe´rez, L., & Benı´tez, T. (1997c). Velum forma- yields during alcoholic fermentation in Saccharomyces tion by flor yeasts isolated from sherry wine. Am. J. Enol. cerevisiae strains overexpressed or disrupted for GPD1 Vitic., 48,55e62. encoding glycerol 3-phosphate dehydrogenase. Yeast, 13, Martıńez, P., Valca´rcel, M. J., Gonza´lez, P., Benı´tez, T., & 783e793. Pe´rez, L. (1993). Consumo de etanol, glicerina y amino- Mortimer, R. K. (2000). Evolution and variation of the yeast aćidos totales en vinos finos durante la crianza biolo´gica (Saccharomyces) genome. Genome Res., 10, 403e409. bajo “velo de flor.” Alimentacioń: Equipos y tecnologıá, 2, Moukadiri, I., & Zueco, J. (2001). Evidence for the attach- 61e65. ment of Hsp150/Pir2 to the cell wall of Saccharomyces Martıńez, P., Valca´rcel, M. J., Pe´rez, L., & Benı´tez, T. (1998). cerevisiae through disulfide bridges. FEMS Yeast Res., 1, Metabolism of Saccharomyces cerevisiae flor yeasts during 241e245. fermentation and biological aging of fino sherry: By- Munõz, D., Peinado, R. A., Medina, M., & Moreno, J. (2006). products and aroma compounds. Am. J. Enol. Vitic., 49, Higher alcohols concentration and its relation with the 240e250. biological aging evolution. Eur. Food Res. Technol., 222, Martıńez de la Ossa, E., Caro, I., Bonat, M., Pe´rez, L., & 629e635. Domecq, B. (1987a). Dry extract in sherry and its evolution Nadal, D., Colomer, B., & Pina, B. (1996). Molecular poly- in the aging process. Am. J. Enol. Vitic., 38, 321e325. morphism distribution in phenotypically distinct pop- Martıńez de la Ossa, E., Pe´rez, L., & Caro, I. (1987b). Vari- ulations of wine yeast strains. Appl. Environ. Microbiol., ation of the major volatiles trough aging of sherry. Am. J. 62, 1944e1950. Enol. Vitic., 38, 293e297. Naumov, G. I., Masneuf, I., Naumova, E. S., Aigle, M., & Martıńez-Llopis, M., Mijares, M. I., & Chirbes, R. (1992). Dubourdieu, D. (2000). Association of Saccharomyces Historia del vino espanõl. Madrid, Spain: Vinoseleccioń. bayanus var. uvarum with some French wines: Genetic Martorell, P., Querol, A., & Fernańdez-Espinar, M. T. (2005). analysis of yeast populations. Res. Microbiol., 151, Rapid identification and enumeration of Saccharomyces 683e691. cerevisiae cells in wine by real-time PCR. Appl. Environ. Naumov, G. I., Naumova, E. S., Antunovics, Z., & Microbiol., 71, 6823e6830. Sipiczki, M. (2002). Saccharomyces bayanus var. uvarum in Mason, A. B., & Dufour, J. P. (2000). Alcohol acetyl- Tokaj wine-making of Slovakia and Hungary. Appl. transferases and the significance of ester synthesis in Microbiol. Biotechnol., 59, 727e730. yeast. Yeast, 16, 1287e1298. Naumova, E. S., Ivannikova Iu, V., & Naumov, G. I. (2005). Mauricio, J. C., Arroyo, M., Millań, C., & Ortega, J. M. Genetic differentiation of the sherry yeasts Saccharomyces (1990). Relationship between the phospholipid and cerevisiae. Prikl. Biokhim. Mikrobiol., 41, 656e661. sterol contents in Saccharomyces cerevisiae and Torulaspora Nyswaner, K. M., Checkley, M. A., Yi, M., Stephens, R. M., & delbrueckii, and their permanence during fermentation of Garfinkel, D. J. (2008). Chromatin-associated genes musts from grapes of the Pedro Ximenez variety. Bio- protect the yeast genome from Ty1 insertional muta- technol. Lett., 12, 265e270. genesis. Genetics, 178, 197e214. Mauricio, J. C., Valero, E., Millań, C., & Ortega, J. M. (2001). Palecek, S. P., Parikh, A. S., & Kron, S. J. (2000). Genetic Changes in nitrogen compounds in must and wine analysis reveals that FLO11 upregulation and cell during fermentation and biological aging by flor yeasts. polarization independently regulate invasive growth in J. Agric. Food Chem., 49, 3310e3315. Saccharomyces cerevisiae. Genetics, 156, 1005e1023. Me´rida, J., Lo´pez-Toledano, A., Ma´rquez, T., Millań, C., Peinado, R. A., Moreno, J., Medina, M., & Mauricio, J. C. Ortega, J. M., & Medina, M. (2005). Retention of (2004). Changes in volatile compounds and aromatic browning compounds by yeasts involved in the wine- series in sherry wine with high gluconic acid levels making of sherry type wines. Biotechnol. Lett., 27, subjected to aging by submerged flor yeast cultures. 1565e1570. Biotechnol. Lett., 26, 757e762. REFERENCES 83

Piper, P. W. (1999). Yeast superoxide dismutase mutants Sipiczki, M. (2003). Candida zemplinina sp. nov., an osmo- reveal a pro-oxidant action of weak organic acid food tolerant and psychrotolerant yeast that ferments sweet preservatives. Free Radic. Biol. Med., 27, 1219e1227. botrytized wines. Int. J. Syst. Evol. Microbiol., 53, Pirino, G., Zara, S., Pinna, G., Farris, G. A., & Budroni, M. 2079e2083. (2004). Diversity of Y region at HML locus in a Saccha- Sipiczki, M. (2008). Interspecies hybridization and recom- romyces cerevisiae strain isolated from a Sardinian wine. bination in Saccharomyces wine yeasts. FEMS Yeast Res., 8, Anton. Leeuw., 85,29e36. 996e1007. Puig, S., & Perez-Ortıń, J. E. (2000). Stress response and Smukalla, S., Caldara, M., Pochet, N., Beauvais, A., expression patterns in wine fermentations of yeast genes Guadagnini, S., Yan, C., et al. (2008). FLO1 is a variable induced at the diauxic shift. Yeast, 16, 139e148. green beard gene that drives biofilm-like cooperation in Puig, S., Querol, A., Barrio, E., & Pe´rez-Ortıń, J. E. (2000). budding yeast. Cell, 135, 726e737. Mitotic recombination and genetic changes in Saccharo- Straver, M. H., & Kijne, J. W. (1996). A rapid and selective myces cerevisiae during wine fermentation. Appl. Environ. assay for measuring cell surface hydrophobicity of Microbiol., 66, 2057e2061. brewer’s yeast cells. Yeast, 12, 207e213. Purevdorj-Gage, B., Orr, M. E., Stoodley, P., Sheehan, K. B., Sua´rez-Lepe, J. A. (1997). Levaduras vıńicas: Funcionalidad y & Hyman, L. E. (2007). The role of FLO11 in Saccharo- uso en bodega. Madrid, Spain: Mundi-prensa. myces cerevisiae biofilm development in a laboratory Sua´rez-Lepe, J. A. (2002). Impacto de levaduras y bacterias based flow-cell system. FEMS Yeast Res., 7, 372e379. en los aromas vıńicos fermentativos. In I Encuentro Rachidi, N., Barre, P., & Blondin, B. (1999). Multiple Ty- internacional de ciencias sensoriales y de la percepcioń (pp. mediated chromosomal translocations lead to karyotype 43e45). Barcelona, Spain: Rubes Editorial, CS2002. changes in a wine strain of Saccharomyces cerevisiae. Mol. Sulo, P., Spirek, M., Soltesova, A., Marinoni, G., & Piskur, J. Gen. Genet., 261, 841e850. (2003). The efficiency of functional mitochondrial Raha, S., & Robinson, B. H. (2000). Mitochondria, oxygen replacement in Saccharomyces species has directional free radicals, disease and ageing. Trends Biochem. Sci., 25, character. FEMS Yeast Res., 4,97e104. 502e508. Tamaki, H., Miwa, T., Shinozaki, M., Saito, M., Yun, C. W., Rando, O. J., & Verstrepen, K. J. (2007). Timescales of genetic Yamamoto, K., et al. (2000). GPR1 regulates filamentous and epigenetic inheritance. Cell, 128, 655e668. growth through FLO11 in yeast Saccharomyces cerevisiae. Remize, F., Cambon, B., Barnavon, L., & Dequin, S. (2003). Biochem. Biophys. Res. Commun., 267, 164e168. Glycerol formation during wine fermentation is mainly Teunissen, A. W., & Steensma, H. Y. (1995). Review: The linked to Gpd1p and is only partially controlled by the dominant flocculation genes of Saccharomyces cerevisiae HOG pathway. Yeast, 20, 1243e1253. constitute a new subtelomeric gene family. Yeast, 11, Reynolds, T. B., & Fink, G. R. (2001). Bakers’ yeast, a model 1001e1013. for fungal biofilm formation. Science, 291, 878e881. Thibon, C., Marullo, P., Claisse, O., Cullin, C., Ristow, H., Seyfarth, A., & Lochmann, E. R. (1995). Chro- Dubourdieu, D., & Tominaga, T. (2008). Nitrogen cata- mosomal damages by ethanol and acetaldehyde in bolic repression controls the release of volatile thiols by Saccharomyces cerevisiae as studied by pulsed field gel Saccharomyces cerevisiae during wine fermentation. FEMS electrophoresis. Mutat. Res., 326, 165e170. Yeast Res., 8, 1076e1086. Rodrigues de Sousa, H., Spencer-Martins, I., & Gonçalves, P. Torrea-Gon˜i, D., & Ancıń-Azpilicueta, C. (2001). Influence of (2004). Differential regulation by glucose and fructose of yeast strain on biogenic amines content in wines: Rela- a gene encoding a specific fructose/Hþ symporter in tionship with the utilization of amino acids during Saccharomyces sensu stricto yeasts. Yeast, 21, 519e530. fermentation. Am. J. Enol. Vitic., 52, 185e190. Roldań, A., Palacios, V., Caro, I., & Pe´rez, L. (2003). Urso, R., Rantsiou, K., Dolci, P., Rolle, L., Comi, G., & Resveratrol content of Palomino fino grapes: Influence of Cocolin, L. (2008). Yeast biodiversity and dynamics vintage and fungal infection. J. Agric. Food Chem., 51, during sweet wine production as determined by 1464e1468. molecular methods. FEMS Yeast Res., 8, 1053e1062. Roldań, A., Palacios, V., Penãte, X., Benı´tez, T., & Pe´rez, L. Valero, E., Cambon, B., Schuller, D., Casal, M., & Dequin, S. (2006). Use of Trichoderma enzymatic extracts on vinifi- (2007). Biodiversity of Saccharomyces yeast strains from cation of Palomino fino grapes in the sherry region. grape berries of wine-producing areas using starter J. Food Eng., 75, 375e382. commercial yeasts. FEMS Yeast Res., 7, 317e329. Sancho, E. D., Hernańdez, E., & Rodrı´guez-Navarro, A. van Dyk, D., Pretorius, I. S., & Bauer, F. F. (2005). Mss11p is (1986). Presumed sexual isolation in yeast populations a central element of the regulatory network that controls during production of sherrylike wine. Appl. Environ. FLO11 expression and invasive growth in Saccharomyces Microbiol., 51, 395e397. cerevisiae. Genetics, 169,91e106. 84 3. YEASTS USED IN BIOLOGICALLY AGED WINES van Mulders, S. E., Christianen, E., Saerens, S. M., Aspergillus nidulans of a Trichoderma longibrachiatum Daenen, L., Verbelen, P. J., Willaert, R., et al. (2009). endoglucanase of enological relevance. J. Agric. Food Phenotypic diversity of Flo protein family-mediated Chem., 48, 951e957. adhesion in Saccharomyces cerevisiae. FEMS Yeast Res., 9, Vriesekoop, F., Haass, C., & Pamment, N. B. (2009). The role 178e190. of acetaldehyde and glycerol in the adaptation to ethanol van Uden, N. (1989). Alcohol toxicity in yeasts and bacteria. stress of Saccharomyces cerevisiae and other yeasts. FEMS Boca Raton, FL: CRC Press. Yeast Res., 9, 365e371. Verstrepen, K. J., Derdelinckx, G., Dufour, J. P.,Winderickx, J., Watanabe, M., Fukuda, K., Asano, K., & Ohta, S. (1990). Pretorius, I. S., Thevelein, J. M., et al. (2003a). The Mutants of bakers’ yeasts producing a large amount of Saccharomyces cerevisiae alcohol acetyl transferase gene isobutyl alcohol or isoamyl alcohol, flavour components ATF1 is a target of the cAMP/PKA and FGM nutrient- of bread. Appl. Microbiol. Biotechnol., 34, 154e159. signalling pathways. FEMS Yeast Res., 4, 285e296. Williams, S. A., Hodges, R. A., Strike, T. L., Snow, R., & Verstrepen, K. J., Jansen, A., Lewitter, F., & Fink, G. R. Kunkee, R. E. (1984). Cloning the gene for the malolactic (2005). Intragenic tandem repeats generate functional fermentation of wine from Lactobacillus delbrueckii in variability. Nat. Genet., 37, 986e990. Escherichia coli and yeasts. Appl. Environ. Microbiol., 47, Verstrepen, K. J., & Klis, F. M. (2006). Flocculation, adhe- 288e293. sion and biofilm formation in yeasts. Mol. Microbiol., 60, Wu, X., & Jiang, Y. W. (2008). Overproduction of non- 5e15. translatable mRNA silences. The transcription of Ty1 Verstrepen, K. J., Reynolds, T. B., & Fink, G. R. (2004). retrotransposons in S. cerevisiae via functional inactiva- Origins of variation in the fungal cell surface. Nat. Rev. tion of the nuclear cap-binding complex and subsequent Microbiol., 2, 533e540. hyperstimulation of the TORC1 pathway. Yeast, 25, Verstrepen, K. J., van Laere, S. D., Vanderhaegen, B. M., 327e347. Derdelinckx, G., Dufour, J. P., Pretorius, I. S., et al. Zara, G., Angelozzi, D., Belviso, S., Bardi, L., Goffrini, P., (2003b). Expression levels of the yeast alcohol acetyl- Lodi, T., et al. (2009). Oxygen is required to restore flor transferase genes ATF1, Lg-ATF1, and ATF2 control the strain viability and lipid biosynthesis under fermenta- formation of a broad range of volatile esters. Appl. tive conditions. FEMS Yeast Res., 9, 217e225. Environ. Microbiol., 69, 5228e5237. Zara, G., Mannazzu, I., Sanna, M. L., Orro, D., Farris, G. A., Vigentini, I., Romano, A., Compagno, C., Merico, A., & Budroni, M. (2008). Exploitation of the semi-homo- Molinari, F., Tirelli, A., et al. (2008). Physiological and thallic life cycle of Saccharomyces cerevisiae for the oenological traits of different Dekkera/Brettanomyces development of breeding strategies. FEMS Yeast Res., 8, bruxellensis strains under wine-model conditions. FEMS 1147e1154. Yeast Res., 8, 1087e1096. Zara, S., Antonio Farris, G., Budroni, M., & Bakalinsky, A. T. Villanueva, A., Ramoń, D., Valles, S., Lluch, M. A., & (2002). HSP12 is essential for biofilm formation by MacCabe, A. P. (2000). Heterologous expression in a Sardinian wine strain of S. cerevisiae. Yeast, 19, 269e276. CHAPTER 4

Non-Saccharomyces Yeasts in the Winemaking Process Paloma Manzanares, Salvador Valle´s, Fernando Viana Departamento de Biotecnologıá de Alimentos, Instituto de Agroquı´mica y Tecnologıá de Alimentos, Consejo Superior de Investigaciones Cientı´ficas, Valencia, Spain

OUTLINE

1. Introduction 85 3.1. Influence of Non-Saccharomyces Yeasts on the Winemaking Process 90 2. Isolation, Enumeration, and Identification 3.1.1. Pectolytic Enzymes 90 of Non-Saccharomyces Yeasts 86 3.1.2. Proteolytic Enzymes 91 2.1. Isolation and Enumeration 86 3.1.3. Cellulolytic and 2.2. Identification 87 Hemicellulolytic Enzymes 93 2.2.1. Chromatographic Analysis 3.2. Influence on Aroma 93 of Long-chain Fatty Acid 3.2.1. Primary Aroma 94 Profile 88 3.2.2. Secondary Aroma 96 2.2.2. Electrophoretic Analysis of Protein Profile 89 4. Design of Mixed Starter Cultures 102 3. The Role of Non-Saccharomyces Yeasts 5. Final Considerations 104 in Vinification 89 Acknowledgments 105

1. INTRODUCTION and winery equipment (Barata et al., 2008b; Bel- tra´n et al., 2002; Combina et al., 2005; Martini & Numerous studies designed to isolate and Vaughan-Martini, 1990; Martini et al., 1996; identify yeasts present on the surface of grapes Raspor et al., 2006; Sabate´ et al., 2002), along

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10004-9 85 Copyright Ó 2011 Elsevier Inc. All rights reserved. 86 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS with quantitative and qualitative analyses of 2. ISOLATION, ENUMERATION, the yeast species present during alcoholic fer- AND IDENTIFICATION OF NON- mentation (Esteve-Zarzoso et al., 2001; Fleet & SACCHAROMYCES YEASTS Heard, 1993; Guillamoń et al., 1998; Renouf et al., 2007; Schu¨ tz & Gafner, 1993; Urso et al., Before we can exploit the potential of non- 2008) have confirmed that in traditional wine- Saccharomyces yeasts and understand their making processes (without the use of starter contribution to the vinification process, we cultures) grape must is transformed into wine must be able to isolate and correctly identify through the sequential activity of different them using appropriate techniques. yeast species. Under these conditions, fermentation generally begins with the growth of 2.1. Isolation and Enumeration weakly fermentative yeast species belonging to the genera Candida, Debaryomyces, Dekkera, Isolation and enumeration of yeast from Hanseniaspora, Metschnikowia, Pichia, Torulas- grapes, must, wine, and wineries has tradition- pora, and Zygosaccharomyces (Heard & Fleet, ally involved plate counts. While spread-plate 1985). The growth of these species, known or pour-plate methods can be used, spread collectively as non-Saccharomyces yeasts, is plates seem to yield better results (King et al., limited to the first 2 or 3 d of fermentation, after 1986). Other systems for the enumeration of which they die as a result of ethanol toxicity. As yeasts include membrane filtration, direct these yeasts disappear, highly fermentative microscopic counts, dye reduction tests, and strains of the species Saccharomyces cerevisiae the most probable number method (reviewed begin to multiply until they become solely in Jay, 1994). responsible for alcoholic fermentation. Despite Various nutrient agars have been described growing only during the first few days of for the isolation of wine yeasts. Examples fermentation, non-Saccharomyces yeasts produce include those based on grape must or malt a large number of compounds that can have agar. In general, these are complex, nutritionally a significant influence on the quality of the rich media containing an energy source (e.g., wine (reviewed in Fleet, 2003). This chapter glucose, fructose, or sucrose), a hydrolyzed will describe the role of non-Saccharomyces protein (e.g., peptone, tryptone, or casitone), yeasts in the winemaking process and their a complex vitamin supplement such as yeast contribution to the final product, both in terms or malt extract, antibiotics to prevent bacterial of their influence on wine aroma and their role growth (oxytetracycline, chlorotetracycline, or in the vinification process itself. We will begin chloramphenicol), and compounds to inhibit by discussing traditional methods for identifi- fungal growth (e.g., rose bengal or dichloran) cation of yeasts before addressing the methods (reviewed in Beuchat, 1998). However, these used to isolate and identify non-Saccharomyces nonselective media allow growth of all species species. We will then describe the role of non- of yeast associated with the winemaking Saccharomyces yeasts in the vinification process. process, and this limits their usefulness. For Finally, we will introduce the design of mixed instance, when the sample contains a mixture starter cultures containing non-Saccharomyces of very different populations derived from yeasts and discuss how they might be used to a range of yeast species, most of the colonies exploit the positive characteristics of these on the plate belong to the predominant species yeasts while minimizing their possible negative and impede recognition of colonies from effects. minority species. Selective and differential ISOLATION, ENUMERATION, AND IDENTIFICATION OF NON-SACCHAROMYCES YEASTS 87 methods have been developed to circumvent of yeasts belonging to the genera Dekkera/ these limitations. Media for use in enology Brettanomyces based on the use of ethanol as must allow S. cerevisiae, spoilage yeasts a carbon source and supplemented with cyclo- belonging to the genera Saccharomyces, and heximide, bromocresol green, and r-cumaric non-Saccharomyces yeasts to be distinguished acid. It has also been possible to demonstrate from each other. The best example of a selective the presence of Dekkera bruxellensis on the medium is lysine agar, which allows S. cerevisiae surface of grape berries thanks to the develop- to be distinguished from non-Saccharomyces ment of an enrichment medium (EBB) com- yeasts. Its selectivity is based on its failure to prising must, ethanol, malt and yeast extracts, support the growth of S. cerevisiae, which is ammonium sulfate, and Tween-80 (Renouf & unable to use lysine as a nitrogen source. Heard Lonvaud-Funel, 2007). and Fleet (1986a) were the first to demonstrate Although yeast plate counts are ideal for the effectiveness of lysine agar for the selective ecologic studies, obtaining reliable results isolation and enumeration of Kloeckera apiculata usually requires incubation of the plates for up and Candida stellata populations during fermen- to 4 d, and this makes the approach too slow tation. This is the medium mainly used in the for use in quality-control procedures. Conse- brewing industry to analyze contamination by quently, more rapid, automated methods based non-Saccharomyces yeasts. Nutrient media con- on impedance, adenosine triphosphate (ATP) taining ethanol and sodium metabisulfite have measurements, or fluorescence microscopy been described for the selective isolation and have been developed for the enumeration of enumeration of Saccharomyces species during yeast populations (reviewed in Lightfoot & winemaking (Kish et al., 1983). In this medium, Maier, 2002). Flow cytometry using fluorescent non-Saccharomyces yeasts, particularly Kloeckera/ dyes has also been employed to obtain rapid Hanseniaspora species, do not grow because of estimates of the number of yeast and bacteria their lower tolerance of sulfur and ethanol. present in wine and in samples obtained during Media containing high concentrations of sor- fermentation (Malacrino et al., 2001). Connell bates and benzoates have also been described et al. (2002) have developed a filter-based for the selective isolation of Zygosaccharomyces chemiluminescent in situ hybridization method bailii, a typical wine spoilage yeast (Pitt & Hock- for the detection, identification, and enumera- ing, 1985), and media containing sulfite and tion of Brettanomyces species in winery air bismuth have been used to analyze the presence samples. of yeasts that produce hydrogen sulfide (Jiranek et al., 1995; Rupela & Tauro, 1984). A differential 2.2. Identification medium containing glucose and formic acid and supplemented with an indicator was recently Standard identification criteria classify yeasts described by Schuller et al. (2000) for the on the basis of their morphological, physiolog- enumeration of Z. bailii in wine samples. This ical, and biochemical characteristics (Barnett medium allowed Z. bailii to be distinguished et al., 1990; Kreger-van Rij, 1984; Kurtzman & from other spoilage yeasts. The same medium Fell, 1998). Physiological characteristics serve has been successfully used to isolate Z. bailii mainly to describe and identify species and, to and Zygosaccharomyces bisporus from damaged a lesser extent, genera. The most commonly grapes, despite the low numbers of these yeasts used tests for routine identification are fermen- present (Barata et al., 2008a). Rodrigues et al. tation capacity on different carbon sources, (2001) described a medium for the detection growth on different carbon and nitrogen 88 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS sources, vitamin requirements, growth at proteins and long-chain fatty acids. We describe different temperatures and in media containing these methods in Sections 2.2.1 and 2.2.2. high concentrations of sugar or sodium chloride, urea hydrolysis, and antibiotic resistance 2.2.1. Chromatographic Analysis (Yarrow, 1998). However, Yarrow (1998) points of Long-chain Fatty Acid Profile out that, in the absence of a standardized This technique involves the extraction of method for identification tests, the results will yeast fatty acids and the analysis of sample depend on the technique used. In addition, the composition by gas chromatography. In the results can vary in different strains of the same past, a number of problems have prevented yeast species, and this can lead to identification this technique from being successfully applied errors. Alongside these drawbacks, traditional to the identification of yeasts in the food identification methods require evaluation of 60 industry. These include the dependence of fatty to 90 tests. This complicated and laborious acid composition on the growth conditions of process can take 2 to 3 weeks to complete. Obvi- the yeast, the lack of differentiation between ously, this methodology cannot be routinely the fatty acid composition of different yeasts, applied in the food industry or indeed in the high cost, the time required to obtain the wineries. Consequently, efforts have been results, and the need for skilled personnel. made to develop simplified and shortened iden- Nevertheless, some of these problems can be tification techniques based on the response of overcome, for instance by standardizing the yeasts to a series of carefully selected tests growth conditions to minimize as far as possible (Dea´k & Beuchat, 1996). Vela´zquez et al. (2001) any variation in lipid composition or by using developed a kit comprising a series of 24 physi- solid media, which allow results to be obtained ological and biochemical tests along with soft- within 2 d of isolating and purifying the ware to analyze the data obtained. Those unknown yeast (Malfeito-Ferreira et al., 1997). authors had previously proposed a kit involving Furthermore, the technique is no more expen- 10 tests for the identification of wine yeasts sive than most available rapid-identification (Vela´zquez et al., 1993). However, these simpli- techniques. According to Malfeito-Ferreira fied techniques are based on the same principles et al. (1997), analysis of the fatty acid profile as the more traditional methods. Consequently, can be used to identify most of the yeasts that although the process may be automated or are important in the food industry, classify the computerized, the time required is the same species according to their potential for spoilage and identification is often incorrect (Loureiro & of a foodstuff, and identify sources of contami- Querol, 1999). nation in wineries and bottling plants. The Recently, new molecular biological tech- yeasts typically associated with foodstuffs can niques have been developed for the identifica- be separated into three groups according to their tion and characterization of yeasts. These fatty acid composition. The yeasts with the high- techniques, which will be described in detail in est potential for spoilage, such as Dekkera anom- Chapter 5, include restriction fragment length ala, D. bruxellensis, and various species of the polymorphism (RFLP) analysis of mitochon- genus Zygosaccharomyces, can be identified by drial DNA (mtDNA), electrophoretic separation the presence of significant quantities of linoleic of chromosomes, restriction analysis of ribo- acid (C18:2) and the absence of linolenic acid somal DNA (rDNA), and random amplification (C18:3). The presence of both fatty acids is of polymorphic DNA (RAPD). Other techniques typical of yeast species belonging to the genera designed to overcome the problems of routine Candida, Cryptococcus, Debaryomyces, Kluyvero- identification involve analyzing the profile of myces, and Pichia, which are associated with THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 89 poor hygiene during the processing of food- for identification and highlight the usefulness stuffs. The lack of C18 polyunsaturated fatty of the approach as a rapid and sensitive method acids suggests the presence of fermentative of identifying strains of this important group of strains that may cause spoilage, such as those industrial yeasts. Electrophoretic analysis of belonging to the genus Saccharomyces. A combi- protein profiles has also been used for taxo- nation of fatty acid profile and typical PCR nomic purposes in the genera Candida (Vancan- methods has also been used as a rapid system neyt et al., 1991) and Zygosaccharomyces (Duarte for the detection of contaminating yeasts et al., 2004). In addition, the technique has been (Sancho et al., 2000). The main drawback of used in a clinical context to identify different this technique in the food industry is the strains of Candida albicans (Boriollo et al., 2003). absence of a database of fatty acid profiles for Electrophoretic analysis of protein profile yeasts typically found in foodstuffs. However, is sensitive, does not require expensive or sop- various research laboratories now have access histicated equipment, and can be completed in to such databases, meaning that it is simply 48 h following isolation of the yeast strain. a question of time before they become available Consequently, it should be used more exten- in the food industry (Loureiro & Querol, 1999). sively in the food industry. However, as in the In some cases, a degree of overlap has been case of long-chain fatty acid profiles, represen- found between the fatty acid profiles of different tative databases of isoenzyme profiles must first yeasts, as was observed for the profile of Kluy- be developed for yeasts used in the wine veromyces, which overlaps with that of Saccharo- industry. myces (Augustyn et al., 1992).

2.2.2. Electrophoretic Analysis 3. THE ROLE OF NON- of Protein Profile SACCHAROMYCES YEASTS IN Analysis of the electrophoretic profile of VINIFICATION extracellular and/or intracellular proteins or of isoenzymes allows differentiation of strains. In The microflora present at the beginning of this technique, extracted proteins are separated fermentation is derived solely from the grapes by polyacrylamide gel electrophoresis. The and essentially comprises species belonging to appearance of specific bands for a given strain genera with very limited ethanol tolerance, is used as a criterion for differentiation. The such as Hanseniaspora/Kloeckera, Hansenula, technique has been used with various yeast Metschnikowia, and Candida, and strains of S. cer- genera and its taxonomic validity has been evisiae, which are more ethanol tolerant but only confirmed. Strains of industrial yeast belonging represent a very small proportion of the micro- to the genus Saccharomyces have been differenti- flora at this stage. This proportion depends on ated using this method (Degre´ et al., 1989; a wide variety of factors, such as the harvesting Duarte et al., 1999; Guillamoń et al., 1993; van method, type of transport, fermentation temper- Vuuren & van der Meer, 1987). Duarte et al. ature, and quantity of sulfur added. The ratio of (1999) analyzed isoenzyme profile and grouped non-Saccharomyces to Saccharomyces yeasts can 35 strains of yeast in the four recognized species contribute to accentuating the chemical and of Saccharomyces sensu stricto (S. cerevisiae, sensory changes that take place during fermenta- Saccharomyces bayanus, Saccharomyces pastorianus/ tion, with clear consequences for the quality of Saccharomyces carlsbergensis, and Saccharomyces the wine obtained. Consequently, irrespective paradoxus). The results of that study confirm of whether or not the fermentations are inocu- the validity of isoenzyme profile as a criterion lated with Saccharomyces yeast, wineries can 90 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS consider enhancing the non-Saccharomyces flora S. cerevisiae (McKay, 1990), non-Saccharomyces in order to take advantage of its fermentative yeasts are notable producers of extracellular characteristics. enzymes (Dizy & Bisson, 2000; Lagace & Bisson, Numerous studies have characterized the 1990; Strauss et al., 2001). non-Saccharomyces yeast strains found on different varieties of grape. In an effort to exploit 3.1. Influence of Non-Saccharomyces the putative causal relationship between the Yeasts on the Winemaking Process presence of these strains and the type of wine produced, fermentations have been attempted Grapes, the raw material for winemaking, with starter cultures that include both S. cerevi- contain numerous different compounds, notably siae and yeasts belonging to the genera Kloeckera, phenols, aromatic precursors, enzymes, and Cryptococcus, Hanseniaspora, Candida, Pichia, and structural components. The structural compo- Hansenula (Fleet & Heard, 1993). The wines nents include pectins, cellulose, glycans, hemi- produced with these mixed starter cultures celluloses, proteins, and lignin. The enzymatic differ significantly in both chemical composi- degradation of this structure can improve the tion and sensory characteristics (Egli et al., different stages of vinification, for instance by 1998). By further analyzing the outcome of these enhancing the yield and clarification of the fermentations, greater insight may be gained must, increasing color extraction, and improving into the particular characteristics of non- filtration of the wine. Not all of the enzyme activ- Saccharomyces yeasts that affect the type of ities of interest can be obtained from the grape wine produced. and those enzymes that are present may not be One characteristic that is thought to differ fully effective under vinification conditions. between non-Saccharomyces yeasts and Saccharo- Therefore, it is of particular interest to control myces species is the production of enzymes the development of non-Saccharomyces yeasts as (esterases, glycosidases, lipases, b-glucosidases, sources of these enzymes. proteases, cellulases, etc.). By interacting with The group of secreted enzymes involved in substrates in the medium, these enzymes can hydrolysis of structural components are improve particular phases of the process (such referred to as macerating enzymes, and include as maceration, filtration, or clarification), pectolytic, proteolytic, cellulolytic, and hemicel- increase yield and color extraction, and enhance lulolytic enzymes. the characteristics of the wine, especially the aroma (Charoenchai et al., 1997). Since grapes 3.1.1. Pectolytic Enzymes produce only small quantities of enzymes with Pectolytic enzymes cleave long-chain pectins very limited activity, exogenous enzymes are to generate shorter, more soluble chains. This often introduced during vinification. However, plays an important role in the changes that if instead we exploit the contribution of occur during grape ripening. Later, during enzymes from yeasts involved in the vinifica- winemaking, it facilitates grape pressing and tion process, we may be able to produce contributes to clarification of the must. The pres- a more natural product and at the same time ence of pectolytic enzymes can also improve improve both the vinification process and the filtration of the wines and increase the extrac- sensory attributes of the wine. In-depth studies tion of substances that contribute to color and are still required to both assess the nature of aroma while the must remains in contact with these yeast-derived enzymes and determine the grape skins. how they favor vinification. However, various The most notable pectolytic enzymes are the studies have already shown that, unlike pectinases. These act on pectins, the major THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 91 constituents of the primary cell wall in higher thought to inhibit polygalacturonase production plants. Pectins are heteropolysaccharides with in non-Saccharomyces yeasts but induce it in S. a backbone made of repeating a-1,4-D galactur- cerevisiae (Strauss et al., 2001). Non-Saccharomyces onic acid units. These galacturonic acid units are yeasts could be used to reinforce the production periodically replaced by b-1,2- and b-1,4-linked of polygalacturonases by S. cerevisiae (McKay, L-rhamnose units (approximately one per 25 1990). The combined activity of these enzymes galacturonic acid units), from which a series of and pectin methylesterase could improve the side chains of varying length and composition degradation of pectins in the medium during branch off. Although these rhamnogalactur- fermentation. onans are the most common pectins, highly branched arabinogalactans also exist. 3.1.2. Proteolytic Enzymes Pectinases are mainly classified according to Proteins are present in varying quantities in their mechanism of action. Pectin methyles- the grape and, along with polysaccharides, are terases, for instance, act via de-esterification, responsible for increasing must and wine releasing methanol and reducing the degree of turbidity. Although these proteins can be elimi- methoxylation of the pectin, whereas the poly- nated with bentonite, this nonspecific process galacturonases, pectin lyases, and pectate lyases also leads to loss of aromas and compounds act through depolymerization. The polygalac- that influence flavor. Protease treatment, on turonases are the most important from a wine- the other hand, specifically hydrolyzes proteins making perspective. There are two types: the and improves the clarity and stability of the exopolygalacturonases, which hydrolyze the wine. The smaller, more soluble peptides and terminal groups and reduce the chain length amino acids generated by this enzymatic hydro- only slightly, and the endogalacturonases, lysis are also nitrogen-containing compounds. which act at random. The endogalacturonases Consequently, in addition to improving clarifi- alter the dimensions of the molecules more cation and stabilization, protease treatment rapidly, reducing the viscosity of the pulp and helps to prevent stuck fermentation caused by generally improving some phases of the vinifi- a lack of assimilable nitrogen in the must. cation process, such as clarification. Yeast proteases play an important role in the Studies published to date indicate that non- process of autolysis during on-lees aging of Saccharomyces yeasts secrete polygalacturonase wines and in the development of protein haze and pectin methylesterase. Polygalacturonase (protein degradation), especially in white wines. is produced by species of the genera Candida, However, not all proteases are active under the Pichia, and Kluyveromyces, whereas pectin meth- particular conditions found in wine. Analysis ylesterase is produced by Candida, Debaryomy- of protease activity from non-Saccharomyces ces, and Pichia (Table 4.1). yeasts has revealed the importance of nitrogen Since it is difficult to perform experiments in sources in the production of the extracellular natural conditions, in most cases activity has enzymes (Charoenchai et al., 1997). been detected by growing the microorganisms As with pectinases, the main problem with on plates. Consequently, little information is proteases is their weak activity under the partic- available on the effect that components of the ular conditions found in wine. As a result, tests media might have on the induction or inhibition to detect proteolytic activity tend to be done on of enzyme production, and, as a result, a defini- plates or using a model wine, even though these tive relationship between the presence of the options do not ensure that the proteases yeast and the secreted activity cannot be detected will be active under vinification condi- assumed. For instance, glucose in the must is tions. Nevertheless, the use of proteases from 92 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

TABLE 4.1 Macerating Enzymes Produced by Non-Saccharomyces Yeasts

Macerating enzymes

Yeasts PG PME CEL GLU XYL PR Reference

Candida albicans X (7) Candida ﬂavus X (1) Candida hellenica X X (1)

Candida krusei X (3) Candida lambica X (1) Candida lipolytica X (4) Candida norvegensis X Candida olea X (1),(4) Candida oleophila X X (1)

Candida pelliculosa X (1) Candida pulcherrima X X X X X (1),(4) Candida silvae X Candida sorbosa X (1) Candida stellata X X X X X (1),(6) Candida tropicalis X (2)

Candida valida X (1) Candida wickerhamii X (8) Cryptococcus sp. X(11) Cryptococcus albidus X (12) Debaryomyces hansenii X (1)

Debaryomyces membranaefaciens X (3) Hanseniaspora guilliermondii X (5) Kloeckera apiculata X (1),(4),(5) Kloeckera thermotolerans X (6) Kluyveromyces marxianus X (10) Metschnikowia pulcherrima X (1),(6)

Pichia anomala X (6) Pichia guilliermondii X (6)

(Continued) THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 93

TABLE 4.1 Macerating Enzymes Produced by Non-Saccharomyces Yeastsdcont’d

Macerating enzymes

Yeasts PG PME CEL GLU XYL PR Reference

Pichia kluyveri XX (9) Pichia membranaefaciens X (6)

CEL ¼ cellulase; GLU ¼ b-glucanase; PG ¼ polygalacturonase; PME ¼ pectin methylesterase; PR ¼ protease; XYL ¼ xylanase. (1) Strauss et al. (2001); (2) Luh and Phaff (1951); (3) Bell and Etchells (1956); (4) Lagace and Bisson (1990); (5) Dizy and Bisson (2000); (6) Fernańdez et al. (2000); (7) Chambers et al. (1993); (8) LeClerc et al. (1984); (9) Masoud and Jespersen (2006); (10) Serrat et al. (2004);(11)Thongekkaew et al. (2008); (12) Servili et al. (1990). non-Saccharomyces yeasts in the vinification high-molecular-weight (1,3)-b-D-glucans, even process has been investigated. Specifically, the though they are only present at low concentra- addition of proteases from K. apiculata has been tions. These polysaccharides can be eliminated used successfully to degrade some of the protein by enzymatic treatment. As in the case of the in Chenin Blanc and Chardonnay wines. It pectins, these compounds are degraded by has even been demonstrated that proteases a series of enzymes, including cellulolytic from Candida olea, Candida lipolytica, Candida (endoglucanases, exoglucanases, cellobiases, pulcherrima,andK. apiculata can produce a and b-glucanases) and hemicellulolytic (xyla- moderate reduction in wine turbidity (Lagace nases) enzymes. & Bisson, 1990). Dizy and Bisson (2000) demon- During cellulose degradation, the endogluca- strated that some species belonging to the genus nases and exoglucanases act at random and Kloeckera/Hanseniaspora are the highest pro- from the ends, respectively, generating a mixture ducers of proteolytic activity in the must and of oligosaccharides, predominantly cellobiose, affect the protein profile of the finished wines. which is hydrolyzed to glucose by cellobiase. Despite these positive properties, in some The b-glucanases are specific enzymes for the cases proteolytic activity does not significantly hydrolysis of the b-glucans mentioned above. reduce the temperature-related turbidity of the Unlike cellulose, the hemicelluloses, notably wines and may even increase it in fermentations xylane, are branched heteropolymers, and as with a high proteolytic activity (Strauss et al., a result of their complexity must be degraded 2001). Table 4.1 shows the species with proteo- by multiple enzymes, such as the xylanases, gal- lytic activity. actanases, and mannanases. To date, the only non-Saccharomyces yeasts 3.1.3. Cellulolytic and Hemicellulolytic that have been described as producers of cellu- Enzymes lolytic or hemicellulolytic enzymes are Candida Given that cellulose and hemicellulose are and Cryptococcus, as shown in Table 4.1. the main structural polysaccharides of the plant cell wall, their enzymatic degradation will allow 3.2. Influence on Aroma extraction and release of pigments and aromas from the grape skins. Treatment with these Aroma is one of the organoleptic characteris- enzymes reduces the maceration time, as the tics that determine the quality of a wine. As with desired results are achieved sooner. many other foodstuffs, wine aroma is deter- When working with grapes infected with mined by hundreds of different compounds the fungus Botrytis cinerea, the clarification with concentrations that can vary between and filtration process is also impaired by 101 and 1010 g/kg (Rapp & Mandery, 1986). 94 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

The concentration of these compounds depends respectively. In the next step, b-D-glucosidase on factors such as grape variety, climate, soil, activity releases terpenes from the glucosides rainfall, and time of harvesting, as well as generated in the first step. numerous variables relating to the fermentation Suboptimal conditions (pH and temperature) process (pH, temperature, nutrients, and micro- or inhibition by glucose and ethanol neverthe- flora) and the operations that it encompasses less result in reduced activity of hydrolytic (filtration, clarification, etc.). The aromatic enzymes derived from the grapes or from S. quality of the wine is determined by the balance cerevisiae. Consequently, these precursors are and interaction of these compounds. commonly hydrolyzed in very small propor- It is important to distinguish between three tions during fermentation (Gunata, 1984). The different types of wine aroma: the varietal or degree of inhibition depends on the species primary aroma, determined by the grape and strains of organism involved (Aryan et al., variety; the fermentation or secondary aroma; 1987; Delcroix et al., 1994; LeClerc et al., 1987; and the bouquet or tertiary aroma resulting Rosi et al., 1994). For instance, Grossman et al. from the transformation of aromas during (1987) showed that the b-glucosidase of Hanse- aging. Non-Saccharomyces yeasts can influence nula sp., isolated from fermented must, was both the primary and secondary aroma, as able to release aromatic substances when added described below. to the wine but was less effective in the must. Studies of yeast glycosidases indicate that 3.2.1. Primary Aroma some specific enzymes can influence the varietal The varietal aroma is mainly determined by aroma of the wine (Laffort et al., 1989), espe- the quantity and chemical nature of the volatile cially when fermentation is carried out under secondary metabolites present in the grape (van natural conditions (Fugelsang, 1997), where Rensburg & Pretorius, 2000). Of these, the non-Saccharomyces yeasts predominate during terpenes have the greatest influence on flavor the initial stages. This apparent influence of and aroma, particularly in wines derived from non-Saccharomyces yeasts may be explained by Moscatel grapes but also in other less aromatic their marked hydrolytic activity, which is absent varieties (Marais, 1983; Rapp & Mandery, 1986). in most Saccharomyces strains (Charoenchai Terpenes are volatile compounds that are et al., 1997; Fernańdez et al., 2000; Gunata present in the grape as free molecules or non- et al., 1994; Manzanares et al., 1999, 2000; aromatic glycosylated precursors. In general, Mendes-Ferreira et al., 2001; Strauss et al., the precursors are glycosides formed from 2001; U´ beda & Briones, 2000; Zoecklein et al., a disaccharide and a terpene, with a-L- 1997). arabinofuranosyl-b-D-glucopyranosides, a-L- Rosi et al. (1994) showed that yeasts of rhamnopyranosyl-b-D-glucopiranosides, and the genera Candida, Debaryomyces, Hansenias- b-D-apiosyl-b-D-glucopyranosides of geraniol, pora/Kloeckera, Kluyveromyces, Metschnikowia, nerol, and linalool among the most abundant Pichia, Saccharomycodes, Schizosaccharomyces,and (Gunata et al., 1988). The hydrolysis that Zygosaccharomyces can produce b-glucosidase, releases the volatile aromatic compounds and this was later confirmed by other authors occurs in two steps (Figure 4.1). Firstly, the (Charoenchai et al., 1997; Manzanares et al., glycosidic bonds are cleaved to release specific 2000; McMahon et al., 1999; Strauss et al., 2001). sugars according to the substrate and the enzyme However, the analogues used as substrates for involved. For instance, arabinose, rhamnose, and the selection of b-glucosidase can equally detect apiose are released by a-L-arabinofuranosidase, exoglucanase, and the activities of these enzymes a-L-rhamnosidase, and b-D-apiosidase enzymes, can therefore be confused (Strauss et al., 2001). THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 95

O CH TERP A O 2 O OH O ABF HO2HC OH OH HO OH

TERP TERP B O CH2 CH OH O 2 O CH OH O O O Arabinose 2 API BGL Apiose Terpene* + Glucose OH OH + Rhamnose OH HO HO OH OH

TERP O C HO O O CH2 O CH3 O O O OH RAM HO OH HO OH *Geraniol *Nerol *Linalool

FIGURE 4.1 Enzymatic hydrolysis of glycosylated precursors. A ¼ a-L-arabinofuranosyl-(1,6)-b-D-glucopyranoside; B ¼ b-D-apiosyl-(1,6)-b-D-glucopyranoside; C ¼ a-L-rhamnopyranosyl-(1,6)-b-D-glucopyranoside; ABF ¼ a-L-arabinofuranosidase; API ¼ b-D-apiosidase; BGL ¼ b-D-glucosidase; RAM ¼ a-L-rhamnosidase.

As a result, the putative production of b-glucosi- produce extracellular b-glucosidases that are dase by those yeasts should be analyzed in not inhibited by glucose; in particular, Candida greater detail. Likewise, natural substrates such peltata b-glucosidase remains unaffected by as precursor extracts should be used, since the glucose at concentrations of up to 250 mg/mL b-glucosidases selected as a result of their (Saha & Bothast, 1996). activity with artificial substrates could prove In contrast to the limited information avail- ineffective when it comes to hydrolyzing aro- able on macerating enzymes, b-glucosidases matic precursors in grape must. This is the case are quite well characterized. In some cases these for the b-glucosidase from Brettanomyces bruxel- enzymes have even been purified; for example, lensis, which was found to be unable to hydro- in Debaryomyces hansenii strains (Riccio et al., lyze an extract of precursors from grape must 1999; Yanai & Sato, 1999). D. hansenii b-glucosi- (Mansfield et al., 2002). The potential effective- dase maintains its activity in the presence of ness of yeast-derived b-glucosidases is even ethanol concentrations of up to 15% (vol/vol) further reduced in most cases by the fact that and releases terpenes, not only from extracts of the enzymes are intracellular and released only glycosylated precursors but also when added in very small amounts into the culture medium to the must during fermentation, where it (McMahon et al., 1999). increases the concentrations of linalool and Another limitation of these enzymes is their nerol by 90 and 116%, respectively (Yanai & very weak activity in the presence of glucose in Sato, 1999). Extracellular b-glucosidase has the must or wine, making it especially necessary also been purified from Debaryomyces vanrijiae to analyze their inhibition by this sugar. Candida, and found to be active in the presence Debaryomyces, Kluyveromyces, and Pichia species of 200 mM glucose (80% activity) and 15% 96 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

(vol/vol) ethanol (64% residual activity). Its 3.2.2. Secondary Aroma addition to Moscatel grape must during fermen- Yeasts are responsible for the secondary or tation also increases the concentration of fermentation aroma of the wine. This aroma terpenes (Belancic et al., 2003). A b-glucosidase arises during alcoholic fermentation and is from Candida molischiana has also been shown determined by compounds produced as part to release terpenes and alcohols both from of wine yeast metabolism. Although ethanol, a Moscatel glycoside extract and from the wine glycerol, and carbon dioxide are quantitatively itself (Genove´s et al., 2003). the most abundant of these compounds and b-D-xylosidase is also involved in releasing play a fundamental role in wine aroma, their aromas, although data are limited on its hydro- contribution to the secondary aroma is relatively lytic capacity. Manzanares et al. (1999) selected limited. Volatile fatty acids, higher alcohols, eight yeast strains belonging to the genera Han- esters, and, to a lesser extent, aldehydes, have seniaspora and Pichia as the best producers of b- a greater contribution to secondary aroma D-xylosidase from a total of 54 species of wine ( Rapp & Versini, 1991). These compounds, yeast. Of these two genera, only the species shown in Table 4.3, are generated through the Hanseniaspora osmophila, Hanseniaspora uvarum, conversion of directly fermentable sugars and and Pichia anomala exhibited b-D-xylosidase also of long-chain fatty acids and nitrogenated activity. Although this is not a new observation and sulfur compounds, among others. These for Pichia species, which have previously been components of the must are able to penetrate described to hydrolyze xylane (Lee et al., the cell wall and participate in a variety of chem- 1986), it has not been reported previously for ical reactions that generate a range of volatile Hanseniaspora species. Another genus that is compounds (Boulton et al., 1996). able to produce this enzyme is . Candida Candida Although S. cerevisiae is the wine yeast par produces a b-D-xylosidase that can hydro- utilis excellence and the main yeast responsible for lyze precursors and increase the concentration the fermentation products, the contribution of of terpenes following addition of the purified non-Saccharomyces yeasts should not be enzyme to Moscatel grape must during fermen- forgotten either in natural fermentation or tation (Yanai & Sato, 2001). when a commercial strain of S. cerevisiae is inoc- There is only one purified a-L-arabinofurano- ulated. In the latter case, the influence of non- sidase from that effectively Pichia capsulata Saccharomyces yeasts is reduced, although it releases terpenes from precursors obtained has been shown that the use of starter cultures from Moscatel grape must. Its main characteris- does not prevent the growth or metabolic tics are that it is not inhibited by glucose and its activity of other natural strains of S. cerevisiae activity is stimulated by the presence of ethanol or of K. apiculata, H. uvarum, C. stellata,orToru- (Yanai & Sato, 2000). Another species from this laspora delbrueckii (Egli et al., 1998; Heard & genus, P. anomala, has been described as Fleet, 1986b, 1985; Henick-Kling et al., 1998; a producer of a-L-arabinofuranosidase activity Lema et al., 1996). (Spagna et al., 2002). Below, we describe the influence of the In a recent study addressing glycosidase different yeast species on the formation of the activities, a strain of Candida guilliermondii was main aromatic compounds originating from found to produce a-L-rhamnosidase (Rodrı´guez wine yeast metabolism that determine the et al., 2004). Table 4.2 shows the glycosidase secondary aroma. The biosynthesis of these activities produced by different species of non- compounds has been reviewed by Lambrechts Saccharomyces yeast. and Pretorius (2000). THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 97

TABLE 4.2 Glycosidases Produced by Non-Saccharomyces Yeasts

Glycosidases

Yeasts BGL XYL RAM ARA Reference

Brettanomyces bruxellensis X (1) Candida stellata X (2),(3) Candida pulcherrima X (3),(4)

Candida cacaoi X (20) Candida cantarelli X (5) Candida colliculosa X (3) Candida dattila X (5) Candida domerquiae X (5) Candida famata X (3)

Candida guilliermondii X X (4),(6) Candida hellenica X (2) Candida krusei X (3) Candida molischiana X (7),(21) Candida parapsilosis X (6) Candida peltata X (8)

Candida utilis X (9) Candida vinaria X (5) Candida vini X (5) Candida wickerhamii X (21) Cryptococcus albidus X (19)

Debaryomyces hansenii X (10),(11) Debaryomyces vanrijiae X (12),(13) Hanseniaspora guilliermondii X (5) Hanseniaspora osmophila X X (5),(14) Hanseniaspora uvarum X X (5),(14),(15) Kloeckera apiculata X (2),(3),(4),(6)

Metschnikowia pulcherrima X (5),(6) Pichia anomala X X X (3),(5),(14),(16) Pichia capsulata X (17)

(Continued) 98 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

TABLE 4.2 Glycosidases Produced by Non-Saccharomyces Yeastsdcont’d

Glycosidases

Yeasts BGL XYL RAM ARA Reference

Pichia membranaefaciens X (5) Pichia stipitis X (18) Zygosaccharomyces bailii X (5) Zygosaccharomyces mellis X (5) Zygosaccharomyces rouxii X (5)

ARA ¼ a-arabinofuranosidase; BGL ¼ b-glucosidase; RAM ¼ a-rhamnosidase; XYL ¼ b-xylosidase. (1) Mansfield et al. (2002); (2) Strauss et al. (2001); (3) Charoenchai et al. (1997); (4) Rodrı´guez et al. (2004); (5) Manzanares et al. (2000); (6) McMahon et al. (1999); (7) Genove´s et al. (2003); (8) Saha and Bothast (1996); (9) Yanai and Sato (2001); (10) Yanai and Sato (1999);(11)Riccio et al. (1999); (12) Belancic et al. (2003); (13) Garcıá et al. (2002); (14) Manzanares et al. (1999); (15) Fernańdez-Gonza´lez et al. (2003); (16) Spagna et al. (2002); (17) Yanai and Sato (2000); (18) Lee et al. (1986); (19) Peciarova´ and Biely (1982); (20) Drider et al. (1993); (21) Gunata et al. (1990).

3.2.2.1. VOLATILE FATTY ACIDS metabolism. It should be remembered that the Acetic acid is responsible for 90% of the volatile production of these fatty acids is also associated acidity of wines (Radler, 1993). The remaining with bacterial growth (Riberau-Gayon et al., 1998). fatty acids, such as propanoic and butanoic acid, Long-chain fatty acids (C16 and C18) are are present in small quantities as products of yeast essential precursors for the synthesis of many

TABLE 4.3 Principal Volatile Fatty Acids, Higher Alcohols, Esters, and Carbonyl Compounds Produced During Alcoholic Fermentation

Volatile fatty acids Higher alcohols Esters Carbonyl compounds

Acetic acid Propanol Ethyl acetate Acetaldehyde Butyric acid Butanol 2-Phenylethyl acetate Benzaldehyde Formic acid Isobutyl alcohol Isoamyl acetate Butanal Isobutyric acid Amyl alcohol Isobutyl acetate Diacetyl

Isovaleric acid Isoamyl alcohol Hexyl acetate Propanal Propionic acid Hexanol Ethyl butanoate Isobutanal

Valeric acid Phenylethanol Ethyl caprate Pentanal Hexanoic acid Ethyl caprylate Isovaleraldehyde Heptanoic acid Ethyl caproate 2-Acetyl tetrahydropyridine Octanoic acid Ethyl isovalerate

Nonanoic acid Ethyl 2-methylbutanoate Decanoic acid Tridecanoic acid

The most abundant compounds in wines are shown in boldface. THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 99 lipid compounds in yeast. These fatty acids aromatic compounds (Amerine et al., 1980). At occur in the cell membrane as esters, specifically the concentrations normally present in wine palmitoleic and oleic acids, which constitute (<300 mg/L), they contribute to the aromatic 70% of the cell membrane in S. cerevisiae complexity of the product. When their concen- (Ratledge & Evans, 1989). In general, these acids trations exceed 400 mg/L, they are considered do not appear in wines, but they are found in to have a negative effect on aroma (Rapp & products distilled in the presence of yeast lees. Mandery, 1986). Among the aliphatic higher In contrast, intermediate-chain fatty acids (C8, alcohols, the most predominant is isoamyl C10, and C12) do appear alongside their ethyl alcohol, although in this group propanol, isobu- esters as components of wine. tyl alcohol, and amyl alcohol are also produced. Although acetic and lactic acid bacteria can The aromatics form another class of higher alco- generate high levels of acetic acid, yeasts are hols, notable among which is 2-phenylethanol also involved in its production. Delfini and (Nyka¨nen et al., 1977). The importance of these Cervetti (1991) classified Saccharomyces yeast compounds is also related to their role as strains into three groups according to their precursors for the formation of esters (Soles production of acetic acid: low (0e0.3 g/L), inter- et al., 1982), compounds that are very important mediate (0.31e0.60 g/L), and high (>0.61 g/L). in wine aroma. Studies of acetic acid production in non-Saccha- Production of higher alcohols is a strain- romyces yeasts have generated highly variable specific characteristic and can be used to select results, and the concentrations reached may be yeasts for industrial applications (Giudici greater than or less than those produced by S. et al., 1990, 1993). The proportions of isoamyl cerevisiae. For instance, K. apiculata produces and amyl alcohol, isobutanol, and propanol between 1 and 2.5 g/L, C. stellata between 1 (Herraiz et al., 1990), and the production of and 1.3 g/L, Metschnikowia pulcherrima between dodecanol and tetradecanol (Longo et al., 0.1 and 0.15 g/L, Candida krusei 1 g/L, Hansenula 1992), are also specific to each strain. In general, anomala between 1 and 2 g/L, and T. delbrueckii studies of higher alcohol production in non- between 0.01 and 1.07 g/L (Fleet & Heard, Saccharomyces yeasts highlight the influence 1993; Renault et al., 2009). that these yeasts can have on the chemical In general, C8 and C10 fatty acids and their composition and quality of the wine (Gil et al., esters are produced in lower quantities by 1996; Herraiz et al., 1990; Longo et al., 1992; non-Saccharomyces yeasts than by S. cerevisiae Mateo et al., 1991). In fermented musts, the total (Herraiz et al., 1990; Ravaglia & Delfini, 1993; production of higher alcohols by pure cultures Renault et al., 2009; Rojas et al., 2001; Viana of non-Saccharomyces yeasts is lower than that et al., 2008). It is notable that the concentrations found with S. cerevisiae (Moreira et al., 2008; of short-chain fatty acids produced by non- Rojas et al., 2003; Viana et al., 2008, 2009). Saccharomyces yeasts are substantially below However, when non-Saccharomyces yeasts are the levels that can inhibit the growth of S. cerevi- used in mixed cultures, the difference is reduced siae and stop fermentation (Edwards et al., and the total quantity of higher alcohols is 1990). similar in all wines.

3.2.2.2. HIGHER ALCOHOLS 3.2.2.3. ESTERS The term “higher alcohol” encompasses Esters are the most abundant compounds those alcohols with more than two carbon atoms found in wine, with around 160 identiﬁed to and a higher molecular weight and boiling point date. In general, the concentration of esters in than ethanol. They are the largest group of wine is above the perception threshold (Salo, 100 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

1970a, 1970b), and some of the sensory descrip- 2-phenylethyl acetate. In addition to its produc- tors used in wine evaluation correspond to ester tion of ethyl acetate, the genus Pichia also stands aromas (Etievant, 1991; Maarse & Visscher, out as a good producer of isoamyl acetate. In 1989). The fresh, fruity aroma of young wines, terms of the ethyl esters, the genus Saccharo- for instance, is due to a mixture of esters gener- myces was the best producer of ethyl caproate, ated during fermentation, in particular acetate whereas the genus Torulaspora was the strongest esters (Ferreira et al., 1995; Marais, 1990). producer of ethyl caprylate. Although various esters can be formed The species and strain of yeast are, among during fermentation, the most abundant are other variables, determinants of the levels of those derived from acetic acid and higher alco- esters produced (Lambrechts & Pretorius, hols (ethyl acetate, isoamyl acetate, isobutyl 2000). For instance, species of the genus Hanse- acetate, and 2-phenylethyl acetate) and ethyl niaspora (H. guilliermondii, H. osmophila, and esters of saturated fatty acids (ethyl butanoate, H. uvarum) produce significant quantities of ethyl caproate, ethyl caprilate, and ethyl 2-phenylethyl acetate and isoamyl acetate caprate). (Moreira et al., 2005, 2008; Plata et al., 2003; Various genera of non-Saccharomyces yeasts Rojas et al., 2001, 2003; Viana et al., 2008, 2009), have been described as good producers of esters although there are notable differences among (Table 4.4). Candida, Hansenula, and Pichia strains (Viana et al., 2008). species have a greater capacity to produce ethyl acetate than wine strains of S. cerevisiae (Nyka¨- 3.2.2.4. CARBONYL COMPOUNDS nen, 1986; Ough et al., 1968). The genus Rhodo- Due to their low perception threshold and the torula has been reported to produce isoamyl characteristics that they confer on the wine acetate (Suomalainen & Lehtonen, 1979), (apple, lemon, and nutty aromas), volatile alde- whereas the genus Hanseniaspora, specifically hydes are among the most interesting carbonyl the species H. uvarum, is reported to be a good compounds. Acetaldehyde constitutes more producer of esters in general (Mateo et al., than 90% of the total aldehyde content of wines 1991; Romano et al., 1997; Sponholz, 1993). In (Nyka¨nen et al., 1977). Other carbonyl com- a recent study, Viana et al. (2008) grouped ester pounds of interest include diacetyl, which indi- production according to the yeast genus and cates growth of lactic acid bacteria when present demonstrated the capacity of the genus Hanse- at high concentrations (1e4 mg/L) (Sponholz, niaspora to produce acetate esters, particularly 1993), and the tetrahydropyridines responsible for the acetamide (mousy) aroma and tightly TABLE 4.4 Non-Saccharomyces Yeast Genera That linked to the growth of lactic acid bacteria and Produce Esters Brettanomyces (Heresztyn, 1986; Rapp, 1998). Data are available on the effect of non- Ester produced Saccharomyces yeasts on the total concentration Ethyl Isoamyl 2-Phenylethyl Ethyl of aldehydes in wine. The species K. apiculata, Genus acetate acetate acetate caproate C. krusei, C. stellata, H. anomala, and M. pulcher- Candida þ rima produce quantities ranging from undetect- able to 40 mg/L, whereas . produces þþ þ S cerevisiae Hanseniaspora between 6 and 190 mg/mL (Fleet & Heard, Pichia þþ 1993; Then & Radler, 1971). In a study des- Rhodotorula þ cribing the aromatic profile of different species of yeast, Romano et al. (2003) found little varia- þ Torulaspora tion in the production of acetaldehyde by 52 THE ROLE OF NON-SACCHAROMYCES YEASTS IN VINIFICATION 101 strains of S. cerevisiae (with a mean of approxi- caffeic acid) present in the grapes through the mately 50 mg/L), whereas there were signifi- sequential action of two enzymes. First, hydroxy- cant differences among the 59 strains of cinnamate decarboxylase converts hydroxy- H. uvarum studied (mean acetaldehyde concen- cinnamic acids into vinylphenols (4-vinyl tration of approximately 25 mg/L). Data have guaiacol and 4-vinylphenol), and these are then also been reported on the maximum production reduced to ethylphenols (4-ethylguaiacol and of benzaldehyde (1200 mg/L) achieved by 4-ethylphenol) by vinylphenol reductase. Their Schizosaccharomyces and Zygosaccharomyces spe- concentrations vary between 0 and 6047 mg/L, cies (Delfini et al., 1991). and when they exceed the perception threshold they are responsible for the phenolic aroma of 3.2.2.5. VOLATILE PHENOLS AND SULFUR wines. The presence of volatile phenols is always COMPOUNDS undesirable, since even at concentrations below From a quantitative point of view, volatile the perception threshold they are reported to phenols and sulfur compounds (Table 4.5) mask the fruity notes of white wines (Chatonnet make a lesser contribution to wine aroma than et al., 1992; Dubois, 1983). the compounds described above. However, Although it was thought that only species of qualitatively they are very important, since their the genera Brettanomyces/Dekkera were able to perception thresholds are very low and in transform hydroxycinnamic acids into ethyl- general they have a negative contribution to phenols, more recent studies have identified wine aroma. other non-Saccharomyces yeast strains with this Volatile phenols are generated by microbio- capacity, although only some strains of Pichia logical transformation of hydroxycinnamic guilliermondii displayed the same conversion acids (trans-ferulic, trans-r-coumaric, and capacity as Dekkera species (Dias et al., 2003; Renault et al., 2009; Shinohara et al., 2000). TABLE 4.5 Principal Phenolic and Sulfur Compounds However, the initial decarboxylation step of Produced During Alcoholic Fermentation hydroxycinnamic acids into vinylphenols is much more common in both the non-Saccharo- Phenolic compounds Sulfur compounds myces yeasts found in wine (e.g., Hanseniaspora, 2-Vinylphenol Hydrogen sulfide Pichia,andZygosaccharomyces species) and in 4-Vinyl guaiacol Dimethyl sulfide wine strains of S. cerevisiae (Chatonnet et al., 1992). Table 4.6 shows the main species of 4-Ethylphenol Diethyl sulfide yeast that have been identified as producers 4-Ethyl guaiacol Dimethyl disulfide of 4-ethylphenol, although it should be remem- Diethyl disulfide bered that this is a strain-dependent characteristic. The production of the ethylphenols found Methyl mercaptan in wine has been reviewed by Sua´rez et al. Ethyl mercaptan (2007). S-methyl thioacetate The sulfur compounds present in wine can be divided into various groups according to their S-ethyl thioacetate chemical structure: sulfides, heterocyclic poly- 4-Mercapto-4-methylpentan-2-one sulfide compounds, thioesters, and thiols. The (4MMP) sensory properties of these compounds vary 3-Mercaptohexan-1-ol (3MH) extensively, and, although most of them are associated with negative aromatic descriptors, 3-Mercaptohexyl acetate (3MHA) they can have a positive contribution to wine 102 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

TABLE 4.6 Species of Yeast that Produce 2008) and, although there were significant 4-Ethylphenol differences between the two species, the concentrations produced were similar to those of Yeasts 4-Ethylphenol S. cerevisiae. Brettanomyces lambicus þþ þ Candida cantarelli 4. DESIGN OF MIXED STARTER Candida halophila þ CULTURES Candida mannitofaciens þ The variability in the yeast flora present in Candida versatilis þ grape musts can be controlled by routine inocu- þ Candida wickerhamii lation with cultures that predominate and there- Debaryomyces hansenii þ fore standardize the initial flora, leading to homogeneous fermentation year after year. Dekkera anomala þþ This is the reason why nowadays fewer wineries þþ Dekkera bruxellensis produce their wines by natural or spontaneous Dekkera intermedia þþ fermentation and instead tend to induce fermentation of the must with selected strains Kluyveromyces lactis þ of . . The use of selected yeasts is þþ S cerevisiae Pichia guilliermondii common in large wineries and in the most tech- Torulaspora delbrueckii þ nologically advanced appellations, since it

þ¼weak producer; þþ ¼ strong producer. ensures rapid initiation of fermentation and, as a result, reduces the risk of oxidation and aroma through the introduction of fruity notes contamination. Prior selection of yeasts also (reviewed in Swiegers et al., 2005). allows wines with improved organoleptic prop- The most extensively studied sulfur erties to be produced. It should not be forgotten, compound is hydrogen sulfide, since it often then, that non-Saccharomyces yeasts can make occurs in musts low in nitrogen. Almost all a positive contribution to the winemaking studies have focused on S. cerevisiae and very process. These yeast species could be used as limited information is available on the produc- part of a strategy to obtain different types of tion of hydrogen sulfide by non-Saccharomyces wines, especially in terms of aromatic profile. yeasts. Strauss et al. (2001) included this char- The combination of different S. cerevisiae and acteristic, along with the production of extra- non-Saccharomyces strains in starter cultures cellular enzymes, in a study characterizing could be used to produce wines with unique non-Saccharomyces yeasts isolated from musts aromatic characteristics. Thus, the presence of and grapes from South Africa. Almost all of S. cerevisiae would prevent premature termina- the strains studied produced hydrogen sulfide, tion of fermentation and non-Saccharomyces with Candida species displaying the highest yeast species would introduce aromatic production. Species of the genus Hanseniaspora complexity. This proposal has been supported have also been reported to produce hydrogen by a number of authors in studies addressing sulfide, as has T. delbrueckii (Renault et al., the effect of non-Saccharomyces yeasts on the 2009; Viana et al., 2008). organoleptic characteristics of wines (Egli Recently, the capacity of H. uvarum and H. et al., 1998; Gil et al., 1996; Henick-Kling et al., guilliermondii to produce heavy sulfur com- 1998; Lema et al., 1996; Mateo et al., 1991; pounds has been evaluated (Moreira et al., Moreno et al., 1991; Romano et al., 1997; Zironi DESIGN OF MIXED STARTER CULTURES 103 et al., 1993). In two of those studies, emphasis described the influence of K. apiculata and C. was placed on the sensory characteristics of pulcherrima in mixed starter cultures with S. cer- the wines obtained. In a study involving wines evisiae. They concluded that the wines obtained produced from Riesling grapes, Henick-Kling had a different aromatic profile to those et al. (1998) concluded that the intensity of the obtained with S. cerevisiae alone and that none fruity aroma generated with non-Saccharomyces of the compounds produced had a negative yeasts was greater than that obtained with S. influence on the organoleptic quality of the cerevisiae starter cultures. In a similar study wine. Similar studies using musts with high involving Riesling and Chardonnay grapes, sugar contents indicated that wines obtained Egli et al. (1998) analyzed the organoleptic pro- with mixed starter cultures of H. uvarum, T. perties of wines produced by spontaneous fer- delbrueckii,orKluyveromyces thermotolerans and mentation in the presence of non-Saccharomyces S. cerevisiae had comparable or improved yeasts or with starter cultures from two strains analytic profiles compared to those obtained of S. cerevisiae. According to a panel of tasters, with S. cerevisiae alone (Ciani et al., 2006). the wines of both varieties achieved a higher However,thesesameyeastshaltedfermenta- score as a result of the greater intensity of their tion when inoculated sequentially. floral and fruity aromas when produced by As with S. cerevisiae strains currently used for spontaneous fermentation. controlled fermentations, the most rational In contrast to these spontaneous fermenta- approach with the greatest likelihood of success tions, recent studies have addressed the effects would be to select non-Saccharomyces yeasts on of mixed starter cultures on the aroma and the basis of their production both of enzymes structure of the wines. Soden et al. (2000) relevant for the winemaking process and of reported that wines produced using sequential metabolites that influence the quality of the fermentations with C. stellata AWRI 1159 and wine. Selection on the basis of enzyme produc- S. cerevisiae AWRI 838 had higher concentrations tion has been analyzed in studies addressing of succinic and acetic acid, glycerol, and ethyl the terpene fraction of a Moscatel wine acetate and lower concentrations of acetalde- produced with a mixed culture of S. cerevisiae hyde and ethanol than wines derived from and D. vanrijiae, which was chosen for its b- pure cultures of S. cerevisiae. The wines glucosidase activity. It was found that the wines produced with the mixed starter cultures also obtained with a mixed culture differed in their had a distinct aromatic character. Similar results concentrations of certain volatile compounds, were obtained in pilot-scale fermentations with in particular geraniol (Belancic et al., 2003; immobilized C. stellata and subsequent inocula- Garcıá et al., 2002). To address selection on the tion with S. cerevisiae (Ferraro et al., 2000). basis of metabolite production, Rojas et al. The effect of mixed and sequential starter (2003) studied the effect of mixed starter cul- cultures of C. cantarellii and S. cerevisiae has also tures using S. cerevisiae and non-Saccharomyces been analyzed during fermentation of musts yeast strains selected for their capacity to from the Syrah grape variety (Toro & Vazquez, produce 2-phenylethyl acetate (H. guilliermondii 2002). The main differences in the analytic profile CECT 11104) and ethyl acetate (P. anomala CECT of these wines, compared with those obtained 10590) in microbiological culture medium. using S. cerevisiae, were in the quantities of ace- However, when tested under winemaking toin, propanol, and succinic acid, as well as the conditions, 2-phenylethyle acetate was pro- higher concentrations of ethanol and glycerol duced by H. guilliermondii but isoamyl acetate (between 7.8 and 10% and between 44.3 and was not produced by P. anomala. In addition, 52.8%, respectively). Zohre and Erten (2002) although the concentration of ethyl acetate in 104 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS the wines obtained with mixed cultures was which are obtained from grapes infected with lower than in those obtained with pure cultures B. cinerea and can reach sugar concentrations of non-Saccharomyces yeasts, it was nevertheless of up to 450 g/L, is that conventional yeasts excessive and surpassed acceptable limits. produce excessive concentrations of acetic Based on the balanced production of secondary acid. Bely et al. (2008) have shown that a mixed metabolites and the organoleptic characteristics culture of T. delbrueckiieS. cerevisiae at a 20:1 of the wines, Mingorance-Cazorla et al. (2003) ratio is the most appropriate for improving the selected a strain of Pichia fermentans as a good analytic profile of these sweet wines. The wines candidate for use in mixed starter cultures. obtained have approximately half the volatile Subsequent studies showed that it was effective acidity and acetaldehyde concentration of sweet in musts sequentially inoculated with S. cerevi- wines produced with S. cerevisiae alone. That siae and gave rise to wines with higher levels study also addressed the influence of combined of esters, alcohol, and glycerol (Clemente- or sequential inoculation of the two yeasts and Jimeńez et al., 2005). the proportions of the species in the starter A simple solution to avoid the possible nega- culture. tive effects of non-Saccharomyces yeasts is to Studies of this type confirm the potential include both positive and negative features in offered by selecting non-Saccharomyces yeasts the selection criteria. Viana et al. (2008) included for use in mixed starter cultures. Taking into in the selection criteria not only the formation of account their general characteristics, it would acetate esters with positive effects on wine be possible to design mixed starter cultures aroma (2-phenylethyl acetate and isoamyl based on non-Saccharomyces yeasts that produce acetate) but also the excessive formation of ethyl macerating enzymes (with possible conse- acetate as a negative characteristic. In addition, quences on the technological aspects of wine- hydroxycinnamate decarboxylase activity and making) and/or glycosidases and acetate the production of hydrogen sulfide, acetalde- esters (with possible effects on wine aroma). hyde, acetic acid, and short-chain fatty acids Exploiting the enzymatic potential of non- were included as negative selection criteria. Saccharomyces yeasts could even represent an These criteria allowed selection of H. osmophila alternative to the use of exogenous enzymes in 1471 as a strain for use with S. cerevisiae in mixed the winemaking process, currently a common starter cultures. In a later study, the same practice in many wineries. Good selection of authors showed that wines with increased non-Saccharomyces yeasts based on enological levels of 2-phenylethyl acetate could be criteria may therefore help in the design of opti- produced using the mixed starter culture and mized mixed starter cultures without compro- that the concentrations of the ester could be mising wine quality. controlled by changing the proportions of the two yeasts in the starter culture (Viana et al., 2009). 5. FINAL CONSIDERATIONS The species T. delbrueckii, which is characterized by its low production of acetaldehyde, ace- It is clear that non-Saccharomyces yeasts have toin, acetic acid, and ethyl acetate (Cabrera et al., considerable unexploited potential as microor- 1988; Herraiz et al., 1990), has been proposed for ganisms able to influence both wine aroma inclusion in a mixed starter culture for fermen- and the winemaking process itself. More tation of musts with a high sugar content (Bely in-depth studies are now required to select et al., 2008). The problem with these musts, those with characteristics that exert a positive REFERENCES 105 influence on winemaking and to minimize Belancic, A., Gunata, Z., Vallier, M.-J., & Agosin, E. (2003). b-glucosidase from the grape native yeast negative effects. The use of non-Saccharomyces Debaryomyces yeasts in mixed starter cultures could extend vanrijiae: Purification, characterization, and its effect on monoterpene content of a Muscat grape juice. the range of the current market by allowing J. Agric. Food Chem., 51, 1453e1459. wines with defined organoleptic characteristics Bell, T. A., & Etchells, J. L. (1956). Pectin hydrolysis by to be produced. Finally, as non-Saccharomyces certain salt-tolerant yeasts. Appl. Microbiol., 4, 196e202. yeasts are natural components of the grape Beltrań, G., Torija, M. J., Novo, M., Ferrer, N., Poblet, M., microflora, these proposals are in line with the Guillamoń, J. M., et al. (2002). Analysis of yeast populations during alcoholic fermentation: A six year current trend in the food industry to use as follow-up study. Syst. Appl. Microbiol., 25, 287e293. few additives as possible and satisfy consumer Bely, M., Stoeckle, P., Masneuf-Pomare`de, I., & demand for natural products. Dubourdieu, D. (2008). Impact of mixed Torulaspora delbrueckii-Saccharomyces cerevisae culture on high- sugar fermentation. Int. J. Food Microbiol., 122, 312e320. Acknowledgments Beuchat, L. R. (1998). Progress in conventional methods for the detection and enumeration of foodborne yeasts. Food The work of the authors in the areas discussed in this Technol. Biotechnol., 36, 267e272. chapter has been supported by grants VIN01-018-C2-1 and Boriollo, M. F. G., Rosa, E. A. R., Bernardo, W. L. D. C., VIN03-007-C2-1 from the Spanish National Institute for Gonçalves, R. B., & Ho¨fling, J. F. (2003). Electrophoretic Agricultural and Food Science Research and Technology protein patterns and numerical analysis of Candida albi- (Instituto Nacional de Investigacioń y Tecnologıá Agraria cans from the oral cavities of healthy children. Rev. I. y Alimentaria [INIA]) and by grants AGL2003-01295 and Med. Trop., 45, 249e257. AGL2004-00978/ALI from the Spanish Ministry of Science Boulton, R. B., Singleton, V. H., Bisson, L. F., & Kunkee, R. E. and Education-FEDER. (1996). Principles and practices of winemaking. New York, NY: Chapman & Hall. Cabrera, M. J., Moreno, J., Ortega, J. M., & Medina, M. References (1988). Formation of ethanol, higher alcohols, esters and terpenes by five yeast strains in must from Pedro Amerine, A. M., Berg, H. V., Kunkee, R. E., Ough, C. S., Ximenez grapes in various degrees of ripeness. Am. Singleton, V. L., & Webb, A. D. (1980). The technology J. Enol. Vitic., 39, 283e287. of winemaking (4th ed.). Wesport, CT: AVI Technical Chambers, R. S., Broughton, M. J., Cannon, R. D., Carne, A., Books Inc. Emerson, G. W., & Sullivan, P. A. (1993). An exo-b-(1, 3)- Aryan, A. P., Wilson, B., Strauss, C. R., & Williams, P. J. glucanase of Candida albicans: Purification of the enzyme (1987). The properties of glycosidases of Vitis vinifera and and molecular cloning of the gene. J. Gen. Microbiol., 139, a comparison of their b-glucosidase activity with that of 325e334. exogenous enzymes. An assessment of possible appli- Charoenchai, C., Fleet, G. H., Henschke, P. A., & cations in enology. Am. J. Enol. Vitic., 38, 182e188. Todd, B. E. N. (1997). Screening of non-Saccharomyces Augustyn, O. P. H., Kock, J. L. F., & Ferreira, D. (1992). wine yeasts for the presence of extracellular hydrolytic Differentiation between yeast species, and strains within enzymes. Aust. J. Grape Wine Res., 3,2e8. a species by cellular fatty acid analysis. 5. A feasible Chatonnet, P., Dubourdieu, D., Boidron, J., & Pons, M. technique? Syst. Appl. Microbiol., 15, 105e115. (1992). The origin of ethylphenols in wines. J. Sci. Food Barata, A., Gonza´lez, S., Malfeito-Ferreira, M., Querol, A., & Agric., 60, 178e184. Loureiro, V. (2008b). Sour rot-damaged grapes are Ciani, M., Beco, L., & Comitini, F. (2006). Fermentation sources of wine spoilage yeasts. FEMS Yeast Res., 8, behaviour and metabolic interactions of multistarter 1008e1017. wine yeast fermentations. Int. J. Food Microbiol., 108, Barata, A., Seborro, F., Belloch, C., Malfeito-Ferreira, M., & 239e245. Loureiro, V. (2008a). Ascomycetus yeast species recov- Clemente-Jimeńez, J. M., Mingorance-Cazorla, L., Martıńez- ered from grapes damaged by honeydew and sour rot. Rodrı´guez, S., Las Heras-Va´zquez, F. J., & Rodrı´guez- J. Appl. Microbiol., 104, 1182e1191. Vico, F. (2005). Influence of sequential yeast mixtures on Barnett, J. A., Payne, R. W., & Yarrow, D. (1990). Yeasts: wine fermentation. Int. J. Food Microbiol., 98, 301e308. Characteristics and identification (2nd ed.). Cambridge, Combina, M., Mercado, L., Borgo, P., Elia, A., Jofre, V., UK: Cambridge University Press. Ganga, A., et al. (2005). Yeasts associated to Malbec 106 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

grape berries from Mendoza, Argentina. J. Appl. Micro- yeast populations and their effect on the sensory char- biol., 98, 1055e1061. acter of Riesling and Chardonnay wines. J. Appl. Micro- Connell, L., Stender, H., & Edwards, C. G. (2002). Rapid biol., 85, 779e789. detection and identification of Brettanomyces from Esteve-Zarzoso, B., Peris-Torań, M. J., Garcıá-Maiquez, E., winery air samples based on peptide nucleic acid anal- Uruburu, F., & Querol, A. (2001). Yeast population ysis. Am. J. Enol. Vitic., 53, 322e324. dynamics during the fermentation and biological aging Dea´k, T., & Beuchat, L. R. (1996). Handbook of food spoilage of sherry wines. Appl. Environ. Microbiol., 67, yeasts. Boca Raton, FL: CRC Press. 2056e2061. Degre´, R., Thomas, D. Y., Ash, J., Mailhiot, K., Morin, A., & Etievant, P. X. (1991). Wine. In H. Maarse (Ed.), Volatile Dubord, C. (1989). Wine yeast strain identification. Am. compounds in foods and beverages (pp. 483e587). New J. Enol. Vitic., 40, 309e315. York, NY: Marcel Dekker Inc. Delcroix, A., Gunata, Z., Sapis, J. C., Salmon, J. M., & Fernańdez, M., U´ beda, J. F., & Briones, A. I. (2000). Typing of Bayonove, C. (1994). Glycosidase activities of three non- Saccharomyces yeasts with enzymatic activities. enological yeast strains during winemaking: Effect on Int. J. Food Microbiol., 59,29e36. the terpenol content of Muscat wine. Am. J. Enol. Vitic., Fernańdez-Gonza´lez, M., di Stefano, R., & Briones, A. 45, 291e296. (2003). Hydrolysis and transformation of terpene Delfini, C., & Cervetti, F. (1991). Metabolic and technological glycosides from Muscat must by different yeast species. factors affecting acetic acid production by yeasts during Food Microbiol., 20,35e41. alcoholic fermentation. Vitic. Enol. Sci., 46, 142e150. Ferraro, L., Fatichenti, F., & Ciani, M. (2000). Pilot scale Delfini, C., Gaia, P., Bardi, L., Mariscalco, G., & Conteiro, M. vinification process using immobilized Candida stellata (1991). Production of benzaldehyde, benzyl alcohol and cells and Saccharomyces cerevisiae. Process Biochem., 35, benzoic acid by yeasts and Botrytis cinerea isolated from 1125e1129. grape musts and wines. Vitis, 30, 253e263. Ferreira, V., Fernańdez, P., Penã, C., Escudero, A., & Dias, L., Dias, S., Sancho, T., Stender, H., Querol, A., Mal- Cacho, J. F. (1995). Investigation on the role played by feito-Ferreira, M., et al. (2003). Identification of yeasts fermentation esters in the aroma of young Spanish wines isolated from wine-related environments and capable of by multivariate analysis. J. Sci. Food Agric., 67, 381e392. producing 4-ethylphenol. Food Microbiol., 20, 567e574. Fleet, G. H. (2003). Yeast interactions and wine flavour. Int. Dizy, M., & Bisson, L. F. (2000). Proteolytic activity of yeasts J. Food Microbiol., 86,11e22. strains during grape juice fermentations. Am. J. Enol. Fleet, G. H., & Heard, G. M. (1993). Yeast-growth during Vitic., 51, 155e167. winemaking. In G. H. Fleet (Ed.), Wine microbiology and Drider, D., Chemardin, P., Arnaud, A., & Galzy, P. (1993). biotechnology (pp. 27e54). Chur, Switzerland: Harwood Isolation and characterization of the exocellular b-gluco- Academic Publishers. sidase of Candida cacaoi: Possible use in carbohydrates Fugelsang, K. C. (1997). Wine microbiology. New York, NY: degradation. Lebensm. Wiss. U. Technol., 26, 198e204. Chapman & Hall. Duarte, F. L., Pais, C., Spencer-Martins, I., & Leaõ, C. (1999). Garcıá, A., Carcel, C., Dulau, L., Samson, A., Aguera, E., Distinctive electrophoretic isoenzyme profiles in Agosin, E., & Gunata, Z. (2002). Influence of a mixed Saccharomyces cerevisiae sensu stricto. Int. J. System. culture with Debaryomyces vanriji and Sacharomyces Bacterial, 49, 1907e1913. cerevisiae on the volatiles of a Muscat wine. J. Food Sci., Duarte, F. L., Pais, C., Spencer-Martins, I., & Leaõ, C. (2004). 67, 1138e1143. Isoenzyme patterns: A valuable molecular tool for the Genove´s, S., Gil, J. V., Manzanares, P., Aleixandre, J. L., & differentiation of Zygosaccharomyces species and detec- Valle´s, S. (2003). Candida molischiana b-glucosidase tion of misidentified isolates. Syst. Appl. Microbiol., 27, production by Saccharomyces cerevisiae and its application 436e442. in winemaking. J. Food Sci., 68, 2096e2100. Dubois, P. (1983). Volatile phenols in wine. In J. R. Piggot Gil, J. V., Mateo, J. J., Jimeńez, M., Pastor, A., & Huerta, T. (Ed.), Flavour of distilled beverages, origin and development (1996). Aroma compounds in wine as influenced by (pp. 110e119). Chichester, UK: Ellis Horwood. apiculate yeasts. J. Food Sci., 61, 1247e1249. Edwards, C. G., Beelman, R. B., Bartley, C. E., & Grossmann, M. K., Rapp, A., & Rieth, W. (1987). Enzyma- McConnell, A. L. (1990). Production of decanoic acid tische freisetzung gebundener aromastoffe in wein. and other volatile compounds on the growth of yeast Deut. Lebensm.-Rundsch, 83,7e12. and malolactic bacteria during vinification. Am. J. Enol. Guillamoń, J. M., Querol, A., Jimeńez, M., & Huerta, T. Vitic., 41,48e56. (1993). Phylogenetic relationships among wine yeast Egli, C. M., Edinger, W. D., Mitrakul, C. M., & Henick- strains based on electrophorectic whole-cell protein Kling, T. (1998). Dynamics of indigenous and inoculated patterns. Int. J. Food Microbiol., 18,115e125. REFERENCES 107

Guillamoń, J. M., Sabate´, J., Barrio, E., Cano, J., & Querol, A. Microbiologıá moderna de los alimentos (3rd ed.). Zaragoza, (1998). Rapid identification of wine yeast species based Spain: Editorial Acribia, SA. pp. 113e219. on RFLP analysis of the ribosomal internal transcribed Jiranek, V., Langride, P., & Henschke, P. A. (1995). Validation spacer (ITS) region. Arch. Microbiol., 169, 387e392. of bismuth-containing indicator media for predicting Guiudici, P., Romano, P., & Zambonelli, C. (1990). A H2S producing potential of Saccharomyces cerevisiae wine biometric study of higher alcohol production in yeast under enological conditions. Am. J. Enol. Vitic., 46, Saccharomyces cerevisiae. Can. J. Microbiol., 36,60e64. 269e273. Guiudici, P., Zambonelli, C., & Kunkee, R. E. (1993). King, A. P., Pitt, J. I., Beuchat, L. R., & Corry, J. (1986). Increased production of n-propanol in wine yeast strains Methods for the mycological examination of food. New York, having an impaired ability to form hydrogen sulphide. NY: Plenum Press. Am. J. Enol. Vitic., 44, 123e127. Kish, S., Sharf, R., & Margalith, P. (1983). A note on a selec- Gunata, Y. Z. (1984). Rećherches sur la fraction liee de nature tive medium for wine yeasts. J. Appl. Bacteriol., 55, glycosidique de l’arome du raisin: Importance des terpenyl- 177e179. glycosides, actino des glycosidases. Montpellier, France: Kreger-van Rij, N. J. W. (1984). The yeasts: A taxonomic study Doctorial thesis, Universite des Sciences et Techniques (3rd ed.). Amsterdam, the Netherlands: Elsevier du Languedoc. Science B.V. Gunata, Y., Bitteur, S., Brillouet, J. M., Bayonove, C. L., & Kurtzman, C. P., & Fell, J. W. (1998). The yeasts: A taxonomic Cordonnier, R. E. (1988). Sequencial enzymatic hydro- study (4th ed.). Amsterdam, the Netherlands: Elsevier Inc. lysis of potentially aromatic glycosides from grape. Laffort, J. F., Romat, H., & Darriet, P. (1989). Les levures et Carbohydr. Res., 184, 139e149. le`xpression aromatique des vins blancs. Revue des Gunata, Z., Bayonove, C. L., Cordonnier, R. E., Arnaud, A., & Oenologues, 53,9e12. Galzy, P. (1990). Hydrolysis of grape monoterpenyl Lagace, L. S., & Bisson, L. F. (1990). Survey of yeast acid glycosides by Candida molischiana and Candida wickerhamii protease for effectiveness of wine haze reduction. Am. b-glucosidases. J. Sci. Food Agric., 50,499e506. J. Enol. Vitic., 41, 147e155. Gunata, Y. Z., Dugelay, I., Sapis, J. C., Baumes, R., & Lambrechts, M. G., & Pretorius, I. S. (2000). Yeast and its Bayonove, C. (1994). Role of enzymes in the use of the importance to wine aroma. A review. S. Afr. J. Enol. Vitic., flavor potential from grape glycosides in winemaking. 21,97e129. In P. Schreier, & P. Winterhalter (Eds.), Progress in flavor LeClerc, M., Arnaud, A., Ratomahenina, R., & Galzy, P. precursor studies. Wurzburg, Germany: Proceedings of (1987). Yeast b-glucosidases. Biotechnol. Genet. Engin. the International Conference. Rev., 5, 269e295. Heard, G. M., & Fleet, G. H. (1985). Growth of natural yeast LeClerc, M., Arnaud, A., Ratomahenina, R., Galzy, P., & flora during the fermentation of inoculated wines. Appl. Nicolas, M. (1984). The enzyme system in a strain of Environ. Microbiol., 50, 727e728. Candida wickerhamii Meyer and Yarrow participating in Heard, G. M., & Fleet, G. H. (1986a). Evaluation of selective the hydrolysis of cellodextrins. J. Gen. Appl. Microbiol., media for enumeration of yeasts during wine fermen- 30, 509e521. tation. J. Appl. Bacteriol., 60, 477e481. Lee, H., Biely, P., Latta, R. K., Barbosa, M. F. S., & Heard, G. M., & Fleet, G. H. (1986b). Occurrence and growth Schneider, H. (1986). Utilization of xylan by yeasts and of yeast species during the fermentation of some its conversion to ethanol by Pichia stipitis strains. Appl. Australian wines. Food Technol. Australia, 38,22e25. Environ. Microbiol., 52, 320e324. Henick-Kling, T., Edinger, W., Daniel, P., & Monk, P. (1998). Lema, C., Garcia-Jares, C., Orriols, I., & Angulo, L. (1996). Selective effects of sulfur dioxide and yeast starter culture Contribution of Saccharomyces and non-Saccharomyces addition on indigenous yeast populations and sensory populations to the production of some components of characteristics of wine. J. Appl. Microbiol., 84, 865e876. Albarinõ wine aroma. Am. J. Enol. Vitic., 47, 206e216. Heresztyn, T. (1986). Formation of substituted tetrahy- Lightfoot, N. F., & Maier, E. A. (2002). Me´todos cuantitativos dropyridines by species Brettanomyces and Lactobacillus y evaluacioń de procedimientos. In Ana´lisis micro- isolated from mousy wines. Am. J. Enol. Vitic., 37, 127e132. biolo´gico de alimentos y aguas. Directrices para el asegur- Herraiz, T., Reglero, G., Herraiz, M., Martıń-Alvarez, P. J., & amiento de la calidad. Zaragoza, Spain: Editorial Acribia Cabezudo, M. D. (1990). The influence of the yeast and SA. pp. 95e133. type of culture on the volatile composition of wines Longo, E., Velazquez, J. B., Siereiro, C., Cansado, J., Calo, P., & fermented without sulphur dioxide. Am. J. Enol. Vitic., Villa, T. G. (1992). Production of higher alcohols, ethyl 41, 313e318. acetate, acetaldehyde and other compounds by 14 S. Jay, J. M. (1994). Identificacioń de los microorganismos y/o cerevisiae strains isolated from the same region (Salnes, de sus productos metabo´licos en los alimentos. In NW Spain). World J. Appl. Microbiol. Biotech., 8,539e541. 108 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

Loureiro, V., & Querol, A. (1999). The prevalence and Mendes-Ferreira, A., Clı´maco, M. C., & Mendes Faia, A. control of spoilage yeasts in foods and beverages. Trends (2001). The role of non-Saccharomyces species in realising Food Sci. Technol., 10, 356e365. glycosidic fraction of grape aroma components. A Luh, B. S., & Phaff, H. J. (1951). Studies on poly- preliminary study. J. Appl. Microbiol., 91,67e71. galacturonase of certain yeast. Arch. Biochem. Biophys., Mingorance-Cazorla, L., Clemente-Jimeńez, J. M., Martıńez- 33, 212e227. Rodrı´guez, S., Las Heras-Va´zquez, F. J., & Rodrı´guez- Maarse, H., & Visscher, C. A. (1989). Volatile compounds in Vico, F. (2003). Contribution of different natural yeasts to alcoholic beverages equalitative and quantitative data. Zeist, the aroma of two alcoholic beverages. World J. Microbiol. the Netherlands: TNO-CIVO Food Analysis Institute. Biotechnol., 19, 297e304. Malacrino, P., Zapparoli, G., Torriani, S., & Dellaglio, F. Moreira, N., Mendes, F., Guedes de Pinho, P., Hogg, T., & (2001). Rapid detection of viable yeasts and bacteria in Vasconcelos, I. (2008). Heavy sulphur compounds, wine by flow cytometry. J. Microbiol. Methods, 45, higher alcohols and esters production profile of Hanse- 127e134. niaspora uvarum and Hanseniaspora guilliermondii Malfeito-Ferreira, M., Tareco, M., & Loureiro, V. (1997). Fatty grown as pure and mixed cultures in grape must. Int. acid profiling: A feasible typing system to trace yeast J. Food Microbiol., 124, 231e238. contamination in wine bottling plants. Int. J. Food Moreira, N., Mendes, F., Hogg, T., & Vasconcelos, I. (2005). Microbiol., 38, 143e155. Alcohols, esters and heavy sulphur compounds Mansfield, A. K., Zoecklein, B. W., & Whiton, R. S. (2002). production by pure and mixed cultures of apiculate Quantification of glycosidase activity in selected strains wine yeasts. Int. J. Food Mocrobiol., 103, 285e294. of Brettanomyces bruxellensis and Oenococcus oeni. Am. Moreno, J. J., Millań, C., Ortega, J. M., & Medina, M. (1991). J. Enol. Vitic., 53, 303e307. Analytical differentiation of wine fermentations using Manzanares, P., Ramoń, D., & Querol, A. (1999). Screening pure and mixed yeast cultures. J. Indust. Microbiol., 7, of non-Saccharomyces wine yeasts for the production of 181e190. b-D-xylosidase activity. Int. J. Food Microbiol., 46, Nyka¨nen, L. (1986). Formation and occurrence of flavour 105e112. compounds in wine and distilled alcoholic beverages. Manzanares, P., Rojas, V., Genove´s, S., & Valle´s, S. (2000). A Am. J. Enol. Vitic., 37,84e96. preliminary search for anthocyanin-b-D-glucosidase Nyka¨nen, L., Nyka¨nen, I., & Suomalainen, H. (1977). activity in non- Saccharomyces wine yeasts. Int. J. Food Distribution of esters produced during sugar fermenta- Sci. Technol., 35,95e103. tion between the yeast cell and the medium. J. Inst. Brew., Marais, J. (1983). Terpenes in the aroma of grapes and 83,32e34. wines. A review. S. Afr. J. Enol. Vitic., 4,49e58. Ough, C. S., Cook, J. A., & Lider, L. A. (1968). Rootstock- Marais, J. (1990). Wine aroma composition. Food Rev., 17, scion interactions concerning wine making. II. Wine 18e21. compositional and sensory changes attributed to root- Martini, A., Ciani, M., & Scorzetti, G. (1996). Direct stock and fertilizer differences. Am. J. Enol. Vitic., 19, enumeration and isolation of wine yeasts from grapes 254e265. surfaces. Am. J. Enol. Vitic., 47, 435e440. Peciarova´, A., & Biely, P. (1982). b-xylosidases and Martini, A., & Vaughan-Martini, A. V. (1990). Grape must a nonspecific wall-bound b-glucosidase of the yeast fermentation, past and present. In J. F. T. Spencer, & Cryptococcus albidus. Biochim. Biophys. Acta (BBA) e D. M. Spencer (Eds.), Yeast technology (pp. 105e123). General Subjects, 716, 391e399. Berlin, Germany: Springer-Verlag. Pitt, J. I., & Hocking, A. D. (1985). Fungi and food spoilage. Masoud, W., & Jespersen, L. (2006). Pectin degrading Sydney, Australia: Academic Press. enzymes in yeasts involved in fermentation of Coffea Plata, C., Millań, C., Mauricio, J. C., & Ortega, J. M. (2003). arabica in East Africa. Int. J. Food Microbiol., 110,291 e296. Formation of ethyl acetate and isoamyl acetate by various Mateo, J. J., Jimeńez, M., Huerta, T., & Pastor, A. (1991). species of wine yeasts. Food Microbiol., 20, 217e224. Contribution of different yeasts isolated from musts of Radler, F. (1993). Yeasts-metabolism of organic acids. Monastrell grapes to the aroma of wine. Int. J. Food In G. H. Fleet (Ed.), Wine microbiology and biotechnology Microbiol., 14, 153e160. (pp. 165e182). Chur, Switzerland: Harwood Academic McKay,A. M. (1990). Degradation of polygalacturonic acid by Publishers. Saccharomyces cerevisiae. Lett. Appl. Microbiol., 11,41e44. Rapp, A. (1998). Volatile flavour of wine. Correlation McMahon, H., Zoecklein, B. W., Fugelsang, K., & Jasinski, Y. between instrumental analysis and sensory perception. (1999). Quantification of glycosidase activities in selected Nahrung, 42, 351e363. yeasts and lactic acid bacteria. J. Ind. Microbiol. Rapp, A., & Mandery, H. (1986). Wine aroma. Experientia, 42, Biotechnol., 23, 198e203. 873e884. REFERENCES 109

Rapp, A., & Versini, G. (1991). Influence of nitrogen Romano, P., Suzzi, G., Comi, G., Zironi, R., & Maifreni, M. compounds in grapes on aroma compounds of wine. In (1997). Glycerol and other fermentation products of J. M. Rantz (Ed.), In: Proceedings of the international apiculate wine yeasts. J. Appl. Microbiol., 82, 615e618. symposium on nitrogen in grapes and wines (pp. 156e164). Rosi, I., Vinella, M., & Domizio, P. (1994). Characterization Davis, CA: American Society for Enology and Viticulture. of b-glucosidase activity in yeasts of oenological origin. Raspor, P., Milek, D. M., Polanc, J., Mozina, S. S., & J. Appl. Bacteriol., 77, 519e527. Cadez, N. (2006). Yeasts isolated from three varieties of Rupela, O. P., & Tauro, P. (1984). Isolation and character- grapes cultivated in different locations of the Dolenjska ization of low hydrogen sulfide producing wine yeast. vine-growing region, Slovenia. Int. J. Food Microbiol., 109, Enzyme Microb. Technol., 6, 419e421. 97e102. Sabate´, J., Cano, J., Esteve-Zarzoso, B., & Guillamon, J. M. Ratledge, C., & Evans, C. T. (1989). Lipids and their (2002). Isolation and identification of yeasts associated metabolism. In The yeasts, Vol 3. London, UK: Academic with vineyard and winery by RFLP analysis of ribo- Press. pp. 367e455. somal genes and mitocondrial DNA. Microbiol. Res., 157, Ravaglia, S., & Delfini, C. (1993). Production of medium 267e274. chain fatty acids and their ethyl esters by yeast strains Saha, B. C., & Bothast, R. J. (1996). Production, purification, isolated from musts and wines. Ital. J. Food Sci., 5,21e36. and characterization of a highly glucose-tolerant novel Renault, P., Miot-Sertier, C., Marullo, P., Hernańdez-Orte, P., b-glucosidase from Candida peltata. Appl. Environ. Lagarrigue, L., Lonvaud-Funel, A., et al. (2009). Genetic Microbiol., 62, 3165e3170. characterization and phenotypic variability in Tor- Salo, P. (1970a). Determining the odor thresholds for some ulaspora delbrueckii species: Potencial applications in compounds in alcoholic beverages. J. Food Sci., 35,95e99. the wine industry. Int. J. Food Microbiol., 134, 201e210. Salo, P. (1970b). Variability of odour thresholds for some Renouf, V., Claisse, O., & Lonvaud-Funel, A. (2007). compounds in alcoholic beverages. J. Sci. Food Agric., 21, Inventory and monitoring of wine microbial consortia. 597e600. Appl. Microbiol. Biotechnol., 75, 149e164. Sancho, T., Gimeńez-Jurado, G., Malfeito-Ferreira, M., & Renouf, V., & Lonvaud-Funel, A. (2007). Development of an Loureiro, V. (2000). Zymological indicators: A new enrichment medium to detect Dekkera/Brettanomyces concept applied to the detection of potential spoilage bruxellensis, a spoilage wine yeast, on the surface of yeast species associated with fruit pulps and concen- grape berries. Microbiol. Res., 162, 154e167. trates. Food Microbiol., 17, 613e624. Ribereau-Gayon, P., Glories, Y., Maugean, A., & Schuller, P., Corte-Real, M., & Leao, C. (2000). A differential Dubordieu, D. (1998). Chimie du vin, stabilisation et medium for the enumeration of the spoilage yeast traitements, Vol. 2. Paris, France: Dunod. Zygosaccharomyces bailii in wine. J. Food Protect, 63, Riccio, P., Rossana, R., Vinella, M., Domizio, P., Zito, F., 1570e1575. Sansevrino, F., et al. (1999). Extraction and immobiliza- Schu¨ tz, M., & Gafner, J. (1993). Analysis of yeast diversity tion in one step of two b-glucosidases released from during spontaneous and induced alcoholic fermenta- a yeast strain of debaryomyces hansenii. Enzyme Microb. tions. J. Appl. Bacteriol., 75, 551e558. Technol., 24, 123e129. Serrat, M., Bermudez, R. C., & Villa, T. G. (2004). Poly- Rodrigues, N., Gonçalves, G., Pereira-da-Silva, M., Malfeito- galacturonase and ethanol production in Kluyveromyces Ferreira, M., & Loureiro, V. (2001). Development and use marxianus: Potential use of polygalacturonase in food- of a new medium to detect yeasts of the genera Dekkera/ stuffs. Appl. Biochem. Biotechnol., 117,49e64. Brettanomyces. J. Appl. Microbiol., 90, 588e599. Servili, M., Begliomini, A. L., Montedoro, G. G., Rodrı´guez, M. E., Lopes, C. A., van Broock, M., Valle´s, S., Petruccioli, M., & Federico, F. (1990). Use of an endo- Ramoń, D., & Caballero, A. C. (2004). Screening and polygalacturonase preparation from Cryptococcus albidus typing of Patagonian wine yeasts for glycosidase activ- var. albidus in manufacture of red wines. Industrie delle ities. J. Appl. Microbiol., 96,84e95. Bevande., 19, 332e336. Rojas, V., Gil, J. V., Pinãga, F., & Manzanares, P. (2001). Shinohara, T., Kubodera, S., & Yanagida, F. (2000). Distribu- Studies on acetate ester formation by non-Saccharo- tion of phenolic yeasts and production of phenolic off- myces wine yeasts. Int. J. Food Microbiol., 70, 283e289. flavours in wine fermentation. J. Biosci. Bioeng., 90,90e97. Rojas, V.,Gil, J. V.,Pinãga, F., & Manzanares, P.(2003). Acetate Soden, A., Francis, I. L., Oakey, H., & Henschke, P. A. (2000). ester formation in wine by mixed cultures in laboratory Effects of co-fermentation with Candida stellata and fermentations. Int. J. Food Microbiol., 86, 181e188. Saccharomyces cerevisiae on the aroma and composition Romano, P., Fiore, C., Paraggio, M., Caruso, M., & of Chardonnay wine. Aust. J. Grape Wine Res., 6,21e30. Capece, A. (2003). Function of yeast species and strains Soles, R. M., Ough, C. S., & Kunkee, R. E. (1982). Ester in wine flavour. Int. J. Food Microbiol., 86, 169e180. concentration differences in wine fermented by various 110 4. NON-SACCHAROMYCES YEASTS IN THE WINEMAKING PROCESS

species and strains of yeasts. Am. J. Enol. Vitic., 33, van Vuuren, H. J. J., & van der Meer, L. (1987). Finger- 94e98. printing of yeasts by protein electrophoresis. Am. J. Enol. Spagna, G., Barbagallo, R. N., Palmeri, R., Restuccia, C., & Vitic., 38,49e53. Giudici, P. (2002). Properties of endogenous b-glucosidase Vancanneyt, M., Pot, B., Hennebert, G., & Kersters, K. of a Pichia anomala strain isolated from Sicilian musts and (1991). Differentiation of yeast species based on electro- wines. Enzyme Microb. Technol., 31, 1036e1041. phoretic whole-cell protein patterns. System. Appl. Sponholz, W. R. (1993). Wine spoilage by microorganisms. Microbiol., 14,23e32. In G. H. Fleet (Ed.), Wine microbiology and biotechnology Vela´zquez, E., Cruz-Sańchez, J. M., Mateos, P. F., Monte, E., & (pp. 395e420). Chur, Switzerland: Harwood Academic Chordi, A. (1993). Prolewine: A program for wine- Publishers. making yeast identification. Binary-Comput. Microb., 5, Strauss, M. L. A., Jolly, N. P., Lambrechts, M. G., & van 58e61. Rensburg, P. (2001). Screening for the production of Vela´zquez, E., Cruz-Sańchez, J. M., Rivas-Pala´, T., Zurdo- extracellular hydrolytic enzymes by non-Saccharomyces Pin˜eiro, J. L., Mateos, P. F., Monte, E., et al. (2001). wine yeasts. J. Appl. Microbiol., 91, 182e190. YeastIdent-Food/ProleFood, a new system for the Sua´rez, R., Sua´rez-Lepe, J. A., Morata, A., & Calderoń, F. identification of food yeasts based on physiological and (2007). The production of ethylphenols in wine by yeasts biochemical tests. Food Microbiol., 18, 637e646. of the genera Brettanomyces and Dekkera: A review. Food Viana, F., Gil, J. V., Genove´s, S., Valle´s, S., & Manzanares, P. Chem., 102,10e21. (2008). Rational selection of non-Saccharomyces wine Suomalainen, H., & Lehtonen, M. (1979). The production of yeasts for mixed starters based on ester formation and aroma compounds by yeast. J. Inst. Brew., 85, 149e156. enological traits. Food Microbiol., 25, 778e785. Swiegers, J. H., Bartowsky, E. J., Henschke, P. A., & Viana, F., Gil, J. V., Valle´s, S., & Manzanares, P. (2009). Pretorius, I. S. (2005). Yeast and bacteria modulation of Increasing the levels of 2-phenylethyl acetate in wine wine aroma and flavour. In R. J. Blair, M. E. Francis, & I. through the use of a mixed culture of Hanseniaspora S. Pretorius (Eds.), Advances in wine science (pp. osmophila and Saccharomyces cerevisiae. Int. J. Food 159e187). Urrbrae, Australia: The Australian Wine Microbiol., 135,68e74. Research Institute. Yanai, T., & Sato, M. (1999). Isolation and properties of b- Then, R., & Radler, F. (1971). Vergleichende untersuchung glucosidase produced by Debaryomyces hansenii and its der acetaldehydbilding bei der aeroben verga¨rung von application in winemaking. Am. J. Enol. Vitic., 50, glucose bei verschiedenen sta¨mmen von Saccharomyces 231e235. cerevisiae und Saccharomyces carlsbergensis. Mschr. Yanai, T., & Sato, M. (2000). Purification and characteriza- Brauerei, 24, 127e130. tion of a novel a-L-arabinofuranosidase from Pichia Thongekkaew, J., Ikeda, H., Masaki, K., & Iefuji, H. (2008). capsulata X91. Biosci. Biotechnol. Biochem., 64, 1181e1188. An acidic and thermostable carboxymethyl cellulase Yanai, T., & Sato, M. (2001). Purification and characteriza- from the yeast Cryptococcus sp. S-2: Purification, char- tion of a b-D-xylosidase from Candida utilis IFO 0639. acterization and improvement of its recombinant Biosci. Biotechnol. Biochem., 65, 527e533. enzyme production by high cell-density fermentation of Yarrow, D. (1998). Methods for the isolation, maintenance Pichia pastoris. Protein Express. Purif., 60, 140e146. and identification of yeasts. In C. P. Kurtzman & Toro, M. E., & Vazquez, F. (2002). Fermentation behaviour of J. W. Fell (Eds.), The yeasts. A taxonomic study (4th ed.). controlled mixed and sequential cultures of Candida San Diego, CA: Elsevier Inc. pp. 77e100. cantarellii and Saccharomyces cerevisiae wine yeasts. World Zironi, R., Romano, P., Suzzi, G., Battistutta, F., & Comi, G. J. Microbiol. Biotechnol., 18, 347e354. (1993). Volatile metabolites produced in wine by mixed ´ Ubeda, J., & Briones, A. (2000). Characterization of differ- and sequential cultures of Hanseniaspora guilliermondii or ences in the formation of volatiles during fermentation Kloeckera apiculata and Saccharomyces cerevisiae. Bio- within synthetic and grape must by wild Saccharomyces technol Lett., 15, 235e238. strains. Lebensm.-Wiss. Technol., 33, 408e414. Zoecklein, B. W., Marcy, J. E., Williams, J. M., & Jasinski, Y. Urso, R., Rantsiou, K., Dolci, P., Rolle, L., Comi, G., & (1997). Effect of native yeasts and selected strains of Cocolin, L. (2008). Yeast biodiversity and dynamics Saccharomyces cerevisiae in glycosyl glucose, potential during sweet wine production as determined by volatile terpenes, and selected aglycones of White Ries- molecular methods. FEMS Yeast Res., 8, 1053e1062. ling (Vitis vinifera L.) wines. J. Food Composition Anal., 10, van Rensburg, P., & Pretorius, I. S. (2000). Enzymes in 55e65. winemaking: Harnessing natural catalysts for efficient Zohre, D. E., & Erten, H. (2002). The influence of Kloeckera biotransformations. A review. S. Afr. J. Enol. Vitic., 21, apiculata and Candida pulcherrima yeasts on wine 52e73. fermentation. Process Biochem., 38, 319e324. CHAPTER 5

Molecular Identification and Characterization of Wine Yeasts M. Teresa Fernańdez-Espinar 1, Silvia Llopis 1, Amparo Querol 1, Eladio Barrio 2 1 Departamento de Biotecnologıá, Instituto de Agroquı´mica y Tecnologıá de Alimentos (CSIC), Valencia, Spain and 2 Unitat de Gene`tica Evolutiva, Institut “Cavanilles” de Biodiversitat i Biologia Evolutiva, Universitat de Vale`ncia, Edificio de Institutos, Valencia, Spain

OUTLINE

1. Introduction 112 3.3. Restriction Analysis of Mitochondrial DNA 126 2. Methods for Species Identification 115 3.4. Polymerase Chain Reaction 2.1. Methods Based on Analysis (PCR)-based Methods 126 of Ribosomal DNA (rDNA) 115 3.4.1. Random Amplification of 2.1.1. Sequencing of Ribosomal DNA Polymorphic DNA (RAPD) 127 (rDNA) 115 3.4.2. Polymerase Chain Reaction 2.1.2. Restriction Analysis of (PCR) Analysis of Repetitive Ribosomal DNA (rDNA) 120 Genomic DNA (Microsatellites 2.2. Real-time Polymerase Chain Reaction and Minisatellites) 127 (PCR) 120 3.4.3. Amplification of d Sequences 128 2.3. Polymerase Chain Reaction (PCR)- 3.5. Amplified Fragment Length denaturing Gradient Gel Polymorphism (AFLP) 129 Electrophoresis (DGGE) 124 4. Applications 129 3. Methods to Differentiate Between 4.1. Analysis of Variation in Yeast Saccharomyces Cerevisiae Strains 125 Populations During Natural 3.1. Hybridization Techniques 125 Fermentation: Wine Ecology 129 3.2. Pulsed-field Gel Electrophoresis (PFGE) of Chromosomes 125

4.2. Analysis of Population Variation in 4.4. Identiﬁcation of New Species and Inoculated Fermentations: Monitoring Hybrids Involved in Wine Fermentation 132 Establishment 130 4.5. Detection of Wine Spoilage Yeasts 132 4.3. Characterization of Commercial Yeasts 131 Acknowledgments 133

1. INTRODUCTION for alcoholic fermentation. Although S. cerevisiae is only found at low levels on grapes, it The transformation of grape must into wine multiplies rapidly and displaces other microor- is a complex microbiological process involving ganisms present in the grape must. As a result yeasts and lactic acid bacteria, though only of its ability to tolerate high concentrations of yeasts of the genus Saccharomyces (principally alcohol and to thrive at higher temperatures Saccharomyces cerevisiae) are responsible for than other yeasts, S. cerevisiae comes to domi- alcoholic fermentation. Wine has traditionally nate the fermentation environment. During been produced by natural fermentation caused vinification, Saccharomyces strains can survive by the growth of yeasts derived from the and continue fermentation at temperatures of grapes and winery environment. The composi- up to 38C, whereas growth of most of the flora tion of the microflora on the surface of the present in the grape must is inhibited at grape is affected by a variety of factors, temperatures close to 25C. Despite these char- including temperature, rainfall, and other acteristics of S. cerevisiae, fermentation often climatic variables (Longo et al., 1991; Querol stops at high temperatures. et al., 1990); the ripeness of the crop (Martıńez Although S. cerevisiae is the most common et al., 1989; Rosini et al., 1982); the use of fungi- species in wine fermentations and has been cides (Bureau et al., 1982); physical damage the subject of most of the studies performed caused by fungi, insects, etc. (Longo et al., to date, other species belonging to the so-called 1991); and the grape variety. The surfaces of Saccharomyces sensu stricto complex, due to their winery equipment (presses, tanks, fermenters, phylogenetic proximity to S. cerevisiae, may also pumps, etc.) are another source of yeasts as be present during alcoholic fermentation and they come into contact with the grape must. even become the predominant species. For Apiculate yeasts of the genera Kloeckera and instance, Saccharomyces bayanus predominates Hanseniaspora (predominant species on the in wines from regions with a continental surface of the grape that account for 50% to climate and Saccharomyces paradoxus has been 70% of the total yeast population) and anaer- described recently to predominate in Croatian obic yeast of the genera Candida (Candida stel- wines (Redzepovic et al., 2002). Furthermore, lata and Candida pulcherrima), Cryptococcus, although Saccharomyces species form the Hansenula, Kluyveromyces, Pichia, and Rhodotor- majority of the flora resident in the winery ula grow during the initial phases of fermenta- (Fleet & Heard, 1993; Martini & Vaughan- tion, but increasing alcohol concentration and Martini, 1990), species belonging to the genera anaerobic conditions later favor the growth of Brettanomyces, Candida, Hansenula, and Pichia yeasts belonging to the genus Saccharomyces, have also been isolated in the winery environ- specifically S. cerevisiae, which are responsible ment and in finished wines. These may be INTRODUCTION 113 responsible for organoleptic changes that result on analysis of total cell proteins (Vacanneyt in wine spoilage, as has been observed for et al., 1991; van Vuuren & van der Meer, 1987), species belonging to the genera Pichia and Bret- isoenzyme profiles (Duarte et al., 1999), and tanomyces (Dias et al., 2003). analysis of fatty acids by gas chromatography Many of the studies designed to identify (Cottrell et al., 1986; Moreira da Silva et al., different species of yeast and different strains 1994; Tredoux et al., 1987). The reproducibility within the same species have been based on of these techniques, however, is somewhat ques- morphological and physiological criteria tionable, since in many cases they depend on the (Barnett et al., 1990; Kreger-van Rij, 1984). Exam- physiological state of the yeasts (Golden et al., ples of the characteristics that distinguish the 1994). Molecular biological techniques circum- main yeast species are shown in Table 5.1. These vent these difficulties by allowing direct anal- characteristics can vary according to the culture ysis of the genome, irrespective of the conditions (Scheda & Yarrow, 1966, 1968; physiological state of the cell. Many such tech- Yamamoto et al., 1991), and species are some- niques have now been developed and success- times defined by a single physiological char- fully applied to the identification and molecular acter that may even be controlled by a single characterization of yeasts. In this chapter, we gene. Consequently, identification depends upon describe the main techniques that have been the physiological state of the yeast. An example used for the analysis of wine yeasts. The princi- is seen in galactose fermentation, which has ples on which some of them are based have been traditionally been used by enologists to differ- described previously (Giudici & Pulvirenti, entiate between the species S. cerevisiae and 2002). Although some molecular studies of S. bayanus (Kurtzman & Phaff, 1987; Price non-Saccharomyces wine yeasts have been per- et al., 1978). More recently, methods have been formed (see Tables 5.1e5.8), research has developed to differentiate between yeasts based focused mainly on yeasts of the genus

TABLE 5.1 The Most Relevant Physiological and Morphological Characteristics for the Identiﬁcation of Predominant Species in Winemaking

Assimilation Fermentation Morphology Gal Glc Lac Mal Raf Sac Tre Gal Glc Lac Mal Raf Sac Tre

Candida stellata Globose/ovoid þþ/þ þ þ þ Dekkera bruxellensis Ellipsoid e oval þþ þ þþw þþ and elongated

Hansiniaspora Lemon-shaped þ uvarum Metschnikowia Globose/ellipsoid þþþþþ/þ pulcherrima Saccharomyces Ovoid v vv vvv þvv cerevisiae Zygosaccharomyces Ellipsoid/ovoid v (þ)v v þ v bailli

þ¼positive; ¼negative; v ¼ variable; Gal ¼ galactose; Glc ¼ glucose; Lac ¼ lactose; Mal ¼ maltose; Raf ¼ rafﬁnose; Sac ¼ saccharose; Tre ¼ trehelose. 114 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

TABLE 5.2 Studies Using Restriction Analysis of the 5.8S-ITS Ribosomal DNA Region for Species Identiﬁcation

Species studied Reference Application Observations

Non-Saccharomyces Constantı´ et al. Population dynamics, controlled Other techniques used: mtDNA yeasts; (1998) fermentation (effect of sulfite Saccharomyces dioxide and inoculum) cerevisiae Non-Saccharomyces Guillamoń et al. Identification of collection strains yeasts; S. cerevisiae (1998)

Non-Saccharomyces Granchi et al. (1999) Ecological study of spontaneous yeasts; S. cerevisiae fermentation

S. cerevisiae flor Fernańdez-Espinar Identification of collection strains et al. (2000) (from the flor of Jerez wines) Non-Saccharomyces Pramateftaki et al. Ecological study in natural Other techniques used: mtDNA; yeasts; S. cerevisiae (2000) fermentation d elements Non-Saccharomyces Torija et al. (2001) Population dynamics in a natural Other techniques used: mtDNA yeasts; S. cerevisiae fermentation Non-Saccharomyces Esteve-Zarzoso Population study in Jerez wines Other techniques used: mtDNA; yeasts; S. cerevisiae et al. (2001) karyotyping Non-Saccharomyces Sabate et al. (2002) Identification of species Other techniques used: mtDNA yeasts; S. cerevisiae associated with vines and wineries

Non-Saccharomyces Ganga and Martı´nez Ecological study of controlled yeasts (2004) fermentations

S. cerevisiae Capello et al. (2004) Ecological study in natural Other techniques used: mtDNA; fermentations d elements

Non-Saccharomyces Rodrı´guez et al. Ecological study in natural Other techniques used: mtDNA, yeasts; S. cerevisiae (2004) fermentation (analysis of b- karyotyping glucosidase activity)

Non-Saccharomyces Gonza´lez et al. Ecological study in natural Other techniques used: mtDNA yeasts; S. cerevisiae (2007) fermentations Non-Saccharomyces Lo´pes et al. (2007) Analysis of culture establishment Other techniques used: mtDNA yeasts; S. cerevisiae (effect on non-Saccharomyces ﬂora) Non-Saccharomyces Romancino et al. Ecological study in natural Other techniques used: restriction yeasts (2008) fermentation analysis of 26S rRNA

Non-Saccharomyces Zott et al. (2008) Population dynamics Other techniques used: yeasts sequencing of 5.8S-ITS rDNA

Non-Saccharomyces Tofalo et al. (2009) Population dynamics in a natural Other techniques used: yeasts; S. cerevisiae fermentation sequencing of D1/D2; RAPD (strain level)

ITS ¼ internal transcribed spacer; mtDNA ¼ mitochondrial DNA; RAPD ¼ random ampliﬁcation of polymorphic DNA; rDNA ¼ ribosomal DNA. METHODS FOR SPECIES IDENTIFICATION 115

TABLE 5.3 Studies Using Hybridization to Characterize Saccharomyces cerevisiae Strains

Species studied Reference Application Observations

S. cerevisiae Degre´ et al. (1989) Characterization of commercial Other techniques used: karyotyping strains

S. cerevisiae Querol et al. (1992a) Comparative study of Development of a new method for characterization techniques analysis of mtDNA Other techniques used: karyotyping

S. cerevisiae Lieckfeldt et al. (1993) Development of a PCR technique for Validation of the PCR technique by characterization comparison with hybridization

S. cerevisiae Lavalle´e et al. (1994) Quality control in the production of Other techniques used: d elements commercial yeasts

S. cerevisiae ﬂor Ibeas and Jime´nez (1996) Analysis of chromosome polymorphism

S. cerevisiae Nadal et al. (1999) Analysis of chromosome polymorphism in sparkling wines mtDNA ¼ mitochondrial DNA.

Saccharomyces, particularly S. cerevisiae,as gene is not part of the transcription unit, in a result of their importance in the winemaking yeasts it is located adjacent to it. The sequence process. Some of these studies have concluded conservation and concerted evolution of the that a combination of techniques is required 5.8S, 18S, and 26S ribosomal genes and the ITS for definitive characterization of individual and NTS spacers means that the similarity strains (Baleiras Couto et al., 1996; Fernańdez- between repeated transcription units within Espinar et al., 2001; Pramateftaki et al., 2000). a given species is greater than between units from different species. This sequence similarity within speciesdwhich arises through mecha- 2. METHODS FOR SPECIES nisms such as unequal crossing-over and gene IDENTIFICATION conversion (Li, 1997)dmakes these ribosomal DNA (rDNA) regions powerful tools with 2.1. Methods Based on Analysis which to identify species and establish phyloge- of Ribosomal DNA (rDNA) netic relationships between them (Kurztman & Robnett, 1998). Ribosomal genes (5.8S, 18S, and 26S) are Various methods have been developed to grouped in tandem to form transcription units identify yeast species based on information that are repeated 100 to 200 times throughout contained within these regions, as described the genome. Each transcription unit contains below. another two regions, the internal transcribed spacer (ITS) and the external transcribed spacer 2.1.1. Sequencing of Ribosomal DNA (ETS), both of which are transcribed but not pro- (rDNA) cessed. The coding regions are separated by Yeast species can be identified by comparison intergenic spacers (IGSs), also known as non- of nucleotide sequences from rDNA regions. transcribed spacers (NTSs). Although the 5S The two most commonly used regions are the 116 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

TABLE 5.4 Studies Using Electrophoretic Separation of Chromosomes or Electrophoretic Karyotyping for Strain Characterization

Species studied Reference Application Observations

Saccharomyces Blondin and Characterization of commercial cerevisiae Vezinhet (1988) strains S. cerevisiae Degre´ et al. (1989) Characterization of commercial Other techniques used: strains hybridization of total DNA

S. cerevisiae Vezinhet et al. (1990) Characterization of commercial Other techniques used: mtDNA strains

S. cerevisiae Yamamoto et al. Characterization of commercial (1991) and collection strains

S. cerevisiae Bidenne et al. (1992) Chromosomal polymorphism S. cerevisiae Querol et al. (1992a) Comparative study of Development of a new method for characterization techniques analysis of mtDNA. Other techniques used: hybridization

S. cerevisiae Frezier and Ecological study in natural Dubourdieu (1991) fermentation

S. cerevisiae van der Westhuizen Genetic improvement and Pretorius (1992)

Non-Saccharomyces Schu¨ tz and Gafner Population dynamics in Strain and species level yeasts; S. cerevisiae (1993) controlled and natural fermentations

S. cerevisiae Grando et al. (1994) Comparison of molecular Other techniques used: RAPD techniques

S. cerevisiae Schu¨ tz and Gafner Population dynamics in natural (1994) fermentations

S. cerevisiae ﬂor Martı´nez et al. Population dynamics during Other techniques used: mtDNA (1995) aging of Jerez wines

S. cerevisiae Versavaud et al. Ecological study and analysis of Other techniques used: mtDNA; (1995) geographical distribution in d elements natural fermentations

S. cerevisiae Briones et al. (1996) Ecological study of controlled fermentation

S. cerevisiae Nadal et al. (1996) Ecological study in natural Other techniques used: mtDNA fermentation

S. cerevisiae Egli et al. (1998) Population dynamics of natural Other techniques used: d elements and controlled fermentations

S. cerevisiae ﬂor Mesa et al. (1999) Ecological study during aging of Other techniques used: mtDNA Jerez wines

(Continued) METHODS FOR SPECIES IDENTIFICATION 117

TABLE 5.4 Studies Using Electrophoretic Separation of Chromosomes or Electrophoretic Karyotyping for Strain Characterizationdcont’d

Species studied Reference Application Observations

S. cerevisiae; S. Esteve-Zarzoso Population dynamics in Jerez Other techniques used: 5.8S-ITS cerevisiae ﬂor et al. (2001) wines (species level) mtDNA

S. cerevisiae Ferna´ndez-Espinar Authentication of commercial Other techniques used: mtDNA; et al. (2001) yeasts d elements

S. bayanus Naumov et al. Identiﬁcation in Hungarian wines Species level (2002)

S. cerevisiae Martı´nez et al. Geographic origin of native Other techniques used: mtDNA (2004) isolates

Non-Saccharomyces Rodrı´guez et al. Ecological study in natural Other techniques used: 5.8S-ITS yeasts; S. cerevisiae (2004) fermentation (analysis of b- (species level); mtDNA glucosidase activity)

S. cerevisiae Schuller et al. (2004) Characterization of commercial Other techniques used: mtDNA; strains d elements; microsatellites

ITS ¼ internal transcribed spacer; mtDNA ¼ mitochondrial DNA; RAPD ¼ random ampliﬁcation of polymorphic DNA.

D1 and D2 regions at the 50 end of the genes deoxynucleotides that interfere with the encoding the 26S (Kurtzman & Robnett, 1998) sequencing reaction. During automated and 18S (James et al., 1997) ribosomal subunits. sequencing, four fluorescent dyes are used to The availability of sequences in DNA databases, label the bases (A, G, C, and T). Dye incorpora- particularly for the D1/D2 region of the 26S tion is carried out by a second round of PCR gene, makes this technique particularly useful amplification using the same primers. Fine for assigning an unknown yeast to a specific capillaries are then used to separate the labeled species when the homology of the sequences is DNA fragments according to size. Laser excita- greater than 99% (Kurtzman & Robnett, 1998). tion of the dyes results in emission of a signal Sequence comparison with the DNA databases at a specific wavelength, and software can be is performed using the program WU-BLAST2, used to transform the signals into peaks, with available from http://www.ebi.ac.uk/Blas2/ each color corresponding to a nucleotide. This index.html. process is rapid and allows approximately 600 The use of direct sequencing of the regions of nucleotides to be read in 2 or 3 h, depending interest by polymerase chain reaction (PCR) has on the model of the sequencer. been combined with technology for automated Given the technological advances that have sequencing to make this a relatively rapid tech- been made and the widespread availability of nique. In this process, the target region is ampli- sequencing data via the Internet, sequencing fied by PCR from a total DNA template. The has become an extremely useful tool that PCR products are purified using commercial complements the other molecular techniques kits to remove the primers and excess described in this chapter. 118 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

TABLE 5.5 Studies Using Restriction Analysis of Mitochondrial DNA for Strain Characterization

Species studied Reference Application Observations

Saccharomyces Vezinhet et al. (1990) Characterization of commercial Other techniques used: cerevisiae strains karyotyping S. cerevisiae Querol et al. (1992a) Comparative study of Development of a new method for characterization techniques analysis of mtDNA. Other techniques used: karyotyping; hybridization

S. cerevisiae Querol et al. (1992b) Monitoring establishment of yeasts in controlled fermentations

S. cerevisiae Querol et al. (1994) Population dynamics in a natural fermentation Martı´nez et al. Population dynamics during Other techniques used: (1995) aging of Jerez wines karyotyping

S. cerevisiae; S. Versavaud et al. Ecological study and analysis of Other techniques used: cerevisiae (1995) geographical distribution in d elements; karyotyping ﬂor natural fermentations S. cerevisiae Guillamo´n et al. Geographical study Other techniques used: (1996) karyotyping

S. cerevisiae Nadal et al. (1996) Ecological study in natural Other techniques used: fermentation karyotyping

S. cerevisiae Gutie´rrez et al. Ecological study in spontaneous (1997) and controlled fermentations

S. cerevisiae Constantı´ et al. Population dynamics in Other techniques used: 5.8S-ITS (1998) a controlled fermentation (effect (species level) of sulfur dioxide and inoculation)

S. cerevisiae Sabate´ et al. (1998) Ecological study in natural fermentation

S. cerevisiae ﬂor Mesa et al. (1999) Molecular characterization Other techniques used: karyotyping

S. cerevisiae; Torriani et al. (1999) Ecological study in natural Other techniques used: RAPD Saccharomyces fermentation bayanus var. uvarum S. cerevisiae Comi et al. (2000) Diversity and geographic distribution on grapes

S. cerevisiae Pramateftaki et al. Ecological study in natural Other techniques used: 5.8S-ITS (2000) fermentation (species level)

S. cerevisiae; S. Esteve-Zarzoso Population dynamics in Jerez Other techniques used: 5.8S-ITS cerevisiae et al. (2001) wines (species level); karyotyping ﬂor

(Continued) METHODS FOR SPECIES IDENTIFICATION 119

TABLE 5.5 Studies Using Restriction Analysis of Mitochondrial DNA for Strain Characterizationdcont’d

Species studied Reference Application Observations

S. cerevisiae Ferna´ndez-Espinar Authentication of commercial Other techniques used: et al. (2001) yeasts karyotyping, d elements

S. cerevisiae Ferna´ndez- Population dynamics in Other techniques used: PCR- Gonza´lez et al. a controlled fermentation TTGE (2001)

S. cerevisiae Torija et al. (2001) Population dynamics in a natural Other techniques used: 5.8S-ITS fermentation (species level)

S. cerevisiae Beltra´n et al. (2002) Ecological study of controlled Other techniques used: 5.8S-ITS fermentation (species level)

S. cerevisiae Lo´pes et al. (2002) Ecological study and analysis of Other techniques used: d elements geographical distribution in natural fermentations

S. cerevisiae Sabate´ et al. (2002) Genetic diversity of strains on the Other techniques used: 5.8S-ITS vine and in the winery (species level)

S. cerevisiae Granchi et al. (2003) Ecological study of a natural fermentation (effect of nitrogen)

S. cerevisiae Torija et al. (2003) Population dynamics in the laboratory

S. cerevisiae Capello et al. (2004) Ecological study in natural Other techniques used: 5.8S-ITS; fermentations d elements

S. cerevisiae (flor) Esteve-Zarzoso Authentication and identification et al. (2001) of natural and collection isolates from Jerez wine flor S. cerevisiae Martıńez et al. Geographic origin of native Other techniques used: (2004) isolates karyotyping

Non-Saccharomyces Rodrı´guez et al. Ecological study in natural Other techniques used: 5.8S-ITS; yeasts; S. cerevisiae (2004) fermentation (analysis of b- karyotyping glucosidase activity)

S. cerevisiae Schuller et al. (2004) Characterization of commercial Other techniques used: yeasts d elements; karyotyping; microsatellites

S. cerevisiae Lo´pes et al. (2006) Biodiversity study

Non-Saccharomyces Gonza´lez et al. Ecological study in natural Other techniques used: 5.8S-ITS yeasts; S. cerevisiae (2007) fermentations Non-Saccharomyces Lo´pes et al. (2007) Analysis of yeast establishment Other techniques used: 5.8S-ITS yeasts; S. cerevisiae mtDNA ¼ mitochondrial DNA; ITS ¼ internal transcribed spacer; PCR-TTGE ¼ polymerase chain reaction-temperature gradient gel electrophoresis; RAPD ¼ random ampliﬁcation of polymorphic DNA; rDNA ¼ ribosomal DNA. 120 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

2.1.2. Restriction Analysis of Ribosomal and ITS2, which are amplified using the primers 0 0 DNA (rDNA) its1 (5 -TCC GTA GGT GAA CCT GCG G-3 ) 0 In an effort to develop techniques for use in and its4 (5 -TCC TCC GCT TAT TGA TAT GC- 0 industrial applications, other simpler methods 3 ), described by White et al. (1990). This tech- have been designed based on PCR amplification nique was used by Guillamoń et al. (1998) for of rDNA regions followed by restriction anal- the rapid identification of wine yeasts, and was ysis of the amplified products. The principles later extended to 191 yeasts associated with of PCR are described in Section 3.4. Various foodstuffs and beverages (de Llanos et al., studies have used small quantities of isolated 2004; Esteve-Zarzoso et al., 1999; Fernańdez- yeast colonies added to the PCR reaction Espinar et al., 2000). The amplified fragments directly without prior purification of DNA. In and restriction profiles for these species with this approach, an initial incubation step at HaeIII, HinfI, CfoI, and DdeI are available online 95C for 15 min is included in the amplification at http://yeast-id.com/. The technique has been protocol to release the DNA into the reaction used in numerous studies for the identification mixture. Thus, by removing the requirement of wine yeasts (see Table 5.2). for DNA purification, the time required is Restriction analysis of other rDNA regions reduced substantially. The amplification prod- has also been used to identify other yeast ucts are visualized following electrophoresis in species, particularly those belonging to the 1.4% agarose gels. Differently sized amplifica- Saccharomyces sensu stricto complex. This is the tion products correspond to different species; case for the NTS region (Baleiras Couto et al., when the amplicons are of the same size, 1996; Capece et al., 2003; Caruso et al., 2002; however, they do not always correspond to the Nguyen & Gaillardin, 1997; Pulvirenti et al., same species, and digestion of these fragments 2000), the 18S gene (Capece et al., 2003), and with restriction enzymes is required for defini- various domains of the 26S gene (Baleiras Couto tive identification. Digestion of PCR products et al., 1996, 2005; Romancino et al., 2008; Smole- is performed directly in the reaction mixture Mozina et al., 1997; van Keulen et al., 2003). without prior purification. The fragments gener- However, the absence of a sequence database ated are separated by electrophoresis in 3% for these regions means that their use cannot agarose gels and their size is assessed by be generalized for the identification of yeasts. comparison with appropriate DNA markers. This technique is both uncomplicated and 2.2. Real-time Polymerase Chain reproducible. Dlauchy et al. (1999) used this Reaction (PCR) method to amplify the 18S ribosomal gene and the ITS1 intergenic region from 128 species asso- Real-time PCR was developed in 1996 and ciated mainly with foodstuffs, wine, beer, and since then its use has increased almost exponen- soft drinks using the primers NS1 (50-GTA tially across a range of applications (Wilhelm & GTC ATA TGC TTG TCT C-30) and its2 (50- Pingoud, 2003). In this technique, the appearance GCT GCG TTC TTC ATC GAT GC-30) and of the amplification products is monitored digesting the PCR products with AluI, HaeIII, during each PCR cycle. It is based on the detec- MspI, and RsaI. This method was later used by tion and quantification of a signal generated by Redzepovic et al. (2002). Another very useful a fluorescent donor dye. The signal is in direct rDNA region that can be used to differentiate proportion to the quantity of PCR product in between species is that containing the 5.8S the reaction. The process is carried out in a ther- gene and the adjacent intergenic regions ITS1 mocycler coupled to a detector that can acquire METHODS FOR SPECIES IDENTIFICATION 121

TABLE 5.6 Studies Using Random Ampliﬁcation of Polymorphic DNA for Strain Characterization

Species studied Reference Application Observations

Saccharomyces cerevisiae; Baleiras Couto et al. (1994) Identiﬁcation of spoilage Species level Zygosaccharomyces species yeasts in alcoholic beverages S. cerevisiae Grando et al. (1994) Comparison of molecular Other techniques used: techniques karyotyping

S. cerevisiae Quesada and Cenis (1995) Identiﬁcation Genus, species, and strain level

S. cerevisiae Baleiras Couto et al. (1996) Characterization of wine Other techniques used: spoilage strains 5.8S-ITS (species level); microsatellites

S. cerevisiae; Saccharomyces Torriani et al. (1999) Ecological study in natural Other techniques used: bayanus var. uvarum fermentation mtDNA

S. cerevisiae Echeverrigaray et al. (2000) Characterization of commercial strains

S. cerevisiae Pe´rez et al. (2001a) Comparative study of Other techniques used: molecular techniques microsatellites; CAPS

Non-Saccharomyces Lopandic et al. (2008) Population study of natural Species level. Other yeasts; S. cerevisiae fermentation techniques used: sequencing of D1/D2; AFLP (Saccharomyces strains) S. cerevisiae Urso et al. (2008) Analysis of establishment Other techniques used: PCR- DGGE (biodiversity study)

Non-Saccharomyces Tofalo et al. (2009) Population dynamics in Other techniques used: yeasts; S. cerevisiae a natural fermentation sequencing of D1/D2 (species level); 5.8S-ITS (species level)

AFLP ¼ amplified fragment length polymorphism; CAPS ¼ cleaved amplified polymorphic sequence (digestion of RAPD products); ITS ¼ internal transcribed spacer; mtDNA ¼ mitochondrial DNA; PCR-DGGE ¼ polymerase chain reaction-denaturing gel gradient electrophoresis. and quantify the signal emitted by the donor in The fluorescent signal may be derived from each sample at the end of each cycle. The data intercalating agents or probes. The intercalating obtained are represented as an amplification agent SYBR green binds to double-stranded curve with the point at which the intensity of DNA, leading to an increase in fluorescence the signal from the donor becomes greater than with increasing amounts of PCR product. Three the background noise indicated. This is known types of probe can be used: hydrolysis probes, as the threshold cycle ( ) and it is inversely Ct hairpin probes, and hybridization probes. The proportional to the number of copies of the target most widely used hydrolysis probe is the Taq- sequence in the sample (DNA or cells). Conse- man probe, which has both donor and acceptor quently,it can be used to assess the starting quan- fluorochromes. When both fluorochromes are tity of target DNAwith a high degree of accuracy bound to the probe, the donor does not emit over a wide range of concentrations. a signal. When the probe is bound to a sequence 122 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

TABLE 5.7 Strain Characterization Studies Using Polymerase Chain Reaction (PCR) Ampliﬁcation of Variable Regions of the Genome (Microsatellites and Minisatellites)

Species studied Reference Application Observations

Saccharomyces Lieckfeldt et al. (1993) Development of Validation by comparison with cerevisiae a characterization technique hybridization techniques. Visualization in agarose gels

S. cerevisiae Baleiras Couto Characterization of wine Other techniques used: 5.8S- et al. (1996) spoilage strains ITS; RAPD

S. cerevisiae Gallego et al. (1998) Development of Visualization in an automatic a characterization technique sequencer

S. cerevisiae Hennequin et al. (2001) Development of Visualization using a characterization technique radioactivity following (wine, beer, baking, and acrylamide gel electrophoresis clinical isolates)

S. cerevisiae Pe´rez et al. (2001a) Comparative study of Other techniques used: molecular techniques microsatellites; CAPS

S. cerevisiae Pe´rez et al. (2001b) Development of Visualization in an automatic a characterization technique sequencer

S. cerevisiae Gonza´lez Techera Development of Visualization by silver staining et al. (2001) a characterization technique of acrylamide gels (commercial strains)

Kloeckera apiculata; Caruso et al. (2002) Ecological study in natural Microsatellites (based on S. cerevisiae fermentation Baleiras Couto et al., 1996). Other techniques used: ampliﬁcation and restriction of NTS region

Non-Saccharomyces Capece et al. (2003) Analysis of effectiveness of the Microsatellites (based on yeasts technique in non- Baleiras Couto et al., 1996. Saccharomyces strains Visualization in agarose gels.). Other techniques used: ampliﬁcation and restriction of 18S gene; ampliﬁcation and restriction of NTS region

S. cerevisiae Howell et al. (2004) Monitoring of laboratory Based on the protocol of fermentations Gonza´lez Techera et al. (2001)

S. cerevisiae Marinangeli et al. (2004) Development of Visualization on agarose gels a characterization technique

S. cerevisiae Shuller et al. (2004) Characterization of Microsatellites (based on Pe´rez commercial strains et al., 2001b). Other techniques used: mtDNA, d elements, karyotyping

S. cerevisiae Ayoub et al. (2006) Biodiversity of natural wine isolates

(Continued) METHODS FOR SPECIES IDENTIFICATION 123

TABLE 5.7 Strain Characterization Studies Using Polymerase Chain Reaction (PCR) Ampliﬁcation of Variable Regions of the Genome (Microsatellites and Minisatellites)dcont’d

Species studied Reference Application Observations

S. bayanus var. Masneuf-Pomare`de et al. Characterization of natural uvarum (2007) wine isolates

S. cerevisiae Richards et al. (2009) Characterization of Creation of a database on 246 commercial strains and genotypes natural isolates

AFLP ¼ amplified fragment length polymorphism; CAPS ¼ cleaved amplified polymorphic sequence (digestion of RAPD products); ITS ¼ internal transcribed spacer; mtDNA ¼ mitochondrial DNA; NTS ¼ non-transcribed spacer; RAPD ¼ random amplification of polymorphic DNA. of interest during the PCR reaction, the exonu- not require additional analyses such as electro- clease activity of Taq polymerase activates the phoresis following PCR. The lack of requirement donor fluorochrome in the rest of the probe, for additional procedures and the shorter reaction leading to emission of a fluorescent signal. This times and amplification cycles make real- signal is monitored as it accumulates during time PCR a very rapid technique. This is particu- successive PCR cycles. Hairpin probes (Molec- larly useful for routine analysis and applications ular Beacons, Scorpions) contain inverted requiring corrective measures. Nevertheless, tandem repeats (ITRs) at their 50 and 30 ends. In despite all the advantages offered by this type of the absence of the target sequence, this design system, designing the probes and primers is allows them to form a stem-loop structure very demanding, since it is this that determines through sequence complementarity between the specificity and sensitivity of the method. Soft- the two ITR regions. When the probe is bound ware is available to help in the design of appro- to the target DNA sequence, the separation of priate primers and probes for use in real-time the fluorochromes results in fluorescence. PCR. These are normally designed based on Finally, hybridization probes consist of two sequence data for genes or genomic regions that probes, a donor and an acceptor, both designed have demonstrated effectiveness for the establish- to bind to the amplified region and each labeled ment of phylogenetic relationships between yeast with a fluorophore. Resonance energy transfer species. These sequences also have the advantage only occurs when both probes are bound to the of being easily available via the Internet. Specifi- target DNA in close proximity. The choice of cally, they correspond to the ITS (James et al., which of these fluorescence systems to use is 1996) and D1/D2 (Kurztman & Robnett, 1998) influenced by the advantages and disadvantages rDNA regions, the mitochondrial gene COX2 that they each present. For instance, SYBR green (Belloch et al., 2000; Kurztman & Robnett, 2003), is most appropriate if a simple, cheap, and easy- and the nuclear gene actin (Daniel & Meyer, to-use system is required. However, during the 2003). These have been applied in real-time PCR PCR reaction it can bind primer dimers and other systems developed for the detection and quantifi- nonspecific products and lead to overestimation cation of total yeasts in wine (Hierro et al., 2006a) of the concentration of target DNA. The need for and for the monitoring of populations of Saccharo- greater specificity calls for the use of a system myces species and Hanseniaspora species during involving probes. alcoholic fermentation (Hierro et al., 2007). Occa- Real-time PCR has a number of advantages sionally, the differences in the nucleotide over other identification techniques. It is highly sequence between some species are insufficient specific and sensitive, quantitative, and does to allow design of primers, and molecular 124 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

TABLE 5.8 Studies Using Polymerase Chain Reaction (PCR) Ampliﬁcation of d Elements for Saccharomyces cerevisiae Strain Characterization

Species studied Reference Application Observations

S. cerevisiae Ness et al. (1993) Development of a novel method for Speciﬁc for S. cerevisiae strains characterization of wine strains

S. cerevisiae Lavalle´e et al. (1994) Quality control in the production of Other techniques used: commercial yeasts hybridization of total DNA

S. cerevisiae Versavaud et al. (1995) Ecological study and analysis of Other techniques used: mtDNA; geographical distribution in natural karyotyping fermentations

S. cerevisiae Egli et al. (1998) Population dynamics in Other techniques used: karyotyping spontaneous and controlled fermentations

S. cerevisiae Pramateftaki et al. (2000) Ecological study in natural Other techniques used: mtDNA fermentation

S. cerevisiae Ferna´ndez-Espinar et al. Authentication of commercial Other techniques used: mtDNA; (2001) yeasts karyotyping

S. cerevisiae Lo´pes et al. (2002) Ecology and geographical Other techniques used: mtDNA distribution study in spontaneous fermentation

S. cerevisiae Legras and Karst (2003) Characterization of commercial Optimization of the method yeasts described by Ness et al. (1993)

S. cerevisiae Capello et al. (2004) Ecological study in natural Other techniques used: mtDNA; fermentations 5.8S-ITS (species level)

S. cerevisiae Ciani et al. (2004) Analysis of the origin of strains responsible for spontaneous fermentation

S. cerevisiae Schuller et al. (2004) Characterization of commercial Other techniques used: mtDNA; yeasts karyotyping; microsatellites

S. cerevisiae Le Jeune et al. (2006) Population dynamics in spontaneous fermentations

ITS ¼ internal transcribed spacer; mtDNA ¼ mitochondrial DNA. markers with greater variability are needed. An 2.3. Polymerase Chain Reaction (PCR)- approach of this type was used by Martorell denaturing Gradient Gel Electrophoresis et al. (2005) and more recently by Salinas et al. (DGGE) (2009) using the random amplification of polymorphic DNA (RAPD) technique to design Recently, a genetic fingerprinting technique specific probes for species of S. cerevisiae.Cloning based on PCR amplification and denaturing and sequencing of a band obtained by RAPD gradient gel electrophoresis (DGGE) has been allowed specific primers for the species to be introduced into microbial ecology (Muyzer designed. et al., 1993). This technique allows DNA METHODS TO DIFFERENTIATE BETWEEN SACCHAROMYCES CEREVISIAE STRAINS 125 fragments of the same length to be separated on does not contribute to the phenotype of the the basis of sequence differences. Separation of yeast. Although mutations in these noncoding DNA amplicons is based on the decreased electro- DNA regions do not affect phenotype, they phoretic mobility of a partially melted double- can eliminate or create restriction sites. These stranded DNA molecule in polyacrylamide gels variations in restriction sites can be detected containing a linear gradient of denaturing agents by hybridization of DNA probes corresponding (a mixture of urea and formamide). DNA migra- to the affected regions. In this technique, the tion is retarded when the DNA strands dissociate restriction fragments obtained by digestion at a specific concentration of denaturing agent. of the DNA are separated on agarose gels Complete strand separation is prevented by the and transferred to nylon or nitrocellulose presence of a high-melting-point domain, created membranes by Southern blotting prior to by DNA amplification using particular groups of hybridization with the probes (Sambrook et al., universal primers. A sequence containing 1989. A similar technique can be applied to guanines (G) and cytosines (C) is added to the 50 chromosomes separated by pulsed-field gel end of one of the PCR primers, coamplified, and electrophoresis (PFGE). The probes are labeled thus introduced into the amplified DNA radioactively with 32P or non-radioactively fragments. with digoxigenin or biotin. The usefulness of A related technique is temperature gradient this technique for the characterization of S. cere- gel electrophoresis (TGGE), which is based on visiae strains or strains from other yeast species a linear temperature gradient for separation of has been demonstrated using probes against DNA molecules. DNA bands in DGGE and PFK2, PY30, and PDC1, which encode glycolytic TGGE profiles can be visualized using ethidium enzymes (Seehaus et al., 1985); TRP1 and TRP3, bromide or a more recent alternative, SYBR which code for enzymes involved in amino acid Green I. PCR fragments can be extracted from synthesis (Braus et al., 1985; Pedersen, 1983, the gel and used in sequencing reactions for 1985, 1986a, 1986b); and repetitive DNA regions species identification. such as the retrotransposons Ty1 and Ty2 Although the use of DGGE and TGGE in (Walmsley et al., 1989). However, few studies microbial ecology is still in its infancy, prelimi- have applied the technique in wine yeasts (see nary results are encouraging (Muyzer & Smalla, Table 5.3). Other repetitive DNA regions (Degre´ 1998). The methods have only recently been et al., 1989; Ibeas & Jimeńez, 1996; Lavalleé et al., used, however, for yeast identification in wine 1994; Lieckfeldt et al., 1993; Nadal et al., 1999) fermentations (Andorra` et al., 2008; Cocolin and genes encoding metabolic proteins such as et al., 2000; di Maro et al., 2007; Prakitchaiwat- Ura3 and TRP1 (Querol et al., 1992a) have also tana et al., 2004; Renouf et al., 2007; Stringini been used as probes. et al., 2009; Urso et al., 2008). 3.2. Pulsed-field Gel Electrophoresis (PFGE) of Chromosomes 3. METHODS TO DIFFERENTIATE BETWEEN SACCHAROMYCES In PFGE, the alternating application of two CEREVISIAE STRAINS transverse electric fields means that the chromosomes are continually forced to change the 3.1. Hybridization Techniques direction of their migration. As a result, large fragments of DNA are no longer detained in A large proportion of the S. cerevisiae genome the agarose gel matrix and can be separated. is not transcribed or translated and, therefore, The yeasts are grown in liquid medium and 126 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS then combined with molten agarose and placed which is 75% AT-rich but nevertheless contains in small molds. The cells are then lysed in situ some 200 GT-rich regions (Gray, 1989). GCAT- and the released DNA is immobilized in the type enzymes do not recognize either GC- or agarose matrix. The blocks are inserted into AT-rich regions in digestions of total DNA. agarose gels, which are then exposed to electric Consequently, given the small number of fields. The parameters that determine the reso- restriction sites in the mtDNA and the large lution of the bands are the electric-field switch- number of sites in the nuclear DNA, the latter ing intervals, the agarose concentration, the is digested into small fragments and the temperature, and the angle between the electric mtDNA bands can be clearly visualized over fields. the background shadow of the digested nuclear This method of karyotype analysis has been DNA. Not all enzymes reveal the same degree demonstrated to be highly efficient for the of polymorphism, and digestion patterns are differentiation of S. cerevisiae strains. The poly- highly species-dependent. In the specific case morphism revealed is the result of the addition of S. cerevisiae, the most appropriate enzymes or elimination of long fragments of DNA in to differentiate between strains are HinfI and homologous chromosomes during the evolution HaeIII (Guillamoń et al., 1994). Lo´pez et al. of the yeast genome (Casaregola et al., 1998; (2001) simplified this method with a modified Keogh et al., 1998; Wolfe & Shields, 1997). protocol that reduced the time required from Numerous studies that have used karyotype 77 to 25 h. This rapid technique enables a greater analysis to characterize wine strains of S. cerevi- number of strains to be analyzed in a shorter siae (see Table 5.4) have shown that it is a power- amount of time and is ideal for industrial appli- ful technique for differentiating between these cations due to its speed, safety, and low cost, strains. and because it does not require sophisticated equipment or highly trained personnel. 3.3. Restriction Analysis of Mitochondrial DNA 3.4. Polymerase Chain Reaction (PCR)-based Methods The mitochondrial DNA (mtDNA) of S. cerevisiae is a small, highly variable molecule of The quickest molecular techniques that have between 60 and 80 kb. The high degree of poly- been used to differentiate strains of wine yeast morphism revealed by restriction analysis of are those based on PCR (Saiki et al., 1985, mtDNA has made it one of the most commonly 1988). Some PCR-based techniques have been applied techniques for the characterization of developed to detect DNA polymorphisms wine strains of this species (see Table 5.5). without the use of restriction enzymes. The tech- Several methods have been developed to niques most frequently used to differentiate isolate yeast mtDNA (Aigle´ et al., 1984; Gar- between yeast strains are RAPD and microsatel- gouri, 1989; Querol & Barrio, 1990). However, lite analysis. Other techniques such as the the use of cesium chloride gradients and amplification of d sequences and intron splice ultracentrifugation make many of them inap- sites have been developed specifically to differ- propriate for industrial applications. To circum- entiate between wine strains of the species vent these difficulties, Querol et al. (1992a) S. cerevisiae. developed an approach to mtDNA analysis All of these techniques use oligonucleotide that does not require either technique. This primers, which bind to target sequences on simplified protocol relies on the composition each strand of the yeast DNA. The sequence of of A-T and G-C base pairs in the yeast mtDNA, the primers varies according to the technique, METHODS TO DIFFERENTIATE BETWEEN SACCHAROMYCES CEREVISIAE STRAINS 127 as discussed below. Amplification is carried out of the reactions can be considered. Given that with a thermostable DNA polymerase, and the the results obtained with several oligonucleo- protocol comprises a variable number of cycles tides must be combined to achieve good resolu- (generally between 25 and 45) that always tion, the technique is not appropriate for routine include denaturation of the DNA followed by industrial application. Consequently, it has not hybridization and extension. The result is been used extensively for the characterization amplification of the DNA, doubling the quantity of S. cerevisiae strains (see Table 5.6) and is of target DNA in each cycle. The amplification more widely applied in taxonomic studies (Fer- conditions, especially the hybridization temper- nańdez-Espinar et al., 2003; Molnar et al., 1995). ature, also differ. The amplification products are visualized in 1.4% agarose gels, and the strain- 3.4.2. Polymerase Chain Reaction (PCR) specific nature of the profiles allows differentia- Analysis of Repetitive Genomic DNA tion between strains. Below we discuss each of (Microsatellites and Minisatellites) these techniques in detail. The extensive variability of repetitive regions of genomic DNA makes them suitable targets 3.4.1. Random Amplification of Polymorphic for the molecular identification of yeast strains. DNA (RAPD) These motifs, known as microsatellites and The RAPD technique (Williams et al., 1990)is minisatellites, vary substantially in length and characterized by the use of a single short primer are present as tandem repeats distributed (around 10 nucleotides) that has a random randomly throughout the genome. Microsatel- sequence. The RAPD-PCR reaction is carried lites are usually shorter than 10 base pairs out at a low hybridization temperature (37C). whereas minisatellites are between 10 and 100 Thus, the pairings between the oligonucleotide base pairs in length. The variability found in and the DNA are determined by the short and these regions can be demonstrated by PCR random sequence of the primer and favored amplification using specific oligonucleotides, by the low hybridization temperature, leading such as (GTG)5, (GAG)5, (GACA)4, or M13. to the amplification of a range of DNA frag- The capacity of these oligonucleotides to reveal ments distributed throughout the genome. The polymorphism among strains of S. cerevisiae has result is a pattern of amplified products of been demonstrated previously by Lieckfeldt different molecular weight that can be charac- et al. (1993) using the hybridization techniques teristic of the species or of different strains or described in Section 3.1. The same authors isolates within the same species (Bruns et al., were the first to use these sequences as primers 1991; Paffetti et al., 1995). in the PCR reaction, and they demonstrated the The main advantage of RAPD is that no prior usefulness of the technique for strain character- sequence information is required in order to ization. The technique was later used by other design a primer. Furthermore, because the tech- authors (see Table 5.7), and Baleiras Couto nique allows analysis of variability throughout et al. (1996) used it successfully to characterize the entire genome, it reveals more polymor- spoilage strains of S. cerevisiae in alcoholic bever- phism than techniques that analyze specific ages. The amplified products obtained are regions. However, the low hybridization approximately 700 to 3500 base pairs long and temperature means that the amplification can therefore be visualized using agarose gels. profiles are unstable and difficult to reproduce, Recently, protocols similar to those used in and multiple reactions are required for each paternity testing and assessment of ancestry in sample using DNA from different extractions humans have been developed. Here, sequence as the template. Only the bands present in all data from S. cerevisiae databases is assessed to 128 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS identify repetitive regions. Sequences contain- S. cerevisiae wine strains based on the amplifica- ing microsatellite motifs are then used to design tion of these genes. In this technique, the high primers. This technique was applied for the first annealing temperatures used (60e65C) in the time in wine strains of S. cerevisiae by Gallego PCR reaction ensure stable and reproducible et al. (1998), although only four strains were amplification profiles. analyzed. Subsequently, Gonza´lez Techera et al. (2001) and Pe´rez et al. (2001b) designed 3.4.3. Amplification of d Sequences new primers to differentiate S. cerevisiae wine Delta sequences are 0.3 kb elements that flank strains. The primers designed by those groups Ty1 retrotransposons (Cameron et al., 1979). have recently been used to characterize Around 100 copies of the d element are present commercial wine strains of S. cerevisiae. Howell in the yeast genome as part of Ty1 retrotranspo- et al. (2004) used the primers designed by sons or as isolated elements. However, the Gonza´lez Techera et al. (2001) to monitor d sequences are concentrated in genomic commercial strains of S. cerevisiae during labora- regions adjacent to the transfer RNA genes tory-scale fermentations and Schuller et al. (Eigel & Feldmann, 1982). The number and (2004) used the method described by Pe´rez localization of these elements display a degree et al. (2001b) to characterize 23 commercial of intraspecific variability that Ness et al. wine strains. The study by Schuller et al. (1993) exploited to develop specific primers (d1 (2004) showed that the resolution with this tech- and d2) for the differentiation of S. cerevisiae nique is comparable to that obtained with strains. They demonstrated that the d elements d elements and restriction analysis of mtDNA. were stable enough for this technique to be Recently, various authors have proposed useful used on an industrial scale, and this was later methods for the identification of S. cerevisiae confirmed by other groups (see Table 5.8). based on PCR amplification of polymorphic Comparison with other high-resolution tech- regions of the genome using combinations of niques, such as restriction analysis of more than two primers in a single PCR reaction mtDNA and chromosome electrophoresis, has (Richards et al., 2009; Vaudano & Garcıá-Mor- shown that analysis of d elements can reveal uno, 2008). The system proposed by Richards extensive variability among S. cerevisiae isolates et al. (2009) is of particular interest since the (Fernańdez-Espinar et al., 2001; Pramateftaki authors have generated a database containing et al., 2000). 246 genotypes including 78 commercial wine Recently, Legras and Karst (2003) optimized strains along with other natural isolates from the technique by designing two new primers various different regions of the world. Clinical (d12 and d21) that are located very close to d1 isolates of S. cerevisiae have also been character- and d2. The use of d12 and d21 or of d12 and d2 ized using this method (Hennequin et al., 2001). reveals greater polymorphism, which is reflected Amplification products are usually visual- by the appearance of a larger number of bands. ized in acrylamide gels, although automatic Consequently, the new primers are able to differ- sequencers can also be used. Consequently, the entiate more strains, with 53 commercial strains technique is of little use in routine applications unequivocally differentiated in their study. despite its high resolution and reproducibility. Schuller et al. (2004) confirmed this finding by Marinangeli et al. (2004) observed that some showing that the combined use of d2 and d12 genes encodingcell-wall proteins from S. cerevisiae identified twice as many strains as the set of contain variable numbers of microsatellites that primers designed by Ness et al. (1993). lead to strain variation in gene size. Those authors An important drawback of this technique is the developed a method for the characterization of influence of DNA concentration on the profile APPLICATIONS 129 obtained, as shown by Fernańdez-Espinar et al. & Samaranayake, 2003; Theelen et al., 2001; (2001) and discussed later by Schuller et al. Trilles et al., 2003). The technique has neverthe- (2004). Although this problem is avoided by stan- less been used for the characterization of dardizing the concentration of DNA, comparison different species of wine yeast (Azumi & Goto- of results between laboratories is complicated. Yamamoto, 2001; Curtin et al., 2007; de Barros Another problem is the appearance of “ghost” Lopes et al., 1999; Flores Berrios et al., 2005; bands due to the low annealing temperature Lopandic et al., 2008). (42C) used during the amplification reaction. Recently, Ciani et al. (2004) used an annealing temperature of 55C to characterize wine strains 4. APPLICATIONS of S. cerevisiae. Although this resulted in much more stable amplification profiles, fewer bands In this section we will discuss industrial were obtained. applications of identifying species and strains of wine yeast. 3.5. Amplified Fragment Length Polymorphism (AFLP) 4.1. Analysis of Variation in Yeast Populations During Natural Although amplified fragment length poly- Fermentation: Wine Ecology morphism (AFLP) is fundamentally based on PCR amplification, we will consider it in a sepa- The microbiological fermentation that is used rate section, owing to its complex methodology to produce wine involves the growth of a series involving the use of other techniques. of microbial populations that have a direct influ- AFLP involves the restriction of genomic ence on the final product. It is therefore of interest DNA followed by the binding of adapters to to characterize the microbial ecology of this the fragments obtained and their selective process in an effort to control fermentation and, amplification by PCR. The adapter sequence ultimately, the final quality of the wine. This and restriction sites are collectively used as the requires techniques that can differentiate targets for the primers during PCR amplifica- between species. The molecular techniques that tion. The fragments are separated in DNA have been most widely used for the differentia- sequencing gels and visualized by autoradiog- tion of wine yeasts present in natural fermenta- raphy or automated sequencing (Vos et al., tions include electrophoretic karyotyping 1995). As in the case of RAPD, no prior sequence (Nadal et al., 1996; Schu¨ tz & Gafner, 1993); restric- information is required in order to design tion analysis of the 5.8S-ITS region (Granchi et al., primers. Furthermore, the technique is easily 1999; Pramateftaki et al., 2000; Rodrı´guez et al., reproduced and yields extensive information. 2004; Torija et al., 2001); restriction analysis of However, although AFLP is a useful technique other rDNA regions (van Keulen et al., 2003); to discriminate between yeast strains (de Barros and a combination of techniques, such as RAPD Lopes et al., 1999), it is very laborious, it requires and mtDNA restriction analysis (Torriani et al., automatic sequencers (which are not appro- 1999) or repetitive intergenic consensus PCR priate for routine industrial applications), and and PCR of intron splice sites combined with the data produced are difficult to interpret. restriction fragment length polymorphism Although the technique has been very widely (RFLP) and sequence analysis of the 5.8S-ITS used to study bacteria, plants, and animals, and D1/D2 rDNA regions (Hierro et al., 2006b). fewer studies have addressed its use in yeasts The results of these studies have (Boekhout, 2001; Borst et al., 2003; Dassanayake revealed microbial diversity not only between 130 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS wine-growing regions but also from year to year analysis (Caruso et al., 2002; Howell et al., in the same winery. Furthermore, the use of 2004), fluorescent in situ hybridization (FISH) these techniques has allowed the identification (Xufre et al., 2006), or a combination of tech- of yeasts that have not been described previ- niques (Egli et al., 1998; Lo´pes et al., 2002; Nadal ously in studies using conventional identifica- et al., 1996; Rodrı´guez et al., 2004; Torriani et al., tion methods. Sabate´ et al. (2002) undertook 1999; Versavaud et al., 1995). These studies have a study of the yeasts present on the vine and shown that different strains of S. cerevisiae domi- grape, in the winery, and during fermentation nate fermentation in different appellations. in the Spanish Priorat appellation using restric- Determining which strains dominate in each tion analysis of the 5.8S-ITS rDNA region. They region is essential in order to select appropriate found that the soil of the vineyards contained strains for use in starter cultures, since the use of strains of Hanseniaspora uvarum, a species that autochthonous strains to achieve controlled had previously been described as associated fermentations will help to maintain the sensory with the grapes and particularly as present characteristics typical of the region. during the initial phases of fermentation. They also isolated species of the genus Cryptococcus 4.2. Analysis of Population Variation in (Cryptococcus uniguttulatum, Cryptococcus lauren- Inoculated Fermentations: Monitoring tii, and Cryptococcus ateren) in the soil and on the Establishment stems and leaves of the vines, while other species, such as Aureobasdium pullulans, which Variability of the yeast flora in musts can be are typically associated with soil environments, reduced by addition of a microbial inoculum were found in the must. Studies undertaken in year after year. This inoculum normalizes the the Jerez region in the south of Spain by initial flora and gives rise to a homogeneous Esteve-Zarzoso et al. (2001) also identified soil fermentation irrespective of vintage. Over the species (Issatchenkia terricola) in the initial phases course of fermentation, the winery should check of fermentation. Those authors also reported that the inoculated yeast displaces the existing that the flor of biologically aged wines from flora and dominates fermentation as a result of the Jerez region contained not only S. cerevisiae its numerical superiority. In order to monitor but also Candida cantarrelli (in 91.6% of samples) the establishment of an inoculated strain, it is and Dekkera bruxellensis (which in this case did necessary to be able to differentiate it from the not cause wine spoilage). Finally, this technique other yeasts present over the course of fermenta- has been used to show that strains of Pichia guil- tion. This task is complicated by the fact that the liermondii produce 4-ethylphenol, which is inoculated yeast belongs to the same species as responsible for the aroma of stables in wines most of the other yeasts present in the must, (Dias et al., 2003). Prior to this study, only D. namely S. cerevisiae. The main techniques that bruxellensis and Dekkera anomala were known to have been used to analyze the ecology of produce this compound. controlled fermentations are electrophoretic Studies have also been undertaken to analyze karyotyping (Briones et al., 1996), restriction molecular variation in natural populations of analysis of mtDNA (Beltrań et al., 2002; S. cerevisiae using restriction analysis of mtDNA Constantı´ et al., 1998; Gutie´rrez et al., 1997), or (Granchi et al., 2003; Gutie´rrez et al., 1997; a combination of different techniques (Egli et al., Pramateftaki et al., 2000; Querol et al., 1994; 1998; Esteve-Zarzoso et al., 2001; Fernańdez- Sabate´ et al., 1998; Torija et al., 2001), electropho- Gonza´lez et al., 2001; Martıńez et al., 1995; retic karyotyping (Frezier & Dubourdieu, 1992; Mesa et al., 1999). Studies have also addressed Schu¨ tz & Gafner, 1993, 1994), microsatellite the effect of inoculating a commercial strain of APPLICATIONS 131

S. cerevisiae on the population of non-Saccharo- was originally selected and not a contaminant, myces yeasts in the must (Beltrań et al., 2002; and, secondly, to detect fraud. Constantı´ et al., 1998; Ganga & Martıńez, 2004; Given that most active dried yeasts belong to Urso et al., 2008). the species S. cerevisiae, the techniques used Although it is usually assumed that when must be able to differentiate clearly between a yeast starter culture is used growth of the strains. Most of the techniques described in autochthonous yeasts is suppressed, some Section 3 are useful for this purpose, as was studies have shown that these yeasts can still recently shown by Schuller et al. (2004) in participate in fermentation (Querol et al., a comparative study of 23 commercial strains 1992b; Schu¨ tz & Gafner, 1993, 1994), and in by electrophoretic karyotyping, restriction anal- some cases commercial yeasts only account for ysis of mtDNA, amplification of d elements, 50% of the population (Esteve-Zarzoso et al., and microsatellite analysis. Electrophoretic kar- 2000). Consequently, it is important to develop yotyping (Blondin & Vezinhet, 1998; Yamamoto simple methods for routine analysis of commer- et al., 1991), amplification of d elements (Legras cial yeasts in industrial fermentations. Querol & Karst, 2003; Ness et al., 1993), and microsatel- et al. (1992b) used restriction analysis of mtDNA lite analysis (Gonza´lez Techera et al., 2001) had to analyze the population dynamics of Saccharo- previously been used for this purpose. Other myces yeast strains during wine fermentation studies have been reported in which more than involving inoculation with an industrial strain. one technique was used to characterize commer- Lo´pez et al. (2002) developed a method for cial isolates: mtDNA analysis and karyotyping monitoring establishment of inoculated yeasts (Schuller et al., 2004; Vezinhet et al., 1990), ampli- based on PCR amplification of variable regions fication of d elements and DNA fingerprinting in the mtDNA. The method was based on vari- (Lavalleé et al., 1994), and karyotyping with ability in the number and position of introns in hybridization (Degre´ et al., 1989). In fact, the COX2 gene between strains of S. cerevisiae. Fernańdez-Espinar et al. (2001) showed that This method is particularly useful since it can definitive characterization of commercial strains be used to assess whether the inoculated yeast requires a combination of various molecular has become established within just 8 h, thus techniques. The techniques applied in that study allowing wineries to initiate corrective measures were restriction analysis of mtDNA with HinfI, to prevent stuck fermentation. electrophoretic karyotyping, and PCR amplification of genomic d elements. One of the most inter- 4.3. Characterization of Commercial esting findings reported by Fernańdez-Espinar Yeasts et al. (2001) was the large number of errors or fraudulent practices by companies that produce Dried wine yeasts were developed in the commercial yeasts. Commercial strains have 1950s when laboratories in Canada (Adams, also been characterized by Echeverrigaray et al. 1954) and the United States (Castor, 1953) inde- (2000) using the RAPD technique and by pendently carried out selection of wine strains Manzano et al. (2006) using TGGE-PCR and that were subsequently used in directed fermen- restriction analysis. De Barros Lopes et al. tations. More than 100 different strains are (1996) have developed a technique based on currently marketed, mainly by six companies. amplification of introns for the characterization Molecular characterization of commercial yeast of commercial strains. However, this technique strains is necessary for two reasons. Firstly, it has not been applied subsequently by other is needed for quality-control purposes to authors, possibly as a result of the complexity confirm that the obtained yeast is the one that of the profiles generated. 132 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

4.4. Identification of New Species and instance, Saccharomyces pastorianus is the result Hybrids Involved in Wine Fermentation of a cross between S. cerevisiae and Saccharomyces monacensis (currently included in the S. bayanus Clearly, only certain yeasts, selected to taxon) (Hansen & Kielland-Brandt, 1994; produce desirable organoleptic properties Yamagishi & Ogata, 1999) and the hybrid S. following fermentation of the must, are appro- bayanus-type strain derived from S. uvarum priate for use in starter cultures. Most of the (wine yeasts included in the S. bayanus taxon) yeast strains used to date belong to the species and S. cerevisiae (Nguyen & Gaillardin, 1997; S. cerevisiae and, to a lesser extent, S. bayanus. Nguyen et al., 2000). Although less frequently, These species have clearly different metabolic hybrids have also been identified in cider and characteristics, and, as a result, the choice of wine, as in the case of the cider strain CID1 which one to use in the fermentation depends mentioned earlier and the wine strain S6U on the desired outcome (Giudici et al., 1995; (Masneuf et al., 1998). Naumov et al., 1993, 2000; Torriani et al., 1999). Molecular techniques such as electrophoretic S. bayanus, for example, is cryotolerant and karyotyping, AFLP, and RAPD can reveal therefore used in fermentations carried out at hybrid character by analysis of the fraction of low temperatures. It would be of particular bands shared between a hybrid strain and interest to identify other species with new eno- each of the parental strains (Azumi & Goto- logic properties that are able to complete Yamamoto, 2001; de Barros Lopes et al., 2002; fermentation. Recently, in a study involving Fernańdez-Espinar et al., 2003; Masneuf et al., restriction analysis of the 18S-ITS1 rDNA 1998). These techniques are time-consuming region, S. paradoxus was isolated in Croatian and laborious. In contrast, the method of restric- wines, where it was found to predominate tion analysis and sequencing of the nuclear gene during fermentation (Redzepovic et al., 2002). Met2 developed by Masneuf et al. (1996) has As discussed below, the presence in alcoholic been used successfully in combination with fermentations of natural hybrids resulting analysis of mitochondrial genes (ATP8, ATP9, from crosses between different species of the or SSU) for the identification of double and S. cerevisiae sensu stricto complex suggests that triple hybrids (Groth et al., 1999; Masneuf some species normally associated with natural et al., 1998). Other more recent approaches environments may be present in fermentations, have involved the amplification of a nuclear as in the case of S. paradoxus. For instance, DNA region (YBR033w) without the need to Saccharomyces kudriavzevii was identified as one resort to restriction analysis of the amplicons of the parent species of the hybrid cider strain (Torriani et al., 2004), and restriction analysis CID1, along with S. cerevisiae and Saccharomyces of five nuclear genes (CAT8, CYR1, GSY1, uvarum (Groth et al., 1999), and other hybrid MET6, and OPY1) from different chromosomes strains that predominate in wine fermentations and the 5.8S-ITS rDNA region alongside from central European regions (Gonza´lez et al., sequence analysis of the mitochondrial gene 2006; Lopandic et al., 2007). COX2 (Gonza´lez et al., 2006). In addition to identifying these and other, as yet unidentified, species in winemaking 4.5. Detection of Wine Spoilage Yeasts contexts, it would be of interest to identify hybrids that are better adapted than their parent Wine is a highly appropriate culture medium strains to those winemaking conditions in for the growth of a large number of microorgan- which they have arisen. The presence of these isms, in part due to its richness in organic acids, hybrid species is common in brewing. For amino acids, residual sugars, growth factors, REFERENCES 133 and mineral salts. The main negative effects of to release of the products onto the market. Ibeas yeasts in wine are the generation of undesirable et al. (1996) have developed a system for detec- aromas and flavors during winemaking and the tion of species of the genera Dekkera/Bretanomy- formation of biofilms or turbidity, or the produc- ces based on two consecutive PCR reactions tion of gas, during storage. (nested PCR). Using this system, they were Various species of wine spoilage yeast have able to detect contaminations of less than 10 been described. The typical contaminating cells in samples of Jerez wine. Currently, real- species found during winemaking belong to time or quantitative PCR represents a good the genera Pichia and Candida or the species alternative technique with which to resolve Saccharomycodes ludwigii. Zygosaccharomyces bai- these types of problem, since it is both rapid lii, S. cerevisiae, and S. ludwigii are the principal and highly sensitive. Systems of this type have contaminants during bottling, whereas species been developed by Phister and Mills (2003) belonging to the genera Dekkera/Brettanomyces and Delaherche et al. (2004) for the detection are found during barrel aging. The high sugar and quantification of D. bruxellensis strains. content of sweet and sparkling wines favors Recently, Hayashi et al. (2007) developed a set the growth of Zygosaccharomyces species, partic- of primers for the detection and identification ularly Z. bailii. Few studies have addressed the of Brettanomyces/Dekkera species using the ITS identification and molecular characterization rDNA region. The technique employed a novel of these species, and those that have done so loop-mediated isothermal amplification used standard techniques such as analysis of method, which appears to be more specific, mtDNA or microsatellites, restriction analysis sensitive, and straightforward than standard of rDNA, or RAPD, alone or in combination PCR techniques. The use of such systems for (Baleiras Couto et al., 1994, 1996; Cocolin et al., other species of wine spoilage yeast would be 2004). Less common techniques have also been of particular interest for wineries in order to used, such as those based on restriction analysis avoid spoilage during wine storage prior to sale. combined with PFGE (Miot-Sertier & Lonvaud- Funel, 2007; Oelofse et al., 2009) or FISH (Ro¨der et al., 2007; Stender et al., 2001). Acknowledgments There are no legal limits in terms of the number Research by our group is funded by the Spanish Interminis- of yeasts permitted in wine, but recommenda- terial Commission on Science and Technology tions do exist. The International Organization of (Ref. AGL2006-12703-CO2 and AGL2009-12673-CO2) and Vine and Wine (OIV) recommends a maximum by the Regional Government of Valencia (Generalitat Valenciana; Ref. PROMETEO/2009/019). of 102e105 colony-forming units (CFU) per mL. Wineries apply their own criteria regarding acceptable levels of contamination, and these References are much stricter than the OIVrecommendations. For instance, they recommend no more than Adams, A. M. (1954). A simple continuous propagator for yeast. 1 CFU/mL in sweet wines (Loureiro & Malfeito- Rept. Ontario Hort. Expt. Sta., and Products Lab 102e103. Aigle´, M., Erbs, D., & Moll, M. (1984). Some molecular Ferreira, 2003). structures in the genome of larger brewing yeast. Am. Clearly, techniques for the detection of Soc. Brew. Chem., 42,1e7. spoilage yeasts are essential. These techniques Andorra`, I., Landi, S., Mas, A., Guillamoń, J. M., & Esteve- must be very sensitive and allow quantification Zarzoso, B. (2008). Effect of oenological pratices on microbial populations using culture-independent tech- of the number of microorganisms present. They niques. Food Microbiol., 25, 849e856. must also be rapid to allow the application of Ayoub, M. J., Legras, J. L., Saliba, R., & Gaillardin, C. (2006). corrective measures on the production line prior Application of multi locus sequences typing to the 134 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

analysis of the biodiversity of indigenous Saccharomyces fermenting musts according to their karyotype patterns. cerevisiae wine yeasts from Lebanon. J. Appl. Microbiol., Int. J. Food Microbiol., 28, 369e377. 100, 699e711. Bruns, T. D., White, T. J., & Taylor, J. W. (1991). Fungal Azumi, M., & Goto-Yamamoto, N. (2001). AFLP analysis of molecular systematics. Annu. Rev. Ecol. Syst., 22, 524e564. type strains and laboratory and industrial strains of Bureau, G., Brun, O., Vigues, A., Maujean, A., Vesselle, G., & Saccharomyces sensu stricto and its application to Feuillat, A. (1982). Etude d’une microflore levurienne phenetic clustering. Yeast, 18, 1145e1154. champenoise. Connaiss. Vigne Vin., 16,15e32. Baleiras Couto, M. M., Eijsma, B., Hofstra, H., Huis in’t Cameron, J. R., Loh, E. Y., & Davis, R. W. (1979). Evidence Veld, J. H. H., & van der Vossen, J. M. B. M. (1996). for transposition of dispersed repetitive DNA families in Evaluation of molecular typing techniques to assign yeast. Cell, 16, 739e751. genetic diversity among strains of Saccharomyces cer- Capece, A., Salzano, G., & Romano, P. (2003). Molecular evisiae. Appl. Environ. Microbiol., 62,41e46. typing techniques as a tool to differentiate non- Baleiras Couto, M. M., Reizinho, R. G., & Duarte, F. L. (2005). Saccharomyces wine species. Int. J. Food Microbiol., 84, Partial 26S rDNA restriction analysis as a tool to charac- 33e39. terise non- Saccharomyces yeasts present during red wine Cappello, M. S., Bleve, G., Grieco, F., Dellaglio, F., & fermentations. Int. J. Food Microbiol., 102,49e56. Zacheo, G. (2004). Characterization of Saccharomyces Baleiras Couto, M. M., van der Vossen, J. M., Hofstra, H., & cerevisiae strains isolated from must of grape grown in Huis in’t Veld, J. H. (1994). RAPD analysis: A rapid experimental vineyard. J. Appl. Microbiol., 97, 1274e1280. technique for differentiation of spoilage yeasts. Int. Caruso, M., Capece, A., Salzano, G., & Romano, P. (2002). J. Food Microbiol., 24, 249e260. Typing of Saccharomyces cerevisiae and Kloeckera apiculata Barnett, J. A., Payne, R. W., & Yarrow, I. J. (1990). Yeast: strains from Aglianico wine. Lett. Appl. Microbiol., 34, Characterization and identification (2nd ed.). London, UK: 323e328. Cambridge University Press. Casaregola, S., Nguyen, H. V., Lepingle, A., Brignon, P., Belloch, C., Querol, A., Garcia, M. D., & Barrio, E. (2000). Gnedre, F., & Guillardin, C. (1998). A family of labora- Phylogeny of the genus Kluyveromyces inferred from the tory strains of Saccharomyces cerevisiae carry rearrange- mitochondrial cytochrome-c oxidase II gene. Int. J. Syst. ments involving chromosomes I and III. Yeast, 14, Evol. Microbiol., 50, 405e416. 551e561. Beltrań, G., Torija, M. J., Novo, M., Ferrer, N., Poblet, M., Castor, J. G. (1953). Experimental development of Guillamoń, J. M., et al. (2002). Analysis of yeast pop- compressed yeast fermentation starters. Wines & Vines, ulations during alcoholic fermentation: Six year follow- 34,8e27. up study. Syst. Appl. Microbiol., 25, 287e293. Ciani, M., Mannazzu, I., Marinangeli, P., Clementi, F., & Bidenne, C., Blondin, B., Dequin, S., & Vezinhet, F. (1992). Martini, A. (2004). Contribution of winery-resident Analysis of the chromosomal DNA polymorphism of Saccharomyces cerevisiae strains to spontaneous grape wine strains of Saccharomyces cerevisiae. Curr. Genet., 22, must fermentation. Anton. Leeuw., 85, 159e164. 1e7. Cocolin, L., Bisson, L. F., & Mills, D. A. (2000). Direct Blondin, B., & Vezinhet, F. (1988). Identification de souches profiling of the yeast dynamics in wine fermentations. de levures oenologiques par leurs caryotypes obtenus en FEMS Microbiol. Lett., 189,81e87. e´lectrophore`se en champ pulse´. Rev. Fr. Oenol., 115,7e11. Cocolin, L., Rantsiou, K., Iacumin, L., Zironi, R., & Comi, G. Boekhout, T., Theelen, B., Diaz, M., Fell, J. W., Hop, W. C., (2004). Molecular detection and identification of Bretta- Abeln, E. C., et al. (2001). Hybrid genotypes in the nomyces/Dekkera bruxellensis and Brettanomyces/Dekkera pathogenic yeast Cryptococcus neoformans. Microbiology, anomalus in spoiled wines. Appl. Environ. Microbiol., 70, 147, 891e907. 1347e1355. Borst, A., Theelen, B., Reinders, E., Boekhout, T., Fluit, A. C., Comi, G., Maifreni, M., Manzano, M., Lagazio, C., & & Savelkoul, P. H. (2003). Use of amplified fragment Cocolin, L. (2000). Mitochondrial DNA restriction length polymorphism analysis to identify medically enzyme analysis and evaluation of the enological char- important Candida spp., including C. dubliniensis. acteristics of Saccharomyces cerevisiae strains isolated J. Clin. Microbiol., 41, 1357e1362. from grapes of the wine-producing area of Collio Italy. Braus, G., Furter, R., Prantl, F., Niederberger, P., & Hu¨ tter, R. Int. J. Food Microbiol., 58,117e121. (1985). Arrangement of genes TRP1 and TRP3 of Constantı´, M., Reguant, C., Poblet, M., Zamora, F., Mas, A., & Saccharomyces cerevisiae strains. Arch. Microbiol., 142, Guillamoń, J. M. (1998). Molecular analysis of yeast 383e388. population dynamics: Effect of sulphur dioxide and Briones, A. I., Ubeda, J., & Grando, M. S. (1996). Differen- inoculum on must fermentation. Int. J. Food Microbiol., 41, tiation of Saccharomyces cerevisiae strains isolated from 169e175. REFERENCES 135

Cottrell, M., Kovk, J. K. F., Lategam, P. M., & Britz, T. Z. Echeverrigaray, S., Paese-Toresan, S., & Carrau, J. L. (1986). Long chain fatty acid composition as an aid in the (2000). RAPD marker polymorphism among comercial classification of the genus. Saccharomyces. Syst. Appl. winery yeast strains. World J. Microbiol. Biotechnol., 16, Microbiol., 8, 166e168. 143e146. Curtin, C. D., Bellon, J. R., Henschke, P. A., Godden, P. W., & Egli, C. M., Edinger, W. D., Mitrakul, C. M., & Henick- de Barros Lopes, M. A. (2007). Genetic diversity of Kling, T. (1998). Dynamics of indigenous and inoculated Dekkera bruxellensis yeasts isolated from Australian yeast populations and their effect on the sensory char- wineries. FEMS Yeast Res., 7, 471e481. acter of Riesling and Chardonnay wines. J. Appl. Micro- Daniel, H. M., & Meyer, W. (2003). Evaluation of ribosomal biol., 85, 779e789. RNA and actin gene sequences for the identification of Eigel, A., & Feldmann, H. (1982). Ty1 and delta elements ascomycetous yeasts. Int. J. Food Microbiol., 86,61e78. occur adjacent to several tRNA genes in yeast. EMBO Dassanayake, R. S., & Samaranayake, L. P. (2003). Amplifi- J, 1, 1245e1250. cation-based nucleic acid scanning techniques to assess Esteve-Zarzoso, B., Belloch, C., Uruburu, F., & Querol, A. genetic polymorphism in Candida. Crit Rev Microbiol., 29, (1999). Identification of yeast by RFLP analysis of the 1e24. 5.8S rRNA gene and the two ribosomal internal tran- de Barros Lopes, M., Bellon, J. R., Shirley, N. J., & scribed spacers. Int. J. Syst. Bacteriol., 49, 329e337. Gaute, P. R. (2002). Evidence for multiple interspecific Esteve-Zarzoso, B., Gostıńcar, A., Bobet, R., Uruburu, F., & hybridization in Saccharomyces sensu stricto species. Querol, A. (2000). Selection and molecular character- FEMS Yeast Res.,1, 323e331. isation of wine yeasts isolated from the “El Penede´s” de Barros Lopes, M., Rainieri, S., Henschkle, P. A., & area (Spain). Food Microbiol., 17, 553e562. Longridge, P. (1999). AFLP fingerprinting for analysis of Esteve-Zarzoso, B., Peris-Toran, M. J., Garcıá-Maiquez, E., yeast genetic variation. Int. J. Syst. Bacteriol., 49, 915e924. Uruburu, F., & Querol, A. (2001). Yeast population de Barros Lopes, M., Soden, A., Henschke, P. A., & dynamics during the fermentation and biological aging Langridge, P. (1996). PCR differentiation of commercial of sherry wines. Appl. Environ. Microbiol., 67, 2056e2061. strains using intron splice site primers. Appl. Environ. Fernańdez-Espinar, M. T., Barrio, E., & Querol, A. (2003). Microbiol., 62, 4514e4520. Genetic variability among species of the Saccharomyces Degre´, R., Thomas, D. Y., Ash, J., Mailhiot, K., Morin, A., & sensu stricto. Yeast, 20, 1213e1226. Dubord, C. (1989). Wine yeasts strain identification. Am. Fernańdez-Gonza´lez, M., Espinosa, J. C., Ubeda, J. F., & J. Enol. Vitic., 40, 309e315. Briones, A. I. (2001). Yeast present during wine Delaherche, A., Claisse, O., & Lonvaud-Funel, A. (2004). fermentation: Comparative analysis of conventional Detection and quantification of Brettanomyces bruxellensis plating and PCR-TTGE. Syst. Appl. Microbiol., 24, and ropy Pediococcus damnosus strains in wine by real- 634e638. time polymerase chain reaction. J. Appl. Microbiol., 97, Fernańdez-Espinar, M. T., Esteve-Zarzoso, B., Querol, A., & 910e915. Barrio, E. (2000). RFLP analysis of the ribosomal internal de Llanos, R., Fernańdez-Espinar, M. T., & Querol, A. (2004). transcribed spacers and the 5.8S rRNA gene region of Identification of species of the genus Candida by RFLP the genus Saccharomyces: A fast method for species analysis of the 5.8S rRNA gene and the two ribosomal identification and the differentiation of flor yeasts. internal transcribed spacers. Anton. Leeuw., 85, 175e185. Anton. Leeuw., 78,87e97. Dias, L., Dias, S., Sancho, T., Stender, T., Querol, A., Mal- Fernańdez-Espinar, M. T., Lo´pez, V., Ramoń, D., Bartra, E., & feito-Ferreiro, M., et al. (2003). Identification of yeasts Querol, A. (2001). Study of the authenticity of commercial isolated from wine-related environments and capable of wine yeast strains by molecular techniques. Int. J. Food producing 4-ethylphenol. Food Microbiol., 20, 567e574. Microbiol., 70,1e10. di Maro, E., Ercolini, D., & Coppola, S. (2007). Yeast Fleet, G., & Heard, G. (1993). Yeast growth during fermen- dynamics spontaneous wine fermentation of the Cata- tation. In G. Fleet (Ed.), Wine microbiology and biotech- lanesca grape. Int. J. Food Microbiol., 117, 201e210. nology (pp. 27e57). Chur, Switzerland: Harwood Dlauchy, D., Tornai-Lehoczki, J., & Ga´bor, P. (1999). Academic Publishers. Restriction enzyme analysis of PCR amplified rDNA as Flores Berrios, E. P., Alba Gonza´lez, J. F., Arrizon a taxonomic tool in yeast identification. System. Appl. Gavinõ, J. P., Romano, P., Capece, A., & Gschaedler Microbiol., 22, 445e453. Mathis, A. (2005). The uses of AFLP for detecting DNA Duarte, F. L., Pais, C., Spencer-Martins, I., & Leao, C. (1999). polymorphism, genotype identification and genetic Distinctive electrophoretic isoenzyme profiles in diversity between yeasts isolated from Mexican agave- Saccharomyces sensu stricto. Int. J. Syst. Bacteriol, 49, distilled beverages and from grape musts. Lett. Appl. 1907e1913. Microbiol., 41, 147e152. 136 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

Frezier, V., & Dubourdieu, D. (1992). Ecology of yeast Gray, M. W. (1989). Origin and evolution of mitochondrial Saccharomyces cerevisiae during spontaneous fermenta- DNA. Annu. Rev. Cell. Biol., 5,25e50. tion in a Bordeaux winery. Am. J. Enol. Vitic., 43, Groth, C., Hansen, J., & Piskur, J. (1999). A natural chimeric 375e380. yeast containing genetic material from three species. Int. Gallego, F. J., Pe´rez, M. A., Martıńez, I., & Hidalgo, P. (1998). J. Syst. Bacteriol., 49, 1933e1938. Microsatellites obtained from database sequences are Guillamoń, J. M., Barrio, E., Huerta, T., & Querol, A. (1994). useful to characterize Saccharomyces cerevisiae. Am. Rapid characterization of four species of the Saccharo- J. Enol. Vitic., 49, 350e351. myces sensu stricto complex according to mitochondrial Ganga, M. A., & Martıńez, C. (2004). Effect of wine yeast DNA patterns. Int. J. Bacteriol., 44, 708e714. monoculture practice on the biodiversity of non- Guillamoń, J. M., Barrio, E., & Querol, A. (1996). Charac- Saccharomyces yeasts. J. Appl. Microbiol., 96,76e83. terization of wine yeast strains of the Saccharomyces Gargouri, A. (1989). A rapid and simple method for the genus on the basis of molecular markers: Relationships extraction of yeast mitochondrial DNA. Curr. Genet., 15, between genetic distance and geographic or ecological 235e237. origin. System. Appl. Microbiol., 19, 122e132. Giudici, P., & Pulvirenti, A. (2002). Molecular methods for Guillamoń, J. M., Sabate, J., Barrio, E., Cano, J., & identification of wine yeasts. In M. Ciani (Ed.), Biodi- Querol, A. (1998). Rapid identification of wine yeast versity and biotechnology of wine yeasts (pp. 35e52). Kerala, species based on RFLP analysis of the ribosomal India: Research Signpost. internal transcribed spacer ITS region. Arch. Microbiol., Giudici, P., Zambonelli, C., Passarelli, P., & Castellari, L. 169, 387e392. (1995). Improvement of wine composition with cry- Gutie´rrez, A. R., Lo´pez, R., Santamarıá, M. P., & Sevilla, M. J. otolerant Saccharomyces strains. Am. J. Enol. Vitic., 46, (1997). Ecology of inoculated and fermentations in Rioja 143e147. Spain musts, examined by mitochondrial DNA restric- Golden, D. A., Beuchat, L. R., & Hitchcock, H. L. (1994). tion analysis. Int. J. Food Microbiol., 20, 241e245. Changes in fatty acid composition of Zygosacchar- Hansen, J., & Kielland-Brandt, M. C. (1994). Saccharomyces omyces rouxii as influenced by solutes, potassium carlsbergensis contains two functional MET2 alleles sorbate and incubation temperature. Int. J. Food Micro- similar to homologues from S. cerevisiae and S. mon- biol., 21, 293e303. acensis. Gene, 140,33e40. Gonza´lez, S. S., Barrio, E., Gafner, J., & Querol, A. (2006). Hayashi, N., Arai, R., Tada, S., Taguchi, H., & Ogawa, Y. Natural hybrids from Saccharomyces cerevisiae, Saccharo- (2007). Detection and identification of Brettanomyces/ myces bayanus and Saccharomyces kudriavzevii in wine Dekkera sp. yeasts with a loop-mediated isothermal fermentations. FEMS Yeast Res, 6, 1221e1234. amplification method. Food Microbiol., 24, 778e785. Gonza´lez, S. S., Barrio, E., & Querol, A. (2007). Molecular Hennequin, C., Thierry, A., Richard, G. F., Lecointre, G., identification and characterization of wine yeasts iso- Nguyen, H. V., Gaillardin, C., et al. (2001). Microsatellite lated from Tenerife (Canary Island, Spain). J. Appl. typing as a new tool for identification of Saccharomyces Microbiol., 102, 1018e1025. cerevisiae strains. J. Clin. Microbiol., 39, 551e559. Gonza´lez Techera, A., Jubany, S., Carrau, F. M., & Hierro, N., Esteve-Zarzoso, B., Gonza´lez, A., Mas, A., & Gaggero, C. (2001). Differentiation of industrial wine Guillamoń, J. M. (2006a). Real-time quantitative PCR yeast strains using microsatellite markers. Lett. Appl. (QPCR) and reverse transcription-QPCR for detection Microbiol., 33,71e75. and enumeration of total yeasts in wine. Appl. Environ. Granchi, L., Bosco, M., Messini, A., & Vincenzini, M. (1999). Microbiol., 72, 7148e7155. Rapid detection and quantification of yeast species Hierro, N., Esteve-Zarzoso, B., Mas, A., & Guillamoń, J. M. during spontaneous wine fermentation by PCR-RFLP (2007). Monitoring of Saccharomyces and Hanseniaspora analysis of the rDNA ITS region. J. Appl. Microbiol., 87, populations during alcoholic fermentation by real-time 949e956. quantitative PCR. FEMS Yeast Res.,7, 1340e1349. Granchi, L., Ganucci, D., Viti, C., Giovannetti, L., & Hierro, N., Gonza´lez, A., Mas, A., & Guillamoń, J. M. Vincenzini, M. (2003). Saccharomyces cerevisiae biodiver- (2006b). Diversity and evolution of non-Saccharomyces sity in spontaneous commercial fermentations of grape yeast populations during wine fermentation: Effect of musts with “inadequate” assimilable-nitrogen content. grape ripeness and cold maceration. FEMS Yeast Res, 6, Lett. Appl. Microbiol., 36,54e58. 102e111. Grando, M. S., Ubeda, J., & Briones, A. I. (1994). RAPD Howell, K. S., Bartowsky, E. J., Fleet, G. H., & analysis of wine Saccharomyces strains differentiated by Henschke, P. A. (2004). Microsatellite PCR profiling of pulsed field gel electrophoresis. Biotechnol. Tech., 8, Saccharomyces cerevisiae strains during wine fermenta- 557e560. tion. Lett. Appl. Microbiol., 38, 315e320. REFERENCES 137

Ibeas, J. L., & Jimeńez, J. (1996). Genomic complexity and Lopandic, K., Gangl, H., Wallner, E., Tscheik, G., Leitner, G., chromosomal rearrangements in wine-laboratory yeast Querol, A., et al. (2007). Genetically different wine yeasts hybrids. Curr. Genet., 30, 410e416. isolated from Austrian vine-growing regions influence Ibeas, J. I., Lozano, I., Perdigones, F., & Jimeńez, J. (1996). wine aroma differently and contain putative hybrids Detection of Dekkera-Brettanomyces strains in sherry by between Saccharomyces cerevisiae and Saccharomyces a nested PCR method. Appl. Environ. Microbiol., 62, kudriavzevii. FEMS Yeast Res.,7, 953e965. 998e1003. Lopandic, K., Tiefenbrunner, W., Gangl, H., Mandl, K., James, S. A., Cai, J., Roberts, I. N., & Collins, M. D. (1997). Berger, S., Leitner, G., et al. (2008). Molecular profiling of Phylogenetic analysis of the genus Saccharomyces based yeasts isolated during spontaneous fermentations of on 18S rRNA gene sequences: Description of Saccharo- Austrian wines. FEMS Yeast Res.,8, 1063e1075. myces kunashirensis sp. nov., and Saccharomyces martiniae Lo´pes, C. A., Lavalle, T. L., Querol, A., & Caballero, A. C. sp. nov. Int. J. Syst. Bacteriol., 47, 453e460. (2006). Combined use of killer biotype and mtDNA- James, S. A., Collins, M. D., & Roberts, I. N. (1996). Use of an RFLP patterns in a Patagonian wine Saccharomyces cer- rRNA internal transcribed spacer region to distinguish evisiae diversity study. Anton. Leeuw., 89, 147e156. phylogenetically closely related species of the genera Lo´pes, C. A., Rodrı´guez, M. E., Sangorrıń, M., Querol, A., & Zygosaccharomyces and Torulaspora. Int. J. Syst. Bacteriol., Caballero, A. C. (2007). Patagonia wines: Implantation of 46, 189e194. an indigenous strain of Saccharomyces cerevisiae in Keogh, R. S., Seoighe, C., & Wolfe, K. H. (1998). Evolution fermentations conducted in traditional and modern of gene order and chromosomal number in Saccharo- cellars. J. Ind. Microbiol. Biotechnol., 34, 139e149. myces, Kluyveromyces and related fungi. Yeast, 14, Lo´pes, C. A., van Broock, M., Querol, A., & Caballero, A. C. 443e457. (2002). Saccharomyces cerevisiae wine yeast populations in Kreger-van Rij, N. J. W. (1984). The yeast, a taxonomic study. a cold region in Argentinean Patagonia. A study at Amsterdam, the Netherlands: Elsevier Science B.V. different fermentation scales. J. Appl. Microbiol., 93, Kurtzman, C. P., & Phaff, H. J. (1987). In A. H. Rose, & 608e615. J. S. Harrison (Eds.), The yeast. Molecular taxonomy, Vol. 1 Lo´pez, V., Fernańdez-Espinar, M. T., Barrio, E., Ramoń, D., & (pp. 63e94). London, UK: Academic Press. Querol, A. (2002). A new PCR-based method for moni- Kurtzman, C. P., & Robnett, C. J. (1998). Identification and toring inoculated wine fermentations. Int. J. Food Micro- phylogeny of ascomycetous yeasts from analysis of biol., 81,63e71. nuclear large subunit 26S ribosomal DNA partial Lo´pez, V., Querol, A., Ramoń, D., & Fernańdez- sequences. Anton. Leeuw., 73, 331e371. Espinar, M. T. (2001). A simplified procedure to analyse Kurtzman, C. P., & Robnett, C. J. (2003). Phylogenetic rela- mtDNA from industrial yeasts. Int. J. Food Microbiol., 68, tionships among yeasts of the ‘Saccharomyces complex’ 75e81. determined from multigene sequence analyses. FEMS Loureiro, V., & Malfeito-Ferreira, M. (2003). Spoilage yeasts Yeast Res.,3(4), 417e432. in the wine industry. Int. J. Food Microbiol., 86,23e50. Lavalleé, F., Salvas, Y., Lamy, S., Thomas, D. Y., Degre´, R., & Manzano, M., Medrala, D., Giusto, C., Bartolomeoli, I., Dulau, L. (1994). PCR and DNA fingerprinting used as Urso, R., & Comi, G. (2006). Classical and molecular quality control in the production of wine yeast strains. analysis to characterize commercial dry yeasts used in Am. J. Enol. Vitic., 45,86e91. wine fermentations. J. Appl. Microbiol., 100, 599e607. Legras, J.-L., & Karst, F. (2003). Optimisation of interdelta Marinangeli, P., Angelozzi, D., Ciani, M., Clementi, F., & for Saccharomyces cerevisiae strain characterization. FEMS Mannazzu, I. (2004). Minisatellites in Saccharomyces cer- Microbiol. Lett., 221, 249e255. evisiae genes encoding cell wall proteins: A new way Le Jeune, C., Erny, C., Demuyter, C., & Lollier, M. (2006). towards wine strain characterisation. FEMS Yeast Res, 4, Evolution of the population of Saccharomyces cerevisiae 427e435. from grape to wine in a spontaneous fermentation. Food Martıńez, C., Gac, S., Lavin, A., & Ganga, M. (2004). Microbiol., 23, 709e716. Genomic characterization of Saccharomyces cerevisiae Li, W. H. (1997). Molecular evolution. Sunderland, MA: strains isolated from wine-producing areas of South Sinauer Associates. America. J. Appl. Microbiol., 96, 1161e1168. Lieckfieldt, E., Meyer, W., & Bo¨rn, T. (1993). Rapid identi- Martıńez, J., Millań, C., & Ortega, J. M. (1989). Growth of fication and differentiation of yeasts by DNA and PCR natural flora during the fermentation of inoculated fingerprinting. J. Basic Microbiol., 33, 413e426. musts from “Pedro Ximenez” grapes. S. Afr. J. Enol. Longo, E., Cansado, J., Agrelo, D., & Villa, T. (1991). Effect of Vitic., 10,31e35. climatic conditions on yeast diversity in grape musts Martıńez, P., Codon, A. C., Pe´rez, L., & Benitez, T. (1995). from northwest Spain. Am. J. Enol. Vitic., 42, 141e144. Physiological and molecular characterization of flor 138 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

yeasts: Polymorphism of flor yeast populations. Yeast, 11, Nadal, D., Colomer, B., & Pinã, B. (1996). Molecular poly- 1399e1411. morphism distribution in phenotypically distinct pop- Martini, A., & Vaughan-Martini, A. (1990). Grape must ulations of wine yeast strains. Appl. Environ. Microbiol., fermentation: Past and Present. In J. F. T. Spencer, & 62, 1944e1950. D. M. Spencer (Eds.), Yeast technology (pp. 105e123). Naumov, G. I., Masneuf, I., Naumova, E. S., Aigle, M., & Berlin, Germany: Springer-Verlag. Dubourdieu, D. (2000). Association of Saccharomyces Martorell, P., Querol, A., & Fernańdez-Espinar, M. T. (2005). bayanus var. uvarum with some French wines: Genetic Rapid identification and enumeration of Saccharomyces analysis of yeast populations. Res. Microbiol., 151, cerevisiae cells in wine by real-time PCR. Appl. Environ. 683e691. Microbiol., 71, 6823e6830. Naumov, G. I., Naumova, E. S., Antunovics, Z., & Masneuf, I., Aigle, M., & Dubourdieu, D. (1996). Develop- Sipiczki, M. (2002). Saccharomyces bayanus var. uvarum in ment of a polymerase chain reaction/restriction frag- Tokaj wine-making of Slovakia and Hungary. Appl. ment length polymorphism method for Saccharomyces Microbiol. Biotechnol., 59, 727e730. cerevisiae and Saccharomyces bayanus identification in Naumov, G. I., Naumova, E. S., & Gaillardin, C. (1993). enology. FEMS Microbiol. Lett., 138, 239e244. Genetic and karyotypic identification of wines Saccha- Masneuf, I., Hansen, J., Groth, C., Piskur, J., & romyces bayanus yeast isolated in France and Italy. Syst. Dubourdieu, D. (1998). New hybrids between Saccharo- Appl. Microbiol., 16, 272e215. myces sensu stricto yeast species found among wine and Ness, F., Lavalleé, F., Dubordieu, D., Aigle, M., & Dulau, L. cider production strains. Appl. Environ. Microbiol., 64, (1993). Identification of yeast strains using the poly- 3887e3892. merase chain reaction. J. Sci. Food Agric.,62,89e94. Masneuf-Pomare`de, I., Le Jeune, C., Durrens, P., Lollier, M., Nguyen, H. V., & Gaillardin, C. A. (1997). Two subgroups Aigle, M., & Dubourdieu, D. (2007). Molecular typing of within the Saccharomyces bayanus species evidenced by wine yeast strains Saccharomyces bayanus var. uvarum using PCR amplification and restriction polymorphism of the microsatellite markers. Syst. Appl. Microbiol., 30,75e82. Non-transcribed spacer 2 in the ribosomal DNA unit. Mesa, J. J., Infante, J. J., Rebordinos, L., & Cantoral, J. M. Syst. Appl. Microbiol., 20, 286e294. (1999). Characterization of yeasts involved in the bio- Nguyen, H. V., Lepingle, A., & Gaillardin, C. A. (2000). logical ageing of Sherry wines. Lebensm.-Wiss. u.-Technol., Molecular typing demonstrates homogeneity of Saccha- 32,114e120. romyces uvarum strains and reveals the existence of Miot-Sertier, C., & Lonvaud-Funel, A. (2007). Development hybrids between S. uvarum and S. cerevisiae, including of a molecular method for the typing of Brettanomyces the S. bayanus type strain CBS 380. Syst. Appl. Microbiol., bruxellensis (Dekkera bruxellensis) at the strain level. 23,71e85. J. Appl. Microbiol., 102, 555e562. Oelofse, A., Lonvaud-Funel, A., & du Toit, M. (2009). Molnar, O., Messner, R., Prillinger, H., Stahl, U., & Molecular identification of Brettanomyces bruxellensis Slavikova, E. (1995). Genotypic identification of Saccha- strains isolated from red wines and volatile phenol romyces species using random amplified polymorphic production. Food Microbiol., 26, 377e385. DNA analysis. System. Appl. Microbiol., 18, 136e145. Paffetti, C., Barberio´, C., Casalone, E., Cavalieri, D., Fani, R., Moreira da Silva, M., Malfeito-Ferreira, M., Aubyn, S., & Fia, G., et al. (1995). DNA fingerprinting by random Loureiro, V. (1994). Long chain fatty acid composition as amplified polymorphic DNA and restriction fragment criterion for yeast distinction in the brewing industry. length polymorphism is useful for yeast typing. Res. J. Inst. Brew., 100,17e22. Microbiol., 146, 587e594. Muyzer, G., de Waal, E. C., & Uitterlinden, A. G. (1993). Pedersen, M. B. (1983). DNA sequence polymorphisms in Profiling of complex microbial populations by dena- the genus Saccharomyces. I. Comparison of the HIS4 and turing gradient gel electrophoresis analysis of poly- ribosomal RNA genes in larger strains, ALE strain and merase chain reaction-amplified genes encoding for 16S various species. Carlsberg Res. Commun.,48, 485e503. rRNA. Appl. Environ. Microbiol., 59, 695e700. Pedersen, M. B. (1985). DNA sequence polymorphisms in Muyzer, G., & Smalla, K. (1998). Application of denaturing the genus Saccharomyces. II. Analysis of the genes RDN1, gradient gel electrophoresis (DGGE) and temperature HIS4, LEU2 and Ty transposable elements in Calrsberg, gradient gel electrophoresis (TGGE) in microbial Tubog and 22 Babarian brewing strains. Carlsberg Res. ecology. Anton. Leeuw., 73, 127e114. Commun., 50, 263e272. Nadal, D., Carro, D., Fernandez-Larrea, J., & Pinã, B. (1999). Pedersen, M. B. (1986a). DNA sequence polymorphisms in Analysis and dynamics of the chromosomal comple- the genus Saccharomyces. III. Restriction endonuclease ments of wild sparkling-wine yeast strains. Appl. fragment patterns of chromosomal regions in brewing Environ. Microbiol., 65, 1944e1950. strains. Carlsberg Res. Commun.,51, 163e183. REFERENCES 139

Pedersen, M. B. (1986b). DNA sequence polymorphisms in Redzepovic, S., Orlic, S., Sikora, S., Majdak, A., & the genus Saccharomyces. IV. Homologous chromosomes Pretorius, I. S. (2002). Identification and characterization III of Saccharomyces bayanus, S. carlsbergensis, and S. uva- of Saccharomyces cerevisiae and Saccharomyces paradoxus rum. Carlsberg Res. Commun.,51, 185e202. strains isolated from Croatian vineyards. Lett. Appl. Pe´rez, M. A., Gallego, F. J., & Hidalgo, P. (2001a). Evaluation Microbiol., 35, 305e310. of molecular techniques for the genetic characterization Renouf, V., Claisse, O., & Lonvaud-Funel, A. (2007). of Saccharomyces cerevisiae strains. FEMS Microbiol. Lett., Inventory and monitoring of wine microbial consortia. 205, 375e378. Appl. Microbiol. Biotechnol., 75, 149e164. Pe´rez, M. A., Gallego, F. J., Martıńez, I., & Hidalgo, P. Richards, K. D., Goddard, M. R., & Gardner, R. C. (2009). A (2001b). Detection, distribution and selection of micro- database profile genotypes for Saccharomyces cerevisiae. satellites SSRs in the gemome of the yeast Saccharomyces Anton. Leeuw., 96, 355e359. cerevisiae as molecular markers. Lett. Appl. Microbiol., 33, Ro¨der, C., Ko¨nig, H., & Fro¨hlich, J. (2007). Species-specific 461e466. identification of Dekkera/Brettanomyces yeasts by fluo- Phister, T. G., & Mills, D. A. (2003). Real-time PCR assay for rescently labeled DNA probes targeting the 26S rRNA. detection and enumeration of Dekkera bruxellensis in FEMS Yeast Res.,7, 1013e1026. wine. Appl. Environ. Microbiol., 690, 7430e7434. Rodrı´guez, M. E., Lopes, C. A., van Broock, M., Valles, S., Prakitchaiwattana, C. J., Fleet, G. H., & Heard, G. M. (2004). Ramon, D., & Caballero, A. C. (2004). Screening and Application and evaluation of denaturing gradient gel typing of Patagonia wine yeasts for glycosidase activies. electrophoresis to analyse the yeast ecology of wine J. Appl. Microbiol., 96,84e95. grapes. FEMS Yeast Res.,4, 865e877. Romancino, D. P., di Maio, S., Muriella, R., & Oliva, D. Pramateftaki, P. V., Lanaridis, P., & Typas, M. A. (2000). (2008). Analysis of non- Saccharomyces yeast populations Molecular identification of wine yeasts at species or isolated from grape musts from Sicily (Italy). J. Appl. strains level: A case study with strains from two vine- Microbiol., 105, 2248e2254. growing areas of Greece. J. Appl. Microbiol., 89, 236e248. Rosini, G., Federichi, F., & Martini, A. (1982). Yeast flora of Price, C. W., Fuson, G. B., & Phaff, H. J. (1978). Genome grape berries during ripening. Microbiol. Ecol., 8,83e89. comparison in yeast systematic: Delimitation of species Sabate´, J., Cano, J., Esteve-Zarzoso, B., & Guillamoń, J. M. within the genera Schwanniomyces, Saccharomyces, (2002). Isolation and identification of yeasts associated Debaryomyces and Pichia. Microbiol. Rev.,42, 161e193. with vineyard and winery by RFLP analysis of ribo- Pulvirenti, A., Nguyen, H., Caggia, C., Giudici, P., somal genes and mitochondrial DNA. Microbiol. Res., Rainieri, S., & Zambonelli, C. (2000). Saccharomyces 157, 267e274. uvarum, a proper species within Saccharomyces sensu Sabate´, J., Cano, J., Querol, A., & Guillamoń, J. M. (1998). stricto. FEMS Microbiol. Lett., 192, 191e196. Diversity of Saccharomyces strains in wine fermentations: Querol, A., & Barrio, E. (1990). A rapid and simple method Analysis for two consecutive years. Lett. Appl. Microbiol., for the preparation of yeast mitochondrial DNA. Nucleic 26, 452e455. Acids Res.,18, 1657. Saiki, R. K., Gelfand, D. H., Stoffel, S., Schaf, S. J., Querol, A., Barrio, E., Huerta, T., & Ramoń, D. (1992b). Higuchi, R., Hoprn, G. T., et al. (1988). Primer-directed Molecular monitoring of wine fermentations conducted enzymatic amplification of DNA with a thermostable by active dry yeast strains. Appl. Environ. Microbiol., 58, DNA polymerase. Science, 239, 487e491. 2948e2953. Saiki, R., Sharf, S., Faloona, F., Mullis, K. B., Horn, G. T., & Querol, A., Barrio, E., & Ramoń, D. (1992a). A comparative Erlich, H. A. (1985). Enzymatic amplification of beta study of different methods of yeast strain characteriza- globin genomic sequences and restriction gene analysis tion. Syst. Appl. Microbiol., 15, 439e446. for diagnostic of sickle-cell anaemia. Science, 230, Querol, A., Barrio, E., & Ramoń, D. (1994). Population 1350e1354. dynamics of natural Saccharomyces strains during wine Salinas, F., Garrido, D., Ganga, A., Veliz, G., & Martıńez, C. fermentation. Int. J. Food Microbiol., 21, 315e323. (2009). Taqman real-time PCR for the detection and Querol, A., Jimeńez, M., & Huerta, T. (1990). Microbiological enumeration of Saccharomyces cerevisiae in wine. Food and enological parameters during fermentation of musts Microbiol., 26, 328e332. from poor and normal grape-harvest in the region of Sambrook, J., Fritsch, E. F., & Maniatis, T. (1989). Molecular Alicante Spain. J. Food Sci., 55, 1603e1606. cloning: A laboratory manual (2nd ed.). Cold Spring Quesada, M. P., & Cenis, J. L. (1995). Use of random Harbor, NY: Cold Spring Harbor Laboratory. amplified polymorphic DNA RAPD-PCR in the Scheda, R., & Yarrow, D. (1966). The instability of physio- characterization of wine yeasts. Am. J. Enol. Vitic., 46, logical properties used as criteria in the taxonomy of 204e208. yeast. Arch. Microbiol., 55, 209e225. 140 5. MOLECULAR IDENTIFICATION AND CHARACTERIZATION OF WINE YEASTS

Scheda, R., & Yarrow, D. (1968). Variations in the fermen- Torriani, S., Zapparoli, G., & Suzzi, G. (1999). Genetic and tative pattern of some Saccharomyces species. Arch. phenotypic diversity of Saccharomyces sensu stricto strains Microbiol., 61, 310e316. isolated from Amarone wine. Diversity of Saccharomyces Schuller, D., Valero, E., Dequin, S., & Casal, M. (2004). strains from Amarone wine. Anton. Leeuw., 75, 207e215. Survey of molecular methods for the typing of wine Tredoux, H. G., Kock, J. L. F., Lategan, P. M., & Muller, H. B. yeast strains. FEMS Microbiol. Lett., 231,19e26. (1987). A rapid identification technique to differentiate Schu¨ tz, M., & Gafner, J. (1993). Analysis of yeast diversity between S. cerevisiae strains and other yeast species in during spontaneous and induced alcoholic fermenta- the winery industry. Am. J. Enol. Vitic., 38, 161e164. tions. J. Appl. Bacteriol., 75, 551e558. Trilles, L., Lazera, M., Wanke, B., Theelen, B., & Boekhout, T. Schu¨ tz, M., & Gafner, J. (1994). Dynamics of the yeast strain (2003). Genetic characterization of environmental population during spontaneous alcoholic fermentation isolates of the Cryptococcus neoformans species complex determined by CHEF gel electrophoresis. Lett. Appl. from Brazil. Med Mycol.,41, 383e390. Microbiol., 19, 253e257. Urso, R., Rantsiou, K., Dolci, P., Rolle, L., Comi, G., & Seehaus, T., Rodizio, R., Heinisch, J., Aguilera, H. D., & Cocollin, L. (2008). Yeast biodiversity and dynamics Zimmerman, F. K. (1985). Specific gene probes as tools in during sweet wine production as determined by yeast taxonomy. Curr. Genet., 10, 103e110. molecular methods. FESM Yeast Res.,8, 1053e1062. Smole-Mozina, S., Dlauchy, D., Deak, T., & Raspor, P. (1997). Vacanneyt, B. P., Hennebert, G., & Kersters, K. (1991). Identification of Saccharomyces sensu stricto and Tor- Differentiation of yeast species based on electrophoretic ulaspora yeasts by PCR ribotyping. Lett. Appl. Microbiol., whole-cell protein patterns. Syst. Appl. Microbiol., 14, 24,311e315. 23e32. Stender, H., Kurtzman, C., Hyldig-Nielsen, J. J., Sørensen, D., Vaduano, E., & Garcıá-Moruno, E. (2008). Discrimination of Broomer, A., Oliveira, K., et al. (2001). Identification of Saccharomyces cerevisiae wine strains using microsatellite Dekkera bruxellensis (Brettanomyces) from wine by fluo- multiplex PCR and band pattern analysis. Food Micro- rescente in situ hybridization using peptide nucleic acid biol., 25,56e64. probes. Appl. Environ. Microbiol., 67, 938e941. van der Westhuizen, T. J., & Pretorius, I. S. (1992). The value Stringini, M., Comitini, F., Taccari, M., & Ciani, M. (2009). of electrophoretic fingerprinting and karyotyping wine Yeast diversity during tapping and fermentation of palm yeast breeding programmes. Anton. Leeuw., 61, 249e257. wine from Cameroon. Food Microbiol., 26, 415e420. van Keulen, H., Lindmark, D. G., Zeman, K. E., & Theelen, B., Silvestri, M., Gueho, E., van Belkum, A., & Gerlosky, W. (2003). Yeasts present during spontaneous Boekhout, T. (2001). Identification and typing of Malas- fermentation of Lake Erie Chardonnay, Pinot Gris and sezia yeasts using amplified fragment length poly- Riesling. Anton. Leeuw., 83, 149e154. morphism (AFLP), random amplified polymorphic van Vuuren, H. J., & van der Meer, L. (1987). Fingerprinting DNA (RAPD) and denaturing gradient gel electropho- of yeast by protein electrophoresis. Am. J. Enol. Vitic., 38, resis (DGGE). FEMS Yeast Res.,1,79e86. 49e53. Tofalo, R., Chaves-Lo´pez, C., di Fabio, F., Schirone, M., Versavaud, A., Courcoux, P., Roulland, C., Dulau, L., & Felis, G. E., Torriani, S., et al. (2009). Molecular identifi- Hallet, J. N. (1995). Genetic diversity and geographical cation and osmotolerant profile of wine yeasts that distribution of wild Saccharomyces cerevisiae strains from ferment a high sugar grape must. Int. J. Food Microbiol., the wine-producing area of Charentes, France. Appl. 130, 179e187. Environ. Microbiol., 61, 3521e3529. Torija, M. J., Rozes, N., Poblet, M., Guillamoń, J. M., & Vezinhet, F., Blondin, B., & Hallet, J.-N. (1990). Chromo- Mas, A. (2001). Yeast population dynamics in sponta- somal DNA patterns and mitochondrial DNA poly- neous fermentations: Comparison between two different morphism as tools for identification of enological strains wine-producing areas over a period of three years. of Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol., Anton. Leeuw., 79, 345e352. 32, 568e571. Torija, M. J., Rozes, N., Poblet, M., Guillamoń, J. M., & Vos, P., Hogers, R., & Bleeker, M. (1995). AFLP: A new Mas, A. (2003). Effects of fermentation temperature on technique for DNA fingerprinting. Nucleic Acids Res, 23, the strain population Saccharomyces cerevisiae. Int. 4407e4414. J. Food Microbiol., 80,47e53. Walmsley, R. M., Wikinson, B. M., & Kong, T. H. (1989). Torriani, S., Zapparoli, G., Malocrino`, P., Suzzi, G., & Genetic fingerprinting for yeast. Biotechnol., 7, Dellaglio, F. (2004). Rapid identification and differenti- 1168e1170. ation of Saccharomyces cerevisiae, Saccharomyces bayanus White, T. J., Bruns, T., Lee, E., & Taylor, J. (1990). Amplifi- and their hybrids by multiplex PCR. Lett. Appl. Micro- cation and direct sequencing of fungal ribosomal RNA biol., 38, 239e244. genes for phylogenetics. In M. A. Innis, D. H. Gelfand, REFERENCES 141

J. J. Sninsky, & T. J. White (Eds.), PCR protocols: A guide to hybridisation (FISH) to the analysis of yeast populations methods and applications (pp. 315e322). San Diego, CA: dynamics in winery and laboratory grape must Academic Press. fermentations. Int. J. Food Microbiol., 108, 376e384. Wilhelm, J., & Pingoud, A. (2003). Real-time polymerase Yamagishi, H., & Ogata, T. (1999). Chromosomal structures chain reaction. Review. Chem. Bio. Chem., 4, 1120e1128. of bottom fermenting yeasts. Syst. Appl. Microbiol., 22, Williams, J. G. K., Kubelik, A. R., Livak, K. J., Rafalski, J. A., 341e353. & Tingey, S. V. (1990). DNA-polymorphism amplified by Yamamoto, N., Yamamoto, N., Amemiya, H., Yokomori, Y., arbitrary primers are useful as genetic markers. Nucleic Shimizu, K., & Totsuka, A. (1991). Electrophoretic Acids Res.,18, 6531e6535. karyotypes of wine yeasts. Am. J. Enol. Vitic., 42, Wolfe, K. H., & Shields, D. C. (1997). Molecular evidence for 358e363. an ancient duplication of the entire yeast genome. Zott, K., Miot-Sertier, C., Claisse, O., Lonvaud-Funel, A., & Nature, 387, 708e713. Masneuf-Pomarede, I. (2008). Dynamics and diversity of Xufre, A., Albergaria, H., Inaćio, J., Spencer-Martins, I., & non-Saccharomyces yeasts during the early stages in Gı´rio, F. (2006). Application of fluorescence in situ winemaking. Int. J. Food Microbiol., 125, 197e203. CHAPTER 6

Genomic and Proteomic Analysis of Wine Yeasts Jose´ E. Pe´rez-Ortıń, Jose´ Garcıá-Martıńez Departamento de Bioquı´mica & Biologıá Molecular & Laboratorio de Chips de DNA, SCSIE, Universitat de Vale`ncia, Spain

OUTLINE

1. Introduction 143 4.2. Effects of Drugs and Other External Factors 152 2. Genomic Characteristics of Wine Yeasts 144 4.3. Use of DNA Microarrays in the 3. Comparative Genomics and the Origin Analysis of Wine Yeasts 153 of the S. Cerevisiae Genome 147 4.4. Genomic Studies 157 4. The Use of S. Cerevisiae as a Model 5. Proteomic Analysis of Wine Strains 158 Organism for the Development of DNA 6. Other Global Studies 161 Microarray Technology 150 4.1. Metabolic Studies 150 7. Future Directions 163

1. INTRODUCTION microorganisms, however, are restrained by a lack of sufﬁcient knowledge regarding the The selection of suitable microorganisms for metabolic and regulatory processes occurring use in industrial processes is a key issue in within the cells. These shortcomings may, food biotechnology. One of the key challenges however, be short-lived, considering the contin- in this area is to improve the properties of starter uous advances being made in functional geno- cultures, such as the ability to establish repro- mics and proteomics. Studies in these areas ducible growth. Many of the programs aimed will help, for example, to identify the effects of at enhancing the properties of industrial genetic alterations on ﬁnal products, generate

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10006-2 143 Copyright Ó 2011 Elsevier Inc. All rights reserved. 144 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS desirable pleiotropic effects through mutations et al., 2003; Liti et al., 2009; Souciet et al., 2000). in regulatory genes, predict stress responses in DNA microarray analysis is a very useful tool the different environments to which microor- for comparing genomes from different strains ganisms are exposed, and identify genomic of S. cerevisiae, including wine strains (Carro variations associated with adaptation to the et al., 2003; Hauser et al., 2001; Schacheter particular conditions of winemaking. et al., 2009) and similar species. This chapter focuses exclusively on the yeast S. cerevisiae has also been used in the develop- species Saccharomyces cerevisiae. In addition to ment of the more recent field of proteomics. being the main microorganism involved in Proteomic studies have generated vast amounts wine fermentation, it has been used as a model of data on protein expression profiles and organism in molecular biology for many years variability in laboratory strains of S. cerevisiae (Miklos & Rubin, 1996) and is the only wine (Washburn et al., 2001), and these have yeast species for which abundant genomic and recently been extended to include wine strains proteomic information is available. It was the (Rossignol et al., 2009; Trabalzini et al., 2003; first eukaryote to have its complete genome Zuzua´rregui et al., 2006). Important advances sequenced (Goffeau et al., 1997), and, since have also been made in metabolomics, a new then, numerous functional analysis projects field in which S. cerevisiae is practically the have uncovered enormous amounts of informa- only eukaryote to have been studied to date tion on the biology of this microorganism (Raamsdonk et al., 2001; Rossouw et al., 2008). (Dujon, 1998). It can safely be said that S. cerevi- The integration of different types of “omic” data siae is currently the best understood of all into predictive models has provided the basis eukaryotic organisms. Most of the techniques for new research strategies in systems biology currently used in functional genomics and pro- (Borneman et al., 2007; Pizarro et al., 2007). teomics were initially developed in this yeast. Most of the information that has been gath- DNA chip, or microarray, technology, for ered in all of the above areas is related to labora- example, was primarily developed using S. cer- tory strains of S. cerevisiae, although more recent evisiae (DeRisi et al., 1997; Schena et al., 1995; studies have been extended to other strains Wodicka et al., 1997), and all the latest advances (particularly wine strains) and industrial processes in this field have also been tested using this (Bisson et al., 2007). Knowledge generated yeast (see Section 4). Vast amounts of data from the analysis of laboratory strains may be have thus been compiled on gene expression helpful in understanding the results of studies in S. cerevisiae. Indeed, the information on S. cer- conducted with wine strains during industrial evisiae far exceeds that available for any other fermentation, and it is extremely simple to prokaryotic or eukaryotic organisms. As a result, apply techniques used with laboratory strains it has been possible to propose global models for to their industrial counterparts. This chapter genetic and metabolic regulation (Gasch et al., will therefore also look at the methods used 2000). and results obtained for non-wine strains of The fact that S. cerevisiae was the first microor- S. cerevisiae. ganism to be widely used in the development of genome technology allowed other phylogenetically related yeasts to be analyzed subsequently 2. GENOMIC CHARACTERISTICS in global sequencing projects, and the use of OF WINE YEASTS comparative genomics has since led to important conclusions regarding gene functionality The history of wine yeasts is as old as the (Butler et al., 2009; Cliften et al., 2003; Kellis earliest civilizations in the Mediterranean GENOMIC CHARACTERISTICS OF WINE YEASTS 145 region, with the first references to winemaking important point is that all of today’s laboratory dating back to 7400 years ago. Reports of wine strains are derived from natural isolates. The production were limited to this geographical best-documented case is that of the most area for many centuries, until the practice was popular yeast among molecular biologists: the spread to other parts of the world with suitable S288c strain, which was derived from a hetero- climate conditions, as Europe embarked on its thallic (ho), diploid strain isolated in a rotten conquest of other continents in the fifteenth fig in California in 1938 (Mortimer & Johnston, century (reviewed in Mortimer, 2000 and Pre- 1986). It is very likely that the strain had been torius, 2000). Must fermentation was consid- transported from a winery by insects. ered to occur spontaneously until 1863, when Most laboratory strains of S. cerevisiae are ho, Louis Pasteur discovered that yeasts were haploid or diploid, and have a set of 16 fixed- responsible for the process. Although numerous length chromosomes (see Figure 6.1). The yeasts and bacteria contribute to must fermen- majority of wine strains, in contrast, are diploid, tation (see Chapters 2e6 and 9), the principle aneuploid, or polyploid (Bakalinsky & Snow, microorganisms responsible for this biotrans- 1990; Codoń et al., 1995). They are also homo- formation belong to the genus Saccharomyces, thallic (HO), variably heterozygous (Barre principally S. cerevisiae. This is why S. cerevisiae et al., 1993; Butler et al., 2009; Carreto et al., is often referred to as the wine yeast (Pretorius, 2008; Codoń et al., 1995), and characterized by 2000). a high level of polymorphism in chromosome The origin of S. cerevisiae has been much length (Bidenne et al., 1992; Rachidi et al., debated. While some authors are of the opinion 1999). Many strains are trisomic or tetrasomic that it is naturally present in fruit (Mortimer & for certain chromosomes (Guijo et al., 1997; Polsinelli, 1999), others believe that its origin is Bakalinsky & Snow, 1990). The above character- more recent and that it is the result of hybridiza- istics have numerous practical implications, tion with other natural species and subsequent including highly variable sporulation capacity natural selection in artificial environments (0e75%) (Bakalinsky & Snow, 1990; Barre (Martini, 1993). This second hypothesis is sup- et al., 1993; Mortimer et al., 1994) and spore ported by the fact that S. cerevisiae is found viability (0e98%) (Barre et al., 1993; Codoń only in areas close to human activity. According et al., 1995; Mortimer et al., 1994). The ability to this theory, all the modern isolates of S. cerevi- of S. cerevisiae to alter its genome is enhanced siae would have been transported by insects by the existence of mitotic and meiotic cycles. from the winery back to the vineyards (Naumov, Genome ploidy and plasticity provide wine 1996). While this debate is central to determining yeasts with certain advantages that facilitate the true origin of the S. cerevisiae genome, what is their adaptation to changing external environ- known for certain is that the genomic constitu- ments and perhaps also increase the dosage of tion of this species has been molded by the genes that have an important role in fermenta- severe fermentation-related stresses to which tion (Bakalinsky & Snow, 1990; Salmon, 1997). it has been exposed throughout the centuries. This genomic plasticity, however, is not Proof of this are the genomic differences restricted to S. cerevisiae and even allows stable between primary and secondary fermentation hybridization with closely related species. wine strains and between brewing strains and Several natural strains, such as S6U and CD1, bread-making strains, whose genotypes have for example, are hybrids of S. cerevisiae and been unknowingly selected over hundreds of Saccharomyces bayanus. S6U is an allotetraploid years with the continual improvements made (Naumov et al., 2000), which probably explains to these biotechnological processes. Another its stability despite having two distinct 146 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

genomes. The same has been observed with (a) brewing strains (Kielland-Brandt et al., 1995). The formation of interspecific hybrids between members of the Saccharomyces sensu stricto group appears to be one of the adaptive mechanisms employed by industrial yeasts (Belloch VIII et al., 2009; Querol et al., 2003). This genome plasticity, which is inherent in wine strains, is not a desirable property in model organisms used in genetic studies, and laboratory strains XVI used for such purposes are selected precisely for their lack of plasticity. Laboratory strains are also capable of adapting to changing environmental conditions, normally via point muta- (b) tions (Ferea et al., 1999), although in certain circumstances large regions or entire chromo- IV somes may also be modified (Hughes et al., 2000b). Wine strains, unlike laboratory strains, are X capable of chromosomal rearrangement during mitosis (Longo & Ve´zinhet, 1993). In an experiment by Puig et al. (2000), URA3 was replaced with an exogenous marker gene, KanMX,in the natural wine strain T73 and used to monitor genetic variation in a series of consecutive must FIGURE 6.1 The genome of the reference Saccharomyces fermentations. The authors found that URA3 cerevisiae laboratory strain has 16 chromosomes whose lengths homozygotes appeared at a rate of 2 105 are shown to scale (a). The centromeres are shown as white dots. The haploid genome is shown in this figure. Diploid per generation in a process they attributed to strains have two, probably identical, copies of each chromo- mitotic recombination or gene conversion. some. Many variants of this reference genome have been Phenotypically, the Ura cells were at a selective found in wine strains. The T73 strain (b), for example, isolated disadvantage to the Uraþ cells (heterozygotes in musts from the Alicante appellation (Querol et al., 1992)has, [ / ] and homozygotes [ / at least, the following variations: (1) a reciprocal translocation URA3 ura3 URA3 between chromosomes VIII and XVI, which generates two URA3]). Chromosomal changes were also variants of each chromosome in T73 (Pe´rez-Ortıń et al., 2002a) detected in some cells. Because of their strong (the site of the translocation is shown by grey arrows) (a); (2) an tendency towards genomic changes, wine additional, presumably identical, copyof chromosomesIVand strains do not display the same genetic unifor- X(Pe´rez-Ortıń, unpublished results); (3) many variations in mity as that used to define laboratory strains the copy number of genes from subtelomeric families, shown (Pretorius, 2000; Snow, 1983). This problem is by arrowheads (b) (Garcıá-Martıńez & Pe´rez-Ortıń, unpublished results); and (4) markedly fewer copies of Ty trans- further compounded by the HO nature of these posons (Hauser et al., 2001). The T73 genome shown probably strains. Haploid cells produced by sporulation has two copies of each chromosome except for chromosomes can change their mating type and conjugate to IV and X. For simplicity, we have shown just a single copy of form new diploid cells. The frequent use of chromosomes with two copies. For chromosomes with three copies, we show the name and just two copies. We have such mechanisms during vinification would included the two copies of chromosomes VIII and XVI to show lead to the generation of multiple genome the translocation between these chromosomes. combinations and very rapid changes. This COMPARATIVE GENOMICS AND THE ORIGIN OF THE S. CEREVISIAE GENOME 147 particular evolutionary mechanism has been by meiosis and conjugation. Consequently, all termed “genome renewal” (Mortimer et al., the homologous loci in a particular strain (two 1994; Mortimer, 2000). The proponents of this or more, depending on the case) must be manip- theory suggest that this renewal would give ulated in an identical fashion to ensure the rise to highly homozygous strains and eliminate phenotypic stability of the strain (Puig et al., deleterious mutations by natural selection. 1998, 2000). Natural strains are known, however, to be typically aneuploid (Bakalinsky & Snow, 1990; Guijo et al., 1997) and heterozygous for many loci 3. COMPARATIVE GENOMICS AND (Barre et al., 1993; Kunkee & Bisson, 1993), and THE ORIGIN OF THE S. CEREVISIAE such properties are inconsistent with the GENOME genome renewal hypothesis (Puig et al., 2000). While the possible influence of meiotic changes Although the origin of S. cerevisiae is cannot be entirely ruled out, there are other unknown, that of its genome can be investigated mechanisms that might explain the natural vari- by comparing genomes from natural strains of ation observed in wine strains. For instance, this species with those from other more-or- translocations mediated by Ty transposons less-related species. A better understanding of (Rachidi et al., 1999), mitotic crossing-over the origin and evolution of the S. cerevisiae (Aguilera et al., 2000), and gene conversion genome will have a positive impact in have all been described as mechanisms capable numerous areas. It will greatly improve our of causing the most rapid adaptive changes knowledge of the origin of the species and the (Puig et al., 2000). ways in which it has adapted to industrial The practice of inoculating must with pure processes over the years, and also shed light wine yeast cultures to improve the quality and on the mechanisms underlying the evolution homogeneity of wines produced from one year of its genome, and, by extension, that of other to the next dates back to the 1970s (Pretorius, eukaryotic organisms. 2000). Pure cultures have been obtained from Comparative genomics studies in yeasts have natural strains in wine-producing countries been performed by partial or complete around the world. In the first half of the twen- sequencing followed by bioinformatic compar- tieth century, these strains were selected and ison of sequence data and chromosomal organi- modified by more or less empirical methods. zation of genes. The first complete genome The selection techniques were improved in later sequence for S. cerevisiae was published for years, however, with the emergence of classical a laboratory strain in 1997 (Goffeau et al., genetic tools (reviewed in Pretorius, 2000). The 1997). The corresponding sequences for natural end of the twentieth century brought genetic wine strains were made available about 12 years engineering methods that opened up a world later (Borneman et al., 2008; Novo et al., 2009). of possibilities and further improved the quality Today, full genome sequences are available for of the selection methods used (see Chapter 8). several dozen S. cerevisiae strains, including The plasticity of the wine strain genome, laboratory, wine, and other strains (Liti et al., however, poses a new challenge, as there is a 2009; Schacherer et al., 2009). risk of genetically engineered changes becoming In 1997, it was suggested that the S. cerevisiae unstable with successive generations. Mutations genome was the result of an ancient duplication, or insertions in a single locus, for example, dating back approximately 108 years, of an could eventually be eliminated by gene conver- ancestral genome followed by the elimination sion, homologous recombination, or even perhaps of duplicated genes and the acquisition of new 148 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS functions for other genes (Wolfe & Shields, a much more recent role in facilitating adapta- 1997). This theory would explain the genetic tion to specific industrial processes. Indeed, redundancy detected in this species. S. cerevisiae various subtelomeric gene families are of has 2458 genes from 722 families containing immense importance to the biology of these between two and 108 members (Herrero et al., yeast strains. Based on the results of a compara- 2003). Part of the redundancy would be due to tive genomic study of multiple wine and non- ancestral duplication and part to smaller dupli- wine strains, Carreto et al. (2008) proposed cations that took place later (Llorente et al., that the diversity observed in the strains 2000). The existence of large numbers of gene analyzed was mainly the result of Ty element families is a common feature of hemiascomyce- insertions and subtelomeric recombination. tous yeasts. In a comparative genomic study of The fact that the subtelomeric regions of these yeasts, Malpertuy et al. (2000) found different chromosomes contain many members a substantial number of genes that do not exist of gene families involved in hexose transport in other organisms. The genes, which are (Bargues et al., 1996), use of natural carbon sour- specific to ascomycetes, seem to have evolved ces such as sucrose (Carlson et al., 1989), more rapidly and are perhaps responsible for maltose (Chow et al., 1989), and melibiose (Nau- the biological differences that characterize this mova et al., 1997), flocculation (Teunissen & group of yeasts. When this ancient duplication Steensma, 1995), and resistance to the toxicity actually took place in S. cerevisiae is a subject of molasses in which industrial yeasts are of debate. Langkjaer et al. (2003) postulated cultured (Ness & Aigle, 1995) suggests that that it was before the divergence of Saccharo- these regions might act as reservoirs of vari- myces and Kluyveromyces but other authors ability for rapid adaptations to the changing have suggested that it was later (Fares & Wolfe, environments to which industrial yeasts are 2003). In a study of collinearity (synteny) exposed. This mechanism may indeed still be between different hemiascomycete species, very active in certain strains such as Cava Llorente et al. (2000) proposed that the primary strains, in which high rates of subtelomeric vari- evolutionary mechanism (apart from global ability have been detected (Carro et al., 2003; genome duplication) was the duplication of Carro & Pinã, 2001). Small and large duplica- small regions (the length of a few genes) of the tions and translocations may also have contrib- genome followed by specialization or gene uted to speciation due to reproductive isolation loss. In related species, such as S. cerevisiae and in the Saccharomyces genus (Delneri et al., 2003; S. bayanus, the duplication sites tend to be Fischer et al., 2000, 2001). There may be other located close to copies of Ty transposons or in cases where the selection of one particular chro- subtelomeric regions where families of repeated mosomal rearrangement rather than another is genes are concentrated (Fischer et al., 2001). A random. Nonetheless, it is reasonable to think genomic comparison of S. cerevisiae, Saccharo- that many of the combinations produced by myces paradoxus, S. bayanus, and Saccharomyces the different genomic rearrangement mecha- mikatae found the greatest variability in subtelo- nisms discussed above have been selected meric regions, particularly in terms of repeated because they provide the organism with a partic- gene families (Kellis et al., 2003). These regions ular selective advantage. Our group found range in size from 7 to 52 kb and their function a case in which reciprocal translocation between might be to facilitate rapid changes via duplica- chromosomes VIII and XVI gave rise to a new, tion and translocation. While these mechanisms more efficient promoter for the sulfite resistance have played a part in the evolution of the gene SSU1 (Pe´rez-Ortıń et al., 2002a). As sulfite Saccharomyces genus, they have also had has been used as a treatment in vineyards, COMPARATIVE GENOMICS AND THE ORIGIN OF THE S. CEREVISIAE GENOME 149 wineries, and wines for thousands of years, ecosystem of human tissues (Schacherer et al., resistance to this substance was probably 2009). selected by wine strains as a useful survival Another interesting point worth noting is the mechanism. In an extensive study of transloca- discovery of hybrid wine yeasts derived from S. tion between various wine and non-wine cerevisiae and other Saccharomyces species. It has strains, our group found that the reciprocal been known for some time that certain lager translocation between chromosomes VIII and brewing strains have genomes derived from XVI was present in some but not all of the more than one species (Rainieri et al., 2006). wine strains, but was absent from all the non- These strains are partial allotetraploids that wine strains, providing evidence that this trans- arose from a natural hybridization event location is associated with the use of sulfite in between S. cerevisiae and a yeast similar to S. winemaking (Pe´rez-Ortıń et al., 2002a). In that bayanus (Nakao et al., 2009; Rainieri et al., study, we also detected a close phylogenetic 2006). More recently, however, there have also relationship between wine strains from been descriptions of wine strains with a genome geographically distant countries such as South containing chromosomes from more than one Africa, France, Japan, Spain, and the United species and wine yeast hybrids of S. bayanus States, suggesting that strains that had origi- and Saccharomyces kudriavzevii (Gonza´lez et al., nated in Europe were spread to other parts of 2006). Genomic analysis showed that all the the world with the expansion of winemaking. hybrids arose from a single hybridization event. The recent development of high-resolution The resulting genome would then have evolved genome mapping techniques such as mass through successive chromosome rearrange- sequencing and tiling array analysis (see Section ments resulting in the generation of hybrid 4) has permitted the genomic sequencing of chromosomes and the loss of several chromo- several dozen S. cerevisiae strains and the formu- some copies (mostly corresponding to S. lation of hypotheses regarding the origin of this kudriavzevii). Such rearrangements affected not species and that of other strains used for only sequences of transposons (as in the cases biotechnological purposes (brewing, bread described above) but also other conserved making, sake production) and pathogenic regions such as ribosomal DNA (rDNA) and strains isolated in immunosuppressed patients protein-encoding genes (Belloch et al., 2009). (Liti et al., 2009; Schacherer et al., 2009). Single The study of these hybrids is of practical nucleotide polymorphism (SNP) analysis has interest because they might have useful proper- shown that the genomes of different strains of ties for biotechnological applications. It is S. cerevisiae tend to represent a mosaic generated known, for example, that S. bayanus var. uvarum by recombination between lineages with is responsible for the fermentation of must at different geographical and/or ecological origins low temperatures and the production of large (Liti et al., 2009). What seems clear is that this quantities of glycerol and b-phenylethanol (Sol- species has been domesticated on various sepa- ieri et al., 2008). In an attempt to obtain yeast rate occasions, at least once in the case of wine strains with improved winemaking properties, fermentation and another time in the case of Solieri et al. (2008) constructed artificial hybrids sake fermentation (Liti et al., 2009). Today’s between S. cerevisiae and Saccharomyces uvarum strains would thus be derivatives and combina- by spore conjugation and found that the tions of those initial domesticated strains. Path- hybrids contained mitochondria from only one ogenic strains, however, seem to have arisen on of the two species and that the fermentative multiple occasions from wild and domesticated properties of the hybrid depended on these strains opportunistically adapted to the new mitochondria. 150 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

over previous techniques in terms of sensitivity, 4. THE USE OF S. CEREVISIAE AS transcript quantification, and, to some degree, A MODEL ORGANISM FOR THE resolution (Nagalakshmi et al., 2008). DEVELOPMENT OF DNA DNA microarrays have been widely used to MICROARRAY TECHNOLOGY investigate many aspects of S. cerevisiae metabolism (Figure 6.2). The technology has other uses, There are a number of reasons why many of however. Apart from providing valuable infor- the technologies used in the field of genomics mation on metabolic activity in different condi- were developed using S. cerevisiae, but the main tions and mutants, it has also been used to one is probably that it was the first organism to investigate the effects of many drugs and toxic be analyzed in a genomic sequencing project products on gene expression and to analyze that generated numerous functional genomics genomic variations in S. cerevisiae and related studies even before the full sequence was pub- species. All of these uses have also been applied lished (Goffeau et al., 1997). The fact that S. cere- to wine yeast strains. visiae has been used as a model organism for genetics and molecular biology since the 1940s 4.1. Metabolic Studies has given rise to an enormous number of very powerful tools for these types of analysis. As Given the vast information already available a result of these developments, our knowledge on yeast regulatory pathways, global expression of the genetics and biology of this yeast is unpar- studies should be able to provide sufficient data alleled. The only other organism that has been so to allow individual genes to be linked to one or thoroughly investigated is perhaps Escherichia more phenotypes or metabolic pathways. It coli. Even before the emergence of DNA microar- should also theoretically be possible to deter- ray technology, S. cerevisiae was used in the mine the components of each of these pathways, development of numerous methods for the to provide, for the first time, a global view of global analysis of gene expression such as Serial a eukaryotic cell. The first global gene expres- Analysis of Gene Expression (SAGE) technology, sion study, performed by Pat Brown’s group, which was used to perform the first analysis of used DNA microarray analysis to study gene the entire messenger RNA (mRNA) complement expression in S. cerevisiae during growth in (baptized transcriptome) of a cell (Velculescu glucose and during the shift from fermentative et al., 1997). As with many other technologies, to respiratory growth (DeRisi et al., 1997). The SAGE was later used to analyze other organisms study has already become a classic in its field with great success (Velculescu et al., 2000). While and has been cited over 2500 times (as of August SAGE is an extremely powerful tool, capable of 2009). Similar studies have analyzed other accurately quantifying the number of copies of processes or situations that involve metabolic mRNA present in a cell, it has largely been changes. Transcriptional changes in S. cerevisiae, replaced by DNA microarray analysis, which is for example, have been analyzed in the change a much simpler and less costly technology. In from a fermentable to a nonfermentable carbon recent years, however, the development of source (Kuhn et al., 2001), in aerobic compared high-throughput sequencing techniques (also to anaerobic conditions in a continuous-culture developed using S. cerevisiae) has led to study (ter Linde et al., 1999), in the lag phase a renewed interest in tag-sequencing technolo- prior to active culture growth (Brejning et al., gies. RNA-seq, for example, has been success- 2003), during sporulation (Chu et al., 1998), fully used to characterize the transcriptome of and during the cell cycle (Cho et al., 1998). S. cerevisiae with considerable improvements Another major research focus is the functional THE USE OF S. CEREVISIAE AS A MODEL ORGANISM FOR THE DEVELOPMENT 151

Control sample Test sample

1. DNA or RNA extraction

2. Labeling with fluorescent dyes or radioactive isotopes

3. Hybridization

4. Arrays

5. Laser scanning Control image Test image

6. Image analysis and quantification

Control data Test data

7. Comparison

8. Diagnosis

FIGURE 6.2 DNA microarray analysis. 1. RNA or DNA is extracted from a test and a control sample using conventional methods. 2. In the case of microarrays on glass slides, the probes are labeled with fluorescent dyes (using a different fluorophore for the test and control sample). The probes used in macroarrays on nylon filters are labeled with radioactive isotopes. 3. Just one hybridization step is used in glass-slide microarrays as these involve the use of a single array with both the test and control samples mixed together prior to hybridization. Two hybridization steps are required for nylon-filter macroarrays. These steps are preferably performed on the same filter but they need to be sequential as the probes are labeled with radioactive isotopes. 4. Following hybridization, the arrays are washed to allow detection of the different hybridization signals. 5. The hybridization images are captured using a laser scanner. This is done directly using two different lasers (one for each fluorophore) in the case of microarray analysis. In macroarray analysis, however, latent images are generated on special screens and later scanned by laser. 6. The readings generate an image for each sample. The intensity is then quantified using special software that generates hybridization intensity data that allows comparison of the samples. 7. Statistically significant differences are analyzed using purpose-designed programs. 8. The final stage involves the formulation of corresponding hypotheses and conclusions. analysis of transcription factors via overexpres- Clusters of genes that display identical or sion or analysis of null or conditional mutants similar expression patterns under the different (Carmel-Harel et al., 2001; DeRisi et al., 1997; conditions studied have been used to identify Holstege et al., 1998). the functions of individual genes based on the 152 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS assumption that coregulated genes must be inducing haploinsufficiency, which consists of involved in the same metabolic pathways. The studying growth deficiencies caused by the most common way to conduct a study of this loss of one of the two gene copies in a diploid type is to use clustering algorithms to group cell. To perform a systematic, comparative genes by expression profiles (reviewed in study, it is necessary to have a full collection of Hughes & Shoemaker, 2001 and Brazma & a diploid strain in which each gene has been Vilo, 2000) in order to identify groups that deleted and replaced with a specific sequence have putative functional relationships. Another tag (Winzeler et al., 1999). In these studies, the way is to search for transcription-factor-binding full collection of approximately 6000 strains sites in gene promoters. Two types of study with single deletions is grown together under have been used for this purpose: in silico particular conditions (such as the presence of comparison of promoter sequences (Brazma a drug) and strains that exhibit delayed growth et al., 1998; Bussemaker et al., 2000; Hampson compared to wild-type strains indicate genes et al., 2000; Roth et al., 1998) and in vivo studies that are necessary for resistance to certain drugs of genome-wide transcription-factor-binding or culture conditions (Giaever et al., 1999, 2002). sites using a technique called Chip-ChIP, which This technique can uncover subtle growth is a combination of DNA microarray analysis differences that would otherwise remain unde- (Chip) and chromatin immunoprecipitation tected. Up to 6000 strains can be compared (ChIP). simultaneously thanks to the sequence tags present in each strain, which enable an accurate 4.2. Effects of Drugs and Other count to be made of the cells in a strain at any External Factors moment using special DNA microarrays containing probes for each sequence tag. These DNA microarray technology can be used to studies open new perspectives not only for measure, in a single experiment, an organism’s pharmacogenomics but also for the study of global transcriptional response to treatment the effect on wine yeasts of toxic substances with an external factor such as a drug or envi- such as alcohol, pesticides, and treatments ronmental agent (Gasch et al., 2000; Hughes such as copper and sulfite. Although most of et al., 2000a; Jelinsky et al., 2000; Jelinsky & the studies to date have been conducted using Samson, 1999). Because the response of genes standard laboratory strains, the results can be to experimental conditions is a dynamic process easily extrapolated to industrial strains. characterized by multiple interactions, analyzing The fermentation of sugars by wine yeasts is responses to external agents can reveal func- followed by rapid growth and carbon dioxide tional relationships within or between metabolic production, which can be interrupted with the pathways. Such techniques have been used to depletion of carbon or nitrogen sources or the analyze, for example, the transcriptional appearance of growth inhibitors (reviewed in response to inhibition of translation or amino Pretorius, 2000). An improved understanding acid biosynthesis, or to compounds with anti- of the metabolic changes that occur in the shift fungal activity (Bammert & Fosel, 2000; Hard- from one carbon source to another (DeRisi wick et al., 1999; Jia et al., 2000). Molecular et al., 1997; ter Linde et al., 1999) and of meta- targets of specific drugs can also be identified bolic signal transduction pathways (Hardwick by comparing expression profiles induced et al., 1999; Ogawa et al., 2000) will contribute by a particular drug with those induced in to improving the technical aspects of fermenta- mutants for specific genes (Hughes et al., tion processes in wineries and help to prevent 2000a). Similar results can be achieved by stuck fermentations. THE USE OF S. CEREVISIAE AS A MODEL ORGANISM FOR THE DEVELOPMENT 153

In an extensive large-scale experiment that industrial strains. Since this yeast plays a key analyzed response to many of the stress condi- role in winemaking and has an enormous influ- tions to which yeasts are exposed, Gasch et al. ence on the final product, it is important to (2000) found that the transcriptional response understand the molecular events underlying to almost all of the stress factors tested was fermentation and the influence of the winery practically identical across a large group of and vintage, and of the physical, biological, genes. The authors termed this the “environ- and chemical properties of the must, on this mental stress response” (ESR). The experiment process. Such an understanding would be provided a basis for further tests with wine greatly enhanced by analysis of the gene expres- strains exposed to general or specific stresses sion profiles of these yeasts in different growth associated with wine fermentation. Indeed, the conditions. Before the emergence of DNA first experiments of this type have already microarray technology, the expression profiles been performed (see below). The ultimate aim of only a small number of genes at a time could of such studies is to identify the most suitable be analyzed in wine yeasts (see Chapter 2). Most strains for the various fermentation conditions of the DNA microarray experiments described found in different wineries and wines. Similar so far in this chapter, however, have analyzed experiments involving a limited group of genes laboratory strains, which are incapable of wine have also been performed (Ivorra et al., 1999). A fermentation. more detailed discussion is given in Chapter 2 Various strategies have been employed in of this book. On investigating the effect of studies using DNA microarrays to analyze ethanol on laboratory yeast strains by DNA expression profiles in wine yeasts. Two studies microarray analysis, Alexandre et al. (2001) have been conducted using laboratory media concluded that cells used ionic homeostasis, and culture conditions (Cavalieri et al., 2000; heat protection, and antioxidant defense, in Hauser et al., 2001), whereas others have used addition to previously described mechanisms, synthetic musts that reproduce the conditions to respond to stress. In a study of the effect of found in a natural environment but provide copper excess and deficiency on laboratory the means to accurately determine and repro- strains, also using DNA microarrays, Gross duce the composition of the must (Backhus et al. (2000) found that a small number of genes et al., 2001; Rossignol et al., 2003). Another were differentially expressed and that some of strategy has involved the use of grape juice these were involved in the iron uptake system. medium sterilized by filtration (Marks et al., This finding suggests that the copper and iron 2003; Mendes-Ferreira et al., 2007a, 2007b). uptake systems might be related. Because The use of standard laboratory conditions has copper is commonly used to inhibit bacterial the advantage of allowing comparison of data and fungal growth in wines, wine yeast strains from wine strains with those from the more must be able to endure elevated copper concen- extensively studied laboratory strains. The enor- trations and it would be useful to determine mous amounts of information available on refer- how they have achieved this capacity. ence strains can thus be used to undertake a much more in-depth investigation of the meta- 4.3. Use of DNA Microarrays in the bolic pathways and molecular mechanisms Analysis of Wine Yeasts underlying wine yeast fermentation. Cavalieri et al. (2000), for example, detected at least two- S. cerevisiae was also the first microorganism fold variability in global expression levels for in which genomic tools such as DNA microar- 6% of the genome between progeny of a natural ray analysis were used to analyze natural and wine strain isolate. Their findings indicate that 154 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS wine strains are highly heterozygous. Because known markers for metabolic phenotype as most of the metabolic differences segregated as they are connected with cell-cycle progression a suite of traits, the authors concluded that they (Patton et al., 2000). The effect of nitrogen avail- were the result of changes in a small number of ability on the growth of wine yeasts has been regulatory genes. One specific example would analyzed in two recent studies. One of these be the genes involved in the biosynthesis of compared global gene expression profiles in amino acids. There have also been descriptions synthetic media containing high and low of other phenotypes caused by changes in struc- concentrations of arginine (a source of nitrogen) tural rather than regulatory genes, explaining (Backhus et al., 2001), whereas the other why these changes are not associated with other compared expression profiles in a Riesling phenotypes. Examples include the YHB1 gene must with normal concentrations of nitrogen (Hauser et al., 2001), genes involved in resis- and another to which diammonium phosphate tance to sulfite (Pe´rez-Ortıń et al., 2002a) and (DAP) was added during the late fermentation copper, and the filigreed phenotype (Cavalieri phase, when yeast growth is no longer active et al., 2000). (Marks et al., 2003). In the first study, it was It is important to conduct experiments in real- found that nitrogen limitation induced genes life conditions as, although laboratory culture that would normally be repressed by the high conditions greatly facilitate analysis, they do not concentrations of glucose in the must. This fully reproduce the conditions found in natural suggests that, in the growth conditions that environments. Given the variability of natural characterize the fermentation of must contain- musts, one option is to use synthetic musts, ing high concentrations of sugars and nitrogen, which mimic natural conditions but can be the use of glucose might be diverted, at least easily reproduced in different laboratories. In partly, to a respiratory metabolism (Backhus a study of this type, using macroarrays and et al., 2001). This effect would be similar to various wine strains with different fermentative what is known as the Pasteur effect, which is capacity, Zuzua´rregui and del Olmo (2004) the inhibition of fermentation in the presence found that the expression levels of certain of oxygen. Although this effect has been stress-response genes were similar across the reported to be irrelevant for yeast in laboratory strains. They also found that the mRNA levels growth conditions (Lagunas, 1986), it might of many of these genes remained very high in occur in the fermentation of musts with low the strains with weaker fermentative capacity. levels of nitrogen, and, accordingly, cause slug- Their results demonstrated that it is possible to gish or stuck fermentations. Indeed, it is stan- establish a correlation between stress resistance dard practice in wineries to add DAP in such and fermentation capacity. cases. A study by Marks et al. (2003) found The amount of available nitrogen is consid- that the addition of DAP affected the expression ered to be one of the main limiting factors for of 350 genes. The 185 genes that were found to yeast growth in musts (reviewed in Pretorius, be downregulated encoded small-molecule 2000). Studies performed with wine yeasts transporters and nitrogen catabolic enzymes, have generally found high expression levels including enzymes involved in the synthesis of for genes linked to amino acid and purine urea, which is a precursor of ethyl carbamate. biosynthesis (Backhus et al., 2001; Cavalieri The other 165 genes affected were all upregu- et al., 2000; Hauser et al., 2001), which are indic- lated. These included genes involved in the ative of high growth rates. Activation of the biosynthesis of amino acids, purines, and ribo- methionine biosynthesis pathway and alter- somal proteins (suggesting a more active metab- ations in sulfate and nitrogen assimilation are olism despite an absence of cell proliferation) THE USE OF S. CEREVISIAE AS A MODEL ORGANISM FOR THE DEVELOPMENT 155 and assimilation of inorganic sulfate (necessary sluggish or stuck fermentations (Mendes-Ferre- for the elimination of hydrogen sulfide). The ira et al., 2007b). The study also demonstrated results of the study by Marks et al. provide that the main transcriptional effect of adding a possible explanation for why the addition of nitrogen was an upregulation in genes involved DAP reduces the production of ethyl carbamate in glycolysis, thiamine metabolism, and energy and hydrogen sulfide, two undesirable compo- pathways (Mendes-Ferreira et al., 2007a), find- nents in wines. They are also consistent with ings that are similar to those reported by Marks results from a study that analyzed samples et al. (2003) following DAP addition. A study taken at different time points during fermenta- performed by Jimeńez-Martı´ and del Olmo tion of a synthetic must with a relatively low (2008) showed that the effect of nitrogen refeed- level of nitrogen (300 mg/L). The authors ing depended on the source of nitrogen used, reported that the gene expression pattern as they detected differences in gene expression observed could be explained by entry into the reprogramming depending on whether ammonia stationary phase (cell proliferation arrest) in or amino acids were added. The addition of response to nitrogen depletion; they also ammonia resulted in higher levels of genes reported that the process was regulated by the involved in amino acid biosynthesis, whereas TOR pathway (Rossignol et al., 2003). that of amino acids directly prepared cells for A more comprehensive and realistic study protein biosynthesis. of transcriptional response in S. cerevisiae to Global gene response has also been analyzed different nitrogen concentrations during alco- in low-temperature winemaking conditions, holic fermentation was published more recently which are widely considered to improve the (Mendes-Ferreira et al., 2007a, 2007b). The sensory quality of wine. In experiments carried authors, using real grape must, compared 11 out at 13 and 25C, Beltrań et al. (2006) observed samples from different time points of a series of that the lower temperature induced cold stress- control vinifications, nitrogen-limiting fermen- response genes at the initial stage of fermenta- tations, and fermentations to which DAP was tion and increased levels of genes involved in added. They found alterations in approximately cell cycle, growth control, and maintenance in 70% of the yeast transcriptome in at least one of the middle and late stages of fermentation. the fermentation stages and also showed a clear Furthermore, several genes involved in mito- association between these changes and nitrogen chondrial short-chain fatty acid synthesis were concentrations. In agreement with earlier find- found to be overexpressed at 13C compared ings published by Backhus et al. (2001), their to 25C. These transcriptional changes were results indicated that early response to nitrogen correlated with higher cell viability, improved limitation involved the induction of genes asso- ethanol tolerance, and increased production of ciated with respiratory metabolism and a subse- short-chain fatty acids and associated esters. quent general decrease in the levels of genes The natural environment of S. cerevisiae has associated with catabolism. Curiously, they shaped the evolution of this organism’s metabo- also found a slight increase in the expression lism to allow it to exploit the anaerobic condi- level of genes encoding ribosomal proteins tions and high ethanol levels that characterize and involved in ribosome biogenesis during fermentation and to tolerate high levels of nitrogen depletion. In total, 36 genes were found certain compounds that are common during to be overexpressed when nitrogen levels were alcoholic fermentation. All these situations, low or absent compared to when DAP was however, are causes of stress for S. cerevisiae added. These signature genes might be useful and are reflected in the yeast’s gene expression for predicting nitrogen deficiency and detecting pattern, even though the organism is capable 156 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS of responding effectively to these stresses. As FSR was found to overlap only partially with has already been discussed, differential expres- the ESR (Gasch et al., 2000). Interestingly, 62% sion of certain stress-response genes has been of the FSR genes were novel, suggesting that detected in wine yeasts. The expression levels the stress conditions in wine fermentation of genes involved in oxidative metabolism, for were rather different from those observed in example, are low (Backhus et al., 2001). The laboratory conditions. Also of interest was the results of the fermentation monitoring study fact that respiratory and gluconeogenesis genes conducted by Rossignol et al. (2003) indicate were expressed even in high glucose concentra- that anaerobic stress is a characteristic of wine tions and that ethanol accumulation, at least in fermentation and that the absence of ergosterol the experiment by Gasch et al., was the main synthesis, one of the main growth-limiting reason for entry into the stationary phase. factors for yeasts in musts with low oxygen Because compounds such as copper sulfate and high ethanol levels (see Pretorius, 2000), and sodium bisulfate have been used for many is due to the continuous decrease in the expres- years to inhibit fungal and bacterial growth on sion levels of genes involved in ergosterol vines and grapes and in wines, wine strains biosynthesis. might very well respond more efficiently than Ethanol stress is another major pressure that other strains to these stresses thanks to the over- S. cerevisiae has to deal with during vinification. expression of certain detoxifying genes. Indeed, Ethanol tolerance is not fully understood (Pre- wine strains have been found to overexpress torius, 2000) but it is known to partly depend genes involved in the transport of sulfur on alterations in the plasma membrane. Genes (SUL1-2) and sulfite (SSU1)(Cavalieri et al., encoding enzymes involved in the synthesis 2000; Hauser et al., 2001). It can be concluded of fatty acids, phospholipids, and ergosterol that the pressures to which wine strains have are highly expressed (Backhus et al., 2001)in been exposed over thousands of years have S. cerevisiae yeasts but decrease towards the led to the selection of strains that are better stationary phase (Rossignol et al., 2003). Using adapted to the fermentation conditions found microarray analysis to identify target genes in wineries. Strains that have developed resis- and analyze ethanol sensitivity in knockout tance to treatments such as copper sulfate and strains, Hirasawa et al. (2007) found that the sodium bisulfate are a good example of this biosynthesis of tryptophan can confer ethanol adaptation. tolerance. Ethanol stress, however, does not Finally, two studies have analyzed the appear to be the main pressure in vinification. genomic response in a commercial wine yeast The greatest effect on gene expression is strain to rehydration and adaptation to osmotic produced upon entry into the stationary phase stress at the beginning of vinification. In the first (Rossignol et al., 2003). The changes in gene study, rehydration was carried out in a complete expression seen in this phase, however, appear glucose medium to identify events related to to differ from those observed under laboratory re-establishment of fermentation (Rossignol conditions (Gasch et al., 2000). et al., 2006). The authors reported substantial In a comprehensive study of the transition transcriptional changes. The expression profile from the exponential to the stationary phase in observed in the dried yeasts was characteristic wine fermentation, Marks et al. (2008) discov- of cells grown under respiratory conditions ered 223 genes that were dramatically induced and exposed to nitrogen and carbon starvation at various points during fermentation. They and considerable stress during rehydration. called this the “fermentation stress response” Furthermore, many genes involved in biosyn- (FSR). The most interesting point was that the thetic pathways (transcription or protein synthesis) THE USE OF S. CEREVISIAE AS A MODEL ORGANISM FOR THE DEVELOPMENT 157 were coordinately induced while those subject Gene expression profiling of industrial to glucose repression were downregulated. strains may also help to uncover as-yet- While expression of general stress-response unknown functions of numerous genes in the genes was repressed during rehydration, S. cerevisiae genome, as these genes might only despite the high sugar levels, that of acid-stress have relevant functions in industrial fermenta- genes was induced, probably in response to the tion conditions. For instance, 130 genes from accumulation of organic acids. In the second various subtelomeric families of unknown func- study, rehydration was carried out in water to tion (PAU, AAD, COS) have been found to be separate this process from adaptation to osmotic induced during wine fermentation (Rossignol pressure (Novo et al., 2007). The results of the et al., 2003), indicating that they probably have study showed that rehydration for an additional an important role in this process. It should hour (following an initial period of 30 min) did also be noted that 28% of the FSR genes detected not induce any relevant changes in global gene in the experiment by Marks et al. (2008) expression. The incubation of rehydrated cells described above had an unknown function. in a medium containing fermentable carbon sources activates genes involved in the fermen- 4.4. Genomic Studies tation pathway, the nonoxidative branch of the pentose phosphate pathway, ribosomal biogen- DNA microarray analysis is also a promising esis, and protein synthesis. tool for the study of wine strain genomes. This Erasmus et al. (2003) analyzed yeast response technology forms the basis for various types of to high sugar concentrations by inoculating study in this area, including Affymetrix oligo- rehydrated wine yeast in Riesling grape juice nucleotide microarray analyses. These microar- containing equimolar amounts of glucose and rays consist of a very large number of short fructose to a final concentration of 40% (wt/ oligonucleotide sequences derived from the vol) and comparing global gene expression reference S. cerevisiae laboratory strain S288c. with that observed in yeasts inoculated in the The oligonucleotides represent all the open same must containing 22% sugar. Although reading frames (ORFs) distributed throughout the sugar concentration used is not generally the yeast genome. In this method, hybridization found in winemaking conditions, some of the is highly dependent on the identity of the results coincided with those reported by Rossi- sequence, and a single nucleotide change will gnol et al. (2003), with sugar stress resulting in alter the hybridization signal. Thus, the signals the apparent upregulation of glycolytic and produced by a particular strain can be pentose phosphate pathway genes and struc- compared with those from a reference strain to tural genes involved in the formation of acetic identify sequence changes, including SNPs. acid from acetaldehyde and succinic acid from The method has been successfully used to study glutamate and the downregulation of genes polymorphisms in various strains (Primig et al., involved in the de novo biosynthesis of purines, 2000; Winzeler et al., 1998). Affymetrix also pyrimidines, histidine, and lysine. The authors manufactures tiling arrays, another type of also reported considerable changes in the oligonucleotide microarray system that covers expression levels of stress-response genes. the entire sequence of the yeast genome. Tiling These changes affected, among others, genes arrays are used for transcriptome mapping involved in the production of the compatible and to identify transcripts that do not corre- osmolyte glycerol (GPD1) and genes encoding spond to annotated genes (Royce et al., 2005). the heat shock proteins HSP104/12/26/30/42/ These arrays have also been used for detailed 78/82 and SSA3/4. genomic analysis. As described in Section 3, 158 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

Schacherer et al. (2009) used this method to rese- gene CUP1. Curiously, CUP1 has a deletion in quence 63 yeast strains, including 14 wine the genomic region of the wine strain (Pe´rez- strains. Ortıń et al., 2002b), which reduces its expression There also exist tiling arrays with long oligo- levels (Hauser et al., 2001). The study by Hauser nucleotides (manufactured by Agilent, for et al. found that the number of Ty transposons example) and arrays containing probes spotted (Ty1, Ty2, Ty3, and Ty4) was greatly reduced at a lower density than that seen in tiling arrays in the T73 wine strain compared to the S288c (oligonucleotides over 60 bases long or double- laboratory strain. This finding was consistent strand fragments of 300 or more bases). These with less-complete previously published results tools, however, are not suitable for detecting iso- (Jordan & McDonald, 1999), with later results lated sequence variations. Microarrays consist- (Carreto et al., 2008), and with results for brewing ing of long oligonucleotides or double-strand strains (Codoń et al., 1998) and suggests that the fragments are, however, useful for genomic colonization of the genome of laboratory strains comparisons designed to identify increases or by these molecular parasites may be recent. The decreases in the number of copies of a particular strong selective pressure exerted on wine strains gene or chromosomal region. The first study of might have prevented the excessive accumula- this type was conducted by Hughes et al. tion of sequences of this type (Jordan & (2000b) using laboratory strains. A similar study McDonald, 1999). by Infante et al. (2003) that analyzed S. cerevisiae The flexibility of DNA chip technology flor yeast strains found that two natural strains means that purpose-designed arrays can be had differences in the copy number of 38% of created for specific studies. In a study of chro- their genes, which illustrates the enormous mosomal rearrangements in Cava strains genomic variability that characterizes yeasts of (secondary fermentation), Carro et al. (2003) this type. In many cases, the differences were used specially designed and constructed macro- in regions flanked by Ty transposons and other arrays containing 14 chromosome I probes and regions with a high recombination rate, which hybridized them with DNA from chromosome would explain the amplification or deletion I isolated from various Cava strains with length events observed. The authors suggested that variations in this chromosome. Their results such regions were the site of double-strand indicated the existence of a subtelomeric region breaks responsible for free ends capable of that tends to be deleted in the right arm of chro- recombination with short homologous regions mosome I of this highly variable strain. (10e18 base pairs). A similar mechanism has been described for the SSU1 gene region in wine strains (see Section 4.3). In the case of flor 5. PROTEOMIC ANALYSIS OF WINE yeast strains, the continuous presence of acetal- STRAINS dehyde and ethanol in the medium would increase the frequency of double-strand breaks, DNA microarray technology allows the conferring a selective advantage on strains that expression of all the genes in a particular have adapted to this hostile environment. organism (the transcriptome) to be analyzed. DNA macroarray analysis has also been used Global analyses can thus be used to assess the to study gross gene expression profiles in the T73 effects of physical, chemical, and biological wine strain (Pe´rez-Ortıń et al., 2002b). The study agents, and even specific mutations, on gene revealed numerous copy-number variations for expression. Nonetheless, analysis of mRNA genes from subtelomeric families and a number levels is not sufficient for a complete description of other genes such as the copper resistance of biological systems. This also requires accurate PROTEOMIC ANALYSIS OF WINE STRAINS 159

FIGURE 6.3 Standard proteomic analysis by two- dimensional (2D) gel electrophoresis and mass spectrometry (MS). The method consists of three fully integrated steps. In the first step, the proteins are separated on 2D gels, stained, and then individual spots isolated. The protein spots are then digested with trypsin and the resulting peptides are separated by high-performance liquid chromatography (HPLC). In the second step, each eluted peptide is ionized by elec- trospray ionization. It then enters the mass spectrom- eter through the first quadrupole mass filter (Q1) and is fragmented in a collision cell (Q2). The resulting spectrum is recorded (Q3). In the third step, the tandem MS spectrum of a selected ionized peptide contains sufficient specific sequencing information to identify the peptide and its associated protein. m/z ¼ mass to charge ratio. measurement of the expression and activity of of the state of a biological system. In other the corresponding proteins (the proteome). words, proteome analysis provides a better Furthermore, even though expression levels of picture of an organism’s phenotype than does different mRNA species and the proteins they the analysis of mRNA levels. encode are correlated, this correlation is not While there are vast amounts of genomic perfect for all genes (Futcher et al., 1999; Ideker data available for yeasts (including sequence et al., 2001). Of even greater importance, and gene expression data obtained by DNA however, is the level of correlation between microarray analysis), the yeast proteome is changes in mRNA and protein levels. While still largely undefined (Fey et al., 1997). This changes in the proteome and transcriptome is particularly true for yeasts of industrial generally occur in parallel (homodirectional and biotechnological interest, as most of the changes), the multiple effects caused by post- studies to date have analyzed laboratory transcriptional regulation justify the need for strains (Link et al., 1999; Washburn et al., proteomic studies (Griffin et al., 2002; Ideker 2001). The first comparative study in this et al., 2001). Thus, proteomics, which is the anal- area, performed using three haploid strains ysis of the full complement of proteins derived from laboratory strains, led the authors expressed by a genome (Pennington et al., to conclude that differences in protein expres- 1997; see Figure 6.3), is considered to be the sion level and post-translational modifications best tool for obtaining a quantitative description influenced the molecular and biochemical 160 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS characteristics of cells and were possibly corroborated by Fauchon et al. (2002), who responsible for the different mutant pheno- related cadmium stress with sulfur metabolism. types observed in these strains (Rogowska- As GSH is essential for the detoxification of Wrzesinska et al., 2001). cadmium, when exposed to this substance, cells Several studies have analyzed the effect of convert most sulfur into GSH. The cells change environmental stresses on proteome-level their proteome to reduce the production of responses in laboratory strains. These studies sulfur-rich proteins to permit optimal GSH are similar to those conducted in the area of turnover and ensure optimal levels of this essen- genomics analyzing the influence of environ- tial compound. It has been estimated that this mental factors on global gene expression in change allows for a 30% reduction in sulfur laboratory strains. One such proteomic study amino acid incorporation into proteins, which analyzed oxidative stress caused by hydrogen would enable a considerable increase in GSH peroxide (Godon et al., 1998) leading to the production and thus ensure cell survival. This expression of batteries of genes referred to by is a clear example of the important role of pro- the authors as “stimulons.” The expression of teome plasticity in yeast cell adaptation to 115 proteins with different functional roles was adverse conditions and agents. observed. These included proteins linked to Little information is available on the proteo- antioxidant activity, heat shock response, and mic profiles of industrial yeasts as most of the protease activity. The expression of 52 proteins, studies in this area have been carried out using including metabolic enzymes and proteins laboratory strains. In two studies involving the involved in translation, was repressed. In analysis and identification of over 200 proteins, another study of S. cerevisiae, sorbic acid was Joubert et al. (2000, 2001) concluded that the K11 found to produce slightly different and less brewing strain was a hybrid of S. cerevisiae and drastic effects, although it did reveal expression Saccharomyces pastorianus (S. bayanus). Their of stress-response proteins (mainly linked to work also led them to postulate that the physio- oxidative stress) and several molecular chaper- logical properties required by top-fermenting ones (Hsp12, 26, 42, and some isoforms of (ale) strains (flocculation and fermentation at Hsp70) (de Nobel et al., 2001). Analysis of low temperatures) might have been acquired mRNA levels following the induction of sorbic by hybridization. Their reasoning was based acid stress showed that these were poorly corre- on the fact that, unlike bottom-fermenting lated with protein abundance. (lager) strains, which are all hybrids, top-fer- In another proteome analysis, the addition of menting strains are not hybrids and are very þ2 cadmium (Cd ) induced expression of 54 closely related to S. cerevisiae laboratory strains. proteins and repressed that of a further 43 The two types of brewing strain also have very (Vido et al., 2001). Of these, nine enzymes different physiological properties. involved in the sulfur amino acid biosynthesis Trabalzini et al. (2003) studied the proteomic pathway and glutathione (GSH) synthesis response in a wine strain of S. cerevisiae (k310) were strongly induced, as were proteins with isolated during spontaneous wine fermentation. antioxidant activity. Although Cdþ2 is not an Wine strains are exposed to numerous hostile active redox ion, it can cause oxidative stress conditions during fermentation. Unlike other and lipid peroxidation and also affect cellular studies, which have analyzed isolated effects of thiol redox balance. These data suggest that environmental stress on yeasts, the study by the two cellular thiol redox systemsdGSH and Trabalzini et al. investigated physiological thioredoxindare essential protection mecha- response to fermentation stress; in particular, nisms against cadmium stress, a theory later depletion of the main carbon source and glucose, OTHER GLOBAL STUDIES 161 and increasing ethanol levels. They found that two wine strains with different fermentative specific proteins, which differed from those capacities and found that one of the strains observed for other S. cerevisiae strains (such as was incapable of completing fermentation. those used in bread making), were either Although the transcriptome and proteome anal- induced or repressed in response to these phys- yses revealed specific differences, they both iological stresses. The proteomic response also indicated that the strain with fermentation diffi- involved the induction of intracellular proteo- culty had defects, namely excess proton uptake lysis, which appeared to be directed towards (a sign of ethanol intolerance) and increased certain classes of protein. The main inference oxidative damage due to elevated levels of acet- from this study is that the proteomic response aldehyde. In the second study, Rossignol et al. to fermentation stress in a wine strain of S. cere- (2009) compared proteomic changes in a wine visiae is largely directed at mitigating the effects strain between the exponential growth phase of increasing ethanol levels. Ethanol stress has and the stationary phase during wine fermenta- been associated with both oxidative damage tion. They found major changes in the abun- (due to an increased production of free radicals) dance of proteins related to glycolysis, ethanol and cytotoxic effects (due to acetaldehyde production, and amino acid metabolism. The production). Ethanol also induces the expression most interesting finding was that these changes of heat shock proteins and proteins involved in were very poorly correlated with previously trehalose metabolism, whose purpose is to stabi- observed transcriptional changes (Rossignol lize membranes and proteins and suppress et al., 2003), which suggests that post-transcrip- protein aggregation. It is extremely important tional regulatory mechanisms are very impor- to further investigate proteomic responses in tant in the late stages of wine fermentation. A fermentation yeasts as a good wine strain must recent study involving laboratory strains and be capable of overcoming the hostile conditions laboratory culture conditions with various it is faced with in industrial processes. Addition- nutrient deficiencies indicated that the response ally, the cell changes that occur in S. cerevisiae to nitrogen depletion was fundamentally during fermentation (autoproteolysis) and aging controlled at a translational and not a transcrip- (autolysis) are responsible for the organoleptic tional level (Kolkman et al., 2006). properties of wine. Accordingly, the amount The importance of gaining a comprehensive of nitrogen in autolysates together with free understanding of proteomic response in amino acid concentrations, which differ greatly fermentation yeasts is thus clear: it will greatly depending on the yeast strain, can have a consid- contribute to improving the organoleptic prop- erable influence on the flavor, composition, and erties associated with high-quality wines. quality of the final product (Martıńez-Rodrı´guez et al., 2001a, 2001b). Proteolytic enzymes might be involved in the turnover of nitrogenous 6. OTHER GLOBAL STUDIES compounds before and during autolysis in winemaking conditions. It has also been proposed One of the aims of large-scale studies is to that yeasts might use amino acids not only as provide a global view of living systems. Geno- sources of nitrogen but also to restore the redox mics, for example, focuses on the full genome balance in critical environmental conditions to help understand the relevance of individual (Mauricio et al., 2001). genes, while transcriptomics and proteomics Two recent studies have compared the tran- analyze the link between physiological changes scriptome and proteome of wine yeasts. In the and changes in transcript and protein expres- first of these, Zuzua´rregui et al. (2006) compared sion levels with respect to total RNA or protein 162 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS expression levels. Most of the large-scale func- group (Allen et al., 2003) to permit large-scale tional studies conducted to date have been analyses by optimizing the experimental condi- based on transcriptomic and proteomic anal- tions and surmounting the technical difficulty of yses. A more recent “omic” approach, metabolo- measuring intracellular metabolites, which have mics, aims to characterize the physiological a rapid turnover and need to be separated from state of a cell by determining the concentration the extracellular space. The optimization of mass of all of the small molecules that comprise the spectrometry has allowed the analysis of extra- metabolism and identifying metabolic path- cellular metabolites in spent culture medium. ways and fluxes. This approach may provide It is also possible to study and define specific the best and most direct measurement of an metabolic pathways by integrating and incorpo- organism’s physiological activity and bring us rating data obtained using the technologies dis- a little closer to a true approximation of its cussed in this chapter into biological models to phenotype since, as stated by Delneri et al. predict cell behavior that can then be tested (2001), “mRNA molecules are not functional experimentally. Ideker et al. (2001), for example, entities within the cell, but simply transmitters used a combined genomic and proteomic of the instructions for synthesising proteins. approach to elucidate the galactose utilization proteins and metabolites [in contrast] represent metabolic pathway. They followed a typical true functional entities within cells” (p. 87). strategy used in systems biology. The steps Furthermore, the use of metabolomic data in they described are summarized in the following the systematic analysis of gene function has points: (1) definition of all the genes in the the added advantage that there are considerably pathway of interest; (2) perturbation of each fewer metabolites than genes or gene products. pathway component through a series of genetic Nevertheless, unlike proteins, metabolites are or environmental manipulations and quantifica- not directly related to genes. tion of global cellular response; (3) integration of Metabolomic studies have emerged in an the observed mRNA and protein responses with attempt to assign functions to genes on the basis the current, pathway-specific model; and (4) of metabolic analyses. The primary aim is to formulation of new hypotheses to explain obser- discover biochemical reactions catalyzed by vations not predicted by the model. Although enzymes encoded by genes of unknown func- metabolomics is a relatively new field, a study tion (Martzen et al., 1999). The difficulty with by Eglinton et al. (2002), using metabolomic such an approach is that it assigns mechanisms analysis of mutant laboratory strains, showed rather than biological functions. how genetic modification affects the production An alternative approach would be to study of several secondary metabolites of fermenta- changes in the metabolome induced by the deletion including acids (such as acetic acid), esters, tion or overexpression of a specific gene and to aldehydes, and higher alcohols. Many of these then assign functions by comparing the changes metabolites make an important contribution to induced with those observed in similar manipu- the flavor and aroma of the wine. A recent study lations of known genes. Such an approach, by Rossouw et al. (2008) investigating the rela- referred to as metabolic footprinting, was used tionship between the transcriptomes of five by Raamsdonk et al. (2001) in S. cerevisiae. wine strains and the aroma profile produced Measuring concentrations of specific metabo- during fermentation found that the expression lites in a cell, however, is a very costly process. levels of five genes were related to differences The approach used by Raamsdonk et al. was in aroma. They then constructed wine strains extended in a subsequent study by the same overexpressing these genes and found that the REFERENCES 163 changes in the exo-metabolome corresponded Alexandre, H., Ansanay-Galeote, V., Dequin, S., & to the predicted changes. Blondin, B. (2001). Global gene expression during short- term ethanol stress in Saccharomyces cerevisiae. FEBS Lett., 498,98e103. 7. FUTURE DIRECTIONS Allen, J., Davey, H. M., Broadhurst, D., Heald, J. K., Rowland, J. J., Oliver, S. G., et al. (2003). High-throughput classification of yeast mutants for functional genomics The use of genomic and proteomic methods to using metabolic footprinting. Nat. Biotech., 21, 692e696. study wine yeasts is still in its infancy. Although Backhus, L. E., DeRisi, J., Brown, P. O., & Bisson, L. F. (2001). the results achieved so far have begun to provide Functional genomic analysis of a commercial wine yeast molecular explanations to problems related to strain of Saccharomyces cerevisiae under differing nitrogen conditions. FESM Yeast Res., 1,1e15. wine yeast physiology, we are still far from the Bakalinsky, A. T., & Snow, R. (1990). The chromosomal level of detail available for laboratory strains. constitution of wine strains of Saccharomyces cerevisiae. It is important to discover what makes Yeast, 6, 367e382. wine strains capable of must fermentation in Bammert, G. F., & Fostel, J. M. (2000). Genome-wide circumstances in which the much-better-known expression patterns in Saccharomyces cerevisiae: Compar- ison of drug treatments and genetic alterations affecting laboratory strains are not. Laboratory strains of biosynthesis of ergosterol. Antimicrob. Agents Chemother, S. cerevisiae are indeed nothing more than simpli- 44, 1255e1265. fied genomic derivatives of natural strains. Deci- Bargues, M., Salom, D., Go´mez, A., Paricio, N., Pe´rez- phering the genome of wine strains is also Alonso, M., & Pe´rez-Ortıń, J. E. (1996). Sequencing interesting from a basic scientific perspective. analysis of a 4.1 kb subtelomeric region from yeast chromosome IV identifies , a new member of the Gaining a deeper understanding of the genome, HXT15 hexose transporter family. Yeast, 12, 1005e1011. transcriptome, and proteome of wine yeasts and Barre, P., Ve´zinhet, F., Dequin, S., & Blondin, B. (1993). integrating this information into mathematical Genetic improvement of wine yeast. In G. H. Fleet (Ed.), models capable of predicting physiological Wine Microbiology and Biotechnology (pp. 421e447). UK: changes will allow carefully constructed Harwood Academic, Reading. Belloch, C., Pe´rez-Torrado, R., Gonza´lez, S. S., Pe´rez- improvements in the characteristics of these Ortıń, J. E., Garcıá-Martıńez, J., Querol, A., et al. (2009). strains and the biotechnological processes in Chimeric genomes of natural hybrids of Saccharomyces which they participate. cerevisiae and Saccharomyces kudriavzevii. Appl. Environ. Although we have been making wine for Microbiol., 75, 2534e2544. over 7000 years, we only very recently discov- Beltrań, G., Novo, M., Leberre, V., Sokol, S., Labourdette, D., Guillamoń, J. M., et al. (2006). Integration of tran- ered, thanks to Louis Pasteur, that S. cerevisiae scriptomic and metabolic analyses for understanding the was the main driving force behind the process. global responses to low temperature winemaking Since then, this yeast has been the focus of fermentations. FESM Yeast Res., 6, 1167e1183. much basic and applied research. Nowadays, Bidenne, C., Blondin, B., Dequin, S., & Ve´zinhet, F. (1992). the in-depth information that large-scale studies Analysis of the chromosomal DNA polymorphism of wine strains of. Saccharomyces cerevisiae. Curr. Genet., 22, can provide on the full complement of macro- 1e7. molecules found in this microorganism will Bisson, L. F., Karpel, J. E., Ramakrishnan, V., & Joseph, L. help us to fully understand its physiology and (2007). Functional genomics of wine yeast Saccharomyces elucidate the manner in which it makes wine. cerevisiae. Adv. Food Nutr. Res., 53,65e121. Borneman, A. R., Forgan, A. H., Pretorius, I. S., & Chambers, P. J. (2008). Comparative genome analysis of References a Saccharomyces cerevisiae wine strain. FESM Yeast Res., 8, 1185e1195. Aguilera, A., Cha´vez, S., & Malagoń, F. (2000). Mitotic Borneman, A. R., Chambers, P. J., & Pretorius, I. S. (2007). recombination in yeast: Elements controlling its inci- Yeast systems biology: Modelling the winemaker’s art. dence. Yeast, 16, 731e754. Trends Biotechnol., 25, 349e355. 164 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

Brazma, A., Jonassen, I., Vilo, J., & Ukkonen, E. (1998). industrial environments of Saccharomyces strains. Appl. Predicting gene regulatory elements in silico on Microbiol. Biotechnol., 49, 154e163. a genomic scale. Genome Res., 8, 1202e1215. Codoń, A. C., Gasent-Ramı´rez, J. M., & Benı´tez, T. (1995). Brazma, A., & Vilo, J. (2000). Gene expression data analysis. Factors which affect the frequency of sporulation and FEBS Lett., 480,17e24. tetrad formation in Saccharomyces cerevisiae baker’s Brejning, J., Jespersen, L., & Arneborg, N. (2003). Genome- yeasts. Appl. Environ. Microbiol., 61, 630e638. wide transcriptional changes during the lag phase of de Nobel, H., Lawrie, L., Brul, S., Klis, F., Davis, M., Saccharomyces cerevisiae. Arch. Microbiol., 179, 278e294. Alloush, H., et al. (2001). Parallel and comparative Bussemaker, H. J., Li, H., & Siggia, E. D. (2000). Building analysis of the proteome and transcriptome of sorbic a dictionary for genomes: Identification of presumptive acid-stressed Saccharomyces cerevisiae. Yeast, 18, regulatory sites by statistical analysis. Proc. Natl. Acad. 1328e1413. Sci. USA, 97, 10 096e10 100. Delneri, D., Brancia, F. L., & Oliver, S. G. (2001). Towards Butler, G., Rasmussen, M. D., Lin, M. F., Santos, M. A., a truly integrative biology through the functional geno- Sakthikumar, S., Munro, C. A., et al. (2009). Evolution of mics of yeast. Curr. Opin. Biotech., 12,87e91. pathogenicity and sexual reproduction in eight Candida Delneri, D., Colson, I., Grammenoudi, S., Roberts, I. N., genomes. Nature, 459, 657e662. Louis, E. J., & Oliver, S. G. (2003). Engineering evolution Carlson, M., Osmond, B. C., & Botstein, D. (1989). SUC to study speciation in yeasts. Nature, 422,68e72. genes of Yeast: A dispersed gene family. Cold Spring DeRisi, J. L., Iyer, V. R., & Brown, P. O. (1997). Exploring the Harbor Symp. Quant. Biol., 45, 799e803. metabolic and genetic control of gene expression on Carmel-Harel, O., Stearman, R., Gasch, A. P., Botstein, D., a genomic scale. Science, 24, 680e686. Brown, P. O., & Storz, G. (2001). Role of thioredoxin reduc- Dujon, B. (1998). European Functional Analysis Network tase in the Yap1p-dependent response to oxidative stress (EUROFAN) and the functional analysis of the Saccha- in Saccharomyces cerevisiae. Mol. Microbiol., 39, 595e605. romyces cerevisiae genome. Electrophoresis, 19, 617e624. Carreto, L., Eiriz, M. F., Gomes, A. C., Pereira, P. M., Eglinton, J. M., Heinrich, A. J., Pollnitz, A. P., Langridge, P., Schuller, D., & Santos, M. A. (2008). Comparative Henschke, P. A., & de Barros Lopes, M. (2002). genomics of wild type yeast strains unveils important Decreasing acetic acid accumulation by a glycerol over- genome diversity. BMC Genomics, 9, 524. producing strain of Saccharomyces cerevisiae by deleting Carro, D., Garcıá-Martıńez, J., Pe´rez-Ortıń, J. E., & Pinã, B. the ALD6 aldehyde dehydrogenase gene. Yeast, 19, (2003). Structural characterization of chromosome I size 295e301. variants from a natural yeast strain. Yeast, 30, 171e183. Erasmus, D. J., van der Merwe, G. K., & van Vuuren, H. J. J. Carro, D., & Pinã, B. (2001). Genetic analysis of the karyo- (2003). Genome-wide expression analyses: Metabolic type instability in natural wine yeast strains. Yeast, 18, adaptation of Saccharomyces cerevisiae to high sugar 1457e1470. stress. FESM Yeast Res., 3, 375e399. Cavalieri, D., Townsend, J. P., & Hartl, D. L. (2000). Manifold Fares, M. A., & Wolfe, K. (2003). Accelerated evolution of anomalies in gene expression in a vineyard isolate of some gene copies after genome duplication in Saccharo- Saccharomyces cerevisiae revealed by DNA microarray myces. Yeast, 20, S277. analysis. Proc. Natl. Acad. Sci. USA, 97, 12369e12374. Fauchon, M., Lagniel, G., Aude, J. C., Lombardia, L., Cho, R. J., Fromont-Racine, M., Wodicka, L., Feierbach, B., Soularue, P., Petat, C., et al. (2002). Sulfur sparing in the Stearns, T., Legrain, P., et al. (1998). Parallel analysis of yeast proteome in response to sulfur demand. Mol. Cell, genetic selections using whole genome oligonucleotide 9, 713e723. arrays. Proc. Natl. Acad. Sci. USA, 95, 3752e3757. Ferea, T. L., Botstein, D., Brown, P. O., & Rosenzweig, R. F. Chow, T. H. C., Sollitti, P., & Marmur, J. (1989). Structure of (1999). Systematic changes in gene expression patterns the multigene family of MAL loci in Saccaromyces. Mol. following adaptive evolution in yeast. Proc. Natl. Acad. Gen. Genet., 217,60e69. Sci. USA, 96, 9721e9726. Chu, S., DeRisi, J., Eisen, M., Mulholland, J., Botstein, D., Fey, S. J., Nawrocki, A., Larsen, M. R., Gorg, A., Brown, P. O., et al. (1998). The transcriptional program of Roepstorff, P., Skews, G. N., et al. (1997). Proteome sporulation in budding yeast. Science, 282, 699e705. studies of Saccharomyces cerevisiae: A methodological Cliften, P., Sudarsanam, P., Desikan, A., Fulton, L., outline. Electrophoresis, 18, 1361e1372. Fulton, B., Majors, J., et al. (2003). Finding functional Fischer, G., James, S. A., Roberts, I. N., Oliver, S. G., & features in Saccharomyces genomes by phylogenetic Louis, E. J. (2000). Chromosomal evolution in Saccharo- footprinting. Science, 301,71e76. myces. Nature, 405, 451e454. Codoń, A. C., Benı´tez, T., & Korhola, M. (1998). Chromo- Fischer, G., Neuveglise, C., Durrens, P., Gaillardin, C., & somal polymorphism and adaptation to specific Dujon, B. (2001). Evolution of gene order in the REFERENCES 165

genomes of two related yeast species. Genome Res., 11, Hirasawa, T., Yoshikawa, K., Nakakura, Y., Nagahisa, K., 2009e2019. Furusawa, C., Katakura, Y., et al. (2007). Identification of Futcher, B., Latter, G. I., Monardo, P., McLaughin, C. S., & target genes conferring ethanol stress tolerance to Garrels, J. I. (1999). A sampling of the yeast proteome. Saccharomyces cerevisiae based on DNA microarray data Mol. Cell. Biol., 19, 7357e7368. analysis. J. Biotechnol., 131,34e44. Gasch, A. P., Spellman, P. T., Kao, C. M., Carmel-Harel, O., Holstege, F. C., Jennings, E. G., Wyrick, J. J., Lee, T. I., Eisen, M. B., Storz, G., et al. (2000). Genomic expression Hengartner, C. J., Green, M. R., et al. (1998). Dissecting programs in the response of yeast cells to environmental the regulatory circuitry of a eukaryotic genome. Cell, 95, changes. Mol. Biol. Cell., 11, 4241e4257. 717e728. Giaever, G., Chu, A. M., Ni, L., Connelly, C., Riles, L., Hughes, T. R., Marton, M. J., Jones, A. R., Roberts, C. J., Ve´ronneau, S., et al. (2002). Functional profiling of the Stoughton, R., Armour, C. D., et al. (2000a). Functional Saccharomyces cerevisiae genome. Nature, 418, 387e391. discovery via a compendium of expression profiles. Cell, Giaever, G., Shoemaker, D. D., Jones, T. W., Liang, H., 102, 109e126. Winzeler, E. A., Astromoff, A., et al. (1999). Genomic Hughes, T. R., Roberts, C. J., Dai, H., Jones, A. R., profiling of drug sensitivities via induced hap- Meyer, M. R., Slade, D., et al. (2000b). Widespread loinsufficiency. Nat. Genet., 21, 278e283. aneuploidy revealed by DNA microarray expression Godon, C., Lagniel, G., Lee, J., Buhler, J. M., Kieffer, S., profiling. Nat. Genet., 25, 333e337. Perrot, M., et al. (1998). The H2O2 stimulon in Saccharo- Hughes, T. R., & Shoemaker, D. D. (2001). DNA micro- myces cerevisiae. J. Biol. Chem., 273, 22 480e22 489. arrays for expression profiling. Curr. Opin. Chem. Biol., Goffeau, A., et al. (1997). The yeast genome directory. 5,21e25. Nature, 387(Suppl. 1), 1e105. Ideker, T., Thorsson, V., Ranish, J. A., Christmas, R., Gonza´lez, S. S., Barrio, E., Gafner, J., & Querol, A. (2006). Buhler, J., Eng, J. K., et al. (2001). Integrated genomic and Natural hybrids from Saccharomyces cerevisiae, Saccharo- proteomic analyses of a systematically perturbed meta- myces bayanus, and Saccharomyces kudriavzevii in wine bolic network. Science, 292, 929e934. fermentations. FESM Yeast Res., 6, 1221e1234. Infante, J. J., Dombek, K. M., Rebordinos, L., Cantoral, J. M., & Griffin, T. J., Gygi, S. P., Ideker, T., Rist, B., Eng, J., Hood, L., Young, E. T. (2003). Genome-wide amplifications caused et al. (2002). Complementary profiling of gene expres- by chromosomal rearrangements play a major role in the sion at the transcriptome and proteome levels in adaptive evolution of natural yeast. Genetics, 165, Saccharomyces cerevisiae. Mol. Cell Proteomics, 1, 323e333. 1745e1759. Gross, C., Kelleher, M., Iyer, V. R., Brown, P. O., & Ivorra, C., Pe´rez-Ortıń, J. E., & del Olmo, M. (1999). An Winge, D. R. (2000). Identification of the copper regulon inverse correlation between stress resistance and stuck in Saccharomyces cerevisiae by DNA microarrays. J. Biol. fermentations in wine yeasts. A molecular study. Bio- Chem., 275, 32 310e32 316. technol. Bioeng., 64, 698e708. Guijo, S., Mauricio, J. C., Salmon, J. M., & Ortega, J. M. (1997). Jelinsky, S. A., Estep, P., Church, G. M., & Samson, L. D. Determination of the relative ploidy in different Saccha- (2000). Regulatory networks revealed by transcriptional romyces cerevisiae strains used for fermentation and “flor” profiling of damaged Saccharomyces cerevisiae cells: Rpn4 film ageing of dry sherry-type wines. Yeast, 13, 101e117. links base excision repair with proteasomes. Mol. Cell. Hampson, S., Baldi, P., Kibler, D., & Sandmeyer, S. B. (2000). Biol., 20, 8157e8167. Analysis of yeast’s ORF upstream regions by parallel Jelinsky, S. A., & Samson, L. D. (1999). Global response of processing, microarrays, and computational methods. Saccharomyces cerevisiae to an alkylating agent. Proc. Natl. Proc. Int. Conf. Intell. Syst. Mol. Biol., 8, 190e201. Acad. Sci. USA, 96, 1486e1491. Hardwick, J. S., Kuruvilla, F. G., Tong, J. K., Shamji, A. F., & Jia, M. H., Larossa, R. A., Lee, J. M., Rafalski, A., Derose, E., Schreiber, S. L. (1999). Rapamycin-modulated transcrip- Gonye, G., et al. (2000). Global expression profiling tion defines the subset of nutrient-sensitive signaling of yeast treated with an inhibitor of amino acid pathways directly controlled by the Tor proteins. Proc. biosynthesis, sulfometuron methyl. Physiol. Genomics, 3, Natl. Acad. Sci. USA, 96, 14 866e14 870. 83e92. Hauser, N. C., Fellenberg, K., Gil, R., Bastuck, S., Jimeńez-Martı´, E., & del Olmo, M. (2008). Addition of Hoheisel, J. D., & Pe´rez-Ortıń, J. E. (2001). Whole ammonia or amino acids to a nitrogen-depleted medium genome analysis of a wine yeast strain. Comp. Funct. affects gene expression patterns in yeast cells during Genom., 2,69e79. alcoholic fermentation. FESM Yeast Res., 8, 245e256. Herrero, E., de la Torre, M. A., & Valentıń, E. (2003). Jordan, I. K., & McDonald, F. (1999). Tempo and mode of Ty Comparative genomics of yeast species: New insights evolution in Saccharomyces cerevisieae. Genetics, 151, into their biology. Int. Microbiol., 6, 183e190. 1341e1351. 166 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

Joubert, R., Brignon, P., Lehmann, C., Monribot, C., exploration of hemiascomycetous yeasts: 19. Ascomy- Gendre, F., & Boucherie, H. (2000). Two-dimensional gel cetes-specific genes. FEBS Lett., 487,113e121. analysis of the proteome of lager brewing yeasts. Yeast, Marks, V. D., Ho Sui, S. J., Erasmus, D., van der 16,511e522. Merwe, G. K., Brumm, J., Wasserman, W. W., et al. Joubert, R., Strub, J. M., Zugmeyer, S., Kobi, D., Carte, N., (2008). Dynamics of the yeast transcriptome during wine van Dorsselaer, A., et al. (2001). Identification by mass fermentation reveals a novel fermentation stress spectrometry of two-dimensional gel electrophoresis- response. FESM Yeast Res., 8,35e52. separated proteins extracted from lager brewing yeast. Marks, V. M., van der Merwe, G. K., & van Vuuren, H. J. J. Electrophoresis, 22, 2269e2282. (2003). Transcriptional profiling of wine yeast in fer- Kellis, M., Patterson, N., Endrizzi, M., Birren, B., & menting grape juice: Regulatory effect of diammonium Lander, E. S. (2003). Sequencing and comparison of yeast phosphate. FESM Yeast Res., 3, 269e287. species to identify genes and regulatory elements. Martıńez-Rodrı´guez, A. J., Carrascosa, A. V., & Polo, M. C. Nature, 423, 241e254. (2001b). Release of nitrogen compounds to the extracel- Kielland-Brandt, M. C., Nilsson-Tillgren, T., Gjermnsen, C., lular medium by three strains of Saccharomyces cer- Holmberg, S., & Pedersen, M. B. (1995). Genetics of evisiae during induced autolysis in a model wine brewing yeasts. In A. Wheals, A. Rose & J. Harrison system. Int. J. Food Microbiol., 68, 155e160. (Eds.), The yeasts (pp. 223e254). London, UK: Academic Martıńez-Rodrı´guez, A. J., Polo, M. C., & Carrascosa, A. V. Press. (2001a). Structural and ultrastructural changes in Kolkman, A., Daran-Lapujade, P., Fullaondo, A., yeast cells during autolysis in a model wine system Olsthoorn, M. M., Pronk, J. T., Slijper, M., et al. (2006). and in sparkling wines. Trends Biochem. Sci., 15, Proteome analysis of yeast response to various nutrient 305e309. limitations. Mol. Syst. Biol., 2, 2006, 0026. Martini, A. (1993). The origin and domestication of the Kuhn, K. M., DeRisi, J. L., Brown, P. O., & Sarnow, P. (2001). wine yeast Saccharomyces cerevisiae. J. Wine Res., 4, Global and specific translational regulation in the 165e176. genomic response of Saccharomyces cerevisiae to a rapid Martzen, M. R., McCraith, S. M., Spinelli, S. L., Torres, F. M., transfer from a fermentable to a nonfermentable carbon Fields, S., Grayhack, E. J., et al. (1999). A biochemical source. Mol. Cell. Biol., 21, 916e927. genomics approach for identifying genes by the activity Kunkee, R. E., & Bisson, L. F. (1993). Wine-making yeasts. In of their products. Science, 286, 1153e1155. A. H. Rose & J. S. Harrison (Eds.), The yeasts: Yeast Mauricio, J. C., Valero, E., Millan, C., & Ortega, J. M. (2001). Technology (pp. 69e125). London, UK: Academic Press. Changes in nitrogen compounds in must and wine Lagunas, R. (1986). Misconceptions about the energy during fermentation and biological aging by flor yeast. metabolism of. Saccharomyces cerevisiae. Yeast, 2, 221e228. J. Agric. Food Chem., 49, 3310e3315. Langkjaer, R. B., Cliften, P. F., Johnston, M., & Piskur, J. Mendes-Ferreira, A., del Olmo, M., Garcıá-Martıńez, J., (2003). Yeast genome duplication was followed by Jimeńez-Martı´, E., Leaõ, C., Mendes-Faia, A., et al. asynchronous differentiation of duplicated genes. (2007b). Signature genes for predicting nitrogen defi- Nature, 421, 848e852. ciency during alcoholic fermentation. Appl. Environm. Link, A. J., Eng, J., Schieltz, D. M., Carmack, E., Mize, G. J., Microbiol., 73, 5363e5369. Morris, D. R., et al. (1999). Direct analysis of protein Mendes-Ferreira, A., del Olmo, M., Garcıá-Martıńez, J., complexes using mass spectrometry. Nat. Biotech., 17, Jimeńez-Martı´, E., Mendes-Faia, A., Pe´rez-Ortıń, J. E., 676e682. et al. (2007a). Transcriptional response of Saccharomyces Liti, G., Carter, D. M., Moses, A. M., Warringer, J., Parts, L., cerevisiae to different nitrogen concentrations during James, S. A., et al. (2009). Population genomics of alcoholic fermentation. Appl. Environm. Microbiol., 73, domestic and wild yeasts. Nature, 458, 337e341. 3049e3060. Llorente, B., Durrens, P., Malpertuy, A., Aigle, M., Miklos, G. L. G., & Rubin, G. M. (1996). The role of the Artiguenave, F., Blandin, G., et al. (2000). Genomic genome project in determining gene function: Insight for exploration of the hemiascomycetous yeasts: 20. Evolu- model organisms. Cell, 86, 521e529. tion of gene redundancy compared to Saccharomyces Mortimer, R. K. (2000). Evolution and variation of the yeast cerevisiae. FEBS Lett., 487, 122e133. (Saccharomyces) genome. Genome Res., 10, 403e409. Longo, E., & Ve´zinhet, F.(1993). Chromosomal rearrangements Mortimer, R. K., & Johnston, J. R. (1986). Genealogy of during vegetative growth of a wild strain of Saccharomyces principal strains of the yeast genetic stock center. cerevisiae. Appl. Environ. Microbiol., 59,322e326. Genetics, 113,35e43. Malpertuy, A., Tekaia, F., Casare´gola, S., Aigle, M., Mortimer, R. K., & Polsinelli, M. (1999). On the origins of Artiguenave, F., Blandin, G., et al. (2000). Genomic wine yeast. Res. Microbiol., 150, 199e204. REFERENCES 167

Mortimer, R. K., Romano, P., Suzzi, G., & Polsinelli, M. Pe´rez-Ortıń, J. E., Querol, A., Puig, S., & Barrio, E. (2002a). (1994). Genome renewal: A new phenomenon revealed Molecular characterization of a chromosomal rear- from a genetic study of 43 strains of Saccharomyces cer- rangement involved in the adaptive evolution of yeast evisiae derived from natural fermentation of grape strains. Genome Res., 12, 1533e1539. musts. Yeast, 10, 1543e1552. Pizarro, F., Vargas, F.A., & Agosin, E. (2007). A systems biology Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., perspective of wine fermentations. Yeast, 24,977e991. Gerstein, M., et al. (2008). The transcriptional landscape Pretorius, I. S. (2000). Tailoring wine yeast for the new of the yeast genome defined by RNA sequencing. millennium: Novel approaches to the ancient art of Science, 320, 1344e1349. winemaking. Yeast, 16, 675e729. Nakao, Y., Kanamori, T., Itoh, T., Kodama, Y., Rainieri, S., Primig, M., Williams, R. M., Winzeler, E. A., Tevzadze, G. G., Nakamura, N., et al. (2009). Genome sequence of the Conway, A. R., Hwang, S. Y., et al. (2000). The core lager brewing yeast, an interspecies hybrid. DNA Res., meiotic transcriptome in budding yeasts. Nat. Genet., 26, 16,115e129. 415e423. Naumov, G. I. (1996). Genetic identification of biological Puig, S., Querol, A., Barrio, E., & Pe´rez-Ortıń, J. E. (2000). species in the Saccharomyces sensu stricto complex. J. Ind. Mitotic recombination and genetic changes in Saccharo- Microbiol., 17, 295e302. myces cerevisiae during wine fermentation. Appl. Environ. Naumov, G. I., Naumova, E. S., Masneuf, I., Aigle, M., Microbiol., 66, 2057e2061. Kondratieva, V. I., & Dubourdieu, D. (2000). Natural Puig, S., Ramoń, D., & Pe´rez-Ortıń, J. E. (1998). An opti- polyploidization of some cultured yeast Saccharomyces mized method to obtain stable food-safe recombinant sensu stricto: Auto- and allotetraploidy. Syst. Appl. wine yeast strains. J. Agr. Food Chem., 46, 1689e1693. Microbiol., 23, 429e442. Querol, A., Barrio, E., Huerta, T., & Ramoń, D. (1992). Naumova, E. S., Turakainen, H., Naumov, G. I., & Molecular monitoring of wine fermentations conducted Korhola, M. (1997). Superfamily of alpha-galactosidase by active dry yeast strains. Appl. Environ. Microbiol., 58, MEL genes of the Saccharomyces sensu stricto species 2948e2953. complex. Mol. Gen. Genet., 253, 111e117. Querol, A., Belloch, C., Fernańdez-Espinar, M. T., & Ness, F., & Aigle, M. (1995). RTM1 a member of a Barrio, E. (2003). Molecular evolution in yeast of new family of telomeric genes in yeast. Genetics, 113, biotechnological interest. Int. Microbiol., 6, 201e205. 35e43. Raamsdonk, L. M., Teusink, B., Broadhurst, D., Zhang, N., Novo, M., Beltran, G., Rozes, N., Guillamon, J. M., Sokol, S., Hayes, A., Walsh, M. C., et al. (2001). A functional Leberre, V., et al. (2007). Early transcriptional response of genomics strategy that uses metabolome data to reveal wine yeast after rehydration: Osmotic shock and meta- the phenotype of silent mutations. Nat. Biotech., 19, bolic activation. FESM Yeast Res., 7, 304e316. 45e50. Novo, M., Bigey, F., Beyne, E., Galeote, V., Gavory, F., Rachidi, N., Barre, P., & Blondin, B. (1999). Multiple Mallet, S., et al. (2009). Eukaryote-to-eukaryote gene Ty-mediated chromosomal translocations lead to transfer events revealed by the genome sequence of the karyotype changes in a wine strain of Saccharomyces wine yeast Saccharomyces cerevisiae EC1118. Proc. Natl. cerevisiae. Mol. Gen. Genet., 261, 841e850. Acad. Sci. USA., 106, 16 333e16 338. Rainieri, S., Kodama, Y., Kaneko, Y., Mikata, K., Nakao, Y., & Ogawa, N., DeRisi, J., & Brown, P. O. (2000). New compo- Ashikari, T. (2006). Pure and mixed genetic lines of nents of a system for phosphate accumulation and pol- Saccharomyces bayanus and Saccharomyces pastorianus and yphosphate metabolism in Saccharomyces cerevisiae their contribution to the lager brewing strain genome. revealed by genomic expression analysis. Mol. Biol. Cell., Appl. Environ. Microbiol., 72, 3968e3974. 11, 4309e4321. Rogowska-Wrzesinska, A., Larsen, P. M., Blomberg, A., Patton, E. E., Peyraud, C., Rouillon, A., Surdin-Kerjan, Y., Go¨rg, A., Roepstorff, P., Norberk, J., et al. (2001). Tyers, M., & Thomas, D. (2000). SCF(Met30)-mediated Comparison of the proteomes of three yeast wild type control of the transcriptional activator Met4 is required strains: CEN.PK2, FY1679 and W303. Comp. Funct. for the G(1)-S transition. EMBO J., 19, 1613e1624. Genom, 2, 207e225. Pennington, S. R., Wilking, M. R., Hochstrasser, D. F., & Rossignol, T., Dulau, L., Julien, A., & Blondin, B. (2003). Dunn, M. J. (1997). Proteome analysis: From protein Genome-wide monitoring of wine yeast gene expression characterization to biological function. Trends Cell Biol., 7, during alcoholic fermentation. Yeast, 20, 1369e1385. 168e173. Rossignol, T., Kobi, D., Jacquet-Gutfreund, L., & Blondin, B. Pe´rez-Ortıń, J. E., Garcıá-Martıńez, J., & Alberola, T. (2002b). (2009). The proteome of a wine yeast strain during DNA chips for yeast biotechnology. The case of wine fermentation, correlation with the transcriptome. J. Appl. yeasts. J. Biotechnol., 98, 227e241. Microbiol., 107,47e55. 168 6. GENOMIC AND PROTEOMIC ANALYSIS OF WINE YEASTS

Rossignol, T., Postaire, O., Storaı¨, J., & Blondin, B. (2006). Teunissen, A. W., & Steensma, H. Y. (1995). The domi- Analysis of the genomic response of a wine yeast to nant flocculation genes of Saccharomyces cerevisiae rehydration and inoculation. Appl. Microbiol. Biotechnol., constitute a new subtelomeric gene family. Yeast, 11, 71, 699e712. 1001e1013. Rossouw, D., Naes, T., & Bauer, F. F. (2008). Linking gene Trabalzini, L., Paffetti, A., Scaloni, A., Talamo, F., Ferro, E., regulation and the exo-metabolome: A comparative Coratza, G., et al. (2003). Proteomic response to transcriptomics approach to identify genes that impact physiological fermentation stresses in a wild-type on the production of volatile aroma compounds in yeast. wine strain of Saccharomyces cerevisiae. Biochem. J., 370, BMC Genomics, 9, 530. 35e46. Roth, F. P., Hughes, J. D., Estep, P. W., & Church, G. M. Velculescu, V. E., Vogelstein, B., & Kinzler, K. W. (2000). (1998). Finding DNA regulatory motifs within unaligned Uncharted transcriptomes with SAGE. Trends Genet., 16, non-coding sequences clustered by whole-genome 423e425. mRNA quantitation. Nat. Biotech., 16, 939e945. Velculescu, V. E., Zhang, L., Zhou, W., Vogelstein, J., Royce, T. E., Rozowsky, J. S., Bertone, P., Samanta, M., Basrai, M. A., Bassett, D. E., et al. (1997). Characteriza- Stolc, V., Weissman, S., et al. (2005). Issues in the analysis tion of the yeast transcriptome. Cell, 88, 243e251. of oligonucleotide tiling microarrays for transcript Vido, K., Spector, D., Lagniel, G., Lopez, S., Toledano, M. B., mapping. Trends Genet., 21, 466e475. & Labarre, J. (2001). A proteome analysis of the Salmon, J.-M. (1997). Enological fermentation kinetics of an cadmium response in Saccharomyces cerevisiae. J. Biol. isogenic ploidy series derived form an industrial Chem., 276, 8469e8474. Saccharomyces cerevisiae strain. J. Ferment. Bioeng, 83, Washburn, M. P., Wolters, D., & Yates, J. R. (2001). Large- 253e260. scale analysis of the yeast proteome by multidimen- Schacherer, J., Shapiro, J. A., Ruderfer, D. M., & Kruglyak, L. sional protein identification technology. Nat. Biotech., 19, (2009). Comprehensive polymorphism survey elucidates 242e247. population structure of Saccharomyces cerevisiae. Nature, Winzeler, E. A., Richards, D. R., Conway, A. R., 458, 342e345. Goldstein, A. L., Kalman, S., McCullough, M. J., et al. Schena, M., Shalon, D., Davis, R. W., & Brown, P. O. (1995). (1998). Direct allelic variation scanning of the yeast Quantitative monitoring of gene expression patterns with genome. Science, 281, 1194e1197. a complementary DNA microarray. Science, 270, 467e470. Winzeler, E. A., Shoemaker, D. D., Astromoff, A., Liang, H., Snow, R. (1983). Genetic improvement of wine yeast. In Anderson, K., Andre, B., et al. (1999). Functional char- J. F. T. Spencer, D. M. Spencer & A. R. V. Smith (Eds.), acterization of the S. cerevisiae genome by gene deletion Yeast genetics e Fundamental and applied aspects (pp. and parallel analysis. Science, 285, 901e906. 439e459). New York, NY: Springer-Verlag. Wodicka, L., Dong, H., Mittmann, M., Ho, M. H., & Solieri, L., Antu´ nez, O., Pe´rez-Ortıń, J. E., Barrio, E., & Lockhart, D. J. (1997). Genome-wide expression moni- Giudici, P. (2008). Mitochondrial inheritance and toring in Saccharomyces cerevisiae. Nat. Biotechnol., 15, fermentative: Oxidative balance in hybrids between 1359e1367. Saccharomyces cerevisiae and Saccharomyces uvarum. Yeast, Wolfe, K. H., & Shields, D. C. (1997). Molecular evidence for 25, 485e500. an ancient duplication of the entire yeast genome. Souciet, J., Aigle, M., Artiguenave, F., Blandin, G., Bolotin- Nature, 387, 708e713. Fukuhara, M., Bon, E., et al. (2000). Genomic exploration Zuzua´rregui, A., & del Olmo, M. L. (2004). Expression of of the hemiascomycetous yeasts: 1. A set of yeast species stress response genes in wine strains with different for molecular evolution studies. FEBS Lett., 487,3e12. fermentative behavior. FESM Yeast Res., 4, 699e710. ter Linde, J. J., Liangm, H., Davism, R. W., Steensma, H. Y., Zuzua´rregui, A., Monteoliva, L., Gil, C., & del Olmo, M. van Dijken, J. P., & Pronk, J. T. (1999). Genome-wide (2006). Transcriptomic and proteomic approach for transcriptional analysis of aerobic and anaerobic che- understanding the molecular basis of adaptation of mostat cultures of Saccharomyces cerevisiae. J. Bacteriol., Saccharomyces cerevisiae to wine fermentation. Appl. 181, 7409e7413. Environ. Microbiol., 72, 836e847. CHAPTER 7

Improvement of Wine Yeasts by Genetic Engineering Daniel Ramoń 1, Ramoń Gonza´lez 2 1 Bio´polis SL, Valencia, Spain and 2 Instituto de Ciencias de la Vid y del Vino (CSIC-UR-CAR), Logronõ, Spain

OUTLINE

1. Wine, Biotechnology, and Genetics 169 4.2. Improvements Affecting the 1.1. Wine and Classical Genetics 169 Physicochemical Properties of Wine 180 1.2. Wine and Genetic Engineering 171 4.3. Improvement of Organoleptic, Nutritional, and Safety-related 2. Systems for the Genetic Transformation Properties 181 of Wine Yeasts 172 2.1. Transformation Procedures 172 5. Legislation Affecting the Use of Genetic 2.2. Selection Markers 173 Engineering in the Wine Industry 183 5.1. European Legislation 183 3. Regulation of Gene Expression: 5.2. Labeling 184 Promoters of Interest in Biotechnology 177 5.3. The Situation in Other Countries 184 4. Transgenic Wine Yeasts 177 6. The Future 184 4.1. Improvements Affecting the Industrial Winemaking Process 179

1. WINE, BIOTECHNOLOGY, AND GENETICS produce a commercial product. Fermenting a microscopic fungus such as Penicillium chryso- 1.1. Wine and Classical Genetics genum in order to produce cephalosporin or penicillin, for instance, is a form of biotech- To a scientist, biotechnology is simply the use nology, since the metabolites synthesized by of a living organism or any of its parts to a living organism are subsequently sold as

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10007-4 169 Copyright Ó 2011 Elsevier Inc. All rights reserved. 170 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING pharmaceutical products. When the living domestication events are unlikely to have been organism is used to produce a food or beverage, particularly complicated, since the evolutionary the process is referred to as food biotechnology. ancestor of this plant displayed a series of char- Since at least two different living organisms are acteristicsdsuch as the natural ability to climb, required to make wine (the vines that produce minimal requirements for water and minerals, the grapes and the yeasts that are responsible and a high propagative capacitydthat facili- for fermentation of the grape must), there is tated cultivation without significant effort. more than one justification for classifying wine However, as some experts have suggested, the as a product of food biotechnology. This is the most important adaptations were undoubtedly scientist’s view, however, and it is quite the switch to functional bisexuality and the different from that held by the consumer, partic- increase in the size of the fruit (Carbonneau, ularly in the European Union (EU). In the minds 1983; Vivier & Pretorius, 2002). All of these of most consumers, food biotechnology is the phenotypic changes occurred when growers use of genetics to produce foods and beverages, empirically employed natural selection of spon- or, put more crudely, it is about putting genes in taneous mutations; in other words, they were your soup. generated by the application of a genetic tech- Even if we apply this incorrect definition of nique. Improvement programs have continued biotechnology, however, such activities have to be implemented through the use of deliberate been linked with wine production for thousands cross-pollination in an effort to achieve greater of years, starting with the application of genetics resistance to pests and environmental stresses. to the cultivation of vines. The first archeolog- This genetic history and the effect of human ical evidence of vine cultivation dates back migration on the spread of vine cultivation to 7000 BC in Mesopotamia. From this point have together resulted in almost nine million onwards, multiple references can be found to hectares currently planted with vines world- the production of wine in the Egyptian and wide and more than 24 000 different extant Phoenician cultures, and, later, of course, to cultivars, of which 5000 belong to the species the spectacular winemaking practices of the V. vinifera. Greek and Roman civilizations. Since then, Genetics has also been used empirically for wine has been produced in many regions of many years to improve wine yeast, a popular the world using different vine cultivars, all but incorrect term used to refer to the microbial belonging to the genus Vitis. These plants are species Saccharomyces cerevisiae. There are classified into two subgeneradEuvitis and currently hundreds of strains of the species Muscadiniadthat each have specific morpholog- that over centuries have adapted through ical characteristics as a result of their genomic a process of mutation and selection to the differences. Most of the species currently used different ecological niches of each must and in viticulture belong to the subgenus Euvitis, winegrowing region. Whereas laboratory specifically Vitis vinifera, which is the most strains of S. cerevisiae have 16 chromosomes widely cultivated species (Antcliff, 1992). The and a genome containing around 12 megabases domestication of this plant took place around of DNA, industrial wine strains tend to be 5000 years ago in the region now occupied by diploid, aneuploid, or occasionally polyploid parts of Azerbaijan, Georgia, the north of Iraq, (Snow, 1983). As a result, the genetic origin of and the northeast of Turkey, although some existing wine strains has generated much authors have suggested that an independent debate among research groups. While some domestication event occurred in Spain suggest that S. cerevisiae is a natural organism (Nu´ n˜ez & Walker, 1989; Stevenson, 1985). These present on fruits (Mortimer & Polsinelli, WINE, BIOTECHNOLOGY, AND GENETICS 171

1999), there is increasing evidence that many Genetic engineering relies on the ability to strains are the result of hybridization between isolate genes from a genome and to then intro- natural strains of Saccharomyces species that duce them into cells through the use of genetic have been spontaneously selected in winemak- transformation techniques. The S. cerevisiae ing environments (Querol & Bond, 2009). What genome has been completely sequenced thanks seems clear is that the genome of existing wine to the efforts of an international, publicly funded yeasts has arisen as a result of powerful selec- research project (Goffeau et al., 1997). The tion pressure over millions of generations primary structure has been determined for the (Querol et al., 2003). In addition, in the last first 6000 genes that make up the genome and 20 years efforts have been made to modify the projects have been initiated to determine their genome of wine strains by mutagenesis and function. As a result, the molecular make-up of selection, hybridization, cytoduction, and S. cerevisiae is better understood than that of any protoplast fusion. In the vast majority of cases, other eukaryotic organism. In recent years, these approaches have been unsatisfactory genomes have been decoded not only for the prin- from an industrial point of view, although cipal laboratory strain but also for strains of they have increased our understanding of the interest in industry and medicine, or for research genome of these yeast strains (Barre et al., into the mechanisms of evolution. The genomes 1993; Gonza´lez et al., 2003; Gonza´lez-Ramos of two industrial wine yeasts have been fully et al., 2009; Hammond, 1996; Pretorius, 2000; sequenced and annotated (Borneman et al., Pretorius & van der Westhuizen, 1991; Quiro´s 2008; Novo et al., 2009). Others have been et al., 2010; Rainieri & Pretorius, 2000). sequenced with a lower coverage for the purpose of evolutionary studies (Liti et al., 2009). In addi- 1.2. Wine and Genetic Engineering tion to their relevance for the understanding of how strains have adapted to winemaking envi- Clearly, then, both vines and wine yeasts ronments during evolution, these sequence have been subject to genetic improvement. data have enormous potential for use in From a scientific point of view, we can say that biotechnology. no vine cultivars or strains of wine yeast are It can be reasoned that any gene introduced free of genetic modification. Nevertheless, since into a yeast that is then inoculated in a winery these changes have been introduced through the fermenter, if expressed during fermentation, use of classical genetics, they appear not to be will lead to accumulation of the protein that it linked with biotechnology in the minds of encodes as vinification advances and will there- consumers. For some years now, genetic engi- fore introduce the technological activity of neering techniques have been available to allow interest. Extensive biochemical and physiolog- specific genes to be isolated, modified in the ical data have been accumulated over many laboratory, and reintroduced into the original years in studies of the growth of laboratory organism or a different one to produce so-called strains of S. cerevisiae in defined media and under transgenic or genetically modified organisms controlled laboratory conditions. Furthermore, (GMOs). It is these targeted molecular changes many of the genes associated with the metabolic that consumers think of as real biotechnology. generation of physicochemical or organoleptic Enology and viticulture have not been exempt properties that are relevant to winemaking from these developments, and genetic engi- have been cloned and sequenced. Since effective neering has begun to be applied to both vines methods are available for the transformation of and wine yeasts, although to varying extents S. cerevisiae, including most of the wine strains and with differing outcomes. studied to date, it has been possible to use all of 172 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

Gene of interest (a) 2. SYSTEMS FOR THE GENETIC TRANSFORMATION OF WINE YEASTS (b) 2.1. Transformation Procedures Unlike other microorganisms such as (c) Bacillus subtilis and Haemophilus influenzae, S. cerevisiae does not undergo natural genetic transformation. In order for DNA to be introduced into Phenotype of interest the cells of this organism, reach the nucleus, and become stably expressed generation after FIGURE 7.1 Schematic diagram showing the approach generation, cells must be made competent used to obtain improved industrial wine yeasts through for the entry of exogenous DNA. These arti- genetic engineering. Firstly, biochemical information is ficial genetic transformation techniques allow required and cloned genes must be available for the gene products implicated in the process of interest (a). Secondly, an millions of transformants to be obtained with effective genetic transformation system is required in order to just a microgram of exogenous DNA in labora- introduce the gene into the chosen wine yeast (b). Finally, tory strains of S. cerevisiae and are now regulatory sequences must be introduced to allow expression employed routinely by thousands of scientists of the cloned gene to be activated at a specific point (c). around the world. Four different techniques are commonly this information to investigate the generation of used. The first is based on the use of protoplasts improved wine yeasts through the use of genetic or spheroplasts, which are obtained by treating engineering (Cebollero et al., 2007; Dequin, 2001; the yeast cells with a mixture of enzymes to Pretorius, 2000; Pretorius & Bauer, 2002; Querol & digest their cell wall and allow exogenous Ramoń, 1996). This process has also been reliant DNA to enter (Hinnen et al., 1978). Since one on the use of molecular techniques to confirm of the main functions of the cell wall is to main- the establishment and eventual dominance of tain osmotic balance, protoplasts must be main- inoculated wine yeasts during fermentation tained in a medium that is equilibrated with the (Querol et al., 1992). intracellular osmotic pressure. When the proto- Genetic and metabolic engineering strategies plasts are placed in contact with a suspension for the improvement of wine yeasts require of exogenous DNA, treatment with polyeth- three tools (see Figure 7.1): ylene glycol (PEG) will induce DNA entry into the cell. After a period of incubation, the proto- Cloned genes with known biochemical and 1) plasts are transferred to fresh medium contain- genetic links to a given enological problem ing an osmotic stabilizer and they regenerate An effective genetic transformation system 2) the cell wall to yield a normal cell containing that will allow the genes to be introduced the exogenous DNA. The second method also into an industrial wine yeast involves creation of artificial permeability, Regulatory sequences known as promoters 3) which in this case is achieved by treating intact that allow expression of the genes of interest yeast cells and DNA with high concentrations of at the appropriate point during vinification lithium salts, normally lithium acetate, and PEG In the following sections, we will address (Gietz et al., 1992). The third method involves each of these elements. the use of a device known as an electroporator, SYSTEMS FOR THE GENETIC TRANSFORMATION OF WINE YEASTS 173 which subjects intact cells or protoplasts to TABLE 7.1 Transformation Methods Used in the short, high-voltage electrical discharges Generation of Transgenic Wine Yeasts (Delorme, 1989). This treatment opens small pores in the surface of the cell or protoplast Transformation method Reference through which exogenous DNA can penetrate and reach the interior of the cell. Finally, Electroporation Salek et al. (1990) Pe´rez-Gonza´lez et al. (1993) although its use is very limited, the biolistic van Rensburg et al. (2005) technique used for transformation of plant cells Coulon et al. (2006) has been employed by some authors for yeast Husnik et al. (2006) transformation (Armaleo et al., 1990). Here, Protoplasts Lee and Hassan (1988) yeast cultures are bombarded with tungsten Pe´rez-Gonza´lez et al. (1993) microparticles coated with exogenous DNA in Lithium salts Lee and Hassan (1988) an attempt to introduce the DNA into the cyto- Boone et al. (1990) plasm and nucleus of the cell. In the vast Petering et al. (1991) majority of cases, transformation of industrial Laing and Pretorius (1993) wine yeasts has been carried out using the Pe´rez-Gonza´lez et al. (1993) lithium salts protocol, although protoplasts Dequin and Barre (1994) van Rensburg et al. (1994) and electroporation have also been used (see Gonza´lez-Candelas et al. (1995) Table 7.1). Ansanay et al. (1996) Sańchez-Torres et al. (1996) Michnick et al. (1997) 2.2. Selection Markers Puig et al. (1998) Sańchez-Torres et al. (1998) To achieve transformation, the gene of Dequin et al. (1999) interest is usually inserted into a plasmid to Ganga et al. (1999) form a transformation vector. In order to select Remize et al. (1999) cells that have acquired the exogenous DNA, Gonza´lez-Candelas et al. (2000) genes that function as selection markers, usually Vilanova et al. (2000) Pe´rez-Torrado et al. (2002) by conferring resistance to a drug or by comple- Manzanares et al. (2003) menting auxotrophies, are introduced into the Walker et al. (2003) vector. Since the marker gene is included in Fernańdez-Gonza´lez et al. (2005) the same plasmid, detection indicates the pres- Cambon et al. (2006) ence of the gene of interest. In the case of van Rensburg et al. (2007) Swiegers et al. (2007) markers conferring drug resistance, genes are Gonza´lez-Ramos et al. (2008) used to generate resistance to certain antibiotics, Herrero et al. (2008) herbicides, or amino acid analogs that are toxic Gonza´lez-Ramos et al. (2009) for yeast, such as canavanine, chloramphenicol, Ehsani et al. (2009) cycloheximide, diuron, geneticin, hygromicin, kanamycin, methotrexate, sulfometuron, p- fluorophenylalanine, phleomycin, and zeocin. Various resistance mechanisms can be involved, case of cycloheximide, sulfometuron, or p- the most common being the inactivation of the fluorophenylalanine resistance). As shown in corresponding antibiotic. Alternatively, the Figure 7.2a, the cells that take up the exogenous resistance gene may be a mutant version or DNA containing the resistance gene are easily allele of a gene encoding the primary target differentiated from those that do not by of the selective agent (for instance in the growing them in medium containing the 174 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

(a) Antibiotic- resistance gene

antr

Antibiotic- containing Nontransformed cell medium (sensitive to the drug)

antr antr

Antibiotic- Antibiotic- containing medium containing medium Transformed cell (resistant to the antibiotic)

(b) Minimal medium Wild-type gene without amino acid aa- aa- aa+

Nontransformed cell (does not produce) aa- Minimal medium aa- without amino acid aa-

aa+ aa+

Minimal medium without amino acid Transformed cell (produces amino acid)

aa-

FIGURE 7.2 Schematic diagram showing the genetic selection of transformed wine yeasts. Selection based on acquisition of antibiotic-resistance genes (a). The recipient cell cannot grow in medium containing an antibiotic to which it is sensitive (white circle). The gene used for selection encodes a protein that inactivates the antibiotic. Consequently, the transformed cells that contain the gene will grow in selective medium containing the drug while the nontransformed cells will die. Selection by auxotrophic complementation (b). The system is based on the use of cells carrying a mutation in the pathway used to synthesize an essential cell metabolite, in this case an amino acid. The cell is therefore auxotrophic for that nutritional requirement and cannot grow in minimal medium without addition of the metabolite. The gene used for selection is a wild-type version of the mutated gene in the receiving cell. As a result, the mutation is complemented in the transformed cells, which will be converted into prototrophs. The transformed cells can therefore grow in minimal medium without addition of nutrients whereas untransformed cells will not grow in this selective medium. SYSTEMS FOR THE GENETIC TRANSFORMATION OF WINE YEASTS 175 antibiotic. In the case of auxotrophic comple- TABLE 7.2 Selection Markers Used to Generate mentation, the strains used carry mutations Transgenic Wine Yeasts that prevent the synthesis of a particular amino acid, nitrogenous base, or vitamin. For Selection method Reference instance, mutations have been used in the URA3 complementation Dequin and Barre (1994) gene, which encodes an enzyme in the Ansanay et al. (1996) URA3 Pe´rez-Torrado et al. (2002) synthesis pathway for the nucleotide uracil, Gonza´lez-Ramos et al. (2008) or TRP1, which codes for an enzyme involved Herrero et al. (2008) in the synthesis of the amino acid tryptophan. Gonza´lez-Ramos et al. (2009) These mutants are unable to grow in minimal LEU2 complementation Vilanova et al. (2000) medium without the addition of uracil or tryptophan, respectively. Selection involves Cycloheximide Pe´rez-Gonza´lez et al. (1993) resistance Gonza´lez-Candelas et al. (1995) inclusion of the wild-type allele in the transfor- Sańchez-Torres et al. (1996) mation vector so that the receiving cells repair Sańchez-Torres et al. (1998) the molecular lesion and are able to grow in Ganga et al. (1999) minimal medium without the addition of Gonza´lez-Candelas et al. (2000) supplements (see Figure 7.2b). Manzanares et al. (2003) In the case of wine yeasts, selection by auxo- Killer factor resistance Lee and Hassan (1988) trophic complementation is not straightforward Boone et al. (1990) since, as indicated previously, many, if not all, Salek et al. (1990) industrial strains are polyploid and have more Phleomycin resistance Remize et al. (1999) than one copy of each gene. As a result, each Cambon et al. (2006) copy must be mutated, with the corresponding Coulon et al. (2006) Husnik et al. (2006) difficulties that this entails. In contrast, the use of antibiotic-resistance genes does not present G418 resistance Laing and Pretorius (1993) methodological difficulties but may not receive van Rensburg et al. (1994) Puig et al. (1998) consumer support. This public rejection rests Dequin et al. (1999) on the possibility that the marker gene might Walker et al. (2003) be transferred to gut bacteria by conjugation or Cambon et al. (2006) natural transformation and lead to the genera- van Rensburg et al. (2007) tion of antibiotic-resistant pathogenic bacterial Gonza´lez-Ramos et al. (2008) Ehsani et al. (2009) flora. However, although such a view is Gonza´lez-Ramos et al. (2009) supported by certain environmental non- government organizations, it has no scientific Sulfometuron Petering et al. (1991) resistance Fernańdez-Gonza´lez et al. (2005) basis and, according to the World Health Orga- van Rensburg et al. (2005) nization (WHO), evidence suggests that no Swiegers et al. (2007) such risk exists (World Health Organization, van Rensburg et al. (2007) 1993). Nevertheless, as a result of public pres- p-fluorophenylalanine Gonza´lez-Ramos et al. (2008) sure, the use of antibiotic-resistance genes in resistance Gonza´lez-Ramos et al. (2009) foodstuffs has been prohibited in the EU since 2005. Since many of the transgenic wine yeasts that gene following selection (Puig et al., 1998; have been produced carry antibiotic-resistance Walker et al., 2003). Instead of plasmid vectors, genes as selection markers (see Table 7.2), these systems use linear DNA fragments con- systems have been developed to eliminate the taining the gene of interest and the selection 176 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING gene. They also contain fragments of DNA cor- selective pressure. The other construct carries responding to the S. cerevisiae genome that allow the gene of interest and is designed to integrate them to be targeted to specific loci on a chromo- in the yeast genome. This strategy takes advan- some of the wine yeast. Once the transgenic tage of the lack of specificity that is characteristic wine yeast has been selected, the resistance of microbial transformation techniques. In other gene is excised to produce a final transgenic words, when a cell is competent, it is likely to yeast that carries only the gene of interest at take up both types of DNA present in the trans- a defined position in the genome. However, in formation medium. Once the transformants most cases, elimination of the antibiotic-resis- have been selected using the resistance marker, tance gene leaves a trace or scar in the genome integration of the construct at the expected site in the form of a short fragment of exogenous is confirmed. Then, growth of the yeast in the DNA. To prevent this occurring, genes absence of selective pressure for the marker belonging to the same yeast can be used as resis- allows it to be eliminated along with all associ- tance markers. These are generally mutant ated bacterial sequences. This strategy has alleles of the protein targets of antibiotics such been used to generate two yeasts that have as cycloheximide, p-fluorophenylalanine, or been approved by the United States Food and sulfometuron. In order for this technique to be Drug Administration (FDA) and Health Canada effective, however, the transformation methods for use in foodstuffs (Coulon et al., 2006; Husnik must prevent carryover of DNA derived from et al., 2006). These types of yeast are referred to the vectors used to generate the DNA construct. by the acronym GRAS (generally recognized as Alternatively, it is possible to use a cotransfor- safe). Unlike in classical genetics, where little is mation strategy (see Figure 7.3). In this case, known about the changes produced by two types of DNA molecule are used in the mutation or crossing, here the nature of the transformation. One carries the antibiotic-resis- genetic modification is understood in detail, tance gene in a plasmid that replicates in the with complete sequence information available yeast nucleus but is unstable in the absence of for both the inserted fragment and the integration site. A different situation occurs in transgenic wine yeasts in which improvements are generated not by expression of a new gene (or Transformation a change in the pattern or level of expression Cotransformation for the yeast’s own gene) but rather by elimination of a gene. This involves integration of Elimination a marker gene in a specific locus, normally replacing the gene of interest. The difficulty of this approach lies in the diploid or aneuploid character of industrial strains, since two or FIGURE 7.3 Schematic representation of conventional more copies of the same gene must be elimi- transformation and cotransformation systems. In conven- nated in order to generate an improvement in tional transformation, the selection marker (hatched rect- the characteristics of the strain. A recent angle) and the gene or construct of interest (white rectangle) example involved the production of yeasts form part of the same molecule. In cotransformation, these with enhanced release of mannoproteins two elements are present in different molecules, allowing easy elimination (curing) of the marker and other exoge- through the elimination of genes involved in nous sequences without leaving scars or traces in the cell wall biogenesis (Gonza´lez-Ramos et al., genome. 2008, 2009). TRANSGENIC WINE YEASTS 177

said, the expression of some genes, such as 3. REGULATION OF GENE HSP12, can be considered a marker of resistance EXPRESSION: PROMOTERS to the stress imposed by vinification (Ivorra OF INTEREST IN BIOTECHNOLOGY et al., 1999). It has been shown that the promoter for the TDH2/3 gene, which encodes the glycer- To produce a transgenic wine yeast, it is not aldehyde-3-phosphate dehydrogenase enzyme, sufficient to simply introduce a gene of interest drives the highest levels of expression during into the genome of the yeast. For a variety of vinification and that some genes such as SPI1 reasons, it is also important to control the appear to display enhanced expression during expression of the exogenous gene. In many the stationary phase (Puig & Pe´rez-Ortıń, 2000). cases, overexpression can be toxic for the trans- More promoters have become available for formed yeast. At other times, it may be neces- use in controlling transgene expression in wine sary to express the transgenic protein at yeasts as a consequence of DNA microarray a specific point during vinification. For instance, analysis of transcription profiles during vinifica- if the intention is to enhance aroma, it is neces- tion (Backus et al., 2001; Beltrań et al., 2006; sary to express the gene towards the end of vini- Jimeńez-Martı´ & del Olmo, 2008; Marks et al., fication, when the yeast has entered the 2008; Mendes-Ferreira et al., 2007; Rossignol stationary phase. This is because aromatic et al., 2003, 2006; Zuzuarregui et al., 2006). To compounds are volatile and, if the gene is date, however, few promoters have been used expressed from the beginning, much of the and most drive constitutive expression (see aromatic product will be lost. Nevertheless, Table 7.3). there are also situations in which it is appropriate to maintain constant expression. Each of these cases requires the use of different regula- 4. TRANSGENIC WINE YEASTS tory sequences known as promoters, derived from inducible, delayed, or constitutively The tools described earlier have been used to expressed genes, accordingly. produce transgenic wine yeasts that have been A great deal of information is available on the produced to display improved metabolic charac- molecular control of gene expression in labora- teristics or enhanced production of compounds tory strains of S. cerevisiae grown under with interesting organoleptic or nutritional prop- controlled conditions. This information erties (see Table 7.4). Most of these yeasts have continues to be expanded through the use of been developed in the laboratories of the Austra- DNA microarrays (Pe´rez-Ortıń et al., 2002). In lian Wine Research Institute at the University of a pioneering study, Puig et al. (1996) showed Adelaide (http://www.awri.com.au/), the Insti- that genes such as HSP104, POT1, and SSA3, tute for Wine Research at the University of which are expressed during the late stationary Stellenbosch in South Africa (http://academic. phase in laboratory yeast strains, were sun.ac.za/wine_biotechnology/), the Institut de expressed early during vinification. In the Produits de la Vigne del Institut Nacional de la same study, those authors found that the Recherche Agronomique (INRA) in Montpellier, promoter of the gene ACT1 is able to drive France (http://www.montpellier.inra.fr/), the expression during late phases of vinification. Instituto de Agroquı´mica y Tecnologıáde Subsequent, more detailed studies have shown Alimentos del Consejo Superior de Investiga- that gene expression patterns during vinifica- ciones Cientı´ficas in Valencia, Spain (http:// tion depend on the strain of wine yeast analyzed www.iata.csic.es/), the Instituto de Fermenta- (Carrasco et al., 2001; Riou et al., 1997). That ciones Industriales del Consejo Superior de 178 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

TABLE 7.3 Promoters Used in the Generation of Transgenic Wine Yeasts

Gene1 Protein Reference ACT1 Actin Pe´rez-Gonza´lez et al. (1993) Gonza´lez-Candelas et al. (1995) Sańchez-Torres et al. (1996) Sańchez-Torres et al. (1998) Ganga et al. (1999) Gonza´lez-Candelas et al. (2000) Manzanares et al. (2003) ADH1 Aldehyde dehydrogenase Petering et al. (1991) Dequin and Barre (1994) Ansanay et al. (1996) Dequin et al. (1999) Cambon et al. (2006) van Rensburg et al. (2007) PGK1 Phosphoglycerate kinase Vilanova et al. (2000) Fernańdez-Gonza´lez et al. (2005) Coulon et al. (2006) Husnik et al. (2006) Swiegers et al. (2007) TDH3 Glyceraldehyde-3-phosphate Herrero et al. (2008) dehydrogenase Ehsani et al. (2009)

1Gene from which the promoter is derived. Examples are shown only for cases in which the regulatory sequences differ from those normally associated with the gene used in the transformation experiment. Examples involving introduction of multiple copies of a gene belonging to the transformed yeast are not shown.

TABLE 7.4 Phenotypes Improved by Genetic Engineering in Wine Yeasts

Modified phenotype Reference Elevated glycogen to increase resistance to nutritional stress Pe´rez-Torrado et al. (2002) Synthesis of pectinases to improve filtration Gonza´lez-Candelas et al. (1995) Vilanova et al. (2000) Fernańdez-Gonza´lez et al. (2005) van Rensburg et al. (2007) Synthesis of killer factor Lee and Hassan (1988) Boone et al. (1990) Salek et al. (1990) Synthesis of pediocin as an antimicrobial agent Schoeman et al. (1999)

Synthesis of glucose oxidase as an antimicrobial agent Malherbe et al. (2003) Increased acidity Dequin and Barre (1994) Dequin et al. (1999) Reduced acidity Bony et al. (1997) Volschenk et al. (1997a)1 Volschenk et al. (1997b)1 Husnik et al. (2006)

(Continued) TRANSGENIC WINE YEASTS 179

TABLE 7.4 Phenotypes Improved by Genetic Engineering in Wine Yeastsdcont’d

Modified phenotype Reference Synthesis of glucose oxidase and reduced ethanol levels Malherbe et al. (2003) Increased glycerol and reduced ethanol Michnick et al. (1997) Remize et al. (1999) de Barros Lopes et al. (2000) Eglinton et al. (2002) Cambon et al. (2006) Ehsani et al. (2009) Synthesis of b-(1,4)-endoglucanase to improve aroma Pe´rez-Gonza´lez et al. (1993) Synthesis of a-L-arabinofuranosidase to improve aroma Sańchez-Torres et al. (1996) Synthesis of b-glucosidase to improve aroma Sańchez-Torres et al. (1998) van Rensburg et al. (2005) Synthesis of b-(1,4)-endoxylanase to improve aroma Ganga et al. (1999) Synthesis of alcohol acetyltransferase to improve aroma Lilly et al. (2000) Verstrepen et al. (2003) Synthesis of a-L-rhamnosidase to improve aroma Manzanares et al. (2003) Synthesis of phenol decarboxylase to improve aroma Smit et al. (2003)1

Increased resveratrol levels Gonza´lez-Candelas et al. (2000) Becker et al. (2003)1

Synthesis of b-glucuronidase as a marker Petering et al. (1991) Reduction of ethyl carbamate levels Coulon et al. (2006) Increased mannoprotein content Gonza´lez-Ramos et al. (2008) Gonza´lez-Ramos et al. (2009) Synthesis of monoterpenes to improve aroma Herrero et al. (2008) Synthesis of tryptophanase to improve aroma Swiegers et al. (2007)

1Modiﬁcation analyzed only in laboratory strains.

Investigaciones Cientı´ﬁcas in Madrid, Spain Before the cells are dried and packed, feeding (http://www.iﬁ.csic.es/), and the Wine with the carbon source is stopped and the yeast Research Center at the University of British cells are forced to switch to respiratory metabo- Columbia in Canada (http://www.landfood. lism and consume the ethanol produced during ubc.ca/wine/index.html). fermentation. The dried cells are then rehydrated before being added to the fermentation tank. 4.1. Improvements Affecting the However, the prior processes of glucose starva- Industrial Winemaking Process tion and dehydration stress the cells and reduce their viability and fermentative capacity (dis- During the industrial production of active cussed in more detail in Chapter 11). In an effort dried yeast, cells are grown under aerobic condi- to alleviate these problems, transgenic wine tions until a considerable biomass is obtained. yeasts have been produced that overexpress the 180 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

GSY2 gene (Pe´rez-Torrado et al., 2002). This gene strain (Boone et al., 1990; Lee & Hassan, 1988; encodes the enzyme glycogen synthase, and its Salek et al., 1990). More recently, the gene from overexpression leads to increased accumulation Pediococcus acidilactici that encodes the bacte- of glycogen, a cell metabolite linked to the stress riocin pediocin A has been expressed in labora- response. As a result, cell viability is increased. tory strains of S. cerevisiae (Schoeman et al., It is desirable for vinification to occur in 1999). Pediocin A is active against, among the shortest time possible without negatively others, Listeria monocytogenes and is effectively influencing the organoleptic properties of the secreted by the transgenic yeast. According to wine obtained. In an effort to achieve this, Schoeman et al., these developments may transgenic wine yeasts that overexpress the make it possible to produce strains of baker’s, glycolytic enzyme glyceraldehyde-3-phosphate wine, or brewer’s yeast with biocontrol activity. dehydrogenase have been produced. However, Similarly, a gene from the filamentous fungus although these yeasts have been reported to Aspergillus niger, which encodes the enzyme indirectly reduce fermentation time, they also glucose oxidase, when introduced into labora- lead to substantial changes in the volatile profile tory strains results in inhibition of the growth and a reduction in the alcohol content of the of spoilage yeasts such as Acetobacter aceti and wine (Remize et al., 1999). Gluconobacter oxidans when cocultured with the During the winemaking process, pectinolytic transgenic yeast (Malherbe et al., 2003). This enzymes are usually introduced to enhance the effect is mediated by enhanced production of extraction of the must and improve clarification hydrogen peroxide. and filtration of the wine. This practice increases the cost of the process and risks the introduction 4.2. Improvements Affecting the of contaminating activities that affect the aroma Physicochemical Properties of Wine or color of the wine, since pure enzyme preparations are not commercially available. To solve In hot winegrowing regions, it is common to this problem, transgenic wine yeasts have been obtain musts with low acidity and unbalanced produced that carry a gene, pelA from the fila- organoleptic properties. A metabolic engi- mentous fungus Fusarium solani, which encodes neering strategy to increase the acidity of these a pectate lyase (Gonza´lez-Candelas et al., 1995). wines has been developed based on the intro- This transgenic yeast secretes the pectinolytic duction of a lactic acid fermentation pathway enzymes into the must during fermentation. into wine yeasts (Dequin & Barre, 1994; Dequin Vilanova et al. (2000) adopted a different et al., 1999). This was achieved by expressing approach involving the S. cerevisiae polygalac- a gene from the lactic acid bacteria Lactobacillus turonase PGU1. Since expression of PGU1 is casei, which encodes L(þ)-lactate dehydroge- weak or absent in many strains, the authors nase and leads to the conversion of 20% of the engineered a wine strain that constitutively glucose into lactate during fermentation. The expressed the gene. resulting pH reduction of 0.3, however, was Finally, it is worth mentioning the production also associated with a slightly reduced ethanol of various transgenic wine strains that are able content. to eliminate microorganisms that interfere with In contrast, wines from cooler regions have the fermentation process or lead to microbial an excess of malic acid, which must be cor- spoilage. Wine strains have been produced rected by the addition of lactic acid bacteria that contain extra copies of genes coding for such as Oenococcus oeni to produce malolactic killer factors or other genes encoding different fermentation and deacidification. Experiments killer factors from those already carried by the have been performed in both laboratory and TRANSGENIC WINE YEASTS 181 wine strains to coexpress the Schizosaccharomy- andmanipulationofcofactors(Cambonetal., ces pombe gene mae1, which encodes malate 2006; Ehsani et al., 2009). permease, with a malic enzyme from the lactic Chemical stability is the main physicochem- acid bacteria Lactococcus lactis (Bony et al., 1997; ical property of concern to enologists, particu- Volschek et al., 1997a). Under those conditions, larly during the production of white wines. the transgenic S. cerevisiae strain was able to The principal considerations are protein and transport malic acid into the cell and transform tartrate precipitation. The first involves the it into lactate in a reaction catalyzed by the precipitation of complexes containing mainly malic enzyme, thus reducing the acidity of the grape-derived proteins that are unstable in the wine. In an alternative approach, the mae1 presence of ethanol, especially over prolonged and mae2 genes from S. pombe were coex- periods at insufficiently cool temperatures. pressed in S. cerevisiae (Volschenk et al., Tartrate precipitation involves the formation of 1997b). The mae2 gene encodes the S. pombe potassium bitartrate crystals in bottled wine. It malic enzyme and coexpression led to the effec- has been reported that, in both cases, the tive degradation of 8 g of malate per liter of problem is reduced by the presence of higher must,comparedwith4.5g/Ldegradedby concentrations of mannoproteins (Dupin et al., earlier transgenic yeast strains. A recent 2000; Feuillat et al., 1998). Transgenic wine advance has been the production by cotransfor- strains have therefore been developed in which mation of the commercial yeast strain ML01, the secretion of mannoproteins during fermen- whichcoexpressesthemalatepermeaseofS. tation is increased by eliminating all copies pombe with the malolactic enzyme of O. oeni (two or three) of certain genes linked to cell (Husnik et al., 2006). wall biosynthesis (Gonza´lez-Ramos et al., 2008, Genetic engineering has also been used 2009). successfully to produce low-alcohol wines. Strategies have involved expressing the A. niger 4.3. Improvement of Organoleptic, gene encoding the glucose oxidase enzyme Nutritional, and Safety-related (Malherbe et al., 2003), enhancing expression Properties of the glyceraldehyde-3-phosphate dehydrogenase-encoding genes GPD1 or GPD2 (de Barros The fruity aroma of certain wines has been Lopes et al., 2000; Michnick et al., 1997; Remize enhanced through the use of genetically et al., 1999), or deleting the FPS1 gene encoding engineered wine yeasts. This aroma depends the glycerol transporter (Eglinton et al., 2002). mainly on the presence of certain terpenesd However, many of these studies have only particularly geraniol, nerol, and linalooldthat been performed in laboratory strains. Further- are found in the must in two fractions: a free more, reducing the ethanol concentration fraction that produces aroma by virtue of being implies increasing that of glycerol and leads to volatile and a bound fraction, associated via a reduced generation of biomass and changes diglycosidic bonds with fragments of cell wall in the production of volatile compounds such from the grape berries. The bound fraction as acetate, acetoin, 2,3-butanediol, or succinate. does not contribute to wine aroma unless the Its use thus needs to be carefully reviewed glycosidic bonds are cleaved by glycosidases. before implementation (Remize et al., 1999). This process is favored by the prior activity of Efforts have continued to be made in recent cellulases and hemicellulases (Villanueva et al., years to redirect the carbon flux of S. cerevisiae 2000). In order to perform this enzymatic treat- towards pathways that are more compatible ment, genes from filamentous fungi and aerobic with wine quality through metabolic engineering yeast that code for these enzymes have been 182 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING expressed in wine yeasts, with the result that yeasts (Gonza´lez-Candelas et al., 2000). Alterna- concentrations of one or more aromatic terpenes tive strategies have been developed to increase were increased (Ganga, 1999; Manzanares, 2003; resveratrol production in laboratory strains by Pe´rez-Gonza´lez et al., 1993; Sańchez-Torres coexpressing a gene from the poplar tree encod- et al., 1996, 1998). Transgenic wine yeasts have ing a coenzyme A (CoA) ligase with the resver- also been produced that are able to produce atrol synthase gene from grape (Becker et al., terpenes directly as a product of their metabo- 2003). It should be remembered, however, that lism (Herrero et al., 2008). the effect of resveratrol has only been analyzed Alternative strategies have sought to express in vitro and that wine is an alcoholic beverage the PAD C gene from B. subtilis and the PDC gene and therefore not a suitable vehicle for the from the lactic acid bacteria Lactobacillus planta- administration of active pharmaceutical rum, genes that encode decarboxylases for ingredients. phenolic compounds (Smit et al., 2003). Microvi- One of the main food safety concerns associ- nification experiments using these transgenic ated with wine production is the formation of wine yeasts showed that phenolic compounds ethyl carbamate. In fact, there are legal limits present in the must were converted into 4-vinyl for this compound that are applicable for and 4-ethyl derivatives, which are volatile and commercial wine import and export. It is contribute to an enhanced aroma. A more a potential carcinogen mainly derived from complicated strategy involved increasing the a spontaneous reaction between ethanol and levels of alcohol acetyltransferase by overex- urea formed by the yeast as a result of arginine pression of ATF1 in wine yeasts in an attempt metabolism. Kitamoto et al. (1991) approached to induce overproduction of volatile aromatic this problem in a sake-producing yeast strain esters (Lilly et al., 2000). Various genes encode by eliminating both copies of the gene CAR1, enzymes with this activity and comparative which encodes an enzyme that catalyzes the first deletion studies have been performed in labora- step in S. cerevisiae arginine metabolism. More tory strains (Verstrepen et al., 2003). These recently, Coulon et al. (2006) used cotransforma- experiments have shown that ATF1 has a role tion to produce a yeast that constitutively in the synthesis of volatile esters. expresses the urease-encoding gene Dur1,2.In There is an important contribution to the some experiments, the use of this yeast has led quality of traditional-method sparkling wines to a 90% reduction in the ethyl carbamate by products derived from yeast autolysis (see content of the wines produced when compared Chapter 2). Given that this is a slow process, with those produced using the parent strain. transgenic wine yeasts have been produced The commercial name of this strain is EMLo1. that undergo more rapid autolysis as a result There are undoubtedly many more develop- of the manipulation of genes related to autoph- ments underway in genetic engineering labora- agy (Cebollero et al., 2009). tories that will come to light in the next few Finally, a Candida molischiana gene encoding years. The advances made to date suggest a b-glucosidase has been expressed in a wine various options for the industrial application yeast. This enzyme can mobilize resveratrol of these new technologies. However, the use of conjugates, compounds thought to be linked to transgenic approaches will have to be consid- the so-called “French paradox” due to their anti- ered within the appropriate legal framework cholesterolemic and antitumoral effects in vitro. and it will be necessary to convince consumers The transgenic yeast that has been developed of the benefits that are offered. These consider- produces wines with a higher resveratrol ations will be discussed in the following content than that produced with conventional sections. LEGISLATION AFFECTING THE USE OF GENETIC ENGINEERING IN THE WINE INDUSTRY 183

5. LEGISLATION AFFECTING THE The most notable element of this legislation USE OF GENETIC ENGINEERING refers to the marketing of transgenic products, IN THE WINE INDUSTRY which must be considered on a case-by-case basis, with obligatory evaluation by panels of 5.1. European Legislation scientific experts. It is usually a long process that involves submission of a dossier by the Wine yeasts that have been modified by company seeking marketing authorization to genetic engineering are transgenic organisms demonstrate the lack of adverse effects on and their industrial use would thus lead to the human health or the environment associated production of transgenic wines. With the excep- with the product for which authorization is tion of a small production in Moldavia a few sought. The minimum requirement in terms of years ago (I.S. Pretorius, personal communica- health assessment is an analysis of the nutrition), there have been no commercial references tional composition of the transgenic product to the production or sale of transgenic wines. compared with the corresponding conventional Sales of beer labeled as transgenic have recently product, an assessment of allergenicity, and begun, however, in Sweden, although the beer is a preclinical study of toxicity in laboratory not produced with a transgenic brewer’s yeast animals. In the first health assessment of a trans- but rather with a conventional yeast and a barley genic wine yeast, performed in a strain that wort to which fragments of transgenic maize expresses a xylanase from the filamentous have been added as a nitrogen source (Editorial, fungus A. nidulans (Pico´ et al., 1999), it was 2004). What would be the implications of seen that no additional risk was associated deciding to market a wine produced with with ingestion of the transgenic product a transgenic yeast strain? The response to this compared to its conventional counterpart. question will vary according to the legislation Assessment of environmental impact must be on transgenic food and beverages in the country carried out by controlled environmental release in question. Below we discuss the European of the GMO. These processes tend to take an model, which affects most of the producing average of 5 years and require significant finan- countries. cial investment. The results are assessed by If approval is requested for marketing in the a panel of scientific experts. EU, or more specifically in a member state In Europe, evaluation of transgenic foods is such as Spain, the evaluation process would undertaken by the panel on genetically modi- fall under legislation regulating research and fied (GM) foods of the European Food Safety development, environmental release, patent- Authority (EFSA). This panel is compossed of ability, marketing, and labeling of transgenic scientific experts in various fields related to foods produced by genetic engineering. food safety evaluation and molecular biology. Currently, following a recent modification, the Based on their evaluation, a final decision is heart of that legislation is EU Directive 2001/ made on whether to accept or reject the 18/EC on the deliberate release into the envi- proposal. In addition, it is obligatory for the ronment of GMOs, which has been developed public to be informed of all authorizations and and adapted in various documents, particularly for the European Parliament to be consulted. Regulation (EC) 1829/2003 on genetically modi- Finally, the Council of the EU can approve or fied food and feed and Regulation (EC) 1830/ reject by majority vote the commission’s 2003 on the traceability and labeling of GMOs proposal to authorize a transgenic product. and the traceability of food and feed products In this new directive, member states must produced from GMOs. guarantee labeling and monitoring throughout 184 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING all phases of marketing, and initial approval is a complex situation that has affected and will limited to a period of 10 years. Following continue to affect all links in the food production marketing authorization, there is an obligatory chain, from the farmer to the consumer, via the monitoring period during which possible long- food processing and distribution industry. term effects can be assessed, particularly in rela- The situation in other non-EU countries is tion to environmental impact. less complex. In the United States, the FDA made a public declaration in 1992 affirming 5.2. Labeling that it was not necessary to develop specific legislation for the commercialization of GM Regulation 1830/2003 on traceability and foods. In their view, the legislation applicable labeling of GMOs and the traceability of food to the marketing of foods obtained using and feed products produced from GMOs is the conventional genetic methods was sufficient, current point of reference for legislation on as it required detailed analysis of potential labeling. The regulation states that anyone in adverse effects on hygiene, health, and the envi- the EU who plants transgenic grape varieties ronment (Kessler et al., 1992). Currently, the must inform all of their clients in writing and FDA requires prior evaluation of genetic modi- store a copy of this communication for at least fications to be used in food or feeds but does 5 years. The same applies to anyone who not require specific labeling when the food is markets a transgenic yeast in the EU. In addi- sufficiently similar in its nutritional composition tion, the recently approved Regulation 65/2004 to the conventional alternatives. There is a clear established a system for creating and assigning distinction between the American model, unique identifiers for GMOs. which assesses the final product irrespective According to these regulations, in the EU, of the method used to obtain it, and the Euro- a wine would be considered transgenic and pean model, in which both elements are thus require labeling as such when it is prepared considered. from a transgenic grape variety or wine yeast. In In other countries, such as Argentina, contrast, if an enzyme obtained from a GMO is Australia, Canada, and Japan, the system is added to the must or wine, it need not be labeled more similar to the American model. Some as genetically modified so long as the enzyme is other countries still have very loose legislation not active in the bottled wine. This decision, for or are discussing how to proceed while which there is no scientific justification, has been continuing to market GM foods. These differ- questioned by some scientists (Ramoń et al., ences could lead to havens of permissiveness 2004). that ultimately undermine efforts to guarantee the lack of additional risk associated with the 5.3. The Situation in Other Countries commercialization of these products, a possibility that is clearly undesirable. Many consider the extensive legislation affecting the marketing and labeling of GM foods in the EU to be a clear example of over- 6. THE FUTURE legislation. Political pressure, particularly from certain environmental groups and multinational Is there a market for wines produced by food companies, has had more influence on the fermentation with a transgenic yeast? The preparation of these directives and regulations answer to this question is not straightforward. than have common sense and consumer interests It will depend on the country in which the (Ramoń et al., 1998). This has given rise to wine is produced, the country in which it is THE FUTURE 185 sold, and, of course, the benefits offered by the of sales. These wines account for a much higher transgenic modification. percentage of annual sales than the more For many years, there was a clear lack of famous wines and they also have problems interest in these new technologies shown by the that might be resolved through the use of International Organization of Wine and Vine genetic engineering. We should not forget, (OIV). Partly for this reason, the United States though, that wine is a very traditional product has abandoned this organization. Many of the with a strong geographic influence. It is much countries that are most active in wine biotech- more than an alcoholic beveragedit represents nology research, including the United States part of the identity of many countries and and Canada, are not members of the OIV. A regions. In products of this type with a long number of them have undertaken genome tradition, it is much more complicated for new sequencing projects for grape varieties or projects technologies to be introduced and, when they to develop transgenic vines or yeasts that are are, they only succeed when they offer some- financed by public and private funds. However, thing of genuine interest to the consumer. It is the situation in the EU, and in particular the therefore fair to predict that transgenic wines major wine-producing countries such as France produced from vine cultivars modified to be and Spain, is quite different. There are leading resistant to pests are likely to fail, as the benefit groups on the development of transgenic yeasts is for the grower rather than the consumer. In as well as some groups involved in sequencing contrast, a transgenic wine in which improve- the genome of grape varieties, but there is no ments have been made in aroma or color interest in these groundbreaking technologies without affecting price could have a greater like- from the wineries themselves. Some Latin Amer- lihood of acceptance. A transgenic wine contain- ican countries such as Argentina and Chile have ing elevated concentrations of a compound begun to employ these new technologies. Their offering nutritional or health benefits might future direction is more likely to follow that of have an even greater chance of success despite Australia, the United States, and South Africa the fact that, as an alcoholic beverage, it should than that of the EU. never be considered an appropriate vehicle for It is currently unimaginable for transgenic active ingredients designed to improve wines to be sold in the EU. This is different consumer health. As mentioned earlier, some from the situation in the United States, where such examples already exist. it is unlikely that specific labeling would be In the United States and Canada, where required for the internal market. One can administrative procedures are apparently less reasonably assume that the response of costly and in the absence of pressure from orga- consumers to transgenic wines will reflect their nizations such as the OIV, two transgenic wine feelings towards GM foods in general. If this is yeasts have already been approved by the rele- the case, we can define a region of strong rejec- vant authorities (Coulon et al., 2006; Husnik tion in the EU represented by Austria, France, et al., 2006). Furthermore, these countries do Germany, and some Scandinavian countries, not require specific labeling for transgenic prod- and a more receptive region comprising Spain ucts of this kind. To obtain approval, it was and Portugal. essential to provide scientific evidence of func- It remains unlikely that genetic engineering tional equivalence (apart from those characteris- will be applied to wines from famous appella- tics that have been intentionally modified) tions such as Rioja or Bordeaux; it is more likely between the original yeast and the correspond- to be used in those wines for which improve- ing transgenic strain. This entailed the use of ments in value for money are key determinants cotransformation systems and the removal of 186 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING unnecessary sequences from the final construct, transformation of Saccharomyces cerevisiae and other along with the use of tools to analyze the tran- fungi. Curr. Genet., 17,97e103. scriptome and provide additional data to Backhus, L. E., DeRisi, J., Brown, P. O., & Bisson, L. F. (2001). Functional genomic analysis of a commercial wine strain demonstrate equivalence (Coulon et al., 2006; of Saccharomyces cerevisiae under differing nitrogen Husnik et al., 2006). From a commercial point conditions. FEMS Yeast Res., 1,111e125. of view and in terms of public opinion, these Barre, P., Ve´zinhet, F., Dequin, S., & Blondin, B. (1993). yeasts differ substantially. One aims to resolve Genetic improvement of wine yeast. In G. H. Fleet (Ed.), a problem that is partly technological and partly Wine microbiology and biotechnology (pp. 421e447). Singapore: Harwood Academic Publishers. affects the sensory properties of the final Becker, J. V. W., Gareth, O. A., van der Merwe, M. J., product (malolactic fermentation) and for which Lambrechts, M. G., Vivier, M. A., & Pretorius, I. P. (2003). nonrecombinant options are available, such as Metabolic engineering of Saccharomyces cereviseae for the conventional malolactic fermentation involving synthesis of the wine-related antioxidant resveratrol. . . The other addresses a problem for which FEMS Yeast Res., 4,79e85. O oeni Beltrań, G., Novo, M., Leberre, V., Sokol, S., there is no easily implemented conventional Labourdette, D., Guillamon, J. M., et al. (2006). Inte- option and that can also affect the health of the gration of transcriptomic and metabolic analyses for consumer. It is reasonable to assume that wines understanding the global responses of low-tempera- labeled as transgenic are more likely to be ture winemaking fermentations. FEMS Yeast Res., 6, accepted in the latter case than in the former. 1167e1183. Bony, M., Bidart, F., Camarasa, C., Ansanay, V., Dulau, L., There is still some way to go. It will certainly Barre, P., et al. (1997). Metabolic analysis of S. cerevisiae be necessary to generate recombinant yeast strains engineered for malolactic fermentation. FEBS strains that are as “clean” as possible, to under- Lett., 410, 452e456. take extensive assessment of health and envi- Boone, C., Sdicu, A. M., Wagner, J., Degre´, R., ronmental impacts, and to explain clearly to Sanchez, C., & Bussey, H. (1990). Integration of the yeast K1 killer toxin gene into the genome of marked consumers exactly how and why the modifica- wine yeasts and its effect on vinification. Am. J. Enol. tions have been made. It will also be important Vitic., 41,37e42. to learn the lessons provided by the industrial Borneman, A. R., Forgan, A. H., Pretorius, I. S., & use and commercialization of the yeasts Chambers, P. J. (2008). Comparative genome analysis of mentioned above. This is not going to be an a Saccharomyces cerevisiae wine strain. FEMS Yeast Res., 8, 1185e1195. easy task, but it will undoubtedly be easier for Cambon, B., Monteil, V., Remize, F., Camarasa, C., & our colleagues in Australia, North America, and Dequin, S. (2006). Effects of GPD1 overexpression in South Africa. With time and patience, we will Saccharomyces cerevisiae comercial wine yeast strains discover whether or not it was worth the effort. lacking ALD6 genes. Appl. Environ. Microbiol., 72, 4688e4694. Carbonneau, A. (1983). Ste`rilite`s male et female dans le References genre Vitis. II. Conseque`nces en ge`ne`tique et se`lection. Agronomie, 3, 645e649. Ansanay, V., Dequin, S., Camarasa, C., Schaeffer, V., Carrasco, P., Querol, A., & del Olmo, M. (2001). Analysis of Grivet, J. P., Blondin, B., et al. (1996). Malolactic the stress resistance of comercial wine yeast strains. fermentation by engineered Saccharomyces cerevisiae as Arch. Microbiol., 174, 450e457. compared with engineered Schizosaccharomyces pombe. Cebollero, E., Gonzalez-Ramos, D., & Gonzalez, R. (2009). Yeast, 12, 215e225. Construction of a recombinant autolytic wine yeast Antcliff, A. J. (1992). Taxonomy, the grapevine as a member strain overexpressing the csc1e1 allele. Biotechnol. Prog., of the plant kingdom. In B. G. Coombe & P. R. Dry 25, 1598e1604. (Eds.), Viticulture, Vol. 1 (pp. 107e118). Underdale, Cebollero, E., Gonza´lez-Ramos, D., Tabera, L., & Australia: Winetitles. Gonza´lez, R. (2007). Transgenic wine yeast technology Armaleo, D., Ye, G. N., Klein, T. M., Shark, K. B., comes of age: Is it time for transgenic wine? Biotechnol. Sanford, J. C., & Johnston, S. A. (1990). Biolistic nuclear Lett., 29, 191e200. REFERENCES 187

Coulon, J., Husnik, J. I., Inglis, D. L., van der Merwe, G. K., by the use of a recombinant wine yeast strain expressing Lovaud, A., Erasmus, D. J., & van Vuuren, H. J. J. (2006). the Aspergillus nidulans xlnA gene. Int. J. Food Micro- Metabolic engineering of Saccharomyces cerevisiae to biol., 47, 171e178. minimize the production of ethyl carbamate in wine. Gietz, D., St. Jean, A., Woods, R. A., & Schiestl, R. H. (1992). Am. J. Enol. Vitic., 57,113e124. Improved method for high efficiency transformation of de Barros Lopes, M., Rehman, A. U., Gockowiak, H., intact yeast cells. Nucl. Acids Res., 20, 1425. Heinrich, A. J., Langridge, P., & Henschke, P. A. (2000). Goffeau, A., et al. (1997). The yeast genome directory. Fermentation properties of a wine yeast overexpressing Nature, 387(suppl. 1), 1e105. the Saccharomyces cerevisiae glycerol 3-phosphate Gonza´lez, R., Martıńez-Rodrı´guez, A., & Carrascosa, A. dehydrogenase gene (GPD2). Aust. J. Grape Wine Res., 6, (2003). Yeast autolytic mutants potentially useful for 208e215. sparkling wine production. Int. J. Food Microbiol., 84, Delorme, E. (1989). Transformation of Saccharomyces 21e26. cerevisiae by electroporation. Appl. Environ. Microbiol., 55, Gonza´lez-Candelas, L., Cortell, A., & Ramoń, D. (1995). 2242e2246. Construction of a recombinant wine yeast strain Dequin, S. (2001). The potential of genetic engineering for expressing a fungal pectate lyase gene. FEMS Microbiol. improving brewing, wine-making and baking yeasts. Lett., 126, 263e270. Appl. Microbiol. Biotechnol., 56, 577e588. Gonza´lez-Candelas, L., Gil, J. V., Lamuela-Ravento´s, R., & Dequin, S., Baptista, E., & Barre, P. (1999). Acidification of Ramoń, D. (2000). The use of transgenic yeasts express- grape musts by Saccharomyces cerevisiae wine yeast ing a gene encoding a glycosyl-hydrolase as a tool to strains genetically engineered to produce lactic acid. Am. increase resveratrol content in wine. Int. J. Food Micro- J. Enol. Vitic., 50,45e50. biol., 59, 179e183. Dequin, S., & Barre, P. (1994). Mixed lactic acid-alcoholic Gonza´lez-Ramos, D., Cebollero, E., & Gonza´lez, R. (2008). A fermentation by Saccharomyces cerevisiae expressing the recombinant Saccharomyces cerevisiae strain over- Lactobacillus casei L(þ)-LDH. Biotechnology, 12, 173e177. producing mannoproteins stabilizes wine against Dupin, I. V., McKinnon, B. M., Ryan, C., Boulay, M., protein haze. Appl. Environ. Microbiol., 74, 5533e5540. Markides, A. J., Jones, G. P., et al. (2000). Saccharomyces Gonza´lez-Ramos, D., Quiro´s, M., & Gonza´lez, R. (2009). cerevisiae mannoproteins that protect wine from protein Three different targets for the genetic modification of haze: Their release during fermentation and lees contact wine yeast strains resulting in improved effectiveness of and a proposal for their mechanism of action. J. Agric. bentonite fining. J. Agric. Food Chem., 57, 8373e8378. Food Chem., 48, 3098e3105. Hammond, J. R. M. (1996). Yeast genetics. In F. G. Priest & Editorial. (2004). With just a hint of transgenes. New I. Campbell (Eds.), Brewing microbiology (pp. 45e82). Scientist, 2433,6. London, UK: Chapman and Hall. Eglinton, J. M., Heinrich, A. J., Pollnitz, A. P., Langridge, P., Herrero, O., Ramon, D., & Orejas, M. (2008). Engineering the Henschke, P. A., & de Barros Lopes, M. (2002). Saccharomyces cerevisiae isoprenoid pathway for de novo Decreasing acetic acid accumulation by a glycerol overproduction of aromatic monoterpenes in wine. Metabolic producing strain of Saccharomyces cerevisiae by deleting Engin., 10,78e86. the ALD6 aldehyde dehydrogenase gene. Yeast, 19, Hinnen, A., Hicks, J. B., & Fink, G. R. (1978). Transformation 295e301. of yeast. Proc. Natl. Acad. Sci. USA, 75, 1929e1933. Ehsani, M., Fernańdez, M. R., Biosca, J. A., Julien, A., & Husnik, J. I., Volschenk, H., Bauer, J., Colavizza, D., Luo, Z., Dequin, S. (2009). Engineering of 2, 3-butanediol dehy- & van Vuuren, H. J. J. (2006). Metabolic engineering of drogenase to reduce acetoin formation by glycerol- malolactic wine yeast. Metabolic Engin., 8, 315e323. overproducing, low-alcohol Saccharomyces cerevisiae. Ivorra, C., Pe´rez-Ortin, J. E., & del Olmo, M. (1999). An Appl. Environ. Microbiol., 75, 3196e3205. inverse correlation between stress resistance and stuck Fernańdez-Gonza´lez, M., Ubeda, J. F., Cordero-Otero, R. R., fermentations in wine yeasts, a molecular study. Bio- Gururajan, V. T., & Briones, A. I. (2005). Engineering of technol. Bioeng., 64, 698e708. an oenological Saccharomyces cerevisiae strain with Jimenez-Marti, E., & del Olmo, M. L. (2008). Addition of pectinolytic activity and its effect on wine. Int. J. Food ammonia or amino acids to a nitrogen-depleted Microbiol., 102, 173e183. medium affects gene expression patterns in yeast cells Feuillat, M., Charpentier, C., & Nguyen van Long., T. (1998). during alcoholic fermentation. FEMS Yeast Res., 8, Yeast’s mannoproteins: A new possible enological 245e256. adjuvant. Bull. O.I.V., 71, 945e967. Kessler, D. A., Taylor, M. R., Maryanski, J. H., Flamm, E. L., Ganga, M. A., Pinãga, F., Valle´s, S., Ramoń, D., & Querol, A. & Kahl, L. S. (1992). The safety of foods developed by (1999). Aroma improving in microvinification processes biotechnology. Science, 256, 1747e1749, 1832. 188 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

Kitamoto, K., Oda, K., Gomi, K., & Takahashi, K. (1991). wine yeast Saccharomyces cerevisiae EC1118. Proc. Natl. Genetic engineering of a sake yeast producing no urea Acad. Sci. USA, 106, 16 333e16 338. by successive disruption of arginase gene. Appl. Environ. Nu´ n˜ez, D. R., & Walker, M. J. (1989). A review of paleobo- Microbiol., 57, 301e306. tanical findings of early Vitis in the Mediterranean and Laing, E., & Pretorius, I. S. (1993). Co-expression of an of the origins of cultivated grapevines, with special Erwinia chrysanthemi pectate lyase-encoding gene (pelE) reference to prehistoric exploitations in the western and an E. carotovora polygalacturonase-encoding gene Mediterranean. Rev. Paleobot. Palynol., 61, 205e237. (peh1)inSaccharomyces cerevisiae. Appl. Microbiol. Bio- Pe´rez-Gonza´lez, J. A., Gonza´lez, R., Querol, A., Sendra, J., & technol., 39, 181e188. Ramoń, D. (1993). Construction of a recombinant wine Lee, F. J. S., & Hassan, L. H. M. (1988). Stability and yeast strain expressing a b-(1, 4)-endoglucanase activity expression of a plasmid-containing killer toxin cDNA in and its use in microvinification processes. Appl. Environ. batch and chemostat cultures of Saccharomyces cerevisiae. Microbiol., 59, 2801e2806. Biotechnol. Bioengin., 31, 783e789. Pe´rez-Ortıń, J. E., Garcıá Martinez, J., & Alberola, T. M. Lilly, M., Lambrechts, M. G., & Pretorius, I. S. (2000). Effect (2002). DNA chips for yeast biotechnology, the case of of increased yeast alcohol acetyltransferase activity on wine yeasts. J. Biotechnol., 98, 227e241. flavor profiles of wine and distillates. Appl. Environ. Pe´rez-Torrado, R., Gimeno-Alcan˜iz, J. V., & Matallana, E. Microbiol., 66, 744e753. (2002). Wine yeast strains engineered for glycogen Liti, G., Carter, D. M., Moses, A. M., Warringer, J., Parts, L., overproduction display enhanced viability under James, S. A., et al. (2009). Population genomics of glucose deprivation conditions. Appl. Environ. Microbiol., domestic and wild yeasts. Nature, 458, 337e341. 68, 3339e3344. Malherbe, D. F., du Toit, M., Cordero, R. R., van Petering, J. E., Henschke, P. A., & Langridge, P. (1991). The Rensburg, P., & Pretorius, I. S. (2003). Expression of the Escherichia coli b-glucuronidase gene as a marker for Aspergillus niger glucose oxidase gene in Saccharomyces Saccharomyces yeast strain identification. Am. J. Enol. cerevisiae and its potential applications in wine produc- Vitic., 42,6e12. tion. Appl. Microbiol. Biotechnol., 61, 502e511. Pico´, Y., Fernańdez, M., Rodrı´guez, R., Almude´ver, J., Manzanares, P., Orejas, M., Gil, J. V., de Graaff, L., Visser, J., Man˜es, J., Font, G., Marıń, R., Carda, C., Manzanares, P., & Ramoń, D. (2003). Construction of a genetically & Ramoń, D. (1999). Toxicological assesment of xylanase modified wine yeast strains expressing the Aspergillus X22 expressed by transgenic wine yeast. J. Agric. Food aculeatus rhaA gene, encoding an a-L-rhamnosidase of Chem., 47, 1597e1602. enological interest. Appl. Environ. Microbiol., 69, Pretorius, I. S. (2000). Tailoring wine yeast for the new 7558e7562. millennium, novel approaches to the ancient art of Marks, V. D., Sui, S. J. H., Erasmus, D., van der winemaking. Yeast, 16, 675e729. Merwe, G. K., Brumm., J., Wasserman, W. W., et al. Pretorius, I. S., & Bauer, F. F. (2002). Meeting the consumer (2008). Dynamics of the yeast transcriptome during wine challenge through genetically customized wine-yeast fermentation reveals a novel fermentation stress strains. Trends Biotechnol., 20, 426e432. response. FEMS Yeast Res., 8,35e52. Pretorius, I. S., & van der Westhuizen, T. J. (1991). The Mendes-Ferreira, A., del Olmo, M., Garcia-Martinez, J., impact of yeast genetics and recombinant DNA tech- Jimenez-Marti, E., Mendes-Faia, A., Perez-Ortin, J. E., nology on the wine industry, a review. S. Afr. J. Enol. et al. (2007). Transcriptional response of Saccharomyces Vitic., 12,3e31. cerevisiae to different nitrogen concentrations during Puig, S., & Pe´rez-Ortıń, J. E. (2000). Expression levels and alcoholic fermentation. Appl. Environ. Microbiol., 73, patterns of glycolytic yeast genes during wine fermen- 3049e3060. tation. System. Appl. Microbiol., 23, 300e303. Michnick, S., Roustan, J. L., Remize, F., Barre, P., & Puig, S., Querol, A., Ramoń, D., & Pe´rez-Ortıń, J. E. (1996). Dequin, S. (1997). Modulation of glycerol and ethanol Evaluation of the use of phase-specific gene promoters yields during alcoholic fermentation in Saccharomyces for the expression of enological enzymes in an industrial cerevisiae strains overexpressed or disrupted for GPD1 wine yeast strain. Biotechnol. Lett., 18, 887e892. encoding glycerol 3-phosphate dehydrigenase. Yeast, 13, Puig, S., Ramoń, D., & Pe´rez Ortıń, J. E. (1998). Optimized 783e793. method to obtain stable food-safe recombinant wine Mortimer, R. K., & Polsinelli, M. (1999). On the origins of yeast strains. J. Agric. Food Chem., 46, 1689e1693. wine yeast. Res. Microbiol., 150, 199e204. Querol, A., Barrio, E., Huerta, T., & Ramoń, D. (1992). Novo, M., Bigey, F., Beyne, E., Galeote, V., Gavory, F., Molecular monitoring of wine fermentations conducted Mallet, S., et al. (2009). Eukaryote-to-eukaryote gene by dry yeast strains. Appl. Environ. Microbiol., 58, transfer events revealed by the genome sequence of the 2948e2953. REFERENCES 189

Querol, A., & Bond, U. (2009). The complex and dynamic pediocin gene (pedA)inSaccharomyces cerevisiae. Yeast, 15, genomes of industrial yeasts. FEMS Microbiol. Lett., 293, 647e656. 1e10. Smit, A., Cordero, R. R., Lambrechts, M. G., Pretorius, I. S., Querol, A., Fernańdez-Espinar, M. T., del Olmo, M., & & van Rensburg, P. (2003). Enhancing voletile phenol Barrio, E. (2003). Adaptative evolution of wine yeast. Int. concentrations in wine by expressing various phenolic J. Food Microbiol., 86,3e10. acid decarboxylase genes in Saccharomyces cerevisiae. Querol, A., & Ramoń, D. (1996). The application of molec- J. Agric. Food Chem., 51, 4909e4915. ular techniques in wine microbiology. Trends Food Sci. Snow, R. (1983). Genetic improvement of wine yeast. In Technol., 7,73e78. J. F. T. Spencer, D. M. Spencer & A. R. W. Smith Quiro´s, M., Gonzalez-Ramos, D., Tabera, L., & Gonza´lez, R. (Eds.), Yeast genetics, fundamental and applied aspects (2010). A new methodology to obtain wine yeast strains (pp. 439e459). New York, NY: Springer-Verlag. overproducing mannoproteins. Int. J. Food Microbiol., Stevenson, A. C. (1985). Studies in the vegetational history 139,9e14. of S.W. Spain. II. Palynogical investigations at Laguna de Rainieri, S., & Pretorius, I. S. (2000). Selection and los Madres, Spain. J. Biogeogr., 12, 293e314. improvement of wine yeasts. Ann. Microbiol., 50,15e31. Swiegers, J. H., Capone, D. L., Pardon, K. H., Elsey, G. M., Ramoń, D., Calvo, M. D., & Peris, J. (1998). New regulation Sefton, M. A., Francis, I. L., et al. (2007). Engineering for labelling of genetically modified foods, a solution or volatile thiol release in Saccharomyces cerevisiae for a problem? Nature Biotechnol., 16, 889. improved wine aroma. Yeast, 7, 561e574. Ramoń, D., MacCabe, A. P., & Gil, J. V. (2004). Questions van Rensburg, P., Stidwell, T., Lambrechts, M. G., linger over European GM food regulations. Nature Bio- Otero, R. C., & Pretorius, I. S. (2005). Development and technol., 22, 149. assessment of a recombinant Saccharomyces cerevisiae Remize, F., Roustan, J. L., Sablayrolles, J. M., Barre, P., & wine yeast producing two aroma-enhancing beta- Dequin, S. (1999). Glycerol overproduction by engi- glucosidases encoded by the Sacharomycopsis fibuligera neered Saccharomyces cerevisiae wine yeast strains leads BGL and BGL genes. Ann. Microbiol., 55,33e42. to substantial changes in by-product formation and to van Rensburg, P., Strauss, M. L. A., Lambrechts, M. G., a stimulation of fermentation rate in stationary phase. Otero, R. R. C., & Pretorius, I. S. (2007). The heterologous Appl. Environ. Microbiol., 65, 143e149. expression of polysaccharidase-encoding genes with Riou, C., Nicaud, J. M., Barre, P., & Gaillardin, C. (1997). oenological relevance in Saccharomyces cerevisiae. J. Appl. Stationary-phase gene expression in Saccharomyces cer- Microbiol., 103, 2248e2257. evisiae during wine fermentation. Yeast, 13, 903e915. van Rensburg, P., van Zyl, W. H., & Pretorius, I. S. (1994). Rossignol, T., Dulau, L., Julien, A., & Blondin, B. (2003). Expression of the Butyrivibrio fibrisolvens endo-b-1, Genome-wide monitoring of wine yeast gene expression 4-glucanase gene together with the Erwinia pectate liase during alcoholic fermentation. Yeast, 20, 1369e1385. and polygalacturonase genes in Saccharomyces cerevisiae. Rossignol, T., Postaire, O., Storai, J., & Blondin, B. (2006). Curr. Genet., 27,17e22. Analysis of the genomic response of a wine yeast to Verstrepen,K.J.,vanLaere,S.D.M.,Vanderhaegen,B.M.P., rehydration and inoculation. Appl. Microbiol. Biotechnol., Derdelinckx, G., Dufour, J. P., Pretorius, I. P., et al. (2003). 71, 699e712. Expression levels of the yeast alcohol acetyltransferase Salek, A., Schnettler, R., & Zimmermann, U. (1990). Trans- genes ATF1, Lg-ATF1,andATF2 control the formation of a mission of killer activity into laboratory and industrial broad range of volatile esters. Appl. Environ. Microbiol., 69, strains of Saccharomyces cerevisiae by electroinjection. 5228e5237. FEMS Microbiol. Lett., 70,67e72. Vilanova, M., Blanco, P., Corte´s, S., Castro, M., Villa, T. G., & Sańchez-Torres, P., Gonza´lez-Candelas, L., & Ramoń, D. Sieiro, C. (2000). Use of a PGU1 recombinant Saccharo- (1996). Expression in a wine yeast strain of the Asper- myces cerevisiae strain in oenological fermentations. gillus niger abfB gene. FEMS Microbiol. Lett., 145, J. Appl. Microbiol., 89, 876e883. 189e194. Villanueva, A., Ramoń, D., Valle´s, S., Lluch, M. A., & Sańchez-Torres, P., Gonza´lez-Candelas, L., & Ramoń, D. MacCabe, A. P. (2000). Heterologous expression in (1998). Heterologous expression of a Candida molischiana Aspergillus nidulans of a Trichoderma longibrachiatum anthocyanin-b-glucosidase in a wine yeast strain. endoglucanase of oenological relevance. J. Agric. Food J. Agric. Food Chem., 46, 354e360. Chem., 48, 951e957. Schoeman, H., Vivier, M. A., du Toit, M., Dicks, L., & Vivier, M. A., & Pretorius, I. S. (2002). Genetically tailored Pretorius, I. S. (1999). The development of bactericidal grapevines for the wine industry. Trends Biotechnol., 20, yeast strains by expressing the Pediococcus acidilactici 472e478. 190 7. IMPROVEMENT OF WINE YEASTS BY GENETIC ENGINEERING

Volschenk, H., Viljoen, M., Grobler, J., Bauer, F., Lonvaud, A., yeast, construction and evaluation of heterothallic wine Denayrolles, M., et al. (1997a). Malolactic fermentation in strains of Saccharomyces cerevisiae, possesing minimal grape musts by a genetically engineered strain of Saccha- foreign DNA sequences. FEMS Yeast Res., 4, 339e347. romyces cerevisiae. Am. J. Enol. Vitic., 48, 193e197. World Health Organization (WHO). (1993). Health aspects Volschenk, H., Viljoen, M., Grobler, J., Petzold, B., Bauer, F., of marker genes in genetically modiﬁed plants. Geneva, Subden, R. E., et al. (1997b). Engineering pathways for Switzerland: WHO. malate degradation in Saccharomyces cerevisiae. Nature Zuzuarregui, A., Monteoliva, L., Gil, C., & del Olmo, M. L. Biotechnol., 15, 253e257. (2006). Transcriptomic and proteomic approach for Walker, M. E., Gardner, J. M., Vystavelova, A., McBryde, C., understanding the molecular basis of adaptation of de Barros Lopes, M., & Jiranek, V. (2003). Application of Saccharomyces cerevisiae to wine fermentation. Appl. the reusable, KanMX selectable marker to industrial Environ. Microbiol., 72, 836e847. CHAPTER 8

Lactic Acid Bacteria Rosario Munõz 1, M. Victoria Moreno-Arribas 2, Blanca de las Rivas 1 1 Instituto de Ciencia y Tecnologıá de los Alimentos y Nutricioń (ICTAN, CSIC), Madrid, Spain and 2 Instituto de Investigacioń en Ciencias de la Alimentacioń (CIAL, CSIC-UAM), Madrid, Spain

OUTLINE

1. General Characteristics of Lactic Acid 4.8. Protein and Peptide Degradation 205 Bacteria 192 4.9. Catabolism of Amino Acids 205 2. Identifying Lactic Acid Bacteria 193 5. Malolactic Fermentation 207 2.1. Classical Identiﬁcation Methods 193 5.1. Use of Malolactic Starter Cultures 207 2.2. Molecular Identiﬁcation Methods 195 5.2. Contribution of Malolactic Fermentation to the Sensory Properties of Wine 208 3. Population Dynamics in Lactic Acid Bacteria During Winemaking 196 6. Additional Considerations 209 6.1. Formation of Biogenic Amines 209 4. Metabolism 198 6.2. Formation of Ethyl Carbamate 4.1. Carbohydrate Metabolism 198 Precursors 211 4.1.1. Monosaccharide and Disaccharide 6.3. Stress Resistance 213 Metabolism 198 6.4. Bacteriophages 213 4.1.2. Polysaccharide Metabolism 199 6.5. Bacteriocin Production 214 4.1.3. Polyalcohol Metabolism 199 4.2. Organic Acid Metabolism 200 7. Interactions Between Lactic Acid Bacteria 4.2.1. Malic Acid Metabolism 200 and Other Microorganisms 215 4.2.2. Citric Acid Metabolism 201 8. Sensory Changes in Wine Due 4.2.3. Tartaric Acid Metabolism 202 to Lactic Acid Bacteria 216 4.3. Metabolism of Phenolic Compounds 202 8.1. Piquˆre Lactique (Lactic Taint) 216 4.4. Aldehyde Catabolism 203 8.2. Glycerol Degradation and Production 4.5. Glycoside Hydrolysis 204 of Acrolein 216 4.6. Ester Synthesis and Hydrolysis 204 8.3. Production of Extracellular 4.7. Lipid Hydrolysis 204 Polysaccharides 216

8.4. Production of Off-ﬂavors 217 8.4.2. Production of Aromatic 8.4.1. Production of Volatile Phenols 217 Heterocyclic Compounds 217 Acknowledgments 218

1. GENERAL CHARACTERISTICS were subsequently linked to the wine defect OF LACTIC ACID BACTERIA known as tourne´ or tartaric spoilage. Mu¨ ller- Thurgau (1891) and Koch (1900) later attributed Lactic fermentation is a bacterial process that the presence of lactic acid bacteria to a reduction takes place during the production of numerous in the acidity of wines and shortly afterwards, in food products. It provides the final products 1901, Seifert reported that these bacteria were with characteristic aromas and textures and plays capable of degrading malic acid. More recent a crucial role in food safety and hygiene. Among studies, particularly from the 1970s onwards, the bacteria responsible for lactic fermentation confirmed the importance of malolactic fermen- are lactic acid bacteria, which display high tation in reducing acidity (essential in red morphological and physiological diversity. The wines) and ensuring the microbiological term lactic acid bacteria emerged at the beginning stability of the final product by preventing the of the twentieth century to describe a heteroge- onset of fermentation after bottling. neous group of bacteria that are currently defined Another important and particularly relevant as spherical (cocci) or rod-shaped (bacilli), gram- line of research is the study of the negative effects positive, catalase-negative, immobile, nonsporu- that lactic acid bacteria have on the quality and lating, anaerobic, aerotolerant, and producers of composition of wine. The emergence of molec- lactic acid (the main metabolite generated during ular tools based on DNA analysis has provided the fermentation of sugars by these bacteria). greater insights into many known alterations In winemaking, lactic acid bacteria are doubly and also helped to uncover new ones. important as they can both enhance and diminish Finally, advances in recent years have led to the quality of wine. They are responsible for malo- a spectacular improvement in our understanding lactic fermentation but they can also cause of the physiology,metabolism, and genetics of the changes that adversely affect the organoleptic lactic acid bacteria involved in winemaking. properties of the final product. During malolactic Thanks to the wealth of information now avail- fermentation, the concentration of lactic acid bac- able, winemakers are better positioned to control teria reaches approximately 107 colony-forming the activity of these bacteria and to analyze and units (CFU)/mL, which gives an indication of exploit their impact on the quality of wine from their importance in winemaking. a broader, multidisciplinary perspective. In 1886, Louis Pasteur demonstrated that This chapter will review the main aspects of microbial growth was a common feature of all malolactic fermentation and the growth of lactic fermentation processes. Different types of acid bacteria in wine. Following a short review fermentation were defined according to the of the basic, practical principles underlying the predominant organic products present at the metabolism of lactic acid bacteria during malo- end of the process, and each was associated lactic fermentation, we will review key studies with a specific type of microorganism. Pasteur that have analyzed the effect of malolactic was also the first to demonstrate the presence fermentation on the organoleptic properties of of lactic acid bacteria in wines. These bacteria wine, examine certain properties of these IDENTIFYING LACTIC ACID BACTERIA 193 bacteria that are of importance to winemaking, the bacteria that perform malolactic fermentation, and, finally, discuss the main defects that the O. oeni has the greatest capacity to grow in acidic metabolism of these bacteria can cause in wine. pH and in the presence of 10% (vol/vol) ethanol (Versari et al., 1999). 2. IDENTIFYING LACTIC ACID To identify bacteria in fermented foods and BACTERIA beverages such as wine, it is first necessary to isolate them through inoculation in suitable growth media. The most common medium used Only a few species are capable of growing in to isolate lactic acid bacteria is de Man Rogosa grape must and wine because of the hostile condi- Sharpe (MRS) medium. Wibowo et al. (1985) rec- tions that they encounter: mainly low pH, a lack of ommended the addition of grape juice, tomato, nutrients, and the presence of ethanol. The main cysteine, malic acid, and several sugars to this species of lactic acid bacteria that can survive in medium. Cycloheximide (100 mg/L) and pimari- this environment are shown in Table 8.1.The cin (50 mg/L) are also needed to inhibit the species was considered to form Leuconostoc oenos growth of yeast and fungi. The presence of carbon part of the genus until as recently as Leuconostoc dioxide, in turn, favors the growth of bacteria. The 1995, when analysis of the 16S ribosomal DNA colonies are left to grow until sufficient biomass (rDNA) sequence showed that it was different for performing all the tests required is obtained. from the other members of the genus. This led Terrade et al. (2009) recently described a chemi- to the creation of a new genus, , which Oenococcus cally defined medium that satisfies the nutritional includes just two species, the malolactic Oenococ- requirements and favors the growth of lactic acid (Dicks et al., 1995)andthenonmalolactic cus oeni bacteria from wine. (Endo & Okada, 2006). Of all Oenococcus kitaharae For many years, the standard, classical methods for identifying and classifying bacteria TABLE 8.1 Lactic Acid Bacteria in Wine were based on phenotypic characteristics. Impor- Genus Metabolism of sugars Species tant advances, however, have been made in recent years thanks to the continuing developments in Pediococcus Homofermentative P. damnosus the field of molecular biology. Molecular tools P. parvulus can now be used to reveal the genetic diversity of a particular species and to study populations P. pentosaceus of microorganisms in wine without the need for Leuconostoc Heterofermentative Leu. mesenteroides prior isolation and culture. Oenococcus Heterofermentative O. oeni Lactobacillus Homofermentative L. mali 2.1. Classical Identification Methods Facultatively L. casei It is standard practice to perform Gram stain- heterofermentative ing and catalase tests following the isolation and L. plantarum purification of wine bacteria in a suitable solid Heterofermentative L. brevis medium. Bacteria that are found to be both gram-positive and catalase-negative are classified L. buchneri as lactic acid bacteria. This initial classification L. fermentum can be confirmed by growing the corresponding L. fructivorans isolates in a liquid medium containing hexoses (glucose/fructose) but not malic acid and L. hilgardii then testing for the production of lactic acid 194 8. LACTIC ACID BACTERIA using a suitable method such as paper chro- in contrast, are homofermentative or heterofer- matography. mentative; they are rod-like and arranged singly Additional tests are required to identify or in pairs or chains. Some of these morphol- isolates at the genus level. The first step involves ogies are shown in Figure 8.1. observing the bacteria under a microscope to Other tests used to confirm genus include determine whether they are spherical or rod- analysis of the lactic acid isomer formed from shaped. Another test involves the identification hexoses and detection of the hydrolysis of argi- of how lactic acid is produced. This is generally nine to ammonium or the production of performed by observing the production of gas mannitol from fructose (Pilone et al., 1991). from hexoses. The formation of gas indicates These conventional characterization tests are that the bacteria are heterofermentative (capable still used in routine practice but they can of producing multiple products from the occasionally give rise to ambiguous results, fermentation of carbohydrates) while the particularly when assessing fermentable carbo- absence of gas indicates that they are homofer- hydrates. Even in optimal conditions, these mentative (able to produce only lactic acid). carbohydrates can slowly change the color of This basic information can be used to make the medium, making it difficult to distinguish a putative identification of the genus. Oenococ- between positive and negative results. Further- cus and Leuconostoc species are heterofermenta- more, certain lactic acid bacteria often contain tive, spherical, have single cells that form pairs plasmids encoding enzymes involved in impor- or short chains, and often have an elongated, tant biochemical pathways. Because of their lenticular shape, giving them the appearance instability, particularly in the absence of selective of short bacilli. Members of the genus Pediococ- pressure, certain tests that normally yield posi- cus are also spherical and occur in pairs or tive results can give negative results. Characters tetrads; cells arranged singly or in chains are that depend on phages can also cause similar uncommon. Members of the genus Lactobacillus, problems.

(a) (b)

FIGURE 8.1 Micrographs of Oenococcus oeni (a) and Lactobacillus brevis (b) taken using low-temperature scanning electron microscopy. Scales of 5 mm are shown at the top left of both images. Images provided by Dr. A.V. Carrascosa. IDENTIFYING LACTIC ACID BACTERIA 195 2.2. Molecular Identification Methods of genes encoding 16S ribosomal RNA (rRNA) (Guerrini et al., 2003; Sato et al., 2001). It is often difficult to distinguish between Amplified rDNA restriction analysis lactic acid bacteria on the basis of physiological (ARDRA) has been used as a rapid, reliable and biochemical criteria as most have very method of identifying the main lactic acid similar nutritional and growth requirements in bacteria involved in winemaking (Rodas et al., environmental conditions (Vandamme et al., 2003; Ventura et al., 2000). This method, 1996). In recent years, however, molecular however, has certain limitations. It is not suit- biology has increasingly been used to circumvent able, for example, for comparing these difficulties. A wide variety of molecular Lactobacillus plantarum or Lactobacillus pentosus as these techniques have been used to characterize lactic display a high level of similarity in their 16S acid bacteria from wine. Examples are tech- rDNA sequence (Collins et al., 1991; Quere niques based on restriction fragment length poly- et al., 1997). Amplified rDNA fragment analysis morphisms (RFLPs) (Zapparoli et al., 2000), via PCR followed by denaturing gradient gel pulsed-field gel electrophoresis (PFGE) (Gin- electrophoresis (DGGE) has also been used to dreau et al., 1997), DNAeDNA hybridization compare diversity and monitor changes in pop- (Dicks et al., 1995; Lonvaud-Funel et al., 1991; ulations of lactic acid bacteria during the wine- Sato et al., 2001), specific DNA hybridization making process (Lo´pez et al., 2003). probes (Lonvaud-Funel et al., 1991; Sohier et al., A more recent system for identifying species 1999), polymerase chain reaction (PCR) (Groisil- of lactic acid bacteria in wine involves the use of lier & Lonvaud-Funel, 1999; Lonvaud-Funel several stages (Rodas et al., 2005). It has been et al., 1993), randomly amplified polymorphic concluded that RAPD and ribotyping are useful DNA (RAPD) (Zavaleta et al., 1997)(Figure 8.2), for identifying and classifying these bacteria, amplified fragment length polymorphisms whereas ARDRA is useful only for identification (AFLPs) (Cappello et al., 2008), and the study purposes and PFGE-RFLP is useful for distin- guishing between different strains of the same species (Rodas et al., 2003). Because O. oeni is the main species of lactic acid bacteria associated with malolactic fermentation in wines, real-time quantitative PCR methods are currently being developed to enable the rapid detection and quantification of these bacteria in samples obtained during fermentation. The main advantage of methods of this type is that they enable rapid corrective action to be taken in order to control bacterial growth (Pinzani et al., 2004). A molecular typing method that combines RAPD and multiplex PCR has been described for characterizing different strains of O. oeni during winemaking and evaluating the impact of malolactic starter cultures (Reguant & Bordons, 2003). Analysis of the population structures of O. oeni FIGURE 8.2 Genetic diversity among 10 strains of has yielded contradictory results. While molecular techniques such as DNAeDNA hybridiza- Oenococcus oeni revealed by random amplification of polymorphic DNA. tion and sequencing of the genes encoding 16S 196 8. LACTIC ACID BACTERIA and 23S rRNA and the intergenic region between species. The most abundant are L. plantarum, 16S and 23S rDNA have shown that O. oeni is Lactobacillus casei, Lactobacillus hilgardii, Leuconos- highly homogeneous, analysis of metabolic and toc mesenteroides, and Pediococcus damnosus. Less physiological characteristics, such as fatty acid common species include O. oeni and Lactobacillus profile and sugar fermentation patterns, have brevis. The numbers and proportions of species shown quite the opposite. The results of such vary according to the ripeness and condition of studies even led to a proposal to divide the species the grapes at the time of harvesting. Nonetheless, into two separate species or subspecies (Tenreiro as the must is fermented by the yeasts, there is et al., 1994). The recent use of multilocus sequence a gradual reduction in the quantity and composi- typing showed that O. oeni strains can be classi- tion of the microflora. In the case of lactic acid fied into well-differentiated groups and that bacteria, only those with the greatest resistance recombination events play an important role in to ethanol and low pH survive this stage. the genetic heterogeneity of this species (Bilhe`re Grape must contains nutrients that favor the et al., 2009; de las Rivas et al., 2004). One study growth of yeasts, which rapidly proliferate and identified regions of variability in the O. oeni initiate alcoholic fermentation. Once this occurs, genome that were the site of both recombination the bacterial population decreases to between and gene insertion/deletion (Bon et al., 2009). approximately 102 and 103 CFU/mL. The addi- This enormous variability is largely due to the tion of sulfite at concentrations typically used loss of the DNA mismatch repair genes mutS in prefermentation phases reduces the bacterial and mutL, which may also have contributed to population but does not inhibit growth, particu- this species’ greater adaptation to the conditions larly in the presence of high pH. Yeasts, which found in winemaking (Marcobal et al., 2008). are less sensitive to the effects of sulfites, grow rapidly and initiate alcoholic fermentation, which occurs practically in the absence of lactic 3. POPULATION DYNAMICS IN acid bacteria. Lactic acid bacteria have greater LACTIC ACID BACTERIA DURING difficulty growing in this environment as they WINEMAKING are less well adapted to the high sugar concentrations (>210 g/L) and low pH of the must Lactic acid bacteria are present throughout all (3.0e3.3). By the end of alcoholic fermentation, stages of winemaking. They can be isolated on the density of lactic acid bacteria will have fallen many surfaces and environments including to approximately 102 CFU/mL. There is also vine leaves, grapes, winery equipment, and a marked reduction in the variety of species barrels. A number of recent molecular identifi- present. This is logical as the metabolism of the cation studies have detected new species of yeasts responsible for alcoholic fermentation lactic acid bacteria in both musts (Lactobacillus gradually increases ethanol levels and generates bobalius [Man˜es-La´zaro et al., 2008a] and Lacto- compounds that are toxic to bacteria as well as bacillus uvarum [Man˜es-La´zaro et al., 2008b]) fatty acids and sulfur dioxide, thus altering the and wines (Lactobacillus nagelii [Edwards et al., composition of the bacterial wall (Edwards 2000], Lactobacillus vini [Rodas et al., 2006], and et al., 1990). In normal conditions, once alcoholic Lactobacillus oeni [Man˜es-La´zaro et al., 2009]). fermentation is complete, there is a lag phase The density of lactic acid bacteria in the initial lasting between 10 and 15 d during which the phases of winemaking (the must phase and onset population of lactic acid bacteria remains of alcoholic fermentation) ranges from approxi- unchanged as their growth is inhibited by the mately 103 to 104 CFU/mL. The bacteria are presence of live yeasts and inhibitory substances from a variety of mostly homofermentative secreted by these. Once this phase is complete, POPULATION DYNAMICS IN LACTIC ACID BACTERIA DURING WINEMAKING 197 the bacteria begin to multiply until they reach further development of lactic acid bacteria. a density of approximately 106 CFU/mL and Aeration of the wine and light carbon dioxide begin malolactic fermentation. This propagation pressure also favor growth. pH is one of the phase is influenced mainly by pH, temperature, main factors that affects bacterial growth in sulfur dioxide and ethanol levels. Growth is wine (Wibowo et al., 1985) and each species favored by a relatively high pH level (>3.5), has a different pH threshold for growth. a sulfur dioxide concentration of no more than Lactic, succinic, and tartaric acid all inhibit 50 mg/L, an ethanol content of 13% (vol/vol), malolactic activity in bacteria, as do high and a temperature of between 19 and 26C. Other concentrations of malic acid. Fumaric acid, in factors that favor malolactic fermentation are contrast, stimulates activity when present at prolonged contact between the wine and grape low levels but inhibits it once it reaches a level skins after alcoholic fermentation and on-lees of between 0.4 and 1.5 g/L. Fatty acids such as aging, as yeast autolysis generates nutrients decanoic acid, a product of yeast metabolism, that stimulate the growth of lactic acid bacteria have a strong inhibitory effect on malolactic (Guilloux-Benatier et al., 1993). The composition activity (Edwards & Beelman, 1987). of the bacterial population changes during this The lag phase often does not occur when the phase as strains that are better equipped to resist acidity of the grapes used to make the wine is this hostile environment are gradually selected. low or when they have a high pH. In such cases, The first species to disappear are homofermenta- lactic acid bacteria may appear before the end of tive lactic acid bacteria, followed by their hetero- alcoholic fermentation, causing what is known fermentative counterparts and Pediococcus as lactic taint, which is a considerable increase species. The dominant species at the end of alco- in the volatile acidity of the wine as the lactic holic fermentation is O. oeni. This is the species acid bacteria start to metabolize sugars at the that is best adapted to the difficult growth condi- same time as malic acid. tions (low pH and high ethanol content) that The survival of bacteria after malolactic characterize this phase (Davis et al., 1985a; van fermentation depends on the environment, Vuuren & Dicks, 1993), which explains why it is particularly on conditions such as pH, ethanol the primary species responsible for malolactic content, and, above all, sulfur dioxide levels. fermentation in most wines. Certain strains In practice, lactic acid bacteria are eliminated belonging to the Pediococcus and Lactobacillus by the addition of sulfite once malic acid trans- genera, however, can also survive in this phase. formation is complete. O. oeni bacteria disap- Malolactic fermentation starts once the bacte- pear rapidly, leaving bacteria from the genera 6 rial population reaches a density of 10 CFU/mL, Pediococcus and Lactobacillus to dominate. In following a lag phase that can last days or wines to which no sulfites are added, certain even months. During this stage, all the malic strains of lactic acid bacteria can alter the quality acid in the wine is transformed into lactic of the wine by degrading components such as acid by the bacteria present. Malolactic fer- citric acid, tartaric acid, and glycerol. This is mentation usually takes between 5 d and 2 or particularly common in low-acid wines. 3 weeks, depending on the physicochemical While only the free form of sulfur dioxide has conditions of the environment and the amount antiseptic properties against yeast in wine, all of malic acid to be transformed. This acid is forms of sulfur dioxide have activity against thought to play an important role in stimu- bacteria. The antibacterial effect of sulfur lating bacterial growth but not in forming the dioxide depends mainly on the pH of the biomass. The disappearance of malic acid wine. The levels of free sulfur dioxide required ensures microbiological stability by inhibiting to inhibit the activity of lactic acid bacteria range 198 8. LACTIC ACID BACTERIA from 10 to 20 mg/L for wines with a low pH and Homofermentative lactic acid bacteria from 20 to 40 mg/L for wines with a high pH. ferment hexoses via the Embden-Meyerhof- Coccoid species (Pediococcus, Oenococcus, and Parnas (EMP) pathway and produce two moles Leuconostoc) are less resistant than Lactobacillus of lactate and adenosine triphosphate (ATP) per species to the effect of sulfur dioxide. mole of hexose. Heterofermentative lactic acid bacteria (O. oeni, L. brevis, L. hilgardii, and Lacto- bacillus buchneri) and facultative homofermenta- 4. METABOLISM tive bacteria (L. plantarum), in contrast, ferment hexoses and pentoses via the phosphate pentose While degradation of malic and citric acid by or phosphoketolase pathway to produce one lactic acid bacteria has the greatest bearing on mole of lactate, ethanol, carbon dioxide, and the final quality of the wine, these bacteria also ATP per mole of hexose. Fructose can also serve metabolize other substrates to ensure their as an electron acceptor and is reduced to propagation. Of note among these substrates mannitol. Consequently, the acetyl phosphate are sugars, tartaric acid, glycerol, and certain formed during hexose fermentation is con- amino acids. The most importantdand desira- verted to acetate instead of being reduced to bledactivity performed by lactic acid bacteria ethanol, thus generating an additional ATP in wine is malic acid degradation; indeed, malo- molecule (Pilone et al., 1991; Salou et al., 1994). lactic fermentation is recommended only when Heterofermentative lactic acid bacteria can also the aim is to eliminate all traces of this acid use other substances such as oxygen and pyru- from the wine. Lactic acid bacteria are also vate as electron acceptors, leading to the responsible for other enzymatic activities, most production of acetate and additional ATP. of which result in changes that can detract Full-genome analysis of the O. oeni strain from or even completely spoil the quality of PSU-1 led to the identification of all of the genes the wine. encoding the pentose phosphate pathway as well as several sugar transporter systems (Mills et al., 2005). O. oeni, like other heterofermenta- 4.1. Carbohydrate Metabolism tive bacteria, is capable of converting fructose into mannitol in a reaction catalyzed by 4.1.1. Monosaccharide and Disaccharide mannitol dehydrogenase. This, however, can Metabolism cause problems in wines, as excessive mannitol Grape must contains monosaccharides, levels can lead to high concentrations of acetic disaccharides, and oligosaccharides. Monosac- acid. Nonetheless, the gene encoding malate charides include hexoses (glucose, fructose, dehydrogenase was not found in O. oeni PSU-1, galactose, and mannose) and pentoses (arabi- indicating that another dehydrogenase must nose, xylose, ribose, and rhamnose). The most be responsible for the formation of mannitol. abundant sugars are glucose and fructose. The fermentation of disaccharides by lactic Disaccharides (maltose, raffinose, and treha- acid bacteria from wine has not been studied lose) and oligosaccharides occur at lower in depth, and it is not clear whether these concentrations (Liu & Davis, 1994). It has been bacteria metabolize disaccharides by hydrolysis shown that lactic acid bacteria from wine use or by conversion to monosaccharides via the sugars as a source of carbon and energy (Davis action of hydrolases or phosphorylases. Once et al., 1986a, 1986b; Liu et al., 1995a; Salou the disaccharides have been released, the result- et al., 1994) and that they preferentially use ing monosaccharides enter the common sugar glucose and trehalose (Liu et al., 1995a). fermentation pathways. Although sugar METABOLISM 199 transporters have been identified in the O. oeni glucose-6-phosphate followed by cleavage to PSU-1 genome, no genes linked to the transport produce erythrose-4-phosphate and acetyl of sucrose, lactose, maltose, or raffinose have phosphate, reduction of erythrose-4-phosphate been found (Mills et al., 2005). to erythritol-4-phosphate, and finally hydrolysis of erythritol-4-phosphate to form erythritol. 4.1.2. Polysaccharide Metabolism Nonetheless, no specific enzymes (or the genes Polysaccharides can have a detrimental encoding them) for the conversion of eryth- effect on wine, first by increasing viscosity rose-4 and erythritol have been identified in (which requires correction by filtration) and the O. oeni genome (Zaunmu¨ ller et al., 2006). second by altering sensory properties such as The formation of polyalcohols is essentially an body, consistency, and roundness. Excessive alternative pathway for the reoxidation of concentrations of polysaccharides are thus NAD(P)H. Coenzyme A (CoA) deficiency undesirable as they cause a ropy wine, but appears to be responsible for the shift to the moderate levels have a beneficial effect on formation of erythritol, acetate, and glycerol both body and roundness. Enzymes with the from glucose in the absence of pantothenic acid, capacity to degrade polysaccharides can as evidenced by the fact that phosphotrans- improve grape must and wine by breaking acetylase and acetaldehyde dehydrogenase down the walls of plant cells and improving are limiting under conditions of pantothenic the extraction of color and aroma precursors. acid deficiency (Ritcher et al., 2001). Glycerol Not many studies, however, have focused on is a minor product of NAD(P)H reoxidation identifying this capacity in lactic acid bacteria that is formed by the reduction of glyceralde- from wine. Guilloux-Benatier et al. (2000) hyde-3-phosphate to glycerol-1-phosphate fol- demonstrated that O. oeni has extracellular lowed by dephosphorylation. The O. oeni b (1/3) glucanase activity in the stationary genome contains genes that may encode the phase of growth, providing the first evidence enzymes glycerol-1-phosphate dehydrogenase that this species has the capacity to degrade and phosphatase. Because the biochemical glucan-type polysaccharides. Increases observed reactions involved in the formation of glycerol in glucose and fructose levels during malolactic and erythritol are similar, the two compounds fermentation may thus be, at least partly, due to may be synthesized by the same enzymes. this glucanase activity. Mannitol, which is one of the predominant polyalcohols in wine, is formed by reduction 4.1.3. Polyalcohol Metabolism of fructose, as mentioned earlier. Using nuclear magnetic resonance (NMR), Some lactobacilli isolated in wine have the Veiga-da-Cunha et al. (1992) confirmed the capacity to degrade glycerol and mannitol, two synthesis of glycerol and erythritol from glucose of the most abundant polyalcohols found in in O. oeni and reported that the erythritol-to- wine. L. brevis and L. buchneri strains isolated glycerol ratio was dependent on oxygen levels. in a spoiled wine were found to metabolize glyc- Other researchers have also reported the erol in the presence of glucose or fructose, production of glycerol, erythritol, and other leading to the formation of 3-hydroxypropanal polyalcohols by O. oeni and other lactic acid (3-hydroxypropionaldehyde), which in turn is bacteria from wine (Firme et al., 1994; Liu reduced to 1,3-propanediol (Schutz & Radler, et al., 1995a). According to results published 1984a, 1984b). 3-Hydroxypropionaldehyde is by Veiga-da-Cunha et al. (1993), the pathway a precursor of acrolein, a bitter compound responsible for the production of erythritol found in alcoholic beverages such as wine and from glucose involves the isomerization of cider. 200 8. LACTIC ACID BACTERIA

Unlike glycerol, mannitol has been found to be used as the sole source of carbon and energy 4.2.1. Malic Acid Metabolism for growth by L. plantarum isolated in wine Malic acid is a major acid in wines; the (Davis et al., 1988; Liu et al., 1995a). The catabo- conversion of a dicarboxylic acid (L-malic lism of mannitol in L. plantarum, however, acid) into a monocarboxylic acid (L-lactic acid) requires the presence of either oxygen (aerobic increases pH and modifies the sensory proper- metabolism) or compounds such as citrate ties of wine. As ascertained by Seifert in 1901, and a-keto acids that can act either directly or lactic acid bacteria from wine transform L-malic indirectly as electron acceptors (anaerobic acid into L-lactic acid and carbon dioxide via metabolism) (Chen & McFeeters, 1986a, 1986b; a direct reaction, meaning that the intermediate McFeeters & Chen, 1986). pyruvic acid is not formed during this The metabolism of polyalcohols in lactic acid conversion. bacteria in wine has an important contribution The malolactic enzyme, which was purified for in winemaking. The production of polyalcohols thefirsttimeinL. plantarum (Lonvaud-Funel & can influence both the sensory quality of wine Strasser de Saad, 1982), has been found in all (e.g., body,viscosity,and roundness) and techno- species of lactic acid bacteria isolated in wine logical processes such as filtration. The forma- (Batterman & Radler, 1991; Lonvaud-Funel, tion of acrolein from glycerol can confer a bitter 1995; Naouri et al., 1990). This enzyme is taste. As far as microbiological stability is con- dimeric and formed by two identical 60 kDa cerned, L. plantarum can sometimes develop after subunits. The active form is dimeric and the malolactic fermentation due to an increase in the monomeredimer transition is pH-dependent pH of the wine and the production of mannitol (Batterman & Radler, 1991). It catalyzes a redox by O. oeni. This mannitol may then be fermented reaction involving NAD followed by NADH2. by L. plantarum, resulting in high levels of lactate The malolactic enzyme has two NAD-binding and a risk of spoilage. domains, an L-malate binding site, and an amino acid motif with a sequence that is charac- 4.2. Organic Acid Metabolism teristic of malic enzymes (Labarre et al., 1996). Many studies have analyzed the biochemical Lactic acid bacteria are capable of metabo- characteristics of the malolactic enzyme in lizing the main organic acids present in grape numerous bacterial species such as L. casei musts and wines. While they mostly act on (Battermann & Radler, 1991), L. plantarum malic and citric acid, they can also metabo- (Schu¨ zt & Radler, 1974), Leu. mesenteroides lize tartaric acid. Citric acid is only used in (Lonvaud-Funel & Strasser de Saad, 1982), and co-fermentation with hexoses, whereas malic O. oeni (Naouri et al., 1990). These have shown acid and tartaric acid can be degraded without that it functions according to an ordered sequen- a co-substrate. Many of the strains that develop tial mechanism in which the cofactors Mn2þ and after malolactic fermentation can metabolize NADþ are bound before L-malate. This activity malic and citric acid and as a result cause can also be induced by the substrate for the reac- a wide range of organoleptic changes. The tion, malic acid. Figure 8.3 shows a diagram of changes linked to the degradation of malic the mechanism underlying the generation of acid have been studied in the greatest detail metabolic energy by lactic acid bacteria during but, more recently, the metabolization of citric malolactic fermentation. acid and its association with enhanced sensory The genetic locus involved in malolactic properties have started to draw increasing conversion (mle) has been identified in O. oeni attention. and other lactic acid bacteria. In O. oeni, this METABOLISM 201

CITRIC ACID

Citrate MALATE MALATE lyase ACETIC ACID

OXALOACETIC ACID EML CO2 Oxaloacetate decarboxylase LACTATE LACTATE CO2

PYRUVIC ACID LACTIC ACID H+ H+ -Acetolactate synthase CO2

ADP -ACETOLACTIC ACID DIACETYL CHEMICAL H+ H+ OXIDATION -Acetolactate decarboxylase ATP Diacetyl CO2 reductase ACETOIN In Out

FIGURE 8.3 Mechanism of metabolic energy production in lactic acid bacteria during malolactic fermentation. 2,3 BUTANEDIOL

FIGURE 8.4 Main metabolic pathway for citric acid in locus contains the malolactic operon, which in lactic acid bacteria. turn contains three genes: mleA, which encodes the malolactic enzyme; mleP, which encodes malate-permease; and mleR, which possibly and isotope labeling. Like other lactic acid encodes the regulator responsible for activat- bacteria, O. oeni does not use citrate as a sole ing the transcription of the malolactic operon carbon source but metabolizes it together with (Labarre et al., 1996). This gene arrangement is glucose; the resulting biomass is greater than conserved in other wine-related lactic acid that produced when grown in the presence of bacteria such as L. plantarum, L. brevis, L. casei, glucose alone. After being transported to the Leu. mesenteroides, and Pediococcus pentosaceus interior of the cell, citrate is converted to (Makarova et al., 2006). a mixture of lactate, acetate, diacetyl, acetoin, and 2,3-butanediol (see Figure 8.4). The bacteria 4.2.2. Citric Acid Metabolism break down the citrate into oxaloacetic acid in Citric acid, which is one of the acids present a reaction catalyzed by citrate lyase. This acid both on grapes and in must, is generally found is converted by oxaloacetate decarboxylase to at lower concentrations (0.1e1 g/L) than those pyruvate, which is mostly reduced to lactate in of major organic acids such as tartaric acid the presence of NADH. Some pyruvate, how- (2e8 g/L) and malic acid (1e7 g/L). Wine lactic ever, is converted by acetolactate decarboxyl- acid bacteria can metabolize citrate, as shown by ase to acetolactic acid, giving rise to acetoin Ramos et al. (1995) using NRM spectroscopy and 2,3-butanediol following decarboxylation. 202 8. LACTIC ACID BACTERIA

The chemical oxidation of acetoin, in turn, 4.2.3. Tartaric Acid Metabolism yields diacetyl. The precursor of diacetyl (and Certain strains of lactic acid bacteria (particu- acetoin), a-acetolactate, is also an intermediate larly Lactobacillus strains) are also capable of in the biosynthesis of the amino acids valine degrading tartaric acid, although this capacity and leucine from pyruvate. is much less common than that of malic and cit- The co-fermentation of citrate and glucose in ric acid metabolism. Tartaric acid is only O. oeni plays an important role in the physiology degraded in certain conditions after the metabo- of these bacteria, leading to increased growth lism of other organic acids. The catabolism of rate and biomass production, which in turn this acid always alters wine by causing a slight leads to increased ATP production (Ramos & reduction in fixed acidity and an increase in Santos, 1996; Salou et al., 1994). As far as its rele- volatile acidity. L. plantarum employs a dehydra- vance to the winemaking process is concerned, tase to convert tartaric acid into oxaloacetic acid, the co-metabolism of citrate and glucose which, in turn, is decarboxylated to pyruvate. increases the formation of volatile acids Full genome sequencing of L. plantarum (acetate), which can have adverse effects on revealed genes encoding tartrate dehydratase wine aroma if excessive levels are reached. (ttdAB), oxaloacetate/malate decarboxylase The greatest impact that citrate fermentation (mae), and pyruvate dehydrogenase. L. brevis has on wine, however, is linked to the production and other heterofermentative bacteria use of diacetyl, as this imparts a buttery aroma. tartrate differently in that they mostly break it Wines that undergo malolactic fermentation down into succinic acid. The metabolism of generally have a greater concentration of diace- this acid varies from one strain of lactic acid tyl than those that do not (Martineau et al., bacteria to the next and also depends on envi- 1995). While moderate levels of diacetyl have ronmental conditions. a positive effect on aroma, high levels cause an The increase in volatile acidity in wines due unpleasant aroma, leading to spoilage (Davis to tartaric acid was described by Louis Pasteur et al., 1985a; Nielsen & Richelieu, 1999). The final and is known in winemaking as tourne´ or tarta- concentration of diacetyl in wine depends on ric spoilage. It mostly affects wines from warm various factors including bacterial strain, wine climates that have a pH of over 3.5 and a low type, sulfur dioxide, and oxygen (Martineau & level of sulfur dioxide, factors that favor the Henick-Kling, 1995; Nielsen & Richelieu, 1999). growth of certain Lactobacillus species. Indeed, Aeration, high levels of citrate and sugars, low only certain strains of are able to Lactobacillus temperature (18 C), and the elimination of yeast degrade tartaric acid. Wines affected by tartaric cells prior to malolactic fermentation all favor the spoilage generally turn cloudy, become darker, production of diacetyl (Martineau et al., 1995). and change color. This defect also causes notice- Sulfur dioxide, in turn, inhibits diacetyl produc- able alterations in aroma and flavor, creating tion and therefore reduces the impact of this organoleptically unacceptable wines. compound on aroma (Nielsen & Richelieu, 1999). Analysis of the O. oeni genome showed the 4.3. Metabolism of Phenolic presence of the typical cit gene group, which Compounds includes genes that encode citrate lyase (cit- DEF), citrate lyase ligase (citC), oxaloacetate Phenolic compounds are of enormous impor- decarboxylase (mae), and the citrate transporter tance in winemaking as they have both direct (maeP o citP)(Mills et al., 2005). The genome and indirect impacts on final quality. They are also contains genes involved in the butanediol responsible not only for the color and astrin- pathway (ilvB, alsD, butA). gency of wine but also for certain nutritional METABOLISM 203 and pharmacological properties. A recent reported that L. plantarum had an inducible review by Rodrı´guez et al. (2009) analyzed the phenolic acid reductase (also uncharacterized relationship between lactic acid bacteria and to date) that degrades phenolic acids into phenolic compounds. substituted phenylpropionic acids and converts In O. oeni, while malolactic fermentation is p-coumaric acid into phloretic acid. It has been activated in the presence of catechin or quercetin, proposed that these inducible activities may it is inhibited by increasing levels of p-coumaric participate in the phenolic acid stress response acid (Reguant et al., 2000). Gallic acid delays by reducing these acids into less toxic com- or inhibits the formation of acetic acid, mean- pounds (Gury et al., 2004). ing that greater control of malolactic fermen- L. plantarum species are the only lactic acid tation is achieved and increases in volatile bacteria to have been found to have tannase acidity are prevented. In contrast, it has been activity (Vaquero et al., 2004). The biochemical demonstrated in O. oeni that phenolic com- characterization of this enzyme in L. plantarum pounds reduce the consumption of sugars has shown that it hydrolyzes the gallic tannins and increase that of citric acid, resulting in present in wine (Curiel et al., 2009). This activity, greater concentrations of acetic acid (Roze`s thus, is very important in winemaking because et al., 2003). of its impact on color and turbidity. Several phenolic acids such as ferulic acid and p-coumaric are natural components of 4.4. Aldehyde Catabolism grape must and wine, and can be decarboxylated by numerous bacteria, including L. brevis, Wine contains volatile aldehydes that have L. plantarum, and Pediococcus species. In L. plan- an important impact on the sensory quality of tarum, this decarboxylation is accompanied by the final product (de Revel & Bertrand, 1993). the formation of volatile phenols (4-ethylphenol The most abundant of these compounds is acet- and 4-ethylguaiacol) (Cavin et al., 1993; de las aldehyde, which is mostly produced by yeasts Rivas et al., 2009). during alcoholic fermentation and can affect Phenolic acids derived from cinnamic acid aging and color stability. Furthermore, acetal- are generally esterified with tartaric acid in dehyde, hexanal, cis-hexen-3-al, and trans- grape must and wines and can be released as hexen-2-al are all responsible for unpleasant free acids through the action of esterases. odors. L. plantarum possesses two phenolic acid decar- Because aldehydes make such an important boxylases. One of these, p-coumaric acid decar- contribution to wine aroma, excess quantities boxylase (PCD), has been characterized (Cavin must be eliminated. Sulfite has traditionally et al., 1997a) and metabolizes only p-coumaric, been used for this purpose but certain lactic caffeic, and ferulic acids into their correspond- acid bacteria, in particular O. oeni, offer an altering 4-vinyl derivatives (Cavin et al., 1997b; native solution as they can metabolize acetalde- Rodrı´guez et al., 2008). In one study, the PCD hyde and convert it to ethanol and acetate gene was expressed in Saccharomyces cerevisiae (Osborne et al., 2000). Lactic acid bacteria, and to create strains capable of decarboxylating O. oeni in particular, might also be useful for phenolic acids in wines (Smit et al., 2003). metabolizing other aldehydes that can give Barthelmebs et al. (2000), on knocking out this rise to unpleasant aromas. gene, found that L. plantarum had a second While certain lactic acid bacteria are also decarboxylase that was preferentially induced able to produce acetaldehyde, it is not yet by ferulic acid. However, this enzyme has not known whether common wine strains have this yet been characterized. The same study also capacity. 204 8. LACTIC ACID BACTERIA 4.5. Glycoside Hydrolysis abundant evidence that ethyl esters such as ethyl acetate, ethyl lactate, ethyl hexanoate, When found in grapes and wine, most mono- and ethyl octanoate are formed during malo- terpenes (important aromatic compounds) and lactic fermentation (de Revel et al., 1999). It anthocyanidins (the main pigments in red would therefore appear that lactic acid bacteria grapes and red wine) are bound to sugars such from wine are able to synthesize esters, but as glucose (Ebeler, 2001). Glycosylated monoter- further studies are needed to identify and study penes are not volatile and therefore do not the enzyme systems involved. confer aroma. Aroma is released when a glycosi- Esterases are involved in both the synthesis dase such as b-glucosidase hydrolyzes the of esters and their hydrolysis in aqueous solu- sugars bound to the monoterpene and produces tion. Davis et al. (1988) showed that the majority sugar and a volatile monoterpene. In contrast, of O. oeni, Pediococcus, and Lactobacillus strains hydrolysis of the sugar component of anthocya- have esterase activity. Gaining further insight nins by glycosidase enzymes known as antho- into the production and hydrolysis of esters by cyaninases leads to the spontaneous formation lactic acid bacteria in wine will improve our of a brown or colorless compound. In other understanding of the impact of malolactic words, anthocyaninases have a decolorizing fermentation on wine aroma. activity. Certain strains of O. oeni have been reported to metabolize anthocyanins and other compounds 4.7. Lipid Hydrolysis via glycosidase activity, producing compounds Wine contains mono-, di-, and triacylglycer- that have an important impact on wine aroma ols. The lipids found in wines are derived (Bloem et al., 2008; Boido et al., 2002; de Revel from grapes or released during yeast autolysis et al., 2005; d’Incecco et al., 2004; McMahon in alcoholic fermentation (Pueyo et al., 2000). et al., 1999; Ugliano et al., 2003). Considerable These lipids can affect the flavor of wines as variations in b-glucosidase activity have been they form volatile fatty acids with a very low detected between different wine and commer- perception threshold when broken down by cial strains of O. oeni (Barbagallo et al., 2004; lipases. The fatty acids formed, additionally, Grimaldi et al., 2000, 2005), indicating that lactic can give rise to esters, ketones, and aldehydes. acid bacteria from wine have the capacity to Volatile fatty acids are natural components of hydrolyze glycoconjugates that affect aroma alcoholic beverages such as cider and wine and color. (Blanco-Gomis et al., 2001) and excessive levels can have a negative organoleptic effect. There 4.6. Ester Synthesis and Hydrolysis is no information on the lipolytic system in wine lactic acid bacteria. While a study by Davis Esters are volatile compounds that are et al. (1988) showed that certain strains of O. oeni present in wine at concentrations above the displayed esterase and/or lipase activities, perception threshold. Derived from grapes, a later study found no lipolytic activity in 32 yeast metabolism, or the esterification of alco- strains of Lactobacillus, two strains of Leuconos- hols and acids during winemaking, esters toc, and three strains of Lactococcus isolated in make a key contribution to the aroma and there- wine (Herrero et al., 1996). Because lipases are fore the quality of wine. While esters such as generally extracellular or associated with whole ethyl acetate are responsible for the fruity aroma cells, lactic acid bacteria have the potential to of wines (Ebeler, 2001), they can also have nega- modify the lipid content of the musts or wines tive effects at high concentrations. There is in which they grow. METABOLISM 205

4.8. Protein and Peptide Degradation with protease activity in O. oeni. No genes encoding a possible extracellular protease or Wine contains proteins that can be hydro- peptidase containing a clear peptide signal lyzed by bacterial proteases and peptidases to were detected in the O. oeni PSU-1 genome form peptides and amino acids that influence (Mills et al., 2005). Ritt et al. (2009), on studying the flavor and stability of wine. Because lactic the use of peptides by O. oeni by analyzing the acid bacteria need amino acids for growth, activity and biosynthesis of PepN, PepX, and they must necessarily have the corresponding PepI, found that the biosynthesis of these three enzyme activities to obtain the peptides and peptidases depended on the peptides in the amino acids they require. Analysis of the O. culture medium; they also reported that these oeni PSU-1 genome suggests that this strain peptidases, which are specific for proline-con- possesses enzymes for the biosynthesis of eight taining peptides, were important for O. oeni amino acids, namely alanine, aspartic acid, nitrogen metabolism. Finally, it has been asparagine, cysteine, glutamate, lysine, methio- reported that O. oeni has the capacity to trans- nine, and threonine. The enzymes required for port and hydrolyze oligopeptides composed of the synthesis of other amino acids such as two to five amino acids (Ritt et al., 2008). isoleucine, leucine, and valine, however, are not present in the genome (Mills et al., 2005). 4.9. Catabolism of Amino Acids Curiously, O. oeni PSU-1 lacks the ability to synthesize proline and serinedthe two most The first reactions that take place in the metab- abundant amino acids in grape mustdyet olism of amino acids are decarboxylation, trans- conserves the ability to synthesize cysteine and amination, deamination, and desulfation. The methionine, which are only present in very decarboxylation of amino acids leads to the low concentrations in this substrate. formation of carbon dioxide and amines, which Proteases and peptidases have been can have harmful health effects (see the case of described in different genera and species of biogenic amines in Section 6.1). Transamination, lactic acid bacteria commonly isolated in wine. in turn, produces amino acids and a-keto acids, While Davis et al. (1988) did not detect protease while deamination leads to the formation of activity in different strains of lactic acid bacteria ammonia and a-keto acids. Sulfur-containing from wine (including O. oeni, Pediococcus, and amino acids such as methionine and cysteine Lactobacillus strains), a later study by Rollan produce volatile sulfur compounds via desulfa- et al. (1993) detected the production of extracel- tion. Secondary reactions in amino acid catabo- lular proteases in O. oeni. These proteases have lism involve the conversion of amines, a-keto been partially characterized in this species acids, and amino acids to aldehydes. The final (Farias et al., 1996; Rollan et al., 1995a). While reactions in the transformation of amino acids lactic acid bacteria have the potential to hydro- are the reduction of aldehydes to alcohols or their lyze proteins in wine, this ability does not oxidation to acids. seem to be very common among O. oeni strains Various studies have analyzed the metabo- (Leitao et al., 2000). Manca de Nadra et al. lism of amino acids by lactic acid bacteria in (1997, 1999) demonstrated that the O. oeni strain wine, with particular emphasis on arginine, X2L, which produces an extracellular protease, histidine, methionine, ornithine, and tyrosine. is capable of releasing peptides and amino acids These metabolic processes have a major impact during malolactic fermentation in both red and on the quality of wine because of the compounds white wines. Folio et al. (2008), in turn, recently they can produce (e.g., alcohols, aldehydes, and characterized EprA, an extracellular protein amines). 206 8. LACTIC ACID BACTERIA

The amino acid composition of wine is L. brevis, L. hilgardii, L. buchneri, and P. pentosa- complex (Lehtonen, 1996). During malolactic ceus strains as well as in several strains of fermentation, for example, concentrations can L. plantarum and Leu. mesenteroides (Araque increase or decrease depending on the type of et al., 2009). The presence of these genes has amino acid involved (Davis et al., 1986a, 1986b). been associated with the capacity to degrade Arginine is the most abundant amino acid in arginine, although the amount degraded varies wine, and extensive information is available on greatly from one strain to the next (Araque its catabolism (Liu & Pilone, 1998). Lactic acid et al., 2009). Arginine degradation, for example, bacteria degrade arginine via the arginine dei- is influenced by the strain of lactic acid bacteria, minase (ADI) pathway (Liu et al., 1996). This pH, arginine concentration, and type of sugar generates energy (ATP), which favors the (Granchi et al., 1998; Liu et al., 1995b; Mira de survival and growth of these bacteria in wine Ordunã et al., 2000a, 2000b, 2001). When broken and provides them with greater viability during down, arginine secretes citrulline, which can the stationary phase of growth under anaero- subsequently be metabolized by several wine bic conditions (Tonon & Lonvaud-Funel, 2000; lactic acid bacteria (Liu et al., 1994; Mira de Tonon et al., 2001). Three enzymes, which act Ordunã et al., 2000a). sequentially, are involved in this pathway: Several strains of O. oeni can metabolize ADI, ornithine transcarbamylase (OTC), and serine with the generation of ammonia (Granchi carbamate kinase (Liu et al., 1995b). The reac- et al., 1998). The catabolism of this amino acid tions they catalyze are shown in Figure 8.5a. has not been studied in lactic acid bacteria but The ADI pathway was recently characterized it is probably degraded by deamination via the in several lactic acid bacteria (Arena et al., action of serine dehydratase, which converts 2003; Divol et al., 2003; Tonon et al., 2001). This serine to ammonia and pyruvate. Pyruvate can pathway generally involves three genes orga- be metabolized to formate, acetate, carbon nized in an operon, arcABC. The arcA gene dioxide, and ethanol or diacetyl, depending on encodes the ADI enzyme, while the arcB and the enzyme system present. O. oeni strains arcC genes encode OTC and carbamate kinase, have been found to metabolize the amino acid respectively (see Figure 8.5b). All three genes methionine, resulting in the formation of char- have been detected in the majority of O. oeni, acteristic aromas that contribute to the aromatic

(a) ADI L-argine + H2O L-citrulline + NH3 OTC L-citrulline + Pi L-ornithine + carbamyl phosphate

CK Carbamyl phosphate + ADP ATP + NH3 +CO2

(b) arcA arcB arcB ADI = arginine deiminase; CK = carbamate kinase; OTC = ornithine transcarbamylase. ADI OTC CK

FIGURE 8.5 Degradation of arginine via the arginine deiminase pathway. Reactions catalyzed by the enzymes involved in the pathway (a). Genetic organization of the arc locus in Lactobacillus hilgardii X1B (b). The arrows indicate open reading frames. The locations of a possible promoter and possible transcription termination sites are also shown. Adapted from Arena et al. (2003). MALOLACTIC FERMENTATION 207 complexity of wine; examples include methane- described reactivation protocols based on the thiol, dimethyl disulfide, 3-(methylsulfanyl) incubation of bacteria in grape must enriched propan-1-ol, and 3-(methylsulfanyl) propionic with yeast extract for 24 to 48 h followed by the acid (Pripis-Nicolau et al., 2004). The reduction inoculation of this culture as soon as possible of methanethiol is the last stage in the enzymatic after alcoholic fermentation (Lonvaud-Funel, synthesis of methanethiol from methionine. Val- 1995; Maicas, 2001; Nielsen et al., 1996). The let et al. (2009) purified the alcohol dehydroge- greatest advances achieved in this area, however, nase enzyme involved in this conversion. came with the use of O. oeni starter cultures Glutamic acid transport has also been (generally lyophilized) that were ready for direct described in O. oeni. Vasserot et al. (2003) inoculation (Maicas et al., 2000). O. oeni is still the reported that the process is energy-dependent main species used in the many commercial and can be activated by the metabolism of argi- starter cultures available today. Most of the nine and sugars and stimulated by malic acid cultures are prepared with single strains or and acidic pH. In a study of the influence of a mixture of two or three strains. For more infor- aspartic acid on growth and malic acid and mation on malolactic starter cultures, see glucose metabolism in O. oeni, Vasserot et al. Chapter 11. (2001) found that low concentrations (<0.3 The use of selected bacterial strains in starter mM) stimulated the growth of O. oeni, while cultures prevents the development of Lactoba- high concentrations (>6 mM) inhibited growth cillus and Pediococcus spoilage bacteria, which and caused a reduction in the degradation of can produce high concentrations of acetic acid malic acid and an increase in that of glucose. It and affect the quality of the wine (acetic acid is important for winemakers to determine accounts for over 90% of total volatile acidity). whether high concentrations of aspartic acid Nutrients or activators formed by inactive can lead to overproduction of acetic acid and yeasts and substances such as casein and cellu- reduced ethanol production in winemaking lose are often used to activate malolactic conditions. fermentation. These cultures contain amino As more studies appear in this area, we can acids and vitamins that function as growth assume that we will soon have a greater under- factors for lactic acid bacteria and also absorb standing of the biochemical activities involved inhibitory substances such as sulfites and in the metabolism of amino acids by lactic acid medium-chain fatty acids (Lonvaud-Funel bacteria in wine and the corresponding impact et al., 1988). on the sensory and health-related properties of A possible alternative to these activators is wine. the use of O. oeni cells immobilized in different matrices. This strategy can increase the productivity of fermentation because of the higher 5. MALOLACTIC FERMENTATION packing density and the greater protection afforded to cells. Examples of different materials 5.1. Use of Malolactic Starter Cultures used as immobilization matrices in studies analyzing the use of immobilized forms of The idea of using lactic acid bacteria to O. oeni to deacidify wine include alginates, induce malolactic fermentation was first pro- polyacrylamide, wood shavings, and cellulose posed in 1961. Studies at the time highlighted sponges (Crapisi et al., 1987; Maicas et al., the difficulties of inducing malolactic fermenta- 2001). Not all of these agents, however, have tion in wine, mostly because of the poor been accepted by winemakers as they imply viability of the inoculated bacteria. Later studies the use of additional chemical compounds. 208 8. LACTIC ACID BACTERIA

5.2. Contribution of Malolactic fermentation of citrate by lactic acid bacteria Fermentation to the Sensory (see Section 4.2.2). The reduction in vegetative, Properties of Wine grassy aromas, in turn, may be due to the catabolism of aldehydes by lactic acid bacte- In addition to reducing the total acidity of ria (see Section 4.4). Malolactic fermentation wine, malolactic fermentation modifies the or- also results in other changes, however, such ganoleptic properties of the final product as it as increased body, viscosity, and roundness converts malic aciddwhich has a bitter flavord due to the production of polyalcohols and to the smoother-tasting lactic acid. Although this polysaccharides by lactic acid bacteria (see transformation is the main reaction that occurs Section 4.1). in malolactic fermentation, it is not the only one. Other characteristic aromas associated with Recent studies have clearly demonstrated the malolactic fermentation include floral, toasty, existence of other metabolic reactions that can vanilla, sweet, wood, smoky, bitter, and honey have both positive and negative effects on the aromas (Henick-Kling, 1993; Sauvageot & Vivier, quality of wine. 1997). Further studies are required to link wine The results of studies that have analyzed the attributes that are altered during malolactic effect of malolactic fermentation on the sensory fermentation to the production or degradation properties of wine using gas chromatography of specific chemical compounds by lactic acid combined with olfactometry and mass spec- bacteria. With such information, enologists will trometry clearly indicate that malolactic fermen- be able to choose specific strains to obtain desired tation affects aroma and adds complexity to the aromas and flavors. flavor of wine (Henick-Kling, 1993; Rodrı´guez Two studies recently undertook a metabolo- et al., 1990; Sauvageot & Vivier, 1997). They mic characterization of malolactic fermentation have also shown that the effect on flavor varies (Lee et al., 2009; Son et al., 2009). Wine contains according to the strain of lactic acid bacteria metabolites produced during alcoholic fermen- and the type of wine involved. tation, malolactic fermentation, and aging that Wine acquires a new aromatic profile have an important impact on its quality. In their following malolactic fermentation, with study, Son et al. (2009) studied changes in a decrease in varietal aromas due to the degra- metabolites in wine via NMR spectroscopy dation or hydrolysis of the aromatic compounds and statistical analysis and found a clear differ- in grapes and a reduction in the number and ence between wines that underwent malolactic concentration of the volatile compounds pro- fermentation and those that did not. Specifically, duced during alcoholic fermentation. Despite they found low levels of malate and citrate and the enormous influence that malolactic fermen- high levels of lactate in the former. Also contrib- tation has on aroma, only certain changes to uting to this differentiation were metabolites wine attributes that occur at this stage are such as alanine, g-aminobutyric acid, 2,3-buta- related to the production or use of specific nediol, choline, glycerol, isoleucine, lactate, chemical compounds by lactic acid bacteria. leucine, polyphenols, proline, succinate, and According to Henick-Kling (1993), malolactic valine. Using an identical approach, Lee et al. fermentation increases the fruity, buttery aroma (2009) studied the effect of different commercial of wine but reduces vegetal, grassy notes. This strains of O. oeni on variations in metabolites increase in fruitiness is possibly caused by the during malolactic fermentation. They identified formation of esters by lactic acid bacteria (see 17 primary metabolites and 65 secondary Section 4.6) and the increased buttery aroma is metabolites of volatile compounds. The signifi- due to the formation of diacetyl from the cant differences between the wines fermented ADDITIONAL CONSIDERATIONS 209 with different O. oeni strains were determined by Lonvaud-Funel, 2001). Because different secondary rather than primary metabolites, with O. oeni and Lactobacillus strains are capable of the effects of these strains visible only in terms of decarboxylating these amino acids, it would the secondary metabolites. Twelve volatile appear that lactic acid bacteria are responsible compounds (2-butanol, butyl butyrate, diethyl for the formation of biogenic amines in wine. succinate, 2-ethyl-1-hexanol, ethyl hexanoate, The formation of these harmful amines in ethyl octanoate, 9-hexadecanoic acid, hexadeca- wine, thus, probably depends on the presence noic acid, isoamyl alcohol, isobutyric acid, octa- of lactic acid bacteria with the necessary decar- noic acid, and 2-phenylethanol) contributed to boxylation capacity. Another factor that can this differ-entiation. influence the abundance of amines in wine is the presence and concentration of precursor amino acids, which, in turn, are influenced by 6. ADDITIONAL the composition of the must, the type of vinifica- CONSIDERATIONS tion, and factors such as pH and sulfur dioxide, which influence bacterial populations and activ- This section will take an in-depth look at ities (Lonvaud-Funel, 2001). other aspects of bacteria that are relevant to Lactic acid bacteria vary in terms of their winemaking such as the formation of toxic capacity to produce biogenic amines from compounds (e.g., biogenic amines and ethyl amino acids. While some studies have indicated carbamate precursors) and factors that play an that O. oeni are the main bacterial species important role in correct malolactic fermenta- responsible for the formation of histamine tion such as stress resistance, presence of bacte- (Coton et al., 1998b; Lucas et al., 2008), others riophages, and production of bacteriocins. have found that histidine decarboxylase activity is an unstable property in O. oeni (Coton et al., 6.1. Formation of Biogenic Amines 1998b), occurring only in certain strains (Lucas et al., 2008). This would explain why not all Several lactic acid bacteria metabolize amino studies have found histamine-producing O. acids in grape must and wine to form ethyl oeni strains (Constantini et al., 2006; Moreno- carbamate precursors and biogenic amines. Arribas et al., 2003). Lucas et al. (2008) recently The impact on the quality of the wine in both reported that O. oeni strains rapidly lose their cases is important as these compounds can ability to produce histamine because this trait have harmful health effects. is encoded on an unstable 100 kb plasmid. A As discussed in Section 4.9, certain lactic acid similar finding was reported for the plasmid bacteria possess decarboxylases that convert responsible for histamine synthesis in a wine amino acids to amines and carbon dioxide. strain of L. hilgardii (Lucas et al., 2005). Histidine Some of these amines, known as biogenic decarboxylase has been purified and character- amines, are toxic substances associated with ized in the O. oeni strain 9204 (Coton et al., adverse health effects (Shalaby, 1996). They are 1998b; Rollan et al., 1995b). It consists of two found mostly in fermented food and drinks. distinct subunits, a and b, that are synthesized Wine, for example, contains as many as 25 from a single polypeptide that is subsequently different biogenic amines, the most abundant processed. The results of the above studies indi- of which are histamine, tyramine, and putres- cate that the active protein has a hexameric (ab)6 cine, which are produced by the decarboxyl- structure. This protein is a decarboxylase ation of the amino acids, histidine, tyrosine, specific for the amino acid histidine (Coton and ornithine, respectively (Lehtonen, 1996; et al., 1998a). 210 8. LACTIC ACID BACTERIA

Certain strains of L. brevis appear to be respon- Apart from the obvious health implications, sible for the formation of tyramine in wine. Tyro- biogenic amines can also have important sine decarboxylase has been biochemically and commercial repercussions as several countries genetically purified and characterized in L. brevis have established maximum limits for these IOEB 9809 (Lucas et al., 2003; Moreno-Arribas & substances in wine. In recent years, research Lonvaud-Funel, 2001).Thegeneencodingthis efforts have focused on lactic acid bacteria in an enzyme forms part of an operon composed of effort to find rapid, reliable methods for detecting four genes encoding a tyrosyl-tRNA synthetase, strains that synthesize these amines. The tradi- tyrosine decarboxylase, a probable tyrosine tional microbiological method used for this permease, and a Naþ/Kþ transporter (Lucas purpose involved the use of a culture medium et al., 2003). containing the precursor amino acid and a pH It was recently seen that O. oeni, the main indicator. With the production of amine, the species responsible for malolactic fermentation, medium would become alkaline and the pH indi- may be involved in the production of putrescine cator would change color accordingly (see in wines, with the identification of the ornithine Figure 8.7). False negative results, however, may decarboxylase gene in a putrescine-producing sometimes be obtained, as lactic acid bacteria O. oeni strain isolated in wine lees (Marcobal produce large quantities of acid. One solution to et al., 2004). The gene encodes a 745-amino- this problem was the development of modified acid protein containing conserved pyridoxal media (Bover-Cid & Holzapfel, 1999; Maijala, phosphate cofactor binding domains and amino 1993). Another simple, rapid method for avoiding acid residues involved in enzymatic activity (see false results involves thin-layer chromatography Figure 8.6). This gene does not appear to be of culture supernatants from lactic acid bacteria common in O. oeni as it was not detected in (Garcıá-Moruno et al., 2005). any of the 42 other O. oeni strains analyzed by More recently, strategies based on molecular Marcobal et al. (2004). In a later study, the same biology techniques have been designed to detect group showed that this putrescine-producing biogenic amine-producing lactic acid bacteria strain had acquired the genes regulating the (Landete et al., 2007). The most common methods synthesis of this compound by horizontal gene are based on PCR amplification as it is rapid and transfer from an unknown bacteria (Marcobal has high sensitivity and specificity. The principle et al., 2006). underlying these procedures is that all bacteria Most of the studies analyzing amine toxicity that produce biogenic amines possess the gene in humans have been performed using hista- encoding the enzyme responsible for their forma- mine. While data show that a healthy man tion. Other methods described include a multi- can consume relatively high doses of histamine plex PCR method (Marcobal et al., 2005)and (up to 2.75 mg/kg of bodyweight), there have a quantitative PCR method (Nannelli et al., been reports of histamine-induced food intoler- 2008) for the detection of lactic acid bacteria that ance leading to hypotension, digestive and produce histamine, tyramine, and putrescine liver disorders, migraine, and other disorders. (the predominant biogenic amines in wine) (see Tyramine can also cause illness due to its vaso- Figure 8.8). constrictive properties. The presence of other The aim in all cases is to design tools that amines such as putrescine and cadaverine facilitate the control of amine production in favor the passage of histamine and tyramine wine through the early detection of strains into the bloodstream as these amines inhibit with the capacity to produce biogenic amines, the activity of detoxification enzymes in the thus allowing the winemaker to take the neces- body. sary corrective measures. ADDITIONAL CONSIDERATIONS 211

FIGURE 8.6 Comparison of ornithine decarboxylase gene sequences in Oenococcus oeni (OEN) and Lactobacillus 30a (L30) generated by the ClustalW sequence alignment program. The residues involved in binding to the cofactor PLP (;) are shown in boldface in Lactobacillus 30a and are underlined in O. oeni. The dotted vertical lines indicate the separation between the different domains described for ornithine decarboxylase in Lactoba- cillus 30a. Adapted from Marcobal et al. (2004).

6.2. Formation of Ethyl Carbamate Precursors N-carbamyl group, such as urea, citrulline, or carbamyl phosphate. The most abundant precursor Ethyl carbamate is a carcinogen found in fer- is urea, which is produced by yeast during mented food and beverages such as wine alcoholic fermentation. Citrulline and carbamyl (Ough, 1976). It is formed by a chemical reaction phosphate, in turn, are produced by lactic acid between ethanol and a precursor containing an bacteria during malolactic fermentation. Both 212 8. LACTIC ACID BACTERIA

FIGURE 8.7 Detection of biogenic amine-producing bacteria (a) (b) in culture media. Lactic acid bacteria grown in the decarboxylation medium described by Maijala (1993). Solid medium (a). Liquid medium (b).

12 de Ordunã et al., 2000a). Liu et al. (1994) reported a good correlation between citrulline excretion and the formation of ethyl carbamate kb precursors during the degradation of arginine - 3.5 by O. oeni and L. buchneri. Arginine degrada- - 1.9 tion is, thus, a potential source of citrulline. ODC- To date, there have been no reports of the excretion of carbamyl phosphate during the TDC- - 0.9 degradation of arginine. Carbamyl phosphate - 0.5 is also a pyrimidine precursor and can be HDC- synthesized by certain lactic acid bacteria from glutamine, bicarbonate, and ATP (Nicoloff et al., 2001). These compounds thus all represent FIGURE 8.8 Amplification via multiplex polymerase chain reaction of decarboxylase genes in lactic acid bacteria. new sources of ethyl carbamate precursors. Amplification of a 1.4 kb fragment of the gene encoding As with biogenic amines, maximum allow- ornithine decarboxylase (ODC), a 0.9 kb fragment of the able limits have also been established for ethyl gene encoding tyrosine decarboxylase (TDC), and a 0.3 kb carbamate in different countries, reflecting the fragment of the gene encoding histidine decarboxylase importance of keeping these levels at an abso- (HDC) (lane 1). The size (in kb) of some DNA fragments l lute minimum to prevent possible health risks. digested with EcoRI and BamHI (lane 2) is also shown. Adapted from Marcobal et al. (2005). Thanks to the knowledge that has been generated in the area, today’s winemakers are better equipped than ever to take steps to prevent or of these substances are metabolic intermedi- reduce the formation of ethyl carbamate during ates in the degradation of arginine, which is the winemaking process. They can now imple- one of the predominant amino acids in wine. ment in-process controls to monitor levels and The excretion of citrulline is very common also inoculate selected strains of yeasts and during the degradation of arginine by wine lactic acid bacteria that do not produce ethyl lactic acid bacteria (Granchi et al., 1998; Mira carbamate during alcoholic or malolactic ADDITIONAL CONSIDERATIONS 213 fermentation. Araque et al. (2009) recently shown that the gene encoding thioredoxin, described a molecular method for detecting trxA, is expressed under thermal and hydrogen genes responsible for the synthesis of ethyl peroxide stress (Jobin et al., 1999a) and that the carbamate in lactic acid bacteria. It should also homologue of clpX, an ATPase regulator, is also be noted that legislation allows the use of expressed under heat shock conditions and pref- a special adjuvant consisting of an acid urease erentially during the exponential phase of isolated in Lactobacillus fermentum (and currently growth (Jobin et al., 1999b). In later studies, the sold under various trade names) in wines con- expression levels of two proteinsdthe protease taining excess levels of urea. This enzyme is FtsH (Bourdineaud et al., 2003) and the trans- active at a pH of between 3 and 4 and acts by porter OmrA (Bourdineaud et al., 2004)dwere hydrolyzing urea, thus preventing the forma- found to increase in response to stress, suggest- tion of ethyl carbamate without altering the ing their involvement in stress protection. In chemical composition of the wine. a proteomics study, Silveira et al. (2004) demonstrated an adaptive response in O. oeni to the 6.3. Stress Resistance presence of ethanol involving both cytoplasmic and membrane proteins (including those Lactic acid bacteria perform malolactic involved in cell-wall synthesis). fermentation in highly adverse conditions. Genes previously implicated in the stress Strains of L. plantarum and O. oeni display the response in O. oeni (clpX, clpLP, trxA, hsp18, greatest resistance to the pH and ethanol levels ftsH, ormA, and the operons groESL and dnaK) found in this stage of the winemaking process were also found in the fully sequenced PSU-1 (Alegrıá et al., 2004). Stress-inducing factors strain (Mills et al., 2005). As far as oxidative stress such as ethanol, acidic pH, phenolic is concerned, like all lactic acid bacteria, O. oeni is compounds, sulfur dioxide, and fatty acids in microaerophilic and does not possess catalase wine have an inhibitory effect on growth and activity. It does, however, have the genes trxA the duration of malolactic fermentation that and trxB and systems to eliminate reactive has been linked to inhibition of ATPase activity oxygen species (ROS) such as NADH-oxidase (Carrete´ et al., 2002). The expression of the malo- and NADH-peroxidase (Mills et al., 2005). No lactic operon in O. oeni appears to be regulated superoxide dismutase homologues were identi- by another factor linked to metabolic energy fied in the O. oeni PUS-1 strain (Mills et al., 2005). (Galland et al., 2003). Analysis of the complete The potential of other lactic acid bacteria to genome of the O. oeni PSU-1 strain showed the tolerate the hostile environment of winemaking presence of the full atp operon (atpBEFHAGDC), has also been studied, leading to the identification which encodes the F0-F1 ATPase system (Mills of three cold-stress genes (cspL, cspP, and cspC)in et al., 2005). Two proton-translocating ATPases L. plantarum (Derzelle et al., 2000, 2002, 2003). In involved in pH homeostasis in O. oeni have a relatively recent study, Spano et al. (2004) cloned also been identified (Fortier et al., 2003). three genes (hsp18.5, hsp19.3,andhsp18.55) Variations in membrane composition have involved in heat-stress resistance in a wine strain also been observed when cells are exposed to of L. plantarum (Spano et al., 2004, 2005). stress, with a reduction in phospholipids and up to a five-fold increase in protein content. 6.4. Bacteriophages Guzzo et al. (1997) found that O. oeni responded to stress by synthesizing six proteins. Of these, Bacteriophage infection of lactic acid bacteria the 18 kDa membrane-linked protein Hsp18 has has enormous economic repercussions in the been purified. Other studies with O. oeni have fermented food industry. The first infection of 214 8. LACTIC ACID BACTERIA this type to be reported in wine was detected by from infecting sensitive bacteria (Davis et al., electron microscopy in Switzerland in 1976 1985b; Henick-Kling et al., 1986). Based on the (Sozzi et al., 1976). Later studies isolated bacterio- above data, it is generally thought that infection phages in wines from other geographic regions by bacteriophages is not the main factor respon- such as Australia (Davis et al., 1985b), South sible for difficult malolactic fermentations. Africa (Nel et al., 1987), Germany (Arendt & Nevertheless, it is important to bear in mind Hammes, 1992), and France (Poblet-Icart et al., potential lysogeny when selecting O. oeni strains 1998). Some studies have linked difficulties for starter cultures. associated with malolactic fermentation in certain wines to the presence of high levels of 6.5. Bacteriocin Production bacteriophages and interruption of malic acid metabolism (Davis et al., 1985b; Henick-Kling The production of bacteriocins and other anti- et al., 1986). microbial compounds by bacterial strains of Poblet-Icart et al. (1998), on analyzing lysogeny enological origin is another research area that of a large number of O. oeni wine strains, found is drawing increasing attention, particularly in that 45% of the strains analyzed were lysogenic. terms of how to gain greater control over malo- This would suggest that lysogeny is common in lactic fermentation. Bacteriocins are peptide- or this species. protein-based compounds that are ribosomically Some of these bacteriophages have been synthesized and that display antimicrobial analyzed in an attempt to shed greater light on activity against genetically related strains. They aspects such as morphology, protein composition, are odorless, colorless, and nontoxic. and genome size and structure. The most common Several wine strains are capable of pro- bacteriophages in O. oeni are Siphoviridae species, ducing bacteriocins, including L. plantarum J-23 which have a hexagonal, icosahedral head and (Rojo-Bezares et al., 2007), L. plantarum J-51 a long, flexible, noncontractile tail. The diameter (Navarro et al., 2000), P. pentosaceus (Strasser de of the head ranges from 60 to 66 nm and the length Saad & Manca de Nadra, 1993), L. plantarum of the tail from 180 to 260 nm. None of the bacterio- LMG 2379 (Holo et al., 2001), L. delbrueckii subsp. phages studied in O. oeni have a collar-whisker delbrueckii, Leu. mesenteroides subsp. cremoris, and complex (Poblet-Icartetal.,1998). Lactobacillus fructivorans (Yurdugu¨ l & Bozoglu, The genome of the O. oeni PSU-1 strain does 2002). Knoll et al. (2008) recently reported antimi- not contain intact, temperate bacteriophages or crobial activity in 8% of lactic acid bacteria, large fragments of a clear phagic origin (Mills mostly L. plantarum species, isolated in wine. et al., 2005), although the strain may act as Furthermore, all the commercial malolactic a host for bacteriophages. Sao-Jose´ et al. (2004) fermentation starter cultures (containing O. oeni identified bacteriophage integration sites in the and L. plantarum strains) tested displayed O. oeni PSU-1 strain that were generally located activity against wine-related indicator strains, adjacent to transfer RNA genes. Furthermore, suggesting that they produce bacteriocins. the regions closest to these anchor sites had PCR analysis has been used to study genetic open reading frames (ORFs) of a phagic origin. variability in genes involved in the synthesis These genes may be remnants of incomplete of bacteriocins (pln genes) in L. plantarum strains excision of the phage from the PSU-1 genome. (Saénz et al., 2009). The pln locus was present in It has been shown that wine composition can 94% of L. plantarum wine strains and displays affect the infective capacity of bacteriophages. considerable plasticity, with variable regions Low pH levels and sulfur dioxide, for example, associated with the regulation of bacteriocin can inactivate these phages, preventing them production. INTERACTIONS BETWEEN LACTIC ACID BACTERIA AND OTHER MICROORGANISMS 215

Several of the bacteriocins produced by lactic inhibitors such as organic acids or activators acid bacteria isolated in grape musts and wines such as polysaccharides. Their effect varies have been characterized. L. plantarum J-51, for depending on the fungi involved and their level example, produces a heat-resistant bacteriocin of growth during the rotting of the grape. The and has broad-spectrum antibacterial activity most common fungi found in grapes are Asper- (Navarro et al., 2000). The bactericidal effects gillus, Botrytis cinerea, Mucor, Penicillium, and of pediocin N5p, produced by a strain of P. pen- Rhizopus stolonifer species. The growth of these tosaceus isolated in a wine in Argentina, have fungi generally modifies oxalic, succinic, and been studied in strains of O. oeni, P. pentosaceus, fumaric acid concentrations in must and wine, and L. hilgardii (Strasser de Saad & Manca de reducing the viability of O. oeni and slowing Nadra, 1993). Pediocin N5p is resistant to the malolactic fermentation. B. cinerea, in contrast, physicochemical factors associated with wine- increases the degradation of malic acid by making such as pH, temperature, ethanol, and causing a shift from the fermentative metabo- sulfur dioxide (Manca de Nadra et al., 1998; lism of sugars to the formation of glycerol (San Strasser de Saad et al., 1995). This bacteriocin Romao & Lafon-Lafourcade, 1979). may also be useful in controlling the growth of O. oeni and L. hilgardii are capable of using spoilage bacteria during vinification. Pediocin polysaccharides synthesized by grape fungi in PD-1, produced by a strain of P. damnosus iso- the absence of other assimilable organic mole- lated in beer (Green et al., 1997), is active against cules for growth. In wine, however, these fungal a wide range of gram-positive bacteria, polysaccharides absorb long-chain fatty acids including O. oeni. Nel et al. (2002) found that (C8 and C10), which weakens their inhibitory this pediocin was more efficient than either effect on malolactic activity. nisin or plantaricin 423 in eliminating the film The growth of acetic acid bacteria in grape formed by O. oeni in stainless steel tanks in must affects the growth and metabolism of lactic wineries. The mode of action of this bacteriocin acid bacteria, with variations according to the in metabolically active cells in O. oeni involves age of the culture and the species of acetic and the cytoplasmic membrane (Bauer et al., 2005). lactic acid bacteria present. The different species Studies have also investigated how to of lactic acid bacteria in wine also interact with improve the efficiency of malolactic fermenta- each other. In wines with a pH greater than tion using nisin and nisin-resistant strains of 3.5, for example, the growth of Pediococcus and O. oeni (Daeschel et al., 1991; Radler, 1990a, Lactobacillus bacteria, which can reach densities 8 1990b). Nisin has no effect on the organoleptic of 10 CFU/mL, leads to the death of O. oeni properties of wine and could therefore be used due to the presence of pediocins that are toxic to inhibit the growth of undesirable bacteria for gram-positive species (Strasser de Saad by adding it directly or by using bacterial strains et al., 1995). that produce it during malolactic fermentation. As mentioned earlier, the growth of yeasts responsible for alcoholic fermentation in wine inhibits bacterial growth. In other cases, 7. INTERACTIONS BETWEEN however, yeasts can stimulate the growth of LACTIC ACID BACTERIA lactic acid bacteria through the release of nutri- AND OTHER MICROORGANISMS ents such as vitamins and amino acids into the medium. The type of strain that conducts alco- The development of fungi in grapes plays holic fermentation thus has a considerable influ- a very important role in the onset of malolactic ence on malolactic fermentation. Osborne and fermentation because these fungi can generate Edwards (2007) described S. cerevisiae strains 216 8. LACTIC ACID BACTERIA that produced a 5.9 kDa peptide responsible for fermentation. Lactic acid bacteria that appear inhibiting O. oeni growth during malolactic before all the sugar in the must has been trans- fermentation. It can thus be concluded that the formed into ethanol convert hexoses to acetic interactions between yeasts and lactic acid acid as well as to ethanol and carbon dioxide bacteria are complex and greatly depend on (which are also produced by yeasts). The pres- the strains present (Avedovech et al., 1992; ence of acetic acid and excessive amounts of Wibowo et al., 1985). lactic acid in the medium results in a considerable increase in volatile acidity (Lonvaud-Funel, 1999). The D-isomer of lactic acid has been asso- 8. SENSORY CHANGES IN WINE ciated with piquˆre lactique in wine while the DUE TO LACTIC ACID BACTERIA L-isomer is produced by malolactic fermentation. Most of the lactic acid bacteria associated In most cases, malolactic fermentation is prop- with this flaw belong to the species L. hilgardii erly controlled and the propagation of lactic acid or L. fructivorans. bacteria and the biochemical reactions they participate in contribute to improving both the 8.2. Glycerol Degradation and quality and stability of wine. There are, however, Production of Acrolein certain species and strains of lactic acid bacteria that can reduce the quality and acceptability of Several lactic acid bacteria convert glycerol wine and, at times, even make it unfit for into 3-hydroxypropionaldehyde via the activity consumption (Bartowsky & Henschke, 2004; of glycerol dehydratase. This reaction generates Bartowsky, 2009). Although some of these alter- acrolein. Alone, this compound is not problem- ations, or defects, have been known for a long atic, but when it reacts with certain groups of time (such as glycerol degradation or piquˆre lac- phenolic compounds such as tannins it can tique), great advances have been made in our cause bitter flavors. Glycerol is one of the most understanding of these processes with the emer- abundant compounds in wine and is generally gence of new molecular techniques. More recent found at concentrations of 5 to 8 g/L. It is one studies have described “newer” defects such as of the main products of yeast metabolism and undesirable odors caused by the production of plays a key role in wine flavor. The metabolism volatile phenols or aromatic heterocyclic bases, of glycerol thus affects the quality of wine, not but much has still be learned about these. Of only because it reduces glycerol levels but also particular interest in recent years has been the because of the metabolic products it generates. analysis of alterations associated with the metab- Strains of lactic acid bacteria capable of degrad- olism of amino acids in lactic acid bacteria that ing glycerol can be detected in wine using a can have important health repercussions such as special PCR-based molecular method (Claisse & the case of ethyl carbamate and biogenic amines. Lonvaud-Funel, 2001).

8.1. Piquˆre Lactique (Lactic Taint) 8.3. Production of Extracellular Polysaccharides Piquˆre lactique or lactic taint is one of the most common wine flaws and is therefore among the Several strains of lactic acid bacteria can best studied. It can occur during the production synthesize extracellular polysaccharides (exo- or even the storage of wine. It is typically asso- polysaccharides or EPSs) from residual sugars, ciated with conditions that favor bacterial detracting from the quality of the wine. Such growth such as stuck or incomplete alcoholic wines are characterized by abnormal viscosity. SENSORY CHANGES IN WINE DUE TO LACTIC ACID BACTERIA 217 Although this problem can occur during 8.4. Production of Off-flavors production, in most cases it develops gradually and appears weeks or even months after the 8.4.1. Production of Volatile Phenols wine has been bottled. Lactic acid bacteria are responsible for Different species of lactic acid bacteria can a variety of off-flavors in wine, including produce EPSs. Llaube`res et al. (1990) found that animal-like odors attributed to excessive levels Pediococcus and Lactobacillus strains isolated of volatile phenols. The main volatile phenols from spoiled wine and cider produced an iden- in red wines are 4-ethylphenol, 4-ethylguaiacol, tical EPS (a D-glucan consisting of a trisaccharide 4-vinylphenol, and 4-vinyl guaiacol. The origin repeating unit of D-glucose attached by (1/3) of ethylphenols has been a topic of debate for bonds and side branches of D-glucose with many years. While Brettanomyces/Dekkera yeasts (1/2) bonds. In Pediococcus parvulus 2.6 (for- with cinnamate decarboxylase and vinylphenol merly P. damnosus 2.6), P. damnosus IOEB8801, reductase activities are the main species respon- and Lactobacillus diolivorans G77 strains isolated sible for the biosynthesis of these phenols, in cider and wine, the production of EPSs has certain Pediococcus and Lactobacillus strains also been associated with the presence of plasmids have a role (Cavin et al., 1993; Chatonnet et al., (Walling et al., 2001; Werning et al., 2006). The 1995). In a recent study, de las Rivas et al. gene responsible for the production of the EPS (2009) analyzed the capacity of lactic acid in O. oeni, however, appears to be chromosomal bacteria to produce volatile phenols in wine (Dols-Lafargue et al., 2008; Werning et al., 2006). and described a PCR method for detecting The EPS-producing Lactobacillus collinoides bacteria with this potential. L. plantarum, L. bre- IOEB0203 and L. hilgardii IOEB0204 strains, vis, and P. pentosaceus strains produced vinyl however, do not appear to contain sequences derivatives from hydroxycinnamic acids, but similar to the glycosyltransferases responsible only L. plantarum strains produced the corre- for the production of EPSs in the strains discussed sponding ethyl derivatives. O. oeni, L. hilgardii, above (Walling et al., 2005). and Leu. mesenteroides strains, in contrast, did Molecular techniques have been developed not decarboxylate the hydroxycinnamic acids to detect the presence of EPS-producing strains p-coumaric and ferulic acids, meaning that that can alter wine during production (Gin- they are not responsible for the production of dreau et al., 2001; Walling et al., 2004). Our volatile phenols. knowledge of EPS-producing lactic acid bacteria is far from complete. An improved under- 8.4.2. Production of Aromatic standing of how these bacteria behave in wine Heterocyclic Compounds and interact with other microorganisms The production of undesirable aromas together with greater knowledge of the factors and flavors in wine described as “mousy” or involved in the synthesis of polysaccharides “acetamide” has been associated with several will help winemakers to predict their growth. lactic acid bacteria (Costello et al., 2001). A Extensive filtration or heat-treatment methods mousy odor or flavor is specifically attributed are necessary to eliminate these bacteria prior to the production of three volatile hetero- to bottling. It should also be noted that these cyclic compounds: 2-ethyltetrahydropyridine, bacteria are highly tolerant of hostile conditions 2-acetyltetrahydopyridine, and 2-acetylpyrro- and sulfur dioxide, as EPSs exert a protective line. Certain winemaking conditions such as effect on the cell. The most important measure high pH (>3.5) or low sulfur dioxide levels for preventing subsequent contamination, is, can favor the growth of the bacterial strains thus, rigorous cleaning of winery surfaces. involved in the production of these bases 218 8. LACTIC ACID BACTERIA

(Snowdon et al., 2006). A mousy taint can render lactic acid bacterium Lactobacillus hilgardii X1B: Struc- a wine unpalatable and cannot be eliminated. tural and functional study of the arcABC genes. Gene, This flaw has been associated with heterofer- 301,61e66. Arendt, E. K., & Hammes, W. P. (1992). Isolation and char- mentative lactic acid bacteria (most often via acterization of Leuconostoc oenos phages from German the production of N-heterocycles by heterofer- wines. Appl. Microbiol. Biotechnol., 37, 643e646. mentative strains of Lactobacillus and L. hilgardii Avedovech, R. M., McDavid, M. R., Watson, B. T., & Sandine, W. E. (1992). An evaluation of combinations of in particular, followed by O. oeni and Pediococcus strains) and homofermentative wine yeast and Leconoctoc oenos strains in malolactic Lactobacillus fermentation of Chardonnay wine. Am. J. Enol Vitic., 43, species (Snowdon et al., 2006). 256e260. Very few studies have analyzed the origin of Barbagallo., R. N., Spagan, G., Palmeri, R., & Torriani, S. this flaw and little is known about the extent to (2004). Assessment of beta-glucosidase activity in which it affects the quality of the wine, mostly selected wild strains of Oenococcus oeni for malolactic because of the complex nature of the processes fermentation. Enz. Microb. Technol., 34, 292e296. Barthelmebs, L., Divie`s, C., & Cavin, J.-F. (2000). Knockout involved but also because it occurs in conjunc- of the p-coumarate decarboxylase gene from Lactobacillus tion with other defects. The presence of D-fruc- plantarum reveals the existence of two other inducible tose, a fermentable sugar, has been associated enzymatic activities involved in phenolic acid metabo- with the production of volatile heterocyclic lism. Appl. Environ. Microbiol., 66, 3368e3375. compounds and it has been suggested that the Bartowsky, E. J. (2009). Bacterial spoilage of wine and approaches to minimize it. Lett. Appl. Microbiol., 48, formation of these compounds involves orni- 149e156. thine and lysine metabolism in the presence of Bartowsky, E. J., & Henschke, P. A. (2004). The “buttery” ethanol, although much remains to be discov- attribute of wine-diacetyl-desirability, spoilage and ered regarding the mechanisms underlying beyond. Int. J. Food Microbiol., 96, 235e252. this process (Costello & Henschke, 2002). Battermann, G., & Radler, F. (1991). A comparative study of malolactic enzyme and malic enzyme of different lactic acid bacteria. Can. J. Microbiol., 37,211e217. Acknowledgments Bauer, R., Chikindas, M. L., & Dicks, L. M. T. (2005). Puri- fication, partial amino acid sequence and mode of action We thank the Spanish Ministry of Science and Innovation of pediocin PD-1, a bacteriocin produced by Pediococcus (MICINN) (grants AGL2005-00470, AGL2008-01052, Consol- damnosus NCFB 1832. Int. J. Food. Microbiol., 101,17e27. ider INGENIO 2010 CSD2007-00063 FUN-C-FOOD), the Bilhe`re, E., Lucas, P. M., Claisse, O., & Lonvaud-Funel, A. Autonomous Community of Madrid (CAM) (S 2009/AGR- (2009). Multilocus sequences typing of Oenococcus oeni: 1469), Spanish Council for Scientific Research (CSIC) Detection of two subpopulations shaped by intergenic 2009201155, and the Spanish National Institute of Agricul- recombination. Appl. Environ. Microbiol., 75, 1291e1300. tural and Food Research (INIA) RM2008-00002 for the finan- Blanco-Gomis, D., Mangas Alonso, J.-J., Cabrales, I. M., & cial support provided. Abrodo, P. A. (2001). Gas chromatographic of total fatty acids in cider. J. Agric. Food Chem., 49, 1260e1263. References Bloem, A., Lonvaud-Funel, A., & de Revel, G. (2008). Hydrolysis of glycosidically bound flavour compounds Alegrıá, E. G., Lo´pez, I., Ruiz, J. I., Saénz, J., Fernańdez, E., from oak wood by Oenococcus oeni. Food Microbiol., 25, Zarazaga, M., et al. (2004). High tolerance of wild 99e104. Lactobacillus plantarum and Oenococcus oeni strains to Boido, E., Lloret, A., Medina, K., Carrau, F., & Dellacassa, E. lyophilisation and stress environmental conditions of (2002). Effect of b-glycosidase activity of Oenococcus oeni acid pH and etanol. FEMS Microbiol. Lett., 230,53e61. on the glycosilated flavour precursors of Tannat wine Araque, I., Gil, J., Carrete´, R., Bordons, A., & Reguant, C. during malolactic fermentation. J. Agric. Food Chem., 50, (2009). Detection of arc genes related with the ethyl 2344e2349. carbamate precursors in wine lactic acid bacteria. Bon, E., Delaherche, A., Bilhe`re, E., de Daruvar, A., Lonvaud- J. Agric. Food Chem., 57, 1841e1847. Funel, A., & Le Marrec, C. (2009). Oenococcus oeni genome Arena, M. E., Manca de Nadra, M. C., & Mun˜ oz, R. plasticity is associated with fitness. Appl. Environ. Micro- (2003). The arginine deiminase pathway in the wine biol., 75, 2079e2090. REFERENCES 219

Bourdineaud, J. P., Nehme´, B., Tesse, S., & Lonvaud- Collins, M. D., Rodrigues, U. M., Aguirre, M., Funel, A. (2003). The ftsH gene of the wine bacterium Farrow, J. A. E., Martinez-Murcia, A., Phillips, B. A., Oenococcus oeni is involved in protection against envi- et al. (1991). Phylogenetic analysis of the genus Lactoba- ronmental stress. Appl. Environ. Microbiol., 69, cillus and related lactic acid bacteria as determined by 2512e2520. reverse transcriptase sequencing of 16S rRNA. FEMS Bourdineaud, J. P., Nehme´, B., Tesse, S., & Lonvaud- Microbiol. Lett., 123, 241e248. Funel, A. (2004). A bacterial gene homologous to ABC Constantini, A., Cercosimo, M., del Prete, V., & Garcıá- transporters protects Oenococcus oeni from ethanol and Moruno, E. (2006). Production of biogenic amine by other stress factors in wine. Int. J. Food Microbiol., 92, lactic acid bacteria, screening by PCR, thin layer chro- 1e14. matography, and high-performance liquid chromatog- Bover-Cid, S., & Holzapfel, W. P. (1999). Improved screening raphy of strains isolated from wine and must. J. Food procedure for biogenic amine production by lactic acid Prot., 69, 391e396. bacteria. Int. J. Food Microbiol., 53,33e41. Costello, P. J., & Henschke, P. A. (2002). Mousy off-flavor Cappello, M. S., Stefani, D., Grieco, F., Logrieco, A., & of wine: Precursors and byosinthesis of the causative Zapparoli, G. (2008). Genotyping by amplified frag- N-heterocycles 2-ethyltetrahydropyridine, 2-acetylte- ment length polymorphism and malate metabolism trahydropyridine, and 2-acetyl-1-pirroline by Lactoba- performances of indigenous Oenococcus oeni strains cillus hilgardii DSM 20176. J. Agric. Food Chem., 50, isolated from Primitivo wine. Int. J. Food Microbiol., 127, 7079e7087. 241e245. Costello, P. J., Lee, T. H., & Henschke, P. A. (2001). Ability of Carrete´, R., Vidal, M. T., Bordons, A., & Constantı´, M. (2002). lactic acid bacteria to produce N-heterocycles causing Inhibitory effect of sulfur dioxide and other stress mousy off-flavour in wine. Aus. J. Grape Wine Res., 7, compounds in wine on the ATPase activity of Oenococcus 160e167. oeni. FEMS Microbiol. Lett., 211, 155e159. Coton, E., Rollan, G., Bertrand, A., & Lonvaud-Funel, A. Cavin, J.-F., Andioc, V., Etievant, P. X., & Divie`s, C. (1993a). (1998b). Histamine-producing lactic acid bacteria in Ability of wine lactic acid bacteria to metabolize phenol wines: Early detection, frequency and distribution. Am. carboxylic acids. Am. J. Enol. Vitic., 44,76e80. J. Enol. Vitic., 49, 199e204. Cavin, J.-F., Barthelmebs, L., & Divie`s, C. (1997a). Molecular Coton, E., Rollan, G. C., & Lonvaud-Funel, A. (1998a). characterization of an inducible p-coumaric acid decar- Histidine carboxylase of Leuconostoc oenos 9204: Purifi- boxylase from Lactobacillus plantarum: Gene cloning, cation, kinetic properties, cloning and nucleotide transcriptional analysis, overexpression in Escherichia sequence of the hdc gene. J. Appl. Microbiol., 84, 143e151. coli, purification, and characterization. Appl. Environ. Crapisi, A., Spettoli, P., Nuti, M. P., & Zamorani, A. (1987). Microbiol., 63, 1939e1944. Comparative traits of Lactobacillus brevis, Lactobacillus Cavin, J.-F., Barthelmebs, L., Guzzo, J., van Beeumen, J., fructivorans and Leuconostoc oenos immobilized cells for Samyn, B., Travers, J.-F., et al. (1997b). Purification and the control of malo-lactic fermentation in wine. J. Appl. characterization of an inducible p-coumaric acid decar- Bacteriol., 61, 145e151. boxylase from Lactobacillus plantarum. FEMS Microbiol. Curiel, J. A., Rodrı´guez, H., Acebroń, I., Manchenõ, J. M., de Lett., 147, 291e295. las Rivas, B., & Munõz, R. (2009). Production and Chatonnet, P., Dubourdieu, D., & Boidron, J. N. (1995). The physicochemical properties of recombinant Lactobacillus influence of Brettanomyces/Dekkera sp. yeasts and lactic plantarum tannase. J. Agric. Food Chem., 57, 6224e6230. acid bacteria on the ethylphenol content of red wine. d’Incecco, N., Bartowsky, E., Kassara, S., Lante, A., Spettoli, Am. J. Enol. Vitic., 46, 463e468. P., Henschke, P. (2004). Release of glycosidically bound Chen, K. H., & McFeeters, R. F. (1986a). Utilization of elec- flavour compounds of Chardonnay by Oenococcus oeni tron acceptors for anaerobic metabolism by Lactobacillus during malolactic fermentation. Food Microbiol., 21, plantarum. Enzymes and intermediates in the utilization 257e265. of citrate. Food Microbiol., 3,83e92. Daeschel, M. A., Jung, D.-S., & Watson, B. T. (1991). Chen, K. H., & McFeeters, R. F. (1986b). Utilization of elec- Controlling wine malolactic fermentation with nisin and tron acceptors for anaerobic mannitol metabolism by nisin-resistant strains of Leuconostoc oenos. Appl. Environ. Lactobacillus plantarum. Reduction of alpha-keto acids. Microbiol., 57, 601e603. Food Microbiol., 3,93e99. Davis, C. R., Wibowo, D., Eschenbruch, R., Lee, T. H., & Claisse, O., & Lonvaud-Funel, A. (2001). De´tection de bacter- Fleet, G. H. (1985a). Practical implications of malolactic ies lactiques produisant du 3-hydroxypropionalde´hyde fermentation. A review. Am. J. Enol. Vitic., 36, 290e301. (prećurseur d’acroleíne) a` partir du glyce´rol par tests Davis, C. R., Wibowo, D., Fleet, G., & Lee, T. H. (1986a). molećulaires. Lait, 81,173e181. Growth and metabolism of lactic acid bacteria during 220 8. LACTIC ACID BACTERIA

and after malolactic fermentation of wines at different Dols-Lafargue, M., Lee, H. Y., Le Marrec, C., Heyraud, A., pH. Appl. Environ. Microbiol., 51, 539e545. Chambat, G., & Lonvaud-Funel, A. (2008). Character- Davis, C. R., Wibowo, D., Fleet, G., & Lee, T. H. (1986b). ization of gft, a glucosyltransferase gene in the genomes Growth and metabolism of lactic acid bacteria during of Pediococcus parvulus and Oneococcus oeni, two bacterial fermentation and conservation of some Australian species commonly found in wine. Appl. Environ. Micro- wines. Food Technol. Austr., 38,35e40. biol., 74, 4079e4090. Davis, C. R., Wibowo, D., Fleet, G. H., & Lee, T. H. (1988). Ebeler, S. E. (2001). Analytical chemistry: Unlocking the Properties of wine lactic acid bacteria: Their potential secrets of wine flavor. Food Rev. Int., 17,45e64. oenological significance. Am. J. Enol. Vitic., 39, 137e142. Edwards, C. G., & Beelman, R. B. (1987). Inhibition of the Davis, G., Silveira, N. F. A., & Fleet, G. H. (1985b). Occur- malolactic bacterium Leuconostoc oenos (PSU-1), by dec- rence and properties of bacteriophages of Leuconostoc anoic acid and subsequent removal of the inhibition by oenos in Australian wines. Appl. Environ. Microbiol., 50, yeast ghosts. Am. J. Enol. Vitic., 38, 239e242. 872e876. Edwards, C. G., Beelman, R. B., Bartley,C. E., & Mc Connell, A. de las Rivas, B., Marcobal, A., & Munõz, R. (2004). Allelic (1990). Production of decanoic acid and other volatile diversity and population structure in Oenococcus oeni as compounds and the growth of yeast and malolactic determined from sequence analysis of housekeeping bacteria during vinification. Am. J. Enol. Vitic., 41,48e56. genes. Appl. Environ. Microbiol., 70, 7210e7219. Edwards, C. G., Collins, M. D., Lawson, P. A., & de las Rivas, B., Rodrı´guez, H., Curiel, J. A., Landete, J. M., Rodrı´guez, A. V. (2000). Lactobacillus nagelii sp. nov., and & Munõz, R. (2009). Molecular screening of wine lactic organism isolated from a partially fermented wine. Int. J. acid bacteria degrading hydroxycinnamic acids. J. Agric. Syst. Evol. Microbiol., 50, 699e702. Food Chem., 57, 490e494. Endo, A., & Okada, S. (2006). Oenococcus kitaharae sp. nov., de Revel, G., & Bertrand, A. (1993). A method for the a non-acidophilic and non-malolactic-fermenting oeno- detection of carbonyl compounds in wine: Glyoxal and coccus isolated from a composting distilled shochu methylglyoxal. J. Sci. Food Agric., 61, 267e272. residue. Int. J. Syst. Evol. Microbiol., 56, 2345e2348. de Revel, G., Bloem, A., Augustin, M., Lonvaud-Funel, A., & Farias, M. E., Rollan, G. C., & Manca de Nadra, M. C. (1996). Bertrand, A. (2005). Interaction of Oenococcus oeni and Influence of nutritional factors on the protease produc- oak wood compounds. Food Microbiol., 22, 569e575. tion by Leuconostoc oenos from wine. J. Appl. Bacteriol., 81, de Revel, G., Martıń, N., Pripis-Nicolau, L., Lonvaud- 398e402. Funel, A., & Bertrand, A. (1999). Contribution to the Firme, M. P., Leitao, M. C., & Sanromao, M. V. (1994). The knowledge of malolactic fermentation influence on wine metabolism of sugar and malic acid by Leuconostoc oneos. aroma. J. Agric. Food Chem., 47, 4003e4008. Effect of malic acid, pH and aeration conditions. J. Appl. Derzelle., S., Hallet, B., Ferrain, T., Delcour, J., & Hols, P. Bacteriol., 76, 173e181. (2002). Cold shock induction of the cspL gene in Lacto- Firme, M. P., Leitao, M. C., & San Romaò, M. V. (1996). The bacillus plantarum involves transcriptional regulation. J. metabolism of sugar and malic acid by Leuconostoc oenos: Bacteriol., 184, 5518e5523. Effect of malic acid, pH and aireation conditions. J. Appl. Derzelle., S., Hallet, B., Ferrain, T., Delcour, J., & Hols, P. Bacteriol., 76, 173e181. (2003). Improved adaptation to cold shock, stationary- Folio, P., Ritt, J.-F., Alexandre, H., & Remize, F. (2008). phase, and freezing stresses in Lactobacillus plantarum Characterization of EprA, a major extracellular protein overproducing cold shock proteins. Appl. Environ. of Oenococcus oeni with protease activity. Int. J. Food Microbiol., 69, 4285e4290. Microbiol., 127,26e31. Derzelle., S., Hallet, B., Francis, K. P.,Ferrain, T., Delcour, J., & Fortier, L. C., Tourdot-Marechal, R., Divies, C., Lee, B. H., & þ Hols, P. (2000). Changes in cspL, cspP, and cspC Guzzo, J. (2003). Induction of Oenococcus oeni H -ATPase mRNA abundance as a function of cold shock and growth activity and mRNA trasncription under acidic con- phase in Lactobacillus plantarum. J. Bacteriol., 182, dictions. FEMS Microbiol. Lett., 222, 165e169. 5105e5113. Galland, D., Tourdot-Marećhal, R., Abraham, M., Chu, K. S., & Dicks, L. M. T., Dellaglio, F., & Collins, M. D. (1995). Guzzo, J. (2003). Absence of malolactic activity is a charac- þ Proposal to reclassify Leuconostoc oenos as Oenococcus oeni teristic of H -ATPase-deficient mutants of the lactic acid [corrig.] gen.nov., comb. nov. Int. J. System. Bacteriol., 45, bacterium Oenococcus oeni. Appl. Environ. Microbiol., 69, 395e397. 1973e1979. Divol, B., Tonon, T., Morichon, S., Gindreau, E., & Lonvaud- Garcıá-Moruno, E., Carrascosa, A. V., & Munõz, R. (2005). A Funel, A. (2003). Molecular characterization of Oeno- rapid and inexpensive method for the determination of coccus oeni genes encoding proteins involved in arginine biogenic amines from bacterial cultures by thin-layer transport. J. Appl. Microbiol., 94, 738e746. chromatography. J. Food. Prot., 68, 625e629. REFERENCES 221

Gindreau, E., Joyeux, A., de Revel, G., Claisse, O., & Lon- Henick-Kling, T., Lee, T. H., & Nicholas, D. J. D. (1986). vaud-Funel, A. (1997). Evaluation de l’e´tablissement des Characterization of the lytic activity of bacteriophages of levains malolactiques au sein de la microflore bacte´ri- Leuconostoc oenos isolated from wine. J. Appl. Bacteriol., enne indigene. J. Int. Sci. Vigne Vin, 31, 197e202. 61, 525e534. Gindreau, E., Wailling, E., & Lonvaud-Funel, A. (2001). Herrero, M., Mayo, B., Gonzalez, B., & Sua´rez, U. E. (1996). Direct polymerase chain reaction detection of ropy Ped- Evaluation of technologically important traits in lactic iococcus damnosus strains in wine. J. Appl. Microbiol., 90, acid bacteria isolated from spontaneosus fermentations. 535e542. J. Appl. Bacteriol., 81, 565e570. Granchi, L., Paperi, R., Rosellini, D., & Vincenzini, M. (1998). Holo, H., Jeknic, Z., Daeschel, M., Stevanovic, S., & Nes, I. F. Strain variation of arginine catabolism among malolactic (2001). Plantaricin W from Lactobacillus plantarum Oenococcus oeni strains of wine origin. Ital. J. Food. Sci., belongs to a new family of two-peptide lantibiotics. 10, 351e357. Microbiology, 147, 643e651. Green, G., Dicks, L. M. T., Bruggeman, G., Vandamme, E. J., Jobin, M. P., Garmyn, D., Divies, C., & Guzzo, J. (1999a). & Chikindas, M. L. (1997). Pediocin PD-1, a bactericidal Expression of the Oenococcus oeni trxA gene is induced antimicrobial peptide from Pediococcus damnosus NCFB by hydrogen peroxide and heat shock. Microbiology, 145, 1832. J. Appl. Microbiol., 83,13e127. 1245e1251. Grimaldi, A., Bartowski, E., & Jiranek, V. (2005). A survey of Jobin, M. P., Garmyn, D., Divies, C., & Guzzo, J. (1999b). The glycosidase activities of commercial wine strains of Oenococcus oeni clpX homologue is a heat shock gene Oenococcus oeni. Int. J. Food Microbiol., 105, 233e244. preferentially expressed in exponential growth phase. Grimaldi, A., McLean, H., & Jiranek, V. (2000). Identification J. Bacteriol., 181, 6634e6641. and partial characterization of glycosidic activities of Knoll, C., Divol, B., & du Toit, M. (2008). Genetic screening of commercial strains of the lactic acid bacterium Oeno- lactic acid bacteria of oenological origin for bacteriocin- coccus oeni. Am. J. Enol. Vitic., 51, 362e369. encoding genes. Food Microbiol., 25, 983e991. Groisillier, A., & Lonvaud-Funel, A. (1999). Comparison of Koch, A. (1900). Uber die ursachem des verschwidens des partial malolactic enzyme gene sequences for phyloge- saure bei garung und lagerrung des weines. Weibau netic analysis of some lactic acid bacteria species and Weinhandel, 18, 395e419. relationships with the malic enzyme. Int. J. Syst. Bacter- Labarre, C., Guzzo, J., Cervin, J. F., & Divies, C. (1996). iol., 49, 1417e1428. Cloning and characterization of the genes encoding Guerrini, S., Bastianini, A., Blaiotta, G., Granchi, L., the malolactic enzyme and the malate permease of Moschetti, G., Coppola, S., Romano, P., & Vincenzini, M. Leuconostoc oenos. Appl. Environ. Microbiol., 62, (2003). Phenotypic and genotypic characterization of 1274e1282. Oenococcus oeni strains isolated from Italian wines. Int. Landete, J. M., de las Rivas, B., Marcobal, A., & Munõz, R. J. Food Microbiol., 83,1e14. (2007). Molecular methods for the detection of biogenic Guilloux-Benatier, M., Pageault, O., Man, A., & Feuillat, M. amine-producing bacteria on foods. Int. J. Food Micro- (2000). Lysis of yeast cells by Oenococcus oeni enzymes. biol., 117, 258e269. J. Ind. Microbiol. Biotech., 25, 193e197. Lee, J.-E., Hong, Y.-S., & Lee, C.-H. (2009). Characterization Guilloux-Benatier, M., Son, H. S., Bouhier, S., & Feuillat, M. of fermentative behaviours of lactic acid bacteria in (1993). Activite´s enzymatiques: Glycosidases et pepti- grape wine through 1H NMR- and GC-based metabolic dases chez Leuconostoc oenos au cours de la croissance profiling. J. Agric. Food Chem., 57, 4810e4817. bacte´rienne. Influence des macromelećules de levures. Lehtonen, P. (1996). Determination of amines and amino Vitis, 32,51e57. acids in wine: A review. Am.J.Enol.Vitic.,47, Gury, J., Barthelmebs, L., Tran, N. P., Divie`s, C., & Cavin, J.-F. 127e133. (2004). Cloning, deletion, and characterization of PadR, Leitao, M. C., Teixeira, H. C., Barreto Crespo, M. T., & San the transcriptional repressor of the phenolic acid decar- Romao, N. V. (2000). Biogenic amine ocurrence in wine. boxylase-encoding padA gene of Lactobacillus plantarum. Amino acid decarboxylase and proteolytic activities Appl. Environ. Microbiol., 70, 2146e2153. expression by Oenococcus oeni. J. Agric. Food Chem., 48, Guzzo, J., Delmas, F., Pierre, F., Jobin, M. P., Samyn, B., van 2780e2784. Beewmer, I., et al. (1997). A small heat-shock protein from Liu, S.-Q., & Davis, C. R. (1994). Analysis of wine carbo- Leuconostoc oenos induced by multiple stresses and during hydrates using capillary gas-liquid chromatography. stationary growth phase. Lett. Appl. Microbiol., 24,393e396. Am. J. Enol. Vitic., 45, 229e234. Henick-Kling, T. (1993). Malolactic fermentation. In Liu, S.-Q., Davis, C. R., & Brooks, J. D. (1995a). Growth and G. H. Fleet (Ed.), Wine microbiology and biotechnology metabolism of selected lactic acid bacteria in synthetic (pp. 286e326). Berlin, Germany: Springer-Verlag. wine. Am. J. Enol. Vitic., 45, 229e234. 222 8. LACTIC ACID BACTERIA

Liu, S.-Q., & Pilone, G. J. (1998). A review: Arginine Lucas, P. M., Claisse, O., & Lonvaud-Funel, A. (2008). High metabolism in wine lactic acid bacteria and its practical frequency of histamine-producing bacteria in the significance. J. Appl. Microbiol., 84, 315e327. enological environment and instability of the histidine Liu, S.-Q., Pritchard, G. G., Hardman, M. J., & Pilone, G. J. decraboxylase production phenotype. Appl. Environ. (1994). Citrulline production and ethyl carbamate Microbiol., 74,811e817. (urethane) precursor formation from arginine degrada- Lucas, P. M., Wolken, W. A. M., Claisse, O., Lolkema, J. S., tion by wine lactic acid bacteria Leuconsotoc oenos and & Lonvaud-Funel, A. (2005). Histamine-producing Lactobacillus buchneri. Am. J. Enol. Vitic., 45, 235e242. pathway encoded on an unstable plasmid in Lactoba- Liu, S.-Q., Pritchard, G. G., Hardman, M. J., & Pilone, G. J. cillus hilgardii 0006. Appl. Environ. Microbiol., 71, (1995). Occurrence of arginine deiminase pathway 1417e1424. enzymes in arginine catabolism by wine lactic acid Maicas, S. (2001). The use of alternative technologies to bacteria. Appl. Environ. Microbiol., 61, 310e316. develop malolactic fermentation in wine. Appl. Microbiol. Liu, S.-Q., Pritchard, G. G., Hardman, M. J., & Pilone, G. J. Biotechnol., 56,35e39. (1996). Arginine catabolism in wine lactic acid bacteria: Maicas, S., Pardo, I., & Ferrer, S. (2000). The effects of Is it via the arginine deiminase pathway or the arginase- freezing and freeze-drying of Oenococcus oeni upon urease pathway. J. Appl. Bacteriol., 81, 486e492. induction of malolactic fermentation in red wine. Int. Llaube`res, R. M., Richard, B., Lonvaud, A., & Dubourdieu, D. J. Food Sci. Technol., 35,75e79. (1990). Structure of an exocellular b-D-glucan from Maicas, S., Pardo, I., & Ferrer, S. (2001). The potential of Pediococcus sp. A wine lactic acid bacteria. Carbohydr. Res., positively-charged cellulose sponge for malolactic 203,103e107. fermentation of wine using Oenococcus oeni. Enz. Microb. Lonvaud-Funel, A. (1995). Microbiology of the malolactic Technol., 28, 415e419. fermentation: Molecular aspects. FEMS Microbiol Lett., Maijala, R. L. (1993). Formation of histamine and tyramine 126, 209e214. by some lactic acid bacteria in MRS-broth and modified Lonvaud-Funel, A. (1999). Lactic acid bacteria in the quality decarboxylation agar. Lett. Appl. Microbiol., 17,40e43. improvement and depreciation of wine. Anton. Leeuw., Makarova, K., Slesarev, A., Wolf, Y., Sorokin, A., Mirkin, B., 76, 317e331. Koonin, E., et al. (2006). Comparative genomics of the Lonvaud-Funel, A. (2001). Biogenic amines in wine: Role of lactic acid bacteria. Proc. Nat. Acad. Sci. USA, 103, lactic acid bacteria. FEMS Microbiol. Lett., 199,9e13. 15 611e15 616. Lonvaud-Funel, A., Guilloux, Y., & Joyeux, A. (1993). Manca de Nadra, M. C., Farias, M. E., Moreno- Isolation of a DNA probe for identification of glucan- Arribas, M. V., Pueyo, E., & Polo, M. C. (1997). Proteo- producing Pediococcus damnosus in wines. J. Appl. Bac- lytic activity of Leuconostoc oenos. Effect on proteins and teriol., 74,41e47. polypeptides from white wines. FEMS Microbiol. Lett., Lonvaud-Funel, A., Joyeux, A., & Desens, C. (1988). The 150, 135e139. inhibition of malolactic fermentation of wines by prod- Manca de Nadra, M. C., Farias, M. E., Moreno- ucts of yeast metabolims. J. Food Sci. Agric., 44, 183e191. Arribas, M. V., Pueyo, E., & Polo, M. C. (1999). A Lonvaud-Funel, A., Joyeux, A., & Ledoux, O. (1991). proteolytic effect of Oenococcus oeni on the nitrogenous Specific enumeration of lactic acid bacteria in fermenting macromolecular fraction of red wine. FEMS Microbiol. grape must and wine by colony hybridization with non- Lett., 174,41e47. isotopic DNA probes. J. Appl. Bacteriol., 71, 501e508. Manca de Nadra, M. C., Sandino de Lamelas, D., & Strasser Lonvaud-Funel, A., & Strasser de Saad, A. M. (1982). Puri- de Saad, M. C. (1998). Pediocin N5p from Pediococcus fication and properties of a malolactic enzyme from pentosaceus: Adsortion on bacterial strains. Int. J. Food. a strain of Leuconostoc mesenteroides isolated from grapes. Microbiol., 39,79e85. Appl. Environ. Microbiol., 43, 357e361. Man˜es-La´zaro, R., Ferrer, S., Rodas, A. M., Urdiain, M., & Lo´pez, I., Ruiz-Larrea, F., Cocolin, L., Orr, E., Phister, T., Pardo, I. (2008a). Lactobacillus bobalius sp. nov., a lactic Marshall, M., et al. (2003). Design and evaluation of PCR acid bacterium isolated from Spanish Bobal grape must. primers for analysis of bacterial populations in wine by Int. J. Syst. Evol. Microbiol., 58, 2699e2703. denaturing gradient gel electrophoresis. Appl. Environ. Man˜es-La´zaro, R., Ferrer, S., Rossello´-Mora, R., & Pardo, I. Microbiol., 69, 6801e6807. (2008b). Lactobacillus uvarum sp. nov. A new lactic acid Lucas, P., Landete, J., Coton, M., Coton, E., & Lonvaud- bacterium isolated from Spanish Bobal grape must. Syst. Funel, A. (2003). The tyrosine decarboxylase operon of Appl. Microbiol., 31, 425e433. Lactobacillus brevis IOEB 9809: Characterization and Man˜es-La´zaro, R., Ferrer, S., Rossello´-Mora, R., & Pardo, I. conservation in tyramine-producing bacteria. FEMS (2009). Lactobacillus oeni sp. nov. from wine. Int. J. Syst. Microbiol. Lett., 229,65e71. Evol. Microbiol., 59, 2010e2014. REFERENCES 223

Marcobal, A., de las Rivas, B., Moreno-Arribas, M. V., & Moreno-Arribas, M. V., & Lonvaud-Funel, A. (2001). Puri- Munõz, R. (2004). Identification of the ornithine decar- fication and characterization of tyrosine decarboxylae of boxylase gene in the putrescine-producer Oenococcus Lactobacillus brevis IOEB 9809 isolated from wine. FEMS oeni BIFI-83. FEMS Microbiol. Lett., 239, 213e220. Microbiol. Lett., 195, 103e107. Marcobal, A., de las Rivas, B., Moreno-Arribas, M. V., & Moreno-Arribas, M. V., Polo, M. C., Jorganes, F., & Munõz, R. (2005). Multiplex PCR method for the Munõz, R. (2003). Screening of biogenic amine produc- simultaneous detection of histamine-, tyramine-, and tion by lactic acid bacteria isolated from grape must and putrescine-producing lactic acid bacteria in foods. J. Food wine. Int. J. Food Microbiol., 84,117e123. Prot., 68, 874e878. Mu¨ ller-Thurgau, H. (1981). Uber die ergebrisse, neuer Marcobal, A., de las Rivas, B., Moreno-Arribas, M. V., & untersughungen auf den gebite der meinbersitung. Munõz, R. (2006). Evidence for horizontal gene transfer Dtsch. Weinbaukong (Worms), 128. as origin of putrescine production in Oenococcus oeni Nannelli, F., Claisse, O., Gindreau, E., de Revel, G., Lonvaud- RM83. Appl. Environ. Microbiol., 72, 7954e7958. Funel, A., & Lucas, P. M. (2008). Detemination of lactic Marcobal, A. M., Sela, D. A., Wolf, Y. I., Makarova, K. S., & acid bacteria producing biogenic amines in wine by Mills, D. A. (2008). Role of hypermutability in the quantitative PCR methods. Lett. Appl. Microbiol., 47, evolution of the genus Oenococcus. J. Bacteriol., 190, 594e599. 564e570. Naouri, P., Chagnaud, P., Arnaud, A., & Galzy, P. (1990). Martineau, B., & Henick-Kling, T. (1995). Formation and Purification and properties of a malolactic enzyme from degradation of diacetyl in wine during alcoholic Leuconostoc oenos ATCC 23278. J. Basic Microbiol., 30, fermentation with Saccharomyces cerevisiae strain EC1118 577e585. and malolactic fermentation with Leuconostoc oenos strain Navarro, L., Zarazaga, M., Saénz, J., Ruiz-Larrea, F., & MCW. Am. J. Enol. Vitic., 46, 442e448. Torres, C. (2000). Bacteriocin production by lactic acid Martineau, B., Henich-Kling, T., & Acree, T. (1995). Reas- bacteria isolated from Rioja red wines. J. Appl. Microbiol., sessment of the influence of malolactic fermentation on 88,41e51. the concentration of diacetyl in wines. Am. J. Enol. Vitic., Nel, H. A., Bauer, R., Wolfaardt, G. M., & Dicks, L. M. T. 46, 385e388. (2002). Effects of bacteriocins pediocin PD-1, plantaricin McFeeters, R. S., & Chen, K. H. (1986). Utilization of electron 423, and nisin on biofilms of Oenocuccus oeni on a stain- acceptors for anaerobic mannitol metabolism by Lacto- less steel surface. Am. J. Enol. Vitic., 53, 191e196. bacillus plantarum. Compounds which serve as electron Nel, L., Wingfield, B. D., van der Meer, L. J., & van acceptors. Food Microbiol., 3,73e81. Vuuren, H. J. J. (1987). Isolation and characterization of McMahon, H., Zoecklein, B. W., Fugelsang, K., & Jasinski, Y. Leuconsotoc oenos bacteriophages from wine and sugar- (1999). Quantification of glycosidase activities in selected cane. FEMS Microbiol. Lett., 44,63e67. yeasts and lactic acid bacteria. J. Ind. Microbiol. Bio- Nicoloff, H., Hubert, J.-C., & Bringel, F. (2001). Carbamoyl- technol., 23, 198e203. phosphate synthetases (CPS) in lactic acid bacteria and Mills, D. A., Rawsthorne, H., Parker, C., Tamir, D., & other Gram-positive bacteria. Lait, 81, 151e159. Makarova, K. (2005). Genomic analysis of Oenococcus Nielsen, J. C., Prahl, C., & Lonvaud-Funel, A. (1996). oeni PSU-1 and its relevance to winemaking. FEMS Malolactic fermentation in wine by direct inoculation Microbiol. Rew., 29, 465e475. with freeze-dried Leuconostoc oenos cultures. Am. J. Enol. Mira de Ordunã, R., Liu, S.-Q., Patchett, M. L., & Vitic., 47,42e48. Pilone, G. J. (2000a). Ethyl carbamate precursor citrul- Nielsen, J. C., & Richelieu, M. (1999). Control of flavor line formation from arginine degradation by malolactic development in wine during and after malolactic wine lactic acid bacteria. FEMS Microbiol. Lett., 183, fermentation by Oenococcus oeni. Appl. Environ. Micro- 31e35. biol., 65, 740e745. Mira de Ordunã, R., Liu, S.-Q., Patchett, M. L., & Osborne, J. P., & Edwards, C. G. (2007). Inhibition of Pilone, G. J. (2000b). Kinetics of the arginine metabolism malolactic fermentation by a peptide produced by of malolactic wine lactic acid bacteria Lactobacillus Saccharomyces cerevisiae during alcoholic fermentation. buchneri CUC-3 and Oenococcus oeni Lo11. J. Appl. Int. J. Food Microbiol., 118,27e34. Microbiol., 89, 547e552. Osborne, J. P., Mira de Ordunã, R., Pilone, G. J., & Liu, S.-Q. Mira de Ordunã, R., Patchett, M. L., Liu, S.-Q., & (2000). Acetaldehyde metabolism by wine lactic acid Pilone, G. J. (2001). Growth, arginine metabolism of the bacteria. FEMS Microbiol. Lett., 191,51e55. wine lactic acid bacteria Lactobacillus buchneri, Oenococcus Ough, C. S. (1976). Ethyl carbamate in fermented beverages oeni at different pH values and arginine concentrations. and foods. I. Naturally occurring ethyl carbamate. Appl. Environ. Microbiol., 67, 1657e1662. J. Agric. Food Chem., 24, 323e328. 224 8. LACTIC ACID BACTERIA

Pilone, G. J., Clayton, M. G., & van Duivenboden, R. J. Ritt, J.-F., Remize, F., Grandvalet, C., Guzzo, J., Atlan, D., & (1991). Characterization of wine lactic acid bacteria: Alexandre, H. (2009). Peptidases specific for proline- Single broth culture for tests of heterofermentation, containing peptides and their unusual peptide-dependent mannitol from fructose, and ammonia from arginine. regulation in Oenococcus oeni. J. Appl. Microbiol., 106, Am. J. Enol. Vitic., 42, 153e157. 801e813. Pinzani, P., Bonciani, L., Pazzagli, M., Orlando, C., Rodas, A. M., Chenoll, E., Maciań, M. C., Ferrer, S., Pardo, I., Guerrini, S., & Granchi, L. (2004). Rapid detection of & Aznar, R. (2006). Lactobacillus vini sp. nov., a wine Oenococcus oeni in wine by real-time quantitative PCR. lactic acid bacterium homofermentative for pentoses. Lett. Appl. Microbiol., 38,118e124. Int. J. Syst. Evol. Microbiol., 56, 513e517. Poblet-Icart, M., Bordons, A., & Lonvaud-Funel, A. (1998). Rodas, A. M., Ferrer, S., & Pardo, I. (2003). 16S-ARDRA, Lysogeny of Oenococcus oeni (syn. Leuconostoc oenos) and a tool for identification of lactic acid bacteria isolated study of their induced bacteriophages. Curr. Microbiol., from grape must and wine. System. Appl. Microbiol., 26, 36, 365e369. 412e422. Pripis-Nicolau, L., de Revel, G., Bertrand, A., & Lonvaud- Rodas, A. M., Ferrer, S., & Pardo, I. (2005). Polyphasic study Funel, A. (2004). Methionine catabolism and production of wine Lactobacillus strains: Taxonomic implications. Int. of volatile sulphur compounds by Oenococcus oeni. J. Syst. Evol. Microbiol., 55, 197e207. J. Appl. Microbiol., 96, 1176e1184. Rodrı´guez, S. B., Amberg, E., & Thornston, R. J. (1990). Pueyo, E., Martıńez-Rodrı´guez, A., Polo, M. C., Santa- Malolactic fermentation in Chardonnay: Growth and Marıá, G., & Bartolome´, B. (2000). Release of lipids sensory effects of commercial strains of Leuconostoc during yeast autolysis in a model wine system. J. Agric. oenos. J. Appl. Bacteriol., 68, 139e144. Food Chem., 48,116e122. Rodrı´guez, H., Curiel, J. A., Landete, J. M., de las Rivas, B., Quere, F., Deschanps, A., & Urdaci, M. C. (1997). DNA Lo´pez de Felipe, F., Go´mez-Cordove´s, C., et al. (2009). probe and PCR-specific reaction for Lactobacillus planta- Food phenolics and lactic acid bacteria. Int. J. Food rum. J. Appl. Microbiol., 82, 783e790. Microbiol., 132,79e90. Radler, F. (1990a). Possible use of nisin in winemaking. I. Rodriguez, H., Landete, J. M., Curiel, J. A., de las Rivas, B., Action of nisin against lactic acid bacteria and wine yeasts Manchenõ, J. M., & Munõz, R. (2008). Characterization in solid and liquid media. Am. J. Enol. Vitic., 41,1e6. of the p-coumaric acid decarboxylase from Lactobacillus T Radler, F. (1990b). Possible use of nisin in winemaking. II. plantarum CECT 748 . J. Agric. Food Chem., 56, Experiments to control lactic acid bacteria in the 3068e3072. production of wine. Am. J. Enol. Vitic., 41,7e11. Rojo-Bezares, B., Saenz, Y., Navarro, L., Zarazaga, M., Ruı´z- Ramos, A., Lolkema, J. S., Konings, W. S., & Santos, H. Larrea, F., & Torres, C. (2007). Coculture-inducible (1995). Enzyme basis for pH regulation of citrate and bacteriocin activity of Lactobacillus plantarum strain 13 pyruvate metabolism by Leuconostoc oenos,a C nuclear J23 isolated from grape must. Food Microbiol., 24, magnetic resonance study. Appl. Environ. Microbiol., 61, 482e491. 1303e1310. Rollan, G. C., Coton, C., & Lonvaud-Funel, A. (1995b). Ramos, A., & Santos, H. (1996). Citrate and sugar cofer- Histidine decarboxylase activity of Leuconostoc oenos 13 mentation in Leuconostoc oenos,a C nuclear magnetic 9204. Food Microbiol., 12, 455e461. resonance study. Appl. Environ. Microbiol., 62, 2577e2585. Rollan, G. C., Farias, M. E., & Manca de Nadra, M. C. Reguant, C., & Bordons, A. (2003). Typification of Oeno- (1993). Protease production by Leuconostoc oenos strains coccus oeni strains by multiplex RAPD-PCR and study of isolated from wine. World J. Microbiol. Biotechnol., 9, population dynamics during malolactic fermentation. 587e589. J. Appl. Microbiol., 95, 344e353. Rollan, G. C., Farias, M. E., & Manca de Nadra, M. C. Reguant, C., Bordons, A., Arola, L., & Roze`s, N. (2000). (1995a). Characterization of two extracellular proteases Influence of phenolic compounds on the physiology of from Leuconostoc oenos strains isolated from wine. World Oenococcus oeni from wine. J. Appl. Microbiol., 88, J. Microbiol. Biotechnol., 9, 153e155. 1065e1071. Roze`s, N., Arola, L., & Bordons, A. (2003). Effect of phenolic Richter, H., Vlad, D., & Unden, G. (2001). Significance of compounds on the co-metabolism of citric acid and pantothenate for glucose fermentation by Oenococcus sugars by Oenococcus oeni from wine. Lett. Appl. Micro- oeni and for suppression of the erythritol and acetate biol., 36, 337e341. production. Arch. Microbiol., 175,26e31. Saénz, Y., Rojo-Bezares, B., Navarro, L., Dıéz, L., Somalo, S., Ritt, J.-F., Guilloux-Benatier, M., Guzzo, J., Alexandre, H., & Zaragaza, M., et al. (2009). Genetic diversity of the pln Remize, F. (2008). Oligopeptide assimilation and trans- locus among oenological Lactobacillus plantarum port by Oenococcus oeni. J. Appl. Microbiol., 104, 573e580. strains. Int. J. Food Microbiol., 134, 176e183. REFERENCES 225

Salou, P., Loubiere, P., & Pareilleux, A. (1994). Growth and Sozzi, T., Maret, R., & Poulin, J. M. (1976). Mise en evidence energetics of Leuconostoc oenos during cometabolism of de bacte´riophages dans le vin. Experientia, 32, 568e569. glucose with citrate or fructose. Appl. Environ. Microbiol., Spano, G., Beneduce, L., Perrotta, C., & Massa, S. (2005). 60, 1459e1466. Cloning and characterization of the hsp 18.55 gene, San Romao, M. V., & Lafon-Lafourcade, S. (1979). Premieres a new member of the small heat shock gene family iso- observations sur l’action de Botrytis cinerea cultive´ lated from wine Lactobacillus plantarum. Res. Microbiol., sur mouˆ t de raisins, a le´gard du me´tabolisme des bac- 156, 219e224. te´ries lactiques dans les mouts et les vins. Vitis, 18, Spano, G., Capozzi, V., Vernile, A., & Massa, S. (2004). 155e160. Cloning, molecular characterization and expression Sao-Jose´, C., Santos, S., Nascimento, J., Brito Madurro, A. G., analysis of two small heat shock genes isolated from Parreira, R., Vieira, G., et al. (2004). Diversity in the lysis- wine Lactobacillus plantarum. J. Appl. Microbiol., 97, integration region of oenophage genomes and evidence 774e782. for multiple tRNA loci, as target for prophage integra- Strasser de Saad, A. M., & Manca de Nadra, M. C. (1993). tion in Oenococcus oeni. Virology, 325,95e98. Characterization of bacteriocin produced by Pediococus Sato, H., Yanagida, F., Shinohara, T., Suzuki, M., Suzuki, K., pentosaceus from wine. J. Appl. Bacteriol., 74, 406e410. & Yokotsuka, K. (2001). Intraspecific diversity Oeno- Strasser de Saad, A. M., Pasteris, S. E., & Manca de coccus oeni. FEMS Microbiol. Lett., 202, 109e114. Nadra, M. C. (1995). Production and stability of pediocin Sauvageot, F., & Vivier, P. (1997). Effects of malolactic N5p in grape juice medium. J. Appl. Bacteriol., 78, fermentation on sensory properties of four Burgundy 473e476. wines. Am. J. Enol. Vitic., 48, 187e192. Tenreiro, R. M., Santos, M. A., Pavela, H., & Vieira, G. Schultz, H., & Radler, F. (1984a). Anaerobic reduction of (1994). Inter-strain relationships among wine leuco- glycerol to propanediol-1, 3 by Lactobacillus brevis and nostocs and their divergence from other Leuconostoc Lactobacillus buchneri. Syst. Appl. Microbiol., 5, 169e178. species, as revealed by low frequency restriction frag- Schultz, H., & Radler, F. (1984b). Propanediol-1, 2-dehy- ment analysis of genomic DNA. J. Appl. Bacteriol., 77, dratase and metabolism of glycerol of Lactobacillus brevis. 271e280. Arch. Microbiol., 139, 366e370. Terrade, N., Noël, R., Couillaud, R., & de Ordunã, R. M. Schu¨tz, M., & Radler, F. (1974). Das vorkommen von mala- (2009). A new chemically defined medium for wine tenzym und malolactac-enzym bei verschiedenen milch- lactic acid bacteria. Food Res. Int., 42, 363e367. sau¨renbakterien. Arch. Mikrobiol., 96,329e339. Tonon, T., Bourdineaud, J. P., & Lonvaud-Funel, A. (2001). Shalaby, A. R. (1996). Significance of biogenic amines to The arcABC gene cluster encoding the arginine deimi- food safety and human health. Food Res. Int., 29, nase pathway of Oenococcus oeni, and arginine induction 675e690. of a CRP-like gene. Res. Microbiol., 152, 653e661. Silveira, M. G., Baumga¨rtner, M., Rombouts, F. M., & Tonon, T., & Lonvaud-Funel, A. (2000). Metabolism of Abee, T. (2004). Effect of adaptation to ethanol on cyto- arginine and its positive effect on growth and revival of plasmic and membrane protein profiles of Oenococcus Oenococcus oeni. J. Appl. Microbiol., 89, 526e531. oeni. Appl. Environ. Microbiol., 70, 2748e2755. Ugliano, M., Genovese, A., & Moio, L. (2003). Hydrolysis of Smit, A., Cordero Otero, R. R., Lambrechts, M. G., wine aroma precursors during malolactic fermentation Pretorius, I. S., & van Rensburg, P. (2003). Enhancing with four commercial starter cultures of Oenococcus oeni. volatile phenol concentrations in wine by expressing J. Agric. Food Chem., 51, 5073e5078. various phenolic acid decarboxylase genes in Saccharo- Vallet, A., Santarelli, X., Lonvaud-Funel, A., de Revel, G., & myces cerevisiae. J. Agric. Food Chem., 51, 4909e4915. Cabanne, C. (2009). Purification of an alcohol dehydro- Snowdon, E. M., Bowyer, M. C., Grbin, P. R., & Bowyer, P. K. genase involved in the conversion of methional to (2006). Mousy off-flavor: A review. J. Agric. Food Chem., methionol in Oenococcus oeni IOEB 8406. Appl. Microbiol. 54, 6465e6474. Biotechnol., 82,87e94. Sohier, D., Coulon, J., & Lonvaud-Funel, A. (1999). Molec- van Vuuren, H. J. J., & Dicks, L. M. T. (1993). Leuconostoc ular identification of Lactobacillus hilgardii and genetic oenos: A review. Am. J. Enol. Vitic., 44,99e112. relatedness with Lactobacillus brevis. Int. J. Syst. Bacteriol., Vandamme, P., Pot, B., Gillis, M., de Vos, P., Kersters, K., & 49, 1075e1079. Swings, J. (1996). Polyphasic taxonomy, a consensus Son, H.-S., Hwang, G.-S., Park, W.-M., Hong, Y.-S., & approach to bacterial systematics. Microbiol. Rev., 60, Lee, C.-H. (2009). Metabolomic characterization of 407e438. malolactic fermentation and fermentative behaviours of Vaquero, I., Marcobal, A., & Munõz, R. (2004). Tannase wine yeasts in grape wine. J. Agric. Food Chem., 57, activity by lactic acid bacteria isolated from grape must 4801e4809. and wine. Int. J. Food Microbiol., 96, 199e204. 226 8. LACTIC ACID BACTERIA

Vasserot, Y., Dion, C., Bonnet, E., Maujean, A., & Jeandet, P. Walling, E., Gindreau, E., & Lonvaud-Funel, A. (2005). A (2001). A study into the role of L-aspartic acid on the putative glucan synthase gene dps detected in exopo- metabolism of L-malic acid and D-glucose by Oenococcus lysaccharide-producing Pediococcus damnosus and oeni. J. Appl. Microbiol., 90, 380e387. Oenococcus oeni strains isolated from wine and cider. Vasserot, Y., Dion, C., Bonnet, E., Tabary, I., Maujean, A., & Int. J. Food Microbiol., 98,53e62. Jeandet, P. (2003). Transport of glutamate in Oenococcus Werning, M. L., Ibarburu, I., Duenãs, M. T., Irastorza, A., oeni 8403. Int. J. Food Microbiol., 85, 307e311. Navas, J., & Lo´pez, P. (2006). Pediococcus parvulus gtf gene Veiga-da-Cunha, M., Firme, P., San Romaò, M. V., & encoding the GTF glycosyltransferase and its application Santos, H. (1992). Application of 13C nuclear magnetic for specific PCR detection of b-D-Glucan-producing resonance to elucidate the unexpected biosynthesis of bacteria in foods and beverages. J. Food Prot., 69, 161e169. erythritol by Leuconostoc oenos. Appl. Environ. Microbiol., Wibowo, D., Eschenbruch, R., Davis, C. R., Fleet, G. H., & 58, 2271e2279. Lee, T. H. (1985). Occurrence and growth of lactic acid Veiga-da-Cunha, M., Santos, H., & van Schaftingen, E. bacteria in wine: A review. Am. J. Enol. Vitic., 36, (1993). Pathway and regulation of erythritol formation in 302e313. Leuconostoc oenos. J. Bacteriol., 175, 3941e3948. Yurdugu¨ l, S., & Bozoglu, F. (2002). Studies on an inhibitor Ventura, M., Casas, I. A., Morelli, L., & Callegari, M. L. producer by lactic acid bacteria on wines on the control (2000). Rapid amplified ribosomal DNA restriction of malolactic fermentation. Eur. Food Res. Technol., 215, analysis (ARDRA) identification of Lactobacillus spp. 38e41. Isolated from fecal and vaginal samples. System. Appl. Zapparoli, G., Reguant, C., Bordons, A., Torriani, S., & Microbiol., 23, 504e509. Dellaglio, F. (2000). Genomic DNA fingerprinting of Versari, A., Parpinello, G. P., & Cattaneo, M. (1999). Leuco- Oenococcus oeni strains by pulsed-field gel electropho- nostoc oenos and malolactic fermentation in wine: A resis and randomly amplified polymorphic DNA-PCR. review. J. Ind. Microbiol. Biotechnol., 23, 447e455. Curr. Microbiol., 40, 351e355. Walling, E., Gindreau, E., & Lonvaud-Funel, A. (2001). La Zaunmu¨ ller, T., Eichert, M., Ritcher, H., & Under, G. (2006). biosnthe`se d’exopolysaccharide par des souches de Variations in the energy metabolism of biotechnologically Pediococcus damnosus isoleés du vin: Mise au point relevant heterofermentative lactic acid bacteria during d’outils moleculaires de detection. Lait, 81, 289e300. growth on sugars and organic acids. Appl. Microbiol. Walling, E., Gindreau, E., & Lonvaud-Funel, A. (2004). A Biotechnol., 72, 421e429. putative glucan synthase gene dps detected in exopoly- Zavaleta, A. I., Martıńez-Murcia, A. J., & Rodrı´guez- saccharide-producing Pediococcus damnosus and Oeno- Valera, F. (1997). Intraspecific genetic diversity of O. oeni coccus oeni strains isolated from wine and cider. Int. as derived from DNA fingerprinting and sequence J. Food. Microbiol., 98,53e62. analysis. Appl. Environ. Microbiol., 63, 1261e1267. CHAPTER 9

Acetic Acid Bacteria Jose´ M. Guillamoń 1, Albert Mas 2 1 Departamento de Biotecnologıá de los Alimentos, Instituto de Agroquı´mica y Tecnologıá de Alimentos (CSIC), Valencia, Spain, 2 Departamento de Bioquı´mica y Biotecnologıá, Facultad de Enologıá, Universitat Rovira i Virgili, Tarragona, Spain

OUTLINE

1. Introduction 228 6. Factors Determining the Growth of Acetic Acid Bacteria: Bacterial Control 2. General Characteristics 228 Methods 245 3. Nutrition and Metabolism 228 6.1. pH 245 3.1. Carbohydrate Metabolism 229 6.2. Temperature 246 3.2. Metabolism of Ethanol and Other 6.3. Ethanol 246 Alcohols 230 6.4. Sulfite 246 3.3. Metabolism of Organic Acids 231 6.5. Oxygen 247 3.4. Nitrogen Metabolism 232 6.6. Storage and Aging 247 4. Taxonomy 232 7. Changes Occurring in Wine as a Result 4.1. Isolation 235 of the Growth of Acetic Acid Bacteria 248 4.2. Identification 235 7.1. Changes in Grapes and Must as a Result 4.3. Molecular Techniques for the Rapid of the Growth of Acetic Acid Bacteria 248 Identification of Acetic Acid Bacteria 236 7.2. Changes in Wine as a Result of the 4.4. Molecular Techniques for Typing of Growth of Acetic Acid Bacteria 249 Acetic Acid Bacteria 240 8. Interactions With Other Microorganisms 5. Growth of Acetic Acid Bacteria in in Wine 250 Winemaking Processes 241 9. Final Recommendations to Avoid Wine 5.1. Association of Acetic Acid Bacteria Spoilage Due to Acetic Acid Bacteria 250 With Grapes 242 5.2. Growth Dynamics of Acetic Acid Bacteria During Alcoholic Fermentation 242

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10009-8 227 Copyright Ó 2011 Elsevier Inc. All rights reserved. 228 9. ACETIC ACID BACTERIA 1. INTRODUCTION that appear under the microscope as individual cells, in pairs, in chains, or in clumps. Their size Acidophilic bacteria can grow on substrates varies between 0.4 and 1 mm wide and between withapHoflessthan5andarefoundinacidic 0.8 and 4.5 mm long. They are clearly motile foodstuffs such as fruit juice. They comprise two under the microscope and have polar or peritri- main groups: acetic acid bacteria and lactic acid chous flagella. They do not form endospores as bacteria. In grape must and wine, the low pH resistant forms. They are aerobic and usually (between 3 and 4) and the presence of alcohol display respiratory metabolism with oxygen and/or high concentrations of sugar limit the functioning as a terminal electron acceptor. microbial flora to just a few yeasts and bacteria. Nevertheless, in unfavorable conditions (anaer- Among these, lactic acid bacteria play an impor- obic or with low concentrations of oxygen), tant role in winemaking, since they are responsible alternative electron acceptors can be used with for malolactic fermentation. In contrast, acetic acid a considerable associated slowing of bacterial bacteria are only linked to wine spoilage processes, metabolism and, therefore, growth. They are mainly through the production of acetic acid, acet- catalase-positive and oxidase-negative. The aldehyde, and ethyl acetate. This form of wine optimal temperature for growth is 25 to 30C spoilage has been recognized since its initial and the optimal pH is 5 to 6, although they description by Pasteur (1868), but winemakers still can still grow well at pHs below 4 (de Ley remain vigilant towards the risk of these bacteria et al., 1984). Some species produce pigments causing an increase in volatile acidity at some on solid growth medium and can produce point during the vinification process and pro- different types of polysaccharides. ducing what is widely known as “pricked” wine. These bacteria are found in substrates contain- In this chapter, we will review the influence of ing sugar and/or ethanol, such as fruit juices, acetic acid bacteria on winemaking and provide wine, cider, beer, and vinegar. On these sub- up-to-date information that will be of use in strates, bacterial metabolism involves incomplete helping enologists to detect and control their oxidation of the sugars and alcohols and leads growth and, thus, prevent spoilage of the final to accumulation of organic acids as end pro- product. Acetic acid bacteria are found on grapes ducts. The production of acetic acid on ethanol- and in wine and must, and their growth depends containing substrates accounts for the common on the phase in the winemaking process and the name ascribed to these bacteria. However, these treatments that have been used. Although the microorganisms are also able to oxidise glucose literature available on these bacteria and their to gluconic acid, galactose to galactonic acid, effects on winemaking is less extensive than that and arabinose to arabonic acid. Some of these on the other two groups of organisms of relevance reactions are of significant interest to the wine- in winemaking (yeasts and lactic acid bacteria), making industry. The traditional industrial appli- some excellent, highly recommendable reviews cation of acetic acid bacteria is in the production have been published (Bartowsky & Henschke, of vinegar; however, lesser-known applications 2008; Drysdale & Fleet, 1988; du Toit & Pretorius, include the production of cellulose and the 2002; Polo & Sańchez-Luengo, 1991). conversion of sorbitol into sorbose.

2. GENERAL CHARACTERISTICS 3. NUTRITION AND METABOLISM

Acetic acid bacteria are gram-negative or Acetic acid bacteria are obligate aerobes and, as gram-variable, ellipsoid or cylindrical bacteria a result, their growth is highly dependent upon NUTRITION AND METABOLISM 229 the availability of molecular oxygen. Never- for most strains of acetic acid bacteria. Unlike theless, under conditions such as those found in yeasts, however, this glucose is not meta- during winemaking (for instance alcoholic bolized as part of glycolysis. Although most fermentation or aging), alternative terminal of the individual reactions are functional, the electron acceptors such as quinones can be used. complete pathway is inactive as a result of Consequently, acetic acid bacteria can survive the lack of the phosphofructokinase enzyme. under the almost completely anaerobic condi- Consequently, acetic acid bacteria must use tions that are generally present during wine- alternative pathways in order to employ carbo- making. Under those conditions, the bacteria hydrates as sources of carbon and energy (see may also display limited growth. This growth Figure 9.1): will be enhanced by any process involving aeration or oxygenation of the medium as a result of 1) The glucose-6-phosphate dehydrogenase the increased levels of the principal electron system transforms a mole of glucose-6- acceptor. phosphate into a mole of ribulose-5- The other important factor in the growth of phosphate with the formation of two these bacteria is the carbon source. This will moles of reduced nicotinamide adenine determine which metabolic pathways are used dinucleotide phosphate (NADPH) via three and, therefore, which metabolic intermediates reactions. In all cases, the production of and end products will ultimately inﬂuence the energy takes place as a consequence of the quality of the wine. oxidation of NADPH via the respiratory chain and the production of adenosine 3.1. Carbohydrate Metabolism triphosphate (ATP) by oxidative phosphorylation. Acetic acid bacteria can metabolize various 2) The Entner-Doudoroff pathway, which carbohydrates as carbon sources. As in other should be considered as an extension of the microorganisms, glucose acts as a carbon source glucose-6-phosphate system, converts

Glucose-6- GLUCOSE FIGURE 9.1 Schematic dia- Phosphate gram of the metabolism of acetic acid bacteria. ED ¼ Gluconate Entner-Doudoroff; EMP ¼ G. oxydans Gluconate-6-Phosphate Embden-Meyerhoff-Parnas; WD ¼ Warburg-Dickens. Ketoglutarate Ribulose-5-Phosphate Ga. xylinum WD ED LACTATE ETHANOL path way

Glyceraldehyde-3- Pyruvate Acetaldehyde Phosphate EMP

OC 2 Krebs Acetate cycle

Acetobacter Gluconacetobacter 230 9. ACETIC ACID BACTERIA

Although acetic acid bacteria can completely glucose-6-phosphate into gluconate-6- degrade glucose, they characteristically display phosphate. This is then hydrolyzed by incomplete oxidation of carbon sources via one an aldolase to form pyruvate and or two biochemical reactions, resulting in accu- glyceraldehyde-3-phosphate. Both mulation of intermediate metabolites. Although molecules are converted into acetate by these partial oxidations are typical of substrates decarboxylation. This reaction is such as alcohols, they also occur with some characteristic of some strains of Gluconobacter monosaccharides such as glucose. Glucose is . oxydans directly oxidized to glucono-d-lactone, which The Warburg-Dickens pathway or hexose 3) is then oxidized to gluconic acid (see Figure 9.1). monophosphate cycle is the most common Although most acetic acid bacteria can carry out pathway for the metabolism of glucose and is this reaction, it is especially active in . present in all species of acetic acid bacteria. It G oxydans and is characteristic in sugar-containing sub- includes the three reactions that comprise the strates such as grapes and must. Under these glucose-6-phosphate dehydrogenase system conditions, the accumulation of gluconic acid and produces monosaccharides of varying is typical of the growth of acetic acid bacteria. size that are ultimately converted into triose The use of glucose and, presumably, other phosphates, which can then be metabolized sugars, either through the hexose monophos- in the Embden-Meyerhoff-Parnas (EMP) phate pathway or via direct oxidation to glu- pathway. conic acid, depends on the pH and the glucose The end products of these pathways can concentration of the medium. At pHs below be completely oxidized to produce carbon 3.5 or glucose concentrations above 0.9 to dioxide and water through the Krebs cycle. 2.7 g/L, oxidation of glucose via the hexose However, among the acetic acid bacteria monophosphate pathway is inhibited (Drysdale only the genera Acetobacter, Acidomonas,and & Fleet, 1988; du Toit & Pretorius, 2002). The end Gluconacetobacter areabletocarryoutthis result under these conditions would be the overoxidation of ethanol (Cleenwerck & de accumulation of gluconic acid in the culture Vos, 2008). Species of the genus Gluconobacter medium. do not have a functional Krebs cycle and are therefore unable to completely oxidize 3.2. Metabolism of Ethanol and Other the molecules formed in the earlier meta- Alcohols bolic pathways. In this genus, the lack of a-ketoglutarate and succinate dehydroge- Transformation of ethanol into acetic acid is nases prevents completion of the cycle. The the most well-known characteristic of acetic failure to oxidize acetic acid is common to Sac- acid bacteria and by far the most relevant in charibacter and Neoasaia species, whereas other winemaking. This transformation involves two genera display limited oxidation of this mole- biochemical reactions: ethanol is first trans- cule. Those species that can carry out complete formed into acetaldehyde in a reaction cata- oxidation only do so through the Krebs cycle lyzed by alcohol dehydrogenase and the when all sugars or alcohols have been fully acetaldehyde is then transformed into acetic consumed. Consequently, the Krebs or tricar- acid by aldehyde dehydrogenase. Both reactions boxylic acid cycle will never be functional in involve electron transfer to molecular oxygen. winemaking processes, since the pathway is The availability of other terminal electronic inhibited by glucose or fructose and ethanol acceptors such as quinones (Drysdale & Fleet, (Saeki et al., 1997a). 1988) may explain the survival and even limited NUTRITION AND METABOLISM 231 growth of these bacteria under anaerobic condi- fructose is produced from mannitol, sorbose tions such as those found in bottled wine and from sorbitol, erythrose from erythritol, etc. even during alcoholic fermentation. Acetobacter Since most of the enzymes that catalyze these alcohol dehydrogenase activity is more stable reactions are located in the cell membrane, under winemaking conditions than that of Glu- a wide range of substrates accumulate in the conobacter, which may explain the greater medium and, as a result, acetic acid bacteria production of acetic acid by Acetobacter. Alde- are particularly appropriate microorganisms hyde dehydrogenase is more sensitive than for use in biotechnology (Deppenmeier et al., alcohol dehydrogenase to the concentration of 2002). ethanol in the medium, and this could lead to greater accumulation of acetaldehyde in wines 3.3. Metabolism of Organic Acids with a higher alcohol concentration (Muraoka et al., 1983). Drysdale and Fleet (1989a) also Although incomplete oxidation is a common observed increasing concentrations of acetalde- metabolic characteristic of acetic acid bacteria, hyde as the concentration of oxygen dissolved some substrates that are present in a state of in the wine diminished. The presence of excess intermediate oxidation, such as organic acids, sulfite also leads to an increase in the concentra- can continue to be oxidized. Lactic acid is tion of acetaldehyde as a consequence of the a good carbon source for many acetic acid stable bond formed between these molecules, bacteria and it can be oxidized through various which prevents its transformation into acetic pathways with different end products. In one of acid (Lafon-Lafourcade, 1985). Both alcohol these, lactic acid is oxidized to pyruvate, which and aldehyde dehydrogenase are located in is then hydrolyzed to acetaldehyde and carbon the cell membrane with their active sites orien- dioxide in a reaction catalyzed by pyruvate tated outwards, and as a consequence the prod- decarboxylase. Acetaldehyde is then oxidized ucts of the reactions they catalyze are usually to form the end product, acetic acid. The activity found in the medium and not inside the cell of this pyruvate decarboxylase is dependent on (Saeki et al., 1997b). Although a cytoplasmic the predominant substrate in the culture form of alcohol dehydrogenase has been medium. This activity is not detected in the described, it has a much lower specific activity presence of mannitol, whereas it is maximal in (Adachi et al., 1978). the presence of lactic acid (Raj et al., 2001). An Glycerol, the main byproduct of alcoholic alternative to this pathway is the production of fermentation, also acts as a carbon source for acetoin from lactic acid via an acetolactate inter- acetic acid bacteria (de Ley et al., 1984). Most mediate (de Ley, 1959). of the glycerol is transformed into dihydroxyac- Complete oxidation of many organic acids is etone (ketogenesis), although some may be used dependent on the availability of a functional effectively for the production of biomass. Krebs cycle. Since Gluconobacter is unable to In a similar process to that seen with glycerol, completely oxidize acetic acid, it is also unable this metabolic capacity of acetic acid bacteria is to completely oxidize many organic acids. Ace- extended to the direct oxidation of other tic acid is oxidized in a reaction catalyzed by primary alcohols and polyalcohols, which are acetyl-coenzyme A (CoA) synthase, which leads converted by oxidation into their respective to the production of actetyl-CoA (Saeki et al., ketones and ketoses. In this way, acetoin is 1997b). Acetyl-CoA enters the Krebs cycle and produced from 2,3-butanediol or acetol from is converted into intermediate metabolites in 1,2-propanediol. Some polyalcohols are con- the pathway before oxidation is completed verted into their corresponding sugars. Thus, with the generation of carbon dioxide and 232 9. ACETIC ACID BACTERIA water. The presence of acetic acid leads to an fact, while additional nutrients must be added increase in the activity of acetyl-CoA synthase, during the production of vinegars from alcohol but this activity is strongly inhibited by ethanol or substrates low in amino acids (such as cider; and glucose. Valero et al., 2005), this is not generally necessary for wines. 3.4. Nitrogen Metabolism Some species of acetic acid bacteria can fix 4. TAXONOMY atmospheric nitrogen. This was first described some years ago for Gluconoacetobacter diazotrophi- The first nomenclature for the classification of cus (Gillis et al., 1989), and six more species have acetic acid bacteria is attributed to Peerson in recently been reported to display this capacity: 1822 with the proposal of the name Mycoderma Gluconoacetobacter johannae, Gluconoacetobacter for this group of microorganisms. Pasteur azotocaptans, Acetobacter peroxydans, Swaminatha- (1868) carried out the first systematic study of nia salitolerans, Acetobacter nitrogenifigens, and acetic fermentation. He recognized that the Gluconoacetobacter kombuchae, the seventh and “vinegar mother” was a mass of live microorgan- last reported to date (Dutta & Gachhui, 2007). isms that induced acetic acid fermentation and Other species use ammonia as a simpler nitrogen that this was not possible in the absence of Myco- source (de Ley et al., 1984). Thus, these bacteria derma aceti. Later, around 1879, Hansen observed can synthesize all of their amino acids and that the microbial flora that converted alcohol nitrogen compounds from ammonia. The pres- into acetic acid was not pure and comprised ence of amino acids in the growth medium can various bacterial species. The genus Acetobacter have a stimulatory or inhibitory effect on growth, was first proposed by Beijerinck (1899). depending on the amino acid. Thus, glutamate, Bacterial taxonomy has traditionally been glutamine, proline, and histidine stimulate the based on morphological, biochemical, and growth of acetic acid bacteria, whereas valine in physiological criteria. The first to propose a clas- the case of G. oxydans and threonine and homo- sification of acetic acid bacteria based on these serine in the case of Acetobacter aceti appear to criteria was Visser’t Hooft (1925). Asai (1935) inhibit growth (Belly & Claus, 1972). However, formulated the proposal to classify acetic acid no studies have addressed the nutritional bacteria into two genera: Acetobacter and Gluco- requirements of acetic acid bacteria in terms of nobacter. Later, Frateur (1950) proposed a classifi- nitrogenous compounds under winemaking cation based essentially on five physiological conditions. A preference for certain amino acids criteria: catalase activity, production of gluconic has been observed for acetic acid bacteria during acid from glucose, oxidation of acetic acid to the production of vinegar (Maestre et al., 2008; carbon dioxide and water, oxidation of lactic Valero et al., 2005), in some cases leaving substan- acid to carbon dioxide and water, and oxidation tial amounts of ammonia in the medium. The of glycerol into dihydroxyacetone. Based on preferential use of proline is noteworthy, as this these criteria, he proposed the subdivision of amino acid is not used by yeast under fermenta- Acetobacter into four groups: peroxydans, oxydans, tive conditions and is therefore particularly mexosydans, and suboxydans (reviewed by Barja abundant in grape must and wine. The release et al., 2003). of amino acids and other nitrogenous The history of the taxonomic criteria applied compounds following yeast autolysis may also to bacterial species is provided in the different be sufficient to make a significant contribution editions of Bergey’s Manual of Determinative to the growth of acetic acid bacteria in wine. In Bacteriology, which has become a reference for TAXONOMY 233 bacterial taxonomy. The eighth edition of this from this process of reordering genera and manual (Buchanan & Gibbons, 1974) recog- species. Nine new genera of acetic acid bacteria nized two generadAcetobacter (able to convert must be added to the two mentioned earlier: acetate and lactate into carbon dioxide and Acidomonas (Urakami et al., 1989), Gluconaceto- water and motile through peritrichous flagella bacter (Yamada et al., 1997), Asaia (Yamada or nonmotile) and Gluconobacter (unable to et al., 2000), Kozakia (Lisdiyanti et al., 2002), Sac- oxidize lactate and acetate completely and charibacter (Jojima et al., 2004), Swaminathania motile through polar flagella or nonmotile)d (Loganathan & Nair, 2004), Neoasaia (Yukphan and placed the genus Gluconobacter within the et al., 2005), Granulibacter (Greenberg et al., Pseudomonadaceae family. The genus Acetobacter 2006), and Tanticharoenia (Yukphan et al., 2008). was not assigned to any family and was placed In addition, a number of new species have with the genera of unknown affiliation. In 1984, been identified. Thus, the family Acetobactera- a new edition of the manual was published ceae currently comprises 11 genera and 56 under a different title: Bergey’s Manual of species (58 if we include newly proposed Systematic Bacteriology. This new edition species that have yet to be accepted) of acetic included among the taxonomic criteria some acid bacteria (see Table 9.1). Acetobacter and Glu- molecular tests, such as fatty acid composition, conacetobacter, which contain 19 and 16 species, electrophoresis of soluble proteins, guanine- cytosine (GC) content, and DNAeDNA hybrid- TABLE 9.1 Genera and Species of Acetic Acid ization. These techniques suggested that the Bacteria (June 2009) genera Gluconobacter and Acetobacter had extensive phylogenetic similarity and, as Genus Species a result, in this new edition of Bergey’s Acetobacter Acetobacter aceti 5 Manual (de Ley et al., 1984) the two genera Acetobacter pasteurianus were included in the family Acetobacteraceae, Acetobacter pomorum which had previously been included in Acetobacter peroxydans the division a-Proteobacteria (Stackebrandt Acetobacter indonesiensis et al., 1988). The genus included Acetobacter tropicalis Acetobacter Acetobacter syzygii four species: A. aceti, Acetobacter pasteurianus, Acetobacter cibinongenesis Acetobacter liquefaciens, and Acetobacter hansenii. Acetobacter orientalis The genus Gluconobacter included only one Acetobacter orleaniensis species: G. oxydans. The main difference Acetobacter lovaniensis between the two genera continued to be that Acetobacter estuniensis Acetobacter malorum Acetobacter species are able to perform complete Acetobacter cerevisiae oxidation of ethanoldin other words, that Acetobacter oeni ethanol can be oxidized to acetic acid and Acetobacter nitrogenifigens then to carbon dioxide and waterdwhereas Acetobacter senegalensis is unable to oxidize acetic acid Acetobacter ghanensis Gluconobacter Acetobacter fabarum completely to carbon dioxide plus water. The taxonomy of microorganisms has been Acidomonas Acidomonas methanolica continually revised and reorganized, mainly Asaia Asaia bogorensis based on data obtained using molecular tech- Asaia siamensis niques such as DNAeDNA or DNAeRNA hyb- Asaia krugthepensis rization and analysis of 16S rDNA. The Asaia lannensis Acetobacteraceae family has not been exempt (Continued) 234 9. ACETIC ACID BACTERIA

TABLE 9.1 Genera and Species of Acetic Acid respectively, show the greatest diversity among Bacteria (June 2009) (cont’d) the species that have been described. The genus differs biochemically from Genus Species Acetobacter Gluconoa- cetobacter in that it produces ubiquinone-9 (Q-9) Gluconacetobacter Gluconacetobacter liquefaciens 4 rather than ubiqiuinone-10 (Q-10), which is Gluconacetobacter diazotrophicus 5 found in species (Yamada Gluconacetobacter xylinus Gluconoacetobacter Gluconacetobacter hansenii et al., 1997). This ubiquinone Q-10 is common Gluconacetobacter europaeus to all other genera of acetic acid bacteria Gluconacetobacter oboediens (Yamada & Yukphan, 2008). The genus Acidomo- Gluconacetobacter intermedius nas, with its single species Acidomonas methanol- Gluconacetobacter sacchari , was characterized as the only genus able to Gluconacetobacter entanii ica Gluconacetobacter johannae grow in methanol as a unique carbon source Gluconacetobacter azotocaptans (Urakami et al., 1989), although the recently Gluconacetobacter swingsii described Granulibacter also has this capacity Gluconacetobacter kombuchae (Greenberg et al., 2006). The characteristic Gluconacetobacter nataicola features of strains assigned to the genera Gluconacetobacter rhaeticus Asaia Gluconacetobacter saccharivorans and Swaminathania are their production of little 1 Gluconacetobacter persimmonis or no acetic acid from ethanol and their failure 2 to grow in the presence of more than 0.35% acetic Gluconobacter Gluconobacter oxydans Gluconobacter frateurii acid. The strains of these genera have mainly been Gluconobacter assaii isolated from flowers (Yamada et al., 2000)and Gluconobacter cerinus rice (Loganathan & Nair, 2004). Gluconobacter albidus The other genera (Kozakia, Saccharibacter, Neo- Gluconobacter thailandicus , , and ) each Gluconobacter kondonii asaia Granulibacter Tanticharoenia Gluconobacter roseus contain only a single species. These bacteria, Gluconobacter sphaericus which have generally been described in soils, Gluconobacter japonicus flowers, and fruits from Asian countries, display Gluconobacter kanchanaburiensis very limited production of acetic acid from Gluconobacter wancherniae ethanol and have a very weak capacity to over- 3 Granulibacter Granulibacter bethesdensis oxidize acetate or lactate (Cleenwerck & de Vos, Kozakia Kozakia baliensis 2008). An exception worth mentioning is Granu- libacter, for which the only species has been iso- Neoasaia Neoasaia chiangmaiensis lated in hospital environments in the United Saccharibacter Saccharibacter floricola States in patients with chronic granulomatous Swaminathania Swaminathania salitolerans disease, a rare hereditary condition that is characterized by the accumulation of superoxides Tanticharoenia Tanticharoenia sakaeratensis and hydrogen peroxide, which facilitate the 1Newly proposed species that have yet to be accepted. growth of parasites with catalase activity Acetic acid bacteria with sequenced genomes: (Greenberg et al., 2006). Some species are also 2Prust, C. et al. (2005); 3Greenberg, D. E. et al. (2007); highly osmophilic and able to develop in media 4Bertalan et al. (2009); containing high concentrations of sugar equiva- 5Azuma (2008). lent to more than 30% glucose ( , Species described in grapes and wine or wine vinegar are shown in Acidomonas bold. Asaia, Neoasaia, Saccharibacter, and Tanticharoenia) (Cleenwerck & de Vos, 2008; Yukphan et al., 2008). TAXONOMY 235

Finally, the genus Frateuria, which contains fungi and with penicillin (3 U/mL) to inhibit a single species, Frateuria aurantia, belongs to the growth of lactic acid bacteria (Ruiz et al., the g-Proteobacteria, but it has an oxidative 2000). Mannitol medium (2.5% mannitol, 0.5% metabolism similar to that of the acetic acid yeast extract, 0.3% peptone, and 2% agar) yields bacteria and is therefore usually considered to very similar results in the isolation of acetic be a pseudo-acetic-acid bacteria (Yamada & acid bacteria from enological samples. This Yukphan, 2008). It was initially named Aceto- liquid medium is particularly useful for the bacter aurantia and characterized on the basis production of biomass, since it is the best of being positive for ubiquinone Q-8, which is medium to support the growth of acetic acid absent from all acetic acid bacteria. bacteria. The plates are incubated for 2 to 4 d at Our understanding of both the taxonomy 28C under aerobic conditions. Other authors and the biochemical and physiological charac- have reported using similar media for the growth teristics of acetic acid bacteria will be increased of acetic acid bacteria derived from enological substantially when a complete genome seq- samples (Bartowsky et al., 2003; du Toit & uence becomes available for the different Lambrechts, 2002). Bartowsky et al. (2003) also species in this bacterial group. To date, seq- recommend the use of Wallerstein nutrient uences have only been published for G. oxydans agar supplemented with 2% ethanol or 10% (Prust et al., 2005), Granulibacter bethesdensis filter-sterilized wine. (Greenberg et al., 2007), and Ga. diazotrophicus No problems have been reported in the isola- (Bertalan et al., 2009). The sequencing of another tion and culture of these bacteria when derived two species (A. pasteurianus and Gluconaceto- from samples obtained during the winemaking bacter xylinus) was reported by the Osaka Insti- process. However, differences have been tute for Fermentation in Japan at the Second observed in bacterial recovery on plates and Congress on Acetic Acid Bacteria in Nagoia, numbers observed under the microscope, and Japan, in November 2008, but the sequence is these have been attributed to the presence of not publicly available. viable but noncultivable bacteria (Millet & Lonvaud-Funel, 2000). In contrast, a number of 4.1. Isolation studies have reported difficulty culturing these bacteria from samples obtained during vinegar Acetic acid bacteria have traditionally been production (Gonza´lez et al., 2006a; Ilabaca considered to be fastidious due to their poor et al., 2008; Sokollek et al., 1998; Trcek, 2005). recovery from culture media. As a result, various This problem has been partially resolved by media have been developed for their isolation. In the introduction of a double layer of agar in our laboratory, to isolate acetic acid bacteria from the medium (0.5% in the bottom layer and 1% grapes, must, and wine, we mainly use GYC agar in the top layer) and the use of culture media (5% D-glucose, 1% yeast extract, 0.5% calcium that simulate the acetifying environment, such carbonate, and 2% agar), which was first as AE medium containing ethanol and acetic described by Carr and Passmore (1979). The acid (Entani et al., 1985). presence of calcium carbonate gives the medium an opaque appearance. The production of acetic 4.2. Identification acid by the bacteria causes the calcium carbonate to dissolve and a transparent halo to form around Until recently, the classification of bacterial the colony. This medium is also adjusted to a pH species was based on morphological, biochem- of 4.5 and is supplemented with natamycin ical, and physiological criteria. These pheno- (100 mg/L) to inhibit the growth of yeasts and typic characteristics have also been used to 236 9. ACETIC ACID BACTERIA assign isolates to specific genera and species. A good alternative to identification based on The more general characteristics used for the phenotypic characteristics is the use of geno- phenotypic classification of acetic acid bacteria typic and molecular criteria. Developments in include cell and colony morphology, Gram genetics and molecular biology have allowed staining, and catalase and oxidase activity. As polymorphism and variability in certain mole- summarized by Cleenwerck and de Vos (2008), cules, mainly DNA, to be used for taxonomic other tests that have been useful for the identifi- characterization and identification. As a result, cation of acetic acid bacteria and even the differ- there has been a marked increase in the number entiation of genera and species include the of genera and species of acetic acid bacteria in following: (1) the production of acetic acid recent years through the use of molecular tech- from ethanol; (2) overoxidation of lactate and niques such as DNAeDNA hybridization and acetate to carbon dioxide and water; (3) growth sequencing of ribosomal genes. Nevertheless, in the presence of 0.35% acetic acid; (4) growth despite their effectiveness for the classification in 1% nitric acid; (5) formation of 2-ketoglu- of bacterial species, these techniques are not conic, 5-ketogluconic, and 2,5-ketogluconic appropriate for routine use and are too complex acid from glucose; (6) ketogenesis of glycerol; to be applied to the processing of large numbers (7) growth on different carbon sources (e.g., of samples. methanol); (8) formation of brown water-soluble pigments; (9) formation of g-pyrones from 4.3. Molecular Techniques for the glucose or fructose; (10) production of acids Rapid Identification of Acetic Acid from sugars; (11) production of cellulose; (12) Bacteria growth in 30% glucose; (13) presence and position of flagella; and (14) motility. These charac- One of the first techniques used in bacterial teristics can be used to discriminate between taxonomy was the analysis of the percentage different genera and species of acetic acid of GC base pairs. The GC content was already bacteria, as shown in Table 9.2. Many of the tests included for the species of Acetobacteraceae in yield variable results between different species Bergey’s Manual of Systematic Bacteriology (de and even between different strains of the same Ley et al., 1984). However, GC content alone is species. This complicates the identification insufficient for the identification of an isolate. process significantly and makes it necessary to Although the GC content of acetic acid bacteria use a larger number of additional tests in order ranges from 52 to 67%, most species include to reliably identify an isolate. strains with percentages around the middle The same characteristics can be used to iden- of this range (e.g., 59%). Nevertheless, this tify acetic acid bacterial isolates. Bartowsky criterion is used alongside DNAeDNA hybrid- et al. (2003) identified wine isolates using a few ization and various morphological and physio- phenotypic tests that can be considered specific logical characteristics for the classification of for the acetic acid bacteria found in wine. These new species, and it forms the basis of so-called tests, which were designed to discriminate polyphasic taxonomy (Cleenwerck & de Vos, between the five species of acetic acid bacteria 2008). described in Bergey’s Manual of Systematic Bacteri- Sequencing is one of the methods proposed ology (de Ley et al., 1984), include Gram staining; for the identification of acetic acid bacteria and catalase test; growth in ethanol, sodium acetate, generally involves ribosomal genes or the and dulcitol; overoxidation of ethanol; ketogen- region between the 16S and 23S ribosomal genes esis of glycerol; oxidation of lactate; and produc- (Yamada & Yukphan, 2008). The latter involves tion of water-soluble brown pigments. intergenic regions known as internal transcribed TAXONOMY 237

TABLE 9.2 Phenotypic Characteristics of the Different Genera of Acetic Acid Bacteria

Characteristic A G Ac Ga As K S Sa N Gr T Production of acetate from ethanol þþþþþþw/þ w þ

Oxidation of acetate to carbon dioxide and þþþwwww water Growth in 0.35% acetic acid þþþþþþþnd þ

Growth in 1% nitric acid þþnd nd Formation of gluconic acid from glucose þþþþþnd nd nd nd nd Ketogenesis of glycerol þþþwwþw þ Growth in methanol þþ Formation of brown, water-soluble pigments þþnd þ Formation of g-pyrones from glucose þnd þnd nd nd nd nd

Production of acids from fructose þþþþþþnd nd Production of cellulose þnd nd nd nd Growth in 30% glucose þþw þþþnd þ Presence and location of ﬂagella per pol pol per per per Majority ubiquinone Q9 Q10 Q10 Q10 Q10 Q10 Q10 Q10 Q10 nd Q10 GC content (%) 52e64 54e64 62e63 56e67 59e61 56e57 52e53 57e60 63 59 66

A ¼ Acetobacter; G ¼ Gluconobacter; Ac ¼ Acidomonas; Ga ¼ Gluconacetobacter; As ¼ Asaia; K ¼ Kozakia; S ¼ Swaminathania; Sa ¼ Saccharibacter; N ¼ Neoasaia; Gr ¼ Granulibacter; T ¼ Tanticharoenia. þ¼positive; ¼negative; nd ¼ not determined; per ¼ peritrichous; pol ¼ polar; w ¼ weak. Adapted from Cleenwerck and de Vos (2008) and Yukphan et al. (2008). spacers (ITSs). Although these are transcribed as comparison of patterns with those from known part of the group of ribosomal genes, they are samples such as type strains for each species. later eliminated from the final ribosomal RNA These techniques are not useful for precise taxo- molecules. Sequencing of the RecA gene has nomic identification but, when combined with also been proposed (Cleenwerck & de Vos, sequencing of representative samples for each 2008). However, despite the reduction in the group, they provide a highly accurate means of cost of DNA sequencing in recent years, its confirming identification. systematic use in descriptive ecological studies Our group has developed various molecular is not practical given the hundreds of samples techniques for the rapid and reliable identifica- that may easily be involved. Consequently, tion and, in some cases, quantification of most approaches are needed that can be used to detect species of acetic acid bacteria, particularly those and group different microorganisms using present in grapes, wine, and vinegar. The first simple, rapid, and inexpensive techniques. method involves restriction analysis of the 16S These techniques are generally based on the ribsomal gene following amplification by poly- analysis of fragments of ribosomal genes using merase chain reaction (PCR) using a protocol electrophoretic or similar methods that allow known as 16S-ARDRA (Poblet et al., 2000; 238 9. ACETIC ACID BACTERIA

Ruiz et al., 2000). The method was fine-tuned as and during the production of traditional balsamic follows (Ruiz et al., 2000): vinegar (Giudici et al., 2003). Primers have also been designed for amplifi- Design of specific primers for the 1) cation of the 16S-23S ITS (Ruiz et al., 2000). amplification of 16S ribosomal DNA (rDNA) These regions usually display greater sequence based on sequences from acetic acid bacteria variability than the ribosomal genes and, there- present in sequence databases. fore, restriction analysis reveals more extensive PCR amplification of 16S rDNA from acetic 2) polymorphism or variability between the strains acid bacteria. To confirm the specificity, and species analyzed. However, amplification of different reference strains of acetic acid the ITS region from the same reference strains bacteria, lactic acid bacteria, and yeasts were followed by digestion with the same restriction used. Only the acetic acid bacteria displayed enzymes does not result in greater species differ- the characteristic 1450-base-pair amplification entiation. This technique was used by Sievers product predicted from sequence data. et al. (1996) to differentiate between two species Digestion of the amplification products with 3) of . Later, Trcek and Teuber (2002) different restriction enzymes. Of all the Acetobacter used digestion with the same restriction enzymes tested, Iand Iproducedthe Taq Rsa enzymes to characterize 57 strains of acetic acid best results for species identification (see bacteria. Those authors were also unable to Table 9.3). Two groups of species could not differentiate between strains of . and be distinguished with any of the enzymes Ga xylinus . , but they were able to distinguish tested: one formed by Ga europaeus Gluconoacetobacter strains of . . The technique has also , ,and Ga liquefaciens liquefaciens Ga. xylinus Gluconoacetobacter been used to identify isolates from grapes in and the other by europaeus Gluconobacter Chile (Prieto et al., 2007). Gullo et al. (2006) and .Inorderto frateurii Gluconobacter asaii used a combination of the two previous tech- differentiate between these groups and niques to identify isolates from traditional between other newly described species, balsamic vinegar. The entire region comprising a system was proposed involving the the 16S-23S-5S genes and the ITS regions was sequential use of different restriction enzymes amplified. However, the length of the amplicon to allow species to be grouped and (approximately 4500 base pairs) can represent distinguished from others according to the a technological barrier in this method. patterns obtained (Gonza´lez et al., 2006b). All of these techniques have been used on The system described above was designed for plate isolates and are therefore also affected by the identification of isolates of acetic acid bacteria the problems mentioned earlier in relation to derived from musts and wines. Isolates from the culture of acetic acid bacteria. The method industrial fermentations have also been identi- has occasionally been used on noncultured fied by comparison of their restriction profiles samples, but with inconclusive results when with those obtained from reference strains. All multiple species were present (Ilabaca et al., of the isolates had profiles that were identical to 2008). Our group has circumvented the problem those described previously for the different of culturing acetic acid bacteria by directly quan- species. The restriction patterns with TaqIfor tifying acetic acid bacteria on different some of these wine strains are shown in substrates using real-time quantitative PCR. Figure 9.2. The technique has also been used for Briefly, this technique allows the amplification the identification of acetic acid bacteria present process to be monitored continually and ´ in grapes (Prieto et al., 2007), wines (Gonzalez a threshold cycle (CT) to be determined, indi- et al., 2004, 2005), vinegars (Ilabaca et al., 2008), cating the point at which exponential TAXONOMY 239

TABLE 9.3 Size of Restriction Fragments Obtained with TaqI and RsaI Following Digestion of the 16S Ribosomal Gene

Strain TaqI RsaI

T G. oxydans LMG 1408 350 þ190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 400 þ 150 þ 90 G. oxydans CECT 360 350 þ 190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 400 þ 150 þ 90 G. oxydans LMG 1484 350 þ 190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 400 þ 150 þ 90 G. oxydans LMG 1414 350 þ 190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 400 þ 150 þ 90 T G. frateurii LMG 1365 350 þ 190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 300 þ 150 þ 130 T G. asaii LMG 1390 350 þ 190 þ 175 þ 160 þ 120 þ 120 þ 110 400 þ 400 þ 300 þ 150 þ 130 T A. aceti LMG 1261 850 þ 350 þ 210 500 þ 400 þ 300 þ 150 þ 125 T A. aceti CECT 298 850 þ 350 þ 210 500 þ 400 þ 300 þ 150 þ 125 A. aceti LMG 1505 850 þ 350 þ 210 500 þ 400 þ 300 þ 150 þ 125 A. aceti LMG 1372 850 þ 350 þ 210 500 þ 400 þ 300 þ 150 þ 125 T A. pasteurianus LMG 1262 500 þ 350 þ 330 þ 210 500 þ 400 þ 300 þ 150 þ 125 A. pasteurianus LMG 1553 500 þ 350 þ 330 þ 210 500 þ 400 þ 300 þ 150 þ 125 T Ga. hansenii LMG 1527 650 þ 350 þ 210 þ 175 500 þ 400 þ 400 þ 150 T Ga. liquefaciens LMG 1381 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 400 þ 150 Ga. liquefaciens LMG 1347 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 400 þ 150 T Ga. xylinus LMG 1515 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 400 þ 150 Ga. xylinus LMG 1518 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 400 þ 150 T Ga. europaeus DSM 6160 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 400 þ 150 T Ga. diazotrophicus DSM 5601 500 þ 350 þ 210 þ 175 þ 160 500 þ 400 þ 250 þ 150 þ 150 T Ac. Methanolica LMG 1668 450 þ 190 þ 175 þ 160 þ 120 þ 110 500 þ 400 þ 400 þ 150

A. ¼ Acetobacter; Ac. ¼ Acidomonas; G. ¼ Gluconobacter; Ga. ¼ Gluconacetobacter. Adapted from Ruiz et al. (2000). amplification begins. This will be proportional (Gonza´lez et al., 2006a) that have been success- to the amount of DNA present in the original fully used in wine (Andorra` et al., 2008) and sample, and therefore standard curves can be vinegars (Jara et al., 2008). Likewise, specific established to plot the number of cycles against probes have been designed for use in real-time the concentration of DNA (or, in fact, cells) PCR (TaqMan probes) for the most common present in the sample. The specificity or identity species found on grapes and in wine and of the cells or DNA amplified will be determined vinegar (Torija et al., 2009). This method has by the specificity of the primers. We have also allowed simultaneous identification and quanti- designed primers for the simultaneous identifi- fication in these substrates and has been cation and quantification of acetic acid bacteria successfully used for the identification and 240 9. ACETIC ACID BACTERIA

FIGURE 9.2 TaqI restriction profiles following amplification of 16S ribosomal DNA in different strains of acetic acid bacteria isolated during alcoholic fermentation. All strains from the same species cter hansenii displayed the same profile. ter oxydans m ¼ molecular weight marker (100-base-pair ladder; Gibco- BRL). Gluconacetoba Gluconobacter oxydans AcetobacterGluconobac aceti Gluconoacetobacter liquefaciensAcetobacter pasteurianus m m

(bp)

1500

600

100

quantification of the species present in vinegar 4.4. Molecular Techniques for Typing made using traditional methods (Jara, 2009). of Acetic Acid Bacteria We have also used other techniques that do not require culture, such as cloning of amplified The main objective of any form of microbial DNA from the 16S gene in Escherichia coli and classification is to identify isolates in terms of subsequent analysis of transformed colonies by the species to which they belong, this being restriction fragment length polymorphism the fundamental taxonomic unit. However, the (RFLP) (Ilabaca et al., 2008). In this technique, discrimination or typing of different strains or although the problem of culture is resolved, genotypes from a given species plays an increas- that of transformation efficiency is introduced, ingly important role in industrial applications, since not all fragments are amplified similarly. since not all strains of the same species generate Finally, denaturing gradient gel electrophoresis similar changes in the product. Molecular tech- (DGGE) and temperature gradient gel electro- niques have the advantage that they allow phoresis (TGGE) can be used in combination within-species discrimination. In industrial with PCR to separate amplified fragments of winemaking, typing of Saccharomyces strains the 16S gene according to small sequence differ- has been important in the selection of autoch- ences. This technique has been used successfully thonous strains for use in industrial starter for the identification of acetic acid bacteria in cultures during alcoholic fermentation (Querol wine (Andorra` et al., 2008; Lo´pez et al., 2003) et al., 1992). Although acetic acid bacteria do and vinegars (de Vero et al., 2006; Gullo et al., not have desirable biotechnological properties 2009; Haruta et al., 2006; Ilabaca et al., 2008). for use in winemaking, the selection of strains GROWTH OF ACETIC ACID BACTERIA IN WINEMAKING PROCESSES 241 for use in the vinegar industry has been of ABCD particular interest (Sokollek et al., 1998). Strain typing has therefore focused on the microbial I 01 02 20 28 15 16 II 17 09 29 30 32 I characterization of the acetification process (Teuber et al., 1987). In the study by Teuber et al. (1987), and also in subsequent studies (Mariette et al., 1991; Sokollek et al., 1998), the plasmid profile of each of the strains was used as a molecular marker. Ohmori et al. (1982) demonstrated the presence of plasmids in strains isolated during vinegar production, and Teuber et al. (1987) considered the plasmid profile to be characteristic of the strain. Nanda et al. (2001) have also characterized strains of acetic acid bacteria isolated in vinegar, although in this case the samples were from rice vinegar. Two PCR-based methods have been FIGURE 9.3 Patterns obtained for strains of different used in this type of characterization: enterobac- species of acetic acid bacteria by enterobacterial repetitive terial repetitive intergenic consensus (ERIC) intergenic consensus polymerase chain reaction. A ¼ PCR and random amplification of polymorphic Gluconobacter oxydans;B¼ Gluconoacetobacter liquefaciens; DNA (RAPD). RAPD involves the use of short C ¼ Gluconoacetobacter hansenii;D¼ Acetobacter aceti. oligonucleotides (9e10 nucleotides) that hybridize with random sequences in the bacte- in each case (see Figure 9.3). Therefore, these rial genome; the number and size of the frag- techniques can be considered appropriate for ments amplified is different for each strain. analysis of variability beyond the species level This technique has also been used recently by in acetic acid bacteria. This technique has also Bartowsky et al. (2003) to type strains of acetic been used for monitoring the growth dynamics acid bacteria isolated from bottled wine with of acetic acid bacteria in vinegars (Gullo et al., symptoms of being pricked. Our group has 2009). A final technique for the typing of acetic used the ERIC-PCR technique alongside acid bacteria involves the use of repetitive another technique known as repetitive extra- (GTG)5 sequences (abundant in all genomes) genic palindrome (REP) PCR to study the strain for the monitoring and characterization of acetic diversity of acetic acid bacteria during alcoholic acid bacteria present during fermentation of fermentation (Gonza´lez et al., 2004). ERIC and cocoa beans (Camu et al., 2007). REP elements were described as consensus sequences derived from repetitive sequences spread throughout the bacterial chromosome 5. GROWTH OF ACETIC ACID of enterobacter species (Versalovic et al., 1991). BACTERIA IN WINEMAKING However, these sequences have been found in PROCESSES other bacterial groups. The technique involves the use of primers to amplify sequences According to the excellent review by between these repetitive regions. By employing Drysdale and Fleet (1988), “further studies are specific oligonucleotides against the DNA of needed to more accurately determine the strains from different species of acetic acid growth behavior of acetic acid bacteria during bacteria, characteristic patterns can be obtained the different stages of vinification. Such studies 242 9. ACETIC ACID BACTERIA should examine the influences of different 2005). All of the isolates were genotyped and conditions of vinification, the significance of G. oxydans was found to display considerable the yeast-to-bacteria ratio in the must, and vari- strain diversity, whereas the diversity was ations in the behavior of different species and substantially lower in isolates belonging to strains of acetic acid bacteria.” Despite the A. aceti (Gonza´lez et al., 2005). We also recovered time that has passed, few studies have been Ga. hansenii from grapes and must. Other published on this subject. However, a series authors have also observed the presence of of conclusions can be drawn from those that A. pasteurianus and Ga. liquefaciens in fresh musts, have. although these species were in the minority (du Toit & Lambrechts, 2002). Finally, the only 5.1. Association of Acetic Acid Bacteria species of acetic acid bacteria present in a study With Grapes of Chilean grapes were G. oxydans and A. cerevisiae (Prieto et al., 2007). Those species displayed The number of acetic acid bacteria present in a specific distribution according to latitude and must is directly proportional to the health of the only overlapped in a transition zone between grapes. While the numbers are low in musts the two regions. Interestingly, A. cerevisiae had derived from healthy grapes (no more than not been described previously in grapes. 2 3 10 e10 colony-forming units [CFU]/mL), they However, prior to 2002 it was classified as A. increase by various orders of magnitude in pasteurianus (Cleenwerck et al., 2002), and it damaged grapes and grapes infected with is therefore difficult to determine whether the the fungus Botrytis cinerea (Joyeux et al., 1984). species is peculiar to Chile or whether it has Joyeux et al. (1984) also identified the majority previously been identified as A. pasteurianus. species found on the grapes. G. oxydans was the main species isolated from healthy grapes 5.2. Growth Dynamics of Acetic Acid and those not infected with Botrytis, whereas Bacteria During Alcoholic Fermentation A. aceti and A. pasteurianus predominated in unhealthy grapes. As discussed in Section 5.2, Although the growth dynamics of acetic acid G. oxydans has a poor tolerance of ethanol. bacteria will depend on the initial numbers of Therefore, synthesis of this compound by each population in the must, most studies damaged grapes or grapes at various stages of agree that these bacteria are unlikely to grow rotting may explain the replacement of this during alcoholic fermentation and that their 1 3 species with Acetobacter species, which have numbers reduce drastically to 10 e10 CFU/ a much greater alcohol tolerance. In our group, mL (du Toit & Lambrechts, 2002; Gonza´lez we have carried out counts of acetic acid bacteria et al., 2004; Joyeux et al., 1984). The massive in successive harvests and our results support production of carbon dioxide by the yeast and the conclusions of earlier studies. In a harvest therefore the establishment of anaerobic condi- in which the grapes were extraordinarily tions in the fermentation medium make it healthy, the bacterial counts in the must did almost impossible for this group of obligate not exceed 103 CFU/mL and more than 80% of aerobes to proliferate (Drysdale & Fleet, 1988). the identified colonies belonged to G. oxydans. In terms of the growth of different species, all In contrast, after a very wet summer that led to of these studies agree that the main species substantial rotting and Botrytis infection, the present in the must, G. oxydans, gradually disap- counts in the corresponding musts reached pears during the first few days of fermentation around 106 CFU/mL and the predominant and is rarely found in wine. This species is species was A. aceti (Gonza´lez et al., 2004, much more competitive in substrates containing GROWTH OF ACETIC ACID BACTERIA IN WINEMAKING PROCESSES 243 high concentrations of sugar but its intolerance in the population of cultivable acetic acid of ethanol means that its numbers diminish as bacteria caused by both addition of sulfite and alcoholic fermentation advances. The majority inoculation of yeast starter cultures (in this species isolated during alcoholic fermentation case occurring in parallel with the rapid onset are A. aceti and A. pasteurianus (Drysdale & of alcoholic fermentation). Notably, analysis of Fleet, 1985; Joyeux et al., 1984), although du strains revealed that some grape-derived strains Toit and Lambrechts (2002) recently reported survived throughout alcoholic fermentation and a significant presence of the species Ga. liquefa- some were enriched in the must after it entered ciens and Ga. hansenii. We have obtained similar the winery. In other words, there is a significant results in isolates from alcoholic fermentation resident population in the winery but this does (see Figure 9.4)(Gonza´lez et al., 2004, 2005). not solely account for the bacteria recovered at Our group has analyzed the effects of some the end of fermentation (Gonza´lez et al., 2005). winemaking practices on populations of acetic The use of independent culture techniques acid bacteria. We have studied, for instance, showed that both A. aceti and Ga. hansenii are the addition of sulfite, inoculation of selected present throughout alcoholic fermentation and yeasts, and their combined effect in relation to are clearly detectable by DGGE. It should be the growth dynamics of different strains and noted that, since this technique only recovers species (Gonza´lez et al., 2005) as well as on the populations larger than 103 cells/mL, both overall population of acetic acid bacteria using species must have exceeded this population independent culture techniques (Andorra` size during fermentation. This is confirmed by et al., 2008). The predominance of A. aceti during the finding that the total population of acetic alcoholic fermentation was confirmed along acid bacteria measured by quantitative PCR with the observation of a considerable reduction remains close to 104 cells/mL throughout alcoholic fermentation. Notably, neither of these observations was affected by inoculation of 100% yeast starter cultures or addition of sulfite 90% (Andorra` et al., 2008), suggesting that the low recovery of colonies on plates may be due to 80% the induction of viable but noncultivable cell 70% states (Millet & Lonvaud-Funel, 2000). 60% The effects of standard winemaking procedures such as control of temperature, macera- 50% tion, micro-oxygenation, etc., during alcoholic 40% fermentation have not been studied in detail, 30% although in general terms the filtration or clarifi- 20% cation of must is thought to reduce the population of acetic acid bacteria and therefore the 10% risk of their growth. Nevertheless, once alcoholic 0% fermentation is complete, racking causes aera- Gluconoacetobacter liquefaciens Gluconobacter oxydans tion of the wine and can lead to proliferation Gluconoacetobacter hansenii Acetobacter aceti of acetic acid bacteria. Consequently, these populations can reach new titers of around FIGURE 9.4 Species profiles in must and at the begin- 6 ning, middle, and end of alcoholic fermentation of a Gar- 10 CFU/mL in stored wine (Drysdale & Fleet, nacha grape variety in the 2001 and 2002 vintages (Gonza´lez 1985), especially if the wine is not stored et al., 2004). under anaerobic conditions. Joyeux et al. (1984) 244 9. ACETIC ACID BACTERIA demonstrated that these populations of acetic acid bacteria. The latter possibility may be sup- acid bacteria that survive the alcoholic fermenta- ported by the observation that malolactic tion process can proliferate rapidly in stored fermentation also causes the release of carbon wine when enriched with 7.5 mg/L of oxygen. dioxide, although in proportions that are much This growth is accentuated when the storage lower than those seen during alcoholic fermen- temperature and pH of the wine are high. Barrel tation, and therefore creates anaerobic condi- aging of wines can be considered an anaerobic or tions. Nevertheless, the racking processes semi-anaerobic process, and, therefore, acetic performed prior to malolactic and after alco- acid bacteria would be expected to have little holic fermentation cause sufficient aeration to likelihood of proliferating. However, Joyeux stimulate the growth of acetic acid bacteria. In et al. (1984) also reported that oxygen penetrates an effort to answer some of these questions, the barrel in quantities of around 30 mg/L per populations of lactic acid and acetic acid year, which is sufficient for small populations bacteria were analyzed alongside the concentra- of acetic acid bacteria to survive. Drysdale and tions of malic and acetic acid during malolactic Fleet (1985) confirmed this capacity to survive fermentation (Guillamoń et al., 2003) (see and even grow under semi-anaerobic conditions Figure 9.5). Lactic acid bacteria grew to reach by isolating acetic acid bacteria from samples 108 CFU/mL within a few days. Interestingly, taken within the barrel. In addition, the racking this growth coincided with substantial growth performed during aging and even bottling leads of acetic acid bacteria. Subsequently, both popu- to new increases in oxygen concentration and the lations diminished to levels similar to those proliferation of acetic acid bacteria (Millet & present at the beginning of the process. Lonvaud-Funel, 2000). Consumption of malic acid began when the Finally, it is worth mentioning that there has population of lactic acid bacteria reached its been little analysis of the growth of acetic acid maximum, and the growth of acetic acid bacteria during malolactic fermentation. Conse- bacteria was also reflected in an increase in the quently, it is not known whether there is any concentration of acetic acid. Therefore, there degree of synergy between the growth of lactic could be some degree of synergy between lactic acid bacteria and that of acetic acid bacteria. acid and acetic acid bacteria, which could grow Alternatively, as occurs with yeasts, the growth in parallel during malolactic fermentation. In of lactic acid bacteria may inhibit that of acetic a similar study, Joyeux et al. (1984) observed

1.E+08 2.1 1.E+07 1.8 1.E+06 1.5 Lactic acid bacteria 1.E+05 1.2 Acetic acid bacteria

1.E+04 0.9 g/L L-Malic acid CFU/ml Acetate 1.E+03 0.6 1.E+02 0.3 1.E+01 0 0 5 10 15 20 Malolactic fermentation (days)

FIGURE 9.5 Dynamics of population growth in acetic acid and lactic acid bacteria during malolactic fermentation (continuous lines and filled symbols). Consumption of malic acid and production of acetic acid are also shown (dotted lines and empty symbols). CFU ¼ colony-forming units. FACTORS DETERMINING THE GROWTH OF ACETIC ACID BACTERIA: BACTERIAL CONTROL METHODS 245 that prior growth of G. oxydans or A. aceti in the 6. FACTORS DETERMINING THE must could stimulate malolactic fermentation GROWTH OF ACETIC ACID by Oenococcus oeni. However, these data should BACTERIA: BACTERIAL CONTROL be confirmed in studies of malolactic fermenta- METHODS tion and, in particular, in wines with varying degrees of aeration. The main physicochemical properties that There are some differences of opinion influence the growth of acetic acid bacteria regarding the majority species growing in during winemaking are the pH of the must/ stored wines or during aging. In Bordeaux wine, the temperature, and the concentrations wines, A. aceti is reportedly present in larger of ethanol, sulfite, and, most importantly, numbers than A. pasteurianus (Joyeux et al., oxygen dissolved in the medium (Drysdale & 1984). In contrast, Drysdale and Fleet (1985) Fleet, 1988). Although few systematic studies found that the majority of isolates in Australian have addressed the effect of these parameters wines corresponded to A. pasteurianus. In more on the growth of acetic acid bacteria, the recent studies, A. pasteurianus was found to most relevant conclusions of the studies be the predominant species in South African that have been performed are summarized wines, although on occasion Ga. liquefaciens below. was also isolated in large numbers (du Toit & Lambrechts, 2002). Finally, in all of the studies 6.1. pH performed by our group, A. aceti was practically the only species isolated at the end of fermenta- Optimal growth of acetic acid bacteria tion, during malolactic fermentation, and occursatapHofbetween5and6(de Ley during storage of wines (Gonza´lez et al., 2004, et al., 1984). Of course, the pH of wine is 2005; Guillamoń et al., 2003), although Ga. hanse- much lower and it has been clearly demon- nii may also be present in notable quantities strated that these bacteria can grow in this (Andorra` et al., 2008). The presence of A. tropica- medium. Thus, although the low pHs found lis has also been observed in Austrian wines that in wine generally inhibit bacterial growth, in were spontaneously fermented and later aceti- thecaseofaceticacidbacteria,theymaybe fied (Silhavy & Mandl, 2006). limiting for proliferation but not for survival, Although any of the acetic acid bacteria that since these bacteria have been found in wines have been described can produce acetic acid, with a pH of 3 (Drysdale & Fleet, 1985). Never- wine spoilage is most commonly linked to theless, Joyeux et al. (1984) observed that the A. pasteurianus, especially when spoilage occurs growth of A. aceti was lower in a wine with in the bottle, where characteristically a flor or a pH of 3.4 than in one with a pH of 3.8. Studies biofilm can form on the surface of the wine and have not addressed the pH resistance of is easily detected by the presence of a ring of different species of acetic acid bacteria isolated residue (Bartowsky et al., 2003). The development from wine. However, certain strains are known of these rings or biofilms is linked to the for- to be particularly resistant, since, under aerated mation of air pockets beneath the cork and can conditions such as those found during the be avoided by storing the bottles in a horizontal production of vinegars, acetic acid bacteria position (Bartowsky & Henschke, 2008). Recently, survive and proliferate in highly acidic media a new species of acetic acid bacteria, Acetobacter with a pH of 2.0 to 2.3, even with limited aera- oeni, was described in wines from the Dao region tion such as in the traditional Orleans method. in Portugal that had been spoiled by Dekkera Furthermore, the effect of pH is synergistic species (Silva et al., 2006). with that of other growth inhibitors such as 246 9. ACETIC ACID BACTERIA ethanol and sulfite. The antimicrobial effect of 6.3. Ethanol sulfite increases in highly acidic media. When this compound is added to wine, an equilib- Ethanol is the main substrate for acetic acid rium is formed between molecular SO2 and bacteria during growth in wine. However, the the bisulfite form (HSO3 ). Although the bisul- degree of ethanol tolerance depends on the spe- fite form predominates in wine, only the molec- cies and even the strain. G. oxydans is the least ular species has antimicrobial properties. The tolerant of the species commonly isolated dur- lower the pH, the greater the proportion of ing winemaking. According to Bergey’s Manual SO2 versus HSO3 and, therefore, the greater (de Ley et al., 1984), only 42% of G. oxydans the antimicrobial effect (Ribe´reau-Gayon strains proliferate in a medium containing 5% et al., 2000). ethanol. Similarly, strains of A. aceti, Ga. hansenii, and Ga. liquefaciens were unable to grow in 6.2. Temperature medium containing 10% ethanol, and only 20% of A. pasteurianus strains grew in this medium. The optimal temperature for growth of ace- However, it is well known that acetic acid tic acid bacteria is between 25 and 30C(de bacteria may be present in wineries in media Ley et al., 1984), although A. pasteurianus containing higher concentrations of ethanol, may have an optimal temperature of around although it is also known that wines containing 20C(Vaughn, 1955). The maximum tempera- higher concentrations of ethanol (15% or more) ture that can be tolerated is estimated at are less likely to become pricked. Nevertheless, between 35 and 40C, depending on the strain strains of acetic acid bacteria have been isolated and species in question. In fact, a temperature from wines with ethanol concentrations above of between 28 and 32C is maintained during 13%; the limit is considered to be 15 or 15.5%, industrial vinegar production, since the reac- which corresponds to the minimum level recom- tion can be highly exothermic and there may mended for fortified wines. Clearly, then, wild be considerable increases in temperature that strains must be much more resistant than the stop the acetification process. In hot countries, collection strains in which these tests are usually resistance to high temperatures is a positive carried out. factor for the selection of acetic acid bacteria to be used in vinegar making (Ndoye et al., 6.4. Sulfite 2007). In terms of lower temperatures, a significant increase in the population of acetic acid The most widely used antimicrobial agent in bacteria has been observed in wines stored at winemaking is also active against acetic acid 18C(Joyeux et al., 1984). Growth has also bacteria. However, the effects of sulfite on the been detected, though at very low levels, in growth and survival of acetic acid bacteria wines stored at 10C. In fact, populations have not been studied in detail (Ribe´reau-Gayon have been observed to increase from 103 to et al., 2000). According to Lafon-Lafourcade and 105 CFU/mL during standard winemaking Joyeux (1981), the concentrations of sulfite that practices such as cold maceration prior to are generally used in winemaking are insuffi- inoculation (du Toit & Lambrechts, 2002). cient to prevent the growth of acetic acid Thus, while standard temperatures used bacteria. Those authors observed the growth of during wine storage or barrel aging may A. aceti in red wines containing 25 mg/L of slow growth, they do not appear to prevent free SO2. Du Toit et al. (2005) established that it. Of course, the higher the temperature, the 1.2 mg/L free SO2 had an observable effect on greater the growth rate. the viability of A. pasteurianus. On the other FACTORS DETERMINING THE GROWTH OF ACETIC ACID BACTERIA: BACTERIAL CONTROL METHODS 247 hand, complete inhibition of the growth of acetic revealed that micro-oxygenation stimulated acid bacteria in grape must was observed the growth of acetic acid bacteria (du Toit following addition of 100 mg/L total SO2 (Wata- et al., 2006), further studies are required to nabe & Ino, 1984). Acetic acid bacteria have assess how the process affects these bacterial a particular capacity to remain in substrates populations. such as wood; therefore, it is recommended that special measures are taken to clean barrels 6.6. Storage and Aging before they are reused. Following an analysis of the effect of treatment with sulfite, potassium Standard aging and storage processes also carbonate, bleach, and hot water, Wilker and present considerable risks for the growth of ace- Dharmadhikari (1997) concluded that hot water tic acid bacteria and wine spoilage. Aging is (85e88C for 20 min) was the most effective. Use usually carried out in oak barrels. Although of other preservatives (sorbic, fumaric, or these may not contain microorganisms when benzoic acid, etc.) proposed as alternatives to they are new, they are rapidly colonized follow- sulfite has not been assessed in detail for effect ing exposure to wine (Renouf et al., 2006). The on the survival of acetic acid bacteria. The porous nature of the wood that makes it so growing restrictions on the use of sulfite due appropriate for use in wine aging also makes to new legislation and a certain reluctance on it an appropriate habitat for a range of microor- the part of consumers may have undesirable ganisms, including acetic acid bacteria. Never- effects, particularly in terms of sensory quality, theless, under aging conditions, the availability as a result of a failure to control these of oxygen is considerably reduced and most microorganisms. acetic acid bacteria are found in viable but noncultivable states (Millet & Lonvaud-Funel, 6.5. Oxygen 2000). Although wineries usually employ various washing methods to allow barrels to Since acetic acid bacteria are obligate aerobes, be reused, only treatment with water at high oxygen becomes an authentic limiting factor for temperatures (greater than 85C) appears to be their growth. However, a number of acetic acid effective for the elimination of acetic acid bacteria can still grow despite the anaerobic bacteria (Wilker & Dharmadhikari, 1997). Wine conditions present during alcoholic fermenta- aging may be completed in the bottle, and this tion not being favorable for their growth. This offers a new opportunity for the proliferation explains why any wine exposed to air will of acetic acid bacteria. Bottles must be stored rapidly develop a biofilm on its surface that horizontally to prevent the formation of air mainly comprises acetic acid bacteria, although pockets in which acetic acid bacteria can grow yeasts may also grow. Clearly, then, while some (Bartowsky et al., 2003). This occurs particularly oxygen is necessary for wine maturation (Mas in aged red wines, since white wines are increas- et al., 2002), inadequate management of oxygen ingly treated by filter sterilization or similar levels will offer a clear advantage to acetic acid processes that limit the survival of microorgan- bacteria and cause spoilage. These minimal isms significantly (Bartowsky & Henschke, requirements for oxygen during maturation 2008). As mentioned earlier, the survival of ace- are usually managed through the use of micro- tic acid bacteria in wines may be due to the pres- oxygenation, an increasingly common practice ence of quinones that can function as terminal in wineries to accelerate the process of color electron acceptors and the aeration that occurs stabilization in the wines. Nevertheless, during the transfer of liquids between tanks, although the only study performed to date barrels, and bottles, which represent ideal 248 9. ACETIC ACID BACTERIA opportunities for growth of these microorgan- a combined effect of some yeasts and acetic isms (Joyeux et al., 1984). acid bacteria. These populations of acetic acid bacteria remain high in the must. In both substrates (grapes and must), glucose is the main 7. CHANGES OCCURRING IN carbon source for these bacteria. As mentioned WINE AS A RESULT OF THE earlier, this glucose is directly oxidized to glu- GROWTH OF ACETIC ACID conic acid, which is the main compound to BACTERIA accumulate in the medium. Although the presence of high concentrations of this acid in The growth of acetic acid bacteria in wine is grapes infected with Botrytis was until recently rapidly followed by the production of acetic considered a consequence of fungal metabo- acid and an increase in volatile acidity, lism, the main source is now known to be acetic producing what is often referred to as pricked acid bacteria, which are present in large wine. Volatile acidity is considered a defect at numbers in botrytized grapes (Barbe et al., levels above 0.4 to 0.5 g/L, depending on the 2001). Acetic acid bacteria also metabolize fruc- type of wine (Ribe´reau-Gayon et al., 2000); it tose, although in smaller quantities, to form 5- can reach levels up to 1.0 to 1.5 g/L in some oxofructose. sweet wines (botyrized wines or icewines) The capacity of acetic acid bacteria to (Nurgel et al., 2004). However, volatile acidity produce extracellular polysaccharides leads to is directly associated with the formation of another significant change due to sugar metabo- ethyl acetate, which appears rapidly during lism (Kouda et al., 1997). Some strains of the growth of acetic acid bacteria in wine. A. pasteurianus and G. oxydans produce cellulose Ethyl acetate has a notably low perception fibrils or other polysaccharides that impede threshold and is easily recognized by the smell filtration of the wines (Drysdale & Fleet, 1988). of glue or nail polish remover (Ribe´reau-Gayon The main consequence of the production of et al., 2000). Vinegar and glue are probably the these polysaccharides during winemaking is first noticeable aromas in wine spoiled by acetic the wine filtration difficulties that it generates. acid bacteria. However, this is not the only Acetic acid bacteria can also use other carbo- change that occurs as a result of the develop- hydrates such as arabinose, galactose, mannitol, ment of these bacteria. These changes in the mannose, ribose, sorbitol, and xylose (de Ley medium also depend on the point during vinifi- et al., 1984). Although these sugars are found cation at which growth occurs. at low concentrations in the must, some form part of the residual sugars found in wine as 7.1. Changes in Grapes and Must as a result of having not been used by yeasts. a Result of the Growth of Acetic Acid Consequently, these sugars can also be used by Bacteria acetic acid bacteria during growth in wine, though ethanol remains the main substrate. Populations of acetic acid bacteria increase Logically, acetic acid bacteria also produce in size over the course of ripening and large acetic acid in grapes and must through the populations are present in unhealthy grapes, metabolism of sugars, although the quantities especially those infected with B. cinerea. Never- produced are lower than those of gluconic theless, any damage to the skin of the grape can acid. It is believed that the acetic acid produced provide a route of entry for all types of microor- is mainly derived from the ethanol produced by ganism, including acetic acid bacteria. Conse- the yeast in grapes and must than through the quently, acid rotting is also considered to be metabolism of hexose sugars. CHANGES OCCURRING IN WINE AS A RESULT OF THE GROWTH OF ACETIC ACID BACTERIA 249

Another effect of sugar metabolism by acetic 200 mg/L. Its perception threshold is around acid bacteria is the high capacity of gluconic 125 mg/L, and at concentrations above 500 acid and oxofructose to bind SO2; this reduces mg/L it is considered detrimental to the quality the proportion of free SO2 and therefore its anti- of the wine as a result of the oxidized character microbial and antioxidant capacity. Barbe et al. it endows (Margalith, 1981). In addition to its (2001) reported that maintenance of 50 mg/L impact on the aroma and flavor of the wine, acet- of free SO2 in a synthetic must in which G. aldehyde has a greater capacity to bind SO2 and oxydans had grown required 3000 mg/L of total therefore reduce the levels of free SO2 (Ribe´reau- SO2 as a consequence of the high concentrations Gayon et al., 2000). As a consequence, wine with of gluconic acid (51 g/L), 5-oxofructose (6 g/L), a high concentration of acetaldehyde requires and dihydroxyacetone (2 g/L) formed from higher concentrations of SO2 to achieve good glucose, fructose, and glycerol, respectively. protection of wine during aging and bottling. The last two components can also bind SO2 effi- After ethanol, glycerol is the main product of ciently (du Toit & Pretorius, 2002). alcoholic fermentation and reaches concentrations of between 2 and 25 g/L. This metabolite 7.2. Changes in Wine as a Result of the is important in determining the quality of the Growth of Acetic Acid Bacteria wine as it contributes unctuous, syrupy, and viscous characteristics. Acetic acids also oxidize The conversion of ethanol into acetic acid in glycerol to dihydroxyacetone, which does not wine is the most widely recognized form of impart these positive characteristics in the spoilage due to acetic acid bacteria. This ethanol organoleptic qualities of the wine. In addition, represents the principal carbon source and its it binds easily to SO2. conversion into acetic acid is responsible for Acetic acid bacteria can also oxidize the the generation of pricked wine. The biochem- different acids present in wine. Drysdale and istry of this reaction has been widely studied Fleet (1989b) observed a reduction in malic, tarta- due to its importance in the production of ric, and citric acids present in wines contaminated vinegar. Some strains of acetic acid bacteria with acetic acid bacteria. These acids, and others can produce up to 150 g/L of acetic acid during such as lactic and fumaric acid, would be comp- vinegar production (Sievers et al., 1997). Such letely oxidized to carbon dioxide and water via high concentrations are only obtained in highly the Krebs cycle (only in those species other than oxygenated cultures. Although this clearly does Gluconobacter that have a functional Krebs cycle). not occur during the winemaking process, vola- Some strains of Acetobacter and Gluconobacter, tile acidity is easily increased to above 0.8 g/L, particularly A. pasteurianus strains, can oxidize a concentration considered detrimental to the lactic acid to acetoin, which produces a buttery quality of wine since it is well above the detec- aroma and flavor (Drysdale & Fleet, 1988). tion threshold for acetic acid. Ethyl acetate is another compound produced As mentioned, ethanol is first oxidized to acet- by acetic acid bacteria that has a negative effect aldehyde, which is then oxidized to form acetic on the sensory quality of wine. This ester has acid. Since acetaldehyde is a metabolic interme- a very low perception threshold and is detecta- diate in this reaction, it is one of the most impor- ble at concentrations as low as 10 mg/L (Berg tant products after acetic acid. Its levels increase et al., 1955). The growth of acetic acid bacteria in wine with decreasing concentrations of dis- can increase the concentration of ethyl acetate solved oxygen (Drysdale & Fleet, 1989a). Yeasts to 140 mg/L in wine and 30 mg/L in must also produce acetaldehyde during alcoholic (Drysdale & Fleet, 1989a). Finally, acetic acid fermentation at concentrations of around 20 to bacteria can also oxidize higher alcohols such as 250 9. ACETIC ACID BACTERIA isoamyl alcohol, propanol, and 2-phenyl ethanol Little is known about the effect of acetic acid in wine to their corresponding aldehydes and bacteria on the growth of lactic acid bacteria. carboxylic acids (Molinari et al., 1999). Gilliland and Lacey (1964) reported that a strain Importantly, most of the acetic acid is intro- of Acetobacter inhibited the growth of Lactoba- duced into the wine during the stationary and cillus species, and Joyeaux et al. (1984) found death phases of acetic acid bacteria and not that acetic acid bacteria stimulated malolactic during the growth phase (Ko¨sebalan & O¨ zingen, fermentation. 1992). Although most acetic acid is produced in Bradley (1965) demonstrated the presence of an extracellular reaction, a proportion can accu- bacteriophages that were active against Aceto- mulate inside the cell and is then released when bacter species. Subsequently, Sellmer et al. the cells die. Thus, although the numbers of (1992) demonstrated that phages could place these bacteria increase as a result of aeration the viability of acetic acid bacteria cultures at (during specific winemaking processes such as risk during the production of vinegar. It is there- pump-over, racking, etc.), the increase in volatile fore likely that phages can affect the growth and acidity will probably not occur until storage and survival of acetic acid bacteria over the course of aging. wine production.

8. INTERACTIONS WITH OTHER 9. FINAL RECOMMENDATIONS TO MICROORGANISMS IN WINE AVOID WINE SPOILAGE DUE TO ACETIC ACID BACTERIA Grape must contains a wide range of species of yeast, lactic acid bacteria, and acetic acid In summary, although acetic acid bacteria bacteria, and their interactions during alcoholic need oxygen, they can survive and even grow fermentation may be complex (Fleet, 2003). to some extent in an almost completely anaer- Joyeux et al. (1984) showed that prior growth of obic medium such as wine. Clearly, then, it is G. oxydans or A. aceti in grape must could result impossible to remove all risk of contamination in stuck fermentation. Drysdale and Fleet by these bacteria during winemaking. However, (1989b) studied the effect of simultaneous inocu- it is possible to balance the need for certain lation of yeast and bacteria in the must, as processes against the risk of bacterial growth. usually occurs in the winery. In this case, there The following considerations may help to was no substantial inhibition of yeast growth; prevent the undesirable growth of acetic acid however, the yeasts struggled to consume all of bacteria during winemaking: the sugars in the must. A. pasteurianus was found to produce greater inhibition of the fermentative 1) Management of vines and correction of capacity of yeasts. The mechanism by which ace- musts should focus on acidiﬁcation of the tic acid bacteria produce this antagonism of must, as this will help to prevent growth of yeasts has yet to be elucidated. Acetic acid is an acetic acid bacteria. Reducing the pH of the inhibitor of yeasts, but other substances may wine will not only reduce the risk of spoilage also be involved in this effect (du Toit & Lam- during production but also enhance the brechts, 2002). Clearly, this inhibition can only options available for aging. occur if there is a considerable delay in the onset 2) The health of the grapes will determine the of alcoholic fermentation, giving rise to the microbial load corresponding to acetic acid development of an excess of acetic acid bacteria bacteria. Reducing the population of acetic that can produce these inhibitory substances. acid bacteria at the beginning of fermentation REFERENCES 251

will reduce the size of the population at the References end of the process. These populations are Adachi, O., Miyagawa, E., Shinagawa, E., Matsushita, K., & reduced by clearing, filtration, addition of Ameyama, M. (1978). Purification and properties of sulfite, and inoculation of yeast starter particulate alcohol dehydrogenase from Acetobacter aceti. cultures. Musts obtained from unhealthy Agric. Biol. Chem., 42, 2331e2340. grapes should be immediately filtered and Andorra`, I., Landi, S., Mas, A., Guillamoń, J. M., & Esteve- inoculated with a yeast starter culture. Zarzoso, B. (2008). Effect of enological practices on microbial populations using culture-independent tech- However, none of these practices alone will niques. Food Microbiol., 25, 849e856. guarantee the elimination of acetic acid Asai, T. (1935). Taxonomic studies of acetic acid bacteria and bacteria. allied oxidative bacteria in fruits and a new classification 3) Rapid initiation and progression of of oxidative bacteria. J. Agr. Chem. Soc. Jap., 11,50e60. fermentation creates anaerobic conditions Azuma. (2008). Genomic comprehension for evolutional flexibility of Acetobacter pasteurianus. The 2nd Interna- that will impede the proliferation of acetic tional Conference on Acetic Acid Bacteria. Nagoya, acid bacteria. Japan, November 11e14. 4) Any process that involves aeration (pump- Barbe, J. C., de Revel, G., Joyeux, A., Bertrand, A., & Lonvaud- over, racking, etc.) will provide an Funel, A. (2001). Role of botrytized grape microorganisms opportunity for the growth of acetic acid in SO2 binding phenomena. J. Appl. Microbiol., 90,34e42. Barja, F., Mesa, M. M., Macıás, M., Bermudez, I., Cantero, D., bacteria and, therefore, favor an increase in & Lopez, J. (2003). Aspectos bioquı´micos, moleculares y volatile acidity. In addition, the recent morfolo´gicos de las bacterias ace´ticas. In A. Mas, & introduction of micro-oxygenation J. M. Guillamoń (Eds.), Primeras jornadas en IþDþienla techniques must be taken into consideration elaboracioń de vinagre de vino (pp. 17e20). Tarragona, as their effect on the growth of acetic acid Spain: Servei de publicacions. Bartowsky, E. J., & Henschke, P. A. (2008). Acetic acid bacteria has not yet been analyzed in detail. bacteria spoilage of bottled red wine e A review. Int. 5) Maintaining low temperatures during J. Food Microbiol., 125,60e70. fermentation, aging, and storage limits the Bartowsky, E. J., Xia, D., Gibson, R. L., Fleet, G. H., & growth of acetic acid bacteria. However, only Henschke, P. A. (2003). Spoilage of bottled red wine by temperatures below 10C can effectively acetic acid bacteria. Lett. Appl. Microbiol., 36, 307e314. Beijerinck, M. W. (1899). Sur les diverses espe`ces de bac- prevent bacterial growth. Cold maceration te´ries ace´tifiantes. Arch. Neerl. des Sciences exactes et Nat. alone is also insufficient to eliminate acetic Haarlem II 180e189. acid bacteria and can even lead to an increase Belly, R. T., & Claus, G. W. (1972). Effect of amino acids on in their numbers (Couasnon, 1999). the growth of Acetobacter suboxydans. Arch. Mikrobiol., 83, Although further studies are required, the 237e245. 6) Berg, H. W., Filipello, F., Hinreiner, E., & Webb, A. D. presence of residual sugars in wine and also (1955). Evaluation of thresholds and minimum differ- malolactic fermentation may increase the ence concentrations for various constituents of wines. I. likelihood of the growth of acetic acid Water solution of pure substances. Food Technol., 9, bacteria. 23e26. Bertalan, M., Albano, R., de Pa´dua, V., Rouws, L., Rojas, C., 7) Good winery hygiene is absolutely essential Hemerly, A., et al. (2009). Complete genome sequence of to reduce the risk of contamination with the sugarcane nitrogen-fixing endophyte Gluconaceto- acetic acid bacteria, especially in wines with bacter diazotrophicus Pal5. BMC Genom., 10, 450. a lower alcohol content and during barrel Bradley, D. E. (1965). The isolation and morphology of some aging. It should be remembered that solid new bacteriophages specific for Bacillus an Acetobacter species. J. Gen. Microbiol., 41, 233e241. porous supports such as wood provide good Buchanan, R. E., & Gibbons, N. E. (1974). In conditions for the survival of acetic acid Bergey’s manual of determinative bacteriology (8th ed.). Baltimore, MD: bacteria. Williams & Wilkins. 252 9. ACETIC ACID BACTERIA

Camu, N., de Winter, T., Verbrugghe, K., Cleenwerck, I., quality and aerobic wine-spoilage microorganisms of Vandamme, P., Takrama, J. S., et al. (2007). Dynamics and different South African red wines. South Afr. J. Enol. biodiversity of populations of lactic acid bacteria and Vitic., 27,57e67. acetic acid bacteria involved in spontaneous heap du Toit, W. J., & Pretorius, I. S. (2002). The ocurrence, control fermentation of cocoa beans in Ghana. Appl. Environ. and esoteric effect of acetic acid bacteria in winemaking. Microbiol., 73, 1809e1824. Ann. Microbiol., 52, 155e179. Carr, J. G., & Passmore, S. M. (1979). Methods for identi- du Toit, W. J., Pretorius, I. S., & Lonvaud-Funel, A. (2005). fying acetic acid bacteria. In F. A. Skinner, & The effect of sulphur dioxide and oxygen on the viability D. W. Lovelock (Eds.), Identification methods for microbi- and culturability of a strain of Acetobacter pasteurianus ologists (pp. 33e47). London, UK: Academic Press. and a strain of Brettanomyces bruxellensis isolated from Cleenwerck, I., & de Vos, P. (2008). Polyfasic taxonomy of wine. J. Appl. Microbiol., 98, 862e871. acetic acid bacteria: An overview of the currently Dutta, D., & Gachhui, R. (2007). Nitrogen-fixing and cellu- applied methodology. Int. J. Food Microbiol., 125,2e14. lose-producing Gluconacetobacter kombuchae sp. nov., Cleenwerck, I., Vandemeulebroecke, K., Janssens, D., & isolated from Kombucha tea. Int. J. Syst. Evol. Microbiol., Swings, J. (2002). Reexamination of the genus Acetobacter, 57, 353e357. with descriptions of Acetobacter cerevisiae sp. nov., and Entani, E., Ohmori, S., Masai, H., & Suzuki, K. I. (1985). Acetobacter malorum, sp. nov. Int. J. Syst. Evol. Microbiol., Acetobacter polyoxogenes sp. nov., a new species of an 52, 1551e1558. acetic acid bacterium useful for producing vinegar with Couasnon, M. B. (1999). Une nouvelle technique: La high acidity. J. Gen. Appl. Microbiol., 31, 475e490. maceration pre´fermentaire a` froid e Extraction a` la neige Fleet, G. H. (2003). Yeast interactions and wine flavour. Int. carbonique. 1re partie: Resultats oenologiques. Revue des J. Food Microbiol., 86,11e22. Oenologues et des Techniques Vitivinicoles et Oenologiques, Frateur, J. (1950). Trial on classification of. Acetobacter. 92,26e30. Cellule, 53, 287e392. de Ley, J. (1959). On the formation of acetoin by Acetobacter. Gilliland, R. B., & Lacey, J. P. (1964). Lethal action of Ace- J. Gen. Microbiol., 21, 352e365. tobacter on yeasts. Nature, 202, 727e728. de Ley, J., Gillis, M., & Swings, J. (1984). Family VI. Aceto- Gillis, M., Kersters, K., Hoste, B., Janssens, D., bacteriaceaea. In N. R. Krieg, & J. C. Holt (Eds.), Bergey’s Kroppenstedt, M., Stephan, M. P., et al. (1989). Aceto- manual of systematic bacteriology, Vol. 1 (pp. 267e278). bacter diazotrophicus sp. nov., a nitrogen-fixing acetic Baltimore, MD: Williams & Wilkins. acid bacterium associated with sugarcane. Int. J. Syst. de Vero, L., Gala, E., Gullo, M., Solieri, L., Landi, S., & Bacteriol., 39, 361e364. Giudici, P. (2006). Application of denaturing gradient gel Giudici, P., Gullo, M., Pulvirenti, A., Solieri, L., & De electrophoresis (DGGE) analysis to evaluate acetic acid Vero, L. (2003). Microflora y la tecnologıá de produccioń bacteria in traditional balsamic vinegar. Food Microbiol., del vinagre balsa´mico tradicional. In A. Mas, & 23, 809e813. J. M. Guillamoń (Eds.), Primeras jornadas en IþDþienla Deppenmeier, U., Hoffmeister, M., & Prust, C. (2002). elaboracioń de vinagre de vino (pp. 59e70). Tarragona, Biochemistry and biotechnological applications of Gluco- Spain: Servei de publicacions. nobacter strains. Appl. Microbiol. BioTechnol., 60, 233e242. Gonza´lez, A., Guillamoń, J. M., Mas, A., & Poblet, M. Drysdale, G. S., & Fleet, G. H. (1985). Acetic acid bacteria in (2006b). Application of molecular methods for routine some Australian wines. Food Technol. Aust., 37,17e20. identification of acetic acid bacteria. Int. J. Food Micro- Drysdale, G. S., & Fleet, G. H. (1988). Acetic acid bacteria in biol., 108, 141e146. winemaking: A review. Am. J. Enol. Vitic., 2, 143e154. Gonza´lez, A., Hierro, N., Poblet, M., Mas, A., & Drysdale, G. S., & Fleet, G. H. (1989a). The growth and Guillamoń, J. M. (2005). Application of molecular survival of acetic acid bacteria in wine at different methods to demonstrate species and strain evolution of concentrations of oxygen. Am. J. Enol. Vitic., 40,99e105. acetic acid bacteria population during wine production. Drysdale, G. S., & Fleet, G. H. (1989b). The effect of acetic Int. J. Food Microbiol., 102, 295e304. acid bacteria upon the growth and metabolism of yeast Gonza´lez, A., Hierro, N., Poblet, M., Mas, A., & during the fermentation of grape juice. J. Appl. Bacteriol., Guillamoń, J. M. (2006a). Enumeration and detection of 67, 471e481. acetic acid bacteria by real-time PCR and nested-PCR. du Toit, W. J., & Lambrechts, M. G. (2002). The enumeration FEMS Microbiol. Lett., 254, 123e128. and identification of acetic acid bacteria from South African Gonza´lez, A., Hierro, N., Poblet, M., Roze`s, N., Mas, A., & red wine fermentations. Int. J. Food Microbiol., 74,57e64. Guillamoń, J. M. (2004). Population dynamics of acetic du Toit, W. J., Lisjak, K., Marais, J., & du Toit, M. (2006). The acid bacteria in a red wine fermentation. J. Appl. Micro- effect of microoxygenation on the phenolic composition, biol., 96, 853e860. REFERENCES 253

Greenberg, D. E., Porcella, S. F., Stock, F., Wong, A., bacterial cellulose production by Acetobacter in aerated Conville, P. S., Murray, P. R., et al. (2006). Granulibacter and agitated culture. J. Ferment. Bioeng., 82, 124e127. bethesdensis gen. nov., sp. nov., a distinctive pathogenic Lafon-Lafourcade, S. (1985). Roˆle des microorganismes dans acetic acid bacterium in the family Acetobacteraceae. Int. la formation de substances combinant le SO2. Bull. OIV, J. Syst. Evol. Microbiol., 56, 2609e2616. 652, 590e604. Greenberg, D. E., Porcella, S. F., Zelazny, A. M., Lafon-Lafourcade, S., & Joyeux, A. (1981). Les bacte´ries Virtaneva, K., Sturdevant, D. E., Kupko, J. J., et al. (2007). ace´tiques du vin. Bull. OIV, 608, 803e829. Genome sequence analysis of the emerging human Lisdiyanti, P., Kawasaki, H., Widyastuti, Y., Saono, S., pathogenic acetic acid bacterium Granulibacter bethes- Seki, T., Yamada, Y., et al. (2002). Kozakia baliensis gen. densis. J. Bacteriol., 189, 8727e8736. nov. sp. nov., a novel acetic acid bacterium in the Guillamoń, J. M., Gonza´lez, A., Poblet, M., & Mas, A. (2003). a-Proteobacteria. Int. J. Syst. Evol. Microbiol., 52, 813e818. Development of molecular techniques for the analysis of Loganathan, P., & Nair, S. (2004). Swaminathania salitorans acetic acid bacteria during winemaking. In Lallemand. gen. nov., sp. nov., a salt-tolerant, nitrogen-fixing and (Ed.), Yeastebacteria interactions (pp. 45e49). Biarritz, phosphate-solubilizing bacterium from wild rice (Por- France: Lallemand Technical Meetings. teresia coarctata Tateoka. Int. J. Syst. Evol. Microbiol., 54, Gullo, M., Caggia, C., de Vero, L., & Giudici, P. (2006). 1185e1190. Characterisation of acetic acid bacteria in traditional Lo´pez, I., Ruiz-Larrea, F., Cocolin, L., Orr, E., Phister, T., balsamic vinegar. Int. J. Food Microbiol., 106, 209e212. Marshall, M., et al. (2003). Design and evaluation of PCR Gullo, M., de Vero, L., & Giudici, P. (2009). Succession of primers for analysis of bacterial populations in wine by selected strains of Acetobacter pasteurianus and other denaturing gradient gel electrophoresis. Appl. Environ. acetic acid bacteria in traditional balsamic vinegar. Appl. Microbiol., 69, 6801e6807. Environ. Microbiol., 75, 2585e2589. Maestre, O., Santos-Duenãs, I. M., Peinado, R., Jimeńez- Haruta, S., Ueno, S., Egawa, I., Hashiguchi, K., Fujii, A., Ot, C., Garcıá-Garcıá, I., & Mauricio, J. C. (2008). Nagano, M., et al. (2006). Succession of bacterial and Changes in amino acid composition during wine fungal communities during a traditional pot fermenta- vinegar production in a fully automatic pilot acetator. tion of rice vinegar assessed by PCR-mediated dena- Proc. Biochem., 43, 803e807. turing gradient gel electrophoresis. Int. J. Food Microbiol., Margalith, P. Z. (Ed.). (1981). Flavour microbiology. Spring- 109,79e89. field, IL: Charles Thomas Publishers. Ilabaca, C., Navarrete, P., Mardones, P., Romero, J., & Mariette, I., Schwarz, E., Vogel, R. F., & Hammes, W. P. (1991). Mas, A. (2008). Application of culture culture-indepen- Characterization by plasmid profile analysis of acetic acid dent molecular biology based methods to evaluate acetic bacteria from wine, spirit and cider acetators for indus- acid bacteria diversity during vinegar processing. Int. trial vinegar production. J. Appl. Microbiol., 71,134e138. J. Food Microbiol., 126, 245e249. Mas, A., Puig, J., Llado´, N., & Zamora, F. (2002). Sealing and Jara, C. (2009). Desarrollo de me´todos de biologıá molecular storage position effects on wine evolution. J. Food Sci., 67, para el ana´lisis directo de bacterias ace´ticas de vinagre. 1374e1378. Tarragona, Spain: Doctoral thesis, Universidad Rovira i Millet, V., & Lonvaud-Funel, A. (2000). The viable but non- Virgili. culturable state of wine microorgamisms during storage. Jara, C., Mateo, E., Guillamoń, J. M., Torija, M. J., & Mas, A. Lett. Appl. Microbiol., 30, 126e141. (2008). Evaluation of several methods for isolating high Molinari, F., Gandolfi, R., Aragozzini, F., Leon, R., & quality DNA in acetic acid bacteria in wine and vinegar. Prazeres, D. M. F. (1999). Biotransformations in two- Int. J. Food Microbiol., 128, 336e341. liquid phase systems. Production of phenylacetaldehyde Jojima, Y., Mihara, Y., Suzuki, S., Yokozeki, K., Yamanaka, S., & by oxidation of 2-phenylethanol with acetic acid Fudou, R. (2004). Saccharibacter floricola gen. nov., sp. nov., bacteria. Enzyme Microb. Technol., 25, 729e735. a novel osmophilic acetic acid bacterium isolated from Muraoka, H., Watab, Y., Ogasawara, N., & Takahashi, H. pollen. Int. J. Syst. Evol. Microbiol., 54, 2263e2267. (1983). Trigger damage by oxygen deficiency to the acid Joyeux, A., Lafon-Lafourcade, S., & Ribe´reau-Gayon, P. (1984). production system during submerged acetic acid Evolution of acetic acid bacteria during fermentation and fermentation with Acetobacter aceti. J. Ferment. Technol., storage of wine. Appl. Environ. Microbiol., 48,153e156. 61,89e93. Ko¨sebalan, F., & O¨ zingen, M. (1992). Kinetics of wine Nanda, K., Taniguchi, M., Ujike, S., Ishihara, N., Mori, H., spoilage by acetic acid bacteria. J. Chem. Tech. Biotechnol., Ono, H., et al. (2001). Characterization of acetic acid 55,59e63. bacteria in traditional acetic acid fermentations of rice Kouda, T., Naritomi, T., Yano, H., & Yoshinaga, F. (1997). vinegar (Komesu) and unpolished rice vinegar (Kurosu) Effects of oxygen and carbon dioxide pressures on produced in Japan. Appl. Environ. Microbiol., 67, 986e990. 254 9. ACETIC ACID BACTERIA

Ndoye, B., Cleenwerck, I., Engelbeen, K., Dubois- unfavorable phenomena in vinegar fermentation. Biosci. Dauphin, R., Guiro, A. T., van Trappen, S., Willems, A., & Biotech. Biochem., 61, 317e323. Thonart, P. (2007). Acetobacter senegalensis sp nov., a ther- Saeki, A., Theeragool, G., Matsushita, K., Taniguchi, M., motolerant acetic acid bacterium isolated in Senegal (sub- Toyama, H., Lotong, N., et al. (1997). Development of Saharan Africa) from mango fruit (Mangifera indica L.). Int. thermotolerant acetic acid bacteria useful for vinegar J. Syst. Evol. Microbiol., 57, 1576e1581. fermentation at higher temperatures. Biosci. Biotech. Nurgel, C., Pickering, G. J., & Inglis, D. L. (2004). Sensory Biochem., 61, 138e145. and chemical characteristics of Canadian ice wines. Sellmer, S., Sievers, M., & Teuber, M. (1992). Morphology, J. Sci. Food Agric., 84, 1675e1684. virulence and epidemiology of bacteriophage particles Ohmori, S., Uozumi, T., & Beppu, T. (1982). Loss of acetic isolated from industrial vinegar fermentations. System. acid resistance and ethanol oxidizing ability in an Ace- Appl. Microbiol., 15, 610e616. tobacter strain. Agric. Biol. Chem., 46, 381e389. Sievers, M., Lorenzo, A., Gianotti, S., Boesch, C., & Teuber, M. Pasteur, L. (Ed.). (1868). Etudes sur le vinaigre. Paris, France: (1996). 16e23S ribosomal RNA spacer regions of Aceto- V. Masson. bacter europaeus and A. xylinum, tRNA genes and anti- Poblet, M., Roze`s, N., Guillamoń, J. M., & Mas, A. (2000). termination sequences. FEMS Microbiol. Lett., 142,43e48. Identification of acetic acid bacteria by restriction frag- Sievers, M., Stockli, M., & Teuber, M. (1997). Purification ment length polymorphism analysis of a PCR-amplified and properteies of citrate synthetase from Acetobacter fragment of the gene coding for 16S rRNA. Lett. Appl. europaeus. FEMS Microbiol. Lett., 146,53e58. Microbiol., 31,63e67. Silhavy, K., & Mandl, K. (2006). Acetobacter tropicalis in Polo, M. C., & Sańchez-Luengo, A. A. (1991). Las bacterias spontaneously fermented wines with vinegar fermenta- ace´ticas. In C. Llaguno, & M. C. Polo (Eds.), El vinagre de tion in Austria. Mitteilungen Klosterneuburg, 56, 102e107. vino (pp. 25e68). Madrid, Spain: EBCOMP. Silva, L. R., Cleenwerck, I., Rivas, R., Swings, J., Prieto, C., Jara, C., Mas, A., & Romero, J. (2007). Application Trujillo, M. E., Willems, A., et al. (2006). Acetobacter oeni of molecular methods for analysing the distribution and sp. nov., isolated from spoiled red wine. Int. J. Syst. Evol. diversity of acetic acid bacteria in Chilean vineyards. Int. Microbiol., 56,21e24. J. Food Microbiol., 115, 348e355. Sokollek, S. J., Hertel, C., & Hammes, W. P. (1998). Culti- Prust, C., Hoffmeister, M., Liesegang, H., Wiezer, A., vation and preservation of vinegar bacteria. J. Biotechnol., Fricke, W. F., Ehrenreich, A., et al. (2005). Complete 60, 195e206. genome sequence of the acetic acid bacterium Glucono- Stackebrandt, E., Murray, R. G. E., & Tru¨ per, H. G. (1988). bacter oxydans. Nat. Biotechnol., 23, 195e200. Proteobacteria classis nov., a name for the phylogenetic Querol, A., Barrio, E., Huerta, T., & Ramoń, D. (1992). taxon that includes the purple bacteria and their rela- Molecular monitoring of wine fermentations conducted tives. Int. J. Syst. Evol. Microbiol., 38, 321e325. by dry yeast strains. Appl. Environ. Microbiol., 58, Teuber, M., Sievers, M., & Andresen, A. (1987). Character- 2948e2952. ization of the microflora of high acid submerged vinegar Raj, K. C., Ingram, L. O., & Maupin-Furlow, J. A. (2001). fermenters by distinct plasmid profiles. Biotech. Lett., 9, Pyruvate decarboxylase: A key enzyme for the oxidative 265e268. metabolism of lactic acid by Acetobacter pasteurianus. Torija, M. J., Mateo, E., Guillamoń, J. M., & Mas, A. (2009). Arch. Microbiol., 176, 443e451. Design and optimisation of TaqMan-MGB probes for the Renouf, V., Perello, M. C., Strehaiano, P., & Lonvaud- identification and quantification of acetic acid bacteria Funel, A. (2006). Global survey of the microbial species. Food Microbiol. doi. 10.1016/j.fm.2009.10.001. ecosystem during alcoholic fermentation in wine- Trcek, J. (2005). Quick identification of acetic acid bacteria making. J. Int. Sci. Vigne Vin, 40, 101e116. based on nucleotide sequences of the 16S-23S rDNA Ribe´reau-Gayon, P., Dubordieu, D., Done`che, B., & internal transcribed spacer region and of the PQQ- Lonvaud, A. (Eds.). (2000). Handbook of enology, the dependent alcohol dehydrogenase gene. Syst. Appl. microbiology of wine and vinifications. Chichester, UK: Microbiol., 28, 735e745. John Wiley & Sons. Trcek, J., & Teuber, M. (2002). Genetic and restriction anal- Ruiz, A., Poblet, M., Mas, A., & Guillamon, J. M. (2000). ysis of the 16S-23S rDNA internal transcribed spacer Identification of acetic acid bacteria by RFLP of PCR- regions of the acetic acid bacteria. FEMS Microbiol. Lett., amplified 16S rDNA and 16S-23S rDNA intergenic 208,69e75. spacer. Int. J. Syst. Bacteriol., 50, 1981e1987. Urakami, T., Tamaoka, J., Suzuki, K., & Komogata, K. (1989). Saeki, A., Taniguchi, M., Matsushita, K., Toyam, H., Acidomonas gen. nov., incorporating Acetobacter meth- Theeragool, G., Lotong, N., et al. (1997a). Microbiological anolicus as Acidomonas methanolica comb. nov. Int. J. Syst. aspects of acetate oxidation by acetic acid bacteria, Bacteriol., 39,50e55. REFERENCES 255

Valero, E., Berlanga, T. M., Roldan, P. M., Jimeńez, C., Yamada, Y., Hoshino, K. I., & Ishikawa, T. (1997). The Garcıá, I., & Mauricio, J. C. (2005). Free amino acids and phylogeny of acetic acid bacteria based on the partial volatile compounds in vinegars obtained from different sequences of 16S ribosomal RNA: The elevation of the types of substrate. J. Sci. Food Agric., 85, 603e608. subgenus Gluconacetobacter to the generic level. Biosci. Vaughn, R. H. (1955). Bacterial spoilage of wines with Biotech. Biochem., 61, 1244e1251. special reference to California conditions. Adv. Food Res., Yamada, Y., Katsura, K., Kawasaki, H., Widyastuti, Y., 6,67e108. Saono, S., Seki, T., et al. (2000). Asaia bogorensis gen. Versalovic, J., Koeuth, T., & Lupski, R. J. (1991). Distribution nov., sp. nov., an unusual acetic acid bacterium in the of repetitive DNA sequences in eubacteria and applica- a-Proteobacteria. Int. J. Syst. Evol. Microbiol., 50, tion to fingerprinting of bacterial genomes. Nucleic Acids 823e829. Res., 19, 6823e6831. Yamada, Y., & Yukphan, P. (2008). Genera and species in Visser’t Hooft, F. (1925). Biochemical investigations on acetic acid bacteria. Int. J. Food Microbiol., 125,15e24. Acetobacter. Delft, the Netherlands: Thesis, Delft Tech- Yukphan, P., Malimas, T., Muramatsu, Y., Takahashi, M., nical University. Kaneyasu, M., Tanasupawat, S., et al. (2008). Tantichar- Watanabe, M., & Ino, S. (1984). Studies on bacteria isolated oenia sakaeratensis gen. nov., sp. nov, a new osmotolerant from Japanese wines. Part 2. Growth of the Acetobacter acetic acid bacterium in the a- Proteobacteria. Biosci. Bio- sp. A-1 during the fermentation and the storage of grape technol. Biochem., 72, 672e676. must and red wine. Yamanashiken. Shokuhin. Kogyo. Shy- Yukphan, P., Malimas, T., Potacharoen, W., Tanasupawat, S., dojo. Kenkyu. Hokoku, 16,13e22. Tanticharoen, M., & Yamada, Y. (2005). Neoasaia chiang- Wilker, K. L., & Dharmadhikari, M. R. (1997). Treatment of maiensis gen. nov., sp. nov., a novel osmotolerant acetic barrel wood infected with acetic acid bacteria. Am. acid bacterium in the a- Proteobacteria. J. Gen. Appl. J. Enol. Vitic., 48, 516e520. Microbiol., 51, 301e311. CHAPTER 10

Filamentous Fungi (Botrytis cinerea) Jesu´s M. Cantoral 1, Isidro G. Collado 2 1 Laboratorio de Microbiologı´a Enolo´gica , Facultad de Ciencias, Universidad de Ca´diz, Spain and 2 Departamento de Quı´mica Orga´nica, Facultad de Ciencias, Universidad de Ca´diz, Spain

OUTLINE

1. Introduction 257 3.3.1. Morphological Characteristics of B. cinerea Strains and 2. Main Diseases Affecting Relationship with Pathogenicity 266 Grapevines 258 3.3.2. Molecular Analysis of B. cinerea: 2.1. Powdery Mildew 258 Karyotype Polymorphism and 2.2. Grapevine Downy Mildew 259 Cloning and Characterization 2.3. Black Rot 260 of the gdhA Gene 267 2.4. Phomopsis Cane and Leaf Spot 261 3.3.3. Proteomics Study of B. cinerea 269 2.5. Eutypa Dieback 261 3.3.4. Isolation and Characterization 3. B. cinerea as a Model for Studying of Toxins Secreted by B. cinerea 270 Grapevine Fungal Diseases 262 3.3.5. Evidence on the Role of B. cinerea 3.1. Grapevine Infection by B. cinerea 262 Toxins in the Infection 3.2. Chemical Penetration 263 Mechanism 273 3.3. Strategies for the Analysis of B. cinerea 264 Acknowledgments 275

1. INTRODUCTION vines produce 674 million quintals of grapes annually. Almost half of the total production The grapevine is the most extensively (45%) comes from Europe. Of this total volume, planted fruit crop in the world. The total surface 184 million quintals are sold as fresh grapes, area dedicated to vines is estimated at 8 million 12.4 million are used to make raisins, and the hectares, of which 62% are in Europe. These rest is used to make grape juice (and wine). Of

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10010-4 257 Copyright Ó 2011 Elsevier Inc. All rights reserved. 258 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA) the 282 million hectoliters of wine produced in grapevines all over the world, even in tropical the world every year, 170 million are produced regions. The disease is believed to have origi- in Europe. The total annual consumption of nated in North America, from where it was wine is 237 million hectoliters, with the rest of probably spread to Europe by trade. It is caused production being allocated to vinegars and by the filamentous fungus Uncinula necator distilled beverages (data for 2005 from the Inter- (Schw.) Burr., discovered by Schweinitz in national Organisation of Vine and Wine [OIV], North America in 1834. The anamorph state is http://www.oiv.int). called Oidium tuckeri (Berk.). It is an obligate Grapes are grown in both tropical and tempe- parasite of various genera belonging to the rature climates, although the majority of grape- family Vitaceae (Bulit & Lafon, 1978). The fungus vines are located in temperate regions, mostly produces black spherical bodies called cleisto- in Europe. The main producers are Spain, thecia in its sexual structures; these bodies are France, Italy, Russia, Turkey, and Portugal, follo- formed by the fusion of male and female hyphae wed by the United States and countries from the (Pearson & Gadoury, 1978) and contain between southern hemisphere, namely Australia, South four and six asci with four to seven ascospores Africa, Chile, and Argentina. Grapevine each. The hyphae of the fungus develop multi- diseases can have serious consequences such lobed appressoria on which pegs used to pene- as reduced crop yields, diminished quality, trate the host develop. Powdery mildew can and increased production and harvesting costs. infect all the green tissues of the grapevine. The diseases are the result of the interaction The fungus only penetrates epidermal cells, between the vine (the sensitive host) and a live into which it introduces globular structures pathogenic organism (called a biotic or infec- known as haustoria that are used to absorb tious pathogen, or simply a pathogen). Most of nutrients. Diseased plants develop a whitish these organisms are bacteria, fungi, nematodes, grey powder caused by the presence of abun- and viruses. dant conidiophores (Sall, 1980). In this chapter, we will discuss some of the Both sides of leaves of all ages are susceptible diseases produced by filamentous fungi, to infection, which manifests as a fine white namely powdery mildew, downy mildew, gray powder. When young leaves become infected, mold, phomopsis cane and leaf spot, and eutypa they shrink and become distorted, and the infec- dieback, as well as the most serious and most tion of clusters close to inflorescences can cause common symptoms and control measures poor fruit set and considerable reductions in (Ministry of Agriculture, Fisheries and Food, crop yields. Infected shoots develop a dark Government of Spain, 2004; Pearson & Goheen, green patch that may turn brown and then 2007). In Section 3, we will take a closer look at black. Berries are prone to infection until they the case of Botrytis cinerea based on the consider- reach a sugar content close to 8%. Once infected, able experience our research group has acquired however, they continue to be affected by the in this area in recent years. production of fungal spores until a sugar content of 15% is reached. If berries become 2. MAIN DISEASES AFFECTING infected before they have reached their full GRAPEVINES size, the epidermal cells die but the flesh continues to grow, causing the berry to split 2.1. Powdery Mildew open and either dehydrate or rot. Infection by B. cinerea is common in such cases. When Powdery mildew, which is known by almost-ripe grapes (particularly red ones) different names in different regions, affects become infected, they cannot be sold as fresh MAIN DISEASES AFFECTING GRAPEVINES 259 grapes or used to make wine as they have plenty of sunlight. Vine training systems a negative impact on flavor. designed to allow good air circulation through Powdery mildew stunts growth and reduces the canopy and prevent excessive shade are crop yields. In particular, it affects the quality also a highly effective infection control system. of grapes and the plant’s resistance to cold. Good air circulation can also be achieved with U. necator fungi can overwinter as hyphae in suitable pruning. dormant buds or as cleistothecia on the surface of canes (the main shoots). Mycelia and conidia 2.2. Grapevine Downy Mildew survive from one season to the next in the plant tissues. The fungus develops within the bud, Grapevine downy mildew is caused by the where it remains latent until the following fungus Plasmopara viticola (Berk & Curt.) Berl. & season. The conidia, in turn, reproduce in large de Toni, which is an obligate parasite of several numbers on infected shoots and spread rapidly genera of the family Vitaceae (Lafon & Bulit, to nearby parts. 1981). It grows intracellularly in infected grape- The infection cycle begins in spring, when vine tissue, where it forms tubular hyphae with temperatures exceed 15C. When moistened globular haustoria. The fungus generally over- by rain, the cleistothecia open and release asco- winters as oospores in fallen, dead leaves, spores, which germinate and infect green although it can also survive as a mycelium in tissues, forming colonies that produce conidia buds and persistent leaves. Oospores, which for secondary spread. Temperature is the most survive better in the upper layers of damp soil, important environmental factor in terms of germinate in the spring and produce a sporan- fungal growth, with optimal development gium from which primary zoospores are occurring between 20 and 27C. Rain can dispersed by rain. This process requires a rela- disrupt the spread of infection as it washes tive humidity of 95 to 100% and at least 4 h of away the conidia and breaks up the mycelium. darkness; the optimal temperature is between A relative humidity of between 40% and 100% 18 and 22C. The sporangia are wind-dispersed is sufficient for the conidia to germinate. Low to leaves, where they germinate. The resulting indirect sunlight favors the development of zoospores swim through water and encyst powdery mildew, and bright sunlight can near the stomata. Because host penetration inhibit the germination of conidia. occurs exclusively through the stomata, only Disease control is normally achieved with plant structures with these pores are susceptible copper and organic fungicides such as benomyl, to infection. The optimal temperature for the dinocap, and sterol biosynthesis inhibitors. growth of P. viticola is 25 C but rain is the main Sulfur was the first fungicide used to success- factor responsible for epidemics. The worst cases fully treat powdery mildew and it is still the of downy mildew are seen in years with a wet most widely used treatment because of its effi- winter followed by a wet spring and a hot sum- cacy and cost. Copper and organic fungicides mer with intermittent rain storms (Langcake & are also available commercially but their use is Lovell, 1980). not as widespread as that of sulfur, which has Grapevine downy mildew is more common the disadvantage of being less active in wet in regions in which the plant growth phase coin- environments. Cultivation strategies can reduce cides with warm, wet weather. The growth of disease severity and increase the effectiveness of this fungus is limited in areas with little rainfall chemical treatments. It is beneficial, for in the spring and summer and in vineyards example, to plant vines in such a way that located in more northerly regions, where suffi- they will be exposed to good air currents and ciently high temperatures are not reached in 260 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA) spring. The fungus attacks all the green parts of other countries through contaminated material. the vine, particularly the leaves. It appears as This fungus can cause crop losses of between 5 a characteristic oily patch on the upperside of and 80%, depending on the severity of the the leaves (which become a dull green or epidemic, which, in turn, depends on the infec- yellowish color) and a downy white growth on tive capacity of the inoculum, the weather the lower sides. The infection of leaves is conditions, and the susceptibility of the host a very important source of inoculum for the plants. infection of berries. Infected leaves generally The main symptom on leaves is the appear- fall off, leading to a reduction in the accumula- ance of small, dark, circular patches in spring tion of sugar in the fruit and an increase in the and early summer, about 1 to 2 weeks after vulnerability of buds in winter. Young berries infection. The lesions, which are initially are very susceptible to infection. When infected, cream-colored, become progressively darker they turn a grayish color and become covered in and eventually acquire a reddish-brown color gray powder caused by sporulation. Although on the upper side of the leaf. Pycnidia appear berries become less prone to infection as they as small black spots (59e196 mm in diameter) ripen, the infection can spread inwards and in the center of the patches, and long, black attack the older berries. In autumn, mosaic-like canker lesions measuring between several milli- symptoms can appear on old leaves. meters and 2 cm in length develop on the edges Preventive cultivation strategies consist of of young leaves. Infected berries develop small ensuring proper ground drainage, reducing white spots and after a few days begin to dehy- sources of inoculum in winter, and pruning drate, shrivel, and wither. The infection can the tips of infected shoots. The timing of treat- affect whole clusters. ment is extremely important. The fungus needs The fungus overwinters in fallen mummies wet weather and a temperature of between 15 (shriveled, diseased grapes) or in old shoots still and 25C to grow. Optimal growth conditions on the vine. Ascospore release begins in spring, are thus rain, fog, or showers followed by hot, shortly after bud break (opening of buds); this sunny days. Fungicides are the best means of process is favored by frequent rainfall as the control (Lafon, 1985). Nonsystemic chemical ascospores need water in order to germinate. fungicides such as dithiocarbamates and copper The ascospores produce lesions on the leaf and salts, which are used in spring, are preventive infect inflorescences and young fruit. Fruit can only and only protect the treated surfaces. Cym- become infected at any time between midway oxanil and chlorothalonil are specific nonsys- through the flowering period until the grapes temic mildew fungicides and are curative if start to change color. Adult leaves and ripe fruit used in the first few days following infection. are not susceptible to infection. Pycnidia grow Aluminium and phenylamides are two of the in dry and newly rotted grapes and, once most widely used systemic fungicides to treat mature, release conidia when dampened by grapevine downy mildew. rain, leading to the risk of infection of leaves, inflorescences, and young fruit. Infection starts 2.3. Black Rot to decline at the end of July and disappears at the end of August (Spotts, 1980). Black rot is caused by Guignardia bidwellii Chemical control measures involve the use of (Ellis) Viala & Ravaz (anamorphic state, Phyllos- preventive fungicides such as manel and fer- ticta ampelicida [Engleman] Van der Aa) (Sivane- bam. Treatment should be started when the san & Holliday, 1981). The disease originated in shoots are 10 to 16 cm long and maintained until North America and was probably spread to the grapes reach a sugar content of 5%. Curative MAIN DISEASES AFFECTING GRAPEVINES 261 fungicides such as triadimefon are also used 2.5. Eutypa Dieback and should be applied as soon as the disease is detected. Eutypa dieback is caused by the ascomycete fungus Eutypa lata (Persoon: Fries) Tulasne and 2.4. Phomopsis Cane and Leaf Spot C. Tulasne. T. It is one of the most worrying diseases for grape growers because of the devas- Phomopsis cane and leaf spot is caused by tating economic losses it can cause. First the fungus Phomopsis viticola (Sacc.) Sacc. (syno- detected in Australia in 1973, the disease affects nyms, Fusicoccum viticola Reddick and Crypto- many vineyards around the world and is sporella viticola Shear [telemorphic form]). It considered to be the most serious threat to vine- can affect all of the green parts of the grapevine. yard longevity. In Spain, it was discovered for While symptoms are similar in the different the first time in 1979 and is becoming an increas- parts of the plant, the extent of damage varies, ingly serious problem affecting all of Spain’s with the main shoots being the most affected grape-growing regions. (Punithalingam, 1979). This disease has particu- The ascospores of the fungus infect and colo- larly worrying consequences in areas with nize the xylem through pruning wounds and frequent rainfall in spring, as the spread of then spread to the cambium. After an incubation disease is facilitated when the vines remain period of at least 3 years, infected pruning wet for several days after bud break. wounds become surrounded by canker and As the fungus grows, it causes the appearance the first symptoms appear in the green parts of of black spots or patches (which eventually the plant. The main symptoms are stunted crack) on green wood at the base of the buds. growth, withering of new branches, necrosis at A bulge forms at the base of newly sprouted the margins of leaves, dryness of inflorescences, shoots; this cracks longitudinally and signs of and the death of one or more branches. There wood strangulation become visible underneath. are different strains of the fungus, with varying This strangulation makes the canes more fragile. levels of virulence. In autumn, the bark develops whitish patches As mentioned above, the disease can have and black spots and, in winter, the vine becomes serious economic consequences. The most sensi- severely damaged as numerous canes start to fall tive grapevine varieties are Sauvignon Blanc, off. Leaves are also prone to attack and develop Cabernet Sauvignon, Ugni-blanc, Cinsault, and dark patches that mostly affect the petioles. The Chenin, and the most tolerant are Merlot and veins are rarely affected. The disease also attacks Semillon (Deswarte et al., 1994). While eutypa cluster stems, causing partial or total dehydra- dieback has a direct impact on crop yield, it tion of the grapes. can also affect the quality of wine made from One recommended disease control measure grapes from infected vineyards. It is one of the is to burn all pruning debris as this can provide most serious grapevine diseases known and is a home for overwintering fungi. The use of becoming an increasing concern for growers. dichlofluanid, folpet, mancozeb, maneb, or me- No means have yet been found for eliminating triam in winter destroys the pycnidia on the the fungus once it has infected the plant. Precau- canes prior to bud break. These fungicides can tionary control measures involve destroying also protect young shoots when applied after infected trunks or areas, pruning, and treating bud break. The careful application of the lesions with fungicides. product on spurs and canes that need treatment A toxin isolated in culture medium from is more efficient than the use of spray guns E. lata was found to be toxic for grape- (Bugaret, 1986). vines (Fallot et al., 1997). This compound, 262 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

4-hydroxy-3-(3-methyl-3-butene-1-ynyl) benzal- it can reduce crop yield and alters the organo- dehyde (eutypine), has been found in all leptic properties of wine. The frequency and infected vines (raw sap, leaves, inﬂorescences, intensity of attacks have made B. cinerea one of herbaceous stems, etc.) but has yet to be isolated the most feared diseases in the agricultural in a healthy plant. In vitro assays have shown community. The development of rational that eutypine rapidly leads to the development control programs is of major environmental of symptoms on cut grapevine leaves and importance (Rebordinos et al., 2003) and is causes structural alterations similar to those a priority for the agrochemical industry with described for infected grapevine leaves in far-reaching consequences for viticulture. vivo, demonstrating that it is involved in the manifestation of eutypa dieback symptoms. In 3.1. Grapevine Infection by B. cinerea grapevine cells, eutypine is metabolized into eutypinol (4-hydroxy-3-[3-methyl-3-butene-1- The quality of wines made with grapes ynyl] benzyl alcohol), which is not toxic for the infected by B. cinerea is diminished by the reduc- vine. This biotransformation is catalyzed by tion in monosaccharide content (glucose and a NADPH reductase (eutypine reductase) (Col- fructose) and the accumulation of metabolites rat et al., 1999). Tolerant grapevine varieties (glycerol and gluconic acid) and enzymes that have a greater capacity to metabolize eutypine catalyze the oxidation of phenolic compounds. than their more sensitive counterparts. The These wines do not age well either as they are discovery of this mechanism of action opens new perspectives for the development of efﬁ- cient tools to manage this worrying fungal disease. Preventive measures consist of burning dead vines and of pruning and burning diseased shoots and other plant parts. Pruning wounds can be treated with carbendazim, thio- phanate-methyl, or triadimefon paste.

3. B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES

Species from the phytopathogenic fungal genus Botrytis constitute a serious threat to a wide variety of crops. B. cinerea (perfect state, Botryotinia fuckeliana) is a particularly virulent variant and attacks many types of crops including grapevines, causing characteristic necrotic patches on leaves, stems, and fruits, and forming a grayish powdery mold known as gray mold (Snowdon, 1990) (see Figure 10.1). In viticulture, where the disease is commonly referred to as botrytis bunch rot, the fungus FIGURE 10.1 Gray mold caused by Botrytis cinerea on can have particularly serious consequences as a cluster of grapes. B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 263 susceptible to oxidation and bacterial contami- to attach itself to the host tissue via an anchor nation (Bulit & Dubos, 1988; Coley-Smith et al., system (appressorium) attached to the germina- 1980). Although B. cinerea can cause serious tion tubes (branched hyphae). As the fungus is damage in the winemaking industry, it is also not capable of growing by hyphal branching responsible for excellent wines such as alone, it secretes substances that destroy or Sauternes (France), Tokaji (Hungary), and prepare the plant tissue for hyphal penetration. Trockenbeerenauslesen (Germany, Austria) (Coley-Smith et al., 1980). This benevolent 3.2. Chemical Penetration form of B. cinerea, known as noble rot, propor- tionally consumes more acid than sugar, giving Once the fungus is near the cell wall, it rise to smooth, sweet wines with a good body launches a biochemical attack on the plant tissue and a pleasant bouquet. and cells to aid the spread of infection. A large Botrytis is a parasite that ﬁrst establishes itself number of complex interactions take place in the weaker, damaged parts of the host before during the cell invasion process and the deve- spreading to the rest of the plant. It can also, lopment of symptoms (see Figure 10.2). The however, attack plants that are already infested fungus uses two chemical weapons at this point: by other pests or pathogens. B. cinerea infects its high-molecular-weight enzymes that break hosts by physical or chemical penetration (Isaac, down the cell wall and membrane and lead to 1992). In the case of physical penetration, the tissue maceration, and low-molecular-weight fungus exploits natural openings on the plant toxins that kill the plant cells as the hyphae such as the stomata or small wounds that can advance through the host tissue. Because the appear on the surface of leaves or fruits. These production of these chemical weapons appears openings are relatively unprotected and are to be essential for the pathogenicity of B. cinerea, therefore vulnerable to penetration. In both cases, the interruption of this activity may render the for direct penetration to occurs, the fungus needs pathogen harmless.

Enzymes and toxins involved in the infective FIGURE 10.2 Botrytis cinerea processes of B. cinerea infection process. CAT ¼ catalase; POX ¼ peroxidase; SOD ¼ superoxide dismutase; GOD ¼ glucose Humidity Host Plant oxidase; PPO ¼ polyphenoloxidase; Toxins + AOS ¼ active oxygen species. B. cinerea Cutinases OHC CHO Phytotoxic Toxins OH Wild type isolates Role in pathogenicity CAT, H Botrydial (1 ppm) AcO POX, Botrydial SOD light-dependent s e Laccasas, T m detection in vitro and o HO OR y PPO in planta x z

i n n HO O correlation with s E infectivity of strains O HO GOD PGasas phytotoxic mode of Botcinic acid action via AOS (endo) (exo) induce oxidative AOS processes H2O2 AOS

Oxalic acid Oxidative + Fe Host Cell Death burst !

Expanding lesion > Necrosis, HR Disease >Resistance 264 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

Botrytis infections are characterized by the Figure 10.2 shows a simplified, summarized appearance of necrotic lesions and rot accompa- version of the different molecular processes and nied by the secretion of different types of impor- chemical reactions that take place during B. cine- tant substances, mainly toxins (Coley-Smith rea infection. Some of these processes have not yet et al., 1980; Collado et al., 2000, 2007; Elad et al., been studied in detail and much remains to be 2004) and enzymes that break down the cell learned before we can fully understand the infec- wall (polygalacturonases, pectin lyases, cellu- tion mechanism employed by this necrotrophic lases) (Prins et al., 2000) and the cell membrane fungus. Several aspects, however, have recently (phospholipases, lipases) (Prins et al., 2000; been clarified, such as the role played by toxins Shepard & Pitt, 1976). The chemical action of in the infection mechanism, which we will the toxins and enzymes secreted during the discuss in further detail in Section 3.3.5. infection process leads to the production of reactive oxygen species (ROS) (Deighton et al., 1999; Govrin & Levine, 2000; von Tiedemann, 1997). 3.3. Strategies for the Analysis ROS have been detected in all the infectious of B. cinerea processes studied to date and have become the The serious damage caused by this plant focus of much research in recent years (Deighton pathogen calls for physiological, biochemical, et al., 1999; Govrin & Levine, 2000; Prins et al., and molecular studies that will shed light on 2000; von Tiedemann, 1997). It has also been the mechanisms underlying . infection. demonstrated that ethylene is frequently pro- B cinerea Most of the studies performed to date have duced in plant tissues in response to attack by focused on identifying the enzymes involved . (Elad, 1995). B cinerea in the infectious process or the metabolites Enzymes play an important role in the infec- produced during infection. In recent years, tion process. While pectic enzymes are instru- however, considerable efforts have been made mental in the degradation of tissues as they in the area of molecular biology (Rebordinos facilitate access to other enzymes, cell-wall- et al., 1997), though many questions remain degrading enzymes alter osmotic pressure, about the molecular mechanisms underlying causing cell death (Elad, 1995). the pathogenicity of the fungus and the molec- Recent studies have uncovered completely ular basis of resistance to fungicides. Greater new aspects of the mechanisms used by necrotro- knowledge of these aspects will help in the phic fungi such as . to attack and invade B cinerea design of effective control strategies. Much their hosts. There is now evidence, for example, also remains to be learned about the genetics that the fungus produces ROS during the infec- of the fungus, which is difficult to study because tion process through the chemical action of toxins of the rarity with which the sexual stage of and enzymes. ROS are produced by both fungi is seen in nature (see Figure 10.3). In and hosts when the cell walls are broken down, Botrytis order to fill these important gaps in our knowl- triggering a series of free-radical reactions (Dei- edge, our research groups have characterized ghton et al., 1999; Govrin & Levine, 2000; Prins several strains of . isolated in different et al., 2000; von Tiedemann, 1997). The infection B cinerea grape-growing areas of Spain, with particular mechanism by which gray mold is produced is focus on the following aspects: thus complex and involves both external factors, which are necessary for the process to begin, and 1. Morphology. By modifying in vitro culture a series of chemical reactions that damage the conditions and observing the fungus in host cell wall and membrane and aid the spread each of the stages of its life cycle, we were of hyphae through the tissue. able to perform crosses between sexually B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 265

Conidia

Macroconidiophores

ASEXUAL CYCLE

Initial development of micelial ramifications Micelial development

Micelium

Spermatization

Microconidiophores Germination Sclerotia

SEXUAL CYCLE

Fecundation Ascospore

Apotecia

FIGURE 10.3 Life cycle of Botrytis cinerea (perfect state, Botryotinia fuckeliana). 266 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

compatible strains and study offspring and Type Culture Collection (CECT) and strains level of pathogenicity. donated by Dr. F. Faretra (University of Bari, 2. Determination of karyotype by pulsed field gel Italy) for crosses. The first step was to morpho- electrophoresis (PFGE). This revealed logically characterize the strains (according to polymorphism among the strains analyzed, production of resistance structures and growth although several strains had an identical rate) and to analyze levels of pathogenicity. electrophoretic profile. We also developed Morphological differences were observed a novel autonomous transformation system in between the strains, with variations seen B. cinerea, and cloned, sequenced, and according to the time in culture mostly attribut- characterized the gdhA gene, which is involved able to heterokaryosis. Strains with the least in the metabolism and regulation of nitrogen. infective capacity did not produce sclerotia. 3. Proteomic analysis. In this first-ever proteomic Infective capacity was low in strains that mostly study of B. cinerea, we optimized the entire retained a conidial morphology and higher in process, from the extraction of proteins to those with a very fast growth rate. The size of their identification. We also performed the the conidia, determined by scanning electron first differential proteomic study of B. microscopy (see Figure 10.4), was another cinerea, comparing proteomes from two highly variable character, as was the number strains with different virulence. On the basis of nuclei in the conidia (determined by fluores- of our findings, we selected proteins that cence microscopy). No relationship was found were expressed in both of the strains between these variables and pathogenicity analyzed and proteins that were (Vallejo, 1997). overexpressed in the most virulent strain for The optimization of the method used to further analysis. obtain B. cinerea apothecia under controlled 4. Isolation and characterization of compounds from laboratory conditions has resulted in greater culture broths. The compounds were found to be toxins or toxin derivatives, some of which displayed considerable biological, phytotoxic, and antibiotic activity. 5. Role of toxins in infection mechanism of B. cinerea. We found that both families of toxins we had isolated were made redundant during infection by the most virulent strains. 6. Pathways involved in the biosynthesis of toxins. We characterized several of the genes involved in the secondary metabolism of the fungus and several of the main enzymes involved in the biosynthetic pathways of the above toxins.

3.3.1. Morphological Characteristics of B. cinerea Strains and Relationship with Pathogenicity Our group isolated and studied several strains of . from different agricultural B cinerea FIGURE 10.4 Images of Botrytis cinerea conidia obtained regions as well as strains from the Spanish by scanning electron microscopy. B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 267 knowledge of the fungus thanks to genetic fragment length polymorphism (RFLP) analysis studies of its progeny. The first studies designed of numerous strains of B. cinerea isolated in to determine sexual compatibility led to the infected grapes has revealed considerable conclusion that this is controlled by a single genetic variability and the presence of the transgene, MAT 1, with two alleles, MAT-1 and posable elements Boty and Flipper in different MAT-2. Not many studies of this type have populations of this fungus. PFGE, in turn, has been performed, however, because of the diffi- revealed considerable chromosome polymor- culty of and time required to obtain apothecia, phism between different strains of the fungus which largely explains why knowledge of the (Vallejo et al., 1996) as well as the presence of genetic aspects of this fungus is still very minichromosomes, which may correspond to limited. what are referred to as supernumerary or B We performed numerous crosses between chromosomes. These chromosomes are sexually compatible reference and grapevine primarily characterized by their variable length, strains and studied genotype segregation of absence in certain strains, and abnormal segre- sequences encoding ribosomal RNA (rRNA), gation in sexual crosses. Their exact function, as rRNA contains valuable information on basic however, remains to be clarified. species structure and gene function. rRNA can Karyotypic analysis by PFGE has also also be used to perform taxonomic and sistema- revealed extensive polymorphism between tic analyses to establish intra- and interspecific different strains of B. cinerea (see Figure 10.5), relationships. The segregation of the above with between four and eight resolvable bands sequences in single-spore progeny of two measuring between 1.88 and 3.86 Mb detected. crosses between B. cinerea strains was analyzed The minimum genome size of these strains using an rDNA hybridization probe labeled was calculated to range between 11.22 and with digoxigenin on nylon filters containing 22.92 Mb (Rebordinos et al., 2000). In that study, chromosome bands corresponding to electro- however, we did not find an association phoretic profiles for single-spore segregations between any of the bands and level of pathoge- obtained by PFGE (Vallejo et al., 2002). nicity, although the high phenotypic variability In a subsequent study, in vitro bioassays on could be explained by chromosome polymor- Vitis vinifera (grapevine) and Phaseolus vulgaris phism and the heterokaryotic nature of the (common bean) leaves were used to determine fungal cells. The study of nonpathogenic the infective capacity of the progeny (Vallejo mutants created using agents such as ultraviolet et al., 2003). The virulence of both parent strains radiation will possibly contribute to a greater and progeny were analyzed by inoculating the understanding of the infection mechanisms of leaves with a suspension of conidia (of a known B. cinerea. concentration) and observing the development One of the most useful and powerful tools of symptoms over the course of days. We found to emerges in molecular biology in recent that the pathogenicity trait had been passed years is cloning. Thanks to this technique, from the parent strains to their progeny. numerous genes involved in B. cinerea pathogenicity and plant-pathogen interactions 3.3.2. Molecular Analysis of B. cinerea: have been identified. Although numerous Karyotype Polymorphism and Cloning and enzymes associated with the infection process Characterization of the gdhA Gene in this fungus have also been identified, no Molecular biological tools are important for conclusive evidence has yet been obtained gaining a greater understanding of B. cinerea that any are directly responsible for the and its mechanisms of action. Restricted damage caused by B. cinerea. 268 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

FIGURE 10.5 Polymorphism in the karyotype of 68 strains of Botrytis cinerea (pulsed ﬁeld gel electrophoresis).

The shortage of nutrients such as nitrogen creation of mutants, and above all the effect of appears to be linked to pathogenicity and different substrates on the gdhA gene promoter several morphological aspects of B. cinerea. The may help to establish a direct relationship loss of regulatory factors involved in the assimi- between nitrogen metabolism and pathoge- lation of ammonia, for example, signiﬁcantly nicity in B. cinerea. reduces the virulence of certain plant and The gdhA gene was cloned in B. cinerea by animal fungi. Given the absence of studies on ﬁrst screening a genomic library by hybridiza- nitrogen and carbon metabolism in B. cinerea tion with a heterologous probe from Aspergillus and the enormous importance of understanding awamori containing part of the coding sequence the interactions between these two pathways of the gene (Santos et al., 2001). This yielded and their possible association with pathogene- a 3.48 kb DNA fragment from which 2351 sis, our group decided to clone the gdhA gene, nucleotides were sequenced and found to which encodes the NADPH-dependent gluta- contain an open reading frame (ORF) of 1350 mate dehydrogenase (GDH) enzyme involved base pairs coding for a 450-amino-acid protein in the synthesis of the essential amino acid (this sequence is available from the EMBL glutamate and, as such, responsible for the Nucleotide Sequence Database under accession direct assimilation of ammonia. Analysis of the no. 093934). The size of the monocistronic tran- regulatory mechanisms for this gene, the script was estimated at 1.7 kb. The gene was B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 269

AB AB as the post-genomics era. The proteomic analysis of filamentous fungi is, however, still in its I infancy, and most of the studies conducted in I II this area have been of species of industrial VII interest. The characterization of the first pro- VI, IX teome of B. cinerea by Fernańdez-Acero et al. VII IX X 2.05 Mb (2006) represented a breakthrough in terms of X 2.05 Mb optimizing the process from start to finish as their study reported all the stages involved in the analysis, from protein extraction to identification. Other techniques can be used to identify factors involved in the infection processes of different organisms, but most involve the selection of a candidate gene in a previous screening step and the subsequent analysis of its influence on infection processes via expression analysis or UCA 993 2850 directed mutagenesis. Proteomics studies, in FIGURE 10.6 Karyotype (A) of two strains (UCA 993, contrast, do not require an a priori selection of 2850) and localization of gdhA gene (B) in chromosome X genes but rather focus on analyzing the expres- (numbering of bands according to Santos et al., 2001). sion of whole sets of proteins under given conditions. Fernańdez-Acero et al. (2006), using two- located on chromosome X (see Figure 10.6), dimensional gel electrophoresis and Coomassie whose karyotype was analyzed by PFGE; the blue staining, detected between 380 and 400 gene complemented two Aspergillus nidulans protein spots with a molecular weight of mutants and restored NADP-dependent GDH between 15 and 85 kDa and an isoelectric point activity (Santos et al., 2001). Expression anal- of between 5.4 and 7.7. They selected 22 of these yses in the same study indicated that the gene spots for identification by matrix-assisted laser was subject to strong regulation by carbon desorption ionization time of flight or electro- and nitrogen. spray ionization ion trap mass spectrometry. Our group was the first to implement an Among the proteins found were different forms innovative (nonintegrative) transformation of malate dehydrogenase (MDH) and glyceral- system in B. cinerea using the plasmid pUT737; dehydes-3-phosphate dehydrogenase (GADPH), the system had a transformation efficiency of cyclophilin, and other proteins of unknown 25e40 transformants/mg of DNA and these function. transformants maintained their capacity for In a later study, the proteomes from two four generations without selective pressure strains of B. cinerea with different virulence (Santos et al., 1996). were compared and 28 proteins that were either expressed in both strains or overexpressed in 3.3.3. Proteomics Study of B. cinerea the more virulent strain were selected for analy- Important advances have been made in sis (Fernańdez-Acero et al., 2007). The most rele- recent years in the area of proteomics and its vant proteins in terms of their possible application to the study of biological, metabolic, involvement in the infection process were pathological, and other processes. Indeed, the (1) NADPH-dependent MDH, (2) GAPDH, current age of molecular biology is referred to (3) metE/metH, and (4) cyclophilin. The first, 270 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

MDH, was found in three clusters and overex- The first proteomic map of B. cinerea,pu- pressed in the most virulent strain. This blished recently (Fernańdez-Acero et al., substance catalyzes the transformation of 2009), contains over 300 identified proteins malate to oxaloacetate, the main precursor of that have been functionally classified into oxalic acid, which has been described as a path- molecular and biological groups using the ogenicity factor. High levels of oxalic acid PANTHER (protein analysis through evolu- reduce the pH of cultures, which is a prerequi- tionary relationships) database (http://www. site for the production of the toxins botrydial pantherdb.org). This information is freely avail- and dihydrobotrydial by B. cinerea. These data able at the website of the Swiss Institute of Bio- support the hypothesis that MDH is involved informatics through the ExPASy Proteomics in the pathogenicity of this fungus. It was Server (http://world-2dpage.expasy.org/ recently demonstrated that these NADPH- repository/, accession number 0005). Because dependent oxidases play a role in the differentia- cellulose is one of the main components of the tion and pathogenicity of B. cinerea (Segmu¨ ller cell wall, many of the proteins identified play et al., 2008), a finding that seems to be in agree- a crucial role in the pathogenesis of the fungus. ment with the proteomic data available. The Given the enormous difficulty and complexity second protein, GAPDH, was found only in associated with proteomic analysis, more recent the most virulent strain. The role of this enzyme studies have taken a more simplified approach in the pathogenicity of various organisms has by dividing samples into more manageable been widely reported (Alderete et al., 2001). packets called subproteomes (Fernańdez-Acero GAPDH might, thus, in addition to its metabolic et al., 2007). The first of these subproteomes to function, have a role in the infection cycle of B. be studied in B. cinerea, the secretome (Fernań- cinerea, as was recently reported for the hexoki- dez-Acero et al., 2009; Shah et al., 2009), provided nase Hxk1 (Rui & Hahn, 2007). The third promising results that reflect the potential of this protein, the transcriptional regulator metE/ tool for further investigating pathogenicity metH, was specific to and overexpressed by factors and therapeutic targets and conducting the most virulent strain. This protein is involved basic research into B. cinerea. in the synthesis of methionine, a pathway that has been widely used in the design of fungi- 3.3.4. Isolation and Characterization cides. The variability observed between the of Toxins Secreted by B. cinerea different strains of B. cinerea might be the molecu- In addition to producing organic acids and lar basis of the different fungicide resistance high- and low-molecular-weight polysaccha- phenotypes described for this fungus. The rides (from glucose monomers and from fourth protein, cyclophilin, which is associated mannose, galactose, glucose, and ramnose with protein assembly and regulation, was monomers, respectively) (Coley-Smith et al., found only in the most virulent strain. Cyclophi- 1980; Elad et al., 2004), Botrytis fungi also secrete lin has been described as a virulence factor in a series of secondary sesquiterpene metabolites different fungi, including B. cinerea, in relation during the development of necrotic lesions on to tissue invasion and colonization processes, the host plant (Colmenares, 2001; Colmenares confirming the usefulness of proteomics to iden- et al., 2002; Deighton et al., 2001; Rebordinos tify components involved in infection mecha- et al., 1996). Recent studies have shown that nisms. The protein might also be involved in these metabolites are toxins that constitute cell signaling as it forms a complex with a virulence factor in plant pathogens of this calmodulin, which is involved in various cell type (Colmenares, 2001; Colmenares et al., 2002; signaling cascades. Deighton et al., 2001; Rebordinos et al., 1996). B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 271

CHO CHO HO O CHO CHO OH OH 1 1 2 9 89 4 6 H H H AcO AcO AcO 1 Botrydial 2 Dihydrobotrydial 3

HO OH HO OH CHO CHO

456

H OR HO OR O O

HO O OH O

HO R= OAc O 3 7 Botcinic acid O 8 Botcinin A

FIGURE 10.7 Structure of secondary metabolites produced by Botrytis cinerea.

B. cinerea synthesizes a series of metabolites Two structurally distinct families of toxins with a botryane skeletondmainly botrydial synthesized by Botrytis fungi have been iso- and dihydrobotrydial (see Figure 10.7) and lated. The most abundant of these has derivatives of thesedthat are responsible for a botryane skeleton and is a chemical deriva- the symptoms associated with this fungus (see tive of the plant toxin botrydial. Some of the Figure 10.8)(Collado et al., 2007; Colmenares, most relevant toxins isolated are shown in 2001; Colmenares et al., 2002; Durań-Patroń Figure 10.7. A new family of toxins recently et al., 2000). The in vivo detection of botrydial isolated from virulent strains of B. cinerea and in Capsicum annuum (sweet pepper) plants initially called botcinolides have a polyketide infected with B. cinerea showed that these metab- backbone. The structures of these toxins were olites, or toxins, are associated with pathogenesis recently revised (Tani et al., 2005, 2006), and infection (Deighton et al., 2001). The toxins, however, and they have since been renamed which are not host-specific, help the fungus to botcinins, the most abundant of which is bot- penetrate the host and colonize the plant tissue, cinic acid (see Figure 10.7). These toxins act thus increasing the severity of disease. synergically with botrydial and derivatives in 272 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

the infection mechanism of B. cinerea (Siewers et al., 2005), and it was recently demonstrated that both families have a redundant role in the infection mechanism of aggressive strains (Pinedo et al., 2008). Diverse biosynthetic studies have revealed detailed information about the biosynthetic pathway to generate botrydial and its derivatives from farnesyl pyrophosphate (see Figure 10.9) (Bradshaw et al., 1977, 1982; Dura´n-Patro´netal., 2001, 2003). Furthermore, kinetic growth studies of B. cinerea that have analyzed botrydial and three of its derivatives have elucidated the biosynthetic pathways of botrydial derivatives secreted by B. cinerea (Daoubi et al., 2006). There are two main pathways involved in degrading the most active toxin secreted by the fungus, botrydial, which is used as a chemical weapon during the infection mechanism. Figure 10.10 shows a general diagram of the degradation

FIGURE 10.8 Effect of the toxin botrydial on Phaseolus vulgaris leaf.

+ H OPP + +

H + Bcbot 2

H OH

OH H OH H OH

OH OAc OH Bcbot 1 OH OH OH O CHO CHO H OH H OH H OH H OH

OAc OAc OAc OAc 2 1

FIGURE 10.9 Biosynthetic pathway of botrydial and derivatives from farnesyl pyrophosphate. B. CINEREA AS A MODEL FOR STUDYING GRAPEVINE FUNGAL DISEASES 273

OH OH HO OH HO OH CHO CHO

10 15 10 15 OH CHO CHO CHO H 1S OH 11 OH neutral pH 14 14 AcO 11 8S 1 2 9S 2 9 8 5 6 5 4 4 6 HO O 15 acid 10 Aromatic compounds H H OH 12 AcO 13 AcO 12 13 pH 1 9 8 +lactol ethers Botrydial (1) 4 H C-1, C-8, C-9 epimer AcO derivatives Dihydrobotrydial (2)

FIGURE 10.10 Detoxification of botrydial and relationship between toxins with a botryane skeleton produced by Botrytis cinerea. process and the relationship between the main The above studies confirmed that BcBOT1 is metabolites secreted by B. cinerea. involved in one of the final steps in the biosyn- Our knowledge of the B. cinerea genome, thesis of botrydial (see Figure 10.9), while combined with the increasing proteomic data BcBOT2 is involved in the first step of the cycli- available, will help to elucidate the role played zation of farnesyl pyrophosphate to the by genes in the different biological processes botryane skeleton (see Figure 10.9). Presilphi- in the fungus and to identify new molecular perfolan-8-ol synthase, an enzyme with a key targets to be exploited in the design of future role in the biosynthesis of botrydial and a large fungicides. Recent studies have characterized number of polycyclic sesquiterpenes of fungal a cluster of genes involved in botrydial biosyn- and plant origin with diverse biological activi- thesis, namely those coding for a sesquiterpene ties (see Figure 10.12), has also been character- cyclase (BcBOT2), an acyltransferase (BcBOT5), ized (Pinedo et al., 2008). and three monooxygenases (BcBOT1, 3,and4) Finally, the discovery of the polyketide syn- (Pinedo et al., 2008; Siewers et al., 2005)(see thases involved in the biosynthesis of botcinins Figure 10.11). Recent studies of the B. cinerea (Collado, Viaud, Tudzinsky groups, unpu- strains B05-10 ku70 and SAS56 (which produce blished results) will also help in the design of both families of toxins) and the T4 strain fungicides based on hybrid molecules that (which produces just one family, namely botry- behave as double inhibitors of polycyclic sesqui- dial and derivatives), resulted in the identifica- terpenes and polyketide synthases. tion of the first two genes involved in the B cinerea biosynthesis of botrydial: BcBOT1 and 3.3.5. Evidence on the Role of . BcBOT2. These genes encoded a P-450 mono- Toxins in the Infection Mechanism oxygenase (Siewers et al., 2005) and a sesquiter- As indicated previously, B. cinerea, which is pene cyclase (STC) (Pinedo et al., 2008), responsible for the devastating gray mold respectively. disease, produces several biologically active

FIGURE 10.11 Cluster of genes involved in the biosynthesis of botrydial (1) and derivatives. 274 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

H FIGURE 10.12 The cation presilphiperfolan-8-ol is the precursor of a wide variety of triquinanes and sesquiterpenes H OH HO derived from fungi and higher plants. Modhephene Cameroonan-7-ol

Prenopsan-8-ol H lsocomene OH H H Nopsan-4-ol H Silphinene H H OH

H H H β O O Presilphiperfolan-8 -ol O Silphiperfolene α-Terrecyclene Quadrone metabolites, several of which (botrydial and Similar results have also been found on compounds 3 and 4 in Figure 10.7)have applying botrydial to tomatoes, peppers, been found to exhibit high levels of in vitro grapes, and strawberries, with chlorosis toxicity. Similar results have been obtained affecting the pericarp of all the fruit and in vivo during the testing of these metabolites evident depletion and collapse in the treated in B. cinerea-susceptible and -resistant P. vulga- areas (Colmenares, 2001; Colmenares et al., ris plants (genotypes N 90598 and N 90563, 2002). Tests designed to detect botrydial respectively). The in vivo effect, however, during in vivo infection have been conducted occurred much more rapidly, with symptoms to gain further insights into the role played by becoming evident after just a few hours (see this toxin in the infection mechanism. Figure 10.8). Botrydial was capable of repro- Deighton et al. (2001),intestsconductedin ducing the phytotoxic effect of B. cinerea at C. annuum and on P. vulgaris and Arabidopsis a concentration of one part per million. At thaliana leaves, for example, detected and iso- this concentration, the toxin affected 70% of lated botrydial from the inoculated areas the treated leaves, with symptoms appearing during the early stages of infection, proving on 4% of the surface area; leaves started yel- that botrydial is produced in the plant during lowing after 60 h on bright sunny days and the infection process. Although the authors after 120 h during periods of cloudy weather. concluded that the toxin was associated with Thesamephytotoxiceffectwasobservedin pathogenesis, they were unable to confirm Nicotina tabacum, Lactuca sativa, Fragaria vesca, whether or not it was a primary factor in the V. vinifera,andCitrus limon leaves, although infection process. it is interesting that the symptoms took longer Various experiments involving the inoculation to appear in the last three of these plants. of B. cinerea onto P. vulgaris leaves previously REFERENCES 275 treated with botrydial have clearly shown that in the T4 strain but not in the more virulent the toxin facilitates fungal penetration and tissue B05-10 strain. One possible explanation is that colonization. Taken together, these results indi- the B05-10 mutant compensates for the loss of cate that botrydial is a nonspecific toxin that botrydial by producing high quantities of bot- affects a wide range of plant species; it can also cinic acid and botcinin A, thus retaining be considered a virulence factor as it appears to similar levels of virulence to the parent strain. be involved in the development and spread of The T4 mutants, in contrast, whose parent the disease. Whether or not botrydial is a patho- strains did not produce botcinic acid or botci- genicity factor is the focus of ongoing studies nin A (see Figure 10.7) had lost their infective involving mutants (Collado, unpublished capacity as they did not produce either of the results). Furthermore, recent experiments have toxins. shown an interesting relationship between toxin The results described in this section are production and virulence. The fungus is resistant particularly relevant in terms of their impor- to its own toxin thanks to a detoxification mech- tance for the design of new fungicides targeting anism that reduces the compound to other biological pathways involved in the biosyn- inactive, nontoxic products. Daoubi et al. (2006) thesis of toxins that play an important role in recently discovered two detoxification mecha- the mechanism of infection by B. cinerea. Target- nisms regulated by pH (see Figure 10.10). ing these toxins could be the key to controlling B. cinerea toxins have a light-dependent mecha- B. cinerea and its pathogenicity. nism of action. Phytotoxicity studies consisting of exposing treated areas to light, darkness, and Acknowledgments alternating light and darkness clearly showed that the mechanism of action of the most active The research by our group described in this chapter was and abundant toxin, botrydial, was light-depen- funded by grants from the Spanish Ministry of Science of Technology (1FD97-0668-C06, AGL2000-0635-C01 and dent and generated ROS via a type 1 photody- -C02, AGL2002-04388-C01 and -C02, AGL2003-06480-C02 namic reaction involving oxygen activation and -C01, AGL2005-07001-C02, AGL2006-13401-C01 and (Colmenares, 2001). When exposed to light, bot- -C02, and AGL2009-13359-C01 and -C02) and the Autono- rydial may induce lipid peroxidation in plant mous Government of Andalusia (Proyecto de Excelencia cells followed by membrane changes that give grant PO7-FQM-02689). We thank the following people for their help: Lau- rise to chlorosis, death, and collapse in the reana Rebordinos, Inmaculada Vallejo, Francisco Javier affected zone. Fernańdez-Acero, MarıáCarbu´ , and Carlos Garrido It should be noted that, in response to (applied microbiology group) and Rosario Hernańdez- silencing of the gene in a . Galań, Antonio Jose´ Macıás-Sańchez, Josefina Aleu Casa- BcBOT2 B cinerea a mutant whose parent strain, 05-10, tejada, and M Rosa Durań-Patroń(organicchemistry Bc group). produced botcinin acid and botcinin A (see Figure 10.7), the fungus shifted its secondary metabolism to an overproduction of these References two compounds. No such shift was seen in Alderete, J. F., Millsap, K. W., Lehker, M. W., & the T4 mutant strain, which does not produce Benchimol, M. (2001). Enzymes on microbial pathogens botcinin A. and Trichomonas vaginalis: Molecular mimicry and func- Virulence tests on Vicia fabae and Lycopersi- tional diversity. Cellular Microbiology, 3, 359e370. leaves with mutants Bradshaw, A. P. W., Hanson, J. R., & Nyfeler, R. (1982). cum esculentum bcbot2 Studies in terpenoid biosynthesis. The fate of the have provided insights into the role played mevalonate hydrogen atoms in the biosynthesis of the by botrydial in . virulence. The results B cinerea sesquiterpenoid, dihidrobotrydial. J. Chem. Soc. Perkin showed that the toxin is necessary for infection Trans., I, 2187e2192. 276 10. FILAMENTOUS FUNGI (BOTRYTIS CINEREA)

Bradshaw, A. P. W., Hanson, J. R., & Siverns, M. (1977). Use Durań-Patroń, R., Colmenares, A. J., Hernańdez-Galań, R., & of induced 13C-C13 coupling in terpenoid biosynthesis. Collado, I. G. (2001). Some key metabolic intermediates in J.C.S. Chem. Comm., 819. the biosynthesis of botrydial and related compounds. Bugaret, Y. (1986). Donneés nouvelles sur l’epide´miologie Tetrahedron, 57, 1929e1933. de l’excoriose et leurs conse´quences pour la lutte. Durań-Patroń, R., Colmenares, A. J., Montes, A., Phytoma, 375,36e41. Hanson, J. R., Hernańdez-Galań, R., & Collado, I. G. Bulit, J., & Dubos, B. (1988). Compendium of grape diseases. (2003). Studies on the biosynthesis of secobotryane St. Paul, MN: APS Press. skeleton. Tetrahedron, 59(33), 6267e6271. Bulit, J., & Lafon, R. (1978). Powdery mildew of the vine. In Durań-Patroń, R., Hernańdez-Galań, R., & Collado, I. G. D. M. Spencer (Ed.), The powdery mildew (pp. 525e548). (2000). Secobotrytriendiol and related sesquiterpenoids: New York, NY: Academic Press. New phytotoxic metabolites from Botrytis cinerea. J. Nat. Coley-Smith, J. R., Verhoeff, K., & Jarvis, W. R. (1980). The Prod., 63, 182e183. biology of botrytis. London, UK: Academic Press. Elad, Y. (1995). Modern fungicides and antifungal compounds Collado, I. G., Aleu, J., Hernańdez-Galań, R., & Durań- (pp. 217e233). Andover, UK: Intercept Ltd. Patroń, R. (2000). Botrytis species: An intringuing source Elad, Y., Williamson, B., Tudzynski, P., & Delen, N. (2004). of metabolites with a wide range of biological activities. Botrytis: Biology, pathology and control. Dordrecht, the Structure, chemistry and bioactivity of metabolites iso- Netherlands: Kluwer Academic Publishers. lated from Botrytis species. Curr. Org. Chem., 4, 1261e1286. Fallot, J., Deswarte, C., Dalmayrac, S., Colrat, S., & Collado, I. G., Macias-Sańchez, A. J., & Hanson, J. R. (2007). Roustan, J. P. (1997). L’eutypiose de la vigne. Isolement Fungal terpene metabolites: Biosynthetic relationships d’une molećule synthe´tiseé par Eutypa lata et toxique and the control of the phytopathogenic fungus Botrytis pour la vigne. C.R.A. Sci. Paris, 320, 149e158. cinerea. Nat. Prod. Rep., 24, 674e686. Fernańdez-Acero, F. J., Carbu´ , M., Garrido, C., Vallejo, I., & Colmenares, A. J. (2001). Obtencion de nuevos metabolitos del Cantoral, J. M. (2007). Proteomic advances in phyto- Fitopato´geno Botrytis cinerea. Evidencias sobre el papel de pathogenic fungi. Current Proteomics, 4,79e88. las toxinas en la fitopatogenicidad del hongo. Cadiz, Spain: Fernańdez-Acero, F. J., Colby, T., Harzen, A., Cantoral, J. M., & Doctoral thesis. Universidad de Ca´diz. Schmidt, J. (2009). Proteomic analysis of the phytopatho- Colmenares, A. J., Aleu, J., Durań-Patroń, R., Collado, I. G., & genic fungus Botrytis cinerea during cellulose degradation. Hernańdez-Galań, R. (2002). The putative role of botry- Proteomics, 9, 2892e2902. dial and related metabolites in the infection mechanism of Fernańdez-Acero, F. J., Jorge, I., Calvo, E., Vallejo, I., Botrytis cinerea. J. Chem. Ecol., 28(5), 997e1005. Carbu´ , M., Camafeita, E., et al. (2006). Two-dimensional Colrat, S., Latche´, A., Pech, J. C., Bouzayen, M., Fallot, J., & electrophoresis protein profile of the phytopathogenic Roustan, J. P. (1999). Purification and characterization of fungus Botrytis cinerea. Proteomics, 6,88e96. an NADPH-aldehyde reductase from Vigna radiata Govrin, E. M., & Levine, A. (2000). The hypersensitive which detoxifies eutypine, a toxin produced by Eutypa response facilitates plant infection by the necrotrophic lata. Plant Physiol., 119, 621e626. pathogen Botrytis cinerea. Curr. Biol., 10, 751e757. Daoubi, M., Duran-Patron, R., Hernandez-Galań, R., Isaac, S. (1992). Fungaleplant interactions (pp. 147e175). Benharref, A., Hanson, J. R., & Collado, I. G. (2006). The New York, NY: Chapman & Hall. role of botrydienediol in the biodegradation of the ses- Klober, A. (1992). Obstau-Weinbau, 17,1e92. quiterpenoid phytotoxin botrydial by Botrytis cinerea. Lafon, R. (1985). Les fongicides viticoles. In I. M. Smith Tetrahedron, 62(35), 8256e8826. (Ed.), Fungicides for crop protection, Vol. 1 (pp. 191e Deighton, N., Muckenschnabel, I., Colmenares, A. J., 198). Croydon, UK: British Crop Protection Council, Collado, I. G., & Williamson, B. (2001). Botrydial is Monogr. 31. produced in plant tissues infected by Botrytis cinerea. Lafon, R., & Bulit, J. (1981). Downy mildew of the vine. In Phytochemistry, 57, 689e692. D. M. Spencer (Ed.), The downy mildews (pp. 601e614). Deighton, N., Muckenschnabel, I., Goodman, B. A., & New York, NY: Academic Press. Williamson, B. (1999). Lipid peroxidation and the Langcake, P., & Lovell, A. (1980). Light and electron oxidative burst associated with infection of Capsicum microscopical studies of the infection of Vitis spp. by annum by Botrytis cinerea. Plant J., 20, 485e492. Plasmopara viticola, the downy mildew pathogen. Vitis, Deswarte, C., Rouquier, P., Roustan, J. P., Dargent, R., & 19, 321e337. Fallot, J. (1994). Ultrastructural changes produced in Ministry of Agriculture. (2004). Fisheries and Food, plantlet leaves and protoplasts of Vitis vinifera cv. Government of Spain. In Los para´sitos de la vid. Estrategias Cabernet Sauvignon by eutypine, a toxin from Eutypa de proteccioń razonada (4th ed.). Madrid, Spain: Ediciones lata. Vitis, 33, 185e188. Mundi-Prensa. REFERENCES 277

Pearson, R. C., & Gadoury, D. M. (1978). Cleistothecia, the Segmu¨ ller, N., Kokkelink, L., Giesbert, S., Odinius, D., van source of primary inoculum for grape powdery mildew. Kan, J., & Tudzynski, P. (2008). NADPH oxidases are Phytopathology, 77, 1509e1514. involved in differentiation and pathogenicity in Botrytis Pearson, R. C., & Goheen, A. C. (1996). Plagas y enferme- cinerea. Mol. Plant Microbe In., 21, 808e819. dades de la vid. In The American Phytopathological Shah, P., Atwood, J. A., Orlando, R., El Mubarek, H., Society. (Ed.), Madrid, Spain: Mundi-Prensa. Podila, G. K., & Davis, M. R. (2009). Comparative pro- Pinedo, C., Wang, C.-M., Pradier, J.-M., Dalmais, B., teomic analysis of Botrytis cinerea secretome. J. Proteome Choquer, M., Le Pecheur, P., et al. (2008). Sesquiterpene Res., 8(3), 1123e1130. synthase from the botrydial biosynthetic gene cluster of Shepard, D. V., & Pitt, D. (1976). Purification of a phospho- the phytopathogen Botrytis cinerea. ACS Chemical Biology, lipase from Botrytis and its effects on plant tissues. 3(12), 791e801. Phytochemistry, 15, 1465e1470. Prins, T. W., Tudzynski, P., von Tiedemann, A., Siewers, V., Viaud, M., Jimeńez-Teja, D. G., Collado, I., Tudzynski, B., Have, A., Hansen, M. E., et al. (2000). Gronover, C. S., Pradier, J.-M., et al. (2005). Functional Infection strategies of Botrytis cinerea and related ana´lisis of the cytochrome P450 monooxygenase gene necrotrophic pathogens in fungal pathology. In bcbot1 of Botrytis cinerea indicates that botrydial is a strain- J. Kronstad (Ed.), (pp. 123e145). Dordrecht, the specific virulence factor. Microbial Plant Molecular Interac- Netherlands: Kluwer Academic Publishers. tion, 18(6), 602e612. Punithalingam, E. (1979). Phomopsis viticola. Descriptions Sivanesan, A., & Holliday, P. (1981). Guidnardia bidwellii. of pathogenic fungi and bacteria, no 635. UK. Kew: Descriptions of pathogenic fungi and bacteria, no 710. Commonwealth Mycological Institute. UK. Kew: Commonwealth Mycological Institute. Rebordinos, L., Cantoral, J. M., Prieto, M. V., Hanson, J. R., & Snowdon, A. L. (1990). A colour atlas of post harvest diseases Collado, I. G. (1996). The phytotoxic activity of and disorders of fruits and vegetables. London, UK: Wolfe some metabolites of B. cinerea. Phytochemistry, 42(2), Scientific. 383e387. Spotts, R. A. (1980). Infection of grape by Guignardia Rebordinos, L., Santos, M., Vallejo, I., Collado, I. G., & bidwellii e Factors affecting lesion development, conidial Cantoral, J. M. (1997). Molecular characterization of the dispersal, and conidial populations on leaves. Phytopa- phytopathogenic fungus B. cinerea.InRecent research thology, 70, 252e255. developments in phytochemistry, Vol. 1 (pp. 293e307). Tri- Tani, H., Koshino, H., Sakuno, E., Cutler, H. G., & vandrum, India: Research Signpost. Nakajima, H. (2006). Botcinins E and F and botcinolide Rebordinos, L., Vallejo, I., Collado, I. G., & Cantoral, J. M. from Botrytis cinerea and structural revision of botcino- (2003). Control of agricultural diseases and pests: The lides. J. Nat. Prod., 69, 722e725. case of B. cinerea. In J. L. Barredo (Ed.), Useful microor- Tani, H., Koshino, H., Sakuno, E., & Nakajima, H. (2005). ganisms for health care, foods and enzyme production (pp. Botcinins A, B, C and D, metabolites produced by 269e287). Trivandrum, India: Research Signpost. Botrytis cinerea, and their antifungal activity against Rebordinos, L., Vallejo, I., Santos, M., Collado, I. G., Magnaporthe grisea, a pathogen of rice blast disease. Carbu, M., & Cantoral, J. M. (2000). Ana´lisis gene´tico y J. Nat. Prod., 68, 1768e1772. relacioń con patogenicidad en Botrytis cinerea. Rev. Iber- Vallejo, I. (1997). Caracterizacioń de diferentes cepas de Botrytis oam. Micol, 17(1), S37eS42. cinerea por tećnicas microbiolo´gicas, gene´ticas y de biologıá Rui, O., & Hahn, M. (2007). The Botrytis cinerea hexokinase, molecular. Cadiz, Spain: Doctoral thesis. Universidad de Hxk1, but not the glucokinase, Glk1, is required for Ca´diz. normal growth and sugar metabolism, and for patho- Vallejo, I., Carbu, M., Munõz, F., Rebordinos, L., & genicity on fruits. Microbiology, 153, 2791e2802. Cantoral, J. M. (2002). Inheritance of chromosome-length Sall, M. A. (1980). Epidemiology of grape powdery mildew: polymorphism in the phytopathogenic ascomycete A model. Phytophatology, 70, 338e342. Botryotinia fuckeliana (Botrytis cinerea). Mycol. Res., 106, Santos, M., Rebordinos, L., Gutierrez, S., Cardoza, R. E., 1075e1085. Martıń, J. F., & Cantoral, J. M. (2001). Characterization of Vallejo, I., Carbu, M., Rebordinos, L., & Cantoral, J. M. the gdhA gene from the phytopathogen B. cinerea. Fungal (2003). Virulence of Botrytis cinerea strains on two Genet. Biol., 34, 193e206. grapevine varieties in south-western Spain. Biologia, 58 Santos, M., Vallejo, I., Rebordinos, L., Gutierrez, S., (6), 1074e1076. Collado, I. G., & Cantoral, J. M. (1996). An autono- Vallejo, I., Santos, M., Cantoral, J. M., Collado, I. G., & mously replicating plasmid transforms Botrytis cinerea Rebordinos, L. (1996). Chromosomal polymorphism in to phleomycin resistance. FEMS Microbiol. Lett., 137, B. cinerea. Hereditas, 124,31e38. 153e158. von Tiedemann, A. (1997). Physiol. Mol. Plant P., 50, 151e166. CHAPTER 11

Production of Wine Starter Cultures Ramoń Gonza´lez 1, Rosario Munõz 2, Alfonso V. Carrascosa 3 1 Instituto de Ciencias de la Vid y del Vino (CSIC-UR-CAR), Logronõ, Spain, 2 Instituto de Ciencia y Tecnologıá de los Alimentos y Nutricıón (ICTAN, CSIC), Madrid, Spain and 3 Instituto de Investigacioń en Ciencias de la Alimentacioń (CIAL, CSIC-UAM), Madrid, Spain

OUTLINE

1. Introduction 279 3. Lactic Acid Bacteria 292 3.1. Selection and Identiﬁcation of Strains 293 2. Yeasts 280 3.2. Production of Biomass 295 2.1. Historical Notes 280 3.3. Lyophilization, Packaging, and Storage 296 2.2. Isolation and Selection 281 3.4. Use 297 2.3. Production of Biomass 286 Acknowledgments 298 2.4. Drying 290 2.5. Use 291

1. INTRODUCTION The effectiveness of starter cultures in the The quality of fermented foodstuffs and wine industry is based largely on the microbio- beverages is determined in part by the microor- logical control that can be achieved during ganisms used in their preparation. The winemaking. Good cultivation practices that secondary character of wine, for instance, is limit contamination of the fruit with molds or determined by sensory characteristics that arise acetic or lactic acid bacteria prior to harvesting from the direct action of microorganisms on are key to obtaining a must that can be correctly the substrate. Consequently, the exploitation of fermented. Likewise, good manufacturing prac- organisms such as the yeasts and lactic acid tices that include appropriate winery hygiene bacteria responsible for alcoholic and malolactic programs favor the development of the inocu- fermentation, respectively, is a constantly ex- lated microorganisms and reduce the microbial panding branch of biotechnology. competition that they encounter, given that

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10011-6 279 Copyright Ó 2011 Elsevier Inc. All rights reserved. 280 11. PRODUCTION OF WINE STARTER CULTURES both musts and wine are nonsterile substrates. Definition of the Characteristics of the Desired The implementation of quality-control systems Wine for hygiene such as the Hazard Analysis Critical Control Point approach is by far the best guarantee of success when using starter cultures in Isolation the wine industry. Recent years have seen an increasing emphasis on the importance of wine quality. Selection Thus, obtaining a wine with characteristics that can be clearly distinguished from others Production of Biomass or with organoleptic properties that remain consistent year after year tends to be associated with increased competitiveness. To a large Concentration extent, consumers determine which wine is sold, and this must be defined in terms of measurable parameters, be that through the Drying use of instruments or by tasting, in order to produce wines correctly. Packaging The use of starter cultures may not meet expectations if the goals are not first understood in terms of measurable characteristics. Conse- Distribution quently, the first step in any wine production process (see Figure 11.1) is to clearly define the type of wine to be produced. This step is more Use important than usually considered since it guides both the preparation and choice of the FIGURE 11.1 Stages in the production of starter cultures starter culture. for use in winemaking.

2. YEASTS or unconsciously present in the minds of the first microbiologists to propose that the results 2.1. Historical Notes of wine fermentation might depend on the strain of yeast used. Nevertheless, it should be The process of grape-must fermentation for remembered that under industrial production the production of wine has been described since conditions grape must is not a sterile substrate antiquity and has been carried out for thou- and the inoculated yeasts therefore compete sands of years by trusting in the action of the with other yeasts and bacteria, particularly microorganisms naturally present on the grapes during the initial phases of fermentation (see and in the winery. However, the birth of micro- Chapter 1). biology as a scientific discipline led to improved At the beginning of the last century, the La control of the process as it enabled the identifi- Claire Institute in France began the isolation, cation of the microorganisms responsible for maintenance, and small-scale production of fermentation and offered an opportunity for pure strains of yeast (up to 100). These were winemakers to employ pure yeast cultures. used to establish pied de cuve cultures for inocu- Koch’s postulates must have been consciously lation of musts in a process that was gradually YEASTS 281 scaled up until it was possible to inoculate with a single pure strain of Saccharomyces cerevi- industrial fermentation tanks (Kraus et al., siae. To explain the observation that inoculation 1983). Commercially produced cultures that of Saccharomyces ellipsoideus some days after inoc- would be ready for inoculation in large tanks ulation of Candida pulcherrima increased the without prior preparation and scaling up of pied fermentative capacity of S. ellipsoideus, it was sug- de cuve cultures were first proposed in the 1950s. gested that the products of autolysis of the first At this time, the work of German, Californian, inoculated strain (proteins and vitamins) acted and Canadian researchers demonstrated that as growth factors for the true fermentative strain. yeast cells produced under aerobic conditions Nowadays, in fact, this effect is achieved by the were perfectly suitable for the fermentation of addition of products of yeast autolysis (yeast must (reviewed in Kraus et al., 1983). As we will cell walls, etc.). see later, prior growth under aerobic conditions The idea of using selected yeasts was brought also offers some advantages in terms of the to Spain by the Italian researcher Tommaso fermentative capacity of industrially produced Castelli, who in 1955 considered Gino de Rossi cells. the founder of the work he had undertaken The industrial production of wine yeasts as since 1933. The aim was to obtain a collection we know it today began in the 1960s, when of microorganisms from different winegrowing various producers of baker’s yeast began to regions according to the logic that those that produce wine yeasts using a similar procedure. were most abundant during the different phases This product had the usual characteristics of of fermentation (initial, tumultuous, and final) pressed baker’s yeast, including a 70% moisture would be potentially the most interesting for content. However, although it was relatively selection. Castelli claimed that fermentation easy to use, it had a short shelf life. It should using pure cultures did not work miracles but be remembered that the use of wine yeasts is allowed greater consistency, increased produc- essentially seasonal and can be difficult to coor- tion of alcohol, improved yield (alcohol to sugar dinate with production. Around 1964, the first ratio), and reduced volatile acidity, although efforts were made to develop a much more with the disadvantage of being less effective in stable product, active dried yeast, as a solution the warmest and coolest regions, since tempera- to this problem. ture-control systems were not widely available. The first report of a selected yeast being used The collaboration between Castelli and In˜igo led in Spain dates back to 1958 and makes reference to the first microbiological studies of Spanish to various microbiological studies of grape harvests in the regions of La Mancha and La harvests performed since 1956 in Spanish wine- Rioja (Castelli & In˜igo, 1957, 1958). growing regions (In˜igo, 1958). The study advocated the use of selected yeasts to avoid 2.2. Isolation and Selection excessive volatile acidity in Spanish wines. The author recommended the sequential inoculation Yeast inoculation offers a series of advantages of strains with weak, intermediate, and strong over spontaneous fermentation for the produc- fermentative capacity using pied de cuve culture tion of wine. These advantages depend on the to imitate the natural process occurring during type of vinification and the characteristics of spontaneous fermentation and thus allow the the yeast strain used. However, most notable desired wines to be obtained in a controlled among them are a more rapid onset and manner. It was concluded that the final volatile progression of fermentation and a reduction in acidity was lower with a stepwise inoculation volatile acidity associated with greater consis- program than when fermentation was performed tency in the quality of the product between 282 11. PRODUCTION OF WINE STARTER CULTURES different tanks and vintages. Specific contribu- The selection procedure generally begins tions to the organoleptic quality of the wine with a large number of strains (often several are also offered according to the characteristics hundred). From an organizational and of the yeast strain. In addition to these character- economic perspective, it is desirable for the istics, the selection criteria applied to potential initial phases of the process to employ easily new yeast strains should take into account three evaluated selection criteria that allow less inter- main elements: the conditions under which esting strains to be identified and ruled out fermentation will take place (characteristics of more quickly as this will help to reduce the the must and other environmental factors such number of strains that need to be analyzed as fermentation temperature), the quality and during the subsequent phases of the process. character of the wine obtained, and the survival More expensive tests and analytical methods of the strain during industrial production of can be left until the final stages, when only active dried yeast. a small number of strains need to be analyzed. The process of selection of new yeast strains A number of criteria have been proposed for for use in winemaking begins with the isola- the selection of new yeasts for use in winemak- tion of natural yeasts from different substrates ing, not all of which are relevant for the specific (generally from wine during the final phase conditions and characteristics of each wine. of fermentation but also, possibly, from grapes, Below we describe some of the most widely fresh must, or lees). A factor that should be applied. taken into account when isolating new strains Fermentation powerdunderstood as a mix- from must and in particular from wine is ture of rapid fermentation, short lag phases, whether the winery has previously used and almost complete consumption of the sugars commercial starter cultures, since in this case present in the mustdwas the first criterion used there is a high risk of repeatedly isolating to select wine yeasts (Kraus et al., 1983). The a known yeast strain for which use may be original purpose of inoculating selected yeasts restricted by intellectual property rights. As was to ensure that the fermentation process a result, the isolation of new strains should was not excessively long and also to prevent be undertaken in newly constructed wineries. stuck fermentation. This was sufficient to guar- Since during wine fermentation there is a selec- antee that wines of appropriate quality could tive enrichment in strains of the genus Saccha- be obtained year after year without spoilage romyces, a spontaneous fermentation that has due to the growth of undesirable micro- yielded good results in terms of the kinetics organisms as a result of poor fermentation of fermentation and the quality of the product kinetics. Good fermentation power is obviously should be an ideal source for the isolation related to the capacity of the strain to overcome and selection of new industrial strains. the stresses associated with wine fermentation, However, one should not ignore other sub- such as the hyperosmotic environment present strates that are less rich in strains of Saccharo- during the initial phases of the process or the myces, such as must or grapes, as a source of elevated concentrations of ethanol and scarcity new strains with interesting properties. The of assimilable nitrogen and growth or survival selection of new strains is increasingly based factors during subsequent phases. on very specific criteria, such as fermentation Fermentation power estimated under labora- at low temperatures or specific contributions tory conditions, using either natural musts or to secondary aroma, and these are more diffi- culture media designed to mimic their composi- cult to identify without using a wide range of tion, tends to be a poor predictor of the behav- isolates. ior of strains under industrial production YEASTS 283 conditions. As a result, various authors have strain and is highly related to the activation of proposed additional selection criteria based on stress response mechanisms associated with the identification of factors that limit the entry into the stationary phase (see Chapter 1). survival of yeasts during the vinification Consequently, nitrogen demand has also been process. For instance, Ivorra et al. (1999) proposed as a selection criterion for industrial observed a negative correlation between the Saccharomyces strains (Manginot et al., 1998). resistance of strains to various stress factors Recently, Marks et al. (2008) performed a and the likelihood of stuck fermentation. In transcriptomic analysis of S. cerevisiae during that study, analysis of expression levels for alcoholic fermentation and defined a “fermenta- a number of genes led the authors to propose tion stress response” (FSR). According to those HSP12 expression as a marker of resistance to authors, this response will be different from, stress in the corresponding strain. Subsequently, although obviously overlap with, the response Zuzuarregui and del Olmo (2004a) used multi- to the individual stress factors described earlier. variate analysis as a basis for suggesting resis- Of the genes found to participate in the FSR, tance to ethanol and oxidative stress as 62% had not previously been linked to the selection criteria for strains with good fermenta- global stress response and 28% had not yet tion behavior. Interestingly, the response to been functionally annotated in sequence oxidative stress would naturally be expected databases. to have more relevance for the behavior of the As described in Chapters 1 and 6, sulfite strain during production of active dried yeast resistance is a common characteristic of S. cere- than as a predictor of its fermentation power. visiae wine strains that can affect their capacity The same authors showed a correlation between to complete fermentation. Pe´rez-Ortıń et al. the expression level of certain stress-response (2002) and Yuasa et al. (2004) reported that genes and the fermentative capacity of the most of the S. cerevisiae wine strains analyzed wine strains studied (Zuzuarregui & del Olmo, in their studies had at least one copy of the 2004b). A conclusion that can be drawn from SSU1-R allele, which confers greater resistance their study is that, although good induction of to sulfite than the wild-type SSU1 allele. This stress-response genes is necessary for the strain is probably due to continued selection pressure to perform well under adverse conditions, (see Chapter 6), given that the use of sulfite is excessive or prolonged expression may also be almost as old as wine production itself. Since it prejudicial. Increased expression levels of is unlikely that the use of sulfites in winemaking GPD1 have also been observed in various will be abandoned in the short or medium term commercial strains during the initial phase of (Romano & Suzzi, 1993), resistance to sulfites fermentation (Zuzuarregui et al., 2005). This will continue to be a character to take into may improve the response to osmotic stress consideration during the selection process, generated by the high concentration of sugars especially when selection is not limited to S. cer- in the must, since the enzyme encoded by evisiae, the only species identified to date that GPD1 is necessary for intracellular accumula- contains the SSU1-R allele. tion of glycerol. Compared with other yeasts found in must or After ethanol, one of the main causes of slug- used as starter cultures, strains of S. cerevisiae are gish or stuck fermentation tends to be the lack of considered to be relatively heat-tolerant. Ther- assimilable nitrogen caused by modern wine- motolerance, or resistance to thermal stress, making practices. The capacity to continue has been an important characteristic in wine fermentation under conditions of limited strains, particularly before the introduction nitrogen depends on the characteristics of the of temperature control systems, since the 284 11. PRODUCTION OF WINE STARTER CULTURES temperatures reached during exothermic the producing strain to the toxin produced by fermentation could be suboptimal for growth. that strain or by other cells that produce a killer As with other characters linked to fermentation factor of the same type. The toxin is able to kill power, thermotolerance is associated with the strains that do not produce a killer factor or expression of stress-response genes. It is no that produce a killer factor of a different type. coincidence that some of the first stress- In principle, the use of killer strains of S. cerevi- response genes were characterized on the basis siae as starter cultures should contribute to of their expression in response to thermal shock favoring the establishment of the inoculated (heat shock proteins) (Craig, 1985). Resistance to strain during fermentation. However, there is thermal stress can thus be a criterion for selec- no consensus regarding the true relevance of tion, partly because it is linked to fermentative the killer factor under natural conditions. The capacity (Ivorra et al., 1999) and partly because killer factor K2 would seem to be of most heat tolerance may be an important factor interest due to its activity and stability at the when yeasts are dried during preparation for pH found in grape must and wine (Heard & industrial use. Fleet, 1987). However, strains expressing The ability to control fermentation tempera- a wide range of killer factors have been con- ture has also made it possible to develop new structed by genetic engineering in an effort to styles of vinification. For instance, cold fermen- improve their survival capacity or their ability tation has become increasingly popular since it to eliminate undesirable yeasts and thereby confers some very interesting aromatic proper- prevent wine spoilage (see Chapter 7). ties on the wines produced. This occurs as Despite its overall importance, good fermen- a result of greater retention of the volatile tation power is not in itself an adequate criterion substances formed during fermentation, the for the selection of winemaking strains. Many of effect of temperature on yeast metabolism, and the additional selection criteria are based on the the characteristics of the cryotolerant strains positive or negative influence that a strain may used in winemaking, which tend to belong to have on the sensory qualities of the wine (or be hybrids of) the species S. bayanus. Conse- produced. Yeast metabolism has a notable influ- quently, cryotolerance has become an addi- ence on secondary aroma. This refers to the tional selection criterion for yeasts destined for contribution to wine aroma made by microor- use in fermentations performed below 15C. ganisms present during fermentation. Chapter Another hereditary characteristic that can 4 describes the most relevant compounds influence the capacity of a strain to induce involved in the formation of secondary aroma complete fermentation of grape must is the by yeasts of the genus Saccharomyces and other killer factor. Three killer toxins have been genera. The contribution to secondary aroma described in S. cerevisiae: K1, K2, and K28 may differ among Saccharomyces strains and (Magliani et al., 1997). These toxins are encoded can therefore be used as a target for selection by double-stranded satellite RNAs (M1, M2, according to the quality of the aroma and its and M28, respectively) that are found in the appropriateness in a given type of wine. Some cytoplasm of the producing strains and encap- of the most interesting metabolites in this sulated in icosahedral particles similar to respect are esters of acetic acid and ethyl esters. viruses. In turn, these RNAs depend on other The levels and relative abundance of the double-stranded RNA viruses known as L-A different esters that contribute to this secondary for their replication and encapsidation. The M aroma are the result of the activity of groups of RNAs are responsible for the synthesis of the enzymes with antagonistic effects: alcohol ace- corresponding toxin and for the immunity of tyltransferases, which catalyze the synthesis of YEASTS 285 these compounds, and esterases, which catalyze produced by different industrial strains of S. their hydrolysis. Genetic engineering has been cerevisiae may act as major determinants of the used to construct strains of Saccharomyces in quality of the finished wine. The main source which the production of esters is increased by of sulfur for the formation of hydrogen sulfide deleting genes that encode esterases and overex- can be either sulfate or sulfite, and depletion pressing genes that encode alcohol acetyltrans- of nitrogen sources has been identified as one ferases (see Chapter 7). of the main determining factors for its forma- The primary or varietal aroma, which is tion. Various genes have been linked to the essentially derived from the grape, can also be production of hydrogen sulfide, including affected by the strain of yeast used (see Chapter MET17 and NHS5, which can suppress the 4) through the action of hydrolytic enzymes that formation of hydrogen sulfide when overex- release terpenes from their glycosylated precur- pressed in yeast (Spiropoulos & Bisson, 2000; sors or facilitate the extraction of aromas and Tezuka et al., 1992). colors from the cell wall of the grapes. Conse- Another undesirable product of yeast metab- quently, the production of hydrolytic enzymes olism is urea. Although it does not influence with these activities has also been proposed as the sensory quality of wine, its presence in an a selection criterion. aqueous, alcohol-containing medium such as Another product of yeast metabolism with wine can over time give rise to the formation a positive contribution to many of the sensory of ethyl carbamate, a toxic compound that is properties of wine is glycerol. Furthermore, also a suspected carcinogen. Many countries increased glycerol levels go hand in hand with have produced legislation on the maximum reduced concentrations of alcohol. Conse- permitted levels of ethyl carbamate in imported quently, the ever-growing demand for wines wines. In wine, urea is mainly derived from the with lower alcohol content has led to a growing action of arginase produced by Saccharomyces interest in strains that produce higher levels of yeasts. This is the first enzyme in the catabolic glycerol and lower levels of ethanol. Recombi- pathway of arginine and catalyzes its hydro- nant yeast strains have been produced in which lysis to give rise to urea and ornithine. One of more glycerol is produced at the expense of the strategies recommended by the United ethanol. However, the high concentrations of States Food and Drug Administration (FDA) acetic acid produced by these strains made it to minimize the formation of ethyl carbamate necessary to incorporate additional genetic is the selection of strains with low urea produc- modifications to prevent excessive volatile tion. Recombinant strains of sake yeast have acidity (see Chapter 7). even been produced in which the gene CAR1, Volatile acidity (due to excess acetic acid) encoding arginase, has been deleted to allow is one of the most easily detected wine flaws. the generation of sake completely lacking Although excessive volatile acidity can often ethyl carbamate (Kitamoto et al., 1991). Coulon be attributed to the uncontrolled growth of ace- et al. (2006) recently used a different strategy tic acid bacteria, there can also be an appre- to the same end, constructing a wine yeast ciable, strain-dependent contribution of S. that was able to eliminate urea from the wine cerevisiae metabolism (Delfini & Cervetti, 1991), through the constitutive expression of the making this an important selection criterion. gene DUR1,2, which encodes urea amidolyase. Excess hydrogen sulfide can also be an The same group used a similar strategy to important defect in finished wine. Given that produce the yeast ECMo01, which has been the sensory threshold for hydrogen sulfide approved by the FDA and Health Canada for is extremely low, differences in the levels commercial use. 286 11. PRODUCTION OF WINE STARTER CULTURES

In all protocols for the selection of indus- The proliferation phase (or production of trial yeasts, tasting is one of the most difficult biomass) naturally begins with a pure culture and at the same time most important steps in that must have been maintained under appro- identifying the most interesting strains. This priate conditions to ensure both purity and is the phase that allows confirmation of genetic stability (see Chapter 12). The scaling- various indicators obtained in earlier tests. up process allows multiplication of the yeast For instance, the potential contribution of from the few hundred million cells typically the strain to primary and secondary aroma present in the starting culture by gradually can now be confirmed by the detection of increasing the culture volume through fermen- undesirable flavors or aromas. Likewise, ters of increasing capacity (from 5 to 250 L), until tasting can be used to determine whether the fermenters with a capacity of hundreds of thou- finished wine has the typical characteristics sands of liters are reached. Maintaining the of the type of wine that was intended to be purity of the culture is a factor that must be produced. taken into account throughout scale-up and in Specific selection criteria may also be used all phases of the production process, although according to the specific style of vinification. the presence of other yeasts and bacteria is As mentioned earlier, cryotolerance would be normal in the final product. An indicator of a criterion for cold fermentation, but autolytic quality in active dried yeast is that the presence and flocculation capacity can also be considered of bacteria and yeasts other than those for the production of traditional-method spar- belonging to the genus Saccharomyces does not kling wines (Cava and Champagne); flor forma- exceed 0.01% of the concentration of cells tion, ethanol tolerance, and other necessary surviving rehydration. characteristics for the production of biologically As in all microbial growth processes, the aged wines; and autolytic capacity for the function of the culture medium is to guarantee production of wines aged on lees. All of these adequate provision of the nutrients required factors are discussed in more detail in other for growth. For the industrial production of chapters. yeasts, the best carbon source is cane or beet Finally, successful marketing of wine yeast molasses. This is an inexpensive substrate that strains will depend largely on their behavior is very rich in sucrose, a sugar that is easily under industrial production conditions, particu- assimilated by S. cerevisiae thanks to genes that larly in terms of genetic stability, growth on encode various forms of invertase. Under molasses, and survival and metabolic activity production conditions, it is calculated that following drying and rehydration. invertase activity allows sucrose to be hydrolyzed some 300 times faster than the resulting 2.3. Production of Biomass glucose and fructose are assimilated by the cells (Sańchez, 1988). Consequently, the initial The first companies to attempt the commer- paucity of monosaccharides in the substrate cial production of wine yeast strains were does not represent a limitation for the use of producers of baker’s yeast and, in fact, as molasses. In addition to sucrose, molasses mentioned earlier, in the 1960s wine strains usually represents an adequate source of other were still sold as pressed yeast. Nowadays, the essential nutrients for yeast growth, including initial phases in the production of active dried some minerals, oligoelements, and vitamins. yeast are very similar to those involved in the However, molasses does not generally represent production of pressed yeast, and the same a good source of nitrogen or phosphorus. There- factors must be taken into consideration. fore, in addition to dilution, it must also be YEASTS 287 supplemented with these nutrients, usually in A (CoA). Since the subsequent reactions that the form of ammonium salts. Since molasses is would allow transformation of the acetalde- a byproduct of another industrial activity and hyde formed by pyruvate decarboxylase into no standardization processes are applied at acetyl-CoA are limited, this ultimately favors source, it is also necessary to confirm the the formation of ethanol even under aerobic composition of each batch of molasses so that conditions (Potma et al., 1989; Pronk et al., in each case the appropriate quantities of nutri- 1996). Figure 11.2 shows a schematic diagram ents can be added. The presence of potential of the relative metabolic flows occurring in growth inhibitors such as sulfites, organic acids, the Crabtree effect. and nitrites must also be taken into consider- Despite adequate aeration, the high initial ation and any defects rectified prior to use. concentration of sugars means that the yeast The pH of the molasses is also checked prior initially adopts a fermentative metabolism to use as a substrate for fermentation and nor- following inoculation of the molasses. The sugar mally adjusted to a pH of 5, which tends to be concentration also places the yeast under optimal for the growth of S. cerevisiae on this osmotic stress. As the culture grows, the sugars substrate. begin to be depleted and controlled feeding is Production of biomass is carried out under required. To minimize the Crabtree effect, aerobic conditions in order to achieve two however, it is necessary to maintain the sugar main goals. The first is to obtain the greatest concentrations at low levels. A “fed-batch” possible yield from the process, expressed as process is therefore used in which molasses is the quantity of biomass produced for a given added little by little as it is consumed by the quantity of molasses. The incomplete oxidation yeast, thus minimizing the production of of glucose that occurs during fermentation ethanol without halting growth. To optimize leads to a net energy yield per mole of sugar fed-batch feeding, feedback systems are usually consumed that is lower in the case of fermenta- employed. In those that work best, the addition tion (56 kcal/mol) than in respiration (688 kcal/ of the substrates is regulated by the respiratory mol). In order to ensure minimal fermentative metabolism and to maximize sugar consumption by respiration, it is essential to maintain Glucose Ethanol good aeration; however, the Crabtree effect must also be taken into account. This metabolic Glycolysis Alcohol phenomenon seen in many yeasts, including . dehydrogenase S Pyruvate cerevisiae, leads part of the sugar consumed to decarboxylase Acetaldehyde be converted into ethanol via the fermentative pathway. This occurs when glucose concen- Pyruvate Acetaldehyde dehydrogenase tration exceeds a relatively low threshold Pyruvate Acetate (0.1e0.5 g/L), even in the presence of sufficient dehydrogenase quantities of oxygen. The Crabtree effect is complex Acetyl-CoA caused by the generation of high intracellular Tricarboxylic synthase concentrations of pyruvate in the presence of acid cycle glucose. This favors pyruvate degradation via Acetyl-CoA a pathway involving pyruvate decarboxylased which has a high loading capacity and a high Kmdrather than the pyruvate dehydrogenase FIGURE 11.2 Relative metabolic flows associated with complex, which leads directly to acetyl-coenzyme the Crabtree effect in Saccharomyces cerevisiae. 288 11. PRODUCTION OF WINE STARTER CULTURES quotient or RQ (moles of carbon dioxide formed of stress-response genes and genes involved in per mole of oxygen consumed) (Aiba et al., 1976; the synthesis of storage compounds. These Cooney et al., 1977; Wang et al., 1977). These processes are normally repressed under condi- control systems analyze the flow and composi- tions of rapid growth and are induced upon tion of gases entering and leaving the fermenter entry into the stationary phase. As a result, the and regulate the provision of the substrate final step in the production of biomass is carried according to the value obtained. As we will out in the presence of limiting nitrogen while see below, the molecular responses to osmotic maintaining sufficient levels of the carbon and oxidative stress are the main adaptations source to allow the synthesis of glycogen and required of the yeast during production trehalose. The concentration of phosphorus is (Pe´rez-Torrado et al., 2005). also controlled since it plays a role in deter- Another goal during the production of mining the rate of division and the stability of biomass under aerobic conditions is to allow the cells during dehydration (Sańchez, 1988). the yeast cells to synthesize sterols. These Nitrogen limitation plays a role in the entry of compounds are survival factors that cannot be cells into the stationary phase. This is a perfectly synthesized under anaerobic conditions. Conse- regulated cellular process brought about by the quently, if the inoculum contains limiting sterol lack of an essential nutrient (Werner-Washburne reserves, its capacity to complete fermentation et al., 1993). Despite the lack of nutrients, the will depend on the characteristics of the must cells are able to complete the cycle of cell divi- (see Chapter 1). sion that has been initiated, meaning that all Finally, the production of biomass should the cells in the stationary phase lack buds and take into account factors that affect the capacity have the same DNA content. The absence of of the yeasts to survive the drying process and cell division in no way implies metabolic inac- to recover their viability and fermentative tivity, since, as discussed in Chapter 1, most of capacity following rehydration and inoculation the fermentation process occurs without any into the must. The ability of the yeasts to survive increase in the number of viable cells. the drying and rehydration process will depend Actively dividing cells pass through the four on the expression of stress-response genes, different phasesdG1, S, G2, and Mdthat make either directly or via the synthesis of storage up the cell cycle. In G1, which is normally initi- compounds such as trehalose and glycogen ated following cell division, each of the resulting (François & Parrou, 2001). Mobilization of these cells (mother and daughter) increases in size, storage compounds through the action of with a corresponding synthesis of macromole- hydrolytic enzymes is also important for the cules (proteins, cell wall polysaccharides, recovery process (Lillie & Pringle, 1980; Novo RNA, etc.). During the S phase, there is a dupli- et al., 2003; Thevelein, 1994). Trehalose is cation of the genetic material such that at the thought to play an additional role as a protective end of this phase the cells contain twice the agent, helping to maintain the integrity of cell content of DNA and nuclear proteins. G2 membranes and stabilizing the native confor- involves de novo synthesis of macromolecules mation of proteins (Felix et al., 1999; Leslie and morphological changes required for cell et al., 1994). The specific genotype of each strain division or mitosis. The M or mitosis phase will thus be important in determining its ability involves separation of the chromosomes to respond to these treatments and, therefore, to between the dividing cells and corresponds to be sold as active dried yeast. cell division proper, which gives rise to two It is also important to take into consideration cells, mother and daughter. Currently, the most the culture conditions responsible for induction widely accepted view is that, when cells stop YEASTS 289

protein kinase C pathway, and the SNF1 pathway. Cells that have not entered the M stationary phase and are therefore actively G0 proliferating have very little chance of surviving Stationary the subsequent treatments involved in the G2 G1 phase production of active dried yeast. Although the entire production process has S been optimized to obtain maximum yield in the production of biomass, there has been little characterization of the molecular mechanisms FIGURE 11.3 Phases of the Saccharomyces cerevisiae cell used by yeast to adapt to the changing condi- cycle. M ¼ mitosis; G0 ¼ adaptation phase during entry into tions that this entails. One of the main difficul- the stationary phase; G1 ¼ growth phase; S ¼ synthesis ties faced in such studies is the simulation of phase; G2 ¼ preparation phase prior to division. industrial processes under laboratory conditions. Recently, this has begun to be resolved dividing and enter the stationary phase, they through the use of micro fermenters with exist in a phase outside of the normal cell cycle similar biomass yield and microbial growth known as G0 (Figure 11.3)(Werner-Washburne conditions to those found in industrial situa- et al., 1993). In this phase, the cells acquire tions, making it possible to use molecular a series of adaptations that help them to survive techniques to understand and improve the in adverse conditions. Examples of these adap- behavior of yeasts during industrial production tations include thickening of the cell wall, accu- (Pe´rez-Torrado et al., 2005). mulation of storage polysaccharides, expression The shift from fermentative to respiratory of a variety of stress-response genes, and activa- metabolism leads to the inhibition of mitochon- tion of mechanisms to allow recycling of cell drial metabolism and causes changes in the components, particularly proteins, via path- intracellular redox potential that can place the ways such as ubiquitin-dependent proteolysis yeast under oxidative stress. The response of or autophagy (see Chapter 2). S. cerevisiae to oxidative stress has been investi- The entry and maintenance of cells in the gated under different conditions, including the stationary phase involve the expression of addition of agents responsible for the generation various genes with different specific functions, of reactive oxygen species (ROS) (hydrogen including resistance to the different types of peroxide, menadione, and various metal ions) stress mentioned above. These changes in and laboratory conditions that produce changes expression are regulated by both transcriptional in the cytoplasmic redox potential (Gibson et al., and post-transcriptional mechanisms. The 2008). Various genes have been identified as response to the lack of nutrients that leads markers of the response to oxidative stress. ultimately to entry into the stationary phase is Firstly, TRX2 (which encodes thioredoxin, an regulated by a complex network of signal trans- enzyme that protects against the toxic effects duction pathways (Gray et al., 2004) based on of hydrogen peroxide) is expressed strongly protein kinases (some of which can be consid- during the transition in the first fed-batch from ered nutrient sensors while others function at fermentative to oxidative growth and acts as later points in the signal transduction pathway). one of the best defenses against oxidative These include the TOR pathway, the cyclic damage. GSH1, which initiates the synthesis of AMP-dependent, protein kinase A pathway glutathione (GSH), and GRE2, which encodes regulated by Ras (RAS/cAMP/PKA), the a reductase associated with oxidative stress, 290 11. PRODUCTION OF WINE STARTER CULTURES also act as indicators for the molecular response again, this time using a rotary vacuum filter or to oxidative stress (Pe´rez-Torrado et al., 2009). a filter press, to produce a paste containing Activation of the gene GPD1, which is linked 35% dry material. It is then extruded to form to the synthesis of glycerol, has also been fine filaments (2e4 mm in diameter) that are described as a consequence of the osmotic stress further dried (to a moisture content of 4e8%) that can occur during the early phases of the using a counter-current hot-air fluidized-bed production of wine yeast, and to a lesser extent dryer. The granules that are produced by in response to oxidative stress (Pe´rez-Torrado breaking these dried filaments are vacuum et al., 2005). The key role played by the Yap1p packaged in the presence of an inert gas or factor in mediating the induction of a set of carbon dioxide. It is important to adequately genes involved in redox protection (Rodrigues isolate the granules from the oxygen in the air Pousada et al., 2004) is well known, as is the acti- in order to maintain the stability of the product, vation of factors involved in the general which should also be maintained at low temper- response to stress such as Hsf1p and its ature (4e8C). homolog Skn7p. Other enzyme systems Despite all of the precautions taken and the involved in the elimination of proteins and process having been optimized over a number lipids damaged by oxygen have also been of years, the numbers of revivable cells present recently described (Toledano et al., 2007). in different batches of active dried yeast can Growth under conditions involving limited vary substantially according to the yeast strain feeding can lead to the temporary absence of and the production conditions. Not surpris- nutrients, leading to stress as a result of nutrient ingly, one of the fundamental characteristics depletion. Consequently, it may be appropriate considered during the selection of new strains to consider using resistance to these stresses as of wine yeast for commercial use is their a selection criterion. Resistance of yeast to these capacity to survive the process of drying while and other types of stress appears to be corre- maintaining their viability during extended lated with their suitability for use in industrial periods of storage. production (Zuzuarregui & del Olmo, 2004a, Tolerance of desiccation facilitates a process 2004b; Zuzuarregui et al., 2005), and it may known as anhydrobiosis, referring to a state therefore be appropriate to consider using labo- in which metabolism ceases due to the lack ratory tests such as those described by a number of water. Although S. cerevisiae tolerates these of authors (Carrasco et al., 2001; Zuzuarregui & conditions well, the molecular details of the del Olmo, 2004a) as selection criteria for use phenomenon are poorly understood. The loss with these types of yeast. of water leads to collapse of the cytoskeleton, which affects cell physiology, membrane integ- 2.4. Drying rity, etc. Although it has generally been accepted that molecules such as trehalose can Once the required biomass has been actasreplacementsforwatermoleculesand obtained, the yeasts must be separated from stabilize the cell during the desiccation the culture medium, dried, and prepared for process, this model has recently been ques- sale (Papin, 1988). The first step in this process tioned due to the demonstration that deletion involves the production of a paste known as of TPS1 (encoding trehalose-6-phosphate syn- a “cream yeast,” which contains 150 to 200 g/L thase) in an S. cerevisiae mutant did not reduce of dry material. To obtain this, the yeasts are the tolerance of desiccation compared with the separated from the culture medium by centrifu- wild-type strain (Ratnakumar & Tunnacliffe, gation or filtration. The paste is then filtered 2006). YEASTS 291

Other recent studies have shown that the a fundamental role in the survival of these cellular response to drying involves the regula- yeasts during the critical phase of rehydration. tion of the transcription or translation of genes The in vitro presence of this disaccharide results involved in metabolism and in the synthesis of in a reduction of around 20C in the transition proteins and lipids (Novo et al., 2007; Rossignol temperature from dry gel to liquid crystal (the et al., 2006; Singh et al., 2005). Specifically, transition temperature determined for isolated hydrophilic proteins or hydrophilins are now plasma membranes is 60C) (Leslie et al., believed to confer resistance to the process of 1994). Thus, cells containing sufficient trehalose dehydration and rehydration in yeasts as they that are rehydrated at temperatures of around do in other living organisms (Tunnacliffe & 38C can avoid the negative effects of passing Wise, 2007). through a phase transition. Recent studies addressing the molecular 2.5. Use processes that occur during rehydration have revealed a specific role for genes linked to the Dried yeast can be used directly to inoculate synthesis of lipids and proteins (Novo et al., grape must. However, in order to ensure the 2007; Singh et al., 2005). These studies have best results, a prior rehydration step is normally begun to employ the concepts of viability and employed. Various procedures can be used for vitality. Viability refers to the capacity of the the rehydration of dried yeast, and companies cell to divide, whereas vitality relates to the usually advise on the best one for their product. capacity to remain metabolically active. Rodrı´- One of the most widely used options is warm guez-Porrata et al. (2008) recently performed water, although extended periods under these a systematic analysis of the optimal conditions conditions should be avoided in order to for rehydration using fluorescence microscopy prevent loss of viability caused by the hypo- and flow cytometry. They found that one of osmotic medium. Direct rehydration of yeasts the principal causes of reduced vitality during in must tends to be slower and leads to death rehydration is the loss of intracellular sub- as a result of the high osmotic pressure. Other stances and that magnesium is an important media used include mixtures of water and element in limiting this loss and maintaining must or water and sugar. Under these condi- cell vitality. Vaudano et al. (2009) studied the tions, the yeast are provided with a carbon expression of eight genes during the rehydra- source while maintaining an osmotic pressure tion process using real-time quantitative poly- that is more compatible with the recovery of merase chain reaction (PCR). Their results vitality. When must or diluted must is used, it suggested that the yeast reacts immediately to should be as clean as possible. In other words, rehydration only when there is a fermentable it should be reasonably free of microorganisms carbon source present in the medium. Further- and inhibitory substances such as fungicides, more, the expression of MEP2 was modulated pesticides, and sulfites. Some companies that by the concentration of ammonia, suggesting produce products for use in winemaking sell that catabolic repression by nitrogen is active media with a specific formulation, generally during the rehydration phase. a commercial secret, that guarantees the best During the process of rehydration and inocu- results during the rehydration phase. Rehydra- lation into fresh must, the yeast cells must tion is generally performed at temperatures respond to the presence of nutrients by exiting between 30 and 38C for 20 to 30 min. The the stationary phase. As in the entry into the protective capacity of trehalose in maintaining stationary phase, exit from this phase involves the integrity of the cell membrane can play reprogramming of gene expression driven by 292 11. PRODUCTION OF WINE STARTER CULTURES signal transduction pathways involving protein the most tolerant of the adverse conditions phosphorylation. Although the exit from the found in wine (Alegrıá et al., 2004; Guerzoni stationary phase has been much less extensively et al., 1995). studied than entry into that phase, it appears Why inoculate wine with selected bacteria? that the essential signal involves the availability The induction of malolactic fermentation by of a carbon source (whereas the availability of inoculation offers a series of advantages, a nitrogen source is insufficient to signal exit including greater control over the timing and from G0) (Granot & Snyder, 1991). length of malolactic fermentation and also Once fermentation has been initiated with the over the strain of lactic acid bacteria that carries dried yeast, it is possible to continue using it in out the process. This inoculation allows: the same winery as a pied de cuve. Here, the fermenting must is diluted in fresh must. In prac- 1) Rapid onset of malolactic fermentation at the most tice, this is performed by progressive filling of appropriate moment. If the bacterial population the fermentation tank or by use of the pied de has been adequately controlled, at the end of cuve to inoculate must at a proportion of 10 to alcoholic fermentation the wine will contain 30% during fermentation. When the yeast is very few bacteria and, therefore, may require used in this way, some of the advantages weeks or even months before an adequate relating to the production process or the purity spontaneous bacterial population is present. of the yeast may be lost. The loss of viable bacteria in the wine during this period and the requirement for increased temperatures in order for malolactic 3. LACTIC ACID BACTERIA fermentation to occur correctly creates a costly problem for the winery. The use of an Malolactic fermentation was discovered at inoculum containing 106 cells/mL can help the end of the nineteenth century by Mu¨ ller- to avoid significant delays. Thurgau. He described the transformation of 2) Maintenance of wine quality. The bacterial L-malic acid into L-lactic acid and carbon population never comprises a single dioxide. Malolactic fermentation occurs natu- microorganism. Spontaneous malolactic rally around the end of alcoholic fermentation. fermentation is carried out by different Sometimes this fermentation can be delayed or strains of O. oeni and, often, other bacterial may not even occur as a result of an inappro- species. Greater variability in this population priate temperature, pH, or sulfite or alcohol increases the risk of negative effects on content, or even due to the presence of phages. fermentation or of undesirable metabolites Starter cultures are preparations containing being produced. cultures of one or more strains belonging to 3) Control over the type of wine produced. The use one or more species of microorganism that are of selected bacterial cultures ensures that the used to inoculate a substrate in order to initiate quality of wine sought by the producers can fermentation. Since the beginning of the 1980s, be obtained. This last point is very important starter cultures have been available to induce since malolactic fermentation is not only malolactic fermentation. These contain strains a process of deacidification of the wine but of lactic acid bacteria belonging to the species also, depending on the strain used, an Oenococcus oeni, Lactobacillus plantarum, and opportunity to obtain additional advantages Lactobacillus hilgardii as preparations of one or by preventing the production of secondary more strains (Hammes, 1990). It has been metabolites that can have a negative effect on demonstrated that strains of these species are the wine. LACTIC ACID BACTERIA 293

The use of spontaneous malolactic fermenta- laboratory-, pilot-, and industrial-scale develop- tion to some extent makes the results unpredict- ment before freezing or lyophilization in prepa- able, delays the process considerably, and can ration for commercial use. Experiments are also ultimately prevent the complete degradation performed under semi-production and produc- of L-malic acid. Consequently, the use of com- tion conditions to assess the quality of the starter mercial starter cultures is recommended. cultures. Generally, to assess whether a strain is As occurs with yeasts, the use of malolactic suitable for use as a starter culture, the survival, starter cultures means that numerical superi- time required for malolactic fermentation, and ority of the bacteria over potentially competing sensory and chemical characteristics of the native strains is achieved immediately through wine obtained are assessed. displacement as a result of mechanisms such as competitive exclusion. Compared with 3.1. Selection and Identification spontaneous malolactic fermentation by the of Strains microflora present in the must, appropriately prepared commercial starter cultures produce Three groups of selection criteria have been more predictable results in terms of the rate of identified that should be met by starter cultures degradation and the final concentration of L- used to induce malolactic fermentation in wine malic acid, and in the production of metabolites (Buckenhu¨ skes, 1993); these are summarized in that ensure and improve sensory quality. Table 11.1. The selection of strains that are well Nowadays, many commercial starter cultures adapted to carrying out malolactic fermentation are available to induce malolactic fermentation. under specific winemaking conditions is of Most consist of strains of lactic acid bacteria, particular importance since wine production mainly O. oeni, which have a high malolactic varies from one region to another, as does the activity and a high tolerance of low pH and pH of the wine and the temperature. Malolactic high ethanol content. Although these starter bacterial strains vary in their tolerance of the cultures have been commercialized in various different stresses associated with the wine envi- forms, including fresh, frozen, and lyophilized ronment (Gindreau et al., 2003; Gockowick & cultures, from a commercial point of view, Henschke, 2003; Maicas et al., 1999a). As a result, lyophilized preparations are preferable. This is strains are usually isolated from samples of because fresh starter cultures must be produced wine in which active malolactic fermentation and sold directly in the producing regions, or is taking place. used immediately in the producing wineries. It has been known for some time that expo- In the case of frozen starter cultures, transport sure to stressful conditions such as heat, cold, over long distances is complicated by the diffi- ethanol, acid pH, etc. can protect against hostile culty of guaranteeing that the required temper- environment conditions (van de Guchte, 2002). ature is maintained. Problems are also This adaptive response requires the activation associated with lyophilized cultures, since there of certain defense mechanisms, and in this is a marked loss of viability when they are inoc- way the bacteria become more tolerant of ulated directly into the wine (Krieger et al., adverse conditions through exposure to condi- 1993). Consequently, most efforts are now tions of moderate stress. For instance, acclimati- focused on the development of lyophilized zation to cold temperatures can be used to malolactic starter cultures that can be directly obtain cryotolerant lactic acid bacteria (Panoff inoculated into wine without prior treatment. et al., 2000). Another exploitable defense mech- After isolation, the starter cultures are sub- anism is the accumulation of osmoprotective jected to various selection stages prior to organic compounds as a response to osmotic 294 11. PRODUCTION OF WINE STARTER CULTURES

TABLE 11.1 Criteria for the Selection of Lactic Acid the stressful conditions found in wine (Bourdi- Bacteria to Induce Malolactic Fermenta- neaud et al., 2003; Delmas et al., 2000; Guzzo tion in Wine et al., 2000; Jobin et al., 1999a, 1999b; Tourdot- 1. FIRST-ORDER CRITERIA Marechal et al., 2000). In addition, the response of . to osmotic stress, cold shock, 1.1. Resistance to low pH L plantarum and redox potential has been analyzed along 1.2. Resistance to ethanol with its behavior following freezing and drying 1.3. Tolerance of low temperatures (Carvalho et al., 2002; Molina-Gutierrez et al., 2002; Ouvry et al., 2002; Smelt et al., 2002). 1.4. Reduced metabolism of hexose and pentose sugars O. oeni is the principal microorganism involved in malolactic fermentation under the 2. SECOND-ORDER CRITERIA stressful conditions habitually found in wine. 2.1. High viability following propagation in Nevertheless, the inoculation of O. oeni starter a standardized medium cultures leads to significant cell death and, 2.2. Short propagation time in a standardized consequently, failure of malolactic fermentation. medium As a result, in order to achieve better control 2.3. High production of biomass in a standardized over malolactic fermentation in the winemaking medium industry, it is essential to understand the mechanisms involved in stress and ethanol tolerance. 2.4. Rapid survival kinetics in a standardized medium One of the most widely studied aspects of the response to ethanol stress is the change occur- 2.5. Rapid degradation of malic acid in a tartaric ring in the composition of cell proteins, acid buffer (pH 4.5) and in standardized wine including heat shock proteins (Jobin et al., 3. THIRD-ORDER CRITERIA 1997). Bourdineaud et al. (2003) found that the 3.1. Production of appropriate organoleptic O. oeni gene ftsH, which encodes a protease characteristics in the wine belonging to the ABC family of proteins, is 3.2. Resistance to phages responsible for this stress, since its expression increases at high temperatures and in response 3.3. Sulfite resistance to osmotic shock. O. oeni cells have also been 3.4. No formation of biogenic amines found to express the 18 kDa protein Lo18 upon 3.5. Potential to form diacetyl and acetoin exposure to various stresses and during the stationary phase (Guzzo et al., 1997), and this 3.6. Limited formation of volatile acids represents a general marker of stress in this 3.7. No degradation of glycerin bacteria. More recently, Silveira et al. (2004) 3.8. No production of extracellular polysaccharides used proteomic analysis to show evidence of an active adaptive response to ethanol both in 3.9. Little formation of D-lactic acid cytoplasmic and membrane proteins. Those Adapted from Buckenhu¨skes (1993). authors reported that ethanol induces changes in the patterns of cellular proteins expressed stress in certain lactic acid bacteria (Baliarda by O. oeni. They found a variation in the levels et al., 2003; Romeo et al., 2001) and a survival of proteins involved in the maintenance of mechanism in lactic acid bacteria subjected to redox balance, suggesting that this process desiccation (Carvalho et al., 2003). Studies plays an important role in the adaptation to have been performed to analyze the mecha- ethanol. Coucheney et al. (2005) measured nisms underlying the resistance of O. oeni to malolactic and ATPase activity along with LACTIC ACID BACTERIA 295 expression of the Lo18 protein in three strains of 3.2. Production of Biomass O. oeni selected as malolactic starter cultures. The strain of O. oeni that showed the highest The production of bacteria for malolactic malolactic activity in complete cells at pH 3 fermentation began many years after that of and the highest expression of Lo18 protein also wine yeast. The process is simpler than the had the highest rates of growth and malic acid method used for yeasts since batch fermentation consumption. As a result, Coucheney et al. is used with a smaller number of successive (2005) suggested that these techniques could stages. During this process, special care must be more rapid and reliable than the standard be taken regarding the cleanliness of the equip- techniques used for the selection of strains as ment since O. oeni grows very slowly and can be malolactic starter cultures. easily and rapidly contaminated. A procedure involving a turbidostat has also Although the technology used for the been described for the selection of strains to be commercial production of O. oeni is similar to used as starter cultures (Nielsen et al., 1996). that used for the production of starter cultures The procedure involves the use of a 1 L used in dairy products, the complex nutritional fermenter containing filter-sterilized wine with requirements of these organisms precludes the an initial ethanol concentration of 11.5% (vol/ use of conventional media. This represents the vol) and a pH of 3.4. The fermenter is inoculated main difference compared with wine yeasts, with a mixture of 30 or 40 isolates of O. oeni that which are propagated in a medium that is iden- have been precultured in de Man Rogosa Sharpe tical to that used for baker’s yeast and to which (MRS) culture medium. The medium is incu- the same principles are applied. bated at 18 C with shaking. The biomass in the Strains of O. oeni are generally stored frozen fermentation is maintained constant at an or lyophilized. Unlike in yeasts, the propagation optical density (600 nm) of 0.10, measured in of O. oeni involves a batch process throughout a spectrophotometer that controls the addition the sequence of fermentations. The first fermen- of sterile wine enriched with yeast extract and tations are carried out in small vessels over 3 to containing increasing concentrations of ethanol 5 d at temperatures that vary between 20 and and decreasing pH. After 4 to 6 weeks of culture, 25 C, depending on the strain. Since O. oeni the pH and ethanol concentration in the fer- grows very slowly, only a small number of menter reach values that prevent bacterial scale-ups are used due to the risk of contamina- growth. At this point, a sample is taken and tion. The final propagation is carried out in cultured in MRS medium, and representative conditions of reduced oxygen and in vessels isolates are then obtained. This method allows ranging in volume from 500 to 2000 L. The entire large numbers of isolates to be subjected to the process requires 20 to 30 d. Samples are taken at desired selection conditions at the same time. varying intervals for the purposes of quality The gradual increase in selection pressure in control and monitoring. the turbidostat allows selection of strains that During the preparation of the malolactic are better adapted to growth in wine with starter culture, conditions should be chosen to a low pH and a high ethanol concentration. produce the highest quantity of biomass with Following selection of strains in this way, the highest viability and malolactic activity. studies are performed to assess their capacity The inoculation of wines or musts with an to maintain their adaptation to the adverse optimal number of cells to perform malolactic conditions found in the wine during the subse- fermentation requires prior production of large quent processes, including concentration, quantities of biomass. The growth rate, the freezing, and lyophilization of the bacteria. biomass produced, and the ability to perform 296 11. PRODUCTION OF WINE STARTER CULTURES malolactic fermentation can be optimized greater malolactic activity following inoculation through the use of an appropriate combination into wine. If the cells are collected sooner, they of medium and culture conditions. Studies die quickly and cannot induce malolactic have been performed to assess the influence of fermentation (Krieger et al., 1993). Nevertheless, different culture variables such as the composi- the studies of Kole et al. (1982) addressing pilot- tion of the medium, the concentration of malic scale production of O. oeni showed that centri- acid, the pH, and the temperature on the fuged cultures obtained during the middle of biomass produced and the malolactic activity the log phase displayed greater viability of cultures of O. oeni and some strains of Lacto- following lyophilization. In addition, these bacillus (Champagne et al., 1989; Naomi et al., lyophilized cultures exhibited greater viability 1989). when packaged under a nitrogen atmosphere Differences in the sugar composition of the and stored in cold, dry conditions. culture medium influence the growth kinetics and production of biomass in O. oeni. Maicas 3.3. Lyophilization, Packaging, et al. (1999b) suggested the use of a combination and Storage of two sugars (glucose and fructose) in the culture medium for O. oeni. This co-fermenta- A major problem in the development of tion allows the production of up to eight times malolactic starter cultures has been the sensi- more biomass than fermentation with a single tivity of the bacterial cells to damage occurring sugar (glucose). Controlling pH, which prevents during lyophilization. The aim is to dehydrate acidification of the culture medium, also leads the bacteria in order to maintain a high level to a 38% increase in the production of biomass. of cell viability. The culture should remain stable MLO medium (medium for Leuconostoc oenos) during storage periods of various months so is the most appropriate for the easy and rapid that it can be commercially prepared in large growth of O. oeni under controlled laboratory quantities prior to the harvest and then appro- conditions (Maicas et al., 2000). However, in priately stored in the winery until required. this medium, the cells lose their natural resis- The biomass produced is frozen almost tance to the adverse environmental conditions instantaneously and vacuum dried at low found in wine and fail as starter cultures for temperature. Careful lyophilization leads to the induction of malolactic fermentation the survival of most cells and helps to preserve (Krieger et al., 1993). Growth of the bacteria in fermentative capacity. Addition of cryoprotec- an appropriate preculture medium reduces the tive agents and careful modification of the preparation time required for the starter culture, freezing and drying process can allow a viability prepares the cells to survive following storage, of more than 95% to be obtained. The residual and allows their subsequent growth in wine. water content of the final product is around 4 The cells grown in these media are adapted to 5%. and are able to perform malolactic fermentation Zhao and Zhang (2009a, 2009b) studied the immediately following inoculation in the wine. influence of lyophilization conditions on the This also reduces the problems of contamination survival of malolactic cultures of O. oeni. They linked to the use of starter cultures. analyzed the effects of a cell-washing step, the It has been reported that the best moment to pH of the resuspension medium, preincubation collect the biomass produced by O. oeni is 18 with sodium glutamate, initial cell concentra- to 24 h after the culture enters the stationary tion, and lyophilization temperature (Zhao & phase. Centrifuged cells obtained at this point Zhang, 2009a). The cell viability in samples in time have a higher rate of survival and that were not washed in potassium phosphate LACTIC ACID BACTERIA 297 buffer was significantly lower than that that strains recently isolated from wine display observed in washed samples. Survival was greater viability and malolactic activity than maximal when the pH of the resuspension strains that have been maintained for some medium was 7. Cell viability was also increased time in culture collections and subcultured when cells were preincubated at 25C prior to successively in synthetic media. freezing. When 2.5% sodium glutamate was The variables measured to assess the quality used as a protective agent in the suspension of the end product are the concentration of medium, the optimal initial cell concentration viable cells, the capacity to degrade malic acid, was 109 colony-forming units (CFU) per mL. and the microbiological quality. Molecular bio- Cell viability increased by 21.6% when the logical techniques are used for the identification thawing temperature was reduced from 20C of the strains. Commercial preparations gener- to 65C. However, survival was markedly ally contain fewer than 103 contaminating reduced in cells frozen in liquid nitrogen bacteria per gram. Quality control is also per- (196C). Zhao and Zhang (2009b) also demon- formed in the marketed product to confirm strated that the survival of O. oeni following that it has been appropriately stored. lyophilization depends on the protective medium, the rehydration medium, and the 3.4. Use storage medium used, making it important to choose these elements carefully in order to Direct inoculation of rehydrated cultures into obtain maximum cell viability. The addition of wine leads to significant bacterial cell death that polysaccharides and disaccharides to the sus- can reduce the cell population from 107 to pension medium significantly increases cell 104 CFU/mL. To compensate for this loss, the viability. Rehydration in the disaccharide solu- cell density of the inoculum would need to be tions tested, however, led to a significant reduc- increased 100-fold. However, such quantities tion in cell viability. Viability was reduced after 6 are not economically viable and the lyophilized months of storage at 4C; the loss of viability was bacteria must therefore be reactivated. The reac- dependent upon the protective agents used, tivation medium generally used is grape must sodium glutamate being the most effective. without sulfite diluted 50% with water, with At the end of the lyophilization process, the a pH adjusted to 4.0 to 4.5 (with calcium biomass is reduced to a powder that is packaged carbonate) and containing 3e5 g/L of yeast under sterile conditions in gas- and vapor- extract. King and Beelman (1986) demonstrated impermeable polylaminated aluminum to the importance of using diluted grape must prevent contact between the lyophilized rather than undiluted must for the culture of bacteria and oxygen or additional moisture. lactic acid bacteria. Cavazza et al. (1999) demon- Exposure to high temperatures for prolonged strated that, during reactivation of stored periods can also kill the cells. cultures in diluted grape must, the time Each batch of lyophilized bacteria must required for malolactic fermentation is consi- undergo rigorous quality control; this process derably reduced. They also showed that the applies not only to the final product but also presence of ethanol at a concentration below to each phase of production. Quality control 2% (vol/vol) acted as a stimulant for the propa- begins with maintaining a stock of the strain in gation of starter cultures. Hayman and Monk liquid nitrogen or in lyophilized form to ensure (1982) reported that the best results were that each industrial production process begins obtained with addition of one volume of sterile with a pure strain that maintains all of the orig- wine (without sulfite) to five volumes of must. inal characteristics. It has also been reported Consequently, addition of wine during 298 11. PRODUCTION OF WINE STARTER CULTURES propagation increases the acclimatization both wine), and it should be maintained at an to alcohol and to the pH of wine, and therefore optimal temperature of 23C. increases the number of viable lactic acid The use of commercial malolactic starter bacteria during the final phase of addition to cultures is dependent on the effectiveness of the wine. In addition, various supplement the technology used for storage in order to guar- mixtures are commercially available for addi- antee a high rate of cell survival and a high tion to the media, although these are more degree of functionality during the processing frequently used during subsequent malolactic stages and following storage and rehydration. fermentation in wines with low concentrations Commercial lyophilized malolactic starter of nutrients (Pilatte & Nygaard, 1999). cultures generally contain dead cells, undam- The first malolactic starter culture for direct aged cells, and live damaged cells. It is thus inoculation into wines was introduced in 1993. important to monitor the proportions of these Since then, cultures have been described that cells during treatments prior to storage (freezing are specifically adapted to the analytical and and lyophilization), after rehydration, and organoleptic characteristics of red, white, and during initial establishment of the cultures. To rose´ wines. The capacity to survive following this end, it has recently been reported that quan- direct inoculation in wine and the maintenance titative data can be obtained in real time (no of this capacity when strains are prepared as more than 1 h) using flow cytometry to deter- lyophilized cultures are of major practical mine the numbers of metabolically active and importance in winemaking. Consequently, dead cells in a sample (Quiro´s et al., 2009). lyophilized preparations of selected strains of O. oeni that display 100% survival following Acknowledgments direct inoculation and that induce malolactic fermentation reliably and rapidly under stan- We thank the Spanish Ministry of Science and Innovation dard vinification conditions are particularly (MICINN) (grants AGL2006-02558, AGL2008-01052, AGL2009-07894 and Consolider INGENIO 2010 CSD2007- useful. Both the immediate survival and the 00063 FUN-C-FOOD) and the Comunidad de Madrid lag period of the inoculated bacteria are critical (CAM) (S2009/AGR-1469) for financial support. factors since they determine the total duration of malolactic fermentation. These commercial References lyophilized preparations thus demonstrate that it is possible to produce malolactic starter Aiba, S., Nagai, S., & Nishizawa, Y. (1976). Fed batch culture cultures that do not require the usual reactiva- of Saccharomyces cerevisiae: A perspective of computer control to enhance the productivity in baker’s yeast tion or preadaptation steps prior to use. This cultivation. Biotechnol. Bioeng., 18, 1001e1016. eliminates the risk of contamination and Alegrıá, E.-G., Lo´pez, I., Ruiz, J. I., Saénz, J., Fernańdez, E., reduces the time required during vinification. Zarazaga, M., et al. (2004). High tolerance of wild These cultures are easy to use. The lyophilized Lactobacillus plantarum and Oenococcus oeni strains to product is simply added directly to the wine lyophilisation and stress environmental conditions of acid pH and ethanol. FEMS Microbiol. Lett., 230,53e61. following alcoholic fermentation. This can be Baliarda, A., Robert, H., Jebbar, M., Blanco, C., done during racking or, alternatively, with Deschamps, A., & le Marrec, C. (2003). Potential osmo- a pump or by recirculation. Certain optimal protectants for the lactic acid bacteria Pediococcus pen- conditions in the wine to be inoculated have tosaceus and Tetragenococcus halophila. Int. J. Food been established for the use of these cultures: Microbiol., 84,13e20. Bourdineaud, J.-P., Nehme´, B., Tesse, S., & Lonvaud- it should contain no free SO2, it should not Funel, A. (2003). The gene of the wine bacterium contain added sulfites (with a maximum SO ftsH 2 Oenococcus oeni is involved in protection against envi- of 40 mg/L in red wine and 30 mg/L in white ronmental stress. Appl. Environ. Microbiol., 69, 2511e2520. REFERENCES 299

Buckenhu¨skes, H. J. (1993). Selection criteria for lactic acid François, J., & Parrou, J. L. (2001). Reserve carbohydrates bacteria to be used as starter cultures for various food metabolism in the yeast Saccharomyces cerevisiae. FEMS commodities. FEMS Microbiol. Rev., 12, 253e272. Microbiol. Rev., 25, 125e145. Carrasco, P., Querol, A., & del Olmo, M. (2001). Analysis of Gibson, B. R., Lawrence, S. J., Boulton, C. A., Box, W. G., the stress resistance of comercial wine yeast strains. Graham, N. S., Linforth, R. S. T., et al. (2008). The Arch. Microbiol., 175, 450e457. oxidative stress response along a lager brewing yeast Carvalho, A. S., Silva, J., Ho, P., Teixeira, P., Malcata, F. X., & strain during industrial propagation and fermentation. Gibbs, P. (2002). Survival of freeze-dried Lactobacillus FEMS Yeast Res., 8, 574e584. plantarum and Lactobacillus rhamnosus during storage in Gindreau, E., Keim, H., de Revel, G., Bertrand, A., & the presence of protectans. Biotechnol. Lett., 24, Lonvaud-Funel, A. (2003). Use of direct inoculation 1587e1591. malolactic starters: Setting efficiency and sensorial Carvalho, A. S., Silva, J., Ho, P., Teixeira, P., Malcata, F. X., & impact. J. Int. Sci. Vig. Vin, 37,51e57. Gibbs, P. (2003). Effect of various growth media upon Gockowick, H., & Henschke, P. A. (2003). Interaction of pH, survival during storage of freeze-dried Enterococcus faecalis ethanol concentration and wine matrix on induction of and Enterococcus durans. J. Appl. Microbiol., 94, 947e952. malolactic fermentation with commercial “direct inocu- Castelli, T. (1955). Biologia e vino. Humus, 10,1e5. lation” starter culture. Austral. J. Grape Wine Res., 9, Castelli, T., & In˜igo, B. (1957). Los agentes de la fermenta- 200e209. ciońvıńica en la regioń manchega y limı´trofes. Ann. Fac. Granot, D., & Snyder, M. (1991). Glucose induces cAMP- Agr. Univ. Perugia, 13, 17. independent growth-related changes in stationary-phase Castelli, T., & In˜igo, B. (1958). Los agentes de la fermentacioń cells of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. vıńica de la Rioja. Ann. Fac. Agr. Univ. Perugia, 14, 15. USA, 88, 5724e5728. Cavazza, A., Vavassori, G. L., & Volonterio, G. (1999). Lactic Gray, J. V., Petsko, G. A., Johnston, G. C., Ringe, D., acid bacteria starter cultures in oenology: Direct-vat Singer, R. A., & Werner-Washburne, M. (2004). “Sleeping inoculation or reactivation? Rivista di Viticoltura e di Beauty”: Quiescence in Saccharomyces cerevisiae. Enologia, 52,41e52. Microbiol. Mol Biol. Rev., 68, 187e206. Champagne, C., Gardner, N., & Doyon, G. (1989). Produc- Guerzoni, M. E., Sinigaglia, M., Gardini, F., Ferruzzi, M., & tion of Leuconostoc oenos biomass under pH control. Appl. Torriani, S. (1995). Effects of pH, temperature, ethanol, Environ. Microbiol., 55, 2488e2492. and malate concentration on Lactobacillus plantarum and Cooney, C. L., Wang, H. Y., & Wang, D. I. (1977). Computer- Leuconostoc oenos: Modelling of the malolactic activity. aided material balancing for prediction of fermentation Am. J. Enol. Vitic., 46, 368e374. parameters. Biotechnol. Bioeng., 19,55e67. Guzzo, J., Delmas, J. F., Pierre, F., Jobin, M.-P., Samyn, B., Coucheney, F., Desroche, N., Bou, M., Tourdot-Marećhal, R., van Beeumen, J., et al. (1997). A small heat shock protein Dulau, L., & Guzzo, J. (2005). A new approach for from Leuconostoc oenos induced by multiple stresses and selection of Oenococcus oeni strains in order to produce during stationary growth phase. Lett. Appl. Microbiol., 24, malolactic starters. Int. J. Food Microbiol., 105, 463e470. 393e396. Coulon, J., Husnik, J. I., Inglis, D. L., van der Merwe, G. K., Guzzo, J., Jobin, M. P., Delmas, F., Fortier, L. C., Garmyn, D., Lonvaud, A., Erasmus, D. J., et al. (2006). Metabolic engi- Tourdot-Marechal, R., et al. (2000). Regulation of stress neering of Saccharomyces cerevisiae to minimize the response in Oenococcus oeni as a function of environ- production of ethyl carbamate in wine. Am. J. Enol. Vitic., mental changes and growth phase. Int. J. Food Microbiol., 57,113e124. 55,27e31. Craig, E. A. (1985). The heat shock response. CRC Crit. Rev. Hammes, W. P. (1990). Bacterial starter cultures in food Biochem., 18, 239e280. production. Food Biotechnol., 4, 383e397. Delfini, C., & Cervetti, F. (1991). Metabolic and technological Hayman, D. C., & Monk, P. R. (1982). Starter culture prep- factors affecting acetic acid production by yeasts during aration for the induction of malolactic fermentations in alcoholic fermentation. Vitic. Enol. Sci., 46, 142e150. wine. Food Tech. Aust., 34,14e18. Delmas, F., Divies, C., & Guzzo, J. (2000). Biochemical and Heard, G. M., & Fleet, G. H. (1987). Occurrence and growth physiological studies of a small heat shock protein from of killer yeasts during wine fermentation. Appl. Environ. a lactic acid bacterium. Sci. Aliments., 20,111e117. Microbiol., 53, 2171e2174. Felix, C. F., Moreira, C. C., Oliveira, M. S., Sola-Penna, M., In˜igo, B. (1958). Vinificacioń del mosto de uva con asocia- Meyer-Fernandes, J. F., Scofano, H. M., et al. (1999). cioń escalar de levaduras seleccionadas. Revista de Protection against thermal denaturation by trehalose on Ciencia Aplicada, 63, 318e324. þ the plasma membrane H -ATPase from yeast. Eur. J. Ivorra, C., Pe´rez-Ortıń, J. E., & del Olmo, M. (1999). An Biochem., 266, 660e664. inverse correlation between stress resistance and stuck 300 11. PRODUCTION OF WINE STARTER CULTURES

fermentations in wine yeasts. A molecular study. Bio- Manginot, C., Roustan, J. L., & Sablayrolles, J. M. (1998). technol. Bioeng., 64, 698e708. Nitrogen demand of different yeast strains during Jobin, M. P., Delmas, F., Garmyn, D., Divies, C., & Guzzo, J. alcoholic fermentation. Importance of the stationary (1997). Molecular characterization of the gene encoding phase. Enz. Microb. Technol., 23,511e517. an 18-kilodalton small heat shock protein associated Marks, V. D., Sui, S. J. H., Erasmus, D., van der with the membrane of Leuconostoc oenos. Appl. Environ. Merwe, G. K., Brumm, J., Wasserman, W. W., et al. Microbiol., 63, 609e612. (2008). Dynamics of the yeast transcriptome during wine Jobin, M. P., Garmyn, D., Divies, C., & Guzzo, J. (1999). fermentation reveals a novel fermentation stress Expression of the Oenococcus oeni trxA gene is induced response. FEMS Yeast Res., 8,35e52. by hydrogen peroxide and heat shock. Microbiology, 145, Molina-Gutierrez, A., Stippl, V., Delgado, A., Ganzle, M. G., 1245e1251. & Vogel, R. F. (2002). In situ determination of the intra- Jobin, M. P., Garmyn, D., Divies, C., & Guzzo, J. (1999). The cellular pH of Lactococcus lactis and Lactobacillus planta- Oenococcus oeni clpX homologue is a heat shock gene rum during pressure treatment. Appl. Environ. Microbiol., preferentially expressed in exponential growth phase. 68, 4399e4406. J. Bacteriol., 181, 6634e6641. Naomi, P., Chagnaud, P., Arnaud, A., Galzy, P., & King, S. W., & Beelman, R. B. (1986). Metabolic interactions Mathleu, J. (1989). Optimization of the conditions for between Saccharomyces cerevisiae and Leuconostoc oenos in preparing bacterial cultures for malolactic bioconver- a model grape juice/wine system. Am. J. Enol. Vitic., 37, sion. J. Biotechnol., 10, 135e150. 53e60. Nielsen, J. C., Prahl, C., & Lonvaud-Funel, A. (1996). Kitamoto, K., Oda, K., Gomi, K., & Takahashi, K. (1991). Malolactic fermentation in wine by direct inoculation Genetic engineering of a sake yeast producing no urea with freeze-dried Leuconostoc oenos cultures. Am. J. Enol. by successive disruption of arginase gene. Appl. Environ. Vitic., 47,42e48. Microbiol., 57, 301e306. Novo, M. T., Beltran, G., Torija, M. J., Poblet, M., Rozes, N., Kole, M., Altosaar, I., & Duck, P. (1982). Pilot scale Guillamon, J. M., et al. (2003). Changes in wine yeast production and preservation of a new malolactic culture storage carbohydrate levels during preadaptation, Leuconostoc oenos 44.40 for use in secondary wine rehydration and low temperature fermentations. Int. J. fermentation. Biotechnol. Lett., 4, 695e700. Food. Microbiol., 86, 153e161. Kraus, J. K., Reed, G., & Villettaz, J. C. (1983). Levures Novo, M. T., Beltran, G., Torija, M. J., Poblet, M., Rozes, N., se`ches actives de vinification. 1ere partie: Fabrication et Guillamon, J. M., & Mas, A. (2007). Early transcrip- caracte´ristiques. Conn. Vig. Vin, 17,93e103. tional response of wine yeast after rehydration: Krieger, S. A., Hammes, W. P., & Henick-Kling, T. (1993). Osmotic shock and metabolic activation. FEMS Yeast How to use malolactic starter cultures in the winery. Res., 7,304e316. Austral. New Zeal. Wine Indust. J., 8, 153e160. Ouvry, A., Wache, Y., Tourdot-Marechal, R., Divies, C., & Leslie, S. B., Teter, S. A., Crowe, L. M., & Crowe, J. H. (1994). Cachon, R. (2002). Effects of oxidoreduction potential Trehalose lowers membrane phase transitions in dry combined with acetic acid, NaCl and temperature on yeast cells. Biochim. Biophys. Acta, 1192,7e13. the growth, acidification and membrane properties of Lillie, S. H., & Pringle, J. R. (1980). Reserve carbohydrate Lactobacillus plantarum. FEMS Microbiol. Lett., 214, 257e261. metabolism in Saccharomyces cerevisiae: Responses to Panoff, J. M., Thammavongs, B., & Gueguen, M. (2000). nutrient limitation. J. Bacteriol., 143, 1384e1394. Cryoprotectants and cold stress in lactic acid bacteria. Magliani, W., Conti, S., Gerloni, M., Bertolotti, D., & Sci. Aliments, 20, 105e110. Polonelli, L. (1997). Yeast killer systems. Clin. Microb. Papin, J. P. (1988). Les levures se`ches en oenologie. Fabri- Rev., 10, 369e400. cation et controˆles de qualite´.InApplication a` l’oenologie Maicas, S., Gil, J. V., Pardo, I., & Ferrer, S. (1999a). Improve- des progres recents en microbiologie e en fermentation ment of volatile composition of wines by controlled (pp. 331e371). Montpellier, France: OIV-ENSA. addition of malolactic bacteria. Food Res. Int., 32,491e496. Pe´rez-Ortıń, J. E., Querol, A., Puig, S., & Barrio, E. (2002). Maicas, S., Gonza´lez-Cabo, P., Ferrer, S., & Pardo, I. (1999b). Molecular characterization of a chromosomal rear- Production of Oenococcus oeni biomass to induce malo- rangement envolved in the adaptive evolution of yeast lactic fermentation in wine by control of pH and strains. Genome Res., 12, 1533e1539. substrate addition. Biotechnol. Lett., 21, 349e353. Pe´rez-Torrado, R., Bruno-Ba´rcena, J. M., & Matallana, E. Maicas, S., Pardo, I., & Ferrer, S. (2000). The effects of (2005). Monitoring stress-related genes during the freezing and freeze-drying of Oenococcus oeni upon process of biomasa propagation of Saccharomyces cer- induction of malolactic fermentation in red wine. Int. J. evisiae strains used for wine-making. Appl. Environ. Food Sci. Technol., 35,75e79. Microbiol., 71, 6831e6837. REFERENCES 301

Pe´rez-Torrado, R., Go´mez-Pastor, R., Larson, C., & Smelt, J. P. P. M., Otten, G. D., & Bos, A. P. (2002). Modelling Matallana, E. (2009). Fermentative capacity of dry active the effect of sublethal injury on the distribution of the lag wine yeast requires a specific oxidative stress response times of individual cells of Lactobacillus plantarum. Int. J. during industrial biomasa growth. Appl. Microbiol. Bio- Food. Microbiol., 73, 207e212. technol., 81, 951e960. Spiropoulos, A., & Bisson, L. F. (2000). MET17 and hydrogen Pilatte, E., & Nygaard, M. (1999). Development of a specific sulfide formation in Saccharomyces cerevisiae. Appl. activator for malolactic bacteria. Rev. Oenol. Techniq. Environ. Microbiol., 66, 4421e4426. Vitivin. Oenol., 92,31e33. Tezuka, H., Mori, T., Okumura, Y., Kitabatake, K., & Potma, E., Verduyn, C., Scheffers, W. A., & van Dijken, J. P. Tsumura, Y. (1992). Cloning of a gene suppressing (1989). Enzymatic analysis of the crabtree effect in hydrogen sulfide production by Saccharomyces cerevisiae glucose-limited chemostat cultures of Saccharomyces cer- and its expression in a brewing yeast. J. Am. Soc. Brew. evisiae. Appl. Environ. Microbiol., 55, 468e477. Chem., 50, 130e133. Pronk, J. T., Steensma, H. Y., & van Dijken, J. P. (1996). Thevelein, J. M. (1994). Regulation of trehalose mobilization Pyruvate metabolism in Saccharomyces cerevisiae. Yeast, in fungi. Microbiol. Rev., 48,42e59. 12, 1607e1633. Toledano, M. B., Kumar, C., Le, M. N., Spector, D., & Quiro´s, C., Herrero, M., Garcıá, L. A., & Dıáz, M. (2009). Tacnet, F. (2007). The system biology of thiol redox Taking advantage of the flow cytometry technique for system in Escherichia coli and yeast: Differential functions improving malolactic starters production. Eur. Food Res. in oxidative stress, iron metabolism and DNA synthesis. Technol., 228(4), 543e552. FEBS Lett., 581, 3598e3607. Ratnakumar, S., & Tunnacliffe, A. (2006). Intracellular Tourdot-Marechal, R., Gaboriau, D., Beney, L., & Divies, C. trehalose is neither necessary nor sufficient for dessica- (2000). Membrane fluidity of stressed cells of Oenococcus tion tolerante in yeast. FEMS Yeast Res., 6, 902e913. oeni is involved in protection against environmental Rodrigues Pousada, C. A., Nevitt, T., Menezes, R., stress. Appl. Envirom. Microbiol., 69, 2512e2520. Azevedo, D., Pereira, J., & Amaral, C. (2004). Yeast acti- Tunnacliffe, A., & Wise, M. J. (2007). The continuing vator proteins and stress response: An overview. FEBS conundrum of the LEA proteins. Naturwissenschaften, 94 Lett., 567,80e85. (10), 791e812. Rodriguez-Porrata, B., Novo, M., Guillamoń, J., Roze´s, N., van de Guchte, M., Serror, P., Chervaux, C., Smokvina, T., Mas, A., & Cordero, R. (2008). Vitality enhancement of Ehrlich, S. D., & Magıń, E. (2002). Stress responses in the rehydrated active dry wine yeast. Int. J. Food Micro- lactic acid bacteria. Anton. Leeuw., 82, 187e216. biol., 126,116e122. Vaudano, E., Constantini, A., Cercosimo, M., del Prete, V., & Romano, P., & Suzzi, G. (1993). Sulfur dioxide and wine Garcıá-Moruno, E. (2009). Application of real-time microorganisms. In G. H. Fleet (Ed.), Wine microbiology RT-PCR to study gene expression in active dry yeast and biotechnology (pp. 373e393). Chur, Switzerland: during the rehydration phase. Int. J. Food Microbiol., 129, Harwood Academic Publishers. 30e36. Romeo, Y., Bouvier, J., & Gutierrez, C. (2001). Osmotic stress Wang, H. Y., Cooney, C. L., & Wang, D. I. (1977). Computer- response of lactic acid bacteria Lactococccus lactis and aided baker’s yeast fermentations. Biotechnol. Bioeng, 19, Lactobacillus plantarum. Lait, 81,49e55. 69e86. Rossignol, T., Postaire, O., Storaı¨, J., & Blondin, B. (2006). Werner-Washburne, M., Braun, E., Johnston, G. C., & Analysis of the genome response of a wine yeast to Singer, R. A. (1993). Stationary phase in the yeast rehydratation and inoculation. Appl. Microbiol. Bio- Saccharomyces cerevisiae. Microbiol. Rev., 57, 383e401. technol., 71, 699e712. Yuasa, N., Nakagawa, Y., Hayakawa, M., & Limura, Y. Sa ńchez, C. (1988). Levures se`ches actives. Fabrication et (2004). Distribution of the sulfite resistance gene SSU1-R controˆles de qualite´.InApplication a` l’oenologie des progres and the variation in its promoter region in wine yeasts. recents en microbiologie e en fermentation (pp. 315e330). J. Biosci. Bioeng., 98, 394e397. Montpellier, France: OIV-ENSA. Zhao, G., & Zhang, G. (2009). Influence of freeze-drying Silveira, M. G., Baumga¨rtner, M., Rombouts, F. M., & conditions on survival of Oenococcus oeni for Abee, T. (2004). Effect of adaptation to ethanol on cyto- malolactic fermentation. Int. J. Food Microbiol., 135, plasmic and membrane protein profiles of Oenococcus 64e67. oeni. Appl. Environ. Microbiol., 70, 2748e2755. Zhao, G., & Zhang, G. (2009). Influences of protectants, Singh, J., Kumar, D., Ramakrishnan, N., Singhal, V., Jervis, J., rehydration media and storage on the viability of freeze- Garst, J. F., et al. (2005). Transcriptional response of dried Oenococcus oeni for malolactic fermentation. World Saccharomyces cerevisiae to desiccation and rehydratation. J. Microbiol. Biotechnol. doi: 10.1007/s11274e009e Appl. Environ. Microbiol., 71, 8752e8763. 0080e9. 302 11. PRODUCTION OF WINE STARTER CULTURES

Zuzuarregui, A., Carrasco, P., Palacios, A., Julien, A., & del suitable criterion for wine yeast selection. Anton. Leeuw., Olmo, M. (2005). Analysis of the expression of some stress 85, 271e280. induced genes in several commercial wine yeast strains at Zuzuarregui, A., & del Olmo, M. (2004b). Expression the beginning of viniﬁcation. J. Appl. Microbiol., 98, 299e307. of stress response genes in wine strains with Zuzuarregui, A., & del Olmo, M. (2004a). Analyses of stress different fermentative behavior. FEMS Yeast Res., 4, resistance under laboratory conditions constitute a 699e710. CHAPTER 12

Preservation of Microbial Strains in the Wine Industry Marıá Dolores GarcıáLo´pez, Jose´ M. Lo´pez-Coronado, Laura Lo´pez-Ocanã, Federico Uruburu Fernańdez Coleccioń Espanõla de Cultivos Tipo (CECT), Universitat de Vale`ncia, Valencia, Spain

OUTLINE

1. Introduction 304 3.1. Long-term Preservation 309 3.1.1. Preservation Protocols 310 2. Methods for the Preservation of 3.2. Short-term Preservation 311 Microbial Strains 304 2.1. Long-term Preservation: The Preferred 4. Preservation of Bacteria 311 Approach 304 4.1. Long-term Preservation 311 2.1.1. Freezing 304 4.1.1. Guidelines for the Preservation 2.1.2. Lyophilization 306 of Lactic Acid Bacteria 311 2.2. Short-term, Alternative Preservation 4.1.2. Guidelines for the Preservation Methods 307 of Acetic Acid Bacteria 314 2.2.1. Preservation by Periodic Transfer 307 4.1.3. Protocols for Long-term 2.2.2. Preservation by Suspension in Preservation 314 Sterile Distilled Water or 4.1.4. Recovery of the Preserved Cells 315 Seawater 307 4.2. Short-term Preservation 315 2.3. Other Methods of Preservation 307 4.2.1. Maintenance by Periodic 2.3.1. Drying on Filter Paper 308 Transfer 315 2.3.2. Drying in Earth, Sand, Silica 5. Preservation of Filamentous Fungi Gel, etc 308 From Wine 316 2.3.3. Drying on Alginate Beads 308 5.1. Methods for the Preservation of 2.4. Recovery 308 Filamentous Fungi 316 3. Preservation of Yeasts in the Wine 5.1.1. Freezing 316 Industry 308 5.1.2. Lyophilization 316

5.1.3. Preservation by Subculture or Appendix 1 318 Periodic Transfer 317 Culture Media 318 5.1.4. Preservation in Sterile Distilled Cryoprotectants 318 Water 317

1. INTRODUCTION In this chapter, we will provide a general description of the most commonly used methods for the Three objectives must be attained to ensure preservation of microbial strains. This will be good preservation of microbial strains in micro- divided into three sections, in each case discus- biology laboratories: the culture must be pure sing the advantages and disadvantages of the (without contamination during the preservation different approaches. We will then go on to process), at least 70 to 80% of the cells must discuss the specific considerations applicable to survive storage, and the cells must remain the preservation of the yeasts, lactic acid bacteria, genetically stable. The first two objectives are acetic acid bacteria, and filamentous fungi that not particularly difficult to achieve when good are relevant to the wine industry. Further infor- microbiological technique is used. The third, mation can be found in Day and Stacey (2007), however, can present difficulties. Consequently, Hatt (1980), Hill (1981), Hunter-Cevera and Belt various methods have been developed for the (1996), Kirsop (1980), and Kirsop and Doyle preservation of microorganisms and no single (1991). method is applicable to all situations. 2.1. Long-term Preservation: The 2. METHODS FOR THE Preferred Approach PRESERVATION OF Long-term preservation methods are consid- MICROBIAL STRAINS ered the most appropriate option wherever possible, since they involve stopping the growth Many different microorganisms influence the of the microbial cells and keeping them in winemaking process. At the earliest stages, vines a viable state. This guarantees maximum can be damaged by the growth of phytopatho- genetic stability by preventing the appearance genic fungi (see Chapter 10). Later, in the winery, of successive generations. Nevertheless, the yeasts and lactic acid bacteria will transform the possibility that the preparation method itself grape must into wine (see Chapters 2, 3, 4, 5, leads to changes cannot be ruled out. There and 9) or have a negative influence on the wine- are two preservation methods belonging to making process (see Chapters 5, 9, and 10), some- this group: freezing and lyophilization. times even growing in unexpected places such as the corks used to seal the bottles (A´ lvarez-Rodrı´- 2.1.1. Freezing guez et al., 2002). Analysis and monitoring of In the first long-term preservation method, these three microbial groups inevitably requires the cells are frozen suspended in a liquid appropriate preservation of pure cultures. Such medium containing a cryoprotective agent and preservation is perhaps even more important stored at temperatures below 0C. As a result, when commercial starter cultures must be stored intracellular and extracellular water is main- in the winery so as not to have to request them tained in a solid state. The reduced cell metabo- from the supplier every time they are needed. lism caused by the low temperature and the METHODS FOR THE PRESERVATION OF MICROBIAL STRAINS 305 absence of liquid water prevents growth. carbohydrates such as glucose, lactose, Cells preserved in this way are recovered by sucrose, and inositol can also be used. The increasing the temperature prior to use. This is choice of cryoprotective agent is influenced the best method of preservation in almost all by the type of microorganism to be respects, although it has the drawback of preserved. requiring special apparatus and carries with it 4) Storage temperature. Storage temperature the risk that system failure could result in an should be as low as possible in order to unintentional increase in temperature during prevent intracellular recrystallization of storage. It is also the most inconvenient method water, which occurs at temperatures between for the distribution of strains. 0 and approximately 130C. It is best to Four factors influence the viability and store the microbial cells in sealed tubes stability of cells preserved in this way: submerged in liquid nitrogen (195C) or in liquid nitrogen vapor (140C). 1) Age of the cells. In most cases, it is best to use mature cells from the beginning of the Although various types of freezer are available stationary phase of the growth curve. on the market, the most appropriate reach However, in the case of microorganisms with temperatures below 70C. Those that only a stage in their life cycle that prepares them to reach temperatures of between 20 and 40C, resist adverse conditions, it is preferable to as is applicable to most of those currently avail- use cells from this stage. This occurs in able in microbiology laboratories, are less recom- microorganisms that sporulate, in some mendable. Among other reasons, this is because pleomorphic microorganisms, and even in the high concentration of solutes in the cell some simpler microorganisms. suspensions reduces their freezing point and 2) Rate of freezing and thawing. Although there cell damage occurs as a result of the frequent are standardized freezing protocols for use freezing and thawing that occurs under these in certain contexts, it is generally best for conditions. Use of a nonionic cryoprotectant changes in temperature to be rapid, both such as glycerol reduces the quantity of ice during freezing and thawing, in order to produced and helps to prevent increases in the minimize the formation of ice crystals. It is ionic concentration. For preservation in freezers, normally appropriate, therefore, to thaw cells the cells are stored in cryotubes (sterilizable at 37C. plastic tubes that are resistant to freezing and 3) Use of cryoprotective agents. Cryoprotectants can be hermetically sealed). Batches of tubes are are substances that protect against the prepared for each strain and then a single tube damage that can occur in microbial cells is used completely each time the culture is during freezing, mainly by favoring the required. This avoids the repeated freezing and vitrification of extracellular water rather than thawing of the strains. The use of cryoballs is its crystallization, which causes cell damage now widespread. However, they have a number and loss of viability of the preserved culture. of disadvantages that make them inappropriate Cryoprotective agents can also stabilize large for use in preserving cells in optimal conditions intracellular molecules but only if the over long periods of time. The main disadvantage cryoprotectant can cross the cell membrane. of this method is that the cells are maintained in Although many compounds can be used as an extremely thin layer on the surface of the cryo- cryoprotectants, the most commonly used is balls and in the absence of cryoprotective agents. glycerol at a concentration of between 15 and This thin layer of cells thaws very rapidly when 20%. Dimethyl sulfoxide, skimmed milk, and a cryoball is taken for inoculation and freezes 306 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY again when the tube is returned to the freezer. various cryoprotectants can be used according Consequently, the cells that remain in the tube to the type of microorganism. However, glycerol are subjected to frequent freezeethaw cycles, should not be used when the cells are to be and this problem is further aggravated when lyophilized due to its high evaporation point the freezers used for storage do not reach temper- and hygroscopic properties, which can lead to atures below 70C. Furthermore, since the vials highly viscous lyophilized samples. The use contain numerous cryoballs, the risk of contami- of dimethyl sulfoxide is also inappropriate nation during handling is very high. The method because it is slightly toxic and can cause damage is also particularly harmful to anaerobic microor- to the microbial cells as a result of concentration ganisms such as bacteria of the genus Clostridium, following the evaporation of water. As a result, since the cells on the surface of the cryoballs are inositol is recommended as a cryoprotective exposed to oxygen in the environment and, as agent for use during the lyophilization of most a result, their viability is reduced. bacteria and skimmed milk for use with fungi and actinomycetes. Other cryoprotectants may 2.1.2. Lyophilization be more appropriate for certain microorgan- Lyophilization is a gentle process in which isms, such as glutamate for lactic acid bacteria, water is removed from the cells to stop their mixtures of glucose and liver broth or chopped growth. The genetic stability obtained with this meat medium (without meat) for anaerobic method is high, although not always as high as bacteria, etc. that seen with freezing, since lyophilization is The factors that specifically influence the effi- achieved by sublimation of the ice in the cells. cacy of lyophilization as a means of preservation First, the free water in the cells must be frozen are as follows: and then eliminated by vacuum without increasing the temperature, as this would affect 1) Type of microorganism. Some microbes cannot the viability of the microorganism. The process tolerate lyophilization; these are logically uses an apparatus known as a lyophilizer or freeze microorganisms that contain more drier, and many different models are available on intracellular water. Some filamentous fungi, the market. The microbial cells that are preserved particularly nonsporulating strains, cannot in this way are subjected to a more complex treat- be stored as lyophilized preparations and ment, since freezing is followed by sublimation of other methods must therefore be used. water. However, it is a highly recommendable 2) Cell concentration. It is best to lyophilize cell method since these lyophilized samples can be suspensions at a concentration of between maintained at room temperature, thus making 108 and 109 cells/mL in the case of bacteria storage and distribution much easier. and at slightly lower concentrations for The factors that must be taken into account in filamentous fungi and yeasts. order to achieve good lyophilization are of 3) Temperature during sublimation. Sublimation course the same as those that influence freezing should be performed at the lowest possible plus additional factors relating to the subse- temperature, always below 50C. quent dehydration of the cells. However, before 4) Extent of dehydration achieved. Dehydration discussing dehydration-related factors, we should always be as extensive as possible, should briefly consider those mentioned earlier although the concentration of solutes may in relation to freezing. Freezing can be per- lead to small traces of water that are not formed rapidly, by submerging the tubes in harmful. liquid nitrogen, or slowly, using freezers with 5) Presence of oxygen in the tube. The lyophilized or without programming. As we saw earlier, cells are stored in tubes sealed under vacuum METHODS FOR THE PRESERVATION OF MICROBIAL STRAINS 307

to prevent both rehydration and entry of produces toxic byproducts), and by storing oxygen, which can damage the cells. cultures at temperatures of between 4 and 8C. 6) Storage conditions. The storage temperature Sometimes the culture is also covered with must be constant, preferably between 4 a layer of sterile mineral oil. This helps to and 18C, and must not fall below 0C. prevent the toxic effects associated with the Lyophilized samples must be stored in the increase in the concentration of the culture dark. medium caused by evaporation. Highly aerobic microorganisms such as filamentous fungi cannot be stored in completely closed vials. 2.2. Short-term, Alternative Finally, an additional drawback of periodic Preservation Methods transfer is the increased risk of contamination due to handling of the vials over time, as well When the strain does not tolerate the treat- as the possibility of mites entering the vials. ments required for long-term preservation or the necessary equipment is unavailable for this 2.2.2. Preservation by Suspension in Sterile type of preservation, alternative methods must Distilled Water or Seawater be used. In such cases, a combination of short- Suspension in sterile distilled water or term methods should always be employeddno seawater is a widely used alternative that main- single short-term method should ever be used in tains a high percentage of viability in a number isolation. of different microorganisms, including filamentous fungi, yeasts, and some bacteria. It involves 2.2.1. Preservation by Periodic Transfer suspending an aliquot of cells from the culture to In preservation methods based on periodic be preserved in sterile water. The samples can be transfer, the microbial strain is stored as an prepared in cryotubes. In this case, the cell active culture in the culture medium in which concentration must not exceed 104e105 cells/ it was grown. However, the strain cannot be mL in the case of bacteria and yeasts. In the stored indefinitely in the same vial. Because case of nonsporulating filamentous fungi, the cells remain active, they continue to excrete suspensions can be made with small pieces of toxic metabolic byproducts that accumulate and agar containing the growing fungus. In the case lead to cell aging and death. It is therefore neces- of marine microorganisms, suspensions are sary to transfer cells to another vial containing prepared in diluted seawater. fresh culture medium. This is the worst method Studies performed by the Spanish Type in terms of genetic stability, since continued cell Culture Collection (CECT) to assess the preser- growth implies ongoing turnover of genera- vation of microorganisms using this method tions, and over time the distant descendents of have revealed high percentages of viability, the initial cells may not retain some of the orig- sometimes for periods of more than 15 years. inal characteristics. If this method is to be used, The stability of morphological and physiolog- it is advisable to delay aging and extend the ical characteristics is also good, although this periods between reinoculation. This can be has not been tested for specific characteristics achieved in various ways. For instance, by such as virulence, fermentation power, etc. reducing the size of the inoculum or reducing the concentration of some nutrients in the 2.3. Other Methods of Preservation culture medium, by using stab inoculation for facultative anaerobes (since growth in the pres- Some methods that are not widely used are ence of oxygen is more rapid and generally required in order to preserve very specific 308 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY groups of microorganisms that do not tolerate has also been used for the preservation of algae lyophilization or freezing, such as the bacterial and plant cells. genera Spirillum and Rhodospirillum. The three methods described here are based on halting 2.4. Recovery growth by eliminating the availability of water for the cells. Whatever the method used to preserve microbial strains, the cells are placed under stress 2.3.1. Drying on Filter Paper (particularly during lyophilization). Conse- In this method, a highly absorbent filter quently, the stored cells are not suitable to be paper (Whatman No. 3) is impregnated with used directly. They must first be revitalized or a concentrated solution of cells and allowed to rejuvenated by seeding in nonselective medium; dry under sterile conditions. The adsorption to that is, a medium that ensures maximum the paper favors the dispersion of the cells in growth. After this step has been performed, it is this matrix and the cells do not form imperme- possible to work with the cells and culture able films that hinder drying. Similar principles them in selective media if necessary. Likewise, apply to other substrates used for drying. It is we should remember that some microorganisms also possible to dry the cells using a procedure will tolerate a given preservation technique known as liquid drying (L-Dry) because it better than others and that special precautions involves the use of a lyophilizer without prior may be necessary for the preservation of certain freezing of the cells. The vacuum created by strains. As mentioned at the beginning of this the lyophilizer dries the cells but excessive chapter, there is no single method appropriate vacuum must be avoided to prevent rapid evap- for the preservation of all microorganisms, but oration with boiling or too great a reduction in it is not difficult to identify the most appropriate temperature, which would lead to uncontrolled method in each case. freezing of the cells. An important element in the efficacy of the preservation method used is the recovery of the 2.3.2. Drying in Earth, Sand, Silica Gel, etc preserved culture. Cultures should be thawed Cells can be added to substances such as rapidly (37C water bath), since slow thawing earth, sand, or silica gel to protect them during causes recrystallization. Lyophilized or dried drying. Spore-producing microorganisms can cultures must be rehydrated for a few minutes be preserved for extended periods using this in an appropriate liquid medium. In both cases method. the reconstituted cells must be inoculated as soon as possible into appropriate culture media. 2.3.3. Drying on Alginate Beads The use of alginate beads is an effective procedure in which the cells are placed in an 3. PRESERVATION OF YEASTS IN alginate matrix and water is eliminated by THE WINE INDUSTRY sequential immersion in increasingly hyper- tonic solutions before air drying to achieve We hardly need repeat here the role played by a 70% reduction in water content. The alginate yeasts in the production and storage of wine. beads can be stored in hermetically sealed tubes However, it is worth stressing the importance at a temperature of between 4 and 18C. They of their preservation, particularly in the case of can even be stored at 80C due to the low starter cultures. These cultures have generally water content of the cells and the protection been obtained after many years of work invested provided by the alginate support. This method in the selection and improvement of the strains. PRESERVATION OF YEASTS IN THE WINE INDUSTRY 309

It is therefore important that they maintain the loss of viability. An alternative to prevent recrys- characteristics for which they were originally tallization during storage of strains at 20Cis selected. As described earlier, the most appro- to preserve them in a final glycerol concentra- priate methods are therefore long-term preservation of 50%, which will not freeze at this temper- tion, mainly by freezing or lyophilization. In ature, thus preventing recrystallization of addition to these long-term methods, however, extracellular water. Although recrystallization we will describe the short-term preservation of of intracellular water cannot be avoided, the yeast strains, as well as approaches that present damage will occur at a slower rate. greater risks for the maintenance of phenotypic The other method of long-term preservation and genotypic stability of the isolates but are is lyophilization. This technique does not yield useful to ensure availability over the period of such good results with yeasts as it does with time that they are being used. bacteria, since the damage inside these larger and more structurally complex eukaryotic cells 3.1. Long-term Preservation reduces the viability of the yeast strains to around 5 years, depending on the strain in ques- Long-term preservation is normally used for tion. One way of increasing the usable life of cultures that can later be employed in the lyophilized samples is to store them at 5C production of the biomass necessary for fermen- instead of 18C (optimal storage temperature tations. Freezing has proved to be the most for lyophilized bacteria and small-spored fila- effective method. As mentioned, various factors mentous fungi). However, as mentioned, lyoph- can influence the stability and in particular the ilization offers certain advantages over freezing, viability of the preserved material. Storage since the infrastructure required for storage of temperature is perhaps the most important the preserved strains is less extensive and distri- factor and should be as low as possible in order bution of samples is much easier. mainly to prevent the water recrystallization Another factor to take into account regarding that can occur at temperatures above 130C the viability of the frozen or lyophilized cultures and that can result in reduced viability of the is their recovery, which is crucially important, frozen material. As would be expected, the since poor recovery can negate any advantages lower the temperature, the lower the degree of obtained through the use of a good preservation recrystallization and therefore the greater the method. As mentioned at the beginning of this length of time that strains can maintain their chapter, frozen cultures should be thawed as viability. In general, at 80C, yeasts remain rapidly as possible to prevent cell rupture and stable for many years (usually more than 5), death caused by water recrystallization. The although this, of course, varies from one strain tube containing the cells should be immersed to the next. To conserve cultures at these in a 37C water bath or warmed in the hand temperatures, 15 to 20% glycerol is added by until completely thawed. It is important not to mixing one volume of 20% glycerol with 0.5 refreeze the tube once it has been thawed, since volumes of the cell suspension obtained using this will inevitably lead to further loss of the protocol described below. When a tempera- viability in addition to that caused by the first ture of 20C is used, the method is considered round of freezing and thawing. Once inoculated to be medium-term storage, since the strains in the appropriate medium, any leftover cell only remain viable for shorter periods of time. suspension obtained from the lyophilized prep- This is mainly due to the variations in tempera- aration cannot be stored since it is not viable for ture associated with these freezers that lead to more than a few hours. In all cases, the culture frequent freezeethaw cycles with a consequent medium and temperature used to recover the 310 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY strains from the lyophilized or frozen sample 3.1.1.1. FREEZING PROTOCOL must be taken into consideration. These must 2) Adjust the number of cells to a final be the most appropriate for growth of the micro- concentration of 2e6 106 cells/mL (or the organism in question without being too nutri- desired concentration). tionally rich, since the cells may not be able 3) Add appropriate cryoprotectant (glycerol) at to assimilate nutrients to any great extent an optimal concentration (15% for storage following the period of metabolic inactivity to at 80C or 50% for storage at 20C). which they have been subjected. 4) Aliquot the suspension into cryotubes, The age of the cells used for preservation of hermetically seal them, and maintain them at the strain is another important consideration. ambient temperature for 15 to 30 min. In most cases, cells develop some natural resis- 5) Transfer the tubes to their final storage tance to adverse environmental conditions, location: 20C freezer, ultra-freezer (80 and this resistance can be exploited in order to or 145C), or liquid nitrogen (196C). achieve greater viability following preservation. Some authors have advocated reducing the These natural resistance mechanisms usually temperature at a rate of between 1 and 3C/ appear in cultures towards the end of the expo- min until 30C, followed by a reduction nential growth phase or at the beginning of the of 15 to 30C/min until the final storage stationary phase. However, some strains may temperature is reached. have produced toxic waste products by the time they reach this point on the growth curve. Once again, it is important to remember that In such cases, it would be more appropriate to cryovials should only be used once and should use younger cultures. In the case of wine yeasts, not be refrozen after thawing to prevent prob- the cells should be collected after 48 h of growth lems such as loss of viability or contamination. in an appropriate culture medium (almost always Glucose Peptone Yeast extract Agar 3.1.1.2. LYOPHILIZATION PROTOCOL [GPYA]; CECT 140 medium). 2) Resuspend the cells (after washing if grown in liquid medium) in one part 15% glucose and two parts sterile skimmed milk until 3.1.1. Preservation Protocols a homogeneous suspension is obtained that For the long-term preservation of yeasts we does not appear saturated with cells under recommend cryopreservation or lyophilization. the microscope. The preparation of the culture (step one) is Aliquot the suspension into tubes common to both, as well as to other preservation 3) (approximately 200e250 mL per tube) that methods: will later be frozen prior to lyophilization. 1) Grow the cultures until the end of the log The literature on the preservation of phase in the most appropriate medium, microorganisms contains a range of opinions depending on the nutritional requirements of on the rate at which samples should be the microorganism, and under appropriate frozen prior to lyophilization. The most incubation conditions of temperature, widely accepted approach involves slow presence or absence of oxygen, and shaking. freezing (around 1C/min) or rapid freezing If the microorganisms are only able to grow by immersion in liquid nitrogen. The most in liquid medium, they will need to be important element in the end, however, is centrifuged before the cell suspension is to obtain a frozen cell suspension that will prepared in order to eliminate the culture undergo lyophilization for a period of 16 medium and concentrate the cells. to 18 h. PRESERVATION OF BACTERIA 311

4) Once the process is finalized, the vials must Although various bacterial species can influ- be hermetically sealed and stored at 5Cin ence the winemaking process, most belong to the dark in order to achieve good the lactic acid bacteria: Lactobacillus brevis, Lactoba- preservation of the strains. cillus hilgardii, Lactobacillus mali,andLactobacillus ; , The same cell suspension obtained in step plantarum Pediococcus damnosus Pediococcus par- ,and ; two could be used for preservation of the yeast vulus Pediococcus pentosaceus Leuconostoc , and in particular . in sterile water by adding approximately mesenteroides Oenococcus oeni The acetic acid bacteria are also an important 0.5 mL of the suspension to 1.0 mL of sterile group that can appear in wine and cause the water. However, this method is not widely conversion of ethanol into acetic acid. There are used as it is less effective than freezing yet five genera that can be included in this group: involves essentially the same amount of work. Ace- , , , ,and Further information on the preservation of tobacter Acidomonas Asaia Gluconacetobacter . However, usually only some yeasts can be found in Beech and Davenport Gluconobacter species of and (1971). Acetobacter Gluconacetobacter xylinus are found in wine. Gluconobacter is isolated in grapes and must, though not in wine. 3.2. Short-term Preservation The most common method used for the short- 4.1. Long-term Preservation term preservation of yeast strains is periodic transfer. This is carried out using GPYA slant As mentioned, the long-term preservation cultures stored at 5C. These conditions slow methods associated with the greatest guarantee growth and allow the strains to be stored for of stability in important physiological character- a few weeks but longer periods of storage are istics are freezing at temperatures below 70 C not recommended. This method tends to be and lyophilization, and both methods are also used for the maintenance of working stocks. the most appropriate for preserving wine An alternative short-term preservation method strains. However, the preparation of the cells is the use of stab cultures in semisolid media. will differ between the different microbial groups (see Tables 12.1 and 12.2).

4. PRESERVATION OF BACTERIA 4.1.1. Guidelines for the Preservation of Lactic Acid Bacteria As we have seen in other chapters, the action 4.1.1.1. CULTURE MEDIA of certain bacterial species influences the aroma Lactic acid bacteria require highly complex and flavor of wine and can increase the quality culture media containing specific growth factors. of the finished product. It is important, however, MRS medium (CECT 8), specifically designed by to use selected bacterial strains that improve de Man, Rogosa, and Sharpe in 1960 for use with the organoleptic character without intro- lactic acid bacteria, supports the growth of most ducing biogenic amines or other undesirable lactic acid bacteria, although some species may compounds. When a strain with the required have additional requirements. For instance, O. metabolic characteristics has been selected, it oeni requires the calcium pantothenate present must be preserved without losing those charac- in tomato juice to stimulate its growth (CECT teristics so that it can be used reliably to prepare 85 medium). Species of the genus Pediococcus iso- starter cultures as part of a controlled produc- lated from beer or wine grow better if 40% beer tion process. or wine, respectively, are added to the medium. 312 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY

TABLE 12.1 Growth Conditions for Wine Bacteria

Culture Incubation Microorganism medium pH Temperature Aeration time Observations

Lactobacillus brevis MRS 6.2e6.5 30C Aerotolerant 24e36 h anaerobe

Lactobacillus hilgardii MRS 6.2e6.5 30C Aerotolerant 48 h Grows in 15e18% anaerobe ethanol

Lactobacillus mali MRS 6.2 30C Aerotolerant 24e36 h anaerobe

Lactobacillus plantarum MRS 6.2e6.5 30C Aerotolerant 24 h anaerobe

Leuconostoc MRS 6.2e6.5 26C Microaerophilic 24e48 h mesenteroides

Oenococcus oeni MLO 4.8e5.2 26C Microaerophilic More than 3 d Grows with wine and at high concentrations of ethanol

Pediococcus damnosus MRS 5.8 26C Anaerobe More than 48 h Grows with 40% must or wine

Pediococcus parvulus MRS 6.2 26C Anaerobe More than 48 h Acetobacter aceti MYP/GYC 5.5e6.0 26C Aerobe 24e36 h Acetobacter pasteurianus MYP 5.5 26C Aerobe 24e48 h

Gluconacetobacter MYP/GY 5.5 26C Aerobe 24e36 h xylinus

Gluconobacter MYP/GYC 5.5 26C Aerobe 24e48 h

All of the conditions shown are the most generally applicable for each species but there is always the possibility that individual strains may differ in their physiological behavior. GYC ¼ CECT 287 medium; GY ¼ CECT 217 medium; MLO ¼ CECT 85 medium; MRS ¼ CECT 8 medium; MYP ¼ CECT 10 medium.

4.1.1.2. INCUBATION CONDITIONS of the stationary phase to prevent aging or Almost all lactic acid bacteria are meso- death of the cells caused by their own philes; they are aerotolerant anaerobes, but metabolites. they can display varying degrees of oxygen sensitivity, and it is therefore advisable to 4.1.1.3. PREPARATION OF THE CELLS FOR culture them in anaerobic chambers or use PRESERVATION recently prepared or degassed medium to Cells grown in liquid medium are collected ensure the absence of dissolved oxygen. If by centrifugation and resuspended in cryopro- the culture is static, the bacteria will grow at tective solutions at a concentration of app- the bottom, where no oxygen is present, and roximately 108 cells/mL. As mentioned, these in most cases further precautions will not be cryoprotective solutions are intended to prevent necessary to achieve anaerobic conditions. the formation of ice crystals that can damage The cultures should be collected in the expo- the cells and they tend to be aqueous solutions nential phase of growth or at the beginning of small molecules (monomeric sugars such as TABLE 12.2 Long-term Preservation of Wine Bacteria

Microorganism Freezing Lyophilization

Freezing and Incubation storage Freezing time Cryoprotectant temperature Cryoprotectant temperature Storage1

Lactobacillus brevis 24e36 h Glycerol þ glutamate or milk þ glucose 80C Glutamate 196C (liquid nitrogen) 5 to 22C 24e36 h Glycerol þ glutamate or milk þ glucose 80C Glutamate 196C (liquid nitrogen) 5 to 22C Lactobacillus hilgardii BACTERIA OF PRESERVATION Lactobacillus mali 48 h Glycerol þ glutamate or milk þ glucose 80C Glutamate 196C (liquid nitrogen) 5 to 22C Lactobacillus plantarum 24 h Glycerol þ glutamate or milk þ glucose 80C Glutamate 196C (liquid nitrogen) 5 to 22C

Leuconostoc mesenteroides 24e36 h Glycerol þ glutamate or inositol 80C Glutamate 196C (liquid nitrogen) 5 to 20C or inositol

Oenococcus oeni 2e3 d Glycerol þ glutamate 80C Glutamate 196C (liquid nitrogen) 5 to 20C

Pediococcus damnosus 24e48 h Glycerol þ glutamate 80C Glutamate 196C (liquid nitrogen) 5 to 20C Pediococcus parvulus 24e48 h Glycerol þ glutamate 80C Glutamate 196C (liquid nitrogen) 5 to 20C Acetobacter aceti 24e36 h Glycerol or glycerol þ inositol 80C Inositol 196C (liquid nitrogen) 5 to 20C Acetobacter pasteurianus 24e48 h Glycerol or glycerol þ inositol 80C Inositol 196C (liquid nitrogen) 5 to 20C Gluconacetobacter xylinus 24e36 h Glycerol or glycerol þ inositol 80C Inositol 196C (liquid nitrogen) 5 to 20C Gluconobacter 16e18 h Glycerol or glycerol þ inositol 80C Inositol 196C (liquid nitrogen) 5 to 20C

1All stored in the dark in vacuum-sealed vials. Incubation time refers to the recommended age of cells to better tolerate the freezing process. 313 314 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY glucose, inositol, etc.) or complex mixtures such The most appropriate culture medium is as skimmed milk. YMA (CECT 209), and GYC medium (CECT During lyophilization, the protective effect of 287) can be used as an alternative for strains glucose and small sugars is also to prevent total that cannot use mannitol. The acid produced loss of water from the cell. The viability of the from glucose is controlled by dissolution of the cells during recovery is greater if 1 to 2% mois- calcium carbonate contained in the medium. ture has been retained during lyophilization. The optimal pH is 5.5 to 6.3. The reducing effect of glucose can also be beneficial during the preservation of anaerobic 4.1.2.2. INCUBATION CONDITIONS microorganisms. Acetic acid bacteria are mesophilic but they The most widely used cryoprotective agents cannot grow at temperatures above 35C. They during lyophilization of lactic acid bacteria are are strict aerobes and can therefore be grown glutamic acid and skimmed milk. These are inex- on the surface of solid media. The cells to be pensive and guarantee good viability, particu- preserved are collected at the end of the expo- larly in the case of glutamic acid, which has nential phase or at the beginning of the been confirmed to confer greater survival over stationary phase after an incubation period of long periods. Some investigators recommend 24 to 36 h, or even 48 h if the strain requires it. the use of other sugars (for instance, adonitol for lactobacilli) but, although they offer certain 4.1.2.3. PREPARATION OF THE CELLS FOR advantages, they have not been widely tested. PRESERVATION To preserve the cells by freezing, suspensions Cells are collected from a solid medium in in glutamic acid or skimmed milk are mixed cryoprotective solution to obtain suspensions with glycerol to achieve a final glycerol concen- of approximately 108 cells/mL. Meso-inositol tration of 12 to 15%. Good results are also (5%) is used as a cryoprotectant for the lyophil- obtained with a mixture of skimmed milk and ization of acetic acid bacteria. When acetic acid glucose. bacteria are preserved by freezing, the suspen- 4.1.2. Guidelines for the Preservation sion is mixed with glycerol to obtain a final of Acetic Acid Bacteria concentration of 12 to 15%, as with lactic acid bacteria. 4.1.2.1. CULTURE MEDIA The species of acetic acid bacteria that have been isolated from wine can grow in very 4.1.3. Protocols for Long-term Preservation simple media but need a non-nitrogenated 4.1.3.1. FREEZING PROTOCOL carbon source. Peptones and amino acids cannot 1) Culture the microorganism under the be used as carbon sources and the options in conditions recommended for each species. descending order of preference are glycerol, 2) Prepare a cell suspension containing ethanol, glucose, and mannitol, followed by a cryoprotective agent at an appropriate other carbon compounds. All of the acetic acid concentration. bacteria produce acid from glucose but only 3) Prepare sterile cryotubes (sterilizable some species can produce it from mannitol. screwtop plastic tubes with a rubber seal that Peptones and amino acids can be used as are appropriate for freezing) containing 1 mL nitrogen sources, but ammonium sulfate is of 20% glycerol. also widely used. The bacteria do not require 4) Add approximately 0.5 mL of cell suspension essential amino acids but their growth is stimu- to each tube to obtain a final concentration of lated by yeast extract. approximately 3 107 cells/mL and 12 to PRESERVATION OF BACTERIA 315 14% glycerol. Prepare batches of at least five 4.2. Short-term Preservation tubes for each strain. 5) Freeze at a temperature of at least 80C and 4.2.1. Maintenance by Periodic Transfer maintain the samples at this temperature Maintenance by periodic transfer keeps the without allowing cycles of freezing and cells in an active and easily available form but, thawing. Remove a single tube each time the as mentioned, the method is not advisable for culture is to be used and discard any use over extended periods. remaining cells. Bacteria associated with winemaking are not easy to maintain in live culture without changes Under these conditions, the cultures can occurring. The strong reduction in pH that occurs remain viable for more than 5 years. The viable in the medium causes serious damage to the cells period is reduced in 40 C freezers. Freezers and, although this can be avoided in some cases at 20C should not be used with this method. by adding insoluble calcium carbonate, which Microorganisms can be stored at 20 C if a final will gradually dissolve in the medium as the acids concentration of 50% glycerol rather than 12 to are produced, this is neither sufficient nor advan- 14% glycerol is used in order to prevent freezing tageous for some strains, and as a consequence of the cell suspension and damage due to recrys- the cells must be subcultured regularly to prevent tallization of free water. them dying. In other cases, cells age as a consequence of excessive growth in the nutrient-rich 4.1.3.2. LYOPHILIZATION PROTOCOL media that some of these bacteria require. The The protocol is similar to that used with microorganisms should therefore only be main- freezing but the cell suspensions obtained in tained in live culture for a defined period while step two are placed directly into sterile tubes they are in use. Once this period has passed, we or vials for lyophilization. The CECT recom- should once again use cells preserved by lyophil- mends that the vials contain a small rectangle ization or freezing. of absorbent paper (Whatman No. 3). Following To maintain the cells by periodic transfer, the lyophilization, the vials are closed under growth conditions are as described in Table 12.1. vacuum and stored in the dark. They can be The cultures obtained should be stored at 5 to stored at ambient temperature (18e22C) or in 8C and the time that they can survive under the cold (5C). these conditions will vary according to the microorganism. Most lactobacilli should be 4.1.4. Recovery of the Preserved Cells reseeded every 10 to 15 d. Acetic acid bacteria The frozen cultures are thawed at 37C, inoc- grown in MYA (CECT 63) or GYC (CECT 287) ulated into appropriate media as soon as media remain viable for 1 to 2 months. possible, and incubated under appropriate However, all strains can be altered after conditions. Lyophilized cultures stored under a certain number of subcultures, and methods vacuum in vials or other recipients should be should be tested to extend the length of time rehydrated in liquid nutrient medium as soon between reseeding. These methods are based as they are opened and then cultured under on reducing cell activity: appropriate conditions. Viability should be at least 80% and the stability and authenticity of 1) Nutrient limitation. Cells grown in the strain in terms of its essential characteristics appropriate conditions (as described earlier) should be confirmed. In strains containing plas- are resuspended in distilled water at a mids, it must be ensured that these are present concentration of approximately 106 cells/mL. and that they multiply. They are stored at 5 to 8C. 316 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY

2) Reducing storage temperature. As mentioned, 5.1. Methods for the Preservation of the cells can be stored at 20C without Filamentous Fungi freezing if mixed with 50% glycerol (ﬁnal concentration). 5.1.1. Freezing A distinction must be made between sporu- Alternative preservation methods such as lating and nonsporulating fungi. The follow- dessication are not recommended for bacteria ing protocol can be used in nonsporulating used in winemaking. strains:

1) Begin with a plate culture of the fungus using 5. PRESERVATION OF the most appropriate medium and FILAMENTOUS FUNGI FROM WINE temperature for the microorganism. 2) Cut blocks of approximately 0.5 0.5 cm containing the growing vegetative mycelium Various factors must be taken into consider- of the fungus. ation when choosing a method for the preserva- Place the blocks in groups of five in screwtop tion of filamentous fungi. These include the 3) vials containing 1 mL of 10% (vol/vol) sterile number of strains, the degree of genetic and glycerol and seal. phenotypic stability required, and the storage Maintain the vials at 4C for 30 min and then period. However, one of the most important is 4) freeze at temperatures below 80C. the type of reproduction used by the strain to be preserved. Fungi that reproduce both asexu- In the case of sporulating fungi: ally (through conidia) and sexually (using Prepare a suspension of spores in an spores) are relatively easy to preserve, since all 1) appropriate cryoprotective solution, which in of them survive lyophilization well. However, the case of filamentous fungi is usually sterile in those species that do not produce spores skimmed milk. under laboratory conditions but rather repro- Adjust the concentration to 106e107 duce asexually by fragmentation of hyphae, 2) spores/mL. lyophilization is impossible and preservation Aliquot 0.5 mL of the suspension into via freezing is difficult to achieve. This is 3) cryotubes containing 1 mL of 15% (vol/vol) because both freezing and lyophilization glycerol. depend on the presence of structures that confer Freeze at temperatures below 80C. resistance, at least in eukaryotic cells. These 4) fungi can only be preserved by periodic transfer As mentioned earlier for other microbial or suspension in sterile distilled water, a surpris- groups, to recover the fungus, the thawing ingly simple method that has proved to be one process should be rapid, with the immersion of the best for use in those strains that do not of the cryotubes in a 37C water bath for 30 s. tolerate long-term preservation methods. Obli- Once thawed, the cells are seeded in an appro- gate parasites such as Plasmopara viticola or priate culture medium. Uncinula necator (which cause mildew) cannot The CECT has studied the effects of different be cultured in laboratory media and must freezing and thawing methods (Juarros et al., therefore be preserved in the host by freezing 1993). pieces of the infected plant in liquid nitrogen (Dahmen, 1983). For more general information 5.1.2. Lyophilization on the preservation of filamentous fungi, see The use of lyophilization is restricted to the Onions (1971) and Smith and Onions (1994). preservation of sporulating filamentous fungi. PRESERVATION OF FILAMENTOUS FUNGI FROM WINE 317

Prepare a suspension of spores in an 1) fully, it is neither necessary nor desirable appropriate cryoprotective solution, which to obtain abundant growth, as this will can be sterile skimmed milk or a combination accelerate aging due to accumulation of toxic of equal parts sterile skimmed milk and 5% metabolites. inositol. Adjust the concentration to Once the microorganism has begun to grow, 106e107 spores/mL. 3) the culture should be kept at a temperature Aliquot equal volumes into lyophilization 2) of 5 to 7C, since reducing the storage vials. temperature will help to reduce the Freeze the tubes and place them in the 3) metabolic activity and increase the time lyophilizer. between subcultures. Once the spores are lyophilized, recovery 4) It is difficult to predict the maximum time involves rehydration of the cells in an 4) between transfers, since this can vary appropriate liquid culture medium, which is according to the species and even the strain. then used to inoculate the recommended There are strains of nonsporulating fungi solid medium. that need to be subcultured every 3 months, each time with the risk of contamination, 5.1.3. Preservation by Subculture or loss of characteristics, and even loss of the Periodic Transfer strain itself. In the case of sporulating In general terms, preservation by subculture filamentous fungi, the time between or periodic transfer is the oldest, simplest, and subcultures can be extended to up to 12 most accessible method with which to preserve months. small culture collections for relatively short periods. It can be used in both sporulating and nonsporulating filamentous fungi. However, 5.1.4. Preservation in Sterile Distilled the viability and stability of the cultures is rela- Water tively poor compared with those obtained with Despite its simplicity, preservation in sterile freezing or lyophilization. distilled water is highly effective for nonsporu- The aim of this method is to maintain pure, lating filamentous fungi. The method was orig- active, stable, and immediately recoverable inally described by Castellani (1939, 1967) for cultures. To this end, it is important to prevent the preservation of pathogenic fungi from aging or phenotypic or genotypic changes in humans. Boeswinkel (1976) achieved reasonable the culture. This requires a specific protocol to success using this method to preserve a collec- be designed for each microorganism in which tion of 650 pathogenic fungi from plants, a careful study is performed of the specific including representatives of the divisions culture medium to be used (including its water Oomycota, Ascomycota, Basidiomycota, and activity), the temperature, light, aeration, pH, imperfect fungi. and incubation time. The method is used extensively by the CECT for genera belonging to the division Basidiomy- 5.1.3.1. GENERAL RECOMMENDATIONS FOR cota such as Agaricus, Agrocybe, Armillaria, and THE PROCEDURE Coprynus, and for some belonging to the divi- 1) Culture media should be alternated, since sion Oomycota, such as the phytopathogenic the microorganisms can degenerate if genera Pythium and Phytophthora. Using this maintained consistently in the same medium. method, we have succeeded in maintaining 2) The inoculum should be small. Although the viable cultures for more than 5 years and have intention is for the microorganism to develop even recovered strains after 15 years. 318 12. PRESERVATION OF MICROBIAL STRAINS IN THE WINE INDUSTRY

5.1.4.1. PROCEDURE References 1) Grow the fungus on a plate using the most A´ lvarez-Rodrı´guez, M. L., Lo´pez-Ocanã, L., Lo´pez- appropriate culture medium and Coronado, J. M., Rodrı´guez, E., Martıńez, M. J., temperature for each microorganism. Larriba, G., et al. (2002). Cork taint of wines: Role 2) Cut blocks of approximately 0.5 0.5 cm of the filamentous fungi isolated from cork in the containing the growing vegetative mycelium formation of 2, 4, 6-trichloroanisole by o methylation of 2, 4, 6-trichlorophenol. Appl. Environ. Microbiol., 68, of the fungus. 5860e5869. 3) Place the blocks in groups of five in screwtop Beech, F. W., & Davenport, R. R. (1971). Isolation, purifica- microvials containing 1 mL of sterile water tion and maintenance of yeasts. In C. Booth (Ed.), and hermetically seal them. Methods in Microbiology, Vol. 4 (pp. 153e182). London, Store the vials in the dark at a controlled UK: Academic Press. 4) Boeswinkel, H. J. (1976). Storage of fungal cultures in water. temperature of 18 to 20C. T. Brit. Mycol. Soc., 66, 183e185. 5) In order to recover the microorganism, it is Castellani, A. (1939). Viability of some pathogenic fungi in sufficient to place the growing region of the distilled water. J. Trop. Med. Hyg., 42, 225e226. fungus in contact with an appropriate culture Castellani, A. (1967). Maintenance and cultivation medium. of common pathogenic fungi of man in sterile distilled water. Further researches. J. Trop. Med. Hyg., 70, 181e184. APPENDIX 1 Dahmen, H., Staub, T., & Schwinn, F. T. (1983). Technique for long-term preservation of phytopathogenic fungi in liquid nitrogen. Phytopathology, 73, 241e246. Culture Media Day, J. G., & Stacey, G. N. (2007). Cryopreservation and freeze- drying protocols. Totowa, NJ: Humana Press. All of the culture media referred to in this de Man, J. D., Rogosa, M., & Sharpe, M. E. (1960). A chapter appear with the number assigned to medium for the cultivation of lactobacilli. J. Appl. Bac- them by the Spanish Type Culture Collection teriol., 23, 130e135. (CECT). Their composition can be found on Hatt, H. (Ed.). (1980). American type culture collection methods: I. Laboratory manual on preservation, freezing and freeze-drying. the CECT webpage (http://www.cect.org). Rockville, MD: American Type Culture Collection. Hill, L. R. (1981). Preservation of microorganisms. In Cryoprotectants J. R. Norris & M. H. Richmond (Eds.), Essays in applied microbiology (pp. 2/1e2/31). Chichester, UK: John Wiley 1. Glutamic acid (0.067 M in aqueous solution). & Sons. 2. Drops of 1N NaOH are added to help Hunter-Cevera, J. C., & Belt, A. (1996). Maintaining cultures dissolve the glutamic acid with shaking. The for biotechnology and industry. London, UK: Academic Press. solution is sterilized for 20 min at 1 Juarros, E., Tortajada, C., Garcıá, M. D., & Uruburu, F. atmosphere pressure. (1993). Storage of stock cultures of filamentous fungi at 3. Glucose (7.5%, wt/vol). 80C: Effects of different freezing-thawing methods. 4. Inositol (5%, wt/vol). Microbiologıá SEM, 9,28e33. Skimmed milk (10%). Kirsop, B. E. (Ed.). (1980). The stability of industrial organisms. 5. Kew, UK: Commonwealth Mycological Institute. Homogenized, skimmed milk from any high- 6. Kirsop,B.E.,&Doyle,A.(1991).Maintenance of microor- quality producer without additives or ganisms and cultured cells.London,UK:Academic preservatives. Sterilize for 25 min at 112C Press. without allowing it to caramelize. It can also Onions, A. H. S. (1971). Preservation of fungi. In C. Booth be prepared from dried skimmed milk. (Ed.), Methods in microbiology, Vol. 4 (pp. 113e151). London, UK: Academic Press. Dried skimmed milk (10%, wt/vol) þ 7. Smith, D., & Onions, A. H. S. (1994). The preservation and glucose (3%, final concentration). maintenance of living fungi. Wallingford, UK: CAB 8. Skimmed milk (10%) þ 2M adonitol. International. CHAPTER 13

Application of the Hazard Analysis and Critical Control Point System to Winemaking: Ochratoxin A Adolfo J. Martıńez-Rodrı´guez, Alfonso V. Carrascosa Santiago Instituto de Investigacioń en Ciencias de la Alimentacioń (CIAL, CSIC-UAM), Madrid, Spain

OUTLINE

1. Introduction 320 Appropriate to these Principles and their Application 322 2. General Considerations Regarding 2.2. Prerequisite Programs 322 The HACCP System 320 2.2.1. Good Agricultural Practice 322 2.1. HACCP Principles 320 2.2.2. Good Manufacturing Practice 322 2.1.1. Principle 1: Conduct a Hazard Analysis 321 3. Application of the HACCP System 2.1.2. Principle 2: Determine the to Winemaking 323 Critical Control Points 321 3.1. Background 323 2.1.3. Principle 3: Establish Critical 3.2. Applying HACCP to the Control Limits 321 of Ochratoxin A (OTA) in Wine 324 2.1.4. Principle 4: Establish a 3.2.1. Mycotoxins in Wine: Monitoring System 321 Ochratoxin A (OTA) 324 2.1.5. Principle 5: Establish Corrective 3.2.2. Ochratoxin A (OTA)-producing Actions 321 Microorganisms in Wine 326 2.1.6. Principle 6: Establish Veriﬁcation 3.2.3. The 12 HACCP Tasks Applied Procedures 321 to Ochratoxin A (OTA) 2.1.7. Principle 7: Establish in Wine 327 Documentation Concerning All Acknowledgments 336 Procedures and Records that are

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10013-X 319 Copyright Ó 2011 Elsevier Inc. All rights reserved. 320 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING 1. INTRODUCTION 20th session, in 1993, and eventually issued guidelines regarding what became known as The Hazard Analysis and Critical Control the twelve tasks (Food and Agriculture Organi- Point (HACCP) system, which emerged in the zation [FAO], 1998b) aimed at correctly imple- 1970s as a systematic way of ensuring food menting seven principles. Both the tasks and hygiene and protecting consumer health, is principles are described in some detail below. based on controlling microorganisms that Application of the HACCP system has since constitute a hazard for consumers (International become a mandatory standard in countries Commission on Microbiological Specifications such as Spain, where it is regulated by Royal for Foods [ICMSF], 1991). The system was orig- Decree 2207/1995 of 28 December, establishing inal in that it adopted a preventive rather than hygiene standards for food products. a retrospective approach. Previous food safety The HACCP is the most comprehensive systems had been based on end-product testing system for preventing risks to consumers from aimed at preventing pathogen-contaminated foods. Its scientific approach makes it superior products from entering the food chain to any other system in terms of its efficacy and anddrather more difficultdassessing the origin breadth of coverage of all aspects that can and source of the contamination. The retrospec- contribute to removing or reducing food tive approach, however, could not fully guar- hazards for humans. Consequently, it is the antee food safety, given the impossibility of only such system underpinned by legislation. analyzing entire production batches. Tolerable The HACCP system is designed to identify levels of pathogens in foods were therefore hazards (potentially harmful microorganisms established by legislation and food processing that can affect food), assess risk (i.e., the proba- sites were routinely inspected by public bodies. bility that these hazards are present in the In the HACCP system, the emphasis is on production system), and guide the establish- monitoring and maintaining production condi- ment of appropriate control measures. Success- tions that prevent undesirable or hazardous ful application of the system largely builds on microorganisms from contaminating and compliance with health and sanitation stan- growing in food, from the raw material stage dards and on the use of well-established quality to the consumer. The system is now widely management systems such as Good Agricul- applied to the control of all kinds of biological, tural Practice (GAP), Good Animal Husbandry physical, and chemical hazards (Mortimore & Practice (GAHP), Good Storage Practice (GSP), Wallace, 2001). Its wide acceptance and success- Good Manufacturing Practice (GMP), and ful application led to its incorporation into food Good Hygiene Practice (GHP). It is also compat- safety legislation in the European Union (EU) ible with quality assurance systems such as ISO (Directive 93/43/EEC on the hygiene of food- 9000. stuffs) and the United States (Food and Drug Administration [FDA] regulation CPR-123). 2. GENERAL CONSIDERATIONS The United States National Advisory REGARDING THE HACCP SYSTEM Committee on Microbiological Criteria for Foods issued HACCP guidelines that included 2.1. HACCP Principles generic plans (National Advisory Committee on Microbiological Criteria for Foods Seven different activities, referred to as the [NACMCF], 1992). The Codex Alimentarius seven principles in the Codex Alimentarius Commission adopted the HACCP system at its Guideline (FAO, 1997), are necessary to establish, GENERAL CONSIDERATIONS REGARDING THE HACCP SYSTEM 321 implement, and maintain an HACCP system. boundary between safe and unsafe products. These seven principles are described below. These limits are sometimes referred to as absolute tolerance or safety limits. Control parame- 2.1.1. Principle 1: Conduct a Hazard ters used for this purpose should be variables Analysis that are directly related to the presence of unde- Hazards should be identified and the associ- sirable microorganisms and can be measured ated risks assessed at each phase of the produc- rapidly (e.g., pH and temperature rather than tion system. Measures for controlling hazards the more time-consuming microbiological tests). and risks should also be described. Since this Indirect measures can also be used if they are book is concerned with wine microbiology, we known to be reliably associated with the pres- will focus exclusively on microbiological ence of the microorganism. hazards that affect the quality of wine and have potential harmful effects on human health. 2.1.4. Principle 4: Establish a The hazard in this case is the presence, Monitoring System survival, and growth of microorganisms or the Monitoring is the systematic, scheduled production of substances (toxins, metabolites, measurement or observation of the parameters etc.) in wine at levels that are unacceptable in established for all the CCPs to check that these terms of ensuring the health of the consumer. are under control; that is, that they are within the critical limits described in Principle 3. Appli- 2.1.2. Principle 2: Determine the cation of this principle requires the definition of Critical Control Points monitoring activities and frequencies and the Critical control points (CCPs) are steps at designation of a person with a supervisory role. which essential control measures designed to prevent or eliminate a food safety hazard or to 2.1.5. Principle 5: Establish Corrective reduce it to an acceptable level are applied. In Actions other words, they are specific production stages When monitoring activities indicate a devia- where the implementation of appropriate tion from an established critical limit at a CCP, control measures will ensure the elimination or specific corrective actions or procedures need minimization of a specific hazard. to be implemented to restore control. It is neces- The initial classification of CCPs distin- sary to both establish these actions and designate guished between a CCP1, which was an opera- a person responsible for implementing them and tional or production phase in which a hazard deciding what to do with the affected product. could be eliminated, and a CCP2, which was an operational or production phase in which 2.1.6. Principle 6: Establish Verification the hazard was only partially eliminated (i.e., it Procedures was minimized but not brought under control) To verify the effectiveness of the HACCP (ICMSF, 1991). Despite the usefulness of this system, periodic checks should be performed classification, however, it is no longer applied. by the persons responsible for the control operations to evaluate deviations and product disposi- 2.1.3. Principle 3: Establish Critical Limits tion and to analyze samples to confirm whether Each control measure associated with a CCP or not the CCPs are under control. The analyses should have an associated critical limit that should incorporate tests (including microbiolog- distinguishes between what is acceptable ical tests) other than those used for monitoring and unacceptable. Critical limits delimit the purposes, even though incubation times may 322 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING mean that results will not be immediate. These 2.2.1. Good Agricultural Practice analyses more or less correspond to standard The primary production process should quality-control checks. ensure that foods are safe for the consumer. In the case of wine, grape growers should manage 2.1.7. Principle 7: Establish Documentation production in such a way that crop contamina- Concerning All Procedures and Records tion, pest proliferation, and animal and plant that are Appropriate to these Principles diseases do not pose a threat to food safety. and their Application The land used for cultivation should be fit for Application of this principle is essential in purpose and not have been previously contam- that it facilitates verification audits and subse- inated with heavy metals, industrial chemicals, quently ensures that the system is kept up to or environmental waste, as such hazards will date and continually evaluated. It is recom- enter the food chain and render the correspond- mendable to create an HACCP manual containing commodities unfit for human consumption. ing a written description of the application of Where appropriate, GHP should be designed to the HACCP principles (in terms of hazards, ensure that the harvested commodity will not risks, critical points, critical limits, corrective represent a food hazard to the consumer. If actions, etc.) to the process in question. The appropriate, GSP should be applied to ensure manual should also be used to keep records of that hazards are eliminated during harvesting, the different operations implemented during after harvesting, and throughout the entire routine functioning of the system. production process. A number of works have been published on the application of the HACCP system to the 2.2.2. Good Manufacturing Practice food industry. As introductory reading, we 2.2.2.1. DESIGN AND CONSTRUCTION OF recommend the books published by ICMSF FACILITIES (1991) (available in Spanish) and Mortimore The structure and location of the facilities and Wallace (2001), both of which provide required to produce wine must meet certain extensive additional information on the princi- requirements and comply with legislation ples of the system and its application to specific specific to wineries. The following general food sectors. issues should be considered:

2.2. Prerequisite Programs 1. Premises should be designed to minimize the risk of commodity contamination. Prerequisite programs such as GAP, GSP, 2. The design and layout of premises should GMP, and GHP need to be correctly imple- enable maintenance, cleaning, and mented before the HACCP system can be disinfection operations that minimize applied to the production of a plant-based airborne contamination. product such as wine (FAO, 2003). The introduc- 3. All surfaces that come into contact with tion of the HACCP system will be complicated if food should be nontoxic and easy to these programs are not functioning effectively maintain and clean to prevent and the outcome will be a cumbersome, over- contamination. documented system. In the interest of avoiding 4. When required, there should be adequate confusion, it should be emphasized that, means for controlling temperature and although GAP, GSP, GMP, and GHP cover some humidity. of the elements of the HACCP system, they do 5. Effective pest control measures should be in not replace it, as will become evident below. place. APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 323

2.2.2.2. CONTROL OF OPERATIONS commensurate with their duties. Food handlers Measures aimed at reducing the risk of should also be overseen by suitably trained contamination of commodities and foods and supervisors. Ongoing training for food handlers ensuring that they are safe and fit for purpose is essential to the success of a food safety should be implemented. These include the management system. following: 2.2.2.7. PRODUCT INFORMATION AND Adequate temperature, time, and humidity 6. CONSUMER AWARENESS controls The end product should be accompanied by Food-grade packaging 7. sufficient information to ensure that the Potable water supplies 8. personnel at the next stage in the food chain Equipment maintenance 9. will handle, store, prepare, and display the product safely and in a way that does not 2.2.2.3. MAINTENANCE AND SANITATION increase hazards. This is particularly important Procedures and work instructions that ensure for foods that are consumed fresh. an adequate level of maintenance of the premises and effective cleaning, waste management, and pest control practices should exist. These operations will help in the constant monitoring 3. APPLICATION OF THE HACCP of potential hazards that could cause food SYSTEM TO WINEMAKING contamination. 3.1. Background 2.2.2.4. PERSONNEL HYGIENE The application of the HACCP system to Measures should be put in place to ensure winemaking is the subject of a number of books that food handlers do not contaminate food (Federacioń de Industrias de Alimentacioń and that they maintain an appropriate level of y Bebidas [FIAB], 1997; Hyginov, 2000) and arti- personal hygiene and comply with relevant cles in specialist journals (Briones & U´ beda, guidelines. 2001; Kourtis & Arvanitoyannis, 2001; Morassut & Cecchini, 1999). Not all of these publications 2.2.2.5. TRANSPORTATION cover microbiological hazards, and none Measures should be put in place to prevent consider that the consumption of wine may deterioration of the commodity during trans- constitute a microbiological health risk. port. Raw materials or products to be trans- A fairly common error, which is essentially ported should be properly monitored. due to a lack of in-depth knowledge of the Examples include products that need to be refrig- HACCP system, is to consider GAP and GMP erated, frozen, or stored at specific humidity to be equivalent to HACCP. This downplays levels. Transport means should be kept in good the true significance of the HACCP system as condition and be easy to clean. Containers used a comprehensive approach consisting of many for bulk transport should be used exclusively elements whose goal is to gain true control of for food. an entire process. At best, GAP and GMP include just some of the preventive measures 2.2.2.6. TRAINING provided by an HACCP system. This confusion All food handlers should be trained in is typicaldmany winery managers, wine personal hygiene and in the specific operations experts, and even scientists are unaware of the for which they are responsible, to a level HACCP system. 324 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING

Wine is contaminated naturally during Organization of Vine and Wine (OIV) (OIV, production. While more typical in foods of 2002), this limit is now mandatory for all EU animal origin, foodborne pathogens such as member states. Since OTA is the only toxic Aeromonas hydrophila, Bacillus cereus, Campylo- substance of microbial origin in wines that is bacter jejuni, enterotoxigenic and enterohemor- regulated internationally, this chapter will rhagic Escherichia coli, Listeria monocytogenes, describe the application of the HACCP system Salmonella enteritidis, and Shigella dysenteriae,as to this mycotoxin. well as toxin-producing bacteria such as Clos- tridium botulinum and Staphylococcus aureus can 3.2. Applying HACCP to the Control contaminate wine throughout the production of Ochratoxin A (OTA) in Wine phases, although they are unable to thrive because of the pH levels that characterize this Although the HACCP system was conceived medium (ICMSF, 1996). to improve and ensure hygiene and sanitation in The combined effects of the ethanol and pH both the agricultural and food processing levels that characterize wine can cause loss of sectors (ICMSF, 1991), it has mainly been viability of pathogens such as Salmonella typhi- applied in the latter. One of the reasons is that, murium, Salmonella sonnei, and enterotoxigenic whereas a processing facility like a winery E. coli (Bellido et al., 1996; Sheth et al., 1988; might have just one owner, the vineyards Weisse et al., 1995), and even of viruses such supplying the winery may have many owners. as hepatitis A (Desenclos et al., 1992). Certain Hence, fully preventing or eliminating a food polyphenols present in the wine can also inhibit hazard or reducing it to an acceptable level is the growth of S. enteritidis (Marimoń et al., 1998) generally more difficult in the primary process- and Campylobacter jejuni (Ganãń et al., 2009). ing of plant-based foods (FAO, 2003). It is also Furthermore, it has been known for some time more difficult to control parameters outdoors that moderate wine consumption increases than in indoor production facilities. gastric secretion and intestinal motility Nonetheless, since the application of the (Bujanda, 2000; Pfeiffer et al., 1992), making it HACCP system to OTA in wine is closely asso- more difficult for pathogens to invade the intes- ciated with the grape growing phase, to all tine. It has also been postulated that moderate effects and purposes, we consider this phase to wine consumption reduces the infectious poten- be an integral part of the winemaking process. tial of intestinal pathogens such as Helicobacter Indeed, in the case of mycotoxins such as pylori, the main cause of chronic gastritis and OTA, it is crucial to implement production duodenal ulcers (Brenner et al., 1999; Ruggiero controls aimed at protecting grape berries et al., 2006). These data show that wine is a func- from fungal infection in the vineyards. tional food and has a role in defending the intes- tine from pathogens. 3.2.1. Mycotoxins in Wine: While the literature contains no reports of Ochratoxin A (OTA) outbreaks of illness caused by toxins in wine, While it has long been acknowledged by it is acknowledged that wines may contain many that the presence of OTA in wine is at least a range of toxic substances of microbial origin. a potential hazard, it was not described as such A maximum limit of 2 mg/L for ochratoxin A in the wine literature until the late 1990s (OTA) levels in musts and wines produced after (Gottardi, 1997). the 2005 harvest, for example, was recently Although there are over 300 known myco- established by the EU (European Union [EU], toxins, only a few are recognized as representing 2005). Previously proposed by the International a level of risk that requires the implementation APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 325 of strict controls (FAO, 2003). The fact that rec- most important effect of certain mycotoxins is ommended limits for OTA levels in wines have their capacity to block the immune response been established is, in itself, a good enough and reduce resistance to infectious diseases. reason for developing an HACCP system for OTA is a mycotoxin with nephrotoxic, carcino- its control. genic, teratogenic, immunotoxic, and possibly Mycotoxins are toxic substances of fungal neurotoxic effects (Turner et al., 2009). It has origin. When ingested, inhaled, or absorbed been associated with Balkan endemic nephrop- through the skin, they cause illness or death in athy (Turner et al., 2009), although the evidence both humans and animals (Pitt, 1996). They are for this has been contested in certain sectors secondary metabolites that appear to have no (Delage et al., 2003; Soleas et al., 2001; Zimmerli specific function in the growth of the species & Dick, 1996). The tolerable daily intake of OTA that produce them. They pass to humans is very low, ranging from just 0.3 to 0.89 mg/d through food that has been contaminated by for a person weighing 60 kg, with acute toxicity mycotoxigenic filamentous fungi. Some of the likely to occur at a dose of between 12 and mycotoxins that have been found in grapes 3000 mg for a person of that weight (Rousseau, and grape products are listed in Table 13.1. 2004). The Joint FAO/World Health Organiza- Exposure to mycotoxins can result in acute or tion (WHO) Expert Committee on Food Addi- chronic toxicity and ultimately lead to delete- tives has established a provisional tolerable rious effects on a range of body organs and weekly intake of OTA as 100 ng/kg of body- systems, and even death. It is widely believed, weight (FAO, 2002), which corresponds to particularly in developing countries, that the 14 ng/d/kg of bodyweight.

TABLE 13.1 Mycotoxins Isolated in Grapes and Grape Products

Mycotoxin Substrate Fungus References

Byssochlamic acid Grape Byssochlamys fulva Samson et al. (1996) Byssochlamys nivea Citrinin Must Penicillium citrinum Vinas et al. (1993) Penicillium expansum Patulin Must Byssochlamys fulva Frisvad and Thrane (1996) B. nivea

Penicillium expansum Ochratoxin A Grapes, must, wine Aspergillus carbonarius Bau et al. (2005); Caban˜es et al. (2002); Gallo et al. (2009); Selma et al. (2008)

Aspergillus fumigatus Battilani and Pietri (2002)

Penicillium pinophilum Aspergillus tubingensis Oliveri et al. (2008) Aspergillus japonicus

Adapted from Carrascosa (2005). 326 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING

After cereals, wine is the next most common they analyzed and estimated a daily intake of source of OTA for humans (Caban˜es et al., 0.01 ng/d/kg of bodyweight (Blesa et al., 2002). OTA was first detected in wine in 1995 2004). In Spain, OTA has been detected at (Zimmerli & Dick, 1996). The Codex Alimentar- concentrations of up to 11.7 ng/mL and up to ius Commission has even conceded that grapes 4 ng/mL in the plasma of patients with chronic and grape products were the source of over 15% kidney failure and healthy individuals, respec- of OTA intake in Europe (FAO, 1998a). tively (Pe´rez de Obanos et al., 2001). These The EU has approved a regulation establish- values are similar to those reported for other ing tolerable intake limits for OTA in cereals European countries. The European Food Safety (5 mg/kg), cereal products (3 mg/kg), and raisins Authority’s (EFSA) Scientific Panel on Contam- (10 mg/kg) (EU, 2002a). An upper limit of 2 mg/L inants in the Food Chain (CONTAM) has estab- has also been proposed for musts and wines lished a tolerable weekly intake of 120 ng/kg of (FAO, 2003). This limit is the same as that recently bodyweight for OTA (European Food Safety established in the EU (2005). The mean content Authority [EFSA], 2006). It is estimated that of OTA in red wine in Europe is 0.19 mg/L, and the true intake of OTA in Europe is between 15 the total daily intake of OTA in Europe has and 60 ng/kg bodyweight, and that OTA levels been estimated as 171 g (FAO, 1998a), which in wines from Africa, America, Australia, and would correspond to an OTA concentration of Japan are lower than in European wines (Mateo between 0.01 and 3.4 mg/L. It should be noted, et al., 2007). however, that the presence of this mycotoxin is Based on the above data and the results of more common and its concentration greater in studies examining OTA levels in wines, it can warmer, wetter years; in temperate climates; in be concluded that the limit of 2 mg/L recom- the south; and in sweet wines made from over- mended by the EU and the OIV is not often ripe or raisined grapes; it is generally more exceeded. Given that a level of 3.4 mg/L of common in red wines, followed by rose´ and OTA is reached in certain kinds of wines then white wines (Battilani & Pietri, 2002; produced in the EU, however, it has to be Burdaspal & Legarda, 1999). Although OTA is acknowledged that the risk existsdas has been detected in over 50% of wines, there are very suggested by a number of authors (Mateo few cases where the maximum allowable limit et al., 2007; Olivares-Marıń et al., 2009)deven of 2 ng/mL is exceeded. Levels also fall with if it is moderate given the infrequent presence increasing latitude (Mateo et al., 2007). of this myotoxin. Nonetheless, in order to Studies on the occurrence of OTA in wines achieve a true reduction in OTA levels in wine indicate that it remains stable in this substrate via the application of an HACCP system, the for at least 12 months (Mateo et al., 2007). theoretical framework has to assume this level They also point to the extremely important of risk to be unacceptable. role played by factors such as the year of harvest as different weather conditions can result in 3.2.2. Ochratoxin A (OTA)-producing enormous differences. In one study, for Microorganisms in Wine example, it was found that the percentage of While Aspergillus ochraceus and Penicillium wines containing OTA ranged from one year verrucosum produce OTA in cereals (ICMSF, to the next from 86% (with OTA concentrations 1996), these fungi are not commonly isolated of 0.056e0.316 ng/mL) to 15% (range 0.074e in grapes or on vines. The presence of OTA in 0.193 ng/mL) (Lo´pez de Cerain et al., 2002). grapes and grape products is attributed funda- The authors of a similar study found OTA levels mentally to Aspergillus carbonarius (Figure 13.1) of between <0.01 and 0.76 ng/mL in the wines (Bau et al., 2005; Caban˜es et al., 2002; Gallo APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 327

perform 12 consecutive tasks (FAO, 1998b). (a) Below we brieﬂy describe each of these tasks both in general terms and in regard to their speciﬁc application to OTA in wine.

3.2.3.1. TASK 1: ESTABLISH AN HACCP TEAM The application of the HACCP system should start with the appointment of a team to perform the tasks necessary to implement the seven principles. The team leader should be familiar with the HACCP system and methods, ensure that the concept is properly applied, be a good listener, and encourage all (b) the team members to become involved. Given that OTA constitutes a microbial hazard, the team should include a microbiologist, preferably with expertise in mycology and mycotox- icology. In order to develop the commodity flow diagram (CFD)ddescribed belowdthe team should also be able to call on individuals familiar with both viticulture and winemaking processes. To avoid possible conflicts of interest, it is advisable to have wine industry representatives from both the public and private sectors on the team. As far as OTA is concerned, the scope of the study should cover the entire production chain, FIGURE 13.1 Aspergillus carbonarius from infected grapes grown in solid medium (a) and close-up of conidia from the vineyard to the bottle. (fresh mount) (phase contrast microscopy) (b). Scale bar ¼ 10 mm. Images kindly provided by Dr F.J. Caban˜es of the Veterinary Mycology Group at Universitat Auto`noma de 3.2.3.2. TASK 2: DESCRIBE THE PRODUCT Barcelona, Spain. A complete description of the product, including client specifications, is necessary to begin the hazard analysis. A basic generic et al., 2009; Go´mez et al., 2006; Selma et al., 2008) product description formdlike the one shown and, to a lesser degree, to other species from the in Figure 13.2dshould be used for this purpose. genus Aspergillus section Nigri (e.g., Aspergillus It should include relevant safety information niger)(Bau et al., 2005; Serra et al., 2003) and to regarding OTA and data on the limits recom- Aspergillus fumigatus and Penicillium pinophilum mended by law, as well as information on (Battilani & Pietri, 2002). packaging, storage, and recommended temperatures. Where appropriate, labeling information 3.2.3. The 12 HACCP Tasks Applied and a sample label should be included. This to Ochratoxin A (OTA) in Wine information will assist the HACCP team in iden- To apply the HACCP principles to the prepa- tifying the real hazards associated with the ration of a food product, it is recommended to process. 328 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING

Product Description Form FIGURE 13.2 Sample product description form. Adapted from FAO, (1998b, 2003). 1. Product name(s) Wine (type, appellation, etc.)

2. Description and key pH, alcohol content, SO2 levels, characteristics of the end product tasting notes, etc. <2 µg/L of OTA

3. Intended use Human consumption

4. Packaging

5. Shelf life

6. Sales point

7. Labeling instructions

8. Special storage and distribution conditions

Date: Approved by:

3.2.3.3. TASK 3: IDENTIFY THE INTENDED USE Examples of CFDs for different wine types are OF THE PRODUCT provided in a number of HACCP studies The product is intended for human consump- applied to winemaking (see Section 3.2.1). A tion. Given the nature of wine, recommenda- common feature of these CFDs is to exclude tions should be included on aspects such as the pre-harvest stages from the winemaking safe consumption levels; incompatibilities, if process. This underlying assumption that wine any, with medication; and precautions to be is prepared exclusively in the winery is taken by individuals with certain health condi- a genuine conceptual errordas we show in tions that would imply limitations on intake. this chapterdand is even contrary to the principles of HACCP, which indicate that each and 3.2.3.4. TASK 4: DRAW UP THE COMMODITY every stage involved in the preparation of FLOW DIAGRAM a food product should be included in the The CFD describes in detail all the stages system. involved in production of each kind of wine In view of the above considerations, we and the order in which these stages occur. believe that all HACCP plans for the control of APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 329

sites and by interviewing enologists, vineyard GRAPE CULTIVATION managers, and winery managers and observing their practices. It is important to bear in mind that the CFD HARVESTING will form the basis for the hazard analysis and establishment of the CCPs. It is, therefore, generally recommended to provide, along SORTING with the CFD, a detailed description of operations with information such as lists of raw materials, additives, the packaging in which these CRUSHING are delivered, storage conditions, activities to be performed throughout the entire process, time and temperature profiles for the different FERMENTATION stages, equipment, design characteristics, a blueprint of the facilities, warehousing conditions, customer and distribution problems, etc. (Martıńez-Rodrı´guez & Carrascosa, 2009; CLARIFICATION Mortimore & Wallace, 2001). For special kinds of wines, such as sweet wines, where the probability of A. carbonarius growth is high (Go´mez BOTTLING et al., 2006), the CFD should include data on grape over-ripening times, environmental conditions, etc. DISTRIBUTION The CFD should also cover the post-production stages right up to delivery to the consumer, in order to identify and draw attention to factors CONSUMPTION that could potentially affect the safety of the product. FIGURE 13.3 Generic commodity flow diagram for winemaking that includes the viticulture stages (growing 3.2.3.5. TASK 5: CONFIRM THE COMMODITY and harvesting) that should be included in any Hazard FLOW DIAGRAM ON SITE Analysis and Critical Control Point system aimed at Once the CFD is completed, the HACCP team controlling ochratoxin A. should visit the vineyard and the winery to check the data collected against the real opera- OTA in wine should take into account the pre- tional conditions. winery phases. To this end, we recommend This operation, known as “walking the that each HACCP team should use a generic line,” consists of checking, step by step, that CFD (like the one shown in Figure 13.3) to guide all the information regarding materials, prac- the creation of a CFD adapted to the wine in tices, controls, etc., has been taken into question. We also recommend that the model consideration by the HACCP team. Where shown in Figure 13.3 be completed by a viticul- appropriate, additional information such as ture expert from the HACCP team, if possible, time of harvest, maximum transportation in collaboration with a representative of the time, transportation conditions to the winery, Ministry of Agriculture. The CFD should be and temperature at, and duration of, the verified further by visiting grape production different stages should be collected and 330 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING included in the CFD. For special kinds of 3.2.3.6.1. HAZARD IDENTIFICATION As men- wines, over-ripening practices at the produc- tioned, this chapter focuses exclusively on OTA tion site (for botrytized wines) and post- as this is the microbiological hazard that is harvest processing (for sweet wines) should most widely regulated internationally in wine. be taken into consideration. The site for which An unacceptable risk is posed when the the HACCP plan is being designed should be permitted level of 2 mg/L of OTA is exceeded. visited as often as necessary to ensure that all This particular hazard should, thus, be evaluated information relevant to the hazard in question at each harvest and production phase. has been collected. In order to simplify the description of how the HACCP system should be applied to OTA, 3.2.3.6. TASK 6: IDENTIFY AND ANALYZE we will assume that this myotoxin is produced HAZARDS (HACCP PRINCIPLE 1) by A. carbonarius, even though other species of Success in applying the HACCP system relies A. section Nigri can also produce it, though to on properly identifying and analyzing the a much lesser extent (Carrascosa, 2005; Go´mez hazards that can arise in association with the et al., 2006; Martıńez-Rodrı´guez & Carrascosa, raw materials and in any of the CFD phases. A 2009). To treat OTA as a chemical hazard merely correctly applied hazard analysis requires the because it is a chemical substance would, in our compilation and evaluation of all data available opinion, make it more difficult to identify on the hazard in question and the factors that control measures, most of which will be related contribute to its occurrence. to the growth of A. carbonarius. A hazard is any factor that might render a food unsafe for consumption. A microbiolog- 3.2.3.6.2. IDENTIFYING COMMODITY FLOW ical hazard is a hazard caused by a microor- DIAGRAM PHASES WHERE OCHRATOXIN A ganism. We consider microbiological hazards (OTA) CONTAMINATION IS MOST LIKELY TO to include hazards established as such on the OCCUR Since A. carbonarius is an opportu- basis of epidemiological data or widely applied nistic pathogen and not highly infectious, the regulations. quantity of OTA produced increases with grape Once a hazard has been identified, the associ- skin damage, temperature, and relative ated riskdthat is, the probability that it will humidity (Bellı´ et al., 2007; Carrascosa, 2005; occurdshould be assessed. Like probability, Kapetanakou et al., 2009; Martıńez-Rodrı´guez risk is rated between 0 and 1, but it is often & Carrascosa, 2009). A. carbonarius generally described qualitatively as low, medium, or develops at harvest time, when grapes are high. Only hazards considered by the HACCP most likely to be damaged (Serra et al., 2003). team to constitute unacceptable risks are carried This is therefore the period when OTA levels forward to Task 7 (HACCP Principle 2). will be highest in the grape. If the grapes have Once risk has been assessed, appropriate suffered extensive damage at an earlier stage, control measures need to be considered. Control the probability of A. carbonarius invasion will measures are actions or procedures used to be higher, as will the risk of higher levels of bring the identified hazard under control, OTA in the final product. The use of damaged whether by preventing or eliminating it or by grapes to make wine will thus increase the reducing it to an acceptable level. The imple- risk of exceeding maximum recommended mentation of control measures requires suitable levels of OTA (Serra et al., 2005). The winemak- training of personnel for specific operations ing process itself can also favor the growth of A. already included or to be included in GAP, carbonarius and hence the production of OTA GMP, and GHP. (Go´mez et al., 2006). For this reason, the CFD APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 331 should, as far as possible, be adapted to each since A. carbonarius can grow at a pH below 4.5, particular case. and indeed does when the skin breaks and Summing up, the main factors that contribute spores attached to the bloom of the fruit or the to biological deterioration by fungi in a vineyard vine germinate. ecosystem are humidity, temperature, and pests. The grape’s skin barrier can be weakened by Fungal growth, for example, is greater in more insects (e.g., wasps, mealybugs, fruit and humid, warmer conditions and insects can vinegar flies, and pyralid caterpillars), phyto- cause considerable damage to grape skin, pathogenic fungi (which cause diseases such leading to the release of nutrients and the as esca and powdery and downy mildew), spread of fungal mycelia through the pulp of birds, and physiological and meteorological the grape. conditions (which can cause water stress and OTA generally appears before harvest time other environmental stresses). All these factors (Serra et al., 2003, 2005). First, however, A. carbo- contribute to skin damage, thus allowing A. car- narius has to develop, and this will only occur in bonarius to access the nutrients in the pulp and phases with sufficient oxygen supplies, as the begin OTA production. fungus is strictly aerobic. These phases occur Attacks on the grapevine by phytopathogenic prior to the crushing of the grapes, as aerobic fungi are more successful when meteorological growth conditions are generally avoided in conditions are propitious. Temperatures of subsequent winemaking stages to prevent the between 20 and 27C, rainy summers, and deterioration of the sensory properties of the damp autumns, for example, all favor the final product. A. carbonarius, therefore, is likely germination of spores and reduce the effective- to develop during the cultivation phase, mainly ness of fungicides (which are most effective in at the grape ripening stage (Bellı´ et al., 2007; dry conditions). Caban˜es et al., 2002; Kapetanakou et al., 2009). Damage to grapes during harvesting by In other plant-based products, contamination rough handling or excessive weight in by mycotoxins can also occur in the storage containers is less likely to lead to the production period between harvest and processing. Indeed, of OTA as the berries are almost immediately storage for periods longer than 48 h at tempera- crushed. This kind of damage, however, should tures of 10C or above is not allowed (FAO, receive particular attention in plant-based foods 2003). In the case of winemaking, the normal such as cereals that are generally placed in practice is to crush the grapes immediately after storage until distribution or processing (FAO, harvesting. If this is not done, the storage 2003). control measures described above should be Further studies are necessary to determine implemented. This is particularly important numerous aspects such as the level of grape for sweet wines (Go´mez et al., 2006). damage required for A. carbonarius to develop, Grape ripening generally coincides with the the time it takes for OTA to be produced and withering of the grape vine, which occurs after the intervening environmental factors, and the veraison. A. carbonarius does not appear to be relationship between fungal growth and OTA capable of attacking the skin of grapes and concentrations in wine. invading the pulp (Bellı´ et al., 2007), which suggests that colonization is strongly favored 3.2.3.6.3. POSSIBLE OCHRATOXIN A (OTA) by pre-existing skin damage (Kapetanakou CONTROL MEASURES OTA control measures et al., 2009). The grape berry has two natural must aim to both prevent and reduce OTA barriers to A. carbonarius: its thick skin and pH. contamination. Preventive control measures Only intact skin, however, can prevent invasion, consist of preventing the development of A. 332 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING carbonarius and, consequently, the synthesis of guidelines that recommend these practices OTA. The main strategies are those designed (OIV, 2005). to prevent grape skin damage. This requires Other preventive control measures include the implementation of a GAP program the use of transgenic grape strains that are resis- involving phytosanitary plans aimed at tant to water stress and damage by phytopatho- ensuring optimally healthy plants that have genic fungi (Colova-Tsolova et al., 2001; Kikkert good defenses against possible parasites and et al., 2001; Vivier & Pretorius, 2000). However, measures to prevent water stress and damage given the time required to adapt the grapevines from fungi, insects, and birds. (3e8 years) and the fact that transgenic foods Insecticides (or alternatives such as chemical are more difficult to market, it is likely to be or biological treatments)dprovided they are some time before preventive measures of this safe for use with foodstuffs and comply with kind are implemented. the legislation underpinning the HACCP sys- If, despite preventive measures, berries temdcan be used to protect against moths become damaged and fungal growth is such as Lobesia botrana, Cryptoblabes gnidiella, detected, the need for OTA reduction measures and Eupoecilia ambiguella. A good preventive must be analyzed. To decide whether or not GAP strategy against birds is to eliminate such measures are necessary, it must be deter- natural shelters or to use optical or acoustic mined, firstly, whether A. carbonarius invasion devices to frighten them away. Fungicidal treat- has occurred, and, secondly, whether unaccept- ments such as sulfur, copper products, and able levels of OTA are being produced. To test organic fungicides can be used to protect against for the presence of A. carbonarius, it is necessary phytopathogenic fungi (Varga & Kozakiewicz, to identify and characterize the species present 2006). These treatments will also protect the using fast, sensitive, and accurate molecular grape skin from damage and subsequent inva- methods (Oliveri et al., 2008). OTA production sion by A. carbonarius. The effectiveness of bio- should also be analyzed using methods that logical control measures based on the use of provide rapid results (Turner et al., 2009; epiphytic yeasts with inhibitory effects on unde- Varga & Kozakiewicz, 2006). sirable fungi by competitive exclusion has also Current recommendations for certain plant- been studied (Bleve et al., 2006), although it is based foods state that fruit damaged by toxi- not entirely clear whether such measures are genic fungi should be discarded. To prevent truly viable. contamination of apple juice by patulin or of Factors that contribute to biological deterio- corn or copra meal by aflatoxin, for example, ration caused by fungi in vineyards (humidity, the recommendation is to discard 99% of all fruit temperature, and pests) are uncontrollable since whose color indicates infection (FAO, 2003). they are dictated by weather conditions. For this Despite the fact that the presence of A. carbonar- reason, preventive measures aimed at signifi- ius is visible (Figure 13.4), no visual selection cantly reducing OTA levels in grapes and wines method has yet been developed to separate need to focus on minimizing damage to the healthy and infected fruit prior to the crushing grapes; such measures will include control of stage. Before removing infected fruit, thus, it is insects and phytopathogenic fungi and the elim- necessary to perform laboratory tests on batches ination of visibly damaged berries before, of grapes from contaminated vineyards to test during, and after harvesting (Bellı´ et al., 2007; for the presence of A. carbonarius and OTA. Carrascosa, 2005; Martıńez-Rodrı´guez & Traceability is obviously an important aspect Carrascosa, 2009). Indeed, in its Resolution in relation to OTA control measures. It is crucial VITI-OENO 1/2005, the OIV issued GAP that the origin of all batches of grapes entering APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 333

can be enhanced by previous heat treatment (Nun˜ez et al., 2008). Nonetheless, these and many other proposed methods lead only to a small reduction in OTA levels; they can also interfere with the organoleptic properties of wine and are of questionable viability (Ame´z- queta et al., 2009). In brief, pre-harvest control measures should be preventive while those implemented during and after harvesting should aim at reducing OTA levels (principally via the sorting and elimination of damaged berries); finally, detoxification methods should be used if OTA levels are detected after the crushing stage (Carrascosa, 2005; Martıńez-Rodrı´guez & Carrascosa, 2009). FIGURE 13.4 Grape infected with Aspergillus carbonarius. Photograph kindly provided by Dr Venancio from the Department of Biological Engineering at the University of 3.2.3.7. TASK 7: DETERMINE THE CRITICAL Minho, Portugal. CONTROL POINTS (HACCP PRINCIPLE 2) The decision tree provided in the Codex the crushing phase is recorded. Each batch Alimentarius (FAO, 1993, 1997) can be used by should be visually inspected for mold and, the HACCP team to help determine the CCPs where necessary, the presence of A. carbonarius to be included in the CFD. The CFD should or OTA should be confirmed so that contami- indicate all phases where hazards are likely to nated batches can be eliminated. arise and all phases with control measures Once the presence of OTA has been detected, should be considered CCPs. If suitable control contaminated grapes should not be mixed with measures cannot be established for a particular uncontaminated grapes (EC, 2002). If an unac- phase or subsequent phases, the corresponding ceptable level of OTA is subsequently detected product should be classified as unfit for human in the must, despite the precrushing control consumption. measures, detoxification measures such as In the example we are analyzing, the grape those recommended for other plant-based sorting phase in the CFD would be a CCP, foods can be used (Coker, 1997). Understand- because it is a point at which a control measure ably, such methods should not compromise can be applied; namely, the elimination of either the safety or the sensory properties of berries contaminated by A. carbonarius and the wine. A range of OTA detoxification thus OTA. Although, as a general rule, preven- methods suitable for application to wine have tive and reduction measures should be applied been studied. One method involves the use during the growing and sorting phases, respec- of activated carbon to reduce OTA levels tively, adaptation of the HACCP system to (Olivares-Marıń et al., 2009). It has also been a particular site may mean that infected berries demonstrated that lactic acid bacteria (del Prete are eliminated by pruning during the ripening et al., 2007) and yeasts (Garcıá-Moruno et al., phase in the vineyard. We have not included 2005) can adsorb OTA. With reference to yeasts, the use of detoxification methods as a CCP, it has been shown that mannoproteins play an given that these methods are still in an experi- important role in OTA adsorption and that mental stage. However, if such methods or any the adsorption capacity of the yeast cell wall other new methods are included in the HACCP 334 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING system, the corresponding phase should logi- agents required to reduce or eliminate different cally be considered a CCP. levels of OTA from contaminated wines. In other fruit sectors, for example, high-pressure 3.2.3.8. TASK 8: ESTABLISH CRITICAL LIMITS water jets are used to remove parts of the fruit FOR EACH CRITICAL CONTROL POINT damaged by mycotoxigenic fungi, but critical (HACCP PRINCIPLE 3) limits need to be set to ensure that the pressure Critical limits are usually established on the used is sufficient to remove the damaged tissue basis of readily measurable CCP parameters without causing further damage to the fruit such as temperature and pH that indicate the (FAO, 2003). presence of a hazard. In the case of OTA, temperature, rainfall, and relative humidity 3.2.3.9. TASK 9: ESTABLISH A MONITORING during the cultivation phase are factors that SYSTEM (HACCP PRINCIPLE 4) are beyond human control, and grape pH does Monitoring activities are essential for check- not necessarily inhibit infection. For this reason, ing whether or not critical limits are being met critical limits will need to be based on parame- at each CCP. The methods used should be both ters that are directly related to grape damage, sensitive and rapid to ensure that any loss of the presence of fungi, and OTA concentration. control is detected by trained personnel at as No studies to date have adequately docu- early a stage as possible. This is crucial to implemented the relationship between OTA concen- menting appropriate corrective measures aimed tration in grape pulp and grape berry damage at preventing or reducing product loss. Moni- or fungal growth. There are, however, such toring activities include the analysis of samples studies for other plant-based products that can collected according to a sampling plan based potentially be affected by mycotoxins. For apple on statistical principles. The most common juice, copra cake, copra meal, and pistachios, for measurements used to monitor mycotoxin example, the guidelines recommend that no formation in plant-based products are storage more than 1% of infected fruit (defined as a fruit time, temperature, and humidity, as these all with >10% surface damage) should enter the provide rapid results and allow suitable correc- processing stage. In other words, 99% of all tive measures to be taken quickly (FAO, 2003). damaged fruit is eliminated, thereby elimi- Visual inspection methods aimed at deter- nating the mycotoxin or at least reducing mining the level of grape damage and fungal concentrations to an acceptable level (FAO, growth in the vineyard should be the first step 2003). in the monitoring of OTA in wine. If the critical If control measures have previously resulted limits are exceeded, tests should then be per- in the detection of A. carbonarius or OTA in formed to determine the presence of A. carbonar- grapes, we recommend that the same critical ius and OTA. Tests for the detection of limits should be applied initially and subse- A. carbonarius should be performed by experts quently adapted on the basis of information in the taxonomy of filamentous fungi using from studies of the winery’s processes. rapid molecular microbiological techniques As detoxification tests are deployed on a large (Oliveri et al., 2008), while OTA tests should be scale and conclusive data become available, performed by chemical analysis experts, prefer- particularly regarding the effect of detoxifica- ably with experience in mycotoxins, using the tion on the sensory quality of wine, and as most validated methods available. wineries put these methods to the test, it should Rapid detection of OTA can be achieved with become possible to establish critical limits based commercial kits based on immunoaffinity and on data showing the quantity of detoxifying similar methods (Varga & Kozakiewicz, 2006). APPLICATION OF THE HACCP SYSTEM TO WINEMAKING 335

These should be sufficiently sensitive to detect associated with the period when the CCP was 2 mg/L of OTA (Turner et al., 2009) and should out of control and modifying product disposi- preferably have been validated in must or tion (by discarding, downgrading, or reprocess- wine. Other analytical methods are equally ing the product; for example, mixing wines to valid, although they do not produce results as reduce the OTA concentration or, if possible, quickly or require samples to be sent to a labora- detoxifying the wine). tory. Such tests will allow batches of grapes to be If grape sorting prior to crushing is not classified as acceptable or unacceptable on the a routine practice, the most effective corrective basis of the level of OTA detected. In other action will be to remove damaged grapes or words, OTA tests will help to ensure that the grapes containing OTA detected by the moni- grape sorting phase is performed correctly. toring system to prevent these from entering All wineries and vineyard holdings should the winemaking process. thus devise an inspection program to periodically check for visible signs of contamination 3.2.3.11. TASK 11: VERIFY THE HACCP SYSTEM and establish the frequency with which labora- (HACCP PRINCIPLE 6) tory tests for A. carbonarius and OTA should be One of the most important ways to check the performed. In all other stages of the production effectiveness of an HACCP system is through chain, application of GMPdjust one element in a verification audit, which consists of a system- the HACCP systemdshould be sufficient to atic, independent inspection to check that all prevent the proliferation of A. carbonarius and actions are being correctly documented (anal- thus protect against undesirable levels of OTA. ysis of documentation) and that the system is being implemented as documented (analysis of 3.2.3.10. TASK 10: ESTABLISH CORRECTIVE HACCP records). Accordingly, procedures for ACTIONS (HACCP PRINCIPLE 5) validating each CCP should be established and If the monitoring activities in place determine the effectiveness of the overall system checked that the critical limits are not being met (indi- on the basis of quantitative analyses of OTA cating that the process is out of control), correc- content in representative samples taken from tive actions should be implemented. These batches of grapes prior to crushing and wine actions should assume a worst-case scenario after production. Three-monthly audits are rec- yet be based on an evaluation of hazards, risks, ommended for other plant-based products at severity, and the intended use of the product. risk of contamination by mycotoxins (FAO, Personnel should receive suitable training in 2003). the application of corrective measures, which The HACCP system should be verified peri- should ensure that control of the CCP is regained odically by an individual designated for this and that the affected raw materials or products purpose. Microbiological or chemical tests can are isolated and discarded if necessary. Wher- be used to ensure that the system is under ever possible, an alarm system should be put control and that the product meets customer in place to warn personnel that a critical limit specifications. These tests will enable verifica- is being approached. Suitable corrective actions tion of the suitability of the CCPs and control applied at this point should avoid deviation measures in place and of the scope and efficacy from the critical limits and prevent product loss. of the monitoring procedures. An internal audit- There are two kinds of corrective actions: ing plan, as well as being an essential tool for those aimed at regaining control (e.g., discard- verifying the effectiveness of the HACCP ing batches of grapes with excessive OTA levels) system, will also document ongoing efforts to and those aimed at isolating the product keep the HACCP up to date. 336 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING

The HACCP system can be verified in the improvements. The implementation of trace- following ways: ability systems that include batch identification right back to the vineyard enables specific By taking samples for analysis using 10. OTA-contaminated batches of product to be a method other than that used for located for elimination purposes and so avoids monitoring purposes losses associated with the unnecessary elimina- By talking to personnel, especially the 11. tion of uncontaminated batches. Proper docu- person in charge of monitoring the CCPs mentation can also provide legal evidence By observing operations at the CCPs 12. of due diligence regarding food safety By commissioning external audits from an 13. management. independent auditor Records should at least include all documen- It is important to emphasize that the applica- tation related to processes, GMP and GHP, CCP tion of a generic HACCP system is not viable as monitoring, compliance with critical limits, each HACCP system must be adapted to the deviations, and corrective actions. specific formulation, handling, and preparation methods for the product in question. Periodic product tests aimed at checking that Acknowledgments acceptable limits have not been exceeded We thank the Spanish Ministry of Science and Innovation should be performed. If limits are exceeded, it (MICINN) (grants AGL2006-02558, AGL2009-07894 and should be possible to detect where the system Consolider INGENIO 2010 CSD2007-00063 FUN-C-FOOD) failed and to identify at which point control and the Comunidad de Madrid (CAM) (S2009/AGR-1469) was lost. In such a case, it may be necessary to for financial support. change critical limits or to validate and introduce new control measures. Changes should References also be made if a study of deviations and product dispositions reveals an unacceptable Ame´zqueta, S., Gonza´lez-Penãs, E., Murillo-Arbizu, M., & degree of control at a particular CCP. Lo´pez de Cerain, A. (2009). Ochratoxin A decontami- nation: A review. Food Control, 20, 326e333. If OTA concentrations in the end product Battilani, P., & Pietri, A. (2002). Ochratoxin A in grapes and exceed the limits established by law, the trace- wine. Eur. J. Plant Pathol., 108, 639e643. ability and record-keeping system will enable Bau, M., Bragulat, M. R., Abarca, M. L., Minguez, S., & the defective batch to be traced and will also Caban˜es, F. J. (2005). Ochratoxigenic species from indicate the CCP where control was lost. This Spanish wine grapes. Int. J. Food Microbiol., 98, 125e130. Bellı´, N., Marıń, S., Coronas, I., Sanchis, V., & Ramos, A. J. CCP should, if necessary, be modified. (2007). Skin damage, high temperature and relative humidity as detrimental factors for Aspergillus carbonar- 3.2.3.12. TASK 12: KEEP RECORDS (HACCP ius infection and ochratoxin A production in grapes. PRINCIPLE 7) Food Control, 18, 1343e1349. Record-keeping is essential to the correct Bellido, J. B., Goza´lez, F., Arnedo, A., Galiano, J. V., Safont, L., Herrero, C., et al. (1996). Brote de infeccioń application of the HACCP system, as it demon- alimentaria por Salmonella enteritidis. Posible efecto strates that procedures have been followed protector de las bebidas alcoho´licas. Med. Clin., 107, appropriately, critical limits have been 641e644. respected, monitoring has been adequate, and Blesa, J., Soriano, J. M., Molto´, J. C., & Man˜es, J. (2004). corrective actions have been implemented Concentration of ochratoxin A in wines from super- markets and stores of Valencian Community (Spain). where necessary. Record-keeping also enables J. Chrom. A., 1054, 397e401. problematic aspects of the system to be docu- Bleve, G., Grieco, F., Cozzi, G., Logrieco, A., & Visconti, A. mented with a view to implementing continual (2006). Isolation of epiphytic yeasts with potential for REFERENCES 337

biocontrol of Aspergillus carbonarius and A. niger on European Union (EU). (2002a). Commission regulation EC No grape. Int. J. Food Microbiol., 108, 204e209. 472/2002 of 12 March 2002 amending regulation (EC) No. Brenner, H., Rothenbacher, D., Bode, G., & Adler, G. (1999). 466/2001 setting maximum levels for certain contami- Inverse graded relation between alcohol consumption nants in foodstuffs. Off. J. Europ. Comm. March, 16. and active infection with Helicobacter pylori. Am. J. Epi- European Union (EU). (2002b). Reglamento (CE) no 257/ demiol., 149, 571e576. 2002 de la Comisioń de 12 de Febrero de 2002. Off. J. ´ Briones, A. I., & Ubeda, J. F. (2001). Elaboracioń de un plan Europ. Comm., L 41/12e15. APPCC en una bodega. Tecnol., Vin 89e93, May/June. European Union (EU). (2003). EU food law news. Contam- Bujanda, L. (2000). The effects of alcohol consumption upon inants-Mycotoxins-EC Permitted levels. Retrieved the gastrointestinal tract. Am. J. Gastroenterol., 95, August 26, 2010 from . Burdaspal, P. A., & Legarda, T. M. (1999). Ocratoxina A en European Union (EU). (2004). EU rules on ochratoxin A vinos, mostos y zumos de uva elaborados en Espanãy extended to coffee, wine and grape juice. Off. J. Europ. en otros paı´ses europeos. Alimentaria, 36, 107e113. Comm., IP/04/1215. Caban˜es, F. J., Accensi, F., Bragulat, M. R., Abarca, M. L., European Union (EU). (2005). Reglamento (CE) No 123/ Castella´, G., Minguez, S., et al. (2002). What is the source 2005 de la comisioń de 26 de enero de 2005 por el que of ochratoxin A in wine? Int. J. Food Microbiol., 79, se modifica el reglamento (CE) no 466/2001 con 213e215. respecto a la ocratoxina A. Off. J. Europ. Comm., L 25/ Carrascosa, A. V. (2005). APPCC en la elaboracıón del vino: 3e5. Ocratoxina A. In Microbiologia del vino. A. V. Carrascosa, Federacioń de Industrias de Alimentacioń y Bebidas (FIAB). R. Munõz, R. Gonza´lez coord. Chap 13, 361e380. Ed. (1997). Aplicacioń del sistema de ana´lisis de riesgos y control AMV, Madrid. de puntos crı´ticos en vinos. Valencia, Spain: Semana Coker, R. D. (1997). Mycotoxins and their control: Constraints Vitivinıćola. and opportunities. NRI Bull. 73. Chatham, UK: Natural Food and Agriculture Organization (FAO). (1993). Guide- Resources Institute. lines for the application of the HACCP system. In Colova-Tsolova, V., Perl, A., Krastanova, S., Tsvetkov, I., & Training considerations for the application of the HACCP Atanassov, A. (2001). Genetically engineered grape system to food processing and manufacturing, 11 Geneva, for disease and stress tolerance. In K. A. Roubelakis- Switzerland: FAO/WHO/FNU/FOS/93.3. Angelakis (Ed.), Molecular Biology & Biotechnology of the Food and Agriculture Organization (FAO). (1997). HACCP Grapevine (pp. 411e432). Amsterdam, the Netherlands: system and guidelines for its application. Annex to Kluwer Publ. Co. CAC/RCP 1e1969, Rev. 3. In Codex Alimentarius food del Prete, V., Rodrı´guez, H., Carrascosa, A. V., de las hygiene basic texts. Rome, Italy: FAO/OMS. Rivas, B., Garcıá-Moruno, E., & Munõz, R. (2007). In vitro Food and Agriculture Organization (FAO). (1998a). Position removal of OTA by wine lactic acid bacteria. J. Food. paper on ochratoxin A, Codex Alimentarius Comission, Protect., 70, 2155e2160. Codex Committee on Food Aditives and Contaminants, Delage, N., d’Harlingue, A., Colonna Ceccaldi, B., & thirty-first session. The Hague, the Netherlands, March Bompeix, G. (2003). Occurrence of mycotoxins in fruit 22e26, 1999, CX/FAC 99/14. juices and wine. Food Control, 14, 225e227. Food and Agriculture Organization (FAO). (1998a). Food Desenclos, J. C. A., Klontz, K. C., Wilder, M. H., & Jun, R. A. quality and safety systems e A training manual on food (1992). The protective effect of alcohol on the occurrence hygiene and the Hazard Analysis and Critical Control Point of epidemic oyster-borne hepatitis A. Epidemiol., 3, (HACCP) system. Rome, Italy: Food Quality and Stan- 371e374. dards Service Food and Nutrition Division. European Commission (EC). (2002). Directive 2002/32/EC Food and Agriculture Organization (FAO). (2002). Report on of the European Parliament and of the Council of 7 May meetings of expert committees and study groups. Evaluation 2002 on undesirable substances in animal feed. Coun. of certain micotoxins. EB110/6, 110th session, provisional Stat. Off. J., L140,10e22. agenda item 7.1, joint FAO/WHO Expert Committee on Food European Food Safety Authority (EFSA). (2006). Opinion of Additives. Fifty-sixth report. Geneva, Switzerland, the Scientific Panel on Contaminants in the Food Chain on February 6e15. a Request from the Commission related to ochratoxin A Food and Agriculture Organization (FAO). (2003). Manual in food. Question N EFSA-Q-2005e154. Adopted April sobre la aplicacioń del sistema de ana´lisis de peligros y de 4, 2006. EFSA J, 365. Retrieved August 26, 2010 from puntos crı´ticos de control (APPCC) en la prevencioń y control . cioń y Nutricioń no 73. 338 13. APPLICATION OF THE HAZARD ANALYSIS AND CRITICAL CONTROL POINT SYSTEM TO WINEMAKING

Frisvad, J.C. & Thrane, U. (1996). Mycotoxin production by Kourtis, L. K., & Arvanitoyannis, I. S. (2001). Implementa- food-borne fungi. (5th ed.). In R. A. Samson, E. S. tion of Hazard Analysis Critical Control Point (HACCP) Hoekstra, J. C. Frisvad & O. Filtenborg (Eds.), Introduc- system to the alcoholic beverages industry. Food Rev. tion to food-borne fungi (pp. 251e260). Centraalbureau Inter., 17,1e44. voor Schimmelcultures, Baarn, the Netherlands. Lo´pez de Cerain, A., Gonza´les-Peoàs, E., Jimenez, A. M., Gallo, A., Perrone, G., Solfrizzo, M., Epifani, F., Abbas, A., & Bello, J. (2002). Contribution to the study of ochra- Dobson, A. D. W., et al. (2009). Characterisation of a pks toxin A in Spanish wines. Food Addit. Contam., 19, gene which is expressed during ochratoxin A produc- 1058e1064. a tion by Aspergillus carbonarius. Int. J. Food Microbiol., 129, Marimoń, J. M., Bujanda, L., Gutie´rrez-Stampa, M A., 8e15. Cosme, A., & Arenas, J. (1998). Antibacterial activity of Ganãń, M., Martıńez-Rodrı´guez, A. J., & Carrascosa, A. V. wine against Salmonella enteritidis pH or alcohol? J. Clin. (2009). Antimicrobial activity of phenolic compounds of Gastroenterol., 27, 179e180. wine against Campylobacter jejuni. Food Control, 20, Martıńez-Rodrı´guez, A. J., & Carrascosa, A. V. (2009). 739e742. HACCP to control microbial safety hazards during Garcıá-Moruno, E., Sanlorenzo, C., Boccaccino, B., & di winemaking: Ochratoxin A. Food Control, 20(5), 469e475. Stefano, R. (2005). Treatment with yeast to reduce the Mateo, R., Medina, A., Mateo, E. M., Mateo, F., & concentration of ochratoxin A in red wine. Am. J. Enol. Jimeńez, M. (2007). An overview of ochratoxin A in beer Viticult., 56,73e76. and wine. Int. J. Food Microbiol., 119,79e83. Go´mez, C., Bragulat, M. R., Abarca, M. L., Mıńguez, S., & Morassut, M., & Cecchini, F. (1999). HACCP in oenological Caban˜es, F. J. (2006). Ochratoxin A-producing fungi industry: A system of self-control not only for hygienic- from grapes intended for liqueur wine production. Food sanitary hazards. It. Food Bever. Technol., 16,8e12. Microbiol., 23, 541e545. Mortimore, S. & Wallace, C. (2001). HACCP: Enfoque Gottardi, G. (1997). Come utilizzare in modo semplice il praćtico. Acribia, Zaragoza, Spain. sistema HACCP nel settore enologico. Enotecnico, 33, National Advisory Committee on Microbiological Criteria 20e27. for Foods (NACMCF). (1992). Hazard analysis and crit- Hyginov, C. (2000). Elaboracioń de vinos. Seguridad-calidad- ical control point system. Int. J. Food Microbiol., 16,1e23. me´todos. Introduccioń al HACCP y al control de los defectos. Nu´ n˜ez, Y. P., Pueyo, E., Carrascosa, A. V., & Martıńez- Acribia, Zaragoza, Spain. Rodrı´guez, A. J. (2008). Effects of aging and heat treat- International Commission on Microbiological Specifications ment on whole yeast cells and yeast cell walls and on the for Foods (ICMSF). (1991). El sistema de ana´lisis de adsorption of OTA in a wine model system. J. Food riesgos y puntos crı´ticos. Acribia, Zaragoza, Spain. Protect., 71, 1496e1499. International Commission on Microbiological Specifications Olivares-Marıń, M., del Prete, V., Garcia-Moruno, E., for Foods (ICMSF). (1996). Microorganisms in Foods Fernańdez-Gonza´lez, C., Macıás-Garcıá, A., & Go´mez- V. Microbiological specifications of food pathogens. London, Serrano, V. (2009). The development of an activated UK: Chapman & Hall. carbon from cherry stones and its use in the removal International Organisation of Vine and Wine (OIV). (2002). of ochratoxin A from red wine. Food Control, 20, CST 1/2002 adoptada en 2002 por los Estados miembros 298e303. de la OIV en ocasioń de la Asamblea General de Oliveri, C., Torta, L., & Catara, V. (2008). A polyphasic Bratislava (Eslovaquia). approach to the identification of ochratoxin A-producing International Organisation of Vine and Wine (OIV). (2005). black Aspergillus isolates from vineyards in Sicily. Int. Resolution VITI-OENO 1/2005. Paris, France, J. Food Microbiol., 127, 147e154. October 25. Pe´rez de Obanos, A., Lo´pez de Cerain, A., Jimeńez, A. M., Kapetanakou, A. E., Panagou, E. Z., Gialitaki, M., Gonza´lez-Penãs, E., & Bello, J. (2001). Ocratoxina A en Drosinos, E. H., & Skandamis, P. N. (2009). Evaluating plasma humano: Nuevos datos de exposicioń en Espanã. the combined effect of water activity, pH and tempera- Rev. Toxicol., 18,19e23. ture on ochratoxin A production by Aspergillus ochraceus Pfeiffer, A., Holgl, B., & Kaess, H. (1992). Effect of ethanol and Aspergillus carbonarius on culture medium and and commonly ingested alcoholic beverages on gastric Corinth raisins. Food Control, 20, 725e732. emptying and gastrointestinal transit. Clin. Investig., 70, Kikkert, J. R., Thomas, M. R., & Reisch, B. I. (2001). Grape- 487e491. vine genetic engineering. In K. A. Roubelakis-Angelakis Pitt, J. I. (1996). What are mycotoxins? Australian Mycotoxin (Ed.), Molecular biology and biotechnology of the grapevine Newsletter, 7(4), 1. (pp. 387e404). Dordrecht, the Netherlands: Kluwer Rousseau, J. (2004). Ocratoxina en los vinos: Estado de los Academic Publishers. conocimientos. Vinidea. Net. Internet Tech. J., 1e4. REFERENCES 339

Ruggiero, P., Tombola, F., Rossi, G., Pancotto, L., Lauretti, L., chromatography mass selective detection. J. Agric. Food del Giudice, G., et al. (2006). Polyphenols reduce gastritis Chem., 49, 2733e2740. induced by Helicobacter pylori infection or VacA toxin Turner, N. W., Subrahmanyamb, S., & Piletsky, S. A. (2009). administration in mice. Antimicrob. Agents Chemother., 50 Analytical methods for determination of mycotoxins: A (7), 2550e2552. review. Anal. Chim. Act., 632, 168e180. Samson, R.A., Hoekstra, E.S., Frisvad, J.C., & Filtenborg, O. Varga, J., & Kozakiewicz, Z. (2006). Ochratoxin in grapes (1996). Introduction to food-borne fungi (5th ed.). Cen- and grape-derived products. Trends Food Sci. Technol., 17, traalbureau voor Schimmelcultures, Baarn, the Nether- 72e81. lands, p. 322. Vinas, I., Dadon, J., & Sanchis, V. (1993). Citrinin-producing Selma, M. V., Martı´nez-Culebras, P. V., & Aznar, R. (2008). capacity of Penicillium expansum strains from apple Real-time PCR based procedures for detection and packinghouses of Lerida (Spain). Int. J. Food Microbiol., quantiﬁcation of Aspergillus carbonarius in wine grapes. 19, 153e156. Int. J. Food Microbiol., 122, 126e134. Vivier, M. A., & Pretorius, I. S. (2000). Genetic improvement Serra, R., Abrunhosa, L., Kozakiewicz, Z., & Venanci, A. (2003). of grapevine: Tailoring grape varieties for the third Black Aspergillus species as ochratoxin A producers in millennium. S. Afr. J. Enol. Vitic., 21,5e26. Portuguese wine grapes. Int. J. Food Microbiol., 88,63e68. Weisse, M. E., Eberly, B., & Person, D. A. (1995). Wine as Sheth, N. K., Wisniewski, T. R., & Franson, T. R. (1988). digestive aid: Comparative antimicrobial effects of Survival of enteric pathogens in common beverages: An bismuth salicylate and red and white wine. BMJ, 311, in vitro study. Am. J. Gastroenterol., 83, 658e660. 1657e1660. Soleas, G. J., Yan, J., & Goldberg, D. M. (2001). Assay of Zimmerli, B., & Dick, R. (1996). Ochratoxin A in table wine ochratoxin A in wine and beer by high-pressure and grape-juice: Occurrence and risk assessment. Food liquid chromatography photodiode array and gas Add. Contam., 13, 655e668. CHAPTER 14

Applied Wine Microbiology Braulio Esteve-Zarzoso 1, Mireia Martıńez 2, Xavier Rubires 3, Marıá Yuste-Rojas 2, Mireia Torres 2 1 Dpt Bioquimica i Biotecnologia, Universitat Rovira i Virgili, Tarragona, Spain, 2 Miguel Torres, Barcelona, Spain and 3 Pago Jean Leoń, Barcelona, Spain

OUTLINE

1. Introduction 342 7.1. Steps to Take in the Event of a Stuck Fermentation (Anchor Yeast, 1994) 348 2. Microbiological Control of Grapes 342 8. Monitoring Contamination by Undesirable 3. Inoculation Methods 343 Microorganisms 348 3.1. Direct Inoculation 344 8.1. Yeast-induced Wine Alterations: 3.2. Preparation of Pied de Cuve Cultures Precautionary Measures 348 and Calculation of Inoculation Rates 344 8.2. Spoilage by Lactic Acid Bacteria: 3.2.1. Preparation of a Pied de Cuve Precautionary Measures 350 Culture for Alcoholic 8.3. Spoilage by Acetic Acid Bacteria: Fermentation 344 Precautionary Measures 350 3.2.2. Preparation of a Pied de Cuve Culture for Malolactic 9. Microbiological Aspects of the Production Fermentation 345 of Typical Wines (Sherry, Cava) 350 9.1. Fino Wines 351 4. Molecular Methods for Analyzing the 9.2. Cava 352 Microorganisms used in the Winery 345 10. Microbiological Quality Control of the 5. Quality Control Analysis of Commercial Final Product 352 Yeasts and Inoculation 346 11. New Challenges Facing the Scientiﬁc 6. Monitoring the Establishment of Community: Genetically Modiﬁed Inoculated Lactic Acid Bacteria 347 Organisms (GMOs) 353 7. Rapid Solutions for Stuck Fermentations 347

Molecular Wine Microbiology Doi: 10.1016/B978-0-12-375021-1.10014-1 341 Copyright Ó 2011 Elsevier Inc. All rights reserved. 342 14. APPLIED WINE MICROBIOLOGY 1. INTRODUCTION To do this, it is necessary to carefully inspect and protect the grapes from the moment they start to Earlier chapters of this book have explored grow to the moment they are harvested and varying aspects of wine microbiology in excep- transported to the winery. The optimal time tional detail. Although the information covered for microbial growth in grapes is the ripening in those chapters provides a good indication of stage as this is when the grape’s protective what is actually happening in today’s wineries, barrier, the skin, is most likely to be broken, it is not always readily applicable to the “real leading to the release of sugars onto the surface world.” Small and medium-sized wineries often of the berry and the proliferation of different follow their intuition or continue to do what has types of microorganism. worked for them in the past. The aim of this The most relevant filamentous fungi are chapter is to provide some general guidance those that produce metabolites with a negative on the practical application of microbiology to effect on wine quality, irrespective of their the winery but without losing sight of the abundance. The best known grapevine fungus extreme importance of keeping facilities and is Botryotinia fuckeliana (anamorphic state, equipment as clean as possible and of closely Botrytis cinerea), which can have both positive following protocols. Many problems will be and negative effects on the wine. When this avoided if these two principles are followed. fungus grows inside the berries (producing The two main cornerstones of the practices of what is known as noble rot), it removes water any winery should be to ensure careful, proper from the fruit and, thus, increases the concen- handling throughout the process, from the vine- tration of compounds that determine the yard to the bottle, and to never lose sight of the primary aroma of the wine. When the fungus fact that wine is intended for human consump- affects the surface of the grape, however, it is tion; any departure from these basic premises known as gray mold. This form of the fungus can cause problems (Garijo, 2008). Another produces b-glucans, which ultimately interfere important consideration is that wineries are with wine clarification and filtration opera- interested in quick, simple, and affordable solutions. Other filamentous fungi found on grapes tions that do not significantly impact the quality are Cladosporium, Mucor,andRhizopus species, of the final product. By adhering to the recom- but they do not have a significant bearing on mendations of the International Organisation the fermentation process (Fleet, 1992). Special of Vine and Wine (OIV) regarding additives attention should also be paid to fungi that and to legislative requirements regarding produce toxins that can be passed into wine. contaminants (pesticides, heavy metals, toxic For example, ochratoxin A, which is produced substances, etc.) in countries to which the wine by Aspergillus and Penicillium species, is is to be exported, wineries will avoid many becoming an increasing concern in the wine- problems related to the sale of their products. making community. This metabolite has been detected at different concentrations in a range of wines from different regions (Belli et al., 2. MICROBIOLOGICAL CONTROL 2005; Solfrizzo, 2008). Chapter 13 describes OF GRAPES how to apply a hazard analysis and critical control point plan to control ochratoxin A One of the keys to avoiding problems in the levels in wine. This procedure is particularly winemaking process is to minimize the grape relevant in view of the recent European Union microflora, as this will prevent the development (EU) legislation establishing maximum allow- of undesirable microorganisms from the outset. able levels for this toxin in food products. INOCULATION METHODS 343

Numerous studies have analyzed the pres- recommended course of action on the detection ence of yeast on the surface of grapes (de of bacteria is to isolate and clean the affected Andre´s-de Prado et al., 2007; Fleet & Heard, material with suitable disinfectants (alkaline 1992) and many have indicated that Saccharo- disinfectants followed by water and acid-based myces cerevisiae is present only in very small disinfectants). numbers on healthy grapes (Martini, 1993; Pre- In any case, the best strategy for preventing torius, 2000). Most of the yeasts in such cases the proliferation of undesirable microorganisms are aerobic species, the most common of which in the winery is to prevent their growth in the are those belonging to the genera Candida, Hanse- vineyard. Vineyards should thus be designed niaspora, Kluyveromyces, Pichia, and Rhodotorula. or adapted to facilitate the application of phyto- The grape microflora, however, varies greatly sanitary protection products and prevent over- depending on factors such as geographical loca- crowding of grape clusters. Another effective tion, rainfall, and temperature (Longo et al., 1991; measure is to remove grapes before they become Parrish & Carroll, 1985), fungicide use (Monteil too big as this is when the risk of skin breakage et al., 1986), soil type (Farris et al., 1990; Poulard is greatest. et al., 1980), vineyard age, grape variety, and harvesting method (Martini et al., 1980; Pretorius et al., 1999; Rosini et al., 1982). Other variations 3. INOCULATION METHODS in microflora can be introduced by the sampling procedure used. Vaughan-Martini and Martini Yeasts, whether naturally present or deliber- (1995), for example, reviewed the differences ately added to the must, have been an essential generated by sampling methods according to part of the winemaking process since time whether or not the sample was enriched prior immemorial. While wine can certainly be to analysis. made with naturally occurring populations of What is certain is that grapes should be yeast, the demand for greater control over phys- handled as gently as possible, as the slightest ical, chemical, and indeed microbiological prop- pressure on the berry could cause the release erties has led to the increasing use of inoculated of juice containing sugars that will favor the strains. Spontaneous, or natural, fermentation is growth of the yeast that come into contact performed primarily by S. cerevisiae yeast strains with it. Adequate hygiene and sanitation stan- (Amerine & Kunkee, 1968), although other dards in the winery are extremely important species may participate in the process (Torija as any lapses will lead to the instant prolifera- et al., 2001) and alter the properties of the final tion of S. cerevisiae yeasts that come into contact product. with these sugars. Accordingly, efforts should To activate fermentation, wineries can use be made to ensure that harvesting equipment either commercial preparations of yeast (or and grape reception facilities are kept as clean bacteria in the case of malolactic fermentation) as possible. To this end, clean containers should or pied de cuve cultures. The use of commercial be used to transport the grapes to the winery, cultures is now widespread as they are conve- and grape reception facilities should be hosed nient and easy to use, and provide guarantees down if used continually or cleaned with disin- about the origin of the yeasts. These prepara- fectants if the arrival of grapes is intermittent. tions contain large numbers of viable cells and Bacteria grow in the same conditions as are used to ensure the rapid establishment of yeast, which explains why acetic and lactic the selected species during fermentation. They acid bacteria can proliferate in the musts of may, however, also contain a number of contam- wineries with poor hygiene conditions. The inating microbes (Radler & Lotz, 1990). 344 14. APPLIED WINE MICROBIOLOGY

By seeding the must with known microor- rehydration should be twice the volume of the ganisms, today’s winemakers have succeeded water used as the mix increases in size during in ensuring that fermentation will primarily be the process. The next step is to vigorously mix conducted by strains with desirable properties. the solution until it is uniform and ready to be To ensure quality, however, and indeed consis- inoculated. The temperature of the must is crit- tency from one year to the next, it is necessary ical at this stage as yeast viability can be seri- to control microbial activity during this process ously compromised by shifts in temperature of (Martini & Vaughan-Martini, 1990). more than 10C. To safeguard against problems of this nature, an intermediate thermal condi- 3.1. Direct Inoculation tioning step is recommended. Using standard commercial preparations, The success of inoculation, whether per- which have a viability of approximately formed with active dried yeast or starter 1010 colony-forming units (CFU)/g, the above cultures, depends largely on correct rehydra- procedure will give rise to a yeast population tion, but there are also a number of other factors of 2 Â 106 CFU/mL, which is used to initiate to bear in mind prior to inoculation. The grape the exponential growth phase (Degre, 1992). juice must, for example, contain only a small Commercial Oenococcus oeni starters are also population of resident microorganisms. This available for activating malolactic fermentation can be achieved through static clarification and via direct inoculation. The rehydration phase flotation (which achieve a 50e80% reduction is also critical in this case, and it is important in the native yeast population), centrifugation to follow the manufacturers’ instructions. (60e90% reduction), and vacuum filtration (99% reduction). The inoculation tank must 3.2. Preparation of Pied de Cuve also be clean, as any microorganisms present Cultures and Calculation of Inoculation could proliferate and compete with the inoculated strains. Another important factor is the Rates time that elapses between harvesting and inocu- The must or wine needs to be inoculated lation, as, the longer this time, the greater the quickly with the selected microorganisms to proportion of autochthonous yeasts that will prevent the growth of unwelcome competitors. thrive and the fewer the nutrients that will be This is achieved by omitting the conditioning available for the inoculated strains. Further- stage (in which the yeasts acclimatize to the more, the temperature of the must should be new culture medium) and adding the yeasts kept as low as possible as high temperatures directly to the must at the height of their meta- can favor the proliferation of naturally occur- bolic activity; that is, during the exponential ring populations. growth phase (Fleet & Heard, 1992). The yeast rehydration process is generally similar across different commercial prepara- 3.2.1. Preparation of a Pied de Cuve tions, with seeding always performed at Culture for Alcoholic Fermentation a density of 20 g/hL. The dry yeast is rehy- A pied de cuve culture is simply a continuous drated in a volume of very warm water (37C) fermenter in which the aim is to maintain a large at a concentration of 10% weight by volume. population of yeast in the growth phase. To do This mixture is then shaken gently for 10 min this, it is necessary to periodically remove liquid and left to rest for an additional 10 min from the fermenter tank and add must (as (maximum 30 min) to allow time for the yeast a source of nutrients) so that the yeast popula- to become rehydrated. The vessels used for tion can continue to proliferate. The liquid that MOLECULAR METHODS FOR ANALYZING THE MICROORGANISMS USED IN THE WINERY 345 is removed can be used to inoculate the must in available nutrient preparations that can help to the fermentation tank. As mentioned above, the stimulate bacterial growth. The traditional prac- fermentation conditions in the tank to which the tice in wineries has been to use wine already active dried yeast has been added must be undergoing malolactic fermentation to inoculate maintained to ensure that the yeast population wine that has not yet entered this stage. Calcu- is constantly in the exponential growth phase. lating the inoculation rate in malolactic fermen- The temperature of the tank must be maintained tation is more complicated than in alcoholic at approximately 17C and it is often advanta- fermentation, however, as there is no quick geous to add nutrients or fermentation activa- method available for counting the number of tors at the moment of inoculation, particularly viable cells (Blackburn, 1984). Instead, fermenta- in the case of musts with depleted ammoniacal tion kinetics must be used as an indicator for the nitrogen supplies. calculation of the appropriate inoculation rate. Strict monitoring of temperature in the pied de cuve tank is essential, and population numbers should be monitored using a Neubauer 4. MOLECULAR METHODS FOR chamber until a level of between 5 Â 107 and ANALYZING THE 1 Â 108 cells/mL is reached. This is the moment MICROORGANISMS USED IN THE at which the inoculation rate required to achieve WINERY an initial culture of 2 Â 106 cells/mL can be calculated. A population level of between 5 Â Advances in technology have greatly 7 8 10 and 1 Â 10 cells/mL in the pied de cuve improved the control that winemakers have tank corresponds to an inoculation rate of over the yeasts that participate in the fermenta- between 2 and 5% in the fermentation tank. tion process. Modern molecular methods, in Microscopic observation of the physiological particular, offer rapid results, but other impor- state of yeasts in the pied de cuve culture can tant features are ease of use and affordability provide valuable information as a very large (Andorra`, 2008). Of the range of techniques population with a very low proportion of described in the literature, mitochondrial DNA actively budding yeasts indicates that the yeasts (mtDNA) restriction analysis and random are nearing the lag phase, whereas a very large amplification of polymorphic DNA (RAPD) population containing a high proportion of are becoming increasingly common in the wine- actively budding yeasts indicates that the yeasts making industry. are in the middle of their exponential growth mtDNA restriction analysis, which is widely phase. used to analyze yeasts, does not require extensive technical skills or a significant outlay of 3.2.2. Preparation of a Pied de Cuve capital, and has the added advantage that it Culture for Malolactic Fermentation provides rapid results. This means that correc- The procedure for preparing a pied de cuve tive measures can be taken in the early stages. starter for malolactic fermentation is much the mtDNA restriction profiles, for example, are same as that used for alcoholic fermentation available on the same day because the DNA but a longer conditioning stage is required can be isolated directly from the tank, with no (Champagne et al., 1989). To ensure that the need for prior culture. lactic acid bacteria in the pied de cuve tank RAPD, which is used to analyze bacteria, is continue to proliferate, it is necessary to add slightly more complicated as it requires the use wine from the tank awaiting malolactic fermen- of amplification products and thermocyclers, tation. There are numerous commercially which add somewhat to the cost. A further 346 14. APPLIED WINE MICROBIOLOGY disadvantage is that the material needs to be the yeasts. The same methods can also be used correctly identified and kept separate from other to check batch uniformity and to test for loss material, which is not very practical in a winery of viability during storage. laboratory. While RAPD has been described as The yeast strains should also be character- an ideal method for typing O. oeni strains ized, as different strains behave differently (Reguant & Bordons, 2003), it requires strict during the fermentation process and thus lend adherence to standardization protocols, which different organoleptic properties to the end again is not very practical in a winery setting. product. A range of methods exists for identi- Furthermore, results take several days to be fying and differentiating between yeast strains processed. at the genus and species level (Esteve-Zarzoso Samples taken for analysis must be statisti- et al., 1998). Several of these methods have cally representative, regardless of the type of been described in previous chapters, but the microorganism being studied. In other words, most reliable, rapid, and economic option for the information obtained using the sampling wineries is the mtDNA restriction analysis tech- technique must provide a true picture of what nique described by Querol et al. in 1992 and is happening in the fermentation tank. In the modified by Lo´pez et al. in 2001. Furthermore, case of alcoholic fermentation, must samples the test does not require skilled personnel. collected 48 h after inoculation will provide Several studies have described how different sufficient information with which to assess the yeast preparations supplied by different success of the operation. In the case of malo- providers with different instructions for use all lactic fermentation, however, samples need to contained the same strain that had been charac- be taken at different time points, particularly terized using different molecular methods in the early stages of fermentation, as this is (Fernańdez-Espinar et al., 2001). There have when bacteria proliferate. even been cases in which the same strain was found to be sold under different names by different suppliers and at considerably different 5. QUALITY CONTROL ANALYSIS prices. In certain cases, there was a price differ- OF COMMERCIAL YEASTS AND ence of 30%, which is by no means insignificant INOCULATION when it comes to determining the price of the final product. In such cases, in addition to Quality control procedures for use with yeast mtDNA restriction analysis, it is necessary to products obtained from commercial suppliers karyotype the strains to conclusively demon- are becoming increasingly important, particu- strate full uniformity between different commer- larly in view of the wide range of products cial products. Uniformity of both mitochondrial available on the market. The first test that and nuclear markers between samples indicates should be performed when a winery receives that they correspond to the same strain. a new batch of active dried yeast is to check Figure 14.1 shows the chromosomal profile the viability of the strains following rehydra- obtained via pulsed field gel electrophoresis tion. This is done by performing serial dilutions (PFGE) for three commercial preparations with in an appropriate culture medium or by staining identical restriction patterns obtained using with vital dyes. These tests are used to check the restriction enzymes associated with a high level number of viable yeast that will be inoculated of variability in the profiles obtained (AluI, into the must. Viability can vary by up to 30% HinfI, RsaI). from one product to the next, depending on Similar studies have revealed cases in which the drying methods used and the sensitivity of the same supplier was selling the same strain RAPID SOLUTIONS FOR STUCK FERMENTATIONS 347

123 performed by autochthonous yeasts is much more common than is generally thought (Esteve- Zarzoso et al., 2000).

6. MONITORING THE ESTABLISHMENT OF INOCULATED LACTIC ACID BACTERIA

The establishment of inoculated lactic acid bacteria is generally controlled by monitoring the consumption of malic acid and the formation of lactic acid. Nowadays, however, molec- Ref Yeast strainManufacturer €/kg ular biology techniques can also help to 1 A α 32 analyze the establishment of these bacteria β 2 B 25 during malolactic fermentation. Zapparolli χ 3 C 27 et al. (1998) described the use of specific primers FIGURE 14.1 Pulsed field gel electrophoresis (PFGE) of for the gene encoding the malolactic enzyme to three commercial yeasts. The figure shows the chromosomal monitor the establishment of O. oeni strains profile obtained by PFGE for three commercial preparations during malolactic fermentation. In a later study, with an identical restriction pattern obtained using restric- Reguant and Bordons (2003) used multiplex tion enzymes associated with a high level of variability in RAPD-polymerase chain reaction (PCR) to char- the profiles obtained (AluI, Hinf I, RsaI). acterize O. oeni and monitor the population dynamics of the different strains of this species under different names and at different prices, during malolactic fermentation. primarily as part of a marketing strategy. The choice of “one strain or another” can, however, influence the price of the final product, which 7. RAPID SOLUTIONS FOR STUCK is why several wineries perform quality checks FERMENTATIONS on all the yeast products they receive. These controls are performed yearly as the batches The interruption of fermentation that occurs supplied vary from one year to the next. when yeasts stop converting the fermentable Another important aspect of winemaking is sugars in the must is generally known as stuck the analysis of inoculation success. Wines some- fermentation. Treatment (which essentially times develop unexpected organoleptic charac- involves the restoration of ideal fermentation teristics because fermentation is actually conditions) can be difficult, as stuck fermenta- conducted by a native yeast rather than the inoc- tion has a number of possible causes. ulated strain. To prevent this from happening, it is According to Lourens and Reid (2003), the standard practice to take regular, statistically most significant factors that affect yeast viability representative samples throughout the fermenta- are osmotolerance; ethanol tolerance; fermentation process to determine, using mtDNA restriction temperature; availability of nutrients; and tion analysis, whether or not the inoculated presence of medium-chain fatty acids (hexanoic, strain has become established. Studies of this octanoic, and decanoic acids), which have an type have shown that not all inoculated strains inhibitory effect on sugar transport. Other survive in all cases and that fermentation factors include the presence of pesticide traces 348 14. APPLIED WINE MICROBIOLOGY and the effects of prefermentation clarification added during the conditioning phase as they treatments. Musts that undergo extensive clari- may contain toxic substances that could fication have great difficulty fermenting because inhibit the growth of the new culture. they lack sterols and long-chain fatty acids, 3) In red wines in which fermentation has which are both considered survival factors for become stuck, the grape skins must be yeasts. removed if maceration has not been completed to prevent unwanted bacterial 7.1. Steps to Take in the Event of contamination. a Stuck Fermentation (Anchor Yeast, 4) The new tank may be aerated but the old tank 1994) may not, as aeration could trigger the growth of acetic acid bacteria. 1) For every hectoliter of wine, rehydrate 60 to 5) The probability of restarting a stuck 100 g of active dried Saccharomyces bayanus fermentation with a residual sugar level of or S. cerevisiae strains that are capable of under 10 g/L or an alcohol content of 14% or fermentation under difficult conditions and greater is low. The conditioning phase should of degrading fructose. If it is not possible to be started as soon as there is any indication of find a yeast with both properties, use two stuck fermentation. separate types of yeast. 2) When rehydrated, 2 kg of active dried yeast will produce a volume of 20 L. Add 10 L (half the volume of the starter mix) of the 8. MONITORING problematic wine and wait for fermentation CONTAMINATION BY to be activated. UNDESIRABLE MICROORGANISMS 3) The total volume will now be 30 L. Add an additional 15 L of the problematic wine Yeasts, lactic acid bacteria, and acetic acid and wait again for fermentation to start. bacteria can all alter the quality of wine. Most The success of this method will depend on of these microorganisms are already present in whether or not fermentation is activated after the wine and can grow in this ecological niche. each addition of the problematic wine. If it is, Their proliferation at the wrong time, however, this will mean that the new yeast has adapted can lead to the production of metabolites that successfully to the alcohol level of the new can alter sensory quality or even cause adverse medium. health effects. 4) The total volume will now be 45 L. Add equal volumes (e.g., 45, 90, 180, 360 L) of the 8.1. Yeast-induced Wine Alterations: problematic wine to the new culture until all Precautionary Measures the stuck wine has been transferred to the new tank. This process takes at least 2 d. Certain yeasts have been associated with refermentation in sweet wines (Enrique et al., The following aspects should be taken into 2007), namely strains from the species . account: S cerevisiae, Zygosaccharomyces bailii, and Saccharomyces 1) The fermentation temperature should be ludwigii, all of which have high resistance to maintained at between 18 and 22C for white sulfur dioxide and ethanol. The best way to wine and at between 20 and 25C for red wine. prevent refermentation is to ensure appropriate 2) The lees in the tank with fermentation storage conditions; in particular, temperatures problems should under no circumstances be of below 15C, adequate sulphur dioxide levels, MONITORING CONTAMINATION BY UNDESIRABLE MICROORGANISMS 349 protection against the formation of large by phenolic compounds. The richest winesdthat volumes of air in the tanks, and, of course, care- is, those made from ripe grapes and therefore ful cleaning of the tanks prior to use. characterized by high alcohol levels, lower Biofilm-forming yeasts from the genera acidity, and longer maceration timesdare often Pichia, Candida, and Hansenula that grow on the those that are richest in assimilable substrates surface of wines or following contact with air and therefore potentially more amenable to the affect young wines and wines with a low growth of Brettanomyces yeasts. The following alcohol level. To prevent these yeasts from form- points are important for preventing contamina- ing in the bottle, in addition to the measures tion by this yeast: described above, it is necessary to subject the wine to sterile filtration, add 30 mg of free sulfur 1. Particular care should be taken to keep the dioxide, and leave a minimum ullage (air grape reception area clean during harvest pocket) in the bottle. time. Certain wines can develop organoleptic flaws 2. Sulfur dioxide is the only effective antiseptic such as unpleasant odors attributable to sulfur against Brettanomyces species that is compounds during the production process. authorized for use in wine. Levels of There are a number of reasons for this. A molecular (active) sulfur dioxide must be shortage of nitrogen sources in the must, for controlled as, the higher the pH, the more example, leads to the production of hydrogen sulfur dioxide will be needed to maintain sulfide (Jiranek et al., 1995; Park, 2008), while adequate levels of molecular sulfur dioxide. the genotypes of certain yeasts are associated Increases in alcohol content and aging with an increased capacity to produce sulfites. temperatures lead to increased molecular These yeasts can also produce appreciable sulfur dioxide levels and hence greater amounts of sulfur compounds as byproducts protection against Brettanomyces yeasts. The of their metabolism of pesticides. growth of these species is inhibited at levels The acetoin produced by yeasts during of 0.3 parts per million (ppm) of molecular fermentation contributes to the bouquet of the sulfur dioxide, and levels of over 0.5 ppm wine but it is also the precursor of 2,3-butane- lead to the rapid elimination of these yeasts. diol and diacetyl. While 2,3-butanediol can Levels of molecular sulfur dioxide must be contribute to aromatic balance, diacetyl is maintained at between 0.5 and 0.8 ppm to considered a flaw. Saccharomyces yeasts produce protect against Brettanomyces species and only small quantities of acetoin, unlike the apic- other organisms that could alter the quality ulate yeasts Kloeckera and Hanseniaspora and of the wine. Zygosaccharomyces species, which produce 3. Factors that can increase the risk of considerable amounts. contamination and subsequent growth of Among the best-studied contaminating wine Brettanomyces species and other undesirable yeasts are Brettanomyces species, which produce microorganisms include the inadequate four byproducts during growth: esterases, vola- treatment of previously used barrels, the tile fatty acids (acetic acid), volatile phenols (4- storage of wine in unsuitable conditions ethylphenol and 4-ethylguaiacol), vinyl phenols (temperature shifts), the presence of air (4-vinylphenol and 4-vinylguaiacol), and tetra- pockets (excessive oxidation), and hydropyridines. These yeasts are more common insufficient racking (frequency of barrel in red wines, which are rich in cinnamic acid disinfection, etc.). Wood is very porous and precursors and in which the cinnamate decar- can house yeasts in its different layers, which boxylase activity of Brettanomyces is not inhibited makes them difficult to eliminate. 350 14. APPLIED WINE MICROBIOLOGY

Brettanomyces species and lactic acid bacteria 8.3. Spoilage by Acetic Acid Bacteria: are also likely to grow in new barrels. Several Precautionary Measures of the phenolic compounds that are extracted from new wood may provide substrates for Acetic acid bacteria contaminate grapes, musts, and wines (Bartowsky & Henschke, Brettanomyces oxidation or reduction reactions that can give rise to unwanted 2008). To prevent the proliferation of these compounds. In new barrels, there are also microorganisms in stored wine, the correspond- large quantities of cellobiose that are ing tanks and barrels should be filled to the degraded by b-glucosidases, giving rise to maximum and the wine treated with additional sulfites as the levels added during the produc- glucose molecules used by Brettanomyces yeasts for growth. Cellobiose forms in barrels tion process are not sufficient to prevent growth. subjected to toasting. It is also important to A total level of 100 ppm of sulfur dioxide in the know the origin of any new barrels to be used must is necessary for this purpose. The optimal by the winery. temperature for the growth of acetic acid bacteria is 25 to 30C, although activity has been detected at temperatures of close to 10C in certain wineries. Extra caution should be 8.2. Spoilage by Lactic Acid Bacteria: taken to monitor residual populations of these Precautionary Measures bacteria in the wine as they represent a perma- Lactic acid bacteria are common in wineries. nent risk of spoilage. One solution is to refill However, the inoculation of selected strains is tanks and barrels regularly. actually the best strategy for preventing the The excessive growth of acetic acid bacteria proliferation of bacteria capable of producing on grapes can lead to changes in the must that secondary metabolites that can have harmful can interfere with the growth of yeast during health effects (the case of biogenic amines) or alcoholic fermentation and the course of malo- diminish the organoleptic quality of the wine. lactic fermentation. Examples are organoleptic To prevent the growth of unwanted lactic acid changes (caused by the production of undesirable metabolites) and physical alterations to bacteria (Lactobacillus and Pediococcus species) in musts, low sulfite and high pH levels must the wine (e.g., some species are capable be avoided (Pfannebecker & Fro¨hlich, 2008). of producing polysaccharides that interfere Inoculation with high concentrations of with filtration). Aeration is another important commercial malolactic starter cultures can be factor that should be controlled as acetic acid used to displace autochthonous populations bacteria preferentially grow in the presence already present on winery equipment or in of low levels of oxygen. The slightest aeration aging barrels. Similar treatment is also neces- following alcoholic fermentation, for example, 2 3 sary for wines that have already completed can lead to the growth of 10 e10 CFU/mL malolactic fermentation as bacteria that survive of these bacteria. through aging can produce biogenic amines during secondary metabolism. Malolactic fermentation, while common in 9. MICROBIOLOGICAL ASPECTS red wines, is not desirable in certain wines OF THE PRODUCTION OF TYPICAL (especially young whites). To prevent the WINES (SHERRY, CAVA) growth of lactic acid bacteria in such cases, the wine should be filtered and treated with sulfites The types of microorganism used to produce after alcoholic fermentation. a particular wine will depend on the end MICROBIOLOGICAL ASPECTS OF THE PRODUCTION OF TYPICAL WINES (SHERRY, CAVA) 351 9.1. Fino Wines product. Even though the base wine might be the same, different strains of yeast, each with The process used in the fermentation of white unique characteristics, will be used to conduct base wines to make Fino wines is similar to that alcoholic fermentation depending on whether used when activating alcoholic fermentation by the aim is to produce a standard wine, or Cava inoculation. In the former case, however, the or Fino, for example. wine is subsequently fortified and left to age Table 14.1 shows the main properties a yeast (Ibeas et al., 1997; Martıńez et al., 1997; Moreno- should have, depending on the wine being Arribas & Polo, 2008). Although there is contro- made. Below we describe some of the properties versy regarding the molecular differences sought in yeasts used to make Cava and Fino between the four races of flor yeasts used to wines. It should, however, be borne in mind age Fino wine, Esteve-Zarzoso et al. (2004) that all wineries have their own practices that detected a flor yeast strain with a very similar lend their wines their distinctive character. A karyotype to that described by Ibeas et al. key issue in all wineries, however, is the (1997) in a different winery in the same region. viability of the starter cultures used. It is there- This may be because of the high levels of ethanol fore important to ensure a high rate of viability and acetaldehyde present or because of the following rehydration as this is the key to oxidative metabolism of these yeasts. Both successful inoculation. studies concluded that, while all wineries have

TABLE 14.1 Main Properties Sought in Fermentation Yeasts Shown by Type of Wine

Fermentation Fermentation Secondary Biological aging of white wine of red wine fermentation in Cava of Fino wines Fermentation Tolerance of low Moderate resistance Moderate resistance Moderate resistance temperature temperatures preferable Color No production of No degradation No production No production of colored metabolites of color of colored colored metabolites metabolites

Volatile acidity Low production Low production Low production Low/moderate production Degradation of malic High Low High High acid Ethanol tolerance Moderate/high High Moderate/high High

Filtration Necessary Desirable Necessary Not desirable Formation of biofilm Undesirable Undesirable Undesirable Necessary Production of aromas Desirable Desirable Desirable Not necessary Production of/ Low Low Low High tolerance of acetaldehyde Fermentation kinetics Moderate Moderate Low Low 352 14. APPLIED WINE MICROBIOLOGY their own distinctive strains, each barrel is Emsweiler-Rose et al., 1984), molecular tech- a world of its own. niques based on DNA probes (Fitts, 1985; Flowers et al., 1987), and quantitative measure- 9.2. Cava ment systems based on PCR and electrical impedance. Because Cava is produced using two separate The need for rapid microbial detection fermentation steps, efforts should be taken to systems in the food industry is closely linked ensure that the yeasts that participate in each to the benefits that such systems could offer of these steps are compatible. The inoculation companies in terms of improving production of yeasts that participate in secondary fermenta- processes and time to market. The quantitative tion (which occurs in the bottle) must be care- and qualitative analysis of microorganisms by fully conducted to ensure that the yeasts are electrical impedance is based on the indirect perfectly adapted to the high alcohol levels measurement of metabolic activity. These present. The importance of these measures measurements can then be used to calculate cannot be overemphasized. The use of a yeast theoretical population numbers long before the with poor flocculation ability requires longer colony would have become visible on solid filtration times. If a filtration step is not used, culture medium (Deak & Beuchat, 1993). higher quantities of bentonite will be necessary, Electrical impedance is based on the principle resulting in a greater absorption of the compo- that the molecules in the culture medium nents of the wine. If the yeasts are not properly (proteins, carbohydrates, etc.) are electrically acclimatized, the secondary fermentation will neutral or only weakly ionized. The activity of take longer as the yeasts will need more time microorganisms, however, converts these mole- to consume the sugars added for secondary cules into numerous smaller molecules with fermentation. a greater electrical charge and electrical mobility (amino acids, lactates, etc.). These changes can be measured by submerging two electrodes in 10. MICROBIOLOGICAL QUALITY the culture medium (Futschik et al., 1988). CONTROL OF THE FINAL Classical quantification of microbial load PRODUCT requires approximately 3 d but faster results are needed in certain industrial applications. The advances made in classical microbiology Electrical impedance, in this sense, has been methods have mainly involved the develop- described as a rapid, reliable means of moni- ment of specific culture media and biochemical toring the presence of viable microorganisms tests capable of discriminating, both quantita- in food. Martıńez et al. (2004) showed that the tively and qualitatively, between different indirect measurement of carbon dioxide microorganisms. Although widely used, these production by electrical impedance was an ideal techniques have the disadvantage that reliable way of detecting and quantifying yeast popula- results (e.g., from colony counts) are not avail- tions in wine samples; the correlation coefficient able for 2 to 3 d (Pless et al., 1994). This can between this and the classical microbiology increase the price of the final product as stocks method (plate count on specific solid media) of wine are accumulated while the results of was 0.98, with the added advantage that results quality checks on the final product are being (confirmation of the absence of microorganisms) processed. Methods designed to overcome this were available in under 21 h. In the case of problem include bioluminescence imaging, aerobic bacteria, electrical impedance was also immunoassays (d’Aoust & Sewell, 1986; more suitable for detection and quantification REFERENCES 353 purposes, with a correlation coefficient of 0.99 GMOs. This legislation came into force on between this method and the plate count November 7, 2003, with a 6-month adaptation method, and results available in under 15 h period for all operators. (Martıńez et al., 2004). From a commercial perspective, the obligatory labeling of the presence of GMOs or elements from GMOs in the final product deliv- 11. NEW CHALLENGES FACING ered to the consumer, combined with the exis- THE SCIENTIFIC COMMUNITY: tence of strong media and public resistance, GENETICALLY MODIFIED places food companies interested in benefiting ORGANISMS (GMOS) from the advantages of GMOs in a difficult position. From a strictly scientific viewpoint, geneti- The world of wine is strongly traditional. cally modified organisms (GMOs) undoubtedly Major modifications that require considerable offer many health and agricultural advantages changes in age-old practices are not well viewed (Fleet, 2008). by critics or consumers, which makes it very Considering that the full sequence of the S. difficult to start using GMOs. Although these cerevisiae genome is available and that this yeast organisms have enormous potential, they are can be modified with relative ease using the not yet fully accepted by the international increasingly sophisticated genetic modification community. tools that are now widely available, research The future of GMOs in winemaking will centers around the world are in a position to fundamentally depend on public opinion and create a`-la-carte yeasts to meet the requirements ultimately the end consumer. Despite the of the base must or to achieve particular charac- advances that could be made, media campaigns teristics in the end product. have put a stop to the possible industrial break- Within the scope of a strictly scientific, throughs in this field. Historically, basic science enological project, researchers in South Africa has always been one step ahead of the rest of have developed genetically modified yeasts society and today’s winemakers need to be capable of improving the quality of wine ready to deliver products with genetically (Pretorius, 2000). It would be advantageous for enhanced characteristics to the market should winemakers to avail of yeasts capable of a change of opinion come about. increased production of exogenous pectolytic or glycosidic enzymes; bacteria with a low production of histamines; yeasts with low nutri- References tional requirements, low methanol production, Amerine, M. A., & Kunkee, R. E. (1968). Microbiology of or extracellular release of mannoproteins; heat- winemaking. Ann. Rev. Microbiol., 22, 323e358. sensitive microorganisms; etc. Anchor Yeast. (1994). The cellarmaster’s wine yeast and In any case, given the enormous importance fermentation handbook. ` ´ attached to the presence of transgenic products Andorra, I., Landi, S., Mas, A., Guillamon, J. M., & Esteve- Zarzoso, B. (2008). Effect of oenological practices on in food, it should not be forgotten that the microbial populations using culture-independent tech- Spanish state bulletin (BOE, L268/24) published niques. Food Microbiol., 25(7), 849e856. Regulation (EC) No 1830/2003 on the trace- Bartowsky, E. J., & Henschke, P. A. (2008). Acetic acid ability and labeling of GMOs and the traceability bacteria spoilage of bottled red wine e A review. Int. J. of food and feed products produced from Food. Microbiol., 125(1), 60e70. Belli, N., Marıń, S., Duaigu¨ es, A., Ramos, A. J., & Sanchis, V. GMOs, establishing strict regulations regarding (2005). Ochratoxin A in wines, musts and grape juices the approval, labeling, and monitoring of from Spain. J. Sci. Food Agric., 84(6), 591e594. 354 14. APPLIED WINE MICROBIOLOGY

Blackburn, D. (1984). Present technology in the use of (Ed.), Wine microbiology and biotechnology (pp. 1e25). malolactic bacteria. Practical Winery, 5,64e72. Chur, Switzerland: Harwood Academic Publishers. Champagne, C., Gadner, N., & Doyon, G. (1989). Production Fleet, G. H. (2008). Wine yeasts for the future. FEMS Yeast of Leuconostoc oenos biomass under pH control. Appl. Res., 8(7), 979e995. Environ. Microbiol., 55, 2488e2492. Fleet, G. H., & Heard, G. M. (1992). Yeast-growth during d’Aoust, J. Y., & Sewell, A. M. (1986). Detection of salmo- fermentation. In G. H. Fleet (Ed.), Wine microbiology and nellae by the enzyme immunoassay (eia) technique. biotechnology (pp. 27e54). Chur, Switzerland: Harwood J. Food. Sci., 51, 484e507. Academic Publishers. de Andre´s-de Prado, R., Yuste-Rojas, M., Sort, X., Andre´s- Flowers, R. S., Mongola, M. A., Curiale, M. S., Gabis, D. A., Lacueva, C., Torres, M., & Lamuela-Ravento´s, R. M. & Silliker, J. H. (1987). Comparative study of a DNA (2007). Effect of soil type on wines produced from V itis hybridization method and the conventional culture vinifera L. Cv. Grenache in commercial vineyards. procedure for detection of salmonellae in foods. J. Food. J. Agric. Food Chem., 55(3), 779e786. Sci., 52, 781e785. Deak, T., & Beuchat, L. R. (1993). Comparision of con- Futschik, K., Pfu¨ tzner, H., Doblander, A., & Asperger, H. dutimetric and traditional plating techniques for (1988). Automatical registration of microorganism detecting yeast in fruit juice. J. Appl. Bacteriol., 75, growth by a new impedance method. Abst. Int. Meet. 546e550. Chem. Eng. & Biotechnol., Achema, 88,3. Degre, R. (1992). Selection and commercial cultivation of Garijo, P., Santamarıá, P., Lo´pez, R., Sanz, S., Olarte, C., & wine yeast and bacteria. In G. H. Fleet (Ed.), Wine Gutie´rrez, A. R. (2008). The occurrence of fungi, yeast microbiology and biotechnology (pp. 421e447). Chur, and bacteria in the air of a Spanish winery during Switzerland: Harwood Academic Publishers. vintage. Int. J. Food Microbiol., 125(2), 141e145. Emsweiler-Rose, B., Gehle, W. D., Johnston, R. W., Ibeas, J. L., Lozano, I., Perdigones, F., & Jimeńez, J. (1997). Okred, A., Moran, A., & Bennet, B. (1984). An enzyme Dynamics of flor yeast populations during the biolog- immunoassay technique for the detection of salmonellae ical ageing of sherry wines. Am. J. Enol. Vitic., 48, in meat and poultry products. J. Food Sci., 49, 1010e1020. 75e79. Enrique, M., Marcos, J. F., Yuste, M., Martıńez, M., Valle´s, S., Jiranek, V., Langridge, P., & Henchke, P. A. (1995). Regula- & Manzanares, P. (2007). Antimicrobial action of tion of hydrogen sulphide liberation in wine-producing synthetic peptodes towards wine spoilage yeasts. Int. J. Saccharomyces cerevisiae strains by assimilable nitrogen. Food Microbiol., 118(3), 318e325. Appl. Environ. Microbiol., 61, 461e467. Esteve-Zarzoso, B., Fernańdez-Espinar, M. T., & Querol, A. Longo, E., Cansado, J., Agrelo, D., & Villa, T. G. (1991). (2004). Authentication and identification of Saccharo- Effect of climatic conditions on yeast diversity in grape myces cerevisiae flor yeast races involved in sherry musas from northwest Spain. Am. J. Enol. Vitic., 42, ageing. Anton. Leeuw., 85, 151e158. 141e144. Esteve-Zarzoso, B., Gostıńcar, A., Bobet, R., Uruburu, F., & Lo´pez, V., Querol, A., Ramoń, D., & Fernańdez- Querol, A. (2000). Selection and molecular character- Espinar, M. T. (2001). A simplified procedure to analyse isation of wine yeasts isolated from the El Penede`s area mitochondrial DNA from industrial yeasts. Int. J. Food (Spain). Food Microbiol., 17, 553e562. Microbiol., 68,75e81. Esteve-Zarzoso, B., Manzanares, P., Ramoń, D., & Querol, A. Lourens, K., & Reid, G. (2003). Stuck fermentation (1998). The role of non-Saccharomyces yeasts in industrial management. Austral. New Zeal. Grape. Wine. Feb. 72e74. wine making. Int. Microbiol., 1, 143e148. Martıńez, M., Torres, M., Bobet, R., Cantarero, X., Farris, G. A., Budroni, M., Vodret, T., & Deiana, P. (1990). A´ lvarez, J. F., & Rubires, X. (2004). Application de Sull’origine dei lieviti vinari i lieviti dei terreni, delle l’impe´dance e´lectrique pour le controˆle microbiologique foglie e degli acini di alcun vigneti sardi. L’Enotecnico, 6, des vins embouteille´s. Rev. Fran. Oenol., 205,34e36. 99e108. Martıńez, P., Pe´rez-Rodriguez, L., & Benitez, T. (1997). Fernańdez-Espinar, M. T., Lo´pez, V., Ramoń, D., Bartra, E., Evolution of flor yeast population during the biological & Querol., A. (2001). Study of the authenticity of ageing of fino Sherry wine. Am. J. Enol. Vitic., 48, 160e168. commercial wine yeast strains by molecular techniques. Martini, A. (1993). Origin and domestication of the wine Int. J. Food Microbiol., 70,1e10. yeast Saccharomyces cerevisiae. J. Wine Res., 4, 165e176. Fitts, R. (1985). Development of DNAeDNA hybridization Martini, A., Federici, F., & Rosini, G. (1980). A new test for the presence of salmonellae in foods. Food Tech- approach to the study of yeast ecology of natural nol., 39, 95. substrates. Can. J. Microbiol., 26, 856e859. Fleet, G. H. (1992). The microorganisms of winemaking. Martini, A., & Vaughan-Martini, A. (1990). Grape must Isolation enumeration and identification. In G. H. Fleet fermentation: Present and past. In J. F. T. Spencer, & REFERENCES 355

D. M. Spencer (Eds.), Yeast technology (pp. 105e123). Querol, A., Barrio, E., & Ramoń, D. (1992). A compartive Berlin, Germany: Springer-Verlag. study of different methods of yeast strain characteriza- Monteil, H., Blazy-Mangen, F., & Michel, G. (1986). Influ- tion. Syst. Appl. Microbiol., 15, 439e446. ence des pesticides sur la croissance des levures des Radler, F., & Lotz, B. (1990). The microflora of active dry raisins et des vins. Sci. Alim., 6, 349e360. yeast and the quantitative changes during fermentation. Moreno-Arribas, M. V., & Polo, M. C. (2008). Occurrence of Wein Wissenschaft, 45,114e122. lactic acid bacteria and biogenic amines in biologically Reguant, C., & Bordons, A. (2003). Typification of Oeno- aged wines. Food Microbiol., 25(7), 875e881. coccus oeni strains by multiplex RAPD-PCR and study of Park, S. K. (2008). Development of a method to measure population dynamics during malolactic fermentation. hydrogen sulfide in wine fermentation. J. Microbiol. J. Appl. Microbiol., 95(2), 344e353. Biotechnol., 18(9), 1550e1554. Rosini, G., Federici, F., & Martini, A. (1982). Yeast flora of Parrish, M. E., & Carroll, D. E. (1985). Indigenous yeast grape berries during ripening. Microb. Ecol., 8,83e89. associated with muscadine (Vitis rotundifolia) grapes and Solfrizzo, M., Panzarini, G., & Visconti, A. (2008). Determi- musts. Am. J. Enol. Vitic., 36, 165e169. nation of Ochratoxin A in grapes, dried vine fruits and Pfannebecker, J., & Fro¨hlich, J. (2008). Use of a species- winery byproducts by high-performance liquid chro- specific multiplex PCR for the identification of ped- matography with fluorometric detection (HPLC-FLD) iococci. Int. J. Food Microbiol., 128(2), 288e296. and immunoaffinity cleanup. J. Agric. Food Chem., 56(23). Pless, P., Futschik, K., & Shoff, E. (1994). Rapid detection of 11 081e11 086. salmonellae by means of a new impedance-splitting Torija, M. J., Roze`s, N., Poblet, M., Guillamoń, J. M., & method. J. Food. Protec., 57, 369e376. Mas, A. (2001). Yeast population dymanics in sponta- Poulard, A., Simon, L., & Cuinier, C. (1980). Variabilite de la neous fermentations: Comparision between two microflore levurienne de quelques terroirs viticoles du different wine producing areas over a period of three pays. Nantais. Conn. Vig. Vin, 14, 219e238. years. Anton. Leeuw. Int. J.G., 79, 345e352. Pretorius, I. S. (2000). Tailoring wine yeast for the new Vaughan-Martini, A., & Martini, A. (1995). Facts, myths and millennium: Novel aproaches to the ancient art of legends on the prime industrial microorganism. J. Ind. winemaking. Yeast, 16, 675e729. Microbiol., 14, 514e522. Pretorius, I. S., van der Westhuizen, T. J., & Zapparolli, G., Torriani, S., Pesente, P., & Dellaglio, F. (1998). Augustyn, O. P. H. (1999). Yeast diversity in vineyards Design and evaluation of malolactic enzyme gene tar- and wineries and its importance to the South African geted primers for rapid identification and detection of wine industry. S. Afr. J. Enol. Vitic., 20,61e74. Oenococcus oeni in wine. Lett. Appl. Microbiol., 27, 243e246. Index

Acetic acid bacteria Acetoin, production by genetic improvement, 46e47 general characteristics, 228 contaminating yeasts, 349 overview, 42e43 growth Acetyl-CoA C-acyltransferase, biochemical changes, 40e42 alcoholic fermentation, 243e245 expression in influence on quality, 45e46 factors affecting fermentation, 21 yeast morphological changes, aging and storage, 247e248 Acidomonas, see Acetic acid bacteria 42e43 ethanol levels, 246e247 Acrolein, lactic acid bacteria Alcohol acetyltransferase, genetic oxygen, 247 generation, 216 engineering, 182 pH, 245e246 AFLP, see Amplified fragment length Amontillado, 52, 54e55, 59, 61 sulfite levels, 247 polymorphism Amplified fragment length temperature, 246 Aging polymorphism (AFLP) grapes, 242 acetic acid bacteria effects, 247e248 lactic acid bacteria identification, identification biological aging 195 classical tests, 236 flor Saccharomyces strain identification, molecular tests composition, 58, 69 129 identification, 236e240 formation, 61e64 Amplified rDNA restriction analysis typing, 241e242 function, 54 (ARDRA) isolation, 235e236 genetic improvements acetic acid bacteria identification, metabolism aging yeasts, 77e78 238 alcohols, 230e231 fermentation yeasts, 75e76 lactic acid bacteria identification, 195 carbohydrates, 229e230, 249 grape varieties and wine types, ARDRA, see Amplified rDNA nitrogen, 232 53e55 restriction analysis organic acids, 231e232, overview, 51e56 Arginase, expression in 249e250 phases, 55 fermentation, 20e21 preservation yeasts Arginine deaminase, lactic acid freezing, 314e315 chromosomes, 71e72 bacteria, 206 incubation conditions, 314 environmental influences, Aroma lyophilization, 315 69e73 acetic acid bacteria effects, 248 media, 314 evolution of genomes and compounds, 15e16 periodic transfer, 315e316 populations, 73e75 lactic acid bacteria contributions, preparation of cells for genetic characteristics, 64e69 208 preservation, 314 membrane and cell wall, non-Saccharomyces yeast influences spoilage avoidance, 250e251, 350 72e73 primary aroma, 94e96 taxonomy, 232e235 mitochondria, 69e71 secondary aroma wine effects physiological characteristics, carbonyl compounds, 100e101 aroma, 248 56e64 components, 96, 98 ethyl acetate, 250 sparkling wine from traditional esters, 99e100 grapes and must, 248e249 method higher alcohols, 99 spoilage, 249e250 autolysis phenols, 101 yeast interactions, 250 acceleration, 46 sulfur compounds, 101e102 Acetobacter, see Acetic acid bacteria autophagy genes, 43e45 volatile fatty acids, 98e99

357 358 INDEX

Cvt pathway, autophagy, 44e45 EPS, Extracellular polysaccharide Aroma (Continued) see starter culture considerations Cytochrome c oxidase, expression in ESR, see Environmental stress primary aroma, 285 fermentation, 21 response secondary aroma, 284e285 Ethanol tolerance yeast metabolism influences, 14, DAP, see Diammonium phosphate acetic acid bacteria, 246 16e17 Dekkera, see Non-Saccharomyces yeasts DNA microarray analysis of gene Delta sequence, amplification for response, 156 Asaia, see Acetic acid bacteria Autolysis Saccharomyces strain improvement in fermentation, 24 acceleration, 46 identification, 124, starter cultures, 283 autophagy genes, 43e45 128e129 Ethyl acetate, acetic acid bacteria genetic improvement of yeasts, Denaturing gradient gel generation, 250 46e47 electrophoresis (DGGE), Ethyl carbamate overview, 42e43 yeast species precursor formation by lactic acid identification, 124e125, bacteria, 211e213 Bacteriocin, lactic acid bacteria 196, 240 starter culture considerations, 285 production, 214e215 DGGE, see Denaturing gradient gel Eutypa dieback, 261e262 Bacteriophages, lactic acid bacteria, electrophoresis Eutypa lata, see Filamentous fungi 213e214 Diammonium phosphate (DAP) Extracellular polysaccharide (EPS) Biofilm, prevention, 349 DNA microarray analysis of gene acetic acid bacteria, 248e249 Biolistic transformation, 173 response, 154e155 lactic acid bacteria, 216e217 supplementation, 6 Biological aging, see Aging Black rot, 260e261 DNAeDNA hybridization, lactic Fermentation, see also Malolactic Botrydial acid bacteria fermentation; identification, 195 Botrytis cinerea synthesis and Saccharomyces function, 271e273 DNA microarray aeration, 8e9 detoxification, 273 chromatin immunoprecipitation biochemistry combination, 152 alcoholic fermentation, 11e12 Botrytis cinerea, see Filamentous fungi serial analysis of gene expression, nitrogen metabolism, 12e14 Brettanomyces contamination, prevention, 349e350 150 carbon dioxide generation, 9 Saccharomyces cerevisiae gene clarification, 9 Cadmium, proteomics analysis of expression analysis pH, 9 yeast response, 160 external agent response studies, population variation analysis with 152e153 molecular techniques Candida, see Non-Saccharomyces yeasts genomic studies, 157e158 identification of new species and Cava, 352 metabolic studies, 150e152 hybrids in fermentation, 132 Cell cycle, yeast, 288e289 wine yeast analysis, 153e157 inoculated fermentation, 130e131 ,56 natural fermentation, 129e130 CFD, see Commodity flow diagram Dorado Rueda Chromatin immunoprecipitation, Drying refermentation prevention, 348e349 DNA microarray starter cultures sparkling wine production combination, 152 lactic acid bacteria, 296e297 primary fermentation, 36e37 Commodity flow diagram (CFD), yeast, 290e291 secondary fermentation, 38e39 Hazard Analysis and strain preservation stuck fermentation solutions, Critical Control Point, alginate beads, 308 347e348 328e330 bacteria, 315 temperature, 8 filter paper, 308 Fermentation power, starter cultures, Condado Viejo,55 Crabtree effect, 287 lyophilization, 306e307 282e283 silica gel, 308 Fermentation stress response (FSR), Criadera, 55, 60, 66, 69 yeast lyophilization, 309e310 156 Critical control point, see Hazard Analysis and Critical Filamentous fungi Control Point Electroporation, transformation, Botrytis cinerea Cryotolerance 172e173 analysis cryoprotectants, 318 Environmental stress response (ESR), molecular analysis, 267e269 starter cultures, 284 153, 156 morphology, 266e267 INDEX 359

proteomics, 269e270 Genetically modified organism, see Gluconacetobacter, see Acetic acid strategies, 264, 266 Genetic engineering bacteria toxin isolation and Genetic engineering Gluconobacter, see Acetic acid bacteria characterization, 270e273 classical genetics, 169e171 Glucose transporters, expression in chemical penetration, 263e264 ethanol tolerance in fermentation, fermentation, 24, 75 grapevine infection, 262e263 24 Glutamate dehydrogenase, Botrytis life cycle, 264e265 fermentation efficiency, 23e24 cinerea regulation, 268e269 toxins in infection mechanism, legislation Glyceraldehyde-3-phosphate 273e275 European Union, 183e184 dehydrogenase diseases labeling, 184 expression in fermentation, 18 black rot, 260e261 United States and other countries, genetic engineering, 181 eutypa dieback, 261e262 184 Glycerol-3-phosphate grapevine downy mildew, overview, 171e172 dehydrogenase 259e260 phenotype improvements in wine expression in fermentation, phomopsis cane and leaf yeasts 19, 22 spot, 261 industrial winemaking genetic engineering, 181 powdery mildew, 258e259 process improvements, Glycerol transporter, genetic overview, 257e258 179e180 engineering, 181 preservation organoleptic, nutritional, and Glycogen synthase, expression in freezing, 316 safety property fermentation, 21, 23 lyophilization, 316e317 improvements, 181e182 Good agricultural practice, periodic transfer, 317 overview, 177e179 322 suspensions, 317e318 wine physicochemical property Good manufacturing practice, Fino, 52, 54e56, 58e61, 351e352 improvements, 180e181 322e323 Flocculation genes, flor formation promoters and gene expression Grapevine downy mildew, 259e260 role, 62e65, 67 regulation, 177e178 Guignardia bidwelllii, see Filamentous Flor prospects, 184e186, 353 fungi composition, 58, 69 sparkling wine yeasts formation, 61e64 aging yeasts, 77e78 HACCP, see Hazard Analysis and function, 54 autolysis improvement, Critical Control Point yeasts 46e47 Hanseniaspora, see Non-Saccharomyces chromosomes, 71e72 fermentation yeasts, 75e76 yeasts environmental influences, tools, 172 Hazard Analysis and Critical Control 69e73 transformation Point (HACCP) evolution of genomes and selection markers, 173e176 commodity flow diagram, populations, 73e75 techniques, 172e173 328e330 genetic characteristics, 64e69 Genomics prerequisite programs genetic improvements, 77e78 comparative genomics and good agricultural practice, 322 membrane and cell wall, 72e73 Saccharomyces cerevisiae good manufacturing practice, mitochondria, 69e71 genome origins, 147e149 322e323 physiological characteristics, DNA microarray analysis of principles 56e64 Saccharomyces cerevisiae gene corrective actions, 321, 335 Freeze-drying, see Drying expression critical control point Freezing external agent response studies, determination, 321, 333e334 bacteria, 315 152e153 critical limit establishment, 321 cryoprotectants, 318 genomic studies, 157e158 documentation, 322, 336 cryotolerance of starter cultures, 284 metabolic studies, 150e152 hazard analysis, 321, 330 strain preservation, 304e306 wine yeast analysis, 153e157 monitoring system establishment, yeast, 309e310 prospects, 163 321, 334e335 Fructose bisphosphate aldolase, yeast characteristics, overview, 320e321 expression in 144e147 verification procedure fermentation, 18 Glucokinase, expression in establishment, 321e322, FSR, see Fermentation stress response fermentation, 23 335e336 360 INDEX

Hazard Analysis and Critical Control glycosides, 204 Metschnikowia, see Non-Saccharomyces Point (HACCP) (Continued) lipids, 204 yeasts product description, 327e328 organic acids Microbiological control, grapes, winemaking application citric acid, 201e202 342e343 background, 323e324 malic acid, 200e201 Microsatellite, Saccharomyces strain ochratoxin A control, 327e335 tartaric acid, 202 identification, 122e123, Heat shock proteins (HSPs) phenolic compounds, 127e128 aging yeast expression, 77 202e203 Minisatellite, Saccharomyces strain fermentation yeast expression, 19, proteins and peptides, 204 identification, 122e123, 21e22 population dynamics during 127e128 Hexokinase, expression in winemaking, 196e198 Mitochondrial DNA, restriction fermentation, 23 preservation analysis in Saccharomyces Histamine, lactic acid bacteria freezing, 314e315 strain identification, formation, 210 incubation conditions, 312 118e119, 126, 345 HSPs, see Heat shock proteins lyophilization, 315 Molecular testing Hydrogen sulfide, starter culture media, 311 acetic acid bacteria considerations, 285 overview, 311 identification, 236e240 periodic transfer, 315e316 typing, 241e242 Inoculation preparation of cells for Botrytis cinerea, 267e269 pied de cuve culture preparation preservation, 312e314 commercial yeast characterization, alcoholic fermentation, 344e345 sensory effects in wine 131 malolactic fermentation, 345 acrolein, 216 population variation analysis quality control analysis, 346e347 aroma, 208 identification of new species starter cultures off-flavors and hybrids in fermentation, lactic acid bacteria, 297e298, 347 aromatic heterocyclic 132 yeast, 291e292 compounds, 217e218 inoculated fermentation, volatile phenols, 217 130e131 Jerez wine, basic characteristics, 52, 57 piquˆre lactique, 216 natural fermentation, 129e130 ‘extracellular polysaccharides, Saccharomyces strain identification Killer factor, starter cultures, 284 216e217 amplified fragment length starter cultures polymorphism, 129 Lactic acid bacteria biomass production, 295e296 delta sequence amplification, 124, bacteriocin production, 214e215 drying, packaging and storage, 128e129 bacteriophages, 213e214 296e297 hybridization, 115, 125 biogenic amine formation, 209e210 functions, 292e293 mitochondrial DNA restriction classical tests, 193e194 inoculation, 297e298, 347 analysis, 118e119, 126 ethyl carbamate precursor strain selection criteria, pulsed-field gel electrophoresis of formation, 211e213 293e295 chromosomes, 116e117, general characteristics, stress response, 213 125e126 192e193 types, 193 rapid amplification of identification yeast interactions, 215e216 polymorphic DNA, 121, 127 malolactic fermentation Lactobacillus, see Lactic acid bacteria repetitive DNA analysis with sensory property contribution, Leuconostoc, see Lactic acid bacteria polymerase chain reaction, 208e209 Lithium, transformation, 122e123, 127e128 starter cultures, 207 172e173 spoilage yeast detection, metabolism Lyophilization, see Drying 132e133 aldehydes, 203 yeast species identification amino acids, 205e207 Malolactic fermentation polymerase chain reaction- carbohydrates sensory property contribution, denaturing gradient gel disaccharides, 198e199 208e209 electrophoresis, 124e125 monosaccharides, 198 starter cultures, 207 real-time polymerase chain polyalcohols, 199e200 Manzanilla,54 reaction, 120e121, 123 polysaccharides, 199 Metabolomics, wine yeast, 162 ribosomal DNA esters, 204 INDEX 361

restriction analysis, 114, 120 ODC, see Ornithine decarboxylase rapid amplification of sequencing, 115, 117 Oenococcus, see Lactic acid bacteria polymorphic DNA, 121, 127 Must Off-flavors, lactic acid bacteria repetitive DNA analysis with acetic acid bacteria effects, 248e249 aromatic heterocyclic compounds, polymerase chain reaction, composition 217e218 122e123, 127e128 lipids, 7 volatile phenols, 217 Powdery mildew, 258e259 mineral salts, 7 5Oloroso,5 Preservation of microbial strains nitrogenous compounds, 5e6 Ornithine decarboxylase (ODC), drying organic acids, 5 lactic acid bacteria, alginate beads, 308 polyphenols, 7 211e212 filter paper, 308 sugars, 5 Ornithine transcarbamylase, lactic silica gel, 308 inoculation, see Inoculation acid bacteria, 206 filamentous fungi OTA, see Ochratoxin A freezing, 316 Non-Saccharomyces yeasts lyophilization, 316e317 aging role in winemaking, 57e58 Pa´lido Rueda,56 periodic transfer, 317 aroma influences Palo Cortado,55 suspensions, 317e318 primary aroma, 94e96 PCR, see Polymerase chain reaction long-term preservation secondary aroma Pediococcus, see Lactic acid bacteria acetic acid bacteria carbonyl compounds, PEG, see Polyethylene glycol media, 314 100e101 PFGE, see Pulsed-field gel incubation conditions, 314 components, 96, 98 electrophoresis preparation of cells for esters, 99e100 Phomopsis cane and leaf spot, 261 preservation, 314 higher alcohols, 99 Phomopsis viticola, see Filamentous freezing, 314e315 phenols, 101 fungi lyophilization, 315 sulfur compounds, 101e102 Phosphoadenylylsulfate reductase, freezing, 304e306 volatile fatty acids, 98e99 expression in lactic acid bacteria identification, see also Molecular fermentation, 22 freezing, 314e315 testing Phosphoglycerate mutase, incubation conditions, 312 long-chain fatty acid profile, expression in lyophilization, 315 88e89 fermentation, 18 media, 311 overview of tests, 87e88, 113 Phosphoribosyl pyrophosphate overview, 311 protein profile, 89 aminotransferase, preparation of cells for isolation and enumeration, 86e87 expression in preservation, 312e314 mixed starter culture design, fermentation, 22 lyophilization, 306e307 102e104 Pichia, see Non-Saccharomyces yeasts yeast morphology, 113 Pie de cuba, 52, 58 freezing, 309e310 overview, 85e86, 112 Pied de cuve, 292, 344e345 lyophilization, 309e311 vinification role Piquˆ re lactique, 216 objectives, 304 cellulolytic and hemicellulolytic Plasmopara viticola, see Filamentous recovery, 308 enzymes, 93 fungi short-term preservation glycosidases, 97e98 Polyethylene glycol (PEG), bacteria, 315e316 overview, 89e90 transformation, 172 periodic transfer, 307, 315e316 pectolytic enzymes, 90e93 Polymerase chain reaction (PCR) suspensions, 307 proteases, 91, 93 polymerase chain reaction- yeast, 311 denaturing gradient gel Presiphiperfolan-8-ol synthase, Ochratoxin A (OTA) electrophoresis, 124e125 273e274 food sources, 325e326 real-time polymerase chain reaction, Primera criadera,55 Hazard Analysis and Critical 120e121, 123 Prise de mousse, sparkling wine Control Point tasks in Saccharomyces strain identification production, 37e38 control, 327e335 amplified fragment length Proline, metabolism, 6 intake limits, 326 polymorphism, 129 Promoter, gene expression regulation microbial sources, 325e327 delta sequence amplification, 124, in genetic engineering, toxicity, 325 128e129 177e178 362 INDEX

Proteomics Saccharomyces strains in fermentation, 1e3 Botrytis cinerea, 269e270 aging yeasts, see Aging SAGE, see Serial analysis of gene wine yeast strains, 158e161 autophagy, see Autolysis expression Protoplast, transformation, 172 cellular organization, 3e4 Serial analysis of gene expression Pulsed-field gel electrophoresis comparative genomics and (SAGE), 150 (PFGE) Saccharomyces cerevisiae Sobretabla,55 Botrytis cinerea analysis, 267 genome origins, 147e149 SOD, see Superoxide dismutase chromosome separation in genetic characteristics, 4 Solera, 55, 66, 69 Saccharomyces strain growth during fermentation Sparkling wine identification, 116e117, biochemistry aging 125e126 alcoholic fermentation, autolysis commercial yeast quality control, 11e12 acceleration, 46 346e347 nitrogen metabolism, 12e14 autophagy genes, 43e45 lactic acid bacteria identification, 195 gene expression genetic improvement, 46e47 glycolytic genes, 18e19 overview, 42e43 Quality control osmotic stress response genes, biochemical changes, 40e42 analysis of commercial yeasts and 19e20 influence on quality, 45e46 inoculation, 346e347 overview, 17e18 yeast morphological changes, final product microbiological stationary phase genes, 20e22 42e43 quality control, 352e353 stress response, 22e23 overview, 34 genetic improvement of production techniques, 34e35 RAPD, see Rapid amplification of efficiency, 23e24 traditional method for production polymorphic DNA kinetics, 9e10 primary fermentation, 36e37 Rapid amplification of polymorphic must composition prise de mousse,37e38 DNA (RAPD) lipids, 7 secondary fermentation, 38e39 acetic acid bacteria typing, 241 mineral salts, 7 stages, 35e36 lactic acid bacteria identification, 195 nitrogenous compounds, 5e6 types, 35 overview, 345e346 organic acids, 5 Spoilage Saccharomyces strain identification, pesticide effects, 8 acetic acid bacteria, 249e250, 350 121, 127 polyphenols, 7 avoidance, 250e251, 350 Raya,55 sugars, 5 lactic acid bacteria, 350 Real-time polymerase chain reaction, sulfite inhibitors, 7e8 yeast detection, 132e133 see Polymerase chain metabolism and wine aroma, 14e17 Starter cultures reaction morphological changes in sparkling drying, 290e291 Restriction analysis wine aging, 42e43 historical perspective, 280e281 acetic acid bacteria identification, 240 secondary fermentation, see inoculation, 291e292 Botrytis cinerea analysis, 267 Sparkling wine isolation of yeasts, 281e282 commercial yeast quality control, strain identification lactic acid bacteria 347 amplified fragment length biomass production, 295e296 lactic acid bacteria identification, 195 polymorphism, 129 drying, packaging and storage, mitochondrial DNA for delta sequence amplification, 124, 296e297 Saccharomyces strain 128e129 functions, 292e293 identification, 118e119, hybridization, 115, 125 inoculation, 297e298 126, 345 mitochondrial DNA restriction strain selection criteria, ribosomal DNA for species analysis, 118e119, 126 293e295 identification, 114, 120 pulsed-field gel electrophoresis of malolactic fermentation, 207 Ribosomal DNA (rDNA), yeast chromosomes, 116e117, non-Saccharomyces yeasts in mixed species identification 125e126 starter culture design, acetic acid bacteria, 238 rapid amplification of 102e104 lactic acid bacteria, 195 polymorphic DNA, 121, 127 production stages, 280 restriction analysis, repetitive DNA analysis with quality control analysis, 346e347 114, 120 polymerase chain reaction, selection criteria for yeasts sequencing, 115, 117 122e123, 127e128 aroma considerations INDEX 363

primary aroma, 285 starter culture stress tolerance, Torulaspora, see Non-Saccharomyces secondary aroma, 284e285 283e284 yeasts cryotolerance, 284 sulfite, 22 Transformation, see Genetic ethanol tolerance, 283 Stuck fermentation, solutions, engineering ethyl carbamate, 285 347e348 Transgenic yeast, see Genetic fermentation power, 282e283 Sulfite engineering hydrogen sulfide, 285 acetic acid bacteria effects, 247 Tryptophan synthase, selection use, killer factor, 284 addition and inhibitor activity, 175 nitrogen demand, 283 7e8 Tyramine, lactic acid bacteria stress tolerance, 283e284 resistance genes, 154 formation, 210 sulfite resistance, 283 starter cultures resistance, 283 tasting, 286 stress response, 22 Unciunula necator, see Filamentous volatile acidity, 285 Superoxide dismutase (SOD), aging fungi yeast biomass production, yeast expression, 71, 78 URA3, complementation in selection, 286e290 175 Strain preservation, see Preservation Temperature gradient gel Urease, genetic engineering, 182 of microbial strains electrophoresis (TGGE), Stress response acetic acid bacteria Vitamins, synthesis, 6 Environmental stress response, 153, identification, 240 156 TGGE, see Temperature gradient gel Yema,52 Fermentation stress response, 156 electrophoresis lactic acid bacteria, 213 Thioredoxin, expression changes in Zygosaccharomyces, see Saccharomyces,19e20, 22e23 oxidative stress, 289 Non-Saccharomyces yeasts