UNS W

>013910086 U N SW - 8 AUG 2008 LIBRARY Investigation of yeasts and yeast-like fungi associated with Australian wine grapes using cultural and molecular methods

by Ai Lin Beh

A thesis submitted as a fulfillment for the degree of Doctor of Philosophy

University of New South Wales School of Chemical Sciences and Engineering Sydney, Australia

2007 PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Beh

First name: Ai Lin Other name/s:

Abbreviation for degree as given in the University calendar: PhD

School: University of New South Wales Faculty: Faculty of Engineering School of Chemical Sciences and Title: Investigation of the yeasts and yeast-like Engineering fungi associated with Australian wine grapes (Food Science and Technology) using cultural and molecular methods

Abstract 350 words maximiim: (PLEASE TYPE) This thesis presents a systematic investigation of yeasts associated with wine grapes cuJtivated in several Australian vineyards during the 2001-2003 vintages. Using a combination of cultural and molecular methods, yeast populations of red (Cabernet sauvignon, Merlot, Tyrian) and white (Sauvignon blanc. Semillen) grape varieties were examined throughout grape cultivation.

The yeast-like ftingus, Aiireohasidiumpiillulans, was the most prevalent species found on grapes. Various species of Cryptococcus, Rhodolorula and Sporobohmyces were frequently isolated throughout grape maturation. Ripe grapes showed an increased incidence of Hameniaspora and Melschnikowia species for the 2001-2002 season, but not for the drought affected, 2002-2003 season. Atypical, hot and dry conditions may account for this difference in yeast flora and have limited comparisons of data to determine the influences of vineyard location, grape variety and pesticide applications on the yeast ecology. More systematic and controlled studies of these variables are required. Damaged grape berries harboured higher yeast jropulations and species diversity than intact healthy berries.

PCR-DGGE analysis was less sensitive than plate culture for describing the diversity of yeast species on grapes; it detected prevalent species, but subdominant populations below 10^ CFU/g were not detected. In some cases, PCR-DGGE revealed the presence of yeasts {Candida galii, C. zemphnim) not isolated by culture. Fermentative wine species (Kluyveromyce.s, Tonilaspora, Saccharomyces) were rarely isolated, and only detected by enrichment cultures.

Significant morphological and genetic variability were detected among A. pullulam and other black yeasts isolates from grapes. Taxonomic characterization of 61 strains by ITS-RFLP and rDNA sequencing revealed that they belonged to several distinct species within the generic groupings of Aureohasidium, Hormonema and Kahatiella. Isolates were strong producers of extracellular enzymes and polysaccharides that could have oenological significance, and, using a plate assay, some were antagonistic towards Badllus ihuringiemis, several wine yeasts, and some spoilage and mycotoxigenic fiingi found on grapes. Growth of Saccharomyces cerevisiae was not inhibited by these organisms in grape juice.

A species-specific probe was developed for the identification of the wine s{X)ilage yeast, Zygosaccharomyces hailii in a microtitre plate hybridization assay. The probe detected 10^ cells/ml in wine, reliably differentiating Z bailii from other Zygosaccharomyces and other wine-related yeasts.

Declziration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing to the Registrar. Requests for a longer period of restriction may be considered in exceptional circumstances if accompanied by a letter of support from the Supervisor or Head of School. Such requests must be submitted with the thesis/dissertation.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

h rr / , Registrar and Deputy . ^[n/^ Principal

THIS SHEET IS TO BE GLUED TUOO THE INSIDE FRONT COVER OF TTH1 E THESIS

N:\FLORENCE \ABSTRACT DECLARATION

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of materials which have been accepted for the award of any other degree of diploma at UNSW or any other educational institution, except where due acknowledgment is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

Ai Lin Beh COPYRIGHT STATEMENT

'I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed

Date

AUTHENTICITY STATEMENT

'I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.'

Signed

Date ACKNOWLEDGEMENTS

The following is an expression of thanks and an acknowledgement to the people and institutions who have helped me make this work possible. I would like to express my gratitude to my supervisor, Prof Graham Fleet, for his guidance on this project, and for my development as a scientist. I thank him for his support, patience and his many anecdotes that inspired and amused me throughout my studies. He has given me so many opportunities to learn and grow, and I feel very privileged to have been a part of his team. I would like to give special thanks to my co-supervisor. Dr. Gillian Heard, who offered her perceptiveness, words of wisdom and encouragement when at times, all seemed lost, but also for the cracking of the whip when needed.

This study was made possible by the financial support of the Australian Grape and Wine Research Development and Corporation (GWRDC) and in part, by the Australian Department of Agriculture, Fisheries and Forestry (DAFF). I am grateful to these organizations for the opportunity to undertake in this project.

The brief visit to the Rosenstiel School of Marine and Atmospheric Science (RSMAS), University of Miami, was a major highlight in my project. I would like to express my warmest thanks to Mara Diaz, Sara Cotton, Traci Kiesling and Jack Fell for sharing with me their expertise in, and enthusiasm for molecular diagnostics. I greatly appreciate all of your patient and cheerful instruction in the development of capture probes for the microplate hybridization assay. I am grateful to the yeast culture curators at the Australian Wine Research Institute (AWRI), Centraalbureau voor Schimmelcultures (CBS), Food Science Australia (FRR), National Agricultural Utilization Research (NRRL) and the ZIM Culture Collection of Industrial Microorganisms; they generously provided me with reference cultures for this work. I would also like to acknowledge the viticulturalists at Rosemount Estate and McWilliam's Wines for their kind assistance with sampling grapes for our study.

To my lab mates Sungsook, Pete, Hugh, Lidia, Pat, Victoria and Michelle, I thank you for your friendship over the years. Your warmth and good humour have made the long days in the Food Micro lab filled with fun, memorable moments. I will miss your happy Ill

faces and our discourses (both fruitful and frivolous) that were held over many samplings of fermented foods and beverages. Thanks are extended to the staff of Food Science and Technology for providing valuable ideas, advice and assistance during my studies.

Finally, I would like to thank my family. Without their love, support and encouragement, 1 would not have been able to complete this work. I am indebted to my parents for the sacrifices they have made in ensuring that I receive a fulfilling education. And to my brothers, Liang and Siang, thank you for inspiring courage and confidence in me. TABLE OF CONTENTS

DECLARATION i ACKNOWLEDGEMENTS jj TABLE OF CONTENTS ¡v ABSTRACT

PUBLICATIONS AND PRESENTATIONS FROM THIS THESIS xi!

CHAPTER 1 INTRODUCTION 1

CHAPTER 2 LITERATURE REVIEW

2.1 VITICULTURE 5 2.1.1 Grapes 5 2.1.1 Grapes 5 2.1.2 Grape cultivation 5 2.2 THE PROCESS OF WINE MAKING 9 2.2.1 Grapes 9 2.2.2 Crushing and pre-fermentation treatments 9 2.2.3 Alcoholic fermentation 11 2.2.4 Malolactic fermentation 12 2.2.5 Post fermentation processes 12 2.3 ROLES OF MICROORGANISMS IN WINEMAKING 13 2.3.1 Fungi 13 2.3.2 Yeasts 14 2.3.3 Bacteria 16 2.4 THE GRAPE SURFACE AS A MICROBIAL ECOSYSTEM 17 2.4.1 The physical and chemical properties of grape surfaces 17 2.4.2 Nutrients for yeast growth on grape surfaces 18 2.5 YEAST ECOLOGY OF WINE GRAPES 21 2.5.1 Origin of yeasts on grapes 21 2.5.2 Yeast species associated with wine grapes 23 2.5.2.1 Grape cultivar 23 2.5.2.2 Immature grapes 23 2.5.2.3 Mature grapes 23 2.5.2.4 Overripe and damaged grapes 25 2.5.3 Saccharomyces cerevisiae 26 2.5.4 Vineyard environmental influences 27 2.5.5 Agrichemical influences 28 2.5.7 Yeast properties 29 2.6 AUREOBASIDIUM PULLULANS 31 2.6.1 Taxonomy 33 2.6.2 Ecology 36 2.6.3 Biotechnological applications 40 2.7 MOLECULAR METHODS FOR THE ANALYSIS OF YEASTS ASSOCIATED WITH WINE GRAPES 41 2.7.1 Methods for yeast species identification 42 2.7.1.1 Ribosomal DNA sequencing 42 2.7.1.2 PCR-RFLP 43 2.7.1.3 Species-specific primers and probes 43 2.7.2 Methods for yeast strain differentiation 44 2.7.3 Molecular strategies for monitoring yeast communities 45 2.7.4 Factors affecting the performance of molecular methods for analysis of yeasts 48 2.7.4.1 DNA extraction 48 2.7.4.2 PGR 49 2.7.4.3 DGGE 49 2.8 SUMMARY 50

CHAPTER 3 YEAST ECOLOGY OF WINE GRAPES DURING CULTIVATION AT SEVERAL VINEYARDS IN AUSTRALIA 3.1 INTRODUCTION 51 3.2 MATERIALS AND METHODS 52 3.2.1 Grape samples 52 3.2.2 Analysis of yeasts on grapes 56 3.2.3 Isolation and enumeration on agar media 58 3.2.4 Isolation by enrichment culture 58 3.2.5 DNA extraction for PGR amplification 59 3.2.6 PGR amplification of DNA for sequence identification and PGR- DGGE analysis 60 3.2.7 Denaturing gradient gel electrophoresis (DGGE) 61 3.2.8 DNA sequencing 63 3.2.9 Species identification 63 3.3 RESULTS 64 3.3.1 Glimatological data for the vineyard regions in the years of 2001-2003 64 3.3.2 Evaluation of plating medium, sample processing and enrichment methods for the isolation and enumeration of yeasts 66 3.3.2.1 Gomparison of MEA and WLNA 66 3.3.2.2 Effect of grape rinsing and maceration 66 3.3.2.3 Gomparison of enrichment and autoenrichment 68 3.3.3 Population of yeasts on wine grapes 69 3.3.4 The diversity of yeast species associated with wine grapes during cultivation 73 3.3.5 The diversity of yeast species associated with undamaged and damaged wine grapes 84 3.3.6 Frequency of isolation of yeasts from wine grapes 93 3.4 DISCUSSION 95 3.4.1 Analytical strategies for characterising yeasts associated with wine grapes 95 3.4.2 Population and species diversity 101 3.4.3 Factors affecting the yeast ecology of wine grapes 105

CHAPTER 4 THE DIVERSITY AND SIGNIFICANCE OF AUREOBASIDIUM AND RELATED SPECIES ON WINE GRAPES 4.1 INTRODUCTION 108 4.2 MATERIALS AND METHODS 109 4.2.1 Gultures 109 4.2.2 DNA extraction for PGR amplification 110 vil

4.2.3 rDNA amplification 112 4.2.4 PCR-based RFLP of the ITS regions 112 4.2.5 rDNA sequencing 113 4.2.6 Microsatellite PGR fingerprinting 113 4.2.7 Physiological characterization 114 4.2.8 Utilisation of malic and tartaric acids 115 4.2.9 pH, ethanol and SO2 tolerance 115 4.2.10 Exopolysaccharide production 116 4.2.11 Interactions between Aureobasidium and related isolates, and other microorganisms 117 4.2.12 Growth oi Aureobasidium-Wke organisms with S. cerevisiae •• • 118 4.2.13 Screening of extracellular enzymes 119

PART 1. MOLECULAR, MORPHOLOGICAL AND BIOCHEMICAL CHARACTERISATION OF AUREOBASIDIUM AND RELATED ISOLATES 4.3 RESULTS 122 4.3.1 Grouping isolates based on PCR-ITS RFLP patterns 122 4.3.2 Identification of Aureobasidium and related species from grapes by rDNA sequencing 127 4.3.3 Morphological features 129 4.3.4 Analysis of all Aureobasidium and related isolates by ITS-RFLP133 4.3.5 Biochemical profiles o1 Aureobasidium and related isolates 134 4.3.6 PGR microsatellite profiling 137 4.4 DISCUSSION 140

PART 2. PROPERTIES OF OENOLOGICAL SIGNIFICANCE 4.5 RESULTS 144 4.5.1 pH, osmo-, ethanol and SO2 tolerance 144 4.5.2 Utilisation of grape organic acids 145 4.5.3 Production of exopolysaccharides 146 4.5.4 Interactions oi Aureobasidium and related isolates with other microorganisms 148 4.5.4.1 Filamentous fungi and yeasts 148 vili

4.5.4.2 Bacteria 150 4.5.5 Mixed culture fermentation with S. cerevisiae 153 4.5.6 Enzymatic profiles 156 4.6 DISCUSSION 159

CHAPTER 5 A MICROTITRE PLATE DNA PROBE HYBRIDIZATION ASSAY FOR THE DETECTION OF WINE YEAST SPECIES 5.1 INTRODUCTION 174 5.2 MATERIALS AND METHODS 174 5.2.1 Reference cultures 177 5.2.2 Design and synthesis of species-specific capture probes 178 5.2.3 Attachment of probes to microtitre plates 179 5.2.4 Extraction and preparation of DNA for microtitre plate hybridization assay 179 5.2.5 PGR conditions to produced biotinylated amplicons 180 5.2.6 Plate hybridization assay 181 5.3 RESULTS 182 5.3.1 Design of probes 182 5.3.2 Elimination of background and non-specific interactions 183 5.3.2.1 Plate surface chemistry 186 5.3.2.2 Blocking procedures 186 5.3.2.3 Removal of excess biotin-labeled primers 186 5.3.3 Evaluation of probe specificities in the microtitre plate hybridization assay 187 5.3.4 Further evaluation of the Z bailii microplate hybridization assay 189 5.3.5 Application of microtitre plate hybridization assay to detect Z bailii in wine samples 191 5.4 DISCUSSION 192

CHAPTER SIX CONCLUSIONS 196

CHAPTER SEVEN BIBLIOGRAPHY 201 APPENDIX 1 240 APPENDIX 2 242 APPENDIX 3 243 APPENDIX 4 244 ABSTRACT

This thesis presents a systematic investigation of yeasts associated with wine grapes cultivated in several Australian vineyards during the 2001-2003 vintages. Using a combination of cultural and molecular methods, yeast populations of red (Cabernet sauvignon, Merlot, Tyrian) and white (Sauvignon blanc, Semillon) grape varieties were examined throughout grape cultivation.

The yeast-like fungus, Aureobasidiumpullulans, was the most prevalent species found on grapes. Various species of Cryptococcus, Rhodotorula and Sporobolomyces were frequently isolated throughout grape maturation. Ripe grapes showed an increased incidence of Hanseniaspora and Metschnikowia species for the 2001-2002 season, but not for the drought affected, 2002-2003 season. Atypical, hot and dry conditions may account for this difference in yeast flora and have limited comparisons of data to determine the influences of vineyard location, grape variety and pesticide applications on the yeast ecology. More systematic and controlled studies of these variables are required. Damaged grape berries harboured higher yeast populations and species diversity than intact healthy berries.

PCR-DGGE analysis was less sensitive than plate culture for describing the diversity of yeast species on grapes; it detected prevalent species, but subdominant populations below 10^ CFU/g were not detected. In some cases, PCR-DGGE revealed the presence of yeasts {Candida galli, C. zemplinina) not isolated by culture. Fermentative wine species {Kluyveromyces, Torulaspora, Saccharomyces) were rarely isolated, and only detected by enrichment cultures.

Significant morphological and genetic variability were detected amongpullulans and other black yeasts isolates from grapes. Taxonomic characterization of 61 strains by ITS-RFLP and rDNA sequencing revealed that they belonged to several distinct species within the generic groupings oí Aureobasidium^ Hormonema and Kabatiella. Isolates were strong producers of extracellular enzymes and polysaccharides that could have oenological significance, and, using a plate assay, some were antagonistic towards Bacillus thuringiensis, several wine yeasts, and some spoilage and mycotoxigenic fungi found on grapes. Growth of Saccharomyces cerevisiae was not inhibited by these organisms in grape juice.

A species-specific probe was developed for the identification of the wine spoilage yeast, Zygosaccharomyces bailii in a microtitre plate hybridization assay. The probe detected 10" cells/ml in wine, reliably differentiating Z bailii from other Zygosaccharomyces and other wine-related yeasts. Xll

PUBLICATIONS AND PRESENTATIONS FROM THIS THESIS

Beh, A.L., Fleet, G.H., Prakitchaiwattana, C. and Heard, G.M. (2006) Evaluation of molecular methods for the analysis of yeasts in foods and beverages. Adv. Exp. Med. Biol 571,69-106.

Fleet, G.H., Prakitchaiwattana, C., Beh, A.L. and Heard, G.M. (2002) The yeast ecology of wine grapes. In Biodiversity and biotechnology of wine yeasts ed. Ciani, M. pp. 1-17. Kerala: Research Signpost.

Beh, A.L. (2004) The interactions of Aureobasidium pullulans with microorganisms significant in wine production. Oral presentation at the Joint Australian Society for Microbiology/Australasian Mycological Society National Conference, Sydney, 27-28 September.

Beh, A.L., Fleet, G.H., and Heard, G.M. (2003) The interaction oiAureobasidium pullulans with microorganisms occurring on wine grapes. Poster presentation at the 23*^^ International Specialised Symposium on Yeasts, Interaction between Yeasts and Other Organisms, Budapest, 26-29 August.

Fleet, G.H., Beh, A.L., Pratehaiwattana, C., Bae, S.S., and Heard, G.M. (2002) The yeast and bacterial ecology of wine grapes. Poster presentation at the International Union of Microbiological Societies Xth International Congress of Bacteriology and Applied Microbiology, The World of Microbes, Paris, 27 July-1 August. CHAPTER ONE

INTRODUCTION

Yeasts are the principal microorganisms of wine production. They conduct the alcoholic fermentation of grape juice where, in addition to ethanol production, they produce a vast array of metabolic end products that substantially contribute to the basic structure and individuality of wine flavour and aroma. The dimensions of this contribution vary with the species and strain of yeast that grow during this fermentation. In later stages of the winemaking process, various yeast species can grow in the product to cause wine spoilage (Pretorious 2000; Fleet 2001, 2003). Consequently, yeasts have a profound impact on wine quality and process efficiency. Therefore, it is important to understand the sources and origins of those yeasts that contribute to wine production.

The yeasts of winemaking are considered to originate from three main sources, namely, (1) the microflora that are associated with the surface and tissue of grapes, (2) the microflora that are associated with the surfaces of wine processing equipment- but most likely originate from the grapes, and (3) the starter or inoculum cultures of specific yeast strains that many wineries now use to conduct the alcoholic fermentation. Such cultures are generally strains of Saccharomyc es cerevisiae or Saccharomyces hay anus that are widely accepted as the principal organisms of alcoholic fermentation. Prior to the introduction of starter cultures to wine production, some 50 years ago, all wine fermentations were spontaneous or wild fermentations that developed from the growth of indigenous yeast species coming from the grapes and winery equipment. Usually, strains of S. cerevisiae and S. bayanus dominate these fermentations, hence their recognition and selection as the main wine yeasts (Fleet and Heard 1993).

The yeast ecology of wine (grape juice) fermentation has been the subject of extensive study during the last 20-25 years. It has been consistently observed, globally, that a succession of yeast species and strains contributes to the fermentation, whether it is allowed to develop spontaneously, or guided by inoculation with a starter culture. The early stages of fermentation are dominated by the growth of species of Hanseniaspora or its anomorph, Kloeckera, followed by various species of Candida^ Metschnikowia, Pichia and Kluyveromyces. These species are indigenous to the grape juice, having originated from the grape raw material or contact of the juice with winery equipment. These species die off after 2-4 days, giving way to a predominance of the more ethanol tolerant S. cerevisiae or S. bayaniis that complete the fermentation. The strains of S.cerevisiae/S.bayanus that complete the fermentation are either those that have been inoculated or those that have arisen as part of the wild or indigenous microflora (Fleet 2003). As mentioned already, different yeast species and strains contribute different amounts of flavour/aroma metabolites and, in this way, their growth impacts on the final character of the wine (Romano et al. 2003; Swiegers et al. 2005). Grapes are a primary source of the yeasts that contribute to the fermentation. Thus, it is important to have a sound knowledge of their yeast ecology and any factors that affect this ecology.

A vast amount of research has been done over the past 100 years to describe the yeast species associated with wine grapes (reviewed in Kunkee and Amerine 1970; Vaughn- Martini and Martini 1995; Martini et al. 1996; Fleet et al. 2002). Most of this work, however, is simple qualitative description of the identity of yeast isolates that were recovered from grapes, with little quantitative focus and connection to an ecological understanding. Indeed, Martini et al. (1996) have questioned the soundness and reliability of much of this eariy study because of inadequacies in methods used to sample and analyse the grapes. Nevertheless, some general conclusions have emerged. Grapes at the time of harvest harbour a surface population of 10"^-10^ CFU/cm^, with species of Hanseniaspora/Kloeckera, Metschnikowia, Candida, Pichia, Cryptococcus and Rhodotorula being frequently isolated. The Hanseniaspora/Kloeckera species can represent as much as 50-70% of the flora. Reports on the occurrence of S. cerevisiae and S. bayaniis are inconsistent; some researchers report its frequent isolation from grapes, while others are unable to detect its presence (Martini et al. 1996; Tôrôk et al. 1996; Polsinelli and Mortimer 1999). Reasons for these discrepancies include the method of analysis (e.g. enrichment culture or plate culture), maturity of the grape berries at the time of analysis, and whether or not the berries have been damaged. Based on the need to better understand the origins of yeasts in wine production and the inadequacies of existing knowledge, there is strong justification for further research that gives a detailed, systematic investigation of the yeast ecology of wine grapes. Studies on the yeast ecology of grapes, to date, have been based on the use of standard culture methods; enrichment and plate culture for yeast isolation, and the application of an extensive range of morphological, biochemical and physiological tests for genus and species identification (Fleet 1993; Martini et al. 1996). The limitations of these culture based approaches in the study of microbial ecology are now well recognised. Enrichment and plate culture methods may not recover all of the viable microbial species associated with natural habitats and may underestimate the "real" populations by as much as 90-99% (Head et al. 1998). A range of molecular methods based on the analysis of habitat DNA has been developed to address these limitations. One such method that has found good application in the analysis of food and beverage microorganisms is PCR-denaturing gradient gel electrophoresis (PCR-DGGE). In this laboratory, PCR-DGGE has been used to study the yeasts and bacteria of wine grapes (Prakitchaiwattana et al. 2004; Bae et al 2006). The identification and biotyping of yeast isolates is now routinely done by a range of molecular methods, including sequencing of ribosomal DNA and other genes, restriction fragment length polymorphism (RFLP) of DNA amplicons, and profiling of microsatellite DNA (Kurtzman et al 2003; Kurtzman 2006).

The availability of specific DNA probe assays for yeasts associated with wines and grapes would greatly facilitate the ecological study of these organisms as well as find useful application in quality control and assurance programs. In contrast to the field of food bacteriology, there has been little progress in developing specific DNA probe assays for food and beverage yeasts that could be used in a routine simplified format.

With the background just described, the main aim of this thesis is to conduct a detailed, systematic investigation of the yeast populations and species associated with grapes cultivated in several geographically distinct vineyards located in New South Wales, Australia. Grapes will be analysed at several stages of maturity throughout their cultivation, and specific attention will be given to differentiate healthy, undamaged berries from damaged berries for these analyses. To minimise the possibility of underestimating the biodiversity of yeasts associated with the habitat, grape samples will be analysed by plate culture, enrichment culture and by PGR- DGGE of DNA extracts (Chapter Three). A combination of phenotypic and molecular methods will be used to identify and characterize the yeast isolates.

The investigation reported in this thesis is part of a much larger study on the yeast and bacterial ecology of Australian wine grapes, from which two other theses and publications have emerged (Prakitchaiwattana 2005; Prakitchaiwattana et al 2004; Bae 2006; Bae et al. 2004, 2006). During the course of this project, it became evident that the yeast-like fungus, Aureobasidiumpullulans, was the most prevalent organism associated with Australian wine grapes. Because of this prevalence and the extreme heterogeneity in the isolates obtained, it was considered that a more focused study of these isolates was necessary to the overall goals of the project. Consequently, Chapter Four reports a detailed study of the occurrence and oenological significance of Aureobasidium and related species on wine grapes. It should be noted that Aureobasidium is now considered as a yeast rather than a yeast-like fungus in modem taxonomy (Kurtzman and Fell 2006).

It became increasingly evident throughout the overall study that simple, rapid DNA- probe assays would greatly facilitate the ecological study of grape and wine yeasts, as has been described for marine yeasts (Kiesling et al. 2002a). Chapter Five reports the results of a study to develop a rapid microtitre plate DNA-probe assay for several grape and wine yeasts. CHAPTER TWO LITERATURE REVIEW

This thesis is concerned with the yeasts associated with wine grapes and the application of molecular methods to their analysis. Key studies associated with these topics will be presented in the Introduction and Discussion sections of subsequent experimental Chapters. This Chapter will provide a brief background to the role of yeast in winemaking, followed by a more detailed review of the literature describing yeasts on wine grapes. Specific discussion of Aureobasidiumpullulans is included because it was found to be the most prevalent yeast on wine grape in this study. A final section of this Chapter gives some background information on molecular methods used to identify and analyse yeasts in wine ecosystems. Some aspects of this discussion have already been published (Fleet et al 2002; Beh et al. 2006).

2.1 VITICULTURE

2.1.1 Grapes Viticulture is the study of grapes and grape cultivation for wine production. Numerous grape varieties of the species, Vitis vinifera L., are used in winemaking, with the particular variety determining the fruity or floral characteristics of the wine product. Some main varieties used in white winemaking are Riesling, Traminer, MUller- Thurgau, Chardonnay, Semillon and Sauvignon blanc, while those used in red winemaking include Cabernet sauvignon, Cabernet franc, Merlot, Pinot noir, Shiraz, Camay, Grenache and Barbera (Dry and Gregory 1992; Boulton et al 1995; Fleet 2001).

2.1.2 Grape cultivation Most grape varieties are propagated by cutting and are usually grown in a nursery for a year to produce rooting, and then they are planted in the vineyard. Grapevines require three to six years after planting to reach full economic production. A considerable amount of viticultural research has identified strategies for optimising the cultivation of grapes for winemaking. These include improvement of vine genotypes, carefully managed practices in grapevine propagation, irrigation, fertigation, canopy cover, and pest and disease control. There is extensive literature in the field of viticulture and the reader is referred to some key texts for further information (Jacquelin and Polulain 1962; Winkler 1973; Coombe and Dry 1992; Jackson and Lombard 1993; Ribereau- Gayon et al. 2000; Wade et aL 2004).

The development of wine grapes in the vineyard follows an annual cycle as they are a deciduous plant. The formation of leaf buds on the grapevine begins in spring. As the vine develops, mature grape berries are produced through the stages of inflorescence, flowering, pollination and fertilisation of the ovary, fruit and berry setting, growth of the green berry and finally ripening of the berry. Details of the key developmental stages are given in Table 2.1 and a pictorial illustration is given in Figure 2.1 (Pratt 1961; Coombe 1992, 1995; Coombe and McCarthy 1997, 2000; Kennedy 2002). The timeframe from budburst to harvesting of mature grapes is about 4-6 months. Various populations and species of yeasts grow on grapes throughout berry development and this ecology will be discussed later in this thesis.

Table 2.1 Key stages in the development of wine grapes in the vineyard

Budburst Dormancy of the vine bud is broken; generally recognised as formation of a green tip or shoot, and visible formation of leaf tissue. Usually occurs in March-May in the northern hemisphere and September-November in southern hemisphere Shoot and leaf development Progressive unfolding of leaves; beginning of inflorescence and flower formation Flowering Loosening and detachment of lower caps, beginning of inflorescence and flower formation Setting and development of grape berries (fruiting)

Berries usually hard and green, progressively increase in size from about 2 mm in diameter (peppercorn size) to about 7 mm in diameter (pea size); accompanied by closure of bunch of berries. This stage lasts about 60-70 days Veraison Berries begin to soften, change colour and increase in size; sugar concentration increases, acidity decreases. Generally starts about August-September in northern hemisphere and December-January in southern hemisphere and lasts about 40-50 days Engustment Occurs just prior to harvest, often late in the ripening process, sudden development and accumulation of grape flavour Harvesting Berries are harvested when fully ripe and proceed to senescence after this stage Pratt 1971; Coombe 1992, 1995; Coombe and McCarthy 1997, 2000; Fleet et al. 2002 Shoot and inflorescence development Flowering Berry Development Ripening E-L number 1 2 3 4 5 7 9 11 12 13 14 15 16 17 18 19 20 21 23 25 26 27 29 31 32 33 34 35 36 37 38

t hf4 38. Harvest r 27. Setting 35 Veraison 31. Berries pea size , , 23. Full bloom

19. Flowering begins

4. Budburst

(setting^

Days after flowering t> A 100 120

Figure 2.1 Adoption of a system for identifying grapevine growth stages as produced by Coombe (1995) and Kennedy (2002) E-L E-L number ALL STAGES number ALL STAGES 1. Winter bud 27. Setting young berries enlarging (>2mm diam.), ow 2. Buds well bunch at right angles to setem (o zr 3. Woolly bud-brown wool visible 29. Berries pepper-corn size (4 mm diam.); o D o 4. Green tips first leaf tissue visible

Figure 2.1 (continued) Adoption of system for identifying grapevine growth stages as produced by Coombe (1995) and Kennedy (2002) 8 2.2 THE PROCESS OF WINEMAKING The process of winemaking varies with the type of wine being produced, the technological innovations that are applied, and with the region or country. Detailed descriptions can be found in Amerine (1985), Boulton et al. (1995) and Ribéreau-Gayon et al. (2000). Red wine production is specifically described by Boulton (2003) and white winemaking is described by Ewart (2003). Some basic principles and operations are outlined in the following sections as a background to understanding the role of microorganisms in wine production. The main steps in the production of white and red wines are given in Figure 2.2.

2.2.1 Grapes The varieties of grapes used in winemaking have been mentioned already in Section 2.1. The grapes for wine production are harvested at an appropriate stage of maturity. Particularly important are their concentrations of sugars and acids, which are major constituents of the juice and have an important impact on its fermentation properties (Fleet 1998, 2001).

2.2.2 Crushing and pre-fermentation treatments Grapes intended for red wine are initially processed in the crusher-destemmer. The resulting juice and skins are transferred to tanks for prefermentation and maceration. Red grapes are fermented for about 7 days, at temperatures ranging from 20-30 The maceration of red grapes and facilitates extraction of colour and tannin components from the skin into the juice (Rankine 1989; Boulton et al. 1995).

For the production of white wines, grapes are crushed and destemmed, and then the juice is separated from the skins by settling and pressing, and transferred to fermentation tanks (Ewart 2003). Different pre-fermentation treatments may be carried out on juice and musts. These include cold settling, clarification of the juice, addition of sulfur dioxide as an antioxidant and preservative, and the introduction of oxygen to minimise browning of white musts (Boulton 2003; Ewart 2003). RED WINES WHITE WINES

Grapes Grapes

Crushing; addition of sulphur dioxide

Juice and Skins

Wine Wine

FINAL PRODUCT FINAL PRODUCT

Figure 2.2 Outline of process for red wine and white wine production (Fleet 2001) 2.2.3 Alcoholic Fermentation White wines are generally fermented at 10-18 °C, for 7-14 days or more, where the lower temperature and slower fermentation rate favour the retention of desirable volatile aroma compounds. Red wines are fermented for about 7 days at 20-30 °C, with higher temperatures required for colour extraction from the grape skins (Rankine 1989; Boulton etaL 1995).

Alcoholic fermentation may be conducted by the natural microflora origniating from the grapes and winery equipment. Fermentations of this style are referred to as either natural, indigenous, wild, spontaneous or uninoculated fermentations. Generally, the initial population of yeast is in the order of lO'^-lO^ cells/ml. These fermentations are generally slow to begin sugar utilisation, and are less vigorous (Fleet and Heard 1993; Fleet 1998). Spontaneous fermentations are prone to arrest before all grape juice sugars have been metabolised, leaving the partially fermented wine susceptible to spoilage by other undesirable microorganisms. However, it is recognised that wines made by this method of fermentation may have more complex and interesting sensory characteristics (Heard 1999; Romano etal. 2003). Alternatively, fermentations may be guided by inoculation or seeding with selected strains of yeasts. Generally strains of Saccharomyces cerevisiae or Saccharomyces bayanus are inoculated at initial populations of 10^-10^ cells/ml of juice. Inoculated fermentations are typically of shorter duration and are fairly vigorous, and allow a greater degree of control and predictability over the process and wine quality (Boulton et al. 1995). However, indigenous yeasts are always present in grape juice and will contribute to the fermentation according to their ability to compete with the inoculated yeasts (Fleet 2003).

Alcoholic fermentation is considered complete when fermentable sugars, glucose and fructose of the juice, are completely utilised (final concentrations less than 2-5 g/1). The wine is then drained or pumped (racked) from the sediment of yeast and grape material (lees) and transferred to stainless steel tanks or wooden barrels for malolactic fermentation, if desired, and aging (Fleet 1998; 2001). 2.2.4 Malolactic fermentation (MLF) Malolactic fermentation is a secondary fermentation caused by growth of lactic acid bacteria. MLF may occur naturally by the growth of lactic acid bacteria resident in wines. However, inoculation with commercial cultures of Oenococcus oeni (formerly Leuconostoc oenos) is widely used to encourage MLF (Dicks et al 1995; Costello et al 2003). The principal reaction in MLF is the decarboxylation of L-malic acid to L-lactic acid, and one of its main effects is the reduction of total acidity and mellowing of the tartness in the wine. Additionally, MLF produces bacterial metabolites (acetaldehyde, acetoin, diacetyl, volatile acids and 2, 3-butanediol) that may enhance the flavour complexity of wines (Davis et al. 1985; Lonvaud-Funel 1999; Bartowsky et al. 2002).

MLF is not necessarily beneficial to all wines. Wines produced from grapes grown in warm climates have lower concentrations of malic acid. Further reduction in acidity by MLF is deleterious to overall sensory balance, and it also increases wine pH to values where spoilage bacteria are more likely to grow. However, preventing the natural occurrence of MLF in these wines is an extra technical burden. Consequently, many winemakers prefer to encourage MLF and later adjust wine acidity, if necessary (Fleet 2001).

2.2.5 Post fermentation processes Most wines are not stored for lengthy periods after completion of fermentation. If storage is necessary, it is generally done in stainless steel tanks. Some white wines (e.g. Chardonnay) may be aged in wooden barrels. Premium quality red wines are aged for periods of 1-2 years by storage in oak barrels. During this time, chemical reactions that contribute to flavour development occur between wine constituents and components extracted from the wood of the barrels. Critical points for control during storage and aging are exclusion of oxygen and addition of sulfur dioxide to free levels of 20-25 |ig/ml. These controls are necessary to prevent the growth of spoilage bacteria and yeasts and to prevent unwanted oxidation reactions (Fleet 2001).

Before bottling, wines may be cold stored at 5-10 °C to precipitate excess tartrate and then clarified by application or more processes which include the addition of fining agents (bentonite, albumen, isinglass, gelatin), centrifugation, pad filtration, and membrane filtration. For some white wines with residual sugar, potassium sórbate may be added to control yeast growth (Rankine 1989; Ewart 2003).

2.3 ROLES OF MICROORGANISMS IN WINEMAKING The surface of grape berries represents a phyllospheric habitat for a diversity of filamentous fungal, yeast and bacterial species that have various impacts on the efficiency and quality of wine production.

2.3.1 Filamentous fungi Fungal diseases of grapes result in major losses of juice yield, reduced vine productive life, and increased vineyard management costs. Moreover, they can produce off- flavours and taints in the juice and wine, and can be responsible for the presence of mycotoxins (e.g. Ochratoxin A) in wines (Fleet 2003).

The filamentous fungi responsible for significant grapevine diseases in Australia have been described by Emmett et al. (1992). Infections of grapevines with Plasmopara vitícola, Uncinula necator (syn. Oidium tuckeri), Elsinoe ampelina, Eutypa lata and Phomopsis vitícola, are responsible for causing downy mildew, powdery mildew, black spot, Phomopsis cane and leaf spot, respectively. In other regions such as Portugal, South Africa and California, the decline of young vineyards, known as young esca, is associated with infections by Phaeomoniella chlamydospora and species of the Phaeoacremonium (Oliveira et al. 2004). Fungal species which attack fruit tissues and cause rotting of ripening berries, include Botrytis cinerea (bunch rot). Pénicillium spp. (blue rot), Aspergillus niger (black mould rot), Rhizopus stolonifer and Rhizopus nigricans (Rhizopus rot), Cladosporium herbarum má Alternaría tenuis (sour rot) (Emmett et al. 1992; Pitt and Hocking 1997). Growth of fungi on grapes, in juice, wooden storage containers and corks may impact on the sensory properties of wine through the formation of potent sensory metabolites. For example, the volatile compound, l-octen-3-one, described as an intense, characteristic mushroom odour, has been identified from berries infected by Uncinula necator (Darriet et al. 2002). Other fungi. Pénicillium, Aspergillus, Mucor, are also known to cause off-flavors with bitter, straw and plastic sensory descriptors. The production of chloroanisoles (2,4,6- trichloroanisole) by Trichoderma longibrachiatum is thought to be responsible for the musty/earthy cork taints in wine (Lee and Simpson 1993; Álvarez-Rodriguez et al. 2002; Coque et al. 2003). In addition to the spoilage of grapes and wines, the presence of mycotoxins in wines is a significant problem that is now recognised worldwide. Grapes contaminated by the hldiok Aspergillus species (A. carbonarius, A. niger) are known to be major sources of Ochratoxin A in wines (Sage et al. 2002; Serra et al. 2005, 2006; Leong et al 2006a, 2006b).

Although fungal colonization of grapes potentially lowers grape and wine quality, infection with Botrytis cinerea can be unique among problematic fungi. In certain circumstances, grapes infected with B. cinerea results in a condition known as 'noble rot', in which berry desiccation and grape sugar concentration is facilitated, allowing the production of highly prized sweet wines (Donèche 1993; Ribéreau-Gayon et al. 2000).

2.3.2 Yeasts The role of yeasts in winemaking has been the subject of extensive studies that have been reviewed in literature (Kunkee and Amerine 1970; Amerine 1985; Benda 1982; Kunkee and Bisson 1993; Heard and Fleet 1993; du-Toit and Pretorius 2000; Lambrechts and Pretorius 2000; Pretorius 2000; Fleet 2003). It has been recognized since the work of Louis Pasteur that yeasts are responsible for conducting the key process of alcoholic fermentation of the grape juice, during which a vast array of flavour contributing metabolites are produced. Yeasts catalyze the transformation of neutral grape components into flavour active components and also influence the sensory profile of wines by this mechanism (Pretorius 2000; Fleet 2003; Sweigers et al. 2005). The contamination and growth of undesirable species during alcoholic fermentation or during storage and bottling may cause unacceptable appearance, flavour and aroma properties, leading to loss of product. Flavour defects include ester taints by Hanseniaspora uvarum, mousy taints by Dekkera/Brettanomyces and Pichia guilliermondii, and film and off-flavours by Candida, Metschnikowia and Pichia species. Zygosaccharomyces bailii, Schizosaccharomyces pombe and S. cerevisiae may referment wines during storage and after packaging. Species of Saccharomyces and Brettanomyces may lead to haziness, turbidity and the production of volatile acidity in packaged wines (Sponholz 1993; du-Toit and Pretorius 2000; Loureiro and Malfeito- Ferreira 2003). Yeast may also affect wine flavour and other properties through their autolysis and bioadsorption of grape juice components (Fleet 2003; Moruno et al. 2005). Additionally, they may influence the growth of malolactic and spoilage bacteria positively or negatively by producing bacterial growth nutrients and antagonistic substances (Fleet 2003; Alexandre et al. 2005; Viljoen 2006).

These influences vary with the species and strains of yeasts that grow throughout the wine making process. Consequently, the individuality of wine character and quality depends, in part, on the yeast ecology of the grape juice fermentation and wine maturation processes. It is, therefore, important to have sound knowledge of this ecology and in particular, how these yeasts enter the winery environment.

Saccharomyces cerevisiae and S. bay anus have become universally accepted as the principal yeasts of alcoholic fermentation. Many fermentations are, today, conducted by inoculation with cells prepared from commercial cultures of these yeasts. Nevertheless, a diverse population of up to 15 species of non- Saccharomyces yeasts may be involved in the fermentation. These include species and strains of Candida, Hanseniaspora (anamorph Kloeckera), Khtyveromyces, Metschnikowia and Pichia, which grow during the initial stages of fermentations before dying off and giving way to dominance by S. cerevisiae or S. bayanus (Fleet and Heard 1993; Pretorius 2000; Fleet 2001). These represent indigenous yeasts that originate from the grape surface and winery equipment. It is often assumed that any inoculated S. cerevisiae and S. bayaniis will overwhelm and diminish the growth of these indigenous yeasts. However, even in these inoculated ferments, indigenous flora will always be present and will make variable contributions to the process, depending on their competitive success with the inoculated S. cerevisiae or S. bayanus. Even though the indigenous non-Saccharomyces yeasts die off in the early stages of fermentation, they will have previously grown to significant populations and put their imprint on wine character (Heard and Fleet 1985; Heard 1999). Also, there will be contributions from indigenous strains of S. cerevisiae and S. bayanus (Fleet 2001,2003). Ecological studies of wine fermentations conducted in different regions around the world have confirmed the vast biodiversity of yeast species and strains associated with this process. Whereas winemakers once saw this biodiversity in a largely negative context, they now have a clearer understanding of its significance and seek innovation in using this knowledge to enhance wine value. Such innovation includes more strategically managed indigenous fermentations and the use of novel species and strains in controlled, inoculated fermentations. The link to such novel innovations, as well as better understanding and management of existing technology, is the yeast ecology of grapes. As indicated already, grapes are a primary source of yeasts for wine fermentation, but little quantitative detail is known about the populations and species principally associated with grapes. It has been noted that some of the finest wines produced in Europe originate from indigenous fermentations (Heard 1999; Fleet 2001; Pretorius 2000).

2.3.3 Bacteria Lactic acid bacteria are significant in winemaking because they are responsible for malolactic fermentation (MLF) and because they cause spoilage. The growth of lactic acid bacteria, primarily Oenococcus oeni functions to decrease wine acidity by decarboxylation of L-malic acid into L-lactic acid. MLF may also enhance wine flavour and complexity and increase the microbial stability of wine by the removal of residual nutrients and production of bacteriocins (Fleet 1998, 2001). Various species of Pediococcus and Lactobacillus can also conduct MLF, but they often give unpleasant off-flavours (Bartowsky et al. 2002). The growth of lactic acid bacteria in the early stages of alcoholic fermentation may inhibit wine yeasts, leading to stuck or sluggish fermentation (Bisson 1999; du Toit and Pretorius 2000).

Acetic acid bacteria (Acetobacter and Gluconobacter spp.) cause the vinegary spoilage of wines through the oxidation of ethanol to acetaldehyde and acetic acid. Growth of these organisms on grapes before fermentation releases metabolites which affect wine flavour and could also retard alcoholic fermentation (Drysdale and Fleet 1988).

Spoilage by acid- and ethanol-tolerant species of Bacillus (B. coagulans, B. circulans and B. subtilis) has previously been reported in desert wines and brandy. On rare occasions, juices and wines of high pH (e.g. pH 4.0) have been spoiled by the growth of Clostridium butyricum, causing acidic (butyric) off-flavours. Growth of Actinomyces and Streptomyces spp. in corks and wooden barrels contributes to musty, earthy or corky taints sometimes found in wines (Fleet 1998, 2001).

2.4 THE GRAPE SURFACE AS A MICROBIAL ECOSYSTEM

The process of grape surface colonization by microorganisms follows the stages of contamination, attachment and immobilization of cells and accessing nutrients for survival and growth. As a basis to understanding the colonization process, the following section provides some background information about the physical and chemical properties of the grape surface and the development of the grape berry. In addition, the growth and survival of microorganisms on grape surfaces will also depend on their physiological and biochemical properties. Colonization events will also be influenced by extrinsic, vineyard environmental and management practices. These factors will be discussed in section 2.5.

2.4.1 The physical and chemical properties of grape surfaces Grape berries are covered by a cuticular material, which consists of a cutin layer and waxes that are embedded within the cuticle (intracuticular wax) or deposited on the outer surface of the cuticle (epicuticular wax). Cutin, the main structural component of the cuticle, is a polymer matrix primarily composed of hydroxyl and hydroxyepoxy fatty acids (10, 16-dihydroxy hexadecanoic acid and 9,10,18-trihydroxy octa-decanoic acid). Cuticular waxes, both intra- and epicuticular, are mainly formed by oleanolic acid (C30H48O3), constituting about 60-80% of the total wax (Commenil et al 1997). The remaining components are hydrocarbons (C14-C32), free alcohols (C20-C34) and free acids (C14-C32), as well as their esters and aldehydes (Radler and Horn 1965; Commenil et al 1997; Casado and Heredia 1999). Some original data on the chemical composition and physical organisation of the cuticle layers can be found in the studies of Chambers and Possingham (1963), Radler and Horn (1965), Baker (1970), Gmcarevic and Radler (1971), Considine and Knox (1979), Rajaei (1987), Rosenquist and Morrison (1989), Hardie et al (1996) and Cassado and Heredia (1999). The cuticular membrane controls water movement from grape berries (Chambers and Possingham 1963), and also shields against microbial infection or from insect attack (Commenil et al. 1997). Its physiochemical characteristics determine the relative adhesion of water, retention of agrichemical sprays and influence adhesion of microbial spores and airborne particles. These functions are related to the content, structural arrangement of the wax and its lipophilic composition (Marois et al. 1985; Rosenquist and Morrison 1989).

The amount, structure, thickness and density of the epicuticular wax layer vary with grape cultivar. For example. Cabernet franc grapes have about three times more epicuticular wax than Riesling grapes. Compared with Riesling grapes. Cabernet franc grapes show a relatively larger size and are more distinct in platelet structure. The structure of waxes also changes with fruit age; initially appearing as a folded layer with distinct ridges, it eventually evolves into a continuous uninterrupted layer of decreased thickness as berries expand (Rajaei 1987; Rosenquist and Morrison 1989; Percival et al. 1993; Commenil et al 1997). The proportion of berry surface covered with wax is increased when berries are exposed to sunlight, and at higher temperatures and humidity (Rosenquist and Morrison 1989; Percival et al. 1993; Spayd et al. 2002).

The cuticle serves as the interface where interactions between microorganisms and the grape berry occur. The waxy-cuticle layer may present an effective barrier to yeast and fungal colonization (Martin et al. 1957; Blakeman and Sztejnberg 1973) or could influence the specificity of yeast association with grapes.

2.4.2 Nutrients for yeast growth on grape surfaces The chemical composition of the waxy-cuticle surface has already been mentioned. The plant lipidic, hydrocarbon components do not appear to be a nutritive source for most microorganisms (Leone and Breuil 1999; Belding et al. 2000). However, leachates or materials exuded materials from the inner grape tissue (endocarp) to the surface may be adequate to support yeast growth. For example, exudates including a variety of organic and inorganic compounds, such as sugars, organic acids, amino acids, methanol and various salts have been shown to support large microbial populations on plant leaves (Mercier and Lindlow 1999). Electron microscope observations reveal that yeasts (Belin 1972) and the bunch rot-fungus Botrytis cinerea (Bessis 1977) are densely localised in the region around the peristomatic regions, where there is increased amounts of exudates.

Table 2.2 shows the main components of grape tissue and how they change with grape maturity. Immature berries contain very little sugars (2.0 mg/g) such as glucose and fructose, and high contents of malic and tartaric acids (20-30 mg/g), giving the internal tissue a pH of 2.5-3.0. As the berry matures, especially during veraison, the total sugar concentration increases to 150-300 mg/g, while concentrations of malic and tartaric acids decease 5-10 mg/g, and the internal pH increases to 3.0-3.5 (Ribereau-Gayon et al 2000). Assuming some leaching or translocation of tissue components to the grape surface, higher amounts of sugars will be available for yeast growth in more mature grapes.

Table 2.2 Main chemical components of grape tissues

Stages of grape Berry tissue composition maturity Sugars Acidity Bnx° Glucose Fructose pH TA" (%) Malic Tartaric acid (mg/g) (mg/g) acid (mg/g) (mg./g) Berry formation 4-6 2-4 0-2 2.5-3.0 1.99-2.75 20-30 15-30 (weeks 1-7) Veraison 8-12 60-70 55-65 3.0-3.3 1.29-1.69 5-10 6-10 (weeks 7-8) Ripening stages 18-22 105-120 115-120 3.3-3.5 5-6 5-6 (weeks 9-12) ^ titratable acidity Data abstracted from several references (Esteban et al. 1999; Bamavon et al. 2000; Varandas et al. 2004)

In unripe berries, more than half of the total nitrogen is represented by the ammonium cation. As grapes mature, the ammonium concentration decreases, while the organic nitrogen content increases (Boulton et al 1995). The total nitrogen in the seeds and skins is relatively high compared with the pulp. Complex forms of nitrogen compounds are found in the grape juice of mature grapes.

Several authors have detected sugars, organic acids and amino acids in distilled water rinses of the surfaces of healthy, intact grape berries and have suggested that these originate as exudates from the inner tissue. Kosuge and Hewitt (1964) determined that glucose and fructose increased 10 fold on the surface of grape berries as they progressed toward maturity. They also reported amino acids in the surface rinses, but their total concentration remained relatively constant throughout the maturation process. In a similar study by Padgett and Morrison (1990), a 10 fold increase in grape surface sugars was found during maturation. They also reported a high concentration of malic acid and phenolic compounds on the surfaces of immature grapes that decreased substantially during maturation, thereby decreasing their inhibitory influences on the growth of spoilage fungi, such as B. cinerea. Interestingly, they could not detect tartaric acid on the surfaces of grapes, despite its high concentration in inner tissues. They speculated that the process of exudation was regulated and compound specific. More recent studies by Varandas et al. (2004) also demonstrated the presence of glucose and fructose on the grape surfaces, and their increase in concentration at this location during berry ripening.

It is evident from the few studies conducted to date, that some constituents of the inner grape tissue fmd their way to the grape surface. The mechanism of this translocation is not clear, but could involve general diffusion and leakage through pores, fissures or stomata. Irrespective of the mechanism, these surface constituents may either serve as nutrients that support the growth of microorganisms such as yeasts, or they may be inhibitory to microbial growth. Their influence could possibly be selective, leading to the evolution of particular yeast species on the grape surface during berry development. Factors affecting the permeability of exudates include grape cultivar, their cuticular and skin properties, environmental conditions such as rainfall, irrigation and temperature fluctuation. The application of agrichemicals is another factor that requires consideration. Pesticide preparations after reconstitution in farm water have been shown to support rapid and significant microbial growth (T^Jg et al. 2005). Guan et al. (2001) demonstrated that adjuncts present in agrichemical preparations could also serve as nutrients for microbial growth. 2-5 THE YEAST ECOLOGY OF WINE GRAPES During the past 100 years, many researches have described the association of yeasts with wine grapes. The reviews of Amerine and Kunkee (1968), Kunkee and Goswell (1977), Benda (1982) and Kunkee and Bisson (1993) give historical accounts of these studies and listings of the yeast species found. However, most studies represent simple, qualitative descriptions of the yeasts isolated and identified, with little attempt to ask why certain species predominate and what factors affect their occurrence (Fleet et al. 2002). This section summarises data from key surveys (Table 2.3) and reviews the various intrinsic, extrinsic and implicit factors that affect the growth and survival of yeasts on the grape surface.

2.5.1 Origin of yeasts on grapes In the vineyard, grapes are exposed to yeasts that occur in the soil, air, wind-blown dust, aerial vine parts and other vegetation, rain, irrigation water, fertilisers and other agrichemical sprays. Vineyard soil serves as a natural or temporary reservoir for a wide range of microorganisms, and wind-blown infested soil is likely to be one of the main sources of microorganisms on the surface of grape berries. The yeast composition of vineyard soils has been reported by Capriotti (1960), Benda (1964), Parle and di Menna (1966), Davenport (1973, 1974), Poulard et al (1980) and Sabate et al (2002). Generally, yeasts were present at populations of 10^-10^ CFU/g soil and comprised basidiomycetous yeasts such as Cryptococcus albidus, Cryptococcus tereus, Cryptococcus curvatiis, Rhodotorula mucilaginosa and species of Bullera, Sporobolomyces and Trichosporon. Ascomycetous yeasts were represented by species oi Candida, Aureobasidium pullulans, Hanseniaspora uvarum and Metschnikowia pulcherrima. Fermentative yeasts such as Saccharomyces cerevisiae were notably rare or absent (Benda 1964; Davenport 1974), although they have been isolated by enrichment culture from vineyard soil (Poulard et al 1980).

The atmosphere plays a role in redistribution of microorganisms in the phyllosphere, acting as both source and sink. Air samples collected around vineyards and horticultural sites contain few yeasts (<100 CFU/m^) and represent about 5% of the total air flora. However, A. pullulans, Kloeckera apiculata, Candida pulcherrima and species Brettanomyces, Rhodotorula and Sporobolomyces have been isolated from air and wind-blown dust particles in vineyards (Ingram and Liithi 1961; Adams 1964; Benda 1964; Davenport 1974).

There is a strong view that insects are probably the most significant source of contamination of grapes and fruits in general, although specific details of this relationship require definition (Mortimer and Polsinelli 1999). The general relationships between insects, yeasts and plants have been reviewed by Phaff and Starmer (1987) and Ganter (2006). Many insects including honey bees, halictus bees, syrphid flies (Scaeva pyrastri) and wasps visit grape blossom. They are attracted to flowers, primarily for pollen and nectar, and sometimes gather honeydew from grape vine leaves. Various species of Candida, Kloeckera, Metschnikowia, Saccharomyces and A. pullulans have been isolated from the bodies of bees, wasps and beetles and their intestinal contents (Svejcar 1968; Davenport 1976; reviewed in Benda 1982). Drosopholid fruit flies (predominantly Drosophiloa melanogaster) feed on berries, and females may oviposit within grapes. Yeasts represent an important part of the drosophilid diet, and in some cases, similar yeast species can be recovered from both fruits and crop of the insect from which they feed (de Carmago and Phaff 1957; Moráis et al 1995). In addition to the insects just mentioned, there is a diversity of moths {Lobesia botrana, Eupoecilia ambiguerra), mealybugs {Pseiidococcus spp.), beetles {Harmonía axyhdis, Carpophilus spp.), aphids, mites, leafrollers and leafhoppers that visit, feed or infect grapevines at various stages of their annual development (Buchanan and Amos 1988). It is likely that these visitors will contribute to the yeast flora of grapes, but no information is available on this topic. Birds also feed on grape berries; significantly, they puncture the fruit and together with attracting insects, promote the contamination and development of yeasts, filamentous fungi and bacteria (Bourdreau 1972; Buchanan and Amos 1988; Tracey and Saunders 2003).

Depending on the proximity of the vineyard to the winery, grape pomace and water run off from the winery sites have been identified as sources of S. cerevisiae found in the vineyard (Valero et al 2005). 2.5.2 Yeast species associated with wine grapes As mentioned previously, the chemical composition of grapes changes as the berry matures, and this is reflected in the diversity of nutrients available at the surface. Such changes are likely to impact on the populations and species of yeasts that colonize the grape phyllosphere. Table 2.3 lists the yeast species isolated from grapes at different stages of maturity.

2.5.2.1 Grape cultivar Several authors have observed that the species of yeasts associated with grapes may vary with cultivar (Benda 1962, 1964; Minarik 1965; Goto and Yokotsuka 1977; Suarez et al. 1994; Raspor et al. 2006). However, no definitive patterns can be drawn from these data (Table 2.3) and more systematic investigation is needed.

2.5.2.2 Immature grapes At the stages of budding and flowering, small populations of yeasts, mainly Rhodotorula spp. are found (van Zyl and du Plessis 1961). Grapes prior to veraison do not harbour many yeasts; generally at populations of less than 10^ CFU/cm^. The predominating species at this stage include A. pullulans, species of Cryptococcus {Cr. albidus, Cr. laurenti), Rhodotorula {Rh. glutinis, Rh. mucilaginosa, Rh. rubra), species of Sporobolomyces, Sporidiobolus, Rhodosporidium and Candida (C. diffluens, C. famata) (van Zyl and du Plessis 1961; Parle and di Menna 1966; Davenport 1974, 1976; Rosini et al 1982; Suarez et al. 1994; de la Torre et al. 1998a). In some cases, Hs. uvarum/K. apiculata have also been isolated from immature grapes, but not consistently (Goto and Yokotsuka 1977; Yanagida et al 1992; Suarez et al 1994).

2.5.2.2 Mature grapes As grapes mature to harvest ripeness, yeasts become more abundant, increasing to populations of lO'^-lO^ CFU/cm^. Most species present on immature grapes can be isolated from mature grapes, but at this stage, K. apiculata and its sporulating anamorph, Hs. uvarum are frequently reported as the dominant species and can account for over 70% of the total flora. Nevertheless, there are studies that have reported the absence of K. apiculata/Hs. uvarum on mature grapes (Parish and Carrol 1985; Yanagida et al. 1992; de la Torre et al. 1999; Jolly et al. 2003), and the reasons for these discrepant observations are not evident. Other prominent yeast species occurring on ripe grapes include M pulcherrima, and species of Candida, Klicyveromyces and P/c/z/<3 (Barnett a/. 1972; Davenport 1976; Rosini et al 1982; Martini et al 1996; Sabate et al 2002; Combina et al 2005; Raspor et al 2006).

Table 2.3 Yeast species isolated from grapes at different stages of maturity

Immature grapes Location and grape Yeast species and population Reference cultivar New Zealand: Chardonnay Cr. albidus (6%), Cr. difjfluens (9%), Rh. mucilaginosa (5%), Rh. minuta {\2%l A. puUulans (\4%) Parle and di Menna (1966) Pinotage C. mycoderma (18%), C. scottii (3%), Cr. albidus (33%), Rh. mucilaginosa (34%), Rh. minuta (1%), A. pullulans (] 1%) Spain: Gamacha C. pulcherrima (14%), C. solam (14%), K. apiculata (57%), K. lafarii {Ì4%) Suareze/a/. (1994)

Cabernet sauvignon C. pulcherrima (33%), K. lafarii (67%) Spain: Sp roseus (3-115 CFU/g), Cr. albidus (2-24 CFU/g). Cr.flavm (0-6 CFU/g). Cr. macerans (0-2 CFU/g), Cr. laurentii (0-1 CFU/g), Rh. minuta (0-2 CFU/g), C. apis (0-1 CFU/g), Schizobl. starkeyi-henricn (0-1 CFU/g) Tempranillo de Rioja de la Torre et al. Sp. roseus (2-461 CFU/g), Cr. albidus {\-\ 16 CFU/g), Cr. flavus (0-6 CFU/g), (1999) Rh. rubra (0-3 CFU/g), Rh. glutinis (0-13 CFU/g), Rh. minuta (0-10 CFU/g), Pedro Ximenez Cr. macerans (0-7 CFU/g), C. apis (0-1 CFU/g), C. famata (0-8 CFU/g), C. vanderwaltii (0-7 CFU/g), Hs. valbyensis (0-5 CFU/g). Schizobl. .starkeyi- henricn (0-0.2 CFU/g) Mature grapes New Zealand: Chardonnay K. apiculala (65%), cerevisiae (17%) Parle and di Menna (1966) Pinotage K^^ici^ata (90%), S. cerevisiae QV^ Japan: Apiculate yeast (58-76%): K. aptculaia, K. africana Koshu Saccharomyces (9-12%): S. cerevisiae, S. bayanus Candida (13-19%): C. stellata, C./amata Film yeasts (3-10%): P. membranifaciem, P. ohmeri, C. valida, C. krusei, C. guilliermondii, P. anomala Rhodotorula (1-3%): Rh. rubra, Rh. glutinis Other yeast: (1-4%): M. pulcherrima, Cr. luteous, Cr. laurentii, S'codes ludwigii Goto and Yokotsuka Muscat Bailey Apiculate yeast (40-54%): K. apiculala, K. africana (1977) Saccharomyces (8-18%): S. cerevisiae, S. bayanus Candida: (14-16%): C. stellala, C. famata Film yeasts (13-22%): P. membranaefaciens, P. ohmeri, C. valida, C. krusei, C. guilliermondii, P. anomala Rhodotorula ( 1-3%): Rh. rubra, Rh. glutinis Other yeast: (2-6%): M pulcherrima, Cr. luteous, Cr. laurentii, S'codes ludwigii Japan: K. apiculata (0-100%), Cr. laurentii (0-6%) Niagara Cr. laurentii (70-75%), Rh. glutinis (17-30%), C. pulcherrima (0-8%) Chardonnay Yanagida et at. (1992) K. apiculata (0-76%), Cr. albidus (0-20%), Cr. laurentii (4-40%), Rh. glutinis Zenkoji (0-50%), Rh. minuta (0-9%), C steatolytica (0-20%)

Koshu K. apiculata (0-100%), Cr. albidus (0-70%), Cr. laurentii (0-30%) Table 2.3 continued

Location and grape Yeast species and population Reference cultivar Mature grapes Spain: Tempranillo Rh. rubra (29%), C. mycoderma {\5%X C. pulcherrima (29%), K. apiculaia (29%) Suarez et al. (1994) Garnacha C. pulcherrima {\A%\ C. solani (38%), K. apiculaia (38%), K. lafarii {\3%)

Cabernet saiivignon C. pulcherrima (\3Vo), K. lafarii (50%), T. rosei (13%), X veronae (25%) Spain: Sp. roseus (4-610 CFU/g), Cr. albidus\ (0-68 CFU/g), Rh. rubra (0-47 CFU/g) de la Torre et al. Tempranillo de Rioja (1999) Sp roseus (2-540 CFU/g), Cr. albidus (Ì-290 CFU/g), C apis (0-1 CFU/g), PedrSpaino :Ximéne z Cr. uniguttulatum (65%), Cr. oler {15%), Cr. laurentii (10%), A. pullulans Garnacha, 1995 (10%) Sabate et al. (2002) Garnacha, 1996 A. pullulans (100%)

Carinyena, 1995 Hs. uvarum (75%), C. zeylanoides (\0%), Cr. unigutlulatum (10%), A. pullulans (5%) Cariyena, 1996 C. zeylanoides (90%), A. pullulans {\QVo) Slovenia: Zametovka; Basidiomycetes (70-85%): Rhodotorula spp., Sp. roseus, Cryplococcus spp., sites A,D,E A. pullulans {\5-2>\%)

Zametovka;site C A. pullulans (78%), Hs. uvarum (17%), M. pulcherrima (5%)

Zametovka; site B Hs. uvarum (84%), M. pulcherrima (6%), F. kiuyveri (6%), F. memhranifaci ens (4%) Raspor et ai (2006) Kraljevina; sites B, C, E Basidiomycetes (100%); Rhodotorula spp., Sp. roseus, Cr. laurentii

Kraljevina; Basidiomycetes (72%); Rh. glutinis, sites A Ascomyctes (28%); Hs. uvarum, M. reukaufn, A. pullulans

Kraljevina site D Basidiomycetes (94%); Rh. ^lutinis Overrioe granes Germany: Z. glohiformis {T. delbrueckii), S. stellatus (C stellata), S. hacillaris {C. Kroemer and various stellala) Krumbholz(1931), England: S. hailii (Z. baila), S. rowcii (Z. rouxii), P. membrawfaciens, F. anomala, Hs. Davenport (1974) various valbyensis, Hs. uvanim, K. africana Spain: Sp. roseus (13-465 CFU/g), Cr. albidus (11—107 CFU/g), Rh. rubra (0-47 Tempranillo de Rioja CFU/g), Cr. macerans (0-6 CFU/g), C.famata (0-5 CFU/g), S. cerevisiae (0-2 CFU/g), Rh. glutinis (0-2 CFU/g), Cr. laurentii (0-1 CFU/g) de la Torre et al. (1999) Pedro Ximénez Sp. roseus (25-1229 CFU/g), Cr. albidus (4-260 CFU/g), Cr.Jlavus (0.3-190 CFU/g), Rh. rubra (0-91 CFU/g), C.famata (0-30 CFU/g), Rh. glutmis (0-0.3 CFU/g), Schizobl. slarkeyi-henncii (0-0.3 CFU/g)

2.5.2.4 Overripe and damaged grapes Overripe berries (shrivelled or mummified grapes on vines) and damaged grapes have lost the integrity of the exocarp and nutrients of the inner tissue are freely available for microbial growth. Yeast populations exceeding 10^ CFU/g are found, in combination with similar populations of acetic acid bacteria, lactic acid bacteria and filamentous fungi. The development of noble rot and other spoilage forms (bunch and sour rots) are accompanied by the increased populations of species sucli as Candida stellata, Candida krusei, M. pulcherrima, and K. apiculata. Fermentative species of Saccharomyces, Torulaspora and Zygosaccharomyces are also prominent in these grapes (Le Roux et al. 1973; Davenport 1976; Giferzoni and Marchetti 1987; Yanagida g/a/. 1992; Laurent 1998; Bene and Magyar 2004).

2.5.3 Saccharomyces cerevisîae Special mention is made of .S. cerevisiae here, because of its principal role in the alcoholic fermentation of grape juice into wine, and because there is much debate and controversy in the literature about its occurrence on grapes. While some authors believe that the grape berry is the natural origin of this species in wine fermentation, others do not agree with this view and believe the winery equipment to be the main source (Martini et al. 1996; Török et al. 1996). Many authors have not been able to isolate S. cerevisiae from either immature or mature grapes using direct agar plating methods (Bamett et al. 1972; Rosini et al. 1982; Yanagida et al. 1992; Suarez et al 1994; Sabate et al. 2002; Raspor et al. 2006), or have found it infrequently compared with other species (van Zyl and du Plesis 1961 ; Parle and di Menna 1966; Goto and Yokotsuka 1977; Parish and Carroll 1985; de la Torre et al 1999). Overall, the many studies suggest that S. cerevisiae does occur on grapes in the vineyard, but its population is too low (<10-100 cells/cm^) to be detected by direct plating techniques, or that there is variability in the yeast ecology within the cluster or bunch of grapes (Rosini et al 1982; Mortimer and Polsinelli 1999). The possibility of its isolation by this method increases as the grapes mature and progress to overripe, damaged condition. Mortimer and Polsinelli (1999) estimated that one in four damaged berries contain cells of S. cerevisiae, compared with one in a thousand healthy berries. Saccharomyces cerevisiae is readily isolated from mature grapes by enrichment culture (Mrak and McClung 1940; Török et al 1996; van der Westhuizen et al 2000a, 2000b) and can frequently be isolated from natural/spontaneous fermentations with aseptically collected grapes (healthy or damaged berries) (Mortimer and Polsinelli 1999; Lopes et al. 2002; Antunovics et al 2005; Schuller et al 2005). Other species of Saccharomyces (S. bayaniis, S. paradoxus, S. uvarum) have been reported on grapes (Demuyter et al 2004; Redzepovic et al 2002). Additionally, genetically diverse strains within S. cerevisiae and S. bayaniis have been isolated from grapes (Pramateftaki et al 2000; Capello et al 2004; Schuller et al 2005). Some strains may be specific to particular geographic regions (Versavaud et al. 1995; Pramateftaki et al. 2000; van der Westhuizen et al. 2000a). Nevertheless, there are reports where S. cerevisiae has not been found after enrichment, and these data suggest that climatic factors such as rainfall, temperature and viticultural factors such as agrichemical applications may be significant in affecting its survival and occurrence (Martini et al. 1996; Khan et aL 2000; van der Westhuizen et al. 2000a, 2000b).

2.5.4 Vineyard environmental factors The surface of wine grapes could be considered to be a hostile habitat for microorganisms. The grape berry surface is exposed to a range of interacting environmental stresses including rapidly fluctuating temperature, repeated alteration between the presence of free moisture due to rain and dew, solar dehydration and radiation, and the stresses of agrichemical applications (Lindiow and Brandl 2003). These environmental factors are likely to influence the yeast ecology of grapes (Castelli 1954, 1957; Minarik 1965; Minarik and Ragala 1975). The extremes of temperature, water availability and radiation vary on a large scale with geographical location (latitude, altitude, topography), and on a smaller scale within the vineyard and vine canopy, and on a microscopic scale over the grape berry surface (Jackson and Lombard 1993). Maritime influences (Longo et al. 1991; Khan et al. 2000), wind (Davenport 1974, 1976) and soil type (Benda 1964) are further influencing factors.

Many authors have reported the influences of environmental factors on the grape microflora, but caution is needed in extrapolating these data to general conclusions due to the difficulty of conducting vigorously controlled field experiments. Castelli (1954) noted the more frequent isolation of Hanseniaspora species in warmer, southern European countries, while its asporogeneous form, Kloeckera, was found to be widespread in cooler, northern European locations. It might be concluded that sporulation enhances the resistance of this yeast to higher temperatures as generally reported for yeasts. Grapes from cooler regions of central Japan were found to favour the predominance of basidiomycetous yeasts {Cryptococciis spp., Rhodotorula spp.) than ascomycetous yeasts (Sasaki and Yoshida 1950; Goto and Yokotsuka 1977). Grapes harvested in colder, rainy vintages tend to have a higher population of yeasts and a greater proportion of oxidative, basidiomycetous species over fermentative, ascomycetous yeasts (Parish and Caroll 1985; Querol et al. 1990; Longo et al 1991; de la Torre et al 1998a, 1999). Minarik (1965) and Poulard (1984) more frequently recovered S. cerevisiae from grapes during warm, dry and sunny conditions than in cold, wet seasons. Pretorius and co-workers have reported extensive studies on the influence of seasonal and geographic factors on the incidence of S. cerevisiae and non- Saccharomyces yeasts on grapes harvested from regions of the West Cape of South Africa (Khan et al 2000; van der Westhuizen et al. 2000a, 2000b; Jolly et al. 2003). Rainfall decreased the occurrence of S. cerevisiae on grapes, but it was difficult to disassociate this effect from the greater use of fungicides on these occasions. The most prominent non-Saccharomyces yeast species isolated from these regions were K. apiculata and C. pulcherrima. However, regional and seasonal influences on these species were noted, where some samples gave no isolation of K. apiculata and others gave no isolation of C. pulcherrima (Jolly et al. 2003).

2.5.5 Agrîchemîcal influences A diversity of agrichemicals is routinely applied throughout the grape growing season to control insect and fungal grapevine pests (Emmet et al. 1992; Cabras and Angioni 2000). These applications are systematically managed from about the time of bud burst until about two weeks prior to grape harvest, to maximise their effectiveness and minimise carry-over of residues into the juice and wine. Generally, these sprays consist of either chemical or biological agents active against particular insects or fungi, a wetting agent, and water.

It has been known for many years now that some pesticides (e.g. folpet, captan, captafol and dichlofuanid) can retard the growth of yeasts and lead to stuck or sluggish fermentations (van Zyl and du Plessis 1961; Parle and di Menna 1966; Minarik and Ragala 1975; Monteil et al. 1986; Mlikota et al. 1996; Bisson 1999; Cabras et al. 1999; Cabras and Angioni 2000). Moreover, there are various reports that their application could affect the yeast flora associated with grapes, including the incidence of S". cerevisiae (Mangiarotti et al. 1987; Regueiro et al 1993; Guerra et al. 1999; van der Westhuizen et al 2000a, 2000b). Spray adjuvants (wetting agents) may alter the waxy structural layer of the berry surface and induce damage to the berry cuticle. Adjuvant application has been found to reduce the yeast populations on berries, and increase susceptibility of grapes to infection by the fungus Botrytis cinerea either through the removal of the indigenous flora, or by disrupting the surface wax morphology (Marois et al 1985; Rogiers et al 2005).

2.5.6 Yeast properties As mentioned already, the grape surface presents an environment of unique chemical and physical attributes; essentially it is a waxy, hydrophobic surface that is subject to significant diurnal and seasonal fluctuations in temperature, sunlight, irradiation, and water and nutrient availability (Andrews and Harris 2000; Lindlow and Brandl 2003). To successfully colonise this habitat, the yeasts will need to express some equally unique survival and growth mechanisms. Factors that influence the growth and survival of yeasts on the surface of grapes include attachment to the phyllosphere surface, tolerance to UV radiation and desiccation, ability to access nutrients for growth, tolerance to naturally occurring inhibitors and agrichemicals and competition with other microbial species.

The formation of extracellular mucilage is a common characteristic of phylloplane yeast species of Rhodotorula, Cryptococcus and Sporobolomyces and A. pullulans. It has been speculated that this property promotes adhesion of yeasts to aerial plant surfaces, preventing cells from being dislodged by rainfall, wind, dew deposition or irrigation waters (Dickinson 1986). The attachment oiA. pullulans to leaf surfaces has been correlated with the production of an extracellular polysaccharide (EPS) capsule (Andrews et al. 1994), while in Rhodospohdium toruloides (anamorph, Rh. glutinis), adhesion to leaves appears to be mediated by cell surface mannoproteins (Buck and Andrews 1999). Mucilage deposition is also thought to protect cells from desiccation (Dickinson 1986). Pseudohyphal growth on solid substrates has been observed in a large proportion of S. cerevisiae strains originating from grapes, and this property has been related to its ability to invasively colonise grapes (Khan et al. 2000) and grapevine plantlets (Gognies et al. 2002, 2006). Yeast species found on the phyllosphere are frequently pigmented; species of Rhodotorula and Sporobolomyces produce red, pink or yellow carotenoid pigments, cells OÍA. pullulans produce melanin, and M. pulcherrima produces pulcherrimin pigments (Davenport 1973; Dickinson 1986; Lindlow and Brandi 2003). A new class of UV-inducible compounds, mycosporins, has been found in non-pigmented yeasts {Cryptococcus species) that commonly occur on leaves (Libkind et al. 2004). These compounds are believed to protect cells against sunlight and UV radiation.

The production of extracellular hydrolytic enzymes such as pectinases, xylanases, proteases and lipases by yeasts (Charoenchai et al 1997; Strauss et al. 2000, Buzini and Martini 2002) could contribute to their ability to colonise grapes. Many Cryptococcus and Rhodotorula species are significant producers of such enzymes. De la Torre et al. (1999) have suggested a connection between xylanase production by Cr. albidus and its prevalence on wine grapes. The ability to assimilate and tolerate a diversity of phenolic and terpenic compounds may be a selective factor on the species that can grow on plant surfaces (Fonseca and Inacio 2006).

The grape surface is also a habitat for a diversity of bacteria and filamentous fungi whose presence may affect the occurrence of yeasts (Fleet 2003). The biological insecticide. Bacillus thuringiensis can be inhibitory to yeast species found on grapes, in vitro (Bae et al. 2004). Botrytis cinerea produces "botryticin" glycoproteins that can be inhibitory to some yeasts (Ribereau-Gayon et al 2000). Sipiczki (2006) reported the antagonistic activity of Metschnikowia species on a diversity of fungi occurring on grapes. There is increasing literature which demonstrates that the epiphytic yeast flora associated with grapes {Aureobasidium, Candida, Cryptococcus, Metschnikowia and Rhodotorula, species) have strong antagonistic activity against Aspergillus niger, Botrytis cinerea. Pénicillium expansum and other fruit spoilage fungi (Castoria et al. 1997; Castoria et al. 2001 ; Kurtzman and Droby 2001 ; Bleve et al 2006). These, as well as other yeast species, are being explored as potential agents for the biocontrol of fruit spoilage and mycotoxigenic fungi. 2.6 AUREOBASIDIUM PULLULANS

As mentioned previously, A. pullulans is frequently isolated from wine grapes. It was found to be the most prevalent yeast on grapes examined in this study. The significance of this association has not been fully recognised or discussed in previous literature. Because of its dimorphic, yeast/mycelial behaviour, it has not always been considered to be a true yeast. Consequently, it has been overlooked in studies by yeast researchers and equally ignored by specialists in filamentous fungi. Generally, it has fallen into the category of a yeast-like fungus, or the complex group of black yeasts. To address this gap or oversight, this section considers some background literature on^. pullulans and related species, focusing on its occurrence and significance on wine grapes.

Aureobasidium pullulans is a highly polymorphic fungal species with a complex lifecycle. When grown on agar media, it forms cream, yellow, pink, olive or black colonies, which are often covered with slimy masses of conidia and scanty tufts of aerial mycelium. The morphological variability of its colonial biomass has been widely reported (Cooke 1959; Hermanides-Nijhof 1977; Kocková-Kratochvílová et al. 1980). Sketches of gross morphology exemplifying differences in their colony pigmentation, texture and periphery are documented in Kocková-Kratochvílová et al. (1980), Figure 2.3.

The species may grow as budding cells or as filamentous mycelium. Other unicellular forms including blastoconidia, swollen cells, and melanized, thick walled chlamydospores may also develop. The expression of, and transition between these cellular forms is determined by various environmental and cultural factors such as carbon and nitrogen sources, pH, light, aeration and cell densities. Several descriptions of these lifecycles have been provided by Lingappa et al (1963), Ramos and Garcia Acha (1975a, 1975b), Petrak and Crang (1977), Kocková-Kratochvílová et al. (1980), Kocková-Kratochvílová and Hronská (1980), Bermejo et al. (1981a,b) and Park (1984). 10

••-^rp^Tfi

7 11

8

Figure 2.3 Variation in colony morphologies of^. pullulans, from Kockova-Kratochvilova etal. (1980)

1 a, margin smooth, light brown; b, aerial mycelium, white hyphae 2 a, mucous mycelium in agar, light-brown to whiteish, b; light-browtn smooth hyphae; c, wrinkled centre, pink 3 a, mycelium in agar, light-brown; b, narrow olive band with pigmentation in intercallar cells; c, creamy band (hyphae and conidia) with narrow line of chlamydospores; d, smooth, mucous, creamy centre 4 a, hyphae in agar, light-brown; b, skin-like olive mycelium; c, centre wrinkled, cinnamon 5 a, plae mucous hyphae; b, dark cinnamon with conidia; d, wrinkled centre, cinnamon 6 a, mycelium in agar, light-brown; b, skin-like olive mycelium; c, centre wrinkled, black (both hyphae and chlamydospores) 7 a, white mycelium in agar; b, black smooth band with intercalar black hypahe and chlamydospores; c. black skin-like centre with white aerial mycelium 8 a, sharp white margin of hyphae; b, black band of chylamydospores; c, white to creamy band of hyphae and blastoconidia, black wrinkled centre. Alternation of white and black bands rapidly disappears in air as the colony becomes covered witha black mucilaginous material 9 a, hyphae in agar, light-brown, white aerial hyphae; b, mucous olive band (intercalar pigmentation); c, mucous light-bro\\'n and later black band; d, black-white centre spotted 10 a, hyphae in agar; b, black skin-like wrinkled wide centre 11 a, black sharp margin; b, white to creamy band; c, brown-olive band; d, mucous black centre. The colony is rapidly covered with a black mucous material in air 12 a, sector growth, alteration of black and white sectors; b, white mucous centre with olive sectors 2.6.1 Taxonomy Aureobasidium pullulans is an asexual fungus. It belongs to a large and diverse group known informally as the black yeasts or dematiaceous hyphomycetes. These fungi are characterized by the production of darkly pigmented hyphae and their asexual states (Ellis 1971). The form genus, Aureobasidium, is morphologically related to species of Hormonema and Kabatiella. Known or suspected teleomorphs within this complex occur in the Dothideales order, Dothideaceae family of filamentous ascomycetes (Sivanesan 1984; Barr 1987; de Hoog and McGinnis 1987, de Hoog and Yurlova 1994). There are 14 accepted species in Aureobasidium, of which only A. pullulans occurs ubiquitously in nature and in food products (Hermanides-Nijhof 1977; Kockova- Kratochvilova et al. 1980). Three varieties in A. pullulans have been described; var. melanigenum, var. pullulans and var. aubasidani (Yurlova and de Hoog 1997).

Primarily, the black yeasts are classified by the structural and developmental features of their asexual states. For example, Aureobasidium and Kabatiella are distinguished from other black yeast genera by the simultaneous development of blastic conidia along the hyphae (synchronous conidiogenesis). However, in the genus, Hormonema, conidia arise in a successional manner, growing through a scar left by the release of previous conidia (percurrent conidiogenesis). Drawings, detailed descriptions of cellular and conidial morphology, and identification keys of this taxonomic group are found in Hermanides-Nijhof (1977) and de Hoog (1998). Illustrations of Aureobasidium pullulans and Hormonema dermatioides are shown in Figure 2.4.

Species and generic assignments in the Aureobasidium/Hormonema/Kabatiella complex however, are often confused because of the similarities they share in colony and cellular morphologies. The species Hormonema dermatioides and Aureobasidium pullulans both produce pinkish, slimy and rapidly expanding colonies that later become dark as chlamydospores develop. Both species are similar in thallus morphology and karyology (Takeo and de Hoog 1991; Yurlova et al. 1996). de Hoog and Yurlova (1994) also noted that percurrent development of conidia in A. pullulans could also occur, as it does in H dermatioides. The key character for differentiating these species is based on the number of conidiogeneous loci present on each expanding hyphae; 1-2 Figure 2>4a Aureobasidiumpullulans wdiv. pullulans\ A, conidial aparatus; B, dark hyphae; C, conidia (Hermanides-Nijhof 1977)

Figure 2.4b Aureobasidium pullulans var. melanigenum\ A, conidial aparatus; B, endoconidia; C, dark hyphae; D, dark arthroconidia (Hermanides-Nijhof 1977) oooooQonno

U

Figure 2.4c Hormonema dermatioides (Hermanides-Nijhof 1977) are found in H. dermatioides, while 2-14 are observed in A. pullulans (Yurlova et al. 1999). The lack of species diagnostic criteria is also confounded by the morphological heterogeneity and plasticity exhibited in this group.

An identification key based on the biochemical profiles of^. pullulans and closely related species can be found in de Hoog and Yurlova (1994). As with morphological characteristics, the authors reported that none of the biochemical or physiological tests adequately discriminated^, pullulans from H. dermatioides, with more variability found within species than between the two species (Table 2.4).

Table 2.4. Key biochemical features differentiating^, pullulans from H. dermatioides

Characteristics A. pullulans H. dermatioides D-Glucosamine w,- - R-Ribose w, + + D-Arabinose w,+ w,- methyl-a-Glucoside w,+ w,- Inulin w,+ w,- D-Gluconate w,+ D-Galacturonate + Ethanol + w, weak response; negative reaction; +, positive reaction Data from de Hoog and Yurlova (1994)

Molecular tools have now been applied to the systematic study of the black yeasts; phylogenetic studies in the 18S and 26S rDNA have been useful in defining ordinal/familial placements, and in determining teleomorph-anamorphic connections (Haase et al 1995; Spatafora et al. 1995; Berbee 1996; Eriksson et al 2001; Lumbusch and Lindemuth 2001). Taxonomic relationships within the Aureobasidium complex have also been explored using sequences in the ITS region (Yurlova et al 1999).

Central to these studies are the development of diagnostic molecular markers to reliably delineate the black yeasts at the species and sub-species level. The distinction of A. pullulans from closely related species in Hormonema and Kabatiella, including the separation of A. pullulans from H. dermatiodes was achieved by the RFLP analysis of the ITS-partial LSU regions (Yurlova et al 1995, 1996) or of the ITS region (Matteson- Heidenreich 1997). PCR-RAPD/hybridization enabled the characterization of varieties and populations within A. pullulans (Yurlova et al 1997). The development of probes specific to A. pullulans also facilitated molecular ecological studies on leaves (Li et al. 1996, 1997). SCAR- and Scorpion PGR assays have enabled the detection of specific strains oí A. pullulans of biotechnological interest (Schena et al. 2002).

2.6.2 Ecology The SX^QQXQS Aureobasidium pullulans was first described for isolates from golden spots of leaves and grapes (Cooke 1959; Hermanides-Nijhof 1977) or from seeds of grapes (Durrell 1967), although the original culture is no longer available for study. The species has since been a subject of multiple descriptions and has undergone several taxonomic revisions. Among others, (obsolete) synonyms include Dematium pullulans, Pullularia pullulans, P. fermentans, Torula schoeni, Candida malicola and Oospora pullulans (Wyne and Gott 1956; Cooke 1962, 1963; Hermanides-Nijhof 1977).

Cooke (1959) and Domsch et al. (1980) provide comprehensive summaries of its ecological history. Table 2.5 draws some examples from these accounts, and the following section summarises the presence of^. pullulans in vineyard and winery environments (without reference to its true origin/source).

Aureobasidium pullulans occurs in diverse habitats and has a worldwide distribution. It is commonly encountered in natural environments, such as soil and water, and on many natural and synthetic materials used in the household and in industry. It is a major colonizer of aerial plant surfaces, including leaves, barks, flowers and fruits (Andrews and Harris 2000; Dickinson 1976) and has accordingly, been one of most widely used organisms for studying ecological processes on this habitat (Andrews et al. 1994, Buck and Andrews 1999; Andrews et al. 2002; Woody et al. 2003). Table 2.5 Occurrence and significance of Aureobasidiumpullulans

Substrates Significance Reference Surface layers of soil; agricultural and Nutrient cycling? Cooke ( 1959); Domsch et al. ( 1980) uncultivated

Water; rivers, lakes, hypersaline Nutrient cycling? Vadkertiova and Slavikova (1995), waters, marine and esturine, Slavikova and Vadkertiova (1997), mangrove Gunde-Cimerman et ai (2000) Air; indoor and outdoor Allergenic response, Taylor et al. (2006) pulmonary infection Marble, limestone, sandstone, rock Deterioration, Urzi et al. (1999), Sterflinger and monuments and artworks discolouration Prillinger (2001) Household items; clothing, leather Contamination Cooke (1959) Wood and wood products Sapstain Schoeman and Dickinson (1996), Croan (2000) Municipal water pipe system Biofilm Doggett (2000) Synthetic polymeric substances; PVC, Biodeterioration Webb et al. (2000), Lugauskas et al. poly(tetrafluorine ethylene) (2004) Human skin, blood and other tissues Rare, opportunistic Boglinano and Criseo (2003), Hawkes mycoses et al. (2005) Fruits and vegetable, postharvest Spoilage, decay Cooke (1959), Dennis and Buhagiar (1980), Matteson-Heidenreich et al. (1997), Morgan and Michailides (2004) Seafood, cold stored meat, cheese, Spoilage, Pitt and Hocking (1997) frozen food products discolouration

In ecological surveys of wine grapes and musts, this species is one of the most commonly isolated. Its occurrence has been reported in wine-making regions throughout the world (Table 2.6).

Vineyard A. pullulans is ubiquitous in the vineyard. Martini (1993) cites the early work of MUller- Thurgau (1896) and Berlese (1896), in v/h\ch Aureobasidium {Dermatium) was consistently present in the vineyard air, soil and aerial plant parts, together with Kloeckera apiculata and other non-fermenting yeasts. Studies by Davenport (1974, 1976), Poulard et a/. (1981) and Sabate et al. (2002) also confirm this observation where pullulans was collected from the atmosphere (air, rain, dew), on surfaces of the grapevine (leaves, fruit, bark, flowers), fresh water and from vineyard soil.

This species generally grows at populations of lO^-lO"* CFU/g on grapes and occurs in 10-100% of grape samples. Its incidence and population have been shown to vary with grape varieties (Benda 1962; Parle and di Menna 1966; Raspor et al. 2006), grape maturity (Renouf et al. 2005), geographical region and season/climate (Davenport 1974, 1976).

Table 2.6 Occurrence oiAureobasidiumpullulam and Aureobasidium species on wine grapes and grape must Countr> Reference \\ ine grapes Australia Barbetti (1980), Nair (1985). Prakitchaiwattana e/o/. (2004) New Zealand Parle and di Menna(1966) Japan Sasaki and Yoshida (1959), Goto and Oguri (1983) South Africa Le Roux et al. (1973) France reviewed in Cooke (1959), Barnett ei al. (1972). Poulard et al. (1980), Sage et al. (2002), Renouf e/ al. (2005), Slovenia Raspor et al. (2006) England Davenport (1974, 1976) Germany Benda(1962, 1964) Italy Guerzoni and Marchetti (1987) Portugal Abrunhosa et al. (2001), Serra et al. (2005, 2006) Spain Sabate et al. (2002) Former Czechoslovakia Minarik(1965) Former USSR Svejcar (1968) Canada Holloway et al. (1990). Chamberlain et al. (1997). Subden et al. (2003) USA Duncan et al. (1995), Dugan et al. (2002) Grape must Spain Pardo etal. (1989) Italy Capriotti 1956 (in Cooke 1959) France Poulard and Simon (1979). Simon and Poulard (1979). Poulard et al. (1980) Germany Benda (1962. 1964), Dittrich (1977). Canada Holloway et al. (1990), Subden et al. (2003)

Data for damaged grapes is inconsistent; some studies indicate that damaged berries harbour similar populations of A. pullulans as sound grapes do (reviewed in Cooke 1959; Amerine and Kunkee 1968; Prakitchaiwattana et al. 2004). Barbetti (1980) lists this species as the most frequently isolated from whole berries affected by bunch-rot. In other studies, it was almost always absent in rotten grape samples (Nair 1985; Guerzoni and Marchetti 1987). A substantial decline in its incidence from 73% to 33% was observed on botrytis-infected grapes (Le Roux et al 1973). These inconsistent observations suggest that the interactions between pullulans with other fungi and bacteria on grapes are important and needs to be better understood.

Internal colonization of asymptomatic berries, buds and grapevine bark by A. pullulans have been demonstrated, where it was one of the most frequent/common species detected following surface disinfection of samples (Schweigkofler and Prillinger 1997; Dugan et al 2002; Serra et al 2005). This species lives endophytically in other plant hosts, including maize (Fischer et al 1992), Eucalyptus trees (Bettuci et al 1999), and other fruit trees (in Pugh and Buckley 1971 ; Schweigkofler and Prillinger 1997).

Plant-host relationship in early ecological studies, Ä. pullulans is often described as a primary saprophyte on the phyllosphere (Dickinson 1976). Fluorescence in situ hybridization studies revealed that the species is an active colonizer on the surface leaves, where its cells densely occupy the surfaces overlying veins and in wounded regions (Andrews et al 1994).

Culture filtrates of pullulans have been shown to stimulate rhizomorph development in Vitis riparia (Blaich and Ruster 1979; Blaich 1977) SLud Armillaha mellea (Pentland 1967). The compounds responsible for modulating plant growth are thought to be indole 3-acetic acid, amino acids or ethanol and other short-chained alcohols (Pentland 1967; Buckley and Pugh 1971). Conversely, this species has been implicated in causing diseases/lesions of several agricultural crops, including apple russet (Matteson- Heidenreich et al 1997), flax (Sanderson 1964) and oak (Bakry and Abourouh 1996). It causes postharvest spoilage and decay of a range of postharvest crops such as apples, peaches and pears (in Cooke 1959), strawberries (Nemec 1970) and citrus fruits (Parish and Higgins 1989). The species is associated with grapevine leaf scalding (in Cooke 1959) and melting decay of grapes (Morgan and Michailides 2004). Blaich and Ruster (1979) provided evidence of parasitism of root and plantlet of Vitis riparia. The literature provides seemingly contradictory clues as to its host-plant relationship. The reasons for its dual behaviour are obscure, but suggests that there may be more interaction with host plants than a purely commensalist relationship.

Winery A. pullulans is frequently and readily recovered from grape must (reviewed in Dittrich 1977). Freshly crushed grapes and grape juice during the early stages of wine fermentation contain 10^-10^ cfu/ml, accounting for 13-91% of the total microflora composition. Its population generally declines to non-detectable levels within 1-2 days of fermentation (Pardo et al 1989; Holloway et al 1990; Subden et al 2003). Other studies however, demonstrated the ability of this species to persist much longer during vinification; Benda (1964) showed that the species could survive in the fermentation of several grape samples after 30 days. Poulard ( 1980) recovered^, pullulans during the tumultuous phase of fermentation, and Poulard and Simon (1979) and Simon and Poulard (1979) noted its presence in both red and white wines, although their populations were not reported.

The species is a common component of the airborne fungi detected in the fermentation, storage and bottling areas of the winery. It is widespread in the wine cellar, where the high humidity provides favourable conditions for its growth. This species has also been isolated from winery walls, floors, surfaces of fermentation vats, and bottling equipment (Gandini 1969; Dittrich 1977; Belin 1969; van Keulen et al. 2003; Picco and Rodolfi 2004).

Species of Aureobasidium and A. pullulans are associated with oak. They have been isolated from cork samples {Quercus suber) (Álvarez-Rodríguez et al. 2003). Chantonent et al. (1994) described its frequent recovery from oak wood {Quercus sp.) during the ageing and seasoning of cask wood in open air.

The impact of A. pullulans in winemaking is unknown. It is not believed to be mycotoxigenic (Pitt and Hocking 1997; Roukas 2000) and has not been associated with wine spoilage nor cork taints (Lee and Simpson 1993). As with other non-fermenting yeasts, its early departure from the wine fermentation and its oxidative metabolism have led many workers to expect a negligible contribution in winemaking.

2.6.3 Biotechnological applications Aureobasidium pullulans is of commercial interest because of its ability to produce the extracellular polysaccharide, pullulan. There are over 150 patents on the production and potential uses of pullulan. This a-glucan polymer is non-toxic, biodegradable, impermeable to oxygen, and its physical properties make it attractive for a range of applications in the food, pharmaceutical and cosmetics industry, such as edible films/coatings, as fat replacers in low-calorie foods, adhesives, thickening or extending agents (reviewed in Seviour et al. 1992; Leathers 2002).

This species is known to produce a wide array of extracellular enzymes. Many reports have described its utility of its polysaccharide degrading enzymes in treating plant material, for example, in processing feedstock, in biobleaching and pulping of paper and in the conversion of lignocellulose to fuel ethanol. Important applications in the food industry include the application of pectinases in the extraction and clarification of juices, and processing of sugar and syrups with sucrases (reviewed in Deshpande et al. 1992).

Aureobasidium piillulans is a major colonizer of the phylloplane and it competes well against other microorganisms that share this habitat. Several strains have demonstrated strong antagonism against plant pathogenic fungi such as Alternaria zinnae, Botrytis cinerea. Pénicillium digitatum. Pénicillium expansum and Rhizopus stolonifer (van den Heuvel 1969; Lima et al. 1997; Schena et al. 1999; Ippolito et al. 2000; Castoria et al. 2001; Anikaram et al. 2002; Schena et al. 2003). These observations have led to the development and application of^. pullulans for the biocontrol of fungal spoilage of a range of agricultural crops and produce.

2.7 MOLECULAR METHODS FOR THE ANALYSIS OF YEASTS ASSOCIATED WITH WINE GRAPES

As mentioned earlier, there are numerous inconsistencies between studies on the yeast ecology of wine grapes. According to Martini et al. (1996), variations in analytical methods used to examine the grapes and failure to recognise the limitations inherent in these methods probably account for many of the discrepant observations. Generally, the methods used to analyse yeasts on grapes and in wines follow the standard approach used for all foods and beverages (Fleet 1993; Deak 2003; Kurtzman et al. 2003; Kurtzman 2006). Until recently, these followed recognised cultural methods. Essentially, grapes were rinsed or macerated, and the rinses and macerates were cultured on various agar media. In some cases, grapes, or their rinses and macerates were enriched in liquid media before plating onto agar media. Colonies that developed on agar plates were differentiated by their morphology, and then isolated and purified for subsequent generic and species identification. In the last 5-10 years, molecular methods have been increasingly used in the analyses of grape and wine yeasts to facilitate species identification, strain differentiation, and more recently, to examine the total yeast community in grape and wine habitats.

In this thesis, molecular methods are extensively used to identify and differentiate yeasts isolated from grapes, and as a culture independent approach for investigating the total yeast community. In addition, one section is concerned with the development of a rapid DNA probe method for the identification of grape and wine yeasts. As a prelude to these Chapters, this section gives an overview and background discussion of the molecular techniques used. The reader is referred to reviews by Giudici and Pulvirenti (2002), Kurtzman et al. (2003), Lourerio and Malfeito-Ferreiro (2003), Beh et al (2006) and Fernández-Espinar et al (2006) which outline the diversity of molecular methods widely used in studying food and beverage yeasts.

2.7.1 Methods for species identiffcation

2.7.1.1 Ribosomal DNA sequencing Sequencing of the D1/D2 domain of the large subunit (26S) ribosomal DNA in yeasts is currently the most definitive and universal approach for speciation of yeasts. This region has been sequenced for almost all known yeasts species and has been shown to contain sufficient variation to allow reliable differentiation of most ascomycetous and basidiomycetous yeasts (Kurtzman and Robnett 1998, Fell et al 2000, Scorzetti et al. 2002). The ribosomal spacer regions (ITS) often show higher rates of sequence divergence than the D1/D2 domain of the 26S rDNA. Sequencing of the ITS regions has proven useful for distinguishing closely related yeast species, where they may appear pooriy resolved or indistinguishable using the D1/D2 region (James et al. 1996; Naumov et al 2000; Cadez et al 2003). 2.7.1.2 PCR-RFLP Restriction fragment length polymorphism (RFLP) analysis of the ITS regions is emerging as one of the most useful methods for rapidly identifying food and beverage yeasts. The ITS region (ITS1-5.8S-ITS2) is amplified by PGR then cleaved with specific restriction endonucleases, and the resulting fragments are separated by gel electrophoresis. The number and size of the restriction fragments as determined by banding patterns on the gel are used to discriminate between yeast species. The restriction enzymes commonly used are: Cfo I, Hae III, Hinf\, Dra I and Dde I (Guillamón et al. 1998; Esteve-Zarzoso et al 1999; Granchi et al. 1999). This method has been frequently applied to identify yeast species associated with wine grapes and wine fermentation (Granchi et al 1999; Sabate etal. 2002; Clemente-Jimenez et al. 2004; Ganga and Martinez 2004; Combina et al 2005; Renouf et al. 2005; van Keulen et al. 2006). Several hundred species of food and beverage yeasts have now been examined by this method and databases of fragment profiles have been established (www.yeast-id.com). However, until the database becomes more extensive, the combined use of ITS-RFLP with other identification methods, such as DNA sequence analysis has been recommended (Heras-Vásquez et al. 2003).

2.7.1.3 Species-specific primers and probes Individual species of yeasts can be rapidly identified using species-specific techniques. Quantitative real-time PGR assays have been developed for Brettanomyces/Dekkera yeast species (Phister and Mills 2003; Delaherche et al. 2004), S. cerevisiae (Martorell et al. 2005) and Z bailii (Rawsthorne and Phister 2006) and can be applied directly to wine samples. Fluorescently labelled, specific probes have been used in whole cell hybridization protocols (fluorescent in situ hybridization) to identify isolates of Z). bruxellensis from wine (Stender et al. 2001; Dias et al 2003), and to monitor various non-Saccharomyces yeast species during wine fermentation (Xufre et al 2006). The application of these procedures however, is not yet widespread, because of need for expensive instrumentation. Specific probes may be immobilised onto a solid substrate, such as the wells of a microtitre tray, then modified ELISA protocol can be used to detect the hybridization event with target DNA. Using this format, a panel of probes may be developed to facilitate the simultaneous detection of multiple yeast species (Kiesling et al 2000a). Chapter 5 investigates this approach for identifying yeast species associated with grape and wine fermentations, A recent advance on this approach uses specific yeast probes attached to microspheres, and automated detection of the hybridization signal by flow cytometric technology (Diaz and Fell 2005; Page and Kurtzman 2005).

2.7.2 Methods of yeast strain differentiation Methods of strain typing have been applied in examining genetic diversity of yeast strains and their contributions to food and beverage fermentations (reviewed in Beh et al 2006; Fernández-Espinar et al. 2006). Potentially, these tools in genetic differentiation can be used to characterise yeast strains by their ecological or geographical origins (Ganter and de Barros Lopes 2000) and also allow the study of evolutionary relationships within species (de Barros Lopes et al. 1999). Techniques to differentiate strains of yeasts have also been useful in quality control measures, for example, in authenticating commercial starter strains (de Barros Lopes et al. 1996; Fernández-Espinar et al. 2001) and in tracing sources of spoilage yeast contamination (Pina et al. 2005).

A number of methods are able to differentiate yeast isolates without specific knowledge of the genome sequence. Karyotyping of genomic DNA using pulse field gel electrophoresis has been widely used to describe the diversity of S. cerevisiae in the vineyard and in wine fermentations (Vezinhet et al. 1993; Versavaud et al. 1995; Török et al. 1996). RFLP analysis of mitochondrial DNA has been used to discriminate S. cerevisiae populations from grapes and in wine fermentations (Pramateftaki et al. 2000; López et al. 2001 ; Granchi et al. 2003; Capello et al. 2004; Schuller et al. 2005). A method to study the genetic variation among wine yeast strains is Amplified Fragment Length Polymorphism (AFLP). The genome structure revealed by AFLP patterns has demonstrated the occurrence of hybrid genotypes within the Saccharomyces sensu stricto group (de Barros Lopes et al. 1999) and the Cr. neoformans complex (Boekhout et al. 2001). AFLP has been widely applied to fingerprint yeast strains with good sensitivity and confidence, but it generally requires pure DNA and significant sample preparation. Although the degree of discrimination is not as high as achieved with AFLP and kaiyotyping, simple methods such as micro- and minisatellite fingerprinting and Random Amplified Polmorphic DNA (RAPD) have been successfully applied in studying the genetic variation of yeasts from various genera with minimal sample preparation (Baleiras Couto et al 1996a, 1996b; Vasdinyei and Deák 2003; Pina et al. 2005). However, reproducibility of these methods can be problematic and it requires attention to the purity and concentration of template DNA and the PCR annealing temperature. Primers targeting intron splice sequences (de Baros Lopes et al. 1996), sequence-tagged or microsatellite markers (Howell et al. 2004) and transposons (Fernández-Espinar et al. 2001 ; Capello et al. 2004; Demuyter et al. 2004) have provided consistent results for differentiating strains of S. cerevisiae, but are not yet applicable to other species for which there is no available sequence information.

2.7.3 Molecular strategies for monitoring yeast communities The standard approach to study the yeasts associated with wine grapes has been to rinse the surface of the grapes or macerate the grapes, then culture for yeasts in the rinses or macerates by plating onto agar media. In some cases, grape samples have been incubated in liquid enrichment media prior to plating onto agar medium. Common media used for the isolation of yeasts from fruit surfaces include Malt Extract Agar (MEA), Sabouraud medium, Wallerstein Nutrient Agar (WLNA) and Yeast-Malt (YM) medium. The use of two or more isolation media is often recommended because no single medium reliably recovers all species of yeasts (Fleet 1993).

Differences between total cell and culturable cell counts have been observed in microbial ecological studies of many natural environments. The ratio of culturable organisms is reportedly less than 1% of the total cell population. This phenomenon, known as the great plate anomaly, is thought to occur because many organisms may not be readily cultivated using conventional approaches (Head et al. 1998; Fleet 1999). Cells that remain viable but not culturable are believed to do so in response to physiological stresses, such as starvation, exposure to temperatures outside of growth range, radiation, osmolarity, aeration/oxidation or the inhibitory effect of toxic compounds (Oliver 1993). Yeasts associated with the surface of grapes, for example, would encounter many of these stresses. In addition to physiological status, cultivation difficulties may also arise because of unknown growth requirements or interactions with other members of the community. Regardless of the cause of the low culturability and viable but non culturable states of cells, current knowledge of microbial communities of many ecosystems obtained by culture-based methods alone is likely to be over- simplified and may represent a significant underestimation of the true diversity.

Culture-independent molecular techniques such as denaturing gradient gel electrophoresis (DGGE), temperature gradient gel electrophoresis (TGGE), terminal restriction fragment length polymorphism (T-RFLP) and single stranded conformation polymorphism (SSCP) have been developed to overcome the limitations of culture- based methods. These methods generate fingerprint profiles that indicate the structural diversity of the microbial community.

The PCR-DGGE technique has been increasingly used to explore the microbial communities of microorganisms in complex environments (e.g. soil, plant and marine ecosystems) including food ecosystems (Muyzer and Smalla 1998; Ercolini 2004; Giraffa 2004). The basic strategy for this approach is outlined in Figure 2.5. Total DNA is first extracted from samples of the product. Using primers targeting conserved regions, yeast ribosomal DNA within the extract is specifically amplified with PGR. Generally, the D1/D2 domain of the 26S rDNA is targeted, but other regions such as the 18S rDNA and ITS regions have been used. Amplicons of up to 500 bases, with one or more nucleotide differences are separated from each other by DGGE. The basis of the technique is separation of DNA fragments of similar length but different sequence composition, according to their melting behaviour in a polyacrylamide gel containing a linear gradient of DNA dénaturants (Fischer and Lerman 1983; Myers et ai 1987). Usually, each DNA band found in the gel corresponds to a yeast species. The band is excised from the gel, reamplified and sequenced to give the species identity. Thus, a profile of the species associated with the ecosystem is obtained, without the need for culture.

It is believed that this method overcomes the biases and limitations of culture methods, and reveals species that might occur as unculturable forms. Consequently, a more accurate representation of the ecology may be obtained. PCR-DGGE has been used to investigate the succession of yeasts in model wine fermentations (Cocolin et al. 2000), to examine the diversity and persistence of yeasts in botrytis-affected wine fermentations (Mills et al. 2002; Divol and Lonvaud-Funel 2005). In several cases, it could be concluded that PCR-DGGE revealed novel insights about the community structure and diversity of food ecosystems. Hence, it can be anticipated that its application to the study of wine grapes offers great potential to increase the understanding of this ecosystem.

Food / Beverage 1 r

Extract and purify DNA/RNA i PCR-amplification of yeast ribosomal DNA i DGGE/TGGE separation of DNA amplicons y

Isolate and sequence bands to give species identity

Figure 2.5 Culture-independent approach for profiling and identifying yeasts in foods and beverages using PCR-DGGE analyses 2.7.4 Factors affecting the performance of molecular methods for the analysis of yeasts

Despite the potentially new and valuable information that can be gained from molecular methods in microbial ecological studies, there are inherent limitations that need to be recognised.

2.7.4.1. DNA extraction Recovery of total, representative DNA is one of the first hurdles in applying molecular methods to study microorganisms in their natural environments. Methods to extract DNA from environmental samples include physical (bead-beating, sonication, boiling, freeze-thawing), chemical lysis (SDS, phenol, various detergents) and enzymatic treatments (lysozyme, proteinase K, pronase) (Kresk and Wellington 1999). Usually, a combination of mechanical and chemical disruption is used to achieve high cell lysis efficiency. Also, the purity, quality and yield of extracted DNA will impact on the reliability of the subsequent PGR assay and, therefore, also needs to be considered.

Environmental samples may contain a number of substances (humic acid, plant polysaccharides, proteins, polyphenols) which inhibit polymerase activity (Wilson 1997; Marshall et al 2003). These compounds co-purify with the target DNA, so that suitable procedures for their removal during the extraction protocol require evaluation.

Microbial cells within their habitat can exist in several physiological forms, including mycelia, spores, vegetative cells and injured cells. Dead and moribound cells may also be found (Kresk and Wellington 1999). More vigorous conditions are required for lysis of filamentous fungal spores and Gram positive bacteria, but such conditions may lead to fragmentation of other nucleic acids, such as mycelia and Gram negative bacterial cells, respectively. Fragmented nucleic acids are sources of artifacts in PGR amplification and may contribute to the formation of chimeric PGR products (Liesack et al 1991). DNA extraction efficiency may be affected by the physiological age of cells (exponential, stationary phase, autolysing). Depending on the culture age, various cell proteins may interact with genomic DNA and affect either the activity of polymerase or primer annealing to the template during PCR amplification (de Barros Lopes et al. 1996).

2.7.4.2 PCR Differential or preferential amplification of ribosomal genes and formation of PCR artefacts are recognised as important sources of errors and biases in assays employing PCR. The pitfalls of PCR-based strategies in ecological studies are discussed in Suzuki and Giovannoni (1996), Wang and Wang (1997), Wintzingerode et al. (1997), Fogel et al. (1999), Speksnijder et al. (2001) and Kurata et al. (2004).

PCR errors and biases influence the detection sensitivity and may skew conclusions about the relative abundance and diversity of the genotypes present in a sample. Often, species contributing less than 1% of the total population are not readily detected by the PCR-DGGE technique (Head et al. 1998; Ogier et al. 2002). There are other cases however, where subdominant or minority genotypes are more favourably represented (Ferris and Ward 1997). Selective amplification of templates from mixed populations may be caused by variation in genome size and copy number, differential GC content in templates, accessibility of templates to primer binding, differential template concentrations (Farrelly et al. 1995; Suzuki and Giovannoni 1996; Polz and Cavanaugh 1998). The formation of chimeric and heteroduplex molecules and the mis- incorporation of bases by Taq polymerase are possible errors that occur during PCR amplification (Speksnijder et al. 2001; Qui et al. 2001).

2.7.4.3 DGGE A limitation of DGGE analysis is that only relatively small PCR amplicon fragments (500 bases or fewer) can be separated. This poses a restriction on the limited number of genotypes that can be resolved on the denaturing gels, as well as the amount of sequence information that can be analysed (Muyzer and Smalla 1998).

Ideally, one species yields one band in the DGGE gel. However, the number of bands generated by DGGE gels may not accurately reflect the unique species present in the sample. DNA fragments from different species may share the same melting behaviour; co-migrating bands in the DGGE gel then leads to an underestimation of the total diversity.

There are cases where two or more bands have been detected for one species and may lead to an overestimation of the species diversity. Multiple banding patterns may be due to heteroduplex bands that are formed as templates with high sequence similarity anneal together (Kowalchuk et al. 1998), or could arise from rDNA operon heterogeneity (Crosby et al. 2003).

2.8 SUMMARY

Yeasts are intimately involved in the winemaking process. They have a profound effect on the quality and efficiency of wine production. Yeasts are naturally associated with the surface of grapes and, consequently, grapes are a primary source of yeasts in wine production. The surface of grapes is a phyllospheric habitat and the yeast, bacterial and fungal flora associated with this ecosystem are likely to evolve and change as the grape berry matures on the vine and in response to environmental conditions and influences of viticultural practices. Previous studies examining yeasts associated with wine grapes have largely relied on the use of standard cultural procedures. These procedures have well known inadequacies and limitations that may underestimate the true microbial ecology in natural habitats, such as wine grapes. The application of molecular, culture- independent methods could reveal entirely new knowledge about the yeast-grape ecosystem, and yeasts associated with wine fermentations. CHAPTER THREE

YEAST ECOLOGY OF WINE GRAPES AT SEVERAL VINEYARDS IN AUSTRALIA

3.1 INTRODUCTION Grapes are a primary source of yeasts associated with wine fermentations. The biodiversity of the yeast ecology has varying impacts on the efficiency and quality of wine production by impacting on the alcoholic fermentation, malolactic fermentation and wine spoilage. Species of yeasts associated with grapes may also affect grape quality prior to harvest, and may be important as natural agents for controlling spoilage and mycotoxigenic grape fungi (Fleet 2001, 2003).

Although the yeast species associated with wines have been studied over many years, knowledge about their ecology and relationship with the grape surface as a natural habitat remains uncertain, and, in some cases, controversial (Martini et al. 1996; Mortimer and Polsinelli 1999). It is not known how yeasts evolve on the grape surface during their development on the vine, or how vineyard geography, climate and viticultural practices affect their ecology. Such information would be important in optimising the contribution of the indigenous yeast flora to wine fermentations and in managing preharvest grape quality through biocontrol programs. Grapes could also be an important source of new strains of yeasts for the development of starter cultures for wine production.

The yeast species associated with wine grapes have been previously examined using standard cultural procedures. Applications of culture-independent, molecular techniques offer alternative approaches to characterize the microbial communities of natural habitats (Muyzer and Smalla 1998). One such method is PCR-denaturing gradient gel electrophoresis (DGGE). PCR-DGGE has been shown to reveal novel and minor components (1% total population) in various habitats and facilitates detection of viable but non culturable organisms in complex microbial populations. Consequently it provides a means of overcoming the biases and limitations known of the plating method (Head et al. 1998). This technique has also proved to be useful in profiling the yeast and bacterial flora in food and beverage fermentations (Ercolini 2004; Giraffa 2004).

This Chapter reports the yeast species and populations associated with wine grapes during their cultivation at several Australian vineyards. In conjunction with a parallel study (Prakitchaiwattana 2005), it represents the first major ecological study of yeasts on Australian wine grapes. Influences of berry damage, vineyard region, seasonal variation and pesticide applications on this ecology are also reported. Cultural isolation and PCR-DGGE are used to examine the yeast flora colonizing the surface of wine grapes.

3.2 MATERIALS AND METHODS

3.2.1 Grape samples Samples of wine grapes were collected from seven commercial vineyards located in the upper and lower Hunter Valley, Mudgee and Griffith regions of New South Wales, Australia (Figure 3.1). The Northern region (upper and lower Hunter Valley), the northwestern region (Mudgee), and the southwestern region (Griffith) represent distinct geographical and climatic zones (i.e. warm and humid in the Hunter Valley; cooler climate of Mudgee; and the hotter, drier climate of Griffith; www.awbc.com.au 2006). Two varieties of white grapes (Sauvignon blanc, Semillon) and three varieties of red grapes (Cabernet sauvignon, Merlot and Tyrian) were collected at five stages of grape maturity during the seasons of October 2001-Febuary 2002 and October 2002-Febuary 2003. These samples consisted of healthy, physically intact, undamaged grape berries. In addition, grapes that were physically damaged (broken skin, shrivelled) were also sampled for each variety, but only at the time of commercial harvest. Table 3.1 shows the vineyards from which particular grape varieties were obtained (see also Figures 3.2 and 3.3). The maturity stages and codes for the grape samples analysed are given in Table 3.2. New South Wales Upper Hunter Valley (276km) • Mudgee * • Lower Hunter Valley (165 km)^ (260 km) O Griffith Sydney (595 km)

Kilometres (X100) 0 12 3 1 I I I distance from Sydney

Figure 3.1 Location of vineyard regions in New South Wales (NSW), Australia

Table 3.1 Sources of wine grapes used for the analysis of yeasts

Code Vineyard Vineyard location (NSW) Grape variety

R Rosemount Upper Hunter Valley Cabernet sauvignon, Sauvignon blanc, Semillon

F Femandes Upper Hunter Valley Shiraz, Semillon

L Lindemans Lower Hunter Valley Semillon

MM Molly Morgan Lower Hunter Valley Shiraz, Semillon

HG Hill of Gold Mudgee Cabemet Sauvignon, Merlot, Shiraz, Sauvignon blanc

Cm Combandry Mudgee Cabemet sauvignon, Merlot, Shiraz

MW McWilliams Griffith Cabemet Sauvignon, Tyrian, Sauvignon blanc, Semillon 10-12 weeks before harvest 8-10 weeks before harvest (veraison)

IV

4-8 weeks before harvest 1- 2 weeks before harvest

harvest harvest (damaged grapes)

Figure 3.2 Maturity stages of Tynan grapes examined for yeasts (also see Table 3.2). 10-12 weeks before harvest 8-10 weeks before harvest (veraison)

m IV

4-8 weeks before harvest 1- 2 weeks before harvest

harvest harvest (damaged grapes)

Figure 3.3 Maturity stages of Semillon grapes examined for yeasts (also see Table 3.2). Table 3.2 Stages of grape maturity when grapes were sampled for yeast analysis

Weeks before Maturity stage Description of grapes^ harvest code 10-12 I Berries hard and green; about 2-5 mm in diameter, accompanied by closure of bunches

8-10 II Berries, 5-10 mm in diameter (pea-sized), (Veraison) begin to change colour and soften

4-8 III Berries, 8-10 mm in diameter,

commencement of ripening

2-4 Illb Berries ripening

1-2 IV Berries, 10-15 mm in diameter

0 V anBerried progress fullys tripeo senescenc, 10-15 mem afte in rdiamete this stagr e a, Pratt 1971; Combe 1992, 1995; Fleet et al. 2002

Each sample consisted of small clusters or bunches of grapes (200 g) that were aseptically removed from at least five different vines in a vineyard plot and combined to give a composite of about 1 kg. Duplicate samples were taken for each variety. Samples were transported to the UNSW laboratory by road or air and stored at 4°C. Grapes were analysed within 24 h of collection from the vine.

3.2.2 Analysis of yeasts on grapes Yeast on grapes were determined by (i) culture plating on agar media, (ii) enrichment culture followed by plating on agar media, and (iii) PCR-DGGE analysis, as outlined in Figure 3.4. Individual grape berries were randomly and aseptically removed from each cluster to give a composite samples of 50 g. Grapes were suspended in 450 ml of sterile 0.1% Bacteriological peptone (Oxoid, Melbourne, Australia), containing 0.01% Tween 80 (Sigma, St. Louis, MO) and shaken at 180 rpm for 30 min at 25°C. The grape rinse was poured from the grapes and examined for yeasts by both cultural and PCR-DGGE methods. Grape berries ^

\ Enrichment culture analysis Diluent 450 ml

Shaking A 30 min 10 g berrie+ s Grape juice (0.1% Bacteriological peptone (90 ml) + Tween 80) \

Supernatant Incubation 25°C / -10 days ( Agar culture analysis

Figure 3.4 Outline of protocol for the analysis of yeasts on wine grapes In some cases, grape samples, after the rinsing process, were placed in sterile Stomacher bags (Seward Lab Systems, London) and macerated for 1 min in a Stomacher (Seward Laboratories, London). Grape macerates were examined for yeasts by plate culture.

In some experiments, grapes were first surface disinfested, following the procedure described by Dugan et al. (2002). Clusters of grapes (20 g) were immersed in 70% ethanol for 30 s, rinsed with water, then immersed in a solution of 0.5% NaOCl (Ajax Chemicals, Sydney Australia) for 5 min with stirring. Grapes were rinsed with sterile distilled water, then homogenized for 2 min in a Stomacher (Seward Laboratories, London) before plating.

3.2.3 Isolation and enumeration on agar media Rinses and macerates from grape samples were serially diluted in 0.1% Bacteriological peptone, from which 0.1 ml aliquots were spread inoculated in duplicate over plates of Malt Extract Agar (MEA, Oxoid) and Wallerstein Laboratories Nutrient Agar (WLNA, Oxoid). Both media were supplemented with filter sterilised oxytetracycline (100 mg/1; Sigma) and biphenyl (100 mg/1; Ajax Chemicals, Australia) to control the growth of bacteria and restrict radial growth of filamentous fungi, respectively. Yeasts and moulds were enumerated after aerobic incubation at 25°C for 4-7 days. Representatives of different colony types were selected, and purified by streaking on MEA to obtain pure cultures. Purified yeast cultures were maintained by regular subculture on MEA and storage with 30% glycerol at -80°C.

3.2.4 Isolation by enrichment culture Grapes (10 g) were aseptically transferred to flasks containing 90 ml of sterile, commercial grape juice (Berri Pty. Ltd., Australia) for enrichment. In some cases, grape samples (200 g) were placed in sterile Stomacher bags (Seward Lab System) and homogenized for 2 min. The resulting homogenate, containing juice and skins (about 100 ml) was transferred into sterile screw-capped containers (120 ml size; Sarstedt, Australia). These cultures were termed autoenrichments. Enrichment and autoenrichment cultures were incubated at 25°C for 10 days or until visibly turbid. Samples were diluted as required in 0.1% Bacteriological peptone solution, and 0.1 ml aliquots were spread inoculated over the surfaces of MEA and WLNA, supplemented with oxytetracycline (100 mg/ml) and biphenyl (100 mg/ml). The plates were Incubated at 25°C for 4-7 days, after which dominant and representative colonies were isolated as pure cultures for identification.

3.2.5 DNA extraction for PGR amplification DNA was extracted from (i) pure cultures of yeasts and (ii) microbial cell pellets from rinses of grapes. Genomic DNA was extracted according to the method described by Cocolin et al. (2000). Yeast cell pellets were prepared from 24 h cultures of isolates in Malt Extract Broth (1 ml, centrifuged at 12,000 g, 10 min, 4°C), and resuspended in 200 ^il of breaking buffer (1% SDS, 2% Triton X-100, 100 mM NaCl, 10 mM Tris, 1 mM EDTA, pH 8.0) containing 0.3 g glass beads (0.5 mm in diameter, BioSpec Products, Inc., Bartlesville, OK). Cells were homogenized at 6,000 rpm for 1 min in a bead beater (Mini-BeadBeater-F'^, BioSpec) in the presence of 200 [il of phenol/chloroform/iso- amylalcohol (25:24:1). Two hundred microliters of TE buffer (10m M Tris, 1 mM EDTA, pH 8.0) were added, and the mixture was centrifuged at 12,000 g for 10 min at 4°C. The aqueous phase was collected and DNA was precipitated by the addition of 2.5 volumes of absolute ethanol and recovered by centrifugation at 12,000 g for 10 min at 4°C. The DNA pellet was washed with 70% ethanol, dried and resuspended in 50 ^il of TE buffer. DNA extracted from pure cultures was used in PGR reactions for the amplification of partial 26S and ITS regions of the rDNA gene.

For DNA extraction from yeasts associated with grapes, approx. 450 ml of grape rinses were sedimented by centrifugation at 16,000 g for 15 min at 4°G. DNA was isolated from the sedimented microbial pellets using the same procedure as described already, with the following modification. After bead beating, cell lysate was applied to a DNeasy Plant kit (Qiagen, Glifton Hill, Australia), according to the manufacturer's instructions, and DNA was eluted with 150 of AE buffer supplied with the kit. This DNA preparation was used for PGR-DGGE analysis. 3.2.6 PGR amplification of DNA for sequence identification and PCR-DGGE analysis

Genomic DNA from yeast cultures was amplified by PGR with universal primers NLl (5'-GCATATCAATAAGCGGAGGAAAAG-3') and NL4 (5'- GGTCCGTGTTTCAAGACGG-3') for identification by partial 26S rDNA sequence analysis (Kurtzman and Robnett 1998). PGR reactions contained Ix PGR Buffer (10 mM Tris-Gl, 50 mM KGl), 0.2 ^M of each primer, 200 ^M of each dNTP (Roche Diagnostics, Indianapolis, IN), 1.5 mM MgGh, 1.25 U Gold Taq DNA polymerase (AmpliTaq'^^*, Roche Molecular Systems, Brachburg, NJ), and 10 ng of DNA template in 50 |il final volumes. Amplification was carried out in a 9600 Thermal Gycler (Applied Biosystems), with the following program: initial denaturation at 95°G for 7 min, 36 cycles at 95°G for 1 min (denaturation), 52°G for 2 min (annealing) and 72°G for 2 min (extension), followed by a final extension for 10 min at 72°G.

A nested PGR strategy was applied to examine yeasts by DGGE (Prakitchaiwattana et al. 2004). Purified DNA extracted from rinses of grapes was first amplified with primers NLl and NL4 described previously. Bunches of grapes collected during the 2002-2003 season were often covered with dust and soil. The T4 gene 32 protein (0.25 |xg; Roche Diagnostics, Germany) was included in the 50 |il volume reaction mixture to improve PGR amplification efficiency in the presence of humic acid (Garol 1996). Products of the first PGR reaction (approximately 600 bp) were diluted 1:100, and used as a template for the second PGR using forward primer NLl with GG clamp, (5' - GGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGATATGAATAAGG-3'), and the reverse primer LS2 (5'-ATTGGGAAAGAACTGGAGTG-3') (Gocolin et ai 2000). PGR reaction mixtures were the same as described already, without the addition of T4 gene 32 protein. Gycling conditions for the second PGR were as follows: initial denaturation at 95°G for 7 min, 30 cycles at 95°G for 1 min (denaturation), 52°G for 2 min (annealing) and 72°G for 2 min (extension), with a final extension for 10 min at 72°G. PGR amplicons were confirmed by 1.5% agarose gel electrophoresis. All primers used in this study were obtained from Proligo Primers & Probes (Sydney, Australia). 3.2.7 Denaturing gradient gel electrophoresis (DGGE) The DCode Universal Mutation Detection System (Bio-Rad, Hercules, CA) was used for sequence-specific separation of GC clamped amplicons. DGGE was performed using 8% polyaciylamide gels (acrylamide:N,N'-methylene bisacrylamide ratio, 19:1, Bio-Rad), with a denaturing gradient from 30-60% of urea and formamide (100% denaturant was defined as 7 M urea plus 40% formamide). The gels were run in Ix TAE buffer (40 mM Tris-acetate, 1 mM EDTA), at a constant temperature of 60°C, initially at 20V for 30 min, then at 120V for 6 h. After the run, gels were stained with SYBR® Gold (Molecular Probes, Eugene, Oregon) in Ix TAE buffer, pH 8.0, and photographed under UV transillumination with a Polaroid DS-34 camera.

Reference cultures of wine yeasts and characterised isolates were observed for their banding profile to determine the limits of separation/resolution of the method (Figure 3.5). Several ascomycetous and basidiomycetous yeast species could not be distinguished from each other on the basis of mobility, for example Saccharomyces bayanus from S. cerevisiae, Hanseniaspora itvarum from Hs. guilliermondii and Hs. opuntiae, Candida stellata from C. zemplinina, and Rhodotorula mucilaginosa from Rh. glutinis. Several migration fragments were observed iox Aureobasidium pullulans and related isolates. PGR amplicons of Metschnikowiapulcherrima appeared as a characteristic DGGE doublet. The reproducibility of DGGE gel gradients and electrophoretic conditions were tested by subjecting DNA bands of reference species to a second round of PCR-DGGE analysis and by comparing the mobility of bands to the original analysis.

Prominent DNA bands in DGGE gels were selected and recovered for sequence analysis. Bands were excised with a sterile razor blade, and gel pieces were incubated in 50 \i\ of TE buffer overnight at 4°C. Eluted DNA was reamplified with primers NL1 (without GC clamp) and LS2, using the same conditions as the second PGR step described above. Amplified products were used as templates for subsequent sequencing reactions. 8 9 10 11 12 13 14 15 16 17

Figure 3.5A DGGE of the partial 26S rDNA amplicons of ascomycetous wine yeast species and grape isolates. Lanes 1-7; 1, Pichia anomala AWRI 1023; 2, Saccharomyces cerexnsiae AWRI 1010; 3, Zygosacchawmyces bailii UNSW 507700; 4, Tonilaspora delbrueckii CBS 1146; 5, Saccharomyces bayanus AWRI 380; 6, Dekkera bruxellensis PN^Kl 1102; 7, Metschnikowia pulcherrima UNSW 514400; S, Aureobasidium pullnlans FRR 4800; 9A2; Aureobasidium-hke isolates 73, 77, 22, 550; 13, Candida stellata A'^Kl 1159; 14, Candida zemplinina isolate E120; 15, Hanseniaspora uvaum isolate E55; 16, Hanseniaspora opuntiae isolate E90; 16, Hanseniaspora gidlliennoindii isolate E103.

B

Figure 3.5B DGGE of partial 26S rDNA amplicons of basidiomycetous yeast isolates from grapes. Lanes 1-7; 1, Cryptococciis saitoi 84; 2, Ciyptococcus laurentii 41; 3, Ciyptococcus oeirensis 87; 4, Rhodotorula mucilaginosa 395; 5, Rhodotorula ghitinis 102; 6, Sporobolomyces roseus 558; 7, Rhodotonda nothofagi E45; 8, Rhodotorula mucilaginosa 335 3.2.8 DNA Sequencing Sequence analysis was performed on PCR amplicons from yeast isolates and from eluates of DGGE gels, using the ABI PRISM® Big Dye™ Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA), according to the manufacturer's directions. DNA fragments were sequenced at the Automated DNA Analysis Facility, School of Biotechnology and Biomolecular Sciences, UNSW. Nucleotide bases were compared with available 26S rDNA and ITS sequences in Genbank using the NCBI BLAST program (Altschul et al. 1990).

3.2.9 Species identification Yeast isolates were differentiated based on types of colony and cellular morphology. Three major groups of yeasts were found; these \v\c\udQ(X Aureobasidium-WkQ organisms (which will be discussed in more detail in Chapter 4), species of Hanseniaspora and species of Metschnikowia. Each group was screened by PCR-RFLP based on the ITS regions. Vox Aureobasidium-Wke isolates, the restriction enzymes Cfo\ and Hae\\\ were used; Hanseniaspora isolates were screened using Dral and Dde\ in RFLP analyses and for growth at 37°C. Isolates belonging to Metschnikowia were grouped using Dral in PCR-RFLP assays.

Amplifications of the ITS (ITS I and ITS2) and 5.8S rDNA regions were performed with primers ITSl (5'- TCCGTAGGTGAACCTGCGG -3') and ITS4 (5'- TCCTCCGCTTATTGATATGC -3') described by White et al. (1990). The PCR conditions were the same as described for amplifying the D1/D2 region. The following cycling conditions were used: initial denaturation at 95°C for 7 min, 36 cycles of 95°C for 1 min (denaturation), 55°C for 2 min (annealing), 72°C for 2 min (extension), followed by a final extension for 10 min at 72 °C. Aliquots (10-15 ^il) of the PCR amplicons were digested without further purification with 5U of restriction enzyme in 15-20 }il volumes, according to the manufacturer's instructions (Promega, Madison, WI). Restriction fragments were analysed by electrophoresis in 3% (w/v) agarose gels. A 100 bp DNA ladder (Bio-Rad, Hercules, CA) was used as a size standard. Electrophoresis was carried out in 15 x 25 cm gels on a Sub-Cell GT Agarose Gel Electrophoresis system (Bio-Rad, Hercules, CA), at lOOV for 2 h. After electrophoresis, gels were stained with ethidium bromide and photographed under transilluminated UV light. Restriction banding profiles were compared to the restriction databases reported by Guillamon et al. (1998), Esteve-Zarzoso et al (1999), Granchi et al (1999) and Sabate et al (2002).

Representative isolates, grouped by morphology and by ITS-RFLP, were identified to the species level by sequence analysis of the partial 26S rDNA and in some cases, by analysis of ITS sequences. Accession numbers retrieved in BLAST matches for the different species analysed are shown in Appendix 1.

3.3 RESULTS

3.3.1 Climatological data for the vineyard regions in the years 2001-2003

Climate and rainfall data were obtained from the archives of the Australian Bureau of Meteorology, New South Wales Regional Office. The mean daily maximum temperatures and mean monthly rainfall of vineyard regions for the overall period of 2002-2004 are shown in Figure 3.6. Abnormally/uncharacteristically dry conditions were experienced in NSW during the season 2002-2003, where much of the state (82- 99%) was officially drought-declared. Rainfall data for the periods 2001-2002 and 2002-2003 are indicated separately (Figure 3.6). Lower rainfalls were generally recorded for vineyards in the Upper Hunter, Lower Hunter and Mudgee regions in the 2002-2003 season. In most cases, temperatures during the 2002-2003 drought-affected season were typically higher than average, especially prior to the start of veraison. Temperatures in the vineyards often reached 45°C during the sampling periods in November and December. The Griffith region however, represented a characteristically drier and warmer area, consistently receiving lower average rainfalls and experiencing higher average temperatures. Rainfall Temperature

Aug Sep Oct Nov Dec Jan Feb Aug Sep Oct Nov Dec Jan Feb Month Month

Figure 3.6 Rainfall (mm) and mean daily maximum temperature (°C) of vineyard regions during the period of grape development (a) Upper Hunter Valley, 2001-2002 and 2002- 2003; (b) Lower Hunter Valley, 2001-2002 and 2002-2003; (c) Mudgee, 2001-2002 and 2002-2003; (d) Griffith, 2002-2003. Arrows indicate time when wine grapes were sampled , ( [1 ) Mean rainfall (mm) during 2000-2004 ( • ), Mean daily maximum temperature during 2000-2004; ( O A | ), 2001-2002; ( • A j ), 2002-2003. 3.3.2 Evaluation of plating medium, sample processing and enrichment methods for the isolation and enumeration of yeasts from wine grapes

3.3.2.1 Comparison of MEA and WLNA for the enumeration of yeasts from grapes

Enumeration of yeast species and populations from grape rinses was performed on MEA and WLNA plates. No significant differences in total yeast populations, number of species detected and individual species composition were detected between media, as determined by a one-way analysis of variance at the probability level of 0.05 (Table 3.3). However, to aid the differentiation and enumeration of colony types, both media were used in all further experiments. Counts reported in subsequent Tables and Figures in this Chapter are averages from both media.

Table 3.3 Comparison of yeast populations on grapes as determined by plate culture on MEA and WLNA

Media (n=75) Yeast group MEA WLNA Total yeast population ^ 4.71'' 4.76' Total number of yeast species 2.49^ 2.42^ Population of yeast groups t Aureobasidium pullulans 5.06^ 5.18^ Hanseniaspora spp. 6.02'' 6.05^ Metschnikowia spp. 6.02^ 6.02^ Candida spp. 5.06^ 4.91^ Cryptococcus spp. 3.83^ 3.71^ Rhodotorula spp. 3.34^ Mean colony counts (log CFU/g) Population counts in the same row with the same superscript are not significantly different (a=0.05) n = number of grape samples examined

3.3.2.2 The effect of rinsing and grape maceration on the recover/ of yeast species and populations on wine grapes

To examine the efficiency of the rinsing process to release yeast cells from surfaces of the grapes, samples were processed first by rinsing, followed by maceration of the rinsed grapes. Table 3.4 shows that yeast cells were not completely removed by the rinsing procedure alone. About 60% of the total yeast counts was recovered by rinsing alone. Generally, the yeast species detected from rinses were also found from macerates. In several cases (18%, 14 of 75 samples), maceration of rinsed grapes revealed the presence of one to two additional species, which were not recovered by the rinsing method. Samples from which additional species were recovered by maceration were usually those of damaged grapes.

Table 3.4 Recovery of yeasts from wine grapes by rinsing and maceration Sample treatment (n=75) Rinse Macerate Total*' Yeast population^ 4.73 (58%) 4.59 (42%) 4.97(100%) Number of yeast species 2.45 (90%) 2.23 Additional species 0 0.27(10%) 2.72(100%) Mean colony counts (log CFU/g) rinse plus macerate n = number of grape samples examined

Samples of mature, healthy grapes were surface disinfested and the resulting macerate plated, to determine the contribution of the internal yeast flora to counts recovered from the maceration procedure. The experiment should also confirm whether yeasts recovered by maceration of grapes were due to the inefficiency of the rinsing method. Filamentous fungi were recovered from all samples of surface-sterilised grapes examined. Yeast-like fungi (black yeasts) were isolated from three samples, and yeasts (pink yeasts) were detected once (Table 3.5).

Table 3.5 Populations of yeasts from surface-sterilised grapes Fungal group* Sample Yeasts Yeast-like fungi Filamentous fungi 1 2.65 2.30 2.00 2 nd 2.00 2.91 3 nd nd 2.97 4 nd nd 4.08 5 nd nd 4.17 6 nd 2.17 1.85 ^ Mean colony counts (log CFU/g) nd, not detected 3.3.2.3 Comparison of enrichment in grape juice and autoenrichment for the isolation of yeasts from grapes

The occurrence of the principal wine yeast, Saccharomyces cerevisiae, in natural environments, including wine grapes/vineyards, has been widely debated. Two enrichment cultures were initially performed to investigate whether recovery of this species could be favoured by different conditions. Table 3.6 shows the proportion of yeast species found for enrichment in grape juice (enrichment culture) and in grape macerates (skin and juice; autoenrichment culture).

Table 3.6 Frequency of isolation of yeasts from wine grapes by enrichment and autoenrichment cultures

Enrichment method (n^36) Yeast species Enrichment Autoenrichment Aureobasidium pullulans 4 Hanseniaspora spp. 30 28 Metschnikowia spp. 19 21 Cryptococcus spp. 0 1 Rhodotorula spp. 7 3 Candida spp. 3 5 Pichia spp. 4 0 Issatchenkia orientalis 3 3 Kluyveromyces thermotolerans 5 6 Tonilaspora delbrueckii 1 3 Saccharomyces cerevisiae 0 2 Zygosaccharomyces bailii 0 1 Zygoascus hellenicus 0 2 number of times isolated n = number of grape samples examined

The species belonging to Hanseniaspora and Metschnikowia were frequently isolated from grapes, and similar isolation frequencies for these groups were observed for both enrichments. There were however, examples where species were isolated by one method and not the other. These were usually of yeast species that were infrequently encountered from grapes. Of the 36 samples of grapes examined, S. cerevisiae was isolated twice only (5.5%), and in both cases, by autoenrichment. Similarly, Zygosaccharomyces bailii and Zygoascus hellenicus were rarely detected from grapes, but both species were only isolated from autoenrichments. Species belonging to Pichia were found from enrichment cultures but not autoenrichment. Aureobasidhim-WkQ organisms, including^, pullulans were more frequently detected from enrichment cultures.

3.3.3 Population of yeasts on wine grapes

Yeast populations during grape nnaturation The population and diversity of yeast species occurring on wine grapes were examined from five different grape varieties at five phenological stages throughout cultivation from four commercial vineyards for the 2002-2003 season. Grapes from two vineyards in the Upper Hunter Valley were also examined throughout the 2001-2002 season.

Figure 3.7 shows the development of yeast populations on healthy grapes of several varieties during cultivation. Populations are the mean counts of yeasts enumerated from duplicate grape samples plated on MEA and WLNA in duplicate. The populations ranged from 10^-10^ CFU/g. Lower counts of 10^-10^ CFU/g were generally found during berry development (stages I to IV). As berries approached harvest ripeness (stage V), a 10-100 fold increase in yeast counts was often observed.

Hunter Vallev, 2001-2002 and 2002-2002 seasons Grapes generally showed higher yeast populations during the cooler, wetter season of 2001-2002 in the Hunter Valley (Figure 3.7a,b). Counts obtained from the Lower Hunter Valley (vineyard L; Figure 3.7c,d) tended to be lower, compared with those from Upper Hunter Valley (vineyard R), despite similar patterns for rainfall and temperature. There were no obvious influences of red or white grape varieties on yeast populations.

Mudgee. 2002-2003 Similar to grapes in the Hunter Valley, grapes from vineyard HG in the Mudgee region also showed an increase in yeast counts towards beny maturity. Lower populations were detected on all three grape varieties at veraison, after which increases occurred to 10^-10^ CFU/g (Figure 3.7e). Griffith, 2002-2003 Grapes from the MW vineyard gave very low populations of yeasts throughout 2002- 2003. Generally, populations were less than 10^ CFU/g, until the stage of harvest (Figure 3.7f)- However, no yeasts (<50 CFU/g) were detected on Semillon grapes two weeks prior to harvest and they were not detected on Sauvignon blanc grapes at harvest. At harvest ripeness, the two red grape varieties gave yeast populations of 10"*-10^ CFU/g.

Yeast populations on undannaged and damaged grapes Figure 3.8 compares the yeast populations on healthy, undamaged berries with those on damaged berries, sampled from the same vineyard at the same time. Compared with undamaged berries, damaged berries harboured 10-1000 fold higher populations of yeast. For the 25 samples examined, 22 samples gave significant differences in yeast populations (Student's t-test, a=0.05). (a) Upper Hunter Valley 2001-2002 (b) Upper Hunter Valley 2002-2003

I II III IV V I II III IV V Phenological stage of berry development Phenological stage of berry development

(c) Lower Hunter Valley 2001-2002 (d) Lower Hunter Valley 2002-2003

6

5

CT 4 u P 2

eI II III IV V I II III IV V Phenological stage of berr/ development Phenological stage of berry development

(e) Mudgee 2002-2003 (f) Griffith 2002-2003 1

T J :•:• iz H 1 if — i\ I II III IV V I II III IV V Phenological stage of berry development Phenological stage of berry development

Figure 3.7 Total populations of yeasts on grapes during cultivation at vineyards in different regions of New South Wales (a) vineyard R, 2001-2002, (b) vineyard R, 2002-2003, (c) vineyard L, 2001-2002, (d) vineyard L, 2002-2003, (e) vineyard HG, 2002-2003, (f) vineyard MW, 2002-2003. Phenological stages of berry development are given in Table 3.2. Cabernet sauvignon (cs,H); Meriot (m,[xl); Tynan (tyr,[[n]); Sauvignon blanc (sab,g); Semillon (sm,H);. (a) undamaged, Upper Hunter Valley (b) damaged, Upper Hunter Valley

2001- 2002 iiiiiiiiiinimmn iinii'rTTTTO sab HIIIIMIIIII 'iiiiimiiiTr en anni e 1

2002- 2003 c 012345678012345678

(c) undamaged, Lower Hunter Valley (d) damaged, Lower Hunter Valley

MM 2001- 2002

2002- 2003 C so 1

(e) undamaged, Mudgee (f) damaged, Mudgee S Cm

2001- 2002 llllll iiilllliliiliiiillll TTTTTI""' iiiiiiiiiiiiiiMiiiii llllll llllll lili III I liuti" HG m Qi m*

TTTTTTT nTFH-4is.b 2002- TTTTt Mill III IIIIII 11 II MM II llPftl— HG 3m» 2003

(g) undamaged, Griffith (h) damaged, Griffith

sab I sab 2002- urn MW 2003 ityr Ityr*

0 1 2 3 4 5 6 0 1 2 3 4 5 6 Log CFU/g Log CFU/g

Figure 3.8 Populations of yeasts on undamaged and damaged grapes at maturity stage V from different vineyards (R, L, F, MM, Cm, HG, MW), during the 2001-2002 and 2002-2003 seasons, (a, c, e, g) undamaged grapes; (b, d, f, h) damaged grapes. Cabernet sauvignon (cs,E); Meriot (m, E]); Tyrian (tyr^); Shiraz (sz,H); Semillon (sm, •); Sauvignon blanc (sab,[l). The student's Mest was conducted and asterisks (*) indicate significant differences between undamaged and damaged samples at a 95% level. 3.3.4 The diversity of yeast species associated with wine grapes during cultivation

Six different grape varieties from seven vineyards in NSW were routinely examined for yeasts throughout cultivation from October 2001 to February 2002 and from October 2002 to February 2003. Yeast populations and species were simultaneously examined by plate culture, PCR-DGGE and enrichment methods.

3.3.4.1 Hunter Vallev Region, 2001-2002 Aureobasidium pullulans was the most prevalent species on both Cabernet sauvignon and Semillon grapes at every maturity stage in vineyards R and L during the 2001-2002

0 "X season (Tables 3. 8 and 3.10). Its population ranged between 10-10 CFU/g at the early stages of grape maturity (stages I to II), where it was frequently the dominating species. The prevalence of A. pullulans was also demonstrated by PCR-DGGE analysis and by enrichment cultures. A diversity of colony morphotypes and DGGE banding patterns corresponding to A. pullulans were found; sequence analysis of isolates and DGGE bands gave between 91-100% homology to A. pullulans, which suggested the presence of additional or multiple species oiAureobasidium. Various species of Cryptococcus, Rhodotorula and Sporobolomyces were frequently, but not consistently, isolated from maturing berries. Generally, their populations were low, at lO'-lO^ CFU/g. These species were also infrequently isolated from enrichment cultures, but were undetected by PCR-DGGE. Enrichment cultures of developing berries showed the presence of the snow fungus, Tremella fuciformis, that was neither detected on plate cultures nor PCR- DGGE.

On mature grapes (stages IV and V), populations of A. pullulans increased to lO"^ CFU/g. Metschnikowia pulcherrima and Hanseniaspora uvarum were prevalent on Semillon grapes at the time of harvest, with populations between 10"*-10^ CFU/g. Although these species were not recovered from mature Cabernet sauvignon grapes by plate culture, their presence was detected by PCR-DGGE and enrichment. A diversity of species belonging to Metschnikowia and Hanseniaspora was isolated by enrichment of mature Cabernet sauvignon grapes. Sequences corresponding to unidentified species of Metschnikowia and an uncultured Saccharomycete (related to Metschnikowia) were also detected by PCR-DGGE (Figure 3.9). Pichia guilliermondii was isolated once by enrichment of grapes of Cabernet sauvignon.

3.3.4.2. Hunter Valley Region, 2002-2003 For grapes collected at vineyards R and L during the 2002-2003 season, A. pullulans was the most prevalent species detected throughout all stages of grape development (Tables 3.8 and 3.10). Its prevalence was confirmed by all three analytical methods; plate culture, PCR-DGGE and enrichment culture. However, its population tended to be lower than that observed for the 2001-2002 season. Species that were isolated by plate culture in this season, but not encountered in the 2001-2002 season, included Pseudozyma antartica, Psz. prolifica, Graphiola cylindrica and Debaryomyces hansenii. Species belonging to Cryptococcus and Rhodotorula were commonly isolated by plate cuhure and enrichments. Although PCR-DGGE often failed to detect these yeast species, it did reveal the presence of two Candida species (C. galli and C. zemplinina) that were not found by cultural isolation (vineyard R, Table 3.9 and Figure 3.9). PCR-DGGE detected the black yeast, Hortaea werneckii once, on grapes at maturity stage III from vineyard L (Table 3.10, Figure 3.10).

Unlike the 2001-2002 data, Hanseniaspora and Metschnikowia species were not observed on grapes at maturity stages IV and V in the 2002-2003 season, by any method.

During both seasons, PCR-DGGE revealed the presence of several species of filamentous fungi on the Hunter Valley grapes. The sequences from DGGE bands corresponded to Phiala scopiformics, Raciborskiomyces longisetosum. Species belonging to Phoma and Alternaria and an uncultured fungal clone were also recovered (Figures 3.9 and 3.10). Table 3.7 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard R during the 2001-2002 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture

Grape Detection Maturity of grapes variety method I II III IV V Cabernet A. pullulans (6.3x100 A. pullulans (6.0x10') A. pullulans (5.4x100 A.pullulans (2.6x10') A.pullulans (4.3x10") sauvignon A sp. (5.0x10') F. globisporum (8.0x10') Cr. chernovii (1.8x10') Rh. laryngis (2.5xl0') Cr. laurentii (2.0x10') Cr. laurentii (2.3x10') Sp- ruberrimus ( 1.5x10^) A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Metschnikowia spp. Hanseniaspora Spp. A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Tr. fuciformis P. guilliermondii Hs. uvarum Rh. nothofagi Hs. guilliermondii Rh. sloofiae Hs. va I bye ns is M. pulcherrima Metschnikowia spp. Semillon A. pullulans (1.4x100 A. pullulans (3.7xl00 A. pullulans (8.5x100 A.pullulans (5.0x100 A.pullulans (1.7x100 Cr. laurentii (5.0x10') Cr. chernovi (2.1x10') Cr. laurentii (1.0x10') Hs. uvarum (2.9x1 OO Rh. glutinis (5.6x10') Sp. rose us (5.6x10') Cr. chernovi (5.0x10') M. pulcherrima (2.1x1 OO Rh. glutinis (3.0x10') A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Hanseniaspora spp. Hanseniaspora spp. Metschnikowia spp. Metschnikowia spp. A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Tr. fuciformis Hs. uvarum Hs. uvarum M. pulcherrima M. pulcherrima Metschnikowia spp. Table 3.8 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard R during the 2002-2003 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture.

Grape Detection Maturity of grapes variety method I II III IV V Cabernet a A. pullulans (1.1x10') A. pullulans (5.0x10') A. pullulans (6.0x10') A. pullulans (1.3x10') A. pullulans (1.9x10') sauvignon Cr. sp. (5.0x10') D. hansenii (5.0x10') A. sp. (5.3x10^) Psz. antartica (3.0x10^) Gr. cylindrica (5.0x10 ) Cr. dimmenae (1.0x10^) Rh. slooffiae (2.8x10^) A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Cgallli C. gallli A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Rh. mucilaginosa A. sp. Rh. nothofagi Cr. heveanensis Cr. heveanensis Semillon A.pulltdans (2.5x10^) A. pulltdans (5.0x10') A. pullulans (9.4x10') A. pullidans (4.6x10') A. pullidans (4.4x100 Psz. prolifica (1.0x10^) Cr. laurentii (5.0x10') Cr. dimennae (9.0x10^) Cr. magnus (9.3x10^) Cr. sp. (5.0x10') Rh. slooffiae (5.0x_10')_ Rh. glutinis (2.6x10') Cr. saitoi (1.5x10^) Rh. slooffiae (2.0x10') A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans C. gain C. galli C. zemplinina C. galli C. zemplinina A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans A. sp. Rh. glutinis Rh. nothofagi Rh. mucilaginosa Rh. nothofagi Cr. heveanensis Cr. heveanensis Table 3.9 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard L during the 2001-2002 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture.

Grape Detection Maturity of grapes variety method I II III IV V Semillon A. pullulans (1.0x100 A. pullulans (1.0x100 nd A. pullulans (1.1x10') A. pullulans (5.0x100 Cr. oeirensis (1.0x10^ Rh. laryngis (l.OxlOO Hs. uvarum (l.OxlOO M pulcherrima (4.0x10^) Cr. laurentii (l.OxlOO A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Hanseniaspora spp. A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Hs. uvarum Hs. uvarum M. pulcherrima M. pulcherrima

nd, not detected

Table 3.10 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard L during the 2002-2003 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture.

Grape Detection Maturity of grapes variety method II III IV _V Semillon A. pullulans (l.OxlO') A. pullulans (1.0x100 A. pullulans (7.5x100 A. pullulans (9.0x100 A. pullulans (1.6x100 Rh. glutinis (5.0x100. A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans A. sp. Hort. werneckii Cryptococcus spp. A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Cryptococcus sp. Rh. nothofagi 3.3.4.3 Mudgee region. 2002-2003 For vineyard HG in Mudgee, Aureobasidium pullulans was the predominant species on all grape varieties throughout cultivation (Table 3.11). A diversity of colony morphotypes and DGGE banding profiles was observed in this species (Figure 3.10). It was consistently detected by plate culture, PCR-DGGE and enrichment. Various species belonging to the genera Cryptococcus, Rhodotorula and Rhodosporidum were isolated by plate and enrichment culture throughout grape development. These yeasts were not detected by PCR-DGGE, even when their populations reached 10^ CFU/g.

The wine yeasts Hs. uvarum and M pulcherrima were absent on grapes at harvest. However, enrichment cultures of Merlot and Sauvignon blanc grapes revealed the presence of M. pulcherrima from immature berries (stage III). Species belonging to Pichia (P. acaciae, P. guilliermondii) were also isolated by enrichment (Merlot, stage III and IV, respectively).

Filamentous fungal species belonging to Phoma, Alternaria md Aspergillus were detected by the PCR-DGGE method in this region (Figure 3.10).

3.3.4.4 Griffith region, 2002-2003 Table 3.12 shows the yeast species and their populations found on four grape varieties in vineyard MW. Overall, yeast populations on grapes were relatively low throughout cultivation, at about 10^-10^ CFU/g. In some cases, no yeast colonies were isolated by plate or enrichment culture. The species, A. pullulans was less frequently encountered from plate and enrichment cultures, and it was notably absent from two or more stages of all grape varieties. However, PCR-DGGE detected this species in every stage of eveiy variety examined (Figure 3.10). DGGE bands corresponding to Cryptococcus and Rhodotorula species were also obtained, even when these yeasts were subdominant, or absent as determined by plate cultures. Isolates obtained by plating included Ustilago sp. and Cintracta cf limitata. Enrichment cultures commonly gave isolates of A. pullulans, and species belonging to Rhodotorula. In the early stages of grape development (stage I and II), species of yeasts detected by enrichment, not found by other methods were P. guilliermondii, D. hansenii, M. pulcherrima and T. delbrueckii. Tremella fuciformis, which was also found by enrichment cultures of grapes from the Hunter Valley, was detected once in Griffith.

At harvest, some differences in populations of red and white varieties were observed. Yeast populations on red grape varieties increased to CFU/g at harvest maturity while white grapes remained yeast poor. Enrichment cultures of red grape varieties gave P. giiilliermondii. Additionally, Cabernet sauvignon grapes gave a diversity of wine fermentative yeasts, including K. thermotolerans, T. delbrueckii and notably, S. cere\isiae.

Bands corresponding to species of Phoma were recovered from PCR-DGGE from grapes in this region (Figure 3.10). Table 3.11 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard HG during the 2002-2003 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture.

Grape Detection Maturity of grapes variety method I II III IV V Cabernet a A.pullulans (3.1x10^) A. pullulans (5.0x10') A. pullulans (1.5x10') A.pullulans (2.1x10') A. pullulans (3.5x10') sauvignon Cr. oeirensis (1.5x10^) Cr. chernovii (1.0x10') Cr. laurentii (5.0x10') A. sp. (3.0x10') Cr. sp. (5.0x10') Cr. sp. (5.0x10') Cr. phenolicus (5.0x10*) Cr. laurentii (1.0x10') b A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans c A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Merlot a A. pullulans (6.3x10') A. pullulans (1.0x10') A. pullulans (3.3x10') A.pullulans (1.7x10') A. pullulans (2.8x10') Cr. oeirensis (2.0x10^) Rh. laryngies (5.0x10') Cr. laurentii (5.0xl0') Cr. laurentii (1.3x10') Cr. sp. (1.0x10^) Cr. sp. (5.0x10') Rh.glutinis (1.0x10') b A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans c A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Cr. heveanensis M. pulcherrima P. guilliermondii P. acaciae Cr. heveanensis Sauvignon a A. pullulans (5.6x10^) A.pullulans (1.4x10') A. pullulans (1.7x10') A.pullulans (3.7x10') A. pullulans (5.3x10') blanc Cr. sp. (1.9x10') Cr. heveanensis {\.lx\&) Cr. magnus (7.0x10') Cr. oerensis (5.0x10') Cr. saitoi (5.0x10') Cr. saitoi (5.0x10') b A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans c A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Rhds. sphaerocarpum M. pulcherrima C. parapsilosis Rh. nothofagi Cr. heveanensis Table 3.12 Yeast species and populations (CFU/g) on grapes at different stages of maturity in vineyard MW during the 2002-2003 season as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture.

Detection Maturity of grapes variety method I II III IV V Cabernet a A. pullulam (5.0x10') Cr. chernovii (5.0x10') Ustilago sp. (5.0x10') Cr. saitoi (5.0x10') A. pullulans (3.2x10') sauvignon Rh. pallida (5.0x10') Cr. flavus (5.0x10') Cr. albidus (1.8x10^) Cr. heveanensis (1.0x103) b A. pullulans A. pullulans A. pullulans A. pullulam A. pullulam Rh. spp. c A. pullulam C. cf. etschellsii A. pullulans nd P. guilliermondii Rh. glutinis P. guilliermondii Rh. glutinis K. thermotoleram T. delbrueckii S. cerevisiae Tyrian a A. pullulans (1.5x10') Cr. saitoi (5.0x10') A. pullulans (5.0x10') Cin. cf limitata (5.0x10') A. pullulam (2.0x10") Cr. magnus (5.0x10') Cr. chernovii (2.0x10') Cr. sp. (8.0x10') b A. pullulans A. pullulans A. pullulam A. pullulam A. pullulans Rh. spp. c. A. pullulans A. pullulans nd nd P. guilliermondii Rh. glutinis D. hansenii Sauvignon a A. pullulans (1.0x10') A. pullulam (1.0x10') Rh. glutinis (5.0x10') Cr. magnus (5.0x10') nd blanc Cr. magnus (5.0x10') Cr. chernovii (2.5x10') Rh. minuta (5.0x10') Ustilago sp. (2.0x10') Cr. saitoi (5.0x10') Cr. saitoi (1.0x10') b A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Cr. spp. Cr. spp. c A. pullulans A. pullulans Cr. sp. A. pullulans A. pullulans D. hamenii Tr. fuciformis Cr. laurentii Rh. glutinis Rh. glutinis Semillon a A. pullulans (1.0x10') A. pullulam (1.3x10') Cr. saitoi (5.0x10') Cr. magnus (5.0x10') A. pullulans (5.0x10') Cr. saitoi (5.0x10') Cr. chernovii (2.0x10') Cr.Jlavus (1.0x10') b A. pullulans A. pullulans A. pullulans A. pullulans A. pullulans Cr. spp. Cr. spp. c T. delbrueckii A. pullulans A. pullulans A. pullulans A. pullulans Rh. glutinis M. pulcherrima Rh. nothofagi Rh. glutinis Cr. heveanensis nd, not detected Cabernet sauvignon, 2001-2002 Semillon, 2001-2002

Cabernet sauvignon, 2002-2003 Semillon, 2002-2003

I II III IV ^ Cg

Ap Cz

Ph Alt

Figure 3.9 PCR-DGGE analysis of yeast DNA in rinses of wine grapes from vineyard R at different stages during cultivation in the seasons 2001-2002 and 2002-2003. Lanes: 1,11, EI, IV and V represent the stages of maturity referred to in Table 3.2. Ap, AiireobasidhunpuJhilans-, Alt Alternana spp.; Cg. Candida gallr, Cz, Candida zenipJinina; Hs, Hanseniaspora spp.; M/?, Metschnikowia puJcherrima; M. Metschnikowia spp.; Ph, Phoma spp.. Phi s, PhiaJocephaJa scopiformis: Rc I Raciborskiomyces longisetosimi; U f c, Uncultured fungal clone; U S, Uncultured Saccharomycete. Vineyard L r

Semillon 2001-2002 Semillon, 2002-2003

I II III IV V I II III IV V • Ap . Hs • US ^mt mm iufc • Phi s • Rcl Ufc 'Ufc y Ufc Cr Ap - -m m M Ap Ap Ap A Ap Ap Ph Hm' Alt

Vineyard HG Vineyard MW Merlot, 2002-2003 Sauvignon blanc 2002-2003

Figure 3.10 PCR-DGGE analysis of yeast DNA in rinses of wine grapes at different stages from several vineyards during cultivation in the 2002-2003 season. Lanes: 1,11, EI, IV and V represent the stages of maturity referred to m Table 3.2. A, Aiireobasidhim sp.; A p, Aiireobasiditim pullulans; Alt Alternaria spp.; Asp, Aspergillus spp.; Ck Ciyptococcus spp.; Hs, Hanseniaspora spp.; Hm\ Hortaea wemeckii-, Ph, Phoma spp.. Phi s, Phialocephala scopiformis; Rc I Raciborskiomyces longisetosiim; Ufc, Uncultured fungal clone; U S, Uncultured Saccharomycete. 3.3.5 The diversity of yeast species associated with undamaged and damaged wine grapes

Comparisons of total yeast populations on undamaged and damaged grapes have been presented in Section 3.3.2. With minor exceptions, damaged grapes of both red and white grape varieties harboured significantly higher populations of yeasts than healthy, undamaged grapes. Comparisons of the yeast species on undamaged and damaged grapes are shown in Tables 3.13 and 3.14. The Tables 3.15 and 3.16 summarise incidences of yeast species occurring on these grapes by the each method.

There were no obvious influences of grape variety on yeast species and populations. Also few vineyard or geographical associations were observed. Seasonal influences on the yeast flora of healthy and damaged grapes were evident.

During the 2001-2002 season, damaged, ripe berries gave a predominance of^. pullulans, Hs. uvarum and M. pulcherrima. In most cases, higher populations of pullulans (lO'^-lO^ CFU/g) were detected on damaged berries. The exceptions were found in only four samples of damaged grapes from the Hunter Valley region (vineyard R; Cabernet sauvignon and Semillon, vineyard L; Semillon, vineyard F; Semi lion), in which this species was either absent or diminished. More frequently recorded on damaged, rather than healthy grapes were the species of Hs. uvarum and M. pulcherrima. On two occasions, C zemplinina was detected on damaged grapes, but not on undamaged grapes, at 10^ CFU/g.

Yeast species detected by PCR-DGGE were generally represented by those which were most abundant on plate culture. Bands from PCR-DGGE often included^, pullulans, M. pulcherrima, an uncultured Saccharomycete related to Metschnikowia, and two undescribed species of Metschnikowia. Several species of Hanseniaspora (Hs. uvarum, Hs. opuntiae, Hs. guilliermondii, Hs. clermontiae) were recovered by this method, but these were detected only as a single DGGE band (Figure 3.11).

There were, however, several examples in which the results from PCR-DGGE and plate culture were not in good agreement. PCR-DGGE revealed the presence of Metschnikowia on damaged grapes from vineyard CM in Mudgee (Cabernet sauvignon. Merlot and Shiraz grapes), where plate culture did not. A PCR-DGGE band corresponding to Hanseniaspora spp. was detected from damaged Shiraz grapes (vineyard MM, Hunter Valley), where it was not recovered by plate culture. Damaged grapes (Cabernet sauvignon, Merlot, Shiraz) from the vineyard HG in Mudgee showed the presence of C. zemplinina by PCR-DGGE but not by plate culture. However, on other damaged samples (Shiraz grapes; vineyard Cm and Semillon grapes, vineyard F), high populations of C. zemplinina were evident by plating, but its corresponding band was not obtained by PCR-DGGE (Tables 3.13 and 3.14).

A diversity of species belonging to Hanseniaspora {Hs. idvarum, Hs. opuntiae, Hs. guilliermondii, Hs. valbyensis) and Metschnikowia were often prominent in enrichment cultures of both damaged and undamaged grapes in the 2001-2002 season. Their presence was at times revealed by enrichment culture, but not by plate culture or by PCR-DGGE. Enrichment cultures also revealed the presence of various other fermentative yeast species on damaged and undamaged grapes. These included H thermotolerans, T. delbrueckii, I. orientalis, P/c/i/aspp. (Tables 3.13 and 3.14). Candida zemplinina was isolated twice on damaged grapes from enrichment cultures (Shiraz, vineyard Cm; and Sauvignon blanc, vineyard HG), and once from undamaged grapes (Shiraz, vineyard Cm). Saccharomyces cerevisiae was isolated twice from enrichment cultures of damaged grapes from vineyard HG (Merlot and Sauvignon blanc).

In contrast with the 2001-2002 season, Hanseniaspora and Metschnikowia species were scarce on damaged and intact ripe grapes during the 2002-2003 season. Major species recovered from grapes harvested during this hotter, drier season included A. pullulans and basidiomycetous yeasts belonging to Cryptococcus and Rhodotorula. The presence of these yeasts could also demonstrated in enrichment cultures, and occasionally by PCR-DGGE.

In Griffith, S. cerevisiae and other fermentative yeasts were recovered by enrichment from only three samples of red grape varieties (undamaged and damaged Cabernet sauvignon grapes, and damaged Tyrian grapes). These yeasts were absent in white varieties cultivated in the same vineyard. In other regions, Torulaspora delbrueckii was the only other fermentative yeast detected during this season; it was found by enrichment culture from damaged harvest ripe grapes from vineyards R and L of the

Hunter Valley. None of these fermentative yeasts were detected on ripe, undamaged or damaged grapes by either plate culture or PCR-DGGE methods (Tables 3.13 and 3.14).

Table 3.13 Yeast species on undamaged and damaged red grapes at commercial harvest time from several vineyards as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture

Grape Vineyard, Detection Undamaged Damaged variety season method a A. pullulans (2.5x10") A. pullulans (2.5x10') Hs. uvarum (3.8x10') Hs. uvarum (1.8x10^) M. pulcherrima (6.0x10') M. pulcherrima (l.lxlO'')

F' b A. pullulans A. pullulans Metschnikowia spp. Hanseniaspora spp. Metschnikowia spp. c Hs. uvarum Hs. uvarum M. pulcherrima Hs. guilliermondii Metschnikowia spp. M. pulcherrima Metschnikowia spp. a A. pullulans (1.4x10') A. pullulans (5.5x10') Cr. laurentii (7.5x10') Cr. laurentii (3.0x10') Rhds. babjevae (1.1x10") Rhds. babjevae (4.5x10') MM' b A. pullulans A. pullulans Hanseniaspora spp. Hanseniaspora spp. c Hs. uvarum Hs. uvarum Hs. guilliermondii a A. pullulans (4.1x10') A. pullulans (1.6x10^) Cr. laurentii (1.3x10') Hs. uvarum (3.3x10^) Shiraz Cr. magnus (1.8x10') M. pulcherrima (1.9x10')

b A. pullulans A. pullulans HG' Metschnikowia spp. C. zemplinina Hanseniaspora spp. Metschnikowia spp. c Hs. uvarum Hs. uvarum M. pulcherrima M. pulcherrima Metschnikowia spp. Metschnikowia spp. K. thermotolerans K. thermotolerans a A. pullulans (3.0x10') A. pullulans (4.5xi^) Cr. victoriae (8.0x10') Hs. uvarum (9.7x10') C. zemplinina (2.2x10^^) D. hansenii (2.6x10') CM' b A. pullulans A. pullulans Metschnikowia spp. Metschnikowia spp. c Hs. uvarum Hs. uvarum M. pulcherrima M. pulcherrima Metschnikowia spp. Metschnikowia spp. P. anomala K. thermotolerans C. zemplinina C. zemplinina

Combandry; MW McWilliams. Seasons 2001-2002;2002-2003 Table 3.13 extended

Grape variety Vineyard, Detection Undamaged Damaged season method a A. puHulans (4.3x10^) Hs. uvarum (5.0x10') Rh. laryngis (2.5x10^) M. pulcherrima (1.2x10') Sp. ruberrimus (1.5x10^) b A. pullulam Metschnikowia spp. R' Metschnikowia spp. Hanseniaspora spp. Hanseniaspora spp. c A. pullulons Hs. uvarum Hs. uvarum Hs. volbyensis Hs. guilliermondii M. pulcherrima Hs. valbyemis Metschnikowia spp. M. pulcherrima Meischnikowia spp. a A. pullulam (l.IxlO') A. pullulons (I.6XLÔY~ Cr. laurentii (5.2x10^) Hs. uvarum (4.0x10^) M. pulcherrima (1.4x10') b A. pullulons A. pullulons HO' Metschnikowio spp. Hanseniaspora spp. C. zemplinina c A. pullulons Hs. uvarum Hs. uvarum M. pulcherrima Cabernet M. pulcherrima Metschnikowio sp. sauvignon K. thermotolerans T. delbrueckii a A. pullulons (3.1x10') A. pullulons (2.0x10^) Sp. roseus (6.8x10^) Cr. sp. (1.0x10') Cr. sp. (3.8x10^) CM' b A. pullulons A. pullulons Meischnikowia sp. Metschnikowio sp. c Hs. uvarum Hs. uvarum M. pulcherrima K. thermotolerans T. delbrueckii a A. pullulons (1.9x10^) A. pullulons '"(376~XÏÔY Psz. aniartica (3.0x10^) Cr. saitoi (7.2x10^) Rh. slooffioe (2.8x10^) R^ b A. pullulons A. pullulam C. gain c A. pullulons A. pullulons Rh. glutinis a A. pullulons (3.5x10') A. pullulam (1.2x10^) Cr. laurentii (1.0x10^) Cr. laurentii (3.0x10^) HG^ A. sp. (3.0x10^) A. sp. (7.0x10^) b A. pullulons A. pullulons c A. pullulons M. pulcherrrimo a A. pullulons (3.2x10') A. pullulons (6.2x10"^ A. sp. (3.0x10^) A. sp. (7.0x10^) Cr. albidus (1.8x10^) Cr. albidus (1.5x10^) Cr. heveonensis (1.0x10^) Cr. chernovii (2.5x10^) MW^ b A. pullulons A. pullulam Rhodotorula spp. Rhodotorula spp. c P. guilliermondii S. cerevisiae K. thermotolerans P. guilliermondii T. delbrueckii S. cerevisiae Vineyards R, Rosemount; F, Fernandes; L, Lindemans; MM, Molly Morgan; HG, Hill of Gold; Cm, Combandry; MW McWilliams Seasons 2001-2002; \ 2002-2003 Table 3.13 extended

Grape Vineyard, Detection Undamaged Damaged variety season method a A. pullulans (2.6x10') A. pullulans (4.5x10') Rh. glutinis (2.5x10') Hs. uvarum (5.4x10') M. pulcherrima (1.7x10^) b A. pullulans A. pullulans HG' Metschnikowia spp. Hanseniaspora spp. Metschnikowia spp. C zemplinina c A. pullulans Hs. uvarum Hs. uvarum Hs. opuntiae Hs. opuntiae P. anomala M. pulcherrima K. thermotolerans P. anomala S. cerevisiae Merlot a A. pullulans (3.2x10') A. pullulans (l.lxToT" Cr. victoriae (7.3x10') Cr. victoriae (2.8x10^) Sp. roseus (2.5x10') Sp. roseus (4.0x10^) CM' b A. pullulans A. pullulans Metschnikowia spp. Metschnikowia spp. c A. pullulans Hs. uvarum Hs. uvarum K. thermotolerans Hs. guilliermondii a A. pullulans (2.8x10') A. pullulans (1.4x10^ Cr. laurentii (1.3x10') HG' b A. pullulans A. pullulans c A. pullulans Hs. opuntiae M. pulcherrima a A. pullulans (2.0x10') A. pullulans "(1.6xl6y Cr. chernovii (2.0x10^) Cr. chernovii (2.5x10') Tyrian Cr. sp. (8.0x10') Cr. sp. (1.0x10') MW^ b A. pullulans A. pullulans Rhodotorula spp. c P. guilliermondii S. cerevisiae Vineyards R, Rosemount; F, Fernandes; L, Lindemans; MM, Molly Morgan; HG, Hill of Gold; Cm, Combandry; MW McWilliams Seasons 2001-2002; ^ 2002-2003 Table 3.14 Yeast species on undamaged and damaged white grapes at commercial harvest time from several vineyards as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment culture

Grape variety Vineyard, Detection Undamaged Damaged season method a A. pullulans (8.2x10') A. pullulans (7.5x10') R' b A. pullulans A. pullulans c M. pulcherrima M. pulcherrima a A. pullulans (5.4x10') A. pullulans (3.5x10^ M. pulcherrima (1.1x10') Hs. uvarum (1.8x10^) M. pulcherrima (5.0x10^) HG' b A. pullulans Hanseniaspora spp. Metschnikowia spp. c Hs. uvarum Hs. uvarum hi. pulcherrima M. pulcherrima Metschnikowia spp. Metschnikowia spp. Sauvignon K. thermotolerans C. zemplinina blanc K. thermotolerans S. cerevisiae a A. pullulans (5.3x10") A. pullulans (1.8x10') Cr. saitoi (5.0x10') Cr. saitoi (4.5x10^) HG' b A. pullulans A. pullulans c A. pullulans A. pullulans Rh. nothofagi Rh. nothofagi a nd Cr. saitoi (7.5x10') Rh. glutinis (5.0x10') MW' b A. pullulans A. pullulans c A. pullulans A. pullulans Rh. glutinis

Vineyards R, Rosemount; F, Femandes; L, Lindemans; MM, Molly Morgan; HG, Hill of Gold; Cm, Combandry; MW McWilliams Seasons 2001-2002; \ 2002-2003 Nd, not detected Table 3.14 extended Grape variety Vineyard, Detection Undamaged Damaged season method a A. pullulons (1.7x10') Hs. uvorum (6.5x10') Hs. uvorum (2.9x10^) M. pulcherrimo (9.4x10') M. pulcherrima (2.1x10^) Rh. gluiinis (3.0x10^) R' b A. pullulons Honseniosporo spp. Hanseniaspora spp. Metschnikowia spp. Metschnikowia spp. c A. pullulons Hs. livarum Hs. uvorum M. pulcherrimo M. pulcherrimo Metschnikowia spp. Metschnikowia spp. a A. pullulons (5.0x10") A. pullulons (2.2x10') Hs. uvorum (1.0x10") Hs. uvorum (2.7x10-') M. pulcherrima (4.0x10^) M. pulcherrimo (3.1x10") Cr. laurentii (1.0x10^) L' b A. pullulons A. pullulons Hanseniaspora spp. Honseniosporo spp. c A. pullulons Hs. uvorum Hs. uvorum Hs. volbyensis M. pulcherrimo M. pulcherrima a A. pullulons (1.9x10") A. pullulans (8.0x10') Hs. uvorum (3.6x10") Hs. uvarum (1.1x10^) M. pulcherrimo (3.1x10") M. pulcherrima (1.4x10^) C. zemplinino (3.0x10^') F' b Hanseniaspora spp. Honseniosporo spp. Semillon Metschnikowio spp. c Hs. uvorum Hs. uvarum M. pulcherrimo M. pulcherrima Metschnikowia spp. Metschnikowio spp. /. orientons /. orientalis a A. pullulons (5.3x10^) A. pullulons (4.2x10-) Hs. uvorum (8.5x10^) Hs. uvorum (4.3x10^) Cr. tnagnus (1.0x10^) M. pulcherrima (1.7x10-') Sp. roseus (1.8x10^) MM' b A. pullulons Honseniosporo spp. Metschnikowio spp. c Hs. uvorum Hs. opuntioe Hs. opuntiae Hs. guilliermondii Hs. guilliermondii C. cf. glabrata a A. pullulons (4.4x10-') A. pullulons (5.9x10") Cr. sp. (5.0x10') Cr. laurentii (1.5x10^) Rh. slooffiae (2.0x10^) Rh. slooffiae (6.0x10-') R^ b A. pullulons A. pullulans C. gai h c A. pullulons A. pullulans Rh. muciloginosa C. paropsilosis T. delbruekii a A. pullulons (1.6xlÔy~ A. pullulons (3.3x10'^) Rh. glutinis (5.0x10^) Rh. sp. (1.0x10^) L' b A. pullulons A. pullulans c A. pullulons A. pullulons T. delbruekii Rhodotorulo spp. a A. pullulons (5.0x10') A. pullulans (2.9x10')

MW^ b A. pullulons A. pullulans A. pullulons Cr. heveanensis Rh. glutinis Vineyards R, Rosemount; F, Fernandes; L, Lindemans; MM, Molly Morgan; HG, Hill of Gold; Cm, Combandry; MW McWilliams Seasons 2001-2002; ^ 2002-2003 Cabernet sauvignon, 2001-2002

Vineyard R Vineyard HG Vineyard Cm

U D

Hs US

Ap

Mp M M M

Semillon, 2001-2002

Vineyard R Vineyard L Vineyard F

Figure 3.11 PCR-DGGE analysis of yeast DNA in rinses of undamaged (U) and damaged (D) wine grapes from several vineyards at commercial harvest time in the season 2001-2002. Lanes U, undamaged grapes at maturity stage V; D, damaged grapes at maturity stage Y Ap, Aureobasidhimpullulans; Cz, CandidazempJmina; Hs, Hanseniaspora spp.; Mp, Metschnikowiapulcherrima; M, Metschnikowia spp.; U S, Uncultured Saccharomycete. Cabernet sauvignon, 2002-2003

Vineyard R Vineyard HG Vineyard MW

U D

•Ap

Ap

.Ufc • Ap

Ph Alt

Semillon, 2002-2003

Vineyard R Vineyard L Vineyard MW

U D Cg Ap

•Ap

m Ap

• Ph 'Alt

Figure 3.12 PCR-DGGE analysis of yeast DNA in rinses of undamaged (U) and damaged (D) wine grapes from several vineyards at commercial harvest time. Lanes U, undamaged grapes at matunty stage V; D, damaged grapes at maturity stage V A p, Aureobasidhim piilluJans; Alt, AUemaria spp.; C g, Candida gallr, Ph Phoma spp.; Rh, Rhodotorula spp.; Ufc, Uncultured fungal clone. 3.3.6 Frequency of isolation of yeasts from wine grapes

Aureobasidium pullulans was prevalent on both undamaged and damaged red and white wine grapes, as determined by all three methods, plate culture, enrichment culture and PCR-DGGE (Tables 3.15 and 3.16). It was much less frequently recovered from enrichment cultures, than plate culture or PCR-DGGE. Damaged berries showed an increased incidence of species of Hanseniaspora and Metschnikowia, by plate culture and PCR-DGGE. In addition to Hs. uvarum and M. pulcherrima, a diversity of apiculate yeasts and pulcherrimin-producing colonies were more readily recovered by enrichment culture.

Basidiomycetous yeasts were commonly isolated on both damaged and undamaged grapes by plate culture. However, these were only occasionally detected by PCR-DGGE or enrichment. The species Candida galli was found rarely, and only by PCR-DGGE. Candida zemplinina was rarely encountered; it was detected by all three methods but with little agreement between the three analytical methods on the same sample.

Saccharomyces cerevisiae and other wine-related yeast species (K. thermotolerans, T. delbrueckii, Pichia spp.) were more commonly found on damaged berries. These yeasts were isolated only by enrichment. Table 3.15 Main yeast species isolated from undamaged and damaged red grape varieties at the time of commercial harvest from all vineyards during the 2001-2002 and 2002-2003 seasons, as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment methods

Yeast species Undamaged grapes (n= 14) Damaged grapes (n= 14) Detection method a b c a b c A. pullulans 14 14 7 13 13 1 Hanseniaspora spp. 1 2 9 6 6 10 Metschnikowia spp. 1 7 7 5 8 7 Cryptococcus spp. 13 0 0 9 0 0 Rhodotorula spp. 4 2 0 1 1 1 C. gain 0 1 0 0 0 0 C. zemplinina 0 0 1 2 2 1 Pichia spp. 0 0 4 0 0 2 K. thermotolerans 0 0 2 0 0 5 T. delbrueckii 0 0 2 0 0 2 S. cerevisiae 0 0 1 0 0 3 n= number of samples examined

Table 3.16 Main yeast species isolated from undamaged and damaged white grape varieties at the time of commercial harvest from all vineyards during the 2001-2002 and 2002-2003 seasons, as determined by (a) plate culture, (b) PCR-DGGE and (c) enrichment methods

Yeast species Undamaged grapes (n= 11) Damaged grapes (n= 11) Detection method a b c a b c A. pullulans 10 10 7 9 7 5 Hanseniaspora spp. 4 5 5 5 5 5 Metschnikowia spp. 4 4 5 5 4 5 Cryptococcus spp. 4 0 0 2 0 0 Rhodotorula spp. 3 0 2 2 0 3 C. gain 0 1 0 0 0 0 C. zemplinina 0 0 0 1 0 1 Pichia spp. 0 0 0 0 0 0 K. thermotolerans 0 0 1 0 0 1 T. delbrueckii 0 0 0 0 0 2 S. cerevisiae 0 0 0 0 0 1 n= number of samples examined 3.4 DISCUSSION

3.4.1 Analytical strategies for characterising yeasts associated with wine grapes

Previous literature has discussed the importance of sound sampling and analytical procedures to reliably describe the yeast communities associated with wine grapes. It is now known that sample variation occurs within vineyards (Davenport 1973, 1974, 1976; van der Westhuizen et al. 2000b), within bunches or clusters of berries (Rosini et al. 1982; Torok et al. 1996; Mortimer and Polsinelli 1999), and within a single berry (Belin 1972; Rosini et a/. 1980). The influences of berry maturity, and the differences between healthy and damaged grape berries on the yeast ecology have also been recognised (Rosini et al. 1982; Torok et al. 1996; Mortimer and Polsinelli 1999). Martini and coworkers have reviewed historical aspects of this field, and have provided a critique on previously applied sample preparation and isolation procedures (Martini 1993; Martini et al 1980; Vaughn-Martini and Martini 1995; Martini 1996). The limitations of traditional, cultural methods in microbial ecological studies have already been documented, where only a small portion (an estimated 0.1-10%) of organisms in natural environments is thought to be cultivated. (Head et al 1998). As a consequence, molecular ecological profiling techniques, such as PCR-DGGE have been increasingly applied to provide a cultivation-independent assessment of microbial communities, including yeasts in foods and beverages (Beh et al 2006; Fernandez-Espinar et al 2006).

The studies presented in this Chapter were designed to consider the influences of some of the variables just mentioned and to maximise the reliability of the data obtained. Grapes representing six different varieties and harvested from seven different vineyards were examined. Vineyard and grape sampling variables were addressed by collecting grapes from at least five different vines within each vineyard plot. Duplicate samples were taken so that grapes from at least ten different locations within the vineyards were examined. At the time of sampling, care was taken in differentiating and separating sound, physically intact berries from damaged berries. Three analytical methods were used to investigate yeasts associated with wine grapes in order to maximize the recovery of the biota. These were plate culture, enrichment culture and PCR-DGGE.

Sample preparation The standard procedure for the isolation of yeasts from foods involves rinsing or macerating food samples to release cells from the food matrix (Fleet 1999). Martini et al. (1980), Rosini et al (1982) and Martini et al. (1996) previously reported that the rinsing procedure does not efficiently detach yeast cells from the surface of grapes and advocated the application of more vigorous sample pretreatments, such as sonication and jet streaming. These methods often revealed higher yeast counts compared with rinsing, but did not result in the recovery of additional species (Bisson and Kunkee 1991). Jet streaming is not practical to use on a routine basis and sonication introduces a risk of inactivating or killing species with weaker cell walls. Combina et al (2005) showed that blending grapes produced significantly higher counts than those obtained by rinsing or jet streaming treatments. However, the composition of yeast species obtained by each method was not evaluated. In this study, rinsing of grapes was shown to release about 60% of the total yeast population, with the remaining cells obtained upon subsequent maceration of the rinsed grapes. In many cases, similar yeast species were found by rinsing and maceration of grapes. Nevertheless, additional species were recovered after maceration in a few grape samples (Table 3.4). Thus, some species of yeasts may have been missed or overlooked by the rinsing procedure alone. Maceration brings grape constituents into the homogenate and some of these (e.g. organic acids, polyphenols) could however, have a negative impact on the survival of some yeasts. Such influences would also vary with the maturity of grape. Preliminary studies in this laboratory also revealed a negative influence of grape homogenates on the performance of PGR assay as part of the DGGE analyses (Prakitchaiwattana, 2005). This effect varied with maturity of the grape and was especially problematic with damaged grape berries. The inhibition of PGR assay by plant cell constituents (lipids, polyphenols and polysaccharides) has been reported before (Wilson 1997). The release of plant DNA by maceration may have also interfered with the PGR amplification of yeast DNA. However, a number of measures to improve PGR amplification from must or wine matrices have since been reported (see for example, Phister and Mills 2003; Delaherche et al 2004; Martorell et al 2005). The occurrence of yeasts within intact tissues of fruit has been previously documented (Leben 1972). There are several records of yeasts isolated from sound, surface disinfested grapes (Dugan et al 2002). Yeasts appeared to be a relatively minor component of the internal mycoflora of grapes (Table 3.5) but these would contribute to the total counts obtained by maceration.

Plate culture and PCR-DGGE Plating of grape rinses onto MEA or WLNA enabled the enumeration of total populations and individual species of yeasts (detection limit 10 CFU/ml), while PCR- DGGE provided a culture-independent assessment of the yeast diversity, and might reveal the presence of viable but non-culturable species.

The plating media, MEA and WLNA supported equivalent populations and numbers of yeast species from grape rinses. Overall, plate culture revealed a greater diversity of yeast species compared with banding profiles obtained by PCR-DGGE. Aureobasidium pullulans was the most dominant species found on wine grapes. Its prevalence was demonstrated by both plate culture and PCR-DGGE. Both methods revealed significant heterogeneity for this organism as observed in colony morphology on agar plates, and variability in DGGE banding profiles. On ripe grapes, particularly those of damaged berries, both methods detected the presence of Hanseniaspora and Metschnikowia yeasts.

There were, however, many occasions where PCR-DGGE analyses did not give data that were consistent with plate count results. Various species of Cryptococcus, Rhodotorula, Rhodosporidium and Sporobolomyces were sporadically isolated on wine grapes, often at lO'-lO^ CFU/g. The corresponding PCR-DGGE bands for these species were seldom found. In evaluating the PCR-DGGE methodology, Prakitchaiwattana et al (2004) showed that yeast populations as low as 10^ CFU/g could be detected from grapes, but that the detection limit of individual yeast species was determined by their relative proportions in mixed populations. Observations from the present study also suggested that minor and subdominant populations of yeasts on grapes were not detectable by PCR-DGGE (Tables 3.13 and 3.14). Several authors have previously reported that some yeast species may not be resolved by PCR-DGGE analyses (see for example, Gildemacher et al. 2004). In this study, DGGE bands for Rhodotorula glutinis and Rh. mucilaginosa occurred in the same migrating position, and species of Hanseniaspora {Hs. uvarum, Hs. opuntiae, Hs. guilliermondii) were recovered as a single DGGE band. These findings might explain the failure of PCR-DGGE to recover some yeast species. Prominent bands of filamentous fungal species were often recovered from grape rinses, and it is possible that the abundance of non-target organisms compromised detection of the targeted yeasts.

On other instances, PCR-DGGE revealed the presence of yeast species that were not isolated by culture on agar media. Candida galli and C. zemplinina were notable examples of unculturable species found on berries in vineyard R, during the drought- affected season. PCR-DGGE analysis has previously demonstrated the presence of unculturable organisms in food and wine fermentations (Ampe et al. 1999; Millet and Lonvaud-Funel 2000; Roling et al. 2001; Mills et al. 2002; Divol and Lonvaud-Funel 2005). Aerial plant surfaces are hostile environments for most microorganisms, where they are exposed to wide and repeated fluctuations in temperature, radiation, water availability, sparse nutrients or exposure to xenobiotic agents. The VBNC state has been reported for phylloplane bacterial species such as Pseudomonas fluorescens, Pseudomonas syringae, Ralstonia solanacearum and Xanthomonas campestris (Lindlow and Brandi 2003). Conceivably, some phylloplane yeasts could be induced into the VBNC in response to similar stresses.

PCR-DGGE analyses of grape rinses sometimes yielded multiple banding patterns for A. pullulans, although pure cultures of this species only gave one prominent band (for each colony morphotype). The pure, reference culture of M. pulcherrima gave rise to two bands on DGGE gels (Figure 3.5), and multiple DGGE bands were also observed for Metschnikowia spp. and M pulcherrima from grape rinses. Such phenomena have been widely documented in DGGE literature (Hernán Gómez et al. 2000; Mills et al. 2002; Gadanho et al. 2004; Masoud et al. 2004; Prakitchaiwattana et al. 2004). In pure cultures, the observations may be explained by microheterogeneity of rDNA copies within a single strain. In environmental samples, multiple bands could suggest the presence of strain variability, as observed with^. pullulans. In both cases, these may reflect PGR amplification artifacts such as heteroduplex/chimeric molecules, incomplete GC extensions or priming infidelity, and also artifacts derived from the DGGE (e.g. single stranded DNA fragments, secondary structures in DNA) (reviewed in Beh et al. 2006).

Enrichment cultures Saccharomyces cerevisiae is thought to occur rarely or at very low populations on grapes (Martini 1993; Martini et al. 1996; Mortimer and Polsinelli 1999). Enrichment cultures were included in this study to encourage the development of fermentative yeast species and increase the prospects of detecting S. cerevisiae from grapes.

The yeast species expected to proliferate in enrichment cultures were those capable of tolerating low pH and high sugar concentrations (Vaughn-Martini and Martini 1995). Dominant organisms found in enrichment cultures of intact grape berries were weakly fermentative species belonging to Hanseniaspora and Metschnikowia, which were less frequently detected by plate culture or PGR-DGGE, particularly on undamaged red grapes. Strongly fermentative yeasts such as K. thermotolerans, T. delbrueckii and S. cerevisiae were detected only by enrichment. Although some researches have readily isolated S. cerevisiae on wine grapes by enrichment cultures (Mrak and McGlung 1940; Torok et al. 1996; van der Westhuizen et al. 2000a, 200b), it was very rarely found in the present study. Aureobasidium pullulans and related isolates were also commonly detected by enrichments, but much higher incidence of these organisms was revealed by plate culture and PGR-DGGE analyses, and this difference was notable for red grapes (Tables 3.16 and 3.17).

Various approaches to enrichment have been used, such as incubating individual berries in enrichment media (Mortimer and Polsinelli 1999), incubating samples of grape macerates in enrichment media (Mrak and McGlung 1940) and allowing grape macerates to self/autoenrich (Benda 1964; Martini 1993; van der Westhuizen et al 2000a, 2000b). These different approaches to enrichment may have lead to seemingly anomalous or discrepant observations reported in the literature for isolation of S. cerevisiae and other species.

Some differences were noted between intact grape enrichments and autoenrichment cultures; from the samples examined, it appeared that the autoenrichment procedure provided a more favourable environment for the proliferation of fermentative yeast species, including S. cerevisiae (Table 3.6). However, isolation of iS". cerevisiae was also successful by intact grape enrichments, on five occasions (Tables 3.15 and 3.16). Given the rarity and infrequency with which S. cerevisiae is encountered on grapes, the differences observed here could be an effect of sample variation. Overgrowth of, and the antagonistic interactions by dominant species could also provide another explanation for the difficulty of isolating S. cerevisiae from enrichment cultures.

Higher incidences of A. pullulans and Pichia spp. were recorded from intact grape enrichments compared with macerated autoenrichment. The chemical composition of grape juice and that of the grape epicuticular wax, and its influence on yeast growth has been previously reviewed (Ribereau-Gayon et al. 2000). Macerated autoenrichments may have affected the levels of yeast nutrients or antinutrients present. Oleanolic acid, the main component of the epicuticular wax surrounding the grape, is thought to serve as growth factors for some yeasts. Conversely, the release of phytoalexins, polyphenolic and terpenic compounds or pesticide residues may have adversely affected the growth of some yeasts.

Although it was not isolated here, the spoilage yeast Dekkera brnxellensis is known to be associated with wine grapes (Loureiro and Malfeito-Ferreira 2003). Its isolation usually involves a more selective/differential approach to support its slow growing and fastidious nature (Couto et al. 2005). Such enrichment cultures were not performed here, and it is a limitation of this study. Similarly, it may be necessary to design highly selective conditions to isolate S. cerevisiae from grapes; for example, the inclusion of ethanol (7.6%) in enrichment media (Sniegowski et al 2002). 3.4.2 Population and species diversity Yeast populations on developing berries were 10^-10^ CFU/g and increased to 10^-10^ CFU/g as grapes progressed to harvest ripeness. Seasonal variability on yeast populations was observed, with lower populations observed in the warmer, drier vintage. Damaged berries supported higher yeast densities, often harbouring 10^-10^ CFU/g. These population trends are in the same order reported in other ecological studies of leaves, wine grapes and other phylloplane surfaces (Parle and di Menna 1966; Yanagida et al. 1992; Suarez et al. 1994; de la Torre et al. 1998a; Sabate et al 2002; Raspor et al 2006).

Aureobasidium pullulans was the most dominant component of the mycoflora of wine grapes. It was isolated from almost all samples of unripe, ripe and with some minor exceptions, also from damaged grapes. Distinct and heterogeneous colony morphotypes and DGGE banding patterns corresponding to A. pullulans and other Aureobasidium spp. were consistently recovered. In addition to strain variability, sequence comparisons of the partial 26S rDNA showed low identities (91-96%) to^. pullulans, and strongly indicated that species oiAureobasidium (possibly novel lineages), other than^. pullulans were present. Chapter Four presents a more detailed taxonomic study of these organisms. Despite their widespread association with grapes, few studies have considered the implications of their growth in the winemaking process; this topic is examined and discussed in Chapter Four.

After A. pullulans, various species belonging to Cryptococcus, Rhodotorula, Rhodosporidium and Sporobolomyces were also widely encountered on grapes. The isolation of these basidiomycetous yeasts, along with A. pullulans was not unexpected, as they are regarded as strong colonisers of leaves and fruit surfaces of many plants, including the grapevine (Cooke 1959; Benda 1964; Parle and di Menna 1966; Davenport 1974, 1976; Andrews et al 1994; Sabate et al 2002). Physiological features common to these organisms include the production of photoprotective compounds (mycosporines, carotenoids) and extracellular polysaccharides. These properties are thought to confer microorganism resistance to UV and desiccation, respectively (reviewed in Fonseca and Inacio 2006) and could select for their growth on surfaces of wine grapes. The ability of these organisms to produce a range of hydrolytic enzymes has also been recognised and may also play an important role in the successful colonisation of the grape surface (Cenakova et al. 1980; Federici 1982; Middelhoven 1997; Buzzini and Martini 2002). In addition, the species, Aureobasidiumpullulans grows well in conditions of low nutrient availability (Andrews and Harris 2000). It is also a known producer of antimicrobial compounds (McCormack et al 1994, 1995). Such properties could also determine their competitive fitness on the phylloplane habitat.

Many of the species associated with unripe grapes have also been isolated from ripe grapes, but at this later stage, Hs. uvarum was frequently reported as the dominant species and could account for up to 70-80% of the total yeast flora (Martini et al. 1996; Raspor et al 2006). The prevalence of Metschnikowia pulcherrima on ripe grapes has also been widely reported (van Zyl and Du Plessis 1961; Minarik 1965; Bamett et al. 1972, Davenport 1974, 1976; Sabate et al. 2002). The prevalence of these two species on healthy mature grape berries was only partially supported by the findings of this thesis, more so by data from enrichment analysis than data by agar plate culture. Various species of Hanseniaspora, in addition to Hs. uvarum were isolated. These included Hs. opuntiae, a recently described apiculate yeast; Hs. guilliermondii and Hs. valbyensis, species frequently associated with grape/musts. Hanseniaspora opuntiae was described from cacti (Cadez et al 2003) and this is the first finding of the species on grapes.

Distinct colony morphotypes among the pulcherrimin producing strains of Metschnikowia have been recovered from botrytis-infected grapes (Sipiczki 2006) and from native wine fermentations (Pallmann et al 2001). In common with these observations, M. pulcherrima and several undescribed Metschnikowia spp., were isolated from grapes, and these were distinguishable by colony and ITS-RFLP restriction patterns. Although not isolated here, M. viticola (Péter et al 2005) M. fructicola (Kurtzman and Droby 2001) and M. reukaufii (Raspor et al 2006) have been previously reported from grapes.

Grapes from the warmer, drier season of 2002-2003 did not give any isolation of apiculate yeasts, and while M. pulcherrima was isolated twice, these were recovered from immature berries (development stage II and III). Other studies have also recorded an absence of Hs. uvarum and M. pulcherhma from some samples of mature grapes (Parish and Carroll 1985; Yanagida et al. 1992; de la Torre et al 1999; Sabate et al 2002; Jolly et al 2003), but the reasons for such variations are unclear.

The association of the principal wine yeast, Saccharomyces cerevisiae with grapes has been a widely debated topic. Some authors have reported difficulty in isolating S. cerevisae (Martini 1993, Combina et al 2005), while others have readily isolated this yeast by enrichment culture (Mrak and McClung 1940; Török et al 1996; van der Westhuizen et al 2000a, 2000b) and in microvinifications (Redzepovic et al 2002; Capello et al 2004; Demuyter et al 2004; Antunovics et al 2005; Schuller et al 2005). It has been successfully isolated by enrichment from other natural habitats such as soils and tree exudates (Naumov et al 1998; Sniegowski et al 2002; Johnson et al 2004). The species was isolated only five times from 115 grape samples examined in this study, and only from enrichment cultures. Pseudohyphal development has been observed in S. cerevisiae, and is believed to be a mechanism to aid its colonisation on grapes (Khan et al 2000; Pretorious 2000; Goignes et al 2001, 2006). However, inoculation of S. cerevisiae on wine grapes in the vineyard gave poor growth (Comitini and Ciani 2006) and these observations here would suggest that S. cerevisiae is not favourably associated with natural environments such as wine grapes (Phaff and Starmer 1987).

The findings of several yeast and yeast-like fungi on grapes are novel observations of this study; the occurrence of Candida stellata on damaged, overripe and botyrized grapes is widely documented (Kroemer and Krumbholz 1931; Le Roux et al 1973; Rosini et al 1982; Guerzoni and Marchetti 1987). Candida stellata was not recovered from any grape samples in this study. However, the yeast SI^QCIQS Candida zemplinina was detected, occasionally by enrichment culture and PCR-DGGE and rarely by plate culture. This species is closely related to C. stellata-, it is an osmotolerant species, recently described from botrytized grape musts (Mills et al 2002; Sipiczki 2003; Magyar and Bene 2006). Candida galli was found on grapes only from vineyard R, in the 2002-2003 season. It was detected only by PCR-DGGE, but never isolated by cultural isolation methods. This species phenotypically resembles Yarrowia lipolytica, and has been recently described from poultry meat (Péter et al. 2004). The association of Y. lipolytica with grapes (Renouf et al 2005) and wine fermentations (Strauss et al. 2001) has been previously noted, but this is the first report of C galli on wine grapes.

It cannot be concluded from the data obtained, if C. zemplinina or C. galli occur as dead, non-viable cells or as viable but non-culturable cells. The discrimination of active versus dead, injured or VBNC populations, however, may be further improved by applying RNA templates, rather than DNA, in the PCR-DGGE strategy (Mills and Cocolin 2002).

Many of the newly described species (C. zemplinina, C. galli, Hs. opuntiae) and potentially novel species {Aureobasidium spp., Metschnikowia spp.), isolated here for the first time on wine grapes, are phenotypically or morphologically similar to their known, characterised relatives. Their detection and differentiation in this study highlights the value of molecular analyses in the identification of close physiological relatives.

Plate cultures of grapes during the warmer, drier season of 2002-2003, revealed the presence of species of Cintractia, Graphiola, Pseudozyma and Ustilago. These fungi belong to the large heterogeneous class of Ustilaginomycetes, and are ecologically well characterised by their parasitism of plants (Begerow et al. 2000). Species of Pseudozyma {Psz. flocculosa, Psz. tsukubaensis) are known to be antagonistic towards a wide range of ascomycetous and basidiomycetous fungi (Golubev 2006), and are considered promising candidates for biocontrol (Avis and Bélanger 2002). The role of these fungi on wine grapes therefore, requires further study.

Several species of filamentous fungi were detected by PCR-DGGE; Phialocephala scopiformis and Raciborskiomyces longisetosum are common endophytes of spruce (Barklund and Kowalsi 1996). DGGE bands corresponding to various species of Alternaria, Aspergillus and Phoma were recovered from grapes, but they were not isolated on DG18 or DRBC agar plates. Their detection by PCR-DGGE may reflect the presence of dead biomass. 3.4.3 Factors affecting the yeast ecology of wine grapes As discussed in Fleet ct ol. (2002) and Chapter 2 of this thesis, several intrinsic and extrinsic factors have the potential to influence the occurrence of microorganisms on wine grapes. These include grape cultivar, climate effects, region, application of agrichemicals, berry damage. It is difficult to attribute the populations and species diversity to any given factor, and it is likely that a combination of several factors are involved. A systematic, controlled investigation of these variables was beyond the time frame and scope of this study, but tentative observations and conclusions can be advanced.

The increased yeast densities observed on ageing and senescing leaves are thought to be related to changes to the surface properties of the phyllosphere, such as increased leaching of nutrients or to diminished concentrations of antimicrobial compounds (reviewed in Fonseca and Inacio 2006). As described in Chapter 2, the increased concentrations of sugars and decline in phenolic compounds on the surface of maturing grapes are likely to provide conditions conducive to yeast proliferation.

Grape berry damage has significant influence on the population and diversity of yeasts associated with grapes. Damaged berries harbour a much higher population of yeasts, and are likely to have increased association of the typical fermentative wine yeasts such as Candida, Hameniaspora, Metschnikowia, Torulaspora, Kluyveromyces and Saccharomyces species (Davenport 1976; Guerzoni and Marchetii 1987; Yanagida et al 1992; Tôrôk et al. 1996; Mortimer and Polsinelli 1999). These conclusions were more evident for grapes sampled during the season of 2001-2002, but were less supportive for the drought-affected 2002-2003 vintage, and demonstrate the importance of climatic factors. Damage to the physical integrity of grape skin can arise from insect and bird activities (Buchanan and Amos 1992), fungal infections (Donèche 1993), and can be provoked by climatic conditions such as prolonged wind and hail. No attempt was made to select grapes on the basis of damage type. The type of berry damage and the factors which initiate this damage appear to be important considerations that require more careful description and evaluation with respect to microbial ecology. Further research is needed to address these concepts as they may have an important influence on the yeast species that occur in grape juice or must, and contribute to wine fermentations. Various authors have observed specific association between yeast species with different grape cultivars (Benda 1962, 1964; Davenport 1976; Yanagida et al. 1992; Suarez et al 1994; Guerra et al 2002; Sabate et al. 2002; Rementeria et al 2003; Raspor et al 2006). Grape juice pH or acidity (Benda 1964; Deak and Beuchat 1993), grape sugar content (Rementeria et al 2003) and thickness of grape skins (Bisson and Kunkee 1991) have been considered selective factors in determining the grape microflora. Geographical or regional and vineyard effects have also been mentioned before (van der Westhuizen et al. 2000a, 2000b). However, carefully controlled studies that eliminated the influence of other factors have not been reported. In this study, key yeast species identified {A. pullulans, species of Cryptococcus, Rhodotorula, Hanseniaspora, Metschnikowia) were unrestricted to a particular variety or sampling location, so that associations between the yeast flora and varieties or vineyards examined were not particularly evident.

The year or vintage gave an obvious effect on the yeast species and population of grapes for the two vineyards (R and L) that were studied over two consecutive vintages, 2001- 2002 and 2002-2003. The 2001-2002 season was characterised by lower temperatures and higher rainfall than the 2002-2003 season, which was atypically hot and dry. Temperatures as high as 45°C were frequently recorded in the vineyards during the December-February sampling period of this drought affected season. Some general climate data for the regions where the grapes were cultivated are given in Figure 3.6, but they do not necessarily reflect the unpredictable daily fluctuations in conditions that occurred from vineyard to vineyard. Additionally, meteorological archives were not complete, and data for some climatic indicators were not available. Thus, from a microbiological perspective, only broad conclusions can be drawn about climate influences.

Climatic parameters, such as temperature, relative humidity, water availability are significant in determining the microbial populations (abundance and species richness) of the leaf surfaces (Talley et al. 2002). It is generally reported that grapes from higher rainfall vintages have higher and more diverse microbial populations (Longo et al. 1991; de la Torre et al. 1998a, 1999; Combina et al. 2005), including greater incidences of fermentative wine yeasts (Benda 1964). Data from the present investigation appear to be consistent with these findings. Population trends in this study are also supported by ecological and modeling or predictive studies that considered weather variables singly (de la Torre et al. 1998b; Rousseau and Donèche 2001; Talley et al 2002). The hot, dry conditions of the 2002-2003 season probably resulted in grapes with very low populations of yeasts (10^-10^ CFU/g) and with some exceptions, the absence of fermentative wine yeasts (Hanseniaspora, Metschnikowia, Kluyveromyces, Torulaspord) that are widely reported in literature to be found on wine grapes (Fleet et al. 2002). Yeasts were more abundant on grapes from the 2001-2002 season when conditions were cooler and there were intermittent periods of rain in the weeks prior to harvest (December-January). Typical fermentative yeasts were more frequently encountered in this season. Conversely, the studies by Jolly et al. (2003) and Rementiera et al. (2003) showed a decrease in yeast populations during cooler, rainy vintages. The application of fungicides is common practice during rainy seasons and this was thought to be a factor contributing to the reduction in yeast populations.

With the exception of vineyard F, all vineyards followed a stringent regime of pesticide applications to control a diversity of pests. The program of pesticide application for each vineyard is given in Appendix 3. Appendix 4 provides more details about the individual pesticides applied and the pests to be controlled. There are numerous suggestions in the literature that residues of pesticides can affect the survival and growth of microorganisms, either favourably (Ng et al. 2005) or adversely (Cabras et al. 1999; Cabras and Angioni 2000), the incidence of wine yeasts on grapes (Guerra et al. 1999; Regueiro et al. 1993; van der Westhuizen et al. 2000a, 2000b) and the alcoholic fermentation (Bisson 1999). However, no obvious impacts on yeast populations and species diversity of grapes from pesticide application were observed, and it is likely that a combination of factors is involved in influencing the qualitative and quantitative yeast composition of grapes. CHAPTER FOUR

THE DIVERSITY AND SIGNIFICANCE OF AUREOBASIDIUM AND RELATED SPECIES ON WINE GRAPES

4.1. INTRODUCTION

The yeast-like organism, A. pullulans and related species are prominent components of the mycoflora of wine grapes. As described in Chapter Three, these organisms are frequently recovered from both unripe and ripe berries of all grape cultivars, from both healthy and damaged berries, and in every wine-growing region investigated in New South Wales, Australia.

Several hxmdvQ^ Aureobasidium and Aureobasidium-Wke isolates were collected during this ecological survey (Chapter Three). The isolates exhibited notable diversity in colony morphology, which presented a problematic task of confidently and efficiently organising them for identification. More accurate species diagnosis of these organisms is important, given their involvement in biodeterioration, potential role in plant pathology, potential as opportunistic human pathogens, and their emerging status as allergens. Yurlova et aL (1995, 1999) have shown that colony morphology, cellular features and biochemical characteristics do not readily permit taxonomic differentiation of species within this group. Hermanides-Nijhof (1977) reported that species belonging to the closely related genus Hormonema have frequently been incorrectly assigned to the g^nus Aureobasidium. The identification of ecological isolates on the basis of cultural properties causes some uncertainty and ambiguity, and now requires assessment using genetic criteria (Uzunovic et al. 1999; Ray et al. 2004).

The first aim of this chapter is to use morphological and physiological characteristics, in combination with molecular tools, to examine the taxonomic and genetic diversity of Aureobasidium and related species, isolated from Australian wine grapes. Through this study, it is hoped to obtain more accurate and meaningful ecological understanding of their association with grapes. The second part of this study examines several biochemical and physiological properties that enable A. pullulans and related species to successfully colonize the surface of wine grapes and grow in grape musts. It also investigates some potential consequences of their growth on the quality and processing of wines. The physiological features of pullulans and relatives are important not only because they colonise many habitats (Cooke 1959; Andrews and Harris 2000), but also because they can be exploited in many areas of biotechnology (Roukas 2000; Leathers 2003a, 2003b).

The impact of the grape fungus, Botrytis cinerea on wine quality has been the subject of numerous studies. Some notable examples include its metabolism of grape organic acids (Doneche 1993), production extracellular glucans (Dubourdieu 1981; Zoecklien et al. 1995), production of laccases (Jeandet et al 2002), and secretion of antimicrobial "Botryticine" (Ribereau-Gayon et al 2000). Few studies have investigated the analogous effects for A. pullulans, despite its frequent association with grapes and musts.

This chapter describes properties that are expected to determine the ability of Aureobasidium and Aureobasidium-like isolates to persist in wines. These traits include growth at low pH, tolerance to ethanol and SO2. Metabolism of organic acids, interactions with other microbial species, elaboration of exopolysaccharides and extracellular enzymes are properties of these organisms that could significantly impact on wine production, and are also reported here.

4.2 MATERIALS AND METHODS

4.2.1 Cultures

Aureobasidium and related species examined in this Chapter were isolated during an ecological survey of wine grapes in New South Wales, as described in Chapter Three. The strains studied, and the grape samples from which they were obtained are listed in Table 1. Reference strains of^. pullulans, CBS 584.75 and FRR 4800 were obtained from the Centraalbureau voor Schimmelcultures (CBS), Utrecht, the Netherlands, and Food Science Australia (FRR), North Ryde, Australia, respectively. Filamentous fungi and yeasts listed in Tables 4.7 and 4.8 as FRR or AWRI strains were obtained, respectively from Food Science Australia, CSIRO, North Ryde, NSW, and the Australian Wine Research Institute, Adelaide, South Australia. They were subcultured and maintained on MEA (Oxoid, Hampshire, England). Yeast and bacterial cultures listed in Tables 4.7-4.9 as UNSW strains were obtained from the culture collection of the School of Biotechnology and Biomolecular Sciences, UNSW. Bacterial cultures were subcultured and maintained on MRS agar (Oxoid) for lactic acid bacteria. Some cultures were isolated and purified from commercial industrial preparations (eg. S. cerevisiae Maurivin, Primeur; Bacillus thuringiensis Delfm, Dipel; Oenococcus oeni Oeni, CH35).

4.2.2 DNA extraction for PGR amplification Genomic DNA was extracted from pure cultures for use in PCR reactions for ITS- RFLP, microsatellite fingerprinting and sequence analysis. The extraction method followed that described by Cocolin etal (2002). Microbial cell pellets were prepared from 24 h cultures in Malt Extract Broth (1 ml, centrifuged at 12,000 g, 10 min, 4°C), and resuspended in 200 ^il of breaking buffer (1% SDS, 2% Triton X-100, 100 mM NaCl, 10 mM Tris, 1 mM EDTA, pH 8.0) containing 0.3 g glass beads (0.5 mm in diameter, BioSpec, Bartlesville, OK). Cells were homogenised at 6,000 rpm for 1 min in a bead beater (Mini-BeadBeater-F^^, BioSpec) in the presence of 200 ^L of phenol/chloroform/iso-amylalcohol (25:24:1). Two hundred microliters of TE buffer (10 mM Tris, 1 mM EDTA, pH 8.0) were added, and the mixture was centrifuged at 12,000 g for 10 min at 4°C. The aqueous phase was collected and DNA was precipitated by the addition of 2.5 volumes of absolute ethanol and recovered by centrifugation at 12,000 g for 10 min at 4°C. The DNA pellet was washed with 70% ethanol, dried and resuspended in 50 of TE buffer. Table 4.1 Origin of Aureobasidium and related strains used in this study

Isolate/ Viiix vinifera, cv. Grape maturity, physiological status Geographical region* strain no. 16 Chardonnay Infloresence Hunter Valley 28 Shiraz Infloresence Hunter Valley 30 Shiraz Infloresence Hunter Valley 33 Cabernet sauvignon Infloresence Hunter Valley 36 Chardonnay Infloresence Hunter Valley El Semiilon Infloresence Hunter Valley E2 Semillon Infloresence Hunter Valley 66 Chardonnay Pea-sized berries Hiuiter Valley E12 Semillon Pea-sized berries Hunter Valley 101 Chardonnay Pea-sized berries Hunter Valley 126 Shiraz Pea-sized berries Hunter Valley 134 Shiraz Veraison Hunter Valley E20 Cabernet sauvignon Veraison Hunter Valley E21 Cabernet sauvignon Veraison Hunter Valley 167 Cabernet sauvignon Berries intermediate brix Hunter Valley 169 Cabernet sauvignon Berries intermediate brix Hunter Valley 174 Cabernet sauvignon Berries intemiediate brix Hunter Valley E26 Cabernet sauvignon Berries intermediate brix Hunter Valley E31 Semillon Berries intermediate brix, damaged Hunter Valley E33 Chardonnay Berries intermediate brix Hunter Valley E38 Shiraz Berries intermediate brix Hunter Valley E39 Shiraz Berries intermediate brix Hunter Valley 201 Chardonnay Ripe Hunter Valley 202 Chardonnay Ripe Hunter Valley 207 Chardonnay Ripe Hunter Valley E49 Semillon Ripe Hunter Valley 262 Chardonnay Harvest ripe Hunter Valley 263 Shiraz Harvest ripe Hunter Valley 264 Semillon Harvest ripe Hunter Valley 268 Cabernet sauvignon Harvest ripe Hunter Valley 283 Shiraz Harvest ripe Hunter Valley 293 Shiraz Harvest ripe Hunter Valley 295 Semillon Harvest ripe Hunter Valley 496 Semillon Harvest ripe Hunter Valley E262 Semillon Harvest ripe Hunter Valley E97 Cabernet sauvignon Harvest ripe, damaged Mudgee E99 Cabernet sauvignon Harvest ripe, damaged Mudgee ElOO Cabemet sauvignon Harvest ripe, damaged Mudgee 313 Shiraz Harvest ripe Mudgee E160 Cabemet sauvignon Harvest ripe Mudgee E185 Cabemet sauvignon Harvest ripe Mudgee 311 Merlot Harvest ripe Mudgee 314 Merlot Harvest ripe Mudgee 315 Chardonnay Harvest ripe, damaged Mudgee 316 Chardonnay Harvest ripe, damaged Mudgee 335 Merlot Harvest ripe, damaged Mudgee 550 Cabemet sauvignon Harvest ripe, damaged Mudgee 437 Sauvignon blanc Harvest ripe GrifTith E274 Tyrian Harvest ripe Gnffith E278 Semillon Harvest ripe Griffith 472 Tyrian Harvest ripe Griffith 475 Cabemet sauvignon Harvest ripe Gnffith E303 Cabemet sauvignon Harvest ripe Griffith E304 Cabemet sauvignon Harvest ripe Griffith E305 Semillon Harvest ripe Griffith E343 Semillon Harvest ripe, damaged Griffith 362 Cabemet frane Harvest ripe Monaro E238 Cabemet frane Harvest ripe Monaro 374 Merlot Harvest ripe Monaro 365 Traminer Harvest ripe Monaro 367 Trami ner Harvest ripe Monaro CBS 584.75" ? France FRR 4800" Cordo Raisin Mildura Within New South Wales, Australia ^ Reference strains of Aureobasidiumpullulans 4.2.3 rDNA amplification The D1/D2 domain of the 26S rDNA was amplified using the primers NLl (5'- GCATATCAATAAGCGGAGGAAAAG-3') andNL4 (5'- GGTCCGTGTTTCAAGACGG-3') (Kurtzman and Robnett, 1998). PGR reactions were carried out in 50 final volumes containing Ix PGR Buffer (10 mM Tris-Cl, 50 mM KCl), 0.2 }iM of each primer, 200 ^iM of each dNTP (Roche Diagnostics, Indianapolis, IN), 1.5 mM MgCb, 1.25U Gold Taq DNA polymerase (AmpliTaq^^«, Roche Molecular Systems, Brachburg, NJ) and 10 ng of DNA template. Amplifications were done in a 9600 Thermal Cycler (Applied Biosystems) with the following temperature program: initial denaturation at 95°C for 7 min, 36 cycles of 95°C for 1 min (denaturation), 52°C for 2 min (annealing) and 72°C for 2 min (extension), followed by a final extension for 10 min at 72°C.

The ITS (ITSl and ITS2) and 5.8S rDNA gene regions were amplified with primers ITSl and ITS4 (White et al. 1990). PGR reaction mixtures were the same as described above. Cycling conditions were as follows: initial denaturation at 95®C for 7 min, 36 cycles of 95°C for 1 min (denaturation), 55°C for 2 min (annealing) and 72°C for 2 min (extension), followed by a final extension for 10 min at 72°C.

All primers used in this study were synthesized by Sigma-Proligo (Proligo Primers & Probes, NSW, Australia).

4.2.4 PCR-based RFLP of the ITS regions PGR amplification of the ITS-5.8S rDNA region was performed as described previously. Restriction patterns of the PGR amplicons were obtained by cleaving with each of the following seven endonucleases; AM, Cfo\, Dde\, Hae\\\, Hinfl, Rsa\ and Taq\ (Promega, Madison, WI). Aliquots (10-15 ^il) of PGR products were digested with 5U of each restriction enzyme separately, according to the manufacturer's recommendations. Restriction fragments were separated on a 3% agarose gel and stained with ethidium bromide. A 100 bp DNA ladder (Bio-Rad, Hercules, CA) was used as a molecular weight marker. The REBASE program (http://rebase.neb.com) was used to identify potential restriction sites and estimate expected fragment sizes from available sequences (Roberts et al 2005). Restriction fragments for each enzyme were visually scored as present (1) or absent (0), and compiled in a binary matrix. Dice coefficients of similarity were calculated with NTSYS-PC software version 2.2 (Exeter Software, Setauket, New York). Using the similarity matrix, cluster analysis was performed with UPGMA to generate a dendogram.

4.2.5 rDNA sequencing Sequence analysis of the partial 26S rDNA and ITS amplicons was performed using the ABI PRISM® Big Dye^^* Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA), according to the manufacturer's directions. Primers used for sequencing both strands of the D1/D2 domain and ITS regions were the same as used for PCR amplification. DNA fragments were sequenced at the Automated DNA Analysis Facility, School of Biotechnology and Biomolecular Sciences, UNSW. DNA similarity searches were performed with the NCBI BLAST program using sequences retrieved from the Genbank database (Altschul et al. 1990). Sequences for the D1/D2 region of the A. pullulans neo type reference strain CBS 584.75 are deposited in Genbank (accession number DQ321374).

4.2.6 Microsatellite PCR fingerprinting In microsatellite-primed PCR amplifications, the primers M13 (GAG GGT GGC GGT TCT), M13 phage core wild sequence (GAG GGT GGX GGX TCT), (GACA)4 and (GTG)5 were used. PCR was performed using a Corbett Research PC-960 air cooled thermal cycler (Corbett Research, NSW, Australia) in 25 volumes containing IX PCR Buffer (10 mM Tris-Cl, 50 mM KCl), 1.0 ^M of primer, 200 ^M of each dNTP (Roche Diagnostics, IN), 2.0 mM MgCl2, 1.25U Gold Taq DNA polymerase (AmpliTaq'^'^, Roche Molecular Systems, Brachburg, NJ) and 25 ng of DNA template. Forty amplification cycles were performed with M13 primers using the following program: 1 min at 94°C, 2 min at 52°C, 3 min at 74°C, with and a final extension step of 10 min at 74°C. The thermal cycler was programmed for 40 cycles of 1 min at 95°C, 2 min at 48°C for primer (GACA)4 or 50°C for primer (GTG)5, and 2 min at 72T, with an initial denaturation step of 7 min at 95°C and a final extension step of 10 min at 72°C. To minimize inconsistencies, PGR reactions for all samples were prepared from a single batch of master mix and processed in one thermocycler run. Negative controls (no template DNA) were included for each amplification to test for non-specific reactions.

PGR products were separated by electrophoresis in 1.4% agarose gels, in 0.5X TBE (Tris-borate-EDTA) buffer at 80 V for 3.5 h and stained with ethidium bromide. DRIgestTMlll, a Hindm digest oiX DNA with HaeXW digest of^Xm (Amersham Biosciences, NJ) was used as a molecular size marker.

Fingerprint patterns were analysed with GelCompar 4.0 (Applied Maths, Kotrijk, Belgium). Similarities between fingerprints were calculated using the Dice coefficient, and a dendogram was constructed using the UPGMA algorithm. Profiles obtained from each primer were combined in a composite fingerprint and co-phenetic correlation coefficients were calculated to evaluate how accurately the dendograms represented the estimates of similarity among the genotypes.

Genetic stability and reproducibility of PGR patterns were assessed using three repeats of six arbitrarily selected strains. Each set of replication consisted of three separate subcultures of each strain. Extraction of genomic DNA and PGR amplifications were prepared in independent assays from each set of isolations. PGR amplicons were resolved on a single gel. The reproducibility of PGR patterns for each primer was estimated using Sneath and Johnson's formula for probability of error, p = 1- Vl-4s^, where (s^) was calculated as the proportion of unmatched amplified bands to the total number of bands obtained in the set of triplicate repeats (Sampaio et al. 2001).

4.2.7 Physiological characterisation Aureobasidium and related isolates, as well as reference strains, were characterised using selected phenotypic features described by de Hoog and Yuriova (1994). Physiological properties were determined by standard yeast classification methods (van der Walt and Yarrow 1984; Yarrow 1998). Cultures (24-48 h) grown on MEA plates were suspended in quarter strength Ringer solution (Oxoid, Hampshire, England) to a density of 10^ cells/ml, estimated with a turbidity meter (Bio Mérieux Vitek, Australia). The suspension was used to inoculate tests in liquid media for assimilation of: D-galactose, D-xylose, D-ribose, lactose, L- sorbose, DL-lactate, citrate, myoinositol, galactitol, D-mannitol, methyl a-D-glucoside, D-glucosamine hydrochloride, D-glucuronic acid and inulin as sole carbon source; growth at 5°C, 10°C, 30°C, 34°C and 37°C; growth in 10% NaCl, 5% glucose plus 10% NaCl and in 5% glucose plus 16% NaCl; fermentation of glucose and fructose. The basal medium used for these growth tests was Yeast Nitrogen Base (Difco, Detroit, MI).

Hydrolysis of urea was detected on Christensen's urease agar (Yarrow 1998). The medium contained (per liter); 24 g Christensen's urea agar base (Oxoid) and 50 ml sterile 40% (w/v) urea solution. A positive result is evidenced by the development of a deep pink colour in the test medium. The tests for casein and gelatin hydrolysis were performed as described in a later section. The biochemical and physiological profiles (27 characters) were converted into a binary matrix and further processed with NTSYS- pc software, version 2.2 (Exeter Software, Setauket, New York). Similarity between isolates was estimated with the simple-matching coefficient and used in cluster analysis with UPGMA.

4.2.8 Utilisation of malic and tartaric acid Assimilation of L-malic acid or L-tartaric acid was tested in YNB medium according to van der Walt and Yarrow (1984). The pH of the media was adjusted to 5.6 with 5N NaOH. The inoculum cultures were prepared as described in the previous section, under physiological characterisation 4.2.7.

4.2.9 pH, ethanol and SO2 tolerance Growth of isolates was examined in YNB (Difco, Detroit, MO) + 5% glucose at pH 3.0, 3.5 and 4.0. Citrate phosphate buffer was used to adjust the basal medium to the required pH. The effect of ethanol and sulfur dioxide on growth was examined in YNB + 5% glucose at pH 3.5. Ethanol was added to the basal medium after sterilization to give final concentrations of 2, 4, 6 and 8% (v/v). Sulfur dioxide at concentrations of 50, 100, 150 and 200 ppm were prepared by the addition of 1, 2, 3 and 4 ml respectively, of a stock solution of sodium metabisulphite (16.7 g/1) into 200 ml of basal medium, after sterilization.

Growth tests were performed in triplicate wells of microtitre plates (Sarstedt, Technology Park, SA, Australia). Test media (245 fil) were dispensed into each well, and were inoculated with 5 ¡il of cell suspension (5.0 x 10^ cells/ml). Microtitre plates were sealed and incubated at 25°C and examined for the presence or absence of turbidity in plate wells at weekly intervals for 3 weeks.

4.2.10 Exopolysaccharide production Aureobasidium and related isolates were screened for mucoid phenotypes on ME A plates. Ten isolates with slimy, spreading or mucoid colony appearance and ten isolates with dry, matt or non-mucoid, phenotypes were selected for this study.

Yeast Nitrogen Base, modified for wine fermentations (WYNB; 6.7 g/1 YNB, 100 g/1 glucose, 100 g/1 fructose, pH 3.5) was used to screen for exopolysaccharide production. Inoculum cultures were prepared by growing isolates in YEPD broth at 25°C for 24 h with shaking (120 rpm). Flasks containing 100 ml of the basal medium were inoculated with 1% (v/v) of the inoculum cultures and fermentation was carried out at 25°C for 5 days without shaking. Cells were separated from the extracellular material by centrifugation for 30 min at 10 000 x g at 4°C. The biomass was determined by washing cells with distilled water twice, then dried at 90°C until the weight remained constant.

Polysaccharide instability Ten ml of WYNB culture filtrates were subject to glucan instability tests for juice and wine samples (Dubourdieu and Ribéreau-Gayon 1981; Zoecklein et al. 1995) by precipitation with 5 ml of 96% (v/v) ethanol. The presence of glucans at concentrations above 15 mg/1 is indicated by the formation of white filaments or gelatinous mass.

Exopolymer collection and analysis Crude polymers were precipitated from the remainder of the WYNB supernatant (approx. 80 ml) by the addition of two volumes of 96% (v/v) ethanol and held at 4°C for 24 h. The polymers were collected by centrifugation for 10 min at 14 000 x g at 4°C. The resulting pellet was washed in acetone and ether, then dried under vacuum and weighed. Crude products were dissolved in distilled water to a concentration of 0.1% (w/v). Total carbohydrate amount in the polymer was estimated by the phenol-sulfuric acid method (Dubois et al. 1956), by the addition of 4% (w/v) phenol and 96% (v/v) sulfuric acid. Optical density of the solution was measured with a UV spectrophotometer (Shimadzu, Australia) at 490 nm. D-glucose was used to generate a calibration curve.

4.2.11 Interactions between Aureobasidium and related isolates, and other microorganisms Isolates were characterized for their interaction with other yeasts, filamentous fungi and bacterial species associated with grapes and wine. Microbial interactions were assayed in-vitro, using the spot on lawn technique (Foldes et al. 2000; Addis et al 2001) and in some cases, using the deferred culture assay (Bae et al 2004). The appropriate basal medium was chosen for the sensitive strain under investigation. Interactions with yeasts and acetic acid bacteria were tested on YEPD (2% glucose, 1% peptone, 1 % yeast extract 2% agar), buffered to pH 4.5 with 0.1 M citrate phosphate buffer. Lactic acid bacteria were inoculated in MRS agar (Oxoid). Other bacterial species and filamentous fungi were seeded in PDA (Oxoid). For the spot on lawn assay, plates were prepared by seeding 10^ cells/ml of freshly cultured test strain to the molten basal medium, mixed and allowed to solidify. After allowing plates to dry, Aureobasidium cultures (48-72 h) were spot inoculated onto the surface of the seeded agar. For the deferred culture assay, Aureobasidium spot cultures grown on ME A plates (96 h) were overlaid with 5 ml of the appropriate molten agar that was seeded with approximately 10^ cells/ml of the test species. Inoculated plates were incubated at 25°C for 4-7 days (yeast and fungi) or at 30°C for 2-3 days (bacteria).

Inhibition of the test organism was recognised by clear zones surrounding the growth of spot cultures. Stimulated or enhanced growth of the seeded organism was indicated by increased biomass density surrounding the spot culture. The S. cerevisiae starter SB-1 (Lallemand, Montreal, Canada) and S. cerevisiae strains 3002 and 516 (Heard and Fleet 1987) were included as indicators of sensitive and inhibitory phenotypes, respectively. 4.2.12 Growth of Aureobasidium-Wke organisms with S. cerevisiae

Two strongly antagonistic ^wreo^a^/û^/wm-like isolates and two commercial strains of wine S. cerevisiae were selected for these studies. Single and mixed culture of were conducted in two media; a chemically defined grape juice medium, and YNB-glucose medium (YNB, Difco Laboratories, Detroit, MI, with 5% glucose, pH 5.5). The chemically defined medium was prepared according to Henschke and Jiranek (1993). The medium contained (per liter); 100 g glucose, 100 g fructose, 2.5 g KH tartarate, 3.0 g L-malic acid, 0.2 g citric acid, 1.14 g K2HPO4, 1.23 g MgS04.7H20, 0.44 g CaCl2.2H20, 200 ^ig MnCl2.4H20, 135 \ig ZnCh, 30 ng FeCb, 15 ^g CuCb, 5 \ig H3B03, 30 fig Co(N03)2.6H20, 25 ng NaMo04.2H20, 10 ^g KIO3, 500 mg (NH4)2HP04, 100 mg L-alanine, 750 mg L-arginine, 150 mg L-asparagine, 350 mg L- aspartic acid, 500 mg L-glutamic acid, 50 mg L-glycine, 150 mg L-histidine, 200 mg L- isoleucine, 300 mg L-leucine, 250 mg L-lysine, 150 mg L-methionine, 150 mg L- phenylalanine, 500 mg L-proline, 400 mg L-serine, 350 mg threonine, 100 mg L- tryptophan, 20 mg L-tyrosine, 200 mg L-valine, 100 mg myo-inositol, 2 mg pyridoxine.HCl, 2 mg nicotinic acid, 1 mg panthothenate.Ca, 0.5 mg thiamine. HCl, 0.2 mg p-aminobenzoic acid, 0.2 mg riboflavin, 0.12 5mg biotin and 0.2 mg folic acid. The lipid component comprised (per liter); 10 mg ergosterol dissolved in 1 ml Tween 80 and ethanol (50:50, v/v), which was added to the filter sterilized medium.

Inoculum cultures for these experiments were prepared by growing them in MEB broth with agitation (120 rpm) at 25°C for 24 h. Defined grape juice and YNB media (80 ml) were inoculated with 1.0 X 10"^ cells/ml of the relevant species. Fermentation flasks were fitted with air locks and incubated at 25°C for 10 days without agitation.

Samples were aseptically removed from fermentations at various intervals for çnumQratïng Aureobasidium-WkQ and S. cerevisiae. The enumeration of Aureobasidium- like populations in bicultures was performed on Lysine agar, (66 g/1 lysine medium, Oxoid Melbourne; supplemented with 10 ml/1 potassium lactate and 1 ml/1 10% lactic acid, pH 4.8), and S. cerevisiae cell densities were estimated on MEA plates. 4.2.13 Screening of extracellular enzymes Aureobasidium and related isolates were grown on MEA plates at 25°C for 48 h and inoculated onto the test media containing suitable substrates. Unless otherwise indicated, all tests were performed on solid media. Plates were incubated at 25°C in the dark and noted for reactivity at 3 and 5 days after inoculation. Zones of clearing from the colony edge, or pigmentation of media where colourimetric reagents were used, were taken as evidence of enzyme activity.

Protease activity Casein hydrolysis was tested on medium containing 20 g casein, 3 g Lab Lemco powder (Oxoid), 5 g tryptone, 15 g agar and 1 g glucose per liter (Aheam et al. 1968). Gelatin liquefaction was detected on Wickerham's medium (per liter; 100 g gelatin, 5 g glucose, 6.7 g YNB). Inoculated gelatin plates were incubated at 18°C and were observed for evidence of liquefaction, after the culture was cooled to 5°C (Yarrow 1998).

Lipolytic activity Tween 20® (sorbitan monolaurate) and Tween 80® (sorbitan monooleate) were used as lipid substrates to detect the production of lipases, as described by Sierra, in Hankin and Anagnostakis (1975). The medium contained (per liter); 10 g peptone, 5 g NaCl, 0.1 g CaCl2.2H20, 20 g agar and 10 ml of Tween ester. A positive result was indicated by the formation of calcium salt crystals of the liberated lauric- or oleic acid around the colonies. In some cases, hydrolysis of Tween 80 appeared as zones of clearing at the periphery of colonies as the salt of the fatty acid is degraded.

Degradation of polysaccharides Degradation of various polysaccharides was tested using modified Melin-Norkrans medium (MMN, 1.0 g glucose, 2.0 g malt extract, 1.0 g yeast extract, 0.5 g KH2PO4, 0.25 g (NH4)2HP04, 0.15 g MgS04.7H20, 0.05 g CaCb, 0.025 g NaCl, 0.012 g FeCl3.6H20, 15.0 g agar and 1 1 dH20) supplemented with suitable substrates. Pectate lyase (PL) and polygalacturonase (PG) screening media each contained 5 g/1 of citrus pectin in MMN, buffered to pH 7.0 and pH 5.0, respectively. Plates were prepared in the manner described by Hankin and Anagnostakis (1975). PG and PL activity were detected on these media by the addition of 1% (w/v) cetyltrimethyl ammonium bromide (CTAB), which precipitated the undegraded substrate. Hemicellulase and cellulase activities were measured using birchwood xylan (5 g/1) and carboxy methyl cellulose (CMC; 5 g/1), respectively, as substrates. For both media, clearing zones were visualised by staining with 1 mg/ml Congo Red for 30 min, then destained with 1 M NaCl for 15 min. A second test for cellulose degradation was the filter paper test, using a strip of Whatman No. 1 filter paper (1.0 x 6.0 cm, 50 mg) as the substrate, submerged in MMN liquid medium. The cellulase-producing strain, Cellulomonas sp. CSl-1 (UNSW 068900) was used as a positive control for CMC and filter paper degradation. p-glucosidase activity p-glucosidase activity was tested using arbutin (Yarrow 1998) and esculin (Hernández e/ al. 2002) as substrates. The test media contained (per liter); 10 g yeast extract, 20 g agar, 5 g arbutin or esculin, and 20 ml of 1 % (w/v) sterile ferric ammonium citrate. Hydrolysis of the substrate results in the development of a dark brown colour around the colony.

Cutinase activity Cutinase activity was detected using polycaprolactone (PCL) and p-nitrophenyl butyrate (PNB) as substrates. Screening plates using PCL were adapted from Murphy et al. (1996) and Sanchez et al (2000); PCL suspension was prepared by dissolving 1 g polycaprolactone pellets (molecular weight 14000, Sigma Aldrich) in 50 ml of acetone. The suspension was emulsified in 1 1 of MMN with the addition of 100 mg surfactant Plysurf A210G (Daiichi Kogyo Seiyaku, Japan). Acetone was removed from the emulsion by rotary evaporation at 50°C and mixed with 20 g of agar before autoclaving. Cultures were examined periodically over 4 weeks. Degradation of PCL is observed as clearing zones against an opaque background.

A second assay using PNB followed the screening procedure devised by Dantzig et al (1986) and Middelhoven (1997). Isolates were grown as spot cultures on plates of MMN containing 0.5% acetate, 0.011% Triton X-100 and 1.5% agarose for 5 d. For isolates that failed to grow on acetate as a sole source of carbon, acetate was replaced with glucose. Plates were overlaid with 1.26 mM PNB substrate (Sigma) in 1.5% agarose containing 0.011% Triton X-100. Isolates which stained bright yellow within 15-30 min of overlay were taken as positive for cutinase activity. Cutinase activity was demonstrated in the yeasts Cryptococcus laurentii and Debaromyces hansenii in studies by Jarret et al. (1985) and Middelhoven (1997), respectively. Grape isolates of both species {D. hansenii E282, E285; Cr. laurentii 430, 489) were also included in the assays as positive controls.

Polyphenol oxidase activity The formation of polyphenol oxidases (catechol oxidase, laccase, monophenol monooxygenase), was collectively screened using the Bavendamm substrates, which detect the oxidative polymerization or degradation of phenolic acids. Staining reaction on tannic acid medium and gallic acid medium were observed using MEA and MMN agar supplemented with 0.5% (w/v) tannic acid or 0.5% (w/v) gallic acid, respectively. Cultures were incubated for up to 4 weeks. The ability to degrade tannic acid and gallic acid was recognized by the formation of a dark brown pigment surrounding the colony (Bending and Read 1997).

Tests for laccases using guaiacol as a substrate was performed according to Thorn et al. (1998). Guaiacol was added to MEA and MMN agar at 0.01% (v/v). A positive reaction is indicated by red-brown colour development in the medium. The assay with a- naphthol (0.1 M in 95% ethanol; 1 drop) was applied to 4 week old cultures on MEA and MMN agar. Development of a blue to violet colour on the biomass, after 1-3 h incubation at room temperature indicated a positive reaction.

Lignin degradation was tested by the clearing of a model lignin compound, polymeric dye Poly R-478, which was incorporated into MEA and MMN agar at 0.04% (w/v) {Thorn etal 1998). PART 1. MOLECULAR, MORPHOLOGICAL AND BIOCHEMICAL CHARACTERISATION OF AUREOBASIDIUM AND RELATED ISOLATES

4.3 RESULTS The ecological survey of wine grapes reported in Chapter Three yielded a predominance of Aureobasidium-likQ organisms. These isolates (about 200) exhibited significant colonial and cellular morphology on MEA plates. Moreover, these properties changed with culture age on MEA, adding to the difficulty of their identification. To facilitate the classification, identification and characterisation of these isolates, a preliminary process of selection and molecular screening was undertaken. Approximately 30% of the isolates (61 strains) were chosen on the basis of morphological and ecological criteria. The ecological criteria included grape variety, grape maturiy, and vineyard location (Table 4.1). These isolates were "fingerprinted" and grouped on the basis of ITS-RFLP profiles, and then identified to genus and species levels according to sequence and morphological data, as supported by biochemical analyses. Subsequently, all isolates were grouped and identified by their RFLP profile.

4.3.1 Grouping isolates based on PCR-ITS RFLP patterns DNA extracted from the isolates was amplified by PGR using the primers ITS 1 and ITS4. A single amplicon of approximately 580 bp was obtained and cleaved with seven restriction enzymes. The enzymes, Cfol, Ddel, HaelU and Hinjl were selected because they are widely used for the differentiation of yeasts in food and wines (Granchi et al. 1999; Esteve-Zarsoso et al 1999). Additionally, enzymes AM, Rsa\ and Taq\ were chosen because they have been used previously to to differentiate Aureobasidium-VikQ organisms (Yurlova et al 1996; Matteson-Heidenreich et al. 1997). Figure 4.1 shows an example of the banding patterns obtained with the digest of ITS amplicons with four restriction enzymes. Among the 61 isolates and two reference strains, restriction digests with C/oI, HaeWl and Taq\ each gave rise to six profiles. Five profiles were obtained from digests with Dde\ or Hinjl, while the enzymes AM and Rsal produced three patterns each. Table 4.2 shows the profiles of the reference strains of^. pullulans as well as other representative strains. The restriction patterns of many isolates did not match those of the reference strains of^. pulMans, or other yeast-like fungal species reported in literature (Yurlova et al 1996; Matteson-Heidenreich et al 1997; Sabate et al. 2002). Additionally, the RFLP groupings produced with one restriction enzyme did not always correspond to groups generated using other enzymes (Table 4.2). In some cases, the two reference strains of A. pullulans CBS 584.75 and FRR 4800 gave different profiles with the one enzyme {Ddel, Table 4.2 and Figure 4.IB). To depict the relationship between the isolates and reference strains, RPLP data derived from all seven enzymes (64 characters) were used to construct a similarity matrix and phenogram (Figure 4.2). At an arbitary level of 70% similarity, the ITS-RFLP phenogram resulted in clear separation of the strains into two distinct clusters, which are designated as groups 1 and 2. The two reference strains of A. pullulans and the majority of yeEist-like isolates (44) were contained within group 1. This A. pullulans group could be further divided into three subclusters (groups la, lb, Ic), presumably representing several populations within this species. Fourteen isolates resided in group 2, and three isolates (E262, 550, 28) were unclustered at the level of 70% similarity (Figure 4.2). 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 M A

B

w. tmr mm- •ii-' -itr- itm 4m J»»-*» . -mm •'if*^-

D

mmmmrnmrn

Figure 4.1 Banding patterns produced by restriction enzyme digests of the PCR-amplified ribosomal ITS1-5.8S-ITS2 DNA with (A) C/oI, (B) Ddel, (C) Hinji, and (D) Rsal. Lanes labeled M are molecular size markers; lanes 1-14, patterns for isolates 315, 316, 335, 437, E274, E278, 472, 475, E303, E304, E343, E262, 496, 550; lanes 15-16, reference cultures of A. pullulans strains CBS 584.75 and FRR 4800 Table 4.2 Restriction profiles of the Aureobasidium-Wkt organisms isolated from wine grapes

Representative Fragment length(s) (bp) of the 5.8S-ITS PGR amplicons after restriction analysis with: isolate Alu\ Cfo\ Ddel Haelll Hinjl Rsal Taql CBS 584.75 190, 390 50, 80, 170, 180' 50, 130, 190,210 140,440' 130, 170, 280' 580 50, 110, 130

FRR4800 190, 390 50, 80, 170, 180 50, 130, 400 140,440 130, 170, 280 580 50, 110, 130

El, 475 190, 390 50, 80, 160, 290 50, 130, 190,210 140,440 130, 170, 280 580 50, 110, 230

28 50, 150, 390 50, 90, 160, 295 50, 90, 140, 300 130, 460 100, 190, 300 590 60 110, 240

E160 580 280, 320 180, 400 120, 460 280,310 580 50, 230, 250

E12 580 140, 180, 280 180, 400 580 280, 310 580 50, 230, 250

E274 580 140, 180, 280 180, 400 120, 460 280,310 580 50, 120, 140, 220

ElOO 580 280, 320 180, 400 120, 460 280,310 580 50, 120, 140, 220

550 190, 390 280, 320 180, 400 130, 160, 290 140, 160, 280 190, 390 50, 230, 250

E262 190, 390 40, 90, 140, 180 60, 120, 400 50, 80, 450 290 100, 480 50, 240 Sabate et al. (2002) 16 la 36 El 293 475 33 66 101 496 E343 E303 E304 E305 335 437 314 374 E238 362 CBS584.75 316 313 367 295 126 283 268 264 263 262 E49 207 202 201 E39 E38 E33 365 311 174 169 167 315 472 J 134 'FRR4800 — 28 * J 30 E160 K E12 E2 E20 E21 E278 . E274 E26 E31 E97 E99 E185 ElOO 550 * ^E262 * T—r T" I ' T" 0.11 0.33 0.78 1.00 0.55

Similarity

Figure 4.2 Dendogram showing the relationship dimongAureobasidium and related isolates based on the UPGMA cluster analysis of ITS restriction patterns. Outgroup isolates are marked with asterisks. 4.3.2 Identification of Aureobasidium and related species from grapes by rDNA sequencing

To verify the species identity, representatives from each ITS-RFLP group were

sequenced in the D1/D2 and ITS regions. Reference cuhures of A. pullulans (CBS

584.75, FRR 4800) were also included in sequence analyses. Table 4.3 lists the isolates

selected for sequence analysis and species identities resulting from the BLAST search.

D1/D2 domain, 26S rDNA Comparison of the D1/D2 sequences for representatives of all RFLP patterns resulted in highest homologies with Aureobasidium pullulans and to the loculoascomycetous fungus, Discosphaerina fagi, as a described species. Isolates from group 1 (Figure 4.2) shared 97-100% homology with these species and could be confidently assigned to the species A. pullulans. Species identification of isolates in group 2 and the remaining unclustered isolates was difficult because of the lack of fungal sequences in GenBank. Strains from group 2 and the three unclustered isolates produced sequence homologies of 96% or lower to A. pullulans or to Dis.fagi. Blast searches associated these isolates with several fungal species belonging to the Dothideales order.

ITS1-5.8S-ITS2 Isolates belonging to group 1 matched 97-100% identity with^. pullulans in the ITS region, and confirm results of the D1/D2 sequence analysis. The species identity of the isolates in group 2 and the remaining isolates is uncertain, but Blast results show their association with several ascomycetous or anamorphic fungi within the Dothideales. Isolates from group 2 shared high sequence homologies to Selenophoma eucalypti or its teleomorphic counterpart, Sphaerulina eucalypti, originally described from leaf spots on Eucalyptus sp. (Crous et al 1995, 2003). Isolate 550 matched significantly to an undescribed Hormonema species. Isolates 28 and E262 were associated to the genera Kabatiella 2ind Aureobasidium, respectively. Table 4.3 Comparison of the D1/D2 and ITS sequences of Aureobasidium and related isolates with homologous sequences in GenBank

Group Isolate Fungal species, GenBank reference and identity (%) D1/D2 26S rRNA ITS1-5.8S-ITS2

la El Aureobasidium pullulans CBS 584.75 DQ321374 97 Aureobasidium pullulans CBS 584.75 AJ244232 97-98 475 Discosphaerinafagi CBS 171.93 AYO16359

lb 367 Aureobasidium pullulans CBS 584.75 DQ321374 100 Aureobasidium pullulans CBS 584.75 AJ244232 100 496 Discosphaerina fagi CBS 171.93 AYO 16359

Ic 134 Aureobasidium pullulans CBS 584.75 DQ321374 97 Aureobasidium pullulans CBS 584.75 AJ244232 97 Discosphaerinafagi CBS 171.93 AY016359

2 E12 Aureobasidium pullulans CBS 584.75 DQ321374 95-96 Sphaerulina eucalypti STE-U 5247 AY293060 98-99 E160 Discosphaerinafagi CBS 171.93 AY016359 Selenophoma eucalypti STE-U 659 AY293059 ElOO Sydowia polyspora AFTOL-178 AY544675 94 P rings he im ia smilacis CBS 873.71 AJ244257 95 E274 Dothiora cannabinae AFTOL-1359 DQ470984 94 Dothichizapityophila CBS 215.50 AJ244242 94

Aureobasidium pullulans CBS 584.75 DQ321374 95 Kabatiella caulivora CBS 242.64 AJ244251 94 28 Discosphaerinafagi CBS 171.93 AY016359 Dothichiza pityophila CBS 215.50 AJ244242 92 Sydowia polyspora AFTOL-178 AY544675 93 Pringsheimia smilacis CBS 873.71 AJ244257 94 Dothiora cannabinae AFTOL-1359 DQ470984 93 Aureobasidium pullulans CBS 584.75 AJ244232 93

Aureobasidium pullulans CBS 584.75 DQ321374 93 Uncultured ascomycete dfmo0723_052 AY969705 98 550 Discosphaerinafagi CBS 171.93 AY016359 Hormonema sp. F-054,258 AF182378 98 Sydowia polyspora AFTOL-178 AY544675 91 Dothidea insculptia CBS 189.58 AF027764 96 Dothiora cannabinae AFTOL-1359 DQ470984 91 Dothidea hippophaeos CBS 186.58 AF027763 95

Aureobasidium sp. CECT 11965 AY167611 100 Aureobasidium pullulans ATCC 16628 AF121283 99 E262 Delphinella strobiligena AFTOL-1257 DQ470977 92 Aureobasidium pullulans CBS 584.75 AJ244232 94 Aureobasidium pullulans CBS 584.75 DQ321374 91 Discosphaerina fagi CBS 171.93 AY016359 The Blast search for isolate E262 produced a high homology with A. pullulans strain ATCC16628 in the ITS region (accession AF121283). In contrast, low identities with neo-type strain of A. pullulans CBS 584.75 was obtained in both the ITS region (accession AJ244232) and D1/D2 loop (accession DQ321374). The nucleotide divergence of strain ATCC 16628 suggests that it belongs to a species separate from A. pullulans, and needs reclassification.

4.3.3 Morphological features Figures 4,3 and 4.4 show the colony and cellular morphology of several isolates examined in this study. Colony morphology, conidia and conidiogeneous cells of isolates were observed and compared with descriptions detailed in Ellis (1971), de Hoog and Hermanides-Nijhof (1977), Hermanindes-Nijhof (1977), Kockova- Kratochvilova et al. (1980) and Yurlova et al (1999).

Aureobasidium Based on morphology of conidiation, isolates in RFLP group 1 and the isolate E262 were assigned to the genus Aureobasidium; they exhibited synchronous conidiation, where conidia were produced simultaneously in dense groups from lateral, terminal and intercalary cells (Hermanides-Nijhof 1977; Yurlova etal. 1999).

Aureobasidium pullulans; morphological features of isolates belonging to RFLP group 1 (Figure 4.2) were typical of^. pullulans, and the observations here, confirm the identification determined by sequence analyses. Colonies were variable in colour (cream, yellow, pink, brown, black or olive) and periphery, rapidly expanding to 40 mm within 7 days. Growth was often slimy and accompanied by sparse tufts of aerial mycelia. Conidiogeneous cells were undifferentiated, intercalary or terminal. Conidia developed simultaneously, often in dense groups of 2-8 along a hyphal cell. Two varieties could be distinguished; isolates which remained largely light-coloured after three weeks of growth were assigned to var. pullulans. With the exception of three melanized isolates, E38, E39 and 315, the RFLP subgroup lb contained all isolates belonging to var. pullulans. Isolates that turned black within 2-7 days were identified as var. melanigenum and were found within RFLP subgroups la and Ic (Figure 4.3A). B I- -Vs.

Figure 4.3 Colonies oiAureobasidium and related isolates on MEA after 7 days. {A) Aureobasidiumpullulans van pullulans, isolates E33, 126, E49, 202, El, 313; A. pullulans van melanigenum isolates E38, 315, E39 (B) A. pullulans van melanigenum, strain FRR4800, isolate 133, 475; Hormonema sp. isolates E12, ElOO, E274; Kabatiella sp. 28; Hormonema sp. 550; Aureobasidium sp. E262. y I

Figure 4.4. Mycelial conidiogenesis and conidiidioiAureobasidium and related isolates. (A) Aureobasidium pidlulans E238; (B)^. piillulans 367; {C) Hormonema sp. El00; {D)Homionema sp. 550; (E) Aureobasidium sp. E262; (F) Kabatiella sp. 28. Scars on conidiophores are indicated by arrows. Aureobasidium sp.; isolate E262 initially produced mucoid colonies resembling Cryptococcus, but blackening became apparent after 7 days (Figure 4.3B). Cultures of this isolate exhibited morphological characters that were largely consistent with those of A. pullulans. Conidia were formed simultaneously in dense groups of five or more. Conidiogeneous cells were undifferentiated, intercalary or terminal, on darkened hyphae (Figure 4.4). Identical D1/D2 sequences (100%) were observed with Aureobasidium sp. strain CECT 11965 (accession AY 167611) isolated from cork (Alvarez-Rodriguez et al 2003). Low homologies of 91% and 93% with A. pullulans (neo type strain) were obtained in the D1/D2 and ITS regions, respectively, and did not support the assignment to this species.

Hormonema Based on morphological criteria, isolates of RFLP group 2 and isolate 550 are assigned to the genus, Hormonema; conidia were found to develop percurrently or successively, from one to two conidiogeneous cells per hyphal cell (Hermanides-Nijhof 1977). Based on D1/D2 and ITS sequence comparisons, members of group 2 and isolate 550 were also affiliated to a complex of fungal genera within the Dothidiales that have, in common, Hormonema anamorphs/synanamorphs (Dothidea, Dothiora, Dothichiza and Pringsheimia).

Hormonema sp., RFLP group 2; colonies were cream to pink in colour, produced a raised centre, floccose mycelial mat and irregular margins. After 7 days, their colony diameter reached 12-18 mm and cultures turned brown to dark brown, but never blackening. Several isolates (E26, E274) produced a reddish pigmentation that diffused into the agar (Figure 4.3B). Conidiogeneous cells were undifferentiated, mostly intercalary on hyaline hyphae. Conidia were produced typically from one, sometimes two conidiogeneous loci. Detachment scars or collarettes are visible as a small buldge (Figure 4.4C). Hyphae were often covered with granular exudates. ITS sequence homologies of 98-99% suggest a close association with the fungal species Selenophoma eucalypti or its teleomorph Sphaerulina eucalypti, known to produce conidia states of Hormonema in culture (Crous et al. 1995, 2003). Hormonema sp., isolate 550; morphological characters of isolate 550 were consistent with Hormonema states of the Dothiora and Pringsheimia genera (Hermanides-Nijhof 1977; Pelaez et al. 2000). Colonies showed slow radial growth, attaining about 20 mm in 7 days. Mycelia and blackening of colonies were apparent after 72 h of growth (Figure 4.3B). Conidiogeneous cells were undifferentiated and were often intercalary on both hyaline and darkened hyphae. Conidia were often formed from one locus only. Detachment scars were small and inconspicuous, as shown in Figure 4.4D. Liberated conidia were often swollen, melanised and thick walled. Placement in this genus was also supported by sequence analysis in the ITS region, with a 98% match to a Hormonema sp. F-054,258 isolated from leaves of Quercus ilex (Pelaez et al. 2000).

Kabatietia Kabatiella sp. isolate 28; initially cream in colour, gave grey black, deeply submerged colonies after 48 h, and white mycelia was apparent after 7 days (Figure 4.3B). Conidia were solitary, oblong to falcate (sickle-shaped) and aseptate (Figure 4.4F). Conidiophores were hyaline, bearing 1-3 conidia terminally. The conidial characters of this isolate resembled Kabatiella caulivora. Its affiliation to the genus Kabatiella was based on sequence analysis, but placement to the species Kabatiella caulivora was not strongly supported with a 6% sequence divergence in the ITS region (Table 4.3).

4.3.4 Analysis of all Aureobasidium and related isolates by ITS-RFLP

Following the taxonomic identification of the five Aureobasidium-WkQ species by morphological and sequence analyses, a further 140 isolates were screened using ITS- RFLP, to look for new patterns (species) and to determine the proportion of each species among the total collection of 200 isolates.

For broad screening purposes, Cfo\ was chosen as the restriction enzyme to separate all of the five species, except for species in Hormonema. An additional enzyme, Hae\\\ was used to differentiate these isolates. Among 200 isolates, 150 (75%) were represented by the s^qcxqs Aureobasidium pullulans. The next major group, Hormonema sp., RFLP group 2, comprised 38 isolates (19%). Six iolates (3%) of a second Hormonema sp. were found (3%), five isolates of Aureohasidium sp. (2.5%) and one isolate (0.5%) of Kabatiella sp. were recovered. No new profiles were found.

4.3.5 Biochemical profiles of Aureohasidium and related isolates

The biochemical characteristics of isolates and strains are summarised in Table 4.4. All isolates utilized xylose, lactose, lactate, inulin, mannitol, glucuronic-acid and methyl a- D-glucoside, they degraded urea, and grew in the presence of 10% NaCl, and 10% NaCl with 5% glucose. Isolates were variable in their growth response to temperature, 16% NaCl with 5% glucose, and varied in their metabolism of galactose, galactitol, glucosamine, citrate, sorbose, casein and gelatin. Many isolates grew at 30°C and most grew at 5°C. At the higher range, only two strains (E38 and E39) of^. pullulans grew at 37°C and these were also the only two strains that did not grow at 5°C (Table 4.4). Kabatiella sp. did not metabolize myoinositol as the sole carbon source. All members of Hormonema spp. could be distinguished from Aureohasidium and Kabatiella by fermenting glucose and fructose slowly or weakly, often filling one third to three quarters of the Durham tube with gas after one to three weeks.

The Aureohasidium, Hormonema and Kabatiella isolates shared very similar biochemical profiles. All isolates and strains could be linked together at a similarity of 74% (Figure 4.5). At a similarity of 85%, three groups could be recognized within the species OÍA. pullulans. In agreement with the ITS-RFLP grouping, the biochemical data supported the separation of A. pullulans (RFLP group 1) and isolates belonging to Hormonema sp. 1 (RFLP group 2). Also consistent with RFLP data, isolate 550 clustered more closely to the Hormonema group, while, Kabatiella sp. (isolate 28) was found in the A. pullulans group. Isolate E262, while placed as a distinct outgroup by ITS-RFLP, resided within the A. pullulans cluster in the biochemical phenogram. Table 4.4. Biochemical characteristics oiAureobasidium and related isolates from wine grapes

Characteristic A. pullulons A. pullulam A. pullulons Hornionemo sp.. Kabotiello sp., Hormonema sp., Aureohosidium sp., Group la (5)" Group lb (39) Group Ic (2) Group 2(14) Isolate 28(1) Isolate 550 (1) Isolate E262(l) Assimilation: Galactose + 4- v(l) - - + + L-Sorbose + + + + + w + Lactose + + + + + + + Inulin + + + + + + + D-Xylose + + + + + + + D-Ribose + + + + + + + D-Glucosamine v(4) v(22) - - - - - Galactitol v(4) v(24) + v(ll) - + + Mannitol + + + + + + + a-methyl-D glucoside + + + + + + + D-Glucuionate + + + + + + + DL-Lactate + + + + + + + Citrate + + + v(2) + - + Myoinositol + + + + - + + Fermentation: Glucose - - - w - w - Fructose - - - w - w - Hydrolysis: Casein + + - - - - - Gelatin + + - + - - - Urea + + + + + + + Growth at temperature: 5°C + v(37) + + + + + 10°C + + + + + + + 30°C + v(35) + + + + + 34°C - v(6) - v(ll) - - + 37°C - v(2) - - - - - Growth in NaCl/glucose: 10%NaCl + + + + + + + 10% NaCl +5% glucose + + + + + + + 16% NaCl + 5% glucose v,w (3) v,w(18) - w - - w number of strains V, variable within species; numbers in brackets indicates a positive reaction; +, all strains tested positive; -, all strains tested negative; w, weak response 0.74

Figure 4.5 Phenogram showing the results from the biochemical characterization oiAureobasidium and related isolates, using the Simple Matching coefficient, UPGMA clustering method. 4.3.6 PGR microsatellite fingerprinting The reproducibility of PGR patterns for microsatellite primers (GACA)4, (GTG)5, M13 and M13 wild type phage core was estimated by calculating Sneath and Johnson's probability of error,/? for replicate sets of PGR amplifications and subcultures (Sampaio et al. 2001). The lowest similarity value for replicate trials was also determined with the Dice Coefficient using the GelCompar software. For the (GAGA)4 primer, a p value of 0.059 and similarity of 91.7% was obtained from triplicate repeats of the selected isolates. The (GTG)5 primer gave profiles with ap value of 0.107 and similarity of 87.5%. The p value of 0.089 and similarity of 87% were obtained for primer M13. The wild type phage core primer yielded profiles with a p value of 0.209 and similarity of 84.6%, and was not further applied in this study. Examples of banding patterns for reproducibility trials are shown in Figure 4.6.

Amplified DNA fragments with primers (GAGA)4, (GTG)5 and M13 ranged from 300- 4000 bp in size. The number of bands generated ranged from 5-15 per sample. A high level of genetic diversity occurred within the isolates. Except for two isolates, none of the primers gave banding profiles that were common among any of the isolates. For each isolate, a combined profile for the three different primers was assembled. In the composite pattern, 59 profiles were distinguished among the 61 isolates and two reference strains. All strains and isolates investigated could be resolved at a similarity level of approximately 50%. Isolates oiA. pullulans generally clustered together, but these occurred in multiple groups (clusters 1, 5, 6 and 7) at low similarities of <55% (Figure 4.7). The majority of Hormonema isolates clustered together in several groups, in clusters 2, 3 and 8. Three isolates belonging to A. pullulans var. melanigenum (isolates 475, El, 36) occurred in the Hormonema cluster 3, while one isolate belonging to Hormonema (isolate E99) was placed among a group of A. pullulans (cluster 5). These examples showed that the microsatellite primers used in this study did not enable a clear separation of species m Aureobasidium and Hormonema. The taxonomic status of the outgroup isolates 28, 550 and E262 could not be ftirther clarified from the PCR- microsatellite dendogram (Figure 4.7). Otoe (Opti ixm)(T0'0^«^e%)(H>00%s>0i)%)pin(.iû0inq

.y It 16 a t 16c 16 b I ElOOa I I I I ElOOb il! ElOOc iiS! CBS584 75 b CBS584 75 c _ iiii CBS584 .75 a 367 b } iiUi' 367 c I I• t i1)l 3 I 367 a II FRR4800 a FRR4800 b FRR4800 c Mil E12a till E12b i!!,,' MM E12c

Dim (C»ji:1 00%) (Toi DB%0 B*) (H>0 0* S=0 0%) p 0%-IX 0%| a*" gtg

B 367 b i^m 367 c 11 titt 367« I ti rt 16a iU i «I 16 b ifi 16c IMnI CBS684.75 a nil CBS684.75 c 11 CBS684.75 b I Mil FRR4800 a • mn\ FRR4800 b t «tii FRR4800 c i mu\ E100a ElOOc mim E100b tut E12a It li E12b Mil E12c

ElOOa E100b ElOOc FRR4800 a FRR4800 b FRR4800 c C8S5M.75 b œS684.75c C8S584.75 a 367 b ' ill ill) 367 c Il I 111» 367 a 16 b 16 c 16 a liil lS I E12b H t inni i • E12c il« t E12a Figure 4.6. Reproducibility of microsatellite-PCR fingerprinting, for primers (A) (GACA)4, (B) (GTG)5 and (C) Ml 3. Similarités between profiles were calculated using the Dice similarity coefficient and UPGMA cluster analysis. Replicates of each strain (a, b, c) were resolved on the same gel. Isolate Grape variety Region Grape maturity 50 60 70 80 90 100

E38 Stiraz Hunter intermbrix E39 Shiraz Hunter intenm brix

315 Chardonnay Mudgee harvest ripe, D 28 Shiraz Hunter inflorescence 30 Stiraz Hunter inflorescence

E2 Semaor Hunter inflorescence E160 Cab Sauv Mudgee harvest ripe E185 Cab Sauv Mudgee harvest ripe

E20 Cab Sauv Hunter veraison E21 Cab Sauv Hunter veraison

E31 Semi lor Hunter interm brix, D d E97 Cab Sauv Mudgee harvest ripe, D E12 Semilon Hunter pea-sized berries

El Senior Hunter inflorescence E100 Cab Sauv Mudgee har/est ripe, 0 36 Chardonnay Hunter inflorescence

475 Cab Sauv GrifRth harvest ripe E262 Semilon Hunter harvest ripe 550 Cab Sauv Mudgee har/est ripe, D

362 Cab Franc Monaro harvest ripe 367 Traminer Monaro harvest ripe 374 Traniner Monaro harvest ripe

335 Meilot Mudgee harvest ripe, D 313 Shiraz Mudgee harvest ripe 437 SauvMane Griflilh harvest ripe

174 Cab Sauv Hunter interm brix 496 Sen ion Hunter harvest ripe 316 Chard om ay Mudgee harvest ripe, D E306 Semilon Grifnih harvest ripe 101 Chardonnay Hunter pea-sized berries 126 Shiraz Hunter pea-sized berries 66 Chardonnay Hunter pea-sized berries E303 Cab Sauv Griffith harvest ripe

201 Chardonnay Hunter ripe r-E 207 Chardonnay Hunter ripe 262 Chardonnay Hunter harvest ripe 264 Semi on Hunter harvest ripe 314 Merlot Mudgee han«stripe 169 Cab Sauv Hunter intermbrix

295 Semion Hunter harvest ripe 283 Shiraz Hunter hardest r^ 167 Cab Sauv Hunter interm brix

E49 Semilon Hunter ripe -Ef 202 Chardonnay Hunter ripe 268 Cab Sauv Hunter harvest ripe

-c CBS684.75 NT France 365 Trsminer Monaro harvest ripe -E 311 Merlot Mudgee harvest ripe 263 Shiraz Hunter harvest ripe E238 Cab Franc Monars harvest ripe 472 Tyrian Griffith harvest ripe

E304 Cab Sauv GrifTilh harvest ripe 4 E99 Cab Sauv Mudgee harvest ripe, D 134 Shiraz Hunter veraison FRR4800 Gordo Miidura raisin 293 Shiraz Hunter harvest ripe

16 Chardonnay Hunter inflorescence 33 Cab Sauv Hunter inflorescence E33 Chardonnay Hunter interm brix € E343 Semillon Griffith harvest ripe, D E26 Cab Sauv Hunter intermbrix E278 Semion Griffith harvest ripe

E274 Tyrian Griffith harvest ripe

Figure 4.7 UPGMA clustering of Aureobasidium and related isolates derived from the combined PCR-microsatellite fingerprints using the Dice similarity coefficient. The numbers on branches are cophenetic correlation indices. Within the species of A. pullulans, clusters generated by microsatellite profiling did not always reflect groupings defined by biochemical data. The exception found here was the pairing of A. pullulans isolates E38 and E39. The microsatellite dendogram did, however, segregate the non- or lightly pigmented isolates of var. pullulans (cluster 5) from melanized isolates of var. melanigenum (clusters 1, 6, 7) (Figure 4.7).

The grouping of isolates based on geographical origin was rarely apparent. Several isolates (362, 367, 374) from grapes collected from the Snowy Mountains/Monaro region shared a cluster of about 80% genetic similarity (cluster 5). However, isolates from distant geographical regions also fell into the same clusters; for example, isolate 365 from Monaro shared 75% similarity with the neo type strain isolated from France. Isolates E238 and 472 from Monaro and Griffith regions, respectively, were genetically, 80% similar (cluster 5, Figure 4.7).

There was also little evidence of clustering of isolates based on grape cultivar or grape maturity. Consequently, definitive associations between strain genotype and grape cultivar, grape maturity or vineyard location were not observed.

4.4 DISCUSSION

Taxonomy The morphological diversity and variability of yeast-like fungi (often called black yeasts) associated with plant habitats have presented significant challenges to their reliable taxonomic characterisation and classification. Although molecular technologies have greatly assisted these goals, their taxonomic understanding and delineation remains relativelly incomplete (Yuriova et al 1999). However, three generic classifications embrace many of these organisms and these are Aureobasium, Hormonema and Kabatiella (Hermanides-Nijhof 1977; Yuriova et al. 1999).

In this study, isolates of black, yeast-like fungi from grapes were characterised using multiple criteria that included ITS-RFLP fingerprinting, DNA sequencing and morphological, biochemical and physiological properties. Based on this approach, most of the isolates could be clustered into two main groups. The largest group, representing 75% of the isolates, was confidently identified as A. pullulans, although there were genotypic and phenotypic subdivisions within this group. The next most significant group, representing 19% of the isolates, were identified as Hormonema species; no definitive species name could be assigned, but a close relationship with the species Sphaerulina/Selenophoma eucalypti was noted, and more detailed, phylogenetic studies are needed to classify these isolates. The remaining isolates were unclustered individuals that did not match any known species but shared criteria across the Aureobasidium, Hormonema and Kahatiella complex.

Phenotypic characterisation or numerical taxonomy of isolates confirmed the division between Aureobasidium and Hormonema. However, it separated neither Kabatiella sp. noT Aureobasidium sp. from A. pullulans, where a more distant relationship from this species was revealed by molecular analyses. The biochemical similarities among the Aureobasidium complex have been previously noted, with few phenotypical tests able to distinguish species within this complex (Table 2.4; de Hoog and Yurlova 1994). Phylogenetic analyses of the ITS region also supported a close relationship between Aureobasidium and Kabatiella, in which two Kabatiella species {K lini and K. caulivord) were found to cluster within the A. pullulans clade (Yurlova et al. 1999).

In this study, only hyphomycetous cultures were observed. Sequence homologies of the isolates however, provided some clues to their relationship with sexual or asexual coelomyceteous fungi. Although no sexual states have been described for^. pullulans, the genus Discosphaerina is connected to asexual states'm Aureobasidium (Sivanesan 1985). For example, Kabatiella lini (synonymised with A. pullulans) is recognised as the anamorphic species of Dis.fulvida (Yurlova et al 1999). In this study, the partial 26S rDNA sequence of A. pullulans CBS 584.75 was found to be identical to that of Dis.fagi. The phylogenetic studies by Eriksson (2001), and Lumbush and Lindemuth (2001) using the 18S and 26S rDNA also support a close relationship between these two species. The Hormonema isolates (RFLP group 2) shared high (98-99%) homologies to Sphaerulina eucalypti and its coelomyceteous anamorph, Selenophoma eucalypti in the ITS region. These two fungal species share a 98.7% homology in the ITS region, and also produce Hormonema synanamorphs in culture (Crous et al 1995, 2003). Both ITS sequence comparisons and morphological parameters would indicate close taxonomic affiliation of the Hormonema isolates in this study, to the species SphaeruUna/Selenophoma eucalypti. The degree of sequence divergence determined between the pairs of these species falls within the limits of conspecificity (1-2%) generally observed among fungal species in the ITS regions (Yurlova et al 1999; Hsiao etal 2005).

Taylor et al. (1999) describes how understanding of fungal lifecycles (meitotic/mitotic states) can have practical implications in plant pathology. The association of our isolates to the fungal species, Discosphaerina and Selenophoma, may be significant because they are known foliar pathogens (Sanderson 1964; Crous et al. 2003). DNA reassociation experiments have been used to define species boundaries among the yeasts (Kurtzman and Fell 1998), and could be helpful here, to determine whether^. pullulans and Discosphaerina fagi^ or Hormonema sp. and SphaeruUna/Selenophoma eucalypti are members of the same species. The relationships between fungal forms and species also involve testing other biological, morphological and phylogenetic species concepts (see examples in Taylor et al. 1999).

Ecology Detailed discussion on the ecological association of A. pullulans with wine grapes has been given in Chapter Two. Several other studies have also reported the occurrence of black yeast species, other than A. pullulans, in grape and wine systems. For example. Poulard (1983) provides descriptions of Exophiala jeanselmei in grape musts, and Schweigkofler and Prillinger (1997) identified Hormonema dermatioides within woody tissues of the grapevine. The genus Hormonema is commonly associated with plants, as endophytes and epiphytes (see for example Hermanides-Nijhof 1977; Middelhoven 1997; Pelaez et al. 2000). However, its presence on wine grapes has never been reported, and given its morphological similarities, may have been previously confused with Aureobasidium (Hermanides-Nijhof 1977; Ray et al. 2004). Species of Kabatiella, may be saprophytic, but most are pathogens of herbaceous plants (Yurlova et al. 1999). The species K. caulivora is a causal agent of anthracnose (scorch) in clover (Bayliss et al. 2003). The incidence oi Kabatiella species on grapes was extremely rare and has not been previously encountered in the literature. Diversity among isolates of A. pullulans examined using PCR- microsatellite fingerprinting In this study, heterogeneity of microsatellite patterns was not caused by issues of PGR inconsistencies or genetic instability, as evaluated in the reproducibility trials. The differences of magnitude, expressed either as % similarity or p values are within the expected range reported in literature (Sampaio et al. 2001 ; Gente et al 2002).

A high level intraspecific variability was detected among pullulans by microsatellite fingerprinting. The primers used could not differentiate these isolates at the species level, suggesting that microsatellite primed PGR are highly discriminatory and may not be useful for characterising higher ranks of fungal taxa (Bruns étal 1991). However, in common with the study by Urzi et al. (1999), differentiation at the subspecies level could be achieved using this method; strains of A. pullulans var. pullulans were separated from strains belonging to var. melanigenum on the combined microsatellite tree (Figure 4.7).

The degree of genetic polymorphism within this species is consistent with other studies where a large number of strains were investigated. Such studies were performed using microsatellite fingerprinting, PGR-RAPD and fluorescent-AFLP protocols (Schweigkofier and Prillinger 1997; Schena et al. 1999; Schena et al 2003; De Gurtis et al 2004). Similarity levels between 18-35% have been reported among phylloplane isolates of pullulans from various crops from Greece and Southern Italy (De Gurtis et al 2004). In comparison, a lower level of genetic variation (variations of up to 30%) was detected among endophytic grapevine isolates from Austria (Schweigkoffler and Prillinger 1997). The lack of association between genetic similarity of A. pullulans and its geographic or host plant origin was also observed by these authors.

Genetic variation among fungal species and populations can provide insights into their ability to respond and adapt to changing or heterogeneous host and environmental conditions (Ahlholm et al 2002). The level of diversity observed among isolates of^. pullulans is notable, given that the fungus has no known sexual stage. A high level of genetic polymorphism has also been documented for other asexual phytopathogenic fungi, for example Kabatiella caulivora (Bayliss et al 2003), Verticillium dahliae (Dobinson et al. 2000) and species of Fusarium oxysporum (Kistler 1997). Recombination in mitosporic fungi has been shown to occur (reviewed in Taylor et al. 1999), and Schena et al. (1999) suggested that this behaviour is supported by the multinucleate condition observed \nA. pullulans (described by Cooke 1961; Takeo and de Hoog 1991). This species does not have a restricted host range (non host specialised). Efficient mobility via conidia (dispersion units) could facilitate the dissemination of genotypes to new habitats at various levels, for example, different plant hosts/substrates and geographical areas. This hypothesis could be tested by examining a larger collection of isolates from other countries, hosts or substrates.

PART 2. PROPERTIES OF OENOLOGICAL SIGNIFICANCE 4.5 RESULTS

4.5.1 pH, osmo-, ethanol and SO2 tolerance The ability oiAiireobasidium and related isolates to grow at several pH values, osmotic pressure and concentrations of ethanol and total SO2 are recorded in Table 4.5. All isolates grew at pH 3.0, 3.5 and 4.0 in glucose-YNB broth. All isolates were also able to grow in the presence of 10% NaCl and 10% NaCl plus 5% glucose tested in glucose- YNB, pH 5.5. All Hormonema isolates (RFLP group 2) grew in the presence of 16% NaCl plus 5% glucose. Isolates of^. pullulans were less osmotolerant, with less than half the number of strains growing under this same osmotic condition.

All isolates of A. pullulans (100%) grew in the presence of 2% ethanol, and six isolates (13%) grew in the presence of 4% ethanol. All isolates of Hormonema sp. (RFLP group 2) grew in media containing 4% ethanol, and none grew in the presence of 6% ethanol.

Strains of A. pullulans were more resistant to SO2 than members oi Hormonema sp. (Table 4.5). All 14 isolates of Hormonema sp. were sensitive to 50 ppm SO2 at pH 3.5.

About half of the k. pullulans isolates (52%) grew in the presence of 50 ppm SO2, and ten isolates (22%) grew in the presence of 100 ppm SO2. None of the A. pullulans isolates grew in 150 ppm of SO2 at pH 3.5. Table 4.5 Growth of Aureobasidium and related isolates at several pH values,

Growth condition Ap Horm sp. Kab sp. Horm sp. A sp. m (14) (1) (1) (1) pHS.O 46'' 14 1 1 1 pH3.5 46 14 1 1 1 pH 4.0 46 14 1 1 1

Glucose (5%) + NaCl (10%) 46 14 1 1 1 Glucose (5%) + NaCl (16%) 21 14 0 0 1

Ethanol (2%) 46 14 1 I 1 Ethanol (4%) 6 14 0 0 1 Ethanol (6%) 0 0 0 0 0 Ethanol (8%) 0 0 0 0 0

SO2 (50 ppm) 24 0 0 1 0 S02(100ppm) 10 0 0 0 0 S02(150 ppm) 0 0 0 0 0 SO2 (200 ppm) 0 0 0 0 0 A p, Aureobasidium pullulam\ Horm, Hormone ma sp.; Kab, Kabatiella sp.; A sp., Aureobasidium sp. total number of strains examined; ^ data in table indicate number of strains positive for the growth condition

4.5.2 Utilisation of grape organic acids Assimilation of L-malic and L-tartaric acids as sole carbon source was a property common to all of the Aureobasidium and related isolates from grapes examined in this study. These compounds were tested in the form of sodium salts.

Isolates grown in L-tartaric acid were tested for pH after 14 days of growth. The pH of cultures changed from an initial value of 5.6 to a final pH between 5.82-7.63. In some cases, culture turbidity was not an indicator of the degree of change in pH; for example, isolates 134, 36 and 315 grew to a turbidity level Biolog standard of only +1, but elevated the final pH to 6.14, 6.27 and 6.10, respectively.

4.5.3 Production of exopolysaccharides Aureobasidium and related isolates were screened for mucoid and non-mucoid phenotypes by observation of colony morphology on plates of MEA. Ten isolates of each phenotype were selected and investigated for their potential to produce exopolysaccharides in a defined medium with sugar concentration and pH similar to grape juice (WYNB; YNB containing 100 mg/1 of glucose, 100 mg/1 fructose, pH 3.5). After culture for 5 d at 25°C, samples were examined for, biomass production, the glucan instability test (Zocklein et al 1995), and polysaccharide production (Table 4.6).

Biomass production Both Aureobasidium and Hormonema isolates grew in the WYNB medium, producing biomass ranging from 119 mg/1 to a maximum of 590 mg/1 dry weight (Table 4.6). Biomass amounts produced by Aureobasidium species were more variable than those produced by Hormonema isolates. The final pH of supematants from the A. pullulans culture remained stable at about pH 3.5, but a slight decrease in pH to 3.29-3.34 was observed for Hormonema cultures, and to pH 3.28 for^. pullulans isolate E38.

Industrial test for glucan Eight of the ten mucoid isolates of Aureobasidium were positive producers of glucan, forming insoluble gelatinous precipitates or white filaments. One non-mucoid isolate, A. pullulans 475, was positive for glucan production, while Aureobasidium sp. E262 was particularly noted for its highly mucoid/slimey colonies, but did not result in the formation of glucan precipitates. Colonies of Hormonema sp. were non-mucoid in appearance (dry, velvety) and did not give insoluble precipitates in the glucan test (Table 4.6). As a negative control/standard, commercially purchased pullulan (Sigma, St. Louis, MO) gave a cloudy suspension, but was largely soluble when used at concentrations of 15 mg/1 in the glucan instability test. This observation suggested that either higher concentrations of pullulan are required to cause precipitates or that some of the isolates examined here, produced polysaccharides other than pullulan.

Polymer yield and carbohydrate composition Ethanol-precipitated polymers (Materials and Methods) gave yields ranging from 140 mg/1 to 478 mg/1. Higher yields were often obtained from cultures of pullulans. In some cases, it was difficult to assess how much of the crude polymer was attributed to the production of exopolysaccharide. In a few cases, the range of error for both biomass and polymer yields of duplicate cultures could exceed 15%; for example for isolates E38 and 264. Difficulty in separating fungal biomass from polysaccharides was sometimes encountered. In these instances, mucilaginous material strongly adhered to cells, despite repeated washing of pelleted cells, and could account for over/under estimation of these concentrations.

To determine the amount of polysaccharide in the precipitated polymers, analysis was performed for the content of total carbohydrate. The carbohydrate content of crude polymers ranged from 31-86%, and polymers from A. puUulans cultures were often more variable in carbohydrate content. Commercially purchased pullulan (control) consistered entirely of carbohydrate. Notably, polysaccharide material was also recovered from the supematants of Hormonema sp. (RFLP group 2), but the amounts were generally less than those from A. pullulans cultures (Table 4.6).

Table 4.6 Production of exopolymers by Aureobasidium and related isolates in WYNB medium

Isolate Biomass Final Glucan test; Crude Total dry weight pH presence and polymer yield carbohydrate % mg/ml appearance mg/ml Mucoid phenotype A. pullulans E305 0.41±0.01 3.42 + filaments, gelatinous 0.42±0.08 74.4±0.7 134 0.38±0.04 3.51 + filaments, haze 0.32±0.07 38.6il.4 174 0.19i0.09 3.45 - 0.36±0.04 82.5±1.7 E33 0.15±0.00 3.55 + gelatinous 0.27±0.05 44.6J:1.6 E38 0.59±0.09 3.28 + gelatinous 0.34±0.03 66.1±0.8 201 0.18±0.04 3.45 + filaments, haze 0.24±0.03 46.at0.9 264 0.21±0.06 3.47 + filaments, haze 0.48±0.02 85.9±0.7 367 0.19±0.00 3.53 + filaments 0.46±0.03 49.5±0.9 315 0.13±0.03 3.55 + gelatinous 0.36±0.03 45.6±0.5 Aureobasidium sp. E262 0.12±0.00 3.52 - 0.2^0.08 31.3±0.5 Non-mucoid phenotype A. pullulans 475 0.33±0.06 3.46 + filaments 0.19±0.02 54.0±0.6 101 0.22±0.01 3.47 0.32±0.03 38.6±i.l Hormonema sp. E12 0.15±0.06 3.26 0.18±0.01 38.6±0.7 40.6±1.4 E21 0.15±0.01 3.29 0.14±0.01 39.4±0.9 E31 0.18±0.01 3.34 0.18±0.01 0.21±0.01 37.5±0.9 E99 0.21±0.01 3.31 0.19±0.0I 40.7±0.4 ElOO 0.15±0.01 3.33 33.3±0.6 E160 0.19±0.01 3.33 0.20±0.00 0.023±0.00 36.8±0.6 E185 0.18±0.01 3.31 0.]5±0.00 38.5±0.7 E278 0.20±0.01 3.29 Uninoculated WYNB medium was ethanol-precipitated and yielded 0.016 mg/ml ± 0.00. 4.5.4 Interactions of Aureobasidium and related isolates with other microorganisms of oenological significance

4.5.4.1 Filamentous fungi and yeasts In vivo, plate culture assays were used to determine microbial interactions. The number of isolates with antagonistic activity against various species of yeasts and filamentous fungi are listed in Table 7. Only few^. pullulons isolates (3 of 46) exhibited antagonistic activity. By contrast, antagonistic activity was far more prevalent among strains of Hormonema sp., where 9 of 14 isolates exhibited antagonism against the yeast and fungal cultures examined.

Inhibition of fungi {Botrytis cinerea and Pénicillium sp.) significant in the spoilage of wine grapes was observed. In addition, species noted for the production of Ochratoxin A such as Aspergillus carbonarius and Aspergillus niger (rare producers) were also inhibited. Several other species of yeasts isolated from wines and significant in wine production were antagonized. The antagonistic effect against the principal wine yeast Saccharomyces cerevisiae was notable (Figure 4.8B). Well known spoilage yeasts such as Dekkera bruxellensis, Schizosaccharomyces pombe and Zygosaccharomyces bailii were also inhibited (Table 4.7).

There were several instances where non-antagonistic A. pullulans isolates encouraged the growth of indicator species. This activity was particularly noted in lawn cultures of the yeast, Hanseniaspora uvarum (Figure 4.8E).

To determine their response to the growth of other yeast species, the isolates were tested as indicator strains. Patterns of sensitivity are reported in Table 4.8. The Aureobasidium, Hormonema and Kabatiella isolates were inhibited by ascomycetous yeast species oiP. anomala, M. pulcherrima, Metschnikowia sp. and K. thermotolerans. None of the basidiomycetous yeasts inhibited growth of the isolates. Isolates were also insensitive to killer strains oi S. cerevisiae, strains 3002 and 516 (Heard and Fleet 1987). Antagonistic Hormonema isolates were inhibitory toward each other and to A. pullulans (Table 4.8). Table 4.7 Inhibiton of filamentous fungi and yeasts significant in wine production hy Aureobasidium and related isolates

Indicator strain Ap Horm sp. Kab sp. Horm sp. A s| 46)" (14) (1) (1) (1]

Filamentous fungi Alternarla infectoria FRR4821 3 0 0 0 Aspergillus aculeatus FRR5376 1 2 0 0 0 Aspergillus carbonarius FRR5374 0 5 0 0 0 Aspergillus niger FRR5375 1 8 0 0 0 Botrytis cinerea FRR5215 1 7 0 0 0 Cladosporium cladosporiodes FRR4778 2 8 0 0 0 Pénicillium crustosum FRR4775 0 7 0 0 0 Pénicillium expansum FRR4832 0 9 0 0 0 Ascomycetous yeasts Dekkera bruxellensis AWRIl 102 1 9 0 0 0 Dekkera bruxellensis AWRIl205 1 9 0 0 0 Hanseniaspora uvarum E55 1 2 0 0 0 Hanseniaspora uvarum E56 1 3 0 0 0 Issatchenkia orientalis E101 I 2 0 0 0 Issatchenkia orientalis El 13 1 3 0 0 0 Kluyveromyces thermotolerans E170 0 3 0 0 0 Kluyveromyces thermotolerans E209 0 3 0 0 0 Metschnikowia pulcherrima 255 1 3 0 0 0 Metschnikowia pulcherrima 266 0 3 0 0 0 Pichia anomala AWRI1023 1 2 0 0 0 Pichia anomala AWRIl051 1 2 0 0 0 Saccharomyces cerevisiae SB-1 2 8 0 0 0 Saccharomyces cerevisiae AWRllOlO 2 5 0 0 0 Saccharomyces cerevisiae Maurivin 1786 2 6 0 0 0 Saccharomyces cerevisiae Primeur 3 6 0 0 0 Saccharomyces cerevisiae E227 1 4 0 0 0 Saccharomyces cerevisiae E228 0 5 0 0 0 Schizosaccharomyces pombe BS24 6 0 0 0 Schizosaccharomycespombe BS29 5 0 0 0 Torulaspora delbrueckii CBS 1146 5 0 0 0 Torulaspora delbrueckii AWRI925 3 0 0 0 Zygosaccharomyces bailii UNSW507700 3 0 0 0 Zygosaccharomyces bailii EI 12 4 0 0 0 Basidiomycetous yeasts Cryptococcus laurentii 41 4 0 0 0 Cryptococcus laurentii 171 3 0 0 0 Rhodosporidium babjevae 51 9 0 0 0 Rhodosporidium babjevae 200 7 0 0 0 Sporobolomyces rubberrimus 37 3 0 0 0 Sporobolomyces ruberrimus 48 4 0 0 0 A p, Aureobasidium pullulans' Horm, Hormone ma sp.; Kab, Kabatiella sp.; A sp., Aureobasidium sp. total number of strains examined; ^ data in table indicate number of antagonistic strains Table 4.8 Sensitivity of Aureobasidium and related isolates to each other and to other yeast species

Producer strain Ap Horm sp. Kab sp. Horm sp. A sp. {If (2) (1) (1) (1)

Aureobasidium pullulans 101 0 0 1 1 Aureobasidium pullulans 126 2 0 0 1 1 Hormonema sp. E12 2 2 1 1 I Hormonema sp. E21 2 2 1 1 1 Kluyveromyces thermotolerans E170 2 2 1 1 1 Kluyveromyces thermotolerans E209 2 2 1 I 1 Metschnikowia pulcherrima 255 2 2 1 1 I Metschnikowia pulcherrima 266 2 2 1 1 1 Metshnikowia sp. E84 2 2 1 1 1 Pichia anomala E214 2 2 0 1 1 Pichia anomala AWRI1051 2 2 0 1 1 Hanseniaspora uvarum E55 0 0 0 0 0 Hanseniaspora uvarum E56 0 0 0 0 0 Issatchenkia orientalis El03 0 0 0 0 0 Issatchenkia orientalis El 13 0 0 0 0 0 Saccharomyces cerevisiae CY96 0 0 0 0 0 Saccharomyces cerevisiae CY97 0 0 0 0 0 Torulaspora delbrueckii CBS 1146 0 0 0 0 0 Torulaspora delbrueckii AWRI925 0 0 0 0 0 Cryptococcus laurentii 41 0 0 0 0 0 Cryptococcus laurentii 171 0 0 0 0 0 Rhodosporidium babjevae 51 0 0 0 0 0 Rhodosporidium babjevae 200 0 0 0 0 0 Sporobolomyces roseus 56 0 0 0 0 0 Sporobolomyces roseus 140 0 0 0 0 0 Sporobolomyces ruberrimus 37 0 0 0 0 0 Sporobolomyces ruberrimus 48 0 0 0 0 0 A p, Aureobasidium pullulans', Horm, Hormonema sp.; Kab, Kabatiella sp.; A sp., Aureobasidium sp. total number of strains examined; data in table indicate number of antagonistic strains

4.5.4.2 Bacteria ThQ Aureobasidium and related isolates were antagonistic toward bacterial species within Gram positive and Gram negative groups (Table 4.9). Except for Bacillus species, the inhibition of other bacterial species only became apparent or was more pronounced using the deferred plate culture assay.

The biopesticide organism, Bacillus thuringiensis, was widely inhibited, with almost all isolates antagonistic towards commercially applied strains, Delfín and Dipel (Figure 4.8C). Other bacterial species commonly encountered on aerial plant surfaces, Envinia carotovora, Xanthomonas campes tris and the plant pathogen Pseudomonas syringae, were inhibited by several isolates. Generally, bacteria associated with wine fermentation such as lactic acid- and acetic

acid bacteria, were not inhibited. Some exceptions were inhibition of the spoilage

bacterium Acetohacter aceti and the malolactic bacterium Oenococcus oeni by a few

isolates of pullulans and a significant portion of the Hormonema sp. (Figure 4.8D).

Many isolates belonging to^. pullulans and Hormonema sp. inhibited a range of

foodbome bacterial pathogens, notably species of Acinetobacter, Listeria monocytogenes Staphylococcus aureus and Bacillus cereus. Gram negative foodbome pathogenic bacteria, Escherichia coli and Salmonella enteritidis were inhibited by far fewer strains (Table 4.9).

Table 4-9 Inhibition of bacterial species hy Aureobasidium and related isolates A p, Aureobasidium pullulans-. Norm, Hormonema sp.; Kab, Kabatiella sp.; A sp., Aureobasidium sp. Indicator strain s,d Ap Horm sp. Kab sp. Horm sp. A sp. (46)^ (14) (1) (1) (1) Phyllospheric species Bacillus thuringiensis Delfín s 39'' 12 Bacillus thuringiensis Dipel s 32 11 Erwinia carotovora UNSW031700 d 40 12 Pseudomonas syringae UNSW036900 d 33 7 Xanthomonas campestris UNS W031500 d 18 7 Wine-related species Lactobacillus brevis UNSW055100 d G 0 Lactobacillus plantarum UNSW084800 d 0 0 Leuconostoc mese nie roides d G 0 UNSW060700 Oenococcus oeni OENOS d 1 5 Oenococcus oeni CH35 d 1 8 Pediococcus pentosaceus UNSW047200 d G 0 Acetobacter aceti UNSW035801 s,d 4 5 G G G Acetobacterpasteurianus UNSW092200 s,d 0 0 0 G G Gluconobacter oxydans UNSW030300 s,d G 0 0 G G Foodbome pathogens Acinetobacter baumannii ATCC15308 d 26 2 0 G 1 Acinetobacter johnsonii lettuce isolate d 29 12 1 G 1 Bacdlus cereus UNSW052300 s 43 11 Escherichia coli UNSW048200 d G 0 0 0 1 Listeria monocytogenes Tecra 1768 d 31 14 1 G 1 Salmonella enteritidis UNSW031901 d 10 0 G G 1 Staphylococcus aureus UNSW038200 d 41 9 G G 1 ^ total number of strains examined; ^ data in table indicate number of antagonistic strains; s, spot assay; d, deferred assay; blank spaces, assay was not done Figure 8. Inhibitory effects oiAureobasidiumpulliilans md Hormonema sp. against lawn of {A) Aspergillus aculeatus; (B) Saccharomyces cerevisiae; (C) Bacillus thuringiensis, (D) Oenococcus oenii.

Figure 4.8. (E) Stimulated growth of lawn culture Hanseniaspora uvarum by isolates of A. pullulans indicated by arrows. 4.5.5 Mixed culture fermentation with S. cerevisiae Agar plate screening showed clear inhibition of S. cerevisiae by isolates of Hormonema species. To investigate this interaction further, two strongly antagonistic Hormonema isolates were grown in mixed culture with two commercial strains of S. cerevisiae, in various combinations. Growth in single and co-cultures was conducted in defined grape juice medium (pH 3.5; DGJM) and in glucose-YNB (pH. 5.5). The growth of these organisms is shown in Figures 4.9 and 4.10, respectively.

No inhibition of the growth of iS". cerevisiae by Hormonema sp. was observed in the defined grape juice medium (Figure 4.9). When grown together with S. cerevisiae, populations oiHormonema sp. rapidly declined after reaching 5 x cells/ml on day 1, and did not survive after day 2 of fermentation. Similar conclusions were found for combined cultures in glucose YNB medium (Figure 4.10).

The results also show that Hormonema isolates had the potential to grow in both DGJM and glucose YNB media (Figures 4.9 and 4.10, respectively). In the absence of S. cerevisiae, they grew to populations of 10^ cells/ml. The Hormonema isolates displayed a gradual transition from yeast phase to mycelia during fermentation and, progressively, a layer of mycelial cells developed on the glass surface either at the bottom of the flask or at the liquid/air interface. This phenomenon would probably affect estimates of its total population in the growth curve. 4 6 2 4 6 8 10 Incubation time (days) Incubation time (days)

Figure 4.9. Interactive growth of wine strains of Saccharomyces cerevisiae with Hormonema sp. in DGJM at pH 3.5.

(A) Populations of S. cerevisiae Primeur: • single culture; with Hormonema sp. E12; O with Hormonema sp. E21

(B) Populations of S. cerevisiae SB-1: • single culture; with Hormonema sp. E12; O with Hormonema sp. E21

(C) Populations of Hormonema sp. El2: Osingle culture; • with S. cerevisiae Primeur; Q with S. cerevisiae SB-1

(D) Populations of Hormonema sp. E21: Osingle culture; • with S. cerevisiae Primeur; Q with S. cerevisiae SB-1 20 40 60 20 40 60 Incubation time (h) Incubation time (li)

20 40 60 20 40 60 Incubation time (h) Incubation time (h)

Figure 4.10 Interactive growth of wine strains of Saccharomyces cerevisiae with Hormonema sp. in glucose-YNB (pH 5.5) medium.

(A) Populations of S. cerevisiae Primeur: • single culture; with Hormonema sp. E12; O with Hormonema sp. E21

(B) Populations of S. cerevisiae SB-1: • single culture; with Hormonema sp. E12; O with Hormonema sp. E21

(C) Populations of Hormonemo. sp. E12: Osingle culture; • with S. cerevisiae Primeur; with S. cerevisiae SB-1

(D) Populations of Hormonema sp. E21: "^single culture; • with S. cerevisiae Primeur; with S. cerevisiae SB-1 4.5.6 Enzymatic profiles Screening for a range of extracellular enzymes (Table 4.10) was carried out to improve understanding of these species- their biology, ecology and relationship with the host plant. Knowledge of their degradative and oxidative capabilities would also help to predict their impact on wine fermentation and reveal directions for industrial applications. Figure 4.11 gives a pictorial illustration of some of the reactions.

The ability to degrade proteins could be a useful taxonomic character. It was observed in isolates of A. pullulam and Hormonema sp. (Table 4.10), but not for the other three species. Except for the two isolates belonging to RFLP group Ic, all other members of A. pullulam hydrolysed both casein and gelatin equally well. Members of Hormonema sp. did not hydrolyse casein, but were all able to liquefy gelatin.

Pectinolytic activity was found in all species except Kahatiella sp. 28. Generally, pectin degradation was weaker at pH 7.0 (clearing zone 1-2 mm) than at pH 5.0 (clearing zones 5-10 mm). Degradation of xylan and carboxy methyl cellulose was widely detected, but neither filter paper nor lignin were degraded. Cellobiase or p-glucosidase activity, using either cellobiose, arbutin or esculin as substrates was detected in most strains.

All species were strongly lipolytic, with Tween 20 more readily degraded than Tween 80. Most isolates of^. pullulans and Hormonema sp. were cutinolytic when assayed with pNB. However, the ability of Hormonema sp. to hydrolyse pNB did not correspond with depolymerising the cutin analogue PCL (Table 4.10).

All isolates produced polyphenol oxidase, but variations in substrate specificity and basal growth requirements were observed. For example, tannic and gallic acid oxidation by Hormonema sp. was undetected when ME A was used as the basal medium, but was weakly positive on MMN agar. Laccase activity, indicated by guaiacol and a-naphthol, was absent in Hormonema sp. but was observed in A. pullulans, and the three ungrouped species. Table 4.10 Extracellular enzymes produced by Aureobasidium and related isolates

Enzyme activity and substrate Ap Harm sp. Kab sp. Horm sp. A sp. (46)^ (14) (1) (I) (1)

Proteolytic Casein 44^ 0 0 0 0 Gelatin 44 14 0 0 0 Lipolytic Tween 20 46 14 1 1 1 Tween 80 44 14 1 1 0 Cutinolytic p-nitrophenyl butyrate 39 10 1 1 0 Polycaprolactone 39 0 0 0 0 Pectinolytic Pectin pH 5.0 46 14 0 1 1 Pectin pH 7.0 46 14 0 1 1 Hemicellulolytic Xylan 46 14 1 1 1 Cellulolytic CMC 46 8 1 1 1 Filter paper 0 0 0 0 0 P-glucosidase Cellobiose 46 14 1 I 1 Arbutin 39 11 1 1 0 Esculin 46 14 1 1 1 Polyphenol oxidase Tannic acid 46 14 1 1 0 Gallic acid 43 14 1 1 0 Guaiacol 7 0 1 1 1 a-naphthol 14 0 1 1 1 Lignin degradation Polymeric dye R-478 0 0 0 0 0 A p, Aureobasidium pullulans; Horm, Hormonema sp.; Kab, Kabatiella sp.; A sp., Aureobasidium sp. ^ total number of strains examined; ^ data in table indicate number of strains found positive for the enzyme assay Figure 4.11. Enzymatic activities oiAureobasidhm and related isolates: Cutinase activity using (A) p-nitrophenyl butyrate and (B) polycaprolactone. Degradation of (C) pectin pH 5.0, (D) carboxymethyl cellulose. Phenolic oxidation of (E) gallic acid and (F) guaiacol. 4.6 DISCUSSION

This section has QXd^mwQd Aureobasidium and related isolates from wine grapes for properties that relate to their ability to grow in grape juice, and their potential to influence wine fermentation and wine quality.

Growth under conditions of grape juice The isolates in this study showed good potential to proliferate in the grape juice by growing at the pH range and osmotic conditions expected in this environment. Their tolerance to ethanol (generally less than 5% ethanol) is quite low and probably accounts for their disappearance in the early stages of grape juice fermentation (Pardo et al. 1989; Holloway et al 1990; Subden et al. 2003). Strain variation with respect to ethanol tolerance was evident and this may account for occasional reports where A. pullulans has been isolated from the later stages of wine fermentation (Benda 1964; Poulard et al. 1980). They are reasonably tolerant of SO2 (50-100 ppm) and unlikely to be inhibited by the levels of SO2 currently used in wine production (Table 4.5).

The ability to ferment sugars (indicated by gas production) is not a property known among the black yeasts or meristematic fungi (Cooke and Matsuura 1963; Dennis and Buhagiar 1973; Sterflinger 2006). Interestingly, all isolates designated to Hormonema exhibited fermentative metabolism of glucose and fructose, although the reactions were relatively weak or slow. None of the isolates of A. pullulans were fermentative. The fermentative potential of these yeast-like fungi has been noted previously, and it was suggested that^. pullulans can generate up to 2% ethanol (Carolan et al. 1976; reviewed in Dittrich 1977). Some years ago, Clark and Wallace (1958) and found that resting cells oiA. pullulans could catabolize glucose by some reactions of the Embden Meyerhof pathway. Given the difficulties associated with the classification and taxonomy in this group, it is possible that such isolates of A. pullulans were strains of Hormonema. In addition to ethanol, metabolites produced by A. pullulans included glycerol, mannitol, acetic-, succinic- gluconic- oxalic- and lactic acids and acetaldehyde (reviewed in Dittrich 1977). It may be interesting to determine if any other volatile components are generated during their growth as they could impact on wine flavour. However, given the weak growth of these Aureobasidium/Hormonema organisms in grape juice ferments, the concentrations produced may be insignificant.

Utilization of malic and tartaric acids Utilization of tartaric acid was a consistent and notable property among the isolates. This property is not common among the ascomycetous yeasts (Fonseca 1992; Fonseca et al. 2000) but several basidiomycetous yeasts such as Cr. laiirentii and Rh. glutinis have been noted to utilize tartaric acid. The grape fungus Botrytis cinerea is well known for its ability to metabolise tartaric acid (Doneche 1993) and earlier. Poulard et al. (1983), described the potential of the black yeast, Exophiala jeanselmei to degrade tartaric acid.

Tartaric acid and, to a lesser extent malic acid are the predominant organic acids in grape juice, and greatly influence the pH, microbial stability and sensory quality of both juice and wines. Carino-Carina (1929) (in Cooke 1959) and Poulard et al (1983) have observed that total acidity of grape juice is greatly reduced when black yeasts are allowed to proliferate. For example, a 30 day incubation with strains of Exophiala jeanselmei degraded 13-18%, while incubation with A. pullulans degraded 1-23% tartaric acid in grape juice (Poulard 1982; Poulard et al 1983). Consequently, significant growth oiAureobasidium/Hormonema species from grapes could diminish juice or wine acidity.

Production of extracellular polysaccharides Polysaccharides present in juice and wines pose major obstacles during wine filtration. The production of P-D-glucans by the growth of Botrytis cinerea on grapes and its interference with wine filtration is well documented (Dubourdieu et al 1981). Extracellular polysaccharide production by lactic acid bacteria and acetic acid bacteria also has negative impacts in wine processing (du Toit and Pretorius 2000). Many of the A. pullulans and Hormonema isolates from grapes demonstrated the capacity to produce extracellular polysaccharides (Table 4.6), and because many of these organisms are dominant on grapes (Chapter 3), they must be considered as a source of polysaccharides likely to impact on wine processing. Moreover, this property could promote the attachment and survival of these species on the surface of grapes and help to explain their prevalence on this habitat (Andrews et al. 1994).

The production of exopolysaccharides hy A. pullulam during wine fermentations has not been previously demonstrated. However, yields of exopolysaccharides (6-16 g/1) are commonly recovered following its growth in various agro-industrial wastes, such as grape skin pulp extract, beet molasses, whey and brewery wastes (Israilides et al 1994, 1998; Roukas 1999; Lazaridou et al. 2002). In addition to pullulan (an a -l->4; a -1^6 linked glucan), A. pullulam is known to produce several other polysaccharides and polymers that differ in chemical structure and composition. Some examples include ß- linked glucans, heteropolysaccharides and acidic polysacchardies which are based on, or contain residues of L-malic acid (Liu and Steinbüchel 1997), uronic acid or glucuronic acid (Pouliot et al. 2005). The polysaccharide and pullulan composition of these various polymers are sensitive to many factors, including strain, growth conditions (pH, carbon and nitrogen source, agitation) and morphological forms of A. pullulam (reviewed in Leathers 2002). The conditions that govern the expression of exopolysaccharides in this species are not well understood (see for example, Campbell et al. 2004). It is possible that other polysaccharides and polymers may have been elaborated under the experimental conditions used. The acidic units of acidic polymers are connected by strong, stable links that may not be readily hydrolysed by sulfuric acid examined in this study (Verhoef et al. 2002). This property could also help to explain the variability of the carbohydrate component found in crude polymers (Table 4.6).

The potential ior Aureobasidium/Hormonema organisms to impact on the quality and wine process by polysaccharide production has been overlooked in wine science and technology. Given recent interest in wine polysaccharides (Feulliat 2003), further studies are needed on the polysaccharides and other exo-polymers produced by the Aureobasidium/Hormonema species that colonize grape habitats. Interactions of Aureobasidium and related isolates with other organisms

Based on in vitro assays, many isolates showed the ability to inhibit microorganisms across the fungal and bacterial groups (Tables 4.7 and 4.9, respectively). Also, some of these isolates were inhibitory towards each other (Table 4.8).

It is now recognized that the total chain of wine production involves various stages where the properties of one microbial species could affect the behaviour of other species, thereby impacting on the efficiency of the pr|)cess and quality of the final product (Fleet 2003; Viljoen 2006). As a group, the black yeasts are well known for their production of antimicrobial compounds and other bioactive metabolites, and this includes reports for pullulans (McCormack et al. 1994; Vadkertiová and Sláviková 1995), Hormonema sp. (Peláez et al 2000), Hormonema dermatioides (Filip et al. 2003) and Kabatiella caulivora (Bayliss et al. 2003). It was not unexpected, therefore, to fmd that many of the isolates from grapes exhibited antimicrobial action against a diversity of filamentous ñingi, yeast and bacteria (Tables 4.7 and 4.9). At the level of the grape, they could produce substances that affect the yeast and bacterial species that colonize the grape surface, and that carry over into the grape juice to affect the conduct of the alcoholic and malolactic fermentations (Bisson 1999; Fleet 2003).

Filamentous fungi and yeasts The in vitro plate interaction assays showed that many isolates belonging to Hormonema sp. and few isolates of A. pullulans were inhibitory toward Ochratoxin A producing fungi and towards fungi significant in grape spoilage. Botrytis cinerea and various Pénicillium species are major grape spoilage fungi (Pitt and Hocking 1997), and the ability to inhibit these organisms is a significant observation. The inhibition of filamentous fungi by A. pullulans on surfaces of fruits and leaves has been widely documented in the literature (Andrews et al. 1983; Wilson and Chalutz 1989; Köhl 1997; Lima et al. 1997; Ippolito et al. 2000; Castoria et al. 2001 ; Adikaram et al. 2002; Schena et al. 2003). These studies showed the effectiveness of A. pullulans in excluding or preventing fungal colonization on surfaces of fruits and plants, and has since been received much attention for biocontrol, including use in viticulture. The mechanisms of antagonism are thought to be related to their ability to successfully compete for space and nutrients (Andrews et al 1983), the induction of plant defence compounds such as chitinases and p-1,3glucanases (Ippolito et ai 2000; Castoria et al. 2001) or could involve the production of antimicrobial compounds (McCormack et al 1994, 1995).

The ability of Aureobasidium/Hormonema isolates to inhibit various species of wine yeasts is a novel finding and could have several oenological implications. Most notable was the inhibition of S. cerevisiae, one of the major yeasts of wine fermentation. Grapes are a primary source of yeasts in wine fermentation, contributing indigenous species and strains that influence the complexity of wine character (Fleet et al. 2002; Fleet 2003). If the inhibitory action of Aureobasidium/Hormonema species occurrs in vivo at the location of the grape surface, then it could significantly affect the complex of indigenous yeasts contributing to wine fermentation. With respect to S. cerevisiae, there is significant literature and controversy as to its occurrence and association with wine grapes. While some authors have readily isolated this species from the surfaces of wine grapes, others have not been able to recover this yeast from this habitat (Martini et al. 1996; Fleet et al. 2002). Possibly, these discrepancies might be explained by the inhibitory influence of the Aureobasidium/Hormonema grape flora. However, the Hormonema species were not inhibitory to S. cerevisiae in liquid, fermentative culture (Figures 4.9 and 4.10), where the S. cerevisiae exhibited stronger growth, and caused death of Hormonema. In this case, death of Hormonema could be explained by their sensitivity to increasing concentrations of ethanol (Fleet 2003), the lack of available oxygen (Holm Hansen et al. 2001) or cell-cell contact with high S. cerevisiae densities (Nissen et al. 2003). The resuUs, here, indicate that antagonistic Hormonema sp. entering the juice are unlikely to compete with the growth of S. cerevisiae and cause stuck or sluggish wine fermentations.

Killer toxins produced by yeasts such as Pichia spp. have been shown to influence the yeast community structure of rotting fruits in cacti (Starmer et al. 1987; Ganter and Starmer 1992) and amampa fruit (Moráis et al. 1995). Here, A. pullulans and related species were inhibited by several fermentative and wine-associated species oí Pichia anómala, Metschnikowiapulcherrima and Kluyveromyces thermotolerans (Table 4.8). Toxins produced by such killer yeast species may influence the succession of yeasts with oxidative metabolism to yeasts with fermentative metabolism, sometimes observed with ripening and damaged grapes (described in Chapter Three).

Bacteria Agar plate, in vitro assays showed that a large proportion of the Aureobasidium/Hormonema isolates inhibited several phylloplane bacterial species, including the plant pathogenic species Pseudomonas syringae, and the biopesticide. Bacillus thuringiensis (Bt). It is uncertain whether antagonism of these species occurs on the plant surface. McCormack at al. (1995) demonstrated that the suppressed growth oiPs. syringae by A. pullulans could occur on an artificial leaf surface, with the inhibition due to the production of antibiotics. The results here, and observations from McCormack et al. (1994, 1995), suggest that A. pullulans and related species may affect the composition of bacterial species on the phylloplane. Bae et al. (2004) found that populations of B. thuringiensis on grapes ranged between 10-10 CFU/g. It is uncertain whether Bt present of surfaces on grapes occurs as spores, or germinates into vegetative cells. Moreover, the effect of sint^gonistic Aureobasidium/Hormonema on spores of Bt is unknown. The potential of these species to influence the efficacy of the biopesticide warrants further investigations.

A small proportion of A. pullulans and several isolates of Hormonema species clearly inhibited O. oeni (Figure 4.8D), which is an important organism in malolactic fermentation in wines (Fleet 2001, 2003). This antagonism of O. oeni may explain the sporadic observation or the unpredictability of the occurrence of malolactic fermentation in some wines (Fleet 2003; Alexandre et al. 2004). Further, detailed investigations are needed to examine the potential for A. pullulans or Hormonema species to inhibit O. oeni. They may either kill the organism at the site of the grape surface, preventing its access to the winery ecosystem, or they may provide anti-O. oeni substances on the grape that transfer to the wine and interfere with successful growth of O. oeni during malolactic fermentation.

Blotechnologlcal applications

The Aureobasidium and related isolates examined here could represent a promising source of bioactive metabolites. Such compounds could find potential clinical utility in combating opportunistic mycoses (Takesako et al. 1991; Pelaez et al 2000). The isolates (cells or their metabolites) may also prove useful in controlling food spoilage organisms, including Zygosaccharomyces bailii and Dekkera bruxellensis (Table 4,7). Recently, Leverentz et al (2006) reported the use of Discosphaerina fagi (related to A. pullulans) to effectively reduce Listeria monocytogenes and Salmonella enterica serovar Poona on freshly cut apple surfaces. The inhibition of several species of foodborne pathogenic bacteria, Acetinobacter spp.. Listeria monocytogenes. Staphylococcus aureus and Bacillus cereus by many Aureobasidium/Hormonema isolates was observed by in vitro plate assays (Table 4.9), and these isolates may also fmd applications in food to control growth of foodborne pathogenic bacteria. The mechanisms by which ihQ Aureobasidium/Hormonema organisms inhibit spoilage yeasts and various pathogenic bacteria require investigation as they could form the basis of novel control initiatives.

Extracellular enzymes

Occurrence Protein Except for two isolates, all of A. pullulans hydrolysed casein and gelatin. The results are in accord with Aheam et al (1968), Dennis and Buhagiar (1973), Cenakova et al (1980), Federici (1982) and de Hoog and Yurlova (1994) who also found that proteolytic activity is widely distributed among strains of this species, but noted that strain variability in this property does exist.

Lipids In agreement with Federici (1982), lipolytic activity, assayed using Tween substrates, was widespread among isolates of A. pullulans. Triacylglycerol (4-22%) or tributyrin (20%) substrates, however, were less frequently hydrolysed by this species (Cemakova et al 1980, Buzzini and Martini 2002).

Cutln Although p-nitrophenyl esters (pN butyrate or pN caproate) are model substrates for cutinase, pNB esterase activity did not parallel that of polycaprolactone depolymerase activity. This discrepancy was found in isolates of A. pullulans and Hormonema sp., and could indicate differences in substrate spécificités for cutinase (Murphy et al 1996). Cutinase activity was previously reported in A. pullulans when assayed with pNB by Matteson-Heidenreich et al. (1997) and Gildemacher et al (2004), and in species oiHormonema (Middelhoven 1997). The degradation of polycaprolactone by pullulans {Pullularia pullulans) was also described by Fields et al (1974).

Carbon A. pullulans and the Hormonema sp. degraded a variety of carbon sources. Most isolates were able to attack pectin, xylan, CMC and cellobiose, but maceration of filter paper and degradation of lignin was not evident in any isolate. In agreement with studies by Cernakova et al (1980), Federici (1982), Biely and Slavikova (1994), Augustin (2000) and Buzzini and Martini (2002), pectin depolymerase activity is widely demonstrated in A. pullulans. While pectin methyl esterase was not screened here, it was observed in a large proportion (55-100%) of isolates by Cernakova et al (1980) and Augustin (2000). Xylanase activity has also been detected in^. pullulans (Leathers 1986, Augustin 2000).

The ability to hydrolyse cellulose seems to be a variable property in A. pullulans, and its measurement could depend on the form of the substrate, strain or basal media. Studies in which crystalline cellulose or filter paper was used as substrate, generally gave negative results for this species (Denis 1972; Horvath et al 1976; Federici 1982; Leathers 1986). Except for studies by Federici (1982) and Buzzini and Martini (2000), cellulase activity was frequently demonstrated in a large proportion of isolates when CMC, cellulose azure and cellobiose were used as substrates in the assay (Cernakova et al 1980; Augustin 2000; Kudanga and Mwenje 2005; Schultz and Thorman 2005). The presence of endo p glucanase (CMCase) and p-glucosidase (cellobiase) suggests that the remaining component of the cellulase system (exo p glucanase) may be lacking or unstable under the assay conditions used. Kudanga and Muenje (2005) detected exo p glucanase in several isolates of^i. pullulans, but suggested that its activity may be rate- limiting.

Mass loss studies indicated that A. pullulans degrades minor quantities or no lignocellulose; Frankland (1969) reported the reduction of 7% cellulose and 3-4% hemicellulose, but no loss of lignin in 10 weeks. This study supports the observation that isolates of .4. pullulans could not breakdown the lignin structural analogue and filter paper (crystalline cellulose).

Although lignolytic activity was absent, phenol oxidizing ability was detected in all isolates of pullulans on agar media. Polyphenol oxidase activity was a character frequently encountered in A. pullulans, and is in agreement with reports of tyrosinase (monophenol monooxygenase EC 1.14.18.1) and laccase (EC 1.10.3.2) activities, observed by Cemakova et al (1980) Deshpande et al (1992).

The presence of polyphenol oxidase in A. pullulans supported by ability of this species to utilize products of lignin photo- or acid degradation and oxidation in wood (Chantonet et al 1994; Shoeman and Dickinson 1996). This species has been found to metabolise a large range of lignin-related, aromatic and phenolic compounds, including tannic acid, benzoate, p hydroxy benzoate, ferulic acid, syringaldehyde, vanillin, catechol and protocatechuic acid (Horvath et al 1968; Bourbonnais and Paice 1987; Cemakova et al 1980). This study adds gallic acid (3,4,5, tri hydroxybenzoic acid) to this list. The species additionally produces a range of auxiliary enzymes which act synergistically to hydrolyse hemicellulose and lignocellulose. These include xylosidases, a-L-arabinofuranosidase, acetyl esterase, a-glucuronidase and ferulic acid esterase (Cemakova et al 1980; Myburgh et al 1991; Saha and Bothast 1998; Augustin 2000; Rumbold et al 2003). The involvement of these enzymes in the biodégradation of lignin has been recognized (Mayer and Staples 2002; Rumbold et al 2003; Burke and Cairney 2002).

Ecological role The production of a wide array of extracellular hydrolytic enzymes by A. pullulans and Hormonema sp. may have important implications for their nutrition on the phyllosphere, and their interactions with the grapevine host. Schultz and Thorman (2005) express caution in extrapolating the data obtained under in vitro conditions, as such enzyme profiles represent potentials rather than field activity. It is anticipated that future studies using novel apporaches that combine phylogenetics, DNA microarrays, functional genomics and in situ activity measurements will provide better perpectives on their interaction with and function on the grape surface.

Enzymes as virulence/pathogenic factors Aerial plant surfaces are covered by cutin and cuticular waxes, which is the interface where interactions with microorganisms occur. Cutinases are a class of serine esterases, that hydrolyse fatty acid ester bonds in the cutin polymer and a large variety of synthetic esters. These enzymes show activity on short and long chains of emulsified triglycerides and hydrolyse fatty acid esters and emulsified triglycerides as efficiently as lipases, but without showing the enhancement of activity in the presence of a lipid-water interface known of true lipases (Müller and Riederer 2005). They are secreted by several phylloplane microorganisms including plant pathogens, to acquire sources of carbon during their saprophytic growth. The activity of both cutinases and lipases is also believed to play a crucial role in the penetration of the host cuticular barrier during the initial stages of pathogenesis. More recently, the involvement of these enzymes in fungal spore adhesion, surface recognition and the differentiation of fungal infection structures have been described (reviewed in Schäfer 1993; Beattie 2002). Cell wall degrading enzymes (CWDE) such as pectic enzymes, cellulases and xylanases are implicated in the later stages of fungal invasion of plant tissues. These enzymes facilitate further penetration and spread by weakening structural polymers in the primary cell walls and middle lamellae of plants. Although less studied, proteases have also been implicated as a virulence factor in several fungal and bacterial plant pathogens (reviewed in Walton 1994; Annis and Goodwin 1997).

Several studies have associated the cutinolytic, protease and cell wall degrading potential of A. pullulans with their ability to cause diseases in fruits and plants (in Cook 1959; Frankland 1969, Matteson-Heidenreich et al 1997; Gildemacher et al. 2004; Kudanga and Muwenje 2005). However, cutinases and cell wall degrading enzymes may also participate in carbon acquisition for saprophytic growth. Indeed, these enzymes are also produced by many phylloplane yeasts not known to be pathogenic, for example D. hansenii, K. apiculata, M. pulcherrima, Rhodotorula spp. and Cryptococcus spp. (Biely and Sláviková 1994; Middelhoven 1997; Strauss et al 2001; Buzzini and Martini 2002; Ganga and Martinez 2004).

Currently, some contradictions about the role of enzymes in plant pathogenesis are apparent. Strong evidence exists to support their involvement in infection, but many other studies suggest that these enzymes are expendable for disease development, with other mechanisms being responsible. The reviews by Collmer and Keen (1986), Annis and Goodwin (1997) and Walton (1994) have discussed these complexities and difficulties associated with assessing the role of enzymes in pathogenesis.

In this study, the enzymatic profiles of A. pullulans and Hormonema sp. did not clarify their interaction with their host, or potential to cause disease. As with the study by Guerzoni and Marchetti (1987), the production of cutinase, lipase, protease, or CWDE of isolates could not be related to their incidence on damaged berries. Nevertheless, the ability to synthesize a wide range of enzymes enables these species to colonize many substrates and exploit resources that become temporarily available (St. Leger et al. 1997; Schultz and Thorman 2005). This versatility accords with the diverse lifestyle and opportunistic behaviour known of A. pullulans (Cooke 1959; Domsch et al 1980).

Role of enzymes in biodégradation Species of Aureobasidium and Hormonema are colonizers of wood and wood products such as logs, lumber, wood chips and veneer (Shoeman and Dickinson 1996, Croan 2000). The frequent association of A. pullulans in litter and decaying plant material or plant detritus is well known (Hermanides-Nijhof 1977; Cooke 1962; Domsch et al 1980; Sterflinger 2006). Frankland (1969) provides anatomical evidence of hyphal penetration into lignified xylem of bracken by pullulans. Penetration into sapwood of conifers, epithelial cells and invasion of ray parenchyma cells by their pigmented hyphae causes discolouration, known as sapstain (Croan 2000).

However, the structural components of plant and wood (lignin and cellulose) appear to be largely inaccessible to^. pullulans. Their growth and metabolic activity substantially reduces wood extractives and simple, soluble sugars from plants. For example, a 70% reduction in wood extractives was noted in pine wood treated with A. pullulans (Croan 2000). Both plant litter and wood are characteristically rich in phenolic and aromatic compounds which may serve as sources of carbon or energy for A. pullulans (Horvath et al. 1968, Schoeman and Dickinson 1996). The metabolism of these compounds are likely to involve polyphenol oxidases, also known as oxido-reductase enzymes (Bourbonnais and Paice 1987; Chantonet et al 1994).

Polyphenol oxidase activity could facilitate the colonization of A. pullulans in the grapevine, by providing it resistance against plant defence compounds. The detoxification of tannins (constituitive barriers) and resveratrol (induced phytoalexins) by B. cinerea is attributed to phenol oxidases (laccase) (Jeandet et al. 2002). Additionally, phenolic degradation may enable A. pullulans to sequester other organic compounds (proteins or polysaccharides) that are complexed with polyphenols (Bending and Read 1997).

Impact on wine quality and potential applications in winemaking The enzymatic activities of^. pullulans and Hormonema sp. may affect wine quality by directly altering the composition of juice and must. Their hydrolytic activities may indirectly affect wine fermentation by releasing substances, which encourage or inhibit S. cerevisiae and MLF bacteria. Lastly, their enzymes may be useful as processing aids in various aspects of winemaking.

Proteinase Proteolytic activity of wine-associated microorganisms are important because peptides and amino acids liberated during growth or autolysis may serve as nitrogen sources or growth substrates for S. cerevisiae or O. oeni (reviewed in Viljoen 2006). Lagace and Bisson (1990) first showed that extracellular proteolytic activity by wine yeasts may be exploited to alleviate haziness in wines that are caused by the precipitation of unstable proteins originating from grapes. The prominent sources of haze-forming proteins have recently been identified to be grapevine pathogenisis-related (PR) proteins, which include chitinases (CH) and thaumatin-like proteins (TL). The disappearance of foam- causing proteins (TL, CHV5) in wines may be linked to the acid proteases secreted by wine yeast and bacteria (Manteau et al. 2003) or from the growth B. cinerea on grapes (Marchai ^/fl/. 1998). Lagace and Bisson (1990) and Dizy and Bisson (2000) reported that while some protease preparations of wine yeasts successfully reduced protein-induced haze, the severity of haze was increased in others. Further studies are needed to determine the susceptibility of PR proteins to the proteases ofJ. pullulans and Hormonema sp., as well as its effect on the protein stability of wines.

Lipase, fatty acid esterase Lipolytic activity may impact on the aroma profile of wines through the release of volatile fatty acids, and other volatile compounds (esters, ketones, aldehydes) derived from fatty acids (reviewed in Matthews et al. 2004). Free fatty acids (decanoic and octanoic acids) liberated by lipid metabolism may be inhibitory to S. cerevisiae and O. oeni (reviewed in Viljoen 2006). Additionally, lipolysis could result in the removal of growth or survival factors (sterols, oleanolic acid) necessary to sustain S. cerevisiae during alcoholic fermentation.

Polysacchande-degrading enzymes Pectins, hemicellulose (xylans) and cellulose (glucans) present in grapes become increasingly insoluble in wines and impede subsequent filtration and clarification steps, during which, it is estimated that as much as 20% of product can be lost (reviewed in Canal-Llauberes 1993). Polysaccharide degrading enzymes are used primarily to aid the removal of polysaccharide colloidal particles from wines and to aid the release of flavour precursors. Efforts to reduce product loss and processing costs have motivated attempts to genetically engineer wine yeast to inherently express xylanase, glucanase, cellobiase, polygalacturonase and L-arabinofuranase activities (reviewed in Verstrepen et al. 2006). Polysaccharide degradation is widely displayed in A. pullulans and Hormonema sp., shown by the variety of enzymatic systems and their wide substrate activity. As a result, there has been considerable interest in exploiting A. pullulans in solubilising agricultural residues and plant biomass into fermentable sugars, and its conversion into value added products such as pullulan and fuel ethanol (Saha and Bothast 1999; Leathers 2003a, 2003b). To further this cause, molecular and genetic characterisation of several polysaccharase systems in this species are being actively researched (see for example, Tanaka et al 2006). Therefore, these species could be a good source of both enzymatic preparations and genes for winemaking.

p-glucosidase The flavour and aroma compounds in grapes can occur as free volatiles or as glycosidically bound non-volatile precursors. The sugar-bound components are frequently more common than the free volatile fractions and, therefore, represent important sources of flavour and aroma in wines. In Muscat and Riesling grape cultivars, this ratio ranges between 1-5 times, and up to 15 times more abundant in the Gewürtztraminer variety (Gunata et al. 1988). Glycosides that contain flavour aglycones (monoterpenes, norisoprenoids, aliphatic residues) are cleaved from the glycosidic bound forms and converted into free volatiles by the action of glycosidase enzymes. Preparations of p-glucosidases from Debaromyces spp. have been shown to have a major impact on the sensory profiles of wines, including the enhancement of varietal characters of grapes and wines (Belancic et al 2003; Villena et al 2006). Glycosidases may be useful in the making of blush or róse wines (Sánchez-Torres et al 1998), but glycosidase activity on monoglucoside anthocyanins may induce loss of wine colour, causing an unfavourable effect in red wines (Manzanares et al 2000). The occurrence of glycosides in S. cerevisiae appears to be rare, and greater activity has been found in non-Saccharomyces yeasts, including D. hansenii, Hs. tivarum, M. pulcherrima, P. anomala and T. delbrueckii (Rosi et al 1994, Charoenchai et al 1997, Mendes Ferreira et al 2001. Rodríguez et al 2004). The study by McMahon et al (1999) is particularly noteworthy in highlighting the potential role ofJ. pullulans in the flavour development of wines. In addition to p-glucosidase, arabinofuranosidase and rhamnopyranosidase activity was also detected in this species. Significantly, A. pullulans was shown to hydrolyse 68% of the glycosides of Voignier grapes (glucosyl- glucose), where the activity was lacking in other wine yeasts (Saccharomyces or non- Saccharomyces) and wine bacteria.

Polyphenol oxidase Phenolic compounds naturally present in grape (seed, skin and flesh) and must are a diverse group of molecules varying in structure, complexity and reactivity/stability. These compounds range from simple benzoic and cinnamic acid derivatives to more complex molecules such as flavonoids, anthocyanins and tamiins, and are responsible for many important wine attributes, including colour, clarity, flavour, aroma and astringency. Phenol oxidases find applications in wine processing for achieving stability, reducing astringency and flavour modifications. Treatment of wines with phenol oxidase aids in preventing the oxidative discolouration or browning reactions by the removal of phenolic substances (Matthews et al. 2004). Strains of pullulans and Hormonema sp. could be novel sources of enzymatic preparations of polyphenol oxidases. These enzymes may also be elaborated during growth of these organisms on berries and in must, and their potential impact in wine deserves more attention. The cany-over and effect of laccases in wines made from Botrytis cinerea-mÎQciQà grapes clearly demonstrate this potential; it causes the loss of antioxidants and cancer preventative agents such as epicatechin, quercetin, resveratrol in wines (Jeandet et al. 1995a; 1995b), browning of musts, and reduced astringency in sweet botrytised white wines (Ribéreau-Gayon et al. 2000; Canal-Llaubéres 1993).

Finally, many enzymes from non-Saccharomyces yeasts have been identified with potential application in wine making. However, their activities in wine may be limited because of poor enzymatic stability in wine conditions (glucose repression, inhibition by low pH, ethanol) or lack of substrate specificity. Further efforts are needed to characterise properties of the enzymes detected in grapes associated with Aureobasidium/Hormonema species; for example, to determine substrate specificities, kinetics and stability under fermentation parameters, so as to better understand and manage their impact in wine production. CHAPTER FIVE

A MICROTITRE PLATE DNA PROBE HYBRIDIZATION ASSAY FOR THE DETECTION OF WINE YEAST SPECIES

5-1. INTRODUCTION

Yeasts have a prominent role in the production of wines, ranging from their association with grapes, contributions to the alcoholic fermentation and to their potential to spoil the final product (Pretorius 2000; Fleet, 2001, 2003). About 15-20 different species are commonly associated with these grapes and wine ecosystems. The ability to monitor the presence of these species throughout the chain of wine production is a key element of quality assurance in modem wine production (Loureiro and Malfeito-Ferreira 2003).

Cultural procedures for the isolation and identification of yeasts from these sources are well established (Fleet 1993; Deak 2003; Kurtzman 2006), but they are too labour intensive and time consuming for routine use in quality assurance programs.

Commercially available miniaturized kits, (e.g. APÍ 20C AUX, VITEK Yeast

Biochemical Card, RapID Yeast Plus and Biolog systems) have been developed to overcome some of the labour and logistical challenges in this cultural-phenotypical approach to yeast identification, but such systems are mostly targeted to clinical yeasts and do not always give equivalent data for yeasts from industrial or environmental habitats (Deak and Beuchat 1993; Praphailong et al 1997; Sancho et al 2000; Arias et al 2002; Foschino et al 2004).

In recent years, molecular methods based on DNA analyses have been developed to facilitate the identification of yeasts, including wine yeasts (Loureiro and Malfeito-

Ferreira 2003). These methods include sequencing, specific PCR assays, restriction fragment polymorphism (RFLP), use of specific nucleic acid probes and real-time PCR.

The application, advantages and limitations of these methods have been reviewed (Beh et al 2006; Fernández-Espinar et al. 2006). Most wineries do not have the capability to implement sophisticated analytical techniques or to purchase expensive analytical instrumentation required of most molecular methods. However, they would have the skills and resources to apply basic PGR technologies.

Within the food and beverage industries, microtitre plate-ELISA techniques have been widely accepted for the detection of specific pathogenic and spoilage bacteria (Radcliffe and Holbrook 2000; McCarthy 2003; Cox and Fleet 2003). In particular, colourimetric, visual ELISAs are technically simple, convenient, robust and rapid. As yet, such ELISAs have not yet been developed for the routine detection of specific yeasts associated with foods and beverages. This is probably related to lack of sufficient variability and specificity in the immunoreactivity of their surface antigens. However, advances have been made in developing microtitre plate-DNA probe hybridization assays for yeasts, and these follow the simple microtitre plate-ELISA format. In this approach, DNA probes specific for certain yeasts are identified and synthesized. The probe is bound (adsorbed) to the well of a microtitre plate, DNA is extracted from the yeast to be identified, and specific regions that include sequences of the probe DNA, are then amplified by PCR. The primers for the PCR reaction are synthesized to contain covalently bound biotin (a tracer tag or reporter system), so that the resultant amplicons are biotinylated. These amplicons are transferred to the probe coated microtitre plate, where they are immobilized by specific binding (hybridization) to the probe. Next, strepavidin, which has been covalently conjugated to the enzyme, horse radish peroxidase, is added to the plate, and binds with the biotin of the immobilized amplicon. The bound enzyme conjugate is then detected by addition of enzyme substrate that gives a colourimetric reaction (Figure 5.1). A positive colourimetric reaction indicates presence of the target yeast species. This assay format has been successfully developed and applied to several yeast species, including environmental isolates and clinical yeasts (Fujita et al 1995; Elie et al 1998; Kiesling et al 2002a). Yeasts can be specifically identified in 8 hours, which is the time taken for DNA extraction, PCR amplification and hybridization (Kiesling et al. 2002a; Goodwin et al 2005).

In this chapter, a microtitre plate-DNA probe hybridization assay is designed and evaluated for four species of grape and wine associated yeasts, namely, Dekkera briaellensis, Hanseniaspora uvarum, Metschnikowia pulcherrima, and Zygosaccharomyces bailii. Attachment of DNA-probes to microplate wells

5'-CCGTGAATCGCTGGAG ACCGTTTTTTT. -3'

T-tailed species-specific probe

PCR amplification of target DNA

Biotinylated universal primers

Yeast genomic DNA Biotin-labeled target DNA

Hybridization of target DNA to capture probes and colourimetric detection

substrate + stop solution *

V

Figure 5.1 Protocol for the microtitre plate DNA hybridization assay 5.2 MATERIALS AND METHODS

5.2.1 Reference cultures

Reference species and strains (Table 5.1) were obtained from the following culture

collections; Agricultural Research Service Culture Collection, National Center for

Agricultural Utilization Research (NRRL), Peoria, USA; Food Science Australia (FRR),

CSIRO North Ryde, Australia; the Australian Wine Research Institute (AWRI), Glen

Osmond, Australia; the School of Biotechnology and Biomolecular Sciences,

University of New South Wales (UNSW), Sydney, Australia; the Collection of

Industrial Microorganisms (ZIM), University of Ljubljana, Slovenia; and

Centraalbureau voor Schimmelcultures (CBS), Utrecht, The Netherlands.

Table 5.1 Reference cultures used in this study Species Strain Dekkera bruxellensis AWRI 1205 Dekkera anomala AWRI 953 Brettanomyces custersianus AWRI 950 Hanseniaspora uvarum ZIM 2126 Hanseniaspora opuntiae ZIM 2137 Hanseniaspora guilliermondii ZIM 118 Metschnikowia pulcherrima NRRLY-7111 Metschnikowia fructicola NRRL Y-27328 Metschnikowia reukaufii NRRL Y-7112 Zygosaccharomyces bailii UNSW 507700 Zygosaccharomyces bailii AWRI 1462 Zygosaccharomyces bailii AWRI 1467 Zygosaccharomyces bailii AWRI 1471 Zygosaccharomyces bailii AWRI 1472 Zygosaccharomyces bisporus CBS 702 Zygosaccharomyces cidri CBS 4575 Zygosaccharomyces fermentati FRR 3666 Zygosaccharomyces florentinus CBS 746 Zygosaccharomyces kombuchaensis CBS 8849 Zygosaccharomyces mellis CBS 736 Zygosaccharomyces microellipsoides FRR 5434 Zygosaccharomyces mrakii CBS 4218 Zygosaccharomyces roiccii UNSW 709200 Candida stellata AWRI 1159 Pichia anomala AWRI 1051 Pichia membranifaciens AWRI 1148 Saccharomyces cerevisiae AWRI 1010 Saccharomyces bayanus CBS 380 Torulaspora delbrueckii CBS 1146 Lactobacillus plantarum UNSW 084800 5.2.2 Design and synthesis of species-specific capture probes The principles and criteria used to design and synthesize the DNA probes followed those reported by Kiesling et al (2002a), Goodwin et al (2005), Diaz and Fell (2005) and Page and Kurtzman (2005). Probes were designed for three grape and wine-related species, namely Hanseniaspora uvarum, Metschnikowia pulcherrima, and Zygosaccharomyces bailii. A probe sequence for Dekkera brwcellemis has already been reported (Stender et al 2001) and was modified by the addition of five bases to give a new probe sequence of 20 bp, as listed in Table 5.2.

Nucleotide sequences in the D1/D2 domain of the 26S rDNA of wine-related yeast species were obtained from the GenBank database (www.ncbi.nlm.nih.gov/). Comparison of sequences was performed using the multiple sequence alignment program Clustal W version 1.83 (Thompson et al. 1994). From the alignment, regions with sequence variability were selected for probe design. A BLAST search was performed to identify significant similarity of probe sequences with known species in the GenBank database (Altschul et al. 1990). Probe characteristics, melting temperature (Tm), and the potential of forming dimers or secondary structures were predicted with the programs Oligo vesrsion 4.0 (Molecular Biology Insights, Cascade, CO) and DINAMelt Server (Markham and Zucker 2005). The probe sequences for targeted species are listed in Table 5.2. The probes were synthesized and purified by reverse- HPLC by Sigma-Genosys, Sydney, Australia.

Table 5.2 Sequences of species-specific probes

Target species Probe Probe sequence (5' to 3') name

Dekkera bruxellensis Dbru CCG TGA ATC GCT GGA GAC CG

Hanseniaspora uvarum Huv TAG TGT TAG CCG GGA 11G AGG

Metschnikowia pulcherrima Mpul GGC CCT TAC TCC CAC ACC ACC

Zyi^osaccharomyces bailii Zba GAA CTT ATA GTC CAG GGG AAT 5.2.3 Attachment of probes to microtitre plates Previous studies have shown that the signal strength of the plate hybridization assay is significantly enhanced by linking about 200 bp of poly T to the 3' end of the probe (poly T-tails). This modification enables the specific capture sequences of the probe to be positioned above the plate surface, thereby enhancing their accessibility for hybridization (Kiesling et al 2002a, Goodwin et al. 2005). The terminal deoxynucleotidyl transferase (Tdt) was used to add poly T-tails to the probe; Tdt enzyme (4000U; Roche Applied Science, Indianapolis, IN) was added to a final volume of 210 sterile distilled water, 1.2X Tdt buffer, 4.6 mM dTTP (Roche Applied Science, Mannheim, Germany), 2.5 nmol of oligonucleotide probe and 0.25 mg/ml inorganic pyrophosphatase (Sigma-Aldrich, St. Louis, MO). The mixture was incubated at 3TC for 4 h. Following incubation, 42.5 EDTA (0.5 M, pH 8.0) and 2.25 ml of IX TE buffer (pH 7.6) were added, giving a final probe concentration of 1 pmol/^l. To immobilize probes, 1 pmol of a T-tailed probe in 100 of IX TE (pH 7.6), was added into each well of the microtitre plate (polystyrene, amine surface, Costar, Coming, NY). The plates were incubated in a drying oven at 3TC overnight or until dry. Blocking buffer (5x Denhardts, Ix PBS, 200 ^il/well) was added to neutralise non- specific binding sites. The plates were incubated at room temperature for 2 h, and the blocking buffer was discarded. Plates were air-dried for 1-2 h, sealed in a pouch with desiccant and stored at 4°C.

5.2.4 Extraction and preparation of DNA for microtitre plate- hybridization assay

Preparation of genomic DNA from yeast cultures followed the methods described by Kurtzman and Robnett (1998). Yeasts were grown in YEPD broth on an orbital shaker (200 rpm), at 25°C for 24 hours. Culture (1 ml) was centrifuged at 10000 g for 2 min. Pelleted cells were resuspended in 200 |il of breaking buffer (2% Triton X-100, 1% SDS, 100 mM NaCl, 10 mM Tris, 1 mM EDTA, pH 8) and homogenized with glass beads (0.5 mm diameter) in the presence of 200 ^il of phenol/chloroform/isoamyl alcohol (50:48:2). Tris-EDTA buffer (IX; 10 mM Tris, 1 mM EDTA, pH 7.6) was added to the disrupted cells and the suspension was centrifuged at 14000 g for 10 min at 4°C. After collecting the supernatant, 2.5 volumes of absolute ethanol were added to precipitate the DNA. The DNA precipitate was centrifuged at 14000 g for 10 min at 4°C, washed with 70% ethanol, then resuspended in 50 of TE buffer.

In several experiments, DNA templates were prepared from yeast isolates by a simple boiling method, adapted from de Barros Lopes et al (1996). Yeast colonies grown on YEPD plates (24-48 h) were suspended in 30 ^il of distilled water and held at 95°C for 20 min in a thermocycler (Applied Biosystems). Cell lysates were kept on ice or stored at -20°C, from which 4-5 were used in PGR amplification reactions, at a final volume of 50 ^il.

In some experiments, DNA was extracted from yeast cells suspended in wines. Cells of Zygosaccharomyces bailii strain UNSW 507700 were grown in YEPD broth until late exponential phase (24 h). The culture was adjusted to a cell density of 10^ cells/ml, and was serially diluted in wines, (pre-sterilised by filtration through a 0.45 jim membrane filter, Millipore, Bedford MA) to yield final cell concentrations of 10^-10^ cells/ml. The wines used were red (Shiraz) and white (Chardonnay) wines of Rosemount Estate, Australia. Yeast cells were harvested from the wines by centrifugation (1 ml, 12,000 g, 10 min 4°C). Cell pellets were washed twice with distilled water and DNA extraction followed the same protocol as described previously. DNA extracts were further purified by treatment with DNeasy Plant minikit (Qiagen, Clifton Hill, Australia).

5.2.5 PGR conditions to produce biotinylated ampiicons Amplification of the D1/D2 region of the yeast 26S rDNA was performed with biotin- labeled universal primers NLl (5' GCATATCAATAAGCGGAGGAAAAG 3') and NL4 (5' GGTCCGTGTTTCAAGACGG 3'), purchased from Sigma-Genosys, Sydney, Australia. Ampiicons were about 600 bp in length and contained the species-specific probe sequences. The PCR reaction contained Ix PCR Buffer (10 mM Tris-HCl pH 8.3, 50 mM KCl), 0.2 [iM of each primer, 200 ^iM of each dNTP (Roche Diagnostics, Indianapolis, IN), 1.5 mM MgCb, 1.25 U of Gold Taq DNA polymerase (Roche Molecular Systems, Brachburg, NJ) and 10 ng of extracted DNA in 50 pi final volume. PCR amplifications were performed in a 9600 Thermal Cycler (Applied Biosystems) with the following cycling program: initial denaturation at 95°C for 5 min, 36 cycles of denaturation at 95°C for 1 min, primer annealing at 52°C for 1 min, and extension at 72°C for 2 min, followed by final extension at 72°C for 10 min. Successful amplification was confirmed by electrophoretic analysis of PGR products on 1.4% agarose gels.

5.2.6 Plate hybridization assay Detection of biotinylated target amplicons was performed according to procedures described by Kiesling et al (2002a) and Goodwin et al (2005). PGR amplicons (10 ^1) were loaded into each well of the microtitre plate containing immobilized capture probe. Denaturation solution (5 |il; 0.1 M NaOH) was added to each well and the plate incubated at room temperature for 2 min. Following denaturation, 85 |il of hybridization buffer (0.45 M phosphate buffer, pH 6.8, 0.1% SDS, 5X Denhardts) were added to the wells, and plate incubated at 60°G for 30min. Wells were washed three times at room temperature with 200 of IX wash buffer (0.5X SSG, 0.1% Tween 20, pH 7.0). The plates were then incubated with high stringency buffer (3 M TMAC, 2 mM EDTA, 50 mM Tris pH 7.5) at 60°G for 20 min. After incubation, the plate was washed three times at room temperature with IX wash buffer. Strepavidin-HRP conjugate (Roche Applied Science, Penzberg, Germany) was diluted 1:5000 in conjugate buffer (5X SSG, 0.1% SDS, 5X Denhardts) to yield 0.5 U/jil and 100 |il of this diluted conjugate was added to each well. Plates were incubated at 37°G for 10 min and subsequently washed three times with IX wash buffer. Golourimetric HRP substrate (TMB Microwell Peroxidase Substrate, KPL, Gaithersburg, MD; 100 was added to each well. Plates were sealed and incubated at 37°G for 5 min. The reaction was stopped with the addition of 0.5 M HGl (100 |il). Yellow colour resulting from positive reactions was assessed visually, and the absorbance read at 450 nm with a plate reader (Beckman Goulter, Fullerton, GA).

Positive and negative controls were conducted to assess the success and specificity of PGR and hybridization reactions. DNA amplified from a pure culture of the target yeast was taken as the positive control. Two negative controls were included in hybridization assays; a PGR negative control, which contained all PGR reagents, but no DNA template, and a chemistry blank, that contained no PGR materials, but Strepavidin-HRP and substrate were added. Replicate wells were set up for every assay to verify the reliability and reproducibility of the procedure.

Acceptable performance criteria for the assay were defined according to Kiesling et al (2002a), where absorbancy of 1.0 or greater was taken as the positive result. Non-target and background signals were approximately 10% of the positive signal. Positive identification was also indicated by the development of a visible, yellow colour in microplate wells.

5-3 RESULTS

5.3.1 Design of probes

Alignment of the sequences of the D1/D2 region of 26S rDNA was performed for Hanseniaspora uvarum, Metschnikowia pulcherrima, and Zygosaccharomyces bailii. The comparative analyses identified regions containing at least two base pair differences from other yeast species within the genus, and from other species occurring in the grape/wine environment. Probe sequences for Dekkera brtaellensis were adapted from the literature (Stender et al 2001). The probe sequences and names for the probes are given in Table 5.2 (Materials and Methods).

To facilitate hybridization of all probes at a single stringency, the oligonucleotide probes were selected to be of uniform length, of 20-21 bp, and the melting temperature of chosen probes usually ranged between 60-70°C (Tm, %GC method). In the hybridization assay, differential stabilities of A:T versus G:C base pairs in the probe sequence are ameliorated in the presence of TMAC buffer. This condition enables the design of probe sequences without the need to consider A+T and G+C proportions Hybridization of multiple probes under such conditions occurs as a function of probe length, rather than base composition (Bains 1994; Yang et al 2001; Diaz and Fell 2005). Probe sequences contained at least one sequence difference located centrally within the probe sequence. According to Kiesling et al (2002a), hybridization probes give optimum performance when base pair difference are located in the centre, rather than at the end of the probe sequence. Probe sequences showing runs of three or more guanines (Gs) or cytosines (Cs) were avoided. Sequences that showed complimentarity to universal primers and the potential to form secondary structures likely to hinder probe hybridization, were also avoided.

The Z bailii probe, however, contained four consecutive guanines at the 3' end. Additionally, the potential to form hairpin loops involving 2-3 bp were predicted from its sequence (Figure 5.2).

5.3.2 Elimination of background and non-specific interactions In the initial stages of evaluating the probe-hybridization assay, high background signals (absorbancies) were encountered with probes for all four species. Potential causes of these background signals were investigated by conducting a series of assays, in which each PGR component and plate hybridization assay reagents were systematically eliminated.

Probe hybridization assays conducted with chemistry blank wells (no PGR samples) gave negligible signals (absorbance values 0.061 ±0.0079). This control indicated that background absorbance in the assay was not caused by non-specific microplate binding of Strepavidin-HRP or the HRP substrate.

The PGR samples added to plate wells contain biotinylated amplicons and also excess biotinylated primers, both of which could bind non-specifically to the plate surface. To differentiate whether or not the amplicon or excess biotin-labeled primers were binding non-specifically to plate wells, the PGR reactions were conducted with DNA template (PGR positive control), and without the addition of DNA template (PGR negative controls). Figure 5.3 A shows the absorbance values for the assay of a PGR positive control and a PGR negative control for each of the four species, tested according to the described protocol (Materials and Methods). Addition of PGR positive controls to the plates and subsequent hybridization assay gave absorbances of 1.5-2.9, depending on the yeast species being tested. PGR negative controls also gave strong background absorbances of 1.0-1.3, and suggested that excess biotinylated primers might be significant in non-specific binding to the plate. Microplate wells were treated with a a —

/ a ^ t 20

g - g

\ a 10 I a ^ a / ^ 20 . - a^ \ / a ^ ^ \ a ^ ' I 3' a ^ . / 5'

10 t - a

/ t a-« I / c 20 a ~

5'

Figure 5-1 Folding structures of the probe sequence for Zygosaccharomyces bailii, as predicted with the Quikfold program of DNAMelt Server blocking agent (Denhardts-PBS) according to the standard protocol. It was expected, therefore, that this would prevent non-specific binding of assay components.

Several aspects of the hybridization protocol were modified and investigated to overcome the interfering background. These included (i) modifying the microplate surface for probe immobilization and capture, (ii) modifying the blocking agent and (iii) removal of excess biotin-labeled primers (Figure 5.3).

<

0.0 Huv Dbru Mpul Zba Dbru Huv Mpul Zba

E 5 1.5 £ 1.5

" 1.0

5< 0.5

ijJxjJDbru Huv Mpul Zba . Dbru Huv Mpul Zba

E 5 lo

<

0.0 ^JLlDbru Huv Mpul LZba

Figure 5.3 Microtitre plate DNA-probe hybridization assays for four yeast species. Dbru, Dekkera brtaellensis; Huv, Hanseniaspora uvarum; Mpul, Metschnikowia pulcherrima; Zba, Zygosaccharomyces bailii. Positive signals (•) and PGR negative control signals (•). Assays with aminated plate surface (A) and non-aminated plate surface (B); the additional blocking agents, bovine albumin (C) and skim milk (D); PGR samples with PGR amplifications using biotinylated reverse primer and non- biotinylated forward primer (E), and with removal of excess biotinylated primers by a clean up kit (F). Hybridization assays for all probes were performed at 50°G. Data represent the average of triplicate experiments. 5.3.2.1 Plate surface chemistry

Figures 5.3A and 5.3B show the hybridization signals obtained when assays were carried out in plate wells with aminated and non-aminated surfaces (Coming Costar,

Corning, NY). Two conclusions were evident from these trials. First, non-aminated plates gave much lower positive signals than aminated plates (about 0.13-0.20, Figure

5.3B, compared with 1.51-2.93, Figure 5.3A). Background signals (PCR negative controls) for non-aminated surface were also much lower than the aminated plate surface. However, the ratio of absorbance values for positive and negative control assays were little different (2.06 and 1.89 for non-aminated and aminated plates, respectively).

5.3.2.2 Blocking procedures

Bovine albumin (Sigma, St. Louis, MO) and non-fat skim milk (Green and Gold,

Australia) were tested as additional blocking agents. Each compound was added at 1%

(w/v) to the hybridization buffer. The use of bovine albumin, partially reduced background signals to 0.38±0.05 (Figure 5.3C and Figure 5.4), improving the signal to noise ratio to 2.7. Skim milk was a more efficient blocking reagent, as shown in Figure

5.3D. The background absorbance was decreased to levels comparable to that of chemistry blanks (0.07±0.01, signal to noise ratio of 3.4). However, at the concentration of 1%, the use of skim milk in the blocking procedure also caused quenching of positive signals (mean absorbance of 0.22±0.05). The assay might be further improved if skim milk was applied at lower concentrations (e.g. 0.2%, 0.5%).

5.3.2.3 Removal of excess biotin-labeled primers

All four species-specific probes were designed in the forward orientation. The detection of probe hybrids would therefore require the presence of a biotin tag on the strand in reverse orientation, to which the probe is complimentary. Figure 5.3E shows the signal and noise intensities when PCR amplification reactions were performed using an unlabeled forward primer and a biotin-labeled reverse primer. Background levels were partially reduced by this approach (mean values of 0.77±0.10, signal to noise ratio of

2.2). Finally, excess biotinylated primers from PCR samples were removed by passage through a commercial PCR purification kit, QIAquick (Qiagen, Hilden Germany). This approach effectively removed background intensities (0.10±0.02) and improved the signal to noise ratio to 14.5 (Figure 5.3F and Figure 5.4). It also provided evidence that excess primers caused the background absorbance by binding non-specifically to microplate wells.

The optimized conditions for conducting the microtitre plate hybridization assay involved PCR amplification with a single biotin-labeled primer, removal of excess PCR biotin-labeled primers through a clean-up column, and the addition of blocking agent (1% BSA) to the hybridization buffer. All subsequent experiments were carried out with these modifications to the standard probe hybridization protocol.

1% BSA

PCR purification

Figure 5-4 Microtitre plate-hybridization assay with the Z bailii probe; removal of background absorbance with additional 1% BSA blocking agent and PCR purification procedures. PCR amplicons with target DNA (Z bailii UNSW 50700) were loaded into wells of rows C, F. The wells of rows D, G contained PCR negative controls (no DNA template). The wells in rows E and H contained no PCR amplicons (chemistry blanks).

5.3.3 Evalutation of probe specificities in the microtitre plate liybridization assay

The optimized microliter plate assay was applied to each of the target species, D. bruxellensis, Hs. uvarum, M. pulcherrima and Z bailii, and also to two other species closely related to the target species. Moreover, to investigate the effect of hybridization temperature on the assay, reactions were conducted at 50°C, 55°C and 60°C (Figure 5.5). In all cases, the target species gave the strongest assay (absorbance) signal, with the highest signals being obtained for hybridization assays at 50°C, and least at 60°C. Interestingly, the absorbance signal varied with the yeast species; Z bailii gave the highest absorbance (1.60) compared with D. bruxellensis that gave the lowest absorbance (0.35) for the 50°C assay.

Dbru 1.0

0I 0.8 m:o. 6 u 1 0.4

$0.s 2

0.0 50C 55C 60C 50C 55C 60C ID. bruxellensis OD. anomala 16. custersianus \Hs. uvarum DHs. opuntlae \Hs. guillermondii

50C 55C 60C 60C IM. pulcherrima DM. fhjcticola lAf. reukaufii OZ. bisporus IZ. rouxii

Figure 5.5 Microtitre plate-DNA probe hybrdization assay; the reactivities of probes for D. bruxellensis (Dbru), Hs. uvarum (Huv), M. pulcherrima (Mpul) and Z. bailii (Zba) with corresponding target and two closely related, non-target species. Assays were performed at hybridization temperatures of 50°C, 55°C and 60°C. Data represent the mean absorbance units for 3 replicate wells and include mean background signals for four probes (0.070).

The microtitre plates also gave absorbance signals for species closely related to the target yeasts. The signals for D. anomala and B. custersianus were about 30-40% of that for D. bruxellensis. Hybridization of the Hs. uvarum probe (Huv) with Hs. opuntiae and Hs. guilliermondii at 50°C produced signals about 50% of that obtained with its target, Hs. uvarum. The Huv probe failed to discriminate against Hs. opuntiae and Hs. guilliermondii at 55°C and 60°C hybridization temperatures. The M. pulcherrima probe cross-hybridized with DNA from M. fructicola and M. reukaufii at 50°C. At the hybridization temperature of 55°C and 60°C, these non-target species also gave signals about 30% of those with the probe target, M. pulcherhma. The Z bailii probe, operated at 55°C gave an acceptable, specific assay. The related species Z bisporus and Z roiaii gave signals that were 10 fold less than that for the target, Z bailii. However, the specificity of the assay was lost for hybridization temperatures at 50°C and 60°C (Figure 5.5).

5.3.4 Further evaluation of the Z. baiUi microplate hybridization assay

Microtitre plates coated with the Z bailii probe were further evaluated in assays against DNA extracted from other Zygosaccharomyces species and from a diversity of other yeast species commonly found in grape and wine ecosystems. Two wine associated bacteria were also included in the assay (Table 5.3). Performance criteria for the assay were taken as (i) positive assay signal gives absorbance greater than 1.0, and (ii) absorbance signals for non-target DNA to be 10 fold less than that of a positive assay. An absorbance value of 1.0 or more gives a strong visibly clear yellow colour in the microtitre wells, whereas a negative reaction gives clear, colourless wells. Figure 5.4 shows a picture of the microtitre plate for these reactions.

All seven strains of Z bailii gave strong assay signals, with absorbance values in the range 1.0-1.4 (Table 5.3). Nine other species of Zygosaccharomyces gave negative reactions, with absorbance values of 0.1 or less. All other yeast species, as well as two bacteria, gave absorbance values less than 0.1. It was not necessary to conduct a detailed extraction (Kurtzman and Robnett 1998) of the DNA from yeast isolates in order to obtain a strong assay signal. For several isolates and trials, DNA was released from cells of Z bailii by a simple heating step adapted from de Barros Lopes et al. (1996) (see Materials and Methods). Assays with the crude DNA preparations gave absorbance values greater than 1.0 (Table 5.3). Table 5.3 Specificity of Z hailii probe microtitre plate assay against DNA from Zygosaccharomyces yeasts and other species associated with grapes and wine Data represent the mean absorbance values of nine samples.

Species Mean A450 nm ± SD Zygosaccharomyces bailii UNSW 507700 1.19±0.06 Zygosaccharomyces bailii UNSW 507700, crude cell extract 1.29 ±0.07 Zygosaccharomyces bailii AWRI 1462 1.37 ±0.07 Zygosaccharomyces /JO////AWRI 1467 1.25 ±0.05 Zygosaccharomyces bailii AWRI 1471 1.44 ±0.03 Zygosaccharomyces bailii AW^ 1472 1.40 ±0.20 Zygosaccharomyces bailii wine isolate LWS, crude cell extract 1.04 ±0.03 Zygosaccharomyces bailii grape isolate El 12, crude cell extract 1.15±0.10 Zygosaccharomyces bispoms CBS 702 0.11 ±0.04 Zygosaccharomyces cidri CBS 4575 0.11 ±0.00 Zygosaccharomyces fermentati FRR 3666 0.08 ± 0.05 Zygosaccharomyces florentiniis CBS 746 0.09 ± 0.02 Zygosaccharomyces kombuchaensis CBS 8849 0.08 ± 0.03 Zygosaccharomyces mellis CBS 736 0.09 ±0.01 Zygosaccharomyces microellipsoides FRR 5434 0.07 ± 0.02 Zygosaccharomyces mrakii CBS 4218 0.08 ±0.01 Zygosaccharomyces rouxii UNSW 709200 0.10 ±0.02 Candida stellata 1159 0.06 ±0.01 Cryptococcus heveanensis grape isolate E315, crude cell extract 0.08 ±0.01 Debaryomyces hansenii grape isolate E282, crude cell extract 0.06 ± 0.03 Dekkera bruxellensis AWRI 1102 0.06 ± 0.00 Hanseniaspora uvarum CBS 314 0.06 ± 0.00 Kluyveromyces thermotolerans isolate El50, crude cell extract 0.08 ±0.01 Metschnikowia pulcherrima NRRL Y-7111 0.09 ± 0.03 Pichia anomala 1051 0.07 ± 0.02 Pichia membranifaciens AWRI 1148 0.06 ± 0.00 Rhodotorula glutinis grape isolate E318, crude cell extract 0.09 ± 0.03 Saccharomyces cerevisiae AWRI 1010 0.05 ±0.03 Saccharomyces bayanus CBS 380 0.08 ± 0.02 Schizosaccharomycespombe isolate BS24, crude cell extract 0.06 ± 0.00 Torulaspora delbrueckii CBS 1146 0.06 ± 0.00 Lactobacillusplantarum UNSW 084800 0.03 ±0.01 Oenococcus oeni CH35 0.02 ± 0.00 The mean absorbance of PGR negative controls (no template DNA) was 0.058±0.0034 (n=9).

Sequence analysis of both reference cultures and pre-characterized isolates of Z bailii

showed these strains shared 99-100% sequence homology with the Z bailii strain

NRRL-Y2227 (accession number U72161), from which the Zba probe sequence was designed. All strains of Z bailii examined showed no sequence heterogeneity (or intraspecific variation) within the 21 bp diagnostic region. 5.3.5 Application of microtitre plate hybridization assay to detect Z. baUn in wine samples

Zygosaccharomyces hailii UNSW 507700 was inoculated into samples of red and white wines to give fmal cell densities of 10^ to lO^/ml DNA was extracted from samples (1.0 ml) of wine as described in Materials and Methods, and amplified using the standard PCR protocol. Amplicon production was verified by electrophoresis on gels, and detection by the microtire plate probe assay. Figure 5.6 shows detection of the amplicons in agarose gels and absorbance signals obtained by microtitre plate assay. Amplicons were generated from wine samples containing 10 - 10 CFU/ml of Z hailii and these were clearly detected by the microtitre plate probe assay. These data show 10^ CFU/ml as the lower limit of detection by the microtitre plate probe assay.

Figure 5.6 Detection Z bailii in wine samples at 10^ - 10^ CFU/ml. (A) Analysis by microtire plate probe assay; mean background signals for duplicate samples were 0.065. (B) Analysis of DNA amplicons by gel electrophoresis. Lanes 1-6, wine samples with 10^ lO', 10\ 10\ lO' and lO' CFU/ml Z hailii, respectively. Lane C, PCR amplicons from prepared from crude cell extracts of Z. hailii. Lanes M, lOObp molecular weight markers. 5.4 DISCUSSION

Microtitre plate DNA-probe assays have recently been developed for clinical yeasts and yeast-like fungi (Elie et al. 1998; Lindsley et al. 2001), marine yeasts (Kiesling et al 2002a), copepods (Kiesling et al 2002b) and dinoflagellates (Goodwin et al 2005). In this chapter, the concepts and principles described in these studies have been extended to develop and evaluate a microtitre plate DNA-probe assay for four yeast species relevant in wine production, namely D. bruxellensis, H. uvarum, M. pulcherrima and Z bailil A successful assay was achieved for only one of these species, Z bailii.

Initial difficulties were encountered with high background signals in the assay, due to non-specific binding of biotinylated primers to the wells of the microtitre plate, despite the use of standard blocking buffers to prevent this problem (Kiesling et al 2002a). The use of additional blocking reagents bovine serum albumin decreased the extent of these background signals, but not sufficiently for a successful assay. Skim milk protein was more effective as a blocking agent, but adversely influenced the sensitivity of the assay. Aminated plate surfaces have been routinely used in other studies and plates without aminated surface were trialed here as a means of removing background signals. Although background could be minimized, probe reactivity on this surface was also reduced, and the use of aminated plates was continued in the assay. Excess biotinylated primers were considered to be the cause of the problem, since their removal from the PCR mixture, prior to hybridization to probed plates, decreased the non-specific background signals to negligible values. The need to remove these from the PCR amplicons, however, added an extra step and cost to the assay.

Using the optimized assay, there were varying strengths in the signals (absorbance values) obtained with the different probes. This phenomenon is not uncommon in probe literature (e.g. Fujita et al 1995; Elie et al 1998; Diaz and Fell 2004; Page and Kurtzman 2005). Although the underlying reasons for the varying signals are not clear, probe signal intensities are often discussed in relation to probe accessibility and stability of the probe-target duplex, which are affected by various molecular and base interactions. For example, secondary structures in the target DNA, base stacking and interactions and steric effects, impose thermodynamic and kinetic obstacles to probe hybridization and have been shown to significantly reduce probe-duplex yield (Kushon et al 2001; Binder et al. 2004; Diaz and Fell 2004). Secondaiy structures in the ribosomal RNA genes are also known to hinder hybridization of probes to their targets in FISH assays (Kosse et al 1997; Inacio et al. 2003). These observations might explain the very weak signals obtained with the Dbru probe in this study. Stender et al. (2001, 2002) have shown the utility of using nucleic acid analogues. Peptide Nucleic Acids (PNA) in probe hybridization assays. When compared with DNA probes, PNA probes are considered to improve hybridization to highly structured targets, conferring higher affinity, improved specificity and more rapid hybridization kinetics. Therefore, PNA probes may provide an advantage in overcoming adverse molecular and structural interactions in hybridization assays, and help to improve signal strength.

The specificity of the assay was evaluated by challenging species-specific probes with PCR amplicons from closely-related species. Unfortunately for three of the target yeasts {D. briaellensis, H. uvarum and M pulcherrima), relatively high assay signals were obtained with these related species (cross hybridization). Varying the hybridization temperature from 50°C, 55°C and 60°C did affect the extent of cross-reactive signals, but did not sufficiently decrease the signals to acceptable levels (Figure 5.5).

Probe-based assays often pose inherent problems with cross hybridization, and have the potential of yielding false positive results. Cross reactivity has been observed in assays targeting PCR amplicons (Elie et al 1998; Lindsley et al. 2001; Hsiao et al 2005; Page and Kurtzman 2005) and in FISH protocols, where ribosomes are the targets for DNA probes (Kosse et al 1997). Further efforts are needed to redesign probe sequences for D. bruxellensis, H. uvarum and M. pulcherrima, that will favour binding to its target while disfavouring binding to non-target sequences.

In addition to assay stringencies (hybridization temperature and ionic strength), probe specificity is determined by characteristics of the oligonucleotide; its base composition, length, the degree and position of mismatches. Despite careful design, the species- specific probes in this study behaved against obvious expectation of sequence specificity. Diaz and Fell (2004, 2005) and Riesling et al (2002a) showed that it was possible to discriminate species with probes containing a single nucleotide difference. Using similar probe design criteria and stringency conditions in this study, two consecutive centrally positioned, nucleotide differences in the Zba probe was found to be sufficient in differentiating Z bailii from Z. bisporus. However, with seven mismatched nucleotides, the Dbru probe failed to resolve D. bruxellensis from D. anomala. Hsiao et al. (2005) and Page and Kurtzman (2005) also described similar ambiguities in probe hybridization behaviour. Likewise, Naef ^/A/. (2002) and He et al. (2005) noted that mismatched sequences could bind probes with greater affinity than perfect matches. These observations reveal the complicated nature of DNA hybridization reactions, and suggest that the interplay of hybridization variables is also not well understood (Naef et al. 2002, Pozhitkov et al. 2006). Model-dependent, empirical and experimental studies in probe hybridization reveal key parameters which affect probe specificities. For example, Diaz and Fell (2004, 2005) and He et al. (2005) examined the relative contribution of mismatched bases and positional dependence to probe specificity. Binder et al. (2004) investigated duplex stabilities as a function of base-pairings (C > G « T > A). Matveeva et al. (2003) indicated that thermodynamic calculations on probe sequences (Gibbs free energy) correlated with, and predicted the stabilities of probe-target duplexes, but were contradicted by the findings of Pozhitkov etal (2006).

Following the criteria of other studies, probe design in this study was based on analysis of sequences in the D1/D2 region of the 26S rDNA. This region is generally selected because it has been sequenced for all known yeast species and a vast database is publicly available. For some taxonomic groups of yeasts, the design of species-specific probes is made difficult because of insufficient variation in genome sequence. In this study, there were few unique sequences in the D1/D2 region from which alternate probe sequences for the species D. bruxellensis, Hs. nvarum and Z bailii could be selected. The limited sequence variability in this region places constraints in the design of species-specific probes. For example, phylogenetic analyses have shown that species within the Hs. uvarum-Hs. guilliermondii cluster differ by less than 1% sequence divergence in the D1/D2 region (Cadez et al 2003). By comparison, the ITS and IGS regions of the rDNA are characterised by higher rates of nucleotide substitution and such regions could be investigated in future as alternatives for species diagnostic probe design (James et al. 1996; Diaz et al. 2005). However, the existing databases for these regions are not yet sufficiently complete for broad application in probe design (Page and Kurtzman 2005). Finally, where the design of a species-specific probe is not feasible, several studies have sought to develop group- and genus-specific probes (Lindsley et al. 2001; Diaz and Fell 2004; Hsiao et al 2005).

The probe designed for Z hailii was found to be specific for its target, and reliably discriminated against other members of Zygosaccharomyces, and other yeast species commonly found in wine fermentations. The assay gave identification of Z bailii from yeast isolates within 8 h, after obtaining pure cultures. Additionally, PGR amplications could be successfully performed using crude cell lysates, eliminating the need for DNA extractions on yeast isolates. The probe hybridization assay was capable of detecting Z bailii in wine samples, with a detection limit of 10 cells/ml. The assay could be further evaluated for its performance in mixed cultures, to determine its ability to detect Z bailii in the presence of multiple yeast species, as would be expected in grape must and wines.

Z bailii is a major spoilage yeast in the food and beverage industries (Pitt and Hocking 1997; James and Stratford 2003; Kurtzman 2006). Consequently, the microplate-probe assay described in this study has scope for wide application in food and beverage quality assurance programs, in addition to those in the wine industry. Specific primer sequences to identify Z bailii in PGR assays (Sancho et al 2000) and more recently, in real time PGR assays (Rawsthome and Phister 2006) have been published. This Ghapter gives the first report of a microtitre plate-DNA probe assay for Z bailii that could be routinely used in quality assurance program. CHAPTER SIX

CONCLUSIONS

This thesis has reported a systematic investigation of the yeasts and yeast-h'ke fungi associated with wine grapes cuhivated in several Austrahan vineyards. In conjunction with a parallel study (Prakitchaiwattana 2005), it represents the first, detailed report of the yeasts associated with Australian wine grapes. Grape samples were examined for populations and species of yeasts as the grapes progressed through the different stages of development until maturity and harvest. A combination of cultural and molecular methods was used to analyse the yeast flora associated with these grapes. Three red grape varieties (Cabernet sauvignon, Merlot, Tyrian) and two white varieties (Sauvignon blanc, Semillon) were examined, and analyses were conducted over the 2001-2002 and 2002-2003 seasons. Various factors that were likely to affect the yeast ecology of grapes were considered. These factors included vineyard location, climatic influences, pesticide applications and grape berry damage.

Few yeasts were detected on immature grapes (10-10^ CFU/g), but yeast populations increased 10-1000 fold as berries developed to harvest ripeness (lO'^-lO^ CFU/g). Greater populations of yeasts (lO'^-lO^ CFU/g) were found on mature, damaged berries.

The yeast-like fungus, Aureobasidium pullulans was the most prevalent species on grapes as detected by plating, enrichment culture and PCR-DGGE. Its population ranged from 10-10^ CFU/g and it was most frequently the dominating species on ripening grapes. Throughout grape maturation, various species of Sporobolomyces, Rhodotorula and Cryptococcus were frequently, but not consistently recovered by plating or enrichment cultures. Their populations generally varied between lO'-lO^ CFU/g, with Rh. glutinis, Rh. mucilaginosa, Cr. laurentii and Cr. heveanensis being the most frequently isolated species. For the 2001-2002 season, many of the species found on immature grapes were also isolated from mature, ripe grapes. At this stage, however, species of Metschnikowia and Hanseniaspora were more prominent; these species were found at 10^-10^ CFU/g by plating, but also recovered by enrichment culture and PCR- DGGE on both sound and damaged ripe berries. Damaged berries showed an increased incidence of Metschnikowia and Hanseniaspora species during the 2001-2002 season, but were often absent on these grapes in the hot and dry vintage of 2002-2003, where A. pullulans and other species of Aureobasidium, Cryptococcus and Rhodotorula were prominent. Climatic factors were considered to be the main reason for these differences; whereas the 2001-2002 season was characterised by significant rainfall during the time of harvest, the 2002-2003 was an atypically hot and dry vintage. The type of berry damage is another factor that needs to be considered, and this could be related to climactic factors as well. There were no obvious influences of grape variety, vineyard location and pesticide application on the yeast species and populations, and more systematic and controlled studies of these variables are required.

PCR-DGGE revealed the presence of several filamentous fungal species on grape samples {Raciborskiomyces longisetosum, Phialocephala scopiformis, Phoma species), the significance of which requires investigation. The effect of, non-target, fungal DNA templates on the successful PGR amplification and detection of yeasts also requires consideration. Generally, plate culture on MEA and WLNA revealed a greater diversity of yeast species than PCR-DGGE; frequently, species of Rhodotorula or Cryptococcus present at 10^ CFU/g were not detected by PCR-DGGE analysis. In other cases, the PCR-DGGE technique revealed the presence of yeasts (C. gallU C. zemplinind) not found by culture, suggesting the occurrence of viable but non culturable populations of yeasts on the surface of grapes. These observations highlight the need for using a combination of culture and culture-independent methods for microbial ecological studies.

Similar yeast species were recovered from grapes using maceration and rinse pre- isolation methods. However, compared with maceration, the rinsing treatment did not recover the total population of yeasts from grapes. Direct microscopic observation of grape surfaces using fluorescent in situ hybridization or electron scanning microscopy are additional analytical methods that could be used to study the yeast flora of grapes to avoid some of the limitations associated with removing the yeast flora for enumeration. Typical wine fermentative yeasts such as Saccharomyces, Torulaspora, Kluyveromyces and Zygosaccharomyces species were isolated by enrichment culture, from both healthy and damaged berries of both seasons. However, their incidences were rare and A. pullulans was the main species isolated in most cases. More defined enrichment culture conditions may be needed to specifically target their isolation and detection, and to further investigate their association with grapes.

Aureobasidium pullulans was the most prevalent species on both red and white grape varieties from every vineyard, in both the 2001-2002 and 2002-2003 seasons. This species was prominent on immature and mature grapes, in either damaged or undamaged conditions, as determined by plating, enrichment culture and PCR-DGGE analysis. Isolates of this species demonstrated significant heterogeneity in colony and cellular morphology, and in PCR-DGGE banding patterns. Because of their prevalence, further studies were undertaken to examine their diversity and aspects of their oenological significance (Chapter Four). Sixty one isolates were characterized taxonomically, by a range of biochemical and physiological properties and by ITS- RFLP analysis. Over half of the isolates were reliably identified as^. pullulans by sequencing of the partial large subunit rDNA and ITS1-5.8S-ITS2 regions. The remaining isolates could not be clearly identified to species level, but could represent distinct or unique species within the Aureobasidium, Hormonema and Kabatiella generic complex. Microsatellite DNA profiling showed significant genetic variability within the isolates, but strain diversity did not correlate with geographical origin of the grapes, grape cultivar or maturity of grapes.

Aureobasidium pullulans and related isolates were able to grow at pH 3.0, 3.5 and 4.0. All isolates were sensitive to ethanol concentrations of 6% or lower. Over half the strains of A. pulllulans grew in the presence of 50 ppm total SO2 at pH 3.5, while strains belonging to Hormonema were all sensitive to 50 ppm SO2. Aureobasidium and related isolates metabolised malic and tartaric acids, they produced extracellular polysaccharides and were producers of an array of extracellular hydrolytic enzymes (protease, lipase, cutinase, pectinases, p-glucosidase, cellulase, polyphenol oxidase), indicating their potential to grow in, and influence the chemical composition of grape juice or musts. In agar plate culture assays, A. pullulans and related isolates showed the ability to inhibit the biological pesticide, Bacillus thuhngiensis. Strains belonging to Hormonema were antagonistic towards grape spoilage fungi such as Botrytis cinerea and the ochratoxin A producing fungus, Aspergillus carbonahus. Several wine yeasts, including strains of cerevisiae were inhibited by Hormonema isolates on agar plate assays, but this was not demonstrated in grape juice fermentations. Some of the properties of the Aureobasidium and Hormonema isolates may have the prospect for commercial development and further exploratory research is needed to examine these possibilities.

A PCR-probe hybridization assay was developed for the detection of several yeast species involved in wine fermentation (Chapter Five). Species-specific capture probes were designed for D. brwcellensis, H. uvarum, M. pulcherrima and Z bailii from the D1/D2 variable domains of the 26S rDNA gene. The specificity of probes was evaluated with amplified DNA from reference cultures in an enzyme immunoassay format (PCR- ELISA). Probes for H marum, D. britxellensis and M. pulcherrima showed cross reactivity to non-target species. The ITS regions may provide an alternative region from which alternative probe sequences for these species can be designed. Additionally, the use of nucleic acid analogues to improve probe hybridization stability and specificity may be explored. The probe designed for Z. bailii reliably discriminated against other members of the genus Zygosaccharomyces, as well as other species widely encountered in wine fermentation. However, its sensitivity in the PCR-probe hybridization assay was 10^ cells of Z bailii per ml of wine. Because this threshold of detection is higher than the few viable Z bailii cells (one to five cells) able to cause spoilage in soft drinks and wines, the PCR-probe hybridization assay is more suitable for the identification of culture isolates, rather than direct detection of the organism in wines. Nevertheless, this is the first report of a rapid, convenient colourimetric assay for specifically detecting Z.bailii, which is a major food and beverage spoilage yeast. The experimental strategy used to develop this assay could be applied to the development of robust, simple PCR- probe hybridization assays for many other species of yeasts that are significant in causing food and beverage spoilage.

In conclusion, the yeast ecology of the grapes examined in this study differs somewhat from that reported for wine grapes examined in other countries. Principally, there is a greater prevalence of A. pullulans, and a lesser presence of Hanseniaspora/Kloeckera, Metschnikowia, Pichia and Candida species. Possibly, the unusually hot and dry conditions in the vineyards that prevailed during most of this project may account for these differences. It is also possible that there is an important and interesting geographical influence on the data obtained. Further research of yeasts on grapes from vineyards at other locations in Australia and over several vintages will be needed to address these questions. CHAPTER SEVEN

BIBLIOGRAPHY

Abrunhosa, L., Paterson, R.R., Kozakiewicz, Z., Lima, N. and Venâncio, A. (2001) Mycotoxin production from fungi isolated from grapes. Lett. Appl. Microbiol. 32, 240- 242.

Adams, A.M. (1964) Airborne yeasts from horticultural sites. Can. J. Microbiol 10, 641-646.

Aheam, D.G., Meyers, S.P. and Nichols, R.A. (1968) Extracellular proteinases of yeasts and yeastlike inngi.Appl Microbiol. 16, 1370-1374.

Ahlholm, J.U., Helander, M., Henriksson, J., Metzler, M. and Saikkonen, K. (2002) Environmental conditions and host genotype direct genetic diversity of Venturia ditricha, a fungal endophyte of birch trees. Evolution 56, 1566-1573.

Addis, E., Fleet; G.H., Cox, J.M., Kolak, D. and Leung, T. (2001) The growth, properties and interaction of yeasts and bacteria associated with the maturation of Camembert and blue-vein cheeses. Int. J. Food Microbiol. 69, 25-36.

Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J. (1990) Basic local alignment search tool. J. Mol. Biol. 215, 403-410.

Alexandre, H., Costello, P.J., Remize, P., Guzzo, J. and Guilloux-Benatier, M. (2004) Saccharomyces cerevisiae - Oenococcus oeni interactions in wine: current knowledge and perspectives. Int. J. Food Microbiol. 93, 141-154.

Àlvarez-Rodriguez, M.L., Belloch, C., Villa, M., Uruburu, F., Larriba, G. and Coque, J- J.R. (2003) Degradation of vanillic acid and production of guaiacol by microorganisms isolated from cork samples. FEMS Microbiol Lett. 220, 49-55.

Amerine, M.A. (1985) Winemaking. In Worlds Debt to Pasteur eds Koprowski, H. and Plotin S.A. New York NY: Alan R. Liss Inc.

Amerine, M.A. and Kunkee, R.E. (1968) Microbiology of winemaking. Annu. Rev. Microbiol 22, 323-358.

Ampe, F., ben Omar, N., Moizan, C., Wacher, C. and Guyot, J.P. (1999) Polyphasic study of the spatial distribution of microorganisms in Mexican pozol, a fermented maize dough, demonstrates the need for cultivation-independent methods to investigate traditional fermentations, ^pp/. Environ. Microbiol 65, 5464-5473.

Andrews, J.H., Berbee, P.M. and Nordheim, E.V. (1983) Microbial antagonism to the imperfect stage of the apple scab pathogen, Venturia inaequalis. Phytopathol 73, 228- 234. Andrews, J.H., Harris, R.F., Spear, R.N., Lau, G.W. and Nordheim, E.V. (1994) Morphogenesis and adhesion of Aureobasidium pullulons. Can J. Microbiol. 40, 6-17.

Andrews, J.H. and Harris, R.F. (2000) The ecology and biogeography of microorganisms on plant surfaces. Annu. Rev. Phytopathol. 38, 145-180.

Andrews, J.H. Spear, R.M. and Nordheim, E.V. (2002) Population biology of Aureobasidium pullulans on apple leaf surfaces. Can. J. Microbiol. 48, 500-513.

Anikaram, N.K.B., Joyce, D.C. and Terry, L.A. (2002) Biocontrol activity and induced resistance as a possible mode of action for Aureobasidium pullulans against grey mould of strawberry fruit. Australasian Plant Pathol. 31, 223-229.

Annis, S.L. and Goodwin, P.H. (1997) Recent advances in the molecular genetics of plant cell wall-degrading enzymes produced by plant pathogenic fungi. Eur. J. Plant Pathol. 103, 1-14.

Antunovics, Z., Irinyi, L. and Sipiczki, M. (2005) Combined application of methods to taxonomic identification of Saccharomyces strains in fermenting botrytized grape musts. J. Appl. Microbiol. 98, 971-979.

Arias, C.R., Bums, J.K., Friedrich, L.M., Goodrich, R.M. and Parish, M.E. (2002) Yeast species associated with orange juice: evaluation of different identification methods. Appl Environ. Microbiol. 68, 1955-1961.

Augustin, J. (2000) Polysaccharide hydrolyases oiAureobasidium pullulans. Folia Microbiol. 45, 143-146.

Avis, T.J. and Bélanger, R.R. (2002) Mechanisms and means of detection of biocontrol activity of Pseudozyma yeasts against plant pathogenic fungi. FEMS Yeast Res. 2, 5-8.

Bae, S., Fleet, G.H. and Heard, G.M. (2004) Significance and occurrence of Bacillus thuringiensis on wine grapes. Int. J. Food Microbiol. 94, 301-312.

Bae, S.S. (2005) Investigation of bacteria associated with Australian wine grapes using cultural and molecular methods. PhD thesis. University of New South Wales.

Bae, S., Fleet, G.H. and Heard, G. M. (2006) Lactic acid bacteria associated with wine grapes from several Australian vineyards. J. Appl. Microbiol. 100, 712-727.

Baileras Couto, M.M., Eijsma, B., Hofstra, H., Huis in't Veld, J.H.J, and van der Vossen, J.M.B. M. (1996a) Evaluation of molecular typing techniques to assign genetic diversity 3mor\g Saccharomyces cerevisiae strains. Appl. Environ. Microbiol. 62, 41-46.

Baileras Couto, M.M., Hartog, B.J., Huis in't Veld, J.H.J., Hofstra, H. and van der Vossen, J.M.B. M. (1996b) Identification of spoilage yeasts in food production chain by microsatellite polymerase chain reaction fingerprinting. Food Microbiol. 13, 59-67. Bains, W. (1994) Selection of oligonucleotide probes and experimental conditions for multiplex hybridization experiments. GATA 11, 49-62.

Baker, E.A. (1970) The morphology and composition of isolated plant cuticles. New Phytologist 69, 1053-1058.

Bakry, M. and Abourough, M. (1996) New data on Querciis suber decline in Morocco. Annales de la Recherche Forestiere au Maroc. 29, 24-39.

Barbetti, M.J. (1980) Bunch rot of Rhine Riesling grapes in lower south-west of Western Australia. Anst. J. Exp. Agrie. Anim. Hitsb. 20, 247-251.

Bardage, S.L., and Bjurman, J. (1998) Isolation of m Aureobasidium pullulans polysaccharide that promotes adhesion of blastospores to water-borne paints. Can. J. Microbiol. 44, 954-958.

Barklund, P; Kowalski, T (1996) Endophytic fungi in branches of Norway spruce with particular reference to Tryblidiopsis pinastri. Can. J. Bot. 74, 673-678.

Bamavon, L., Doco, T., Terrier, N., Agcorges, A., Romicu, C. and Pellerin, P. (2000) Analysis of cell wall neutral sugar composition, beta-galactosidase activity and related cDNA clone throughout the development of Vitis vinifera grape berries. Plant Physiol. Biochem. 38, 289-300.

Bamett, J.A., Delaney, M.A., Jones, E., Magson, A.B. and Winch, B. (1972) The numbers of yeasts associated with wine grapes of Bordeaux. Arch. MikrobioL 83, 52-55.

Bartowsky, E.J., Costello, P.J. and Henschke, P. (2002) Management of malolactic fermentation - wine flavour manipulation. Aiist. Grapegrower and Winemaker 461a, 7- 12.

Bayliss, K.L., Spindler, L., Lagudah, E.S., Sivasithamparam, K., and Barbetti, M.J. (2003) Variability within Kabatiella caulivora Race 2 revealed by cultural and molecular analyses. Aust. J. Agrie. Res. 54, 77-84.

Beattie, G.A. (2002) Leaf surface waxes and the process of leaf colonization by microorganisms. In Phyllosphere microbiology eds. Lindow, S.E., Hecht-Poinar, E.I. and Elliontt, V.J. pp 3-26. Minnesota: APS Press.

Begerow, D., Bauer, R. and Boekhout, T. (2000) Phylogenetic placements of Ustilaginomycetous anamorphos as deduced from nuclear LSU rDNA sequences. Mycol. Res. 104, 53-60.

Beh A.L., Fleet, G.H., Prakitchaiwattana, C. and Heard, G.M. (2006) Evaluation of molecular methods for the analysis of yeasts in foods and beverages. Adv. Exp. Med. 571, 69-106. Belancic, A., Gunata, Z., Vallier, M.-J. and Agosin, E. (2003) p-Glucosidase from the grape native yeast Debaryomyces vanhjiae: purification, characterization, and its effect on monoterpene content of a muscat grape juice. J. Agric. FoodChem. 51, 1453-1459.

Belding, R.D., Sutton, T.B., Blankenship, S.M. and Young, E. (1999) Relationship between apple epicuticular wax and growth of Peltaster fructicola and Leptodontidium elatius, two fungi that cause sooty blotch disease. Plant Dis. 84, 767-772.

Belin, J.M. (1969) Les levures d'un chai du Mâconnais. Vitis 18, 40-48.

Belin, J.M. (1972) Recherches sur la répartition des levures à la surface de la grappe de raisin. Vitis 1, 135-145.

Benda, L (1962) Okologishe Untersuchungen über die Hefeflora im fränkishen Weinbaugebeit. Bayerisches Landvirschqftliches Jahrbuch 39, 595-614.

Benda, I. (1964) Die Hefeflora des fränkischen Weinbaugebietes. Weinberg und Keller 11, 67-80.

Benda, I. (1982) Wine and brandy. In Prescott and Dunn 's Industrial Microbiology Fourth Edition ed Reed, G. pp. 293-402. Wesport Conn.: AVI Publishing.

Bending, G.D. and Read, D.J. (1997) Lignin and soluble phenolic degradation by ectomycorrhizal and cricoid mycorrhizal fungi. Mycol. Res. 101, 1348-1354.

Bene, Z. and Magyar, I. (2004) Characterization of yeast and mould biota of botrytized grapes in Tokaj wine region in the years 2000 and 2001. Acta Alimentaria 33, 259-267.

Berbee, M.L. (1996) Loculoascomycete origins and evolution of filamentous ascomycete morphology based on 18S rRNA gene sequence data. Mol. Biol Evol. 13, 462-470.

Bermejo, J.M., Dominguez, J.B., Goni, P.M., and Uruburu, F. (1981) Influence of carbon and nitrogen sources on the transition from yeast-like cells to chlamydospores in Aureobasidium pullulans. Antonie van Leetfwenhoek 47, 107-119.

Bermejo, J.M., Dominguez, J.B., Goni, P.M., and Uruburu, P. (1981) Influence of pH on the transition from yeast-like cells to chlamydospores \n Aureobasidium pullulans. Antonie van Leeuwenhoek 47, 385-392.

Bessis, R. (1977) Le raison de table et le froid, I.I.P. Comission C2-LiV, Commissions let 111, Paris, 39.

Bettuci, L., Alonso, R. and Tiscomia, S. (1999) Endophytic fungal mycobiota of healthy twigs and the assemblage of species associated with twig lesions of Eucalyptus globules and E. grandis in Uruguay. Afycol. Res. 103: 568-472.

Biely, P. and Slâvikovâ, E. (1994) New search for pectolytic yeasts. Folia Microbiol. 39, 485-488. Binder, H., Kirsten, T., Hofacker, I.L., Stadler, P.F. and Loeffler, M. (2004) Interactions in oligonucleotide hybrid duplexes on microarrays. J. Phys. Chem. B 108, 18015-18025.

Bisson, L.F. (1999) Stuck and sluggish fermentations. Am. J. Enol. Vitic. 50, 107-119.

Bisson, L.F. and Kunkee, R.E. (1991) Microbial interactions during wine production. In Mixed cultures in biotechnology eds. Zeikus, J.C. and Johnson, E.A. pp. 37-68. New York: McGraw Hill.

Blaich, R. (1977) Versuche zur künstlichen Mykorrhizabildung bei Vitis riparia. Vitis 15, 32-37.

Blaich, R. und Rüster, D. (1979) Der Infektionszeitpunkt als Kriterium für parasitäres oder mykotrophes Verhalten won Aureobasidium pullulans auf Vitis riparia. Vitis 18, 21-30.

Blakeman, J.P. and Sztejnberg, A. (1973) Effect of surface wax on inhibition of germination of Botrytis cinerea spores on beetroot leaves. Physiol Plant Phytol. 3, 269- 278.

Bleve, G., Grieco, F., Cozzi, G., Logrieco, A. and Visconti, A. (2006) Isolation of epiphytic yeasts with potential for biocontrol oiAspergillus carbonarius mdA. niger on grape. Int. J. Food Microbiol. 108, 204-209.

Boekhout, T., Theelen, B., Diaz, M., Fell, J.W., Hop, W.C.J., Abeln, E.C.A., Dromer, F. and Meyer, W. (2001) Hybrid genotypes in the pathogenic yeast Cryptococcus neoformans. Microbiol. 147, 891-907.

Boglignano, G. and Criseo, G. (2003) Disseminated nosocomial fungal infection by Aureobasidium pullulans var. melanigenum: a case report. J. Clin. Microbiol. 41, 4483- 4485.

Boudreau, G.W. (1972) Factors related to bird depredations in vineyards. Am. J. Enol. Vitic. 23, 50-55.

Boulton, R. (2003) Red wines. In Fermented beverage production second edition eds. Lea, A.G.H. and Piggot, J.R. pp. 107-134. Berlin: Springer Verlag.

Boulton, R.B., Singleton, V.L., Bisson, L.F. and Kunkee, R.E. (1995) Principles and practices of wine making. New York, NY: Chapman & Hall.

Bourbonnais, R. and Paice, M.G. (1987) Oxidation and reduction of lignin-related aromatic compounds by Aureobasidium pullulans. Appl. Microbiol. Biotechnol. 26, 164-169.

Bruns, T. D., White, T. J., and Taylor, J.W. (1991) Fungal molecular systematics. Annu. Rev. Ecol. Syst. 22, 525-564. Buchanan, G.A. and Amos, T.G. (1992) Grape pests. In Viticulture Vol. 2 Practices eds. Coombe, B.G. and Dry, P.R. pp. 209-231. Adelaide: Winetitles.

Buck, J.W. and Andrews, J.H. (1999) Role of adhesion in the colonization of barley leaves by the yeast Rhodosporidium toruloides. Can J. Microbiol. 45, 433-440.

Buckley, N.G. and Pugh, G.J.F. (1971) Auxin production by phylloplane fungi. Nature 231, 332.

Burke, R.M. and Caimey, J.W.G. (2002) Laccases and other polyphenol oxidases in ecto- and cricoid mycorrhizal fungi. Mycorrhiza 12, 105-116.

Buzzini, P. and Martini, A. (2002) Extracellular enzymatic activity profiles in yeast and yeast-like strains isolated from tropical environments. J. Appl. Microbiol. 93, 1020- 1025.

Cabras, P. Angioni, A., Garau, V.L., Pirisi, P.M., Farris, G.A., Farris, G.A., Madau, G. and Emoonti, G. (1999) Pesticides in fermentative processes of wine. J. Agric. Food Chem. 47, 3854-3857.

Cabras, P. and Angioni, A. (2000) Pesticide residues in grapes, wine and their processing products. J. Agric. Food Chem. 48, 967-973.

Cadez, N., Poot, G.A., Raspor, P. and Th. Smith, M. (2003) Hanseniaspora meyeri sp. nov., Hanseniaspora clermontiae sp. nov., Hanseniaspora lachancei sp. nov. and Hanseniaspora opuntiae sp. nov., novel apiculate yeast species. Int. J. Syst. Evol. Microbiol. S3, 1671-1680.

Campbell, B.S., Siddique, A.-B.M., McDougall, B.M. and Seviour, R.J. (2004) Which morphological forms of the fungus Aureobasidium pullulans are responsible for pullulan production? FEMS Microbiol. Lett. 232, 225-228,

Canal-Llauberes, R.-M. (1993) Enzymes in winemaking. In Wine microbiologgy and biotechnology, ed. Fleet, G.H. pp. 477-499. Switzerland: Harwood Academic Publishers.

Capriotti, A. (1960) Yeasts of the Miami, Florida area. ArchMikrobiol. 20, 323-342.

Carol, A.K. (1996) Relief of amplification inhibition in PCR with bovine serum albumin or T4 gene 32 protein. Appl. Environ. Microbiol. 62, 1102-1106.

Carolan, G., Catley, B.J. and Kelly, P.J. (1976) Ethanol production during the growth cycle oi Aureobasidium pullulans. Biochem. Soc. Trans. 4, 1021-1022.

Casado, C.G. and Heredia, A. (1999) Structure and dynamics of reconstituted cuticular waxes of grape berry cuticle (Vitis vinifera L.)J. Experimental Botany 50, 175-182.

Castelli, T. (1954) Les Agents de la Fermentation Vinaire. Arch. Mikrobiol. 20, 323- 342. Castelli, T. (1957) Climate and agents of wine fermentation. Am. J. Enol. Vitic. 8,149- 156.

Castoria, R., De Curtis, F., Lima, G., Caputo, L., Pacifico, S., and De Cicco, V. (2001) Aureobasidium pullulans (LS-30) an antagonist of postharvest pathogens of fruits: study on its modes of action. Postharvest Biol. Technol 22, 7-17.

Cemáková, M., Kocková-Kratochvílová, A., Suty, L., Zemek, J., and Kuniak L. (1980) Biochemical similarities among strains of Aureobasidium pullulans (de Baiy) Arnaud. Folia Microbiol. 25, 68-73.

Chaimberlain, G., Husnik, J. and Subden, R.E. (1997) Freeze-desiccation survival in wild yeasts in the bloom of icewine grapes. Food. Res. Int. 30, 435-439.

Chambers, T.C. and Possingham, J.V. (1963) Studies of the fine structure of the wax layer of Sultana grapes, ^wi/. J. Biol. Sci. 16, 818-825.

Charoenchai, C., Fleet, G.H., Henschke, P.A. and Todd, B.E.N. (1997) Screening of non-Saccharomyces wine yeasts for the presence of extracellular hydrolytic enzymes. Aust. J. Grape and Wine Res. 3, 2-8.

Chatonnet, P., Boidron, J. N., Dubourdieu, D., and Pons, M. (1994) Changes in the composition of polyphenolic compounds in oak wood during drying. J. Int. Sci. Vigne et du Vin 28, 337-357.

Clark, D.S. and Wallace, R.H. (1958) Carbohydrate metabolism of Pullulariapullulans. Can. J. Microbiol. 4, 43-54.

Clímente-Jimenez, J.M., Mingorance-Carloza, L., Martínez-Rodríguez, S., Heras- Vázquez, F.J. and Rodriguez-Vico, F. (2004) Molecular characterization and oenological properties of wine yeasts isolated during spontaneous fermentation of six varieties of grape must. Food Microbiol. 21, 149-155.

Cocolin, L., Bisson, L.F. and Mills, D.A. (2000) Direct profiling of the yeast dynamics in wine fermentations. FEMS Microbiol Lett. 189, 81-87.

Cocolin, L. and Mills, D.A. (2003) Wine yeast inhibition by sulfur dioxide: a comparison of culture-dependent and independent methods. Am. J. Enol. Vitic. 54, 125- 130.

Collmer, A. and Keen, N.T. (1986) The role of pectic enzymes in plant pathogenesis. Annu. Rev. Phytopathol. 24, 383-409.

Combina, M., Mercado, L., Borgo, P., Elia, A., Jofré, V., Ganga, A., Martinez, C. and Catania, C. (2005) Yeasts associated to Malbec grape berries from Mendoza, Argentina. J. Appl. Microbiol 98, 1055-1061. Comitini, F. and Ciani, M. (2006) Survival of inoculated Saccharomyces cerevisiae strain on wine grapes during two vintages. Lett. Appl. Microbiol 42, 248-253.

Commenil, P., Brunet, L. and Audran, J.C. (1997) The development of the grape berry cuticle in relation to susceptibility to bunch rot disease. J. Experimental Botany 48, 1599-1607.

Considine, J.A. and Knox, R.B. (1979) Development and histochemistry of the pistil of the grape Vitis vinifera. Annals of Botany 43, 11-22.

Cooke, W.B. (1959) An ecological life history oiAureobasidium pullulans (De Bary) Amaud. Mycopath. Mycol. Appl. 12, 1-45.

Cooke, W.B. (1962) A taxonomic study in the 'black yeasts'. Mycopath. Mycol. Appl. 17, 1-43.

Cooke, W.B. and Matsuura, G. (1963) Physiological studies in the black yeasts. Mycopath. Mycol. Appl 21, 225-21

Coombe, B.G. (1992) Grape Phenology. In Viticulture Vol. 1 Resources eds. Coombe, B.G. and Dry, P.R. pp. 139-153/Adelaide: Winetitles.

Coombe, B.G. (1995) Adoption of a system for identifying grapevine growth stages. Aust J. Grape and Wine Res. 1, 100-110.

Coombe, B.G. and Dry, P.R. (1992) Viticulture Vol. 2 Practices. Adelaide: Winetitles.

Coombe, B.G. and McCarthy, M.G. (1997) Identification and naming of the inception of aroma development in ripening grape berries. Aust J. Grape and Wine Res. 3, 18-20.

Coombe, B.G. and McCarthy, M.G. (2000) Dynamics of grape berry growth and physiology of ripening. Atdst. J. Grape and Wine Res. 6, 131-135.

Coque, J.-J. R., Alvarez-Rodriguez, M. L., Larriba, G. (2003) Characterization of an inducible chlorophenol o-methyltransferase from Trichoderma longibrachiatum involved in the formation of chloroanisoles and determination of its role in cork taint of v/'mQS.Appl. Environ. Microbiol. 69, 5089-5095.

Costello, P.J., Henschke, P.A. and Markides, A.J. (2003) Standardised methodology for testing malolactic bacteria and wine yeast compatability. Aust. J. Grape and Wine Res. 9, 127-137.

Couto, J.A., Barbosa, A. and Hogg, T. (2005) A simple cultural method for the presumptive detection of the yeasts Brettanomyces/Dekkera in wines. Lett. Appl. Microbiol. 41, 505-510.

Cox, J.M. and Fleet, G.H. (2003) New directions in the microbiological analysis of foods. In Foodborne microorganisms of public health significance sixth edition, ed. Hocking, A.D. pp 103-162. Sydney: Aust. Inst, of Food Science and Techno!. Croan, S. C. (2000) Evaluation of white-rot fungal growth on southern yellow pine wood chips pretreated with blue-stain fungi. Conference proceedings of the International Research Group, Wood Preservation (Hawaii, USA), IRG Secretariat, Stockholm.

Crosby, L.D. and Criddle, C.S. (2003) Understanding bias in microbial community analysis techniques due to rrn operon copy number heterogeneity. Biotechniques 34, 2- 9.

Crous, P.W., Lennox, C.L. and Sutton, B.C. (1998) Selenophoma eucalypti and Stigmina robbenensis spp. nov. from Eucalyptus leaves on Robben Island. Mycol. Res. 99, 648-652.

Crous, P.W., Groenewald, J.Z., Wingfield, M.J. and Aptroot, A. (2003) The value of ascospore septation in separating Mycosphaerella from Sphaerulina in the Dothideales: a Saccardoan myth? Sydiowia 55, 136-152.

Dantzig, A.H., Zuckermann, S.H. and Andonov-Roland, M.M. (1986) Isolation of a Fusarium solani mutant reduced in cutinase activity and virulence. J. Bacteriol. 168, 911-916.

Darriet, P., Pons, M., Henry, R., Dumont, O., Findeling, V., Cartolaro, P., Calonnec, A. and Dubourdieu, D. (2002) Impact odorants contributing to the fungus type aroma from grape berries contaminated by powdery mildew (Uncinula necator); Incidence of enzymatic activities of yeast Saccharomyces cerevisiae. J. Agric. Food Chem. 50, 3277- 3282.

Davenport, R.R. (1973) Vineyard yeasts and environmental study. In Sampling- microbiological monitoring of environments eds Board, R.G. and Lovelock, D. pp. 143- 174. London: Academic Press.

Davenport, R.R. (1974) Microecology of yeasts and yeast-like organisms associated with an English vineyard. Vitis 13, 123-130.

Davenport, R.R. (1976) Distribution of yeasts and yeast-like organisms from aerial surfaces of developing apples and grapes. In Microbiology of Aerial Plant Surfaces eds Dickinson, C.H. and Preece, T.F. pp. 325-351. London: Academic Press.

Davis, C.R., Wibowo, D., Eschenbruch, R., Lee, T.H. and Fleet, G.H. (1985) Practical implications of malolactic fermentation: a review. Am. J. Enol. Vitic. 36, 290-301.

Deak, T. (2003) Detection, enumeration and isolation of yeasts. In Yeasts in food: beneficial and detrimental aspects eds. Boekhout, T. and Robert, V. pp. 39-68. Hamburg: Behr's Verlag.

Deak, T. and Beuchat, L.R. (1993) Comparison of the SIM, API 20C, and ID 32C systems for identification of yeasts isolated from fruit juice concentrates and beverages. J. FoodProt. 56, 585-593. Delaherche, A., Claisse, O. and Lonvaud-Funel, A. (2004) Detection and quantification of Brettanomyces brwcellensis and 'ropy' Pediococcus damnosus strains in wine by real-time polymerase chain reaction. J. Appl. Microbiol 97, 910-917.

Demuyter, C., Lollier, M., Legras, J.-L. and Le Jeune, C. (2004) Predominance of Saccharomyces uvarum during spontaneous fermentation, for three consecutive years, in an Alsatian winery. J. Appl. Microbiol. 97, 1140-1148.

Deshpande, M.S., Rale V.B., and Lynch, J.M. (1992) Aureobasidium pullulans in applied microbiology: a status report. Enzyme Microb. Technol. 14, 514-527.

Dennis, C. (1972) Breakdown of cellulose by yeast species. J. Gen. Microbiol 71, 409- 411.

Dennis, C. and Buhagiar, R.W.M. (1980) Yeast spoilage of fresh and processed fruits and vegetables. In Biology and Activities of Yeasts eds. Skinner, F.A., Passmore, S.M. and Davenport, R.R. pp. 123-133. London: Academic Press. de Barros Lopes, M., Soden, A., Henschke, P.A. and Langridge, P. (1996) PGR differentiation of commercial yeast strains using intron splice site primers. Appl Environ. Microbiol 62, 4514-4520. de Barros Lopes, M., Rainieri, S., Henschke, P.A. and Langridge, P. (1999) AFLP fingerprinting for analysis of yeast genetic variation. Int. J. Syst. Bacteriol 48, 279-286.

De Camargo, R. and Phaff, H.J. (1957) Yeast occurring in Drosophila flies and in fermenting tomato fruits in Northern California. Food Res. 22, 367-372.

De Curtis, F., Caputo, L., Castoria, R., Lima, G., Stea, G. and De Cicco, V. (2004) Use of fluorescent amplified fragment length polymorphism (fAFLP) to identify specific molecular markers for the biocontrol agQni Aureobasidium pullulans strain LS30. Postharvest Biol Technol 34, 179-186, de La Torre, M.J., Millan, M.C., Perez-Juan, P.M., Morales, J. and Ortega, J.M. (1998a) Changes in the microbiota during ripening of two Vitis vinifera grape varieties in southern Spain. Microbios 96,165-176. de La Torre, M.J., Millan, M.C. and Ortega, J.M. (1998b). Application of linear regression to time series for predicting microorganism counts on the surface of grapes. Cerevisia 23, 21-25. de La Torre, M.J., Millan, M.C., Perez-Juan, P.M., Morales, J. and Ortega, J.M. (1999) Indigenous yeasts associated with two Vitis vinifera grape varieties cultured in southern Spain. Microbios 100, 27-40.

Dias, L., Dias, S., Sancho, T., Stender, H., Querol, A., Malfeito-Ferreira, M. and Loureiro, V. (2003) Identification of yeasts isolated from wine-related environments and capable of producing 4-ethylphenol. Food Microbiol 20, 567-574. Diaz, M.R. and Fell, J.W. (2004) High-throughput detection of pathogenic yeasts of the genus Trichosporon. J. Clin. Microbiol. 42, 3696-3706.

Diaz, M.R. and Fell, J.W. (2005) Use of a suspension array for rapid identification of the varieties and genotypes of the Cryptococciis neoformans species complex. J. Clin. Microbiol. 43, 3662-3672.

Dickinson, C.H. (1976) Fungi on the aerial sufaces of higher plants. In Microbiology of aerial plant surfaces eds Dickinson, C.H. and Preece, T.F. pp. 293-324. London: Academic Press.

Dickinson, C.H. (1986) Adaptations of micro-organisms to climatic conditions affecting aerial plant surfaces. In Microbiology of the phyllosphere eds Fokkema, N.J. and Heuvel, J.V.D. pp. 77-100. Cambridge: Cambridge University Press.

Dicks, L.M.T., Dellaglio, F. and Collins, M.D. (1995) Proposal to reclassify Leuconostoc eonos as Oenococcus oeni (corrig.) gen. nov., comb. nov. Int. J. Syst. Bacteriol. 45, 395-397.

Dittrich, H.H. (1977) Mikrobiologie des weines. Stuttgart: Ulmer.

Divol, B. and Lonvaud-Funel, A. (2005) Evidence for viable but nonculturable yeasts in botrytis-affected wine. J. Appl. Microbiol. 99, 85-93.

Dizy, M., and Bisson, L.F. (2000) Proteolytic activity of yeast strains during grape juice fermentation. Am. J. Enol. Vitic. 51, 155-167.

Dobinson, K.F., Harrington, M.A., Omer, M. and Rowe, R.C. (2000) Molecular characterization of vegetative compatibility group 4A and 4B isolates of Verticillium dahliae associated with potato early dying. Plant Dis. 84, 1241-1245.

Dogget, M.S. (2000) Characterization of Fungal Biofilms within a Municipal Water Distribution System. Appl Environ Microbiol. 66, 1249-1251.

Domsch, K.H., Gams, W. and Anderson, T.-H. (1980) Compendium of soil fungi. London: Academic Press.

Doneche, B. (1993). Botrytized wines. In Wine microbiologgy and biotechnology, ed. Fleet, G.H. pp. 327-351. Switzerland: Harwood Academic Publishers.

Dry, P.R. and Gregory, G.R. (1992) Grapevine varieties. In Viticulture, Vol. 1, Resources eds. Coombe, B.G. and Dry, P.R. pp. 119-138. Adelaide: Winetitles.

Drysdale, G.S. and Fleet, G.H. (1998) Acetic acid bacteria in winemaking: a review. Am. J. Enol. Vitic. 39, 143-154. Dubois, M., Gilles, K.A., Hamilton, K., Rebers, P.A. and Smith, F. (1956) Colourimetric method for determination of sugars and related substances. Anal. Chem. 28, 350-356.

Dugan, F., Lupien, S.L and Grove, G.G. (2002) Incidence, aggressiveness and in planta interactions oiBotrytis cinerea and other filamentous fungi quiescent in grape berries and dormant buds in central Washington State. J. PhytopathoL 150, 375-381.

Duncan, R.A., Stapleton, J.J. and Leavitt, G.M. (1995) Population dynamics of epiphytic mycoflora and occurrence of bunch rots of wine grapes as influenced by leaf removal. Plant Pathol 44, 956-965.

Durreil, L.W. (1967) Studies of Aureobasidium pullulans (de Barry) Arnaud. Mycopath. Mycol.AppL 35, 113-120.

Dubourdieu, D., Ribereau-Gayon, P., and Foumet, B. (1981) Structure of the extracellular 6 D-glucan from Botrytis cinerea. Carbohydr. Res. 93, 294-299.

du Toit, M. and Pretorius, I.S. (2000) Microbial spoilage and preservation of wine: Using weapons from nature's own aresenal- A review. S. Afr. J. Enol Vitic. 21, 74-96.

Egli, G.M. and Henick-Kling, T. (2001) Identification of Brettanomyces/Dekkera species based on polymorphism in the rRNA internal transcribed spacer region. Am. J. Enol. Vitic. 52, 241-247.

Elie, C.M., Lott, T.J., Reiss, E. and Morrison, C.J. (1998) Rapid identification of Candida species with species-specific DNA probes. J. Clin. Microbiol. 36, 3260-3265.

Ellis, M.B. (1971) Dematiaceons Hyphomycetes. Kew: Commonwealth Mycological Institute.

Emmett, R.W., Harris, A.R., Taylor, R.H. and McGechan, J.K. (1992) Grape diseases and vineyard protection. In Viticulture Vol. 2. Practices eds. Coombe, B.G. and Dry, P.R. pp. 232-278. Adelaide: Winetitles Australia.

Ercolini, D. (2004) PCR-DGGE fingerprinting: novel strategies for detection of microbes in food. J. Microbiol. Methods 56, 297-314.

Eriksson, O.E., Baral, H.-O., Currah, R.S., Hansen, K., Kurtzman, C.P., Rambold, G. and Laess0e T. (2001) Outline of Ascomycota imX.Myconet 7, 1-88.

Esteban, M.A., Villanueva, M.J. and Lissarrague, J.R. (1999) Effect of irrigation on changes in berry composition of Tempranillo during maturation. Sugars, organic acids and mineral elements. Am J. Enol. Vitic. 50, 418-434.

Esteve-Zarzoso, B., Belloch, C., Uruburu, F. and Querol, A. (1999) Identification of yeast by RFLP analysis of the 5.8S rRNA gene and the two ribosomal internal transcribed spacers. Int. J. Syst. Bacteriol. 49, 329-337. Ewart, A. (2003) In Fermented beverage production second edition eds. Lea, A.G.H. and Piggot, J.R. pp. 89-105. Berlin: Springer Verlag.

Farrelly, V., Rainey, F.A. and Stackebrandt, E. (1995) Effect of genome size and rm gene copy number on PGR amplification of 16S rRNA genes from a mixture of bacterial species. Environ. Microbiol 61,2798-2801.

Federici, F. (1982) Extracellular enzymatic activities in Aureobasidium pullulans. Mycologia 74, 738-743.

Fell, J.W., Boekhout, T., Fonseca, A., Scorzetti, G. and Statzell-Tallman, A. (2000) Biodiversity and systematics of basidiomycetous yeasts as determined by large-subunit rDNA D1/D2 domain sequence analysis. Int. J. Syst. Evol. Microbiol. 50, 1351-1371.

Fernández-Espinar, M.T., López, V., Ramón, D., Bartra, E. and Querol, A. (2001) Study of the authenticity of commercial wine yeast strains by molecular tehcniques. Int. J. Food Microbiol. 70, 1-10.

Fernández-Espinar, M.T., Martorell, P., de Llanos, R. and Querol, A. (2006) Molecular methods to identify and characterize yeasts in foods and beverages. In Yeasts in Food and Beverages ed. Querol, A. and Fleet, G.H. pp. 55-82. Berlin: Springer-Verlag.

Ferris, M.J. and Ward, D.M. (1997) Seasonal distributions of dominant 16S rRNA- defmed populations in a hot spring microbial mat examined by denaturing gradient gel electorphoresis. Environ. Microbiol. 63, 1375-1381.

Feuillat, M. (2003) Yeast macromolecules: origin, composition and enological interest. Am. J. Enol. Vitic. 54, 211-213.

Fields, R.D., Rodriguez, F, and Finn, R.K. (1974) Microbial degradation of polyesters: polycaprolactone degraded by P. pullulans. J. Appl. Polymer Sci. 18, 3571-3579.

Filip, P. Weber, R.W.S., Sterner, O. and Anke, T. (2003) Hormonemate, a new cytotoxic and apoptosis-inducing compound from the endophytic fungus Hormonema dematioides. I. Identification of the producing strain, and isolation and biological properties of hormonemate. Zeitschrift fur Naturforschung. Section C, Biosciences. 58, 547-552.

Fischer, S.C. and Lerman, L.S. (1983) DNA fragments differing by single-base pair substitutions are separated in denaturing gradient gels: correspondence with melting theory. Proc. Natl. Acad Sci. USA. 80, 1579-1583.

Fischer, P.J., Petrini, O. and Lappin Scott, H.M. (1992) The distribution of some fungal and bacterial endophytes in maize {Zea mays L.) New Phytol. 122, 299-305.

Fleet, G.H. (1993) The microorganisms of winemaking: Isolation, enumeration and identification. In Wine microbiology and biotechnology ed. Fleet, G.H. pp. 1-25. Switzerland: Harwood Academic. Fleet, G.H. (1998) Microbiology of alcoholic beverages. In Microbiology of fermented foods, Vol. 1 ed. Wood, BJ. pp. 217-262. London: Blackie Academic & Professional.

Fleet, G.H. (1999) Microorganisms in food ecosystems. Int. J. Food Microbiol. 50, 101- 117.

Fleet, G.H. (2001) Wine. In Food microbiology fundamentals andfrontiers eds. Doyle, M.P., Beuchat, L.R. and Montville, T.J. pp. 747-772. Washington DC: ASM Press.

Fleet, G.H. (2003) Yeast interaction and wine flavour. Int. J. Food Microbiol 86, 11-22.

Fleet, G.H. and Heard, G.M. (1993) Yeast growth during fermentation. In Wine microbiology and biotechnology ed. Fleet, G.H. pp. 27-54. Chur, Switzerland: Harwood Academic Publisher.

Fleet, G.H., Prakitchaiwattana, C., Beh, A.L. and Heard, G. (2002) The yeast ecology of wine grapes. In Biodiversity and biotechnology of wine yeasts ed. Ciani, M. pp. 1-17. Kerala: Research Signpost.

Fogel, G.B., Collins, C.R., Li, J. and Brunk, C.F. (1999) Prokaryotic genome size and SSU rDNA copy number: estimation of microbial relative abundance from a mixed population. Microbial Ecol. 38, 93-113.

Foldes, T., Banhegyi, I., Herpai, Z., Varga, V. and Szigeti, J. (2000) Isolation of Bacillus strains from the rhizosphere of cereals and in vitro screening for antagonism against phytopathogenic, food-borne pathogenic and spoilage micro-organisms. J. Appl. Micorbiol 89, 840-846.

Fonseca, A. and Inacio, J. (2006) Phylloplane yeasts. In Biodiversity and Ecophysiology of Yeasts eds. Rosa, C. and Péter, G. pp. 263-301. Berlin: Springer Verlag.

Foschino, R., Gallina, S., Andrighetto, C., Roseeti, L. and Galli, A. (2004) Comparison of cultural methods for the identification and molecular investigation of yeasts from sourdoughs for Italian sweet baked products. FEMS Yeast Res. 4, 609-618.

Frankland, J.C. (1969) Fungal decomposition of bracken petioles. J. Ecology 51, 25-36.

Fujita, S-L, Lasker, B.A., Lott, T.J., Reiss, E., and Morrison, C.J. (1995) Microtitration plate enzyme immunoassay to detect PCR-Amplified DNA from Candida species in blood. J. Clin. Microbiol. 33, 962-967.

Gadanho, M. and Sampaio, J.P. (2004). Application of temperature gradient gel electrophoresis to the study of yeast diversity in the estuary of the Tagus river, Portugal. FEMS Yeast Res. 5, 253-261.

Gandini, A. (1969) Microbiological studies of Asti Muscatel. III. Microbial contamination of sparkling Asti wine during its bottling and the effect of an iodine- containing compound. Vini d'ltalia. 11, 527-544. Ganga, M.A. and Martínez, C. (2004) Effect of wine yeast monoculture practice on the biodiversity of non-Saccharomyces yeasts. J. Appl. Microbiol 96, 76-83. Ganter, P.F. (2006) Yeast and invertebrate associations. In Biodiversity and Ecophysiology of Yeasts eds. Rosa, C. and Péter, G. pp. 303-3370. Berlin: Springer Verlag. Ganter, P.F. and de Barros Lopes, M. (2000) The use of anonymous DNA markers in assessing worldwide relatedness in the yeast species Pichia kluyveri Bedford and Kudrjavzev. Can. J. Microbiol 46, 967-980. Ganter, P.F. and Starmer, W.T. (1992) Killer factor as a mechanism of interference competition in yeasts associated with cacti. Ecology 73, 54-67. Gente, S., Desmasures, N., Panoff, J.-M. and Guéguen, M. (2002) Genetic diversity among Geotrichum candidum strains from various substrates studied using RAM and RAPD-PCR. J. Appl Microbiol 92, 491-501. Gildemacher, P.R., Heijne, B., Houbraken, J., Vromans, T., Hoeksra, E.S. and Boekhout, T. (2004) Can phyllosphere yeasts explain the effect of scab fungicides on russeting of Elstar apples? Eur. J. Plant Pathol 110, 929-937. Giraffa, G. (2004) Studying the dynamics of microbial populations during food fermentation. F EMS Microbiol Rev, 28, 251-260. Giudici, P. and Pulvirenti A. (2002) Molecular methods for identification of wine yeast. In Biodiversity and biotechnology of wine yeast ed. Ciani, M. pp. 35-52. Kerala: Research Signpost. Goignes S., Belarbi, A. and Barka E.A. (2001) Saccharomyces cerevisiae, a potential pathogen towards grapevine, Vitis vinifera. FEMS Microbial Ecol 37, 143-150. Gognies, S., Barka, E.A., Gainvors-Claisse, A. and Belarbi, A. (2006) Interactions between yeasts and grapevines: filamentous growth, endopolygalacturonase and phytopathogenicity of colonizing yeasts Microbial Ecol 51,109-116. Goodwin, K.D., Cotton, S.A., Scorzetti, G. and Fell, J.W. (2005) A DNA hybridization assay to identify toxic dinoflagellates in coastal waters: detection of Karenia brevis in the Rookeiy Bay National Estuarine Research Reserve. Harmful Algae 4, 411-422. Goto, S. and Oguri, H. (1983) Yeast flora in wild grapes from mountainous places around the Kofu Basin of Central Japan. Trans. Mycol Soc. Japan. 24,151-157. Goto, S. and Yokotsuka, I. (1977) Wild yeast populations in fresh grape musts of different harvest times. J. Fermentation TechnoX. 55, 417-422 Golubev, W.I. (2006) Antagonistic interactions among yeasts. In Biodiversity and Ecophysiology of Yeasts eds. Rosa, C. and Péter, G. pp. 197-219. Berlin: Springer Verlag. Granchi, L., Bosco, M., Messini, A., and Vincenzini, M. (1999) Rapid detection and quantification of yeast species during spontaneous wine fermentation by PCR-RFLP analysis of the rDNA ITS region. J. Appl. Microbiol. 87, 949-956.

Granchi, L., Ganucci, D., Viti, C., Giovannetti, L. and Vincenzini, M. (2003) Saccharomyces cerevisiae biodiversity in spontaneous commercial fermentations of grape musts with 'adequate' and 'inadequate' assimilable-nitrogen content. Lett. Appl Microbiol. 36, 54-58.

Grncarevic, M, and Radier, F. (1971) A review of the surface lipids of grapes and their importance in the drying process. Am. J. Enol. Vitic. 22, 80-85.

Guan, T.Y., Blank, G., Ismond, A. and Van Acker, R. (2001) Fate of foodbome bacterial pathogens in pesticide products. J. Sci. FoodAgri. 81, 503-512.

Guerra, E., Sordi, G., Mannazzu, I., Climenti, F. and Fatichenti, F. (1999) Occurrence of wine yeasts on grapes subjected to different pesticide treatments. Ital. J. Food Sci. 3, 221-230.

Guerzoni, E. and Marchetti, R. (1987) Analysis of yeast flora associated with grape sour rot and of the chemical disease markers. Appl. Envi. Microbiol. 53, 571-576.

Guillamon, J.M., Sabaté, J., Barrio, E., Cano, J. and Querol, A. (1998) Rapid identification of wine yeast species based on RFLP analysis of the ribosomal internal transcribed spacer (ITS) region. Arch. Microbiol. 169, 387-392.

Gunata, Z., Bitteur, S., Brillouet, J.-M., Bayonove, C. and Cordonnier, R. (1988) Sequential enzymic hydrolysis of potentially aromatic glycosides from grape. Carbohydr. Res. 184, 139-149.

Gunde-Cimerman, N., Zalar, P., de Hoog, S. and Plemenitas, A. (2000) Hypersaline waters in salterns - natural ecological niches for halophilic black yeasts. FEMS Microbiol. Ecol. 32, 235-240.

Haase, G., Sonntag, L., von de Peer, Y., Uijthof, J.M.J., Podbielski, A. and Melzer- Krick, B. (1995) Phylogenetic analysis of ten black yeast species using nuclear small subunit rRNA gene sequences. Antonie van Leeuwenhoek 68, 19-33.

Hankin, L. and Anagnostakis, S.L. (1975) The use of solid media for detection of enzyme production by fungi. Mycologia 67, 597-607.

Hardie, W.J. O'Brien, T.P and Juadzems, V.G. (1996) Morphology, anatomy and development of the pericarp after anthesis in grape Vitis vinifera. Aust J. Grape and Wine Res.

Hawkes, M., Rennie, R., Sand, C. and Vaudry, W. (2005) Aureobasidium piillulans infection: fungemia in an infant and a review of human cases. Diagn. Microbiol. Infect. Dis. 51,209-213. He, Z., Liyou, W., Xingyuan, L., Fields, M.W. and Zhou, J. (2005) Empirical establishment of oligonucleotide probe design criteria. Appl. Environ. Microbiol 71, 3757-3760.

Head, I.M., Saunders, J.R. and Pickup, R.W. (1998) Microbial evolution, diversity and ecology: a decade of ribosomal RNA analysis of uncultivated microorganisms. Microbial Ecol 35, 1-21.

Heard, G.M. (1999) Novel yeasts in winemaking: looking to the future. Food Australia 51, 347-352.

Heard, G.M. and Fleet, G.H. (1987) Occurrence and growth of killer yeasts during wine fermentation. Appl. Environ. Microbiol. 53, 2171 -2174.

Henschke, P.A. and Jiranek, V. (1993) Yeasts - metabolism of nitrogen compounds. In Wine microbiologgy and biotechnology, ed. Fleet, G.H. pp. 77-145. Switzerland: Harwood Academic Publishers.

Hermanides-Nijhof, E.J. {\911) Aureobasidium and allied genera. Stud. MycoL 15, 141- 177.

Heras-Vásquez, F.J., Mingorance-Carlorza, L., Clemente-Jimenez, J.M. and Rodriguez- Vico, F. (2003) Identification of yeast species from orange fruit juice by RFLP and sequence analysis of the 5.8S rRNA gene and the two internal transcribed spacers. FEMS Yeast Res. 3, 3-9.

Hernán-Gómez, S., Espinosa, J.C. and Ubeda, J.F. (2000) Characterization of wine yeasts by temperature gradient gel electrophoresis (TTGE). FEMS Microbiol. Lett. 193, 45-50.

Hernández, L.F., Espinoza, J.C., Ferández-Gonzáles, M. and Briones, A. (2002) p- Glucosidase activity in a Saccharomyces cerevisiae wine strain. Int. J. Food Microbiol. 80, 171-176.

Holloway, P., Subden, R.E. and Lachance, M-A. (1990) The yeasts in a Riesling must from the Niagara grape-growing region of Ontario. Can. Inst. Food Sci. Technol. 23, 212-216.

Holm Hansen, E., Nissen, P., Sommer, P., Nielsen, J.C. and Ameborg, N. (2001) The effect of oxygen on the survival of non-Saccharomyces yeasts during mixed culture fermentations of grape juice with Saccharomyces cerevisiae. J. Appl. Microbiol. 91, 541-547.

Hoog, G.S. de (1998) A key to the anamorph genera of yeastlike Archi- and Euascomycetes. In The Yeasts, A Taxonomic Study, Fourth edition, eds. Kurtzman, C.P. and Fell, J.W. pp. 123-125. Amsterdam: Elsevier Science B.V. Hoog, G.S. de and McGinnis M.R. (1987) Ascomycetous black yeasts. Stud. MycoL 31,187-199.

Hoog, G.S. de and Yurlova, N.A. (1994) Conidiogenesis, nutritional physiology and taxonomy of Aureobasidium and Hormonema. Antonie van Leeuwenhoek, 65, 41-54.

Horvath, R.S., Brent, M.M., and Cropper, D.G. (1976) Paint deterioration as a result of the growth oiAureobasidium pullulans on wood. Appl Environ. Microbiol. 32, 505- 507.

Howell, K.S, Bartowsky, E.J., Fleet, G.H. and Henschke, P.A. (2004) Microsatellite PGR profiling of Saccharomyces cerevisiae strains during wine fermentation. Lett Appl. Microbiol. 38, 315-320.

Hsiao, C.R., Huang, L., Bouchara, J.-P., Barton, R., Li, H.C. and Chang, T.C. (2005) Identification of medically important molds by an oligonucleotide array. J. Clin. Microbiol. 43, 4706-3768.

Inacio, J., Behrens, S., Fuchs, B.M., Fonseca, A., Spencer-Martins, I. and Amann, R. (2003) In situ accessibility of Saccharomyces cerevisiae 26S rRNA to Cy3-labeled oligonucleotide probes comprising the D1 and D2 domains. Appl. Envi. Microbiol. 69, 2899-2905.

Ingram, M. and Liithi, H. (1961) Microbiology of fruit juices. In Fruit and vegetable juices processing technology eds Tressler, D.K. and Joslyn, M.A. pp. 117. Westport, Conn: AVI Publishing.

Ippolito, A., El Ghaouth, A., Wilson, C.E. and Wisniewski, M. (2000) Control of postharvest decay of apple fruit by Aureobasidium pullulans and induction of defence responses. Postharvest. Biol. Technol. 19, 265-272.

Israilides, C., Scanlon, B., Smith, A., Harding, S.E. and Jumel, J. (1994) Characterization of pullulans produced from agro-industrial wastes. Carbohydr. Polym. 25, 203-209.

Israilides, C., Smith, A., Harthill, J.E., Bamett, C., Bambalov, G. and Scanlon, B. (1998) Pullulan content of the ethanol precipitate from fermented agro-industrial wastes. Appl. Microbiol. Biotechnol. 49, 613-617.

Jacquelin, L. and Polulain, R. (1962) The wine and vineyards of France. London: Paul Hamyln Ltd., Drury House.

Jackson, D.I. and Lombard, P.B. (1993) Environmental and management practices affecting grape composition and quality. Am. J. Enol. Vitic. 4, 409-430.

James, S.A., Collins M.D. and Roberts, I.N. (1996) Use of an rRNA internal transcribed spacer region to distinguish phylogenetically closely related species of the genera Zygosaccharomyces and Torulaspora. Int. J. Syst. Bacteriol. 46, 189-194. James, S.A. and Stratford, M. (2003) Spoilage yeasts with emphasis on the genus Zygosaccharomyces. In Yeasts in Food: Beneficial and Detrimental Aspects ed. Boekhout, T. and Robert, V. pp. 171-187. Hamburg: Behr's Verlag. Jarrett, P., Benedict, C.V., Bell, J.P., Cameron, J.A. and Huang, S.J. (1985) Mechanism of the biodégradation of polycaprolactone. In Polymers as biomaterials eds. Shalaby, S.W., Hoffman, A.S., Ratner, B.D. and Horbett, T.A. pp. 181-192. New York: Plenum Press. Jeandet, P., Bessis, R., Sbaghi, M., Meunier, P. and Trollat, P. (1995a) Resveratrol content of wines of different ages: relationship with fungal disease pressure in the vineyard, yiw. J. Enol. Vitic. 46, 1-4. Jeandet, P., Bessis, R., Maume, B.F., Meunier, P., Peyron, D. and Trollat, P. (1995b) Effect of enological practices on the resveratrol isomer content of wine. J. Agric. Food Chem. 43,316-319. Jeandet, P., Douillet Breuil, A.C., Besssis, R., Debord, S., Sbaghi, M. and Adrian, M. (2002) Phytoalexins from the Vitaceae: biosynthesis, phytoalexin gene expression in transgenic plants, antifungal activity, and metabolism. J. Agric. Food Chem. 50, 2731- 11 A\. Johnson, L.J., Koufopanou, V., Goddard, M.R., Hetherington, R. and Schäfer, S.M. (2004) Population genetics of the wild yeast Saccharomyces paradoxus. Genetics 166, 43-52. Jolly, N., Augustyn, O. and Pretorius, I. (2003) Occurrence of non-Saccharomyces yeast species over three vintages on four vineyards and grape musts from four production regions of the Western Cape, South Africa. S. Afr. J. Enol. Vitic. 24, 35-42. Josepa, S., Guillamon, J.M. and Cano, J. (2000) PCR differentiation of Saccharomyces cerevisiae from Saccharomyces bayanus/Saccharomycespastorianus using specific primers. F EMS Microbiol. Lett. 193, 255-259. Kennedy, J. (2002) Understanding grape berry development. Practical Winery and Vineyard July/August, 14-18. Khan, W., Augustyn, O.P.H., van der Westhuizen, T.J., Lambrechts, M.G. and Pretorius, I.S. (2000) Geographic distribution and evaluation of Saccharomyces cerevisiae strains isolated from vineyards in the warmer, inland regions of the Western Cape in South Africa. S Afr. J. Enol. Vitic. 21,17-31. Kiesling, T., Diaz, M. R., Stazell-Tallman, A. and Fell, J.W. (2002a) Field identification of marine yeasts using DNA hybridization macroarrays. In Fungi in Marine Environments eds. Hyde, K.D., Moss, S.T. and Vrijmoed, L.L.P. pp. 69-80. Hong Kong: Fungal Diversity Press. Kiesling T.L., Wilkinson, E., Rabalais, J., Ortner, P.B., McCabe, M.M. and Fell, J.W. (2002b) Rapid identification of adult and naupliar stages of copepods using DNA hybridization methodology. Mar. BiotechnoL 4, 30-39.

Kistler, H.C. (1997) Genetic diversity in the plant-pathogenic fungus Fusarium oxysporum. PhytopathoL 87, 474-479.

Kockovâ-Kratochvilovâ, A., Cemâkovâ, M. and Slâvikovâ, E. (1980) Morphological changes during the life cycle of Aureobasidiumpullulans (de Bary) Arnaud. Folia Microbiol. 25, 56-67.

Kockovâ-Kratochvilovâ, A. and Hronskâ, L. (1980) Effect of sugars on the morphogenesis oi Aureobasidium pullulans. Folia Microbiol. 26, 217-220.

Köhl, J., Bélanger, R.R. and Fokkema, M.J. (1997) Interaction of four antagonistic fungi with Botrytis aclada in dead onion leaves: a comparitive microscopic and ultrastrucutral study. PhytopathoL 87, 634-642.

Kosse, D., Seiler, H., Amann, R., Ludwig, W. and Scherer, S. (1997) Identification of yoghurt-spoiling yeasts with 18S rRNA-targeted oligonucleotide probes. System. Appt. Microbiol. 20, 468-480.

Kosuge, T. and Hewitt, B.M.W. (1964) Exudates of grape berries and their effect on germination of conidia of Botrytis cinerea. Phytopathology 34, 167-172.

Kowalchuk, G.A., Bodelier, P.L.E., Heilig, G.H.J., Stephen, J.R. and Laanbroek, H.J. (1998) Community analysis of ammonia-oxidising bacteria, in relation to oxygen availability in soils and root-oxygenated sediments, using PCR, DGGE and oligonucleotide probe hybridisation. F EMS Microbiol. Ecol. 27, 339-350.

Kresk, M. and Wellington, E.M. (1999) Comparison of different methods for the isolation and purification of total community DNA from soil. J. Microbiol Meth. 39, 1- 16.

Kroemer, K. and Krumbholz, G. (1931) Untersuchungen über osmophile Sprosspilze. I. Beiträge zur Kenntnis der Gärungsvorgänge und der Gärungserreger der Trockenbeerenauslesen. ^rcÄ. Microbiol. 2, 352-410.

Kudanga, T. and Mwenje, E. (2005) Extracellular cellulase production by tropical isolates of Aureobasidium pullulans. Can. J. Microbiol. 51, 773-776.

Kunkee, R.E. and Amerine, M.A. (1970) Yeasts in wine-making. In The Yeasts First Edition Vol. 1 eds Rose, A.H. and Harrison, J.S. pp. 50-71. London: Academic Press.

Kunkee, R.E. and Bisson, L.F. (1993) Wine-making yeasts. In The Yeasts, Second Edition Vol. 5 eds Rose, A.H. and Harrison, J.S. pp. 69-127. London: Academic Press.

Kunkee, R.E. and Goswell, R.W. (1977) Table wines. \n Alcoholic beverages ed. Rose, A.H. pp. 315-386. London: Academic Press. Kurata, S., Kanagawa, T., Magariyama, Y., Takatsu, K., Yamada, K., Yokomaku, T. and Kamagata, Y. (2004) Réévaluation and reduction of a PCR-bias caused by reannealing of templates. Appl. Environ. Microbiol. 70, 7547-7549.

Kurtzman, C.P. (2006) Zygosaccharomyces and related genera. In Food spoilage microorganisms ed. Blackburn, C. de W. pp. 289-301. Cambridge: Woodhead Publishing Ltd.

Kurtzman, C.P., Boekhout, T., Robert, V., Fell, J.W. and Deak, T. (2003) Methods to identify yeasts. In Yeasts in food: beneficial and detrimental aspects eds. Boekhout, T. and Robert, V. pp. 69-121. Hamburg: Behr's Verlag.

Kurtzman, C.P. and Fell, J.W. (2006) Yeast systematics and phylogeny - implications of molecular identification methods for studies in ecology. In Biodiversity and Ecophysiology of Yeasts eds. Rosa, C. and Péter, G. pp. 11-30. Berlin: Springer Verlag.

Kurtzman, C.P. and Droby, S. {200\) Metschnikowiafructicola, a new ascosporic yeast with potential for biocontrol of postharvest fruit rots. Syst. Appl. Microbiol 24, 395- 399.

Kurtzman, C.P. and Robnett, C. (1998) Identification and phylogeny of ascomycetous yeasts from analysis of nuclear large subunit (26S) ribosomal DNA partial sequences. Antonie van Leeuwenhoek. 73, 331-371.

Kurtzman C.P. and Robnett, C.J. (2003) Phylogenetic relationships among yeasts of the 'Saccharomyces complex' determined from multigene sequence analyses. FEMS Yeast Res. 3,417-432.

Kushon, S.A., Jordan, J.P., Seifer, J.L., Nielsen, H., Nielsen, P.E. and Armitage, B.A. (2001) Effect of secondary structure on the thermodynamics and kinetics of PNA hybridization to DNA hairpins. J. Am. Chem. Soc. 123, 10805-10813.

Lagace, L.S. and Bisson, L.F. (1990) Survey of yeast acid proteases for effectiveness of wine haze reduction, J. Enol. Vitic. 41, 147-155.

Lambrechts, M.G. and Pretorius, I.S. (2000) Yeasts and its importance to wine aroma - A review. S. Afr. Enol. Vitic. 21, 97-129.

Laurent, J. C. (1998) La pourriture acide. Progrès Agricole et Viticole 115, 7-9.

Lazaridou, A., Biliaderis, C.G., Roukas, T. and Izydorczyk, M. (2002) Production and characterization of pullulan from beet molasses using a nonpigmented strain of Aureobasidium pullulans in batch culture. Appl. Biochem. Biotechnol. 97, 1-22.

Leathers, T.D. (1986) Color variants oí Aureobasidium pullulans overproduce xy lañase with extremely high specific activity. Appl. Environ. Microbiol. 52, 1026-1030. Leathers, T.D. (2002) Pullulan. In Biopolymers, vol 6. Polysaccharides II: Polysaccharides from eukaryotes eds. Vandamme, E.J., De Baets, S., and Steinbüchel, A. pp 1-35. Weinheim: Wiley-VCH.

Leathers, T.D. (2003a) Biotechnological production and applications of pullulan. Appl Microbiol. BiotechnoL 62, 468-473.

Leathers, T.D. (2003b) Bioconversions of maize residues to value-added coproducts using yeast-like fungi. FEMS Yeast Res. 3, 133-140.

Le Roux, G., Eschenbruch, R. and De Bruin, S.I. (1973) The microbiology of South African wine-making. Part VIL The microflora of healthy and Botrytis cinerea affected grapes. Phytophylactica 5, 51-54.

Leben, C. (1972) Micro-organisms associated with plant buds. J. Gen. Microbiol. 71, 327-331.

Lee, T.H. and Simpson, R.F. (1993) Microbiology and chemistry of cork taints in wine. In Wine microbiology and biotechnology, ed. Fleet, G.H. pp. 353-369. Switzerland: Harwood Academic Publishers.

Leong S.L., Hocking, A.D., Pitt, J.I., Kazi, B.A., Emmett, R.W. and Scott, E.S. (2006b) Aspergillus species in Australian vineyards: from soil to ochratoxin A in wine. Adv. Exp Med Biol 571,153-171.

Leong, S.L., Hocking, A.D. and Scott, E.S. (2006a) Survival and growth of Aspergillus carbonarius on wine grapes before harvest. Int. J. Food Microbiol. lllSl, S83-S87.

Leone, R. and Breuil, C. (1999) Biodegradation of aspen steryl esters and waxes by two filamentous fungi with or without other carbon sources. World J. Microbiol. BiotechnoL 15, 723-727.

Leverentz, B., Conway, W.S., Janisiewicz, W., Abadias, M., Kurtzman, C.P. and Camp, M.J. (2006) Biocontrol of the food-borne pathogens Listeria monocytogenes and Salmonella enterica Serovar Poona on fresh-cut apples with naturally occurring bacterial and yeast antagonists. Appl. Environ. Microbiol. 72, 1135-1140.

Li, S., Spear, R.N., and Andrews, J.H. (1996) Development of an oligonucleotide probe for Aureobasidium pullulans based on the small-subunit rRNA gene. Appl. Environ. Microbiol 62, 1514-1518.

Li, S., Spear, R.N., and Andrews, J.H. (1997) Quantitative fluorescence in situ hybridization of Aureobasidium pullulans on microscope slides and leaf surfaces. Appl Environ. Microbiol 63, 3261-3267.

Libkind, D., Perez, P., Sommaruga, R., Dieguez-Mdel, C., Ferraro, M., Brizzio, S., Zagarese, H. and van Broock, M. (2004) Constitutive and UV-inducible synthesis of photoprotective compounds (carotenoids and mycosporines) by freshwater yeasts. Photochem. Photobiol Sei. 3, 281-286. Liesack, W., Weyland, H. and Stackebrandt, E. (1991) Potential risks of gene amplification by PGR as determined by 16S rDNA analysis of a mixed culture of strict barophilic bactQr'm. Microbial EcoL 21, 191-198.

Lima, G., Ippolito, A., Nigro, F., and Salerno, M. (1997) Effectiveness of Aureobasidium pullulans and Candida oleophila against postharvest strawberry rots. Postharvest Biol. TechnoL 10, 169-178.

Lindlow, S.E. and Brandl, M.T. (2003) Minireview: Microbiology of the phylloshere. Appl Emiron. Microbiol. 69, 1875-1883.

Lindsley, M.D., Hurst, S.F., Iqbal, N.J. and Morrison, C.J. (2001) Rapid identification of dimorphic and yeast-like fungal pathogens using specific DNA probes. J. Clin. Microbiol. 39,3505-3511.

Lingappa, Y., Sussman, A.S., and Bernstein, LA. (1963) Effect of light and media upon growth and melanin formation in Aureobasidium pullulans (DE By.) ARN {=Pullularia pullulans). Mycopath. Mycol. Appl. 20, 109-128.

Liu, S., and Steinbüchel, A. (1997) Production of poly(malic acid) from different carbon sources and its regulation m Aureobasidium pullulans. Biotechnology Letters 19, 11-14.

Liu, Y. and Rauch, C.B. (2003) DNA probe attachment on plastic surfaces with microfluidic hybridization array channel devices with simple oscillation. Analytical Biochem. 317, 76-84.

Longo, E., Cansado, J., Agrelo, D. and Villa, T.G. (1991) Effect of climatic conditions on yeast diversity in grape musts from northwest Spain. Am. J. Enol. Vitic. 42, 141-144.

Lonvaud-Funel, A. (1999) Lactic acid bacteria in the quality improvement and depreciation of wine. Antonie van Leeuwenhoek 76, 317-331.

Lopes, C.A. van Broock, M., Querol, A. and Caballero, A.C. (2002) Saccharomyces cerevisiae wine yeast populations in a cold region in Argentinean Patagonia. A study at different fermentation scales. J. Appl. Microbiol. 93, 608-615.

López, v., Querol, A., Ramon, D. and Fernández-Espinar, M.T. (2001) A simplified procedure to analyze mitochondrial DNA from industrial yeasts. Int. J. Food Microbiol. 68, 75-81.

Loureiro, V. and Malfeito-Ferreira, M. (2003) Spoilage yeasts in the wine industry. Int. J. Food Microbiol. 86, 23-50.

Lugauskas, A., Prosychevas, I., Levinskaite, L. and Jaskelevicius, B. (2004) Physical and chemical aspects of long-term biodeterioration of some polymers and composites. Environ. Toxicol. 19, 318-328. Lumbusch, H.T. and Lindemuth, R. (2001) Major lineages of Dothideomycetes (Ascomycota) inferred from SSU and LSU rDNA sequences. Mycol. Res. 105, 901-908.

Magyar, I. and Bene, Z. (2006) Morphological and taxonomic study on mycobiota of noble rotted grapes in the Tokaj wine district. Acta Alimentaria 35, 237-246.

Mangiarotti, A.M., Picco, A.M., Crippa, A. and Savino, E. (1987) Fungi on phylloplane of treated and not treated vineyard. Rivisa di Patologia Vegetale. 23, 27-37.

Manteau, S., Lambert, B., Jeandet, P. and Legendre, L. (2003) Changes in chitinase and thaumatin-like pathogenesis-related proteins of grape berries during the champagne winemaking process. Am. J. EnoL Vitic. 54, 267-272.

Manzanares, P., Ramón, D. and Querol, A. (1999) Screening of non-Saccharomyces wine yeasts for the production of P-D-xylosidase activity. Int. J. Food Microbiol 46, 105-112.

Manzanares, P., Rojas, V., Genovés, S. and Vallès, S. (2000) A preliminary search for anthocyanin-P-D-glucosidase activity in non-Saccharomyces wine yeasts. Int. J. Food Sci. Technol 35, 95-103.

Markham, N.R. and Zucker, M. (2005) DINAMelt web server for nucleic acid melting prediction. Nucleic Acids Res. 33, W577-W581.

Martini, A. (1993) Origin and domestication of the wine yeast Saccharomyces cerevisiae. J. Wine Res. 177-185.

Marchai, R., Berthier, L., Legendre, L., Marchal-Delahaut, L., Jeandet, P., and Maujean, A. (1998) Effects of Botrytis cinerea infection on the must protein electrophoretic characteristics. J. Agrie. Food Chem. 46, 4945-4949.

Marois, J.J., Bledsoe, A.M. and Gubler, W.D. (1985) Effect of surfactants on epicuticular wax and infection of grape berries by Botrytis cinerea. Phytopathology 75, 1329.

Marshall, M.N., Cocolin, L., Mills, D.A. and VanderGheynst, J.S. (2003) Evaluation of PGR primers for denaturing gradient gel electrophoresis analysis of fungal communities in compost. J. Appl. Microbiol. 95, 934-948.

Martin, J.T., Batt, R.F. and Burchill, R.T. (1957) Fungistatic properties of apple leaf wax. Nature 180, 796-797.

Martini, A., Federichi, F. and Rosini, G. (1980) A new approach to the study of yeast ecology of natural substrates. Can. J. Microbiol. 26, 856-859.

Martini, A. (1993) Origin and domestication of the wine yeast Saccharomyces cerevisiae. J. Wine Res. 165-176. Martini, A., Ciani, M. and Scorzetti, G. (1996) Direct enumeration and isolation of wine yeasts from grape surfaces. Am. J. Enol. Vitic. 47, 435-440.

Martorell, P., Querol, A. and Fernández-Espinar, M.T. (2005) Rapid identification and enumeration of Saccharomyces cerevisiae cells in wine by real-time PGR. Appl. Environ. Microbiol 71, 6823-6830.

Masoud, W., Gesar, LB., Jespersen, L. and Jakobsen, M. (2004) Yeast involved in fermentation of Coffea arabica in East Africa determined by genotyping and by direct denaturing gradient gel electorphoresis. Yeast 21, 549-556.

Matteson Heidenreich, M.G., Gorral-Garcia, M. R., Momol, E. A., and Burr, T. J. (1997) Russet of Apple Fruit Gaused by Aureobasidium pullulans and Rhodotorula gluiinis. Plant Disease 81, 337-342.

Matthews, A., Grimaldi, A., Walker, M., Bartowsky, E., Grbin, P. and Jiranek, V. (2004) Lactic acid bacteria as a potential source of enzymes for use in vinification. Appl. Environ. Microbiol. 70, 5715-573 L

Matveeva, O.V., Shabalina, S.A., Nemtsov, V.A., Tsodikov, A.D., Gesteland R.F. and Atkins J.F. (2003) Thermodynamic calculations and statistical correlations for oligo- probe design. Nucleic Acids Res. 31, 4211-4217.

Mayer, A.M. and Staples, R.G. (2002) Laccase: new ftmctions for an old enzyme. Phytochem. 60, 551-565.

McGarthy, J. (2003) Immunological techniques: ELISA. In Detecting pathogens in food ed. McMeekin, T.A. pp. 241-258. Gambridge: Woodhead Publishing Ltd.

McGormack, P. J., Wildman, H.G. and Jeffries, P. (1994) Production of antibacterial compounds by phylloplane-inhabiting yeasts and yeastlike fungi. Appl. Environ. Microbiol. 60, 927-931.

McGormack, P.J., Wildman, H.G. and Jeffries, P. (1995) The influence of moisture on the suppression of Pseudomonas syringae by Aureobasidium pullulans on an artificial leaf surface. FEMS Microbiol. Ecol. 16, 159-166.

McMahon, H., Zoecklein, B.W., Fugelsang, K., and Jasinski, Y. (1999) Quantification of glycosidase activities in selected yeasts and lactic acid bacteria. J. Ind. Microbiol. Biotechnol. 23, 198-203.

Mendes Ferreira, A., Glimaco, M.G. and Mendes Faia, A. (2001) The role of non- Saccharomyces species in releasing glycosidic bound fraction of grape aroma components - a preliminary study. J. Appl. Microbiol. 97, 67-71.

Mercier, J. and Lindlow, S.E. (1999) Role of leaf sugars in colonization of plants by bacterial epiphytes. Appl. Environ. Microbiol. 66, 369-374. Middelhoven, W.J. (1997) Identity and biodegradative abilities of yeasts isolated from plants growing in an arid climate. Antonie van Leeuwenhoek, 72, 81-89.

Millet, V. and Lonvaud-Funel, V. (2000) The viable but non-culturable state of wine micro-organisms during storage. Lett. Appl. Microbiol 30, 136-141.

Mills, D.A., Johannsen, E.A. and Cocolin, L. (2002) Yeast diversity and persistence in botrytis-affected wine fermentations. Appl. Environ. Microbiol. 68, 4884-4893.

Minárik, E. (1965) Ecology of natural species of wine-yeasts in Czechoslovakia. Mikrobiologija 2, 29-37.

Minárik, E. und Rágala, P. (1975) Die selective Wirkung von Rebschutzmitteln auf die Mikroflora von Weintrauben. Mitt. Rebe, Wein, Ostbau und Früchteverwertung 25, 187-204.

Mlikota, F., Males, P. and Cvjetkovic, B. (1996) Effectiveness of five fungicides on grapevine grey mould and their effects on must fermentation. J. Wine Res. 7, 103-110.

Moráis, P.B., Martins, M.B., Klaczko, L.B., Mendonça-Hagler, L.C. and Hagler, A.N. (1995) Yeast succession in the amazon fruit Parahancornia amapa as resource partitioning among Drosophila spp. Appl. Environ. Microbiol. 61, 4251-4257.

Morgan, D. P., and Michailides, T. J. (2004) First report of melting decay of'Red Globe' grapes in California. Plant Disease 88, 1047.

Mortimer, R. and Polsinelli, M. (1999) On the origins of wine yeast. Res. Microbiol. 150, 199-204.

Moruno, E.G., Sanlorenzo, C., Boccaccino, B. and Di Stefano, R. (2005) Treatment with yeast to reduce the concentration of Ochratoxin A in red wine. Am. J. Enol. Vitic. 56,1:73-76.

Mrak, E.M. and McClung, L.S. (1940) Yeasts occurring on grapes and in grape products in California. J. Bacteriol. 40, 395-407.

Müller, C. and Riedere, M. (2005) Plant surface properties in chemical ecology. J. Chem. Ecology n, 2621-2651.

Murphy, C.A., Cameron, J.A., Huang, S.J. and Vinopal, R.T. (1996) Fusarium polycaprolactone depolymerase is cutinase. Appl. Environ. Microbiol. 62, 456-460.

Muyzer, G. and Smalla, K. (1998) Application of denaturing gradient gel electrophoresis (DGGE) and temperature graident gel electrophoresis (TGGE) in microbial ecology. Antonie van Leeuwenhoek 73, 127-141.

Myburgh, J., Prior, B.A. and Kilian, S.G. (1991) Production of xylan-hydrolyzing enzymes by Aureobasidium pullulans. J. Fermentation Bioeng. 72, 135-137. Myers R.M., Maniatis, T. and Lerman, L.S. (1987) Detection and localization of single base changes by denaturing gradient gel electrophoresis. Methods Enzymol. 155, 501- 527.

Nair, N.G. (1985) Fungi associated with bunch rot of grapes in the Hunter Valley. Anst. J. Agrie. Res. 36, 435-442.

Naef, F., Lim, D.A., Patil, N. and Magnasco, M. (2002) DNA hybridization to mismatched templates: a chip study. Phys. Rev. 65, 40902-40905.

Navarini, L., Bella, J., Flaibani, A., Gilli, R., and Rizza, V. (1996) Structural characterization and solution properties of an acidic branched (1—>3)-beta-D-glucan from Aureobasidiumpullulans. Int. J. Biol. Macromol. 19, 157-63.

Naumov, G., James, S.A., Naumova, E.S., Louis, E.J. and Roberts, I.W. (2000) Three new species of Saccharomyces sensu stricto complex: Saccharomyces cariocanus, Saceharomyces kudriavzevii and Saccharomyces mikatae. Int. J. Syst. Evol. Microbiol. 50, 1931-1942.

Naumov, G.I., Naumova, E.S., Sniegowski, P.D. (1998) Saccharomyces paradoxus and Saccharomyces cerevisiae are associated with exudates of North American oaks. Can. J. Microbiol. 44,1045-1050.

Nemec, S. (1970) Fungi associated with strawberry root rot in Illinois. Mycopathol. Appl. 41, 331-346.

Ng, P.J., Fleet, G.H. and Heard, G.M. (2005) Pesticides as a source of microbial contamination of salad vegetables. Int. J. Food Microbiol. 101, 237-250.

Nikiforov, T.T., Rendle, R.B., Kotewicz, M.L. and Rogers, Y. (1994) The use of phosphorothioate primers and exonuclease hydrolysis for the preparation of single- stranded PGR products and their detection by solid-phase hybridization. PCR Methods and Applications 3, 285-291.

Nissen, P., Nielsen, D. and Ameborg, N. (2003) Viable Saccharomyces cerevisiae cells at high concentrations cause eariy growth arrest of nonSaccharomyces yeasts in mixed cultures by a cell-cell contact-mediated mechanism. Yeast. 20, 331-341.

Oliveira, H., Rego, M.C. and Nascimento, T. (2004) Decline of young grapevines caused by fungi. Acta Horticulturae, 652, 295-304.

Oliver, J.D. (1993) Formation of viable but non cuhurable cells. In Starvation in bacteria ed. Kjelleberg, S. pp. 239-272. New York: Plenum Press.

O'Connor, L. (2002) Detection of Listeria monocytogenes using a PCR/DNA probe assay. In Methods in Molecular Biology, Vol. 216, PCR Detection of Microbial Pathogens: Methods and Protocols eds. Sachse, K. and Frey, J. pp. 65-84. Totowa, NJ: Humana Press. Padgett, M. and Morrison, J.C. (1990) Changes in grape berry exudates during fruit development and their effect on mycelial growth. J. Am. Soc. Hort. Sci. 115, 269-273.

Page, B. and Kurtzman, C.P. (2005) Rapid identification oi Candida species and other clinically important yeast species by flow cytometry. J. Clin. Microbiol 32, 4507-4514.

Pallman, C.L., Brown, J.A., Olineka, T.L. Cocolin, L., Mills, D.A. and Bisson, L.F. (2001) Use of WL medium to profile native flora fermentations. Am. J. Enol. Vitic. 52, 198-203.

Parle, J.N. and di Menna, M.E. (1966) The source of yeasts in New Zealand wines. N.Z. J. Agri. Res. 9, 98-107.

Pelaez, F., Cabello, A., Platas, G., Diez, M.T., Del Val, A.G., Basilio, A., Martân, I., Vicente, F., Bills, G.F., Giacobbe, R.A., Schwartz, R.E., Onishi, J.C., Meinz, M.S., Abruzzo, G.K., Flattery, A.M., Kong, L., and Kurtz, M.B. (2000) The discovery of enfumafungin, a novel antifungal compound produced by an endophytic Hormonema species: biological activity and taxonomy of producing organism. System. Appl. Microbiol. 23, 333-343.

Péter, G., Dlauchy, D., Vasdinyei, R., Tomai-Lehoczki, J. and Déak, T. (2005) Candida gain sp. nov., a new yeast from poultry. Antonie van Leeuwenhoek 86, 105-110.

Percival, D.C., Sullivan, J.A. and Fisher, K.H. (1993) Effect of cluster exposure, berry contact and cultivar on cuticular membrane formation and occurrence of bunch rot {Botrytis cinerea) with 3 Vitis vinifera L. cultivars. Vitis 32, 87-97.

Pardo, I., Garcia, M.J., Zuniga, M. and Uruburu, F. (1990) Dynamics of microbial populations during fermentation of wines from the Utiel-Requena region of Spain. Appl. Environ. Microbiol. 55, 539-541.

Parish, M.E. and Carroll, D.E. (1985) Indigenous yeasts associated with Muscadine {Vitis rotundifolid) grapes and musts. Am. J. Enol Vitic. 36, 165-169.

Parish, M.E. and Higgins, D.P. (1989) Yeasts and moulds isolated from spoiling citrus products and by-products. J. Food Prot. 52, 261-263.

Park, D. (1984). Population density and yeast mycelial dimorphism in Aureobasidium pullulans. Trans Br. mycol Soc. 82, 39-44.

Parle, J.N. and di Menna, M.E. (1966) The source of yeast in New Zealand wines. NZJ. Agri. Res. 9, 98-105.

Pechak D., and Crang, R.E. (1977) An analysis of Aureobasidium pullulans developmental stages by means of scanning electron microscopy. Mycologia 69, 783- 792.

Pentland, G.D. (1967) Ethanol produced by Aureobasidium pullulans and its effect on the growth oiArmillaria mellea. Can. J. Microbiol 13, 1631-1639. Phaff, HJ., Miller, M.W. and Mrak, E.M. (1978) The life of yeasts. Cambridge: Harvard University Press.

Phaff, HJ. and Starmer, W.T. (1987) Yeasts associated with plants, insects and soil. In The Yeasts. Vol. 1, Biology of yeasts eds. Rose, A.H. and Harrison, J.S. pp. 123-180. New York: Academic Press.

Phister, T.G. and Mills, D.A. (2003) Real-time PGR assay for detection and enumeration of Dekkera bruxellensis in wine. Appi. Envi. Microbiol 69, 7430-7434.

Picco, A.M. and Rodolfi, M. (2004) Assessments of indoor fungi in selected wineries of Oltrepò Pavese (Northern Italy) and Sottoceneri (Switzerland). Am. J. Enol. Vitic. 55, 355-362.

Pina, C., Teixeiró, P., Leite, P., Villa, M., Belloch, C. and Brito, L. (2005) PCR- fingerprinting and RAPD approaches for tracing the source of yeast contamination in a carbonated orange juice production chain. J. Appi. Microbiol 98, 1107-1114.

Pitt, J.I and Hocking, A.D. (1997) Fungi and Food Spoilage, second edition. London: Blackie Academic and Professional.

Polz, M.F. and Cavanaugh, C.M. (1998) Bias in template-to-product ratios in multitemplate PGR. Appi. Environ. Microbiol 64, 3724-3630.

Poulard, A. (1982) Assimilation of tartaric acid by Aureobasidium pullulans (DE BARY) ARNAUD. Rev. Franc. Oenol 22, 23-26.

Poulard, A., (1984) Influence of several factors affecting variability of the yeast microflora of musts and wines. Vignes et Vins. 326, 18-21.

Poulard, A., Leclanche, A. and Kollonkai, A. (1983) Degradation of tartaric acid in grape musts by a new yeast species: Exophiala jeanselmei var. heteromorpha. Vignes et Vins. 434, 33-35.

Poulard, A. and Simon, L. (1979) Etude ecologique et métabolique de quelques levures rares du vignoble nantais. Bull. Soc. Se. nat. Ouest de la France 1, 185-196.

Pouland, A., Simon, L. and Guinier, G. (1980) Variabilité de la microflore levurienne de quelques terrois viticoles du pays Nantais. Connaissance Vigne Vin 14, 219-238.

Pozhitkov, A., Noble, P.A., Domazet-Loso, T., Nolte, A.W., Sonnenberg, R., Staehler, P., Beier, M. and Tautz, D. (2006) Tests of rRNA hybridization to microarrys suggest that hybridization characteristics of oligonucleotide probes for species discrimination cannot be predicted. Nucleic Acids Res. 34, e66.

Prakitchaiwattana, G.J., Fleet, G.H. and Heard, G.M. (2004) Application and evaluation of denaturing gradient gel electrophoresis to analyse the yeast ecology of wine grapes. FEMS Yeast Res. 4, 865-877. Prakitchaiwattana, C.J. (2005) Investigation of yeasts associated with Australian wine grapes using cultural and molecular methods. PhD thesis. University of New South Wales.

Pramateftaki, P.V., Lanaridis, P. and Typas, M.A. (2000) Molecular identification of wine yeasts at species or strains level: a case study with strains from two vine-growing areas of Greece. J. Appl Microbiol. 89, 236-248.

Praphailong, W., Van Gestel, M., Fleet, G.H. and Heard, G.M. (1997). Evaluation of the Biolog system for the identification of food and beverage yeasts. Lett. Appl. Microbiol 24, 455-459.

Pratt, C. (1971) Reproductive anatomy in cultivated grapes: A review. Am. J. Enol. Vitic. 22, 92-109.

Pretorius, I.S. (2000) Tailoring wine yeast for the new millennium: novel approaches to the ancient art of winemaking. Yeast 16, 675-729.

Pugh, G.J.F. and Buckley, N.G. {\91 \) Aureobasidiimpullulans: an endophyte in sycamore and other trees. Trans. Br. mycol. Soc. 57, 227-231.

Querol, A., Jiménez, M. and Huerta, T. (1990) Microbiological and enological parameters during fermentation of musts from poor and normal grape harvests in the region of Alicante (Spain). J. Food Sci. 55, 1603-1606.

Qui, X., Wu, L., Huang, P., McDonel, P.E., Palumbo, A.V., Tiedje, J.M. and Zhou, J. (2001) Evaluation of PCR-generated chimeras, mutations and heteroduplexes with 16S rRNA gene-based cloning. Appl. Environ. Microbiol. 67, 880-887.

Radcliffe, D.M. and Holbrook, R. (2000) Detection of microorganisms in food - principles and application of immunological techniques. In The microbiological safety and quality of food Vol. 1 eds. Lund, B.M., Baird-Parker, T.C., and Gould, G.W. pp. 1791-1809. Maryland: Aspen Publishers Inc.

Radler, F. and Horn, D.H. (1965) The composition of grape cuticle wax. Aust. J. Chem. 18, 1059-1069.

Rajaei, H. (1987) Changments cytochimiques et ultra structurarraux des parois cellularies de la pellicule du raisin, Vitis vinifera durant la croissance et al maturation de labaie. Can. J. Botany 65, 1343-1355.

Ramos, S., Garcia Acha, I., and Peberdy, J.F. (1975) Wall structure and the budding process in Pullulariapullulans. Trans. Br. mycol. Soc. 64, 283-288.

Ramos, S., Garcia Acha, I., and Peberdy, J.F. (1975) A vegetative cycle of Pullularia pullulans. Trans. Br. mycol. Soc. 64, 129-135

Rankine, B.C. (1989) Making Good Wine. South Melbourne, Australia: Sun book Raspor, P., Milek, D.M., Polanc, J., Mozina, S.S. and Cadez, N. (2006) Yeasts isolated from three varieties of grapes cultivated in different locations of the Dolenjska vine- growing region, Slovenia. Int. J. Food Microbiol. 109, 97-102.

Rawsthorne, H. and Phister, T.G. (2006) A real-time PGR assay for the enumeration and detection oí Zygosaccharomyces bailii from wine and fruit juices. Int. J. Food Microbiol. 112, 1-7.

Ray, M., Dickinson, D. J. & Buck, M. (2004). Aureobasidium or Hormonema? A genetic approach. Conference proceedings of the International Research Group, Wood Preservation (Ljubljana, Slovenia), IRG Secretariat, Stockholm.

Redzepovic, S., Orlic, S., Sikora, S., Majdak, A. and Pretorius, I.S. (2002) Identification and characterization oíSaccharomyces cerevisiae and Saccharomyces paradoxus strains isolated from Croatian vineyards. Lett. Appl. Microbiol. 35, 305-310.

Regueiro, L.A., Costas, C.L. and Rubio, J.E.L. (1993) Influence of viticultural and enological practices on the development of yeast populations during winemaking. Am. J. Enol. Vitic. 44, 405-408.

Rementiera, A., Rodriguez, J., Cadaval, A., Amenazar, R., Muguruza, F., Hernando, F. and Sevilla, M. (2003) Yeast associated with the spontaneous fermentations of white wines from the 'Txakoli de Bizkaia' region (Basque Country, North Spain). Int. J. Food Microbiol. 86, 201-207.

Renouf, V., Claisse, O. and Lonvaud-Funel, A. (2005) Understanding the microbial ecosystem on the grape berry surface through numeration and identification of yeast and bacteria. Aust. J. Grape and Wine Res. 11, 316-327.

Ribéreau-Gayon, P., Dubourdieu, D., Donèche, B. and Lonvaud, A. (2000) Handbook of Oenology Vol. 1, The microbiology of wine and vinifications. England: John Wiley & Sons.

Roberts, R.J., Vincze, T., Posfai, J. and Macelis, D. (2005) REBASE - restriction enzymes and DNA methyltransferases. Nucleic Acids Res. 33, D230-D232.

Rodriguez, M.E., Lopes, C.A. van Brock, M., Valles, S., Ramón, D. and Caballero, A.C. (2004) Screening and typing of Patagonian wine yeasts for glycosidase activities. J. Appl. Microbiol. 96, 84-95.

Rogiers, S.Y., Hatfield, J.M., Jaudzems, V.G., White, R.G. and Keller, M. (2004) Grape berry cv. Shiraz epicuticular wax and transpiration during ripening and preharvest weight loss. Am. J. Enol. Vitic. 55, 121-127.

Roling, W.F.M., Kerier, J., Braster, M., Apriyantono, A., Stam, H. and van Verseveld, H.W. (2001) Microorganisms with a taste for vanilla: microbial ecology of traditional Indonesian vanilla curing, ^p/?/. Environ. Microbiol. 67, 1995-2003. Romano, P., Fiore, C., Paraggio, M., Caruso, M. and Capece, A. (2003) Function of yeast species and strains in wine flavour. Int. J. Food Microbiol. 86, 169-180.

Rosenquist, J.K. and Morrison, J.C. (1989) Some factors affecting cuticle and wax accumulation on grape berries. Am. J. Enol. Vitic. 40, 241-244.

Rosi, I., Vinella, M. and Domizio, P. (1994) Characterization of p-glycosidase activity in yeasts of oenological origin. J. Appl. Bacteriol. 77, 519-527.

Rosini, G., Federici, F. and Martini, A. (1982) Yeast flora of grape berries during ripening. Microbial Ecol 8, 83-89.

Roukas, T. (1999) Pullulan production from brewery wastes by Aureobasidium pullulans. World J. Microbiol. BiotechnoL 15, 447-450.

Roukas, T. (2000) Aureobasidium. In Encyclopedia of Food Microbiology eds. Robinson, R.K., Batt, C.A., and Patel, P.D. pp. 109-112. London: Academic Press.

Rousseau, S. and Doneche, B. (2001) Effects of water activity (aw) on the growth of some epiphytic microorganisms isolated from grape beny. Vitis 40, 75-78.

Sabate, J., Cano, J., Esteve-Zarzoso, B. and Guillamón, J.M. (2002) Isolation and identification of yeasts associated with vineyard and winery by RFLP analysis of ribosomal genes. Microbiol. Res. 157, 1-8.

Sage, L., Krivobok, S., Delbos, E., Seigle-Murandi, F. and Creppy, E.E. (2002) Fungal flora and ochratoxin a production in grapes and musts from France. J Agrie. Food C/ze/w. 50,1306-1311.

Saha, B.C. and Bothast, R.J. (1998) Purification and characterization of a novel thermostable a-L-Arabinofuranosidase from a color-variant strain of Aureobasidium pullulans. Appl. Environ. Microbiol. 64, 216-220.

Sampaio, J.P., Gadanho, M., Santos, S., Duarte, F.L., Pais, C., Fonseca, A. and Fell, J.W. (2001) Polyphasic taxonomy of the basidiomycetous yeast genus Rhodosporidium: Rhodosporidium kratochvilovae and related anamorphic species. Int. J. Syst. and Evol. Microbiol. 51, 687-697.

Sanchez, J.G., Tsuchii, A. and Tokiwa, Y. (2000) Degradation of polycaprolactone at 50°C by a \hQnn0X0\tx3nX Aspergillus sp. BiotechnoL Lett. 22, 849-853.

Sánchez-Torres, P., González-Candelas, L. and Rámon, D. (1998) Heterologous expression of a Candida molischiana anthocyanin-p-glucosidase in a wine yeast strain. J. Agrie. FoodChem. 46, 354-360.

Sanderson, F.R. (1965) Description and epidemiology of Guignardia fulvida sp. nov., the ascogenous state of Aureobasidium pullulans var. lini (Lafferty) Cooke. N. Zl. J. Agrie. Res. 8, 131-141. Sancho, T., Giménez-Jurado, G., Malfeito-Ferreira, M. and Loureiro, V. (2000) Zymological indicators: a new concept applied to the detection of potential spoilage yeast species associated with fruit pulps and concentrates. Food Microbiol. 17, 613-624.

Sasaki, Y. and Yoshida, T. (1959) Distribution and classification studies on the wild yeasts or budding fungi on the fresh fruits of Hokkaido. J. Fac. Agr. Hokkaido Univ. 51, 194-220.

Schäfer, W. (1993) The role of cutinase in ftingal pathogenicity. Trends Microbiol. 1, 69-61.

Schena, L., Ippolito, A ., Zahavi, T., Cohen, L., Nigro, F., and Droby, S. (1999) Genetic diversity and biocontrol activity of Aureobasidiumpullulans isolates against postharvest rots. Postharvest Biol and Technol. 17, 189-199.

Schena, L., Sialer, M.F. and Gallitelli, D. (2002) Molecular detection of strain L47 of Aureobasidium pullulans, a biocontrol agent of postharvest diseases. Plant Disease 86, 54-60.

Schena, L., Nigro, F., Pentimone, I., Ligorio, A., and Ippolito, A. (2003) Control of postharvest rots of sweet cherries and table grapes with endophytic isolates of Aureobasidium pullulans. Postharvest Biol. Technol. 30, 209-220.

Schoeman, W., and Dickinson, D. J. {1996) Aureobasidium pullulans can utilize simple aromatic compounds as a sole source of carbon in liquid culture. Lett. Appl Microbiol. 22,129-131.

Scorzetti, G., Fell, J.W., Fonseca, A. and Stazell-Tallman, A. (2002) Systematics of basidiomycetous yeasts: a comparison of large subunit D1/D2 and internal transcribed rDNA regions. FEMS Yeast Res. 2, 495-517.

Schuller, D., Alves, H., Dequin, S. and Casal, M. (2005) Ecological survey of Saccharomyces strains from vineyards in the Vinho Verde Region of Portugal. FEMS Microbial. Ecol. 51, 167-177.

Schultz, M.J. and Thormann, M.N. (2005) Functional and taxonomic diversity of saprobic filamentous fungi from Typha latifolia from central Alberta, Canada. Wetlands, 25, 675-684.

Schweigkofler, W. and Prillinger, H. (1997) Untersuchung von endophytischen und latent pathogenen Pilzen aus Rebholz in Österreich und Südtirol. Mitt. Klosterneuburg 47, 149-158.

Serra, R., Braga, A. and Venáncio, A. (2005) Mycotoxin-producing and other fungi isolated from grapes for wine production, with particular emphasis on ochratoxin A. Res. Microbiol. 156, 515-512.

Serra, R., Mendoza, C. and Venáncio, A. (2006) Fungi and ochratoxin A detected in healthy grapes for wine production. Lett. Appl. Microbiol. 42, 42-47. Seviour, RJ., Stasinopoulos, SJ., Auer, D.P.F. and Gibbs, P.A. (1992) Production of pullulan and other exopolysaccharides by filamentous fungi. Crit. Rev. BiotechnoL 12, 279-298.

Simon, L. and Poulard, A. (1979) Présence de I'Aureobasidium pullulans (de Bary) Arnaud dans le vignoble nantais. Etude microbiologique et écologique. Bull. Soc. Sci. nat. Ouest France 1, 57-68.

Sipiczki, M. (2003) Candida zemplinina sp. nov., an osmotolerant yeast that ferments sweet botrytized wines. Int. J. Syst. Evol. Microbiol. 53, 2079-2083.

Sipiczki, M. (2006) Metschnikowia strains isolated from botrytized grapes antagonize fungal and bacterial growth by iron depletion. Appl. Envi. Microbiol. 72, 6716-6724.

Sivanesan, A. (1984) The bitunicate ascomycetes and their anamorphs. Vaduz: Lubrecht and Cramer.

Slâvikovâ, E. and Vadkertikova, R. (1997) Seasonal occurrence of yeasts and yeast-like organisms in the river Danube. Anionic van Leeicwenhoek 72, 77-80.

Sniegowski, P.D., Dombrowski, P.G. and Fingerman, E. (2002) Saccharomyces cerevisiae and Saccharomyces paradoxus coexist in a natural woodland site in North America and display different levels of reproductive isolation from European conspecifics. FEMS Yeast Res 1, 299-306.

Spatafora, J.W., Mitchell, T.G. and Vilgalys, R. (1995) Analysis of genes coding for small-subunit rRNA sequences in studying phylogenetics of dematiaceous fungal pathogens. J. Clin. Microbiol. 33, 1322-1326.

Spayd, S.E., Tarara, J.M., Mee, D.L. and Ferguson, J.C. (2002) Separation of sunlight and temperature effects on the composition of Vitis vinifera cv. Merlot berries. Am. J. Enol. Vitic. 53, 171-182.

Speksnijder, A.G.C.L., Kowalchuk, G.A., de Jong, S., Kline, E., Stephen, J.R. and Laanbroek, H.J. (2001) Microvariation artefacts introduced by PGR and cloning of closely related 16S rRNA gene sequences. Appl. Environ. Microbiol. 67, 469-472.

Sponholz, W. (1993) Wine spoilage by microorganisms. In Wine microbiology and biotechnology ed. Fleet, G.H. pp. 395-429. Chur, Switzerland: Harwood Academic Publisher.

Stender, H., Fiandaca, M., Hyldig-Nielsen, J.J., and Coull, J. (2002) PNA for rapid microbiology. J Microbiol. Methods 48, 1-17.

Stender H., Kurtzman C., Hyldig-Nielsen, J.J., S0rensen D., Broomer A., Oliveira K., Perry-O'Keefe H., Sage A., Young B. and Coull, J. (2001) Identification of Dekkera burxellensis (Brettanomyces) from wine by fluorescence in situ hybridization using peptide nucleic acid probes. Appl. Environ. Microbiol. 67, 938-941. Starmer, W.T., Ganter, P.F., Aberdeen, V., Lachance, M.-A. and Phaff, HJ. (1987) The ecological role of killer yeasts in natural communities of yeasts. Can. J. Microbiol. 33, 783-796.

Sterflinger, K. (2006) Black yeasts and meristematic fungi: ecology, diversity and identification. In Biodiversity and Ecology of Yeasts eds. Rosa, C.A. and Peter, G. pp 501-514. Berlin: Springer.

Sterflinger, K. and Prillinger, H.J. (2001) Molecular taxonomy and biodiversity of rock fungal communities in an urban environment (Vienna, Austria). Antonie van Leeimenhoek 80, 275-286.

Strauss, M.L.A., Jolly, N.P., Lambrechts, M.G. and van Rensburg, P. (2001) Screening for the production of extracellular hydrolytic enzymes by x\or\-Saccharomyces wine yeasts. J. Appl. Microbiol. 91, 182-190.

Suárez, J.A., González, M.C., Calljo, M.J., Colomo, B. and González, A. (1994) Contribution to the study of varietal wines from Rioja and Navarra. Bulletin de L 'OLIV 759-760, 398-407.

Subden, R.E., Husnik, J.I, van Twest, R., van der Merwe, G. and van Vuuren, H.J.J. (2003) Autochthonous microbioal population in a Niagara Peninsula icewine must. Food Res. Int. 36, 747-751.

St. Leger, R.J., Joshi, L. and Roberts, D.W. (1997) Adaptation of proteases and carbohydrases of saprophytic, phytopathogenic and entomopathogenic fungi to the requirements of their ecological niches. Microbiol. 143, 1983-1992.

Suzuki, M.T. and Giovannoni, S.J. (1996) Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Appl Environ. Microbiol. 62, 625-630.

Svejcar, V. (1968) Die Hefeflora an Trauben der Weinrebe in der Weinbergen der Landwirtschaftlichen Hochschule von lednice na Morave. Die Wein-wissenschaft 6, 251-254.

Swiegers, J.H. and Pretorius, LS. (2005) Yeast modulation of wine flavour. Adv. Appl. Microbiol. 57, 131-175.

Takeo, K., and Hoog, G.S. de (1991) Karyology and hyphal characters as taxonomic criteria in Ascomycetous black yeast and related fungi. Antonie van Leeuwenhoek 60, 35-42.

Takesako, K., Ikai, J., Haruna, F., Endo, M., Shimanaka, K., Sono, E., Nakamura, T. and Kato, 1. (1991) Aureobasidins, new antifungal antibiotics. Taxonomy, fermentation, isolation and properties. J. Antibiotics. 44, 919-924. Talley, S. M., Coley, P. D. and Kursar, T. A. (2002) The effects of weather on fungal abundance and richness among 25 communities in the Intermountain West. BioMed Central Ecol. 2, 7 http://www.biomedcentral.eom/1472-6785/2/7.

Tanaka, H., Muguruma, M. and Ohta, K. (2006) Purification and properties of a family- 10 xylanase from Aureobasidium pullulans ATCC 20524 and characterization of the encoding gene. Appl. Microbiol. Biotechnol. 70, 202-211.

Taylor, P.E., Esch, R., Flagan, R.C., House, J., Tran, L. and Glovskya, M. M. (2006) Identification and possible disease mechanisms of an under-recognized fungus. Int. Arch. Allergy and Immunol. 139, 45-52.

Taylor, J.W., Jacobson, D.J., and Fisher, M.C. (1999) The evolution of asexual fungi: Reproduction, speciation and classification. Annu. Rev. Phytopathol. 37, 197-246.

Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673- 4680.

Thorn, R.G., Ginn, J.H. and Malloch, D.W. (1998) Leucogyrophana lichenicola and the similar!, romellii. Can J. Bat. 76, 686-693.

Török, T., Mortimer, R.K., Romano, R., Suzzi, G. and Polsinelli, M. (1996). Quest for wine yeasts - an old story revisited. J. Ind. Microbiol. 17, 303-313.

Torriani, S., Zapparoli, G., Malacrino, P., Suzzi, G. and Dellaglio, F. (2004) Rapid identification and differentiation of Saccharomyces cerevisiae, Saccharomyces bayanus and their hybrids by multiplex PCR. Lett. Appl. Microbiol. 38, 239-244.

Tracey, J. and Saunders, G. (2003) Bird damage to the wine grape industry. Report to the Bureau of Rural Sciences, Dept. Agriculture, Fisheries and Forestry. Oct, Vertebrate Pest Research Unit NSW Agriculture.

Urzi, C., De Leo, F., Lo Passo C. and Criseo, G. (1999) Intra-specific diversity of Aureobasidium pullulans strains isolated from rocks and other habitats assessed by physiological methods and by random amplified polymorphic DNA (RAPD). J. Microbiol. Methods 36, 95-105.

Uzunovic, A., Yang, D.Q., Gagné, P., Breuil, C., Bemier, L., Byrne, A., Gignac, M., and Kim, S.H. (1999) Fungi that cause sapstain in Canadian softwoods. Can. J. Microbiol. 45, 914-922.

Vadkertikova, R. and Slâvikovâ, E. (1995) Killer activity of yeasts isolated from the water environment. Can. J. Microbiol. 41, 759-766.

Valero, E., Schuller, D., Cambona, B., Casal, M. and Dequin, S. (2004) Dissemination and survival of commercial wine yeast in the vineyard: A large-scale, three-years study. FEMS Yeast Res. 5, 959-969. Vandaras, S., Teixeira, M.J., Marques, J.C., Aguiar, A., Alves, A. and Bastos, M. (2004) Glucose and fructose levels on grape skin: interference in Lobesia botrana behaviour. Analytica Chemica Acta 513, 351-355.

Vasdinyei R. and Deak, T. (2003) Characterization of yeast isolates originating from Hungarian dairy products using traditional and molecular identification techniques. Int. J. Food Microbiol. 86, 123-130.

Vaughn-Martini, A. and Martini, A. (1995) Facts, myths and legends on the prime industrial microorganism. 7. Ind. Microbiol. 14, 514-522.

Van Den Heuvel, J. (1969) Effects oiAureobasidium pullulans on numbers of lesions on dwarf bean leaves caused by Alternaria zinniae. Eur. J. Plant Pathology 75, 300- 307.

van der Walt, J.P. and Yarrow, D. (1984) Methods for the isolation, maintenance, classification and identification of yeasts. In The Yeasts, a Taxonomic Study, Third Edition ed Kreger-van Rij, N.J. pp 45-104. Amsterdam: Elsevier.

van der Westhuizen, T.J., Augustyn, O.P.H. and Pretorius, I.S. (2000a) Geographical distribution of indigenous Saccharomyces cerevisiae strains isolated from vineyards in the coastal regions of the Western Cape in South Africa. S. Afr. J. Enol. Vitic. 21, 3-9.

van der Westhuizen, T.J., Augustyn, O.P.H. and Pretorius, I.S. (2000b) Variation of indigenous Saccharomyces cerevisiae strains isolated from vineyards of the Western Cape in South Africa. S Afr. J. Enol. Vitic. 21, 10-16.

van Keulen, H., Lindmark, D.G., Zeman, K.E. and Gerlosky, W. (2003) Yeasts present during spontaneous fermentation of Lake Erie Chardonnay, Pinot Gris and Riesling. Antonie van Leeuwenhoek S3, 149-2003.

Van Zyl, J. A. and Du Plesis, L. de W. (1961) The microbiology of South African winemaking. Part 1. The yeasts occurring in vineyards, musts and wines. S. Afr. J. Agric. Sci. 4, 393-403.

Verhoef, R., de Waard, P., Schols, H.A., Ratto, M., Siika-aho, M. and Voragen, A.G.J. (2002) Structural elucidation of the EPS of slime producing Brevundimonas vesicularis sp. isolated from a paper machine. Carbohydr. Res. 337, 1821-1831.

Versavaud, A., Courcoux, P., Roulland, C., Dulau, L. and Hallet, J.N. (1995) Genetic diversity and geographical distribution of wild Saccharomyces cerevisiae strains from the wine-producing area of Charentes, France. Appl. Environ. Microbiol 61, 3521- 3529.

Verstrepen, K.J., Chambers, P.J. and Pretorius, I.S. (2006) The development of superior yeast strains for the food and beverage industries: challenges, opportunities and potential benefits. In In Yeasts in Food and Beverages eds. Querol, A. and Fleet, G.H. pp 399-432. Beriin: Springer. Viljoen, B. (2006) Yeast ecological interactions. Yeast-yeast, yeast-bacteria, yeast-fungi interactions and yeasts as biocontrol agents. In Yeasts in Food and Beverages eds. Querol, A. and Fleet, G.H. pp 83-100. Berlin: Springer.

Villena, M.A., Úbeda Iranzo, J., and Briones Pérez, A. (2006) Relationship between Debaryomyces pseudopolymorphus enzymatic extracts and release of terpenes in wine. Biotechnol. Progress. 22, 375-381.

Wade, J., Holzapfel, B., DeGaris, K., Keller, M. and Williams, D. (2004) Nitrogen and water management strategies for wine-grape quality. Acta Hortic. 640, 61-67.

Walton, J.D. (1994) Deconstructing the cell wall. Plant Physiol. 104, 1113-1118.

Wang, G.C.Y. and Wang, Y. (1997) Frequency of formation of chimeric molecules as a consequence of PGR coamplification of 16S rRNA genes from mixed bacterial genomes. Appl. Environ. Microbiol 63, 4645-4650.

Webb, J.S., Nixon, M., Eastwood, I.M., Greenhalgh, M., Robson, G.D., Read, S.J. and Handley, P.S. (2000) Fungal colonization and biodeterioration of plasticized polyvinyl chloride. Appl. Environ. Microbiol. 66, 3194-3200.

White, T.J., Bruns, T., Lee, S. and Taylor, J. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR Protocols: a guide for methods and applications eds. Innis, M.A., Gelfand, D.H., Sninsky, J.J. and White, T.J. pp. 315-352. San Diego: Academic Press.

Wilson, C.L. and Chalutz, E. (1989) Postharvest biological control of Pénicillium rots of citrus with antagonistic yeasts and bacteria. Scientia horticulturae 40, 105-112.

Wilson, I. (1997) Inhibition and facilitation of nucleic acid amplification. Appl. Environ. Microbiol 63,3741-3751.

Winkler, A.J. (1973) General Viticulture. London: University of California Press, Ltd.

Wintzingerode, F.V., Gobel, U.B. and Stackebrandt, E. (1997) Determination of microbial diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiol Rev. 21, 213-229.

Woody, S.T., Spear, R.N., Nordheim, E.V., Ives, A.R. and Andrews, J.H. (2003) Single- leaf resolution of the temporal population dynamics of Aureobasidium pullulans on apple leaves. Appl Environ. Microbiol 69, 4892-4900.

Wyne, E.S. and Gott, C.L. (1956) A proposed revision of the genus Pullularia. J. Gen Microbiol 14, 512-519.

Yanagida, F., Ichinose, F., Shinohara, T. and Goto, S. (1992) Distribution of wild yeasts in the white grape varieties at Central Japan. J. Gen. Appl Microbiol 38, 501-504. Yang, L., Tran, D.K. and Wang, X. (2001) BADGE, beads array for the detection of gene expression, a high-throughput diagnostic bioassay. Methods 11, 1888-1898.

Yarrow, D. (1998) Methods for the isolation, maintenance and identification of yeasts. In The Yeasts, a Taxonomic Study, Fourth Edition eds. Kurtzman, C.P. and Fell, J.W. pp 77-100. Amsterdam: Elsevier.

Yurlova, N.A., Hoog, G.S. de, and Gerrits van den Ende, A.H.G. (1999) Taxonomy of Aweobasidiim and allied genera. Stud. MycoL 43, 53-69.

Yurlova, N.A., and Hoog, G.S. de, (1997) A new variety oíAureobasidium puUulans characterized by exopolysaccharide structure, nutritional physiology and molecular ÍQdAuxQS. Antonie Van Leeuwenhoekll, 141-147.

Yurlova, N.A., Uijthof, J.M.J., and Hoog, G.S. de, (1996) Distinction of species in Aureobasidium and related genera by PCR-ribotyping. Antonie van Leeuwenhoek, 69, 323-329.

Yurlova, N.A., Mokrousov, I.V. and Hoog, G.S. de, (1995) Intraspecific variability and exopolysaccharide production m Aureobasidiumpullulans. Antonie van Leeicwenhoek^ 68, 57-63.

Zoecklein, B.W., Fugelsang, K.C., Gump, B.H. and Nury, F.S. (1995) Wine analysis and production. New York: Chapman and Hall.

Xufre, A., Albergaría, H., Inácio, J., Spencer-Martins, I. Girio, F. (2006) Application of fluorescence in situ hybridization (FISH) to the analysis of yeast population dynamics in winery and laboratory grape must fermentations. Int. J. Food Microbiol. 108, 376-384. APPENDIX 1

1.1 Partial 26S rDNA sequence identification of yeast isolates from wine grapes by plate culture and enrichment culture

Species Accession Sequence similarity Number (%) Aureobasidium pullulans AY185811 90-100 Discosphaehna fagi AY016359 90-100 Aureobasidium sp. AY167611 94 Aureobasidium sp. AF459656 96-99 Candida cf. etschellsii AF313354 98 Candida cf. glabrata AF313362 98 Candida parapsilosis AF485969 98 Candida zemplinina AY 16076 98-100 Cintractia cf. limitata AJ236147 96 Cryptococcus albidus AF335982 97 Cryptococcus chemovii AF181530 100 Cryptococcus dimennae AF075489 98 Cryptococcus flavus AF44338 100 Cryptococcus heveanensis AF406890 97-99 Cryptococcus laurentii AF459661 98-100 Cryptococcus magnus AY242120 97-100 Cryptococcus oeirensis AF181537 100 Cryptococcus phenolic us AF459669 97 Cryptococcus saitoi AF181540 98-100 Cryptococcus victoriae AF406901 99 Cryptococcus sp. AF444707 98 Cryptococcus sp. AF444699 96 Debaryomyces hansenii AJ50850 99 Filobasidium globisporum AF406932 96 Graphiola cylindrica AF487400 95 Hanseniaspora guilliermondii U84230 99 Hanseniaspora opimtiae AY267820 99-100 Hanseniaspora uvarum U84229 99-100 Hanseniaspora valbyensis U73596 100 Issatchenkia orientalis AF335979 99 Kluyveromyces thermotolerans U69581 100 Metschnikowia pulcherrima U45736 98-100 Metschnikowia sp. AFO17403 97-100 Metschnikowia sp. AFO17401 97-100 Pichia acaciae U45767 99 Pichia anomala AF330115 98 Pichia guilliermondii AY233758 99 Pseudozyma antartica AB089377 96 Pseudozyma prolifica AB089369 96 APPENDIX 1 continued

Species Accession Sequence similarity Number (%) Rhodotorula glutinis AF335985 97-100 Rhodotorula laryngis AF189942 99 Rhodotorula minuta AB078501 99 Rhodotorula mucilaginosa AF335987 98-100 Rhodotorula nothofagi AF444736 99-100 Rhodotorula pallida AF 189962 100 Rhodotorula slooffiae AF 189968 100 Rhodosporidium babjevae AF189913 97-99 Rhodosporidium sphaerocarpum AF189919 100 Sporobolomyces roseus AY070003 98 Sporobolomyces ruberrimus AF406930 99 Torulaspora delbrueckii U72156 99 Tremella fuciformis AF042234 96 Sac char omyces cerevisiae AF533067 99 Ustilago sp. AY233734 99 Zygoascus hellenicus U40125 98 Zygosaccharomyces bailii U72161 97

1.2 ITS1-5.8S-ITS2 sequence identification of yeast isolates from wine grapes by plate culture and enrichment culture

Species Accession Sequence similarity Number (%) Cryptococcus magnus AF444341 98 Hanseniaspora. guilliet mondii AJ512433 99 Hanseniaspora opimtiae AJ512440 100 Hanseniaspora uvarum AJ512432 99-100 Hanseniaspora valbyensis AJ512434 100 Rhodotorula glutinis AYl 88373 98 Rhodosporidium babjevae AF444636 98 APPENDIX 2 Sequence identification of partial 26S rDNA DGGE bands obtained from grape rinses

Species Accession Sequence similarity Number (%) Aureobasidium pullulons AY1858I1 95-100 Discosphaerina fagi AYO16359 95-100 Aureobasidium sp. AY167611 94 Candida galli AY346454 100 Candida zemplinina AY 16076 98-100 Cryptococcus laurentii AF49663 97 Cryptococcus aethanolamini AY315663 97 Hanseniaspora guilliermondii U84230 99 Hanseniaspora opuntiae AY267820 99 Hanseniaspora uvarum U84229 99 Hortaea werneckii AB079595 97 Metschnikowia pulcherrima U45736 95-99 Metschnikowia sp. AFO17403 96-100 Metschnikowia sp. AFO17401 96-100 Rhodotorula sp. AFI 89932 96 Uncultured Saccharomycete AJ550988 95 Alternaria sp. AY154791 94-100 Aspergillus carbonarius AF459734 98-100 Aspergillus niger AFI 09344 98-100 Phialocephala scopiformis AF326085 97-99 Phoma glomerata AY293796 95-100 Phoma herbarium AY293791 95-100 Raciborskiomyces longisetosum AYO 16367 96-99 Uncultured fungal clone AY4649371 96 APPENDIX 3

3.1 Schedule of pesticide application for vineyards from which grapes were analysed for yeasts (2001-2002 season)

Vineyard Grape cultivar Date of Pesticide applications^ sampling Cab, Sem 16/10/01 Bravo 720, Bayfidan 250EC, Delfín WG R Cab, Sem 15/11/01 Oxydule, Thiovit DF, Delfm WG Cab, Sem 03/01/02 Kocide Blude DF, Delfín WG, Fortress 500 Cab, Sem 30/01/02 Dithane DF, Rovral Liquid, Delfín WG Cab, Sem 10/02/02 No applications Appplied approximately 1-14 days prior to sampling Cab, Cabernet sauvignon; Sem, Semillon Data supplied by viticulturalist of vineyard sampled

3.2 Schedule of pesticide application for vineyards from which grapes were analysed for yeasts (2002-2003 season)

Vineyard Grape cultivar Date of Pesticide applications' sampling Cab, Sem 15/10/02 Bravo 720, Bayfidan 250EC, Delfín WG R Cab, Sem 27/11/02 Oxydule, Thiovit DF, Delfín WG Cab, Sem 03/01/03 Kocide Blude DF, Delfín WG. Fortress 500 Cab, Sem 24/01/03 Dithane DF, Rovral Liquid, Delfín WG Cab, Sem 31/01/03 No applications Sem 15/10/02 Oxydule, Thiovit Jet MG, Dipel Forte Sem 27/11/02 Oxydule, Thiovit Jet MG, Dipel Forte L Sem 03/01/03 Oxydule, Dipel Forte Sem 24/01/03 Kocide Blue DF< Rovral Liquid Sem 31/01/03 No applications Cab, Mer, Sab 25/10/02 Scala, Topass, Bravo, Polyyram, Dipel Forte Cab7Mer,Sab 05/12/02 Spin, Dipel Forte Cab, Mer, Sab 09/01/03 No applications HG Cab, Mer, Sab 02/02/03 No applications Sab 06/02/03 No applications Cab, Mer 20/02/03 No applications Cab, Tyr, Sab, Sem 29/10/02 Sulfur, Captan, BT Cab, Tyr, Sab, Sem 28/11/02 Flint, Spindo, BT Cab, Tyr, Sab, Sem 06/01/03 Captan, Sulfur MW Cab, Tyr, Sab, Sem 30/10/03 No applications Sab 04/02/03 No applications Cab, Tyr, Sem 19/02/03 No applications ^ Appplied approximately 1-14 days prior to sampling Cab, Cabernet sauvignon; Mer, Merlot; Tyr, Tyrian; Sab, Sauvignon blanc; Sem, Semillon Data supplied by viticulturalist from various vineyards APPENDIX 4

Commercial name, active ingredient and application purpose for various pesticides used during wine grape cultivation

Commercial name Active ingredient Pest control Thiovit Jet Sulfur Mites and powdery mildew Delan 700 WG Dithianol Downy mildew and black spot Kocide Blue Copper hydroxide Frost Topas Penconazole Downy mildew and rusts Polyram DF Metiram Downy mildew and black spot Bravo Chlorothalonil Botrytis and downy mildew Scala Pyrimethanil Botrytis Spin-Flo Benzothiadiazole Botrytis Oxydul Copper oxychloride Downy mildew Rovral Liquid Iprodione Botrytis Spin Phenmedipham Weeds (kochia, lambsquaters, mustard. green foxtail) Fortress 500 Chlorethoxyphos Botrytis Dithane DF Mancozeb Downy mildew Rovral Iprodione Botrytis Proclaim Emamectin benzoate Moths Flint Trifloxystrobin Powdery mildew and black rot Dipel Forte Bacillus thuringiensis subsp. Insects (leaf-eating caterpillars, gypsy moth. kurstaki EG2371 spruce budworm, jack pine bud worm. Success Bacillus thuringiensis hemlock looper, tussock moth, tent BT Bacillus thuringiensis caterpillar, pine processionary moth and Delfm WG Bacillus thuringiensis others) Data in this table was collected from the Agrochemicals navigation web page of the Australian Wine Research Institute, accessed on 15 August 2005 http://www.awrixom.au/agrochemicals/right_chemical/recommendations.asp ALLBOOK BINDERY

91 RYEDALE ROAD WESTRYDE 2114

PHONE: 9807 6026