"Land and Food Systems, Faculty of"@en . "DSpace"@en . "UBCV"@en . "Anderson, Michael James"@en . "2011-10-11T14:57:01Z"@en . "2011"@en . "Master of Science - MSc"@en . "University of British Columbia"@en . "Wine fermentation presents a unique environment in which strains of the budding yeast Saccharomyces cerevisiae have evolved with superior tolerance to a multitude of stressors. Ethanol toxicity has one of the greatest impacts in reducing cellular viability and metabolic function and thus poses a threat in causing slow, stuck and incomplete fermentations. In pursuit of optimizing commercial strains of S. cerevisiae, identification of genes involved in ethanol tolerance has been of recent interest. Genomic resources such as the S288C deletion collection and microarray analysis have been widely utilized and have provided a foundation, albeit incomplete, for understanding ethanol toxicity in yeast. As a new approach, the recently developed molecular barcoded yeast open reading frame (MoBY-ORF) library, in which all S. cerevisiae genes along with native promoter and terminator sequences have been cloned into barcoded high copy 2\u00CE\u00BC plasmid vectors, has been utilized. Both the S288C laboratory strain and the M2 wine strain of S. cerevisiae were transformed with the MoBY ORF library and genes were identified by quantitation of molecular barcodes after 48 hours in 12% ethanol stressed library pools. Five genes were highly ranked in both S288C and M2 screens, two of which, RCN1 and RSA3, improved tolerance to high (16-21% v/v) ethanol toxicity over 1-3 hour incubation periods in both strain backgrounds. RCN1 is a regulator of the stress signalling protein calcineurin whereas RSA3 has a role in ribosome maturation. Additional fitness advantages conferred upon overproduction of RCN1 and RSA3 include increased resistance to cell wall degradation, heat, osmotic and oxidative stress. Neither RCN1 nor RSA3 over-expression in M2 during model fermentations of synthetic wine medium significantly increased the fermentation rate or final ethanol yield. Regulation of calcineurin and ribosomal assembly processes during ethanol stress, however, may still be key targets for improving tolerance to the stressful conditions of wine fermentation."@en . "https://circle.library.ubc.ca/rest/handle/2429/37887?expand=metadata"@en . "Identification of RCN1 and RSA3 as Ethanol Tolerant Genes in the Saccharomyces cerevisiae S288C lab strain and the M2 Wine Strain by MICHAEL JAMES ANDERSON B.Sc. Biology (2008), University of British Columbia. A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (FOOD SCIENCE) THE UNIVERSITY OF BRITISH COLUMBIA (VANCOUVER) OCTOBER, 2011 \u00C2\u00A9 Michael James Anderson, 2011 ii Abstract Wine fermentation presents a unique environment in which strains of the budding yeast Saccharomyces cerevisiae have evolved with superior tolerance to a multitude of stressors. Ethanol toxicity has one of the greatest impacts in reducing cellular viability and metabolic function and thus poses a threat in causing slow, stuck and incomplete fermentations. In pursuit of optimizing commercial strains of S. cerevisiae, identification of genes involved in ethanol tolerance has been of recent interest. Genomic resources such as the S288C deletion collection and microarray analysis have been widely utilized and have provided a foundation, albeit incomplete, for understanding ethanol toxicity in yeast. As a new approach, the recently developed molecular barcoded yeast open reading frame (MoBY-ORF) library, in which all S. cerevisiae genes along with native promoter and terminator sequences have been cloned into barcoded high copy 2\u00CE\u00BC plasmid vectors, has been utilized. Both the S288C laboratory strain and the M2 wine strain of S. cerevisiae were transformed with the MoBY ORF library and genes were identified by quantitation of molecular barcodes after 48 hours in 12% ethanol stressed library pools. Five genes were highly ranked in both S288C and M2 screens, two of which, RCN1 and RSA3, improved tolerance to high (16-21% v/v) ethanol toxicity over 1-3 hour incubation periods in both strain backgrounds. RCN1 is a regulator of the stress signalling protein calcineurin whereas RSA3 has a role in ribosome maturation. Additional fitness advantages conferred upon overproduction of RCN1 and RSA3 include increased resistance to cell wall degradation, heat, osmotic and oxidative stress. Neither RCN1 nor RSA3 over- expression in M2 during model fermentations of synthetic wine medium significantly increased the fermentation rate or final ethanol yield. Regulation of calcineurin and ribosomal assembly iii processes during ethanol stress, however, may still be key targets for improving tolerance to the stressful conditions of wine fermentation. iv Preface This work was performed as part of a collaborative effort with Dr. Charlie Boone and Dr. Sarah Barker at the University of Toronto. The MoBY-ORF 2.0 library was graciously provided by Dr. Charlie Boone. Once the library was received I further amplified it and froze aliquots for future use at -80\u00C2\u00B0C (outlined in Materials and Methods 2.2.1). During my genome-wide screen, following ethanol treatment for 48 hours, I extracted 2\u00CE\u00BC MoBY-ORF plasmids from S. cerevisiae and plasmids were subsequently shipped on ice to Dr. Barker for analysis. Barcode analysis, including PCR amplification, microarray hybridization and determination of cut-off values, (described in Materials and Methods section 2.2.4) was conducted by Dr. Sarah Barker at the University of Toronto Terrence Donnelly Centre for Cellular and Biochemical Research. A version of this research has been submitted for publication as: Anderson, M.J., Barker, S., Boone, C. and Measday, V. (2011) Identification of RCN1 and RSA3 as ethanol tolerant genes in Saccharomyces cerevisiae using a high copy barcoded library. FEMS Yeast Research. I wrote the manuscript with substantial input and editing by Dr. Vivien Measday. v Table of Contents Abstract............................................................................................................................................ii Preface............................................................................................................................................iv Table of Contents.............................................................................................................................v List of Tables ...............................................................................................................................viii List of Figures.................................................................................................................................ix Acknowledgements ........................................................................................................................x Dedication......................................................................................................................................xii 1 Introduction.................................................................................................................................. 1 1.1 A Brief History of Saccharomyces cerevisiae.............................................................. 1 1.2 The Biology of Saccharomyces cerevisiae................................................................... 2 1.2.1 Saccharomyces cerevisiae as a Lab Yeast..................................................... 3 1.2.2 Saccharomyces cerevisiae as a Wine Yeast................................................... 4 1.3 The Stressful Fermentation Environment..................................................................... 7 1.4 Improvement of Wine Yeast Stress Tolerance............................................................. 8 1.4.1 Genetic Influences on the Stress Tolerance of Wine Yeast........................... 9 1.4.2 Genetic Engineering and Targets for Wine Yeast Improvement..................11 1.5 Ethanol Stress and the Response of Saccharomyces cerevisiae..................................13 1.5.1 Microarray Identification of Genes Induced During Ethanol stress.............15 1.5.2 Deletion Collection; Identification of Ethanol Tolerance Genes..................16 1.6 Research Proposal........................................................................................................18 1.6.1The MoBY-ORF Library...............................................................................18 1.6.2 Research Outline...........................................................................................19 2 Materials and Methods................................................................................................................21 2.1 Yeast Strains and Media Conditions............................................................................21 2.1.1 Yeast Strains.................................................................................................21 2.1.2 Genetic Manipulation of the M2 Wine Yeast...............................................21 2.1.3 Yeast Propagation Media..............................................................................22 2.1.4 Escherichia coli Propagation Media.............................................................23 2.1.5 Microtiter Plate Growth Conditions............................................................. 24 2.2 MoBY-ORF Library....................................................................................................24 vi 2.2.1 MoBY-ORF 2.0 Library Transformations....................................................24 2.2.2 Individual MoBY-ORF Plasmid Transformations........................................25 2.2.3 Empty Vector Ethanol Tolerance Assay.......................................................26 2.2.4 MoBY-ORF Genome-wide Ethanol Tolerance Screen................................26 2.3 FunSpec Genome-wide Screen Analysis.....................................................................27 2.4 Ethanol Tolerance Confirmation Assays.....................................................................27 2.4.1 Growth Analysis in 12% v/v Ethanol...........................................................28 2.4.2 Growth Analysis Following 16% v/v Ethanol Shock...................................28 2.4.3 Viability Assay Following 17-21% v/v Ethanol Shock................................29 2.4.4 Deletion Collection Ethanol Sensitivity Spot Assays...................................29 2.5 Zymolyase Cell Wall Digestion Assay........................................................................30 2.6 Environmental Stress Assays.......................................................................................30 2.6.1 Environmental Stress Growth Curve Analysis.............................................30 2.6.2 Cycloheximide Translation Inhibition Assays..............................................31 2.7 Growth Curve Statistical Analysis...............................................................................31 2.8 Model Wine Fermentations..........................................................................................32 2.8.1 Synthetic Grape Juice Media and Growth Conditions..................................32 2.8.2 Fermentation Analysis by High Pressure Liquid Chromatography..............33 3 Results.........................................................................................................................................34 3.1 Genome-wide Ethanol Tolerance Screen in S288C and M2.......................................34 3.1.1 Empty Vector Controls Have Reduced Viability in 12% v/v Ethanol..........34 3.1.2 Twenty Ethanol Tolerant Candidate Genes Identified in S288C and M2....35 3.1.3 Splicing Genes are Enriched in Both S288C and M2...................................37 3.2 Confirmation of Ethanol Tolerance ............................................................................39 3.2.1 Over-expression of Five Genes Common to S288C and M2.......................39 3.2.2 RCN1 and RSA3 Improve Viability Following 16% v/v Ethanol ................40 3.2.3 RCN1 and RSA3 Improve Viability After Extreme Ethanol Shock..............45 3.2.4 Deletion of Candidate Genes Causes Ethanol Sensitivity............................46 3.3 RCN1 and RSA3 Expression Influences Cell Wall Assembly.....................................48 3.3.1 RCN1 and RSA3 decrease Zymolyase Cell Wall Digestion..........................49 3.4 RCN1 and RSA3 Over-expression Improves Environmental Stress Tolerance...........53 vii 3.4.1 RCN1 and RSA3 Over-expression Improves Thermal Tolerance.................54 3.4.2 RCN1 and RSA3 Over-expression Improves S288C Osmo-tolerance..........56 3.4.3 RSA3 Over-expression Improves Oxidative Stress Tolerance......................58 3.5 Cycloheximide Translation Inhibition is Greater in M2 than S288C..........................60 3.5.1 Over-expression of RSA3 Does Not Affect Translation Inhibition..............62 3.6 Model Wine Fermentations..........................................................................................62 4 Discussion...................................................................................................................................67 4.1 Genome-wide Screen Identifies Five Ethanol Tolerant Genes....................................67 4.2 The Role of Mitochondria in Ethanol Tolerance.........................................................68 4.2.1 Acetyl-CoA Production in the Mitochondria...............................................69 4.2.2 Gut2p, Img1p and Sym1p Also function in the Mitochondria.....................70 4.3 The Role of RCN1 Stress Signalling in Ethanol Tolerance.........................................71 4.4 Transcription, Translation and Ribosome Assembly in Ethanol Tolerance................73 4.4.1 Transcription Termination and Splicing Influence Ethanol Tolerance.........73 4.4.2 RSA3 and Translation in Ethanol Tolerance.................................................74 5 Conclusions.................................................................................................................................78 Literature Cited..............................................................................................................................80 viii List of Tables Table 1. Primers utilized in this study for gene disruption and mating type determination in the M2 strain of S. cerevisiae .....................................................................................23 Table 2. Genes identified that improve ethanol tolerance when over-expressed in Saccharomyces cerevisiae strains S288C or M2............................................................ 36 Table 3. Functional classification of enriched genes from ethanol tolerance screen in S288C.....39 Table 4. Functional classification of enriched genes from ethanol tolerance screen in M2..........40 Table 5. Growth curve analysis of S288C strains carrying 2\u00C2\u00B5 over-expression plasmids after 16% v/v ethanol shock............................................................................................41 Table 6. Growth curve analysis of M2 strains carrying 2\u00C2\u00B5 over-expression plasmids After 16% v/v ethanol shock...........................................................................................42 Table 7. Zymolyase cell wall digestion analysis of S288C strains either over-expressing or carrying null mutations of RCN1 and RSA3....................................................................51 Table 8. Zymolyase cell wall digestion analysis of M2 strains either over-expressing or carrying null mutations of RCN1 and RSA3....................................................................53 Table 9. Comparison of lag time, growth rate and O.D.Max values during heat stress with RCN1 and RSA3 over-expression in S288C...............................................................................54 Table 10. Comparison of lag time, growth rate and O.D.Max values during heat stress with RCN1 and RSA3 over-expression in M2....................................................................................54 Table 11. Comparison of lag time, growth rate and O.D.Max values during osmotic stress with RCN1 and RSA3 over-expression in S288C....................................................................56 Table 12. Comparison of lag time, growth rate and O.D.Max values during osmotic stress with RCN1 and RSA3 over-expression in M2..........................................................................56 Table 13. Comparison of lag time, growth rate and O.D.Max values during oxidative stress with RCN1 and RSA3 over-expression in S288C....................................................................58 Table 14. Comparison of lag time, growth rate and O.D.Max values during oxidative stress with RCN1 and RSA3 over-expression in M2..........................................................................58 Table 15. Final ethanol production from model wine fermentations with M2 and S288C...........63 ix List of Figures Figure 1. The glycolysis pathway in Saccharomyces cerevisiae.....................................................6 Figure 2. Schematic representation of the MoBY-ORF 2.0 plasmid.............................................19 Figure 3. Viability of S288C and M2 empty vector strains in 12% v/v ethanol............................34 Figure 4. Quantified barcode values from MoBY-ORF genome-wide ethanol tolerance screen.................................................................................................36 Figure 5. Growth curves of M2 over-expression strains in 12% v/v ethanol................................40 Figure 6. Growth curve of S288C and M2 over-expression strains following 16% ethanol shock.........................................................................................................44 Figure 7. Phloxine B spot assays of S288C and M2 RCN1 and RSA3 over-expression strains following 16% ethanol shock.............................................................................45 Figure 8. Spot assays of S288C and M2 over-expression strains following 17-21% ethanol shock....................................................................................................46 Figure 9. Deletion set spot assays on 6-10% ethanol.....................................................................47 Figure 10. M2 pho80\u00E2\u0088\u0086 and img1\u00E2\u0088\u0086 deletion spot assays on 6% ethanol........................................48 Figure 11. S288C zymolyase digestion with over-expression and deletion of RCN1 and RSA3..................................................................................50 Figure 12. M2 Zymolyase digestion with over-expression and deletion of RCN1 and RSA3 .................................................................................52 Figure 13. Growth of S288C and M2 with RCN1 and RSA3 over-expression during heat stress..........................................................................................................55 Figure 14. Growth of S288C and M2 with RCN1 and RSA3 over-expression during osmotic stress....................................................................................................57 Figure 15. Growth of S288C and M2 with RCN1 and RSA3 over-expression during oxidative stress..................................................................................................59 Figure 16. Cycloheximide inhibition comparison of S288C and M2............................................61 Figure 17. Model Fermentations of 20% glucose/fructose synthetic grape juice with M2 and S288C over-expressing RCN1 and RSA3...............................................64 Figure 18. Model Fermentations of 32% glucose/fructose synthetic grape juice with M2 over-expressing RCN1 and RSA3...................................................................65 x Acknowledgements There are many people who deserve recognition for their various forms of support in bringing this research to completion. First and foremost, I would like to thank my academic supervisor Dr. Vivien Measday, for taking me on both as a graduate student and also during my undergraduate degree when my excitement for wine research first began. Dr. Measday\u00E2\u0080\u0099s guidance, patience and support through my research, coursework and teaching assistantships have made this work possible. I would also like to extend a deep appreciation to Dr. Chris Loewen and Dr. Hennie van Vuuren for their guidance as members of my supervisory committee. Faculty members Dr. Eunice Li-Chan, Dr. David McArthur, and Dr. Margaret Cliff have also been tremendously helpful in affording me opportunities for academic growth outside of the lab through course work, volunteering, teaching and attending conferences. Secondly, without funding there is no research. I would like to acknowledge the UBC University Graduate Fellowship, the American Wine Society, the American Society for Enology and Viticulture, the Canadian Vintners Association, and Dr. Vivien Measday, whose grants all helped in funding my research. Of my colleagues at the UBC Wine Research Centre I would also like to extend a sincere thank you to Samantha Turner, Calvin Adams, Dr. Chris Walkey, Dr. Zongli Luo, Dr. Lina Ma, and Dr. Lina Madilao for their many forms of assistance, sense of humour and friendship over the past two years. I would also like to acknowledge Nina Piggott, who not only helped me through my undergraduate experience at the Wine Research Center, but also keeps our lab organized and running on a daily basis. I would like to thank my fellow wine lovers, Emily Terrell and Jay Martiniuk for our tastings and discussion of all things wine. I look forward to our career paths crossing again. Last and certainly not least, I would also like to thank Jenny xi McQueen and Krystina Ho with whom I shared the many highs and lows of life as a graduate student. Your advice, technical help and lab favours have made many things possible. Thank you. Finally, away from my studies I have far too many friends and family to thank for their support in my personal life. At the very least, I would like to thank my parents Doug and Diane Anderson for their undying support and for teaching me the value in hard work. I can only hope it shows through. Lastly, my personal sanity is forever indebted to Roo Anderson (Phelps) for her constant reminding that there is life outside of the lab. I am truly thankful you took a chance on a \u00E2\u0080\u0095scientist\u00E2\u0080\u0096, and have been proud to call you my girlfriend, my fiance and now my wife as I have pursued my Master\u00E2\u0080\u0099s Degree. xii Dedication This work is dedicated to the Wine Yeast Saccharomyces cerevisiae. I\u00E2\u0080\u0099m sorry for all of the stress I\u00E2\u0080\u0099ve put you through But I couldn\u00E2\u0080\u0099t have done it without you. 1 1 Introduction 1.1 A Brief History of Saccharomyces cerevisiae Wine, beer, bread, and other fermented foods have played a significant role in both ancient and modern day societies. The earliest records of wine date back more than 6500 years in northern Greece, and to 3000 BC in Egypt (McGovern, 2003). Wine not only played an important role in agriculture, economics, and socialization, but also in ceremonial life where it was often revered as a gift from the gods. Over the millennia, wine making technology has advanced from the early grape presses and wooden barrels to the industrialized processes used today. Despite the fact that wine has been inextricable from the global culture for thousands of years, the spontaneous change from grape to wine (now known as alcoholic fermentation) was largely a mystery until relatively recently. Although multiple microorganisms were likely present in early fermentations, analysis of ribosomal DNA extracted from one of the earliest wine vessels found in Egypt has confirmed the presence of one yeast in particular, Saccharomyces cerevisiae (S. cerevisiae) which has now established itself as the predominant yeast in industrial processes (Cavalieri, et al., 2003). The natural environment of S. cerevisiae is a matter of some debate, largely due to its long history of domestication by humans. Sequence analysis of the genetic diversity and geographical origin of S. cerevisiae strains has suggested that geographically distinct strains are closely related having arisen from a single domestication event and subsequently followed the migration of humans through vineyard plantings (Legras, et al., 2007, Liti, et al., 2009, Schacherer, et al., 2009). One line of evidence suggests that S. cerevisiae diverged from the natural or \u00E2\u0080\u0095wild\u00E2\u0080\u0096 populations of its closest relative, Saccharomyces paradoxus in fruit trees, soils, and vineyards (Naumov, 1996, Sniegowski, et al., 2002). Although many studies have 2 isolated S. cerevisiae from vineyard grapes, whether they originate as native vineyard flora remains largely unknown. Some studies have shown that S. cerevisiae is difficult to isolate from intact fruit and claim that S. cerevisiae is an entirely domesticated organism that only inhabits fermentation equipment found in winery cellars (Rosini, et al., 1982, Martini, 1993). Isolating S. cerevisiae is easier using grapes damaged by birds and insects and it has therefore been suggested that S. cerevisiae isolated from these grapes are a result of back transport from winery environments by insect and bird vectors (Naumov, 1996). Regardless of the origin of S. cerevisiae, the microflora of grapes includes many genera of yeasts such as Kloeckera, Pichia, Zygosaccharomyces, Dekkera, Torulopsis, and Brettanomyces (Querol, et al., 2003). Although these native yeasts may influence the early stages of wine fermentation, S. cerevisiae almost always dominates the latter stages due to its tolerance for high concentrations of ethanol, thus earning a reputation as the \u00E2\u0080\u0095wine yeast\u00E2\u0080\u0096. 1.2 The Biology of S. cerevisiae Saccharomyces. cerevisiae is a single-celled eukaryote, round or ovoid in shape and 5-10 \u00CE\u00BCM in diameter. S. cerevisiae exists predominantly in a diploid state with vegetative reproduction carried out by asymmetric budding. After sporulation (meiosis), the ascospore forms containing four haploid cells with two of each complimentary mating types denoted \u00E2\u0080\u0095mat a\u00E2\u0080\u0096 and \u00E2\u0080\u0095mat \u00CE\u00B1\u00E2\u0080\u0096 (Scott, et al., 2004). Wild strains of S. cerevisiae are heterothallic; they are able to switch between mating types if no complimentary mating type is present, which allows them to revert back to the diploid stage. Homothallic strains, artificially generated by deletion of the HO endonuclease, are not able to switch mating types (Scott, et al., 2004). S. cerevisiae is able to produce energy from both anaerobic fermentation of high sugars (where glucose, fructose, and 3 galactose sugars can be converted to ethanol), and aerobic respiration where ethanol or low concentrations of sugars are used as a carbon source. S. cerevisiae can also use a wide range of compounds, such as urea and ammonia as nitrogen sources but it cannot metabolize nitrate. Although it is able to use amino acids, small peptides, and nitrogen bases as nitrogen sources, the lack of protease secretion in S. cerevisiase prevents it from using extracellular proteins. 1.2.1 Saccharomyces cerevisiae as a Lab Yeast Aside from domestication within a food production environment, S. cerevisiae has also been used since the twentieth century as a model organism for investigating molecular and cellular processes in eukaryotes. Laboratory strains of S. cerevisiae, such as the commonly utilized strain S288C, have been selected for ease of genetic manipulation and consistent growth in controlled nutrient-rich environments (Borneman, et al., 2008). The S288C strain reportedly arose from crosses by scientists in the 1940\u00E2\u0080\u0099s and has now been propagated for over seventy years as a pure laboratory culture representing hundreds of thousands of generations of divergence within this environment (Mortimer & Johnston, 1986). S288C was the first eukaryotic organism to have its entire genome sequenced in 1996 and thus much of the available genetic data is based on this specific strain, making it one of the most commonly used strains for genomic studies (Goffeau, et al., 1996). S. cerevisiae has a relatively small genome of about 6200 genes spread over 16 linear chromosomes in its haploid form. S. cerevisiae chromosomes range in size from 250 to 2000 kilobases, comprising a total of 13 megabases of DNA. Overall, the S. cerevisiae chromosomes contain few introns and relatively few repetitive sequences. However, with recent advances in the throughput capabilities of DNA sequencing technology, many other laboratory and industrial yeast strains have also been sequenced. A comparison of 4 the S288C genome with five other commonly used laboratory S. cerevisiae strains found that there was as little as 0.05% divergence between genome sequences, and a maximum of only 0.36% nucleotide difference between S288C and its most divergent counterpart (Schacherer, et al., 2007). This allows almost universal application of the S288C sequence information to other laboratory strains, and provides a foundation for studying other industrial strains of S. cerevisiae. 1.2.2 Saccharomyces cerevisiae as a Wine Yeast The primary role of Saccharomyces cerevisiae in wine fermentation is to convert grape sugars (glucose and fructose) to ethanol via a cellular process known as glycolysis (Figure 1). In glycolysis, glucose and fructose are converted to pyruvate by a complex series of enzymatic reactions. Pyruvate is decarboxylated to acetaldehyde with the release of carbon dioxide. The acetalydehyde is subsequently reduced to ethanol by alcohol dehydrogenase, and the alcohol is then diffuses from the cell into the external environment (Mohammed, 2007). This last reduction step is important as it restores the co-enzyme NAD+ which is consumed during glycolysis. Glycerol formation from the glycolysis intermediate dihydroxyacetone phosphate can also help to restore NAD+ throughout glycolysis (Mohammed, 2007). Glycerol, being the third-highest fermentation by-product following ethanol and CO2, also contributes an important sensory aspect to wine fermentation as it increases the wine's viscosity and perceived sweetness, positively contributing to the wine's overall \"mouth feel\" (Rankine & Bridson, 1971, Noble & Bursick, 1984). In addition to glycerol, S. cerevisiae generates many other metabolic by-products that can impact wine quality in positive and negative ways. Often referred to as S. cerevesiae's secondary role, the production (or lack) of other compounds such as acetic acid, succinic acid, sulphur 5 compounds, higher alcohols, and esters is often regarded as an important yeast trait in wine making. S. cerevisiae often produces small quantities of acetic and succinic acid. In combination with other spoilage microorganisms, acetic acid from S. cerevisiae can contribute to volatile acidity, an unpleasant vinegar aroma considered to be a fault in quality wines (Ribereau-Gayon, et al., 2000). Hydrogen sulphide (H2S) production from S. cerevisiae also negatively affects wine quality by producing a rotten egg smell. This can occur when nitrogen is limited in grape must (Monk, 1986). S. cerevisiae can also synthesize higher alcohols such as propanol, butanol, isoamyl alcohol, and hexanol, all of which can contribute to the fruitiness and complexity of a wine's aroma and flavour (Lambrechts & Pretorius, 2000). Esters are also an important class of volatile aroma compounds. They can be produced directly from S. cerevisae metabolic pathways or by S. cerevisiae mediated enzymatic liberation of grape compounds (Swiegers, et al., 2005). Two classes of esters are produced in wine from different S. cerevisiae-derived precursors: acetate esters are produced from higher alcohol precursors and ethyl esters from fatty acid precursors. Similar to higher alcohols, esters can positively contribute to a wine\u00E2\u0080\u0099s aromatic complexity, often being described as fruity, floral, or honey-like (Lambrechts & Pretorius, 2000) 6 Figure 1. The glycolysis pathway in S. cerevisiae (Mohammed, 2007). . 7 1.3 The Stressful Wine Fermentation Environment Over centuries of wine making, wine strains of S. cerevisiae have become adept at performing alcoholic fermentation in unfavourable extracellular environments. During fermentation, wine yeasts are routinely exposed to osmotic and organic acid stress upon inoculation into grape must (Bauer & Pretorius, 2000). Advances in wine making and agriculture have seen the introduction of antimicrobial agents (such as sulphur dioxide and residual agrochemicals) in grape must, demanding that S. cerevisiae wine strains have an increased resistance to these compounds (Zuzuarregui & del Olmo, 2004). Moreover, in the end stages of fermentation, nutrients such as glucose, nitrogen, and amino acids are often depleted and the concentration of ethanol increases to toxic levels (Bauer & Pretorius, 2000, Zuzuarregui & del Olmo, 2004, Marks, et al., 2008). Over one hundred strains of S. cerevisiae are currently available to wine makers and are mass-produced as active dried strains. This industrial process further exposes S. cerevisiae to osmotic and oxidative stress, nutrient limitation, and desiccation (Zuzuarregui & del Olmo, 2004). The stress that S. cerevisiae encounters during fermentation can be extreme, depending on the type and style of wine. For example, dessert-style wines (such as ice wine produced from frozen grapes) may contain 40-50% glucose and fructose, causing extreme levels of osmotic stress. In order to preserve aromatic compounds in white wine, fermentation is often conducted in stainless steel tanks at temperatures as low as 12-20\u00C2\u00B0C, well below the optimal range for S. cerevisiae. Sparkling wines such as Champagne (produced in the m\u00C3\u00A9thode traditionnelle) require a secondary bottle fermentation in which S. cerevisiae is inoculated into a base wine containing 12% ethanol, limited nitrogen, and minimal glucose for the production of CO2. Lastly, some wine regions have reported an increase in the average daily temperature, which has resulted in an 8 extended growing season for viticulture (Duchene & Schneider, 2005). As a result, in warm climates such as California, there is increasing concern regarding higher sugar levels at harvest. Higher sugar levels yield ethanol concentrations of more than 16% in finished wines, which is a formidable level for S. cerevisiae to withstand and can affect the sensory quality of the wine. When yeast strains fail to perform in challenging fermentation environments, stuck and sluggish fermentations can arise. Stuck and sluggish fermentations terminate the fermentation with higher levels of residual sugar than desired\u00E2\u0080\u0094generally greater than 0.5% glucose or fructose. High levels of residual sugar cause a greater chance of microbial spoilage, therefore addition of higher levels of sulphites may be required in the finished wine (Ribereau-Gayon, et al., 2000). Stuck and sluggish fermentations can arise when a wine yeast fails to establish numerical (cell viability) and metabolic (cell vitality) dominance early in a fermentation. This can lead to reduced wine quality through the production of offensive odours (from spoilage microorganisms) and increased wine oxidation from the loss of the protective CO2 layer in fermentation tanks (Bisson, 1999). Factors that can lead to poor fermentation performance include nutrient limitations, organic acid stress, microbial competition, temperature extremes and ethanol toxicity. Thus, the incidence of stuck fermentations correlates negatively with a yeast strain's fitness vigour and stress tolerance (Bisson, 1999, Ivorra, et al., 1999). In order to prevent stuck and sluggish fermentations, wine yeast strains are consistently selected for their ability to perform well in a stressful fermentation environment (Attfield, 1997, Bauer & Pretorius, 2000). 1.4 Improvement of Wine Yeast Stress Tolerance Wine makers' demands for yeast strains that can survive challenging enological conditions\u00E2\u0080\u0094without the loss of other desirable fermentation traits\u00E2\u0080\u0094have carved out a niche area of S. cerevisae research with a focus on improving stress tolerance of wine strains. Historically, 9 wine yeast selection has relied on relatively simple methods such as strain variant selection, mutagenesis, and hybridization or breeding. In strain variant selection, naturally occurring genetic variants from a population of wine yeasts are selected based on random mutations that provide superior fermentative properties (Pretorius & Bauer, 2002). Mutagenesis relies on the application of a mutagen to increase the rate of random mutation in a wine yeast population and requires subsequent screening for improved fermentation (Pretorius & Bauer, 2002). Lastly, strain breeding or hybridization involves either sexual or forced mating of two strains with promising fermentative properties to blend their genetic material and hopefully combine the desirable traits into one strain. This mechanism is widely used when the underlying molecular nature of the desired traits is poorly understood (Pretorius & Bauer, 2002). Although strain variant selection, mutagenesis, and hybridization have led to improved wine yeast strains, the lack of precision associated with these methods often leads to the loss of other desirable properties in an uncontrolled manner. 1.4.1 Genetic Influences on the Stress Tolerance of Wine Yeasts By studying natural variations within S. cerevisiae we have gained insight into the genetic loci as well as the biological and chemical processes that give rise to desirable fermentative properties (Dunn, et al., 2005, Borneman, et al., 2008). Early work found that most industrial yeasts are diploid, and the presence of chromosome length variation, aneuoploidy, or even polyploidy is not uncommon (Barre, et al., 1993, Codon, et al., 1998). One commonly accepted theory explaining the presence of extra genetic material in industrial yeasts is that the extra genetic material increases the dosage of genes necessary for survival in stressful environments such as fermentation (Bakalinsky & Snow, 1990). In fact, several examples exist 10 of increased gene dosage or gene expression corresponding to specific industrial environments. Duplication of alcohol deyhdrogenase (ADH) genes, hexose transporters, and other glycolysis genes improve S. cerevisiae's ability to survive the rapid production and consumption of ethanol as a strategy to out-compete microorganisms (Thomson, et al., 2005, Gordon, et al., 2009). Moreover, flor yeasts\u00E2\u0080\u0094employed in wine making for their ability to oxidize ethanol to pyruvate\u00E2\u0080\u0094have multiple copies of ADH2 and ADH3 genes necessary for this function (Guijo, et al., 1997). Industrial baker\u00E2\u0080\u0099s and distiller\u00E2\u0080\u0099s yeasts were also found to have widespread copies of SUC family genes, encoding an invertase necessary for sucrose utilization which confers a faster growth rate on molasses compared to laboratory strains (Codon, et al., 1998). Lastly, a chromosomal rearrangement that led to increased expression of the sulphite tolerance gene SSU1 was found to naturally occur in the sulphite-resistant wine yeast T73. This suggests that alterations in gene expression can have a similar effect with increasing gene dosage (Perez-Ortin, et al., 2002). In contrast to these findings, other studies suggest a lesser prevalence of aneuploidy among S. cerevisiae wine strains than previously thought. Genetic analysis of forty-five commercial yeast strains found that only five were actually aneuploid (Bradbury, et al., 2006). Similarly, karyotyping of four commercial S. cerevisiae wine strains found no major aneuoploidy or chromosomal rearrangements, although variations in gene copy number were noted when compared to S288C. Interestingly, although these variations were moderate at best, some correlations between gene copy and drug sensitivity were observed even among intra strain isolates (Dunn, et al., 2005). Moreover, sequencing of one of the most commonly used commercial wine yeasts, EC1118, found that 99% of its predicted genes are common to S288C, with three distinct chromosomal regions that are of non-Saccharomyces origin. Located within 11 these distinct regions are thirty-four genes, many of which have key roles in wine making such as sugar utilization as well as nitrogen-related functions (Novo, et al., 2009). Lastly, proteome comparison of the wine yeast strain AWRI1631 and S288C found that despite nearly 70,000 single nucleotide changes, their predicted proteomes contained over 99% amino acid identity (Borneman, et al., 2008). Taken all together, two conclusions can be drawn from these findings. First, it appears that small changes in genetic material can account for differences in fermentative properties between industrial and laboratory yeasts. Second, due to the high sequence and proteome similarity between commercial wine yeasts and S288C, we can utilize resources developed for the latter to investigate gene function for industrial applications such as wine fermentation. 1.4.2 Genetic Engineering and Targets for Wine Yeast Improvement The most efficient method of introducing a desirable fermentation property to a wine yeast strain without incurring unwanted changes to the remaining yeast genome is by genetic engineering and recombinant DNA methods (Barre, et al., 1993). Despite the current limitations on the use of genetically modified organisms (GMOs) in the wine industry, the vast genomic resources and established genetic transformation methods available for S. cerevisiae make genetic engineering a promising and viable method for wine yeast strain development (Pretorius & Bauer, 2002). Targets for genetic engineering have already been identified for S. cerevisiae wine strains, and some have already been put into practice (Schuller & Casal, 2005). General targets for yeast strain improvement can be divided into those which affect technical attributes of fermentation (such as processing efficiency and fermentation performance) and those that improve the wine qualitatively by changing the sensory qualities or chemical 12 composition of finished wine (Schuller & Casal, 2005, Fleet, 2008). For example, two time- consuming and sometimes difficult processes following wine fermentation are protein and polysaccharide clarification and malic acid reduction by a secondary malo-lactic fermentation with the lactic acid bacterium Oneococcus oeni. Engineering of the commercial strain EC1118 with partial or full deletion of any one of three genes (KNR4, GPI7, or FKS1) showed increased mannoprotein secretion into wine during fermentation and aging and required 20-40% less bentonite for protein haze clarification (Gonzalez-Ramos, et al., 2009). Furthermore, a wine yeast strain (ML01) capable of performing malo-lactic fermentation has been engineered and is currently being utilized in North America. ML01 was created by the introduction of two non- Saccharomyces genes, allowing malate uptake and subsequent decarboxylation to lactate, and has been approved for use in the wine industry by the US FDA, Health Canada and Environment Canada (Volschenk, et al., 1997, Husnik, et al., 2007). Other genetic alterations have been proposed that aim to improve the sensory attributes of a finished wine. For example, increasing the expression of glycosidases, glucanases, and esterases has been suggested as possible mechanisms for increasing enzymatic production of grape terpenoids and volatile esters to improve the aromatic quality of wine (Pretorius & Bauer, 2002). An increase in glycerol metabolism has been suggested to reduce ethanol production by diverting carbon flux away from ethanol production during glycolysis, with the added benefits of increasing the viscosity and \u00E2\u0080\u0095mouth feel\u00E2\u0080\u0096 in finished wines (Remize, et al., 1999). While increased glycerol production and reduction of ethanol through GPD1 over-expression was successful, this resulted in an increase of acetate production (Remize, et al., 1999). GPD1 over- expression in combination with ALD6 gene deletion was subsequently tested and found to reduce acetate production, but this coincided with loss of other desirable aroma compounds (Eglinton, et 13 al., 2002). Lastly, a recent study isolated a naturally occurring allele of the catalytic sub-unit of sulphite reductase, MET10, with a mutation that significantly reduces H2S formation. H2S formation was also significantly reduced during fermentation by cloning the MET10 allele and transforming it into other high H2S producing strains of S. cerevisiae (Linderholm, et al., 2010). One final target area of genetic modification is to improve fermentation performance by increasing tolerance to fermentation-related stresses. For commercial production of dry active strains, the focus has been on generating yeast strains with high levels of sterols and storage carbohydrates (such as trehalose and glycogen) which stabilize the membrane wall and provide a quick energy source during rehydration and activation (Schuller & Casal, 2005). Nitrogen is often a limiting factor during fermentation as yeast cannot utilize proline, the most abundant amino acid in grape must under anaerobic conditions (Pretorius, 2003). During nitrogen limitation, internalization and degradation of hexose transporters occurs, slowing the rate of fermentation. Approaches thus far have tried to increase utilization of grape sugar in nitrogen- limited musts by over-expressing the key glycolitic enzymes HXT1-18, increasing hexose uptake efficiency, and removing the repression on nitrogen catabolism (Pretorius, 2003). Perhaps one of the most complex stress responses elicited by S. cerevisiae is in response to ethanol toxicity. Due to the pleiotropic effects of ethanol on cellular function, the tolerance mechanisms are not fully understood, despite being widely studied. 1.5 Ethanol Stress and the Response of Saccharomyces cerevisiae One of the major targets for improving the stress tolerance of industrial strains of S. cerevisiae (not only for wine fermentation but also for bio-ethanol production) is to improve its ethanol tolerance. Ethanol toxicity in S. cerevisiae largely targets the plasma membrane, disrupting its normal structure and increasing its fluidity, causing an influx of protons. Proton 14 influx subsequently disrupts the electrochemical gradient across the plasma membrane, resulting in intracellular acidification and, ultimately, a decrease in growth rate and cell viability (Casey & Indgledew, 1986, Piper, 1995). To resolve this change in membrane fluidity, S. cerevisiae increases production of ergosterol and increases the ratio of unsaturated fatty acids (UFAs): saturated fatty acids causing membrane stabilization (Swan & Watson, 1998, You, et al., 2003). A comparison of membrane composition between ethanol-adapted and non-adapted S. cerevisiae found that while oleic acid (C18:1) levels were comparable, palmitic acid (C16:0) levels decreased in the non-adapted cells, thus increasing the ratio of UFAs in ethanol-adapted cells. Sterol accumulation genes SUT1 and SUT2 have been proposed as engineering targets to decrease membrane fluidity during ethanol stress, possibly in combination with over-expression of proton pumps PMA1 and PMA2 to restore ion homeostasis within the cell (Pretorius, 2003, Schuller & Casal, 2005). It has long been known that ethanol stress inhibits essential amino acid transport and glycolysis pathways, restricting the synthesis of energy and nutrient utilization (Leao & van Uden, 1982, Leao & Van Uden, 1984). In addition to membrane disruption, ethanol also increases protein denaturation and thus inhibits molecular function (Casey & Indgledew, 1986). Not surprisingly, a correlation between increased trehalose synthesis, expression of chaperone proteins that stabilize misfolded proteins and ethanol tolerance has been noted. Therefore, over- expression of trehalose synthesis genes TPS1 and TPS2 has been suggested for improving ethanol tolerance (Swan & Watson, 1998, Schuller & Casal, 2005). Ethanol has been further found to generate reactive oxygen species (ROS), causing oxidative damage to lipids, proteins, and DNA, and inhibiting mitochondrial function (Costa, et al., 1997, Du & Takagi, 2007). Over- expression of MPR1 decreased intracellular ROS, which increased ethanol tolerance in S. 15 cerevisiae (Du & Takagi, 2007). The mitochondrial superoxide dismutase (MnSOD) has played a key role in the development of ethanol tolerance in exponentially-growing cells as mutants lacking MnSOD showed higher levels of ethanol toxicity (Costa, et al., 1997). 1.5.1 Microarray Identification of Genes Induced During Ethanol Stress Two studies have used a global transcription approach using microarray technology to identify genes whose regulation changes in response to short term (30-60 minute) ethanol stress. Alexandre, et al., (2001) identified 395 ORFs that were either up-regulated or down-regulated in response to 7% ethanol exposure for thirty minutes. Chandler, et al., (2004) found 374 ORFs whose regulation changed after sixty minutes of 5% ethanol stress. Both studies found that heat shock proteins (HSPs) were highly up-regulated in response to ethanol, in particular the HSP70 family of proteins consisting of SSA1, SSA2, SSA3, SSA4, and SSE1 (Alexandre, et al., 2001, Chandler, et al., 2004). Trehalose synthesis genes were also up-regulated during ethanol stress in both studies, confirming their previously identified role in the survival of environmental stress (Gasch, et al., 2000, Alexandre, et al., 2001, Chandler, et al., 2004). Trehalose storage may inhibit protein denaturation and ethanol permeability in the membranes of ethanol-stressed cells. Lastly, despite the presence of ample glucose in the media, ethanol stress in both studies induced a starvation-like response in S. cerevisiae. This caused up-regulation of the genes involved in glycolysis and hexose transport in order to overcome inhibition of glycolysis and glucose sensing imposed by ethanol stress (Alexandre, et al., 2001, Chandler, et al., 2004). Down-regulated genes in both studies included RNA metabolism genes and protein biosynthesis genes, indicating that cell growth is inhibited as a possible mechanism to conserve cellular energy and adapt to ethanol stress conditions (Alexandre, et al., 2001). 16 1.5.2 Deletion Collection; Identification of Ethanol Tolerance Genes Another genomic resource that has been widely utilized for identification of ethanol tolerant genes is the S288C deletion strain collection in which each ORF has been systematically replaced with a drug-selectable marker and a unique molecular barcode. All deletion mutants can be screened for a particular phenotype through use of one of two methods. In the first method, individually-arrayed strains are grown under various conditions on solid media and colonies with reduced growth are identified. In the second method, the deletion collection is pooled, grown under a stress condition in liquid media, and deletion mutants that are enriched or depleted from the pool are identified. The latter requires isolation of genomic DNA, PCR amplification of barcode tags, and hybridization to an oligonucleotide array carrying the tag complements to determine strain abundance (Giaever, et al., 2002) A plethora of studies have utilized the S288C deletion set to screen for gene deletions that result in ethanol sensitivity. Kubota et al., (2004) screened the haploid deletion collection on solid 8% (v/v) and 11% (v/v) ethanol media and scored for colony size. A total of 256 genes were required for growth on 11% ethanol, with enrichment in functional categories of biosynthesis, protein transport, vacuole function, transcription, and mitochondria function. On 8% ethanol, 181 of the 256 gene deletion strains still showed a growth defect (Kubota, et al., 2004). Similarly, van Voorst et al., (2006) grew the haploid deletion collection on solid 6% (v/v) ethanol media; only 46 genes were required for growth, many of which functioned in the mitochondria, vacuole, and cell integrity pathway (which regulates cell wall stress and the DNA replication transition). Lastly, using liquid 5% (v/v) ethanol media with limited aeration, 175 genes were identified as required for growth equivalent to the parental strain (Kumar, et al., 17 2008). Of the 175 identified genes, many functioned within the vacuole and mitochondria, as well as in protein synthesis and ribosome and transcription processes. Surprisingly, only four ethanol tolerant genes were in common among the above studies,: GIM4, GIM5, GCN5, and KAR3 (Kumar, et al., 2008). These strikingly variable results suggested that the requirement for a particular gene\u00E2\u0080\u0099s function in ethanol tolerance was dependent upon experimental conditions, and thus further studies were warranted. More recently, Yoshikawa et al., (2009) reported a quantitative analysis of yeast deletion mutant growth rate in liquid media with and without 8% (v/v) ethanol. They identified 359 mutants sensitive to ethanol with 70% overlap to previously reported literature. In addition to functional categories previously identified, peroxisomal genes were also found to be largely ethanol sensitive when deleted. Teixeira et al., (2009) utilized both diploid and haploid yeast deletion mutant collections at 12% (v/v) and 8% (v/v) ethanol, respectively, and identified over 250 genes important for ethanol tolerance. Among these, vacuole, peroxisome, and endosome genes were also enriched. Given the widespread effects of ethanol toxicity on S. cerevisiae, it\u00E2\u0080\u0099s not surprising that many of the genes identified for ethanol tolerance also protect against other stressors. Aeusukaree et al., (2009) utilized the deletion collection to screen a variety of environmental stresses. They found 95 gene deletions that resulted in sensitivity to 10% ethanol. Of these genes, 35 are required for thermo-tolerance, 12 for tolerance to NaCl, and 7 for tolerance to H2O2 (Auesukaree, et al., 2009). The findings that ethanol and heat stress genes overlap has also been previously reviewed (Piper, 1995). In addition to the 359 ethanol-sensitive gene deletions identified in the quantitive analysis, 87 more genes were required for tolerance to both ethanol and osmotic stress (Yoshikawa, et al., 2009). Despite the advances in understanding ethanol 18 tolerance that these studies, and others, have established, they are not entirely without limitations. For example, the deletion collection may not always be accurately constructed or arrayed and does not allow for essential genes to be included in genome wide screens. Because some of these studies were conducted at low (5-8%) ethanol concentrations atypical of industrial fermentations, we suspect more genes may remain unidentified. 1.6 Research Proposal 1.6.1 The MoBY-ORF Library Each strain in the yeast deletion library contains a unique molecular barcode enabling all deletion mutants to be screened for a particular phenotype. This is done by growing the deletion pool under certain conditions, isolating genomic DNA, PCR amplification of barcode tags and hybridization to an oligonucleotide array carrying the tag complements (Giaever, et al., 2002). Recently, a molecular bar-coded yeast open reading frame (MoBY-ORF) plasmid library was generated where all S. cerevisiae bar-coded genes (along with their native promoters and terminators) were cloned into a centromere-based plasmid (Figure 2) (Ho, et al., 2009). The bar- codes enable quantification of the plasmids using microarray technology. A high copy based (2\u00EF\u0081\u00AD) plasmid library has been derived from the MoBY-ORF centromere base library (MoBY- ORF 2.0), such that dosage suppression studies can be conducted and plasmids conferring a growth advantage can be identified by quantification (Magtanong, et al., 2011). 19 Figure 2. Schematic representation of the 2\u00CE\u00BC MoBY-ORF plasmid. Each individual open reading frame (ORF) has been inserted with unique molecular barcodes flanking a KanMX drug resistance marker allowing for plasmid selection. The LEU2 gene also allows for selection with a leucine auxotrophic strain. Adapted from Ho et. al., 2009. 1.6.2 Research Outline In the above literature review many examples have been cited, including both naturally occurring and genetically engineered strains of S. cerevisiae, that demonstrate how gene over- expression can confer an adaptational advantage and improved tolerance to stressful environments. Therefore, I hypothesized that by utilizing the 2\u00EF\u0081\u00AD MoBY-ORF plasmid collection, I could identify genes that improve ethanol tolerance when over-produced, as they should be more abundant in an ethanol-stressed pool. The objectives of this study were three-fold. The first objective was to transform the MoBY-ORF 2.0 library into two strains of S. cerevisiae (S288C and the commercial wine yeast M2) and subject both strains to 12% ethanol stress to identify genes that confer a growth advantage. The second objective was to select a sub-set of 20 genes with which to conduct a follow-up analysis to confirm their role in ethanol tolerance. Two genes, RSA3 and RCN1, when over-produced were found to enable S. cerevisiae to tolerate the high ethanol concentrations typical of a wine fermentation. The last objective was to conduct model wine fermentations with S. cerevisiae strains over-expressing these genes of interest, to investigate whether they could improve the efficiency of a stressful wine fermentation. 21 2 Materials and Methods 2.1 Yeast Strains and Media Conditions 2.1.1 Yeast Strains Two strains of Saccharomyces cerevisiae were used for the majority of this research, an S288C strain YPH499 (MAT a his3\u00E2\u0088\u0086200 ade2-101 ura3-52 lys2-801 leu2\u00E2\u0088\u00861 trp1-\u00E2\u0088\u008663) and an industrial wine yeast M2 (WYM31 MAT a ho\u00E2\u0088\u0086::HygRhph leu2\u00E2\u0088\u0086::NatRMX4) derived from a diploid parental strain of M2 obtained as a gift to the UBC Wine Research Centre from Dr. Richard Gardner and outlined by (Bradbury, et al., 2006). Full details of the M2 wine strain including genetic manipulation are outlined below. The S288C deletion collection strain BY4741 (Mat a his3\u00E2\u0088\u00861 leu2\u00E2\u0088\u00860 met15\u00E2\u0088\u00860 ura3\u00E2\u0088\u00860) was also utilized when analysis of gene deletion was required in the S288C background. The two strains YPH499 and WYM31 are further referred to simply as S288C and M2 for the purpose of this study. 2.1.2 Genetic Manipulation of the M2 Wine Yeast To utilize the parental M2 strain with the 2\u00C2\u00B5 MoBY-ORF library it was necessary to first construct an M2 strain that is auxotrophic for leucine and sensitive to the drug G418 allowing for plasmid selection. A KanMX6 marker cassette was PCR amplified with primers containing 70 base pairs upstream of the LEU2 start codon and downstream of the LEU2 stop codon (Longtine, et al., 1998). Amplified PCR products were transformed via standard yeast lithium acetate (LiAc) methods and proper integration of the cassette at the LEU2 locus was confirmed by PCR with one primer annealing in the terminator region of the KanMX6 cassette and the other annealing downstream of the LEU2. Because the 2\u00C2\u00B5 MoBY-ORF 2.0 plasmids contain a KanMX6 marker, disruption of the LEU2 locus with KanMX6 would not allow G418 to be used 22 in plasmid selection. Thus, a NatMX4 marker with homologous promoter and terminator sequences to KanMX6 was used to switch markers via homologous recombination and selection on YPD containing 100 \u00CE\u00BCg mL-1 clonat (Tong, et al., 2001). Following sporulation, tetrad analysis confirmed proper integration and disruption of LEU2 as 2:2 segregation of the leucine auxotrophy and clonat resistance was observed in all tetrads. Mating types were determined by PCR as described (Huxley, et al., 1990) and a haploid M2 isolate, WYM31 (MATa ho\u00E2\u0088\u0086::HygRhph leu2\u00E2\u0088\u0086::NatRMX4) was selected for further utilization with the MoBY-ORF library in this study. Knockouts of genes identified in this study in the M2 background were also constructed as described above with 70 base pairs of homology specific to the flanking regions of each gene. Because gene knockout strains were not used in combination with 2\u00CE\u00BC over expression plasmids the KanMX6 marker did not have to be replaced. For sequences of primers utilized in this work see Table 1. 2.1.3 Yeast Propagation Media Both strains of S. cerevisiae were initially cultured and maintained with Yeast Peptone Dextrose (YPD) broth containing 1% yeast extract, 2% peptone and 2% glucose. After transformation with 2\u00CE\u00BC MoBY-ORF plasmids, transformants were selected and maintained with either YPD media supplemented with 100 \u00CE\u00BCg mL-1 G418 or with synthetic complete media containing 0.67% yeast nitrogen base without amino acids, 1% ammonium sulphate, 2% glucose and supplemented with all necessary amino acids except for leucine (SC -Leu ) (Guthrie & Fink, 1991). Solid agar plates were made by adding 2% agar (Difico Laboratories) to the above media. 23 Cultures were aerobically grown in an orbital shaker at 175 rpm or in a horizontal roller drum at a temperature of 25 \u00C2\u00B0 C unless otherwise stated. Table 1. Primers utilized in this study for gene disruption and mating type determination in the M2 strain of S. cerevisiae. Primer Name Sequence Reference LEU2 Knockout (Forward) 5\u00E2\u0080\u0099-ctttctaacttttcttaccttttacatttcagcaatatatatatatatatttcaa ggatataccattctacggatccccgggttaattaa-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 LEU2 Knockout (Reverse) 5\u00E2\u0080\u0099-tcgtgtcgtttctattatgaatttcatttataaagtttatgtacaaatatca taaaaaagagaatctttgaattcgagctcgtttaaac-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 Universal Integration confirmation (Forward) 5\u00E2\u0080\u0099-gagtactgacaataaaaagattc-3\u00E2\u0080\u0099 Longtine et al. 1998 LEU2 Integration confirmation (Reverse) 5`- cttgataaatgtatgtagattgc -3` This Study RCN1 Knockout (Forward) 5\u00E2\u0080\u0099-ttccatgaaaaaaaagagggccaaaaagatcaagcaataaaccaac cgatatataaaacacagaactgcagcggatccccgggttaattaa-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 RCN1 Knockout (Reverse) 5\u00E2\u0080\u0099-atattttatattttaaagcaaccacccgtaagcatttaagtctcttaagc caacaaatcgcctcgccatcgaattcgagctcgtttaaac-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 RCN1 Integration Confirmation (Reverse) 5\u00E2\u0080\u0099-ggaggtagactcggacgag-3\u00E2\u0080\u0099 This Study RSA3 Knockout (Forward) 5\u00E2\u0080\u0099-atatatacaagaatttaaactgcccacttgaaatagcacggaaagaa ctagccataatatattatacaggcggatccccgggttaattaa-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 RSA3 Knockout (Reverse) 5\u00E2\u0080\u0099-aatgtgcacgtcaatatattctccgcggaaacatgacaaacttttaga aaaagatttaaagcaatattttgaattcgagctcgtttaaac-3\u00E2\u0080\u0099 This Study; Longtine et al. 1998 RSA3 integration confirmation (Reverse) 5\u00E2\u0080\u0099-gcacaacttggatcaggacc-3\u00E2\u0080\u0099 This Study MAT primer \u00E2\u0080\u0093 non specific 5\u00E2\u0080\u0099-agtcacatcaagatcgtttatgg-3\u00E2\u0080\u0099 Huxley et al. 1990 MAT primer \u00E2\u0080\u0093 \u00CE\u00B1 specific 5\u00E2\u0080\u0099-gcaccggaatatgggactacttcg-3\u00E2\u0080\u0099 Huxley et al. 1990 MAT primer \u00E2\u0080\u0093 a specific 5\u00E2\u0080\u0099-actccacttcaagtaagagtttg-3\u00E2\u0080\u0099 Huxley et al. 1990 2.1.4 Escherichia coli Propagation Media Escherichia coli was amplified by growth overnight at 37 \u00C2\u00B0 C in 2YT media containing 1.6% Bacto-tryptone, 1% Bacto-yeast extract, 0.5% NaCl, (pH 7.0) and supplemented with 0.4% glucose, 200 \u00C2\u00B5g ml-1 carbenicillin and 100 \u00C2\u00B5g ml-1 G418 for selection of 2\u00CE\u00BC MoBY-ORF 24 plasmids. Cultures were grown aerobically in an orbital shaker at 175 rpm for at least 12-16 hours. 2.1.5 Microtiter Plate Growth Conditions For experiments where cells were grown in 96 well microtiter plates, an Infinite\u00C2\u00AE Pro micro-plate reader (Magellan v7.0 TM Software, Tecan Group Ltd.) was used as follows. Cultures were diluted from log phase to an optical density (O.D. 600nm) of 0.1 and a 200 \u00C2\u00B5L volume was added to each well. The plate was sealed with a transparent film (Beckman Coulter\u00C2\u00AE USA) to minimize evaporation. Air exchange was facilitated by puncturing the film near the edge of each well with a 27.5 gauge hypodermic needle. O.D. 600nm was read after kinetic cycles of 20 minutes with equal amounts of orbital and linear shaking at an amplitude of 2.5 mm. The temperature of the growth chamber was maintained at 25\u00C2\u00B0C except where otherwise noted. 2.2 MoBY-ORF Library 2.2.1 MoBY-ORF 2.0 Library Transformations The 2\u00C2\u00B5 MoBY-ORF 2.0 library was received frozen in a 1mL aliquot as a gift from Dr. Charlie Boone. Prior to utilization, the library was first amplified in 2YT media as described above in section 2.1.4. Cells from 1.5 mL aliquots (O.D.600nm ~ 20) were collected by centrifugation and suspended in 50 \u00C2\u00B5l of 2YT containing 15% (v/v) glycerol and stored at -80 o C until needed. Plasmids from one aliquot were extracted with a Genejet Plasmid Miniprep Kit (Fermentas) following the manufacturer\u00E2\u0080\u0099s instructions. The entire 50 \u00CE\u00BCL elution was used for transformation into S288C using a LiAc method as described by Butcher and Schrieber (2006). For transformation of the 2\u00C2\u00B5 MoBY-ORF 2.0 library into M2, electroporation was used to obtain sufficiently high yields of transformants. 50 mL of log phase M2 cells (~O.D.600nm 1.0) 25 were harvested by centrifugation at 4\u00C2\u00B0C and resuspended in 10mL of TE/LiAc buffer (pH 7.5). After 45 minute incubation at 30\u00C2\u00B0C, 250 \u00C2\u00B5L of 1M DTT was added and further incubated for 15 minutes at 30\u00C2\u00B0C. The yeast suspension was then diluted to 50mL with dH20 and washed and concentrated three times (4000 Xg, 4\u00C2\u00B0C) with 25 mL dH20, 3 mL 1M sorbitol, and 0.5 mL 1M sorbitol respectively after which 40 \u00C2\u00B5L of concentrated yeast (O.D.600nm ~50) was mixed with a single aliquot of plasmid DNA obtained from an extraction as described above. The mixture was transferred to a 0.2 cm gap electroporation cuvette and the cuvette was pulsed at 1.5 kV, 25 \u00C2\u00B5F, 200 ohms, after which cells were recovered in YPD for 2 hours at 25 \u00C2\u00B0 C. In both transformation protocols, cells were spread onto 150mm SC -Leu plates and grown for 2-3 days at 25\u00C2\u00B0C until single colonies were obtained. Approximately 40 X 10 3 colonies were scraped off the plates and pooled into 20 mL YPD containing 15% (v/v) glycerol and 100 \u00C2\u00B5g mL -1 G418 and frozen in 200 \u00CE\u00BCL aliquots at ~O.D.600nm 20 at -80 o C. 2.2.2 Individual MoBY-ORF Plasmid Transformations Individual MoBY-ORF plasmids containing genes of interest were received in E. coli on solid 2YT agar plates with selective drugs as described (2.1.4). Single colonies were amplified and frozen with 15% (v/v) glycerol at -80 \u00C2\u00B0 C until needed. To transform individual MoBY-ORF 2\u00CE\u00BC plasmids extraction of plasmids was conducted as described (2.2.1), however only 3-5 \u00CE\u00BCL of plasmid elution was needed per transformation reaction. A standard LiAc protocol was utilized for both S288C and M2 with selection on 100 mm SC -Leu plates. 26 2.2.3 Empty Vector Ethanol Tolerance Assay To determine the concentration of ethanol required to stress cells for the MoBY-ORF ethanol tolerance screen, strains of S288C and M2 carrying an empty vector were incubated in SC -Leu media supplemented with of 8-14% (v/v) ethanol. Over a course of 70 hours, O.D.600nm was recorded at intervals of roughly 16 hours and appropriate serial dilutions were performed to spread 200 cells onto SC -Leu plates. After 2-3 days of growth at 25 \u00C2\u00B0 C, viable colonies were counted to determine a viability percentage. 2.2.4 MoBY-ORF Genome-wide Ethanol Tolerance Screen To screen for genes that improve ethanol tolerance when over expressed, the following approach was used for both the M2 and S288C MoBY-ORF 2.0 (2\u00C2\u00B5) plasmid libraries. A single aliquot of each MoBY-ORF library (see 2.2.1) was diluted into 10 mL of SC -Leu (to select for the plasmid) and grown overnight. After 12-16 hours of growth the plasmid pools were again diluted and grown to log phase (O.D.600nm ~1.0) in 20 mL SC -Leu . The pool was then split into two 10 mL cultures at an O.D.600nm ~0.2, one containing SC -Leu with 12% v/v ethanol (stress condition) and one with SC -Leu containing no ethanol (control). S. cerevisiae in both the stress and control conditions were grown aerobically for 48 hours at 25\u00C2\u00B0C. The control pool was maintained in log phase by dilution every 12 hours in fresh SC -Leu media. To prevent exhaustion of glucose and nutrients over 48 hours in the ethanol stress pool, cells were pelleted and re- suspended in fresh SC -Leu 12% ethanol media after 24 hours. After 48 hours, cells from both the stress and control conditions were harvested and plasmids were extracted using a Genejet Plasmid Miniprep Kit (Fermentas) following manufacturer\u00E2\u0080\u0099s directions (Butcher & Schreiber, 2006). 27 Plasmids were stored at -20 o C until being shipped (on ice) for barcode analysis. A competitive hybridization was performed for each screen as outlined previously (Magtanong, et al., 2011). Briefly, barcodes from the 12% (v/v) ethanol condition were amplified with biotinylated universal TAG4 primers whereas non-biotinylated universal TAG4 primers were used to amplify barcodes from the 0% ethanol control condition. Each hybridization mixture contained 3:1 (v/v) non-biotinylated:biotinylated PCR amplified barcodes. Barcodes were hybridized to a TAG4 microarray as previously described (Pierce, et al., 2006). 2.3 FunSpec Genome-Wide Screen Analysis A web based program FunSpec (http://funspec.ccbr.utoronto.ca) was utilized to query ORFs associated with the 200 highest barcode raw average values from the M2 and S288C screens. FunSpec compared these subsets of ORFs against existing Munich Information Centre for Protein Sequencing (MIPS) and Gene Ontology (GO) databases for functional classification, protein complexes and protein localization annotation (Robinson, et al., 2002). Gene categories with a p value < 0.01 were considered significant. 2.4 Ethanol Tolerance Confirmation Assays A variety of assays were utilized in attempt to confirm ethanol tolerance due to over expression of genes identified from the MoBY-ORF library screens. Ethanol stress levels ranged from 12-20% (v/v) depending upon the length of exposure for each assay. Individual plasmids were transformed into both S288C and M2 and all assays were conducted identically regardless of strain background. A control strain carrying an empty vector was used in all assays. 28 2.4.1 Growth Analysis in 12% v/v Ethanol In order to replicate conditions of the initial 12% ethanol stress environment of the MoBY-ORF library screens, 5 mL cultures of S288C or M2 carrying over expression plasmids were grown to log phase (O.D.600nm ~ 0.6-0.8). Each strain was diluted (in quadruplicate) to an O.D.600nm of 0.15 in minimal media containing 12% v/v ethanol in a 96 well microtiter plate. Growth was assessed by monitoring the changes in O.D.600nm as described (2.1.5) over a period of up to 60 hours. Mean values for each strain were plotted versus time for graphical analysis, with the standard deviation for O.D.600nm given at 10 hour intervals to indicate statistical significance. 2.4.2 Growth Analysis Following 16% v/v Ethanol Shock To test strains carrying individually transformed MoBY-ORF 2\u00C2\u00B5 plasmids for improved tolerance to ethanol, 5 mL log phase cultures (O.D.600nm ~ 0.6-0.8) were diluted to an O.D.600nm 0.1 in SC -Leu media containing 16% (v/v) ethanol and incubated for 3 hours in a horizontal roller drum at 25\u00C2\u00B0C. After incubation, cells were collected by centrifugation and resuspended in SC -Leu media containing no ethanol. Recovery of MoBY-ORF 2\u00C2\u00B5 strains after 16% ethanol shock was monitored in multiple ways. First, 200 \u00C2\u00B5L of each strain in SC -Leu media were added in quadruplicate to 96 well microtiter plates and growth was measured by recording O.D.600nm for 24 hours as described above. Second, serial dilutions of each strain were spotted onto solid SC - Leu media plates with and without 5 \u00CE\u00BCM phloxine B (SC-Leu/PhloxB) in order to visually assess viability. Phloxine B is a pink dye that only stains inviable S. cerevisiae cells as it passively diffuses through the membrane. As a control, initial cultures at O.D.600nm = 0.1 were serially diluted and spotted onto SC -Leu and SC -Leu/PhloxB plates prior to ethanol exposure. Cell viability 29 was visually assessed by relative density of each spot after an incubation period of 2-3 days at 25 \u00C2\u00B0 C. 2.4.3 Viability Assay Following 17-21% v/v Ethanol Shock As a final ethanol tolerance confirmation assay, 5 mL log phase cultures of each strain carrying over-expression plasmids were again diluted to an O.D.600nm of 0.1 and exposed to extremely toxic levels of ethanol. In addition to a 0% ethanol control, strains were exposed to 17, 18, 19, 20, and 21% v/v ethanol for 1 hour. Following the ethanol treatment, 5 \u00CE\u00BCL of each strain were spotted onto SC -Leu plates and incubated for 2-3 days at 25 \u00C2\u00B0 C after which cell viability was visually assessed. This assay was carried out in duplicate for confirmation of results. 2.4.4 Deletion Collection Ethanol Sensitivity Spot Assays As another method to infer a role in ethanol tolerance for genes of interest, strains containing null mutations were grown to log phase (O.D.600nm 0.6-0.8) in YPD prior to dilution to O.D.600nm of 0.1. Serial dilutions were then spotted onto YPD media supplemented with 0, 6, 8, or 10% (v/v) ethanol. Rather than constructing gene disruption strains in our lab S288C strain background, null mutants were selected from the S288C yeast deletion strain BY4741 (Mat a). Deletions in the M2 background were constructed as described above (2.1.2). Strains were visually assessed for lack of growth after 2-7 days at 25 \u00C2\u00B0 C in a sealed plastic container to minimize ethanol evaporation from the solid media. 30 2.5 Zymolyase Cell Wall Digestion Assay Differences in sensitivity to cell wall perturbing agents were tested using a zymolyase cell wall digestion assay adapted to a 96 well plate microtiter format as described (Ovalle, et al., 1999) with exception to the utilization of 20T Zymolyase in this study rather than 100T. Briefly, overnight cultures were grown to log phase and resuspended in TE with 5% polyethylene glycol 8000 (PEG 8000) at O.D.600nm 2.0. In a 96 well plate, 200 \u00C2\u00B5L aliquots of each strain were added to individual wells in quadruplicate followed by 50 \u00CE\u00BCl of 20T Zymolyase solution. Different concentrations of 20T Zymolyase were used for each strain background (200 \u00C2\u00B5g mL -1 for S288C/BY4741 and 500 \u00C2\u00B5g mL -1 for M2) due to the intrinsic differences in cell wall strength between S288C and M2. Changes in O.D.600nm were measured by reading absorbance at 600nm once a minute for up to 60 minutes. The O.D.600nm/O.D.600nm initial was calculated for each replicate and mean values were determined for each time point. Log transformed values were plotted versus time with error bar values calculated using the formula log(SDx)=0.5[Log(X+SDx)- Log(X-SDx)] where X is the mean O.D.600nm/O.D.600nm initial . The maximum lysis rate (MLR) and lag time for each strain were also calculated as previously described (Ovalle, et al., 1999). 2.6 Environmental Stress Assays 2.6.1 Environmental Stress Growth Curve Analysis Growth in the presence of a variety of other environmental stress media was monitored by O.D.600nm in a 96 well microtiter plate as described above. Under all stress conditions, cells from log phase cultures (O.D.600nm ~ 0.6-0.8) grown at 25 \u00C2\u00B0 C in SC -Leu were added in triplicate to a 96 well plate and immediately transferred to the growth chamber. Due to inherent stress tolerance differences between M2 and S288C, different stress levels were utilized to elicit a 31 similar level of stress response in each strain for heat, osmotic and oxidative stresses. For heat stress, log phase cultures (O.D.600nm ~ 0.6-0.8) grown at 25 \u00C2\u00B0 C in SC -Leu were added to a 96 well plate and immediately transferred to the growth chamber preheated to 37 \u00C2\u00B0 C for S288C and 42 \u00C2\u00B0 C for M2. For osmotic stress, log phase cells were grown directly in SC -Leu supplemented with 0.4 M NaCl (S288C) or 1.0 M NaCl (M2). For oxidative stress, growth in 0.75 mM H2O2 (S288C) or 1.0 mM H2O2 (M2) SC -Leu media was utilized. 2.6.2 Cycloheximide Translation Inhibition Assays To test sensitivity to translation inhibition, log phase cells (O.D.600nm 0.6-0.8) grown in SC -Leu were suspended in SC -Leu supplemented with the drug cycloheximide at a range from 25 to 100 ng mL -1 . Growth was measured by monitoring O.D.600nm in a 96 well microtiter plate as outlined above. Additionally, log phase cells diluted to an O.D.600nm of 0.1 as above were also serially diluted and spotted onto SC -Leu agar plates with and without 100 and 500 ng mL -1 cycloheximide and growth was visually assessed after 3 days at 25 \u00C2\u00B0 C. 2.7 Growth Curve Statistical Analysis Growth curves generated from the automated plate reader from 16% (v/v) ethanol shock assays (2.4.2) and environmental stress assays (2.6.1) were analyzed for statistical significance as follows. The O.D.600nm of each time point was averaged amongst replicates and the mean values were graphed versus time. We analyzed three distinct growth phases for individual replicates: The lag phase, the logarithmic growth rate, and the maximum O.D.600nm (O.D.600nm max). The logarithmic growth rate was determined to be the slope of a linear trend line fit to the exponential portion of each curve. Where the exponential growth phase was less distinct, the growth rate was 32 taken to be the slope of a linear trend line fit to the maximal growth increase consisting of at least 12 time points (4 hours). Minimum O.D.600nm values (O.D.600nm min) were calculated by averaging the first six time points of each curve, and the lag phase was determined to be time of intersection between O.D.600nm. min values and an extrapolated tangent line from the exponential growth phase. Finally, O.D.600nm max values were calculated by averaging the last six time points of each curve. For each of the parameters measured a Student\u00E2\u0080\u0099s t-test was used to determine if the mean values for strains carrying 2\u00EF\u0081\u00AD MoBY-ORF plasmids were significantly different from a control strain carrying an empty 2\u00CE\u00BC vector. 2.8 Model Wine Fermentations 2.8.1 Synthetic Grape Juice Media and Growth Conditions Model fermentations were conducted in filter sterilized synthetic wine media (SWM) initially containing 10% glucose, 10% fructose, 4.5 g L -1 of both tartaric and malic acid, 0.3 g L -1 citric acid, 0.1% tween 80, 1.7g L -1 yeast nitrogen base without amino acids, 2g L -1 ammonium sulphate, and supplemented with all necessary amino acids except for leucine. High osmolarity (for increased alcohol yield) fermentations were conducted with SWM as formulated above with the addition of glucose and fructose to 32% rather than 20%. The pH of SWM was adjusted to 3.0 with 5M KOH followed by storage at 4 \u00C2\u00B0 C until needed. Model fermentations were conducted in 80 mL of SWM in sterile 100 mL media jars each containing a magnetic stir bar and sealed with a sterilized (70% v/v ethanol) vapour lock to maintain an anaerobic environment while allowing release of CO2. Each strain of S. cerevisiae was first grown to log phase (O.D.600nm 0.6- 0.8) in SC -Leu prior to inoculation into SWM at an O.D.600nm of 0.2. Fermentations were allowed to proceed at room temperature and samples of the SWM were withdrawn from the fermentation 33 every two days with a sterile hypodermic needle. Prior to sampling, yeast cells were evenly resuspended by magnetic stirring for 10 minutes. Fermentation progression was also monitored on a daily basis by weight loss due to CO2 evolution, and fermentations were deemed completed when no further weight loss was noted for at least 2 days. 2.8.2 Fermentations Analysis via High Pressure Liquid Chromatography Samples of fermenting SWM were filtered with a 0.22 \u00CE\u00BCm sterile filter to remove yeast cells from the medium. Ethanol, glucose and fructose levels in the SWM supernatant were quantified with an Agilent 1100 series high pressure liquid chromatography (HPLC) apparatus (Agilent Technologies, Palo Alto, CA) using a Supelcogel C-61OH 30 cm x 7.8 mm column (Sigma Aldrich, Oakville USA) maintained at 50 \u00C2\u00B0 C over a 22 minute isocratic run with 0.1% phosphoric acid at 0.75 mL/minute (Adams & Van Vuuren, 2010). Compounds were monitored with an Agilent G1362A refractive index detector with positive polarity and a 35 \u00C2\u00B0 C optical unit temperature. Quantification was based on standard curves using Agilent LC-MS ChemStation revision A.09.03 software. 34 3 Results 3.1 Genome-Wide Ethanol Tolerance Screens in S288C and M2 3.1.1 Empty Vector Controls Have Reduced Viability in 12% v/v Ethanol To determine an appropriate level of ethanol toxicity for our haploid strains of S. cerevisiae, M2 and S288C, strains carrying 2\u00CE\u00BC MoBY-ORF empty vectors were incubated for up to 72 hours in minimal media containing 8-14% (v/v) ethanol. During incubation, 200 cells were plated directly onto minimal media plates containing no ethanol for viability counts. Between 48- 60 hours, both M2 and S288C vector control strains showed less than 40% viability when incubated in 12% (v/v) ethanol media (Figure 3). At lower concentrations of ethanol cell viability was greater than 60% (data not shown). In the 14% (v/v) ethanol treatment, although viability was also low, no changes in O.D.600nm were observed suggesting that transcription and translation may have been fully inhibited. Therefore, 12% (v/v) ethanol was selected as an inhibitory level of stress to conduct the genome wide screens with as it allowed for minimal growth and a substantial reduction in cell viability. Figure 3. Both S288C and M2 strains of S. cerevisiae show reduced viability in 12% (v/v) ethanol after 48 hours. (a) S288C and (b) M2 strains carrying an empty vector were grown in minimal media with 12% (v/v) ethanol; 200 cells were periodically plated for viability counts and reported as % viability. Error bars indicate +/- standard deviation. 0 20 40 60 80 100 120 140 24 42 48 66 72 % V ia b il it y Hours a. 0 20 40 60 80 100 120 6 27.5 50 72 % V ia b il it y Hours b. 35 3.1.2 Twenty Ethanol Tolerant Candidate Genes Identified in S288C and M2 To identify genes involved in S. cerevisiae ethanol tolerance, two strains of S. cerevisiae \u00E2\u0080\u0093 the lab strain S288C and the industrial wine yeast strain M2 \u00E2\u0080\u0093 were independently transformed with the 2\u00C2\u00B5 MoBY-ORF 2.0 2\u00EF\u0081\u00AD plasmid library and each set of transformants was grown in 12% (v/v) ethanol as the stress condition and 0% ethanol as the control condition. After 48 hours of incubation, plasmid barcodes from the 12% ethanol stress pool were quantified relative to the 0% control pool. Quantified raw average values for all uptag and downtag barcodes identified by microarray from the S288C and M2 (Figure 4 a,b) screens are shown. To validate the results of these screens, cut off barcode values were set arbitrarily at 600 for M2 and 800 for S288C, as these yielded raw hybridization values significantly above background (boxed area, Figure 4 a,b). Barcode values exceeding the respective cut-off values from each screen are shown collectively in Figure 4c. With this criterion 14 plasmids were identified in the M2 screen and 11 plasmids in the S288C screen (Table 2) that were more abundant in 12% ethanol versus control cultures. Five of the plasmids from the M2 and S288C screen had genes in common, which also happened to be the top ranked genes from the S288C screen. These genes are: RCN1, MAP2, EMI5, RSA3 and YAT1 (Figure 4c). 36 Figure 4. Quantified abundance of MoBY-ORF barcode tags after prolonged ethanol stress (12% v/v) in S. cerevisiae. Descending order raw average values of all barcode tags are shown for parallel screens with (a) S288 and (b) M2. (c) Genes associated with barcode values above 600 in M2 (black) and 800 in S288C (white) are presented. Boxed area indicates genes that exceed cut off values. 0 200 400 600 800 1000 1200 1400 1 8715 R aw A v er ag e V al u es Ordered Barcode Tags a. 0 100 200 300 400 500 600 700 800 900 1 8715 R aw A v er ag e V al u e Ordered Barcode Tags b. 0 200 400 600 800 1000 1200 1400 Q u an ti fi ed B ar co d e V al u e c. S288C M2 37 Table 2. Genes identified that improve ethanol tolerance when over-expressed in Saccharomyces cerevisiae strains S288C or M2. ORF Gene (Aliases) Biological Process ** Cellular Component ** Strain YKL159c RCN1 Calcium Mediated Signalling Cytoplasm M2 / S288C YBL091c MAP2 Protein Maturation (Methionine Removal) Cytoplasm M2 / S288C YOL071w EMI5 (SDH5) Cellular Respiration / TCA cycle Mitochondrion M2 / S288C YLR221c RSA3 Ribosome Large Subunit Assembly Nucleolus M2 / S288C YAR035w YAT1 Alcohol / Carnitine Metabolism Mitochondrion M2 / S288C YLR251w SYM1 Ethanol Metabolism Mitochondrion M2 YMR283c RIT1 Charged tRNA amino acid modification Cytoplasm M2 YLR271w Unknown Cytoplasm/Nucleus M2 YOL001w PHO80 Regulation of cyclin dependant protein kinase activity Nucleus M2 YDR399w HPT1 GMP/IMP salvage Cytoplasm/Nucleus M2 YIL155c GUT2 Glycerol Metabolism / NADH oxidation Mitochondrion M2 YNL317w PFS2 mRNA cleavage/ polyadenylation Cytoplasm M2 YGR216c GPI1 Glycosylphosphatidylinositol anchor biosynthesis Membrane M2 YLR456w Unknown Unknown M2 YCR046c IMG1 Mitochondrial Translation Mitochondrion S288C YML079w Unknown Cytoplasm/Nucleus S288C YKL043w PHD1 Positive Regulator RNA Polymerase II promoter/ Pseudohyphal growth Nucleus S288C YDR072c IPT1 (SYR4, KTI6) Sphingolipid Biosynthesis Membrane Fraction S288C YML050w AIM32 Unknown Unknown S288C YNL277w-A Unknown Unknown S288C 3.1.3 Splicing Genes are Enriched in Both S288C and M2 Enrichment of biological processes in the dataset was determined by taking the top 200 genes (an arbitrary cut-off) from the M2 and S288C dataset and performing FunSpec 38 (http://funspec.ccbr.utoronto.ca) analysis (Robinson, et al., 2002). Significantly enriched gene categories were defined as those with p-values of less than 0.01 as determined by FunSpec. The MoBY ORF 2.0 collection, however, contains only ~4500 of the 6200 predicted S. cerevisiae ORFs. As the FunSpec p-value calculation is based on every gene from a given functional category, the given p-values (Table 3,4) are only estimates, and true p-values would be reduced and therefore have increased significance . The S288C ethanol screen was found to be enriched for genes involved in splicing and for genes with a role in transcription termination (Table 3). Genes identified in the M2 ethanol screen were also enriched in the functional category of splicing in addition to mitochondrial transport genes and single carbon (C1) catabolism (Table 4). There is an overlap of 134 genes between the top 200 genes from the M2 and S288C screen \u00E2\u0080\u0093 this group of genes is enriched in the categories of mitochondrial transport, splicing and transport facilities (data not shown). Table 3. Functional classification of gene enrichment from ethanol tolerance screen in S288C Functional category a Genes Total genes observed Total genes in dataset b p value b Transcription Termination 11.02.03.01.07 ESS1, PTA1, SUB1 3 14 (10) 7.68 x 10 -3 Splicing 11.04.03.01 CPD1, DBR1, LEA1, LSM4, LSM5, LSM7, PFS2, PML1, PTA1 10 139 (108) 9.0 x 10 -3 a Functional categories are defined by FunSpec and corresponding category numbers are given. b p-values are estimates only based on all genes in data set. Actual number of genes in the MoBY-ORF library is indicated in brackets. 39 Table 4. Functional classification of enriched genes from ethanol tolerance screen in M2 Functional category a Genes Total genes observed Total genes in dataset b p value b Mitochondrial Transport 20.09.04 ATM1, ATP15, CTP1, CYC2, HSP78, MIR1, PET8, TIM22, TIM44, YAT1, YDL119C 11 104 (86) 2.72 x 10 -4 Splicing 11.04.03.01 CPD1, DBR1, LEA1, LSM2, LSM4, LSM5, MTF2, PFS2, PML1, RSE1, SYF1 11 139 (108) 3.0 x 10 -3 C-1 compound catabolism GCB3, YJL068C 2 5 (3) 8.51 x 10 -3 a Functional categories are defined by FunSpec and corresponding category numbers are given. b p-values are estimates only based on all genes in data set. Actual number of genes in the MoBY-ORF library is indicated in brackets. 3.2 Confirmation of Ethanol Tolerance 3.2.1 Over-expression of Five Genes Common to S228C and M2 Because five highly ranked genes (RCN1, MAP2, RSA3, EMI5 and YAT1) were common hits to both the M2 and S288C ethanol tolerant screens, strains carrying high copy plasmids expressing these genes were tested for improved growth upon exposure to toxic levels of ethanol by a variety of assays as outlined (Materials and Methods, 2.4). Both M2 and S288C strains were individually transformed with RCN1, MAP2, RSA3, EMI5 and YAT1 plasmids in addition to an empty vector (EV) control plasmid. First, growth of over-expression strains was monitored in SC -Leu media containing 12% ethanol, the same conditions utilized in genome wide screens. Growth curves were characterized by two distinct growth phases separated by a short lag period, suggesting a shift in metabolism known as the diauxic shift may be occurring following depletion of glucose. Growth following this metabolic shift was highly variable and statistical significance could not be determined between any strain and the vector control. An example of 40 M2 growth curves is shown in Figure 5. Similar data was obtained from S288C strains and no statistical relevance was found (data not shown). Figure 5. Growth of M2 over-expression strains in minimal 12% ethanol media. Strains of the M2 wine yeast carrying 2\u00CE\u00BC plasmids over producing RCN1 (red), MAP2 (purple), RSA3 (green), EMI5 (yellow), and an empty vector control (blue) were monitored in 96 well plates over 60 hours. Error bars indicate +/- standard deviation at each time point. 3.2.2 RCN1 and RSA3 Over-expression Improves Viability Following 16% v/v Ethanol Variability in S. cerevisiae growth in 12% ethanol could have been from variable ethanol evaporation from liquid media cultures over the prolonged incubation periods. Moreover, it has been suggested that during ethanol exposure changes in O.D. may be attributed, at least in part, to changes in cell size and not entirely to active cell division (Kubota, et al., 2004). Thus to circumvent this potential bias, an alternative method to characterize ethanol tolerance was 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 10 20 30 40 50 60 O .D . 6 0 0 n m Hours M2+Vector M2+RCN1 M2+MAP2 M2+RSA3 M2+EMI5 41 employed. Cells were exposed to 16% (v/v) ethanol in minimal media for three hours, then growth of resuspended cells in SC -Leu media was monitored using an automated 96 well plate reader. Compared to results from conditions similar to our primary ethanol tolerance screen (12% ethanol for 48 hours, Figure 5) exposure to 16% ethanol for 3 hours yielded higher reproducibility in the resulting growth curves, which was likely due to the rapid ethanol shock over a shorter period of time. Growth curve analysis consisted of measuring the lag phase, logarithmic growth rate, and the maximum O.D.600nm (O.D.600nmmax) following ethanol shock. Mean values for growth curve parameters of each strain and plasmid combination following ethanol shock are shown in Table 5 (S288C) and Table 6 (M2). In S288C, over-expression of RCN1, RSA3 and MAP2 (Figure 6a) resulted in growth curves characterized by significantly faster growth rates and higher O.D.600nm max compared to the vector control strain (Table 5). Neither EMI5 nor YAT1 over expression significantly improved growth following 16% ethanol shock in S288C. Table 5. Growth curve analysis a of S288C strains carrying 2\u00C2\u00B5 over-expression plasmids after 16% ethanol shock. Strain + plasmid Mean lag time (hours) Mean growth rate (10 5 cells mL -1 /hr) Mean O.D.max (10 7 cells mL -1 ) S288C + vector 6.84 2.07 0.388 S288C + RCN1 8.87 3.56 * 0.571 * S288C + MAP2 10.0 5.18 *** 0.722 ** S288C + EMI5 6.94 1.93 0.410 S288C + RSA3 8.34 5.74 ** 0.765 ** S288C + YAT1 8.81 1.43 0.335 a analysis was conducted with a Student\u00E2\u0080\u0099s t-test comparing mean values for lag phase, logarithmic growth rate and O.D.600nm max values for each plasmid with the vector control strain. * significance of p<0.05, ** significance of p<0.01, *** significance of p<0.001 42 Table 6. Growth curve analysis a of M2 strains carrying 2\u00C2\u00B5 over-expression plasmids after 16% ethanol shock. Strain + plasmid Mean lag time (hours) Mean growth rate (10 5 cells mL -1 /hr) Mean O.D.max (10 7 cells mL -1 ) M2 + vector 14.4 12.1 1.41 M2 + RCN1 10.5 *** 12.6 1.46 M2 + MAP2 12.7 *** 8.41 ** 1.17 ** M2 + EMI5 14.1 10.7 1.24 * M2 + RSA3 11.5 *** 11.4 1.39 M2 + YAT1 13.7 * 13.1 1.45 a analysis was conducted with a Student\u00E2\u0080\u0099s t-test comparing mean values for lag phase, logarithmic growth rate and O.D.600nm max values for each plasmid with the vector control strain. * significance of p<0.05, ** significance of p<0.01, *** significance of p<0.001 In the M2 background, RCN1 and RSA3 (Figure 6b) over-expression significantly improved recovery from 16% ethanol shock by reducing the lag phase prior to logarithmic growth (Table 6). As the length of lag phase can be effected by the physiological competence of cells damaged from prior events (McMeekin, et al., 2002) the shorter lag phases observed upon RCN1 and RSA3 over-expression may be due to less cellular damage induced by 16% ethanol exposure. Compared to the vector control, the logarithmic growth rate and O.D.600nm max values did not significantly differ from the EV with RCN1 or RSA3 gene over-expression and were actually reduced with MAP2 over-expression (Figure 6b) contrary to the results in S288C (Table 5). The lag phase with MAP2 and YAT1 over-expression in M2 was also reduced compared to the EV control however YAT1 showed no improvement in S288C and MAP2 showed a decrease in log phase and O.D.600nm max in M2. No significant differences upon EMI5 over expression were observed in M2 or S288C. To further confirm ethanol tolerance of RCN1 and RSA3 after 16% ethanol shock, serial dilutions of strains shocked with ethanol for two or three hours were spotted directly onto SC -Leu to assay cell viability. Because no differences in viability were detected between over expression strains and the vector control on SC -Leu plates (data not shown), plates were 43 supplemented with 5\u00CE\u00BCM phloxine B to indicate viability. Phloxine B is a pink dye that is passively absorbed by inviable cells but does not enter viable cells, allowing subtle differences in viability to be determined by differences in color. Unexpectedly, following 16% ethanol shock, a drastic change in viability between cells plated on SC -Leu and SC -Leu + Phloxine B was noted (Figure 7a). Because ethanol targets the cell membrane and weakens the cell wall in S. cerevisiae, membrane damage may have allowed influx of the dye in otherwise viable cells resulting in increased cell death. Although no differences in coloration were noted between strains on Phloxine B plates, over expression of RSA3 and RCN1 in both M2 and S288C (Figure 7 b,c) improved the viability of cells following 16% (v/v) ethanol shock compared to the vector control. 44 Figure 6. High copy expression of RCN1 and RSA3 improves recovery following 16% (v/v) ethanol shock in both S288C and M2 strains. 2\u00CE\u00BC MoBY-ORF 2.0 vector (blue), RCN1 (red), RSA3 (green), MAP2 (purple), EMI5 (yellow) and YAT1 (orange) plasmids were individually transformed into (a) S288C and (b) M2 and incubated in minimal 16% (v/v) ethanol media for 3 hours. Growth of strains following release from ethanol shock into minimal media was monitored in quadruplicate using an automated 96 well plate reader by measuring O.D.600nm every 20 minutes. Mean O.D.600nm is plotted over a 24 hour period. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 5 10 15 20 25 O .D . 6 0 0 n m Hours a. S288C+Vector S288C+RCN1 S288C+MAP2 S288C+RSA3 S288C+EMI5 S288C+YAT1 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 5 10 15 20 25 30 O . D . 6 0 0 n m Time (Hours) b. M2+Vector M2+RCN1 M2+MAP2 M2+RSA3 M2+EMI5 M2+YAT1 45 Figure 7. M2 and S288C strains over producing RCN1 and RSA3 show increased viability following 16% (v/v) ethanol shock on Phloxine B plates. (a) S288C and M2 vector control strains show reduced viability on SC -Leu plates supplemented with Phloxine B following 16% (v/v) ethanol shock compared to SC -Leu without Phloxine B. Over-expression of RCN1 and RSA3 improves the viability of (b) S288C and (c) M2 when plated onto SC -Leu + Phloxine B plates following 16% (v/v) ethanol shock. 3.2.3 RCN1 and RSA3 Over-expression Improves Viability After Extreme Ethanol Shock As a final measure to evaluate the ethanol tolerance of RCN1, MAP2, RSA3, EMI5 and YAT1 over-expression strains, viability following a more extreme ethanol shock was tested. Strains were incubated in 17-21% ethanol for 1 hour following which they were spotted onto SC - Leu media containing no ethanol. Lethal levels of ethanol toxicity were determined to be 21% for S288C and 20% ethanol for M2, as each respective EV strain showed severely reduced viability (Figure 8). Interestingly, although normally considered to be more tolerant to ethanol, the wine 46 strain M2 showed lower tolerance to 20% ethanol compared to S288C. Tolerance in wine yeasts may need to accumulate over an adaptation period of slowly increasing ethanol concentrations to be fully effective. In S288C, over expression of both RCN1 and RSA3 drastically improved viability compared to the vector control at 21% ethanol (Figure 8a) and in M2, over expression of RSA3 slightly improved viability as suggested by the denser population of viable colonies (Figure 8b). Based on the cumulative evidence above, over-expression analysis was continued with RCN1 and RSA3 which increased ethanol resistance of both S288C and M2 when overproduced. Figure 8. Increased dosage of RCN1 or RSA3 improves viability following high ethanol shock in S288C. (a) S288C and (b) M2 over-expression strains were subjected to 20-21% (v/v) and 19- 20% (v/v) ethanol shock, respectively, (in duplicate) for one hour and spotted onto minimal media plates. Viability was assessed following growth for 2 days at 25\u00C2\u00B0C. 3.2.4 Deletion of Candidate Genes Causes Ethanol Sensitivity Although genome wide screens with the S288C deletion collection had not previously identified any of RCN1, MAP2, RSA3, EMI5 or YAT1 null mutations as causing ethanol sensitivity, each null mutant was individually tested for growth on ethanol plates. Serial 47 dilutions of null mutants were spotted onto YPD plates containing 0%, 6%, 8%, and 10% ethanol and incubated for up to 8 days at 25\u00C2\u00B0C in a sealed plastic container to reduce evaporation of ethanol from media. Deletion of RCN1, MAP2 and YAT1 resulted in sensitivity at 10% ethanol compared to wildtype (Mat a) strains although no differences were noted on 8% ethanol (Figure 9) or 6% ethanol plates (data not shown). Deletion of RSA3 and EMI5 did not result in ethanol sensitivity. Figure 9. Deletion of RCN1, MAP2 and YAT1 in S288C causes sensitivity to 10% (v/v) ethanol. Strains with null mutations were serially diluted and spotted onto YPD plates containing 0%, 8% (v/v) and 10% (v/v) ethanol and viability was assessed after growth for 4-7 days at 25\u00C2\u00B0C. Gene disruptions of RCN1, MAP2, EMI5, RSA3, and YAT1 were also constructed in the M2 background to test for sensitivity to ethanol. As outlined above, gene deletion strains were spotted onto 6-10% ethanol plates, however none of the null mutations in M2 conferred an ethanol sensitivity phenotype (data not shown). Deletion of two other genes identified in our genome-wide screens, PHO80 and IMG1, had been previously identified as causing ethanol sensitivity by S288C deletion collection screens (van Voorst, et al., 2006). Because their role in ethanol tolerance in the industrial strain M2 may be of future relevance, confirmation of this 48 phenotype in the M2 background was conducted. Similar to results in S288C, deletion of both genes caused extreme ethanol sensitivity at 6% ethanol when spotted onto YPD (Figure 10). Figure 10. Deletion of IMG1 and PHO80 in the wine yeast M2 causes sensitivity to 6% (v/v) ethanol. M2 mutant strains were serially diluted and spotted onto YPD with 0% and 6% (v/v) ethanol and viability was assessed after growth for 4 days at 25\u00C2\u00B0C. 3.3 RCN1 and RSA3 Expression Influences Cell Wall Assembly When grown in the presence of cell membrane or cell wall perturbing agents S. cerevisiae alters the constituents of its cell wall by increasing membrane proteins, chitin and mannoproteins bound to chitin through \u00CE\u00B2 1,3 and \u00CE\u00B21,6 glucans (de Groot, et al., 2001). Based on the increased viability of ethanol shocked RCN1 and RSA3 strains in the presence Phloxine B, it is possible that the increased viability may be directly related to a stronger cell wall preventing or limiting uptake of the dye. A correlation between ethanol tolerance genes and genes involved in cell wall maintenance has been previously noted, therefore it seemed reasonable that RCN1 and RSA3 could also influence cell wall assembly (Teixiera, et al., 2009). To test this theory, strains of both M2 and S288C over-expressing RCN1 and RSA3 were first challenged with calcuflour white (CFW), a fluorescent stain that interferes with the cell wall by binding chitin, and the membrane perturbing agent sodium dodecyl sulphate (SDS), with no significant results found (data not shown). Although it is possible that RCN1 and RSA3 expression may not effect either chitin 49 distribution or the cell membrane directly, it is equally likely that spot assays are not sensitive enough to detect subtle differences in viability due to changes in cell wall constitution. 3.3.1 RCN1 and RSA3 Decrease Zymolyase Cell Wall Digestion Zymolyase, an enzyme mixture comprised of proteinases and \u00CE\u00B21,3 glucanases, causes digestion of cell wall constituents and subsequently cell lysis. As a third and more sensitive cell wall assay, decrease in O.D.600nm due to lysis was monitored using an automated microplate reader. Semi-log plots of normalized O.D.600nm values were graphed for analysis and the maximum lysis rate (MLR) and lag time prior to lysis were used as parameters for quantifying Zymolyase resistance or sensitivity as described by Ovalle, et al. (1999). A 10% or greater difference in MLR represented a significant difference between strains and was used in conjunction with a Student\u00E2\u0080\u0099s t-test to determine statistical differences. Over-expression of RCN1 and RSA3 in S288C increased the lag time by 5.5 and 4.0 minutes respectively, both of which were statistically different from the vector control (Figure 11a, Table 7). Following the lag time, curves were characterized by a steady decline in optical density. RCN1 over-expression decreased the MLR by over 10% compared to the vector control strain whereas over-expression of RSA3 had no effect (Table 7). These data suggest that high copy expression of RCN1 can improve resistance to zymolyase by influencing cell wall assembly in S288C. Mean lysis curves were also generated using rcn1 and rsa3 deletion strains in the BY4741 (MAT a) background (Figure 11b). Deletion of RCN1 resulted in an increased MLR to 117% of the wildtype MLR, whereas the MLR decreased by about 10% in the rsa3 deletion strain (Table 7). 50 Figure 11. Over-expression of RCN1 in S288C decreases the cell lysis rate during zymolyase digestion. Log O.D.600nm values of S288C strains with (a) the 2\u00CE\u00BC over expression plasmids and (b) gene deletions of vector/WT (blue), rcn1\u00E2\u0088\u0086 (red) and rsa3\u00E2\u0088\u0086 (green) are shown as Zymolyase digestion progresses; 200 \u00CE\u00BCg mL -1 Zymolyase was added to individual wells (in quadruplicate) of a 96 well plate and O.D.600nm was automatically measured every minute for up to 1 hour. -0.12 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0 10 20 30 40 50 L o g ( O .D . 6 0 0 n m /O .D . 6 0 0 n m in it ia l ) Minutes a. S288C+Vector S288C+RCN1 S288C+RSA3 -0.12 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0 10 20 30 40 50 L o g O .D . 6 0 0 n m /O .D . 6 0 0 n m In it ia l Minutes b. S288C WT S288C \u00E2\u0088\u0086rcn1 S288C \u00E2\u0088\u0086rsa3 51 Table 7. Zymolyase cell wall digestion analysis of S288C strains either over-expressing or carrying null mutations of RCN1 and RSA3. Strain + Plasmid Mean Lag Time (Minutes) Mean Lysis Rate (Log 10 4 Cells mL - 1 /minute) % MLR Vector/WT S288C+Vector 23.5 1.78 S288C+ RCN1 29.0 * 1.60 * 89.9% S288C+RSA3 27.5 * 1.73 97.1% BY4741 Wildtype 20.6 3.15 BY4741 \u00E2\u0088\u0086rcn1 20.2 3.70** 117% BY4741 \u00E2\u0088\u0086rsa3 19.4 2.83* 89.8% Significantly different from control at p value *<0.05, **<0.01 Over-expression of RSA3 in M2 significantly decreased the MLR by nearly 18%, from 5.88x10 4 log cells mL -1 /minute (vector control) to 4.83x10 4 log cells mL -1 /minute, whereas RCN1 over expression did not decrease the MLR by a significant margin (Figure 12a, Table 8). The lag time of strains over-expressing RCN1 and RSA3 was again increased, albeit to a lesser extent than in S288C, from 18.4 minutes (vector control) to 19.8 minutes for RCN1 and 20.4 minutes for RSA3 (Table 8). Statistical analysis found that the lag time upon RSA3 over- expression was significantly different from the vector control however the lag time upon RCN1 over-expression was not (Table 8). Mean lysis curves were again generated for deletion strains in the M2 background (Figure 12b) and it was found that null mutants of rcn1 and rsa3 increased the MLR to 123% and 118% of the wildtype value respectively, supporting a role for both genes being required for Zymolyase resistance in M2. Although the MLR of rsa3 mutants increased by greater than 10% compared to the control strain, a Student\u00E2\u0080\u0099s t-test, found that the MLR of the rsa3 strain was not statistically different from the EV control, leaving its significance in question (Table 7). Whereas in S288C increased expression of RCN1 resulted in greater resistance to Zymolyase than with RSA3 over expression, it is interesting that the reverse holds true for the M2 strain, and thus each respective gene appears to influence cell wall assembly. 52 Figure 12. Over-expression and deletion of RSA3 in M2 changes the cell lysis rate during zymolyase digestion. Log O.D.600nm values of M2 strains carrying (a) the 2\u00CE\u00BC over-expression plasmids and (b) gene deletions of vector/WT (blue), rcn1\u00E2\u0088\u0086 (red) and rsa3\u00E2\u0088\u0086 (green) are shown as Zymolyase digestion progresses; 500 \u00CE\u00BCg mL-1 Zymolyase was added to individual wells (in quadruplicate) of a 96 well plate and O.D.600nm was automatically measured every minute for up to 1 hour. -0.2 -0.15 -0.1 -0.05 0 0 10 20 30 40 50 L o g ( O .D . 6 0 0 n m /O .D . 6 0 0 n m in it ia l ) Minutes a. M2+Vector M2+RCN1 M2+RSA3 -0.2 -0.15 -0.1 -0.05 0 0 10 20 30 40 50 60 L o g ( O .D . 6 0 0 n m /O .D . 6 0 0 n m in it ia l ) Minutes b. M2 WT M2 \u00E2\u0088\u0086rcn1 M2 \u00E2\u0088\u0086rsa3 53 Table 8. Zymolyase cell wall digestion analysis of M2 strains either over-expressing or carrying null mutations of RCN1 and RSA3. Strain + Plasmid Mean Lag Time (Minutes) Mean Lysis Rate (Log 10 4 Cells mL - 1 /minute) % MLR Vector M2+Vector 18.4 5.88 M2+RCN1 19.8 5.68 96.6% M2+RSA3 20.4 * 4.83 * 82.1% M2 Wildtype 10.9 2.55 M2 \u00E2\u0088\u0086rcn1 6.9 3.13* 123% M2 \u00E2\u0088\u0086rsa3 4.4* 3.00 118% *Significantly different from control at p value <0.05 3.4 RCN1 and RSA3 Over-expression Improves Environmental Stress Tolerance It has been suggested that genes that respond to thermo (Piper, 1995), osmotic (Yoshikawa, et al., 2009), and oxidative stress (Costa, et al., 1997) also play a role in ethanol tolerance, thus the growth rates of M2 and S288C over-expressing RCN1 and RSA3 were analyzed in the presence of these stressors. A 96 well automated growth chamber was used to monitor the growth of both S288C and M2 over expression strains in SC -Leu media with a range of inhibitory but sub lethal levels of thermo stress (~37 \u00C2\u00B0 C), osmotic stress (~0.4 M NaCl) and oxidative stress (0.75 mM H2O2) according to previously published works (Izawa, et al., 1995, Rep, et al., 1999, Gasch, et al., 2000). To observe a reduction in the growth rate of the more stress tolerant M2 strain, a higher range of heat (40-42 \u00C2\u00B0 C), osmotic (0.7-1.0 M NaCl) and oxidative, (1.0 mM H2O2) stressors were used to challenge the M2 over expression strains. Prior to stress exposure, growth curves of both M2 and S288C with each plasmid were compared to the vector control in non stress SC -Leu media at 25\u00C2\u00B0C with the finding that growth rates and O.D. max did not differ significantly (data not shown). Over all, increased expression of RCN1 and 54 RSA3 did not significantly change the lag phase period during exposure to any of the environmental stressors tested and thus only the growth rate and O.D. max are reported. 3.4.1 RCN1 and RSA3 Over-expression Improves Thermal Tolerance When subjected to heat stress at 37\u00C2\u00B0C, over-expression of both RCN1 and RSA3 in S288C conferred a significantly faster growth rate and a higher O.D.600nm max compared to the vector control strain suggesting that high copies of both genes can improve thermo tolerance in S288C (Figure 13a, Table 9). RCN1 and RSA3 over-expression in M2 during 42\u00C2\u00B0C heat stress resulted in statistically higher growth rates and O.D.600nm max values (Figure 13b, Table 10). Whereas the improvement in growth was more subtle with RCN1 over-expression, RSA3 over-expression caused a drastic improvement of growth under heat stress in the M2 strain. Table 9. Comparison of lag time, growth rate and O.D.Max values during heat stress with RCN1 and RSA3 over-expression in S288C. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) S288C + Vector 4.42 6.94 1.11 S288C + RCN1 4.17 7.19 ** 1.17 ** S288C + RSA3 4.33 7.24 ** 1.24 *** Significantly different at from vector control at p value < *0.05, **0.01, ***0.001 Table 10. Comparison of lag time, growth rate and O.D.Max values during heat stress with RCN1 and RSA3 over-expression in M2. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) M2 + Vector 4.11 2.32 0.53 M2 + RCN1 3.78 3.94 0.66 * M2 + RSA3 3.89 6.74 ** 0.84 *** Significantly different at from vector control at p value < *0.05, **0.01, ***0.001 55 Figure 13. Over-expression of RSA3 improves growth rate at elevated temperatures in S288C and M2. (a) S288C and (b) M2 strains carrying vector (blue), RCN1 (red) and RSA3 (green) 2\u00EF\u0081\u00AD plasmids were grown in minimal media at 37\u00C2\u00B0C (S288C) or at 42\u00C2\u00B0C (M2). O.D.600nm was recorded in an automated 96 well plate reader and mean values for each strain shown for up to 24 hours. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 5 10 15 20 25 O .D . 6 0 0 n m Hours a. S288C+Vector S288C+RCN1 S288C+RSA3 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 2 4 6 8 10 12 O .D . 6 0 0 n m Hours b. M2+Vector M2+RCN1 M2+RSA3 56 3.4.2 RCN1 and RSA3 Over-expression Improves S288C Osmo-tolerance When exposed to osmotic salt stress, over-expression of RCN1 and RSA3 in S288C led to increased growth rates compared to vector control, but did not significantly alter any of the growth curve parameters when either gene was over-expressed in M2 (Figure 14 a,b Table 11,12). The mean values for growth rate and O.D. 600nm max upon RSA3 over-expression in M2 during osmotic stress were notably higher, however high variability in growth in 1.0 M NaCl medium did not allow for statistical significance (Table 12). Growth at lower NaCl concentrations also yielded no significant differences between strains (data not shown). Since the M2 strain has adapted to a high osmolarity environment during wine fermentation, inherent osmo-tolerance is naturally high and thus over-expression of a single gene may have a lesser effect than in S288C. Table 11. Comparison of lag time, growth rate and O.D.Max values during osmotic stress with RCN1 and RSA3 over-expression in S288C. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) S288C + Vector 2.71 5.15 0.663 S288C + RCN1 3.18 6.64 ** 0.754 S288C + RSA3 3.78 7.22 ** 0.727 **Significantly different from vector control at a p value < 0.01 Table 12. Comparison of lag time, growth rate and O.D.Max values during osmotic stress with RCN1 and RSA3 over-expression in M2. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) M2 + Vector 8.50 2.94 0.689 M2 + RCN1 7.93 3.16 0.699 M2 + RSA3 8.05 3.62 0.786 57 Figure 14. Increased dosage of RCN1 and RSA3 improves osmotolerance in S288C but not M2. (a) S288C and (b) M2 strains carrying vector (blue), RCN1 (red) and RSA3 (green) 2\u00CE\u00BC plasmids were grown in minimal media containing 0.4M NaCl (S288C) or 1.0M NaCl (M2) in triplicate. O.D.600nm was recorded in an automated 96 well plate reader and mean values for each strain are shown for up to 35 hours. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 5 10 15 20 25 O .D . 6 0 0 n m Hours a. S288C+Vector S288C+RCN1 S288C+RSA3 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 5 10 15 20 25 30 35 O .D . 6 0 0 n m Hours b. M2+Vector M2+RCN1 M2+RSA3 58 3.4.3 RSA3 Over-expression Improves Oxidative Stress Tolerance The final ethanol related stress tested in this study was sensitivity to oxidative stress by exposing cells to hydrogen peroxide. Over-expression of RCN1 in either S288C or M2 did not significantly improve growth during incubation with hydrogen peroxide, suggesting that RCN1 does not have a significant role in response to oxidative stress (Figure 15 a,b, Table 13,14). To the contrary, over-expression of RSA3 led to a higher O.D.600nm max in S288C and a faster growth rate in M2, thus increasing the dosage of RSA3, but not RCN1, improves tolerance to oxidative stress (Figure 15 a,b, Table 13.14). Table 13. Comparison of lag time, growth rate and O.D.Max values during oxidative stress with RCN1 and RSA3 over-expression in S288C. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) S288C + Vector 5.42 5.11 0.841 S288C + RCN1 5.50 5.28 0.866 S288C + RSA3 5.14 5.93 0.981 * * Significantly different from vector control at a p value < 0.05 Table 14. Comparison of lag time, growth rate and O.D.Max values during oxidative stress with RCN1 and RSA3 over-expression in M2. Strain + Plasmid Mean Lag Time (Hours) Mean Growth Rate (10 5 cells mL -1 /hr) Mean O.D.Max (10 7 cells mL -1 ) M2 + Vector 9.49 7.81 1.29 M2 + RCN1 9.66 8.20 1.28 M2 + RSA3 9.43 9.95 * 1.39 * Significantly different from vector control at a p value < 0.05 59 Figure 15. Over-expression of RSA3 but not RCN1 improves oxidative stress tolerance in S288C and M2. (a) S288C and (b) M2 strains carrying vector (blue), RCN1 (red) and RSA3 (green) 2\u00EF\u0081\u00AD plasmids were grown in minimal media containing 0.75mM H2O2 (S288C) or 1.0mM H2O2 (M2) in triplicate. O.D.600nm was recorded in an automated 96 well plate reader and mean values for each strain are shown for up to 40 hours. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 5 10 15 20 25 O .D . 6 0 0 n m Hours a. S288C+Vector S288C+RCN1 S288C+RSA3 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 10 20 30 40 O .D . 6 0 0 n m Hours b. M2+Vector M2+ RSA3 M2+RCN1 60 3.5 Cycloheximide Inhibition of Translation is Greater in M2 Increasing RSA3 production imparts a growth advantage to S288C and M2 strains under ethanol, heat, osmotic, and oxidative stress, with exception of the M2 strain during osmotic stress. Given that Rsa3p has a role in 60S ribosomal subunit assembly, RSA3 over-expression may improve resistance of cells to a variety of stressors by affecting protein translation (Rosado, et al., 2007). To test this hypothesis, sub-lethal concentrations of cycloheximide, a drug that inhibits translation in S. cerevisiae, were determined for M2 and S288C strains. By analyzing growth of S288C and M2 vector control strains on SC -Leu plates supplemented with 100 and 500 ng/mL (Figure 16a), it was first noticed that the M2 strain background is more sensitive to cycloheximide than S288C at 500 ng/mL. The hyper sensitivity of M2 to cycloheximide was surprising because M2 is an industrial strain and our analysis here demonstrates that the haploid strain of M2 is inherently more tolerant to other stresses than the S288C lab strain. This surprising phenotype was further confirmed by monitoring S288C and M2 growth in an automated plate reader. As the translation stalling effects of cycloheximide were more efficient in liquid media, a concentration range of 25-100 ng/mL cycloheximide was utilized in conjunction with the plate reader assays. Both strains showed a concentration dependent reduction of growth in cycloheximide, however M2 growth was drastically inhibited at 50 ng/mL and 100 ng/mL cycloheximide, whereas S288C was still able to grow (Figure 16b-d). Reduced translation may explain why the M2 wine yeast was earlier found to be more sensitive to 20% ethanol shock for 1 hour when compared to S288C (Figure 8). . 61 Figure 16. The S. cerevisiae wine yeast M2 is more sensitive to translation inhibition than S288C. S288C (purple) and M2 (yellow) vector control strains were grown on SC -Leu plates with (a) 100 or 500 ng/mL cycloheximide or in liquid media with (b) 25 ng/mL, (c) 50 ng/mL, and (d) 100 ng/mL cycloheximide. Strains grown in liquid broth were grown in an automated 96 well plate reader in quadruplicate and mean O.D. 600nm values are shown over 40 hours. 0 0.2 0.4 0.6 0.8 1 1.2 0 20 40 O .D . 6 0 0 n m Hours b. M2 S288C 0 0.2 0.4 0.6 0.8 1 1.2 0 20 40 O .D . 6 0 0 n m Hours c. M2 S288C 0 0.2 0.4 0.6 0.8 1 1.2 0 20 40 O .D . 6 0 0 n m Hours d. M2 S288C 62 3.5.1 Over-expression of RSA3 Does Not Affect Translation Inhibition To determine if RSA3 over-expression influences translation, the growth of both S288C and M2 upon over-expression of RSA3 in cycloheximide was monitored. The growth rate of strains carrying the RSA3 plasmid did not significantly differ from the vector control when grown in a range of 100-150 ng mL -1 cycloheximide for S288C and 25-50 ng mL -1 for M2 (data not shown). Although RSA3 over-expression may increase tolerance to ethanol and other stressful environments by subtly influencing translation rates, high copy expression of a single trans-acting ribosome assembly factor was not enough to conclusively improve growth in the presence of a translation inhibitor. 3.6 Model Wine Fermentations Since increasing RSA3 or RCN1 expression improved the resistance of yeast to a variety of stresses, it seemed plausible that S288C or M2 carrying the RSA3 or RCN1 over-expression plasmids could outperform vector control strains during a wine fermentation. Model wine fermentations were conducted with synthetic wine media (SWM), initially containing 20% equimolar glucose/fructose which potentially yields ~12% ethanol if the fermentation completes. All strains in the S288C strain background struggled to complete fermentation over 21 days leaving over 8% (w/v) residual glucose/fructose whereas fermentations with the M2 strain completed in 17 days with less than 0.3% (w/v) residual sugars. Despite differences in fermentation completion between strain backgrounds, neither the utilization of glucose and fructose (data not shown) nor the production of ethanol (Figure 17) was significantly different between RCN1 or RSA3 over-expression strains and the vector control in S288C or M2. Although not statistically different, RSA3 over-expression led to the highest mean ethanol yield 63 at 6.38% (w/v) compared to 6.28% (w/v) for the vector control in S288C and 12.82% (w/v) ethanol compared to 12.60% (w/v) for the vector control in M2 (Table 15). Therefore it was concluded that SWM with 20% glucose/fructose may have created too much fermentation stress for the S288C strain to elicit a difference in fermentation rate whereas the SWM may not have been a stressful enough environment for the M2 wine yeast. Table 15. Final ethanol production from model wine fermentations with M2 and S288C. Mean Ethanol % (w/v) (n=3) Strain Background % Glucose/ Fructose Vector Control RCN1 RSA3 S288C 20 6.28 5.96 * 6.38 M2 20 12.60 12.67 12.82 M2 32 13.79 13.88 14.22 * *Significantly different from vector control at p<0.05 Because fermentation with M2 is industrially more relevant and S288C lacks the ability to ferment wine efficiently, fermentation analysis was only continued with the M2 strain using a high osmolarity SWM containing 32% glucose/fructose. In addition to imposing high osmolarity stress, if fermented to completion the synthetic wine would yield over 16% ethanol and should pose a formidable level of ethanol toxicity to the haploid M2 strain. After 24 days, the M2 strains were not able to complete fermentation and there was approximately 7% (w/v) residual sugar (Figure 18a). Again, RSA3 over-expression resulted in the highest ethanol yield, with an average of 14.22% (w/v) compared to 13.79% (w/v) for the vector control, a statistically significant difference as determined by a t-test (Figure 18c, Table 15). Despite this, levels of residual glucose and fructose did not differ significantly at any of the fermentation sampling points, nor was ethanol production significantly different at other time points prior to 24 days. Because of these observations, it cannot conclusively be determined if RSA3 over production significantly improves fermentation rate during high alcohol fermentations. It is interesting however that 64 under all fermentation conditions RSA3 over-expression strains yielded the highest mean ethanol yield. Figure 17. Quantification of ethanol from model fermentations conducted with RCN1 and RSA3 over-expression. Model fermentations of synthetic grape juice consisting of 20% glucose: fructose were conducted with (a) S288C and (b) M2 strains carrying RCN1 (red), RSA3 (green) and vector (blue) high copy plasmids. Ethanol was quantified by HPLC from the fermentation supernatant which was periodically withdrawn over a 24 day fermentation period. Error bars are +/- standard deviation. 0 2 4 6 8 10 12 14 0 5 10 15 20 % E th an o l (w /v ) Days Fermentation a. S288C+Vector S288C+RCN1 S288C+RSA3 0 2 4 6 8 10 12 14 0 5 10 15 % E th an o l (w /v ) Day of Fermentation b. M2+Vector M2+RCN1 M2+RSA3 65 Figure 18. Model fermentations conducted with RCN1 and RSA3 over-expression in the M2 wine yeast. Model fermentations of synthetic grape juice consisting of 32% glucose:fructose were conducted with M2 strains carrying RCN1 (red), RSA3 (green) and vector (blue) high copy plasmids. (a) Glucose, (b) fructose (c) ethanol were quantified by HPLC from the fermentation supernatant which was periodically withdrawn over a 24 day fermentation period. Error bars are +/- standard deviation. 0 2 4 6 8 10 12 14 16 18 0 5 10 15 20 % G lu co se ( w /v ) Day of Fermentation a. M2+Vector M2+RCN1 M2+RSA3 0 2 4 6 8 10 12 14 16 0 5 10 15 20 % F ru ct o se ( w /v ) Day of Fermentation b. M2+Vector M2+RCN1 M2+RSA3 66 Figure 18 (Continued): 0 2 4 6 8 10 12 14 0 5 10 15 20 % E th an o l (w /v ) Day of Fermentation c. M2+Vector M2+RCN1 M2+RSA3 67 4 Discussion 4.1 Genome Wide Screen Identifies Five Ethanol Tolerance Genes The first genome wide screen for ethanol tolerance with the 2\u00CE\u00BC MoBY-ORF 2.0 plasmid library has been conducted in the S. cerevisiae wine yeast, M2, as well as in the common laboratory yeast S288C. By setting stringent arbitrary cut offs for quantified barcode values a small subset of genes was identified for follow up analysis. The top five ranked genes in S288C also exceeded the cut off in M2 and thus were likely candidates for conferring ethanol tolerance when over produced on high copy plasmids. Over-expression of two of these genes, RCN1 and RSA3, demonstrated improved tolerance to 16% ethanol across both strain backgrounds. Although over expression of MAP2, YAT1 or EMI5 did not show significantly improved ethanol tolerance in both strains by multiple confirmation assays, their putative roles in ethanol tolerance have not been dismissed. MAP2 and YAT1 over-expression improved growth post ethanol shock in S288C and M2 backgrounds respectively, and thus the relative importance of their molecular function may vary between strains. MAP2 is an aminopeptidase that functions in the cotranslational removal of N-terminal methionine from newly synthesized polypeptides. Therefore, over-expression of MAP2 may improve ethanol tolerance by influencing translation, and may warrant future investigation. YAT1, which transports acetyl-carnitine into the mitochondria for oxidation, and EMI5 which functions with the succinate dehydrogenase complex in succinate oxidation, both feed into the TCA cycle for energy production in the mitochondria (Franken, et al., 2008, Cao, et al., 2010). The original genome-wide screen was conducted with 12% v/v ethanol over 48 hours whereas a more efficient strategy of short term, high 16% ethanol exposure was utilized to confirm the screen which may be why EMI5 and YAT1 were not confirmed to be ethanol tolerant genes. Mitochondrial ATP production may not 68 be severely affected during a three hour ethanol shock whereas 48 hours of ethanol exposure may deplete ATP and over-expression of EMI5 and YAT1 may be more beneficial under these circumstances. Functional analysis of the top 200 ranked genes identified in the M2 screen showed an enrichment for mitochondrial associated genes suggesting that the mitochondria is important for ethanol tolerance in S. cerevisiae. Functional analysis also showed enrichment for genes involved in splicing from both the M2 and S288C screens, and transcription termination genes were also enriched in S288C. Based on the above results, combined with the confirmation of RSA3 and RCN1 mediated ethanol tolerance, four important processes for surviving ethanol stress highlighted by this study are mitochondrial energy production, calcium stress signalling, and transcription and translation processes. 4.2 The Role of Mitochondria in Ethanol Tolerance Mitochondria play an important role in energy production, particularly through oxidative respiration, and in supporting anabolic pathways through tricarboxacylic acid (TCA) cycle generated products. Enrichment of mitochondrial transport genes was noted in the M2 commercial wine yeast ethanol tolerance screen. This study is not the first to suggest the importance of mitochondria in ethanol tolerance however. Regulation of heat shock proteins, which widely respond to both thermo stress and ethanol stress, has been suggested to occur within the mitochondria (Rikhvanov, et al., 2005) and mitochondria associated genes have also been identified by some genome wide deletion studies as required for ethanol tolerance (Kubota, et al., 2004, van Voorst, et al., 2006, Yoshikawa, et al., 2009). Many genes involved in mitochondrial function and oxidative respiration are repressed in the presence of glucose in favour of anaerobic fermentation. Despite this, recent studies have 69 found that during fermentation stress glucose repression is attenuated and mitochondrial genes are up regulated (Mendes-Ferreira, et al., 2007, Marks, et al., 2008). One explanation for the lack of glucose repression during fermentation was attributed to ethanol stress, which may disrupt membrane glucose sensors and inhibit glycolysis, thereby requiring induction of respiratory genes (Marks, et al., 2008) In this study, aerobic conditions combined with high ethanol stress could alleviate repression of mitochondrial genes even in the presence of 2% glucose, and some of these genes could similarly function in high ethanol wine fermentations. 4.2.1 Acetyl-CoA Production in the Mitochondria Recent microarray analysis of two ethanol tolerant strains of S.cerevisiae (generated by adaptive evolution to ethanol) during ethanol stress showed elevated levels of acetyl-CoA production suggesting these strains may increase flux through the TCA cycle as a survival mechanism (Stanley, et al., 2010). Acetyl-CoA is produced in S. cerevisiae either by oxidation of pyruvate, a glycolysis intermediate, which then directly enters the TCA cycle, or by fatty acid \u00CE\u00B2- oxidation which occurs in the peroxisome in yeast. Interestingly, two of the top five genes, EMI5 and YAT1, are both indirectly involved energy production from peroxisomal acetyl-CoA. In one pathway, acetyl-CoA generated in the peroxisome is converted to succinate, which is transported to the mitochondria and further oxidized in the TCA cycle by succinate dehydrogenase (SDH). EMI5 (SDH5) interacts with the succinate dehydrogenase complex adding the cofactor flavin to SDH1, necessary for its function in oxidizing succinate. Alternatively, acetyl-CoA can also be converted to acetylcarnitine, which is transported into the mitochondria via the carnitine shuttle, where YAT1 can catalyze the conversion of acetylcarnitine back into carnitine and acetyl-CoA. 70 Over-expression of either YAT1 or EMI5 may increase efficiency of acetyl-CoA production from fatty acid \u00CE\u00B2-oxidation utilization via alternative pathways that both feed into the TCA cycle, improving survival during ethanol stress. Although genes involved in peroxisome transport and import machinery are required for ethanol tolerance, deletion of genes involved in fatty acid \u00CE\u00B2-oxidation regulation were not found to cause ethanol sensitivity (Teixiera, et al., 2009, Yoshikawa, et al., 2009) suggesting that acetyl-CoA utilization is not absolutely required for tolerance. Peroxisome function may still improve tolerance to ethanol without being absolutely required, especially at concentrations greater than 8%, and over-expression of genes required for utilization of acetyl-CoA may further facilitate this process. 4.2.2 Gut2p, Img1p and Sym1p Function in the Mitochondria Multiple authors have cited that inhibition of glycolysis during ethanol stress further increases expression of glycolytic enzymes, which require NAD+ for a key enzymatic step (Remize, et al., 1999, Stanley, et al., 2010). This can result in an accumulation of NADH during ethanol stress. Restoration of the redox balance through mitochondrial based oxidation of NADH to NAD+ may alleviate this stress, and act to further stimulate glycolysis. One gene hit in M2, GUT2, functions in NADH oxidation in addition to glycerol metabolism, which can generate the glycolysis intermediate dihydroxyacetone. Thus, over-expression of GUT2 may be acting to stimulate glycolysis through either mechanism, or both, to alleviate ethanol stress. In support of this hypothesis, GUT2 was also found to have sustained induction during a wine fermentation in the commercial yeast Vin13 (Marks, et al., 2008). Two other genes, IMG1 and SYM1, function within the mitochondria and were within the initial 20 candidate genes for ethanol tolerance identified in this study. IMG1 has been previously noted as causing ethanol sensitivity when 71 deleted in S288C (van Voorst, et al., 2006) and this phenotype was also confirmed in the M2 strain background (Figure 10). Induction of the mitochondrial located Sym1p during ethanol stress has previously been reported, and likely functions in metabolizing ethanol as a carbon source during prolonged ethanol stress (Trott & Morano, 2004). 4.3 The Role of RCN1 Stress Signalling in Ethanol Tolerance The calcineurin signalling pathway is induced during heat, osmotic and cell wall stress. RCN1 is a regulator of calcineurin, a calcium and calmodulin dependent protein phosphatase that responds to changes in intracellular calcium levels (Kingsbury & Cunningham, 2000). In this study, RCN1 was initially identified in a screen for ethanol tolerance genes. An overlap between calcineurin signalling and osmotic and ethanol stress is not surprising as high ethanol can induce osmotic shock by binding free water which effectively reduces the water activity of the extracellular environment (Hallsworth, 1998). This overlap was confirmed by data showing that RCN1 over-expression can improve Zymolyase resistance (Figure 11), saline osmotic shock (Figure 14) and ethanol tolerance (Figure 6) in S288C. The role of RCN1 in response to these individual stresses was less clearly defined in M2. Recently, it was shown that the Crz1p transcription factor translocates into the nucleus in a calcineurin dependent manner when exposed to 8% ethanol stress (Araki, et al., 2009). In addition, the calcineurin/Crz1 pathway is required for adaptive tolerance to 18% ethanol (Araki, et al., 2009). This study found that over-expression of RCN1 improves recovery of both M2 and S288C strains after a 3 hour 16% ethanol shock which further suggests that the calcineurin pathway is important for adaptation to ethanol stress (Figure 6). Both over-expression and deletion of RCN1 has been found to stimulate calcineurin signalling, therefore a biphasic model 72 of regulation by RCN1 has been proposed (Hilioti, et al., 2004). When phosphorylated by the Mck1p protein kinase, Rcn1p stimulates calcineurin signalling which activates Crz1p and induces transcription of RCN1. In turn, high levels of Rcn1p bind and inhibit calcineurin by preventing the access of Ca 2+ and calmodulin to the calcineurin active site (Hilioti, et al., 2004). Whether over-expression of RCN1 improves stress resistance under the conditions of this study due to inhibition or stimulation of calcineurin signalling remains to be experimentally determined. If RCN1 over-expression improves ethanol tolerance through the calcineurin/Crz1p pathway, it would likely do so in a stimulatory manner. Crz1p induces expression of the calcium ATPases PMR1, PMC1, the sodium ATPase ENA1, and FKS2, a catalytic component of the \u00CE\u00B2 1, 3 glucan synthase complex. Induction of ENA1 and FKS2 could respectively contribute to the NaCl tolerance and Zymolyase resistance upon over-expression of RCN1 (Figures 11 and 14), and could equally play a role in ethanol tolerance. Deletion of PMR1 causes sensitivity to ethanol (Kubota, et al., 2004) suggesting that calcium ATPases may contribute to ethanol tolerance. Induction of many genes that regulate ion homeostasis in the cell has been previously noted (Alexandre, et al., 2001), and thus may be a plausible mechanism for RCN1 over expression contributing to ethanol tolerance. Interestingly, Pho80 another of the 20 candidate genes that confer ethanol tolerance, interacts with its cyclin-dependent kinase partner Pho85p to negatively regulate Crz1p via phosphorylation. Although PHO80 over-expression was not further tested in this study, the ethanol sensitivity of pho80 was confirmed in the M2 background. As Rcn1p and Pho80p are both capable of negative regulation of the calcineurin/Crz1p pathway it is not yet known if ethanol tolerance conferred by over-expression of RCN1, or the unconfirmed PHO80, acts 73 through stimulation or inhibition of Crz1p. It should be noted that RCN1-mediated inhibition of calcineurin prevents repression of the vacuolar H + /Ca 2+ exchanger Vcx1p (Kingsbury & Cunningham, 2000) which could also contribute to restoring ion homeostasis in ethanol stressed cells. Microarray analysis of RCN1 over-expression in the laboratory W303 strain increased expression of 14 ORFs, some of which are heat shock proteins that act as molecular chaperones for protein folding (HSP60, SSA1), protein trafficking (SAC1) or cell wall stability (HSP150) (Hilioti, et al., 2004). Heat shock proteins that repair denatured proteins, in particular the HSP70 family including SSA1, were widely induced during short term ethanol stress (Alexandre, et al., 2001). Other genes induced by RCN1 over-expression act to link glycolysis and the TCA cycle (LPD1, LAT1), and the remaining ORFs have unknown functions or may be dubious (Hilioti, et al., 2004). Microarray analysis of RCN1 over-expression in the M2 wine strain during ethanol stress would thus be an interesting follow up study. 4.4 Transcription, Translation and Ribosome Assembly in Ethanol Tolerance 4.4.1 Transcription Termination and Splicing Influence Ethanol Tolerance Altering gene transcription in S. cerevisiae has long been known as a mechanism to respond to changes in the extracellular environment. In response to ethanol stress, nearly 400 genes have been identified as being induced or repressed (Alexandre, et al., 2001, Chandler, et al., 2004). Ethanol stress however, was also found to cause accumulation of mRNA with hyperadenylation of the 3\u00E2\u0080\u0099 tail, preventing export from the nucleus for translation, and thus inhibiting gene function (Izawa, et al., 2008). Multiple mRNA processing reactions such as capping, splicing and polyadenylation are closely coupled with transcription (Proudfoot, et al., 74 2002). In the S288C screen, highly ranked genes were enriched for both transcription termination and splicing biological processes. Increased expression of genes affecting transcription termination during ethanol stress may facilitate more efficient recycling of RNA polymerase II and expedite nuclear export of mRNA transcripts during ethanol stress. Furthermore, only ~5% of S. cerevisiae genes contain introns and genes containing introns are significantly enriched for ribosomal protein genes - 102 out of the 139 ribosomal protein genes contain introns (reviewed in Meyer and Vilardell, 2009). Therefore, an increase in expression of splicing genes, as was found in both M2 and S288C, may help ethanol tolerance by affecting ribosomal gene expression. The identification of the ribosomal assembly gene RSA3 in both screens further highlights this role of ribosomal associated genes during ethanol stress. 4.4.2 RSA3 and Translation in Ethanol Tolerance In S. cerevisiae, relatively little is known about the molecular function of Rsa3p. RSA3 is a nonessential gene that was first identified in a synthetic lethal interaction with dpb6 (Kressler, et al., 1999, de la Cruz, et al., 2004). Dbp6p is a putative ATP-dependent RNA helicase that is required for assembly of the 60S ribosomal subunit (Kressler, et al., 1998). Likewise, deletion of RSA3 results in a deficit of 60S ribosomal subunits and Dbp6p and Rsa3p co-sediment with 60S ribosomal subunits and interact in vivo (de la Cruz, et al., 2004). Both Dpb6p and Rsa3p have been identified in a low molecular mass complex with three other proteins, Npa1p, Npa2p, and Nop8p, all of which function in early 60S ribosomal biogenesis (Rosado, et al., 2007). In this study however, neither DBP6, NOP8, NPA1 nor NPA2 were highly ranked in either of our screens. RSA3 may be a more critical factor for 60S ribosomal 75 assembly than its molecular complex partners, or high copy expression might result in alternative roles for RSA3 function during ethanol stress. S. cerevisiae contains 18 known ATP dependent dead box protein RNA helicases which promote structural rearrangements between rRNA and proteins (de la Cruz, et al., 2004). Members of the dead box protein super family have widespread roles in assembly of the spliceosome, nuclear mRNA export, folding of self-splicing RNA introns, gene expression control mechanisms, and translation initiation (Jarmoskaite & Russel, 2010). RSA3 may be able to functionally interact with DBP RNA helicases other than Dbp6p when over-expressed, and perform a cellular role not yet fully understood during ethanol stress. A slow growth phenotype at 37\u00C2\u00B0C has been observed in rsa3 null mutants which concurs with this study that over-expression of RSA3 improves growth during thermo-stress (de la Cruz, et al., 2004). This data also suggests that over-expression of RSA3 does not confer a dominant negative phenotype, at least with respect to heat stress, since the rsa3 null phenotype is opposite to the phenotype upon increased RSA3 dosage. Given that rsa3 mutants lack normal levels of 60S ribosomal subunits, one hypothesis is that over-expression of RSA3, a trans-acting ribosome assembly factor, increases ribosome levels and subsequently translation rates. It was surprising to discover that the M2 industrial yeast strain is highly sensitive to the translation stalling agent cycloheximide compared to the S288C lab strain because M2 is more tolerant of other stresses (Figure 16). A significantly improved growth rate was not found when RSA3 was over- expressed in cycloheximide treated cultures of M2 or S288C (data not shown). Therefore over- production of RSA3 cannot prevent translation stalling by cycloheximide but could still have a subtle impact on translation rates. Both an increase and decrease in translation has been observed with gene over-expression. Over production of Stm1p was found to decrease growth by affecting 76 the translation elongation step in S. cerevisiae whereas increased levels of the guanine nucleotide exchange factor Elf-2B have also been cited as a mechanism for increased the rate of translational activity per ribosome (Jedlicka & Panniers, 1991, Van Dyke, et al., 2009). Since increasing RSA3 expression improved the resistance of yeast to a variety of stresses, it was investigated whether S. cerevisiae carrying the RSA3 over-expression plasmid could outperform vector control strains during a wine fermentation. Both S288C and M2 strains were inoculated into a range of synthetic grape musts (20-32% glucose/fructose) a very minor increase in ethanol production was detected in RSA3 overproducing strains compared to strains carrying vector alone (Table 11). Monitoring ethanol production may not strictly reflect the robustness of the yeast, however yeast deletion set fitness profiling studies have found that heterozygous diploid mutants of ribosomal protein genes, including RSA3, actually have a fitness advantage during a 14 day wine fermentation (Piggot, et al., 2011 ). Therefore in the high osmolarity, low pH and gradually increasing ethanol conditions of a wine fermentation, decreasing ribosomal production is beneficial to the cell. Ribosome biosynthesis consumes a majority of the cell\u00E2\u0080\u0099s resources in growing yeast cells and a number of environmental conditions can rapidly induce or repress transcription of ribosomal RNA and ribosomal protein genes (Warner, J.R. TIBS 1999). Therefore, it is not surprising that both induction and repression of ribosomal genes has been reported in response to different stresses. During nitrogen limited wine fermentation, microarray analysis revealed an increase in ribosome biogenesis and assembly genes; a similar response has been observed under cold and high salinity environments (Mendes-Ferreira, et al., 2007). Protein synthesis genes are reported to be down regulated in response to short term ethanol stress (7% ethanol for 30 min) but induced in response to long term ethanol stress (7% ethanol for ~20 hrs) (Alexandre, et al., 77 2001, Li, et al., 2010). Moreover, a S. cerevisiae strain previously adapted to 10% (v/v) ethanol induced expression of genes related to ribosomes and protein synthesis during long term ethanol stress (10% ethanol for ~ 150 hours) whereas its un-adapted parent strain largely down regulated these genes (Dinh, et al., 2009). These results suggest that after initial transcriptome changes to conserve energy, survival during prolonged ethanol exposure requires induction of ribosomal genes to enable growth and metabolic function in ethanol stressed environments. In both 16% (v/v) ethanol (2-3 hours) and 19-21% (v/v) (1 hour) ethanol shock assays, over-expression of RSA3 could have aided the recovery of both S288C and M2 yeast strains by elevating ribosome levels or translational activity prior to or following ethanol exposure allowing for fast repair of damaged cells. Similarly, the decrease in cellular lysis with RSA3 over-expression during Zymolyase digestion in M2 could be a product of increased de novo synthesis of cell wall repair proteins. Finally, the hyper-sensitivity of the haploid M2 wine yeast strain to cycloheximide compared to S288C suggests that S. cerevisiae strains may have optimal growth at inherently different translation rates and that perturbation of ribosome biogenesis and ultimately translation will be an important avenue to explore in the creation of stress-resistant yeast strains. 78 5 Conclusions By using the MoBY-ORF 2.0 library transformed into two strains of S. cerevisiae, a set of 20 genes was identified whose quantified barcode values exceeded the established cut off value from each screen in 12% (v/v) ethanol. Five of these genes, RCN1, MAP2, RSA3, EMI5 and YAT1 were common to both the laboratory strain S288C and the industrial wine yeast M2 and thus proposed to be the most likely to confer superior tolerance to ethanol. Indeed, increased dosage of two of these genes, RCN1 and RSA3, led to superior growth and viability following exposure to 16-21% (v/v) ethanol for a period of 1-3 hours in both S288C and M2. Furthermore, a link between ethanol tolerance and cell wall assembly was suggested by the observation that RCN1 and RSA3 over-expression strains showed improved viability in the presence of Phloxine B, a dye that passively diffuses through cell wall and membrane compromised cells, such as those damaged by ethanol. Confirmation of this hypothesis was determined by using a Zymolyase cell wall digestion assay, in which RCN1 over-expression decreased the cell wall lysis rate in S288C, and RSA3 similarly protected M2 against cell wall degradation. The relationship between ethanol tolerance genes, RCN1 and RSA3 with other common environmental stressors was further tested. Over production of RCN1 improved growth during heat stress in both S288C and M2 compared to control strains, and further improved growth under osmotic stress conditions in S288C only. RSA3 over-expression similarly improved performance under heat, osmotic and oxidative stress in S288C and M2 by increasing either the growth rate or maximal O.D.600nm achieved under these conditions. Because Rsa3p functions in ribosome assembly, the next step was to challenge over-expression strains with the translation stalling agent cycloheximide. Although over-expression of RSA3 did not significantly change the growth rate of either strain under translation inhibition, it was unexpectedly noted that the 79 M2 wine yeast was more sensitive to cycloheximide than S288C. As the regulation of ribosomal associated genes is often transient with changes in environmental conditions, the control mechanisms of ribosome assembly could be an important avenue of research for further optimization of stress tolerant industrial yeasts. Although improved growth under individual environmental stressors was found, these conditions do not often emulate the natural environment, in particular wine fermentation. Under the combined stress of high osmolarity, low pH, and increasingly high ethanol concentrations, neither RCN1 nor RSA3 over-expression in S288C or M2 was able to significantly change the rate of sugar utilization or ethanol production during model fermentations. Further investigation of S. cerevisiae fermentation with increased dosage of RCN1 or RSA3 under a variety of conditions may reveal a more significant change in yeast viability or fitness. . 80 Literature Cited Adams C & Van Vuuren HJJ (2010) Effect of Timing of Diammonium Phosphate Addition to Fermenting Grape Must on the Production of Ethylcarbamate in Wine. Am. J. Enol. Vitic. 61: 125-129. Alexandre H, Ansanay-Galeote V, Deguin S & Blondin B (2001) Global gene expression during short term ethanol stress in Saccharomyces cerevisiae. FEBS Lett 498: 98-103. Araki Y, Wu H, Kitagaki H, Akao T, Takagi H & Shimoi H (2009) Ethanol stress stimulates the Ca2+ mediated calcineurin/Crz1 pathway in Saccharomyces cerevisiae. J. Biosc. Bioeng. 107: 1- 6. Attfield PV (1997) Stress tolerance: the key to effective strains of industrial baker's yeast. Nat Biotechnol 15: 1351-1357. Auesukaree C, Damnernsawad A, Kruatrachue M, Pokethitiyook P, Boonchird C, Kaneko Y & Harashima S (2009) Genome-wide identification of genes involved in tolerance to various environmental stresses in Saccharomyces cerevisiae. J. Appl. Genet. 50: 301-310. Bakalinsky AT & Snow R (1990) The chromosomal constitution of wine strains of Saccharomyces cerevisiae. Yeast 6: 367-382. Barre P, Vezinhet F, Dequin S & Blondin B (1993) Genetic Improvement of Wine Yeast. Wine Microbiology and Biotechnology,(Fleet GH, ed.), p.421-447. Hardwood Academic Publishers. Bauer B & Pretorius IS (2000) Yeast Stress Response and Fermentation Efficiency: how to survive the making of wine - a review. S. Afr. J. Enol. Vitic. 21: 27-51. Bisson LF (1999) Stuck and Sluggish Fermentation. Am. J. Enol. Vitic. 50: 107-119. Borneman AR, Forgan AH, Pretorius IS & Chambers PJ (2008) Comparative genome analysis of a Saccharomyces cerevisiae wine strain. FEMS Yeast Res. 8: 1185-1195. Bradbury JE, Richards KD, Niederer HA, Lee SA, Dunbar PR & Gardner RC (2006) A homozygous diploid subset of commercial wine yeast strains. Antonie Van Leeuwenhoek 89: 27- 37. Butcher RA & Schreiber SL (2006) A microarry-based protocol for monitoring the growth of yeast over-expression strains. Nat. Protocols 1: 569-576. Cao H, Yue M, Li S, Bai X, Zhao X & Du Y (2010) The impact of MIG1 and/or MIG2 disruption on aerobic metabolism of succinate dehydrogenase negative Saccharomyces cerevisiae. Appl Microbiol Biotechnol 89: 733-738. 81 Casey GP & Indgledew WM (1986) Ethanol Tolerance in Yeasts. Crit. Rev. Microbiol. 13: 219- 280. Cavalieri D, McGovern PE, Hartl DL, Mortimer R & Polsinelli M (2003) Evidence for S. cerevisiae Fermentation in Ancient Wine. J. Mol. Evol. 57: s226-232. Chandler M, Stanley GA, Rogers P & Chambers P (2004) A genomic approach to defining the ethanol stress response in the yeast Saccharomyces cerevisiae. Ann. Microbiol. 54: 427-454. Codon AC, Benitez T & Korhola M (1998) Chromosomal polymorphism and adaptation to specific industrial environments of Saccharomyces strains. App. Microbiol. Biotech. 49: 154- 163. Costa V, Amorim MA, Reis E, Quintanilha A & Moradas-Ferreira P (1997) Mitochondrial superoxide dismutase is essential for ethanol tolerance of Saccharomyces cerevisiae in the post- diauxic phase. Microbiol. 143 1649-1656. de Groot PW, Ruiz C, Vazquez de Aldana CR, et al. (2001) A genomic approach for the identification and classification of genes involved in cell wall formation and its regulation in Saccharomyces cerevisiae. Comp. Funct. Genomics. 2: 124-142. de la Cruz J, Lacombe T, Deloche O, Linder P & Kressler D (2004) The putative RNA helicase DBP6 functionally interacts with Rpl3p, Nop8p, and the novel trans-acting factor Rsa3p, during biogeneis of 60s ribosomal subunits in Saccharomyces cerevisiae. Genetics 166: 1687-1689. Dinh TN, Nagahisa K, Yoshikawa K, Hirasawa T, Furusawa C & Shimizu H (2009) Analysis of adaptation to high ethanol concentration in Saccharomyces cerevisiae using DNA microarray. Bioprocess. Biosyst. Eng. 32: 681-688. Du X & Takagi H (2007) N-Acetyltransferase Mpr1 confers ethanol tolerance on Saccharomyces cerevisiae by reducing reactive oxygen species. Appl. Microbiol. Biotechnol. 75: 1343-1351. Duchene E & Schneider C (2005) Grapevine and Climate Changes: a glance at the situation in Alsace. Agron. Sust. Dev. 25: 93-99. Dunn B, Levine RP & Sherlock G (2005) Microarray karyotyping of commercial wine yeast strains reveals shared, as well as unique, genomic signatures. BMC Genomics 6: 53. Eglinton JM, Heinrich AJ, Pollnitz AP, Langridge P, Henschkle PA & Lopes MB (2002) Decreasing acetic acid accumulation by a glycerol overproducing strain of Saccharomyces cerevisiae by deleting ALD6 alcohol dehydrogenase gene. Yeast 19: 295-301. Fleet GH (2008) Wine yeasts for the future. FEMS Yeast Res 8: 979-995. 82 Franken J, Kroppenstedt S, Swiegers JH & Bauer FF (2008) Carnitine and carnitine acetyltransferases in the yeast Saccharomyces cerevisiae: a role for carnitine in stress protection. Curr Genet 53: 347-360. Gasch AP, Spellman PT, Kao CM, et al. (2000) Genomic expression programs in the response of yeast cells to environmental changes. Mol Biol Cell 11: 4241-4257. Giaever G, Chu AM, Ni L, et al. (2002) Functional profiling of the Saccharomyces cerevisiae genome. Nature 418: 387-391. Goffeau A, Barrell BG, Bussey H, et al. (1996) Life with 6000 genes. Science 274: 546, 563- 547. Gonzalez-Ramos D, Quiros M & Gonzalez R (2009) Three Different Targets for Genetic Modification of Wine Yeast Strains Resulting in Improved effectiveness of Bentonite Fining. J. Agr. Food Chem. 57: 8373-8378. Gordon JL, Byrne KP & Wolfe KH (2009) Additions, losses and rearrangements on the evolutionary route from a reconstructed ancestor to the modern Saccharomyces cerevisiae genome. PLoS Genetics 5: e1000485. Guijo S, Mauricio JC, Salmon JM & Ortega JM (1997) Determination of the relative ploidy in different Saccharomyces cerevisiae strains used for fermentation and 'flor' film ageing of dry sherry-type wines. Yeast 13: 101-117. Guthrie C & Fink GR (1991) Guide to Yeast Genetics and Molecular Biology. Academic Press, San Diego. Hallsworth JE (1998) Ethanol-induced water stress in yeast. J.Ferm. Bioeng. 85: 125-137. Hilioti Z, Gallagher DA, Low-Nam ST, et al. (2004) GSK-3 kinases enhance calcineurin signaling by phosphorylation of RCNs. Genes Dev 18: 35-47. Ho CH, Magtanong L, Barker SL, et al. (2009) A molecular barcoded yeast ORF library enables mode-of-action analysis of bioactive compounds. Nat Biotechnol 27: 369-377. Husnik JI, Delaquis PJ, Cliff MA & Van Vuuren HJJ (2007) Functional Analysis of the Malolactic Wine Yeast ML01. Am. J. Enol. Vitic. 58: 42-52. Huxley C, Green ED & Dunham I (1990) Rapid assessment of S. cerevisiae mating type by PCR Trends Genetics 6. Ivorra C, Perez-Ortin JE & Del Olmo M (1999) An inverse correlation between stress resistance and stuck fermentations in wine yeasts. A molecular study. Biotech. Bioeng. 64. Izawa S, Inoue Y & Kimura A (1995) Oxidative stress response in yeast: effect of glutathione on adaptation to hydrogen peroxide stress in Saccharmoyces cerevisiae. FEBS Lett 368: 73-76. 83 Izawa S, Kita T, Ikeda K & Inoue Y (2008) Heat shock and ethanol stress provoke distinctly different responses in 3'-processing and nuclear export of HSP mRNA in Saccharomyces cerevisiae. Biochem. J. 414: 111-119. Jarmoskaite I & Russel R (2010) DEAD-box proteins as RNA helicases and chaperones. WIREs RNA 2: 135-152. Jedlicka P & Panniers R (1991) Mechanism of activation of protein synthesis initiation in mitogen-stimulated T lymphocytes. J Biol Chem 266: 15663-15669. Kingsbury TJ & Cunningham KW (2000) A conserved family of calcineurin regulators. Genes Dev 14: 1595-1604. Kressler D, Linder P & De la Cruz J (1999) Protein trans-Acting Factors Involved in Ribosome Biogenesis in Saccharomyces cerevisiae. Mol. Cell. Biol. 19: 7897-7912. Kressler D, De la Cruz J, Rojo M & Linder P (1998) Dbp6p is an Essential Putative ATP- dependant RNA helicase required for 60S-Ribosomal-Subunit Assembly in Saccharomyces cerevisiae Mol. Cell. Biol. 18: 1855-1865. Kubota S, Takeo I, Kume K, et al. (2004) Effect of ethanol on cell growth of budding yeast: genes that are important for cell growth in the presence of ethanol. Biosci Biotechnol Biochem 68: 968-972. Kumar GR, Goyashiki R, Ramakrishnan V, Karpel JE & Bisson LF (2008) Genes Required for Ethanol Tolerance and Utilization in Saccharomyces cerevisiae. Am. J. Enol. Vitic. 59: 401-411. Lambrechts MG & Pretorius IS (2000) Yeast and its importance to wine aroma-a review. S. Afr. J. Enol. Vitic. 21: 97-129. Leao C & van Uden N (1982) Effect of ethanol and other alkanols on the glucose transport system of Saccharomyces cerevisiae. Biotech. Bioeng. 24: 2601-2604. Leao C & Van Uden N (1984) Effects of ethanol and other alkanols on passive proton influx in the yeast Saccharomyces cerevisiae. Biochimica et Biophysica Acta 774: 43-48. Legras J, Merdinoglu D, Cornuet J & Karst F (2007) Bread, beer and wine: Saccharomyces cerevisiae diversity reflects human history. Molecular Ecology 16: 2091-2102. Li BZ, Cheng JS, Ding MZ & Yuan YJ (2010) Transcriptome analysis of differential responses of diploid and haploid yeast to ethanol stress. Journal of Biotechnology 148: 194-203. Linderholm A, Dietzel K, Hirst M & Bisson LF (2010) Identification of MET10-932 and characterization as an allele reducing hydrogen sulfide formation in wine strains of S. cerevisiae. Applied and Environmental Microbiology 76: 7699-7707. 84 Liti G, Carter DM, Moses AM, Warringer J & Parts ea (2009) Population geneomics of domestic and wild yeasts. Nature 458: 342-345. Longtine MS, McKenzie A, Demarni DJ, et al. (1998) Additional Modules for Versatile and Economical PCR-based Gene Deletion and Modification in Saccharomyces cerevisiae. Yeast 14: 953-961. Magtanong L, Ho CH, Barker SL, et al. (2011) Mapping Genetic Networks by Systematic Dossage Supression. Nature Biotechnology. Marks VD, Ho Sui SJ, Erasmus D, et al. (2008) Dynamics of the yeast transcriptome during wine fermentation reveals a novel fermentation stress response. FEMS Yeast Res 8: 35-52. Martini A (1993) Origin and domestication of the wine yeast Saccharomyces cerevisiae. J. Wine Research 4: 165-176. McGovern PE (2003) Ancient Wine: The Scientific Search for the Origins of Viniculture. Princeton University Press, Princeton, New Jersey. McMeekin TA, Olley J, Ratkowsky DA & Ross T (2002) Predictive microbiology: towards the interface and beyond. Int. J. Food Microbiol. 73: 395-407. Mendes-Ferreira A, Del Olmo M, Garcia-Martinez J, Jimenez-Marti E, Mendes-Faia A, Perez- Ortin JE & Leao C (2007) Transcriptional Response of Saccharomyces cerevisiae to different nitrogen concentrations during alcoholic fermentation. App. Environ. Microbiol. 73: 3049-3060. Mohammed I (2007) Gene expression profile of ethanol-stressed yeast in the presence of acetaldehyde. Thesis, Victoria University, Melbourne. Monk PR (1986) Formation, utilization and excretion of hydrogen sulphide by wine yeast. Australian Wine Industry Journal 1: 10-16. Mortimer RK & Johnston JR (1986) Geneaology of principal strains of the yeast genetic stock center. Genetics 113: 35-43. Naumov GI (1996) Genetic Identification of biological species in Saccharomyces sensu stricto complex. J. Indust. Microbiol. 17: 295-302. Noble AC & Bursick GF (1984) The contribution of glycerol to percieved viscosity and sweetness in white wine. Am. J. Enol. Vitic. 35: 110-112. Novo M, Bigey F, Beyne E, et al. (2009) Eukaryote to eukaryote gene transfer events revealed by the genome sequence of the wine yeast Saccharomyces cerevisiae EC1118. Proceedings of the National Academy of Sciences of the United States of America 106: 16333-18338. 85 Ovalle R, Spencer M, Thiwanont M & Lipke PN (1999) The spheroplast lysis assay for yeast in microtiter plate format. Appl Environ Microbiol 65: 3325-3327. Perez-Ortin JE, Querol A, Puig S & Barrio E (2002) Molecular characterization of a chromosomal rearrangement involved in the adaptive evolution of yeast strains. Genome Res 12: 1533-1539. Pierce SE, Fung EL, Jaramillo DF, Chu AM, Davis RW, Nislow C & Giaever G (2006) A unique and universal molecular barcode array. Nat. Methods 3. Piggot N, Cook M, Tyers M & Measday V (2011 ) Genome-wide fitness profile reveals a requirement for autophagy during yeast fermentation. G3: Genes, Genomes and Genetics (in review). Piper PW (1995) The heat shock and ethanol stress responses of yeast exhibit extensive similarity and functional overlap. FEMS Microbiol Lett 134: 121-127. Pretorius IS (2003) The Genetic Analysis and Tailoring of Wine Yeasts. Functional Genetics of Industrial Yeasts, Vol. 2 (Winde JH, ed.), pp. 99-142. Springer, New York. Pretorius IS & Bauer FF (2002) Meeting the consumer challenge through genetically customized wine-yeast strains. Trends Biotechnol 20: 426-432. Proudfoot NJ, Furger A & Dye MJ (2002) Integrating mRNA processing with transcription. Cell 108: 501-512. Querol A, Fernandez-Espinar MT, del Olmo M & Barrio E (2003) Adaptive evolution of wine yeast. Int J Food Microbiol 86: 3-10. Rankine BC & Bridson DA (1971) Glycerol in Australian wines and factors influencing its formation. Am. J. Enol. Vitic. 22: 6-12. Remize F, Roustan JL, Sablayrolles JM, Barre P & Dequin S (1999) Glycerol overproduction by engieered S. cerevisiae wine yeast strains leads to substantial changes in by-product formation and to a stimulation of fermentation rate in stationary phase. Appl Environ Microbiol 65. Rep M, Reiser V, Gartner U, Thevelein JM, Hohmann S, Ammerer G & Ruis H (1999) Osmotic Stress-Induced Gene Expression in Saccharomyces cerevisiae Requires Msn1p and the Novel Nuclear Factor Hot1p. Molecular and Cellular Biology 19: 5474-5485. Ribereau-Gayon P, Glories Y, Maujean A & Dubourdieu D (2000) Handbook of Enology. John Wiley and Sons, Chichester. Rikhvanov EG, Varakina NN, Rusaleva TM, Rachenko EI, Knorre DA & Voinikov VK (2005) Do mitochondria regulate the heat-shock response in Saccharomyces cerevisiae? Curr Genet 48: 44-59. 86 Robinson MD, Grigull J, Mohammad N & Hughes TR (2002) FunSpec: a web based cluster interpreter for yeast. BMC Bioinformatics 3. Rosado IV, Dez C, Lebaron S, Caizergues-Ferrer M, Henry Y & de la Cruz J (2007) Characterization of Saccharomyces cerevisiae Npa2p (Urb2p) reveals a low molecular mass complex containing Dbp6p, Npa1p (Urb1p), Nop8p, and Rsa3p involved in early steps of 60s ribosomal subunit biogenesis Mol. Cell. Biol. 27: 1207-1221. Rosini G, Federici F & Martini A (1982) Yeast Flora of grape berries during ripening. Microbiology Ecology 8: 83-89. Schacherer J, Shapiro A, Ruderfer D & Kruglyak L (2009) Comprehensive polymorphism survey elucidates population structure of Saccharomyces cerevisiae. Nature 458: 337-341. Schacherer J, Ruderfer D, Gresham D, Dolinski K & Botstein D, et al (2007) Genome wide analysis of nucleotide-level variation in commonly used Saccharomyce cerevisiae strains. PLoS ONE 2: e322. Schuller D & Casal M (2005) The use of genetically modified Saccharomyces cerevisiae strains in the wine industry. Appl Microbiol Biotechnol 68: 292-304. Scott MP, Matsudaria P, Lodish H, et al. (2004) Molecular Cell Biology. WH Freeman and Col., New York. Sniegowski P, Dombrowski PG & 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. Stanley D, Chambers PJ, Stanley GA, Borneman A & Fraser S (2010) Transcriptional changes associated with ethanol tolerance in Saccharomyces cerevisiae. Appl Microbiol Biotechnol 88: 231-239. Swan TM & Watson K (1998) Stress tolerance in a yeast sterol auxotroph: role of ergosterol, heat shock proteins and trehalose. FEMS Microbiol. Lett. 169: 191-197. Swiegers JH, Bartowsky EJ, Henschkle PA & Pretorius IS (2005) Yeast and bacterial modulation of wine aroma and flavour. Austral. J. Grape Wine Res. 11: 139-173. Teixiera MC, Raposo LR, Mira NP, Lourenco AB & Sa-Correia I (2009) Genome-wide Identification of Saccharomyces cerevisiae Genes Required for Maximal Tolerance to Ethanol. App. Environ. Microbiol. 75: 5761-5772. Thomson JM, Gaucher EA, Burgan MF, De Kee DW, Li T, Aris JP & Benner SA (2005) Resurrecting ancestral alcohol dehydrogenases from yeast. Nat Genet 37: 630-635. 87 Tong AH, Evangelista M, Parsons AB, et al. (2001) Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science 294: 2364-2368. Trott A & Morano KA (2004) SYM1 is the Stress-induced Saccharomyces cerevisiae ortholog of the mamallian kidney disease gene Mpv17 and is required for ethanol metabolism and tolerance during heat shock. Eukaryotic Cell 3: 620-631. Van Dyke N, Pickering BF & Van Dyke MW (2009) Stm1p alters the ribosome association of eukaryotic elongation factor 3 and affects translation elongation. Nucleic Acids Res 37: 6116- 6125. van Voorst F, Houghton-Larsen J, Jonson L, Kielland-Brandt MC & Brandt. A (2006) Genome- wide Identification of genes required for growth of Saccharomyces cerevisiae under ethanol stress. Yeast 23: 351-359. Volschenk H, Viljoen M, Grobler J, et al. (1997) Engineering pathways for malate degradation in Saccharomyces cerevisiae. Nat Biotechnol 15: 253-257. Yoshikawa K, Tanaka T, Furusawa C, Nagahisa K, Hirasawa T & Shimizu H (2009) Comprehensive phenotypic analysis for identification of genes affecting growth under ethanol stress in Saccharomyces cerevisiae. FEMS Yeast Res 9: 32-44. You KM, Rosenfield CL & Knipple DC (2003) Ethanol tolerance in yeast Saccharomyces cerevisiae is dependent upon cellular oleic acid content. Appl Environ Microbiol 69: 1499-1503. Zuzuarregui A & del Olmo ML (2004) Expression of stress response genes in wine strains with different fermentative behavior. FEMS Yeast Res 4: 699-710. Zuzuarregui A & del Olmo M (2004) Analyses of stress resistance under laboratory conditions constitute a suitable criterion for wine yeast selection. Antonie Van Leeuwenhoek 85: 271-280. "@en . "Thesis/Dissertation"@en . "2011-11"@en . "10.14288/1.0072280"@en . "eng"@en . "Food Science"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "Attribution-NonCommercial-NoDerivatives 4.0 International"@en . "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en . "Graduate"@en . "Identification of RCN1 and RSA3 as ethanol tolerance genes in the Saccharomyces cerevisiae S288C lab strain and M2 wine strain"@en . "Text"@en . "http://hdl.handle.net/2429/37887"@en .