Flavour Compounds in Fungi

Home , Eduard Buchner, Eremothecium, Eremothecium coryli, Eremothecium gossypii, Muscodor vitigenus, Pichia membranifaciens

FACULTY OF SCIENCE UNIVERSITY OF COPENHAGEN

PhD thesis Davide Ravasio

Flavour compounds in fungi Flavour analysis in ascomycetes and the contribution of the Ehrlich pathway to flavour production in Saccharomyces cerevisiae and Ashbya gossypii

Academic advisor: Prof. Steen Holmberg, Department of Biology, University of Copenhagen.

Co-supervisor: Prof. Jürgen Wendland, Yeast Genetics Group, Carlsberg Laboratory

Submitted: 01/10/14

“There is nothing like looking, if you want to find something. You certainly usually find something, if you look, but it is not always quite the something you were after.”

― J.R.R. Tolkien, The Hobbit

Institutnavn: Natur- og Biovidenskabelige Fakultet

Name of department: Department of Biology

Author: Davide Ravasio

Titel: Flavour-forbindelser i svampe. Flavour-analyse i ascomyceter og bidrag fra Ehrlich biosyntesevejen til smagsproduktion i Saccharomyces cerevisiae og Ashbya gossypii

Title: Flavour compounds in fungi. Flavour analysis in ascomycetes and the contribution of the Ehrlich pathway to flavour production in Saccharomyces cerevisiae and Ashbya gossypii

Academic advisor: Prof. Steen Holmberg, Prof. Jürgen Wendland

Submitted: 01/10/14

Table of contents

Preface ...... 1 List of Papers ...... 2 Summary ...... 3 Resumé ...... 5 Acknowledgments ...... 7 Abbreviations ...... 10 1. Introduction ...... 11 1.1 The Fungi Kingdom ...... 11 1.2 Ascomycota: The Saccharomycotina clade ...... 12 1.3 Saccharomyces cerevisiae: the model organism ...... 15 1.4 Yeast carbon metabolism ...... 18 1.5 The Eremothecium genus: Ashbya gossypii and Eremothecium cymbalariae ...... 20 Ashbya gossypii ...... 21 Eremothecium cymbalariae ...... 23 1.6 Fungal system and their contribution to industrial processes ...... 23 1.7 Aroma and flavour definition, chemical type ...... 24 1.8 Flavour additives and natural flavours ...... 26 1.9 Bioflavour production (Ehrlich pathway, FFAs and lactate) ...... 29 Ehrlich pathway ...... 30 Fatty acids as substrates for flavour formation ...... 33 Metabolism of lactate and citrate ...... 34 1.10 Biological properties of quorum sensing molecules and VOCs as signaling molecules . 35 1.11 Biotechnological application of fungal VOCs ...... 37 1.12 Fungal VOC collection and detection ...... 38 VOC collection ...... 38 VOC separation ...... 39 VOC detection ...... 39 2. Aim ...... 40 3. Objectives and state-of-the-art ...... 41 3.1 Functional analysis of the ARO gene family in S.cerevisiae and A. gossypii ...... 41 3.2 Analysis of the different volatile profiles of A. gossypii and E. cymbalariae is correlated to their genetic backgrounds...... 43

3.3 Flavour molecules produced in the Saccharomyces clade...... 44 4. Discussion ...... 44 4.1 Reporter assay for ARO genes in S. cerevisiae and A. gossypii ...... 44 4.2 Use of a lacZ- reporter assay to correlate reporter gene activity with flavour production ...... 46 4.3 General flavour differences between A. gossypii and E. cymbalariae ...... 47 4.4 VOCs produced in the Saccharomyces clade ...... 49 5. Conclusions and Outlook ...... 53 References ...... 55 Paper 1 ...... 64 Paper 2 ...... 65 Paper 3 ...... 66 Appendix ...... 67

Preface

This thesis “Flavour compounds in fungi: Flavour analysis in ascomycetes and the contribution of the Ehrlich pathway to flavour production in Saccharomyces cerevisiae and Ashbya gossypii” represents an overview of my PhD internship carried out at the Carlsberg Laboratory in Denmark. This project was supervised by Prof. Jürgen Wendland, head of the Yeast Genetics Group in the Carlsberg Laboratory and Prof. Steen Holmberg at the Department of Biology, University of Copenhagen. My thesis was funded by the European Union Marie Curie Initial Training Network, Cornucopia.

List of Papers

I. Davide Ravasio, Andrea Walther, Kajetan Trost, Urska Vrhovsek and Jürgen Wendland (2014). An indirect assay for volatile compound production in yeast strains. Scientific report 4: 3707.

II. Davide Ravasio, Jürgen Wendland and Andrea Walther (2014). Major contribution of the Ehrlich pathway for 2-phenylethanol/rose flavour production in Ashbya gossypii. FEMS Yeast Res. doi: 10.1111/1567-1364.12172.

III. Davide Ravasio, Silvia Carlin, Teun Boekhout, Urska Vrhovsek, Jürgen Wendland and Andrea Walther (2014). A survey of flavor production among non-conventional yeasts. Manuscript.

Summary

Fungi produce a variety of volatile organic compounds (VOCs) during their primary or secondary metabolism and with a wide range of functions. The main focus of this research work has been put on flavour molecules that are produced during fermentation processes, mainly esters and alcohols derived from the catabolism of amino acids. These compounds are produced by the Ehrlich pathway. The conversion of amino acids into aroma alcohols is accomplished by three enzymatic steps: i) a transamination, ii) a decarboxylation and iii) a dehydration reaction. The transaminase and decarboxylase enzymes are encoded by the ARO gene family which represents a widely conserved set of genes in the Saccharomyces clade. Comparative genomic analysis revealed conservation of these genes also in the riboflavin over producer Ashbya gossypii, a closely related species belonging to the Eremothecium clade. ARO80 is a transcription factor that represents the key regulator of the ARO gene family. The first part of the thesis will unveil the ARO80-dependent regulation of the Ehrlich pathway in both Saccharomyces cerevisiae and A. gossypii.

Promoter analyses of the ARO genes in S. cerevisiae showed that the ScARO9 promoter region is directly regulated by the ScAro80 transcription factor. This interaction has been used to create a lacZ-reporter system to correlate the formation of two volatile compounds, 2- phenylethanol and 2-phenylethyl acetate in yeast with ARO9 expression levels. This indirect genetic assay also provides a tool for the prediction of volatile production in other Saccharomyces sensu stricto species. It can be used to screen a large number of strains for their flavour production within a short time and with low costs and effort.

In Ashbya single mutations in the ARO genes led to a strong reduction in volatile production, especially in the amount of isoamyl alcohol and 2-phenylethanol. In contrast, the overexpression of the transcriptional regulator AgARO80 did only increase the level of isoamyl alcohol but did not enhance the 2-phenylethanol yield. Promoter analyses of the ARO genes in A. gossypii identified both ARO8 and ARO10 to be activated by Aro80. In this study we further analyzed the aroma profile of another Eremothecium species, Eremothecium cymbalariae. This species lacks most of the ARO genes involved in amino acid catabolism. The only ARO gene present in E. cymbalariae is a homolog of the A. gossypii ARO8a, which is a non-syntenic homolog of ARO8 in yeast. We compared the VOC profiles of both species in order to investigate the consequences of their different gene set up on their

3 flavour profiles. Here we found that in contrast to A. gossypii E. cymbalariae does not produce 2-phenylethanol and 2-phenylacetate.

The last part of this thesis presents the initial characterization of twenty non-conventional yeasts (NCY) and their potential application in fermentative processes. These strains have been selected as they have been previously isolated from various fermented food sources. This selection of strains was used in fermentations with the aim of identifying new interesting flavour producers. Fermentation profiles, volatile analyses, off-flavour identification and resistance to osmotic/oxidative stress have been addressed to highlight new candidates to use for industrial applications. This resulted in the identification of Wickerhamomyces anomalus and Pichia kluyveri as high producers of esters fruity compounds, which contribute to enhance the complexity of wine and beer product. In addition the strain Debaromyces subglobosus showed high yields of aldehydes and fruity ketones, which constitute active aroma compounds in dry cured ham.

Resumé

Svampe producerer en række flygtige organiske forbindelser (VOC, Volative Organic Compounds) under deres primær eller sekundær metabolisme og med en bred vifte af funktioner. Hovedfokus i dette forskningsarbejde er på flavour molekyler, der produceres under gæringsprocesser, primært estere og alkoholer fra aminosyrers catabolisme. Disse forbindelser dannes under Ehrlich reaktionsvejen. Omdannelsen af aminosyrer til alkoholer opnås ved tre enzymatiske trin: i) transaminering, ii) decarboxylering og iii) dehydreringsreaktion. Transaminase- og decarboxylase enzymer kodes af ARO-genfamilien, som repræsenterer et bredt konserveret sæt af gener i Saccharomyces cladus. Komparativ genomisk analyse viste konservering af disse gener også i riboflavin-overproduceren Ashbya gossypii, en nært beslægtet art, der tilhører Eremothecium cladus. ARO80 er en transkriptionsfaktor, der repræsenterer nøgleregulatoren i ARO-genfamilien. Første del af afhandlingen viser ARO80- afhængig regulering af Ehrlich-biosyntesevejen i både Saccharomyces cerevisiae og A. gossypii.

Promotor analyse af ARO-generne i S. cerevisiae viste, at ScARO9 promotor-regionen er direkte reguleret af ScAro80 transskriptionsfaktoren. Denne interaktion er blevet brugt til at konstruere et lacZ-reporter-system for at korrelere dannelsen af to flygtige forbindelser, 2-phenylethanol og 2-phenylethylacetat i gær med ekspressionsniveauet af ARO9. Dette indirekte genetiske assay er også et redskab til at forudsige produktionen af flygtige forbindelser i andre Saccharomyces sensu stricto arter. Det kan anvendes til på kort tid og lave omkostninger at screene et stort antal stammer for deres flavourproduktion.

I Ashbya, førte enkeltmutationer i ARO gene rned til en kraftig reduktion i produktion af flygtige forbindelser, især i mængden af isoamylalkohol og 2-phenylethanol. I modsætning hertil førte overekspression af den transkriptionelle regulator AgARO80 kun til øget isoamylalkoholdannelse men ikke forøget 2-phenylethanol udbytte. Promotoranalyser af ARO generne i A. gossypii viste at både ARO8 og ARO10 aktiveres af Aro80.

I denne undersøgelse analyserede vi yderligere aromaprofilen i en anden Eremothecium art, Eremothecium cymbalariae. Denne art mangler de fleste af ARO generne, involveret i aminosyre catabolisme. Det eneste ARO-gen, der findes i E. cymbalariae er en homolog af A. gossypii ARO8a, som er et ikke-syntenisk homolog af ARO8 i gær. Vi sammenlignede VOC- profilerne i begge arter for at undersøge konsekvenserne af deres forskellige gen-set up på deres smagsprofiler. Her fandt vi, at A. gossypii i modsætning til E. cymbalariae ikke producerer 2- phenylethanol og 2-phenylacetat.

Den sidste del af denne afhandling præsenterer den indledende karakterisering af tyve ikke- konventionelle gærtyper (NCY, Non-Conventional Yeasts) samt deres potentielle anvendelse i gæringsprocesser. Disse stammer er valgt, fordi de tidligere er blevet isoleret fra forskellige fermenterede fødevarer. Formålet var at identificere nye interessante flavour-producenter. Gæringsprofiler, analyse af volatile forbindelser, off-flvour bestemmelse og modstandsdygtighed mod osmotisk/oxidativ stress er blevet undersøgt for at adressere nye kandidater til brug for industriel anvendelse. Dette resulterede i identifikationen af Wickerhamomyces anomalus og Pichia kluyveri som høj-producenter af frugtagtige esterforbindelser. Disse kan bidrage til at øge kompleksiteten af vin og øl produkter. Desuden viste Debaromyces subglobosus høje udbytter af aldehyder og frugtagtige ketoner, der udgør aktive aromastoffer i tørret skinke.

Acknowledgments

Sometimes what you need it is a chance to prove yourself and have the courage to make it happen. My PhD experience and this thesis would not be possible without all the support I received to carry out this astonishing scientific opportunity.

First, I would like to give my honest and profound gratitude to Prof. Jürgen Wendland to offer me this position. Without his acceptance and thoughtful guidance this PhD would not have been possible. You bet on me even though you could have chosen many other more qualified students. I just want to let you know that your inputs, ideas, supervision and support became important for the success of this PhD project.

I am indebted to my academic supervisor Prof. Steen Holmberg because his supervision during my PhD and kindness to examine this thesis. My deepest thanks to the opponents of my PhD defense which have spent time to read this thesis and evaluate my work.

I am particularly grateful also to Andrea. We have done so many things together than I do not know where to start. You helped me to settled down and get around here in Copenhagen. You taught me almost all the techniques I acquired during my PhD. We shared funny moments together during conferences and extra work activities. Our LONG discussions were extremely useful to improve my skills and delineate well organized experiments. I have so many things to be thankful that one page would not be enough. I guess my biggest satisfaction was to introduce you the GC/MS technique. That time you were the student and I was the teacher. This turned around situation made me understand all the progress I made thanks to you and Jürgen.

The next person I would like to give my gratitude is Prof. Laura Popolo, my previous supervisor at the University of Milan. She trained and offered me a wonderful year experience in her lab during my master thesis. The research work with C. albicans opened my future development, clarified my interests and opened the way to my career decision. You told me about the Cornucopia project and you convinced me to send the application. From that moment on my life changed completely and I will never stop to thank you for encouraging me to take this first step.

Klaus, you are like the German Terminator but with less muscles and HUGE brain. Every time Google did not have an answer for me you had one. You are an unlimited source of useful information. If we connected your brain to the server we could also survive without internet connection. Thanks also for bringing up funny stories and topics during our lunches. You have the attitude to have always the right word to say, turning each situation in a pleasant moment.

Lisa, I can be the Superman in the lab but you are definitely the Supermom in the daily life! You managed to get a great success during your PhD even though you had a husband and a small daughter to take care of. Apparently all the early mornings did not bother you at all. On the contrary you looked always resolute and full of motivation. Keep going with this attitude and I wish you a bright future, wherever that will be!

I have to give my special thanks also to “my girls” Ana and Jevgenia. I have started this PhD path with you two in the office. I will never forget our discussions and the funny moments we shared together. Ana thanks for your IT help. If Neo in Matrix was the chosen one you can be our Neo here at Carlsberg. I will not be surprised if one day I will see a plug behind your skull! Jevgenia, “tusind tak” for supporting my/our research every day. I do not know how many selection plates and YPD media I used that were made by you. You made our daily experience in the lab easier, taking care of the needs of everyone in the group. The word “multitasking” fits perfectly to you! 

Therese, thanks for the good time in the lab as well as being an inspiration for starting my thesis writing. The experience in Asilomar was the greatest conference ever thanks to you, Ana and Andrea. We had a lot of fun and I will never forget the pleasant and joyful atmosphere you helped to create.

Claudia, you do not talk too much but when you show your qualities you are among the best. In our last team building event you brought out your innate talent with the bowling. By keeping a low profile and using a questionable technique you defeated everyone in the group (except Jürgen of course). Thank you too for the good company and the nice discussion we had from time to time.

Klara, you were the last one who joined our group but in these few months you shared your honest beer enthusiasm with us. Thanks for helping us with the organization of the FungiBrain workshop. You have been a very irreplaceable resource to coordinate all the activities and to accommodate the students. I wish you all the best with your PhD project and do not stop to use passion in your work!

Claes, thanks a lot with your help with the Danish summary. With my current Danish skills I could have written maximum 2 lines, full of grammar mistakes most likely.

Big thanks to Dr. Urska Vrhovsek and her team for introducing me to the GC/MS technique and for hosting me during my stay at the Fondazione Edmund Mach, Research and Innovation Centre (Italy). Your exceptional and high qualified support was extremely important for the success of this PhD thesis.

Special thanks also to the entire Carlsberg laboratory for providing a high quality research and exciting work environment. The last three years have been memorable and they will always represent an example of genuine and high scientific research work.

I am also deep indebted with the Cornucopia consortium, with all the PIs and organizers of the network. You have done an incredible work with me and all the other PhD students. You have found the right mixture of good scientific spirit and social activities to make this journey precious. A special gratitude to Prof. Jure Piskur who was the coordinator and main organizer of the EU-project. Your scientific mood will always be with us as well as the funny moments you promoted.

A particular note also to all the PhD (and non) fellows in the Cornucopia consortium. Anna, Jiří, Nerve, Honeylet, Md, Alicia, Amparo, Ida, Raquel, Benjamin, Dorota and Jana

8 you helped to make these three years’ experience brilliant and extraordinary. We shared worries, doubts, happiness, stress, connections, interests, information and much more. Thanks to all of you for making every Cornucopia workshop a pleasant event.

Lucia, ci siamo conosciuti all’inizio di questa esperienza. La tua permanenza doveva essere breve ma il destino ha voluto che restassi e condividessimo preoccupazioni e gioie. A dire il vero abbiamo pure finito con il condividere la lavatrice ma questa è tutta un’altra storia. Grazie di essere sempre stata un’ottima amica nonostante i miei continui scherni. Sappi che sei stata di grande aiuto in tutto, anche se a volte posso non averlo dimostrato come dovevo. Spero vivamente che tutti i tuoi sogni si realizzino e che in un modo o nell’altro tu riesca a capire che non sei proprio “brava” come pensi nelle attività manuali! :-P

Solfa, Andrea, Sara e Conzi che dire... voi siete senz’altro i numeri 1. Mi avete assistito e supportato fin dal principio di questo viaggio. Mi siete stati accanto nelle lunga distanza ma soprattutto nei miei ritorni a casa. Senza di voi restare qui sarebbe stato impossibile. I bei trascorsi passati insieme mi sono stati di conforto negli ultimi anni, aiutandomi a superare momenti “bui” che avrebbero potuto farmi vacillare. Restate come siete, vi voglio bene.

Frank, io e te siamo cresciuti insieme e abbiamo condiviso tante di quelle cose che manco me le ricordo tutte. Comunque per dovere di cronaca io me ne ricordo più di te di sicuro! Ad un certo punto le nostre strade si sono divise...ma è più forte di noi, non esiste che ci perdiamo di vista. Finchè al cinema ci saranno film inutili da vedere noi dobbiamo andarci. Siamo come Frodo e Sam, non esiste che ci separino. Grazie di essere come sei e volerti bene mi sembra quasi scontato dopo tanti anni!

Papà, mamma, sorellina, nonne, Piero e bodolino. Nessuno più di voi ha supportato questa insolita esperienza. Non so neanche da che parte cominciare con i ringraziamenti. Siete le colonne portanti della mia vita, quindi PhD o meno voi avrete sempre il mio amore e la mia gratitudine. Mi siete mancati moltissimo in questi anni e Dio solo sa quanto vi avrei voluto qui accanto a me. Niente di tutto ciò che sono sarebbe stato possibile senza il vostro aiuto. Grazie di T-U-T-T-O!!!!!!

Abbreviations

ADP Adenosine diphosphate ATP Adenosine triphosphate bp Base pair ChIP Chromatin Immunoprecipitation DNA Deoxyribonucleic acid FAD Flavin adenine dinucleotide FFA Free fatty acid FID Flame ionization detector GC-MS Gas chromatography–mass spectrometry kb Kilo base pair NAD Nicotinamide adenine dinucleotide NADP Nicotinamide adenine dinucleotide phosphate NCY Non-conventional yeast PCA Principal component analysis PCR Polymerase chain reaction pFA plasmid for Functional Analysis pKa Logaritmic acid dissociation constant PUFA Polyunsaturated fatty acid RNA Ribonucleic acid SPME Solid-phase microextraction TCA Tricarboxylic acid cycle TCD Thermal conductivity detector TF Trascriptional factor VOC Volatile organic compund X-Gal 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside

1. Introduction

1.1 The Fungi Kingdom Fungi are eukaryotic, heterotrophic organisms, encompassing both single-celled yeasts and multi-cellular filamentous fungi. Their existence is strictly linked to the decomposition of organic material. Many fungal species can survive in oligotrophic environments, scavenging nutrients from the substrate which they colonize [1]. With over 900 million years of evolution the Fungal Kingdom shows an enormous evolutionary and biological diversity [2] (Figure 1). Mycologists have classified fungi into four groups according to their sexual reproduction and molecular structure: chytridiomycetes, zygomycetes, ascomycetes and basidiomycetes. chytridiomycetes occupy a wide range of natural habitats and today represent the most ancient phylum of fungi present on earth [3]. Unlike the others, chytridiomycetes produce zoospores, motile spores propelled by a flagellum which allows them to move towards light or chemical stimuli [4]. Basidiomycetes contain about 30,000 described species and are best known for the production of large fruiting bodies. Most of the fungal body is made of the mycelium, which constitutes the vegetative part of the fungus. The fruiting bodies, the mushrooms, are the results of sexual reproduction and are normally the visible part of the fungus. zygomycetes, which encompass most of the organisms we define as “molds”, typically grow inside their food. They can reproduce sexually and asexually, through a process that is light governed. Their rapid asexual reproduction involves the formation of sporangia and sporangiospores. The presence of sporangia and unseptate hyphae represents a common feature in this class of fungi [5]. The ascomycota is the most studied phylum, and accounts for approximately 75% of all described fungi. Their defining feature is the ascus, a sexual structure in which non-motile ascospores are produced. However, asexual reproduction is the most common form of propagation, responsible for the rapid spread of these fungi into new areas or the conquest of short-lived [6]. The phylum harbors very diverse organisms and includes a number of industrially relevant fungi (i.e. Saccharomyces cerevisiae, Penicillium chrysogenum) but also plant pathogens (i.e. Magnaporthe grisea, Cryphonecrita parasitica, Fusarium oxysporum) and human pathogens (i.e. Candida albicans, Aspergillus fumigatus, Coccidioides immitis). Ascomycetes have gained great importance especially as producers of antibiotics and in food production (bread, beer, wine, cheese, etc.).

Figure 1. Phylogeny of the fungal kingdom. The diamonds represent evolutionary branch points and indicate approximate time points. The colored lines define the major fungal groups [7].

1.2 Ascomycota: The Saccharomycotina clade The ascomycota, or sac fungi, are the largest monophyletic phylum of Fungi, with over 64,000 species [8]. Molecular phylogenetic analyses of nuclear and mitochondrial ribosomal RNA divided them into three main groups: 1) Schizosaccharomyces and Protomyces; 2) ‘filamentous fungi’ (Pezizomycotina); and 3) budding yeast (Saccharomycotina) [7]. This last group comprises yeasts and is home to the most commonly known fungi. Numerous studies have examined the phylogenetic relationships among yeasts of the ‘Saccharomyces complex’ (Figure 2). Multigene sequence analysis resolved 75 species into 14 clades [9].

Figure 2. Phylogenetic tree resolving species of the ‘Saccharomyces complex’ into 14 clades [9].

S. cerevisiae (Clade 1) is part of the Saccharomyces sensu stricto complex, which includes S. paradoxus, S. mikatae, S. cariocanus, S. kudriavzevii, S. arbicola, S. eubayanus and S. uvarum. Among them stable interspecies hybrids are common and widely used in wine, cider, and beer industry. Yeast hybrids share multiple origins, harboring chimeric allopolyploid genomes with different ratios of their parental genomes. The hybridization event may have taken place in the production environments during the domestication of the strain under man- made selection conditions or could occur in nature (Figure 3) [10] [11].

S. cerevisiae S. cerevisiae x S. paradoxus S. kudriavzevii

S. mikatae S. pastorianus S. kudriavzevii S. carlsbergensis S. arboricola

S. eubayanus S. bayanus S. uvarum

Figure 3. Saccharomyces sensu stricto complex and hybrids [12]

The genetic constitution of these hybrids offers the advantage to acquire physiological properties from both parents, generating strains better suited for extreme environmental or production conditions. Yeast breeding and selection processes have been used to improve industrial strains , e.g. by generating strains with novel flavours, cold adapted hybrids with higher fermentation rates or osmotolerance [13] [14]. The first Saccharomyces interspecies hybrid identified was the lager brewing yeast S. pastorianus (S. carlsbergensis). Initially named Unterhefe No. 1 by Emil Chr. Hansen the hybrid has been used in beer production since 1883. The introduction of this hybrid as pure strain in beer industry revolutionized the production process completely, stabilizing and improving fermentation performance [15]. Later, other yeast hybrids have been discovered and characterized. S. cerevisiae × S. kudriavzevii hybrids are widely used in wine and brewing production [16] [17]. S. cerevisiae × S. bayanus yeasts often appear in wine, cider and brewing production. They favor to preserve the S. bayanus-like genome and reduce the S. cerevisiae part [18]. Furthermore, S. cerevisiae × S. paradoxus natural and artificial crosses have been documented [19] [20]. In the natural isolates introgression occurred and rare progeny was found due to post zygotic isolation [21]. In conclusion, Saccharomyces yeasts and their hybrids represent key resources in different industrial processes. Their existence inspired new molecular genetic studies, offering new insight into the evolution of genomic variation and genomic architecture.

This study will focus not only on representatives of clade 1 but also on other clades: clade 13 with Hanseniaspora as representative genus; clade 9 with Torulaspora; clade 7 with Zygosaccharomyces species and clade 12 with Eremothecium species. The Eremothecium genus contains attractive study models and industrially important species, which will be described below in more detail (Paragraph 1.4). Among the Saccharomycotina I will also consider other species such as Candida, Kazachstania, Pichia, Wickerhamomyces and Kluyveromyces, which are not represented in Figure 2.

1.3 Saccharomyces cerevisiae: the model organism Saccharomyces cerevisiae with its intrinsic ability to ferment sugar into carbon dioxide and ethanol has gained a central role in food industry in being a leavening agent for bread and a producer of alcoholic beverages [22]. The use of fermentation, particularly for beverages, has existed since the Neolithic era. The oldest archeological evidence has been documented dating from 7000–6600 B.C. in Jiahu, China (Figure 4) [23]. Even without any deeper knowledge about the process, early human societies could use the combined leavening action of yeasts and bacteria (i.e. sourdough). Around the 19th century bakers had started to use the yeasts from beer brewers by collecting them after wort fermentation [24]. As consequence of that the bread started to have a sweet note, lacking the sourness acquired by co-fermenting with Lactobacilli. In the late 1700 Antoine Lavoisier was the first who tried to describe the fermentation process identifying the raw material necessary for the process: sugar, water, and “ferment” paste [25]. At that time the fermentative process was considered a simple chemical reaction where the “living material” had only a secondary role. In the second half of the 19th century, Louis Pasteur proved that alcoholic fermentation was the result of a microbial process (Figure 4) [25]. In 1857 Pasteur showed that lactic acid fermentation is also caused by living organisms and some years later he defined fermentation as a “form of life without air” (“Pasteur effect”). By the early 1900s still no one could exactly explain the biochemistry of the process until the German chemist Eduard Buechner could ferment a sugar solution by using the first “yeast-extract” ever made. He proved that “cellular machineries” inside microorganisms are able to catalyze all the chemical reactions that occur inside the cell [26]. Enzymatic studies followed and scientists began to analyze and purify the components of cell-free extract. Eduard Buchner detected the active components of the cells extract, coined them as “zymase” in 1897 (Figure 4). Some years later Arthur Harden and William Young could distinguish two fractions of yeast extract: i) A high molecular weight and heat sensitive fraction, which contained mainly enzymes; and ii) a low molecular not heat

15 sensitive fraction, which was enriched in cofactors (i.e. NAD, NADH, NADP,ATP, ADP, etc.). A major breakthrough in enzyme research was then reached after the 1940s when Otto Meyerhof and Luis Leloir resolved details about the glycolytic pathway, showing the complexity of an enzymatic pathways for the first time [27] [28].

Another important milestone in the history of fermentation was the definition of the Crabtree effect by the English biochemist Herbert Grace Crabtree [29]. This effect can be described as the production of ethanol in aerobic condition instead of the production of biomass via the tricarboxylic acid cycle (TCA). This mechanism clearly denies the “Pasteur effect” described above. At high glucose concentrations and in the presence of oxygen, S. cerevisiae and other Crabtree positive yeasts first consume the glucose and then start consuming the by-product ethanol when the glucose is depleted. This mechanism is now known to be controlled by glucose repressing respiratory enzymes (e.g. Mig1), which control sugar metabolism and glucose transport activity [30].

Several authors speculated that the origin of the fermentation mechanism and the relative molecular controls are linked with the origin of modern plants with fruits, at the end of the Cretaceous age, more than 125 million years ago [31]. In this environment the competition for fruit sugars started among the microbial communities. The Crabtree effect could have been a strategic solution for yeast to outcompete bacteria. By sacrificing biomass, the production of ethanol and the increased glucose uptake would have given yeasts an advantage to secure the carbon source and inhibit the environment from microbial competitors [32].

Figure 4. Timeline showing the major discoveries which have contributed to define the fermentation process.

In the 1930s advances in microbiology and biochemistry made it possible to obtain the first mutated yeast by physical and chemical treatment. The easy propagation of yeast cells and manipulation of the genome combined with the simple cellular physiology made S. cerevisiae a model system. Sequencing technologies provided the possibilities to sequence entire genomes, unveiling the comprehensive biological information contained therein. In this genomic era S. cerevisiae was the first eukaryotic species to be completely sequenced [33]. With this knowledge about the genetic material questions about gene regulation, structure-function relationships of proteins and chromosome structure could be answered [34]. Moreover, the finding that 40% of yeast proteins share part of their primary amino acid sequence with the corresponding human protein indicate that genes and functions are widely conserved in certain processes [35]. For instance, human and yeast HMG-CoA reductase proteins are 66% identical. They are responsible for starting the steroid biosynthesis in both species. Even though yeast does not produce cholesterol the yeast proteins can function in the human cholesterol biosynthesis pathway [36].

The yeast receptor Ste2 is involved in the recognition of α-mating factor. A similar protein is found in human called β-adrenergic receptor which functions as a regulator of the blood

17 pressure [36]. Overall many secretary proteins, heat-shock proteins, G-proteins and transcriptional factors of yeast show homologs in the human genome and have similar function in both species. This relationship can be used as a powerful tool to unveil the function of unknown sequences of human and other higher eukaryotic genes [37].

Even though S. cerevisiae has played a central role as a model organism as far back as the early 1900s it is just one representative of the entire yeast community. Most of these species were described in the last 20 years (i.e. sensu stricto species), but the yeast variety is enormous and new species are discovered continuously [38].

Therefore, studying new species represent a unique opportunity to open and create new research directions, intended to understand more about biological processes or to unveil new industrial applications.

1.4 Yeast carbon metabolism A tightly regulated carbon metabolism is a fundamental requirement for in the fermentation process. Yeast can thrive on different carbon sources but glucose and fructose are favored over others [39]. In yeasts the glucose uptake occurs by facilitated diffusion through the plasma membrane by hexose transporters. These plasma membrane proteins belong to the HXT family where each of them exhibits different glucose affinity depending on the glucose concentration. Their expression is tightly regulated by glucose sensor proteins placed in the plasma membrane. Due to these sensors the expression of the optimal hexose transporters is achieved for the respective concentration of glucose available outside the cell [40] [41]. When yeast is grown on carbon sources other than glucose (e.g. maltose) the sugar is imported into the cell by a proton symport mechanism. This transport requires the establishment of an electrochemical gradient of protons which can only be achieved by depletion of internal ATP [42] [43].

The first step of the glycolytic pathway is the conversion of glucose into glucose-6-phosphate by hexokinases (Hxk1/Hxk2) or the glucokinase Glk1 (Figure 5; 1). This step is the first irreversible step of the glycolysis and results in the retention of glucose inside the cell. In particular, Hxk2 is highly expressed during the growth phase when fermentable carbon sources (i.e. glucose, fructose…) are provided. Hk1 and Glk1 on the other hand are de-repressed when a non- fermentable carbon source (i.e. ethanol, glycerol…) is present [39].

After phosphorylation a series of chemical reactions lead to the formation of pyruvate [44]. Once pyruvate is formed it can be fed either into the respiratory or fermentative pathways (Figure 5; 2). In the respiratory pathway, pyruvate enters the mitochondria where it is converted into acetyl CoA by the mitochondrial pyruvate dehydrogenase enzyme complex [45].

The acetyl CoA is then further oxidized in the tricarboxylic acid cycle (TCA cycle) where it generates reducing power (i.e. NADH and FADH) necessary for the final production of ATP (Figure 5; 3). Since TCA intermediates are also used as building blocks to generate biomass, several anaplerotic reactions are available to ensure a proper flux in the TCA cycle. An example of an anaplerotic reaction is the enzymatic conversion of pyruvate into oxaloacetate, which can directly enter the mitochondria [44] (Figure 5; 4).

In fermentative conditions the pyruvate is decarboxylated into acetaldehyde and - depending on the redox status of the cell - it can follow two fates. Acetaldehyde can be reduced to ethanol by the enzyme alcohol dehydrogenase [46] or oxidized to acetate by aldehyde dehydrogenase [47] (Figure 5; 5 and 6). If we look at the energetic yields, respiration is far more efficient. Fermentation generates only a net of 2 ATP per molecule of glucose, whereas respiration produces 38 ATP in S. cerevisiae. On the other hand, fermentation does not utilize an electrochemical gradient to form ATP, which simplifies the overall redox balance of the cell. Moreover to counterbalance this low ATP yield yeasts increase the glycolytic flux by up regulating glucose transporters. This mechanism ensures a rapid conversion to pyruvate and ﬁnal fermentation products, producing enough ATP to allow cell growth [48].

Figure 5. Schematic overview of the carbon metabolic pathways in yeasts.

1.5 The Eremothecium genus: Ashbya gossypii and Eremothecium cymbalariae The Eremothecium genus was instated by Antonino Borzi (1852-1921) with the description of Eremothecium cymbalariae in 1888 [49]. The genus has been assigned to clade 12 (Figure 2) of the ‘Saccharomyces complex’ [9] and shows interspecific divergence [50]. It contains seven members: Eremothecium gossypii (better known as Ashbya gossypii; [51]; Eremothecium cymbalariae [49]; Nematospora/Eremothecium coryli [52] [53]; Nematospora/Holleya/Erem othecium sinecaudum [52], [54]; Eremothecium ashbyi [55]; Ashbya aceri [56].

Ashbya gossypii Ashbya gossypii is a filamentous fungus, originally isolated from cotton as a pathogen causing stigmatomycosis. The fungus uses insects of the Pyrrohocoridae family as vectors [51]. With today’s use of insecticides fungal infections via insect vectors are less problematic. A. gossypii is mainly used in industry for the production of riboflavin (vitamin B2) [57] [58]. Several studies have elucidated the molecular pathways involved in riboflavin production [59] and aimed at improving the riboflavin yield by utilizing different substrates [60] [61] [62] [63]. The life cycle of A. gossypii starts with the germination of needle-shaped spore and the formation of a spherical germ cell (Figure 6). This stage is characterized by DNA replication, accompanied by an isotropic growth phase. After a switch from isotropic to polar growth a hyphal tip is formed that grows to a juvenile mycelium by lateral branching. The hyphae are compartmentalized into multinucleated cells whose borders are separated by chitin-rich septa. About 20-24 hours post germination hyphal maturation results in the formation of a mature mycelium, which is characterized by an accelerated growth speed and dichotomous branches at the hyphal tips. Upon nutrient limitation, initiated in older segments of a mycelium, hyphae generate sporangia which contain needle shaped spores. The life cycle ends with the formation of 26-30 µm long spores, clumping together via terminal filaments [57].Due to the biotechnological interest in riboflavin production molecular analyses were already started in the 1990s. Molecular studies revealed the identification of an efficient homologous recombination system in A. gossypii [64]. Based on that, efficient gene targeting methods for molecular manipulation have been successfully developed by the Philippsen group [65]. In 2004 the A. gossypii complete genome sequence was published. Its annotation revealed a large degree of synteny with the genome of the close relative S. cerevisiae [66]. About 95% of all the 4718 protein-coding genes in A. gossypii have a homologue in yeast, indicating that the basic cellular functions and morphogenetic machineries are well conserved [57]. However, the Eremothecium and Saccharomyces genera diverged more than 100 million years ago before the whole genome duplication occurred in S. cerevisiae. This duplication event created about 5000 twin ORFs in the duplicated S. cerevisiae genome, raising the possibility to diverge the function of the duplicates. A. gossypii with its 8.8 million base pairs evolved to a very compact genome caused by the reduction of intergenic regions [50].

Figure 6. A. gossypii life cycle. Isotropic growth phase of the germinated spore (a) is followed by the generation of the first germ tube (b). The formation of the second germ tube is opposite to the first (c). The growing hyphae start the lateral branching pattern (d) before they mature and switch from lateral to dichotomous tip branching with an increased growth speed (from 10 µm/h to 200 µm/h) (e). In older hyphae sporangia formation and hyphal fragmentation occur, and sporulation and breakdown of the ascus walls takes place (f) [57]. The picture in the background of the life cycle shows a juvenile mycelium of A. gossypii stained with Calcofluor white. Septal sites and hyphal cell walls contain chitin and are, therefore, highlighted with the dye.

Additionally, it could be shown that the average length of introns (107 bp) is less than half of the average intron size in S. cerevisiae (244 bp) [56]. The ease of genetic manipulation combined with the available genome sequence and industrial interest of the strain made A. gossypii an attractive model for studying the formation and maintenance of filamentous growth and a model for comparative genomics and evolutionary studies [57].

Eremothecium cymbalariae Eremothecium cymbalariae is a close relative of A. gossypii, which was recently sequenced and annotated. Its genome consists of 9,7 million bp encoding for 4712 [50].Both fungi share about 97%of homologous genes. Nevertheless, there are significant differences in their genomes. A. gossypii has only 7 chromosomes whereas E. cymbalariae has 8. The high GC content of 52% in A. gossypii was not found in E. cymbalariae. With 40% GC content it is more similar to that of S. cerevisiae (38.3%) [50]. As for A. gossypii, Eremothecium cymbalariae is a homothallic fungus. [67] In contrast to A. gossypii the spores are produced in aerial sporangia. Additionally, E. cymbalariae is not over producing riboflavin like A. gossypii [50] (Figure 7).

Figure 7. Juvenile mycelium of A. gossypii (A) and E. cymbalariae (B). In the upper right corner are displayed the different spores shape of A. gossypii and E. cymbalariae.

1.6 Fungal system and their contribution to industrial processes In the history of mankind fungi have been involved in many processes such as the production of fermented food and drinks. The use of fungi in spontaneous fermentations led to an increase in the shelf life of the food products and to the production of fermented beverages. The consumption of alcoholic drinks may have started as a religious procedure but became very popular and was a way of avoiding water-borne diseases [68]. Later, biotechnological advances provided new ways to use fungi as cell factories [69]. Discoveries and the development of molecular biology provided new insights into biological pathways offering wide applications in the food, chemical, cosmetic and pharmaceutical sector [70]. Nowadays, biotechnological applications are generally preferred over chemical processes. Among others the main reasons are that chemical syntheses often involve environmentally unfriendly production processes coupled with an undesirable racemic mixture of compounds [71]. Advantages of bio-production

23 are the reduced material and energy consumption, increased use of renewable and biodegradable materials, reduced waste generation and the production of more environmentally friendly stereospecific products [72]. Table 1 contains a list of some common industrial application of fungi.

1.7 Aroma and flavour definition, chemical type The word “aroma” derives from Latin and was coined in the 13 century to define “sweet odors”. Also the Greek word “ἄρωμα“describes any “seasonal spice or sweet herb”. “Flavour” on the other hand, derived from the vulgar Latin word flator , literally "that who blows”, and the old French word flaour "smell, odor" (http://www.etymonline.com/). Today flavours are defined as a complex combination of the olfactory, gustatory and trigeminal sensations perceived during tasting (according to the International Organization for Standardization). Nevertheless, the term "flavour" is used to describe different things. Sensory experts use flavour to describe the combination of taste and smell. Flavour chemists typically mean a single aroma, while chefs tend to use it not only on the food per se, but also for its overall presentation on the plate. In many cases the flavouring agents have a defined molecular structure, making food attractive and eating or drinking a pleasure. Flavour substances can be either volatile or non-volatile. The latter consists mainly of carbohydrates, amino acids and fruit acids, and includes taste substances only. The volatile part includes both taste and odor substances, and contains alcohols, esters, acids, aldehydes, amines and sulfur and nitrogen compounds (Table 2). When we compare the relative sensitivity of our senses we realize the big difference between our smell and taste sensitivities. The sense of smell is approximately 10.000 fold more sensitive than our sense of taste [73]. Furthermore, without an appropriate volatile bouquet our food would taste very alike. Hence studying the vast variety of volatile constituents of food and fermented products is of great importance.

Application List of compounds Organism Description Alcohols, esters, acids, Saccharomyces sp., Lactobacillus sp., Bread, beer, wine, cheese are just a sample of the fermented Food, drinks and fodder pyrazines, aldehydes and Brettanomyces sp., Zygosaccharomyces products produced by yeast fermentation. ketones sp., Hanseniospora sp. and Candida sp. Ashbya gossypii , Eremothecium ashbyi, B group vitamins Brewer’s yeast

Antibiotics: Cephalosporin C, Penicillium notatum, Cephalosporium Penicillin, Griseofulvin sp., Penicillium griseofulvum

Fungal secondary metabolites are extremely important to our health and nutrition. More and more metabolites are produced Immuno- suppressives: Tolypocladium inflatum by fungal species, some of them modified and optimized by Metabolic products Cyclosporin A genetic engineering (for istance organic acids, vitamins, antibiotics, steroids). In this category we also find production of Statins: Lovastatin, Mevastatin Aspergillus terreus, Penicillium citrinum enzymes and recombinant polypeptides (insulin, hirudin, statins, taxol) . Polyunsaturated fatty acids Morteriella isabellina, Mucor (PUFAs) circinelloides

Antitumor: Taxol Taxomyces andreanae

Pigments Blakeslea trispora, Monascus purpureus

Mucor sp., Paecilomyces sp., Penicillium Metal-containing ores are percolated with the microrganism. The Citrate, Succinate, Oxalate, Microbial leaching sp., Trichoderma sp., Cladosporium sp., metal is then solubilized by the fungal metabolism and then Oxalacetate Alternaria and Apergillus sp. extracted or precipitated.

In oder to reduce the need for chemical compounds is now Naphthalene, Citronellal, Phoma sp., Muscodor sp., Rhizopus sp., Biological control of pathogen possible to prepare a mixture of fungal species to control insect Nitrosoamide Fusarium sp. pests. Apergillus sp., Pestalotiopsis sp., Mucor Gentically modied fungi have been shown to be able to degrade Environment protection Depolymerases sp. plastic materials and highly toxic hydrocarbon species

Table 1. Sample list of the more common biotechnological application of fungi.

1.8 Flavour additives and natural flavours Adding substances to food for conservation, flavouring, or appearance is not a modern-day invention [74]. Before the first refrigeration system was invented, salts were used to preserve meats and fish. Sugar was added to preserve fruits and cloves were placed in hams to inhibit the growth of bacteria [75]. Ancient cultures added sulfites to preserve wine and spices and colorings were used to enhance flavours of foods [74]. Today, there are thousands of food additives found in foods.

Functional group Source Example Smell Alcohol Geraniol, linalool Fresh, floral -OH Plants Menthol Mint Aroma-active alcohols > c3 Sweet or pungent Aldehyde; ketone Fat Diacetyl Like butter -CHO; >C=O Milk products Acid (C1-C12) Formic acid Pungent Cheese -COOH Capric acid Like goats’ milk Solvent (these chemicals Ester, lactone Ethyl acetate Glue are used as solvents) -COOR Fruit Methyl/ethyl butyrate Pineapple Amyl/butyl acetate Banana Pentyl butyrate Apricot 2-isobutyl-3- Pyrazine Earth, spice, green pepper Roasted, cooked, methoxypyrazine aromatic =N- fermented foods 2-acetyl-tetrahydro-pyridine Popcorn Diallyldisulphide Garlic S-compounds: aliphatic, Vegetables 1,2-dithiolane-4-carboxylic aromatic Asparagus acid Guaiacol Wood smoke Phenols (mono-, poly-) Smoked food Cresol Tarry

Table 2. List of the most common volatile compounds and relative perception descriptors (www.scienceinschool.org).

The increasing demand for high quality food products made a radical change in food production and technology necessary. Most of the global market is now dominated by processed food which requires food additives to reestablish flavour and taste characteristics. A food additive is defined as “any substance the intended use of which results (…), in its becoming a component or otherwise affecting the characteristics of any food” [76]. Direct food additives are added

26 intentionally to a food for a specific purpose like the sweetener aspartame, sugar or salt which are used to give taste and texture to the products. Indirect food additives are those substances which become part of the food product during the packaging, storage and handling [77]. Direct food additives have different functions: they improve the texture appearance, keep freshness, increase the nutritional contents and improve flavour and taste of the products [78] (Table 3). Due to the increasing number of food additives, the European Economic Community (EEC) established code names to identify food additives. All food additives used in the European Union are identified by an E-number and each food additive is assigned a unique number. The numbering scheme follows precise rules which place the food additive in the respective category (for more details look at the European commission website: http://ec.europa.eu/food/food/fAE F/additives/lists_authorised_fA_en.htm). Typical flavour additives are fruit flavours, sweeteners, cheesy flavours, buttery or roasted flavours, dosed in either natural or artificial concentrations. The difference between them is not at the chemical level. The natural process and the chemical synthesis lead to the same chemical compound. The difference between natural and chemical flavours lies in the source of the flavour. The so called “natural” flavours” are substance obtained by appropriate physical, enzymatic or microbiological processes from material of vegetable, animal or microbiological origin [76]. Food manufacturers often use natural flavours simply because the term "natural" appears healthier to consumers. The result of this trend was a remarkable increase in natural flavour production, reaching a market of $3.5 billion in 2011 and a predicted market of $5 billion in 2017 (http://www.marketsandmarkets.com).

Types of food Function Examples Names Found on Product Labels additivies Fruit sauces and jellies, beverages, Ascorbic acid, citric acid, sodium benzoate, calcium propionate, Prevent food spoilage from bacteria, molds, fungi, or baked goods, cured meats, oils and Preservative sodium erythorbate, sodium nitrite, calcium sorbate, potassium yeast margarines, cereals, dressings, sorbate, BHA, BHT, EDTA, tocopherols (Vitamin E) snack foods, fruits and vegetables Beverages, baked goods, Sucrose (sugar), glucose, fructose, sorbitol, mannitol, corn syrup, high Sweeteners Add sweetness with or without the extra calories confections, table-top sugar, fructose corn syrup, saccharin, aspartame, sucralose substitutes, many processed foods Offset color loss due to exposure to light, air, FD&C Blue Nos. 1 and 2, FD&C Green No. 3, FD&C Red Nos. 3 and 40, Many processed foods, (candies, temperature extremes, moisture and storage FD&C Yellow Nos. 5 and 6, Orange B, Citrus Red No. 2, annatto Color Additives snack foods margarine, cheese, soft conditions; correct and enhance natural variations in extract, beta-carotene, grape skin extract, cochineal extract or carmine, drinks)) color paprika oleoresin, caramel color, fruit and vegetable juices, saffron Pudding and pie fillings, gelatin dessert mixes, cake mixes, salad Flavours and Spices Add specific flavours (natural and synthetic) Natural flavouring, artificial flavour, and spices dressings, candies, soft drinks, ice cream Enhance flavours already present in foods (without Monosodium glutamate (MSG), hydrolyzed soy protein, autolyzed Flavour Enhancers Many processed foods providing their own separate flavour) yeast extract, disodium guanylate or inosinate Baked goods, dressings, frozen Olestra, cellulose gel, carrageenan, polydextrose, modified food starch, Provide expected texture and a creamy "mouth-feel" in Fat Replacers desserts, confections, cake and microparticulated egg white protein, guar gum, xanthan gum, whey reduced-fat foods dessert mixes, dairy products protein concentrate Replace vitamins and minerals lost in processing Flour, breads, cereals, rice, Thiamine hydrochloride, riboflavin (Vitamin B2), niacin, niacin amide, Nutrients (enrichment), add nutrients that may be lacking in the macaroni, margarine, salt, milk, folate or folic acid, beta carotene, potassium iodide, iron or ferrous diet (fortification) fruit beverages, energy bars, sulfate Salad dressings, peanut butter, Soy lecithin, mono- and diglycerides, egg yolks, polysorbates, sorbitan Emulsifiers Allow smooth mixing of ingredients, prevent separation chocolate, margarine, frozen monostearate desserts Frozen desserts, dairy products, Stabilizers and Produce uniform texture, improve "mouth-feel" cakes, pudding and gelatin mixes, Gelatin, pectin, guar gum, carrageenan, xanthan gum, whey Thickeners dressings, jams and jellies, sauces Beverages, frozen desserts, pH control agents and Control acidity and alkalinity, prevent spoilage chocolate, low acid canned foods, Lactic acid, citric acid, ammonium hydroxide, sodium carbonate acidulants baking powder Leavening Agents Promote rising of baked goods Breads and other baked goods Baking soda, monocalcium phosphate, calcium carbonate Keep powdered foods free-flowing, prevent moisture Salt, baking powder, confectioner's Anti-caking agents Calcium silicate, iron ammonium citrate, silicon dioxide absorption sugar Shredded coconut, marshmallows, Humectants Retain moisture Glycerin, sorbitol soft candies, confections Yeast Nutrients Promote growth of yeast Breads and other baked goods Calcium sulfate, ammonium phosphate Dough strengtheners Produce more stable dough Breads and other baked goods Ammonium sulfate, azodicarbonamide, L-cysteine and conditioners Firming agents Maintain crispness and firmness Processed fruits and vegetables Calcium chloride, calcium lactate Enzyme preparations Modify proteins, polysaccharides and fats Cheese, dairy products, meat Enzymes, lactase, papain, rennet, chymosin Oil cooking spray, whipped cream, Carbon dioxide, nitrous oxide Gases Serve as propellant, aerate, or create carbonation carbonated beverages

Table 3. Summary list of the most common food ingredients (http://www.fda.gov)

1.9 Bioflavour production (Ehrlich pathway, FFAs and lactate) Up to date, plants and animals represent the main source of bioflavours. Nevertheless, these compounds are often produced in low concentration, making extraction, purification and formulation too expensive for the market. Therefore, microbial processes represent a promising frontier, offering de novo flavour synthesis or bioconversion of natural precursors [79]. The ability to produce flavour compounds can be applied in two different ways: i) in situ flavour generation, where the compounds become an integral part of food or beverage (i.e. cheese, yogurt, beer, wine) and ii) specifically designed processes where the obtained aroma compounds are isolated and used later as ‘natural’ additives in food production [80]. Several fungi are able to produce various mixtures of so-called gas-phase or generally named volatile organic compounds (VOCs). Diverse definitions of the term VOCs are in use, depending on laws and government agencies. Biologically generated VOCs can be defined as any organic chemical that has a high vapor pressure at room temperature. Fungal VOCs are mainly derived from secondary metabolism pathways. Due to their small sizes they can be very difficult to detect. Furthermore, VOC profiles are species or strain dependent and can vary according to different parameters such as temperature, micronutrients, oxygen, redox status of the cell, pressure, incubation time, carbon and nitrogen sources and others environmental factors [81]. At the beginning of the 20th century, researchers clarified some of the major biochemical pathways involved in flavour formation and unveiled relationships between a specific microorganism and a desirable mix of flavours [79]. Despite that, flavour yields are often quite low and the limited knowledge about the biochemical pathways and the regulation of key enzymes still represents a bottleneck for novel technologies.

In the following paragraphs I provide a brief description of the three main molecular pathways involved in the flavour synthesis in fungi (Figure 8), with emphasis on the catabolism of amino acids and related events in yeast.

Figure 8. Schematic representation of the three main pathways responsible for the flavour formation in yeast (modified from [82]).

Ehrlich pathway In 1907 the German biochemist Felix Ehrlich (1877-1942) proposed that amino acids were split by a “hydrating” enzyme activity to form the corresponding fusel alcohols, along with carbon dioxide and ammonia [83]. Following work by Lampitt [84], Yamada [85], and [86] [87] could show that the higher alcohols produced by yeast are derived from amino acid catabolism. This pathway, which later was named after its discoverer, Ehrlich, utilizes valine, leucine, isoleucine, methionine, phenylalanine, tyrosine and tryptophan to generate higher volatiles with distinctive flavours (Table 4).

Amino a-Keto acid Fusel aldehyde Fusel alcohol Fusel acetate acid 4-Methyl-2-oxo- Leu 3-Methylbutanal Isoamyl alcohol Isoamyl acetate pentanoate 3-Methyl-2-oxo- Val 2-Methylpropanal Isobutanol Isobutyl acetate butanoate 3-Methyl-2-oxo- Ile 2-Methylbutanal 2-Methylbutanol Ethyl pentanoate pentanoate 3-Phenyl-2-oxo- Phe 2-Phenylethanal 2-Phenyl ethanol 2-Phenylacetate propanoate 2-(4- 2-(4- 2-(4- 3-(4-Hydroxyphenyl Tyr Hydroxyphenyl) Hydroxyphenyl) Hydroxyphenyl) 2-oxopropanoate ethanal ethanol ethanoate 3-(Indol-3-yl)-2- 2-(Indol-3-yl)- 2-(Indol-3-yl) Trp Tryptophol oxopropanoate ethanal ethanoate 4-Methylthio-2- 3-(Methylthio) Met Methional Methionol oxobutanoate propanoate

Table 4: The Ehrlich pathway intermediates (modified from [88]).

The catabolism of the amino acids is characterized by three enzymatic steps: including a transaminase, a decarboxylase and an aldehyde dehydrogenase reaction (Figure 9). Although this pathway has been discovered a century ago little is known about the metabolic and genetic regulation of each step. Furthermore, it is still unclear why these secondary metabolites are produced since yeast could use the amino acids for its protein synthesis instead. Several publications consider the higher volatile production as a way to remove toxic compounds from the cell, to help maintaining the NADH/NAD+ ratio and redox balance [89] or to balance the cells nitrogen sources [90] [91]. Furthermore, volatiles are also known to be important signal molecules used as attractants to insect which are then used as vectors for dispersal [92].

Figure 9: The Ehrlich pathway. The three enzymatic reactions responsible for converting aromatic and branched-chain amino acids into aromatic compounds (modified from [88]). Respective genes of S. cerevisiae are listed aside for each enzymatic reaction.

Increasing demand for natural flavour compounds, such as 2-phenylethanol, isoamyl alcohol and their respective esters has led to a new interest in research to investigate this pathway in order to control the flavour production in fermented products.

The first transamination step of the catabolic degradation of amino acids via the Ehrlich pathway is controlled by four genes, namely BAT1, BAT2, ARO8 and ARO9. BAT1 and BAT2 encode mitochondrial and cytosolic amino acid aminotransferases, respectively. The expression of the mitochondrial enzyme Bat1 is higher in the exponential phase of growth whereas the cytosolic Bat2 is preferably expressed during stationary phase [93]. Recent studies demonstrated the fundamental importance of BAT2 during flavour production. Overexpression and deletion of this gene resulted in a drastic increase or reduction of the level of higher volatile compounds, respectively [94]. The two additional aminotransferases ARO8 and ARO9 are not specifically expressed in distinct growth phases only. However, little is known about the regulation of ARO8 or ARO9 [88]. The second reaction of the Ehrlich pathway is a decarboxylation step. This step can be performed by five thiamine diphosphate (TPP)- dependent decarboxylases, namely Pdc1, Pdc5, Pdc6, Aro10, and Thi3. Presumably, each

32 decarboxylase has an α-keto acid specificity but to date there are no convincing results available. Only Aro10 seems to have a broad activity towards α-keto acids [88].

The third step of the Ehrlich pathway is a reduction/oxidation. This step is strictly dependent on the growth condition. Under anaerobic condition yeast cells have excessive production of NADH. The Ehrlich pathway may then guarantee an efficient NAD(P)+ regeneration. The cellular redox status is most likely the major factor that controls the ratio of fusel acid or fusel alcohol production. In the S. cerevisae genome six aldehyde dehydrogenase and 16 alcohol dehydrogenase genes are found. Unfortunately, glucose-limiting chemostats with phenylalanine, methionine, or leucine as the sole nitrogen source did not show conclusive results of their regulation since these enzymes strongly overlap in their function [95].

Transcriptional and post-transcriptional regulation of the Ehrlich pathway genes is poorly understood. ChIP–chip experiments (chromatin immuno-precipitation on a microarray) revealed that ARO80 is the main transcription factor (TF) that regulates the expression of the ARO9, ARO10 and ARO80 itself [96]. Other approaches e.g. in vivo studies confirmed this regulation pattern [97]. Aro80 belongs to the Zn2Cys6 family of transcription factors known to bind palindromic elements. Recent literature reveals several ARO80 binding sequences which are all different and unique binding site [96] [98] [99]. Although the DNA binding motive is unclear all studies agree on the Aro80-dependent expression of ARO10 and ARO9.

Once volatile compounds are made they are transported out of the cell. The export of higher alcohols might occur by simple membrane diffusion, although the ATP-dependent transporter Pdr12 seems to be involved in the export of weak organic acids [100].

Fatty acids as substrates for flavour formation Free fatty acids (FFAs) are considered important precursors for different varieties of volatile compounds, especially in the flavour development of cheese. Lipolysis is an important biochemical step in cheese ripening. In particular the process is essential in Italian cheese varieties and blue mold cheeses. Indeed, milk-fat contains high concentration of short and intermediate triglycerides which then can become free after hydrolysis. Lipases in cheese can originate from several sources: milk, starter bacteria or fungi but also lipase preparations. Lactococcus sp. has a weak lipolytic capability but if present in high numbers can be responsible for the liberation of high levels of FFAs. In contrast, Penicillium sp. exhibits high levels of

33 lipolytic activity and is responsible for the distinctive lipolysis in mold ripened cheeses [82]. As shown in Figure 10 FFAs are the precursors for a number of flavour molecules.

Figure 10. General pathways for the catabolism of FFAs (modified from [82]).

Floral and fruity lactones are the results of a spontaneous intramolecular esterification of hydroxyl fatty acids. For instance, γ -and δ -lactones contribute to a peculiar peach note in Cheddar cheese. FFAs can also interact with alcohols to yield long chain esters. In Emmenthal cheese fourteen different esters have been found and in the Italian Parmiaggiano Reggiano up to 38 esters, suggesting their importance in the final product [82].

Metabolism of lactate and citrate Lactose is the starting fermenting substrate and the major carbon sources in food industry applications. For example, the dairy processing starts with certain starter bacteria (Leuconostoc sp., Lactococcus sp., Clostridium sp.) able to metabolize the lactose into a racemic mixture or L- or D-lactate (Figure 8). The lactate is then utilized by bacteria to produce mainly propionate, acetate and CO2. Acetate and propionate contribute to the flavour of the cheese whereas carbon dioxide is important for the “eye” formation in the cheese, the typical holes found in Emmenthal and Swiss-type cheeses [82]. Camembert and Brie are the results of mixed-fermentation of bacteria and fungi. Initially lactic acid bacteria release lactate which is rapidly metabolized by fungi such as Debaryomyces hansenii, Geotrichum candidum, Penicillium camemberti [101].

The role of the fungi is to convert the lactate into CO2 and water, and therewith to increase the pH at the cheese surface. Deacidification is an important step during cheese production, because it allows growth of coryneform bacteria involved in the ripening and flavour maturation of many

34 cheeses. Furthermore, the alkaline pH on the surface makes the final maturation step possible that is carried out by Penicillium camemberti in Camembert cheese. In this last part the amino acids released from the casein are converted into NH3 and fruity esters [82].

As mentioned earlier citrate is a key intermediate in the TCA cycle but it is also one of the constituents of milk (ca. 8 mmol∙L-1). Further, citrate is present in fruit juices, vegetables and is generally added as a preservative to foods. In nature, only a limited number of bacteria are able to ferment citrate [101]. Generally, it is not used as a carbon source but it is metabolized together with lactose and other carbon compounds. The products of the citrate metabolism are 4-carbon compounds, diacetyl, acetate, acetoin and 2,3-butanediol. Diacetyl is well studied and usually produced only in small amounts (1-10 µg∙mL-1 in milk) [82]. In cheeses and yogurts it has a very positive effect on the aroma profile because it gives the typical buttery note to the final products. In contrast, diacetyl is considered an off-flavour in lager beer production when it exceeds a concentration higher than 0,15 ppm [102].

1.10 Biological properties of quorum sensing molecules and VOCs as signaling molecules Biofilm formation and morphological development are controlled by the exchange of chemical compounds, better known as quorum-sensing signals. When a quorum sensing molecule reaches a critical concentration, the entire population of a species reacts by activating specific target genes. For instance, yeasts grown in the presence of isoamyl alcohol, one of the major alcohols produced during beer/wine fermentation, induce pseudohyphal growth [103] [104]. The effect is not limited to this alcohol alone, since 2-phenylethanol, another compounds derived from the amino acids catabolism, seems to affect morphogenesis as well and induces invasive growth [105]. The same two compounds are produced in low concentration also in Candida albicans and Candida dubliniensis. In these species they participate in the dynamics of bioﬁlm formation [106]. However tyrosol and farnesol are the main extracellular molecule which controls mycelia development in C. albicans. Tyrosol stimulates germ tube formation, whereas farnesol act as an inhibitor of mycelia development [107] [108]. When both signals are present the farnesol is dominant over tyrosol [109].

Further, volatiles and several flavouring molecules are known to be used as biological signals for communication. Many fungal-insect, fungal-fungal, fungal-bacterial and fungal-plant interactions have been discovered in a broad range of disciplines. For practical reasons very few

35 studies have been done using gas phases, assuming that the physical state of the compounds only matters the critical dosage necessary to trigger an effect. Furthermore VOCs produced by a given fungal species can raise different responses, depending on the environmental context. The classical example is the ubiquitous fungal volatile 1-octen-3-ol. In Aspergillus nidulans this C-8 compounds acts as inhibitor of conidiation when it reaches high concentrations. It is a self- control system to steer the asexual reproduction. The inhibitory effect is reversible since conidiospores could normally germinate when the 1-octen-3-ol was removed [110]. When Arabidospis thaliana is exposed to 1-octen-3-ol the plant’s defense mechanisms become active [111] blocking the expansion of the pathogen, Botrytis cinerea, on infected leaves. In insects, the C-8 alcohols production by wood-decay fungi stimulate the reproduction of the conifer feeding bark beetle, and inhibit the growth of their natural predators [112].

Furthermore, the R enantiomer of the 1-octen-3-ol is commercially known as “roctenol” and is widely used as insect attractant [113] [114]. Even though fungal volatiles are generally perceived as attracting agents for insects, they can also act as repellents. Oviposition of female houseflies is blocked in substrates infected by Fusarium, Phoma, Rhizopus, and other fungi. The VOCs responsible for this reaction are dimethyl disulfide, phenylacetaldehyde, 2-penylethanol, citronellal, and norphytone [115]. Bacillus nematocida is an interesting example of bacterial predation against nematodes. It produces the worm attracting, volatile, organic compounds benzaldehyde and 2-heptanone. The secretion of these bacteria attracts the nematodes and once in the nematode guts they produce proteases leading the host to death [116]. Moreover, several fungi are known to establish symbiotic interactions with plants as mycorrhizae. Gas phase molecules can easily diffuse in cavity of the soil, reach far distances and produce intermediate molecules to promote interspecies interactions. This mutualistic relationship between fungi and plants is dynamic and controlled by different stimuli such as nitrogen to phosphate ratio, abundance in the soil, soil moisture, etc. Advantages of such mycorrhizal interactions are an increased nutrient uptake and resistance to pathogen attacks [117]. In conclusion, these small metabolites produced by a variety of organisms are responsible for a number of biological activities and interactions. Many of them are still poorly understood and need to be investigated. Presumably, technological advances and further research will elucidate more of these chemical interactions, offering new prospective for applied disciplines.

1.11 Biotechnological application of fungal VOCs Beside their role in biological activities VOCs have also an enormous potential in biotechnological applications. In agriculture they can be used as biocontrol compounds, reducing fungicides applied on crop plants. Precisely designed mixtures of volatiles are able to inhibit plant pathogens [118]. The advantage of using small volatile compounds lies in their ability to travel long distances and infiltrate in the soil pores. For instance, the non-pathogenic Fusarium oxysporum MSA35 lives in association with a consortium of ectosymbiotic bacteria. This association is able to suppress Fusarium wilt of pathogenic F. oxysporum. It has been shown that the volatiles emitted by the fungus-bacteria consortium are able to modify the mycelia surface of another pathogenic Fusarium isolates [119]. Another example is Muscodor albus which is able to inhibit the growth of the broccoli pathogen Rhizoctonia solani and the pathogen Phytophthora capsici, which cause root rot in bell pepper [120]. Mixture of volatiles from M. albus can also be used in non-agricultural applications, for instance in controlling and significantly reducing common building molds [121]. Also, it has been discovered that a number of antifungal volatiles are active against several human pathogens [122] [123], and this may offer an additional alternative to fight against drug resistant strains.

Recent evidence suggests that single VOCs are generally not effective [124] [125]. Organic volatile mixtures are more predominant making it necessary to decrypt this chemical complex to achieve the antifungal activity observed in vivo. The analysis of these VOC mixtures could also become useful in medical areas to identify bacterial, viral and fungal infections in the lungs [126]. Characteristic volatile mixture can serve as biomarkers to assess human infections by using rapid and non-invasive techniques [127]. VOC emissions can also be used to mimic flower essences to attract pollinating insects and facilitate pollen transfer [128]. By using mixtures of volatiles that are normally produced by fungi to attract and promote their dispersion nature, scents can be created to increase yields of fruit and seed crops [129]. Botanophila ﬂies, which are the common vectors for the endophytic fungal genus Epichloë, are attracted by a volatile sesquiterpenoid alcohol, and a methyl ester emitted by the fungus [130]. Additionally, fungal VOCs can be applied for their insecticidal activity. An example is naphthalene produced by Muscodor vitigenus, an endophytic fungus which colonizes the Amazonian plant Paullinia paullinioides. This compound was initially used in “moth balls” due to its effective functions as insect repellent [131]. In the last years VOCs have also been considered as biofuel precursors or additives. Many organic volatiles derived from saprotrophic fungi, including alcohols, alkanes, alkenes, esters, ketones, sesquiterpens, propylene, ethane, are similar to biofuel target compounds [132] [133] [134]. But only very little information is available concerning the

37 pathways involved. New studies including metabolomic and transcriptomic data are now trying to correlate and clarify which genes and enzymes are involved in the production of these hydrocarbon compounds [135].

1.12 Fungal VOC collection and detection

VOC collection The intrinsic small nature of VOCs and their low concentrations make them very difficult to characterize and study. In addition, fungi produce very complicated mixtures of volatiles which require powerful separation technologies to unravel this complicated composition of organic compounds. Depending on the application and the chemical nature of the volatile of interest, different sample preparation, separation, identification, and quantification methods can be applied. At present, gas chromatography-mass spectrometry (GC-MS) is the traditional technology applied because it combines efficient separation methods with highly sensitive detection capabilities and precise quantification (up to part-per-billion/trillion) (Figure 11). Sampling and trapping of the VOCs is the initial critical step of the analysis. Each sampling method has its own advantages and disadvantages, the choice should be made according to chemical nature of the analyte(s) and the concentration in the gas mixture. The goal is to prepare an extract that is representative for the original sample. Chemical instability, variation in volatility and matrix composition are the main factors which have to be considered in this process. In general the extraction methods can be separated into two classes: adsorbing and solvent extraction methods. In adsorbing techniques a fiber/membrane is used which traps and collects volatile compounds present in the headspace. No sample preparation is required for the headspace sampling. It can be done either for liquid or solid samples. Besides, gradients of temperature and an inert gas flow can be applied to improve the sampling of the headspace. Various fibers are available on the market coated with different selective properties, depending on the polar character of the analyte(s). Among the different absorbing materials are charcoal filters, Tenax and Super Q fibers are the most frequently used. In all cases the volatile compounds released from the sample are trapped and concentrated into the absorbing material. Later thermal desorption is applied to release the organic compounds into the GC-MS system. Recently, trapping procedures using the solid phase micro-extraction (SPME) fiber technique have become a popular method in several contexts [136]. The technique combines extraction, concentration and introduction of the volatiles into one automated step, offering the advantage to directly profile living fungal cultures [137]. Steam distillation extraction (SD) and liquid-

38 liquid extraction offer the advantage to extract volatiles that show reduced volatility [138]. Therewith they get separated from the aqueous phase containing proteins and carbohydrates which are not wanted in the GC-MS system. Though, the choice of the organic solvents (i.e. dichloromethane, hexane, pentane) can affect the final VOC profile, excluding highly volatile or early eluting compounds [139]. Already in 1995 Larsen and Frisvad showed that different profiles of fungal VOCs can be obtained by applying varying collection techniques.

VOC separation After the extraction step the separation of the volatiles occurs in the GC system. The gaseous compounds fly through a thin column, pushed by an unreactive gas (usually nitrogen or hydrogen). The process is also called ‘mobile-phase’. The chemical components of the VOC mixture have different progression rates into the column (‘stationary phase’), depending on the strength of adsorption. Compounds which have a low affinity with the column are eluted first and vice versa chemicals which establish more interactions with the solid phase are eluted last. The separation process can be optimized by modifying different parameters such as polarity of the column, temperature, pressure of the mobile phase and amount of samples run through the column.

VOC detection Once the separation is completed the organic compounds are received and captured by a detector. The flame ionization detector (FID) and the thermal conductivity detector (TCD) are the most common and robust detectors used in combination with GC systems. They can detect a wide range of compounds with a sensitivity of approximately 0.5ng/µL for carbon compounds. Nowadays most of the GC systems are coupled with a mass spectrometer (MS) which acts as the detector. In the MS the compounds eluting from the column are bombarded with high-energy electrons breaking them into charged fragments. These small fragments are separated by their mass to charge ratios (m/z) in the analyzer and produce a spectral pattern unique to the compound. Compounds can then be recognized by using a library of mass spectra.

The resulting output of the detector is a chromatogram, based on peaks and areas. The retention time taken from the tip of the peak can be used to identify the compounds. The area below the peak is proportional to the amount of analyte in the sample. A mathematical integration of the peak and a calibration curve then allow a quantification of each compound of interest.

Figure 11. Schematic representation of a GC-MS system.

2. Aim

This PhD thesis project is part of the EU Marie Curie Initial Training Network Cornucopia. The general task of the program has been to make use of the yeast biodiversity and find new applications for non-conventional species. The goal of this thesis is to give more insight into the molecular pathways, in particular into the Ehrlich pathway, which drives flavour production in fungi.

The Ehrlich pathway is a linear and well conserved pathway that converts amino acids into either fusel acids or fusel alcohols. In yeast this secondary metabolism is one of the major routes that contribute to the aroma profile of fermented products. One of the goals of this study was to examine the main genes and their corresponding proteins that are involved in the Ehrlich pathway. In particular, I focussed on the ARO gene family of Saccharomyces cerevisiae. As a result I gained insight into the transcriptional regulation of the ARO genes. This knowledge enabled me to develop a tool that allows the prediction of volatile production in Saccharomyces sensu stricto species that can be used to screen a large number of strains within a short time and low effort and investment.

I further analysed the function of the ARO gene family in Ashbya gossypii, a very aromatically smelling Eremothecium species. In particular, I focused my attention on the key transcription factor of the pathway, ARO80. I could show that the deletion of each of the ARO genes except for ARO8a affects the flavour profile of the fungus. Furthermore, I investigated the volatile compound profiles of the two closely related Eremothecium species, A. gossypii and E. cymbalariae. The purpose was to determine differences in flavour production and relate these profiles to the gene set of the species.

Finally, 53 novel flavour producers strains selected within the Saccharomyces clade have been investigated to study their potential for industrial fermentation. For this purpose a high throughput screening of non-conventional yeasts was performed to test fermentation performance and flavour characteristics of these non-Saccharomyces yeasts.

3. Objectives and state-of-the-art

3.1 Functional analysis of the ARO gene family in S.cerevisiae and A. gossypii During the last years several studies have been started to elucidate the genetic regulation of the Ehrlich pathway, however, not with focus on the ARO gene family. ARO8 and ARO9 were initially isolated by the complementation of the phenylalanine/tyrosine auxotrophy [140], and were defined as the aromatic amino acid aminotransferases I and II, respectively [141]. Aro8 is responsible for the phenylalanine and tyrosine biosynthesis and is constitutively expressed. Aro9p, on the other hand, is more a catabolic enzyme and is involved in the tryptophan degradation [142] [140]. However the double mutant aro8/aro9 can be partially complemented by one of the two, suggesting broad substrate speciﬁcity and reversibility of the transaminase reaction [140]. The irreversible decarboxylation step was prematurely attributed to the pyruvate decarboxylases, PDC1, PDC5 and PDC6 [143]. Only years later, after combining genetic, physiological, and biochemical approaches, Aro10p was identified as the major decarboxylases of the Ehrlich pathway, showing an overlapping substrate specificity when phenylalanine, leucine, or methionine was used as a nitrogen source [144] [145].

The transcriptional regulation of the ARO genes is mediated by two factors: the amino acids availability and the quality of the nitrogen source. The amino acid activation is controlled by the transcription factor Aro80 [141] [146], whereas the response to nitrogen is governed by nitrogen

41 catabolite repression (NCR) conveyed by the GATA factors Gat1 and Gln3. Optimal conditions for the ARO9 and ARO10 induction are therefore the presence of inducers, aromatic amino acids and the absence of an optimal nitrogen sources like ammonia or glutamine [146]. Only recently, the interplay of GATA activators and the transcription factor Aro80 has been explained in respect of the amino acids catabolism. As shown by Kyusung Lee and Ji-Sook Hahn Aro80 is constitutively bound to its target promoters and is necessary to recruit Gat1 and Gln3 to the Aro80 target promoters. Nevertheless, the binding of the GATA factors is independent of the Aro80 activity and they do not have any influence on the binding of Aro80. Figure 12 shows the proposed transcriptional activation model of ARO9 and ARO10.

Figure 12. ARO9 and ARO10 transcriptional control by Aro80 and GATA activators (modified from [146]).

Although studies have been done to elucidate the interplay between Aro80 and the GATA factors, still little is known about the precise binding site of Aro80 to the target promoter regions. Several binding sites have been proposed which make the establishment of a consensus sequence difficult [97] [99] [96] [88] [98]. All authors agreed that the binding should be in proximity of the CCG triplets, the typical recognition site for the Zn2Cys6 transcription factor family. In Paper 1 we identified promoter regions of several genes which were more responsive to ARO80 over expression. We co-transformed a plasmid containing ARO80 under control of the A. gossypii TEF-promoter and plasmids containing ARO8, ARO9, ARO10 and ARO80 promoter-StlacZ reporter gene fusions into S. cerevisiae. We measured the β-galactosidase activity and identified the promoter regions which were most responsive to the ARO80 expression level. Based on the assumption that high expression levels of the ARO80 transcription factor results in high activation of the Ehrlich pathway we used the ARO9 promoter to correlate reporter gene activity and flavour formation in yeast. To determine the

42 applicability of the ARo9-StlacZ system as a screening tool we transformed various strains with this ARO9p-StlacZ reporter gene construct. Then we measured the β-galactosidase activities and compared them with the corresponding flavour profiles. The same approach has been used to study the genetic regulation of the conserved ARO genes in A. gossypii. We tested the Aro80 dependent expression of the AgARO genes by applying the same set up described above (Figures 9, Paper 2 and our unpublished data).

3.2 Analysis of the different volatile profiles of A. gossypii and E. cymbalariae is correlated to their genetic backgrounds. The contribution of S. cerevisiae to the aromatic profile of food and beverages is well documented [147]. However the fungal biodiversity has not been explored to its full potential. In Paper 2 we studied the potential use of A. gossypii for flavour production, employing the GC- MS technique to explore the complete aroma profile. Comparative genomic analysis has shown a conservation of the key genes of the Ehrlich pathway in A. gossypii, especially the family of the ARO genes. As shown in the Table 5, A. gossypii does not contain the ARO9 gene but we found a duplication of ARO8. The two ARO8 paralogs were named ARO8a (AGR167w, non-syntenic homolog of S. cerevisiae ARO8) and ARO8b (AFR548w, syntenic homolog of S. cerevisiae ARO8).

Species Transaminase Decarboxylase

S. cerevisiae YGL202W (ARO8) YDR380W (ARO10)

YHR137W (ARO9) YLR044C (PDC1)

YHR208W (BAT1) YLR134W (PDC5)

YJR148W (BAT2) YGR087C (PDC6)

A. gossypii AGR167w (ARO8a) ACR211W (ARO10)

AFR548c (ARO8b) ACL134C (PDC1)

ACL072C (BAT1) AAL073W (PDC6)

E. cymbalariae Ecym_7228 (ARO8a) Ecym_2369 (PDC1)

Ecym_7477 (BAT1) Ecym_3186 (PDC6)

Table 5. Genes contributing to the Ehrlich pathway.

To address the metabolic role of the ARO genes for the Ehrlich pathway in A. gossypii, we generated the respective deletion mutants and investigated the overall impact on flavour production. In addition, we also analysed the VOC formation of the closely related specie E. cymbalariae, which does not show the fruity flavour characteristics of A. gossypii. In line with that, E. cymbalariae possess an even more limited set of genes that encode for enzymes of the Ehrlich pathway (Table 5). The same gene setup was observed in L. waltii, where only the transaminase; ARO8a and ARO8b, are present (see Paper 2 for more details). Here, the objective was to present a comparative flavour profile of A. gossypii and E. cymbalariae, in order to extend our knowledge of the Ehrlich pathway and its role in the production of natural flavouring molecules.

3.3 Flavour molecules produced in the Saccharomyces clade. A wide selection of non-conventional yeasts (NCY) has been studied in this thesis. The list contained more than 50 strains encompassing a wide variety of species within the Saccharomyces clade. Compared to previous works our approach was not limited to certain geographical regions [148]. On the contrary, our selection included strains isolated worldwide from different natural sources (i.e. meat, cereals, fruits, cocoa beans, olive, cheese, etc.). The strain repertoire was also based on the previous identification of these strains in fermentation processes [149]. The complete list of strains is shown in Table 1 (Paper 3). All strains were tested on plate assays for their response to stresses (i.e. ethanol, osmolarity and temperature) but also for the production of off-flavours (H2S and phenolic compounds). Furthermore, all strains were tested in fermentation condition and the produced volatiles were sampled by headspace solid-phase microextraction (SPME) and analyzed by gas chromatography–mass spectroscopy (GC–MS). Data mining and statistical analysis software were employed to cluster and highlight species related flavours.

4. Discussion

4.1 Reporter assay for ARO genes in S. cerevisiae and A. gossypii The genetic regulation of the Ehrlich pathway has gained increasing interest in the last years, with respect to the critical role in fermenting processes. Particular attention was paid to the

ARO genes family, due to the broad substrate activities of the encoding enzymes. The transcriptional control of these genes depends at least in part on the transcription factor Aro80. Recent studies have proposed the molecular mechanism of this network [146]. However, despite computational predictions and chromatin immuno-precipitation (ChIP) experiments, it was not clearly shown which genes were more responsive to increased ARO80 expression. In addition, previously there was no information available on the regulation of ARO genes in A. gossypii. The ß-galactosidase assay presented in this thesis (see also Paper 1 and Paper 2 for detailed assay information) investigated the potential of AgAro80/ScAro80 to activate target promoter regions (Figure 13).

AgARO80 ScARO80 1000 AgARO80 ScARO80 900 1800 800 1600

700 1400

600 1200 500 1000

400 800 Miller Units Miller

300 Units Miller 600 200 400 100 200 0 0 ScARO8 ScARO9 ScARO10 ScARO80 AgARO8a AgARO8b AgARO10 AgARO80

Figure 13. A ß-galactosidase reporter assay was used to examine target promoters of AgAro80 and ScAro80 and to quantify the activation of these promoters. Plasmids containing AgTEF1p driven AgARO80 or ScARO80 sequence and plasmids with an ARO gene promoter StlacZ constructs were co- transformed into S. cerevisiae. StlacZ served as a reporter to quantify the activation of the target promoter.

In A. gossypii, Aro80 positively regulates the expression of all ARO promoters tested. The highest activation was found for the AgARO8a promoter. In S. cerevisiae, ScAro80 predominantly activates the ScARO9 promoter, but shows very low activation of other ARO gene promoters. It was described in the literature that ScAro8o does not regulate the expression of ScARO8 [97]. Our results could confirm this finding (see Figure 13 upper diagram). However, in contrast to the computational results and ChIP experiments, we could not observe a strong activation of the ScARO10 promoter region. Surprisingly, we could observe a higher activation of ARO10p by AgAro80 than by ScAro80. The functional analysis of the AgARO deletion mutants showed no growth phenotype of the mutants but a substantial reduction in the production of the rose flavour 2-phenylethanol (Paper 2). Interestingly, and unlike

S. cerevisiae, neither the double mutant AgARO8a/b nor the single mutants showed any amino acid auxotrophies. This may imply a different regulation in the amino acids biosynthetic pathway or an additional enzyme which can provide such an activity in A. gossypii.

4.2 Use of a lacZ- reporter assay to correlate reporter gene activity with flavour production Flavours are small volatile organic compounds produced during yeast fermentation. Their natural presence in the gas-phase has been used to collect, separate and analyze them by modern GC-MS technologies. Even though high-throughput fermentation screening systems have been improved enormously over the last years, the aroma profiling still requires a lot of manual and analytical work. In addition the fermentation process needs days to be completed, preventing the possibility of fast screening with low material and labour costs. In Paper 1 we discussed an indirect genetic method to identify flavour active strains, exploring their genetic potential to produce high amounts of aromas for an industrial interest. ScAro80 represents one of the key transcription factors involved in the regulation of the Ehrlich pathway in yeast. Based on the results of the ß-galactosidase assay, we can conclude that the ARO80 expression level directly influences the regulation of the ARO9 promoter. Further experiments found a correlation between the β-galactosidase activity of ARO9 and the amount of β-phenylacetate and β-phenylethanol produced during fermentations. This finding was recently supported by other studies which have used ARO9 and ARO80 overexpression to improve significantly the biocatalytic production of β-phenylethanol [150]. Thus, this newly developed fast tool can conveniently streamline the identification of yeast strains with potential flavour characteristics. It is designed to be applied for high throughput screening of yeast collections. However, the genetic tool builds on the conservation of the transcriptional regulatory circuits found in S. cerevisiae. Thus, one limitation of the assay is the range of species or hybrids in which the use of ScARO80 provides reliable correlations with flavour production Nevertheless, new screening tools can be developed based on our genetic tool using other species’ ARO80 target genes. Furthermore, the genetic tool can be used to screen a large variety of S. cerevisiae strains or hybrids for the flavour production as well as large numbers of progeny of these strains.

4.3 General flavour differences between A. gossypii and E. cymbalariae In this study we present the aroma profile analysis of two closely related filamentous ascomycetes, A. gossypii and E. cymbalariae (Paper 2). We used the volatile profiling analysis to provide a quick access to the metabolites they produced. Specifically, we focused on all the VOCs linked to the Ehrlich pathway in which particularly aromatic and branched chain amino acids get catabolized into higher flavour compounds, also known as fusel products.

In general, the amounts of branched-chain aldehydes are rather high in both fungal species, whereas the production of aromatic volatiles was found to be specific for A. gossypii. Comparing the gene set linked to the Ehrlich pathway we observed that E. cymbalariae lacks two important genes that are known to be involved in catabolism of aromatic amino acids, namely ARO9 and ARO80 [151]. A. gossypii lacks only the homolog of ARO9 but contains two paralogs of ARO8. We could show that the transcription factor AgAro80 can activate the transcription of both ARO8 genes as well as ARO10 in A. gossypii (Figures 13). Although A. gossypii lacks the homologous ScARO9 gene, we concluded that one of the ARO8 genes may function as a transaminase in the catabolism of phenylalanine. Though, this has to be confirmed by future experiments.

Branched-chain aldehydes such as 3-methyl butanal, 2-methyl butanal and 2-methyl propanal are derived from branched chain amino acid degradation. These compounds are key flavour metabolites [152]. Mainly perceived as malty or chocolate-like they are usually formed during the fermentation of many food related substrates, especially during the fermentation of cocoa beans. The high reactivity of the aldehyde carbonyl group makes this class of compounds very easy to be reduced to the respective alcohols or oxidized to the corresponding acids. A high flux in the leucine biosynthesis pathway of both fungi might explain the high levels of these intermediates. Having an exaggerated leucine production could maintain the Ehrlich pathway catabolism of leucine and therefore keep a high level of intermediates at all times. A genetic evidence for a high production and flux in the leucine synthesis pathway might support this hypothesis. In A. gossypii a duplication of the LEU4 (AFL229w) gene has been found (ADL015c, non syntenic homolog of ScLEU4, Figure 14).

Figure 14. Gene synteny of the LEU4 loci in S. cerevisiae, A. gossypii and E. cymbalariae. The red stars indicate positions of tRNA in A. gossypii.

LEU4 encodes a α-isopropylmalate synthase, responsible for the first step in the leucine biosynthesis pathway. In S. cerevisiae LEU4 has a paralog, LEU9, which arose from the whole genome duplication. Functional analyses in conjunction with non-fermentable carbon sources demonstrated that only LEU4 is required to maintain normal growth rates of S. cerevisiae in the absence of leucine [153] LEU9 plays an auxiliary role in yeast since about 80% of total α- isopropylmalate synthase activity in wild-type cells is provided by LEU4 [154] The second LEU4 gene in A. gossypii (ADL015c) is not in synteny to the S. cerevisiae LEU4 or LEU9 gene and therefore arose by a unique duplication event in A. gossypii (Figure 14). In E. cymbalariae no duplication of the LEU4 gene occurred. The role of LEU4 in A. gossypii has been investigated by deletion mutants. The single LEU4 deletions did not show any growth delay or difference in the VOCs production. We failed to delete both LEU4 genes of A. gossypii concluding that this double deletion is lethal even in media with excessive amounts of leucine. These results suggest a different regulation of the leucine biosynthesis pathway in A. gossypii.

3-methyl butanal is a key intermediate in the isoamyl alcohol biosynthesis. By continuous feeding of the pathway with this leucine derived intermediate the fungus can then overproduce the respective fusel alcohol. However, the reason why the Eremothecium species produce high yields of this particular alcohol is still unknown. In S. cerevisiae isoamyl alcohol together with 2-

48 phenyl ethanol, and tryptophol can induce pseudohyphal growth and stimulate flocculation [104] [105]. In A. gossypii the production of fruity flavours could provide a benefit in its ecological niche to attract spore-transmitting insects for dispersal to reach new sites of infection. Further experiments are necessary to test this hypothesis.

4.4 VOCs produced in the Saccharomyces clade Non-conventional yeasts (NCY) represent a new frontier for industrial applications. In the last years a lot of studies have explored the yeast biodiversity looking for new traits that can make a breakthrough in biotechnological processes. In this study we focused our attention on novel flavour producers within the Saccharomyces clade. Initial results showed that yeast strains which belong to the same species exhibited similar VOCs pattern (Paper 3). During fermentation we collected various parameters (i.e. fermentation kinetics, ethanol content, pH of the media and sugar content) which allowed us to make a rational selection of the NCY we wanted to continue with. This approach streamlined the strain selection to 18 representative NCY, which were then further tested for specific traits and flavour profiles.

Comparing the fermentation profiles we noticed that almost all the NCY finished fermentation after four days and showed a typical sigmoidal growth curve-like pattern of the CO2 release.

The final Plato at the end of all fermentations was in a range between 3 and 4. It is not yet known why yeast strains stop fermenting before all fermentable sugar is consumed [155] Accumulation of toxic intermediates, oxidative stress, lacking of micronutrients and deplete of all the fermentable sugars could be possible explanations for an incomplete sugar fermentation.

The only exception we noticed was in fermentation of Wickerhamomyces anomalus. The strain did not show a lag-phase of CO2 release but a linear CO2 curve instead, which ended after eleven days. During the fermentation of W. anomalus we observed a water-insoluble layer emerging on top of the fermentation cylinder. This might explain the slower and constant CO2 release. The formation of the film layer has been reported as a typical phenomenon in stored wines. Apparently the residual amount of oxygen in the headspace of the fermenters triggers the formation of a biofilm. The main volatiles produced by this film are acetic acid, acetaldehyde and acetate esters [156].

Nevertheless, Wickerhamomyces anomalus and Pichia kluyveri generated the highest amount of CO2 at the end of the fermentation process (up to 2.8 g/40ml media). On the other hand, they produced the lowest amount of ethanol with 6 and 4% v/v, respectively.

The strain Debaromyces subglobosus showed an opposite fermentation profile. It was able to produce 8% v/v ethanol and utilize most of the sugar available in the media. However, it released the lowest CO2 content (1, 3 g/40ml media). In all the other fermentations we noticed a reasonable relation between ethanol production and sugar consumption.

Ethanol is not the only product of the fermentation process. Volatiles also contribute to CO2 production. Our data showed a clear correlation between the fermentation performance of Wickerhamomyces anomalus and Pichia kluyveri species and the amount of volatiles produced.

The highest CO2 release was coupled with the extremely high production of esters in these strains. In particular, the strains produced large amounts of ethyl acetate, isoamyl acetate and phenyl acetate. Together with the esters high yields of acetic acid were registered. Both ethanol and acetic acid are generated from the common intermediate acetaldehyde by the aldehyde dehydrogenase, reducing one molecule of NAD(P)+. Ethanol and acetate can both inhibit yeast growth if present at high concentrations. Ethanol generally diffuses out from cell and just a very low concentration is retained inside. Acetic acid accumulation inside the cell is toxic. It can generate high turgor pressure and trigger the formation of free oxygen radicals and consequently oxidative stress [157]. Moreover, the passive diffusion through the plasma membrane is very limited due to the largest dissociated form of the acid since the intracellular pH is higher than 4.75, which is the pKa of the acetic acid. Therefore, the cell needs to either pump it out by ATP-hydrolyzing transporter systems (ABC transporters) or convert the acetate into a less toxic compound. The first option of detoxification implies the use of ATP, which is quite limited in fermentative conditions [158]. In the second case an enzymatic reaction is required. Ethyl acetate can be synthetized by two different pathways in which both require ethanol as substrate. Alcohol acetyl transferases can synthetize ethyl acetate by a reaction that involve acetyl co-enzyme A as a cofactor [159]. Generally, this pathway is involved in the production of isoamyl acetate, using isoamyl alcohol instead of ethanol as a substrate for the reaction. The other enzymatic way for the production of ethyl acetate is by esterase activity. This reaction that is a reverse hydrolysis uses the acetate and ethanol to produce ethyl acetate (Figure 15).

Figure 15. Enzymatic routes for the production of ethyl acetate.

W. anomalus differs from S. cerevisiae by producing most of acetate esters by the inverse esterase reaction [160]. More recent studies from Rojas et al. (2002) [161] confirmed the high ester synthase activity in W. anomalus. The authors argue that the natural overproduction ethyl acetate could be related with the anti-fungal activity of this compound. Along with other NCY, W. anomalus could have developed the ability to accumulate damaging compounds to outcompete S. cerevisiae or other microorganism for nutrients, oxygen and space. For this reason, W. anomalus has been studied as a bio control agent, especially against molds. The inhibitory effect against Penicillium roqueforti has been shown to be related to the combined effect of ethanol and ethyl acetate [162].

Esters are also important in food production due to the organoleptic properties that they imply. A balance among them is important to give the typical bouquet to specific product. The high ester production of W. anomalus, P. kluyveri and Z. mellis has been reported in other works and it is known as a common feature for NCY. Together with their lower fermentation abilities the uncontrolled production of esters is the main reason that has prevented their use as starter culture in mixed fermentation. Despite that, recent studies have reevaluated the application of these strains, applying them in different production processes. For instance, Wickerhamomyces sp. are commonly used in the production of soya-sauce, where the high amount of esters, especially ethyl acetate have shown to give the strong and sweet/caramel like aroma to the product [163]. Pichia kluyveri has been extensively tested as multistarter culture for cocoa, tequila and wine fermentation. It has been shown that controlled mixed fermentation of Pichia kluyveri and Kluyveromyces marxianus enhanced the sensory quality of the cocoa beans, offering a valid alternative to the standard spontaneous fermentation used in cocoa production [164]. Moreover the strains have been investigated for industrial application of tequila mixed

51 fermentation. Compare to standard S. cerevisiae fermentations, P. kluyveri and K. marxianus mixtures were more efficient in alcohol and ester production [165]. P. kluyveri strains are also commercialized by Chr. Hansen in products like Viniflora® FrootZen™. The strains are used to start the alcoholic fermentation of must until they reach an alcohol content of 4%. Then a standard S. cerevisiae strain is added to complete the process. Again, the advantage of the second strain is to add additional fruity flavours to the wine resulting in a higher complexity of the wine (for more details look at http://www.pros.co.nz/PDF/PDS/FROOTZEN%20PDS.pdf). Debaromyces subglobosus showed another exceptional profile of volatiles. Characterized by delayed fermentation the strain was yet able to produce large amount of aldehydes. In literature Debaromyces sp. are described as strains with weak fermentation properties [166] but a notable halotolerance [167]. We could confirm this high tolerance in our studies as well by growing D. subglobosus in SD media with 4% NaCl (Paper 3, data not shown). This natural salt tolerance of the genus has been used for production processes such as the synthesis of dairy products such as cheese and fermentation of meat products [166]. Fruity ketones are important chemical compounds in cheese ripening as well. In line with that, the largest amount of pyranone was detected in D. subglobosus. In contrast, we detected extremely low levels of esters by GC-MS. Van den Tempel and Jacobsen (2000) [168] found high esterase activity in different isolates of D. hansenii which could explain this low amount of esters. In meat production Debaryomyces sp. have been reported to generate volatile compounds during the fermentation of dry-fermented sausages. In particular, several species inhibit the formation of lipid oxidation compounds and generate ethyl esters [169]. Also, branch-chained aldehydes are critical components for the maturation of meat products, giving nutty, cheesy and salty notes to the food products [152]. Generally these aldehydes are produced by heat-induced processes (Strecker degradation) but small amounts of aldehydes are also produced enzymatically during food fermentation. D. subglobosus shows the highest production of aldehydes among our yeast selection. As one of them 3-methyl butanal, an intermediate in leucine degradation, is produced in higher amount in fermentations with D. subglobosus and Z. mellis. 3-methyl butanal is a common aldehyde produced in cured-hams by lactic acid bacteria and Micrococcaceae [170]. Since the strains produced detectable amounts of 3-methyl butanal but also other compounds derived from Strecker degradation, i.e. 5- hydroxymethylfurfural, furfural, 5-methylfurfural they might have a possible application in meat processing. In summary we did an initial characterization of 20 NCYs by correlating their fermentation performances with their aroma profiles. In this first screening D. subglobosus, W. anomalus, P.

52 kluyveri and Z. mellis displayed unique fermentation and volatile profiles that may be used for supply of additional flavours, e.g. in mixed fermentations.

5. Conclusions and Outlook

Flavours and volatile compounds have a growing market potential, especially for food and beverages or fragrance industries. Fungi are able to produce various mixtures of VOCs and thus provide various opportunities to generate desirable natural flavours. The Ehrlich pathway is one of the main routes involved in the synthesis of volatile aroma compounds, especially during yeast fermentations of beverages. Previous observations identified the ARO gene family as a target set of genes with a strong impact on the activity of the Ehrlich pathway. Comparative genetic analysis showed a clear conservation of this gene family in the Saccharomyces clade. In our experiments we focused on the key transcriptional factor, ARO80. By using a lacZ-reporter system we quantified the Aro80-dependent expression of key ARO genes in both S. cerevisiae and A. gossypii. Expression of ARO-genes is also known to depend on other transcriptional regulators, e.g. the general biosynthetic amino acids regulator GCN4, which contributes to the complex regulation of ARO genes. At present, the regulation of ARO-genes has only been studied on a transcriptional level. Further studies are needed to elucidate if there also is a regulation of ARO genes on a post-transcriptional stage.

In our promoter analyses we found that the ScARO9 expression is induced by Aro80. This prompted us to generate a ScARO9- lacZ reporter system. This indicated that the expression level of ARO9 correlates with the formation of two aroma alcohol products, 2-phenylethanol and 2-phenylethyl acetate. We went on to transfer this reporter tool into other Sacchamyces sensu stricto species. With these studies we could confirm the correlation of ARO9 gene expression with the production of 2-phenylethanol and 2-phenylethyl acetate. This tool opens the way for high throughput screenings of entire strain collections or the analysis of progeny pools using microtiter plate assays. This inexpensive indirect assay offers new opportunities for the isolation of candidate strains with increased flavour production..

Our data with A. gossypii showed that AgARO8a/b and AgARO10 are regulated by AgAro80. Deletion of almost all the ARO genes resulted in a severe decrease in flavour production, especially in 2-phenylethanol and isoamyl alcohol. However, overexpression of ARO80 resulted

53 specifically increased the production of isoamyl alcohol but not of 2-phenylethanol. Experiments in S. cerevisiae showed an increased flavour production on media with poor nitrogen sources. These growth conditions could be tested in A. gossypii as well to further increase the production of flavour components. Additionally, spiking of media with amino acids could induce flavour product formation. Further experiments will be necessary to identify the underlying causes of natural over production of isoamyl alcohol in A. gossypii. We have begun to analyze the leucine biosynthesis pathway of A. gossypii. Also, the biological function for the production of the high levels of this alcohol should be addressed. Recent studies showed that 2-phenylethanol and isoamyl alcohol emitted by the fungus Aureobasidium pullulans are involved in the attraction of eusocial wasps [92]. Similar experiments could be used in A. gossypii to test the ability to attract spore-transmitting insects on which A. gossypii depends for its dispersal.

The close relative of A. gossypii, Eremothecium cymbalariae, produced a very distinct flavour profile, particularly with regards to its high production of ethyl acetate. The biological function of this is also unknown. Ethyl acetate could represent a strategy to avoid high concentrations of acetic acid in a similar mechanism as described for W. anomalus (see paragraph 4.4). Studies could be done to address the enzymatic activity of esterases and alcohol acetyl transferase in E. cymbalariae.

In collaboration with the EU consortium Cornucopia, which is focusing on the identification of novel potentials of NCY, we characterized a selection of 18 different species that are presented in this thesis. In this characterization we recorded their fermentation performance followed by a flavour profiling. The volatile aroma profiles of W. anomalus, P. kluyveri and Z. mellis and D. subglobosus were unique from all other strains. These four species will be characterized further in various fermentation conditions, i.e. different carbon and nitrogen sources, temperature and pH to explore their potential in food production. Additionally, these strains might be favorable in mixed-fermentations.

By using NCY strains in different co-fermentation regimes new flavour mixes may be generated and to be used in new kinds of beverages.

References 1. Hattenschwiler, S., N. Fromin, and S. Barantal, Functional diversity of terrestrial microbial decomposers and their substrates. C R Biol, 2011. 334(5-6): p. 393-402. 2. Mueller, M.F.G.B.a.G., Biodiversity of Fungi: Inventory and Monitoring Methods. Imprint: Academic Press, 2004. 3. Wubah, D.A., D.E. Akin, and W.S. Borneman, Biology, fiber-degradation, and enzymology of anaerobic zoosporic fungi. Crit Rev Microbiol, 1993. 19(2): p. 99-115. 4. Velez, C.G., et al., Molecular phylogenetic and zoospore ultrastructural analyses of Chytridium olla establish the limits of a monophyletic Chytridiales. Mycologia, 2011. 103(1): p. 118-30. 5. Ferreira, J.A., et al., Zygomycetes-based biorefinery: present status and future prospects. Bioresour Technol, 2013. 135: p. 523-32. 6. Suh, S.O., et al., Phylogenetics of Saccharomycetales, the ascomycete yeasts. Mycologia, 2006. 98(6): p. 1006-17. 7. Galagan, J.E., et al., Genomics of the fungal kingdom: insights into eukaryotic biology. Genome Res, 2005. 15(12): p. 1620-31. 8. Kirk, P.M., Cannon, P.F., David, J.C. and Stalpers, J.A., Ainsworth & Bisby's Dictionary of the Fungi, 9th Edition. CABI Publishing, 2001. 9. Kurtzman, C. and C. Robnett, Phylogenetic relationships among yeasts of the ? complex? determined from multigene sequence analyses. FEMS Yeast Research, 2003. 3(4): p. 417-432. 10. Gonzalez, S.S., et al., Natural hybrids from Saccharomyces cerevisiae, Saccharomyces bayanus and Saccharomyces kudriavzevii in wine fermentations. FEMS Yeast Res, 2006. 6(8): p. 1221-34. 11. Stefanini, I., et al., Role of social wasps in Saccharomyces cerevisiae ecology and evolution. Proc Natl Acad Sci U S A, 2012. 109(33): p. 13398-403. 12. Hittinger, C.T., Saccharomyces diversity and evolution: a budding model genus. Trends Genet, 2013. 29(5): p. 309-17. 13. Gangl, H., et al., Exceptional fermentation characteristics of natural hybrids from Saccharomyces cerevisiae and S. kudriavzevii. N Biotechnol, 2009. 25(4): p. 244-51. 14. Gibson, B.R., et al., Comparative physiology and fermentation performance of Saaz and Frohberg lager yeast strains and the parental species Saccharomyces eubayanus. Yeast, 2013. 30(7): p. 255-66. 15. Walther, A., A. Hesselbart, and J. Wendland, Genome sequence of Saccharomyces carlsbergensis, the world's first pure culture lager yeast. G3 (Bethesda), 2014. 4(5): p. 783-93. 16. John E. Bradbury, K.D.R., Heather A. Niederer, Soon A. Lee, P. Rod Dunbar, Richard C. Gardner, A homozygous diploid subset of commercial wine yeast strains. Antonie van Leeuwenhoek, 2006. 89(1): p. 27-37. 17. Gonzalez, S.S., E. Barrio, and A. Querol, Molecular characterization of new natural hybrids of Saccharomyces cerevisiae and S. kudriavzevii in brewing. Appl Environ Microbiol, 2008. 74(8): p. 2314-20. 18. Gerton, J.L., et al., Global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae. Proc Natl Acad Sci U S A, 2000. 97(21): p. 11383- 90. 19. Hawthorne D, P.P., Genetic and molecular analysis of hybrids in the genus Saccharomyces involving S. cerevisiae, S. uvarum and a new species, S. douglasii. Yeast, 2000. 10(10). 20. Hunter, N., et al., The mismatch repair system contributes to meiotic sterility in an interspecific yeast hybrid. EMBO J, 1996. 15(7): p. 1726-33.

21. Naumov, G.I., E.S. Naumova, and A. Querol, Genetic Study of Natural Introgression Supports Delimitation of Biological Species in the Saccharomyces Sensu Stricto Complex. Systematic and Applied Microbiology, 1997. 20(4): p. 595-601. 22. Donalies, U.E., et al., Improvement of Saccharomyces yeast strains used in brewing, wine making and baking. Adv Biochem Eng Biotechnol, 2008. 111: p. 67-98. 23. McGovern, P.E., et al., Fermented beverages of pre- and proto-historic China. Proc Natl Acad Sci U S A, 2004. 101(51): p. 17593-8. 24. Horsford, E.N., Report on Vienna Bread. In U.S. Commission to the Vienna exhibition, 1873. Reports. 1875-76. v.2, B Volume 2 of Reports, U.S. Commission to the Vienna exhibition, 1873. 1875, U.S. Government Printing Office, 1875. p. 122. 25. Barnett, J.A., A history of research on yeasts 2: Louis Pasteur and his contemporaries, 1850-1880. Yeast, 2000. 16(8): p. 755-71. 26. Barnett, J.A. and F.W. Lichtenthaler, A history of research on yeasts 3: Emil Fischer, Eduard Buchner and their contemporaries, 1880-1900. Yeast, 2001. 18(4): p. 363-388. 27. Kresge, N., R.D. Simoni, and R.L. Hill, Otto Fritz Meyerhof and the elucidation of the glycolytic pathway. J Biol Chem, 2005. 280(4): p. e3. 28. Nicole Kresge, R.D.S., and Robert L. Hill, JBC Historical Perspectives: Glycobiology and Carbohydrates. The Journal of Biological Chemistry, 2002. 277(16). 29. De Deken, R.H., The Crabtree effect: a regulatory system in yeast. J Gen Microbiol, 1966. 44(2): p. 149-56. 30. van den Brink, J., et al., Energetic limits to metabolic flexibility: responses of Saccharomyces cerevisiae to glucose-galactose transitions. Microbiology, 2009. 155(4): p. 1340-1350. 31. Sun, G., et al., A eudicot from the Early Cretaceous of China. Nature, 2011. 471(7340): p. 625-8. 32. Dashko, S., et al., Why, when, and how did yeast evolve alcoholic fermentation? FEMS Yeast Res, 2014. 33. Goffeau, A., et al., Life with 6000 genes. Science, 1996. 274(5287): p. 546, 563-7. 34. Tugendreich, S., et al., Genes conserved in yeast and humans. Hum Mol Genet, 1994. 3 Spec No: p. 1509-17. 35. Foury, F., Human genetic diseases: a cross-talk between man and yeast. Gene, 1997. 195(1): p. 1-10. 36. Bostein, D. Why yeast? 1991; Available from: https://www.princeton.edu/genomics/botstein/publications/1991_Botstein_Why.pdf. 37. Mattiazzi, M., U. Petrovic, and I. Krizaj, Yeast as a model eukaryote in toxinology: a functional genomics approach to studying the molecular basis of action of pharmacologically active molecules. Toxicon, 2012. 60(4): p. 558-71. 38. Kurtzman, C.P., Recognition of Yeast Species from Gene Sequence Comparisons. The Open Applied Informatics Journal, 2011. 5(Suppl 1-M4): p. 20-29. 39. JM., G., Yeast carbon catabolite repression. Microbiol Mol Biol Rev. , 1998. 62(2): p. 334-61. 40. Leandro, M.J., C. Fonseca, and P. Goncalves, Hexose and pentose transport in ascomycetous yeasts: an overview. FEMS Yeast Res, 2009. 9(4): p. 511-25. 41. Filip Rolland, J.W., Johan M. Thevelein, Glucose-sensing and -signalling mechanisms in yeast. FEMS Yeast Research, 2002. 2: p. 183-201. 42. Serrano, R., Energy requirements for maltose transport in yeast. Eur J Biochem, 1977. 80(1): p. 97-102. 43. Sr|an Novak, V.Z.-K.a.V.M., Regulation of Maltose Transport and Metabolism in Saccharomyces cerevisiae. Food Technol. Biotechnol., 2004. 42(3): p. 213–218. 44. Gancedo, C.S., R. , Energy-yielding metabolism. In The Yeasts. New York : Academic Press, 1989. 3: p. 205-259.

45. Kresze GB, R.H., Pyruvate dehydrogenase complex from baker's yeast. 2. Molecular structure, dissociation, and implications for the origin of mitochondria. Eur J Biochem., 1991. 119(3): p. 581-7. 46. Lutstorf, U. and R. Megnet, Multiple forms of alcohol dehydrogenase in Saccharomyces cerevisiae. I. Physiological control of ADH-2 and properties of ADH-2 and ADH-4. Arch Biochem Biophys, 1968. 126(3): p. 933-44. 47. Jacobson, M.K. and C. Bernofsky, Mitochondrial acetaldehyde dehydrogenase from Saccharomyces cerevisiae. Biochim Biophys Acta, 1974. 350(2): p. 277-91. 48. Conant, G.C. and K.H. Wolfe, Increased glycolytic flux as an outcome of whole-genome duplication in yeast. Mol Syst Biol, 2007. 3: p. 129. 49. A., B., Eremotheciuim cymbalariae, nuovoascomicete. Nuovo giornale botanico italiano, 1888. 20: p. 452-456. 50. Wendland, J. and A. Walther, Genome evolution in the eremothecium clade of the Saccharomyces complex revealed by comparative genomics. G3 (Bethesda), 2011. 1(7): p. 539-48. 51. Ashby, S.F.a.N.W., The fungi of stigmatomycosis. Ann Bot, 1926. 40(1). 52. EC, H., Grundlagen zur Systematik der Saccharomyceten. Zbl Bakt II Natur., 1904. 12: p. 529-538. 53. Kurtzman, C.P., Relationships among the genera Ashbya, Eremothecium, Holleya and Nematospora determined from rDNA sequence divergence. J Ind Microbiol, 1995. 14(6): p. 523-30. 54. Prillinger, H., et al., Phytopathogenic filamentous (Ashbya, Eremothecium) and dimorphic fungi (Holleya, Nematospora) with needle-shaped ascospores as new members within the Saccharomycetaceae. Yeast, 1997. 13(10): p. 945-60. 55. Guilliermond, C. r. hebd. Séanc Acad. Sci., Paris, 1935. 200(19): p. 1556. 56. Dietrich, F.S., et al., Genomes of Ashbya fungi isolated from insects reveal four mating- type loci, numerous translocations, lack of transposons, and distinct gene duplications. G3 (Bethesda), 2013. 3(8): p. 1225-39. 57. Wendland, J. and A. Walther, Ashbya gossypii: a model for fungal developmental biology. Nat Rev Microbiol, 2005. 3(5): p. 421-9. 58. Kato, T. and E.Y. Park, Riboflavin production by Ashbya gossypii. Biotechnol Lett, 2012. 34(4): p. 611-8. 59. Stahmann K. P., R.J.L., Seulberger H, Three biotechnical processes using Ashbya gossypii, Candida famata, or Bacillus subtilis compete with chemical riboflavin production. Appl. Microbiol. Biotechnol., 2000. 53: p. 509–516. 60. Demain, A.L., Riboflavin oversynthesis. Annu Rev Microbiol, 1972. 26: p. 369-88. 61. Tijen Özbaş, T.K., Effects of growth factors on riboflavin production by Ashbya gossypii. Enzyme and Microbial Technology, 1991. 13(7): p. 594-596. 62. Özbas, T. and T. Kutsal, Comparative study of riboflavin production from two microorganisms: Eremothecium ashbyii and Ashbya gossypii. Enzyme and Microbial Technology, 1986. 8(10): p. 593-596. 63. Tajima, S., et al., Increased riboflavin production from activated bleaching earth by a mutant strain of Ashbya gossypii. J Biosci Bioeng, 2009. 108(4): p. 325-9. 64. Steiner, S., et al., Homologous recombination as the main mechanism for DNA integration and cause of rearrangements in the filamentous ascomycete Ashbya gossypii. Genetics, 1995. 140(3): p. 973-87. 65. Wendland, J., et al., PCR-based gene targeting in the filamentous fungus Ashbya gossypii. Gene, 2000. 242(1-2): p. 381-391. 66. Dietrich, F.S., et al., The Ashbya gossypii genome as a tool for mapping the ancient Saccharomyces cerevisiae genome. Science, 2004. 304(5668): p. 304-7.

67. Wasserstrom, L., et al., Molecular determinants of sporulation in Ashbya gossypii. Genetics, 2013. 195(1): p. 87-99. 68. Chambers, P.J. and I.S. Pretorius, Fermenting knowledge: the history of winemaking, science and yeast research. EMBO Rep, 2010. 11(12): p. 914-20. 69. Porro, D. and P. Branduardi, Yeast cell factory: fishing for the best one or engineering it? Microb Cell Fact, 2009. 8: p. 51. 70. Adrio, J.L. and A.L. Demain, Fungal biotechnology. Int Microbiol, 2003. 6(3): p. 191-9. 71. Advances in Fungal Biotechnology for Industry, Agriculture, and Medicine. 2004. 72. Zhong, J.J. and J.H. Xiao, Secondary metabolites from higher fungi: discovery, bioactivity, and bioproduction. Adv Biochem Eng Biotechnol, 2009. 113: p. 79-150. 73. Gravina, S.A., G.L. Yep, and M. Khan, Human biology of taste. Ann Saudi Med, 2013. 33(3): p. 217-22. 74. Owen R Fenema, P., Fod aditves-an unending controversy. Am J Clin Nutr, 1987. 46(1): p. 201-203. 75. Terr, G.a.F.o.C., Preserving Food Without Freezing or Canning: Traditional Techniques Using Salt, Oil, Sugar, Alcohol, Vinegar, Drying, Cold Storage, and Lactic Fermentation. 2007. 197. 76. Flavourings and certain food ingredients with flavouring properties for use in and on foods. Official Journal of the European Unions, 2008: p. L 354/34. 77. Geueke, B., C.C. Wagner, and J. Muncke, Food contact substances and chemicals of concern: a comparison of inventories. Food Addit Contam Part A Chem Anal Control Expo Risk Assess, 2014. 31(8): p. 1438-50. 78. ; Available from: http://www.fda.gov. 79. Vandamme, E.J., Bioflavours and fragrances via fungi and their enzymes. Fungal Diversity, 2003. 13: p. 153-166. 80. Berger, R.G., Biotechnology of flavours--the next generation. Biotechnol Lett, 2009. 31(11): p. 1651-9. 81. Saerens, S.M., et al., Parameters affecting ethyl ester production by Saccharomyces cerevisiae during fermentation. Appl Environ Microbiol, 2008. 74(2): p. 454-61. 82. Paul L.H. MCSWEENEY, M.J.S., Biochemical pathways for the production of flavour compounds in cheeses during ripening: A review. Lait 80, 1999. 2000: p. 293–324. 83. Ehrlich, F., Über die Bedingungen der Fuselölbildung und über ihren Zusammenhang mit dem Eiweißaufbau der Hefe. Berichte der deutschen chemischen Gesellschaft, 1907. 40(1): p. 1027-1047. 84. Lampitt, L.H., Nitrogen Metabolism in Saccharomyces cerevisiae. Biochem J, 1919. 13(4): p. 459-86. 85. M, Y., Decomposition of amino acids by yeast I. Nippon Nogeikagaku Kaishi 1932. 8: p. 428-432. 86. S., T.R., The growth and fermentation of a strain of S. cerevisiae with amino acids as nutrients. J. Inst. Brew, 1941. 47(5): p. 255–272. 87. RS., T., Mechanism of nitrogen assimilation from amino-acids by yeast. Nature, 1949. 164(4165). 88. Hazelwood, L.A., et al., The Ehrlich pathway for fusel alcohol production: a century of research on Saccharomyces cerevisiae metabolism. Appl Environ Microbiol, 2008. 74(8): p. 2259-66. 89. Scheffer, J.P.v.D.a.A., Redox balances in the metabolism of sugars by yeasts. FEMS Microbiology Letters, 1986. 32(3-4): p. 199-224. 90. Boulton Roger B., S.V.L., Linda F. Bisson, Principles and Practices of Winemaking. The Chapman & Hall Enology Library, 1995.

91. Vollbrecht D, R.F., The formation of higher alcohols by amino acid auxotrophic mutants of Saccharomyces cerevisiae. I. The conversion of amino acids to higher alcohols. Arch Mikrobiol., 1973. 94(4): p. 351-358. 92. Davis, T.S., K. Boundy-Mills, and P.J. Landolt, Volatile emissions from an epiphytic fungus are semiochemicals for eusocial wasps. Microb Ecol, 2012. 64(4): p. 1056-63. 93. Kispal, G., et al., Mitochondrial and cytosolic branched-chain amino acid transaminases from yeast, homologs of the myc oncogene-regulated Eca39 protein. J Biol Chem, 1996. 271(40): p. 24458-64. 94. Lilly, M., et al., The effect of increased branched-chain amino acid transaminase activity in yeast on the production of higher alcohols and on the flavour profiles of wine and distillates. FEMS Yeast Res, 2006. 6(5): p. 726-43. 95. Dickinson, J.R., L.E. Salgado, and M.J. Hewlins, The catabolism of amino acids to long chain and complex alcohols in Saccharomyces cerevisiae. J Biol Chem, 2003. 278(10): p. 8028-34. 96. Eden, E., et al., Discovering motifs in ranked lists of DNA sequences. PLoS Comput Biol, 2007. 3(3): p. e39. 97. Iraqui, I., et al., Transcriptional induction by aromatic amino acids in Saccharomyces cerevisiae. Mol Cell Biol, 1999. 19(5): p. 3360-71. 98. Zhu, C., et al., High-resolution DNA-binding specificity analysis of yeast transcription factors. Genome Res, 2009. 19(4): p. 556-66. 99. Lee, T.I., et al., Transcriptional regulatory networks in Saccharomyces cerevisiae. Science, 2002. 298(5594): p. 799-804. 100. Hazelwood, L.A., et al., A new physiological role for Pdr12p in Saccharomyces cerevisiae: export of aromatic and branched-chain organic acids produced in amino acid catabolism. FEMS Yeast Res, 2006. 6(6): p. 937-45. 101. Fox, P.F., J.A. Lucey, and T.M. Cogan, Glycolysis and related reactions during cheese manufacture and ripening. Crit Rev Food Sci Nutr, 1990. 29(4): p. 237-53. 102. Duong, C.T., et al., Identification of Sc-type ILV6 as a target to reduce diacetyl formation in lager brewers' yeast. Metab Eng, 2011. 13(6): p. 638-47. 103. Dickinson, J.R. and V. Norte, A study of branched-chain amino acid aminotransferase and isolation of mutations affecting the catabolism of branched-chain amino acids in Saccharomyces cerevisiae. FEBS Lett, 1993. 326(1-3): p. 29-32. 104. Kern, K., et al., Isoamyl alcohol-induced morphological change in Saccharomyces cerevisiae involves increases in mitochondria and cell wall chitin content. FEMS Yeast Res, 2004. 5(1): p. 43-9. 105. Chen, H. and G.R. Fink, Feedback control of morphogenesis in fungi by aromatic alcohols. Genes Dev, 2006. 20(9): p. 1150-61. 106. Martins, M., et al., Morphogenesis control in Candida albicans and Candida dubliniensis through signaling molecules produced by planktonic and biofilm cells. Eukaryot Cell, 2007. 6(12): p. 2429-36. 107. Chen H, F.M., Feng Q, Clardy J, Fink GR., Tyrosol is a quorum-sensing molecule in Candida albicans. Proc Natl Acad Sci U S A., 2004. 101(14): p. 5048-52. 108. Hornby JM, J.E., Lisec AD, Tasto JJ, Jahnke B, Shoemaker R, Dussault P, Nickerson KW., Quorum sensing in the dimorphic fungus Candida albicans is mediated by farneso. Appl Environ Microbiol., 2001. 67(7): p. 2982-92. 109. Kruppa, M., Quorum sensing and Candida albicans. Mycoses, 2009. 52(1): p. 1-10. 110. Herrero-Garcia, E., et al., 8-Carbon oxylipins inhibit germination and growth, and stimulate aerial conidiation in Aspergillus nidulans. Fungal Biol, 2011. 115(4-5): p. 393- 400. 111. Hung, R., S. Lee, and J.W. Bennett, The effects of low concentrations of the enantiomers of mushroom alcohol (1-octen-3-ol) on. Mycology, 2014. 5(2): p. 73-80.

112. Schlyter, Q.-H.Z.a.F., Olfactory recognition and behavioural avoidance of angiosperm nonhost volatiles by conifer-inhabiting bark beetles. Agricultural and Forest Entomology, 2004. 6: p. 1-19. 113. Kline DL, A.S., Bernier UR, Welch CH, Evaluation of the enantiomers of 1-octen-3-ol and 1-octyn-3-ol as attractants for mosquitoes associated with a freshwater swamp in Florida, U.S.A. Med Vet Entomol, 2007. 21(4): p. 323-31. 114. Bohbot, J.D. and J.C. Dickens, Characterization of an enantioselective odorant receptor in the yellow fever mosquito Aedes aegypti. PLoS One, 2009. 4(9): p. e7032. 115. Lam, K., et al., Semiochemical-mediated oviposition avoidance by female house flies, Musca domestica, on animal feces colonized with harmful fungi. J Chem Ecol, 2010. 36(2): p. 141-7. 116. Niu, X.M., et al., Nematodetoxic aurovertin-type metabolites from a root-knot nematode parasitic fungus Pochonia chlamydosporia. J Agric Food Chem, 2010. 58(2): p. 828-34. 117. Sijam, M.M.T.a.K., Arbuscular mycorrhizal fungi and plant root exudates bio- communications in the rhizosphere. African Journal of Microbiology Research, 2012. 6(46): p. 7295-7301. 118. Strobel, G., et al., An endophytic/pathogenic Phoma sp. from creosote bush producing biologically active volatile compounds having fuel potential. FEMS Microbiol Lett, 2011. 320(2): p. 87-94. 119. Minerdi, D., et al., Volatile organic compounds: a potential direct long-distance mechanism for antagonistic action of Fusarium oxysporum strain MSA 35. Environ Microbiol, 2009. 11(4): p. 844-54. 120. Mercier, J. and D.C. Manker, Biocontrol of soil-borne diseases and plant growth enhancement in greenhouse soilless mix by the volatile-producing fungus Muscodor albus. Crop Protection, 2005. 24(4): p. 355-362. 121. Mercier, J. and J.I. Jimenez, Potential of the volatile-producing fungus Muscodor albus for control of building molds. Can J Microbiol, 2007. 53(3): p. 404-10. 122. Mitchell, A.M., et al., Volatile antimicrobials from Muscodor crispans, a novel endophytic fungus. Microbiology, 2010. 156(Pt 1): p. 270-7. 123. Zhang, C.L., et al., Muscodor fengyangensis sp. nov. from southeast China: morphology, physiology and production of volatile compounds. Fungal Biol, 2010. 114(10): p. 797-808. 124. Strobel, G.A., et al., Volatile antimicrobials from Muscodor albus, a novel endophytic fungus. Microbiology, 2001. 147(Pt 11): p. 2943-50. 125. Singh, S.K., et al., An endophytic Phomopsis sp. possessing bioactivity and fuel potential with its volatile organic compounds. Microb Ecol, 2011. 61(4): p. 729-39. 126. Chambers, S.T., A. Scott-Thomas, and M. Epton, Developments in novel breath tests for bacterial and fungal pulmonary infection. Curr Opin Pulm Med, 2012. 18(3): p. 228-32. 127. Zhu, J., et al., Detecting bacterial lung infections: in vivo evaluation of in vitro volatile fingerprints. J Breath Res, 2013. 7(1): p. 016003. 128. Ngugi, H.K. and H. Scherm, Mimicry in plant-parasitic fungi. FEMS Microbiol Lett, 2006. 257(2): p. 171-6. 129. Schiestl, F.P., et al., Evolution of 'pollinator'- attracting signals in fungi. Biol Lett, 2006. 2(3): p. 401-4. 130. Steinebrunner, F., F.P. Schiestl, and A. Leuchtmann, Ecological role of volatiles produced by Epichloe: differences in antifungal toxicity. FEMS Microbiol Ecol, 2008. 64(2): p. 307-16. 131. Daisy BH, S.G., Castillo U, Ezra D, Sears J, Weaver DK, Runyon JB., Naphthalene, an insect repellent, is produced by Muscodor vitigenus, a novel endophytic fungus. Microbiology, 2002. 148(Pt 11): p. 3737-41.

132. Rude, M.A. and A. Schirmer, New microbial fuels: a biotech perspective. Curr Opin Microbiol, 2009. 12(3): p. 274-81. 133. Griffin, M.A., et al., Volatile organic compound production by organisms in the genus Ascocoryne and a re-evaluation of myco-diesel production by NRRL 50072. Microbiology, 2010. 156(Pt 12): p. 3814-29. 134. Tomsheck, A.R., et al., Hypoxylon sp., an endophyte of Persea indica, producing 1,8- cineole and other bioactive volatiles with fuel potential. Microb Ecol, 2010. 60(4): p. 903-14. 135. Gianoulis, T.A., et al., Genomic analysis of the hydrocarbon-producing, cellulolytic, endophytic fungus Ascocoryne sarcoides. PLoS Genet, 2012. 8(3): p. e1002558. 136. Jelen, H.H., Use of solid phase microextraction (SPME) for profiling fungal volatile metabolites. Letters in Applied Microbiology, 2003. 36: p. 263-267. 137. Stoppacher, N., et al., Identification and profiling of volatile metabolites of the biocontrol fungus Trichoderma atroviride by HS-SPME-GC-MS. J Microbiol Methods, 2010. 81(2): p. 187-93. 138. Godefroot, M., P. Sandra, and M. Verzele, New method for quantitative essential oil analysis. Journal of Chromatography A, 1981. 203: p. 325-335. 139. Tait, E., et al., Identification of volatile organic compounds produced by bacteria using HS-SPME-GC-MS. J Chromatogr Sci, 2014. 52(4): p. 363-73. 140. Urrestarazu, A., et al., Phenylalanine- and tyrosine-auxotrophic mutants of Saccharomyces cerevisiae impaired in transamination. Mol Gen Genet, 1998. 257(2): p. 230-7. 141. Iraqui, I., et al., Characterisation of Saccharomyces cerevisiae ARO8 and ARO9 genes encoding aromatic aminotransferases I and II reveals a new aminotransferase subfamily. Mol Gen Genet, 1998. 257(2): p. 238-48. 142. Kradolfer, P., P. Niederberger, and R. Hutter, Tryptophan degradation in Saccharomyces cerevisiae: characterization of two aromatic aminotransferases. Arch Microbiol, 1982. 133(3): p. 242-8. 143. Derrick, S. and P.J. Large, Activities of the enzymes of the Ehrlich pathway and formation of branched-chain alcohols in Saccharomyces cerevisiae and Candida utilis grown in continuous culture on valine or ammonium as sole nitrogen source. J Gen Microbiol, 1993. 139(11): p. 2783-92. 144. Vuralhan, Z., et al., Physiological characterization of the ARO10-dependent, broad- substrate-specificity 2-oxo acid decarboxylase activity of Saccharomyces cerevisiae. Appl Environ Microbiol, 2005. 71(6): p. 3276-84. 145. Romagnoli, G., et al., Substrate specificity of thiamine pyrophosphate-dependent 2-oxo- acid decarboxylases in Saccharomyces cerevisiae. Appl Environ Microbiol, 2012. 78(21): p. 7538-48. 146. Lee, K. and J.S. Hahn, Interplay of Aro80 and GATA activators in regulation of genes for catabolism of aromatic amino acids in Saccharomyces cerevisiae. Mol Microbiol, 2013. 88(6): p. 1120-34. 147. Encyclopedia of Food Microbiology, ed. C.B. Tortorello. 2014: Academic Press. 2955 148. Buzzini, P., et al., A study on volatile organic compounds (VOCs) produced by tropical ascomycetous yeasts. Antonie Van Leeuwenhoek, 2003. 84(4): p. 301-11. 149. Jyoti Prakash Tamang, K.K., Fermented foods and beverages of the world. CRC Press, 2010. 150. Kim, B., B.R. Cho, and J.S. Hahn, Metabolic engineering of Saccharomyces cerevisiae for the production of 2-phenylethanol via Ehrlich pathway. Biotechnol Bioeng, 2014. 111(1): p. 115-24.

151. Vuralhan, Z., et al., Identification and characterization of phenylpyruvate decarboxylase genes in Saccharomyces cerevisiae. Appl Environ Microbiol, 2003. 69(8): p. 4534-41. 152. Smit, B.A., W.J. Engels, and G. Smit, Branched chain aldehydes: production and breakdown pathways and relevance for flavour in foods. Appl Microbiol Biotechnol, 2009. 81(6): p. 987-99. 153. Casalone, E., et al., Identification by functional analysis of the gene encoding alpha- isopropylmalate synthase II (LEU9) in Saccharomyces cerevisiae. Yeast, 2000. 16(6): p. 539-45. 154. Chang LF, G.P., Kohlhaw GB., Total deletion of yeast LEU4: further evidence for a second alpha-isopropylmalate synthase and evidence for tight LEU4-MET4 linkage. Gene, 1985. 33(3): p. 333-9. 155. Huuskonen, A., et al., Selection from industrial lager yeast strains of variants with improved fermentation performance in very-high-gravity worts. Appl Environ Microbiol, 2010. 76(5): p. 1563-73. 156. Pretorius, M.d.T.a.I.S., Microbial Spoilage and Preservation of Wine: Using Weapons from Nature's Own Arsenal- A Review S. Afr. J. Enol. Vitic, 2000. 21(Special Issue): p. 74-96. 157. Peter Piper, C.O.C., † Kostas Hatzixanthis and Mehdi Mollapour, Weak acid adaptation: the stress response that confers yeasts with resistance to organic acid food preservatives. Microbiology, 2001. 147: p. 2635–2642. 158. Van Aelst, L., et al., Molecular cloning of a gene involved in glucose sensing in the yeast Saccharomyces cerevisiae. Mol Microbiol, 1993. 8(5): p. 927-43. 159. Nordström, K., Formation of ethyl acetate in fermentation with Brewer's yeast. Journal of the Institute of Brewing, 1961. 67: p. 173-181. 160. Yoshioka, K.H., N., Ester formation by alcohol acetyltransferase from brewer's yeast. Agricultural and Biological Chemistry, 1981. 45: p. 2183-2190. 161. Rojas, V., Gil, J.V., Manzanares, P., Gavara, R., Piñaga, F. & Flors, A., Measurement of alcohol acetyltransferase and ester hydrolase activities in yeast extract. Enzyme and Microbial Technology, 2002. 30: p. 224-230. 162. Druvefors UA, P.V., Schnürer J, Nutrient effects on biocontrol of Penicillium roqueforti by Pichia anomala J121 during airtight storage of wheat. Appl Environ Microbiol., 2005. 71(4): p. 1865-9. 163. Cao, X., et al., Genome shuffling of Hansenula anomala to improve flavour formation of soy sauce. World J Microbiol Biotechnol, 2012. 28(5): p. 1857-62. 164. Crafack, M., et al., Influencing cocoa flavour using Pichia kluyveri and Kluyveromyces marxianus in a defined mixed starter culture for cocoa fermentation. Int J Food Microbiol, 2013. 167(1): p. 103-16. 165. Amaya-Delgado, L., et al., Performance evaluation of Pichia kluyveri, Kluyveromyces marxianus and Saccharomyces cerevisiae in industrial tequila fermentation. World J Microbiol Biotechnol, 2013. 29(5): p. 875-81. 166. Breuer, U. and H. Harms, Debaryomyces hansenii--an extremophilic yeast with biotechnological potential. Yeast, 2006. 23(6): p. 415-37. 167. Prista C, A.A., Loureiro-Dias MC, Ramos J., Physiological basis for the high salt tolerance of Debaryomyces hanseni. Appl Environ Microbiol. , 1997. 63(10): p. 4005-9. 168. van den Tempel, T. and M. Jakobsen, The technological characteristics of Debaryomyces hansenii and Yarrowia lipolytica and their potential as starter cultures for production of Danablu. International Dairy Journal, 2000. 10(4): p. 263-270. 169. Flores, M., et al., Effect of Debaryomyces spp. on aroma formation and sensory quality of dry-fermented sausages. Meat Sci, 2004. 68(3): p. 439-46.

170. Montel, M.C., et al., Biochemical activities of Micrococcaceae and their effects on the aromatic profiles and odours of a dry sausage model. Food Microbiology, 1996. 13(6): p. 489-499.

Paper 1

This paper investigated the regulatory network of the ARO gene family in S. cerevisiae. The ARO gene set expresses enzymes that belong to the Ehrlich pathway, producing flavoured alcohols and esters by the amino acid catabolism. ScARO80 is a Zn2Cys6 transcription factor and activator of the pathway. The study investigated the ARO80-dependent expression of the ARO target promoter regions. ScARO9 expression is strongly regulated by ScAro80. This established ScARO9 as a potential reporter to correlate the gene activation and flavour formation in yeast. To determine the applicability of the ARO9-lacZ system as a screening tool we transformed various strains with the ARO9p-lacZ plasmid. GC-MS results confirmed the ability of the lacZ- system to predict the yield of two flavour compounds, 2-phenylethanol and 2- phenylacetate.

OPEN An indirect assay for volatile compound

SUBJECT AREAS: production in yeast strains MICROBIOLOGY Davide Ravasio1, Andrea Walther1, Kajetan Trost2, Urska Vrhovsek2 &Ju¨rgen Wendland1 TECHNIQUES FUNGAL GENETICS 1Carlsberg Laboratory; Yeast Genetics Gamle Carlsberg Vej 10 DK-1799 Copenhagen V, Denmark, 2Fondazione Edmund Mach Research and Innovation Centre Food Quality and Nutrition Department Via E.Mach 1, I-38010 S.Michele all’Adige, Italy. Received 9 August 2013 Traditional flavor analysis relies on gas chromatography coupled to mass spectrometry (GC-MS) methods. Accepted Here we describe an indirect method coupling volatile compound formation to an ARO9-promoter-LacZ 6 December 2013 reporter gene. The resulting b-galactosidase activity correlated well with headspace solid phase micro extraction (HS/SPME) GC-MS data, particularly with respect to the formation of rose flavor. This tool Published enables large-scale screening of yeast strains and their progeny to identify the most flavor active strains. 15 January 2014

he organoleptic perception of beer depends mainly on substances produced by yeast during the fermentation process. Flavor active substances are volatile compounds and include higher alcohols, esters, and fatty acids. Correspondence and In the wine industry attempts are made to increase flavour compounds by either simultaneous or sequential requests for materials T co-fermentations using either different yeast strains, i.e Saccharomyces cerevisiae with a non-Saccharomyces yeast, should be addressed to or mixing bacterial strains, e.g. Oenococcus oeni, with wine yeasts5,8,14,17. Research interest in natural flavors J.W. (juergen. produced by yeasts has gained increasing interest, particularly focusing on isoamyl alcohol (banana flavor) wendland@ and b-phenylethanol (flowery, rose flavor). Both compounds are produced during amino acid catabolism in carlsberglab.dk) yeast9. The Ehrlich pathway, a linear pathway requiring three enzymatic activities, is responsible for converting aromatic amino acids (phenylalanine, tyrosine, and tryptophan), branched-chain amino acids (leucine, isoleucine, and valine) and methionine into higher alcohols. The regulation of the Ehrlich pathway depends at least in part on the Zn2Cys6 transcription factor Aro80, which regulates ARO9 and ARO10 in a nitrogen source dependent manner (Fig. 1A)12. One of the key bottle necks in flavor research is the requirement of chemical analytical tools to measure volatile compounds produced during fermentation, which is generally done using HS/SPME extraction methods coupled to GC-MS2,15. This method, however, is time consuming, requires additional quant- itation as well as prior lab scale fermentations and sample preparations, which are often difficult to optimize for high throughput screening. In order to identify a promoter that is most responsive to ARO80 overexpression, we co-transformed ARO80 under the control of the Ashbya gossypii TEF-promoter with plasmids containing ARO8, ARO9, ARO10, and ARO80 promoter-lacZ reporter gene fusions into S. cerevisiae (Fig. 1B). To investigate whether expression of the reporter genes was actually Aro80-dependent we quantified b-galactosidase activity in strains bearing the endo- genous ARO80,anARO80 deletion, or the ARO80 overexpression construct (Fig. 1C). This established the ARO9 as a potential reporter for a strain’s flavor production. To correlate ARO9 reporter gene activity with flavor formation we first determined its activity in a set of strains with S. cerevisiae background expressing ARO80 at wild type levels. This included the laboratory strain CENPK, two hybrid lager yeast strains, collectively known as S. pastorianus as well as a Bordeaux wine yeast. For comparison we used these strains in bench-top fermentation assays and at the end of fermentation volatiles were extracted by HS/SPME and analyzed via GC-MS (Tab. S1). For the comparison of volatile compound formation with b-galactosidase activity we focused our attention to phenylalanine catabolites (rose flavor). This showed that b-galactosidase activity of the ARO9-lacZ reporter correlated well with the amount of b-phenylacetate and b- phenylethanol produced by these strains (Fig. 2). To determine the applicability of this tool beyond S. cerevisiae we used the ARO9-reporter with strains from the Saccharomyces sensu stricto complex including S. bayanus, S. cariocanus, S. eubayanus, S. kudriavzevii, S. mikatae, S. paradoxus,andS. uvarum (Fig. 2). The flavor profiles show that there is a great variability in volatile formation between these strains (Tab. S1). This variability is also reflected in the b-galactosidase activity in these strains indicating that high b-galactosidase activity pairs with increased flavor production. A correlation curve was analyzed comparing b-galactosidase activity with the combined flavor values for 2-phenyl ethanol and 2-phenyl acetate (Fig. 2C). This took into account that Aro9 enzymatic activity is upstream of 2-phenyl ethanol and 2-phenyl acetate production.

SCIENTIFIC REPORTS | 4 : 3707 | DOI: 10.1038/srep03707 1 www.nature.com/scientificreports

Figure 1 | Identification of a reporter gene for Ehrlich Pathway activity. (A) Amino acids (branched-chain amino acids, leucine, isoleucine, and valine, aromatic amino acids, phenylalanine, tyrosine, and tryptphan, or methionine) are converted in the Ehrlich pathway to fusel alcohol or fusel acids in a three step process. The genes encoding enzymes that catalyze single steps are indicated. Oxidation of aldehydes to fusel acids is done by aldehyde dehydrogenases (e.g. ALD1). Reduction of aldehydes to fusel alcohols is done by alcohol dehydrogenases (e.g. ADH1). Transcriptional regulation by Aro80 and co-factor requirement is indicated. (B) Plasmids carrying the ARO80 overexpression and one of the ARO-promoter-lacZ reporter gene constructs were co-transformed into S. cerevisiae (BY4741). (C) Quantitative b-galactosidase assay with strains bearing the indicated ARO-promoter-lacZ constructs in strains in which ScARO80 was either overexpressed or deleted, or contained the wildtype ARO80.

Fermented beverages contain only small amounts of volatile com- been elucidated to a great extent18. The Ehrlich pathway plays a pounds; yet, these are of paramount importance for the flavor profile central role in aromatic and branched-chain amino acid catabolism and organoleptic perception of a beverage19,20. Changes in brewing resulting in the conversion of amino acids to aroma compounds9. technology, e.g. introduction of high-gravity brewing, can drastically Several studies have described an increase in flavor production by alter the flavor composition - in this case - by resulting in an increase selecting for yeast strains resistant to fluoro-amino acids. An in the amount of acetate esters. Consumer preference is towards all increased production of isoamyl alcohol, for example, can be natural flavors and unique flavor signatures10. Based on this non- achieved by selecting mutants resistant to trifluoroleucine3. In such GMO preference, three main roads are currently followed to improve strains a mutation of D578Y in the LEU4 gene releases feedback flavor content of beverages: (i) choice of the starter culture, (ii) mixed inhibition and initiates increased production of leucine and its cat- fermentations using different yeast species or a combination of yeast abolites16. Using a genetic approach it was shown that overexpression and bacterial species, and (iii) selection of strains high in volatile of the alcohol acetyl transferases ATF1 and ATF2 substantially compound formation via yeast breeding approaches1,5,6,22. increased the production of isoamyl acetate20. For example, yeasts belonging to the genera Hanseniaspora and The indirect assay described in this study converts Ehrlich Pichia are good producers of acetate esters, whereas mixed fermenta- pathway activity into a reporter gene readout that can be quan- tions with S. cerevisiae and Lachancea thermotolerans increased the tified as b-galactosidase activity. We base the tool on the ARO9 level of b-phenylethanol4,21. Furthermore, mixed fermentations, promoter as the ARO8 promoter was not responsive to Aro80 and including S. cerevisiae and a bacterial strain e.g. Oenococcus oeni, has been shown to be under general control13. With this method promise to provide novel flavor variations17. we can preferably assay rose flavor. Apparently, however, this With the highly advanced gene function analyses in S. cerevisiae reporter is not discriminatory towards branched chain amino the genetic repertoire involved in volatile compound formation has acids (Tab. S1).

SCIENTIFIC REPORTS | 4 : 3707 | DOI: 10.1038/srep03707 2 www.nature.com/scientificreports

Figure 2 | Comparison of b-galactosidase activity with volatile compound formation. (A) Assay with either the indicated S. cerevisiae strains (A) or with Saccharomyces sensu stricto strains (B). Upper panels depict b-galactosidase activity based on the ARO9p-lacZ reporter construct. Lower panels show b- phenylethanol and b-phenylacetate volatile compounds. Note: Fermentation with the wine strain in (A), was done in YPD due to its lack of MAL-genes. The low amount of flavor produced by S. mikatae, S. cariocanus, and S. cerevisiae in (B) is due to their inability to end-ferment granulated malt used in these fermentations. Correlation of -galactosidase activity and the combined yield of phenylalanine catabolites are shown in (C).

Our tool is fast and convenient and can be adapted for use with 12. Iraqui, I., Vissers, S., Andre, B. & Urrestarazu, A. Transcriptional induction by high throughput microtiter plate assays in yeast7,11. Thus this indirect aromatic amino acids in Saccharomyces cerevisiae. Mol Cell Biol 19, 3360–3371 (1999). flavor assay system is inexpensive and allows screening of large lib- 13. Iraqui, I., Vissers, S., Cartiaux, M. & Urrestarazu, A. Characterisation of raries of yeast strains as well as F1/F2 populations of interbred Saccharomyces cerevisiae ARO8 and ARO9 genes encoding aromatic strains. This will lead to the rapid identification of strains with poten- aminotransferases I and II reveals a new aminotransferase subfamily. Mol Gen tially improved flavor characteristics compared to the parental Genet 257, 238–248 (1998). 14. Kim, D. H., Hong, Y. A. & Park, H. D. Co-fermentation of grape must by strains. Additionally, different growth regimes can lead to altered Issatchenkia orientalis and Saccharomyces cerevisiae reduces the malic acid flavor production. This allows the implementation of changes in content in wine. Biotechnol Lett 30, 1633–1638 (2008). oxygen supply and use of different nitrogen sources. 15. Mauriello, G., Capece, A., D’Auria, M., Garde-Cerdan, T. & Romano, P. SPME- GC method as a tool to differentiate VOC profiles in Saccharomyces cerevisiae wine yeasts. Food Microbiol 26, 246–252 (2009). 1. Arroyo-Lopez, F. N. et al. Yeasts in table olive processing: desirable or spoilage 16. Oba, T., Nomiyama, S., Hirakawa, H., Tashiro, K. & Kuhara, S. Asp578 in LEU4p microorganisms? Int J Food Microbiol 160, 42–49 (2012). is one of the key residues for leucine feedback inhibition release in sake yeast. Biosci Biotechnol Biochem 69, 1270–1273 (2005). 2. Canuti, V. et al. Headspace solid-phase microextraction-gas chromatography- mass spectrometry for profiling free volatile compounds in Cabernet Sauvignon 17. Rossouw, D., Du Toit, M. & Bauer, F. F. The impact of co-inoculation with Oenococcus oeni on the trancriptome of Saccharomyces cerevisiae and on the grapes and wines. J Chromatogr A 1216, 3012–3022 (2009). flavour-active metabolite profiles during fermentation in synthetic must. Food 3. Casalone, E. et al. Genetic and biochemical characterization of Saccharomyces Microbiol 29, 121–131 (2012). cerevisiae mutants resistant to trifluoroleucine. Res Microbiol 148, 613–623 18. Styger, G., Jacobson, D., Prior, B. A. & Bauer, F. F. Genetic analysis of the (1997). metabolic pathways responsible for aroma metabolite production by 4. Comitini, F. et al. Selected non-Saccharomyces wine yeasts in controlled Saccharomyces cerevisiae. Appl Microbiol Biotechnol 97, 4429–4442 (2013). multistarter fermentations with Saccharomyces cerevisiae. Food Microbiol 28, 19. Verstrepen, K. J. et al. Flavor-active esters: adding fruitiness to beer. J Biosci Bioeng 873–882 (2011). 96 (2003). 5. Domizio, P. et al. Outlining a future for non-Saccharomyces yeasts: selection of 20. Verstrepen, K. J. et al. Expression levels of the yeast alcohol acetyltransferase genes putative spoilage wine strains to be used in association with Saccharomyces ATF1, Lg-ATF1, and ATF2 control the formation of a broad range of volatile cerevisiae for grape juice fermentation. Int J Food Microbiol 147, 170–180 (2011). esters. Appl Environ Microbiol 69, 5228–5237 (2003). 6. Garcia Sanchez, R., Solodovnikova, N. & Wendland, J. Breeding of lager yeast with 21. Viana, F., Gil, J. V., Genoves, S., Valles, S. & Manzanares, P. Rational selection of Saccharomyces cerevisiae improves stress resistance and fermentation non-Saccharomyces wine yeasts for mixed starters based on ester formation and performance. Yeast 29, 343–355 (2012). enological traits. Food Microbiol 25, 778–785 (2008). 7. Gietz, R. D. & Schiestl, R. H. Microtiter plate transformation using the LiAc/SS 22. Zott, K. et al. The grape must non-Saccharomyces microbial community: impact carrier DNA/PEG method. Nat Protoc 2, 5–8 (2007). on volatile thiol release. Int J Food Microbiol 151, 210–215 (2011). 8. Gobbi, M. et al. Lachancea thermotolerans and Saccharomyces cerevisiae in simultaneous and sequential co-fermentation: a strategy to enhance acidity and improve the overall quality of wine. Food Microbiol 33, 271–281 (2013). 9. Hazelwood, L. A., Daran, J. M., van Maris, A. J., Pronk, J. T. & Dickinson, J. R. The Acknowledgments Ehrlich pathway for fusel alcohol production: a century of research on This research was supported in part by the European Union Marie Curie Initial Training Saccharomyces cerevisiae metabolism. Appl Environ Microbiol 74, 2259–2266 Network Cornucopia (http://www.yeast-cornucopia.se/). (2008). 10. Hugenholtz, J. Traditional biotechnology for new foods and beverages. Curr Opin Biotechnol 24, 155–159 (2013). Author contributions 11. Ichikawa, K. & Eki, T. A novel yeast-based reporter assay system for the sensitive D.R. carried out the molecular experiments; D.R. and K.T. carried out flavor measurements; detection of genotoxic agents mediated by a DNA damage-inducible LexA-GAL4 A.W., J.W.W. and U.V. designed the experiments, A.W. and D.R. prepared the figures, protein. J Biochem 139, 105–112 (2006). J.W.W. wrote the main manuscript text; all authors reviewed the manuscript.

SCIENTIFIC REPORTS | 4 : 3707 | DOI: 10.1038/srep03707 3 www.nature.com/scientificreports

Additional information How to cite this article: Ravasio, D., Walther, A., Trost, K., Vrhovsek, U. & Wendland, J. An Supplementary information accompanies this paper at http://www.nature.com/ indirect assay for volatile compound production in yeast strains. Sci. Rep. 4, 3707; scientificreports DOI:10.1038/srep03707 (2014). Competing financial interests: The authors declare no competing financial interests. This work is licensed under a Creative Commons Attribution- NonCommercial-NoDerivs 3.0 Unported license. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/3.0

SCIENTIFIC REPORTS | 4 : 3707 | DOI: 10.1038/srep03707 4 Paper 2

Paper 2

The riboflavin over producer Ashbya gossypii produces significant amounts of fruity flavours compare to the close relative Eremothecium cymbalariae. Comparative genomic analysis showed that E. cymbalariae lacks most of the Ehrlich pathway genes, especially members of conserved ARO gene family. In A. gossypii mutation in the ARO genes led to a strong reduction in volatile production compare to the wild-type. On the other hand AgAro80 over expression produced significantly higher yields of VOCs, especially of isoamyl alcohol. Furthermore, in the study we compared the kinetic volatile profiles of the two Eremothecium species. The volatile compound analysis showed that both Eremothecium species produce large amounts of isoamyl alcohol while E. cymbalariae, lacking the major components of the ARO family, does not produce 2-phenylethanol.

RESEARCH ARTICLE Major contribution of the Ehrlich pathway for 2-phenylethanol/ rose ﬂavor production in Ashbya gossypii Davide Ravasio, Jurgen€ Wendland & Andrea Walther

Carlsberg Laboratory, Yeast Genetics, Copenhagen V, Denmark

Correspondence: Jurgen€ Wendland, Abstract Carlsberg Laboratory, Yeast Biology, Gamle Carlsberg Vej 10, DK-1799 Copenhagen V, Aroma alcohols of fermented food and beverages are derived from fungal Denmark. Tel.: +45 3327 5230; amino acids catabolism via the Ehrlich pathway. This linear pathway consists fax: +45 3327 4708; of three enzymatic reactions to form fusel alcohols. Regulation of some of the e-mail: [email protected] enzymes occurs on the transcriptional level via Aro80. The riboflavin overproducer Ashbya gossypii produces strong fruity flavours in contrast to its much Received 13 May 2014; accepted 3 June less aromatic relative Eremothecium cymbalariae. Genome comparisons indi- 2014. cated that A. gossypii harbors genes for aromatic amino acid catabolism DOI: 10.1111/1567-1364.12172 (ARO8a, ARO8b, ARO10, and ARO80) while E. cymbalariae only encodes ARO8a and thus lacks major components of aromatic amino acid catabolism. Editor: Terrance Cooper Volatile compound (VOC) analysis showed that both Eremothecium species produce large amounts of isoamyl alcohol while A. gossypii also produces high Keywords levels of 2-phenylethanol. Deletion of the A. gossypii ARO-genes did not confer ARO80; VOC; volatile flavour compounds; any growth deficiencies. However, A. gossypii ARO-mutants (except Agaro8a) biodiversity. were strongly impaired in aroma production, particularly in the production of the rose flavour 2-phenylethanol. Conversely, overexpression of ARO80 via the AgTEF1 promoter resulted in 50% increase in VOC production. Together these data indicate that A. gossypii is a very potent flavour producer and that amongst the non-Saccharomyces biodiversity strains can be identified that could provide positive sensory properties to fermented beverages.

pathway was already described in 1907 by the German Introduction biochemist Felix Ehrlich, (1907). The pathway consists of Fermented beverages contain a variety of flavour active three reactions: the initial transamination of an amino compounds that convey a fruity aroma, which is very acid, a decarboxylation and a subsequent alcohol dehy- prominent in beer and sake (Verstrepen et al., 2003; Kit- drogenase reaction (Fig. 1; for review see Hazelwood agaki & Kitamoto, 2013). Volatile esters represent the most et al., 2008). In Saccharomyces cerevisiae several genes and important group. Amongst them 2-phenylethanol, isoamyl gene families can perform each reaction. The transamina- alcohol, and ethyl caproate convey rose/flower, banana/fru- tion can be carried out by Aro8/Aro9 and the ity, and apple aromas, respectively. In lager beer high branched-chain amino acid transaminases Bat1/Bat2. The amounts of isoamyl alcohol can be present while other decarboxylation occurs through the Ehrlich pathway gene esters tend to contribute only minor amounts to the overall ARO10 or by the pyruvate decarboxylases Pdc1/Pdc5/ flavour bouquet. Alterations in beer flavour profile have Pdc6. This will convert an amino acid first into an a-keto been experienced when changing the production condi- acid and then into an aldehyde. In S. cerevisiae these alde- tions. The amount of higher alcohols increases when using hydes are then converted either to aroma alcohols or oxygenated wort or higher fermentation temperatures aroma acids via either alcohol dehydrogenases (encoded (Valero et al., 2001). Additionally, high-gravity brewing by the ADH-gene family) or aldehyde dehydrogenases

YEAST RESEARCH (providing increased amounts of sugars) results in (encoded by the ALD-gene family), respectively. Genetic increased production of acetate esters (Saerens et al., 2008). studies of these genes and others, including for example The biochemistry of volatile compound formation is the alcohol acetyl transferase genes ATF1/ATF2,contrib- based on amino acid catabolism. The corresponding uted greatly to the understanding of the molecular basis of

+ 2-oxoglutarate Glutamate CO2 NAD(P)H, H NAD(P)+ Acyl CoA CoA

Amino acid α-keto acid Aldehyde Alcohol Acetate O O R R O R R CH OH R OH OH O R1 O NH2 O Transaminase Decarboxylase Dehydrogenase Acetyltransferase

Amino acid α-Keto acid Fusel aldehyde Fusel alcohol Fusel acetate Fig. 1. Ehrlich pathway conversion of amino acids into fusel alcohols and esters. The 4-Methyl-2-oxo- Leu 3-Methylbutanal* Isoamyl alcohol* Isoamyl acetate** pentanoate relevant co-factors are indicated for each 3-Methyl-2-oxo- reaction. Speciﬁc molecular groups are Val 2-Methylpropanal* Isobutanol* Isobutyl acetate** butanoate highlighted and enzymatic functions for each 3-Methyl-2-oxo- Ile 2-Methylbutanal* 2-Methylbutanol Ethyl pentanoate reaction are provided. For the amino acids pentanoate processed in Ehrlich pathway reactions the 3-Phenyl-2-oxo- Phe 2-Phenylethanal 2-Phenyl ethanol* 2-Phenylacetate** intermediates are noted. Several of these propanoate substances were detected in Eremothecium * Compounds detected in A. gossypii and E. cymbalariae; ** Compound detected in E. cymbalariae only spec. as indicated.

flavour production in S. cerevisiae (Iraqui et al., 1998; pattern, colony morphology and flavour production Dickinson et al., 2003; Verstrepen et al., 2003; Vuralhan between both species. et al., 2005; Lilly et al., 2006; Ravasio et al., 2014). The aim of this report was to study the potential of The need for all-natural flavour production may guide A. gossypii for flavour production and to compare it with the selection of suitable lager or wine yeasts. Alternatively, its relatives E. cymbalariae and S. cerevisiae. For that pur- other yeast strains, collectively termed non-Saccharomyces pose we targeted the Ehrlich pathway ARO-genes for yeasts may be considered as flavour producers (Domizio deletion in A. gossypii and employed gas chromatography et al., 2011). This yeast biodiversity has not been evalu- methods to measure the volatile aroma compound forma- ated to its full extent. Yeast species derived from several tion in these strains. Our results reveal the potential use genera, including Dekkera, Hanseniaspora, Lachancea, of A. gossypii for natural flavour production in beverages, Pichia, Saccharomycodes, Zygosaccharomyces, have been which provides further support for screening the non- shown to contribute to the flavour profile of fermented Saccharomyces biodiversity for novel strains to be used in beverages in interesting new ways either in sequential fer- the food and beverage industry. mentation or when used as co-inoculants (Viana et al., 2008; Ciani et al., 2010; Gobbi et al., 2013). Materials and methods Ashbya gossypii, a known and industrially exploited overproducer of riboflavin/vitamin B is a filamentous ascomy- 2, Strains and media cete within the Saccharomycetaceae. Ashbya gossypii is a protoploid, pre-whole genome duplication species with a Strains generated and used in this study are shown in small genome of just 8.7 Mb + rDNA-repeats (Dietrich Table 1. Yeast strain CEN.PK2 and WS34/70 were grown et al., 2004). Molecular studies have centered on increasing in YPD (1% yeast extract, 2% peptone, 2% glucose) at riboflavin productivity (Stahmann et al., 2000; Jimenez 30 °C. Ashbya gossypii and E. cymbalariae were grown in et al., 2008; Park et al., 2011). With the identification of AFM (1% yeast extract, 2% caseine peptone, 2% glucose) highly efficient homologous recombination gene function at 30 °C. Antibiotic substances (G418 or ClonNAT) were analyses were initiated studying polar hyphal growth and added using a final concentrations of up to 200 lgmL 1 nuclear dynamics in A. gossypii (for review see Wendland & for the selection of transformants. Plates were incubated Walther, 2005; Schmitz & Philippsen, 2011). Comparative for 7 days at 30 °C prior to photography. For the growth genomics revealed marked differences between A. gossypii assays diameter of the mycelium was measured daily. Bio- and its close relative, the filamentous fungus Eremothecium logical triplicates were performed. For sporulation, an cymbalariae. These include genome size (8.7 Mb vs. overnight culture of A. gossypii was further incubated in 9.7 Mb), chromosome number (7 vs. 8), degree of synteny minimal medium (1.7 g L 1 YNB w/o ammonium sul- with the yeast ancestor (higher in E. cymbalariae), and phate and w/o amino acids, 0.69 g L 1 CSM, 20 g L 1 GC-content (52% vs. 40%; Wendland & Walther, 2011). glucose, 1 g L 1 asparagine, and 1 g L 1 myo-inositol) Furthermore, we noticed decisive differences in growth for up to 5 days.

Table 1. Strains used and generated in this study

Strain Collection number Genotype Source ATCC 10895 A. gossypii 71 leu2 Lab Strain Eremothecium cymbalariae C64 Wild-type DBVPG 7215 Saccharomyces pastorianus C44 Weihenstephan Lager Yeast 34/70 Lab Strain CEN.PK2 C598 Prototrophic strain, (CEN.PK113-7d x CEN.PK113-1A) Lab strain A. gossypii C692 Δagr167w (Δaro8a) This study A. gossypii C694 Δafr548w (Δaro8b) This study A. gossypii C703 Δagr167w/afr548w (Δaroa/b) This study A. gossypii C632 Δacr211w (Δaro10) This study A. gossypii C696 Δadr199c (Δaro80) This study A. gossypii C787 ADR199C XL (ARO80XL) This study

Primary heterokaryotic transformants were sporulated and Transformation homokaryotic mutant mycelia were generated from uninu- Transformation of A. gossypii was done by electropora- cleate and haploid spores. For each desired genetic manipulation as described previously (Wendland et al., 2000). tion, two independent transformants were generated. PCR-based cassettes were amplified from pFA-GEN3/ All primers are listed in Table 3. Primers were obtained pFA-SAT1 plasmids (see Table 2) using gene specific S1 from Integrated DNA Technologies (IDT, Leuven, Bel- and S2 primers. For the deletion of ARO80 a disruption gium). Diagnostic PCR was used for verification of correct cassette was cloned containing a 2 kb ORF-fragment integration of a disruption cassette and concomitant using the primer G1-AgARO80 (#6117) and G4-Ag- absence of the target gene in homokaryotic deletion strains. ARO80 (#6118). An internal 1.1 kb EcoRV-NdeI fragment was replaced by kanMX. The ARO80 disruption Fermentation conditions cassette was cleaved from the vector backbone with KpnI and SpeI and contained flanking homology regions of Tall tube cylinders containing 200 mL medium were used 0.45 and 0.6 kb at the 50 and 30 ends, respectively. for fermentation. As fermentation broth YPD (with 15% glu- To create a chromosomal ARO80 overexpression strain cose) was used for CEN.PK2 and WS34/70 at 20 °C, respec- an integration cassette was generated. The ARO80-ORF- tively. The starting cell density was set to OD600 at 0.2. Tall terminator fragment was amplified using the primer tubes containing magnetic stirrers were placed on stirrer A1-AgARO80 (#6170) and A4-AgARO80 (#6171). The pads run at 190 r.p.m. Fermentation progress was moni- resulting 3.2 kb fragment was then cloned downstream tored over 5 days measuring CO2 loss and reduction of sugar of the ScTEF1-promoter of the pFA-GEN3-ScTEF1p content using a DMA 35 Anton Paar densitometer, which vector via XhoI/SacII. The AgARO80-promoter served as determines medium gravity in °Plato. The fermentation was 50-flanking homology region. This fragment was amplified considered finished when the sugar content did not drop by PCR (primers #6287-#6288) and after a PvuII/MscI further. Fermentation supernatants were decanted and used cleavage cloned into the unique PvuII site upstream of for GC/MS, GC/flame ionization detector (FID) analysis. All the GEN3 marker gene in pFA-GEN3-ScTEF1p-Ag- fermentations were carried out in biological duplicates. ARO80, resulting in pFA-AgARO80p-GEN3-ScTEF1p-Ag- ARO80. Two NcoI sites within the flanking homolog Eremothecium growth conditions for flavor regions were used to release the overexpression cassette, profiling which provided flanking homology regions of 0.26 and 2.7 kb at the 50 and 30 ends, respectively. Strains were pre-cultured o/n in AFM. One millilitre of each pre-culture was used to inoculate a 150 mL AFM culture in a 250 mL baffled flasks. At specific time points Table 2. Plasmids used and generated in this study 50 mL samples were collected. In order to preserve the Plasmid Description Source volatiles and stop cell metabolism samples were kept at ° #121 pFA-KanMX6 Philippsen lab 20 C until analysis. #550 pFA-SAT1 Schaub et al. (2006) Analytical methods – GC/MS, GC/FID #C886 pFA-AgARO80p-GEN3-ScTEF1p-AgARO80 This study #C791 pSK-AgARO80::kanMX This study Two different extraction methods were used in this study. #C886 pFA-AgARO80pGEN3-ScTEF1p-AgARO80 This study For the comparison of A. gossypii mutants with yeast

Table 3. Oligos used in this study

Oligo Name Sequence 50-to-30* #6129 S1-AGR167w (ARO8a) TAAGTCAGCAAGATCGCTGGGCGCTAAGGTAGATAACGACAAGAGgaagcttcgtacgctgcaggtc #6130 S2- AGR167w (ARO8a) GCACCACGAGGCAGGGCAGGTGACTGGAGGCTAGTATTTTATGGActgatatcatcgatgaattcgag #6109 G1-AGR167w CTCGGAACCGGGTCAGTTC #6110 G4-AGR167w GACCTTGGAAGTGGACTCC #6111 I1-AGR167w CAGCACACGGAGAAGTTCCAC #6112 I2-AGR167w CCACGAGCGTGCTGCCCATC #6127 S1-AFR548w (ARO8b) CCGCACAGTGGCTCCCGCAGGGCGCTTCTTTGGCTGAGCGCTCCGgaagcttcgtacgctgcaggtc #6128 S2-AFR548w (ARO8b) AAGTCTTGGCCACAAGTGGCAATCGAAGTCGGCACCTATAAGTGActgatatcatcgatgaattcgag #6113 G1-AFR548w CTCGTCAAAGCTGGCTTACTG #6114 G4-AFR548w CTGGAGACGCTCTCCTCGG #6115 I1-AFR548w GACGTGCTGGCGACGGTCG #6116 I2-AFR548w GGAGAACGAGTCCAGACGC #6121 S1-ACR211w (ARO10) AACAGGGGTAGGAGGTTTACAGGAGGTAAGCGAGCGGCACGAGACgaagcttcgtacgctgcaggtc #6122 S2-ACR211w (ARO10) CCCCTGCATGTCTTCTGTTGCTCGTCTGCGGAGTAGCTACGCAGCctgatatcatcgatgaattcgag #6123 G1-ACR211w GTGGCTATTCGTGGGCTGG #6124 G4-ACR211w GCGTAGAACCAGCTCTCTTC #6125 I1-ACR211w GGCTCTCAATGGCGTGGCAG #6126 I2-ACR211w CGCATAGTATTGCAGGCGTGC #6297 50-ADR199c (KpnI) AGAATAggtaccGTCCCAGTGCTTTTGTGACCGTC #6171 30-ADR199c (XbaI) ATAAGAtctagaGTACCAGCTCCATAGTCCATGGTAATC #6117 G1-ADR199c CTCATCACTTGTGTGGAGCC #6118 G4-ADR199c GCCGTAGAGGATGCGCTC #6193 I1-ADR199c GTCGAATCGCGGTTCTCCG #6305 I2-ADR199c CTGAACTCCGGCTCATCGC #1202 G2-KanMX6 GCGTTTCCCTGCTCGCAGGTC #1198 G3-KanMX6 CGCCTCGACATCATCTGCCC #6717 G2-SAT1 GCAATAAATCTTGGTGAGAACAGC #6718 G3-SAT1 GCGGCATTGACCTCTTCACG *Restriction sites are in lower case. strains, a solvent extraction with carbon disulphide (CS2) Clara, CA). Separation of volatiles was carried out on a was used. The samples were stirred for 30 min, centri- ZB-Wax capillary column (Zebron capillary, 30 m 9 fuged and 2 lL of the liquid organic phase was injected 0.25 mm 9 0.5 lm). The mass spectra were recorded in into the gas chromatograph (Agilent 6890 GC). 1-octanol electronic impact mode at 70 eV in a mass/charge ratio was used as an internal standard. Detection of volatiles from 15 to 300 m/z. Data analysis was performed was performed using a FID. The volatiles were separated using the software program MSDCHEMSTATION (Version on a DBWAX capillary column (30 m 9 0.32 mm 9 E.02.02.1431, Agilent Technologies). Identification of 0.25 lm). For comparison of Eremothecium strains vola- compounds was based on comparison with a mass spec- tile compounds were measured using biological and tech- tral database (Nist 1.0.0.23). One characteristic quantifier nical duplicates for dynamic headspace sampling. As an ion and two to three qualifier ions were selected for each internal standard 4 methyl-1 pentanol was added to each compound. The peak area of the quantifier ion was used sample. All samples were equilibrated to 37 °C in a water for quantification. bath and then purged with hydrogen (100 mL min 1) for 20 min while stirring at 200 r.p.m. The volatile com- Results pounds were collected on Tenax-TA traps (250 mg, mesh size 60/80, density 0.37 g mL 1, Buchem bv, Apeldoorn, Comparison of flavour production between The Netherlands) at 37 °C. To eliminate excess water A. gossypii and S. cerevisiae traps were purged by hydrogen flux of 100 mL min 1 for 10 min. The trapped volatiles were desorbed using an The amino acid catabolism of branched-chain amino acids automatic thermal desorption unit (ATD 400, Perkin (Leu, Val, Ile) and phenylalanine via the Ehrlich pathway Elmer, Norwalk, CT) and were automatically transferred generates a-keto acid and fusel aldehyde intermediates to a gas chromatograph-mass spectrometer (789OA GC that are converted to fusel alcohol, which can be esterified system, 5975C VL MSD, Agilent technologies, Santa into fusel acetates (Fig. 1). Ashbya gossypii produces a

ª 2014 Federation of European Microbiological Societies. FEMS Yeast Res && (2014) 1–12 Published by John Wiley & Sons Ltd. All rights reserved Ehrlich pathway in Ashbya gossypii 5 characteristic highly fruity flavour when grown either on highest at day 5. Eremothecium cymbalariae, on the other plates or in liquid culture. For a comparison of flavour pro- hand, produced isoamyl acetate and 2-phenyl acetate but files via GC/FID measurements of volatile compounds we only in the initial period of growth and these compounds used A. gossypii grown in AFM for 5 days, the S. cerevisiae were significantly reduced at day 4 and 5. Thus our analysis strain CEN.PK2, and the Weihenstephan lager yeast produc- indicates that A. gossypii produces strong notes of banana tion strain grown in YPD (15% glucose; Fig. 2). All strains and rose flavour, that is flavour alcohols, while E. cymbala- were incubated up to 5 days in either oxidative (A. gossypii) riae produces higher amounts of esters. Specifically, E. cym- or fermentative conditions (yeast strains). To increase the balariae generated high amounts of ethyl acetate and ethyl volatile production the sugar content of the media was propionate, which at moderate concentration impart pear increased to 15% glucose. Since the growth of A. gossypii and pineapple like flavours, respectively (Table 5). was significantly reduced in media with sugar content higher than 2% we incubated this stain in standard AFM medium. Components of the Ehrlich pathway in This showed that A. gossypii was superior in 2-phenyletha- A. gossypii nol production compared to these yeast strains. We noted previously that E. cymbalariae lacks most components of the Ehrlich pathway which are, however, pres- Comparison of flavour production between ent in A. gossypii (Wendland & Walther, 2011). Here we Eremothecium species provide a more detailed analysis of the Ehrlich pathway Volatile compound formation has not been analysed so far gene set in comparison to other yeasts and the compiled in Eremothecium species. To monitor any change in VOCs yeast ancestral genome prior to the whole genome dupli- formation over time we measured the volatiles produced by cation (Gordon et al., 2009). In contrast to S. cerevisiae liquid cultures at day three, four, and five representing med- and the yeast ancestor pre-duplication yeast species may ium to late stages of growth in both Eremothecium species harbour two ARO8 paralogs, here termed ARO8a and (Tables 4 and 5). Within this comparison between A. gos- ARO8b for A. gossypii, Kluyveromyces lactis and Lachancea sypii and E. cymbalariae we identified only a marginal pro- waltii (Table 6). The E. cymbalariae genome encodes duction of 2-phenylethanol production in E. cymbalariae. ARO8a but lacks ARO8b, ARO9, ARO10, as well as the In contrast E. cymbalariae had low levels of isobutylacetate transcriptional regulator ARO80. Comparison of Aro- which were absent in A. gossypii (Fig. 3a). Both species proteins indicated that the transaminases Aro8a, Aro8b showed a strong production of isoamyl alcohol (Fig. 3b). and Aro9 are sufficiently divergent to cluster into separate Production of both isoamyl alcohol and 2-phenylethanol in groups (Fig. 4). Interestingly, the amino acid sequence A. gossypii increased over the assayed period and were identity between Aro8a and Aro8b paralogs of one species is lower (in the three species evaluated on average 48.1%) than that of Aro8a or Aro8b proteins of different species: 500 50 on average Aro8a proteins show about 55% amino acid WS34/70 450 45 sequence identity, Aro8b proteins about 61.0%. Eremothe- CEN.PK2 cium species are distinguished from other yeasts by their Ag 400 40 lack of ARO9. 350 35

300 30

mg L Deletion of AgARO-genes does not confer –1 250 25

–1 growth phenotypes mg L 200 20 Via PCR-based gene targeting we generated A. gossypii sin- 150 15 gle deletion mutants lacking either ARO8a, ARO8b, ARO10,

100 10 or ARO80 as well as a double mutant with deletions of ARO8a and ARO8b. Growth comparisons of these mutants 50 5 with the parental strain showed no radial growth delay 0 0 Iso butanol Isoamyl 2-phenyl Isoamyl 2-phenyl (Fig. 5). alcohol ethanol acetate ethyl acetate

Fig. 2. Comparison of ﬂavor compounds between Ashbya gossypii (Ag) VOC analysis of A. gossypii Ehrlich pathway and yeast strains. Gas chromatography was employed to measure the mutants indicated fusel alcohols or esters. Strains: Weihenstephan lager yeast (WS34/70) and Saccharomyces cerevisiae laboratory strain CEN.PK2. For The A. gossypii ARO-mutant strains with deletions in culture and growth conditions see ‘Materials and methods’. Ehrlich pathway components were analysed for their

Table 4. Ashbya gossypii VOCs produced over 3 days

3rd day, ppm 4th day, ppm 5th day, ppm Alcohols 1 Butanol 4.46 (0.27) 6.05 (0.49) 0.36 (0.10) 2 Phenylethanol 4.00 (0.56) 32.86 (6.85) 90.82 (12.42) Benzylalcohol 0.11 (0.08) 0.16 (0.03) 0.16 (0.04) Isoamyl alcohol 459.53 (11.27) 754.01 (23.27) 850.24 (37.03) Isobutanol 36.52 (2.47) 74.16 (4.50) 51.63 (2.95) Propanol 10.87 (0.66) 5.79 (0.59) / Esters 2 Phenylacetate 0.03 (0.00) 0.39 (0.13) 1.99 (0.30) Ethyl acetate 173.46 (13.02) 194.17 (2.59) 70.45 (2.39) Ethyl phenylacetate 0.07 (0.01) 0.12 (0.01) 0.18 (0.01) Isoamylacetate 0.41 (0.04) 3.85 (1.46) 2.57 (1.04) S-methylthioacetate 3.32 (0.37) 0.26 (0.02) 0.22 (0.02) Aldehydes 2-Methyl-2 butenal 3.50 (0.05) 0.50 (0.23) 0.43 (0.07) (E-E) 2,4-Hexadienal 0.15 (0.17) / / 2-Butenal 73.01 (2.40) 3.02 (1.91) / Acetaldehyde 0.71 (0.30) 0.69 (0.06) 0.69 (0.05) 3-Methyl benzaldehyde / 0.57 (0.06) / Benzaldehyde 0.22 (0.03) 0.53 (0.02) 1.00 (0.13) Benzeneacetaldehyde 1.90 (1.28) 1.89 (0.19) 7.68 (1.10) Butanal 0.34 (0.09) / / 2-Methyl-1 butanal 49.09 (8.19) 38.40 (5.25) 55.31 (0.45) 2-Methyl-1 propanal 4.31 (0.57) 3.26 (1.03) 3.31 (0.44) 3-Methyl-1 butanal 71.29 (5.21) 24.09 (0.41) 57.37 (2.30) 5 Methyl-2-phenyl hexanal / / 0.13 (0.06) Ketons 2 Pentanone 0.73 (0.02) 1.17 (0.07) 1.00 (0.04) 2-Butanone 5.47 (0.36) 4.69 (0.35) 3.24 (0.37) c-Decalactone 2.09 (1.51) 2.72 (0.73) 14.23 (1.09) Acids Acetic acid 0.15 (0.05) / / Ethyl benzoate 0.23 (0.05) 0.24 (0.02) 0.11 (0.12) Ethyl butanoate 2.23 (0.04) 2.16 (0.14) 0.66 (0.18) Ethyl caproate 0.26 (0.00) / / Ethyl isobutyrate 0.38 (0.37) 0.22 (0.01) / Ethyl propionate 4.47 (0.54) 7.78 (0.13) 3.11 (0.64) Valeric acid 2.36 (1.94) 1.22 (0.63) 7.29 (0.85) Aromatic compounds 2,5-Dymethyl-pyrazine 2.42 (0.13) 2.47 (0.08) 2.52 (0.01) 2-Ethyl-5 methyl-pyrazyne 0.12 (0.08) 0.20 (0.01) 0.22 (0.01) 2-Ethyl-6-methyl-pyrazyne 0.10 (0.03) 0.63 (0.09) 0.94 (0.15) 3-Ethyl-2,5-dimethyl-pyrazine 0.16 (0.02) 0.25 (0.02) 0.21 (0.01) 3-Phenylfuran 0.09 (0.01) 0.08 0.00 0.05 (0.01) 3-Phenylpentane / 0.86 (0.05) / 1-3-Bis-1,1-dymethylethyl-benzene 0.12 (0.09) 0.08 (0.06) 0.06 (0.02) Others E-1-Hydroxy-1,3butadiene 0.89 (0.14) 0.31 (0.27) / Hexane 0.79 (0.49) 0.68 (0.38) 0.37 (0.10) aroma alcohol proﬁle using GC/FID (Fig. 6a). We found reduction of isoamyl ethanol or 2-phenylethanol produc- a striking effect of deletions of ARO8b, ARO10, and tion and the aro8a/aro8b double mutant exhibited the ARO80 in the severe reduction in 2-phenylethanol forma- aro8b phenotype with respect to these ﬂavours. Thus we tion in A. gossypii. Additionally, the amount of isoamly conclude that the Ehrlich pathway in A. gossypii plays a alcohol was decreased in these strains to about 80% of major role in 2-phenylethanol production and also con- the wild type level. Deletion of Agaro8a did not result in tributes to isoamyl alcohol formation.

Table 5. Eremothecium cymbalariae VOCs produced over 3 days

3rd day, ppm 4th day, ppm 5th day, ppm Alcohols 1-Butanol 2.50 (0.02) 7.31 (0.36) 16.34 (0.64) 1-Propanol / 0.83 (0.15) 2.68 (0.13) 2 Phenylethanol 0.26 (0.06) 0.29 (0.01) 0.31 (0.05) 3-Methyl-3-Buten-1-ol / 0.30 (0.02) 0.35 (0.01) Benzylalcohol 0.04 (0.00) 0.04 (0.00) 0.04 (0.00) Citronellol / 0.06 (0.00) 0.05 (0.00) Ethanol 0.22 (0.03) 0.22 (0.07) 0.17 (0.02) Isoamyl alcohol 275.51 (2.28) 468.72 (8.72) 526.42 (19.73) Isobutanol 8.32 (0.92) 22.15 (2.00) 42.60 (2.39) Phenol 0.09 (0.01) 0.10 (0.00) 0.06 (0.01) Esters 2 Phenylacetate 6.78 (0.11) 0.05 (0.02) 0.01 (0.01) 2-Furanmethanol acetate 0.11 (0.01) / / Amylacetate 0.13 (0.00) / / Isoamylacetate 367.37 (8.69) 15.23 (2.41) 7.13 (0.58) Isobutylacetate 2.37 (0.09) 1.68 (0.38) 0.47 (0.00) Ethyl acetate 941.94 (41.37) 720.65 (11.86) 507.13 (28.03) Heptyl acetate 0.09 (0.02) / / Propylacetate 13.63 (1.08) 2.73 (0.30) 2.12 (0.00) Benzylacetate 1.49 (0.06) / / Aldehydes 2-Butenal / 49.04 (3.42) 108.03 (1.35) 2-ethyl-trans-2-butenal / 0.63 (0.07) 2.94 (0.04) 2-Methyl propanal 0.34 (0.12) 1.14 (0.11) 1.20 (0.06) 2-Methyl-2-butenal / 7.80 (0.19) 13.18 (0.25) 2-Methyl-butanal 31.60 (1.75) 47.71 (4.66) 10.58 (0.92) 2-Mehtyl-benzaldehyde / 0.12 (0.01) 0.04 (0.01) 3-Methyl-butanal 42.15 (2.37) 93.29 (4.59) 35.93 (0.90) Acetaldehyde 0.75 (0.03) 0.75 (0.04) 0.76 (0.05) Benzaldehyde 0.21 (0.01) 0.58 (0.04) 0.53 (0.01) Butanal / / 0.87 (0.07) Caproaldehyde 0.05 (0.00) 0.00 (0.00) / Ketons 2 Butanone 1.30 (0.17) 3.31 (0.63) 5.61 (2.03) 2,3-Butanedione 14.24 (0.35) 4.11 (0.48) 4.00 (0.43) Acetophenone 0.12 (0.01) 0.16 (0.03) 0.13 (0.00) c-Decalactone 5.19 (1.60) 9.41 (2.74) 11.26 (0.67) 3-Hydroxy-2-butanone 11.94 (1.32) 14.56 (0.59) 11.77 (0.83) Acids Amyl-propionate 5.69 (0.67) 4.20 (0.45) 1.42 (0.02) Ethyl Butyrate 1.05 (0.03) 2.70 (0.12) 0.76 (0.01) Ethyl Isobutyrate / 0.64 (0.08) 0.19 (0.01) Ethyl propionate 65.65 (1.94) 119.46 (3.26) 66.70 (0.25) Propyl propionate 0.31 (0.01) 0.00 (0.00) / Valeric acid 2.49 (0.04) 7.87 (1.70) 1.23 (0.06) Acetic acid 0.25 (0.02) 0.31 (0.03) 0.19 (0.03) Aromatic compounds 2,5-Dihydro-furan / 1.63 (0.04) 4.00 (0.07) 3-Ethyl-2,5-dimethyl-pyrazyne / 0.17 (0.04) 0.15 (0.01) 2,5-Dymethyl-pyrazine 2.44 (0.04) 2.28 (0.17) 2.40 (0.06) 2-Ethyl-5-methyl-pyrazyne 0.16 (0.00) 0.14 (0.02) 0.15 (0.00) 3-Phenylfuran 0.06 (0.00) 0.06 (0.01) 0.04 (0.00) Benzonitrile 0.08 (0.04) 0.08 (0.02) 0.05 (0.00)

Table 5. Continued

3rd day, ppm 4th day, ppm 5th day, ppm Others (E)1-Hydroxy-1,3-butadiene / 0.70 (0.09) 0.53 (0.09) 2,4,5-trimethyl-3-oxazoline / 1.04 (0.03) 1.24 (0.05) 2-Isopropylbut-2-enal / 0.40 (0.03) 0.33 (0.01) Hexane 0.22 (0.01) 0.22 (0.03) 0.49 (0.09)

(a) 10 100 10 Ashbya E. cymbalariae 8 80 6 60 5 4 40 2 20

ppm expressed as I.S 0 0 0 345 345 345 day Val 2-Methylpropanal Isobutanol Isobutyl acetate Phe 2-Phenylethanal 2-Phenylethanol 2-Phenylacetate

100 10 80 8 60 6 40 4 20 2

ppm expressed as I.S 0 0 345 345 day (b)

100 Ashbya E. cymbalariae 80 60 40 20

ppm expressed as I.S 0 345 day Fig. 3. Time course analysis of ﬂavor Ile 2-Methylbutanal 2-Methylbutanol Ethyl pentanoate production in Ashbya gossypii and Leu 3-Methylbutanal Isoamyl alcohol Isoamyl acetate Eremothecium cymbalariae. Strains were cultured for 5 days and samples were 1000 100 1000 analyzed between days 3–5. Volatile 800 80 800 compounds were analyzed via GC/MS. Ehrlich 600 60 600 pathway products for the amino acids valine, 400 40 400 phenylalanine, isoleucine and leucine were 20 200 200 quantiﬁed. Compounds written in grey are

ppm expressed as I.S 0 0 0 345 345 345 intermediates of the respective amino acid day degradation but were not detected.

Table 6. ARO-gene content in different fungal species

Transaminase Decarboxylase Transcription factor Species* ARO8a ARO8b ARO9 ARO10 ARO80 Ancestor Absent Anc_3.511 Anc_2.102 Anc_5.456 Anc_5.525 S. cerevisiae Absent YGL202W YHR137w YDR380W YDR421W K. lactis A04906 F10021 D11088 E02707 A01804 L. waltii 17462 18982 14582 Absent Absent A. gossypii AGR167W AFR548C Absent ACR211W ADR199C E. cymbalariae Ecym_7228 Absent Absent Absent Absent *Species depicted were: Saccharomyces cerevisiae, Kluyveromyces lactis, Lachancea waltii, Ashbya gossypii, Eremothecium cymbalariae.

before (Fig. 6b). It became apparent that ARO80 overex- EcymAro8a pression resulted in a 2.5 fold increase of isoamyl alcohol AgAro8a LwAro8a and isobutanol levels derived from branched-chain amino KlAro8a acid catabolism, whereas the levels of 2-phenylethanol KlAro8b were basically unchanged, which could be attributable to LwAro8b the limited uptake of phenylalanine from the medium. ScAro8b AgAro8b Discussion LwAro9 KlAro9 Non-conventional yeasts (NCY) harbour a great potential ScAro9 as novel production strains, for example the production KlAro80 of L-lactic acid by Candida sonorensis, or the production AgAro80 of another highly useful chemical building block, D- ScAro80 xylonic acid, by Pichia kudriavzevii, for the production of AgAro10 virus like particles in Kluyveromyces lactis or for the pro- KlAro10 duction of all-natural ﬂavours by a wide variety of NCYs ScAro10 (Goretti et al., 2013; Ilmen et al., 2013; Toivari et al.,

Fig. 4. Phylogenetic tree of transaminase and decarboxylase genes. 2013). The scope of this biodiversity within the Sacchar- Aro-proteins of Ashbya gossypii (Ag), Eremothecium cymbalariae omycetaceae has but just begun to be exploited and will (Ecym), Lachancea waltii (Lw), and Kluyveromyces lactis (Kl) were provide a rich source for biotechnological research in the analyzed using DNASTAR MEGALIGN software. future (Domizio et al., 2011). This group of yeasts is very attractive based on their ease of cultivation and molecular genetic tractability combined with small genome sizes Overexpression of AgARO80 enhances flavour allowing for rapid determination of draft genome production in A. gossypii sequences and the functional analysis of target genes. Since the deletion of ARO80 showed a clear negative In this study we have analysed the flavour profile of effect on flavour production we went on to analyse if two closely related Eremothecium species, which are both overexpression of ARO80 could further increase flavour filamentous fungi belonging to clade 12 of the Saccharo- levels. To this end the chromosomal AgARO80 gene was mycetes (Kurtzman & Robnett, 2003). Both species differ placed under the control of the strong constitutive TEF1 on the genomic level and morphologically, for example promoter from S. cerevisiae (see ‘Materials and methods’). regarding extensive aerial mycelium formation in E. cym- Resulting transformants were compared to the parental balariae compared to synnemata formation in A. gossypii wild type strain analysed under the same conditions as and also in the shape of their spores (Wendland &

Fig. 5. Functional analysis of Ashbya gossypii ARO-genes. Radial colony growth of the indicated strains inoculated at the center of Ag aro8a Ag aro8b Ag aro8a/b Ag aro10 Ag aro80 Ag wt each petri plate was monitored over 8 days. Growth was either at either 30 °Cor37°C. Mycelia were grown on AFM plates and photographs were taken at the end of the growth period.

wt aro8a aro8b aro8a-b aro10 aro80 2002). Ashbya gossypii generated the highest amounts of 600 (a) 2-phenylethanol among the species we tested in our

500 experiments, which makes it a very efﬁcient natural producer of this ﬂavour. Our deletion analyses showed that

) 400 Ehrlich pathway genes are of central importance for the –1 production 2-phenylethanol. The microbial production of 300 2-phenylethanol has recently received considerable atten-

ppm (mg L tion as an all-natural ﬂavour in the food and cosmetics 200 industry. Due to the toxic effect of large amounts of 1 100 2-phenylethanol (>2.5 g L ) in situ removal strategies using organic solvents have been developed to avoid a 0 decrease in growth rate (Stark et al., 2003; Eshkol et al., Isobutanol Isoamyl alcohol 2-Phenyl ethanol 2009; Achmon et al., 2011; Hua & Xu, 2011). wt The production of more fruity notes in beverages could (b) TEF1p-ARO80 600 beneﬁt from for example co-fermentations using S. cerevi-

500 siae and another non-conventional yeast strain. Such co-fermentations have gained increased popularity in the 400 wine industry (Ciani et al., 2010). Using a lacZ-based ) –1 assay system we recently evaluated the flavour profiles of 300 S. cerevisiae and its closely related species (Ravasio et al., 2014). In this study we could show that overexpression of ppm (mg L 200 ARO80 has a profound impact on flavour production.

100 Thus a more broadly applicable assay for non-conventional yeasts could help linking the expression level of 0 Ehrlich pathway genes to the flavour production capabili- Isobutanol Isoamyl alcohol 2-Phenyl Ethanol ties of a previously uncharacterized isolate. Fig. 6. Volatile compound formation in Ashbya gossypii aro-mutants. Ashbya gossyppi is well-known for its riboflavin over- (a) the indicated mutant strains were grown as described in production. Recent studies hypothesized that flavin- ‘Materials and methods’ and fusel alcohols of the valine, isoleucine, dependent detoxification of plant compounds may open and phenylalanine were assayed via GC/FID. (b) Overexpressionof new niches for insects (Sehlmeyer et al., 2010; Dietrich the ARO80 transcription factor enhances isoamyl formation in et al., 2013). Involving Ashbya in such a plant-insect sys- A. gossypii. tem required adaptation to an insect environment to sus- tain viability in milkweed bugs (e.g. during hibernation Walther, 2005). The purpose of this study was to deter- periods) and overproduction of riboflavin to detoxify mine differences in flavour production between these toxic alkaloids produced for example by oleander. These Eremothecium species and relate these to the different environmental stresses may have resulted in alteration genetic makeup of ARO-genes. Our flavor profiling and streamlining of the Ashbya genome compared to results revealed a preference regarding flavour alcohol E. cymbalariae. On the other hand, with such a niche production in A. gossypii vs. ester production in E. cym- found, reinforcing the relationship of Ashbya with certain balariae. Higher ester production in E. cymbalariae insects used for spreading of Ashbya could have occurred resembles that found in lager yeast strains fermenting by using volatile compounds such as 2-phenylethanol as high gravity wort with high concentrations of sugar attractant. Our study opens new aspects of linking gen- (Verstrepen et al., 2003). Considering the slow growth of ome evolution to both biotic and abiotic environmental E. cymbalariae even the relatively low amounts of 2% glu- challenges. cose in our experiments may have triggered an excess amount of pyruvate formation resulting in increased ace- Acknowledgements tate ester formation. In line with this we have observed larger amounts of ethyl acetate and other acetate esters in This research was supported in part by the European E. cymbalariae compared to the faster growing A. gossypii Union Marie Curie Initial Training Network Cornucopia. (see Tables 4 and 5). Ashbya gossypii formed very large Sequence data for Ashbya gossypii was obtained from the amounts of both isoamyl alcohol and 2-phenylethanol. Ashbya Genome Database website at http://agd.vital-it.ch/ From S. cerevisiae it is known that 2-phenylethanol for- index.html. We thank Dr Mikael Agerlin Petersen, Uni- mation is restricted to the growth phase (Stark et al., versity of Copenhagen for help with VOC measurements.

Ilmen M, Koivuranta K, Ruohonen L, Rajgarhia V, Suominen References P & Penttila M (2013) Production of L-lactic acid by the Achmon Y, Goldshtein J, Margel S & Fishman A (2011) yeast Candida sonorensis expressing heterologous bacterial 12 Hydrophobic microspheres for in situ removal of and fungal lactate dehydrogenases. Microb Cell Fact : 53. 2-phenylethanol from yeast fermentation. J Microencapsul Iraqui I, Vissers S, Cartiaux M & Urrestarazu A (1998) 28: 628–638. Characterisation of Saccharomyces cerevisiae ARO8 and Ciani M, Comitini F, Mannazzu I & Domizio P (2010) ARO9 genes encoding aromatic aminotransferases I and II Controlled mixed culture fermentation: a new perspective reveals a new aminotransferase subfamily. Mol Gen Genet 257 – on the use of non-Saccharomyces yeasts in winemaking. : 238 248. FEMS Yeast Res 10: 123–133. Jimenez A, Santos MA & Revuelta JL (2008) Phosphoribosyl Dickinson JR, Salgado LE & Hewlins MJ (2003) The pyrophosphate synthetase activity affects growth and 8 catabolism of amino acids to long chain and complex riboflavin production in Ashbya gossypii. BMC Biotechnol : alcohols in Saccharomyces cerevisiae. J Biol Chem 278: 8028– 67. 8034. Kitagaki H & Kitamoto K (2013) Breeding research on sake Dietrich FS, Voegeli S, Brachat S et al. (2004) The Ashbya yeasts in Japan: history, recent technological advances, 4 gossypii genome as a tool for mapping the ancient and future perspectives. Annu Rev Food Sci Technol : – Saccharomyces cerevisiae genome. Science 304: 304–307. 215 235. Dietrich FS, Voegeli S, Kuo S & Philippsen P (2013) Kurtzman CP & Robnett CJ (2003) Phylogenetic relationships Genomes of Ashbya fungi isolated from insects reveal four among yeasts of the ‘Saccharomyces complex’ determined 3 – mating-type loci, numerous translocations, lack of from multigene sequence analyses. FEMS Yeast Res : 417 transposons, and distinct gene duplications. G3 (Bethesda) 432. 3: 1225–1239. Lilly M, Bauer FF, Lambrechts MG, Swiegers JH, Cozzolino D Domizio P, Romani C, Lencioni L, Comitini F, Gobbi M, & Pretorius IS (2006) The effect of increased yeast alcohol Mannazzu I & Ciani M (2011) Outlining a future for acetyltransferase and esterase activity on the flavour profiles 23 – non-Saccharomyces yeasts: selection of putative spoilage wine of wine and distillates. Yeast : 641 659. strains to be used in association with Saccharomyces Park EY, Ito Y, Nariyama M, Sugimoto T, Lies D & Kato T cerevisiae for grape juice fermentation. Int J Food Microbiol (2011) The improvement of riboflavin production in Ashbya 147: 170–180. gossypii via disparity mutagenesis and DNA microarray 91 – Ehrlich F (1907) Uber€ die Bedingungen der Fuselolbildung€ analysis. Appl Microbiol Biotechnol : 1315 1326. und uber€ ihren Zusammenhang mit dem Eiweissaufbau der Ravasio D, Walther A, Trost K, Vrhovsek U & Wendland J Hefe. Ber Dtsch Chem Ges 40: 1027–1047. (2014) An indirect assay for volatile compound production 4 Eshkol N, Sendovski M, Bahalul M, Katz-Ezov T, Kashi Y & in yeast strains. Sci Rep : 3707. Fishman A (2009) Production of 2-phenylethanol from Saerens SM, Verbelen PJ, Vanbeneden N, Thevelein JM & L-phenylalanine by a stress tolerant Saccharomyces cerevisiae Delvaux FR (2008) Monitoring the influence of high-gravity strain. J Appl Microbiol 106: 534–542. brewing and fermentation temperature on flavour formation Gobbi M, Comitini F, Domizio P, Romani C, Lencioni L, by analysis of gene expression levels in brewing yeast. Appl 80 – Mannazzu I & Ciani M (2013) Lachancea thermotolerans Microbiol Biotechnol : 1039 1051. and Saccharomyces cerevisiae in simultaneous and sequential Schaub Y, Dunkler A, Walther A & Wendland J (2006) New co-fermentation: a strategy to enhance acidity and improve pFA-cassettes for PCR-based gene manipulation in Candida 46 – the overall quality of wine. Food Microbiol 33: 271–281. albicans. J Basic Microbiol : 416 429. Gordon JL, Byrne KP & Wolfe KH (2009) Additions, losses, Schmitz HP & Philippsen P (2011) Evolution of and rearrangements on the evolutionary route from a multinucleated Ashbya gossypii hyphae from a budding 115 – reconstructed ancestor to the modern Saccharomyces yeast-like ancestor. Fungal Biol : 557 568. cerevisiae genome. PLoS Genet 5: e1000485. Sehlmeyer S, Wang L, Langel D, Heckel DG, Mohagheghi H, Goretti M, Turchetti B, Cramarossa MR, Forti L & Buzzini P Petschenka G & Ober D (2010) Flavin-dependent (2013) Production of flavours and fragrances via monooxygenases as a detoxification mechanism in insects: 5 bioreduction of (4R)-()-carvone and (1R)-()-myrtenal new insights from the arctiids (lepidoptera). PLoS ONE : by non-conventional yeast whole-cells. Molecules 18: 5736– e10435. 5748. Stahmann KP, Revuelta JL & Seulberger H (2000) Three Hazelwood LA, Daran JM, van Maris AJ, Pronk JT & biotechnical processes using Ashbya gossypii, Candida Dickinson JR (2008) The Ehrlich pathway for fusel alcohol famata,orBacillus subtilis compete with chemical riboflavin 53 – production: a century of research on Saccharomyces production. Appl Microbiol Biotechnol : 509 516. cerevisiae metabolism. Appl Environ Microbiol 74: 2259– Stark D, Munch T, Sonnleitner B, Marison IW & von Stockar 2266. U (2002) Extractive bioconversion of 2-phenylethanol from Hua D & Xu P (2011) Recent advances in biotechnological L-phenylalanine by Saccharomyces cerevisiae. Biotechnol Prog 18 – production of 2-phenylethanol. Biotechnol Adv 29: 654–660. : 514 523.

Stark D, Kornmann H, Munch T, Sonnleitner B, Marison IW mixed starters based on ester formation and enological & von Stockar U (2003) Novel type of in situ extraction: use traits. Food Microbiol 25: 778–785. of solvent containing microcapsules for the bioconversion of Vuralhan Z, Luttik MA, Tai SL et al. (2005) Physiological 2-phenylethanol from L-phenylalanine by Saccharomyces characterization of the ARO10-dependent, cerevisiae. Biotechnol Bioeng 83: 376–385. broad-substrate-specificity 2-oxo acid decarboxylase activity Toivari M, Vehkomaki ML, Nygard Y, Penttila M, Ruohonen of Saccharomyces cerevisiae. Appl Environ Microbiol 71: L & Wiebe MG (2013) Low pH D-xylonate production with 3276–3284. Pichia kudriavzevii. Bioresour Technol 133: 555–562. Wendland J & Walther A (2005) Ashbya gossypii: a model for Valero E, Millan C & Ortega JM (2001) Influence of oxygen fungal developmental biology. Nat Rev Microbiol 3: 421– addition during growth phase on the biosynthesis of lipids 429. in Saccharomyces cerevisiae (M(3)30-9) in enological Wendland J & Walther A (2011) Genome evolution in the fermentations. J Biosci Bioeng 92:33–38. eremothecium clade of the Saccharomyces complex Verstrepen KJ, Van Laere SD, Vanderhaegen BM et al. (2003) revealed by comparative genomics. G3 (Bethesda) 1: Expression levels of the yeast alcohol acetyltransferase genes 539–548. ATF1, Lg-ATF1, and ATF2 control the formation of a broad Wendland J, Ayad-Durieux Y, Knechtle P, Rebischung C & range of volatile esters. Appl Environ Microbiol 69: 5228–5237. Philippsen P (2000) PCR-based gene targeting in Viana F, Gil JV, Genoves S, Valles S & Manzanares P (2008) the filamentous fungus Ashbya gossypii. Gene 242: Rational selection of non-Saccharomyces wine yeasts for 381–391.

Paper 3

Non-conventional yeasts (NCY) constitute a rich and accessible resource for new traits that can find applications for new biotechnological processes. In this manuscript we investigated a selection of 18 NCYs, encompassing species that represent the entire Saccharomyces clade. The strains have been tested in fermentation condition and fermentation parameters were collected. In addition the strains have been evaluated under stress condition (i.e. osmotic and oxidative stress) and for the production of off-flavours by several plate assays. Even though each species displayed unique characteristic, the volatile profile analysis highlighted four extreme flavour producers: W. anomalus, P. kluyveri and Z. melli showed very high ester production; in contrast D. subglobosus produced high levels of fruity ketones and aldehydes.

third group of five strains produced higher ethanol concentrations than the reference A survey of flavor strain and the fourth group consisting of two production among strains is characterized by either highest or non-conventional no ketone production. We present a comprehensive overview of these strains yeasts identifying those specialized for volatile alcohol and ester formation or ethanol Davide Ravasio, Silvia Carlin, Teun production, respectively. Boekhout, Urska Vrhovsek, Jürgen Introduction Wendland and Andrea Walther Beer is one of the most widely consumed

alcoholic beverages in the world. In 2003 the Abstract worldwide beer production reached around Fungi produce a variety of volatile organic 1.82 billion hectoliters and increased to a compounds (VOCs) during their primary and volume of 1.97 billion hectoliters in 2013 (© secondary metabolism. In particular, in the Statista 2014). beverage industry these volatiles contribute Beer production is divided into two main beer to the flavor and aroma profile of the final styles: Ale and Lager beers. Generally, ale is product. Herein, we evaluated the produced by top fermenting yeasts and fermentation ability and aroma profiles of typically at higher temperatures of 15-30 °C. non-conventional yeasts (NCY) that have This normally results in faster fermentation. been isolated from various food sources. A Ales are known for their fruity aromas which total of 60 strains have been analyzed with are regarded as a distinctive character of top regard to their fermentation and flavor fermenting beers. profile. After a primary screen 18 The word “Lager” actually derives from the representative strains were selected for German word “lagern”, which means “to further strain characterizations based on store”. Lager beers are produced by bottom principal component analysis as well as their fermenting yeasts typically at temperatures fermentation performances. below 15 °C. The aroma of lager beers is more These 18 strains have been further divided neutral compared to Ale type beers since they into four groups: Group 1 consists of six contain lower amounts of fruity flavors. strains with exceptionally high flavor Top fermenting and bottom fermenting production, while group 2 includes five yeasts are two distinct Saccharomyces strains with very low flavor production. A species. Top fermenting yeasts typically belong to the species S. cerevisiae. At the end contribute to the beer by a solvent-like aroma of fermentation these yeasts rise to the which gives a warm mouth feel. The latter two surface of the fermenter creating a thick cell are prevalent for their sweet/rose and layer. Bottom fermenting yeasts are called fruity/banana like aroma, respectively [3]. S. carlsbergensis or S. pastorianus. These Esters that are produced during beer yeasts are interspecies hybrids of two fermentation contribute mainly to the different parental strains that can be fruitiness of the product. They can be divided subdivided into two subgroups namely Saaz into two main groups. The first group of type strains and Frohberg type strains, which acetate esters consists of fruity esters such as derived from 2 independent hybridization ethyl acetate (solvent like, fruity), isoamyl events [1]. acetate (banana) and 2-phenyl acetate (rose). Beer production by spontaneous Ethyl or medium-chain fatty acid esters such fermentation of wort is typical for Belgian as ethyl hexanoate, ethyl octanoate and ethyl Lambic sour beers. Here, the fermentation decanoate belong to the second group of process is not initiated through inoculation esters. They give a fruity apple or wine like with specific yeasts. The wort is filled in a flavor to the beer [4]. shallow open vessel that allows microbial The main aldehyde present in beer is intrusion of bacteria and wild yeasts [2] acetaldehyde. It can be formed of ethanol Beer is a very complex product consisting of when oxygen is present [5]. Other aldehydes both volatile and non-volatile components that contribute to the beer aroma are divided that form the final aroma. The contribution of into three groups: Strecker aldehydes, yeasts to the final flavor bouquet is dependent aldehydes of Maillard reaction and aldehydes on the beer style. In ale beers the contribution of fatty acid oxidation. Strecker aldehydes of the yeast strain to the final aroma is derive from the degradation of leucine, considered to be around 30 % whereas in isoleucine and phenylalanine during wort lager beers the contribution is only 15 %. The boiling and aging. Maillard aldehydes consist volatile compounds produced by the yeast of various heterocyclic compounds which strain belong to different chemical groups. develop due to the reaction of sugars and The most prominent and dominating flavor amino acids. The most abundant Maillard compounds are higher alcohols, esters, aldehyde in beer is furfural [6]. One of the aldehydes, ketones and organic acids. predominant aldehydes of the fatty acid oxidation is trans-2-nonenal (T2N). It creates Among the higher alcohols n-propanol, iso- a stale papery taste that is generally released butanol, 2-phenylethanol and isoamyl alcohol during storage of beer. are the most abundant alcohols. They Further, ketones contribute an important recent years SPME has been widely used as a aroma to the beer. In particular, diacetyl (2,3- solvent-free extraction method for the buanedione) can influence the flavor of beer. analysis of volatiles in beverages. In brewing process it is considered Lately, non-conventional yeasts (NCY) or undesirable due to its unpleasant buttery non-Saccharomyces yeasts have gained flavor and its low taste threshold. Diacetyl importance for fermented alcoholic can be produced during the biosynthesis beverages. They produce various mixtures of pathway of valine and isoleucine. volatile compounds and therewith contribute Acetolactate, an intermediate of the valine to the aroma profile of beverages. [8]. biosynthesis pathway can be oxidized to However, these strains have not been diacetyl preferably under high fermentation characterized in great detail. A vast number temperatures. At the end of fermentation and of non-conventional yeast strains has been during beer maturation diacetyl can be isolated from various food sources and reabsorbed by the yeast cells and converted to collected in a strain collection database such 2,3-butanediol via the intermediate acetoin. as CBS (Centraalbureau voor Schimmel- Both acetoin and 2,3-butanediol have a much cultures, Utrecht, Netherlands). In this study higher taste threshold than diacetyl [7]. we aimed at covering a broad spectrum of Short-chain organic acids can greatly species isolated from different substrates like influence the final beer flavor. They can berries, fruits, cheese, fruit flies or even soil reduce the pH during fermentation and and spanning a broad evolutionary distance therewith give a sour taste to the beer. within the Saccharomycetes. Medium-chain fatty acids such as caproic, Material and Methods caprylic and capric acid can add a rancid goaty flavour to the beer. Since they can Strains and media influence the foam stability of the beer they Strains used in this study are shown in Table are undesirable compounds. 1. Prior to spotting assays yeast strains were grown in YPD (1 % yeast extract, 2 % peptone, Given the low concentration and volatile 2 % glucose) at room temperature o/n. Then, nature of these aroma compounds gas the cells were harvested by centrifugation, chromatography coupled to mass washed in sterile deionized water and finally spectrometry (GC–MS) offers an optimal diluted to an OD of 0.1. Dilution series in technique to analyze the flavor profile of beer. 600 steps of 1:10 were then spotted onto the Coupled to this, several sampling methods plates. For growth tests on granmalt medium such as solid-phase micro extraction (SPME), (150 g/l malt granules, 5 g/l yeast extract, 2% steam distillation extraction (SD) and liquid- agar), plates were incubated for 2 days at liquid extraction have been developed. In the room temperature prior to photography. YPD Analytical methods GC/MS plates were incubated at 10, 20 and 37°C for Volatile compounds were detected and up to 4 days to assess the strains’ thermo analyzed by using a Thermo Scientific TSQ tolerance. SD based medium (20 g/L glucose, Quantum GC Triple Quadropole GC-MS. 2- 6.7 g/l YNB w/o amino acids but with octanol was added as an internal standard to ammonium sulphate, 20 g/l agar) with each sample. The standard was chosen as a addition of cinnamic acid (100 µg/l) was used compound known not to be present in the to test the strains ability to produce phenolic fermentation samples. 2.5 ml of sample were off-flavors. BiGGY agar plates (Bismuth prepared in a 20 ml vial added with Glucose Glycine Yeast Agar, 1 g/l yeast appropriate amounts of sodium chloride, extract, 10 g/l glycine, 10 g/l glucose, 13 g/l 50µl NaN3 0,1%, 25 µl of the internal agar, 3 g/l sodium sulphite, 5 g/l bismuth standard and ascorbic acid. All samples were ammonium citrate) were prepared according incubated for 10 min at 50°C. The volatile to the manufacturer’s instructions (Fluka). compounds were collected and separated on a These plates were then incubated at 20°C for Divinylbenzene/Carboxen/Polydimethylsilox up to 4 days. ane fiber (DVB-CAR-PDMS) for an extraction time of 40 min. A Solgel-wax column, 30 Fermentation conditions m/I.D 0.25 mm/Film 0.25 μm, was used for Fermentation was performed in 50 ml tubes all analyses. filled with 40 ml YPD (16°Plato) at 20°C. Each fermentation was started with a strain The oven was kept at 40 °C for 4 min then density of OD600 = 0.2. The stirring was set at increased by 6 °C/min to 250 °C and kept at 300 rpm using a triangular magnetic stirrer. the final temperature for 5 min. The injector Fermentation performance was monitored up and interface temperatures were kept at 250 to 14 days by daily measurement of the CO2 °C as well. Helium was used as the carrier gas release. The sugar content was measured by a with a flow rate of 1.2 ml/min. The time for DMA 35 Anton Paar densitometer (medium thermal desorption of analytes was 4 min. gravity in °P). The Plato measurements were The MS detector was operated in full scan taken at the beginning and at the end of each mode at 70 eV with a scan range from 35 to fermentation process. The fermentation was 350 m/z. defined as finished when the CO2 loss was not Data analysis was performed using the increasing anymore and the Plato value did software ThermoXcalibur (Version 2.2 not reduce further for 2 days. A supernatant SP1.48, Thermo scientific). Identification of of 35 ml was then used for GC-MS analysis. compounds was based on comparison with a All fermentations were carried out in mass spectral database (NIST version 2.0). biological duplicates. One characteristic quantifier ion and two to three qualifier ions were selected for each compared to a set of brewing and wine yeast compound. The peak area of the quantifier strains. Each selected strain was assigned to a ion was used for quantification. The code based on its coordinates on a 96 well concentration of each volatile was expressed plate. as µg/l 2-octanol I.S. As was expected, most strains harbored good Multivariate data analysis fermentation properties and produced final Multivariate data analysis was performed ethanol concentrations above 5% v/v (see using StatSoft, Inc. STATISTICA version 8.0 below). Only 12 strains produced ethanol (data analysis software system, 2007). A PCA below 5 % v/v. Eight out of these 12 strains model was employed to simplify data have been unable to grow in fermentative interpretation. The matrix contained the conditions and were therefore repeated in initial 60 strains considered in this study and oxidative conditions. However, these strains the average of the relative 62 VOCs detected. have been discarded from our strain selection The data were standardized to the mean of 0 due to their inability to grow under anaerobic and unit standard deviation was scaled. conditions. Three of the four strains that have Eigenvalues and eigenvectors of the matrix produced low ethanol concentrations in were calculated and the relative plot was fermentation are isolates of created. Wickerhamomyces anomalus, the fourth strains is an isolate of Pichia Results and Discussion membranifaciens. Out of this group of strains Strain selection and identification of with a low ethanol production we have representative isolates chosen strain F10, a Wickerhamomyces Via collaboration of the EU-ITN Cornucopia, anomalus strain as representative for our focusing on non-conventional yeasts, with the subset of strains. CBS Fungal Strain Collection we selected 53 Overall we determined 62 different different species of the Saccharomyces clade headspace volatiles in the fermentation (Table 1). We prioritized strains with a proven samples of the 60 yeast strains tested. Key background from fermented liquids, fruits, volatile classes that were detected include vegetables, or meat. These strains, therefore, esters, alcohols aldehydes, ketones and acids. may have evolved superior features in However, the number and amount of VOCs fermenting different sugars and thereby detected varied considerably among the producing ethanol or have been recognized as different species. The complete list of contributing flavor to the samples. All strains identified volatiles for each species is shown were fermented in 16 °Plato YPD. Then flavor in Table 2. Each volatile was grouped into the profiles were analyzed using GC-MS and respective chemical class. Esters were the most prominent group of volatiles. In total 22 the 2 dimensional projection of our data. The different esters could be identified and within first principal component (PC1) explained this group ethyl-esters were particularly 22.44 % of the total variability between the dominating such as ethyl hexanoate and ethyl fermentation results. PC1 was mainly acetate, a fruity wine or apple like flavor and associated with parameters like small ethyl sweet pear drop flavor, respectively. esters. PC2 explained 13.40 % of the total variability. It was more associated with Alcohols comprised the second major aldehydes and ketones. Both principal chemical group produced during these components together explained 35.84 % of fermentations. Besides ethanol we found 14 the variability between the 60 fermentations. different alcohols. Yet, only two compounds, The array of species populated all four 2-phenylalcohol, known for its rose aroma, quadrants of the PCA plot showing the large and isoamyl alcohol, a banana like flavor, diversity amongst these yeasts. However, were produced by all the strains we analyzed. within the dataset different species were well- Fermentations were conducted in tall tube segregated according to the respective genus. cylinders with no aeration. Therefore, oxygen For example, Wickerhamomyces and Pichia depletion occurs rapidly under these sp. occupied the bottom right quadrant and conditions. Formation of aroma alcohols is separate clearly from the other species based favored under oxygen limited conditions over on their production of small fruity esters the production of aroma acids. Thus only six particularly ethyl acetate, ethyl hexanoate, acids were identified in our VOC profiling isoamyl acetate and phenethyl acetate. including acetic acid, butyric acid, decanoic Zygosaccharomyces, Saccharomyces and acid, hexanoic acid, isovaleric acid and Debaryomyces sp. were found in the upper octanoic acid. Of these acids acetic acid and quadrants. These species had the highest butyric acids were produced by almost all scores for aldehydes, acids and ketones which strains and were the most abundant clearly separated them from the other compounds in this chemical class. species. The genera Candida, Lachancea, Kazachstania, Meyerozyma and To provide an overview of the yeast VOC Hanseniaspora species were found to profiles, multivariate data analysis (PCA, produce similar volatile profiles as lager yeast principal component analysis) was applied as strains based on those PCA. Therefore, these a statistical technique to allow visualization species grouped closely together. and grouping of the yeast strains based on the class and amount of volatiles. This PCA was The loading plot (PC1 vs. PC2) represented in based on data derived from biological Figure 1B interprets the relationship between duplicates of all 60 strains. Figure 1A shows the biochemical variables (VOCs). Small esters are well separated by the PC1. Saccharomyces cerevisiae (E11) isolated from Aldehydes, acids and ketones, on the other Spanish sherry produced very high amounts hand, cluster together as they are significantly of fruity alcohols (isoamyl alcohol and 2- present in Zygosaccharomyces, Saccharo- phenyl ethanol) and acids (acetic acid and myces and Debaryomyces species. This butyric acid). Zygosaccharomyces mellis analysis allowed the grouping of our strains (G4) isolated from wine grapes showed an in two major classes based on their overall increased production of volatiles such production of either esters and alcohols or as fruity alcohols and esters as well as acids, aldehydes and ketones. aldehydes and ketones. In contrast, five strains, namely Geotrichum candidum (C3), Based on our results of the PC analysis we Kazachstania servazii (C9), Kluyveromyces continued our experiments with a subset of dobzhanskii (D4), Meyerozyma 18 strains. The selection of these strains was guilliermondii (D11) and Nakaseomyces done by considering the results of the bacillisporus (E2) showed an overall reduced multivariate analysis of the VOCs as well as volatile production in comparison to other factors obtained during and after the Weihenstephan (G10). However, Kazach- fermentation such as ethanol production, stania servazii (C9) produced the highest fermentation speed or efficient sugar ethanol concentration among the stains conversion. As a reference for the evaluation tested. Another five strains were selected due of strains we used the fermentation results of to their higher ethanol concentrations the strain Weihenstephan WS34/70 (G10). compared to the reference strain namely Six strains were selected based on their Clavispora lusitaniae (B8), Lachancea extraordinary volatile profile: In Candida fermentati (D6), Saccharomyces cerevisiae diversa (B4) and Clavispora lusitaniae (B8) (F2), Saccharomycodes ludwigii (F3) and fermentations we measured very high Schwanniomyces occidentalis (F4). The amounts of furfural. Candida diversa was latter is interesting and important for its further recognized for its high production of amylolytic system. This fungus can degrade ethyl hexanoate, a wine or apple like flavor, numerous different starches completely such and butyric acid. Pichia kluyveri (E7) and as potato, barley or wheat starch [9]. Wickerhamomyces anomalus (F10) produced The remaining two strains were chosen exceptionally high amounts of fruity alcohols according to a unique trait found in their and their respective esters such as isoamyl volatile profiles: Hanseniaspora vineae (C7) alcohol, 2-phenylethanol and isoamyl acetate has been chosen since this strain produced and 2-phenyl acetate as well as ethyl the highest acetoin concentration during hexanoate. Further these strains produced fermentation. Acetoin is an intermediate in very high amounts of acetic acids. the conversion of diacetyl to 2,3-butanediol. equal amounts of CO2 every day. During the During fermentation yeast produces diacetyl fermentation of this strain we observed a that is considered a buttery off-flavor. During water-insoluble layer and a biofilm of cells maturation the diacetyl can be taken up and appearing on the fermentation cylinder. This consumed by yeast by an enzymatic suggests that this layer may have reduced the conversion to acetoin and subsequently to evaporation of CO2 and therewith caused the

2,3-butanediol and therewith remove the off- constant and continuous CO2 release. In fact, flavor from the liquid [10]. In contrast, the fermentation liquid of Wickerhamomyces Torulaspora microellipsoides (F7) was the anomalus showed the lowest pH measured only strain of our selection that did not among all stains suggesting that the dissolved produce any ketones in the fermentation CO2 was contributing to the acidification of experiment. the liquid. The formation of such a film has been reported as a typical phenomenon in Fermentation results and strain stored wines. The residual amount of oxygen characterization in the headspace of the fermenters then These 18 strains have been processed further triggers the formation of the biofilm. It has through strain characterization tests on been shown that this bio-layer produced various growth conditions and for the mainly acetic acid, acetaldehyde and acetate production of common off-flavors that are esters, volatiles that have been identified in present in lager beer, namely H2S and high concentrations in W. anomalus phenolic off-flavors. The results are shown in fermentation [11]. Table 3. The table values are represented as a heat map. The most rapid fermentation as represented

by fast CO2 loss was observed with the beer All strains were adjusted to the same optical production strains WS34/70 (G10). However, density prior to fermentation start to allow this strain did not cause the highest CO2 the tracking of the fermentation speed of each release. Pichia kluyveri (E7) finished single strain. Fermentation rates were fermentations after four days with the highest followed by measuring the CO2 release per CO2 release. The lowest CO2 release occurred day (Figure 2A). As prerequisite we only in the fermentation of Debaryomyces selected strains that were able to ferment the subglobosus (B11). Fermentation was finished medium within seven days. As only exception after four days, although only about two we selected Wickerhamomyces anomalus thirds of the average CO2 release occurred. (F10) due to its exceptional volatile profile. Surprisingly, the remaining sugar content of Different to the other 17 strains that showed a this fermentation was rather low with 3.45 °P sigmoidal CO2 release W. anomalus had almost a linear fermentation curve releasing and the ethanol produced was high with Growth on other carbon sources and 7.44 %. temperatures In order to test the strains ability to grow on Since we conducted the fermentation using low or high temperatures such as 10 °C and glucose based medium we did not expect 37°C we setup dilution series on plates and drastic changes in the pH of the final liquid observed their growth after four and two days due to the buffering capacity of the medium respectively (Table 3). We found that 60 % of itself. Nevertheless, we observed the lowest the strains were able to grow on high pH after fermentation with temperature whereas none of the strains was Wickerhamomyces anomalus (F10; pH 4.79). showing full growth at the low temperature of Apart from the suggested dissolution of CO2 10 °C. Further we tested the ability of the in the fermented liquid this strain strains to grow on an alternative carbon additionally produced the highest amounts of source such as maltose. Only two strains of acids which explain this low pH of the final our selection showed full growth on maltose liquid. In contrast, Clavispora lusitaniae (B8) plates namely Candida diversa (B4) and produced the lowest amount of acids and Clavispora lusitaniae (B8). therewith resulted in the highest pH of 6.07. Off-flavor formation Ethanol production We focused our attention on the production For most of the strains we could observe a of off-flavors that are important factors for correlation of the sugar consumption and the the selection of strains in brewing industry produced ethanol (Figure 2B). The highest such as diacetyl, phenolic compounds like 4- ethanol production was observed for vinylguaiacol and hydrogen sulfide (Tables 2 Kazachstania servazii (C9) with 9.2 % v/v. and 3). Wickerhamomyces anomalus (F10), on the other hand, produced the lowest amount of Diacetyl (butanedione or butane-2,3-dione), a ethanol correlating with the highest buttery like flavor, can be detected by GC-MS. remaining sugar content. Hanseniaspora Within our selected strains we found four vineae (C7) and Pichia kluyveri (E7) strains where we could detect diacetyl. Since produced a lower ethanol yield than the diacetyl can be converted to 2,3-butanediol estimation due to the sugar consumption through the intermediate acetoin we would imply. This could be connected to the compared the acetoin production of the increased production of acetic acids in both strains as well. Both acetoin and 2,3- strains. Additionally, Pichia kluyveri butanediol can diffuse out of the cell, but produced vast amounts of ethyl acetate. neither of them contributes to the undesired buttery flavor [7]. Hanseniaspora vineae (C7) produced the highest amounts of acetoin but did not show detectable amounts of diacetyl. a high sulfite reductase activity and produce a

[12]. detectable amount of H2S while white colonies produced no detectable sulfide [15]. Phenolic compounds such as 4-vinylguaiacol However, this assay does not give absolute are produced by enzymatic decarboxylation of quantitative measurements but a tendency of ferulic acid. When found in beers they are sulfite reductase activity. We found three responsible for a pungent clove-like aroma. strains with apparent low sulfite reductase Key enzymes for the ferulic acid conversion activity, namely Nakaseomyces bacillisporus are PAD1 and FDC1. In our screening we used (E2), Saccharomycodes ludwigii (F3) and cinnamic acid as a substrate to identify POF+ Torulaspora microellipsoides (F4). To and POF- strains. POF+ strains express the measure the H2S formation during genes PAD1 and FDC1 and can therewith fermentation the lead acetate method can be decarboxylate cinnamic acid to styrene to used. Here, hydrogen sulfide interacts with circumvent the inhibitory effects of cinnamic the lead in the lead acetate matrix to form a acid [13]. Of the 18 strains in our study only dark precipitate. This allows an estimate of two showed very high production of phenolic the volatile H2S in the headspace of the flavors, namely Debaryomyces subglobosus fermentation [16]. (B11) and Meyerozyma guilliermondii (D11). However, the production of 4-vinylguaiacol Complexity of flavor production could not be detected by GC-MS since our Within our selection of strains we observed fermentation was based on YPD where ferulic different complexities in flavor production acid is not expected. Typically, ferulic acid (Figure 3). The highest complexity of 47 derives from barley germination where it is volatile compounds resulting in the largest released from the aleurone layers as ester- number of different flavors was produced by linked feruloyl arabinoxylans. During Debaryomyces subglobosus (B11). D. mashing and wort boiling it is released from subglobosus, also known as Candida famata its bound form and found in significant var. flareri, is a natural overproducer of concentrations in the wort [14]. riboflavin and D-Arabitol depending on the iron concentration of the media [17] Strain Hydrogen sulfide is a rotten egg flavour that optimization and development have reached is undesired in fermented beverages. a riboflavin production up to 20 g/l with this Hydrogen sulfide is produced by yeast during species [18] [19]. Interestingly, this strain was fermentation and during the maturation also identified as the one with the lowest CO2 phase. On BiGGY-agar the sulfite reductase release during fermentation. Fermentation activity of a strain can be estimates since was delayed but, nevertheless, finished color of colonies can be correlated with levels already after four days. Further, this strain of H2S production. Brown color colonies have was standing out for its high production of and Z. mellis produced detectable amounts of pyranone and furfural. Pyranone was recently 3-methyl butanal as well as other compounds identified as signaling molecules in bacterial derived from Strecker degradation, such as 5- communications similar to quorum sensing. hydroxymethylfurfural, furfural, 5- [20]. Fruit ketones such as pyranone are methylfurfural (Figure 4). Furfural is a known to be important for processes like colorless oily liquid with the odor of almonds. cheese ripening [21]. In contrast, we found It is further known as one of the components very low levels of fruity esters in our GC-MS found in vanilla. Although both strains have analysis of D. subglobosus fermentation. been isolated from fruits they might have a Already in 2000 Van den Tempel and possible application in meat processing. Z. Jacobsen identified high esterase activity in mellis is already known to be involved in various isolates of D. hansenii which may vinegar fermentation. Together with other cause the low concentration of esters in species of this genus such as Z. bailii, Z. D. subglobosus [22]. bisporus and Z. rouxii, Z. mellis belongs to a group of osmotolerant yeasts [9]. Branched-chained aldehydes are favorable components for the maturation of meat The lowest flavor complexity was found for a products, giving nutty, cheesy and salty notes strain of Saccharomyces cerevisiae (E11) to the food products [23] Usually these isolated from Spanish sherry (Figure 3). aldehydes are produced by Strecker Despite the low number of 31 flavor degradation, a heat-induced process compounds this strain was exceptionally high converting α-amino acids into an aldehydes. in its higher alcohol production namely in Small amounts of these aldehydes are butanol and isoamyl alcohol. Additionally, produced enzymatically during food high amounts of acetic acid and butyric acid fermentation. Within our strain set as well as acetaldehyde were produced D. subglobosus (B11) showed the most (Figure 4). abundant production of aldehydes (Figure 4). With 36 different flavors Wickerhamomyces One other prominent aldehyde was 3-methyl anomalus (F10) is among the strains with butanal. This compound is an intermediate of intermediate flavor complexity (Figure 3). the leucine degradation pathway. It was also However, this strain was noticed as a found in high concentrations in prominent producer of high amounts of Zygosaccharomyces mellis (G4). 3-methyl esters such as butyl acetate, ethyl acetate, butanal is a known flavor component isoamyl acetate and 2-phenyl acetate. Within aldehyde associated with the production of the group of higher alcohols cured-hams by lactic acid bacteria and Wickerhamomyces anomalus produced the Micrococcaceae [24] Both, D. subglobosus highest amount of 2-phenyl ethanol. Moreover, the highest yields of acetic acids Conclusion and acetaldehyde were found in this species In our study we present a comprehensive (Figure 4). As mentioned above W. anomalus overview of the volatile compound formation formed a water insoluble layer on top of the potential of a set of 18 strains that were fermentation consisting of a film that could chosen from a selection of 60 non- produce acetic acid, acetaldehyde and acetate conventional yeasts. We found both esters. generalists that produce a broad variety of organic volatiles but also specialist that had a In earlier studies W. anomalus has been low complexity of different volatile but isolated from a range of cereal based sources. produced vast amounts of certain It has been reported from sourdoughs and compounds. Furthermore, a number of was found as the dominating yeast in strains produced higher ethanol sourdough microbial ecosystems next to S. concentrations than the reference strain. cerevisiae. The prevalence of the fungus was Among the generalists in our selection we associated with its osmotolerance and identified Debaryomyces subglobosus (B11). increased acid tolerance in comparison to S. This strain produced the highest complexity cerevisiae [25]. Further, it was shown that W. of volatiles (a total of 47) but low anomalus can outgrow other yeasts and concentrations of esters when compared to become the dominant fungal species most the reference strain Weihenstephan WS34/70 likely due to its ability to assimilate lactate (G10). However it produced high amounts of produced by lactic acid bacteria [26]. aldehydes and ketones.

Pichia kluyveri (E7) produced a total of 41 Zygosaccharomyces mellis (G4) produced volatiles during our fermentations. It was with 44 volatiles almost as many volatiles as recognized by its high production of volatile D. subglobosus. However the concentration esters that reach up to fivefold concentrations of fruity esters and alcohols was significantly found in the reference strain (Figure 4). increased when compared to the reference Pichia kluyveri is found in 'wild ferments' of strain. wine. Chr. Hansen A/S has selected P. kluyveri for its ability to boost fruit flavours Wickerhamomyces anomalus (F10) and and launched products such as Frootzen™ Pichia kluyveri (E7) were noticed because of that can be used in sequential inoculation their enormous production of volatile esters with standard wine yeast. In a recent study P. but low ethanol production. In contrast, kluyveri together with Kluyveromyces Kazachstania servazii (C9) was identified as marxianus were presented as potential strain with the highest ethanol production starter yeasts for controlled cocoa but with a reduced production of volatiles. fermentation [27]. The yeast biodiversity holds a plethora of strains that show useful characteristics such as ethanol production and flavor formation. This requires a detailed evaluation of the initially identified favorable strains under different conditions to ascertain their properties and potentially promote these strains in specific fermentation regimes. Figures and Tables

Table 1: Strains used in this study. Each selected strain was assigned to a code based on its coordinates on a 96 well plate.

Pos. CBS number Taxon name Substrate of isolation Origin B2 CBS 10151 Candida alimentaria Cured ham Norway B3 CBS 12367 Candida alimentaria Brie Régalou cheese

B4 CBS 4074 Candida diversa Grape must Japan B5 CBS 8058 Candida kofuensis Berries of Vitis coignetiae Japan B6 CBS 1760 Candida versatilis Pickling vat with 22% brine USA B7 CBS 2649 Candida stellate Grape juice France B8 CBS 6936 Clavispora lusitaniae Citrus essence Israel B9 CBS 4373 Debaryomyces fabryi Dry white wine South Africa B10 CBS 767 Debaryomyces hansenii

B11 CBS 2659 Debaryomyces subglobosus Apple Italy C2 CBS 8139 Dekkera anomala Netherlands

C3 CBS 615.84 Geotrichum candidum Brie cheese France Hanseniaspora Fermenting bottled C4 CBS 95 Netherlands guilliermondii tomatoes Hanseniaspora occidentalis C5 CBS 6783 Orange juice Italy var. citrica C6 CBS 2585 Hanseniaspora uvarum Sour dough Portugal Drosophila persimilis (fruit C7 CBS 2568 Hanseniaspora vineae fly) C8 CBS 2352 Hyphopichia burtonii Pollen, carried by wild bees

C9 CBS 4311 Kazachstania servazii Soil Finland C10 CBS 3019 Kazachstania spencerorum Soil South Africa C11 CBS 2186 Kazachstania transvaalensis Soil South Africa D2 CBS 398 Kazachstania unispora

D3 CBS 7775 Kluyveromyces aestuarii Neotredo reynei (shipworm) Brazil D4 CBS 8530 Kluyveromyces dobzhanskii Drosophila sp. Canada D5 CBS 1557 Kluyveromyces marxianus Stracchino cheese Italy D6 CBS 7005 Lachancea fermentati Alpechín Spain Drosophila pinicola (fruit D7 CBS 3082 Lachancea kluyveri fly) Either fruit or leaf of fruit D8 CBS 7703 Lachancea waltii tree Berries of Vitis labrusca D9 CBS 5833 Metschnikowia pulcherrima USA (Concord grapes) Insect frass on Ulmus D10 CBS 2030 Meyerozyma guilliermondii USA americana (elm tree) D11 CBS 8417 Meyerozyma guilliermondii Brine bath in cheese factory Netherlands Pos. CBS number Taxon name Substrate of isolation Origin Exudate of Quercus emoryi E2 CBS 7720 Nakaseomyces bacillisporus USA (Emory oak) E3 CBS 2170 Nakaseomyces delphensis Sugary deposit on dried figs South Africa E4 CBS 8255 Pichia Kefyr

Trinidad E5 CBS 2020 Pichia farinosa Fermenting cacao and Tobago E6 CBS 2057 Pichia fermentans Brewers yeast

E7 CBS 188 Pichia kluyveri Olives

E8 CBS 5147 Pichia kudriavzevii Fruit juice

E9 CBS 191 Pichia membranifaciens Wine Italy Fermenting must of E10 CBS 429 Saccharomyces cerevisiae champagne grapes E11 CBS 1250 Saccharomyces cerevisiae Sherry Spain F2 CBS 1782 Saccharomyces cerevisiae Super-attenuated beer

F3 CBS 820 Saccharomycodes ludwigii Grape must Germany Schwanniomyces F4 CBS 2863 Soil of vineyard Spain occidentalis Schwanniomyces F5 CBS 6741 Soil South Africa polymorphus var. africanus F6 CBS 133 Torulaspora delbrueckii Ragi Indonesia F7 CBS 427 Torulaspora microellipsoides Apple juice Germany Wickerhamomyces F8 CBS 248 Red currants Netherlands anomalus Wickerhamomyces F9 CBS 249 Berries anomalus Wickerhamomyces F10 CBS 261 Ragi Indonesia anomalus Wickerhamomyces F11 CBS 262 Beer anomalus Zygosaccharomyces bailii G2 CBS 4689 Grape must Italy var. bailii Zygosaccharomyces G3 CBS 1082 Tea-beer fungus Indonesia bisporus G4 CBS 726 Zygosaccharomyces mellis Wine grapes Germany G5 C1030 Saccharomyces pastorianus Brewers' yeast

G6 C1039 Saccharomyces cerevisiae Wine yeast

G7 C746 Saccharomyces cerevisiae Brewers' yeast

Saccharomyces G8 CBS 1513 Brewers' yeast carlsbergensis G9 C482 Saccharomyces cerevisiae Brewers' yeast

G10 WS34/70 Saccharomyces pastorianus Brewers' yeast

G11 C598 Saccharomyces cerevisiae Laboratory strain

Table 2: Complete list of volatiles detected by the total list of 60 strains

B2 Stdev B3 Stdev B4 Stdev B5 Stdev B6 Stdev B7 Stdev B8 Stdev B9 Stdev B10 Stdev B11 Stdev C2 Stdev C3 Stdev C4 Stdev C5 Stdev C6 Stdev C7 Stdev C8 Stdev C9 Stdev C10 Stdev C11 Stdev D2 Stdev D3 Stdev D4 Stdev D5 Stdev D6 Stdev D7 Stdev D8 Stdev D9 Stdev D10 Stdev D11 Stdev Alcohols Benzyl alcohol 1,1 0,1 1,4 0,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 4,9 0,5 0,0 0,0 3,5 0,1 4,1 1,1 0,0 0,0 1,6 0,4 4,9 0,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3,4 0,6 0,0 0,0 0,0 0,0 0,0 0,0 2,4 0,2 0,0 0,0 2,6 0,1 2,3 0,3 3,0 0,5 4,5 1,2 Butanol 0,0 0,0 0,0 0,0 301,2 63,3 452,6 27,7 69,0 13,3 594,4 29,5 434,7 19,2 243,1 71,6 361,0 78,9 412,9 41,3 146,2 24,0 134,1 83,7 349,7 28,1 192,2 53,5 364,6 20,3 505,8 16,0 172,6 37,9 322,1 53,2 331,7 58,3 299,3 14,4 504,4 10,7 416,7 16,7 337,5 68,4 521,2 65,8 268,2 38,1 385,0 132,0 380,4 87,8 303,7 14,3 288,4 40,4 385,6 86,9 Dodecanol 8,5 0,7 13,2 5,0 13,2 2,1 14,9 6,8 8,1 1,8 23,8 6,4 8,4 3,4 7,4 1,3 11,2 1,7 14,0 5,8 2,9 0,4 3,0 2,0 9,2 0,3 6,0 4,9 5,2 0,0 14,5 1,3 2,2 1,6 10,0 11,0 5,6 4,0 6,1 6,6 4,4 0,3 8,5 2,4 6,5 0,2 9,8 4,4 4,9 1,5 5,5 0,9 5,2 1,0 5,0 1,9 5,1 2,2 11,7 4,3 Fenchyl alcohol 2,5 0,6 2,1 0,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,5 0,2 0,0 0,0 0,0 0,0 1,6 0,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Furaneol 0,0 0,0 0,0 0,0 11,1 6,4 0,0 0,0 0,0 0,0 29,2 0,8 6,5 0,1 12,1 1,7 26,3 1,1 37,2 2,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Isoamyl alcohol 50,0 11,8 20,1 8,6 342,6 26,0 706,6 62,0 206,7 1,4 729,5 271,4 579,8 19,5 244,8 13,6 499,2 61,6 589,9 34,8 523,1 50,1 580,8 86,3 533,5 95,3 414,8 13,2 455,2 37,4 767,1 11,8 527,3 40,4 544,3 53,9 491,1 33,1 406,7 52,2 523,9 93,6 589,6 23,7 695,7 0,3 830,0 101,6 658,9 66,6 597,4 2,4 581,0 19,5 449,1 28,0 477,3 23,6 502,1 64,1 Propanol 0,0 0,0 0,0 0,0 0,0 0,0 15,4 3,1 0,0 0,0 0,0 0,0 12,4 1,1 31,6 2,4 5,7 0,2 4,5 0,7 11,5 1,5 3,3 0,9 11,5 2,4 6,9 0,5 10,2 0,7 29,0 8,7 16,1 4,5 10,3 0,6 7,8 1,3 14,6 8,7 20,6 6,0 31,7 9,1 15,4 0,1 19,6 2,9 11,4 1,7 13,6 1,9 21,4 2,4 7,9 0,5 12,0 0,1 12,7 1,7 2-Ethyl-1-hexanol 0,0 0,0 3,3 1,9 4,7 0,4 7,0 0,6 4,7 0,0 5,3 1,8 4,6 0,0 5,3 0,4 5,4 0,3 4,9 0,5 3,7 0,9 3,6 0,2 6,0 1,6 5,5 2,0 4,4 0,4 7,2 0,6 3,8 0,4 4,4 0,2 3,8 0,2 2,9 1,2 6,4 1,7 4,9 0,2 4,6 0,5 8,0 5,2 3,7 0,7 2,7 0,2 4,4 0,3 4,7 0,3 4,0 0,2 7,0 3,2 2-Furanmethanol 0,0 0,0 0,0 0,0 52,6 8,5 0,0 0,0 0,0 0,0 164,7 14,0 49,6 4,1 81,4 6,1 139,9 6,7 159,4 11,2 14,4 6,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Methyl propanol 0,0 0,0 0,0 0,0 6,0 1,7 42,4 2,5 0,8 0,0 44,1 17,4 31,1 2,4 17,8 3,5 26,2 1,2 24,4 3,9 25,0 2,5 29,7 7,4 17,9 1,1 14,6 3,0 11,5 1,3 12,5 2,9 19,3 6,6 12,8 3,6 7,2 0,3 10,3 1,6 9,7 3,6 18,8 3,1 35,9 3,1 36,5 3,2 30,4 3,9 13,8 3,9 23,5 1,2 13,8 1,3 11,1 0,5 11,9 2,0 2-Nonanol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,9 0,3 2,3 0,3 0,0 0,0 2,2 0,2 1,9 0,1 0,0 0,0 1,2 0,4 0,0 0,0 8,5 0,9 3,1 1,4 2,0 0,1 0,0 0,0 1,6 0,0 2,7 0,5 23,1 20,9 3,9 0,6 21,5 4,5 1,4 0,0 1,2 0,1 1,0 0,1 1,2 0,2 1,9 0,0 16,1 2,1 10,4 1,2 2,1 0,4 2-Phenyl ethanol 239,2 25,0 317,5 103,7 445,0 108,5 373,8 66,3 154,5 15,5 577,4 15,5 425,0 7,9 73,7 5,2 302,6 20,1 365,9 12,9 202,9 11,8 333,7 29,9 415,2 10,0 348,0 46,7 329,7 55,5 608,6 52,0 215,2 85,3 416,5 32,7 314,6 28,5 314,2 15,8 400,2 50,7 430,8 18,9 373,1 36,8 448,5 89,4 382,1 8,2 533,7 25,4 431,0 8,7 329,5 1,3 312,3 27,6 383,8 59,4 3 Ethoxy - 1 Propanol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,6 0,8 0,0 0,0 1,1 0,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3-(Methylthio)-1-propanol 1,2 0,1 0,0 0,0 48,3 3,4 28,3 1,1 0,0 0,0 55,4 5,4 127,5 20,6 0,0 0,0 30,8 9,6 36,8 7,8 63,9 4,7 18,1 4,6 30,8 4,7 54,4 7,5 29,6 2,5 352,6 55,2 33,2 0,9 31,3 18,4 0,0 0,0 0,0 0,0 28,3 7,4 38,4 10,7 8,0 2,6 132,7 50,4 11,8 5,3 45,8 15,6 12,6 0,9 0,0 0,0 19,9 2,1 40,8 7,6 Esters Butyl acetate 0,0 0,0 0,0 0,0 15,8 8,6 5,9 3,5 0,1 0,0 3,3 0,2 70,6 7,1 3,5 2,1 3,7 0,2 3,0 2,0 34,8 0,2 7,4 4,2 49,2 5,5 1,3 0,1 134,3 6,1 47,1 2,0 3,4 1,1 49,4 4,0 9,5 7,0 13,2 1,2 13,3 8,7 60,9 14,4 5,8 1,1 52,5 0,0 5,4 0,4 74,1 4,2 22,1 1,9 19,0 4,0 53,1 4,6 12,4 3,1 Ethyl (4E)-4-decenoate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 30,0 1,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 244,9 36,3 0,0 0,0 173,9 7,7 198,6 9,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 159,7 13,3 10,7 0,1 101,5 6,4 13,5 1,0 17,4 3,9 Ethyl 2-methylbutyrate 0,0 0,0 0,0 0,0 1,8 0,4 0,0 0,0 0,7 0,1 0,5 0,7 0,0 0,0 0,0 0,0 0,9 0,1 0,6 0,1 0,0 0,0 5,4 1,4 0,3 0,0 9,7 3,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,4 0,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,7 0,7 0,6 0,4 0,0 0,0 Ethyl acetate 0,0 0,0 0,0 0,0 148,7 47,0 124,0 0,8 0,0 0,0 0,0 0,0 867,7 58,6 0,0 0,0 198,2 0,4 177,0 30,7 471,4 36,3 303,1 74,6 662,8 21,1 143,7 9,7 1130,9 39,5 985,2 111,3 168,8 2,0 625,0 19,0 321,1 11,5 102,2 37,0 112,3 2,8 524,5 25,9 181,5 32,7 497,0 44,6 183,9 9,5 688,8 94,1 326,5 15,9 619,2 57,3 562,4 67,1 116,5 34,8 Ethyl butanoate 0,0 0,0 0,0 0,0 159,7 80,4 23,3 3,1 10,9 1,1 87,8 28,3 18,4 6,3 32,3 1,6 80,8 10,0 47,8 10,4 22,2 0,1 69,4 15,6 19,2 0,0 94,7 0,7 25,1 5,1 94,6 6,5 56,6 8,0 17,5 9,3 34,2 6,0 35,2 17,2 12,3 5,1 137,8 31,6 36,9 6,7 36,5 1,6 29,4 1,2 55,0 10,6 43,9 4,8 138,4 2,1 88,4 2,1 17,0 3,9 Ethyl decanoate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 124,0 18,4 62,9 28,2 0,0 0,0 0,0 0,0 93,8 1,6 51,5 14,7 154,5 6,6 348,0 46,9 0,0 0,0 254,4 3,8 330,3 27,4 0,0 0,0 158,1 44,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 334,5 38,4 0,0 0,0 382,1 83,4 117,1 14,7 70,5 13,4 Ethyl dodecanoate 0,0 0,0 0,0 0,0 53,5 9,4 24,9 4,2 0,0 0,0 0,0 0,0 11,8 7,0 7,5 1,1 58,6 15,4 46,8 19,8 5,9 2,2 4,7 2,1 163,0 29,5 13,1 0,4 89,4 12,6 123,1 19,5 9,2 1,8 275,4 8,9 253,4 1,9 0,0 0,0 103,1 22,8 36,5 36,2 6,9 0,8 34,5 5,1 4,4 0,4 71,8 7,6 6,2 0,8 85,1 8,5 133,6 19,7 63,9 9,8 Ethyl heptanoate 0,0 0,0 0,0 0,0 37,2 13,9 1,5 0,6 0,9 0,1 1,8 0,0 0,9 0,0 0,0 0,0 39,7 2,1 11,6 5,3 0,2 0,2 0,7 0,4 6,9 0,7 6,4 1,2 3,1 0,5 0,0 0,0 9,8 1,2 5,4 2,3 19,2 3,5 10,1 1,1 3,2 1,3 2,3 0,5 2,0 0,0 1,8 0,1 2,2 0,3 2,6 0,7 1,6 0,1 11,5 4,3 5,7 0,8 2,7 0,5 Ethyl hexadecanoate 0,0 0,0 0,0 0,0 13,7 0,9 0,0 0,0 0,0 0,0 11,9 7,6 10,5 0,9 7,4 0,2 77,0 1,3 23,0 18,6 5,1 0,6 24,5 19,5 11,7 1,4 26,3 9,4 14,6 3,3 0,0 0,0 24,2 2,7 12,1 7,7 10,1 6,6 5,6 1,7 14,4 0,8 5,7 4,8 0,0 0,0 0,0 0,0 0,0 0,0 9,7 9,7 0,0 0,0 35,8 1,8 26,1 6,0 24,2 5,1 Ethyl hexanoate 0,0 0,0 0,0 0,0 294,5 41,3 69,4 23,9 0,0 0,0 102,7 11,0 34,9 3,4 0,0 0,0 133,7 43,6 38,4 6,2 0,7 0,2 18,3 15,8 56,2 4,6 27,7 8,4 55,2 12,7 92,1 19,7 26,4 7,2 58,5 6,5 71,2 14,9 94,8 2,6 49,8 8,6 69,3 7,9 40,2 10,8 44,6 4,9 70,0 61,9 78,7 18,6 24,4 7,7 103,1 28,0 70,5 17,0 45,4 23,4 Ethyl isobutyrate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 18,8 1,2 13,3 0,6 0,0 0,0 76,2 21,7 0,0 0,0 91,4 2,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 63,3 12,2 5,4 4,8 0,0 0,0 Ethyl isovalerate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,3 0,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 4,9 0,1 0,0 0,0 9,6 3,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Ethyl octanoate 2,1 0,3 0,0 0,0 337,0 28,6 89,7 3,4 1,3 0,1 96,9 25,6 63,0 7,4 4,0 0,0 244,4 16,8 110,4 56,4 2,3 0,2 29,8 28,2 185,7 33,8 0,0 0,0 210,7 72,5 244,3 22,9 30,9 5,7 263,5 38,1 259,8 23,9 246,6 31,9 168,5 32,8 191,8 80,7 89,5 7,5 181,9 27,9 90,4 12,4 274,7 10,4 131,5 1,4 358,6 34,7 210,5 21,4 134,4 15,1 Ethyl propanoate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 54,3 2,2 38,3 6,8 26,3 2,1 0,0 0,0 85,0 7,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Ethyl tetradecanoate 0,0 0,0 0,0 0,0 21,5 4,1 0,0 0,0 0,0 0,0 12,1 8,7 0,0 0,0 6,6 0,9 31,4 0,8 15,6 6,5 0,0 0,0 4,9 2,8 15,3 4,6 18,6 2,6 23,6 6,9 0,0 0,0 13,8 1,0 21,9 0,7 29,6 0,5 12,8 1,2 94,8 12,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 15,7 1,8 0,0 0,0 39,8 4,0 105,1 10,2 105,7 29,2 Isoamyl acetate 0,0 0,0 0,0 0,0 68,0 4,4 21,2 14,2 0,4 0,0 8,1 0,0 94,9 6,3 3,9 1,9 8,8 0,4 11,2 9,1 116,9 1,1 75,2 16,5 90,5 3,8 4,7 0,1 165,3 9,9 321,6 54,2 49,8 10,9 157,5 15,1 43,7 22,4 40,1 8,1 57,0 3,0 252,7 62,4 23,7 0,0 299,2 19,9 24,1 1,0 145,9 58,7 109,9 8,4 47,6 12,9 127,5 29,4 46,5 9,6 Isoamyl butyrate 0,0 0,0 0,0 0,0 5,7 2,0 2,5 0,7 2,4 0,2 3,9 0,6 0,0 0,0 3,7 0,1 13,8 1,8 5,0 3,2 4,2 0,3 17,3 7,1 2,0 0,1 0,0 0,0 1,3 0,0 0,0 0,0 14,3 1,0 2,6 0,9 9,0 0,6 6,4 5,8 1,8 1,0 4,2 0,0 6,6 1,9 0,0 0,0 7,2 0,5 6,4 1,4 2,5 0,2 9,1 1,8 7,0 3,1 0,0 0,0 Isobutyl acetate 0,0 0,0 0,0 0,0 0,0 0,0 3,3 1,8 0,0 0,0 1,5 0,2 2,0 0,8 1,1 0,5 1,4 0,1 1,8 1,3 2,5 0,1 14,6 4,9 3,1 0,2 0,9 0,2 16,6 1,9 6,0 1,1 1,2 0,5 4,4 3,7 1,6 0,6 0,7 0,2 2,2 1,3 5,9 1,9 3,3 1,0 9,0 0,8 2,9 0,2 5,0 2,9 9,0 0,7 7,3 1,2 12,3 3,8 1,6 0,6 Isobutyl butanoate 0,0 0,0 0,0 0,0 0,9 0,4 0,4 0,2 0,2 0,0 1,2 0,1 0,2 0,0 1,4 0,2 1,5 0,2 0,9 0,4 0,5 0,0 2,9 0,3 0,2 0,0 0,6 0,0 0,0 0,0 0,0 0,0 0,9 0,2 0,0 0,0 0,6 0,0 0,0 0,0 0,1 0,1 0,5 0,0 0,5 0,0 0,4 0,0 0,5 0,2 0,4 0,2 0,3 0,0 1,0 0,0 0,5 0,2 0,1 0,2 Phenethyl acetate 0,0 0,0 0,0 0,0 175,9 7,5 23,1 8,7 0,0 0,0 0,0 0,0 69,0 27,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 32,0 11,7 0,0 0,0 0,0 0,0 205,1 8,4 0,0 0,0 86,3 3,2 0,0 0,0 0,0 0,0 0,0 0,0 192,7 36,6 618,5 31,4 20,2 0,3 451,4 91,1 19,6 2,6 315,1 73,5 320,0 4,6 22,8 1,8 48,4 8,3 97,7 2,8 S-methyl thioacetate 0,0 0,0 0,0 0,0 0,6 0,1 8,9 4,5 0,0 0,0 0,0 0,0 12,2 0,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,1 0,2 1,7 0,4 0,0 0,0 0,6 0,0 0,0 0,0 0,2 0,3 0,0 0,0 0,0 0,0 0,0 0,0 1,1 0,5 3,7 1,9 2,5 0,4 3,5 0,2 0,4 0,2 0,4 0,1 0,0 0,0 0,0 0,0 0,0 0,0 2-Methyl propanoate 110,9 24,5 0,0 0,0 27,5 4,6 0,0 0,0 42,0 3,2 0,0 0,0 24,8 0,4 30,4 11,5 30,0 0,4 24,2 4,7 10,4 1,4 79,0 11,4 28,4 2,6 79,5 20,6 25,7 2,2 50,6 19,7 23,4 3,8 21,1 4,9 15,5 6,6 31,7 6,9 25,5 3,5 22,5 2,5 64,2 9,0 66,2 18,3 49,0 13,8 34,0 3,6 24,4 1,3 53,4 2,7 23,3 3,0 17,6 3,7 Acids Acetic acid 51,2 9,4 225,6 78,8 709,2 126,6 296,5 45,6 40,1 11,4 511,2 82,9 323,9 33,2 347,3 82,3 491,4 24,1 608,1 27,5 165,1 26,0 547,4 110,9 736,7 64,0 362,2 16,5 610,4 127,2 1115,2 39,1 359,1 1,4 511,0 1,7 596,4 181,9 454,6 28,8 1038,2 85,4 976,2 292,8 317,0 24,6 874,8 70,0 225,3 5,7 375,8 11,3 232,8 6,3 817,1 30,1 1014,2 5,6 958,6 10,4 Butyric acid 227,2 40,5 375,8 188,7 328,6 99,3 138,4 16,5 255,4 7,0 199,9 66,1 132,4 9,1 172,7 24,3 175,9 13,0 147,4 18,1 129,2 8,8 155,2 6,6 0,0 0,0 192,6 71,9 176,2 35,2 0,0 0,0 186,1 14,8 133,0 22,6 162,3 75,7 126,8 1,2 0,0 0,0 355,4 78,3 178,3 12,5 133,2 8,1 109,4 35,8 0,0 0,0 170,4 1,7 326,1 47,8 260,3 13,1 176,5 42,1 Decanoic acid 0,0 0,0 0,0 0,0 14,0 1,0 30,7 6,3 0,0 0,0 59,4 3,2 16,9 8,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Hexanoic acid 0,0 0,0 0,0 0,0 141,1 44,7 121,4 21,0 0,0 0,0 151,6 31,5 174,6 4,6 0,0 0,0 120,6 13,5 112,6 10,7 0,0 0,0 0,0 0,0 133,9 31,7 85,1 26,6 82,0 6,0 169,7 55,6 0,0 0,0 86,9 10,5 82,5 31,0 0,0 0,0 107,0 4,3 109,0 35,3 69,6 7,2 99,1 35,2 37,5 5,3 54,9 8,4 51,4 1,7 48,3 9,3 73,1 8,6 72,9 16,3 Isovaleric acid 193,1 9,7 249,2 36,9 52,5 9,8 0,0 0,0 147,9 6,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Octanoic acid 0,0 0,0 0,0 0,0 96,1 19,0 103,7 26,7 0,0 0,0 128,9 4,7 144,3 16,4 3,9 1,1 65,8 7,4 69,0 6,3 0,0 0,0 0,0 0,0 90,4 16,6 44,5 10,0 0,0 0,0 142,0 36,1 0,0 0,0 0,0 0,0 48,3 18,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 109,4 34,5 0,0 0,0 57,1 19,8 77,8 6,0 0,0 0,0 0,0 0,0 0,0 0,0 Aldehydes Acetaldehyde 0,5 0,1 9,7 13,1 18,9 4,9 21,8 14,2 1,9 0,2 13,3 1,1 12,2 0,7 6,9 0,6 14,8 1,0 17,0 12,3 11,9 0,5 8,4 1,4 12,6 1,6 25,6 8,6 10,6 1,7 11,4 4,9 66,9 3,9 19,2 1,9 49,7 3,9 13,7 2,1 9,5 0,1 22,1 9,7 26,3 2,1 16,2 1,2 23,0 4,9 12,5 0,9 13,9 1,6 17,9 1,9 20,9 4,6 26,5 15,8 Benzaldehyde 14,0 0,5 11,7 5,1 0,0 0,0 8,7 2,7 28,5 3,9 42,5 12,6 29,8 1,1 10,6 3,1 60,7 32,2 31,0 2,0 4,2 0,5 19,4 1,8 24,8 1,3 28,6 1,9 34,9 2,2 26,6 3,3 5,5 0,8 34,1 8,3 27,9 2,0 21,8 2,2 22,9 4,1 19,6 3,2 6,5 2,2 27,7 1,5 8,4 1,5 31,3 5,1 31,9 1,9 37,3 6,0 32,1 3,8 32,7 8,8 Furfural 2,3 0,5 258,8 64,5 192,7 10,9 138,8 29,8 12,8 11,0 395,1 13,3 147,5 2,0 168,9 0,2 282,8 0,4 358,2 54,1 54,6 21,1 22,2 25,9 53,5 1,1 9,2 0,2 9,6 1,7 12,4 1,4 6,4 1,0 6,9 0,6 5,1 2,5 5,0 3,0 5,5 0,5 6,3 1,7 4,6 0,6 5,1 0,8 4,3 1,1 4,3 0,3 4,4 0,2 3,9 0,8 4,2 0,1 4,0 0,2 Phenyl acetaldehyde 79,4 15,1 91,2 15,5 0,0 0,0 0,0 0,0 9,3 0,6 43,5 4,8 0,0 0,0 22,5 5,2 51,6 7,6 56,0 6,1 15,8 2,6 0,0 0,0 73,9 5,8 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1 -Decanal 1,1 0,4 0,0 0,0 5,7 0,0 6,4 1,3 1,3 0,1 5,6 1,5 3,2 1,8 3,9 2,4 6,3 4,3 4,7 1,4 0,6 0,7 2,5 0,7 0,0 0,0 0,0 0,0 1,3 0,3 0,0 0,0 3,7 0,6 2,5 0,6 3,2 1,9 3,3 0,6 0,0 0,0 3,2 0,1 4,6 1,6 0,0 0,0 2,4 1,2 0,0 0,0 0,0 0,0 4,5 0,8 3,4 1,1 0,0 0,0 1-Nonanal 0,0 0,0 0,0 0,0 0,0 0,0 1,1 0,0 1,0 0,1 1,6 0,7 1,9 0,7 1,1 0,1 0,0 0,0 1,7 0,5 0,6 0,1 1,0 0,1 0,0 0,0 0,0 0,0 0,7 0,2 0,0 0,0 1,8 0,4 1,1 0,3 1,5 0,7 1,7 0,1 0,0 0,0 1,3 0,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1,8 2,0 2,4 0,4 1,7 0,1 2,4 1,0 3-Methyl butanal 83,2 28,8 171,9 39,7 65,3 5,7 0,0 0,0 19,9 2,7 176,5 16,8 31,0 0,8 65,0 8,7 161,5 14,8 218,0 33,3 0,0 0,0 8,9 1,3 307,6 33,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 4-Methyl benzaldehyde 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 15,5 0,2 11,5 1,2 18,1 0,4 0,0 0,0 0,0 0,0 6,8 0,4 7,3 2,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 9,3 1,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 10,5 0,8 7,0 0,2 8,7 0,6 6,9 0,6 8,2 1,1 5 Methyl furfural 0,0 0,0 0,0 0,0 78,0 6,7 46,0 38,3 5,4 3,1 153,3 10,7 36,5 0,9 65,0 3,7 136,4 1,8 187,5 7,9 21,9 6,9 8,3 8,7 68,3 6,2 0,0 0,0 11,9 1,0 0,0 0,0 6,3 0,8 6,1 2,0 4,9 0,5 0,0 0,0 6,7 0,3 6,5 1,4 3,9 1,6 6,3 0,9 3,5 0,0 4,5 0,5 3,8 0,1 3,3 0,2 2,9 0,1 2,8 0,5 5-Hydroxymethylfurfural 0,0 0,0 0,0 0,0 283,5 11,4 195,5 31,7 20,7 20,3 554,1 4,1 252,9 4,9 188,7 25,0 287,5 7,7 317,9 70,2 109,7 47,0 0,0 0,0 29,9 1,8 5,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Ketones Acetoin 0,0 0,0 0,0 0,0 2,3 0,1 2,8 0,6 4,1 0,3 47,3 20,0 27,2 3,9 26,7 9,2 29,8 4,2 42,0 34,6 7,6 1,4 5,1 2,6 53,8 9,8 14,4 7,5 43,1 18,8 203,0 8,0 17,9 2,7 94,6 2,1 14,5 3,1 3,7 2,8 2,8 0,7 5,6 3,4 2,3 0,4 2,7 0,4 1,8 0,1 1,0 0,0 0,9 0,1 2,8 0,2 2,9 0,9 5,4 0,9 Diacetyl 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 12,4 1,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3,1 1,0 29,2 5,1 8,0 3,5 17,6 4,1 5,2 1,1 28,2 4,7 7,3 0,4 14,7 3,0 13,6 0,6 32,2 1,3 19,1 2,6 14,8 5,3 Pyranone 0,0 0,0 0,0 0,0 175,4 48,8 133,2 15,9 0,0 0,0 460,3 97,9 163,0 7,5 170,4 20,9 266,2 5,3 319,2 26,3 34,3 18,1 0,0 0,0 190,6 24,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Cyclopentene-1,4-dione 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Dodecanone 12,3 0,9 21,8 3,7 23,4 18,9 0,0 0,0 14,6 2,0 61,2 8,2 0,0 0,0 0,0 0,0 0,0 0,0 40,6 29,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 11,3 0,6 25,4 1,9 8,5 1,7 60,5 6,5 30,8 33,8 0,0 0,0 0,0 0,0 0,0 0,0 13,8 7,9 0,0 0,0 8,4 0,2 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Methyltetrahydrothiophen-3-one0,0 0,0 0,0 0,0 21,3 2,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Nonanone 4,8 0,1 0,0 0,0 1,3 0,3 2,7 0,3 0,6 0,1 4,2 0,3 0,9 0,0 0,4 0,0 5,8 0,6 6,2 1,6 0,6 0,4 1,0 0,1 0,5 0,2 10,8 3,3 0,0 0,0 2,2 0,5 5,4 0,2 2,7 1,3 1,6 0,1 16,2 2,4 11,3 3,7 14,8 1,8 2,5 0,2 0,0 0,0 2,1 0,1 1,2 0,1 6,2 1,7 10,8 0,4 26,8 2,6 3,4 0,3 2-Undecanone 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,2 0,0 0,0 0,0 0,9 0,2 1,5 0,0 1,8 1,0 1,0 0,3 1,4 1,1 1,7 0,7 0,7 0,0 0,4 0,1 0,5 0,1 0,3 0,0 0,3 0,0 1,9 0,2 2,9 1,0 3,5 0,2 1,4 0,1 Pyrazine 2,5-Dimethyl-3-ethylpyrazine 0,4 0,1 0,0 0,0 8,2 1,0 6,2 0,7 6,4 0,3 7,6 0,8 8,8 0,3 0,5 0,1 9,7 0,8 8,1 0,7 0,4 0,0 3,8 1,3 6,9 0,2 8,1 0,4 5,2 0,6 5,4 0,6 0,2 0,0 6,8 1,0 5,6 0,3 4,0 0,4 4,3 0,4 6,6 0,4 5,7 0,4 6,6 0,7 4,9 0,5 6,2 0,2 5,8 0,1 5,9 0,6 4,8 0,6 5,3 0,8 2,6-Dimethylpyrazine 4,5 0,2 2,7 0,2 29,0 0,4 20,4 2,5 34,7 0,2 20,1 3,7 21,0 0,7 4,2 0,4 23,2 1,8 21,4 4,4 4,2 0,2 12,2 3,0 34,6 5,8 22,6 1,6 14,2 0,8 16,5 3,7 2,9 0,4 14,6 0,4 20,9 0,4 9,5 0,8 18,6 3,1 13,7 4,0 15,1 0,5 18,8 1,5 14,3 0,6 15,4 1,5 14,7 1,4 16,1 0,2 17,4 0,6 20,2 4,4

E2 Stdev E3 Stdev E4 Stdev E5 Stdev E6 Stdev E7 Stdev E8 Stdev E9 Stdev E10 Stdev E11 Stdev F2 Stdev F3 Stdev F4 Stdev F5 Stdev F6 Stdev F7 Stdev F8 Stdev F9 Stdev F10 Stdev F11 Stdev G2 Stdev G3 Stdev G4 Stdev G5 Stdev G6 Stdev G7 Stdev G8 Stdev G9 Stdev G10 Stdev G11 Stdev Alcohols Benzyl alcohol 1,6 0,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1,8 0,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,6 0,0 0,0 0,0 5,0 0,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Butanol 360,1 16,0 372,6 6,8 369,5 12,6 449,4 115,4 199,0 76,1 698,6 37,9 91,6 31,4 719,3 6,2 494,1 88,7 1660,4 14,9 383,2 78,4 397,1 21,0 596,8 48,8 437,2 67,6 461,4 27,2 734,0 12,0 668,4 49,1 565,9 76,7 755,4 167,8 1290,6 145,1 518,6 7,7 1826,7 152,2 1189,7 129,3 396,5 14,3 822,7 373,3 476,6 55,8 458,4 1,5 398,8 24,0 399,7 53,4 801,4 123,8 Dodecanol 4,5 0,1 3,7 1,1 2,3 0,0 2,5 0,9 1,0 0,1 6,8 4,6 1,6 0,6 0,0 0,0 0,0 0,0 0,0 0,0 1,3 0,3 0,0 0,0 7,2 3,9 2,2 0,9 1,6 0,1 2,1 0,1 3,2 2,1 3,1 1,2 6,4 0,5 4,7 3,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1,5 0,2 2,0 0,4 2,0 0,2 0,0 0,0 0,0 0,0 Fenchyl alcohol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Furaneol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 54,0 7,4 29,1 3,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 6,1 1,6 56,3 31,7 33,4 7,0 10,7 6,2 0,0 0,0 0,0 0,0 10,9 1,3 8,7 3,5 16,1 1,5 24,7 14,0 Isoamyl alcohol 457,9 16,0 424,2 3,2 375,5 4,6 728,7 138,0 603,8 1,8 870,6 42,4 502,9 22,9 1840,0 151,4 819,3 75,1 3702,4 46,1 565,3 95,8 645,8 21,8 444,8 25,1 740,0 103,4 539,0 64,1 615,5 49,5 632,2 55,3 632,8 99,7 1081,6 52,4 2363,5 256,6 409,1 34,9 1798,2 77,3 1973,2 461,9 709,7 267,2 511,7 641,5 781,6 39,2 748,2 76,1 800,6 19,9 965,7 39,5 726,8 15,9 Propanol 15,2 1,6 16,3 0,5 6,3 5,5 13,5 2,6 5,6 1,3 31,6 7,4 2,3 0,1 0,0 0,0 3,9 0,1 66,4 7,5 0,0 0,0 26,4 5,4 15,0 2,3 39,2 15,4 15,7 2,2 19,8 2,8 16,8 3,6 17,7 5,6 26,0 9,8 18,2 3,0 13,5 3,4 34,0 23,1 36,8 1,6 16,9 6,9 49,7 34,9 25,3 1,5 10,8 1,2 13,7 2,5 15,5 0,2 44,5 13,1 2-Ethyl-1-hexanol 4,5 0,9 4,2 0,1 5,3 2,8 6,6 2,5 3,1 0,2 8,7 4,4 2,7 0,5 3,0 0,7 3,3 0,1 5,1 0,4 4,1 2,0 3,3 0,2 5,5 2,0 4,5 3,9 2,8 1,0 4,5 1,6 2,6 1,7 6,8 5,3 0,0 0,0 0,0 0,0 4,3 0,1 2,8 0,8 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Furanmethanol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 56,1 35,2 37,8 8,7 0,0 0,0 0,0 0,0 0,0 0,0 12,3 0,2 10,2 2,5 18,0 2,8 31,7 8,8 2-Methyl propanol 9,8 0,9 5,1 0,1 16,7 2,3 39,6 7,7 22,0 1,4 26,3 4,2 10,5 2,6 76,5 2,3 40,5 3,5 192,8 28,4 20,8 0,1 16,7 1,7 2,5 0,9 26,8 13,3 20,9 3,0 20,6 2,6 22,4 17,8 30,6 2,6 48,9 30,5 61,9 9,4 26,7 4,2 59,0 38,6 76,7 32,2 16,0 0,0 35,2 20,8 27,9 0,7 15,3 1,8 8,3 1,2 21,2 3,6 6,2 2,8 2-Nonanol 0,8 0,0 1,3 0,2 0,5 0,2 0,5 0,1 370,1 62,0 3,3 0,3 278,3 70,6 0,8 0,5 0,9 0,0 0,7 0,2 2,1 0,6 20,5 2,9 5,1 2,0 16,1 1,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 10,0 2,8 0,0 0,0 12,4 3,8 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Phenyl ethanol 263,6 24,8 298,7 18,7 212,8 2,8 425,8 8,5 342,8 47,0 1178,3 645,5 328,2 0,3 869,2 77,4 413,2 21,1 562,3 41,2 222,3 29,8 456,9 72,4 561,1 95,9 290,7 166,2 424,7 72,4 434,5 88,1 854,3 8,4 963,7 13,4 1908,8 48,1 1693,3 322,7 289,5 13,1 787,3 82,6 771,3 13,9 328,4 18,5 503,2 239,8 248,5 3,2 416,9 80,5 486,1 24,0 397,2 14,0 681,3 12,8 3 Ethoxy - 1 Propanol 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3-(Methylthio)-1-propanol 38,0 21,8 141,9 30,5 20,9 3,0 39,8 7,1 0,0 0,0 0,0 0,0 32,0 13,6 51,4 30,4 0,0 0,0 66,8 7,3 38,1 10,6 39,2 14,8 0,0 0,0 35,8 28,3 0,0 0,0 53,9 11,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 11,6 4,5 0,0 0,0 83,5 42,0 0,0 0,0 273,4 94,2 30,8 6,9 31,3 2,4 35,7 11,9 28,0 3,6 187,6 23,8 Esters Butyl acetate 8,9 2,8 21,1 4,9 56,7 9,5 6,8 1,5 1,9 0,6 1085,5 76,9 12,9 14,2 13,3 1,5 37,4 2,9 21,9 19,7 61,6 9,1 7,8 1,2 20,8 8,2 7,2 3,4 421,1 22,4 13,5 1,1 1387,1 16,7 1035,9 53,0 2796,6 70,0 1992,1 160,2 38,4 4,3 454,3 61,2 157,8 28,4 270,0 18,9 224,4 109,7 207,3 17,8 155,4 76,7 164,8 83,1 118,3 4,8 198,5 31,8 Ethyl (4E)-4-decenoate 26,2 1,6 109,2 20,8 3,9 1,2 6,0 4,5 0,0 0,0 63,0 5,7 39,3 3,7 0,0 0,0 0,0 0,0 0,0 0,0 233,1 11,3 38,3 3,0 1170,5 63,7 100,1 65,8 25,9 4,7 497,5 2,5 15,5 1,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 211,1 13,3 105,5 7,6 422,1 31,6 658,8 263,8 459,3 12,1 119,1 48,6 285,5 57,8 377,1 33,2 290,4 5,8 Ethyl 2-methylbutyrate 0,6 0,5 0,2 0,0 0,6 0,5 1,5 0,6 3,3 1,6 12,6 0,3 5,2 2,8 18,1 9,9 9,2 1,0 0,0 0,0 0,1 0,0 0,5 0,1 0,2 0,1 5,9 2,6 3,7 1,4 0,2 0,0 13,7 1,8 21,8 17,5 17,8 9,1 128,8 12,2 2,2 0,3 1,0 0,3 3,4 1,4 1,1 0,9 0,3 0,1 0,2 0,0 0,5 0,2 0,5 0,2 0,3 0,1 0,3 0,0 Ethyl acetate 281,1 24,4 281,1 45,4 106,0 17,4 262,8 47,6 73,7 5,3 7329,9 645,1 571,8 491,1 671,0 42,2 524,7 40,2 1121,9 121,4 245,1 11,1 515,1 84,6 156,1 7,3 595,6 82,3 2631,8 55,9 498,3 11,9 4627,3 32,6 6045,5 4,4 ##### 153,5 ##### 101,3 1390,4 271,1 4023,9 352,8 1940,9 407,2 616,1 239,4 674,7 34,1 504,7 29,9 392,8 61,8 384,8 56,8 461,7 11,4 501,5 17,1 Ethyl butanoate 56,7 8,3 84,5 15,7 21,9 3,4 65,3 10,1 32,4 0,5 458,6 24,9 27,6 23,2 307,4 12,0 155,6 32,0 158,7 54,8 10,9 2,8 90,1 18,8 27,9 8,5 221,5 10,5 155,2 5,4 71,9 14,1 616,5 46,7 531,2 17,5 950,9 230,4 1543,5 158,6 211,2 51,3 147,9 82,8 88,1 9,7 70,0 10,3 60,7 32,9 34,5 3,7 58,6 27,4 80,8 12,5 57,3 9,5 148,8 20,2 Ethyl decanoate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 487,4 32,4 64,5 40,2 96,8 6,5 76,6 28,7 0,0 0,0 858,3 26,1 218,1 15,1 1191,2 79,8 1033,7 595,4 169,7 16,9 1546,7 49,0 106,5 11,4 108,3 54,3 32,4 7,0 0,0 0,0 0,0 0,0 629,4 66,9 414,8 82,0 916,5 24,9 217,0 35,4 339,1 19,1 563,1 27,6 514,4 356,3 235,7 21,9 1222,9 34,5 Ethyl dodecanoate 12,9 4,7 22,8 3,5 27,5 9,2 34,2 6,6 2,4 0,1 190,7 80,6 4,0 1,9 96,2 49,7 34,9 7,6 0,0 0,0 399,3 45,1 32,5 0,9 487,7 52,4 382,3 294,8 71,8 81,4 418,1 71,6 105,8 5,8 64,6 0,3 66,9 1,0 0,0 0,0 0,0 0,0 197,9 44,4 67,0 4,7 195,6 64,4 69,6 51,7 225,0 44,4 299,7 9,0 250,7 83,0 73,5 9,0 545,2 24,8 Ethyl heptanoate 1,2 0,1 5,4 0,5 0,0 0,0 1,6 0,5 1,0 0,5 17,0 5,1 0,0 0,0 3,2 0,3 5,0 0,6 0,0 0,0 1,5 0,1 2,4 0,1 7,4 2,5 8,2 9,8 6,9 2,8 3,3 0,9 4,3 0,8 7,7 5,5 2,9 0,1 5,3 3,4 3,7 0,6 3,8 2,5 0,8 0,2 4,6 0,1 1,8 0,0 2,0 0,3 0,9 0,5 1,5 0,6 2,4 0,1 1,3 0,1 Ethyl hexadecanoate 4,3 1,8 4,7 0,5 0,0 0,0 10,7 1,7 6,7 1,5 131,1 8,4 0,0 0,0 90,4 9,8 26,7 1,6 23,1 3,9 8,0 1,2 3,6 0,1 11,3 2,5 13,1 10,2 19,3 6,0 0,0 0,0 67,2 4,6 74,4 6,3 99,6 6,7 111,3 5,9 27,1 5,1 89,9 5,0 50,5 10,5 44,5 5,2 0,0 0,0 0,0 0,0 0,0 0,0 6,6 1,2 8,2 0,1 21,6 0,9 Ethyl hexanoate 44,0 5,1 102,9 0,9 32,7 4,1 66,2 13,1 16,4 1,9 238,1 73,8 10,3 1,1 175,1 11,5 90,7 6,2 77,9 3,7 61,1 1,1 40,3 2,4 120,1 20,4 128,5 30,7 127,9 0,4 73,0 22,6 350,8 22,7 323,7 17,9 339,3 152,4 605,0 57,7 57,6 17,1 203,3 9,9 202,6 26,4 125,6 18,5 130,2 46,6 80,1 9,8 94,2 42,8 71,6 33,3 46,9 0,0 177,5 22,2 Ethyl isobutyrate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 8,6 2,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 40,7 30,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 28,0 0,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Ethyl isovalerate 0,0 0,0 0,0 0,0 0,6 0,1 4,8 1,9 1,2 0,3 22,8 5,6 1,5 0,7 36,1 3,4 17,1 7,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 9,3 4,2 7,2 1,4 0,0 0,0 10,1 3,4 9,7 1,5 7,9 1,4 81,1 14,6 2,8 0,7 0,0 0,0 4,4 1,4 0,8 0,2 0,5 0,2 0,2 0,0 1,3 0,6 1,0 0,2 0,7 0,2 0,9 0,0 Ethyl octanoate 98,0 4,1 229,9 11,4 44,7 4,2 66,0 26,4 60,0 7,0 1197,4 63,3 129,9 0,5 347,3 14,9 246,4 17,6 84,7 22,2 528,5 151,8 411,8 23,4 682,7 92,7 881,0 190,6 417,0 127,8 433,4 87,7 239,1 73,8 447,4 40,7 420,0 56,7 192,8 14,5 184,1 37,8 480,5 22,3 447,8 137,3 996,5 62,3 725,5 279,5 498,2 24,3 419,2 76,4 416,6 53,5 754,1 33,3 1146,6 42,7 Ethyl propanoate 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 573,8 18,7 301,0 90,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Ethyl tetradecanoate 8,6 0,6 15,6 3,2 0,0 0,0 10,1 2,6 2,8 0,1 79,4 9,1 2,0 0,0 77,4 3,1 22,0 5,8 0,0 0,0 12,3 5,4 0,0 0,0 60,4 1,1 50,1 54,8 23,7 12,1 0,0 0,0 86,4 34,6 41,2 6,1 63,0 5,6 88,7 13,9 14,2 4,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Isoamyl acetate 43,4 15,2 102,6 35,7 20,1 2,9 45,0 5,9 27,6 3,2 1955,4 136,5 155,6 148,2 88,2 4,9 163,9 22,7 101,2 7,4 224,8 14,4 20,1 1,7 52,9 19,8 44,5 9,6 602,5 26,8 20,5 3,6 1888,9 163,1 1779,0 25,6 5710,3 56,5 3562,4 59,1 79,2 10,2 773,6 39,8 488,2 179,6 777,8 5,7 914,0 462,1 556,1 9,2 440,9 66,1 561,5 194,5 610,1 112,9 676,4 116,3 Isoamyl butyrate 2,9 0,0 8,1 1,0 2,9 0,1 5,8 1,2 9,7 0,8 95,2 4,4 10,1 7,3 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 83,0 32,0 102,1 25,9 192,5 1,1 0,0 0,0 13,7 6,3 0,0 0,0 0,0 0,0 5,0 2,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Isobutyl acetate 1,2 0,3 0,8 0,2 1,9 2,1 4,8 1,1 1,6 0,4 314,9 38,1 11,5 14,5 11,9 2,2 24,1 14,9 21,1 2,3 6,7 4,8 2,1 0,6 0,5 0,2 4,5 0,5 126,6 37,4 1,3 0,3 338,8 53,2 518,8 34,3 1950,3 31,1 759,8 84,8 13,1 1,1 86,1 49,9 53,3 3,0 37,8 16,3 21,0 13,8 28,4 2,0 11,1 5,4 7,8 4,9 12,7 3,9 9,5 1,7 Isobutyl butanoate 0,4 0,0 0,3 0,0 0,4 0,2 0,9 0,2 0,9 0,1 11,5 4,7 0,7 0,5 6,8 6,6 5,4 1,2 1,9 0,4 0,2 0,1 0,6 0,1 0,2 0,1 1,3 0,3 5,3 3,8 0,7 0,2 20,4 14,4 34,7 26,4 81,6 13,3 65,3 3,6 4,0 0,7 4,2 2,2 1,3 0,7 0,5 0,2 0,4 0,2 0,4 0,0 0,3 0,2 0,1 0,1 0,4 0,0 0,3 0,0 Phenethyl acetate 26,5 4,0 322,5 11,9 15,7 1,8 25,7 5,5 19,0 1,8 1144,1 711,6 146,9 121,1 68,6 60,4 53,6 6,0 0,0 0,0 341,2 9,2 37,9 5,6 382,4 39,9 0,0 0,0 384,5 102,4 34,5 17,5 1227,5 97,4 1134,8 88,4 2493,0 195,0 1727,3 113,9 31,2 4,8 188,2 17,0 262,3 8,9 484,0 16,3 529,0 187,7 400,1 37,4 422,9 17,0 350,4 46,1 424,0 21,1 485,9 88,3 S-methyl thioacetate 1,2 0,3 1,1 0,3 4,7 0,7 5,4 1,1 0,8 0,0 0,0 0,0 0,0 0,0 15,7 4,5 0,0 0,0 33,0 1,6 0,0 0,0 1,2 0,3 0,0 0,0 0,0 0,0 0,0 0,0 0,8 0,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1,7 0,9 0,0 0,0 4,8 1,9 0,0 0,0 2,1 1,5 4,1 2,3 2,1 0,9 7,8 2,1 0,0 0,0 2-Methyl propanoate 12,8 2,8 12,6 2,7 24,6 0,8 33,9 5,4 39,2 12,9 38,2 4,7 30,3 1,9 32,7 9,6 48,8 2,2 85,1 8,3 21,1 2,4 66,3 21,3 14,4 0,2 48,7 3,6 14,6 3,3 29,7 1,2 15,2 2,1 37,2 11,0 29,3 1,2 0,0 0,0 74,0 10,9 0,0 0,0 0,0 0,0 23,8 5,2 49,2 32,0 16,6 3,2 21,5 2,2 22,9 2,5 33,0 2,8 0,0 0,0 Acids Acetic acid 257,4 4,0 308,1 43,2 299,9 11,2 466,7 61,9 194,6 5,9 1561,5 99,0 226,6 75,9 495,9 44,3 928,3 126,8 2445,9 25,0 461,2 67,8 565,6 57,9 867,3 41,6 794,6 34,4 1069,5 190,7 562,1 13,6 2264,4 87,9 2465,7 224,2 4941,9 57,7 3801,5 552,2 1283,7 358,8 2812,8 62,2 1425,9 264,7 537,3 79,0 1019,8 482,2 469,6 8,5 590,9 21,4 700,4 126,2 719,7 41,7 1809,9 95,5 Butyric acid 121,0 24,2 136,7 23,5 125,4 35,4 170,0 26,4 120,4 6,8 0,0 0,0 180,1 27,0 349,6 25,2 263,6 1,6 483,0 49,4 0,0 0,0 303,8 58,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 62,6 7,0 183,2 30,1 135,1 10,0 460,2 54,0 349,9 51,7 642,3 50,2 233,6 16,5 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 118,4 3,3 543,9 45,9 Decanoic acid 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 350,1 83,0 0,0 0,0 127,5 6,4 366,6 116,2 0,0 0,0 406,3 75,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 82,2 42,8 105,7 9,0 152,0 76,0 36,7 16,4 62,6 6,2 180,2 13,0 132,9 17,0 28,2 5,6 754,8 22,4 Hexanoic acid 34,6 9,6 33,6 14,6 77,4 22,9 73,2 14,8 0,0 0,0 56,7 17,9 26,8 1,4 147,1 14,3 78,9 10,7 76,0 26,3 0,0 0,0 64,2 27,1 0,0 0,0 59,2 19,6 0,0 0,0 0,0 0,0 0,0 0,0 63,2 7,9 0,0 0,0 0,0 0,0 63,6 2,4 214,6 29,5 153,1 13,3 0,0 0,0 271,4 95,6 164,5 65,9 199,4 6,7 130,8 1,2 210,6 26,2 0,0 0,0 Isovaleric acid 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 139,5 1,2 Octanoic acid 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 132,8 1,1 48,3 3,3 51,2 4,2 46,9 3,5 0,0 0,0 375,2 88,7 74,9 30,5 218,9 35,1 404,7 146,3 60,0 30,0 64,8 6,1 0,0 0,0 34,6 48,8 103,7 6,3 0,0 0,0 29,2 2,5 145,3 8,9 217,0 9,9 377,7 20,5 697,5 298,5 418,0 81,2 451,8 18,3 330,9 26,5 420,9 35,5 1026,0 184,5 Aldehydes Acetaldehyde 8,1 0,3 15,9 0,9 14,8 14,2 22,7 3,7 10,8 0,9 38,9 6,3 12,8 9,6 56,2 4,7 36,4 8,7 103,8 11,7 14,7 4,9 23,4 4,1 18,7 4,8 13,0 7,1 27,4 18,6 11,1 0,0 66,4 21,5 28,4 1,5 135,4 14,5 138,4 18,6 18,6 0,6 20,9 5,8 62,2 20,2 19,0 1,8 11,3 4,2 7,5 0,8 20,7 3,1 11,3 0,2 8,6 0,4 6,9 1,4 Benzaldehyde 28,9 1,0 24,5 2,2 7,7 4,7 5,0 1,2 0,0 0,0 35,3 5,1 0,0 0,0 6,3 4,3 10,4 0,7 16,6 15,4 12,4 0,6 22,9 0,1 31,8 3,3 16,4 7,3 12,0 1,5 26,4 3,4 21,5 7,1 31,4 4,4 29,8 1,1 38,1 24,4 23,5 1,6 40,7 0,5 32,4 4,7 13,5 1,3 15,1 6,8 9,6 0,6 12,1 2,3 11,6 0,7 14,8 0,1 28,3 3,6 Furfural 3,2 0,4 3,1 0,1 2,8 0,6 3,9 0,3 2,1 0,5 7,5 3,8 1,8 0,3 3,6 0,3 36,1 12,4 14,9 19,0 5,6 2,1 6,7 0,8 7,9 3,7 2,9 3,3 6,0 0,3 7,1 0,1 5,6 0,5 11,9 2,5 6,1 3,2 18,3 5,2 15,9 1,3 63,6 8,9 90,1 4,2 28,5 10,2 14,9 8,1 7,4 1,3 32,7 2,2 27,1 5,9 48,0 4,5 84,1 26,8 Phenyl acetaldehyde 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 82,6 16,2 33,6 12,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 19,4 0,5 109,1 6,9 69,9 17,1 0,0 0,0 0,0 0,0 0,0 0,0 29,6 0,5 20,2 7,0 34,0 3,0 55,2 16,1 1 -Decanal 0,0 0,0 4,0 0,1 1,9 1,1 0,0 0,0 0,0 0,0 12,6 3,9 0,9 1,2 0,0 0,0 0,0 0,0 0,0 0,0 5,2 1,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 4,6 1,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 1-Nonanal 1,0 0,1 1,4 0,1 1,6 0,8 1,4 0,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3-Methyl butanal 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 480,9 211,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 82,8 15,9 666,8 35,1 425,8 95,5 0,0 0,0 0,0 0,0 0,0 0,0 154,6 17,9 95,3 53,0 171,0 1,5 238,3 24,0 4-Methyl benzaldehyde 5,1 0,1 4,1 1,3 0,0 0,0 0,0 0,0 3,7 0,2 0,0 0,0 3,9 0,0 0,0 0,0 0,0 0,0 0,0 0,0 3,9 1,2 0,0 0,0 0,0 0,0 0,0 0,0 6,5 1,6 0,0 0,0 0,0 0,0 19,2 1,0 5,6 7,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 4,8 0,6 0,0 0,0 0,0 0,0 0,0 0,0 5 Methyl furfural 2,2 0,6 1,4 0,4 1,3 0,6 1,7 0,3 1,2 0,2 4,1 2,5 0,8 0,0 2,2 1,9 74,0 9,2 94,1 5,6 11,4 5,9 11,9 2,8 6,4 1,5 8,2 2,8 5,0 1,0 6,7 0,8 3,4 0,5 14,6 0,4 3,7 1,3 6,6 0,2 17,4 2,8 142,5 6,2 158,2 5,5 50,8 2,0 28,7 16,9 10,8 1,0 39,3 2,3 39,3 13,7 68,5 6,1 123,2 33,6 5-Hydroxymethylfurfural 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 13,4 0,8 200,0 12,2 186,0 5,0 69,1 4,3 7,7 4,7 0,0 0,0 76,7 1,5 38,5 5,5 127,9 5,0 213,7 8,8 Ketones Acetoin 0,7 0,1 4,1 0,2 6,2 0,3 2,3 0,5 0,0 0,0 5,5 1,9 1,0 0,9 1,1 0,3 0,0 0,0 0,0 0,0 3,4 1,2 19,3 8,9 2,2 0,6 0,0 0,0 4,4 5,1 0,0 0,0 26,7 7,8 6,6 7,7 0,0 0,0 0,0 0,0 0,0 0,0 23,3 4,2 33,1 4,9 2,8 1,8 0,0 0,0 2,0 1,1 3,5 0,3 0,0 0,0 0,0 0,0 0,0 0,0 Diacetyl 10,3 7,9 24,0 1,7 5,6 1,8 3,3 0,4 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Pyranone 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 432,3 30,9 0,0 0,0 6,4 8,2 7,7 1,8 2,6 2,3 2,1 0,7 0,0 0,0 0,0 0,0 0,0 0,0 36,0 5,3 0,0 0,0 24,6 31,7 144,5 66,2 1105,9 65,1 689,7 61,7 150,8 13,4 4,4 2,6 2,4 1,9 282,5 24,1 213,8 7,6 306,4 11,1 465,5 13,5 2-Cyclopentene-1,4-dione 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 61,8 5,9 81,5 20,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 14,0 0,5 4,3 2,7 0,0 0,0 17,1 0,8 134,1 5,9 102,4 22,8 39,5 6,3 15,8 7,1 5,6 0,5 31,4 1,0 27,2 8,9 44,1 3,7 76,2 12,4 2-Dodecanone 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Methyltetrahydrothiophen-3-one0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2-Nonanone 0,8 0,1 1,4 0,0 1,4 0,4 1,5 0,2 314,5 17,1 3,9 1,0 320,8 100,7 3,2 1,4 4,9 5,8 0,0 0,0 2,6 1,4 34,1 4,5 0,0 0,0 14,7 16,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 7,2 5,1 0,0 0,0 19,1 7,2 0,0 0,0 3,4 0,1 0,0 0,0 3,1 0,7 0,0 0,0 25,3 2,9 6,1 1,2 2-Undecanone 0,3 0,1 0,3 0,0 0,5 0,1 0,6 0,2 64,3 3,5 2,1 1,0 43,0 35,6 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Pyrazine 2,5-Dimethyl-3-ethylpyrazine 3,7 0,1 4,3 0,1 4,5 0,7 4,0 1,2 2,9 0,1 9,7 2,5 3,4 0,6 3,0 0,4 4,5 1,9 2,8 0,3 3,2 0,6 6,2 0,5 5,3 2,9 5,2 1,2 3,3 0,5 5,9 0,0 2,9 2,4 6,4 0,8 0,9 0,1 4,3 1,0 3,7 0,1 2,8 0,6 7,8 2,8 3,4 0,8 3,4 1,4 2,5 0,1 4,6 0,9 2,5 0,2 3,7 0,1 4,1 0,0 2,6-Dimethylpyrazine 13,9 0,2 11,3 0,5 16,3 4,6 15,0 3,6 12,9 0,6 12,7 0,4 12,1 1,9 12,4 2,0 27,3 0,8 17,5 2,3 9,1 1,1 13,5 2,6 13,6 1,7 12,2 1,3 9,3 1,2 12,7 1,0 7,4 5,4 10,9 1,6 3,6 0,0 11,9 2,6 12,1 5,9 12,7 9,0 2,8 0,5 7,1 0,2 10,3 1,6 11,1 1,3 11,9 0,5 11,5 0,5 15,1 0,9 9,5 2,1

Table 2B: List of 62 volatiles that have been detected by GC-MS measurement

Alcohols Acids

1 Benzyl alcohol 37 Acetic acid 2 Butanol 38 Butyric acid 3 Dodecanol 39 Decanoic acid 4 Fenchyl alcohol 40 Hexanoic acid 5 Furaneol 41 Isovaleric acid 6 Isoamyl alcohol 42 Octanoic acid 7 Propanol

8 2-Ethyl-1-hexanol Aldehydes

9 2-Furanmethanol

10 2-Methyl propanol 43 Acetaldehyde 11 2-Nonanol 44 Benzaldehyde 12 2-Phenyl ethanol 45 Furfural 13 3 Ethoxy - 1 Propanol 46 Phenyl acetaldehyde 14 3-(Methylthio)-1-propanol 47 1 -Decanal 48 1-Nonanal

Esters 49 3-Methyl butanal

50 4-Methyl benzaldehyde

15 Butyl acetate 51 5 Methyl furfural 16 Ethyl (4E)-4-decenoate 52 5-Hydroxymethylfurfural 17 Ethyl 2-methylbutyrate

18 Ethyl acetate Chetons

19 Ethyl butanoate

20 Ethyl decanoate 53 Acetoin 21 Ethyl dodecanoate 54 Diacetyl 22 Ethyl heptanoate 55 Pyranone 23 Ethyl hexadecanoate 56 2-Cyclopentene-1,4-dione 24 Ethyl hexanoate 57 2-Dodecanone 2-Methyltetrahydrothiophen-3- 25 58 Ethyl isobutyrate one 26 Ethyl isovalerate 59 2-Nonanone 27 Ethyl octanoate 60 2-Undecanone 28 Ethyl propanoate

29 Ethyl tetradecanoate Pyrazines

30 Isoamyl acetate

31 Isoamyl butyrate 61 2,5-Dimethyl-3-ethylpyrazine 32 Isobutyl acetate 62 2,6-Dimethylpyrazine 33 Isobutyl butanoate

34 Phenethyl acetate 35 S-methyl thioacetate 36 2-Methyl propanoate

Table 3: Characterization of fermentation performance. .

The table values are represented as a heat map with grey values. The legend shows the range for each grey value for the respective condition/test. Figure 1A: Principal component analysis of the set of strains and their chemical compounds. The matrix is based on the full set of 60 strains and the average of the 62 VOCs detected. Strains are presented with their assigned coordinates (see Table 1). Species belonging to the same genus are represented with the same color. A higher resolution of the central plot is shown as inset at the upper right corner.

Figure 1B: Loading plot of the principal component analysis of all components. Components were grouped according to the chemical class, numbered and colored similarly. The complete list of volatiles is shown in table 2B.

Figure 2A:

Fermentation profile of the sub-set of 18 strains. The CO2 release was measured daily for each individual strain. Fermentation was followed for 11 days.

Figure 2B: Final Sugar content and ethanol production of the subset of 18 strains. Sugar concentration was measured in °P after 11 days of fermentation. The ethanol concentration is shown as % ethanol (vol/vol).

Figure 3: Volatile organic compounds produced during fermentation by the subset of 18 strains. The compounds were grouped, numbered and colored according to their chemical class. The complete list of compounds is shown in Table 2B. Figure 4: Concentration of selected volatiles in six strains that produced exceptionally high concentrations of organic volatiles

Candida diversa (B4); Debaryomyces subglobosus (B11); Pichia kluyveri (E7); Saccharomyces cerevisiae (E11); Wickerhamomyces anomalus (F10); Zygosaccharomyces mellis (G4); Saccharomyces pastorianus (G10). Appl Microbiol Biotechnol, 2013. 97(15): References p. 6919-30. 11. Pretoriu, M.d.T.a.I.S., Microbial Spoilage and Preservation of Wine: Using 1. Dunn, B. and G. Sherlock, Reconstruction Weapons from Nature's Own Arsenal- A of the genome origins and evolution of Review S. Afr. J. Enol. Vitic., 2000. the hybrid lager yeast Saccharomyces 21(Special Issue): p. 74-96. pastorianus. Genome Res, 2008. 18(10): 12. Viana, F., et al., Monitoring a mixed p. 1610-23. starter of Hanseniaspora vineae- 2. Spitaels, F., et al., The microbial diversity Saccharomyces cerevisiae in natural of traditional spontaneously fermented must: impact on 2-phenylethyl acetate lambic beer. PLoS One, 2014. 9(4): p. production. Int J Food Microbiol, 2011. e95384. 151(2): p. 235-40. 3. Pires, E.J., et al., Yeast: the soul of beer's 13. Hope, C.F.A., Cinnamic acid as the basis aroma--a review of flavour-active esters of a medium for the detection of wild and higher alcohols produced by the yeasts. Journal of the Institute of brewing yeast. Appl Microbiol Biotechnol, Brewing, 1987. 93(3): p. 213-215. 2014. 98(5): p. 1937-49. 14. Ian McMurrough, D.M.D.D., June Hurley, 4. D. Šmogrovičová, Z.D., Beer volatile by- Ann-Marie Doyle, Gerard Hennigan and product formation at different Niall McNllty, Control of ferulic acid and fermentation temperature using 4-vinyl guaiacol in brewing. J. Insl. Brew, immobilised yeasts. Process 1996. 102: p. 327-332. Biochemistry, 1999. 34(8): p. 785-794. 15. E. Acree, E.P.S.a.D.F.S., Effect of Yeast 5. Bart Vanderhaegen, Filip Delvaux, Luk Strain and Type of Sulfur Compound on Daenen, Hubert Verachtert, Freddy R. Hydrogen Sulfide Production. Am. J. Enol. Delvaux, Aging characteristics of different Vitic, 1972. 23(1): p. 6-9. beer types. Food Chemistry, 2007. 103(2): 16. Apostolos Spiropoulos, J.T., Ioannis p. 404-412. Flerianos, and Linda F. Bisson, 6. Vanderhaegen, B., et al., Evolution of Characterization of hydrogen sulfide chemical and sensory properties during formation in commercial and natural aging of top-fermented beer. J Agric Food wine isolates of Saccharomyces. Am. J. Chem, 2003. 51(23): p. 6782-90. Enol. Vitic., 2000. 51: p. 223-248. 7. Chiam Yu Ng, M.-y.J., Jinwon Lee and 17. Dmytruk, K.V. and A.A. Sibirny, Candida Min-Kyu Oh, Production of 2,3-butanediol famata (Candida flareri). Yeast, 2012. in Saccharomyces cerevisiae by in silico 29(11): p. 453-8. aided metabolic engineering. Microbial 18. Harris, F.B.A.a.G., The Production of Cell Factories, 2012. 11(68). Riboflavin and D-Arabitol by 8. Mauriello, G., et al., SPME-GC method as Debaryomyces subglobosu. J. gen. a tool to differentiate VOC profiles in Microbiol, 1963. 33: p. 137-146. Saccharomyces cerevisiae wine yeasts. 19. Seong Han Lim, J.S.C., and Enoch Y. Park, Food Microbiol, 2009. 26(3): p. 246-52. Microbial Production of Riboflavin Using 9. Laluce, J.F.T.S.A.L.R.d.S.C., Non- Riboflavin Overproducers Ashbya conventional yeasts. Appl Microbiol gossypii, Bacillus subtilis, and Candida Biotechnol, 2002. 58: p. 147-156. famate: An Overview. Biotechnol. 10. Krogerus, K. and B.R. Gibson, Influence of Bioprocess Eng., 2001. 6: p. 75-88. valine and other amino acids on total 20. Brachmann, A.O., et al., Pyrones as diacetyl and 2,3-pentanedione levels bacterial signaling molecules. Nat Chem during fermentation of brewer's wort. Biol, 2013. 9(9): p. 573-8. 21. Paul L.H. Mcsweeney, M.J.S., Biochemical pathways for the production of flavour compounds in cheeses during ripening: A review. Lait 80, 2000: p. 293–324. 22. T van den Tempel, M.J., The technological characteristics of Debaryomyces hansenii and Yarrowia lipolytica and their potential as starter cultures for production of Danablu. International Dairy Journal, 2000. 2000(4): p. 263–270. 23. Smit, B.A., W.J. Engels, and G. Smit, Branched chain aldehydes: production and breakdown pathways and relevance for flavour in foods. Appl Microbiol Biotechnol, 2009. 81(6): p. 987-99. 24. M.-C. Montel, J.R., R. Talon, J.-L. Berdagué, S. Rousset-Akrim, Biochemical activities of Micrococcaceae and their effects on the aromatic profiles and odours of a dry sausage model. Food Microbiology, 1996. 13(6): p. 489–499. 25. Daniel, H.M., et al., Wickerhamomyces anomalus in the sourdough microbial ecosystem. Antonie Van Leeuwenhoek, 2011. 99(1): p. 63-73. 26. Hartman, H.R.B.A.P.A., Yeasts in Ensiled High-Moisture Corn. Applied Microbiology, 1966. 14(1): p. 35-38. 27. Crafack, M., et al., Influencing cocoa flavour using Pichia kluyveri and Kluyveromyces marxianus in a defined mixed starter culture for cocoa fermentation. Int J Food Microbiol, 2013. 167(1): p. 103-16.

Appendix

PCA based disruption cassette were used to generate the A.gossypii deletion mutants as described by Wendland, J., et all., 2000 [65] The deletion cassettes were amplified from pFA- GEN3 (C634) and pFA-SAT1(550) using S1- and S2- primers which contained 50 bp homologus flanks to the target gene. The amplified DNA product was precipitated, purified and transformed into A.gossypii by electroporation. Diagnostic PCR was used to verify the correct integration of the cassette into the target gene locus. G1(upstream) and G4(downstream) annealing primers of the marker integration site, and GEN3/SAT1 primers (G2 and G3) were used for the verification. I1/I2 primers were used to confirm the homokaryotic deletion(HOM) of the target gene in the A.gossypii genome. In Figure 1 there is a schematic view of the process.

Figure 1. Schematic representation of the PCA-based target protocol used in A.gossypii.

A first round of transformation was necessary to generate the heterokaryons mutants. Then the heterokarions were let sporulate in selective condition and the haploid spores macromanipulated s into selective media to generate homokaryons mutants. Two independent mutants were generated for each gene deletion.