Field of the invention

The present invention relates to the field of fermentation, more particularly to ethanol production. Even more particularly the present invention relates to altered aroma production during fermentation processes. The present invention provides mutant alleles and chimeric genes useful to develop yeast strains to regulate the production of rose flavor during fermentation. In addition, the invention also relates to the use of such yeast strains for the production of fermented foods and liquids with increased rose flavor.

Background

Flavor is one of the main defining characteristics of alcoholic beverages with critical importance for their commercial value (1, 2). Yeast plays an important role in generating the final aroma profile of alcoholic beverages. Pleasant flavors to the human palate include the fruity aromas that are largely derived from secondary metabolism of sugar during yeast alcoholic fermentation. Esters comprise the largest and most important group. They often have pleasant aromas, low perception thresholds and relatively high concentrations in alcoholic beverages (1, 3, 4). Small changes in ester concentrations can have significant effects on the taste of alcoholic beverages (3). Esters are formed from an alcohol and a carboxylic acid linked with coenzyme A. There are two groups of flavor-active esters in fermented beverages, ethyl esters and acetate esters. Ethyl esters are formed from ethanol and the acyl-coA derivative of medium- chain fatty acids (MCFA). Among this group, the most important are ethyl hexanoate (anise seed, applelike aroma) and ethyl octanoate (apple, pineapple aroma). Acetate esters are formed from acetyl-CoA and an alcohol that can be ethanol (yielding ethyl acetate) or a higher alcohol derived from amino acid metabolism. The most flavor-relevant acetate esters are ethyl acetate (solvent-like aroma), isoamyl acetate (banana-like aroma) and 2-phenylethylacetate (honey, rose-like aroma) (1, 4-7).

The main genes responsible for ester biosynthesis identified up till now are ATFl and ATF2 for acetate esters, and EHT1 and EEB1 for ethyl esters (8-12). Double deletion of ATFl and ATF2 largely abolishes the production of isoamyl acetate and strongly reduces the production of many other flavor esters, including 2-PEAc. The observation that significant production levels are retained in the atflA atf2A strain indicates involvement of other unknown biosynthetic enzymes (8). Double deletion of EEB1 and EHT1 caused a considerable, but also only partial reduction in the levels of all ethyl esters, again indicating the presence of additional biosynthetic enzymes (10). In addition, little is known about the regulation of the biosynthetic pathways and about other factors that may influence the production of specific flavor compounds. Given the large variety of flavor compounds and the many parameters affecting their formation, most of the underlying genetic basis of the natural variation in flavor compound production remains unknown.

In this study, we have investigated the polygenic basis of phenylethyl acetate production (2-PEAc), an important flavor compound in alcoholic beverages, imparting a honey, rose-like flavor. 2-PEAc is an acetate ester made by esterification of acetyl-CoA and 2-phenylethanol (2-PE). 2-PE is synthesized by degradation of the aromatic amino acid L-phenylalanine. We have used two random segregants (BTC.1D and ER18) derived from two non-selected diploid strains as parents and identified offspring with high 2- PEAc production. Several major QTLs were then mapped by pooled-segregant whole-genome sequence analysis and two novel causative genes, TORI and FAS2, that have never previously been connected to production of 2-PEAc, were identified.

TORI encodes a phosphatidylinositol kinase homologue, that regulates multiple cellular processes, including the induction of autophagy (46), the diauxic shift and directly modulates glucose activation and discrimination pathways (47). Tori is well known to be involved in nitrogen-regulated processes, such as nitrogen catabolite repression. (49, 50). Surprisingly and disclosed herein, Tori is also involved in the production of 2-phenylethyl acetate. This is demonstrated by a reduction in 2-PeAc production after exchanging wild-type Tori for a defective mutant Tori allele.

FAS2 encodes the alpha subunit of fatty acid synthetase, a very large, multifunctional protein of 1887 aa, which contributes acyl-carrier, 3-ketoreductase, 3-ketosynthase and phosphopantetheinyl transferase activities to the fatty acid synthetase complex (53) (Fig. 8). The latter is composed of six Fasl and six Fas2 subunits. It is known that the microbial type I fatty acid synthases (FASs) are involved in multiple functions linked to fatty acid synthesis: de novo synthesis of long-chain fatty acids, mitochondrial fatty acid synthesis, acylation of certain secondary metabolites and coenzymes, fatty acid elongation, and synthesis of the vast diversity of mycobacterial lipids (54). In contrast to the Tori mutant, the Fas2 mutant allele herein disclosed is a dominant mutation. Two unique SNPs are disclosed herein as being responsible for high 2-PeAC production.

Summary

It is disclosed herein that two genes, the wild type TORI allele and a superior FAS2BTC.1D allele, affect the level of the important flavor compound 2-PEAc, with highly desirable rose flavor. The FAS2BTC.1D allele allows to strongly enhance the production of 2-PEAc, even in a strain that already produces a higher level than generally found in fermentations by Saccharomyces cerevisiae strains. This is accomplished with additional positive effects on other flavor compounds and apparently without significant effects on compounds with a negative contribution to the flavor profile. Hence, the FAS2BTC.1D allele appears to be highly promising for the creation of cisgenic brewing strains with an attractive and novel flavor profile.

Therefore and in a first aspect of the invention, a mutant Fas2 yeast protein comprising an A1136T mutation and/or a V1624I mutation is provided. The application also provides a nucleic acid molecule encoding said mutant Fas2 protein, as well as a chimeric gene comprising a promoter which is active in a eukaryotic cell and said nucleic acid molecule encoding the mutant Fas2 yeast protein from the application and a 3' end region involved in transcription termination and/or polyadenylation. Also a vector is provided comprising the above described nucleic acid molecule or chimeric gene.

In a second aspect, a microorganism is provided comprising a mutant Fas2 yeast protein comprising an A1136T mutation and/or a V1624I mutation or comprising the nucleic acid molecule, the chimeric gene or the vector above described. In a particular embodiment, said microorganism is an engineered microorganism, even more particularly said microorganism is a yeast. In an additional embodiment, the application provides a fermented solution comprising said yeast.

In a third aspect, the use of the mutant Fas2 protein is provided to increase the flavor of a fermented solution. In particular embodiments, said flavor is rose flavor. Also, the use of the mutant Fas2 protein is provided to increase the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol in a microorganism.

In a fourth aspect, methods are provided to increase the flavor of a fermented solution, said method comprising expressing the Fas2 mutant protein of the application in a yeast, wherein said yeast is used during the production of said fermented solution. Preferably, said flavor is determined by 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol.

The application also provides a C-terminally truncated Tori yeast protein. In particular embodiments, said protein lacks residues 216-2470 of a wild type full length Tori protein as defined by SEQ ID No 5. In another aspect, a microorganism comprising the above described truncated Tori proteins is provided. In particular embodiments, said microorganism is an engineered microorganism, in other particular embodiments, said microorganism is a yeast. In an additional embodiment, the application also provides a fermented solution comprising said yeast.

In another embodiment, the use of a non-functional Tori yeast protein is provided to modulate the flavor profile of a fermented solution. More particularly, said flavor profile modulation is a decrease in the level of rose flavor. Even more particularly, said decrease in rose flavor level is a decreased level of 2-phenylethyl acetate or phenylethanol.

In yet another embodiment, a method is provided of producing a fermented solution with a reduced level of 2-phenylethyl acetate, said method comprising expressing a non-functional Tori yeast protein in a yeast, wherein said yeast is used during the production of said fermented solution. More particularly, said non-functional Tori protein lacks residues 216-2470 of a wild type full length Tori protein as defined by SEQ ID No 5. Also a method is provided to increase the level of rose flavor or more particularly 2- phenylacetate in a yeast fermentation, said method comprising determining the presence of one or more loss-of-function mutations in the TORI allele of said yeast and restoring said one or more mutations in said yeast to wild-type or replacing the non-functional TORI allele by a wild-type TORI allele, if one of more loss-of-function mutations is present in said yeast's TORI gene. In particular embodiments, said method is disclosed comprising the additional step of expressing the mutant Fas2 protein lacking residues 216-2470 of a wild type full length Tori protein as defined by SEQ ID No 5 in said yeast. Brief description of the Figures

Figure 1. Violin density plots of flavor compound production by the segregants from BTC.1D/ER18.

Fermentations were carried out in 100 mL YP250GlulO% and flavors were measured by GC-FID at the end of the fermentation. For each compound, the width of the violin plot on the X-axis represents the kernel density showing the distribution of values within a dataset, and is thus representative of the number of segregants producing a specific level of flavor compound. The flavor production by the parent strains is represented by · (BTC.1D) and + (ER18). The lowest value indicated represents either the lowest value measured or the detection limit.

Figure 2. Genetic mapping of OJLs involved in high phenylethyl acetate production by pooled- segregant whole-genome sequence analysis.

Pooled Fl selected or random segregants (24 segregants in each pool) were subjected to sequence analysis with the lllumina platform at BGI. The SNP variant frequency was used for QTL mapping. Black lines represent the random 'unselected pool' and red lines represent the selected 'superior pool' (row 1). The log-odds ratio with confidence interval is shown in row 2. The most reliable QTLs are indicated by solid rectangles. When the SNP variant frequency > 0.5, linkage is with the BTC.1D parent. When the SNP variant frequency < 0.5, linkage is with the ER18 parent, p-values calculated from the sequencing data were plotted against the respective chromosomal position (row 3). p-values <0.05 (indicated by dotted line) were considered statistically significant. Blue line represents p-values calculated with 0.5 as reference for the whole genome. Green line represents p-values calculated with the actual result of the random pool as reference for the whole genome.

Figure 3. Identification of TORI as causative gene in OJLl located on chromosome X.

A. SNP variant frequency for seven selected SNPs as determined with the 24 superior segregants individually; B. p-values for the same seven SNPs plotted against their chromosomal position, p-values below 0.05 were considered statistically significant. C. overview of all genes present in the region with strongest linkage of the QTL1 on chromosome X. Genes marked with a star contain one or more non- synonymous mutations in the ORF; D. 2-PEAc production in strains derived from superior segregant 442: wild type, torlA and torlE216*. Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated difference in 2-PEAc production was significant by the unpaired Student t-test; E. TORI allele swapping in the parent strains BTC.1D and ER18. Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated difference in 2-PEAc production was significant by the unpaired Student t- test.

Figure 4. Identification of FAS2 as causative gene in QTL2 located on chromosome XVI. A. SNP variant frequency for six selected SNPs as determined with the 24 superior segregants individually; B. p-values for the same six SNPs. p-values below 0.05 were considered statistically significant; C. Q.TL2 was divided into 10 gene blocks, which were each tested for causative character using CRISPR/Cas9 direct replacement with the 2xgRNA approach; D. Block 7 contained the FAS2 gene, which turned out to be causative for Q.TL2; E. 2-PEAc production by the wild type strain 442 and derivatives with a bulk replacement of each gene block. Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated difference in 2-PEAc production was significant by the unpaired Student t-test; F. FAS2 allele swapping in the parent strains BTC.1D and ER18. Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated difference in 2-PEAc production was significant by the unpaired Student t- test.

Figure 5. Allele swapping of both causative alleles from BTC-1D into the ER18 parent strain and test for dominance/recessivity of the FAS2BTC.1D allele.

A. The torlE216* SNP in parent strain ER18 was changed into the wild type TOR1E216 amino acid from parent strain BTC.1D, the FAS2ER18 allele in parent strain ER18 was replaced by the FAS2BTC.1D allele from parent strain BTC.1D, while in the third ER18-derived strain both changes were introduced. 2-PEAc production by the parent strain ER18, the single and double replacement strains, and strain 442, one of the top 2-PEAc producing segregants, is shown. Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated difference in 2-PEAc production was significant by the one-way ANOVA with Dunnet's multiple comparison test.

B. The strains ER18 T0R1*216E and ER18 T0R1*216E FAS2BTC.1D were crossed with ER18 T0R1*216E (white bars) and 16D (black bars). Fermentations were carried out in 100 mL YP250GlulO%. 2-PEAc was measured at the end of the fermentation. The indicated differences in 2-PEAc production was significant by the unpaired t-test (p<0,05). Figure 6. Evaluation of the complete flavor profile of the ER18 engineered strains.

The torlE216* SNP in parent strain ER18 was changed into the wild type TOR1E216 amino acid from parent strain BTC.ID, the FAS2ER18 allele in parent strain ER18 was replaced by the FAS2BTC.1D allele from parent strain BTC.ID, while in the third ER18-derived strain both changes were introduced. The production of different flavor compounds by the parent strain ER18 (1), the single (2, ER18 TOR1E216; 3, ER18 FAS2BTC.1D) and double (4) replacement strains, and strain 442 (5), one of the top 2-PEAc producing segregants, is shown. Fermentations were carried out in 100 mL YP250GlulO%, at 140rpm, 25°C. Flavor compounds were measured at the end of the fermentation. For each flavor compound, the one-way ANOVA with Dunnet's multiple comparison test was applied to determine significance of any difference with the control strain ER18 (* p < 0.05; **p < 0.01; ***p < 0.001).

Figure 7. Evaluation of the flavor profile in brewing wort of ER18 and its TORI, FAS2 double replacement strain.

The T0R1*216E and FAS2BTC.1D alleles were engineered into the ER18 strain to obtain the ER18T0R1*216E FAS2BTC.1D strain. Both strains were used for fermentations in 17°P Brouwland brewing wort with 20 ppm 02 at 20°C. Flavor compounds were measured at the end of the fermentation. For each flavor compound, the unpaired t-test was applied to determine significance of any difference between the two strains (* p < 0.05; **p < 0.01; ***p < 0.001).

Figure 8. Location of the mutant amino acid residues identified in the Fas2 alpha-subunit of the fatty acid synthase complex.

The huge Fas2 protein contains domains for acyl carrier protein (ACP), ketoacyl reductase (KR), ketoacyl synthase (KS), phosphopantetheinyl transferase (PPT), and two structural domains (SD1 and SD2) (86). The mutations identified in this paper, A57T, S565N, A1136T, V1624I, S1800N, are indicated. A1136T and V1624I are unique among sequenced S. cerevisiae strains. Also indicated is the previously identified G1250S mutation, that causes both increased production of ethyl hexanoate and cerulenin resistance.

Detailed description

Definitions

The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. Any reference signs in the claims shall not be construed as limiting the scope. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes. Where the term "comprising" is used in the present description and claims, it does not exclude other elements or steps. Where an indefinite or definite article is used when referring to a singular noun e.g. "a" or "an", "the", this includes a plural of that noun unless something else is specifically stated. Furthermore, the terms first, second, third and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein. The following terms or definitions are provided solely to aid in the understanding of the invention. Unless specifically defined herein, all terms used herein have the same meaning as they would to one skilled in the art of the present invention. Practitioners are particularly directed to Michael . Green and Joseph Sambrook, Molecular Cloning: A Laboratory Manual, 4 ^th ed., Cold Spring Harbor Laboratory Press, Plainsview, New York (2012); and Ausubel et al., Current Protocols in Molecular Biology (Supplement 47), John Wiley & Sons, New York (1999), for definitions and terms of the art. The definitions provided herein should not be construed to have a scope less than understood by a person of ordinary skill in the art.

In this application, two novel mutations were identified in the Fas2 protein of yeast, more particularly of Saccharomyces cerevisiae. Said mutations were shown to be needed and sufficient to enhance the production of the rose-like flavor 2-phenylethyl acetate, a highly wanted flavor in the production of fermented solutions. Therefore and in a first aspect, a Fas2 yeast protein comprising an A1136T mutation and/or a V1624I mutation is provided. This is equivalent as saying that a mutant Fas2 yeast protein is provided wherein said mutant protein comprises an A1136T and/or a V1624I mutation. This means that the amino acid alanine (A) on position 1136 from a wild-type Fas2 protein has been mutated in a threonine (T) and/or that the amino acid valine (V) on position 1624 from a wild-type Fas2 protein has been mutated in an isoleucine (I). Mutant Fas2 proteins that have additional mutations (besides A1136T and/or V1624I) are also envisaged in this application as long as the additional Fas2 mutations do not negatively interfere with Fas2 function. Thus, in one embodiment a mutant Fas2 yeast protein or a mutant Fas2 protein from Saccharomyces cerevisiae is provided, wherein said mutant protein has at least an A1136T mutation or a V1624I mutation. In a particular embodiment, said mutant Fas2 protein has an amino acid sequence with a sequence identity to SEQ ID N° 4 preferably of at least 80%, 81%, 82%, 83%,84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% over a range of at least 1500 amino acids or more particularly over a range of at least 1800 amino acids or more particularly over the full length protein. In a further particular embodiment, said mutant Fas2 protein is the amino acid sequence depicted as SEQ ID No 4. In the rest of this document, the above described mutant Fas2 yeast proteins or mutant Fas2 proteins from Saccharomyces cerevisiae will be referred to as "one of the mutant Fas2 proteins of the application". In another embodiment, a nucleic acid molecule is provided encoding a mutant Fas2 protein, wherein said mutant Fas2 protein comprises an A1136T and/or a V1624I mutation. This is equivalent as saying that a mutant FAS2 allele is provided, wherein said mutant allele encodes a mutant Fas2 protein comprising an A1136T and/or a V1624I mutation. In particular embodiments, said nucleic acid molecule encodes one of the mutant Fas2 proteins of the application. In more particular embodiments, said nucleic acid molecule has a sequence identity to SEQ ID N° 2 preferably of at least 80%, 81%, 82%, 83%,84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% over a range of at least 5000 bases or more particularly over a range of at least 5500 bases or more particularly over the full length nucleic acid molecule. In more particular embodiments, said nucleic acid molecule is depicted in SEQ ID No 2. In the rest of this document, the above described nucleic acid molecules encoding a mutant Fas2 yeast protein will be referred to as "one of the nucleic acid molecules of the application".

The "sequence identity" of two related nucleotide or amino acid sequences, expressed as a percentage, refers to the number of positions in the two optimally aligned sequences which have identical residues (xlOO) divided by the number of positions compared. A gap, i.e., a position in an alignment where a residue is present in one sequence but not in the other is regarded as a position with non-identical residues. The alignment of the two sequences is performed by the Needleman and Wunsch algorithm (Needleman and Wunsch (1970) J Mol Biol. 48: 443-453). The computer-assisted sequence alignment above, can be conveniently performed using standard software program such as GAP which is part of the Wisconsin Package Version 10.1 (Genetics Computer Group, Madison, Wisconsin, USA) using the default scoring matrix with a gap creation penalty of 50 and a gap extension penalty of 3. Sequences are indicated as "essentially similar" when such sequences have a sequence identity of at least about 75%, particularly at least about 80 %, more particularly at least about 85%, quite particularly about 90%, especially about 95%, more especially about 100%, quite especially are identical.

In another embodiment, a chimeric gene is provided comprising a promoter which is active in a eukaryotic cell, one of the nucleic acid molecules of the application and a 3' end region involved in transcription termination and/or polyadenylation. In another embodiment, a vector is provided comprising said chimeric gene or one of the nucleic acid molecules of the application.

As used herein, "nucleic acid" includes reference to a deoxyribonucleotide or ribonucleotide polymer in either single- or double-stranded form, and unless otherwise limited, encompasses known analogues having the essential nature of natural nucleotides in that they hybridize to single-stranded nucleic acids in a manner similar to naturally occurring nucleotides (e.g. peptide nucleic acids).

By "encoding" or "encodes" or "encoded", with respect to a specified nucleic acid, is meant comprising the information for transcription into an NA molecule and in some embodiments, translation into the specified protein or amino acid sequence. A nucleic acid encoding a protein may comprise non-translated sequences (e.g., introns) within translated regions of the nucleic acid, or may lack such intervening non- translated sequences (e.g., as in cDNA). The information by which a protein is encoded is specified by the use of codons. Typically, the amino acid sequence is encoded by the nucleic acid using the "universal" genetic code.

A "chimeric gene" or "chimeric construct" is a recombinant nucleic acid sequence in which a promoter or regulatory nucleic acid sequence is operably linked to, or associated with, a nucleic acid sequence that codes for a mRNA and encodes an amino acid sequence, such that the regulatory nucleic acid sequence is able to regulate transcription or expression of the associated nucleic acid coding sequence. The regulatory nucleic acid sequence of the chimeric gene is not operably linked to the associated nucleic acid sequence as found in nature.

In the present application a "promoter" comprises regulatory elements, which mediate the expression of a nucleic acid molecule. For expression, the nucleic acid molecule must be linked operably to or comprise a suitable promoter which expresses the gene at the right point in time and with the required spatial expression pattern. The term "operably linked" as used herein refers to a functional linkage between the promoter sequence and the gene of interest, such that the promoter sequence is able to initiate transcription of the gene of interest. A promoter that enables the initiation of gene transcription in a eukaryotic cell is referred to as being "active". To identify a promoter which is active in a eukaryotic cell, the promoter can be operably linked to a reporter gene after which the expression level and pattern of the reporter gene can be assayed. Suitable well-known reporter genes include for example beta- glucuronidase, beta-galactosidase or any fluorescent or luminescent protein. The promoter activity is assayed by measuring the enzymatic activity of the beta-glucuronidase or beta-galactosidase. Alternatively, promoter strength may also be assayed by quantifying mRNA levels or by comparing mRNA levels of the nucleic acid, with mRNA levels of housekeeping genes such as 18S rRNA, using methods known in the art, such as Northern blotting with densitometric analysis of autoradiograms, quantitative real-time PCR or RT-PCR (Heid et al., 1996 Genome Methods 6: 986-994).

The term "a 3' end region involved in transcription termination or polyadenylation" encompasses a control sequence which is a DNA sequence at the end of a transcriptional unit which signals 3' processing or polyadenylation of a primary transcript and is involved in termination of transcription. The control sequence for transcription termination or terminator can be derived from a natural gene or from a variety of genes. For expression in yeast the terminator to be added may be derived from, for example, the TEF or CYCl genes or alternatively from another yeast gene or less preferably from any other eukaryotic or viral gene. In a particular embodiment the promoter in the chimeric gene of the invention is active in yeast. In a preferred embodiment, said promoter is selected from the list comprising pTEFl (Translation Elongation Factor 1); pTEF2; pHXTl (Hexose Transporter 1); pHXT2; pHXT3; pHXT4; pTDH3 (Triose-phosphate Dehydrogenase) also known in the art as pGADPH (Glyceraldehyde-3-phosphate dehydrogenase) or pGDP or pGLDl or pHSP35 or pHSP36 or pSSS2; pTDH2 also known in the art as pGLD2; pTDHl also known in the art as pGLD3; pADHl (Alcohol Dehydrogenase) also know in the art as pADCl; pADH2 also known in the art as pAD 2; pADH3; pADH4 also known in the art as pZRG5 or pNRC465; pADH5; pADH6 also known in the art as pADHVI; pPGKl (3-Phosphoglycerate Kinase); pGALl (Galactose metabolism); pGAL2; pGAL3; pGAL4; pGAL5 also known in the art as pPGM2 (Phosphoglucomutase); pGAL6 also known in the art as pLAP3 (Leucine Aminopeptidase) or pBLHl or pYCPl; pGAL7; pGALlO; pGALll also known in the art as pMED15 or pRAR3 or pSDS4 or SPT13 or ABE1; pGAL80; pGAL81; pGAL83 also know in the art as pSPMl; pSIP2 (SNFl-interacting Protein) also know in the art as pSPM2; pMET (Methionine requiring); pPMAl (Plasma Membrane ATPase) also known in the art as pKTHO; pPMA2; pPYKl (Pyruvate Kinase) also known in the art as pCDC19; pPYK2; pENOl (Enolase) also known in the art as pHSP48; pEN02; pPHO (Phosphate metabolism); pCUPl (Cuprum); pCUP2 also known in the art as pACEl; pPET56 also known in the art as pMRMl (Mitochondrial rRNA Methyltransferase); pNMTl (N-Myristoyl Transferase) also known in the art as pCDC72; pGREl (Genes de Respuesta a Estres); pGRE2; GRE3; pSIP18 (Salt Induced Protein); pSV40 (Simian Vacuolating virus) and pCaMV (Cauliflower Mosaic Virus). These promoters are widely used in the art. The skilled person will have no difficulty identifying them in databases. For example, the skilled person will consult the Saccharomyces genome database website (http://www.yeastgenome.org/) or the Promoter Database of Saccharomyces cerevisiae (http://rulai.cshl.edu/SCPD/) for retrieving the yeast promoters' sequences. Yeast, as used here, can be any yeast useful for industrial applications. Preferable, said yeast is useful for ethanol production, including, but not limited to Saccharomyces, Zygosaccharomyces, Brettanomyces and Kluyveromyces. Preferably, said yeast is a Saccharomyces sp., even more preferably it is a Saccharomyces cerevisiae sp. The term "vector" refers to any linear or circular DNA construct comprising the above described chimeric gene of the invention or one of the nucleic acid molecules of the application. The vector can refer to an expression cassette or any recombinant expression system for the purpose of expressing one of the nucleic acid molecules of the application in vitro or in vivo, constitutively or inducibly, in any cell, including, in addition to plant cells, prokaryotic, yeast, fungal, insect or mammalian cells. The vector can remain episomal or integrate into the host cell genome. The vector can have the ability to self-replicate or not (i.e., drive only transient expression in a cell). The term includes recombinant expression cassettes that contain only the minimum elements needed for transcription of the recombinant nucleic acid. The vector of the invention is a "recombinant vector" which is by definition a man-made vector. In a second aspect, a microorganism comprising a mutant Fas2 yeast protein is provided wherein said mutant Fas2 protein comprises an A1136T mutation and/or a V1624I mutation. In particular embodiments, a microorganism is provided comprising any of the mutant Fas2 proteins described in this application. Also, a microorganism is provided comprising one of the nucleic acid molecules of the application or comprising the chimeric gene of the application or comprising the vector of the application. In particular embodiments, a microorganism is provided comprising a mutant FAS2 allele, wherein said mutant allele encodes a Fas2 protein comprising an A1136T mutation and/or a V1624I mutation. In particular embodiments, said microorganism is an engineered microorganism. In other particular embodiments, said microorganism is a yeast. In more particular embodiments, said microorganism is an engineered yeast. In even more particular embodiments, said yeast is useful for ethanol production, including, but not limited to Saccharomyces, Zygosaccharomyces, Brettanomyces and Kluyveromyces. Preferably, said yeast is a Saccharomyces sp., even more preferably it is a Saccharomyces cerevisiae sp. In most particular embodiments, said yeast is not BTC.1D or WLP575. Means and methods to engineer microorganisms, particularly yeasts are well known by the person skilled in the art. The most known techniques involve traditional genetic transformation of yeast and recombinant DNA techniques. Nowadays, the most attractive technique to engineer a microorganism is by the use of nucleases, such as zinc-finger nucleases (ZFNs), Transcription Activator-Like Effector Nucleases (TALENs), meganucleases but especially the C ISP -Cas system. "Nucleases" as used herein are enzymes that cut nucleotide sequences. These nucleotide sequences can be DNA or RNA. If the nuclease cleaves DNA, the nuclease is also called a DNase. If the nuclease cuts RNA, the nuclease is also called an RNase. Upon cleavage of a DNA sequence by nuclease activity, the DNA repair system of the cell will be activated. Yet, in most cases the targeted DNA sequence will not be repaired as it originally was and small deletions, insertions or replacements of nucleic acids will occur, mostly resulting in a mutant DNA sequence. ZFN are artificial restriction enzymes generated by fusing a zinc finger DNA- binding domain to a DNA cleavage domain. Zinc finger domains can be engineered to target desired DNA sequences, which enables zinc-finger nucleases to target a unique sequence within a complex genome. By taking advantage of endogenous DNA repair machinery, these reagents can be used to precisely alter the genomes of simple and higher organisms. Other technologies for genome customization that can be used to knock out genes are meganucleases and TAL effector nucleases (TALENs, Cellectis bioresearch). A TALEN ^® is composed of a TALE DNA binding domain for sequence-specific recognition fused to the catalytic domain of an endonuclease that introduces double strand breaks (DSB). The DNA binding domain of a TALEN ^® is capable of targeting with high precision a large recognition site (for instance 17bp). Meganucleases are sequence-specific endonucleases, naturally occurring "DNA scissors", originating from a variety of single-celled organisms such as bacteria, yeast, algae and some plant organelles. Meganucleases have long recognition sites of between 12 and 30 base pairs. The recognition site of natural meganucleases can be modified in order to target native genomic DNA sequences (such as endogenous genes). Another recent and very popular genome editing technology is the C ISP -Cas system, which can be used to achieve RNA-guided genome engineering. CRISPR interference is a genetic technique which allows for sequence-specific control of gene expression in prokaryotic and eukaryotic cells. It is based on the bacterial immune system-derived CRISPR (clustered regularly interspaced palindromic repeats) pathway and has been modified to edit basically any genome. By delivering the Cas nuclease (in many cases Cas9) complexed with a synthetic guide RNA (gRNA) in a cell, the cell's genome can be cut at a desired location depending on the sequence of the gRNA, allowing existing genes to be removed and/or new one added and/or more subtly removing, replacing or inserting single nucleotides (e.g. DiCarlo et al 2013 Nucl Acids Res doi:10.1093/nar/gktl35; Sander & Joung 2014 Nat Biotech 32:347- 355). As previously explained, production of particular flavors such as 2-phenylethyl acetate or phenylethanol can have positive effects on consumers' appreciation of a certain fermented product. Indeed, for example, fermented beverages (e.g. beer) with high concentrations of 2-phenylethyl acetate are often positively perceived as having a "honey-like" or "rose-like" flavor. Therefore, the yeast strains which are subject of and are disclosed in the current application are extremely useful in the fermentation industry. Fermented products can have applications in food (for example but not limited to chocolate) and beverages (for example but not limited to beer, wine, sake) as well as in general industry (for example but not limited to bioethanol production) where the production of an enjoyable flavor is desired. Thus, also envisaged in this application is a fermented solution comprising the yeast strains of this application, or more particularly comprising a yeast strain that comprises a mutant Fas2 yeast protein, wherein said mutant protein comprises an A1136T mutation and/or a V1624I mutation. In even more particular embodiments, said fermented solution is a fermented beverage or is a fermented solution not suited for consumption. Non-limiting examples of a fermented beverage are beer, wine, sake, ... A non-limiting example of a fermented solution not suited for consumption is bio-ethanol. In an even more particular embodiment, said fermented beverage is beer.

As clearly illustrated in this application, expression of the mutant Fas2 protein in yeast increases the production of 2-phenylethyl acetate but also that of isobutanol, isoamyl alcohol and phenylethanol. The observation that the mutant Fas2 protein has a dominant effect is of particular interest for industrial application because the mutation does not have to be homozygous to lead to the mutant phenotype, i.e. increased production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and phenylethanol. Therefore, and in a third aspect of this application, the use is provided of a mutant Fas2 yeast protein to modulate the flavor profile of a fermented solution, wherein said mutant Fas2 protein comprises an A1136T mutation and/or a V1624I mutation. In a particular embodiment, said mutant Fas2 yeast protein comprises the amino acid sequence depicted as SEQ ID No 4. Also envisaged is the use of any of the nucleic acid molecules of the application or of any of the chimeric genes comprising one of said nucleic acid molecules or of a vector comprising one of said nucleic acid molecules or one of said chimeric genes to modulate the flavor profile of a fermented solution.

In a particular embodiment of the third aspect and of its embodiments, said flavor profile is a rose flavor profile. In other particular embodiments, said modulation is an increase in flavor profile or an increase in flavor, more particularly an increase in rose flavor, even more particularly an increase in the production of 2-phenylethyl acetate and/or phenylethanol. 2-Phenylethyl acetate and phenylethanol are two molecules well-known for their rose-like flavor. The use is thus provided of any of the mutant Fas2 yeast proteins of the application or of the nucleic acid molecule encoding said Fas2 mutant protein or of the chimeric gene comprising said nucleic acid molecule or of the vector comprising said nucleic acid molecule or said chimeric gene to increase rose flavor in a fermented solution. More particularly said rose flavor is determined by 2-phenylethyl acetate and/or phenylethanol. This is equivalent as saying that said use is provided to produce a fermented solution with an increased rose flavor or with an increased level of 2-phenylethyl acetate and/or phenylethanol. In alternative embodiments, said use is provided to produce a high rose flavor producing yeast or a high 2-phenylethyl acetate producing yeast. The application also provides the use of a mutant Fas2 yeast protein to increase the production of 2- phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol during yeast fermentation, wherein said mutant Fas2 protein comprises an A1136T mutation and/or a V1624I mutation. In a particular embodiment, said mutant Fas2 yeast protein comprises the amino acid sequence depicted as SEQ ID No 4. In other embodiments, the use is provided of any of the nucleic acid molecules of the application or of the chimeric gene comprising said nucleic acid molecule or of the vector comprising said nucleic acid molecule or said chimeric gene to increase the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol during yeast fermentation. Also the use of a yeast comprising a mutant Fas2 protein comprising an A1136T mutation and/or a V1624I mutation is provided to increase the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol during fermentation. In a particular embodiment, said mutant Fas2 yeast protein comprises the amino acid sequence depicted as SEQ ID No 4. In other embodiments, the use is provided of any of the mutant Fas2 yeast proteins of the application or any of the nucleic acid molecules of the application or of the chimeric gene comprising said nucleic acid molecule or of the vector comprising said nucleic acid molecule or said chimeric gene to increase the production of at least one molecule from the group consisting of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and phenylethanol in a microorganism. In a particular embodiment, the use of any of the mutant Fas2 yeast proteins of the application or any of the nucleic acid molecules of the application or of the chimeric gene comprising said nucleic acid molecule or of the vector comprising said nucleic acid molecule or said chimeric gene is provided to increase the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and phenylethanol in a microorganism or during yeast fermentation.

In current application a second gene involved in rose flavor production is disclosed, more precisely TORI. It is demonstrated that deleting the expression of a functional Tori protein reduces the production of rose flavor, more precisely that of 2-phenylethyl acetate. Therefore, in a particular embodiment of the above described third aspect and of its embodiments, said use is combined with the use of a functional TORI allele. "Functional" as used herein refers to a wild-type version of TORI.

"Increasing the production of rose flavor" or "increasing the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol" as used herein means that the microorganism (in particular embodiments the yeast), that comprises one of the mutant Fas2 yeast proteins of the application produces more rose flavor or more 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol, compared to a corresponding reference microorganism (in particular embodiments the yeast) lacking said mutant Fas2 protein and comprising a functional or wild-type Fas2 protein. Synonyms for increasing are improving, augmenting, boosting, enhancing. In particular embodiments, "increasing the production of rose flavor" or "increasing the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol" means an at least 5%, at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90, at least 95% or 100% increase or a 2-fold, 3-fold, 5-fold or 10-fold increase in the production of rose flavor or in the production of 2- phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol. In other particular embodiments, "increasing the production of rose flavor" or "increasing the production of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol" means obtaining an increased production of said rose flavor or of said 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol of between 10% and 50%, of between 20% and 60%, of between 30% and 70%, of between 40% and 80% or of between 60% and 150% of the production of said rose flavor or 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol in conditions where the mutant Fas2 yeast protein of the application is not expressed. The skilled person is familiar with methods to analyse and quantify the production of rose flavor or of 2-phenylethyl acetate, isobutanol, isoamyl alcohol or phenylethanol from a sample. In this application, those molecules were quantified using headspace gas chromatography coupled with flame ionization detection (HS-GC-FID) as clearly described in the materials and method section. The odor or flavor is the property of a substance that activates the sense of smell. Odor, smell, scent, stench all refer to sensations perceived through the nose by the olfactory nerves. Alcohols with improved flavors or odors are highly desired for blending fermented or alcoholised beverages but also for the industrial production of non-beverage alcohol.

In a fourth aspect, a method is provided to modulate the flavor profile or more particularly to modulate the rose flavor profile of a fermented solution, said method comprising expressing the Fas2 mutant protein comprising an A1136T and/or a V1624I mutation or expressing the nucleic acid molecule encoding said mutant Fas2 protein or expressing the chimeric gene comprising said nucleic acid molecule in a yeast. In a particular embodiment, said yeast is used during the production of said fermented solution. In another particular embodiment, said flavor profile is determined by 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol.

With "expressing" or "expression" or "functional expression" of the Fas2 mutant protein, it is meant the transcription and production of said mutant Fas2 protein.

In another embodiment, a method is provided to increase the flavor of a fermented solution or to increase the level of rose flavor in a fermented solution or to increase the production of rose flavor in a fermented solution, said method comprising expressing one of the mutant Fas2 proteins of the application or one of the nucleic acid molecules of the application or the chimeric gene comprising one of said nucleic acid molecules in a yeast. In particular embodiments, said yeast is used during the production of said fermented solution. In other particular embodiments, said fermented solution is an alcoholic beverage. In other embodiments, a method is provided to produce a fermented solution with increased level of rose flavor or to produce a flavor enriched alcohol, said method comprising the step of expressing in a yeast any of the mutant Fas2 proteins any of the nucleic acid sequences or any of the chimeric genes described in the first aspect of current application and in the embodiments of the first aspect.

In particular embodiments, said increased level is at least 5%, at least 10%, at least 15%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or 100% more or at least 2-fold, at least 3-fold, at least 5-fold or at least 10-fold more or said increased level is between 10% and 50%, of between 20% and 60%, of between 30% and 70%, of between 40% and 80% or of between 60% and 150% more compared to a control yeast not expressing any of the mutant Fas2 proteins or any of the nucleic acid sequences or any of the chimeric genes described in the first aspect of current application and in the embodiments of the first aspect. Said control yeast expressed a wild-type Fas2 protein. Means and methods to determine the expression of protein- encoding genes or to determine the activity of proteins are well-known in the art.

Also a method is provided to produce a high rose flavor producing yeast, said method comprising the step of expressing in said eukaryotic cell any of the mutant Fas2 proteins or any of the nucleic acid sequences or any of the chimeric genes described in the first aspect of current application and in the embodiments of the first aspect.

A "high rose flavor producing yeast" as used in this application, refers to a yeast strain that produces a higher amount of 2-phenylethyl acetate and/or phenylethanol compared to a wild-type yeast strain. Said wild-type yeast strain is a yeast strain which was not selected and/or engineered to produce a high level of 2-phenylethyl acetate and/or phenylethanol. In particular embodiments, said higher amount is an at least 5%, at least 10%, at least 15%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or 100% higher amount or a between 10% and 50%, or a between 20% and 60%, or a between 30% and 70% or a between 40% and 80% bigger amount compared to said wild-type yeast strain. In alternative embodiments, said higher amount is a statistical significant higher amount of at least 2-phenylethyl acetate or phenylethanol. "Statistical significant" as used here refers to a p-value of less than 0.5, which is a commonly accepted level for statistical significance and well known by the person skilled in the art. In a particular extension of the methods described in the fourth aspect or in one of the accompanying embodiments from aspect four, the methods further include a step of quantifying at least one rose flavor. Said rose flavor can be 2-phenylethyl acetate or phenylethanol. In another particular extension of the methods described in the fourth aspect or in one of the accompanying embodiments from aspect four, said rose flavor is determined by the presence of 2-phenylethyl acetate and/or phenylethanol. In even another particular extension of the fourth aspect and of all embodiments of the fourth aspect, said microorganisms is a yeast, more particularly a yeast useful for ethanol production, including, but not limited to Saccharomyces, Zygosaccharomyces, Brettanomyces and Kluyveromyces. Even more particularly, said yeast is a Saccharomyces sp., most particularly it is a Saccharomyces cerevisiae sp. Said yeast strains are particularly useful for industrial fermentation of beverages at conditions wherein increase flavor (more particularly rose flavor) is desired.

In a particular extension of any of the aspects described above or of any of the embodiments of said aspects, the said uses or methods to increase the production of rose flavor in a eukaryotic cell or to produce a fermented solution with increased levels of 2-phenylethyl acetate, isobutanol, isoamyl alcohol and/or phenylethanol or to produce a flavor enriched alcohol, do not reduce or negatively influence alcohol and more particular ethanol production.

We have also identified and isolated a novel mutant allele from the Saccharomyces cerevisiae TORI gene that when expressed homozygously in industrial S. cerevisiae strains or when expressed in the absence of a wild-type (and thus fully functional) TORI gene limits the production of rose flavor, including 2- phenylethyl acetate. The novel mutant allele comprises a mutated nucleic acid at position 646 of the open reading frame sequence depicted in SEQ ID No. 6, wherein said mutation introduces a stop codon. "Position 646" as used herein refers to the nucleic acid that is 645 positions removed downstream from the first nucleotide (i.e. adenosine) from the start codon. This position is indicated in SEQ ID No. 6 by underlining. As such, said mutant TORI allele encodes the C-terminally truncated Tori protein as depicted in SEQ ID No. 7.

Therefore, in one aspect, this application provides a C-terminally truncated Tori yeast protein. In one embodiment, said protein lacks residues 216-2470 of a wild-type full length yeast Tori protein. The amino acid sequence of this wild-type full length yeast Tori protein is depicted in or defined by SEQ ID No. 8. "Lacking residues 216-2470" is equivalent as saying that the truncated protein lacks residues 216 until 2470 or that the truncated protein is devoid of the amino acid region from 216 until 2470.

The term "truncated protein" refers to a protein which lacks one or more amino acids of the wild-type version of the protein, preferably the protein lacks one or more functional domains present in the wild- type protein. This is typically achieved by a mutation. A "C-terminally truncated Tori yeast protein" as used here means that part of the C-terminus of the wild-type yeast Tori protein (defined by SEQ ID No. 8) is missing. This does not mean that the rest of the truncated Tori protein should be identical in length or sequence to the wild-type yeast Tori protein. A C-terminally truncated Tori protein lacking residues 216-2470 thus also comprises Tori proteins with additional mutations or truncations besides the C- terminal truncation from residue 216 until residue 2470.

The C-terminally truncated Tori protein from the application lacks most of the C-terminal part of the wild-type version of the protein. This can be achieved by a mutation specifically inducing premature termination of messenger NA translation. As a non-limiting example, said C-terminal truncated protein may be created by a point mutation introducing a stop codon in the reading frame of SEQ ID N° 5, or by a deletion or insertion resulting in a stop codon. In the latter case, the deletion or insertion may cause a frame shift, resulting in a mutant sequence at the C-terminal end of the truncated protein. Indeed, in one embodiment, an isolated truncated yeast Tori protein is provided, wherein said protein lacks residues 216-2470 or 215-2470 or 214-2470 or 213-2470 or 212-2470 or 211-2470 or 210-2470 or 200- 2470 or 150-2470 or 100-2470 or 80-2470 or 50-2470 or 20-2470 of a wild-type full length yeast Tori protein defined by SEQ ID No. 8.

In a particular embodiment, an isolated truncated yeast Tori protein is provided, wherein said protein consists of amino acid 1-215 of a wild type full length Tori protein defined by SEQ ID No 8. In a most particular embodiment, a C-terminally truncated Tori yeast protein is provided as depicted in SEQ ID No. 7. In another embodiment, a nucleic acid molecule is provided encoding one of the above described C-terminally truncated Tori yeast proteins. Any of said nucleic acid molecules is thus equivalent to a mutant yeast TORI allele. In a most particular embodiment, said mutant yeast TORI allele is depicted in SEQ ID No. 6.

In another aspect, a microorganism is provided comprising one of the above described C-terminally truncated Tori yeast proteins or comprising a nucleic acid molecule encoding one of said truncated Tori proteins. In a particular embodiment, a microorganism is provided comprising the C-terminally truncated Tori yeast protein as depicted in SEQ ID No. 7 or comprising the mutant Tori allele depicted in SEQ ID No. 6. In another particular embodiment, said microorganism is an engineered or recombinant microorganism. In another particular embodiment, said microorganism is a yeast. In another particular embodiment, said microorganism is an engineered or recombinant yeast. In particular embodiments, said yeast is useful for ethanol production, including, but not limited to Saccharomyces, Zygosaccharomyces, Brettanomyces and Kluyveromyces. Even more particularly, said yeast is a Saccharomyces sp., most particularly it is a Saccharomyces cerevisiae sp. Said yeast strains are particularly useful for industrial fermentation at conditions wherein too much flavor, more particularly too much rose flavor is produced. The application also provides a fermented solution comprising one of the above described yeasts. In a particular embodiment, said fermented solution is an alcoholic beverage, for example but not limited to beer. In other particular embodiments, said fermented solution is a non-food or non-beverage solution, for example but not limited bio-ethanol.

In another aspect, the use of a non-functional Tori yeast protein or of a mutant TORI allele is provided to modulate the flavor profile of a fermented solution. In one embodiment, said flavor profile modulation is a decrease in flavor. In a particular embodiment, said flavor is rose flavor. In a more particular embodiment, the use of a non-functional Tori yeast protein or of a mutant TORI allele is provided to reduce the production of 2-phenylethyl acetate or phenylethanol in a fermented solution. In this application a non-functional Tori yeast protein is a mutant Tori yeast protein with a disrupted, partially deleted or completely deleted function. The one or more mutations that lead to a nonfunctional Tori can be frame shift mutations, nonsense mutations or missense mutations. In a particular embodiment, said mutation is a nonsense mutation, thus leading to a premature stop codon. This is equivalent as saying that the use of a C-terminally truncated Tori protein is provided to modulate the flavor profile of a fermented solution. In an even more particular embodiment, the use of a nonfunctional Tori yeast protein or of a mutant yeast Tori allele encoding said non-functional Tori protein is provided to modulate the flavor profile of a fermented solution, wherein said mutant Tori protein is depicted in SEQ ID No. 7 and said mutant allele is depicted in SEQ ID No. 6. In even more particular embodiments, said modulation of flavor profile is a reduction in flavor, more particularly rose flavor, even more particularly a reduction of 2-phenylethyl acetate and/or phenylethanol.

In another aspect, the use of a mutant yeast TORI allele is provided to limit the production of rose flavor in a microorganism or to produce a fermented solution with a reduced level of rose flavor or to produce a flavor neutral alcohol. In one embodiment, the use of a yeast is provided to limit the production of rose flavor during yeast fermentation or to produce a fermented solution with a reduced level of rose flavor or to produce a flavor neutral alcohol, wherein said yeast comprises a disrupted, partially deleted or completely deleted TORI allele. In another embodiment, the use of a yeast is provided to limit the production of rose flavor during yeast fermentation or to produce a fermented solution with a reduced level of rose flavor or to produce a flavor neutral alcohol, wherein said yeast comprises the mutant yeast TORI allele depicted in SEQ ID No. 6.

In another embodiment, the use of a mutant yeast TORI allele, wherein said allele encodes a nonfunctional Tori protein or the use of a yeast comprising said mutant yeast TORI allele is provided, to produce a low rose flavor producing yeast. In particular embodiments, said rose flavor is selected from the list consisting of 2-phenylethyl acetate and phenylethanol.

In another aspect, the use of a yeast is provided to decrease rose flavor production in a fermented solution or to produce a fermented solution with a reduced level of rose flavor or to produce a rose flavor neutral alcohol, wherein said yeast has a disrupted, partially deleted or completely deleted functional expression of TORI. In particular embodiments, said rose flavor is 2-phenylethyl acetate or phenylethanol.

In yet another aspect, a method to limit the production of rose flavor during yeast fermentation or to produce a fermented solution with a reduced level of rose flavor or to produce a flavor neutral alcohol is provided, said method comprising the step of disrupting, partially deleting or completely deleting Tori function in a yeast, wherein said yeast is used during the production of said fermented solution or said flavour neutral alcohol. In a more particular embodiment, a method to limit the production of rose flavor during yeast fermentation or to produce a fermented solution with a reduced level of rose flavor or to produce a flavor neutral alcohol is provided, said method comprising the step of replacing a wild-type TORI allele with the mutant yeast TORI allele depicted in SEQ ID No. 6 in yeast. In a particular embodiment, a method is provided to decrease, to limit or reduce the 2-phenylethyl acetate production in a yeast-mediated fermentation, said method comprising disrupting, partially deleting or completely deleting Tori function in said yeast.

"Disrupted, partially deleted or completely deleted function" or "disrupting, partially deleting or completely deleting the functional expression" is equivalent as saying partially or completely inhibiting the formation of a functional mRNA molecule encoding Tori. Means and methods to disrupt, partially deleted or completely delete a gene or protein are well known in the art. The skilled person can select from a plethora of techniques to affect the expression or function of Tori. One very efficient technique is the Crispr/Cas technology which has also been used in this application e.g. in Examples 3 and 4. At the DNA level, disruption, partial deletion or complete deletion can for example be achieved by removing or disrupting a gene encoding Tori or by mutations in the promoter of a gene encoding Tori. Non-limiting examples are knock-outs or loss-of-function mutations but also gain-of-function mutations and dominant negative mutations can disrupt the functional expression or inhibit the formation of a functional mRNA molecule. A "knock-out" can be a gene knockdown (leading to reduced gene expression) or the gene can be knocked out by a mutation such as, a point mutation, an insertion, a deletion, a frameshift, or a missense mutation by techniques known in the art. The lack of transcription can e.g. be caused by epigenetic changes (e.g. DNA methylation) or by loss-of-function mutations. A "loss-of-function" or "LOF" mutation as used herein is a mutation that prevents, reduces or abolishes the function of a gene product as opposed to a gain-of-function mutation that confers enhanced or new activity on a protein. The disclosed TORI mutant allele has a loss-of-function effect and is recessive, meaning that the mutation has to be homozygous to lead to the mutant phenotype. Both dominant negative or LOF mutations can be caused by a wide range of mutation types, including, but not limited to, a deletion of the entire gene or part of the gene, splice site mutations, frame-shift mutations caused by small insertions and deletions, nonsense mutations, missense mutations replacing an essential amino acid and mutations preventing correct cellular localization of the product.

"Limiting the production of rose flavor" or "limiting rose flavor production" as used herein means that the microorganism (in particular embodiments the yeast), that comprises a disrupted, partially deleted or completely deleted Tori produces less rose flavor, more particularly less of at least 2-phenylethyl acetate or phenylethanol, compared to a corresponding reference microorganism (in particular embodiments the yeast) lacking a disrupted, partially deleted or completely deleted Tori. Synonyms for limiting are reducing, lowering, restricting, constraining, decreasing, lessening, diminishing. In particular embodiments, "limiting the production of rose flavor" or "a reduced level of rose flavor" means an at least 5%, at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90, at least 95% or 100% reduction in the production of said rose flavor or more particularly of 2-phenylethyl acetate or phenylethanol. In other particular embodiments, "limiting the production of rose flavor" or "a reduced level of rose flavor" means obtaining a reduced production of rose flavor of between 10% and 50%, of between 20% and 60%, of between 30% and 70% or of between 40% and 80% of the production of rose flavor in conditions where Tori is not disrupted, not partially deleted or not completely deleted. The skilled person is familiar with methods to analyse and quantify the production of rose flavor, of 2-phenylethyl acetate, of phenylethanol, or of total and single acetate esters from a sample. In this application, flavor compounds including 2-phenylethyl acetate and phenylethanol were quantified using headspace gas chromatography coupled with flame ionization detection (HS-GC-FID) as clearly described in the materials and method section.

In another aspect, a method is provided to increase the level of rose flavor in a yeast fermentation, said method comprising:

a. determining the presence of one or more loss-of-function mutations in the TORI allele of said yeast;

b. restoring said one or more mutations in said yeast to wild-type or replacing the non- functional TORI allele by a wild-type TORI allele, if one of more loss-of-function mutations is present in said yeast's TORI gene.

In one embodiment, said rose flavor is 2-phenylethyl acetate. In another embodiment, said loss-of- function mutation is a reduce-of-function mutation. In a particular embodiment, said method comprises the additional step of expressing a mutant Fas2 protein in said yeast, wherein said mutant Fas2 protein comprises an A1136T mutation and/or a V1624I mutation.

It is to be understood that although particular embodiments, specific configurations as well as materials and/or molecules, have been discussed herein for cells and methods according to the present invention, various changes or modifications in form and detail may be made without departing from the scope and spirit of this invention. The following examples are provided to better illustrate particular embodiments, and they should not be considered limiting the application. The application is limited only by the claims.

Examples

Example 1: Flavor profile of a set of segregants obtained from unselected parent strains

We have obtained a pool of haploid strains for flavor profile analysis by sporulating a hybrid diploid yeast strain, obtained by crossing two descendants from the unrelated industrial yeast strains WLP575, an ale yeast, and Ethanol Red (ER), a bioethanol production yeast. BTC.1D (MATa), a segregant from WLP575, and ER18 (MATa), a segregant from ER, were crossed to generate a hybrid diploid named BTC.1D/ER18. This hybrid strain was sporulated and 574 segregants were isolated and evaluated in small-scale fermentations for flavor production. We have used violin plots to visualize the data distribution and the probability density for each flavor compound (Fig. 1). Some segregants produced much higher levels of specific flavor compounds than the two parent strains. For isobutyl acetate, one segregant produced 3.14 ppm which is almost 4 times higher than the level produced by the best parent strain (0.85 ppm). Similarly, the highest production of ethyl acetate reached in the population of segregants was 123 ppm, whereas the parent strains produced 53.6 and 47.2 ppm. For 2-PEAc, one of the segregants produced 4 ppm, while the BTC.1D and E 18 parent strains produced only 1.7 and 1.9 ppm, respectively. Because 2-PEAc is an important and highly desirable ester in alcoholic beverages due to its pleasant rose-like and honey-like aroma, and since little is known about factors controlling its synthesis, we have selected this flavor production trait for genetic analysis.

Example 2: Pooled-segregant whole-genome sequence analysis and QTL mapping

We have composed a pool of 24 selected, superior segregants with a 2-PEAc production level of at least 2.8 ppm and a second 'unselected pool' with 24 random segregants. Both pools were subjected to pooled-segregant whole-genome sequence analysis using lllumina HiSeq2000 technology (BGI, Hong Kong).

The SNP variant frequency was used for QTL mapping using the Next Generation Sequencing Eclipse Plugin (NGSEP) program in order to identify the genomic loci linked to high 2-PEAc production. The sequence reads from the parental strains and the pool were first aligned to the reference S288c genome sequence and SNPs between the two parent strains with a coverage of at least 20 times were selected following previous polygenic analysis protocols (17, 38). SNPs with the reference genome sequence but absent between the parent strains were filtered out using the NGSEP program. The SNP variant frequency was plotted against the chromosomal position using the RStudio 0.99.903 program, generating the QTL map (Fig. 2). p-values calculated from the sequencing data were plotted against the respective chromosomal position and p-values <0.05 were considered statistically significant. Because of the low number of segregants and the unexpectedly large variation of the SNP variant frequency in the random pool, we calculated the p-values for the whole genome both with the mapping result of the random pool as reference and with the mean value of 0.5 for the whole genome as reference (Fig. 2). Since both parents could contribute to the high 2-PEAc phenotype in the superior pool, the QTLs could be linked to either parent genome. A high SNP frequency, significantly above 50%, indicated linkage with the genome of the BTC.1D parent strain, while a low SNP frequency indicated linkage with the genome of the ER18 parent strain. The midline obtained fluctuated around 50% inheritance in most areas of the genome. Strong deviation from 50% inheritance was observed on chromosomes VII, X, XV and XVI. The QTLs on chromosomes X and XVI were linked to the BTC.1D parent genome while the QTLs on chromosomes VII and XV were linked to the E 18 parent genome (Fig. 2). We have selected the QTLs on chromosomes X and XVI linked to the BTC.1D parent for further analysis.

Example 3: Identification of TORI as causative gene in the major QTL on chromosome X

First, we have performed fine mapping by allele-specific PCR of the major QTL on chromosome X. Seven SNPs were chosen and scored by allele-specific PCR in the 24 individual segregants of the superior pool. The whole region of 46.5kb showed significant linkage, with a p-value lower than 0.001. The 15 kb-region with the highest linkage (p-value of 3.6x10-5) contained 12 genes, from which five had non-synonymous mutations in the ORF (Fig. 3A, B and C). Among these five genes, three were non-essential genes. Since neither the haploid parents nor the diploid hybrid showed the superior phenotype, the possible involvement of the candidate genes was investigated in a superior segregant, nr. 442. This haploid strain showed a 2-PEAc production of about 4 ppm and contained the superior allele in the region under investigation. The non-essential genes MOG1, BNA2 and TORI were deleted in this superior segregant by replacing them with a nourseothricin resistance marker cassette. The deletion strains were evaluated by fermentation for flavor production and a significant decrease in 2-PEAc production was observed only for the 442 torlA strain (Fig. 3D).

Sequence comparison of the TORI alleles from the two parents revealed several polymorphisms, eight of them leading to non-synonymous mutations (Table 1). The most striking mutation was an early stop codon at position 216 of the amino acid sequence in the ER18 parent, while the wild type Tori protein has a total length of 2470 amino acids. The truncated protein was expected to cause a phenotype similar to that of the 442 torlA strain. We inserted this specific point mutation in the superior segregant 442 by CRISPR/Cas9 methodology and compared the phenotype of the mutant strain with that of the 442 wild type and 442 torlA strains. Figure 3D shows that the 442 torlA and 442 torlE216* strains produced a similarly reduced level of 2-PEAc, which was approximately 30% lower than that of the wild type strain. This indicated that TORI affected 2-PEAc production and that the torlE216* allele, derived from ER18, behaved as a loss of function allele.

The TORI alleles were also exchanged in the parent strains. The stop codon mutation E216* was introduced in the BTC.1D parent, while the E216 residue was restored in the ER18 parent, both via CRISPR/Cas9 methodology. The 2-PEAc production by the BTC.1D torlE216* and ER18 TOR1E216 strains was evaluated in small-scale fermentations. The introduction of the early stop codon into the BTC.1D parent resulted in a decrease with about 40% in 2-PEAc production, while ER18 showed a 35% increase when the E216 residue was restored and the complete wild type Tor protein was therefore expressed (Fig. 3E). These results further confirmed that TORI was the causative gene in the major QTL on chromosome X and that Tori inactivation compromises 2-PEAc production. Example 4: Identification of FAS2 as causative gene in the major QTL on chromosome XVI

We have performed fine mapping with allele-specific PC to confirm and downscale the major QTL on chromosome XVI. Six SNPs were selected covering a region of 54 kb. The region between the first and the fifth SNP showed significant linkage with the BTC.ID parent with p-values ranging from 0.002 to 0.007 (Fig. 4A, B). This area consists of 25 genes, most of them containing non-synonymous mutations in the ORF as well as mutations in the promotor and/or terminator regions. The region was divided in 10 blocks in order to evaluate consecutive sets of genes for location of the causative gene in a process of bulk replacement (Fig. 4C). As for the major QTL on chromosome X, we chose the superior segregant 442, which contained the genomic sequence of the BTC.ID parent over the whole length of this QTL, for exchange of the selected blocks of genes.

Direct replacement of the blocks of genes was required since the superior segregant 442 is haploid and some genes in the blocks are essential. Direct replacement was achieved using CRISPR/Cas9 methodology using two gRNAs, one targeting a sequence upstream of the first SNP of interest and the other a sequence downstream of the last SNP of interest. The donor DNA was PCR-amplified from the ER18 parent, which apparently contained the genomic DNA with the inferior causative allele. The bulk replacement strains for each block of genes were used in small-scale fermentations and 2-PEAc production was measured at the end of the fermentation. This revealed a conspicuous drop in 2-PEAc production specifically with the 442 strain in which block number 7 with the BTC.ID parent sequence had been exchanged for the sequence of the ER18 parent strain (Fig. 4E).

The length of block 7 was 6.4kb and it contained one gene (FAS2) with about 500bp of its promotor and terminator. Since FAS2 encodes the alpha subunit of fatty acid synthetase, it appeared to be the best candidate causative gene in this block (Fig. 4D). Comparison of the FAS2 gene sequence of the BTC.ID and ER18 parent strains revealed 5 non-synonymous mutations in the open reading frame. We checked the occurrence of those amino acid mutations in 39 strains for which the whole genome sequence is available at the Saccharomyces Genome Database. The result is shown in Table 1. Fas2 is a highly conserved protein and among the five amino acid changes found in the BTC.ID parent sequence, two were not found in any sequenced strain. At position 1136 in the amino acid sequence, BTC.ID contained threonine, while all the other strains contained alanine while at position 1624 in the amino acid sequence, BTC.ID contained isoleucine, while it was in all other strains a valine. For practical reasons, the BTC.ID superior allele, FAS2A57T, S565N, A1136T, V1624I, S1800N, will now be named FAS2BTC.1D and the inferior allele from ER18 will be fas2ER18.

In order to confirm FAS2 as the causative allele in the major QTL on chromosome XVI, we performed FAS2 allele swapping in the parent strains by targeting the same cutting sites as used to construct the bulk replacement in segregant 442. The production of 2-PEAc by the newly constructed transformant strains was compared with that of the wild type parents BTC.1D and ER18. Figure 4F shows that the replacement of FAS2BTC.1D by the fas2ER18 allele in the BTC.1D parent led to a 25% reduction in 2-PEAc production, while the replacement of fas2ER18 by the FAS2BTC.1D allele in the ER18 parent raised it by 30%.

Example 5: Establishing the superior phenotype by replacing the TORI and FAS2 alleles in the ER18 parent strain

QTL mapping revealed four major QTLs linked to high production of 2-PEAc. Allele-specific PCR analysis with the two parent strains confirmed that the ER18 parent contained the superior allele for the QTLs on chromosome VII and XV and the inferior allele for the QTLs on chromosome X and XVI, in which we identified TORI and FAS2, respectively, as the causative genes. Hence, we decided to replace both the inferior alleles torlE216* and fas2ER18 by the TOR1E216 and FAS2BTC.1D superior alleles, respectively, in the ER18 parent strain to obtain a strain possessing all superior alleles of the four major QTLs (Fig. 5A). Also the single replacement strains were constructed. The original ER18 parent strain and the modified ER18 strains were used in small-scale fermentations and 2-PEAc production was analyzed. This showed that the replacement of torlE216* was already enough to reach the superior phenotype of the selected segregants (> 2.8 ppm 2-PEAc), while the replacement of both inferior alleles, torlE216* and fas2ER18, led to an increase with about 70% in 2-PEAc production, reaching about 3.5 ppm, almost as high as the level of 4 ppm for the top superior segregant 442 (Fig. 5A).

To test whether the FAS2BTC.1D allele was dominant or recessive in the same and in a different genetic background, we crossed the control strain ER18 T0R1*216E (MATa) and ER18 T0R1*216E FAS2BTC.1D (MATa) with ER18 T0R1*216E (MATa) and with 16D (MATa) to obtain the two pairs of diploid strains ER18 T0R1*216E/ER18 T0R1*216E and ER18 T0R1*216E/ER18 T0R1*216E FAS2BTC.1D, and 16D/ ER18 T0R1*216E and 16D/ER18 T0R1*216E FAS2BTC.1D. Strain 16D is a haploid segregant of JT22689, a strain used in "sturm" must fermentations (Austria) (21). ER18 (MATa) and 16D (MATa) contain the same wild type TORI allele (T0R1*216E or TOR1E216, respectively). Therefore, any difference between diploid strains having either ER18 T0R1*216E or ER18 T0R1*216E FAS2BTC.1D is only due to the effect of the FAS2BTC.1D allele. In both cases, the presence of a wild type FAS2 copy and one copy of the superior allele FAS2BTC.1D in the same strain was sufficient to increase 2-PEAc production significantly above that in the control strain with two FAS2 wild type alleles, showing the dominant character of the FAS2BTC.1D allele.

Example 6: Production of the other flavor compounds in the ER18 strains engineered for high 2-PEAc production

The single and double ER18 replacement strains were also evaluated for the production of other flavor compounds in YP250GlulO% medium (Fig. 6). For none of the other acetate esters, ethyl, isobutyl and isoamyl acetate, was there any significant change in the production level compared with the parent ER18 strain. Hence, in YP250GlulO% medium, the mutations causing the increase in 2-PEAc production do not have a general effect on acetate ester production. On the other hand, for isobutanol, isoamyl alcohol and particularly phenylethanol, there appeared to be a similar increase as observed for phenylethyl acetate. The increase in phenylethanol, which also has a rose flavor, may strengthen the effect of the enhanced 2-PEAc level on total rose flavor impression. The higher increase in phenylethanol compared to isobutanol and isoamyl alcohol may explain why there was only a significant increase in phenylethyl acetate and not in isobutyl acetate and isoamyl acetate. The FAS2 allele caused a strong reduction in the production of ethyl hexanoate and ethyl octanoate, which may be due to reduced activity of Fas2 for synthesis of medium-chain fatty acids and/or indicate competition between the synthesis of these compounds and the synthesis of 2-PEAc. For acetaldehyde, there was no significant difference among the four strains.

The double replacement strain, containing TOR1E216 and FAS2BTC.1D, was also compared with the parent strain in wort fermentations (Fig. 7). This revealed a 60% increase in the level of 2-PEAc, confirming the potency of the superior alleles to enhance rose flavor production. As in YP medium, the level of isobutanol and isoamylalcohol, but not that of phenyl ethanol, was significantly increased. The double replacement strain also showed a significant increase in both isoamyl alcohol and isoamyl acetate, leaving the ratio unchanged, as opposed to YP medium where only isoamyl alcohol increased and not isoamyl acetate. Interestingly, in wort medium there was no significant drop in the production of ethyl hexanoate and ethyl octanoate, as opposed to YP medium.

Table 1. Occurrence of non-synonymous mutations in the proteins Tori and Fas2 in a set of 39 yeast strains of which the complete genome sequence is known.

Tori Fas2

58 133 216 396 547 1117 1640 2414 57 565 1136 1624 1800

BTC.1D D S E N N S F K T N T 1 N

ER18 G N * K S P V R A S A V S

S288c D S E N N s F K A s A V s

AWRI796 - P F T Y - - K - N A V N

BC187 D s E K S s F K T N A V N

BY4741 D s E N N s F K A S A V S

BY4742 D s E N N s F K A S A V S Tori Fas2

58 133 216 396 547 1117 1640 2414 57 565 1136 1624 1800

CBS7960 - - E K S s V R A S A V s

CEN.PK D S E K S s V R A S A V N

D273-10B G N E K s s V R A s A V N

DBVP6044 G N E K s s V R A N A V N

EC1118 D S E K s s V K A N A V N

EC9-8 G N E K s s - - T N A V N

FL100 D S E N N s F K A S A V S

FY1679 D S E N N s F K X N A V N

FostersB P P F T Y - - X A N A V N

FostersO - - E K S - - - A N A V N

JAY291 D S E K S s V R A S A V S

JK9-3d D s E K S s V R A S A V S

Kll G N E K S P V R A S A V S

Kyokai7 G N E K S P V R T N A V N

L1528 D S E K S s F K A N A V N

LalvinQA23 - - - - Y - - - A S A V N

RMll-la D S E K S s F K T N A V N

RedStar G N E K S s F R T N A V N

SEY6210 D S E K S s V R A S A V S

SKI G N E K S s V R A S A V N

UC5 G N E K S P V R A S A V S

VL3 D S E K s s - - A S A V S

Vinl3 D S E K s - - - A S A V S

W303 D S E N N s F K T N A V N

X2180-1A D S E N N s F K X N A - -

Y55 G N E K S s V R A S A V S

YJM269 G N E K S s V R A S A V s

YJM339 G N E K S P V R T N A V N Tori Fas2 58 133 216 396 547 1117 1640 2414 57 565 1136 1624 1800

YJM789 G N E K S P V R A S A V s

YJM499 D S E K s s V R T N A V N

YPS128 G N E K s s V R A S A V S

YPS163 G N E K s s V R A S A V N

YS9 P P F T Y - - R A S A V N

ZTW1 G N E K s s V R T N A V N

Table 2. S. cerevisiae strains used in this study.

Strain Description Reference

WLP575 Industrial ale yeast strain "White Labs Pitchable

Liquid Yeast" ( USA )

BTC.1D Haploid segregant from WLP575, MATct This study

ER18 Segregant of Ethanol Red (Industrial yeast strain used MCB, KU Leuven

for bioethanol production), MATa (21)

BTC.1D/ER18 Hybrid diploid strain obtained by crossing BTC.1D and This study

ER18

442 Haploid segregant from BTC.1D/ER18 with high 2-PEAc This study

production

442 TOR1A Haploid 442, torlA (marker removed) This study

442 MOG1A Haploid 442, moglA (marker removed) This study

442 BNA2A Haploid 442, bna2A (marker removed) This study

442 torl ^E216* Haploid 442, torl ^E216* point mutation inserted via This study

CRISPR/Cas9

442 ChrXVI. bl Haploid 442, block 1 exchanged by ER18 allele (ChrXVI, This study

79486.. 85523)

442 ChrXVI. b2 Haploid 442, block 2 exchanged by ER18 allele This study

(ChrXVI, 85504.. 91015)

442 ChrXVI. b5 Haploid 442, block 5 exchanged by ER18 allele (ChrXVI, This study

101093.. 104721)

442 ChrXVI. b6 Haploid 442, block 6 exchanged by ER18 allele (ChrXVI, This study

104702.. 108350)

442 ChrXVI. b7 Haploid 442, block 7 exchanged by ER18 allele (ChrXVI, This study

108331.. 114729)

442 ChrXVI. b8 Haploid 442, block 8 exchanged by ER18 allele (ChrXVI, This study

114710.. 119641) Strain Description Reference

442 ChrXVI.b9 Haploid 442, block 9 exchanged by ER18 allele (ChrXVI, This study

119622.. 125450)

442 ChrXVI.blO Haploid 442, block 10 exchanged by ER18 allele This study

(ChrXVI, 125432.. 129615)

BTC2 Haploid BTC.ID, torl ^E216* This study

BTC3 Haploid ER18, T0R1 ^E216 This study

BTC4 Haploid BTC.ID, /os2 ^ER18 This study

BTC5 Haploid ER18, FAS2 ^{BTC 1D} This study

BTC6 Haploid ER18, T0R1 ^*216E / FAS2 ^{BJC W} This study

Materials and methods

Microorganisms and cultivation media

Yeasts were grown in YPD medium [2% (w/v) glucose, 2% (w/v) peptone, 1% (w/v) yeast extract] shaking at 200 rpm and at 30 °C. For solid nutrient plates, 1.5% (w/v) Bacto agar was added. The S. cerevisiae yeast strains used in this study are listed in Table 2. Escherichia coli cells (DH5, Invitrogen) were grown at 37 °C in Luria Broth (LB) medium containing 0.5% (w/v) yeast extract, 1% (w/v) Bacto tryptone, and 1% (w/v) sodium chloride (pH 7.5). For solid nutrient plates, 1.5% (w/v) Bacto agar was added. Selection of transformants was performed in the presence of 100 mg/mL ampicillin.

Flavor compound screening was performed in YP250GlulO% (0.27% yeast extract (Merck), 0.54% bacto peptone (Oxoid) to a total predicted nitrogen content of 250 mg/L, which is in the same range as beer and wine fermentations, and 10% (w/v) glucose. The predicted nitrogen content was based on information of titratable nitrogen from the suppliers. For confirmation, unhopped wort was used [17 "Plato, 18.8% (w/v) malt extract, 2.068 ppm ZnSO ₄ .7H ₂ 0].

Fermentation experiments

Yeast cells were first cultivated overnight in 3 mL of YPD2% in a shaking incubator at 30 °C, 200 rpm. This pre-culture was used to inoculate 100 mL of YP250GlulO% (w/v) with an initial OD600 nm of 0.1. Fermentation was carried out at 25°C in 100 ml fermentation tubes, previously validated against European Brewing convention (EBC) tall tubes, fitted with a water lock in order to create semi-anaerobic conditions, mimicking large scale fermentations. Agitation was performed with a magnetic rod at 130 rpm. Fermentation progress was monitored by weight loss due to CO2 release. Samples were taken at the end of the fermentations for headspace GC-FID analysis. Wort fermentations were performed at 20°C and yeast was inoculated to an initial OD600 nm of 2.0.

Headspace GC-FID analysis

Headspace gas chromatography coupled with flame ionization detection (GC-FID) was used to measure flavor compounds at the end of the fermentation. Samples were collected and centrifuged at 3500 rpm for 5 min. Then, 2 ml of the supernatant was collected in 25-ml vials and analyzed using a gas chromatograph with a headspace sampler (HS40; PerkinElmer Life Sciences). The headspace was equilibrated by shaking and incubating for 10 min at 60 °C using a Thermoscience RS Plus auto sampler and then injected into a polyethylene glycol column (Restek Stabilwax, 60m x 0.25mm x 0.25μιτι). Injection block and flame ionization detector temperatures were kept constant at 220 and 250 °C, respectively. Oven temperature was kept at 40 °C for 2 min, then increased to 240 °C at a rate of 15 °C/min. Helium was used as carrier gas at a flow rate of 2.0 ml/s. GC operating conditions were used as follows: injection volume lmL; split rate 1:25; split flow 50mL/min.

Mating type, sporulation and tetrad dissection

Standard procedures were used for sporulation and tetrad dissection (79) and for mating type determination by PCR with primers for the MAT locus and MATa and MATa DNA (80).

Molecular Biology methods

Yeast cells were transformed by electroporation (81). Standard molecular biology protocols were used in this work.

Genomic DNA extraction and whole genome sequence analysis

Segregants were grown separately in 3ml YPD and pooled together by OD600 nm with the purpose of getting approximately equal amounts of DNA for each strain. Genomic DNA was extracted and purified with the "Masterpure Yeast DNA purification kit" from Epicentre in order to obtain high quality genomic DNA. At least 10 μg of DNA per pool was provided to Beijing Genomics Institute (BGI, Hong Kong, China) for whole genome sequence analysis, which was performed with the lllumina platform (HiSeq2000). Assembly and mapping were done with NGSEP (Next Generation Sequencing Eclipse Plugin) (82). Significance tests used to determine quantitative trait loci were carried out as described by Claesen et al. (83).

Allele-specific PCR

Individual SNPs were scored by allele-specific PCR performed with two sets of primers that differed only at the 3' terminus, containing either the BTC.1D or the ER18 nucleotide. Both primers were always applied in separate PCR reactions. The optimal annealing temperature was determined by gradient-PCR using DNA of BTC.1D and ER18 parents. The optimal temperature is the annealing temperature at which only hybridization with primers containing an exact match was observed. The SNP data of the individual segregants were analyzed using the binomial distribution probability.

CRISPR/Cas9 technology

Cas9 plasmid.

The CRISPR-Cas9 technology was applied by using the pTEF-Cas9-KanMX plasmid (CrispR-mediated gene inactivation or replacement, single copy, Cas9 behind TEF-promotor, KanMX marker) which was derived from the p414-TEFlp-Cas9-CYClt plasmid (27). The pTEF-Cas9-KanMX plasmid was created by introducing the KanMX4 marker into Kpnl-digested p414-TEFlp-Cas9-CYClt.

gRNA plasmids.

For each target, specific gRNAs were designed, for lx, 2x or 3xgRNA approach. Specific gRNAs were assembled by using the Gibson Assembly Kit into the Xhol-EcoRV-digested P58 vector. P58 was derived from p426-SNR52p-gRNA.CANl.Y-SUP4t (27). It was created by assembling the HPH marker and the universal CRISPR gRNA cloning site into the p426-SNR52p-gRNA.CANl.Y-SUP4t backbone.

lx gRNA approach - introduction of point mutation

A specific gRNA without PAM-site and flanked by the flanking regions 1 and 2 was designed.

Flanking region 1 (promotor): GCAGTGAAAGATAAATGATC (SEQ ID No 9)

Flanking region 2 (terminator): GTTTTAGAGCTAGAAATAG (SEQ ID No 10)

Forward and reverse oligomers were hybridized and assembled into the Xhol-EcoRV-digested P58 vector with the Gibson Assembly Kit. For hybridization, oligomers were dissolved in STE-buffer (10 mM Tris pH 8.0, 50 mM NaCI, 1 mM EDTA) at a concentration of 500 μΜ. Equimolar concentrations of the forward and reverse primers, were combined and heated at 94°C (4 °C/min) for 3 min and slowly cooled down (to 75 °C at 0.05 °C/min, and then cooled to 10°C at 0.02 °C/min).

pJET-2xgRNA

The pJET-2xgRNA plasmid was derived from pJETl,2-blunt (ThermoScientific) and contains a module consisting of the gRNA terminator and promotor. The amplification of this module, flanked by gRNA sequences, allows the addition of more gRNAs in the regular gRNA plasmids.

2xgRNA approach - allele swapping in haploids

A specific gRNA without PAM-site and flanked by the flanking regions 1 and 2 was designed, one before the first SNP of interest and the other after the last SNP of interest. Forward gRNAl and reverse gRNA2 oligomers were used to amplify the module from pJET-2xgRNA. The amplification of this module flanked with gRNA sequences, allows the addition of more gRNAs in the regular gRNA plasmids. The amplification product (505bp) was assembled into the Xhol-EcoRV-digested P58 vector using the Gibson Assembly Kit. Design guide RNA targets

Proper gRNA targets were selected based on the presence of natural SNPs between inferior and superior allele at the protospacer adjacent motif (PAM) sequence or the 8 bp of DNA preceding the PAM sequence, described as the most important for specificity (84, 85). Next, we analyzed potential off- targets in the genome, which are regions matching the gRNA sequence with 3 or less SNPs followed by a valid PAM-site. Donor DNA

Introducing a point mutation in TORI

Donor DNA was a 71bp-oligomer containing the desired mutation and 35bp identical sequences on each side: TAAAGTCTTGCCTAGAATGGCTTACTGCCTCCACGTAAAAGAATTCATTCTCAAGTTCGA AGCCAGACCAT (SEQ ID No ll)

The forward and reverse sequences were ordered and then hybridized.

Allele swapping

Donor DNA was PCR-amplified from the BTC.1D parent, which contains the superior allele on chromosome XVI. Donor DNA contained 500-800 bp of flanking sequence similarity.

CRISPR/Cas9 application

Point mutations were introduced with the CRISPR/Cas9 technology, using a single gRNA. Bulk replacements on chromosome XVI were performed by using the 2x gRNA approach. In haploid cells the lx or 2xgRNA approach was used. Cells were first transformed with 250 ng of pTEF-Cas9-KanMX plasmid by electroporation. Cas9-transformants were pre-grown in 3mL YPD-geneticin (200 μg/ml), 30 °C, overnight. This culture was used to inoculate 50ml YPD-geneticin to an initial OD 0.2. Cells were transformed by electroporation using 500 ng of the gRNA plasmid and 1000 ng of donor DNA (or 2 μΙ of 250 μΜ duplexed oligomers, for the point mutation). Transformants were selected on solid YPD2% + 200 μg/ml geneticin + 300 μg/ml hygromycin B. Sequences

SEQ ID No 1: wild-type FAS2 allele (present in inferior ER18)

SEQ ID No 2: mutant FAS 2 allele (present in superior BTC.1D)

SEQ ID No 3: wild-type FAS2 protein sequence (present in inferior ER18) SEQ ID No 4: mutant FAS2 protein sequence (present in superior BTC.1D) SEQ ID No 5: wild-type TORI allele (present in superior BTC.1D) SEQ ID No 6: mutant TORI allele (present in inferior ER18) SEQ ID No 7: mutant TORI protein sequence (present in inferior ER18) SEQ ID No 8: engineered TORI protein sequence References

1. Lambrechts MG, Pretorius IS. 2000. Yeast and its Importance to Wine Aroma - A Review. S Afri J Enol Vitic 21:97-129.

2. Swiegers JH, Bartowsky EJ, Henschke PA, Pretorius IS. 2005. Yeast and bacterial modulation of wine aroma and flavour. Australian Journal of Grape and Wine Research 11:139-173.

3. Saerens SM, Delvaux FR, Verstrepen KJ, Thevelein JM. 2010. Production and biological function of volatile esters in Saccharomyces cerevisiae. Microb Biotechnol 3:165-177.

4. Verstrepen KJ, Derdelinckx G, Dufour JP, Winderickx J, Thevelein JM, Pretorius IS, Delvaux FR.

2003. Flavor-active esters: adding fruitiness to beer. J Biosci Bioeng 96:110-118.

5. Peddie HAB. 1990. Ester formation in brewery fermentations. Journal of the Institute of Brewing

96:327-331.

6. Cordente AG, Curtin CD, Varela C, Pretorius IS. 2012. Flavour-active wine yeasts. Applied Microbiology and Biotechnology 96.

7. Meilgaard MC. 1975. Flavor chemistry of beer, part ii: flavour and threshold of 239 aroma volatiles.

Tech Q Master Brew Assoc Am 12:151-168.

8. Verstrepen KJ, Van Laere SDM, Vanderhaegen BMP, Derdelinckx G, Dufour J-P, Pretorius IS, Winderickx J, Thevelein JM, Delvaux FR. 2003. Expression levels of the yeast alcohol acetyltransferase genes ATF1, Lg-ATFl, and ATF2 control the formation of a broad range of volatile esters. Applied and Environmental Microbiology 69:5228-5237.

9. Fujii T, Nagasawa N, Iwamatsu A, Bogaki T, Tamai Y, Hamachi M. 1994. Molecular cloning, sequence analysis, and expression of the yeast alcohol acetyltransferase gene. Appl Environ Microbiol 60:2786-2792.

10. Saerens SM, Verstrepen KJ, Van Laere SD, Voet AR, Van Dijck P, Delvaux FR, Thevelein JM. 2006.

The Saccharomyces cerevisiae EHTl and EEBl genes encode novel enzymes with medium-chain fatty acid ethyl ester synthesis and hydrolysis capacity. J Biol Chem 281:4446-4456.

11. Mason AB, Dufour JP. 2000. Alcohol acetyltransferases and the significance of ester synthesis in yeast. Yeast 16:1287-1298.

12. Lilly M, Lambrechts MG, Pretorius IS. 2000. Effect of increased yeast alcohol acetyltransferase activity on flavor profiles of wine and distillates. Appl Environ Microbiol 66:744-753.

27. DiCarlo JE, Norville JE, Mali P, Rios X, Aach J, Church GM. 2013. Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems. Nucleic Acids Research 41:4336-4343.

46. Noda T, Ohsumi Y. 1998. Tor, a phosphatidylinositol kinase homologue, controls autophagy in yeast.

J Biol Chem 273:3963-3966. 47. Hardwick JS, Kuruvilla FG, Tong JK, Shamji AF, Schreiber SL. 1999. apamycin-modulated transcription defines the subset of nutrient-sensitive signaling pathways directly controlled by the Tor proteins. Proc Natl Acad Sci U S A 96:14866-14870.

49. Loewith R, Hall MN. 2011. Target of rapamycin (TOR) in nutrient signaling and growth control.

Genetics 189:1177-1201.

50. Conrad M, Schothorst J, Kankipati HN, Van Zeebroeck G, Rubio-Texeira M, Thevelein JM. 2014.

Nutrient sensing and signaling in the yeast Saccharomyces cerevisiae. FEMS Microbiol Rev 38:254- 299.

53. Mohamed AH, Chirala SS, Mody NH, Huang WY, Wakil SJ. 1988. Primary structure of the multifunctional alpha subunit protein of yeast fatty acid synthase derived from FAS2 gene sequence.

J Biol Chem 263:12315-12325.

54. Schweizer E, Hofmann J. 2004. Microbial type I fatty acid synthases (FAS): major players in a network of cellular FAS systems. Microbiol Mol Biol Rev 68:501-517, table of contents.

79. Sherman F, Hicks J. 1991. Micromanipulation and dissection of asci. Methods Enzymol 194:21-37. 80. Huxley C, Green ED, Dunham 1. 1990. Rapid assessment of S. cerevisiae mating type by PCR. Trends Genet 6:236.

81. Thompson JR, Register E, Curotto J, Kurtz M, Kelly R. 1998. An improved protocol for the preparation of yeast cells for transformation by electroporation. Yeast 14:565-571.

82. Duitama J, Quintero JC, Cruz DF, Quintero C, Hubmann G, Foulquie-Moreno MR, Verstrepen KJ, Thevelein JM, Tohme J. 2014. An integrated framework for discovery and genotyping of genomic variants from high-throughput sequencing experiments. Nucleic Acids Res 42:e44.

83. Claesen J, Clement L, Shkedy Z, Foulquie-Moreno MR, Burzykowski T. 2013. Simultaneous mapping of multiple gene loci with pooled segregants. PLoS ONE 8:e55133.

84. Hsu PD, Lander ES, Zhang F. 2014. Development and applications of CRISPR-Cas9 for genome engineering. Cell 157:1262-1278.

85. Fu Y, Foden JA, Khayter C, Maeder ML, Reyon D, Joung JK, Sander JD. 2013. High-frequency off- target mutagenesis induced by CRISPR-Cas nucleases in human cells. Nat Biotechnol 31:822-826.

86. Lomakin IB, Xiong Y, Steitz TA. 2007. The crystal structure of yeast fatty acid synthase, a cellular machine with eight active sites working together. Cell 129:319-332.