Specifically,the invention has made it possible to make a composite yeast fermented beverage product such as lager having a predetermined content of one or more fla- vourcompound (s) and/or compound (s) that stabilize flavour compounds e. g. by pro- vidingat least batch of fermented beverage which is produced by using a yeast strain whichis modified to substantially not produce at least flavour compound and/or com- pound stabilizing flavour compounds (base batch of fermentate) combining such a batch with one or more batches of beverage fermentates that contain (s) a level flavour compounds and/or stabilizing which is normal a fermentate produced by using a conventional yeast strain or which is produced by a modified yeast strain having, relativeto conventionally yeast strains, an enhanced production of at least fla- vourcompound or flavour stabilizing compound.

TECHNICALAND PRIOR ART The sensory quality of a yeast fermented beverage such as beer depends largely on the particularbrewing yeast strain which is used. During the beer brewing process, the yeast will,addition to ethanol, a large of minor metabolites affect the taste and flavour the beer. Whereas the presence of certain of such minor metabolites may, at least to certain levels, conferdesirable characteristics to the beer,

several minormetabolites undesired and may even, if present in amounts, re- sult spoilage of the beer.

Yeast fermented beverages including beer contain more than a hundred minor metabo- lites, which some, at least when they are present above certain concentrations, will confer to the beverage an undesirable taste and flavour. comprehensive review of beer flavouringcompounds has been given by Meilgaard to which there is referred. In this connection, the most important metabolites include sulphur such as di- methyl (DMS), hydrogen sulphide (H2S), thiols thioesters which at certain levels will result undesirable sulphur-containing flavours.

DMS is a thioether of great importance for the aroma and flavour of beer. The content of DMS in conventional lager beersregularly the taste threshold level about 30 pg/L (Meilgaard, which explains the focus that has been on this compound. Above its threshold level butbelow 100 ug/L contributes to the distinctive taste of some lager When present at concentrations above 100 pg/L may, however, impart a generally undesiredflavour as"cooked sweet corn". DMS in beer may be derived from thermal degradation of S-methyl-methionine duringkiln-drying wort preparation and it has been suggested that this is the only pathway of significance for the finalDMS content in beer (Dickenson and Anderson, 1981 ; Dickenson, 1983). However, substantial evidence suggests that enzymatic conversion of dimethyl sulphoxide (DMSO) to DMS by the brewing yeast is of great importance and, under some circumstances, even the major source of the final DMS level beer (Leemans et al., 1993).

There is no doubt, that Saccharomyces do contain an enzymatic activity that can reduce DMSO to DMS in a NADPH-dependent manner (Zinder and Brock, 1978a ; Anness et al. , ; Anness, 1980). A multicomponent methioninesulphoxide reductase (EC 1. 8. 4. 5) has been isolated from yeast (Black etal. , ; Porque etal. , and it has been suggested that this activity is identical the DMSO reductase activity (Anness et al. , ; Anness, 1980 ; Bamforth, 1980 ; Bamforth and Anness, 1979 ; 1981). The nor- malof MetSO reductase seems to be reduction of oxidized methionines of cellu- lar which is in agreement with observations showing that the enzyme has a higher affinity for MetSO than for DMSO (Bamforth and Anness, 1979 ; 1981) and that MetSO inhibits DMSO reduction (Anness et al. , ; Anness, 1980 ; Bamforth and An- ness, 1981). The consequence hereof is that the degree of MetSO formation during kiln- dryingof the malt will the degree of DMSO reduction.

The nitrogen content of the growth medium also to affect DMS formation yeast.

Thus, high amounts of highly assimilable keep DMSO activity at a low level, enzyme activity is induced under nitrogen-limiting conditions (Gibson et al., 1985). The high nitrogen content of most worts would appear to keep DMSO reduction at the base level fermentation. See Anness and Bamforth (1982) for a review on DMS formation in beer production. So-called DMSO reductases some prokaryotes have been characterized, as well the genes encoding them (Satoh and Kurihara, 1987 ; Bi- lous etal. , ; Weiner et al. , ; Yamamoto et al. , 1995).While enzymes, in some cases, have MetSO reducing capabilities, their main purpose in the bacterial cell is probably to act as the terminal in DMSO respiration (Zinder and Brock, 1978b). Pep- tide methionine sulphoxide reductases(PMSRs) probablyfulfil therole repairing oxi- dised methionine species in proteins, thereby restoring their biological activity.

Sulphite is another sulphur-containing metabolite formed during yeast fermentation.

Sulphite is a versatile food additive used for preservation of foodstuffs. In beersulphite has a dual in that it acts both as an antioxidant and as an agent that masks certain off-flavours. to its great importance, considerable work has been carried out to eluci- date the physiology sulphur metabolism in Saccharomyces in relation to sul- phite production.

Sulphite produced by the yeast during beer fermentation and is present in the final beer.

The compound is produced via the sulphur assimilation pathway. Inorganic sulphate is taken up through two sulphate permeases encoded by SUL1 andSUL2, respectively (Cherest et al. , 1997).Intracellularly, is converted to adenylylsulphate by the action of ATP sulphurylase 2. 7. 7. 4) encoded by MET3. In subsequent step, the MET14 APS kinase (EC 2. 7. 1. 25) catalyses the formation of phosphoadenylyl- sulphate which in turn is reduced by PAPS reductase (EC 1. 8. 99. 4) (MET16 en- coded) to sulphite.

Hydrogen sulphide from reduction of sulphite in this form inorganic sulphur is incorporated into organic compounds by fusion with O-acetyl homoserineleading the formation of homocysteine. The latter is the precursor for biosynthesis of cys- teine, methionine and S-adenosylmethionine SAM represses transcriptionally all MET-genes (Cherest et al. , ; Sangsoda et al. , ; Langin et al. , ; Thomas et al., ; Thomas et al. , ; Korch etal. , ; Mountain et al. , ; Hansen et al., 1994).

Hydrogen sulphide as it mentioned above, an intermediate in the sulphur assimila- torypathway of Saccharomyces spp.It isalso point of entry into the methionine bio- synthetic pathway of carbon backbones derived from the threonine biosynthetic pathway.

Itderived in four enzymatic steps from inorganic sulphate ions obtained from the growth medium of the yeast.

Besides being an intermediate in the yeast sulphur hydrogen sulphide is an importantflavour compound in beer and its distinct taste of"putrefied eggs"generally ren- ders this compound undesired in beers, except at very low where it may aid in disguising the taste of other flavour compounds such as acetyl esters.

Several have attempted to control the formation of hydrogen sulphide in beer productionby modifying yeast strains by recombinant DNA technology. Thus, overexpres- sion of the MET25 gene encoding the enzyme (EC 4. 2. 1. 22) that catalyses the condensa- tionof hydrogen sulphide and O-acetyl homoserineleads a reduced hydrogen sulphide concentration in beer (Omura et al. , When the gene NSH5 (STR4, encoding the first of two steps (EC 4. 2. 1. 22) in formation of cysteine from homocysteine is overex- pressed, the hydrogen sulphide level also reduced (Tezuka et al. , This effect was suggested by these authors to be due to faster removal of homocysteine and thus of its precursor, hydrogen sulphide. These experiments suggest the existence of metabolic bottlenecks the conversion of hydrogen sulphide into organic sulphur compounds.

Inliterature, there have been speculations hydrogen sulphide may be a precursor forcertain thiols thioesters, e. g. methane-and ethanethiol (MeSH andEtSH) and methyl-and ethylthioacetate and EtSAc). Thus, in the above experiment with NSH5 et al. , the decrease in hydrogen sulphide production was followed by a decrease in methanethiol andethanethioi production.

Anothergroup of flavour compounds which, when present above certain threshold levels, result in undesirable flavours higher alcohols esters such as ethyl acetate, isoamyl and ethyl Above their threshold levels such esters will confer a fruity (apple, banana)flavour beer.

Avariety of higher alcohols is found in beer. As used herein"higher alcohols"indicate other alcohols thanethanol. higher alcohols are also to in the art as fu- sel alcohols are listed inMeilgaard and include isoamyl alcohol.

Some of the higher alcohols produced by metabolism the branched chain amino acids isoleucine, valine andleucine. a-Ketoacids(a-keto-P-methylvalerate, a-ketoiso- valerate a-ketoisocaproate) are important intermediates for branched-chain amino acids as well for higher alcohols. Isoamyl alcohol its corresponding acetate ester are among the distinct beer flavour components.

Three enzymatic steps encoded by the genes LEU4 (Baichwal etal. 1983),LEU1 and LEU2 involved in de novo of a-keto-isocaproate. The enzymes are a- isopropylmalate (alPM) synthase(LEU4p, 4. 1. 3. 12), isopropylmalate isomerase (Leu1p, 4. 2. 1. 33) and ß-isopropylmaiate dehydrogenase EC 1. 1. 1. 85) (Re- viewed by Kohlhaw, The pathway is regulated by a complex of the regulatory protein Leu3p and isopropylmalate. complex acts as an activator and regulates the level ofmRNA produced by binding to a regulatory region in front of the LEU4, LEU1 andLEU2 regions. Conversion of a-ketoisocaproate to leucine is catalyzed by the transaminases encoded by the genes BAT1 and BAT2. of both BAT1 and BAT2 results auxotrophy for the branched-chain amino acids isoleucine, va- line andleucine (Kispal etal. 1996).

In yeast cell, theleucine level isregulated several ways, of which feed-back inhibi- tionof the isopropylmalate by leucine one. It possible to inactivate the feed- back inhibition of LEU4p. feed-back inhibition resistant mutants with a dramati- cally lower for leucine been isolated. The mutants are resistant to the leu- cine analogue 5, 5-trifluoro-DL-leucine (TFL).LEU4 was originally as a mutant <BR> <BR> <BR> <BR> <BR> resistant to TFL (Baichwal etal. Such dominant feed-back resistant (LEU4fbr) mu- tations are known to produce high amounts of isoamyl alcohol andisoamyl in brewers yeast (Lee et al. 1995).

Previous studies have shown that at least threeisopropylmalate are present in the yeast cell (Baichwal etal. Two synthases are produced by LEU4 of the wild-typeactivity). The two forms produced by LEU4 a long form(designated la) that is exported to the mitochondria and a short form designated Ib in the cyto- plasm. The gene coding for the remaining activity was originally designatedLEU5, but later investigations have shown LEU5 be a gene with PET gene similarities (petites : unable grow on non-fermentable sources) (Drain and Schimmel, and not directly involved the leucine synthesispathway. Inactivation the LEU4 product

does not in itself lead toleucine auxotrophy(Baichwal etal. The LEU4 (leaky) phe- notype might be due to other synthase activities. Three other loci (LEU6, LEU7 andLEU8) responsible for leucine have been described, where LEU7 andLEU8 appear to be candidates for a gene or genes that encode an a-IPM synthase (Drain and Schimmel, Linkage to known open reading frames has not yet been established.

During the Saccharomyces cerevisiae project (http ://genome- www.stanford. edu/Saccharomyces/), a putative LEU4 homologue been identified. The open reading frame (ORF) is designated YOR108Wand located on chromosome XV.

The ORF YOR108Wis homologous toLEU4 more than 80% nucleotide in a contiguous sequence of 1806 nucleotides 1860 nucleotides in the LEU4 ORF, whereas the upstream regulatory regions are clearly There appears to be a pos- sibilityfor translation of a long a short form similar to LEU4. of the YOR108WORF tworegulatory and Leu3p binding sequences seem to be present.

An a-ketoisocaproate decarboxylase encoded by the ORF YDL080C been described for S. cerevisiae et al. , This ORF is most likely same as THI3 et al. , The enzyme appears to catalyze the conversion of a- ketoisocaproate to isoamyl alcohol. In ayd/080 a minor activity (40% activity relative to that of the wild appears to be present.

Two alcohol acetyltransferase genesdesignated ATF1 (ORFYOR377Won chromosome XV) and ATF2 (ORFYGR177C chromosome VII) been identified in S. cerevisiae.

Alcohol acetyltransferases (AATases 2. 3. 1. 84) catalyze transfer of the acetyl group from acetyl-CoA toalcohols producing acetate esters. MTase1 encoded by the ATF1 has been purified (Malcorps &Dufour, and Minetoki etal., 1993), while AATase2 isonly as an open reading frame, ATF2 et al., 1996).

MTase1 andAATase2 about 36% amino acid sequence identity. The ATF1 gene has been disrupted (Fujii et al. This mutant produced about 20% isoamyl acetate and about 60% ethyl as compared to the original strain (Fujii et al. 1996b).

The allotetraploid lager yeast,Saccharomyces carlsbergensis, at least different genomic sets. One genomic set is similar that of S. cerevisiae another geno- mic set is similar that of S. monacensis 1986). Allotetraploid lager yeast thus has two copies of a S. cerevisiae-like allele of ATF1 herein as ATF1- CE), two copies of a S. carlsbergensis specificallele herein as ATF1-

CA), presumably originating S. monacensis. Fujii etal. (1996a)cloned two genes ATF1-CE and ATF1-CA (previously designatedATF1 andLg-ATF1, respectively) and determined the nucleotide The amino acid sequences encoded by the two genes are 76% identical. The ATF2-CE from the bottom fermenting yeast Saccha- romyces pastorianus has been cloned and sequenced (Yoshimoto et al. , 1996b). It is not known whether two forms of this gene are found in the S. carlsbergensis lager- yeast. Fermentation of wort with a brewers yeast transformed with a 2p-based plasmid containing the ATF1-CEgene in increased levels ofisoamyl (7. 6-fold) and ethyl (3. 9-fold) the final beer compared to the control et al., 1993).

Anormal may also be associated with various ketones of which diacetyl the most important and with aldehydes such as acetaldehyde, precursor for ethanol, and so-called staling aldehydes which the most important is trans-2-nonenal confer to beer a highly undesired"cardboard"flavour.

Currently, the brewing attempts to control the sensory quality of beer such as la- ger by several measures, including of batches having an undesired content of one or several metabolites batches having a lower content hereof. However, when using conventional of brewer's yeast, it may not be possible to obtain a fully ac- ceptable blend beer by this approach. An alternative is to select, using classical and mutagenization techniques, yeast strains which, relative to the conventionally strains, have a lower production of an undesired metabolite.

The most widely used yeast for brewing is Saccharomyces However, not all Saccha- romyces spp. are suitable for brewing purposes. Typical faults non-brewing Saccharo- myces strains include of phenolic off-flavour, inability utilize maltotriose and low rate at the temperature optimal for the desired aroma of the beer.

In industry, the characteristics of the existing brewing yeasts may be improved by se- lecting strains which e. g. have a higher rate of fermentation, a decrease in beer matura- tion time, better flocculation or increased tolerance to alcohol.

There are, however, two major factors that currently limit in the breeding of brewing yeasts. First, it is often complicated translate the desired change in yeast per- formance into biochemical and genetic terms. The second difficulty that brewing yeasts generally have deficiencies in their sexual and as a consequence hereof it is

difficultcarry out many of the breeding steps and procedures in analysis that are trivial the non-brewing yeasts used as genetic reference strains in academic studies.

Althoughdesignation Saccharomyces carlsbergensis iswidely this common lager is also referred to as S. pastorianus S. uvarum. It hasa- complicated structure, being allotetraploid one chromosome set similar to the S. cerevisiae genomewhile other set is structurally similar that of S. monacensis.

Therefore, it is expected to find two copies of a S. cerevisiae-like allele two copies of a S. monacensis-like allele a certain gene in S. carlsbergensis. the S. cer- evisiae-like allele specified by the gene name followed (e. g. MET2-CE). Simi- larly,S. carlsbergensis-specific (S. monacensis-like) allele designated by the gene name followed by-CA. It generally assumed that this"mixed"genome may have an ef- fect on the characteristics of the lager yeast. Strain improvement by classical ge- netic methods is not straightforward, as most brewing yeasts exhibit poor sporulation and lowsporeviability.

It therefore currently a significant problem in the brewing industry that it is impossible or difficultto provide a range of variants of the same beverage type such as lager which are specifically adapted to consumer preferences in individual marketplaces to different seasons of the year by having specific and predetermined contents of compounds deter- miningthe sensory characteristics of the beverages.

Methods to make fermented beverages having a modified content of aroma or flavour compounds have been developed. One approach which has been used to modify bever- ages is to add isolated and/or aroma or flavour derived from a yeast fermentation process and using such compounds as flavour to beverages.

Thus, as an example, is suggested in US 3, 713, 838 to produce food products and bev- erages to which are added flavour compoundsisolated yeast dregs from a conven- tionalbrewing process. In 96/39480 is disclosed use of kettle hop extracts to pro- vide a fully flavoured beverage Another approach is to provide yeast strains which have been modified to produce an in- creased level particular flavour Thus, EP 574 941-A1 disclosesSaccha- romyces cerevisiae transformed with a plasmid-borne coding for alcohol acetyltransferase (AATase). transformed strains having multiple copies of this

AATase-encoding ATF1 showed an enhanced production of isoamyl and ethyl acetate.

In 94/08019 is disclosed microbial which are transformed with a gene coding for a-acetolactate decarboxylase, enzyme capable converting the diacetyl precursor, a-acetolactate, to acetoin, thus avoiding the formation of diacetyl.

JP 62-92577 and 30-07579 disclose mutants of Saccharomyces cerevisiae have en- hanced production of amyl alcohol andisobutyl alcohol, JP 50-49465 discloses yeast mutant strains producing high amounts of ß-phenethylene alcohol esters thereof.

There have been very few attempts to provide brewing yeast strains including S. carlsber- gensis a reduced production of flavour Thus, it has been attempted to control the production of diacetyl byblocking theacetolactate activity in Saccha- romyces cerevisiae mutation or in Saccharomyces carlsbergensis in vitro deletion and replacement recombination of ILV2 with the mutant alleles. com- pletely blocking this biosynthetic step in Saccharomyces carlsbergensis not yielded brewing strains with satisfactory characteristics, presumably because the parental strains herefor do not take up isoleucine andvaline effectively as Saccharomyces cerevisiae (Kielland-Brandt etal. 1995).

Thus, the prior art is not aware of industrially strains of brewer's yeast in which pathways for production of undesired metabolites been substantially completely in- terrupted or of strains of the lager yeast Saccharomyces carlsbergensis has been modified to produce in the brewing process none or less of a flavour compound and/or a compound that stabilizes a flavour compound.

It now been found that it is possible provide modified Saccharomyces brewer's yeast strains including modified strains of Saccharomyces carlsbergensis which one or several biosynthetic pathways leading to the formation of flavour or com- pounds stabilizing such compounds have been substantially completely and that such modified strains have retained the beverage fermentation capacity and efficiency of the parent strains.

This has provided industrially means for producing yeast fermented beverages including lager which substantially not contain one or more flavour compounds or flavour stabilizing compounds that is/are normally by the lager yeaststrain. In turn, this achievement has made it possible to produce, in an industrial production scale, a fermented base beverage in which one or more flavour-related metabolite normally present in a particular beverage type is/are absent or present at very low levels. Such-a base beverage can in turn be"flavour customized"by blending it with batches of beverage having a normal enhanced content of the respective flavour compounds or by adding isolated compounds to provide a composite beer having a desired, predetermined flavour compound profile.

SUMMARY OF THE INVENTION Accordingly, the present invention relates a first aspect to a method of preparing a composite yeast fermented beverage having a predetermined content of flavour com- pounds or compounds stabilizing the flavour said beverage, the method comprising (i) producing separate batches of fermented beverage using a different strain of yeast for each batch, at least of said batches being a fermented base beverage produced with a yeast strain which is modified by mutation or recombination to have, relative to the par- ent strain from which it is derived, a reduced or lacking production of at least one flavour compound or of at least one compound that is capable stabilizing the flavour of said composite yeast fermented beverage by forming adducts with staling present therein, (ii) optionally followed at least partially one or more of said sepa- rate batches or isolating therefrom a flavour stabilizing compound, and (iii) combining said batches and/or compounds to obtain the composite beverage.

In useful of the above method the composite beverage comprises at least one batch of beverage that is produced by using a yeast strain which is not mutationally or recombinationally in respect of production of flavouring or flavour stabilizing compounds and/or at least batch of fermented beverage that is produced by using a yeast strain which is modified by mutation or recombination to have, relative to the parent strain from which it is derived, an enhanced production of a flavour compound or of a compound that is capable stabilizing the flavour of the composite yeast fermented bev- erage by forming adducts with staling present therein.

In aspects, the invention pertains to a composite beverage obtainable by the above method and to the use of a batch of yeast fermented beverage produced by a modified yeast having an enhanced flavour production, or a flavour compound isolated from such a batch, as a flavouring agent.

In stillaspects, there are provided a genetically modified yeast strain which, rela- tiveto its parent strain, has a reduced or lacking production of sulphite production of a fermented beverage and a method of producing a yeast fermented beverage using such a modified yeast strain, a genetically yeast strain which, relative to its par- ent strain, has a reduced or lacking capability of converting dimethyl sulphoxide di- methyland a method of producing a fermented beverage using such a modified strain.

In aspect there is provided a novel non-recombinantly modifiedSaccharomyces carlsbergensis strain having, beer fermentation conditions, relative to its par- ent strain, an altered production of at least of hydrogen sulphide, athiol a thioester, and a method for producing a lager using such strain.

In stillaspect, the invention pertains to a genetically modifiedSaccharomyces carlsbergensis strain which, relative its parent strain, has a reduced or lacking production under beer fermentation conditions of a fusel alcohol an ester thereof, and to a method of producing a lager comprising fermenting a brewer's wort by such a modified yeast strain.

DETAILED DISCLOSURE OFTHE INVENTION One major objective of the present invention is to provide a method of preparing a com- posite yeast fermented beverage by combining separate batches of fermented beverage ofwhich at least is produced by a yeast strain which is genetically by muta- tionand/or recombination to have, relative its parent strain, a reduced or lacking pro- duction of at least oneflavour or flavour stabilizing compound.

Another objective is to provide a composite beverage having a predetermined content of one or more flavour (s) and/or compounds that stabilizes theflavour the bev- erage.

It will appreciated that the invention also the means of providing novel fer- mented beverages which are based on the production of a single batch using a single yeast strain according to the invention which is genetically modified by mutation and/or recombination to have, relative to its parent strain, a reduced or lacking of at least one flavour or flavour stabilizing compound.

In aparticularly embodiment of the invention, the composite yeast fermented beverage is a beer such as a lager beer.Almost all production world-wide carried out with pure cultures, i. e. single cell of yeast belonging to the genus Saccharo- myces. Beer can also be made with yeast which are not Saccharomyces such as e. g.

Schizo saccharomyces Among Saccharomyces brewing yeast, S. carlsber- gensis described above is a commonly used strain for production of lager beer.

However, the invention is also useful the manufacturing of any other type of yeast fer- mented beverage, of which notable include wine, sake and cider. It within the scope of the invention to provide either separate batches of such beverages having a modified content of flavouring which are manufactured by means of a modi- fied yeast strain according to the invention, or a composite beverage being produced by combining two or more separate batches of the beverage, each of which is made with a different yeast strain, at least one of said batches being a fermented base beverage pro- duced with a yeast strain which is modified by mutation or recombination to have, relative to the parent strain from which it is derived, a reduced or lacking production of at least one flavour compound or of at least one compound that is capable of stabilizing the flavour of said composite yeast fermented beverage by forming adducts with staling compounds present therein.

Whereas the composite beverage may be prepared by combining two or more separate batches of fermented beverage as they are after completion of the fermentation step, it may be convenient to make a composite beverage of which at least one of the compo- nents is a batch which is at least partially dehydrated.In with the invention, it is also possible to isolate from one or more of the separate batches, a flavour compound or a flavour stabilizing with the objective of obtaining the composite beverage by combining the one or more batches of beverage with the thus isolated compound (s).

It will understood that in one convenient embodiment, the invention provides a fer- mented base beverage produced with a yeast strain which is modified by mutation or re- combination to have, relative to the parent strain from which it is derived, a reduced or

lacking of at least flavour compound or of at least compound that is capable of stabilizing the flavour of said composite yeast. As used herein the expression "fermented base beverage"indicates a separately produced beverage which, with respect to one or more flavouring and/or flavour stabilizing as defined above, has a re- duced content relative to the content in the same beverage produced with a yeast strain which, with respect to the particular have not been genetically modified.- Thus, such a base beverage may be one that has a reduced content of at least com- pound that is normally present in that particular type of beverage. In beer, and also in other yeast fermented beverages, such compounds include those mentioned by Meilgaard (1975) to which there is referred. Thus, as an example, a base beer beverage may have a reduced content of one or more of the following compounds : sulphite, sulphide, dimethyl sulphoxide, methanethiol, ethanethiol, methylthioacetate, ethylthioacetate, amyl alcohol andisoamyl acetate.

In with the invention, the desired composite beverage is made by combining a base beverage as defined above with one or more batch components or flavouring and/or stabilizing compound (s) isolated from such a batch. It will understood that these com- ponents which are combined with the base beverage will (i) batches of beverages produced with a yeast strain that has not been genetically (ii) batches of bever- age produced with a strain that is genetically modified by mutation or recombination to have a reduced or enhanced content of one or more particular flavouring or stabilizing compound (s), (iii) batches of (i) or (ii) that has been partially and (iv) one or more flavouring and/or stabilizing compounds isolated (i), (ii) or (iii).

Another objective of the invention is to make it possible to provide a composite beverage by manufacturing, either central orlocally, base beverage by using a genetically modi- fied yeast strain as provided herein and based upon that base beverage make composite beverages by combining the base beverage with one or more of the above components. It may be particularly convenient to use, as the non-base beverage components, rehy- drated, i. e. concentrated, beverages or isolated as this makes it economically feasible to ship such components from a central facility to manufacturers of a base beverage who can then combine their own base beverage with one or more of the provided components to make the final having the desired flavour.

It will also appreciated that it ispossible such a manufacturer to make different composite beverages based on the base beverage each having a specified and prede- termined flavour profile.

Whereas the invention can be applied as described above to facilitate local of composite beverages having a given desired flavour profile, italso it possible to provide a central facility with the means of providing in an economically ad- vantageous manner the possibility to manufacture, based on a bulk production of a base beverage as defined herein, a range of products within a given beverage product which are all with respect to content of flavour i. e. having different flavour profiles.

Sulphite is, as it is described above, an important compound in yeast fermented bever- ages including beer where sulphite acts both as an antioxidant and as an agent that masks certain off-flavours thereby stabilizes beer flavour. In it has been shown that several aldehydes such as trans-2-nonenal are responsible for so-called "stale flavour". It recognized that sulphite may mask the staling of such com- pounds. Thus, a certain level ofsulphite beer and other yeast fermented beverages is essential.

However, whereas it is generally desirable that a fermented beverage has a high content of sulphite order to obtain a high stability of the beverage, there are in many countries regulatory for how much sulphite a beverage is permitted to contain. It iswell-known that even if the same yeast strain is used in the fermentation of the beverage, the amount of sulphite in separate production batches may vary considerably and hence, the amount of sulphite is either too low secure stability it exceeds the permitted level. In for- mer case, sulphite can be added up to the desired level.

The present invention has made it possible to provide a beverage product having an opti- mized, predetermined level sulphite by combining a batch of the beverage having a high content of sulphite with at least other batch made with a genetically modified yeast strain which does not produce sulphite or which, relative to its parent strain, has a significantly reduced production hereof. Accordingly, in one useful embodiment the method makes use of a yeast strain which is modified by mutation or recombination to have a reduced or lacking of sulphite, e. g. a strain that produces at the most 5 ppm of sulphite, as at the most 4 ppm including at the most 3 ppm or even at the

most 2 ppm, e. g. at the most 1 ppm. In embodiments, the strain is substantially incapable producing sulphite. Such a modified yeast strain is preferably a Saccharomy- ces species including Saccharomyces carlsbergensis.

In with the invention, the method of preparing a composite beverage may im- ply at least batch of fermented beverage is used that is produced by using a- yeast strain which is modified genetically have, relative to the parent strain from which it is derived, an enhanced production of a flavour or of a flavour stabilizing com- pound. Accordingly, the invention encompasses that such a batch is produced with a yeast strain which is modified by mutation or recombination to have an enhanced pro- duction of sulphite, such as a sulphite in excess of 30 ppm, including in excess of 40 ppm such as in excess of 50 ppm of sulphite, e. g. in excess of 100 ppm. Such a strain can e. g. be provided as described in Hansen et al., 1996b.

Hydrogen sulphide is, as it is mentioned above, a generally undesiredsulphur compound in yeast fermented beverages including beer, as it confers a strong"putrefied egg"off- flavour the beverage. The threshold level that compound has been reported as being 10 ug/L inlager and 30 ug/L inales. It therefore one objective of the present inven- tion to provide batches of fermented beverage having a low of that compound.

Additionally, it has been shown that there is a direct metabolic linkage formation of hydrogen sulphide and certain thiols as e. g. ethanethiol thioesters including methylthioester andethylthioester. linkage implies that fermenting the beverage batch with a yeast strain having a reduced or lacking ability to produce hydrogen sulphide will result a beverage also a reduced content of thiols esters of such thiols.

Accordingly, in a useful of the invention, there is used a method of producing a composite beverage wherein the yeast strain which is modified by mutation or recombi- nation has a reduced or lacking production of a compound selected from the group con- sisting of hydrogen sulphide, athiol a thioester. However, it is also possible to provide such a modified strain which has an enhanced production of at least one of hydrogen sul- phide, a thiol a thioester.

In embodiments, the method according to the invention provides the means of making a beverage batch which is made by fermenting with a yeast strain which is geneti- cally to have a reduced or lacking production of dimethyl and of making a beverage batch to be used as a component in the composite beverage which has a re-

duced or no production of a fusel alcohol as amyl alcohol orisoamyl alcohol and/or an ester of such a fusel alcohol. In specific embodiment of the latter the modified yeast strain having a reduced or lacking production of fusel alcohol one that is modified in a gene involved in metabolism of branched amino acids, which in the parent strain leads to formation of isoamyl As it is described above, the pathway leading to the formation of isoamyl alcohol andisoamyl in yeast belonging to the SacC4a- romyces involves several enzymes encoded by specific genes. It been found that modified strains of brewer's yeast can be provided which has a reduced or lacking production of isoamyl alcohol the corresponding acetate ester by disrupting or mutating at least of the genes involved this pathway such as a gene selected from the group consisting of LEU1, LEU2, LEU3, LEU4, YOR108W, ATF1, AFT1-CE, ATF1-CA and ORF YDL080CR (THI3). Examples how such genetically modified strains can be provided will described in details in the following examples.

Afurther objective of the present invention is to provide a method of making a composite beverage which includes that the separate batches of beverage which are used for mak- ingthe final beverage can be produced in conventional scale and using conventional methods including, when the beverage is a beer, that a conventional brewer's wort can be used. This implies that the modified strains which are used for fermentation have a fermentation performance under such industrial production conditions which is essentially equivalent to that of a normal, non-modified production strain of yeast. This requirement implies i. a. that the fermentation time under conventional conditions is not extended to any significant degree and that the amount of yeast biomass that is produced during the fermentation is substantially at the same level as with a normal yeast production strain.

Inpresent context, the expression"industrial scale"indicates the volume of a batch of yeast fermented beverage is at least 000 L such as at least 000 litres, in- cludingat least 000 litres.

In specific embodiment, an example of which is described in the following examples, the composite yeast fermented beverage is made by combining at least separate batches of beverage each having a different content of sulphite so as to obtain a compos- ite beverage having a predetermined content of sulphite. The predetermined amount of sulphite will on the type of fermented beverage and may also determined by local regulatoryrules. Typically, predetermined level of sulphite be in the range of 5

to 20 including 10 to 20 ppm such as in range of 12 to 18 ppm including in range of 14 to 16 ppm, e. g. about 15 ppm.

Whereas the invention as described herein is particularly useful the production of a fer- mented beverage product such as beer including lagers, and ales, itwill appre- ciated that genetically modified yeast strain as provided herein can be used in the produc- tion of any fermented beverage where there is a need to control content of one or more flavour compounds. This applies in particular to products like cider and sake, but it is contemplated such modified yeast strains are useful the produc- tion of spirits and beverage products which is made by distillation yeast fermented ce- reals, and vegetables. It isalso that the methods for providing genetically modified yeast strains as provided herein can be applied directly to yeast strains which are particularly adapted to be effective leavening agents in bread doughs such as baker's yeast.

Accordingly, the genetically yeast strains as provided herein include brewer's yeast, wine yeast, distiller's yeast, baker's yeast and a fusion or a hybrid thereof.

In production of conventional fermented beverages, the yeast strain is generally present in the fermentation medium as free, non-immobilized cells, in a floccu- lated orpartially state. However, it may be convenient in a fermentation proc- ess to have the yeast cells an immobilized state. Accordingly, the present invention en- compasses a method as defined herein wherein the yeast strain used for at least of the batches is immobilized during The immobilization may e. g. be on poly- mer particles on any other solid support material or the cells mayalternatively con- tained or entrapped in or by a porous material having pores of a size which retains the yeast cells permits the liquid of the fermentation medium and solutes to come into contact with the thus immobilized yeastcells.

The method of the invention may, as it is mentioned above, comprise a step of at least partially at least of the separate batches of fermented beverage. Such a dehydration step can be carried out using any conventional of removing water from an aqueous medium including asexamples and filtration by reverse os- mosis. The rehydration step may lead a concentrate of the fermented beverage having a dry matter (DM) content of at least 10 wt% including a DM content of at least wt% such as at least wt%, e. g. at least wt% DM such as at least wt% DM.

In afinal of the method according to the invention, the base fermented beverage is combined with one or more separate batches of fermented beverage and/or isolated com- pounds to obtain the final beverage product. The number of batch components that is combined may vary e. g. according to particular preferences. Thus, the number of components which are combined is typically the range of 2 to 10, such as in the range of 2 to 8 including a number in the range of 2 to 5 batches. In specific em- bodiment, the proportion of the fermented base beverage in the composite beverage is at least 25% by volume, such as at least including at least 75%.

It willappreciated it is possible to add further components to at least one of the sepa- rate batches or to the composite fermented beverage with the objective of providing addi- tional desirable characteristics to the beverage product. Thus, as examples further additive components are selected a flavour component, a stabilizing agent and a colouring Such a flavour component may e. g. be a separate yeast fermented bev- erage batch having a high content of one or more particularly desired flavour compounds.

In further aspect, the invention relates to a composite beverage obtainable by the method as described herein. As it has been mentioned above, such a beverage can be a beer including ale, lager stout, a wine, a spirit product, an ethanol-containing bever- age product made by distillation a fermented cereal, or vegetable material. In pre- ferred embodiments, the beverage is a beer, which can be provided as bulk large ship- ment containers such as tanks or barrels as retail in conventional containers forthat purpose including bottles of glass a polymeric material as e. g. PET or PEN, and metal e. g. of tin foil aluminium.

Whereas one major use of the batches of yeast fermented beverage as defined herein is as a component of a composite beverage product, it will appreciated that such batches having a relative content of a flavour compound which is desirable another con- text such as in the manufacturing of a food product other than a beverage, such batches can, optionally after having been at least partially dehydrated, be used directly a flavour compound in food products e. g. in a bakery product, or the flavour may be isolated such batches and used in such isolated optionally after purification, as a flavour compound.

Infurther aspect, the invention relates to a genetically yeast strain which, rela- tive to its parent strain, has a reduced or lacking of sulphite during production of a fermented beverage.

As it is described above, sulphite produced by yeast via the sulphur path- way, the final steps of which include that the MET14 APS kinase catalyses the formation of phosphoadenylylsulphate which in turn is reduced by PAPS reduc- tase (encoded by MET16) tosulphite. sulphite deficient yeast strain according to the invention can be provided by genetically modifying a strain having a wild-type sulphur as- similation in one or more the genes involved in that pathway.

Such modifications are e. g. provided by disruption of one or more genes, e. g. by replacing a wild-type with a deletion allele the gene becomes inactivated. Other pos- sible means of making a strain deficient with respect to a certain phenotype is to mutate the strain in the appropriate target gene including by means of conventional chemical or UV and site-directed mutagenesis according to methods which are well-knownthe art. Alternatively, deficient yeast strains can be provided by se- lecting which are spontaneously in one or more genes involved in the sul- phur pathway. Illustrative of using such techniques for providing sulphite defi- cient yeast strains including S.carlsbergensis described in details the following ex- amples. In present context, useful strains include strains wherein all of the MET14 coding for APS kinase is inactivated.

It been found that is possible to provide such a genetically modified yeast strain whichduring production of a fermented beverage produces less than 5 ppm sulphite in the beverage. In embodiments, the modified strain produces less 4 ppm sul- phite including less 3 ppm sulphite as less 2 ppm sulphite, e. g. less 1 ppm sulphite. In preferred embodiments, the modified yeast strain is substan- tiallyincapable of producing sulphite.

In with the invention, strains of any species of yeast which are used in the production of fermented beverages including Saccharomyces carlsbergensis andSaccha- romyces cerevisiae be provided which are genetically modified to have a reduced or lackingof sulphite.

One typical example such strains is theSaccharomyces carlsbergensis PFJ501 that has been deposited in accordance with the Budapest Treaty on July 1998 with the American Type Culture Collection (ATCC) under the accession number ATCC 74454.

It will appreciated that a sulphite deficient yeast strain can be further modified so as to have, in addition to the reduced or lacking capability to produce sulphite, a reduced or- lacking of one or more of other of the above flavour or flavour sta- bilizingcompounds. Thus, in a further embodiment the sulphite deficient yeast strain ac- cording to the invention is further genetically modified to have under fermentation condi- tions, relative to its parent strain, a modified production of at least one further of the above compounds including a reduced or lacking production of a fusel alcohol an ester thereof and/or an altered production of at least of hydrogen sulphide, a thiol a thio- ester and/or a reduced or lacking to produce dimethyl sulphide fromdimethyl sulphoxide.

It a significant objective of the present invention to provide genetically modified yeast strains as described above which irrespective of the modification have retained the ca- pacityto ferment a medium to a beverage at an industrially acceptable efficiency. Thus, the sulphite deficient strains are strains which are industrially in the production of a fermented beverage including a lager the method comprising fermenting an aque- ous substrate medium with the modified yeast strain as described above.

Dimethylsulphide, is another sulphur the level which in a fermented beverage determines the organoleptic characteristicshereof. In with the in- vention, it is therefore desirable produce batches of yeast fermented beverages having a reduced level this flavour compound. Accordingly, the invention provides in another aspect a genetically modified yeast strain which, relative to its parent strain, has a re- duced or lacking of converting dimethyl sulphoxide (DMSO) dimethyl sul- phide.

The present inventors have confirmed that the S. cerevisiae reading frame (ORF), YER042w, encodes a MetSO reductase and they suggest the designation MXR1 the gene. It discovered that this MetSO reductase is capable reducing DMSO to DMS.

It wasalso that the allotetraploid lager S. carlsbergensis a gene whichis homologous to the above S. cerevisiae As referred to herein, this homolo- gous gene is designated MXR1-CE. It further discovered that S. carlsbergensis con-

tains at least further gene coding for MetSO reductase also showinghomology the YER042w ORF and the MXR1-CE. homologous further gene is designated herein as the MXR1-CA gene.

Based on these discoveries it has become possible to provide yeast strains including strains of Saccharomyces carlsbergensis andSaccharomyces cerevisiae a re-- duced or lacking capability of converting DMSO into DMS. Such can be made by genetically modifying a yeast strain having functional coding for the MeSO reduc- tase including the above genes. Thus, in one useful the yeast strain ac- cording to the invention is a yeast strain such as a Saccharomyces in which the ORF YER042w (MXR1) a homologue which codes for a gene product capable of converting dimethyl sulphoxide intodimethyl is inactivated.

An inactivation of one or more of said genes can be made by use of conventional tech- niques including those mentioned above in connection with the description of a sulphite deficient strain.

It willunderstood that a yeast strain as described above, which is modified to have a reduced production of DMS can be further genetically modified to have under fermentation conditions, relative to its parent strain, a reduced or lacking production of a fusel alcohol and an ester thereof and/or an altered of at least one of hydrogen sulphide, a thiola thioester and/or a reduced or lacking capability produce sulphite.

To be useful a beverage fermenting strain according to the invention, a DMS deficient yeast strain has preferably its capability be an effective fermenting strain in in- dustrial beverage production such as in the production of a lager Accordingly, an objective of the invention is to provide a method of producing a yeast fermented beverage, comprising fermenting an aqueous substrate medium with a DMS deficient yeast strain according to the invention.

As mentioned above, there have been disclosed attempts to control the formation of hy- drogen sulphide beer production by means of DNA recombinant technology. It now been found that it is possible to provide non-recombinantly modifiedSaccharomyces carlsbergensis strains having under beer fermentation conditions, relative to its par- ent strain, an altered production of hydrogen sulphide or of a thiol a thioester. Such non-recombinational include conventional mutagenization procedures and selec-

tion spontaneously occurring combined, when appropriate, with crossing of yeasts of opposite mating types. In thefollowing, illustrative of such procedures and methods are described. These examples, however, are not limiting and any non- recombinational which lead modified brewing yeast strains having, relative to their respective parent strain, a reduced or enhanced production of hydrogen sulphide or sulphur the production of which is linked the hydrogen sulphide production, is encompassed by the invention.

Thus, in accordance with the invention there is provided a mutant S. carlsbergensis yeast strain which, relative to a parent strain from which it is derived, has a reduced hydrogen sulphide production including the strains designated herein as JH441, JH442, JH443 and JH444 as described in the following, which JH441 has been deposited in accordance withthe Budapest Treaty on July 1998 with the American Type Culture Collection (ATCC) under the Accession No. ATCC 74451.

There is also provided S. carlsbergensis having, relative to a parent strain, an en- hanced hydrogen sulphide production, including the strains disclosed herein which are designated JH506, JH515, JH516 and JH517, of which JH506 has been deposited in ac- cordance with the Budapest Treaty on July 8, 1998 with the American Type Culture Col- lection under the Accession No. ATCC 74453.

It willunderstood that a yeast strain as described above, which is modified to have a reduced or enhanced production of hydrogen sulphide can be further genetically modified to have under fermentation conditions, relative its parent strain, a reduced or lacking production of a fusel alcohol an ester thereof and/or a reduced or lacking capability of converting dimethyl sulphoxide intodimethyl and/or a reduced or lacking capabil- ityto produce sulphite.

It another objective of the present invention to provide a genetically yeast strain as described above which irrespective of the modification has retained the capacity to ferment a medium to a beverage at an industrially acceptable efficiency. Thus, such modified strains are preferably which are industrially in respect of a method of producing fermented beverage including alager the method comprising fermenting an aqueous substrate medium with the modified yeast strain as described above.

Inaspect of the invention, there is provided a genetically modifiedSaccharomyces carlsbergensis strain, which, relative its parent strain, has a reduced or lacking productionunder beer fermentation conditions of a fusel alcohol an ester thereof.

The genetic background for the production of these flavour compounds are outlined above. Based on the pathway leading to higher alcohols their corresponding esters, several approaches can be used to provide S. carlsbergensis having a reduced productionof these compounds.

Thus, with respect to e. g. reduced production of isoamyl alcohol esters thereof in- cludingacetate ester, at least the following approaches can be used in an attempt to providesuch modified yeast strains : (a) Inactivation/modification the isopropylmalate in the leucine denovo syn- thesis pathway (encoded by LEU4 ORF YOR108W); (b) Inactivation the TH13 activity responsible the conversion of a- ketoisocaproate to 3-methyl-1-butanol; and (c)Removal of the alcohol acetyl activities encoded by ATF1-CE, ATF1-CA and ATF2 genes.

Examples of suitable methods involved in both approaches are described in the following examples. It from these examples that strains in which one or more of these genes are inactivated can be generated by introducing into the relevant a deletion allele the naturally gene by means of e. g. a disruption plasmid com- prisingthe mutant gene or by means of the two-step deletion as described by Scherer and Davis (1979), the so-called loop-in/loop-out Although these methods are presently preferred, it is contemplated alternative conventional methods for dis- ruptingor inactivating genes can be applied such as e. g. site-directed mutagenesis or dis- ruptionby means of transposable elementsincluding transposons. It isalso conceivable that strains having the reduced production of a higher alcohol and/or corresponding ester can be provided by random mutagenesis or by selecting spontaneously occurring mutants.

In oneuseful the invention provides a S. carlsbergensis having a re- duced or lacking of isoamyl and/or isoamyl including a strain in

which a gene involved the leucine novo synthesis pathway is inactivated or modified such as a strain in which at least of LEU4 ORF YOR108w inactivated or modi- fied.

As mentioned above, a brewer's yeast strain modified in its pathways leading to the gen- eration of esters of higher alcohols be provided as a strain in which at least one gene coding for alcohol acetyl activity including the genes ATF1-CE, ATF1-CA and ATF2 inactivated or deleted.

A genetically S. carlsbergensis in accordance with the invention has a re- duced or lacking production of one or more higher alcohols esters of such alcohols and which is effective in a method of producing a lager comprising fermenting a brewer's wort, can be further genetically to have under fermentation conditions, relative to its parent strain, an altered production of at least one of hydrogen sulphide, a thiola thioester and/or a reduced or lacking capability of converting dimethyl sulphoxide into dimethyl and/or a reduced or lacking capability produce sulphite.

The invention is further illustrated the following and the drawings wherein Fig. 1 is a flow diagram of the different steps involved in the construction of a Met brewer's yeast.

Fig. 2a is a schematic illustration of the restriction map of the S. cerevisiae MET14 geno- mic region and the expected sizes of fragments that will visualized in a Southern analysis employing depicted restriction endonucleases; Fig. 2b shows a Southern blot DNA from T224, allotetraploid yeast, C80- CG65 and C80-CG110, allodiploid yeast, CBS1503, S.monacensis S288C, S. cerevisiae; Fig. 3 illustrates the isolation the S. monacensis MET14 that was done by using the sequence information from the S. cerevisiae database (SGD). The MET14 gene is found on chromosome XI. part of the chromosome, from bp 437133 to 442363, has been enlarged schematically, showingMET14 its flanking open reading frames.

Oligoprimers designed (numbered #1-#6, Table 1. 2) having either an Xbal or

BamHl in the end, enabling PCRamplification the 3'-and 5'-ends of MET14 and also open reading frame of the gene. The PCR fragments A and B digested with either Xbal orBamHI werecloned pRS316 (Sikorski and Hieter, 1989) digested with the same enzymes. Both inserts were using appropriate oligoprimers. frag- ment C was sequenced directly using oligoprimer #5. Based the raw sequences, oligo- primer #7 and#8 designed allowing amplification of fragment D, containing the en- tire MET14-CA including 3'-and 5'-regions. Three individually synthesized D- fragments were digested with BamHl cloned into the BamHl of pUC18 resulting in pPF50, pPF51 and pPF52 ; Fig. 4 shows the sequence (SEQ ID NO: of the MET14-CA The result se- quencing of the three independent clones pPF51 and pPF52. The ATG-start co- don and TAA-stop codon are underlined. The BamHl sites the ends are cloning sites and not part of the sequence ; Fig. 5a shows the construction of a deletion allele ofMET14. insert in clone #42 is approximately 8, 000 bp, the Clal-Sa/l digest leaves behind about 5, 600 bp of the insert ; A A the position of the deletion ; Fig. 5b illustrates construction of a deletion allele ofMET14-CA PCR. The four oli- goprimers (#9-#12, 1. 2) were all with restriction sites in the ends, EcoRl, BamHl, Xbal andHindlll, respectively. Only restriction sites are depicted.

Fig. 6 illustrates Southern hybridizations of brewing yeast strains containing an inactivated allele the S. cerevisiae-like MET14 PFJ439 is C80-CG65 with pPF35 integrated at the MET14-CE locus, Southern blot both the WT (wild-type) and the A (deletion) PFJ442 is the C80-CG65 derived integration strain before loop-out, con- taining a gene conversion, therefore only theA-band present. PFJ448-PFJ452 are loop- outs of PFJ442 being G418-sensitive and having only the A-band. probe was used the Xbal-Ndel from MET14; Fig. 7 shows the fermentation profile of strain M204 PFJ501 and PFJ514 from 1st, and 3rd brew generations, respectively from a 50 L pilot Gravity of the wort is calculated as % Plato;

Fig. 8 shows the number of cells 50 L fermentation of strain M204 (reference), PFJ501 and PFJ514 from three brew generations ; Fig. 9 illustrates sequencehomology methionine sulphoxide genes from different organisms and yeast ORF YER042w. EcPMSR, Escherichia coli peptide methionine sulphoxide (Rahman et al. , ; BtPMSR, Bos taurus peptide- methionine sulphoxide (Moskovitz et al. , ; AtPMSR, Arabidopsis thaliana peptide methionine sulphoxide (Genbank accession X97326) ; FaPMSR, Fra- garia ananassa (strawberry) peptide methionine sulphoxide (Genbank acces- sion Z69596) ; Fig. 10 shows a one-step gene disruption approach employed fordeletion ORF YER042w in S. cerevisiae confirmative PCR reaction ; Fig. 11 shows toxicity of ethionine sulphoxide in wild-type yeast (M1997) in strain JH465 (AYER042w). 200 u ! of 0. 1 M DL-ethioninesulphoxide were ap- plied to the filter and 50 NI the yeasts in water suspension were applied to the ethioninesulphoxide-gradient. Theplates incubated at 30°C 4 days ; Fig. 12 illustrates growth in liquid culture (YPD or SD media) of a wild-type yeast strain and strain JH465 ; Fig. 13 illustrates of DMS from DMSO added to liquid growth media or already present in brewer's wort. A : M1997, ; B : JH465, SD ; C : M1997, 1 ppm DMSO ; D : JH465, SD + 1 ppm DMSO ; E : M1997, SD + 10 ppm DMSO ; F : JH465, SD + 10 ppm DMSO; G : M1997, SD + 100 ppm DMSO ; H : JH465, SD + 100 ppm DMSO ; I: M1997, MP; J : JH465, MP ; K : M1997, MP+ 1 ppm DMSO ; L : JH465, MP + 1 ppm DMSO ; M: M1997, MP + 10 ppm DMSO ; N : JH465, MP + 10 ppm DMSO ; O: MP + 100 ppm DMSO; P : JH465, MP + 100 ppm DMSO ; Q : M1997, wort (autoclaved) ; : JH465, wort (autoclaved); : M1997, (fresh) ; T : JH465, wort (fresh) ; Fig. 14 shows alleles S. cerevisiae in S. carlsbergensis, S. monacensis and the allodiploid of S. carlsbergensis, C80-CG110 C80-CG65 ; Fig. 15 is a schematic showing of feed-back inhibition of threonine its own biosynthetic pathway and its supposed effect on the sulphur metabolism inSaccharomyces ;

Fig. 16 illustrates for hydrogen sulphide on BIG-YNB plates se- lected with reduced hydrogen sulphide (strains, JH442, JH443 and JH444). Strain M204 was used as the reference S. carlsbergensis strain ; Fig. 17 is an outline of the strategy used to make mutants of allotetraploid yeasts with enhanced hydrogen sulphide production ; Fig. 18 illustrates for hydrogen sulphide on BIG-YNB with se- lected strains with enhanced hydrogen sulphide (strains JH506, JH515, JH516 and JH517). Strain M204 was used as the reference S. carlsbergensis strain ; Fig. 19 shows production of methylthioacetate, (Strains M204, JH506, JH515, JH516, JH517, JH442, JH443 and JH444) or ethylthioacetate, (Strains M204, JH506, JH515, JH516 and JH517). Numbers are means of three fermentation experi- ments ; Fig. 20 shows the production of ethanethiol (EtSH),methylthioacetate and eth- ylthioacetate from a reference production strain (Prod. strain), a strain with high hydrogen sulphide production(met2A) a strain without hydrogen sulphide production (met10A); Fig. 21 is a map of the LEU4 promoter and open reading frame (ORF). Relevant restric- tion sites are indicated. DNA sequences used for restriction maps and PCR amplifications were obtained from the Saccharomyces Genome Database (http ://genome-www.- stanford.edu/Saccharomyces/). Sizes are in bp. The shown map was made with pDRAW 1.0 ; Fig. 22 is a map of PCR generated 1. 5 kb LEU40 The fragment is shown in the LEU4 plasmid PMBP9711. final loop-in plasmid an additional 3. 7 kb Hindlil fragment with MET2-CA from S.carlsbergensis and Kielland-Brandt, as selection The MET2-CA was inserted at Hindlil nt 1881). The plas- mid was linearized withAflll. sequences used for restriction maps and PCR amplifi- cations were obtained from the Saccharomyces Genome Database (http ://genome- www. stanford. edu/Saccharomyces/). are in bp. The shown map was made with pDRAW 1. 0;

Fig. 23 shows the amino sequence (SEQ ID NO: of Leu4p. DNA and protein se- quences used for restriction maps and PCR amplifications obtained from the Sac- charomyces Genome Database (http ://genome-www. stanford. edu/Saccharomyces/).

Amino acid residues marked in bold deleted while in underlined have been altered ; Fig 24 is a map of the LEU4 homologue YOR108W on chromosome XV. Useful re- striction sites are indicated. DNA sequences used for restriction maps and PCR amplifica- tions were obtained from the Saccharomyces Genome Database (http ://genome-www.- stanford. edu/Saccharomyces/). Sizes are in bp. The shown map was made with pDRAW 1.0 ; Fig. 25 shows nucleotide homology of LEU40, LEU4 ORF YOR108W. Disrup- tion linkers the LEU4 homologue ORFYOR108W aswell asMET2-CA points are indicated. DNA sequences used for restriction maps and PCR amplifications were obtained from the Saccharomyces Genome Database (http ://genome-www. stanford.- edu/Saccharomyces/). Sizes are in bp. The shown alignment was made with MACAW 2.0. 5 ; Fig. 26 shows a SC-leucine (SC-leu) mediumplate containingLEU4, LEU4, leu4A and leu4A yor108w mutants. Colonies 1, 2, 3, 4 and 5 are the Ieu40 yor108w doublemutants MBP97-51, MBP97-53, MBP97-54 and MBP97-55.

Fig. 27 is a map of part of the integration plasmid shown with selected restric- tion sites. The afflua fragment was ligated the Sacl-Pstl of pMBP96A2 (see Fig. 29). Besides the indicated pUC19, this plasmid an additional 7 kb Hindlil fragment (at nt 1521) comprising the MET2-CA (Hansen and Kielland-Brandt, 1994) as selection marker. Prior to integration, the plasmid partially digested with Nsil; Fig. 28 is a map of part of the integration plasmid shown with selected restric- tion sites. The atf2E fragment was ligated the Sall-Sphl of pMBP96A2 (see Fig. 29). Besides the indicated pUC19, this plasmid an additional 7 kb fragment (at nt 1521) comprising the MET2-CA (Hansen and Kielland-Brandt, 1994) as selection Prior to integration, the plasmid linearized with Mscl;

Fig. shows the plasmid constructed by inserting a 3. 7 kb Hindlil-fragment comprising the MET2-CA from S. carlsbergensis and Kielland-Brandt, 1994) into the Hindi l of pUC19 et al. , The MET2-CA en- codes a homoserine acetyl ; Fig. 30 is a map of the integration plasmids and pSBS97-6 shown with selected restriction sites. The atf1-ceA fragments were ligated the Xbal-Sacl of pCH216 (Hadfield etal. , The only between the two plasmids that the template for the PCR fragment in pSBS97-3 was genomic DNA from C80-CG65 and the template for pSBS97-6 was genomic DNA from C80-CG110. to integration, the plasmids werelinearized with;Sphl Fig. 31 is a map of the integration plasmids and pSBS97-7 shown with selected restriction sites. The atf1-cat fragments were ligated intotheXbal-Sacl of pCH216 (Hadfield etal. , The only between the two plasmids that the template the PCR fragment in pSBS97-4 was genomic DNA from C80-CG65 and the template pSBS97-7 was genomic DNA from C80-CG110. to integration, the plasmids linearized by partial digestion with Sapl; Fig. 32 is a map of the integration plasmids and pSBS97-8. The atf2-ceA PCR fragments were ligated the Xbal-Sacl of pCH216 (Hadfield etal. , The only difference between the two plasmids that the template for the PCR fragment in pSBS97-5 was genomic DNA from C80-CG65 and the template for pSBS97-8 was geno- mic DNA from C80-CG110. to integration, the plasmids werelinearized withStul, and Fig. 33 shows the results a sensory comparative test of batches of beer produced by the following S. carlsbergensis : PFJ501 having a reduced sulphite production (in- activated in all fourMET14 genes),SB130 increased sulphite (inacti- vated in all fourMET10 genes for sulphite reductase), and strain M204, an indu- strial lager a composite beer made by mixing the PFJ501 and SB130 batches.

EXAMPLE 1 Construction of S. carlsbergensis yeast without production of sulphite 1. 1. Summary of the experiments Sulphite, being an antioxidant and flavour stabilizer, plays key role preservation of beer taste. In to be able to control amount of sulphite a final it was de- cided to make a brewer's yeast being deficient with respect to sulphite By combining beer from low high sulphite beer batches, it would bepossible to make beer with a predetermined amount of sulphite. Furthermore, due to a reduced content of the yeast-derived sulphur the low-sulphite yeast strain would an appropriate basis for construction of a yeast strain which could used to produce a taste-neutral Such a beer could form the basis for combination with other beer batches.

Saccharomyces carlsbergensis yeast was inactivated in all copies of the MET14 by disruption combined with classical Beer produced with this yeast contained no measurable sulphite when bottled and was quite satisfactory with re- spect to brewing characteristics.

Saccharomyces carlsbergensis yeast is presumably allotetraploid, two copies of each of two divergent alleles the genes investigated so far, as reviewed by Kielland-Brandt etal. , 1995. In following is described the construction of S. carlsber- gensis yeast strains impaired in their sulphite production by inactivation of all four copies of MET14 APS kinase, EC 2. 7. 1. 25) using targeted genetical disruption as well asclassical methods (Fig. 1). The resulting strains were evaluated with respect to brewing performance. Levels sulphite and trans-2-nonenal were deter- mined, and the beer was evaluated a trained taste panel.

1.2. Materials and methods Strains and media Strains are presented in Table 1. Synthetic sulphur-free B-medium was made according to Cherest and Surdin-Kerjan (1992), but modified by the addition of 10 mM ammonium

sulphate 0. 2 mM DL-homocysteine thiolactone. YPDcomplex medium(1 % ex- tract, 2% peptone, 2% glucose) synthetic complete (SC, Sherman, 1991) was used for all except the 50 L-scale where brewer's wort with a gravityof 14. 5 % Plato used. Geneticin (G418) was added (30 g/L) to YPD plates for selection of transformants of brewer's yeast and the following of transformants.

5 mM of sodiumselenate added to B-medium plates the selection ofselenate- resistant yeast strains.

Table 1. Strains of Saccharomyces in the experiments.

Strain Mating type Genotype Ploidy Origin M204 nm 2+2 Lager production strain C80-CG65 a 1+1 Meiotic segregant of M204 C80-CG110 a1+1 Meiotic segregant of M204 T224 nm 2+2 Lager productiorFstrain CBS1503** nm1 CBS PFJ445 a deficient of chr. XI 1+1 of C80-CG65 PFJ448 a metl4-ca-MET14-CE Derivative of C80-CG65 PFJ449 a do 1+1 do PFJ450 a do 1+1 do PFJ451 a do 1+1 do PFJ452 a do 1+1 do PFJ457 a metl4-ca-metl4'-ce Derivative of PFJ450 PFJ458 a do 1+1 Derivative of PFJ449 PFJ459 a do 1+1 do PFJ460 a do 1+1 do PFJ461 a do 1+1 Derivative of PFJ451 PFJ462 a do 1+1 do PFJ463 a do 1+1 do PFJ464 a do 1+1 do PFJ465 a do 1+1 Derivative of PFJ452 PFJ466 a do 1+1 do PFJ467 a metl4-ce MET14-CA Derivative of C80-CG110 PFJ468 a do 1+1 do PFJ530 a metl4-cemetl4'-ca Derivative of PFJ468 PFJ501 nm met14-ce/met14-ce 2+2Cross PFJ459 and metl4*-ca/mefl4*-ca PFJ530 PFJ502 nm Do 2+2 Cross PFJ460 and PFJ530 PFJ506 nm Do 2+2 Cross between PFJ461 and PFJ530 PFJ509 nm Do 2+2 Cross PFJ462 and PFJ530 PFJ510 nm Do 2+2 Cross between PFJ462 and PFJ530 PFJ514 nm Met14-ce/met14-ce 2+2 between PFJ463 and Me'-ca/me'-ca PFJ530 K396-22B a spoll ura3 adel hisl leu2 lys7 S. Klapholz met3trp5 X2928-3D (1C) aadel gall leul his2 ura3trp1 1 met14 S288C** a MAL GAL2 1 YGSC, K. Mortier

The genotype of all yeast strain are wild-type except for the MET14 locus de- scribed. * This strain seems to have lost S. cerevisiae-like chr.XI an integration attempt (P. F. Johannesen, unpublished). ** All are S. carlsbergensis derivatives except CBS1503 (S. monacensis) S288C (S. cerevisiae). : Central voor Schimmelcultures, Netherlands ; YGSC : Yeast Genetics Stock Centre, University of California, Berkeley, CA, U. S. A. Abbreviations : 1+1, allodiploid; 2+2,;allotetraploid nre non mater.

Yeast transformation Transformation of yeast was carried out according to Schiestl Gietz (1989). When transforming brewer's yeast the transformed cells with Tris-EDTA buffer) were resuspended in 100, ul of YPD and left at room temperature over night allowing the cells recover and express the neomycin phosphotransferase before plating on selective medium.

Extraction of genomic DNA Genomic DNA was extracted from cells broken by use of glass beads ac- cording to Hoffman and Winston (1987). DNA was extracted from 10 ml yeastcultures.

For Southern analysis 10 ug DNAsolution used.

Southern blot hybridization Genomic DNA was digested with appropriate restriction enzymes and separated on 1 % agarose gels. was transferred to Hybond-N nylon by capillary blotting.

Covalent of the DNA to the filter mediated by UV-irradiation. The filters were hybridizedwith suitable 32P-labelled at high stringency (65°C, with 0.1 x SSC). The signals were detected by using a Phosphorlmager Dynamics, Sunnyvale, CA).

Sequencing DNAfor sequencing was prepared using the WizardE Plus DNA Purification System (Promega). Sequence reactions were made according to protocols from the manufacturer using PRISMTM AmpliTaqO Dye Terminator Cycle kit (Ap-

piied in a Perkin Elmer machine. The sequencing reactions were proc- essed in an Applied Biosystems 373A DNA Sequencer, according to the 373A User's manual.

PCR Synthesizing the MET14 using S. monacensis as template and homoeologous primers designed from S. cerevisiae chromosomeXI, requiredlow in the PCR reaction: 50-100 ng genomic DNA, 0. 2 mM of each dNTP, 1.0 uM of eacholigoprimer, re- action buffer including to 4 mM chloride, 2. 5 u Taq-polymerase (Perkin- Elmer), millipore up to 100 NI. was at 94°C 1 min., annealing at 45- 60°C 2 min. (Generally product was only at the lowest temperatures), extension at 72°C 3 min., and 25-30 cycles run. All were started by an initial de- naturation at 94°C 4 min. and terminated by an extension at 72°C 10 min. Amplifi- cation of the S. carlsbergensis-specific MET14 for sequencing was accomplished by using Expand Fidelity PCR System (Boehringer Mannheim), a mixture of Taq- polymerase andPwo-polymerase thelast proof-reading potency. Reaction condi- tions differed from the above as follows : 2. 5 mM magnesium in the buffer, an- nealing at 52°C 2 min., extension at 72°C 3 min., and only 20cycles run. A Stratagene Robocycler0 40 Temperature Cycler used for all reactions.

Reconstitution of allotetraploid brewer's from allodiploid maters Crossing of the allodiploid to reconstitute an allotetraploid yeast was performed according to Gjermansen and Sigsgaard (1981) with some modifications.

Small of cells mixed on a YPD plate. strains were allowed mate for 24 hours at room temperature followed by inoculation into YPGal medium(1 % ex- tract, 2% peptone, 2% galactose and Kielland-Brandt, as carbon source) and left 2 days of growth at room temperature. Some of the cultures were spread for single on YPGal plates. showed a markedly growth compared with the two allodiploid on this carbon source. Hybrids can be isolated aswell-growing yeast colonies.

Fermentation at 50 L scale After propagation, yeast was inoculated in brewer's wort at 1. 5x107 cells/ml 50 L fer- mentation vessels. Fermentation was performed at 13°C. yeast was harvested after 10 days and the beer was left the vessels for a period of 8-12 days of lagering. The harvested yeast was used for the subsequent brewing generations.- Other analyses Either the supernatant from centrifuged samples or bottled were assayed for total S02 headspace with Sulfur Chemiluminescence (SCD) TM and Dreyer, 1997). Acetaldehyde measured by gas liquid chromatography. Trans-2- nonenal measured by GC-MS. Beer was determined using an Anton Paar DMA46 density meter (Anton Paar KG, Graz, Austria).

1.3. Results (i) Evaluation the content of MET14 genealleles S. carlsbergensis In to determine the content of MET14 in S. carlsbergensis yeast a Southern analysis was made using genomic DNA from different Saccharomyces yeasts.

Genomic DNA from S288C (haploid S. cerevisiae strain), CBS1503 (S. monacensis), C80-CG65(allodiploid of S. carlsbergensis yeast), C80-CG 110 (allodiploidof S. carlsbergensis yeast) and T224 (allotet- raploid yeast) were used in the analysis.

DNAwas digested with two different combinations of restriction enzymes, Nrul+Scal and Xbal+Ndel. Southern blot hybridized using a 1150 bp fragment of the MET14 as a probe (Fig. 2). A distinct pattern of bands was seen in both S. cerevisiae S. mona- censis, in T224, C80-CG65 and C80-CG110 pattern which is a combination of those with S. monacensis S. cerevisiae seen. This shows that the T224 produc- tion yeast as well the two allodiploid contain two different alleles ofMET14 (one hereinafter denoted MET14-CE) the gene found in S. cerevisiae, one (hereinafter denoted MET14-CA) like theMET14 found in S. monacensis.

(ii) Isolation the S. cerevisiae-like MET14 from brewer's yeast In to be able make a deletion allele the S. cerevisiae-like MET14 the MET14-CE was cloned brewer's yeast. This was performed as it was assumed that the MET14-CE genewould be exactly to the MET14 from S. cere- visiae, as a deletion construct based on MET14-CE would make it not to-in- troduce foreign DNA into the brewer's yeast.

The gene was cloned bycomplementation a S. cerevisiae met14 using a S. carlsbergensis (Casey, 1986). The complementation cloning has been described (Johannesen, 1994) and the resulting clone#42 an YRp17 Botstein and Davis,was mapped by restriction analysis. It out that the MET14 dif- fered from the MET14-CE by at least the unique Sall which is not present in the MET14-CE gene.

A Clal-Sall of approximately 400 bp from clone #42 wassubcloned vector pRS316 (Sikorski and Hieter, 1989). The Clal is in the MET14-CE whereas the Sall is in YRp17 bp of the vector sequence between the BamHl and the Sall site are brought along). The resulting plasmid pPF28, and the insert can be cut out with a Clal-Sall digest (Fig. 5).

(iii) Construction of a deletion allele the MET14-CE gene.

Adeletion allele the S. cerevisiae-like MET14 was constructed, enabling inactiva- tion of the MET14-CE gene S. carlsbergensis yeast. Plasmid contains an insert of about 2, 400 bp including the S. cerevisiae-like MET14 in the pRS316 vector (Sikorski and Hieter, 1989). This plasmid digested with Hpal and Nrul, both cutting uniquely, thus deleting a piece of 471 bp covering the promoter region and the first part of the open reading frame of MET14-CE. Theplasmid wasreligated, both en- zymes generated blunt resulting in plasmid The insert of pPF33 could cut out with Clal andSall. fragment was blunt endligated the integration vector pCH216 (Hadfield, opened in Sacl blunted. This vector contains the G418 (ge- neticin) resistance gene, usable selection of transformation of brewer's yeast. The re- sulting plasmid, (Fig. 5a), can be linearized cutting with Sacl the MET14-CE coding region. Linearization of pPF35 will the integration preferentially take place through the MET14-CE (according to Orr-Weaver et al., 1981).

(iv) Use of the deletion allele the MET14-CE to inactivate the S. cerevisiae- like MET14 alleles in the allodiploid C80-CG65 and C80-CG110 In to obtain a brewer's yeast without sulphite it was decided to inactivate all copies of MET14 this yeast. The strategy that was chosen was based on the capacity of the yeast to allow forhomologous between a deletion allele-of a gene on a plasmid the wild-type gene on the chromosome using the loop-in/loop-out method (Scherer and Davis, 1979). Each of the two allodiploid from the C80-CG65 and C80-CG110 of brewer's yeast presumably contains one copy of each version of MET14 2). Hybrids made from the two maters result in reconstituted allotetraploid brewer's yeast with a normal brewing performance (Gjermansen and Sigsgaard, 1981).

The MET14-CE in C80-CG65 and C80-CG110, were disrupted using a deletion allele the gene on plasmid (Figure 5a). Construction of this plasmid is described in the section above. Before transformation of C80-CG65 and C80-CG110 with pPF35, the plasmid waslinearized digestion with Sacl order to direct the integration to the MET14-CE locus et al. , Transformants were selected YPD plates geneticin. Geneticin resistant strains were picked and analyzed by Southern hybridizations to confirm correct integrations (data not shown). Genomic DNA was di- gested with Xbal and Ndel the Southern blots probed with a Xbal-Ndel fragment from MET14.

Asecondary recombination event is necessary in order to loop-out vector DNA and either the wild-type or deletion allele ofMET14-CE. integration strains derived from each mater were selected the loop-out (Scherer and Davis, 1979). These were grown to stationary phase in 10 ml medium without geneticin, 100 NI culture was reinoculated 10 ml and grown to stationary phase, and the procedure re- peated once more. Approximately 000 cells each culture were spread onto YPD plates the colonies werereplica-plated YPD supplemented geneticin. Ge- neticin-sensitive colonies wereisolated analyzed by Southern hybridization to identify clones which wild-type MET14-CE been substituted with the deletion allele.

In of the C80-CG65 transformants a gene conversion appeared to have taken place, resulting in two copies of the deletion allele the chromosome, separated by the vector sequences. After successful loop-out the vector sequences in this particular transfor- mant, only thedeletion allele could on the chromosome. Strains PFJ448, PFJ449,

PFJ450, PFJ451 and PFJ452 are derivatives of this strain, all in the S. cere- visiae-like MET14 (Fig. 6). Strain PFJ467 and PFJ468 are the corresponding C80- CG110 derivatives,although were made from a transformant containing both a wild- type and a deletion allele ofMET14-CE not shown).

(v) Inactivation the S. carlsbergensis-specific MET14 alleles PFJ449, PFJ450, PFJ451, PFJ452 and PFJ468 by employing and resistance towards heavy metal ions To inactivate the S. carlsbergensis-specific MET14 alleles PFJ449, PFJ450, PFJ451, PFJ452 and PFJ468, respectively, was used, taking advantage of the toxicityof certain sulphate analogs. It been shown (Breton and Surdin-Kerjan, 1977) that mutant strains of S. cerevisiae are resistant to selenate chromate (sulphate analogs) aremainly in the MET3 but some are also inMET14 orMET16.

The strains described in the previous section only one active copy of MET14 (MET14-CA),to deletion of the other (MET14-CE). Whenscreening mutants from these strains on selenate we expected only find met14-ca as proba- bly twoMET3 andMET16 are present in the allodiploid maters.

The five strains PFJ449, PFJ450, PFJ451, PFJ452 and PFJ468 were cultured to station- ary phase in YPD medium, cells washed in water and aliquots of200 uL culture were spread onto plates B-medium and 5 mM sodium selenate. yeast was muta- genized using UV irradiation corresponding to about 50% killing. Mutants were isolated after 5-8 days growth at 20°C. mutants that were methionine auxotrophs were kept, as only thesecould possible met14-ca Such auxotrophs were then test- crossed to S. cerevisiae carrying mutations in either MET3 orMET14. Comple- mentation of the methionine requirement was seen with the mutant met3-strain, not in the crosses involving the mutant met14 Ten strains having a C80-CG65 back- ground were isolated: PFJ458, PFJ459, PFJ460, PFJ461, PFJ462, PFJ463, PFJ464, PFJ465 and PFJ466. Only oneC80-CG110 was isolated, PFJ530.

The strains were also for reversion to methionine and they showed a lowof reversion. The nature of the mutations obtained via selection onselenate medium are unknown, and might not lead complete inactivation of the MET14-CA genes. Therefore, it would suitable to produce mutants, in which all MET14 are inactivated in a defined manner by recombinant methods.

(vi) Isolation characterization of the S. monacensis MET14 gene As described above, S. carlsbergensis yeast contains two different alleles of MET14 which the S. cerevisiae-like allele been characterized previously (Korch et al., 1991). In to inactivate all MET14 activity in a S. carlsbergensis brewer's yeast, it was necessary to isolate andcharacterize S. carlsbergensis-specific MET ?} gene as well. unknown reasons, it was not possible to isolate gene by genetic complementation an S. cerevisiae met14 strain with an S. carlsbergensis gene library 1986). Although 500, 000 transformed colonies were screened only the S. cerevisiae-like allele found.

Instead, sequence information from S. cerevisiae used to design oligoprimers which enabled PCR amplification of the S. monacensis MET14 from S. monacensis DNA template (Fig. 3). This strategy relied two assumptions : i) that the S. carlsbergen- sis-specific MET14 is homologous the only MET14 found in S. monacensis, and ii) that the gene order in the S. cerevisiae the S. monacensis MET14 is conserved and the DNA sequences sufficiently related toallow annealing the S. cere- visiae-derived to the S. monacensis DNA. <BR> <BR> <BR> <BR> <BR> <BR> <P>Fragment B (Fig. 3) was amplified usingoligoprimers #3 and#4 (Table 2.) based on the sequence of the S. cerevisiae MET14 reading frame. Likewise, fragment A and C (Fig. 3) were synthesized using oligoprimers #1, #2, #5 and#6 1. 2), also designed by using the sequence information from S. cerevisiae chromosomeXI. sequencing of the 3'-and 5'-ends (fragments A and C) of the S. monacensis MET14 gene,oligo- primers were designed from the nucleotide of these regions, only a low degree of homology the same regions in the S. cerevisiae-like MET14 As tem- plate for the synthesis of the full-length S.carlsbergensis-specific MET14 a deriva- tive of C80-CG65, PFJ445 was used. This strain appears to be deficient of the entire S. cerevisiae-like chromosomeXI not shown) on which MET14-CE found, and only to contain the S. carlsbergensis-specific Chr.XI. PCRamplification the S. mona- censis-derived oligoprimers #7 and#8 resulted a fragment of 1160 bp containing the MET14-CA Three individual of PCR fragments were amplified. frag- ments from each of these were inserted into the pUC18 vector giving rise to three individ- ual clones pPF51 and pPF52. The inserts in all threeclones been se- quenced and compared in order to eliminate possible errors due to the PCR reaction. The

three individual was completely The resulting MET14-CA sequence is in Fig. 4.

Table 2. Oligoprimers (SEQ ID : 1-12) used for amplification of fragments A-D in Fig.2 Oligoprimers Sequence # 1:MET14-o)i125'-GCTCTAGAGCATTGGAGTTGGTTATGCG-3' # 2:MET14-o!i75'-GCGGATCCCCGACCAGAAGACGGTTGAAGAATGTGC-3' # 3:MET14-o!i35'-GCGGATCCGCACATTCTTCAACCGTCTTCTGGTCGG-3' # 4:O)igo#45'-GCTCTAGAATGGCTACTAATATTACT-3' # 5:MET14-Oti65'-GCTCTAGAAGTAATATTAGTAGCCAT-3' #6:MET14-o!i175'-GCGGATCCTGTTCATGATTTCCGAAC-3' # 7:Mona-o)i235'-GCGGATCCGGAGTCGGTACTAAATATC-3' #8:Mona-o))205'-GCGGATCCGAAAGGTGGCCTATC-3' # 9:Mona-o))295'-GCGAATTCCCATCGCAAACTGGGG-3' #10: Mona-oti305'-GCGGATCCGTCGAGAACATCTGCG-3' #11: Mona-oli31 5'-GCTCTAGAGCTTGTGCACTGGAAC-3' #12: Monaeti32 5'-GCAAGCTTCTCTCATGGAATCCTG-3' (vii) Construction of a deletion allele the S. carlsbergensis MET14 gene A deletion allele the S. carlsbergensis-specific MET14 was made by PCR, taking advantage of the sequence information obtained from sequencing of the MET14-CA gene.

Two pieces of DNA was synthesized from the 3'-and 5'-ends of MET14-CA a gap of 208 bp, covering the most downstream part (position-85) of the promoter and the first part (position +124) of the open reading frame. The two pieces were ligated pUC19 in two steps. The 5'-end fragment of the gene (348 bp) was inserted as an EcoRl-BamHl fragment into pUC19 digested with the same enzymes. This resulted inplasmid pPF55.

The 3'-end piece of DNA was inserted as an Xbal-Hindlll of 761 bp into pPF55 digested with the same enzymes, giving rise to plasmid The entire deletion allele of MET14-CA be released from pPF56 using EcoRl and Hindlil. The EcoRl-Hindlll fragment from pPF56 was blunted andsubcloned pCH216 digested with EcoRV, re- sulting inplasmid (Fig. 5b). pPF58 can be linearized by a partial digest with Xmnl, which cuts once in the coding region of MET14-CA once in the vector sequence, and thus be used for integration directed to the wild-type MET14-CA gene.

(viii) Use of the MET14-CA to inactivate the S. carlsbergensis-specific MET14 alleles in the allodiploid C80-CG65 and C80-CG110 To obtain a genetically completely brewer's yeast with no sulphite the S. carlsbergensis-specific MET14 alleles be disrupted in a similar way as described above for the S. cerevisiae-like MET14 allele. deletion allele the S. carlsbergensis- specific MET14 on plasmid (see above, Fig. 5) was used to disrupt the corre- sponding wild-type MET14-CA on the chromosome of the strains PFJ448 (C80- CG65 derivative) and PFJ467 (C80-CG110 derivative).

(ix) Evaluation of brewing performance and sulphite productionof met14 mutants of S. carlsbergensis yeast by fermentation at 50 L scale Methionine auxotrophic allotetraploid were constructed by hybridization of individ- ual allodiploid met14 of opposite mating types (see the Materials and methods section). Thus, a and a strains were crossed, e. g. PFJ459 was crossed with PFJ530, re- sulting strain PFJ501 and PFJ463 was crossed with PFJ530 to obtain strain PFJ514. A total six met14 tetraploid were constructed, namely PFJ502, PFJ506, PFJ509, PFJ510 PFJ514 (see Table 1. 1). Fermentation of brewer's wort with strains PFJ501 and PFJ514 were performed at 50 L scale the production strain M204 a reference strain. All were tested in three successive fermentation generations. The production of sulphite and the attenuation were followed the fermentation and after bottling, the beer was analyzed for aroma compounds.

Sulphite was markedly reduced during fermentation compared to the reference strain, but not completely eliminated (data not shown). Though, after bottling no measur- able sulphite was present in the beer from strain PFJ501 and PFJ514 compared to 8-13 ppm sulphite in beer from the reference strain (Table 1. 3). The attenuation from PFJ501 and PFJ514 was slightly compared to the reference strain, although three strains reached approximately the same gravity level the end of fermentation (Fig. 7).

This occurred in spite of the fact that the cell was usually somewhat lower for PFJ501 and PFJ514 as compared to M204 during fermentation (Fig. 8). The bottled beer was assessed by a trained taste panel and after forced aging. Before aging, the non-sulphite were found to be comparable the reference beer, but after forced aging (7 days at 37°C) thenon-sulphite were heavily oxidized,while reference beer was still This correlates nicely the t2n measurements of the low-

sulphite before and after aging, where a 3-5 fold increase were seen compared to the reference beer (Table 3). Another significant difference between non-sulphite beer and the reference beer was that after bottling the acetaldehyde of the non-sulphite beer was only 50% relative to that of the reference beer (Table 3).

Table 1. 3. Levels of acetaidehyde, sulphite andtrans-2-nonenal inbottled from fer-- mentation with non-sulphite in 50 L pilot scale. *trans-2-nonenal measured (ppb) before and after forced aging at37°C for 7days M204 PFJ501 PFJ514l.2. gen 3. gen 1. 2. gen 3. gen l. 2. gen 3. gen Acetaldehyde, ppm 4. 01-2. 04 1. 75-1.05 1.09 Sulfite,ppm 13 11 8 0 0 0 0 0 0 trans-2-nonenal,fresh030.05 0.070.05.-0.07beer*0. 04 0. trans-2-nonanal, aged'0. 04-0.17 0. 16-0. 30 0.3914-0.

1.4. Discussion By Southern analysis was verified that S. carlsbergensis yeast contains two divergent alleles the MET14 This is in accordance with several earlier observa- tions, as reviewed by Kielland-Brandt etal. , These alleles designated MET14- CE (S.cerevisiae-like) andMET14-CA (S.carlsbergensis-specific). TheMET14-CE in al- lodiploidof brewing yeast was inactivated by replacing the wild-type alleles with inactive deletion alleles. TheMET14-CA in the same strains were inactivated by selection for UV-induced mutants resistant to selenate. auxotrophic met14 mu- tants of tetraploid yeast could constructed by crossing of diploid met14 mu- tants, such strains were found to revert to prototrophy at a low Thus, it was de- cidedto isolate MET14-CA order to inactivate this gene by recombinant gene tech- niques. Furthermore, the inactivation of MET14 in this yeast revealed that it har- bours two copies of each divergent allele. S. carlsbergensis-specific (see e.g. Hansen and Kielland-Brandt, ; Hansen et al. , have been isolated by com- plementation with an S. carlsbergensis genelibrary 1986), though we were not able to find MET14-CA bycomplementation an S. cerevisiae met14 strain. Instead, the MET14-CAwas amplified PCR under low-stringency using oligo- primers based on the S. cerevisiae MET14 and template DNA from S. mona- censis. Threeindividual fragments were isolated in order to minimize the impact of

any errors in thenucleotide introduced during PCR amplification. the sequence information, an inactive deletion allele ofMET14-CA constructed.

The reconstituted brewer's yeast strains PFJ501 and PFJ514 that were inactivated in all four copies of MET14 the combination of recombinant genetics and UV-mutagenesis) were tested at 50 L scale. strains showed a final comparable to the refer- ence strain M204, although the number of cells suspension was a little lower during fermentation as compared to the reference. There are no serious faults with respect to technical abilities of the two strains. At bottling no sulphite detectable in the beer, even though low of sulphite were detectable during fermentation (1-3 ppm). This could be due to a leaky MET14-CA or to reduction of thiosulphate might take place in wort (Thomas et al., 1992)).

The fresh beer from strain PFJ501 and PFJ514 was assessed as"satisfactory"or"not quite satisfactory". However, after forced aging (7 days at 37°C) beer was generally assessed as being"not satisfactory"due to heavy oxidation. This corresponds nicely with the found values of t2n. Initially, thelevels of t2n PFJ501, PFJ514 and M204 were comparable, but after forced aging a dramatic increase in t2n content was seen in the non-sulphite Another difference between PFJ501, PFJ514 and the reference strain is the acetaldehyde in the bottled beers which was reduced by about 50% in the non-sulphite This is presumably due to the absence of sulphite during fermentation, allowingfor degradation of more acetaldehyde the yeast.

Astrategy for use of the low-sulphite strains for beer production would be mix- ingof two individual batches of beer, one produced using the low-sulphite and the other using a high-sulphite yeast strain. A yeast strain with the latter character- istics have been constructed and described by Hansen and Kielland-Brandt This strain of brewer's yeast was inactivated in all copies of MET10 (encodingsulphite re- ductase) resulting in highly increased sulphite (60-70 ppm). Under normal cir- cumstances, the amount of sulphite is often not satisfactory and varying from brew to brew. This strategy would result in beer produced with a predetermined sulphite content, sufficient to stabilize the beer without the need for sulphite as an external addi- tive.Due to the reduced production of taste active sulphur compounds(H2S-derived a wild-type),beer produced with the low-sulphite yeastwill a very weak sulphur-profile, and therefore be more"taste-neutral". makes the low-sulphite an appropriate starting point for construction of a yeast strain for production of a"basic beer", which in

turn can be combined with different beer batches containing flavour compounds of inter- est.

EXAMPLE2 Identificationthe genes responsible for reduction of dimethylsulphoxide todimethyl su ; phide and construction of yeast strains inactivated in suchgenes 2. 1. of the experiment Dimethylsulphide, is a sulphur compound of importance for the organoleptic proper- ties of beer, especially somelager Production of DMS during beer production oc- curs partly during wort production and partly during fermentation. Methionine sulphoxide reductases (EC 1. 8. 4. 5 and 1. 8. 4. 6) are enzymes responsible for reduction of oxidized cellular These enzymes have been suggested to be able reduce dimethyl sulphoxide as well, DMS as product.

A gene coding for an enzymatic activity leading methionine sulphoxide reduc- tion in Saccharomyces has recently identified.

Itconfirmed that the Saccharomyces cerevisiae reading frame YER042w in- deed appears to encode an MetSO reductase, and the name MXR1 proposed for the gene. An mxr1 mutant does not have the ability to reduce DMSO to DMS but appears to have unchanged fitness under laboratory conditions.It suggested that dis- ruption of MXR1 brewing yeasts would neutralize the contribution of the yeast to the DMS levels in beer.

2.2. Materials Methods Strains of bacteria and yeast and microbiological methods The following strains were used : S. cerevisiae S288C(MATa mal gal2), M3750 (MATa SUC2 mal mel gal CUP1 ura3),M1997 (MATa SUC2 mal me/gal CUP1), JH465 (MATa SUC2 mal me/gal CUP1 YER042w : : URA3), S.carlsbergensis M204(Carls- berg production strain), S. monacensis CBS1503, and C80-CG110 seg- regants of Carlsberg strain 244 (M204), Gjermansen and Sigsgaard, 1981).

Escherichia coli DH5a (Gibco was used for selection propagation of plasmid DNA.SC (synthetic compiete) SD (synthetic dextrose) were prepared as described by Sherman (1991). MP medium was identical to SD medium, except that yeast nitrogen base without ammonium sulphate was used, and 2 g/L of proline included to serve as nitrogensource. YPD medium contained 1% Bacto yeast extract, 2% Difco Bacto peptone and 2% glucose. B medium was prepared according to Cherest and- Surdin-Kerjan(1992). Brewer's wort had a gravity of 14. 5 % Plato. was grown at roomtemperature, unless otherwise indicated. S. cerevisiae transformed essentially accordingto Schiestl Gietz (1989).

DNAmanipulations Plasmid was prepared from E. coli according Sambrook et al. or using Qiagen maxiprep columns (Qiagen Inc. ). manipulations were performed according to manufacturers of enzymes (Boehringer Mannheim, Promega or New England Biolabs).

Polymerase reaction (PCR) was performed using Amplitaq polymerase (Perkin Elmer)according to the manufacturer.

Fermentations for assaying DMS production Yeastwas inoculated fromfreshly grownplate cultures 200 ml medium in 500 ml polypropyleneErlenmeyer flasks. proceeded through a 10 day period at 30°C rpm shaking) in the flasks with fermentation locks water.

Samples were taken through the side of the flasks syringes into evacuated blood samplingtubes (Sigsgaard and Rasmussen, 1985). DMS was measured using gas liquid chromatography.

Southern analysis Genomicyeast DNA was treated with restriction enzymes and separated on 1 % agarose gels,to Hybond-N membranes, covalently to the membranes by UV irradiation,and hybridized to 32P-labelled (random priming) at low stringency (50°C,in 1. 0 x SSC) essentially to Sambrook et al. Signals were recorded in aPhosphorlmager Dynamics Inc.).

2.3. Results Discussion (i) Sequence alignments No DNA sequences with homology tomicrobial reductases (Escherichia coli and Rhodobacter sphaeroides) found within the open reading frames of the complete- genome of S. cerevisiae. when looking for a homologue peptide methionine sulphoxide (PMSR's) from mammals, plants and bacteria, S. cerevisiae ORF YER042w showed high homology, visualized in Fig. 9, employing the Macaw program (Schuler etal. , Obviously, there appears to be different families of PMSR's, the yeast homologue very similar toall other chosen PMSR's (E. coli, Bostaurus, Arabidopsis thaliana andFragaria ananassa) the majority of its length. It wasconcluded that yeast ORF YER042w was a likely for the gene for yeast PMSR, and the working hypothesis was that the same gene would be responsible for the enzymatic activ- ity leading to reduction of DMSO to DMS during wort fermentation.

(ii) Disruption of ORF YER042w Itdecided to disrupt ORF YER042w in S. cerevisiae M3750. The one-step inte- gration approach of Rothstein (1991) was employed to remove the middle of YER042w and to insert a functional URA3 thus deleting disrupting the ORF at the same time. Oligonucleotide primersMSR1 (SEQ:ID NO (sequence 5'- GGTAAAGCTTGGCGAGTCGAGAAAGGAAATC-3', coveringnucleotide to- 683 relative the start codon of ORF YER042w and containing at the 5'-end 4 arbitrary nucleotides [toallow enzyme cleavage by an Hindlll site) and MSR2 (SEQ ID NO: (sequence 5'-GGTATCTAGAATCGATGGTTTTTGAAATAA- GCGACGAC-3', covering nucleotide +3 to +24 and containing at the 5'-end 4 arbitrary nucleotides by an Xbal site and a Clal were used to PCR amplifya 753 bp DNA fragment of ORF YER042w using S. cerevisiae genomic DNAas template the reaction.

Oligonucleotide MSR3 (SEQ ID NO: (sequence 5'-GGTAGGATCCCATTAT- CTGAGAGAAATGTAG-3', coveringnucleotide +535 to +555 and containing at the 5'-end 4 arbitrary nucleotides followed a BamHl and MSR4 (SEQ ID NO: 16) (sequence 5'-GGTAGAATTCGTCGCCTGGTTAAAGGCTAAC-3', coveringnucleotide po- sitions +1233 to +1253 and containing at the 5'-end 4 arbitrary nucleotides followed an

EcoRl site) were used to PCR amplify 739 bp DNA fragment. The 739 bp DNAfragment was restriction digested with BamHl andEcoRl the cut DNA purified on an agarose gel. this process it was discovered that a previously overlooked EcoRl site was present at position +1014 to +1019. This had the effect that the size of the digested DNA fragment was decreased to 489 bp.

This fragment was ligated a pUC18 (Yanisch-Perron et al. , that had been digested with BamHl andEcoRl purified on an agarose gel, creating pJH118. 753 bp DNA fragment was digested with Hindlll andXbal andligated into pJH118 with the same enzymes, thus creating pJH119. S. cerevisiae URA3 gene from plasmid (Botstein et al. , was cut out with BamHl andClal, and ligated the BamHl andClal sites present between the two YER042w frag- ments in pJH119, creating pJH121.

The URA3-containing cassette was cut out with Pvull purified on an aga- rose gel. S.cerevisiae strainM3750 grown in 50 ml medium to early exponential phase (OD600= 0), and transformed with 10 ug disruption cassette DNA. The transformation mixture was spread onto SC-ura plates allowed to grow for 5 days.

Several hundreds uracil-prototrophic were obtained, of which ten were pure- cultivated. DNA was prepared from these clones, and analytical employing oligonucleotide primersMSR1 andMSR4 performed. A DNA band of about 3. 0 kb size was seen from all and a band of about 2. 0 kb from strain M3750.

These sizes were those expected for disruptants and wild-type yeast, respectively. One disruptant, designated JH465, was chosen for further experiments. The disruption strategy and analytical PCR of strain JH465 is depicted in Fig. 10.

Asample of strain JH465 was deposited on July 1998 in accordance with the Budapest Treaty with the American Type Culture Collection (ATCC) under the Accession No. 74452.

(iii)Phenotype of disruption mutant EthSO resistance Ethionine sulphoxide is a putative analogue of MetSO. Strains M1997 JH465 were both applied in water suspension onto an SC or an SC-met (SC medium without methionine) plate a pre-formed ethioninesulphoxide-gradient 11). After

four days of growth at 30°C aclear zone of about 22 mm proximal the filter strip was observed with strain M 1997, inhibition which was not seen when methionine was present in the medium (SC). Strain JH465, on the other hand, was clearly re- sistant to EthSO on SC-met ; very good growth was found up to 17 mm from the filter strip, and some growth was visible to 5 mm from the strip. Under the conditions chosen, the enzymatic product of ORF YER042w probably converts EthSO into ethionine, a methio-- nine analogue that is toxic to most organisms, including Saccharomyces (Singer et al., 1978).

The same experiment was performed on MP plates (minimal containing proline as the only source). Here, EthSO seemed more toxic to strain JH465, indicating that an alternative system for reduction of EthSO is induced under these conditions (data not shown).

(iv) Strain fitness Strains M1997 JH465 were grown from an OD6ooof 0.15 YPD or SD media and followed byOD600 until early phase. There was no difference between the growth abilities the two strains in either media (Fig. 12). The same strains were streaked onto solidified YPG (rich medium containing glycerol carbon source) and NF media (minimal containing glycerol asonly source). There was no appar- ent difference in colony after 1 week of growth at 30°C YPG, while NF there seemed to be a slight in colony of strain JH465. It concluded that the putative peptide methionine sulphoxide encoded by ORF YER042w is not vital for growth under laboratory conditions.

(v) Growth with MeSO assulphur source Strains M1997 and JH465 were both applied as 20 u ! of suspension onto a plate withsulphur-free medium (B) containing at its center a filter with 200 NI1 M methioninesulphoxide (MetSO).After three incubation at 30°C strains could seen to grow equally well the inner part of the MetSO gradient. The experi- ment was repeated using liquid B medium that contains no sulphur source.M 1997 and JH465 were grown for 12 hours with 0. 02 mM L-methionine, washed in water and starved for 12 hours in B medium without any sulphur source at a dilution toOD600= 01. During this period, the OD600 to 0. 13, most probably to the endogenous glutathione

stock (Elskens etal. , At that point each culture was divided into three aliquots. To the first aliquot added 5 uM to a second aliquot 5 uM L-methionine- sulphoxide to the third nothing was added, and OD6oowas for 9 hours. Dur- ingthis period the density of the wild-type culture methionine as sulphur source in- creased by a factor 3. 1, while culture of the same strain with MetSO increased by a factor 3. 4. The same factors for strain JH465 were 2. 9 and 2. 6, respectively. The experi-_ ment was repeated with basically the same results.

Whileis quite difficult explain why MetSO should be a better sulphur source than methionine, it is also that it is of some advantage for the yeast to have an active ORF YER042w, when it comes to utilization methionine sulphoxide. fact that strain JH465 do grow quite well MetSO both on solid in liquid media indicates that ORF YER042w is not the sole activity capable converting MetSO into methionine.

These data and the fact that EthSO is still slightly for JH465 (see above) could be explained one assumes the existence of a low-affinity for reduction of both com- pounds, and that this system has not been affected by the inactivation of ORF YER042w.

This is in accordance with the results of Moskovitz et al. who showed that a yeast strain disrupted in the same gene had lost only 33% of its reductase activity against free MetSO. Possibly, system could the one that seems to be active in EthSO reduction under nitrogen-limited (MP medium, above) which should then have a very lowactivity under nitrogen-abundant conditions.

(vi)Production of DMS from DMSO by strain JH465 The strains M1997 and JH465 were inoculated in liquid SD or MP medium containing 0, 1.0, 10 or 100 mg/L DMSO or in autoclave fresh wort. Inoculation performed in 200 mi in 500 ml polypropyleneErlenmeyer flasks with fermentation locks and left grow for 1 week at 30°C. 10ml were taken and the DMS content was measured. As can be seen from Fig. 13, quite significant amounts of DMS were formed in the fermentations with strain M1997 DMSO had been added although only about 0.8% of the substrate was converted on a molarity The same pattern could seen using MP medium, but here about 1. 4% was converted. The rather low ratio of DMSOto DMS by this enzyme (4% in SD and 14% in MP medium) is in agreement with earlier (Anness, 1980 ; Anness and Bamforth, 1982). No conversion at all was performed by strain JH465, except for 0. 001% MP medium with 100 ppm DMSO.

The last observation supports the notion that an alternative system that can reduce com-

pounds as MeSO, EtSO DMSO does get active under certain nitrogen-limiting condi- tions. However, this putative system can certain explain the rise in DMSO reduction by the wild-type when going from SD to MP medium (almost a doubling the con- version ratio). This effect is rather an induction of the system encoded by ORF YER042w itself.

Brewer's wort assumingly a natural content of DMSO, either autoclave fresh from production, was also with these strains, and the mixture allowed to grow under the same conditions as described above. A significant amount (0. 02 ppm) of DMS was formed by strain M 1997 fromautoclave wortwhile DMS resulted from fermenta- tion with the mutant strain in this medium. When using fresh wort, 0. 169 ppm DMS was formed from the wild-type and only 5. 3% less the mutant yeast. The increased level DMS probably reflects the endogenous level ofDMS in brewer's wort. It should be noted that S. cerevisiae not ferment brewer's wort very well, conse- quently these data do not reflect the situation in beer production.

It wasconcluded yeast ORF YER042w is responsible the enzymatic activity lead- ing to DMS formation from DMSO. As it was furthermore verified that YER042w is re- sponsible for at least part of the capability of S. cerevisiae reduce methionine sulphox- ide, the name MXR1 (peptide Methionine sulphoxde Reductase is suggested for this gene.

(vii) Occurrence of a gene sequence homologous S. cerevisiae YER042w in a production strain of S. carlsbergensis To verify the theory that a gene, or set of genes, in S. carlsbergensis yeast corre- sponding to MXR1 is for DMSO reduction during wort fermentation, the con- tent of such sequences in this yeast was analyzed. Southern analysis at low of digested genomic DNA from S. carlsbergensis, S.monacensis, S.cerevisiae of two spore-clones an S. carlsbergensis yeast is shown in Fig. 14. The DNA, cut with Bglllhybridized with an S. cerevisiae MXR1 at low to possibly allow the detection of homoeologous present in the brewing yeast. A lower band of 2. 7 kb was observed in all except from S. monacensis, while upper band of about 4.0 kb was observed in all except S. cerevisiae. even higher molecular weight (and somewhat fainter) band was observed for all except S. monacensis may represent a partial digestion of the MXR1 As 2. 7 kb is the expected size for an

MXR1-containing Bgfl 1-fragment 10), and as the 2. 7 kb bands were also present on high-stringency Southern hybridizations using the same filter not shown), these bands supposedly representsMXR1 (S.cerevisiae) a gene basically identical this, MXR1-CE (S.cerevisiae-like MXR1) C80-CG65 and C80-CG110). bands of 4.0 kb were not seen in high-stringency Southern hybridizations (data not shown) and these bands probably represent a gene somewhat diverged from but functionally analo-- gous to MXR1. putative gene (present in Saccharomyces monacensis CBS1503, S. carlsbergensis M204, and C80-CG110) designated MXR1-CA S. carlsbergensis-specific MXR1 S. carlsbergensis a species hybrid a genome which is originally derived from two Saccharomyces and two homoeolo- gous alleles a given gene is usually found in this hybrid organism (for a comprehensive description of the genetics of S. carlsbergensis yeast see Kielland-Brandt etal., 1995). The results presented here is in perfect concord with this.

(viii) Conclusion To conclude, it is suggested that S. carlsbergensis yeast has two homoeologous genes for DMSO reduction, MXR1-CE and MXR1-CA, that these two genes are re- sponsible the activities reducing DMSO to DMS during primary fermentation of brewer's wort.

To summarize, it was discovered that S. cerevisiae YER042w is responsible the enzymatic activity that reduces DMSO to DMS, at least under nitrogen-rich conditions.

There may, however, be another system able tofulfil same task under certain nitro- gen-limiting As the native function of ORF YER042w supposedly is reduction of peptide methionine sulfoxides, name MXR1 (peptide Methionine sulphoJ6de_e- ductase) is suggested for this gene. The S. carlsbergensis yeast appears to con- tain a gene almost identical toMXR1, MXR1-CE in addition an analogous but di- verged gene, MXR1-CA. a S. cerevisiae without MXR1 does not seem to loose or vitality underlaboratory it is contemplated that the inactivation ofthese genes in the brewing yeast is a suitable of decreasing the amount of DMS formed during primary beer fermentation.

EXAMPLE 3 Construction of mutants of brewer's yeast with altered of hydrogen sulphide : In vivofor a direct link formation of H2S certain thiols andthioesters 3. 1. of the experiment Hydrogen sulphide a brewing yeast-derived sulphur of great importance for the taste and flavour beer. While provides some of the character of very young beer and may be desired in low to mask the impression of certain other flavour com- ponents, it is in general at higher concentrations in beer. Furthermore, certain other taste-active sulphur components,like thiols thioesters, may be derived from hy- drogen sulphide. Here it is shown that it is possible, classical breeding methods, to construct well-performing yeasts with an altered of hydrogen sulphide.

Furthermore, evidence was found for a direct metabolic link formation of hydro- gen sulphide and of ethanethiol, methylthioester andethylthioester.

3.2. Materials and methods Strains of yeast and microbiological methods The following strains were used in the experiments : M204 (Carlsberg production strain), C80-CG65 (Mat a) and C80-CG110 a) (spore segregants of strain 244 (M204), see Gjermansen and Sigsgaard, 1981). JH 441, JH442, JH443, JH444, JH506, JH515, JH516, JH517 are all in this study. SC (synthetic complete) and SD (synthetic dextrose) were as described by Sherman (1991). MP medium is iden- tical SD medium, except that yeast nitrogen base without ammonium sulphate was used, and 2 g/L of proline wasincluded serve as nitrogen source. YPD medium con- tained 1% Bacto yeast extract, 2% Difco Bacto peptone and 2% glucose. YP- galactose is identical to YPD medium, except that glucose has been substituted with galactose carbon source. BIG-YNB was prepared according to Rikkerink et al. as modified by Thomas et al. and is basically synthetic complete medium containing sodium sulphite ammonium bismuth citrate. Brewer's wort had a gravity of 14. 5 % Plato. was grown at room temperature, unless indi- cated.

Mutagenesis UV mutagenesis performed by irradiating Petri dishes (YPD) containing spread yeast cultures. The time necessary to ensure 50% killing determined empirically found to be about 6 seconds for strain C80-CG110. used for UV irradiation was two PhilipsTUV 15W, G15T8 tubes in a steel cabinetplaced 20 cm above the Petri dishes irradiated. After irradiation, the plates were kept in the dark for 24 h.

Mating of yeast and screening for mating products S. carlsbergensis segregants of different mating type were mated by overnight co- inoculationthe surface of a YPD plate at20°C, according to Gjermansen and Sigsgaard (1981). Mating products were enriched for by inoculating tiny amount of the cellinto liquid rich medium containing as the carbon source galactose and Kielland-Brandt,After 6 days of growth at 20°C of the culture was streaked outto single colonies onsolidified medium containing galactose carbon source.

Large and fast-forming colonies were picked and streaked in replica-pattern onto YPD plates. 4 days of growth these plates werereplica-plated tosporulation and the plates left at20°C 8 days. The plates then illuminated UV light 302 nm (Brizaet al. , and emission of visible light taken as evidence for spore forma- tionand such clones were assumed to be mating products.

Random spore isolation Brewingyeast was sporulated streaking in a thin layer solid sporulation and the plates at 20°C 6 to 14 days as necessary for spore formation. Spore formationin the culture was detected by malachite green staining and Fulton, 1933; Gjermansen and Sigsgaard, 1981) and the culture enriched for ascospores and these isolated to Gjermansen and Sigsgaard (1981). Spore clones were identi- fiedby the lack ability to emit visible light growth on sporulation and illu- minationwith UV light 302 nm. Spore clones of mating type a was identified by replica platingstreaks or colonies of the spores in question onto a YPD plate a grow- ing"lawn"of strainC80-CG110 a). After 24 h growth at 20°C thisplate further replica-plated ontosporulation and after one week of growth sporulation visu- alized with UV light) the involved spore-clone as one of mating type a.

Fermentation experiments Yeast strains were propagated in brewer's wort at 13°C successive inoculations of larger larger volumes of sterile brewer's wort : Ten mi inoculated with 106 yeast cells; 7 days this culture was diluted with 300 ml which after yet 7 days were poured into 8 litres of wort. After fermentation for 12 days (with occasional stirring), the- cultures were cooled to7°C the yeast was allowed to settle for two days. Fourteen grams of wet yeast was then added to 2 litres of aerated brewer's wort in an EBC tube (EBC-Analytica Microbiologica,The temperature during the first 7 days of the fer- mentation was 13°C which it was lowered to7°C kept there for the last h to allowyeast to settle to the bottom of the tubes. Samples were taken anaerobically each day through silicone Fermentation and sampling performed auto- maticallyin the Multiferm as described by Sigsgaard and Rasmussen (1985). All experiments were performed in duplo.

Chemical other analyses Yeast cell were performed using a Coulter Counter model ZB.Residual ex- tract of the beer was determined using gradient tubes (Atkin et al. , Higher alcohols, esters, acetaidehyde, diacetyl, and dimethylsulphide measured using gas liquid chromatography.SO2, methylthioacetate, ethanethiol andethylthioacetate were measured in a Purge-and-Trap system using a chemiluminescence sulphur detector.

3.3. Results and Discussion (i)Construction of mutants of tetraploid brewing with decreased production of hydrogen sulphide The level not only but also in brewer's wort may influence the sul- phur metabolism the brewing yeast. Ramos-Jeunehomme et al. showed that threonine is taken up by the yeast rather much before methionine. That could possibly ex- plainthe high production of hydrogen sulphide seen in the early of the main fermentation (Nagami et al. , ; Takahashi et al. , ; threonine exerts feed-back control its own formation at the point of conversion of aspartate (see Fig. 15). This means that a high threonine concentration will formation of O-acetylhomoserine, the binding partner for hydrogen sulphide. If the same time, only amounts of

methionine is presentin thecell therewill a high of hydrogen sulphide. Only after the threonine has been used for further metabolism will yeast cell able to util- ize hydrogen sulphide surplus. It been shown that yeast mutants that are resistant to hydroxynorvaline, toxic threonine analogue, has an aspartokinase (EC 2. 7. 2. 4) that is insensitive to threonine feed-back inhibition (Ramos and Calderon, 1992).O-acetyl homo- serine formation in such mutants is consequently no longer controlled by the level of- threonine.

As these mutations are dominant, it was decided to employ this as a strategy to make mutants of brewing yeast that would produce less than normal hydrogen sulphide. The brewing yeast was found to be quite insensitive towards hydroxynorvaline. In con- taining ammonia ions hydroxynorvaline to 100 mM not toxic. However, in MP me- dium, where the only source is proline, 100mM hydroxynorvaline toxic.

Strain M204 was grown in MP medium to stationary phase at 20°C 50 pi this cul- ture added to 10 ml MP medium containing 100 mM hydroxynorvaline. one week of growth at slow shaking,only very week growth could seen. 50 NI this cul- ture was reinoculated yet 10 ml with hydroxynorvaline growth was continued for one more week. After this week significant growth could seen and after one more in- oculation culture was dense after 48 hours. Cells this culture were spread onto solid medium, and single clones assayed for hydrogen sulphide production by applying droplets of yeast in water suspension onto BIG-YNB plates. sulphide has the property of easily forming an insoluble brownsulphide the bismuth in the BIG-YNB plates et al. , ; Thomas et al. , The developed colour was observed after 8 days of growth. The outcome of this experiments was the mutant strains JH439 through JH452 and JH489 through JH505. The ability of these strains to tolerate increased concentrations of hydroxynorvaline confirmed by re-inoculation each strain, along strain M204, the medium used for the selection. The slightly lower hydrogen sulphide production on BIG-YNB plates a selected range of these strains can be seen in Fig. 16.

Asample of S. carlsbergensis JH441 was deposited on July 8, 1998 in accordance withthe Budapest Treaty with the American Type Culture (ATCC) under the Accession No. 74451.

(ii) Selection of allodiploid of S. carlsbergensis yeast with en- hanced production of hydrogen sulphide The BIG-YNB plate for hydrogen sulphide was employed directly a screen for S. carlsbergensis producing increased amounts of hydrogen sulphide.

An exponentially growing culture of the allodiploid S.carlsbergensis strainC80-CG110- was diluted spread onto BIG-YNB plates a concentration suitable obtaining about 1, 600 colonies perplate. yeast cells 50 such plates mutagenized using UVirradiation for a time period sufficient to ensure about 50% killing. it was ex- pected to obtain about 800 surviving colonies on each plate. Theplates wrapped in aluminiumand the yeast left to grow and develop colour for 8 days at 20°C. the 8 days, 4 colonies had developed a brown colour darker than the rest of the colonies. Yeast from these colonies were pure-cultivated by re-streaking on YPD medium.

The four mutant yeast strains were designated JH341, JH342, JH343 and JH344.

(iii)Construction of allodiploid of mating type a the same genetic trait enhancing hydrogen sulphide production A similar was followed in an attempt to isolate the same type of mutants of the strain C80-CG65, an allodiploid S.carlsbergensis of mating type a. However, no mutants were found. As an alternative strategy it was decided to transfer the mutations to yeasts with the opposite mating type by mating of the four original mutants with C80- CG65, followed by random spore isolation and screening for colonies of mating type a withdark brown colour onBIG-YNB Each of the four mutants were mated with C80-CG65 and mating products were selected as described in the Materials and Methods section. The four yeast strains that resulted were designated JH477, JH478, JH479 and JH480, respectively.

Toconfirm the tetraploid and ability to sporulate these strains they were streaked onto sporulation After incubation for 8 days at 20°C formation, showing up as fluorescence when exciting the cells UV light 302 nm (Briza et al., 1986), was taken as evidence for spore formation. After sporulation ten days on solid sporulation at 20°C, followed random spore isolation, petri dishes contain- ingthe spores were replica-plated ontoBIG-YNB and the yeast strains were al- lowedgrow and develop colour for 7 days. Dark brown colonies were pure-cultivated and lack dityrosine formation on sporulation along with ability mate with

strain C80-CG110 taken as evidence that these strains were allodiploid of mat- ingtype a.

Eleven of such strains derived from JH477 and containing the genetic trait from strain JH341 were designated JH481, JH482, JH483, JH484, JH486, JH487, JH518, JH519, JH520 and JH521, respectively. Four strains derived from JH579 and containing- the genetic trait from strain JH343 were named JH522, JH523, JH524 and JH525, re- spectively.strain, derived from JH580 and containing the genetic trait from strain JH344 was designated JH488.

(iv)Reconstitution of brewing yeast with enhanced hydrogen sulphide production Toconstruct tetraploid yeast with enhanced hydrogen sulphide the followingcrosses were made : JH341 were crossed with JH482, JH483, JH484, JH486, JH518, JH519, JH520 and JH521 and hybrids selected good growth on YP-galactose plates. expected sporulation-proficient of these strains was confirmed as dity- rosine formation on sporulation was evident. The strains were designated JH506 throughJH517 and JH526 through JH540. JH343 were crossed with JH522, JH523, JH524 and JH525 and after confirmation of the tetraploid of these strains, they were named JH541 through JH560. The high-H2S phenotype of the tetraploid yeast strains was confirmed by applying droplets of aqueous suspensions of the different strains onto BIG-YNB and allowing growth for 4 days. Fig. 17 depicts the strategy forthe making of these mutants and Fig. 18 illustrates sulphide production of a selected range of these strains.

Asample of S. carlsbergensis JH506 was deposited on July 1998 in accordance withthe Budapest Treaty with the American Type Culture Collection under the Accession No. 74453.

(v) Evaluation strains with altered hydrogensulphide in Multiferm: Fermentation and sulphur profile Toassess the fermentation capabilities production of sulphur compounds by brewing yeast strains having altered of hydrogen sulphide range of these yeast strains was tested in trial in 2 L EBC fermentation tubes in Multiferm (Sigsgaard and Rasmussen, 1985). The outcome of these experiments are summarized in

Table 3. 1. An enhanced formation of 2-methyl-1-propanol seen with the strains JH 441- JH444 while ethylacetate andacetaldehyde formation especially increased with strain JH442. Apart from that the strains with decreased hydrogen sulphide appeared to perform quite normally fermentation : levels of S02 andresidual in the beer were quite normal yeast harvest and the amount of dead cells satisfactory val- ues. The high formation of ethylacetate andacetaldehyde strain JH442 suggests that-_ the mutation or mutations present in this strain differ (s) somewhat from the other strains.

Table 3.1. characteristics of strains of brewing yeast with enhanced or decreased production of hydrogen sulphide v 1 i v v. v. 1.'v ! lM204(1)19.4102 3.5 0. 11 31. 5 4. 8 0.270.150.06 8. 5 2. 7 9.520.21 1. 3 16. JH441 21. 3 1. 3 21. 9 112 3. 9 0. 15 34 3. 8 0. 24 0. 17 0. 06 4. 9 2. 8 10.5 8 JH442 23. 9 2 21. 5 101. 5 4. 9 0. 2 48. 2 7. 1 0. 25 0. 19 0. 08 8. 2 2. 8 9 21.3 JH443 21. 1 1. 3 19. 6 113. 9 4. 3 0. 15 36. 7 4. 6 0. 29 0. 19 0. 06 6 2. 7 10 2 JH444 20. 4 1. 2 19. 0 104 4 0. 14 35. 3 4. 7 0. 3 0. 2 0. 06 5. 8 2. 7 10.5 5 M204 (11) 14. 8 1 16. 1 93. 6 4. 7 0. 15 34. 3 4. 7 0. 17 0. 1 0. 05 8. 5 2. 8 13.3 19 JH506 15. 9 0. 9 27. 2 101. 1 3. 0 0. 18 26. 4 6. 8 0. 06 0. 05 0. 07 14. 2 2. 9 16.7 1 JH515 17. 1 1. 4 23. 8 91. 8 2. 6 0. 15 23. 7 8. 6 0. 07 0. 09 0. 07 12. 8 2. 8 17.5 5 JH516 16. 7 1. 2 24. 4 92. 2 2. 5 0. 15 23. 3 7 0. 07 0. 08 0. 07 12. 8 2. 8 15.8 1 JH517 16. 9 1. 4 23. 2 89. 6 2. 3 0. 13 21. 7 6 0. 07 0. 09 0. 07 13 2. 8 18. 3 23.6 !

Abbreviations: 1-Pro : 1-Propanol; : 1-Butanol; 2M1p: 2-Methyl-1-propanol; 3M1b: 3-Methyl-1-butanol; : 3-Methyl-butyl-acetate (isoamylacetate);! bac: Isobutylacetate; : Ethylacetate; Acal: Acetaidehyde; : Diacetyl; : Penta- dione; DMS : Dimethylsuiphide; : Sulphur ; Ext : Residual in the beer; DC : Dead cells the end of fermentation ; YH : Yeast harvest at the end of fermentation; Numbers are in mg/L except for Ext (%) and DC (X x 109 cells/L).

M204(I) to the reference strains tested together with strains JH506 and JH515-JH517.

As for the strains with enhanced hydrogen sulphide production (JH506, JH515, JH516 and JH517) also strains appeared to behave quite normally fermentation. Also these strains had an enhanced production of 2-methyl-1-propanol andalso thelevel of acetaldehyde wasgenerally increased.In a decreased formation of ethylacetate was observed and very low values the vicinal diketones(diacetyl pentanedione). A strikingfeature of these strains is a significantly enhanced sulphite production (SO2) : to 67 %). This, however, not unexpected, as sulphite is the precursor for hydrogen sul- phide.

The beer from these fermentation trials were also for formation of MeSAc and EtSAc. Fig. 19 depicts some of the results of these analyses. strains having en- hanced formation of hydrogen sulphide (JH506 and JH515-JH517) also a signifi- cantlyformation of MeSAc as well of EtSAc, while strains with decreased hydrogensulphide production (JH 441-JH444) had a decreased MeSAc production. This is evidence for a direct link formation of hydrogen sulphide the thioesters methyl-and ethylthioacetate.

(vi)Further evidence from fermentation at 50 L scale with met2A, met10A andmet14 strains : sulphur and taste evaluation Tofurther confirm the metabolic coupling hydrogen sulphide and the thioesters MeSAcand EtSAc, the formation of these compounds and EtSH strains of brewing yeast with well-described geneticalterations to elimination a large of hydrogensulphide production (Hansen and Kielland-Brandt, ; 1996b) was investi- gated. In of these strains all of the two types of the MET2 are inacti- vated.This has the effect that very high amounts of hydrogen sulphide are accumulated and secreted to the growth medium. The other strain has no MET10 As MET10

encodes part of the sulphite enzyme (Hansen and Kielland-Brandt, this strain cannot produce hydrogen sulphide atall. from 50 L fermentation trials (Hansen and Kielland-Brandt, ; 1996b) were subjected to analysis the above mentioned thiols thioesters. The results depicted in Fig. 20. Clearly, high hydrogen sul- phide production is associated with an increase in the production of all three sulphur compound whereas, when no hydrogen sulphide is produced, almost production of- these compounds takes place.

Thus, we have provided the first biological based on genetic alterations of brewing yeast directed towards up-or down-regulation of hydrogen sulphide production, of a direct metabolic link formation of certain thiols thioesters and the forma- tion of hydrogen sulphide.

(vii) Conclusions Using classical and cross-breeding mutant strains of S. carlsbergensis brewing yeast were constructed which had decreased or enhanced production of hydro- gen sulphide. These yeast strains were subjected to trial and turned out to perform satisfactory in most respects. Analysis sulphur components of the resulting beer showed that the yeast strains having decreased production of hydrogen sulphide also a reduced production of methylthioacetate the yeast strains with in- creased hydrogen sulphide production also increased formation of methyl-and ethylthioacetate. is genuine biological for a metabolic link hydro- gen sulphide and certain thioesters. Using brewing yeast strains with defined mutations resulting in either zero or extremely hydrogen sulphide production, this relationship could confirmed. These experiments furthermore showed that also of ethanethiol hydrogen sulphide dependent.

To summarize, it has been demonstrated that it is possible provide strains of brewing yeast by classical methods having altered production of hydrogen sulphide and inthis way change the production of certain thiols thioesters that are important for the organoleptic of beer.

EXAMPLE 4 Construction of Saccharomyces cerevisiae andSaccharomyces carlsbergensis strains having reduced or no production of higher alcohols acetate esters The purpose of this experiment was to construct brewer's yeast mutants with lowered pro- duction of isoamyl alcohol andisoamyl and to obtain finished beer produced by such mutants which has a reduced level higher alcohols their corresponding ace- tate esters or which does not contain such compounds. The following approaches were used for introducing modifications at two points in the pathway leading higher alcohols and their corresponding acetates : (a) Inactivation/modification the isopropylmalate in the leucine denovo syn- thesis pathway (encoded by LEU4, ORFYOR108W andTHI3 ORF YDL080c) (This Example); (b) Removal the alcohol acetyl activities encoded by ATF1-CE, ATF1-CA and ATF2 (Example 5).

Total removal ofisopropylmalate activity might not be desirable the Leu3p- isopropylmalate is needed for positive regulation of at least thefollowing : LEU1, LEU2, LEU4, ILV2 andGDH1 and Schimmel Hu et al. 1995).

4.1. Experimental (i) Materials andmethods In Tables 1 and 4. 2 are shown the genotypes of relevant S. cerevisiae S. carlsber- gensis used in this study. Agar plates synthetic complete (SC), rich high glucose mediaplates and defined minimal mediaplates complete (SC) and derivatives (SC without methionine (SC-methionine), SC without leucine (SC-leucine) for selection genetic studies were made up according to Sherman, Fink and Hicks (1979). Selection for LEU4 fbr was done on SC-leucine the addition of 1 or 5 mM5, 5, 5-triflouro-DL-leucine (SC-leucine TFL). 5, 5, 5-triflouro-DL-leucine 12890 was obtained from Lancaster Synthesis Ltd.

Separation of chromosomal length DNAmolecules done in aBIO-RAD CHEF- MAPPER system. Chromosomal was purified in FMC Incert according to Schwartz and Cantor (1984) with the modification by Pedersen (1988) and chromosomal length DNA molecules separated in FMC FastLane agarose. DNA sequences used for restriction maps and PCR amplifications were obtained from Saccharomyces Genome Database (http ://genome-www. stanford. edu/Saccharomyces/).

E. coli strain XL1-blue (#200236) fortransformation obtained from STRATAGENE and newly PCR fragments and DNA fragments and plasmids purified using a BIO-RAD DNA Purification Kit 732-6010. Yeast transformations were done with the lithium acetate method according Schiestl Gietz (1989) and Gietz et al. (1992).

The S. carlsbergensis MET2 allele designated herein as MET2-CA (Hansen and Kielland-Brandt, The pUC19 vector containing the 3. 7-kb Hindlll MET2-CA in- sert is designated pMBP96A2. The 3. 7 kb Hindlll MET14-CA-containing was subcloned the plasmid p19-6 and Kielland-Brandt, The plasmid pMBP96A2 was used for cloning of MET2-CA as basis for the YOR108W disruption plasmid.

Construction of strains carrying inactive forms of LEU4 done by the two-step deletion procedure by Scherer and Davis (1979) (loop-in/loop-out method). Inactivation of YOR108Wwas by insertion of a plasmid selected YOR108W sequences and the 3. 7-kb Hindlil MET2-CA as selective marker.

Transformants containing the LEU4 deletionplasmid the YOR108Wwere on selective SC-methionineplates. transfer of agarose gels done onto Schleicher &Schuell 0, 45 mm Ref.-No. 401196. Probes were labelle detected withthe DIG DNALabelling Detection Kit (Boehringer Mannheim 1 093 657). Restric- tion endonucleases obtained from either Boehringer Mannheim or Promega and used according to the manufacturer's specifications. PCR fragments were synthesised withthe Expand HighFidelity System (Boehringer Mannheim 1 732 650) according to specifications. PCR fragments were synthesised in a ROBOCYCLER Gradient 40 or 96 (STRATAGENE).The standard ROBOCYCLER program consists of the following steps and temperatures : 1 cycle at94°C 4 minutes, 20-25 cycles of 1 minute at 94°C, min- utes at 60°C, minutes at 72°C andfinally 1cycle at72°C 10 minutes before storage at 6°C.

Table4. 1. and sources of S. cerevisiae used in this study. TFLR = 5, 5,5-trifluoro-DL-leucine resistant s&aS ! !2S i ffi l ss-; sS ss is. s xm-3S ! : ! : : : H ! ! !i ! i !saM 2n. : :.. : :...P9 ?-s ; ssSM ! iRM !.S s ; ? ?..............-. : :.... g. -5 ! : : ; ; :H : !i ; K : ;ssi ! ! ! ! ! ss ! s ! ! ! ! ! s ! ! ! ! ! ! K ! ! ! ! ! ! ! ! ! ! ! s ! ! ! ! ! ! SssM s ! ;mK :ifl8 pa-sjs ; ;sjs' ! ! ! mss ! :mR ! ! N ! ! : H ! : ! ! ! ! ! ! ! ! ! ! ! M ! ! ! ! ! ! ! ! ! ! ! ! ! BS ! ! ! : n ! : ! : : i...., ;'K : ! ! ! ! ! ! H ! ! ! ! ! !KK ! ! ! ! : ! ! s ! ; ! ! ! M ! ! stt' j ! ! ! ! ! !3 ; KM sss ! !. ! ! !- ; : n K : ! ; ! ; ! s ?-i : a : ! ; s ! i : !$N ! ss ! ! K : ! ! s% !B ! ! ! ! K !s ; in ! ! !H ! ! !. !. ! ! ! ! ! !ss ! ! ! ! ! !sa ! ! s ! ; ! ! ! !s! :I4 : !.s : :K a :s ! :ms ! ! ! S ! ! M ! K !m ! ! ! ! ! !K ! ! ! ! mR3 ? ! ! ; ? ! ! : ssK ! amm ; ss K...... t--s--- -' : :.. : BPSi'H : : ! ! : ! : ! ; ! : ! ! ! ! ? ! ? s ! ! ! ! ; !MasKs j ; ! s ! !MBEgs ! ; ! ; sis ! ! ! i ! !K ! K ! ! ! ! i ! !i8 ! s ! ! sast Ks s ; ! ! 8 ! a !s ; Es ; g ; ;4m_~~n Ms ! M ! ! ! ! jmtM ; ! ! ! ! g ! 8 ! ! ! ! ! m ! ! M ! ! a s s j ! !x.... (.. ! ! !7 t... G !. 2. . tlI9'y ?', ;... :....... :. t a... . rsr...... :. :......::,, : a,, ;, ; :" met. 2 : : eLt4l.t. R 0& : :. 7 4 2 a............ ......... Y... > :. :..... : : : : : : : : : : : :.,.. : ,.... :. :. : : *,. ',7 a... Tha. s. s tu ....... P37 7 :' ::: : a....'' v :.... ; :,.. :.". : <' . a......ozoa ... : :.. a.. h : i. s4.... m .......e. :.'.. rro.. :.R s... :. :.. : : :, P 7. 9. a....::...... h: zs..... irt.,f. ... : R.... ......... d r' T. 9......... :.. :.. :.... hZS..... :............ s. ud......... :. : : ... !. ° :.......'..',,. ; v :::: : .: r::. :.::.. :: > : ;: 4:::. : ::::: 11 a......... 'O. i2Q8fr, r... T'Lt : N : :. I.:......,' :'i9' : 1,. : r... ; Iiil il : l' : : : iri : i9 :,..'.'........ illi : ili : i : it,."...,.'*., ri il-'iii'9i : *,"I,,, r,-, 9... 1, 1,',',', r,................. 9r9... r...... r...... r................. r. 1........ I...... ...... :. :...................... : : :. :. :. :... : : 2 : a... : : :, :. : t :"... :. :,... : ; : :. :.., : ;...' : : :. : : : 7 1 .s4 me 2d,................ ..... $p3 : ... : :. : :'" :' :..... : :.., : : <.. : > : : : : : : : : : : : : : : ? : : : > : : : : : :."r : :' : : : : : : w : : : : : : : < : : : : : : : : ; : : v : : ;.. :.. :. : :.. : : : : : : :7.. 1.... a............ i : : : : :,.. n-.. . ..... 3..... tu...Z.......................... .......... : .: V. 4. : :':::.,.,:: . : BOP9: g. : :. :... ; : ; : : : : : : : :'.' : :. :.,.., ; ; : : :... : : : : : : : : : : : : : ; : : ; : ; ; :. : ; : ; :. : ; : : ; : : : : :.. :. : : : : : : ; : : : ; : :. : : :...... :. :.. : 'J : B R....... : ...................... ...... .... sud.....::' :' :' :' :' :' :'' :' :. :' :. :' :.. r : : :. r. r r...:;:....° : _ ; :... :.... I. I............ I.......... II1 91 1 1 r.. i :''I Ir I 1 1 1 r I I r I r. r 0 : : :. : 9 : : : :' : : : r : :' : : : : : BP.. ,,.... : e' : v". °. : : :, : :. :.... :. :, : : :.. :...,.. : : :........9.'7 51... k, s m 2t.. :., ud or :8 0., 13 wr. MET2.-. : :....., : : : : : : : : :... :. : : : : :.... : : : : : : : : : : : : : : : : : :. :.. :.. ; :.. :. :. :.. .. st................, Y.... (y. :. :........... :. : :. : :. : : : : : : :. ;..... :. :. : : : : :": : : ; ;.... : : : : : : ; < : : ; : : : : : : : :' :. : : : : : : :. : ; ; : : : : :. : :,. : : : : : : : : :' : : : : : : : : : : :< : : :. : : : : : : : : : : : o 8.. _. r.. r.. r.. _ : _ :. :. r-.-. r.. r-r. r.-- I d. 3. l. z lrr... 4 ; t 8..., s..: 4. 13Q : ; : : : : : :..-.'. 4 9. 999 i-9. 183. !.. r. r. r... r. 0... ',: :...... : s m.Z. ;,............., ...., 1.....:. y >.. :.. : :., :.., g : : ;. : : :.. :..,.,. _, :,.,,. ; : : :....'v.. :,.. : : : : :.. : : ;P.. ? 4 e*...... *........ *... * *.... :, I ; : :, : :. :.... r,',,, *.............. I... ............. :, : : :. : : : :' : :' :. : :. : r : : : r :. : : :. : :. : : : : : ; : 1 ; : : :. :' : :.. : :... i : i : ii-r.-r9... 9.. r.':......, .Y Y... : :.. : '' : : :.. :.. : :..... : :-. :. : :. : :. :. : : :. : : : : : : : : : : : : : : : : : ; : : : : : : : : : ; : ; s : :. :.. :.. : :...., : : :.' : : : : P. . 7 : Ss :.. :. a h s rn2 :................... 1...................... ........ s. udTable2. Genotypes and sources of S. carlsbergensis used in this study stria ; train ii ; ! : ! ! ! NKs ! ; i ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! i ! ! ! ! s ! ! M !ss ! sss K : ! ! ! : ! ! ! ! ! ! ! ! ! ! ! ! Qa-ssss' ! : H. ! s ! !iH : ! ! ! ! ! ! ! ! ! ! ! ! ! s ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !ss ! s ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! s ! ! ! ! i § ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! M !aNsMs:.............i,, g gj ii$i ß stt. ar sb ... ........ ...... ........ .. : : :, :.,.,. : : : : : : : : : :... : : : : : : : : : : : : .... . .saMs gs Mg ! ! ! !,.. I. I. I.. 11. 1...s .. ... 2 6.. :. :..,. : : :::: :.......", ......-" :".,..,''., * mus I. : : .......... ...'-... a-. *a--r-.4 : ; : :. : : ; : : : : : : : : : :.. : : : : : : : : : : : : : :. : ; :.,.- ; : : : ; : ; : : : :. : : : : : : : ; : : ; : : : : : : : : : : : : : : : : : : : : : : ; : : : : : : : : < : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :, : : : : : : : : : ; : : : : : : : : : ; 3 <4, : :. ;.... :: :: : :'.. : ; : : : : : : : : ; : :'......... :... : : : 3................ :........... .. :,. : : *,--..,. :, :. :. : : ....... ::: :.......................... :. : .... : .. ..... m o.. 8.....-I.. I-- y. r :. : :. :.. : : :..... :.. :....... :.. :..... L,."$: : :. : :...... :............ ,'', . : : :, f. :.. W.,. gp, g7 . :.., : :.. :..... : : ...... ; : : ; : : : : :.. ". ::. : :. : :. : : : : : : :. : : : : : :.. :. ...C31. 10.. : :,....... :.......... :...:.... :...... :: . r: : WZh MBP ?. I. x :. :. . :.. :.. :. nr.. ir. : :.. :.. ... :.... _... : :....",....... :'..... : : :'' : : : : :.... : : : : : : : : :.. : : : :.. : : : :$P9. 7 .................. ..... P ......... :......... :..... :.fY. : : : : :. :., : :... : : : : : : : : :. : : : :.....> : : w : :... : :.: : : :... :... .rus.... : ....................... :...... : :..................... .... :.......-....... ti2 t'W'...............I. I........ p.... : : :. . :....... ; ......... -.....-.. ' tut..::. .. o........... ............. .. :.......... __- .............. p.. :....................... . I7.......... .. T s.. : :. : :.. : ir : : : : : :' ;. :.. . 5. : :.. te... : th : :.. : a o.... :. : :..... : : : :...::P 9.. #.*: :.., 8 H.. i*i ;'' r. r. : r'..... r."'..-- _ r. __ ........... I I.::... : : -_ , ; : . t. tY .;.. ;.... .....' :.. :... :.-..... :..........P .... :....... :.... x.. 5. c...t a o.... a... :.. :.......... :.,-.-','. r.... -... x...... n. :. ; ;. t- : ,.. SP : 9 : 7.,. : :...,, :. ; : :.. C :. ss h.z.... .... rsr....:.. x. : b ; ; : : : : : :". : :.-.I.. " :::. ; ; : : .. :. 0.... :. :,.' :. : . r :... ... . 5...etd : : w.h las 3n : :. :..... :..... : : :.. : : : : : :..... f... F'9.'...x. ... :....

The 40-mer oligonucleotides (Tables 3 and 4. 4) used as primers for PCR amplification of LEU4 andYOR108W disruption were synthesised by DNA Technology (e- mail oligo@DNAtech. dk). Restriction sites necessary for insertion of the PCR frag- ments into the plasmid andAflll necessary for the linearization of the final plasmids are indicated in Tables 4. 3 and 4. 4. The used procedure is known as splicing by overlap extension. In procedure, two DNA fragments containing sequences that are homolo- gous with a designed fusion point are mixed in equimolar denatured, annealed and extended to a hybrid DNA fragment. The 1. 5-kb/eu4A (Fig. 22) has been made by fusing the two primary 5'and 3'/eu4 fragments described below. 5' fragment of leu4A amplified with the primers"/eu4 97-2EcoRl"and"leu4 97-6 DLT B Sall". fragment covers the LEU4 from-700 to +320. The 3'fragment of leu4A amplified with the primers"/eu4 97-3Pvull Hindlll"and"/eu4 B DLT C Sall". most 3'fragment covers the/eu4 from +1563 to +2020. After PCR amplification the two/eu4 were mixed in equimolar and extended for 4 to 10 cycles in the ROBOCYCLER without the addition outer primers. After 10 extension cycles outer primers"leu4 97-2EcoRl"and"leu4 97-3Pvull HindlIl"were and the amplification continued for another 20-25 cycles. The final 5 kb/eu4E DNA fragment was purified using the Prep-A-Gene matrix. The/eu4A fragment was di- gested with EcoRl and Hindlll, and ligated the dephosphorylated EcoRl-, Hin- dlll-linearized vector (pUC19 : Yanisch-Perron et al. The resulting plasmid was designated pMBP9711.

Table3. Oligonucleotides to construct the pMBP9711 /eu4A plasmid andpMBP97111B/eu4A loop-in/loop-out plasmid t +. : ! ! K : g ;. : ! sHe . H : ' :-m : H : : : ; :iH : ; : : : f ; ; : : g : ; : : : ; : ; « ; : t : : ; : f « g<t : : g ! g ; : Sg ;'fS ; SS :'< < !f : <Sf : KHnsSS''1SN'<<SRS<'S : « : S<f ; < : : : : : : : : : : : : : : : :' : : : : : : :.' :' : : :- : : : : : : : ; : i : : : : : : : : W. : : ;. :. ;.-. :\.. : ; : ; :. : : : : : : : : : : : : : :' :- :1 ! . !' :"Sf'a ! i'' ! i sSmSiKS jTSKSeTS6BTE6'66'§iK<' : : : : : : ; : : : : : : : : : : : : : K : :' : : : iSf ; : :. S ; : f : t : ! S : :'f ! : ff :'< ; : : : : NSff *'TR :f*'&rRfr'ri''f'''<'T'& (' :''t'r'tT : Sr'S'PSNSsSS: :... : : : : : : : : ; : : : : : ; ; : : : : : : : : ; ; ; ;. t :,< : : tq'tH : M" :'' : M ; : : (''S' ! : KSK : S"''f*S" :'""<S :'*<<S<""" ?-H i :*-' : ; ; :'j :f'-"K. H : : : : : ; : i :. Kf :. : i :i : : :::. GCC:.< :H'f :§SiS ; fM:! fg fS : f f ; : & ; :i ; 7 :Mt'fpr Tff :j-'- ! fRPSifSiKS CA :

The 0. 5-kb YOR108W disruption was synthesised the splicing overlap extension procedure.

The 5'fragment was amplified with the primers"YOR108W Link A Aflll" (SEQ:ID NO 17) and"YOR108WLink B:Sall" (SEQ ID NO 18)(Table 4). The fragment covers the YOR108Wsequence +639 to +848. The 3'fragment was amplified with the primers- "YOR108W Link CSall" (SEQ:ID NO 19)and"YOR108W link D Aflll" (SEQ:ID NO 20).

The 3'fragment covers the YOR108Wsequence from+871 +1195. After PCR amplifi- cationthe two yor108w fragments were mixed in equimolar and ex- tended for 10 cycles in the ROBOCYCLER without the addition outer primers. After 10 extension cycles the outer primers"YOR108Wlink CSall"and"YOR108Wlink BSall" were added and the amplification continued for another 20-25 cycles. Thefinal 5-kb YOR108W DNAwas purified by the Prep-A-Gene matrix. The DNA fragment was digested with Sall, and ligated the dephosphorylated Sall-linearized pUC19 vector (pUC19 : Yanisch-Perron et al. The final plasmid designated pMBP97VC. Theplasmid pMBP97VC linearized with Aflll transformation into yeast.

Table4. Oligonucleotides for the construction of ORF YOR108W disruption plasmid Oligonucleotides used for yorlO8w . .. sequence 5'to 3'direction ATC Link A AflII ATC CTT CTG TAG AAA GTG AAG CTG TTA AGG C AflII ''"'.,'AATCTGTCG.ACA'AATCTGTCG.ACA CAA CCA CGG TTG TGA CAA TGC G YOR108WLinkBSail"'TF''' YOR108W Link C SalI CTTGTC GCG CTG ATC GTG TAG AAG GTT GTC TTG G SalI CAGCTTAAG GATCCA ATG GTAAGT ATG GAA TCC ACC G YOR108W Link D AflII AflII

4.2. inactivation/modification the isopropylmalate activities in theleucine de novo pathway (encoded by LEU4 andYOR108Wj (i) Inactivation ofLEU4fbr Inactivation ofLEU4 done by introduction of a PCR produced leu4A into S. cerevisiae S. carlsbergensis sporeclones the two-step deletion procedure by Scherer and Davis (1979). The oligonucleotides to construct the leu4A deletion plasmid shown in Table 3.

A map of the wild typeLEU4 nucleotide with selected restriction sites is shown in Fig. 21. A leu4 mutant has been made by removal the coding region from nucleotide +331 to nt +1563. The leu4/\ plasmid pMBP9711 shown in Fig. 22. The deleted part corresponds to a deletion of amino acid (aa) 111 to aa 521 seen in Fig. 23.

Bold letters deleted amino acid residues. The remaining residues are shown in ordinary font.

The selection system is based on the non-reverting met2 (J. Hansen personal communication). The mutation met2A been chosen as selection marker. The loop-in plasmid pMBP97111B based on the plasmid pMBP9711 it contains an additional 3. 7- kb fragment, inserted at the Hindlll site, the wild type MET2-CA isolated from S. carlsbergensis and Kielland-Brandt, This wild typeMET2-CA gene is used for selection of transformed colonies. The leu4 was made in the strain MBP94-21 containing met2 with the original LEU4 described by Baichwal etal. The strain MBP94-21 was constructed from a cross between JH250 (Hansen and Kielland-Brandt and XK14-13C (Baichwal etal 1983).

Aftertransformation of MBP94-21 with pMBP97111B linearized with Afill selection for transformants was done on SC-methionine selective plates. Incubation ofcells done at 27°C. the recovered colonies, one, MBP97-1, chosen for the further experi- ments.

In to pick up spontaneous loop-out the selective plasmid the adjacent wild typeLEU4 the strain MBP97-1 was serially transferredin liquid highglucose medium and grown to log to relax theselective pressure.

Asuspension of cells the third transfer was diluted and plated rich high glucose solidifiedmedium plated a concentration of 250-400 colonies on each of the 120 plates.

Afterincubation for 4-5 days at 27°C eachplate wasreplica plated ontoplates containing solidifiedSC-methionine and rich high glucose medium, respectively. which did not grow on SC-methionine plates assumed to have looped-out theplasmid con- tainingthe wild typeMET2-CA in the plasmid.

Two met2 loop-out were further tested to confirm that they contained the deleted version of LEU4. were tested by replica to various drop-out plates including SC-leucine andSC-leucine TFL. The loop-out designated MBP97-4 and MBP97- 5 are both auxotrophs for methionine and sensitive to 5, 5, 5-trifluoro-DL-leucine (TFL), thus indicating that the original LEU4 inactivated. PCR amplifications the LEU4 regions in MBP97-4 and MBP97-5 showed that both clones contains the deleted version ofLEU4 (ii) Inactivation the LEU4 homologue ORF YOR108w The ORF YOR108w disruption vector pMBP97VB based on the vector pUC19 (Yanisch-Perron et al. containing the 3. 7-kb Hindlll fragment carryingMET2-CA from S. carlsbergensis asselective The sequences necessary for integration to the ORF YOR108w region are inserted into the Sall of the pUC19 plasmid. Thelink- ers and sequences necessary for integration are shown in Table 4. 4 and Figs. 24 and 25.

Due to the high sequence homology betweenLEU4 ORF YOR108w integration linkers in the disruption plasmid are within the deleted area of leu4A (Table 4. 4, Figs. 24 and 25). This precaution will forceall to happen at YOR108W.

To elucidate the effect of Yor108wp, anYOR108w mutant designated yor108w was introduced in the leu4A MBP97-4. The leu4 yor108w double MBP97- 51, MBP97-52, MBP97-53, MBP97-54 andMBP97-55 exhibitedleaky leucine auxotroph phenotypes (Fig. 26) but they appeared to require isoleucine andvaline aswell. Numbers

1,2, 3, 4 and 5 indicate colonies requiringleucine, the colony designated 6 is growing relatively well onSC-leucine SC-methionine. Colony number 6 was not in- vestigated further.

A sample of S. cerevisiae strainMBP97-52 deposited on July 1998 in accordance withthe Budapest Treaty with the American Type Culture Collection under the- Accession No. 74455.

Southern blot of electrophoretic PCR amplifications and analysis of restriction endonuclease total DNA confirmed the inactivation of LEU4 and YOR108W givingto the/eu4E and yor108w mutants.

(iii) Inactivation ofLEU4 S. carlsbergensis The LEU4 deletion plasmid pMBP97111Bbeen introduced to the two met2A spore clones and JH334 (Table 4. 2). Its to correct position in both their ge- nomes was confirmed by PCR analysis, electrophoretic karyotypeanalysis Southern analysis of digested total DNA.It expected that the deletion plasmid integrated at the cerevisiae-like LEU4-CE allele the spore clones. Three loop-out from the first analyzed strains MBP97-19, MBP97-20 and MBP97-46 were identified SC-methionine drop-out plates. one strain, MBP97-46, showedloop-out the pUC19 sequence.

DuringSouthern analysis of JH334 and JH268 at least oneLEU4-CE one LEU4-CA sequence was identified in each strain. There is at least oneYOR108W-CE sequence present in the spore clones.

The strategy is to introduce the YOR108W disruption plasmid the S. carlsbergensis specific LEU4 and thereby producing a leu4 allodiploid for met2- ceA met2-caAleu4-ceA and leu4-ca:: MET2-CA yor108WpUC19. Therationale that LEU4 andYOR108Ware identical in the chosen region for introduction of the dis- ruption sequence. It therefore assumed that the integration occurs equally at LEU4chromosome XIV andYOR108Won XV. The localisation the integrated disruption plasmid been performed by Southern blot analysis ofelectropho- retickaryotype analysis.

(iv) Fermentation studies S. cerevisiae LEU4, LEU4 and leu4E S. cerevisiae and mutant strains were grown aerobically at20°C five days in 50 ml medium. The cultures were inoculated with 1 ml culture a density of 10 cells/ml. In 4. 5 are shown GC-headspace analysis (cf. 5. 5 (iii) GC-headspace analysis) of3-methyl-1-butanol and2-methyl-1-butanol together(M1 B) corresponding acetate esters (MBA), 2-methyl-1-propanol (2M1P) as isobutanol) isobu- tylacetate (IBAC). background values are used those for the unfermented rich high glucose growth medium.

The LEU4 strains (MBP94-21 MBP97-1) produced very high amounts of M1 Bcompared to the LEU4 wild strains (S288C and JH250. The two leu4A strains MBP97-4 and MBP97-5 produced an amount of M1 B but not less than what is produced by LEU4 wild (S288C and JH250).

Table5. GC-Headspace analysis produced volatiles S. cerevisiae strains and LEU4 from the first fermentation trials<BR> inrich high glucose medium. Values in ppm. M1 B the combined peak containing both 3M1 B (isoamyl alcohol) and2M1 B (ac-<BR> tive amyl alcohol). are corresponding esters 3MBA (isoamylacetate) 2MBA (active amyl acetate) Strain Genotype-M1B ME£ PS medium 0. 16 0.01 S288C MAT a HIS4 NIET2 LELI4-119 0.33 MA,. T 1--0. 07 MBP94-21 r, 397 0. 20 MBP97-1 MAT a h. is4 : 3 ; 2 ;,ErI4, fJl 613 2. 50MBP97-4 RAT a hí-s 4 tm'129 0 27 MBP97-5 HA-T a hís4 X Å'i 13i 0 30 MBP97-7 MAT a-TFLR"178 0. 31 MBP97-8 MAT a hís4 : 0më-e2 f R-"27R-2-TFLR"177 0. 40 MBP97-8 M-A-T ahi. s4 irtet2 leu4 2'FZ-R"27A-3-7'FLR"185 0. 61 fi 0. 40 MBP97-11 AS åi XA-1-TFr ; 151 O 11 MBP 97-11.-a'tn1510. 11 MBP97-11 MA-. Tt.. :, h,, A 1-TFLR 0. 11MBP97-13 IIAT a his4 me. t. 2 1eu24 TF'I-R"41A-3 :-TFZR" 123 0. 04 MBP97-14-RAS a fii s8. 4, ;, mèt L-R- A-is-TrS''11 o o o 9

Leu4 andyor108w mutants Into monitor the cell and the fermentation performed by the leu4 yor108w doublethe double strain MBP97-52 was tested in three different liquid media. Six strains were grown in rich glucose medium (PS), SC and SC-leucine medium, respectively(Table 4. 6). The following S. cerevisiae were used : S288C, JH250,- MBP94-21,MBP97-1, MBP97-4 and MBP97-52 4. 1). In preliminary experi- ments, the double mutant leu4Ayor108w still producedM1 B an amount equivalent to the LEU4 YOR108Wwild typestrains. It however, assumed that the majority of the re- covered M1B from the leucine up by the yeast cell the growth media, as MBP97-52 was unable grow on synthetic complete medium without leucine (SC- leucine).

Table 4.6. GC-Headspaceanalyses produced volatiles fromS. cerevisiae referencestrains and leu4/\yor108w double rnutant MBP97-<BR> 52from the first fermentation trials in SC,SC-leucine andrich glucose medium(PS), respectivety. Values in ppm. M 1 B the com-<BR> binedpeak containing both3M1B (isoamyl alcohol) and2M1B (activeamyl alcohol). are corresponding 3MBA (isoamylacetate) and<BR> 2MBA (active amyl acetate) STRAIN MIB M1B M1B MBA MBA MBA SC SC-LEUPS SC SC-LEU PS S288C 8, 327 0,018 10,088 0,023 130,250,373 JH250 7, 726 0, 017 8,3630,02 123,4310,307 MBP94-21 65,619 0, 041 45,9080, 031 314,811 0,713 MBP97-1 60, 482 0,044 49,4320,065 0,962 MBP97-4 9,223 0,019 343 0, 015 126, 604 0,258 MBP97-52 24,360,025 0, 9720, 013 186,26 0,383 PSbackground 1, 71 0,014 SCbackground 0, 42 0,014 SC-LEUCINE 168 0, 013 background

(v) Inactivation an a-ketoisocaproate decarboxylase by the open read- ing frame YDL080c.

The present invention contemplates to produce yeast strains where the Yd1080cp (Thi3p) enzyme is inactivated either by a deletion or by a gene disruption. The inactivation of YDL080C assumingly will rise to reduced production of 3-methyl-1-butanol (isoamyl-_ alcohol) andconsequently production of 3-methyl-1-butyl-acetate.

The gene is inactivated by a disruption plasmid pMBP981A on the vector pUC19.

The vector is opened by Hindlll a 3. 7-kb Hindl I I containing MET2-CA isli- gated into the opened vector. This plasmid designated pMBP96A2. TheSall in the remaining pUC19 multi-cloning is opened and an approximately 500 bp Sall fragment is ligated the plasmid. 500-bp Sall contains 250 bp of the region nt +1721 to +1971 and 250 bp of the region nt-249 to +1. The restriction enzyme Aflll line- arises the plasmid pMBP981A.

It isalso contemplated use the yd1080c in combination with mutations in any of the following genesLEU4, YOR108W, ATF1 andATF2 strains of any Saccharomyces species including the lager strain S. carlsbergensis.

EXAMPLE 5 Lowered ester content in beer by inactivation of genes coding for alcohol acetyltransfer- ases in Saccharomyces 5.1. Introduction Acetate esters are important flavour components in fermented beverages such as beer, wine and whisky. The acetate esters are produced by the yeast by metabolism of branched chain amino acids such as leucine, isoleucine andvaline are formed from higher alcohols andacetyl A by alcohol acetyltransferases.

The aim of the present study was to inactivate the genes coding for the alcohol acetyl- transferases in S. carlsbergensis lager in an attempt to lower content of the acetate esters isoamyl and ethyl acetate in beer.

5.2. Materials Methods Genotypes of the yeast strains used in this study are shown in Tables 5. 1 and 5. 2. S. carlsbergensis lager strains, C80-CG65 (MATa) andC80-CG 110 (MATa) spore clones ofM204, in-house Carlsberg lager strain. Agar plates rich medium highin glucose or plates with synthetic complete medium (SC) or synthetic- completewithout methionine (SC-methionine) for selection made according toSherman (1991). For selection of G418-resistant transformants of S. carlsbergensis, G418 (geneticin) was used at a concentration of 30 ug/ml YPD (1% extract, 2% peptone, 2% glucose).

Table 5.1. and source of S. cerevisiae strains Yeast straia Mating type Genotype ortype Source S288C HIS4 MET2 LEU4 ATF1 ATF2 JH250 HIS4 met25 LEU4 ATF1 ATF2 and Kielland-Brandt 1994 XK14-13Ca'h.is4MEr2LEC74 (*)ArFArF2'Baichwal al. 1983 MBP94-21ais4jnet2ALEU4'' (*)ATF1 ATF2 This study SBS97-1 a his4 met2# LEU4fbr(*) atf1# ATF2 This study (*)5, 5, 5-trifluoro-DL-leucine resistant mutant containing leucine resistant (fbr) a-isopropylmalate synthase.

Table 2. Genotypes and source of S. carlsbergensis strains YeastMatingGenotype ortypeSource strain type M2045.carisjberyensislager strainAm-CJE/ATfU-CA ATF2-CE/* Carlsberg C80-CG65 Meiotic segregant from M204ATF1-CE/ATF1-CA ATF2-CE/* Gjermansen Sigsgaard 1981 C80-CG110aMeioticsegregantfrom M204ATF1-CE/ATF1-CA ATF2-CE/* Gjermansen Sigsgaard 1981 SBS97-3aC80-CG65at:f.I-ceA/Am-CAATF2-CE/*This study SBS97-6a'C80-CG110at:f.t-ceA/ATFI-CAATF2-CE/*This study SBS97-7 a C80-CG110 ATFI-CE/atfl-ca0 ATF2-CE/* This study SBS97-8 a C80-CG110 atfl-ceA/atfl-ca0 ATF2-CE/* I This study * It not known whether S carlsbergensis have two forms of the ATF2 gene.

Plasmids prepared from E. coli DH5a described by Sambrook et al. or by use of"Wizard Plus DNA Purification System" (Promega A7640). Plots ofplas- mids with restriction analysis were drawn using the program pDRAW (K.Olesen, personal communication E-mail: kol@crc. Restriction endonucleases DNA modifying en- zymes were from Boehringer Mannheim, New England Biolabs Promega and used in accordance with the manufacturers procedures. Oligonucleotides obtained from _ DNA Technology, Aarhus, Denmark. All chemicals were of the highest purity available.

Yeast transformation was done by the lithium acetate method according to Schiestl and Gietz (1989) and Gietz et al. Genomic DNA was isolated in FMC Incert agarose according to the method of Schwartz and Cantor (1984) with the modification by Pedersen (1988). Separation of chromosomal length DNA molecules performed in 1. 3% FMC FastLane Agarose by pulsed fieldelectrophoresis a Bio-Rad CHEF MAPPER System.

The buffer used was 0. 5 x TBE (0. 044 M Tris-borate, 0. 001 M EDTA). Southern blotting of DNA in agarose gels done after depurination (0. 25 M HCI 30 min.), denaturation (0. 4 M NaOH 0. 6 M NaCI 60 min.) and neutralization (1. 0 M Tris + 1. 5 M NaCI, 7. 2 in 120 min.) of the gel Hybond N nylon from Amersham Life Science. The DNA was fixed to the membrane by UV crosslinking a Stratagene UV Stratalinker 2400 and Southern blot hybridization performed with a probe which was either la- belled witha-32P-dCTP Life Science Products) using the"Random Primed DNA La- beling and Detection Kit" Mannheim 1 004 760) or with digoxigenin-dUTP employing the"DIG Labeling and Detection Kit" (Boehringer Mannheim 1 093 657).

The labelling performed according to the manufacturer's procedure.

Inactive of the ATF-genes constructed by three PCR-reactions using a Perkin Elmer DNAThermal Cycler Stratagene RoboCycler 40 or 96. The Expand High Fidelity System (Boehringer Mannheim 1 732 650) was used with genomic DNA as template and oligonucleotide primershomologous sequences in the alcohol acetyltransferase (SEQ ID : 21-40) (Table 5. 3). The DNA sequences were ob- tained from the Saccharomyces Genome database (http ://genome-www. stanford.- edu/Saccharomyces/) or the Genbank database (NCBI) (http ://www. ncbi. nim. nih. gov/). By PCR reactions two DNA fragments, designated PCR1 and PCR2, about 400 to 700 bp in length were synthesized. PCR1 was identical a region upstream or the beginning of the alcohol acetyltransferase in question, whereas PCR2 was identical a region downstream or the end of the gene in question. In third reaction, a PCR fragment com- prising an inactive ATF gene, leaving 500 to 2, 000 bp of the DNA coding for the

transferase gene (see Table 4) was synthesized with restriction sites for cloning the ends using PCR1 and PCR2. The procedure is known as splicing by overlap extension, as two PCR-fragments containing homologous with a predetermined fusion point are mixed in equimolar denatured, annealed extended to a hybrid PCR- fragment. The PCR-fragments were purified from agarose gels employing"Prep-A-Gene DNAPurification System" (Bio-Rad cat. no. 732-6010, USA).

Table5. 3. Oligonucleotide used for synthesis of PCR fragments employed todelete of the alcohol acetyl transferasegenes S.cerevisiae Gene PCR-Restriction Oligonucleotide sequences Position in sequence relative to fragment site initiation codon of gene ATF1 PCR1 SacI 5'-AGQ CTd AGC TCC TCCTGG GTT AAG ACT TTC TC-1219-E-1195 5'-GGG TAGAAC TGT CCA GAG C-640-661 PCR2 5'-GCT CTG GAC AGT GCT ACC CAT GTG CGA TCG TGC CAT CGG G-661-640 and+1300-)'+1320 Pali 5'-AGA TCT CTG CAS ATC AAA TCA ATT AAT+1833-).+1604 ATF2 PCR1 SalI 5'-GCA CGC GTC GAC CTA CAT TGA ACT CTG TAG GCC ACC G-300-276 5'-CTCGAC TTC TCT GTA TTC TGG +384-+ +361 PCR2 5'-CCA GAA TAC AGA GAA GTC ATG GAGCGG CAA CGT TGG AGG TTC GC +361 and +1125-+ +1147 SphI5'-GAGCAG CCACGG CATGCA TCG ACT +1730 -> +1707 S. carlabergensis C80-CG65 andC80-CGl10 ATFl-CE PCR1 XbaI 5'-AGC TCT AGA GTG TGA GGA CTC ATT GGC TTG-230-+-207 5'-ACTCTG TAC TCA GGT TGT TCA +392 +369 PCR2 5'-TGA ACC TGA GTA CAG TGC AGT CGC AGA CCG CTC ACA ACT ACC +369 and +1044-+ +1067 SacI 5'-AGC TGC GAG CTC ACACGA AAT CAT ATT GTC G +1629 +1608 ATFl-CA PCR1Xbal5'-AGC TAG TCTAGA TTG ATT GAT CAA TGT GAA-249--229"" 5'-CCATGC CGG TAC AAT A +315-+ +294 PCR2 5'-TGT ACC GGC AAG ATG GCC ACT TAT TGC ACA TTA TCA T +297-i and +1240 +1260 SacI 5'-AGC GAG CTC CAC TTA CTT ACC TTA CAC ACG TCG TT +1822 +1797 ATF2-CE PCR1 XbaI 5'-AGC TAG AGA AGC GTA CTA CTC TAG CGA AGA GTA +255-+ +278 5'-CAGAGT CTC GAT CAT AGT CAA +709-+ +686 PCR2 5'-TTG ATG ATC GAG ACA CTG GTCGAA TAC TAT GAC CGC TT +686 and +1213 -> +1235 SacI 5'-TGT ACG AGC TCG GCC GAGCTA TAC +1657 -> +1634 <BR> Restriction sites incorporated in the oligonucleotides underlined.<BR> i Table4. Nucleotides by inactivation of genes coding for alcohol acetyltransferases GeneNucleotidesdeletedFragment relative to initiation deleted ATF1 (S. cerevisiae-639-+12991938 bp ATF2 (S. cerevisiae) + 365 -> + 1124 740 bp ATFl-CE (S. carlsbergensis C80-CG65and C80-CG110) +393 o + 1043 651 bp ATF1-CA (S. carlsbergensis C80-CG65and C80-CG110) +315+1239924.bp ATF2-CE(S. carlsbergensis C80-CG65and C80-CG110) +710 1212503 bp

In S.cerevisiae, theselectional was based on the non-reverting meut20 mutation (J. Hansen personal communication). Integration plasmids (Fig. 27) and pSBS97-2 (Fig. 28) for ATF1 andATF2, were constructed by ligation the PCR fragments comprising the inactive forms of the ATF-genes (see into the Sacl- Pstl sites(ATF1) orSall-Sphl sites(ATF2) pMBP96A2. The plasmid pMBP96A2 (Fig.

29) was produced from pUC19 (Yanisch-Perron et al. , by inserting a 3. 7 kb Hindi fragment including S. carlsbergensis specificallele ofMET2 (MET2-CA), en- codes a homoserine acetyltransferase and Kielland-Brandt, into the Hin- dl I Iof pUC19. TheMET2-CA thus conferred prototrophy for methionine in S. cerevisiae (see Table 5. 1) when this strain was transformed with either pSBS97-1 or pSBS97-2.

In S.carlsbergensis, integrationplasmids constructed by ligation the PCR frag- ments comprising the inactive forms of the A TF-genes above) into the Xbal andSacl sites of the integration vector pCH216 containing a G418-resistance cassette as selective marker (Hadfield etal. , 1990).In way the integration plasmids and pSBS97-6 for ATF1-CE 30), pSBS97-4 and pSBS97-7 for ATF1-CA 31) and pSBS97-5 and pSBS97-8 for ATF2-CE 32) were constructed. Genomic DNA from C80-CG65 (MATa) used as template the synthesis of the inactive forms of the ATF-genes into pSBS97-3, pSBS97-4 and pSBS97-5 while DNA from C80-CG110(MATa) used as template for the synthesis of the inactive forms of the ATF-genes into pSBS97-6, pSBS97-7 and pSBS97-8.

5.3. Construction of haploid laboratory with eliminated or reduced alcohol acetyl- transferase activity.

Totest the hypothesis that inactivation of alcohol acetyltransferase in yeast results in lowered of esters in the fermented medium, inactive forms of ATF1 and ATF2 were introduced into the haploid laboratory MBP94-21 by homologous recombina- tionbetween deletion alleles wild-type alleles the genes. MBP94-21 is auxotrophic for methionine and overproduces isoamyl alcohol andisoamyl (Table 5. 1). The two-step loop-in/loop-out of Scherer and Davis (1979) was employed to sub- stitute the wild-type allele each alcohol acetyltransferase in MBP94-21 with dele- tion alleles. construction of the inactive forms of ATF1 andATF2 insertion into the integration plasmids described above. The integration plasmid was line-

arized withNsil 27) while was linearized with Mscl 28) and MBP94- 21 transformed with the linearized plasmids. As Nsil twice in pSBS97-1, the DNA preparation was only digested. Transformants were selected as being able to grow on SC-methionine plates. confirm that integration had taken place the location of the wild-type alleles Southernblots hybridizations with labelle pUC19 DNAas probe were performed on agarose gels, genomic DNA had been separated by pulsed field gel electrophoresis.

In to loop-out theselective plasmid with the adjacent wild typeATF-gene leaving only inactive alcohol acetyltransferase on the chromosome, two inte- grants were grown to stationary phase 3 times in rich medium high in glucose content (PS-medium) with 100-fold dilution each stationary culture into fresh PS-medium. Ten thousand to twenty thousand cells the third stationary culture screened for spon- taneous loop-out the vector DNA and the wild-type of the gene in question by plating on solid (50-100 plates). incubation for 4-5 days at 27°C the colonies on each plate werereplica-plated two plates SC-methionine and PS- medium, respectively. The colonies that were unable grow on SC-methionine were picked and analyzed by Southern blots hybridizations of genomic DNA digested with Stul orKpnl identify those, where an inactive deletion allele substituted the wild- type gene. In way, mutant strains of MBP94-21 with inactive alcohol acetyltransferase genes were isolated.

5.4. Construction of S. carlsbergensis lager strains with eliminated or reduced alco- holacetyltransferase activity.

Totest the hypothesis that inactivation of alcohol acetyltransferase in yeast results inbeer with a lowered of the esters isoamyl and ethyl inactive forms of ATF1-CE, ATF1-CA and ATF2-CE were into the spore clones C80- CG65 and C80-CG110 homologous recombination between deletion alleles andwild- type alleles the genes (the loop-in/loop-out in a procedure similar that de- scribed for S. cerevisiae yeast. The construction of the inactive forms of ATF1- CE, ATF1-CA andATF2-CE described above. The two-step integration procedure of Scherer and Davis (1979) as described above was employed. The integration plasmids were linearized with the restriction endonucleases Sphl pSBS97-6) (Fig. 30), Sapl pSBS97-7) (Fig. 31) or Stul pSBS97-8) (Fig. 32), respec-

tively. SinceSapl twice in pSBS97-4 and pSBS97-7, the DNA preparations were only partially The spore clone C80-CG65 was transformed with each of the line- arized plasmids pSBS97-4 and pSBS97-5 while C80-CG110 transformed withthe linearized plasmids pSBS97-7 and pSBS97-8. Transformants were selected as being geneticin resistant. To confirm that integration had taken place at the location the wild-type alleles, Southernblots hybridizations were carried out. Loop- out of the vector DNA together with the wild gene, leaving only copy of the inac- tive alcohol acetyltransferase on the chromosome, was carried out as described for the laboratory strain MBP94-21. thousand to twenty thousand cells were screened by plating on PS-medium (50-100 plates) after incubation 4-5 days at 20°C the colonies were replica-plated two plates withsolid with geneticin and PS- medium, respectively. The colonies that had become sensitive to geneticin were picked and analyzed as described above by Southern blots hybridizations to identify those, where an inactive deletion allele substituted the wild-type gene. In way, spore clones one of the genes coding for alcohol acetyltransferase an inactive form were isolated.

5.5. Fermentation studies (i) S. cerevisiae strains.

Liquidrich medium high in glucose (10 ml in 25 ml was inocu- lated with the appropriate yeast strain and grown overnight at 30°C. 1-2mi this culture containing 107 cells used to inoculate 50 mi medium in a 250 ml conicalflask.

The culture was incubated at room temperature for 5 days with agitation (90 rpm.). Yeast cells sedimented overnight at 4°C the supernatant discarded. 50 mi me- dium was added to the cells 2. generation fermentation performed as above. 25 ml supernatant from the fermentation was delivered for GC-headspace analysis (seebelow) to determine the content of esters and alcohols. results are shown in Table 5. The yeast strains carrying inactive alcohol acetyltransferase produced a reduced amount of acetate esters such as ethyl and isoamyl in comparison to the MBP94-21 strain. Table5. GC-headspace analysis of volatiles by S. cerevisiae strains.

Valuesin ppm

Compound Yeast strain MBP94-21 SBS97-1 Ethyl acetate60 1. 752. Isoamylacetate + 21 0. 12 active amylacetate Isoamylalcohol 290 257 (ii)S. carlsbergensis lager strains.

Instead rich medium high in glucose content (PS-medium) fermentation of brewers wort was performed as described for the laboratory yeast. Incubation at20°C used in all cases.

(iii) GC-headspace analysis.

GC-headspace analysis carried out on a Perkin-Elmer with HS 101 head- space sampling and Turbochrome analytical A 50 m CP-WAX 52 CB capillary column 32 mm inner diameter and 1. 20 pm thickness) was temperature pro- grammed : 60°C 1 min. -10°C/min. to110°C-110°C 4 min. -30°C/min. to175°C- 175°C 4 min. -30°C/min. to225°C-225°C 1 min. The injection temperature was 140°C the detection temperature 250°C. carrier gas was helium at 15. 0 psi. Vials for the autosampler werefilled 5 ml of fermented medium together with 1.5 g NaCI, magnet and 1 ml standard (n-pentylacetate).

(iv) Results Atpresent, the S. cerevisiae SBS97-1 where ATF1 inactivated (see Tabel 1) has been isolated and loop-in pSBS97-2 in S. cerevisiae is correct. Prelimi- nary fermentation with SBS97-1 to stationary phase in two generations has by GC- headspace analysis revealed 40% decrease in the content of isoamyl compared to a similar performed with S. cerevisiae (Table 5. 5).

The S. carlsbergensis lager strains SBS97-6, where ATF1-CE is and SBS97-7, where ATF1-CA inactivated, have been isolated (Table 2). In mutant strain SBS97-6 it has been attempted to substitute wild-type ATF1-CA a deletion al- lele. possible loop-out colonies been investigated by restriction Southern analy- sis to identify correct loop-out strains. Nine were found to have looped-out leaving the wild-type gene in the chromosome, while the last colony, (Table 5. 2), pre- sumably lost the chromosome carrying ATF1-CA. present, two loop-out colonies are investigated by Southern hybridizations for loop-out of ATF1-CE C80-CG65. One of these presumably has lost the chromosome carrying ATF1-CE, in the mutant strain SBS97-3 (Table 5. 2).

When allodiploid clones of C80-CG65 and C80-CG110 inactive alleles of both ATF1-CE andATF1-CA having lost the chromosomes carrying these genes) are obtained, these will crossed with each other to reconstitute allotetraploid lager yeast strains, which can be tested technologically incylindroconical fermentationvessels.

Asample of S. cerevisiae SBS97-1 was deposited on July 1998 in accordance withthe Budapest Treaty with the American Type Culture Collection under the Accession No. 74450.

EXAMPLE6 Production of a composite beer by combining high-and low-sulphite batches of beer 6. 1. of the experiments Anew approach has been made towards production of a beer containing satisfactory amounts of sulphite and at the same time having a normal beer flavour profile. Two inde- pendent beer batches were produced using two individual strains, both genetically modified.One yeast strain, PFJ501, was inactivated in all copies of the MET14 gene encoding the APS kinase (EC 2. 7. 1. 25) (see Example 1),resulting very low absent sulphite production during fermentation. The other yeast strain, SB130, inactivated in allcopies of the MET10 encoding one subunit of the sulphite en-

zyme (EC 1. 8. 1. 2), (Hansen and Kielland-Brandt, resulting in a dramatically in- creased sulphite production during fermentation.

The two batches of beer brewed using the two individual yeast strains were combined to obtain a predetermined sulphite content of 15 ppm in the final beer. Based on evaluationan experienced taste panel, composite beer obtained higher score after forced aging as compared to the two individual beers, and it was comparable to a refer- ence beer brewed from a normal strain.

6.2. Materials and methods Strains media The three Saccharomyces carlsbergensis used were PFJ501 (see Example 1, Ta- ble 1), SB130 (Hansen and Kielland-Brandt, and M204 as a reference production strain (see Example 1,Table Brewer's wort with a gravity of 14. 6% Plato used for the 50 L-scale fermentations.

Fermentation at 50 L scale Afterpropagation, yeast was inoculated in brewer's wort at 1. 5x107 cells/ml 50 L fer- mentationvessels. Fermentation was performed at 13°C. yeast was harvested after 10 days and the beer was left in the vessels for 8-12 days of lagering. harvested yeast was used as inoculum the following brewing generations.

Other analyses Eitherfermentation samples (the supernatant from centrifuged samples) bottled beer were assayed for total S02 headspace chromatography with a Sulfur Chemilumi- nescence Detector (SCD) (Lowe and Dreyer, 1997).

6. 3. Results (i) Production and evaluation of the flavour stability of a composite beer versus in- dividual of beer with or low sulphite content.

Fermentation of brewer's wort with strains PFJ501 and SB130 performed at 50 L scale, the production strain M204 was used as a reference. The sulphite production was followed during fermentation (Table 1). Fermentation with strain PFJ501 resulted in a very low production, whereas strain SB130 produced high amounts of sulphite as compared to the reference strain M204. With respect to the different brewing parame- ters, strains PFJ501 and SB130 performed normally to the reference strain M204.

Prior to bottling the low-and high-sulphite were combined to obtain a composite beer containing 13 ppm of sulphite. This was done by mixing 11. 4 litres of the beer pro- duced by strain SB130 13. 6 litres the beer produced by strain PFJ501 to obtain a total 25 litres composite beer having a calculated sulphite of 15 ppm (the difference of 2 ppm sulphite disappears during bottling). The composite beer was bottled as were the two individual of beer from which is was made, along the refer- ence beer.

Table 6.1. Sulphite during fermentation at 50 L scale the three strains M204 (reference), SB130 (high sulphite and PFJ501 (low sulphite producer).

The last column the amount of sulphite measured in the bottled beer.All num- bers are from the 2. brewing generation.

S02/ppm Strain day 3. day 6. day 7. day 10. day13. day14. day15. day17. day Bottled beer M204 0 1. 4 5. 7 8. 5 7. 1 6. 4 6. 4 6. 4-4 SB130782328504!40363833 PFJ50100?7007O'T000?7O Mixture... 13 The bottled was judged by an experienced taste panel and after forced aging (7days at 37°C). Theevaluation the beers is presented graphically Fig. 27. Before aging,beer produced with strain PFJ501 assessed as"not quite satisfactory", and so was beer produced with strain SB130, the composite beer was"satisfactory"in

linethe reference beer. After forced aging, the two individual beers from strain PFJ501 and SB130 judged"not satisfactory"and"not quite satisfactory", respective- ly.composite beer was assessed as"not quite satisfactory"along the reference beer (Fig. 33).

6.4. Conclusion By combining two different batches of beer made by individual brewer's yeast strains, one having enhanced production of sulphite and one having no sulphite produc- tion, it was possible to make a composite beer which fulfills criteria, in this case a beer containing a predetermined sulphite level 15 ppm and at the same time being as satisfactory with respect to taste and flavour stability as is a conventionally ref- erence beer.

When the mixture of PFJ501-based and SB1 30-based was evaluated by an experi- enced taste panel composite beer obtained scores comparable the reference beer.

This shows that it is possible produce a composite beer having a desired characteristic by mixing individual beer batches produced by yeast strains modified by recombinant ge- netics to have enhanced and decreased production of the compound causing that char- acteristic.

REFERENCES Anness, B. J. (1980) The reduction of dimethyl sulphoxide todimethyl sulphideduring fer- mentation. J. Inst. 86 : 134-137.

Anness, B. J. and Bamforth, C. W. (1982) Dimethyl sulphide-a J. Inst. Brew.

88: 244-252.

Anness, B. J., Bamforth, C. W. and Wainwright, T. (1979) The measurement of dimethyl sulphoxide in barley andmalt its reduction to dimethyl sulphide yeast. J. Inst.

Brew. : 346-349.

Atkin,L., Stone, I. Gray, P. P. (1947) Use of the specific gravity gradient tube for brewery control. Wallerstein 11 : 281-287 Baichwal,R. ; Cunningham, T. S. ; Gatzek, P. R. andKohlhaw, B. (1983) Leucine bio- synthesis in yeast. Identification two genes (LEU4, LEU5) affect a-isopropylmalate synthase activity and evidence that LEU1 andLEU2 expression is controlled a- isopropylmalate the product of a regulatory Current Genetics 7 : 369-377.

Bamforth, W. (1980) Dimethyl sulphoxide of Saccharomyces spp. FEMS Mi- crobiol.7 : 55-59.

Bamforth, C. W. and Anness, B. J. (1979) Dimethyl Sulphoxide by yeast. Society for General MicrobiologyQuarterly : 160.

Bamforth, C. W. and Anness, B. J. (1981) The role ofdimethyl sulphoxide in the formation of dimethyl during fermentations. J. Inst. 87 : 30-34.

Bilous, T., Cole, T., Anderson, W. F. and Weiner, J. H. (1988) Nucleotide of the dmsABC operon encoding the anaerobic dimethyl sulphoxide of Escher- ichia coli. Mol. Microbiol.: 785-795.

Black, Harte, E. M., Hudson, B. and Wartofsky, L. (1960) A specific enzymatic reduc- tionof L (-) methionine sulphoxide a related nonspecific reduction of disulfides. J.

Biol.235 : 2910-2916.

Botstein, D., Falco, C., Stewart, S. E., Brennan, M. , S., Stinchcomb, D. T., Struhl, and Davis, R. W. (1979), Sterile host yeasts (SHY) : a eukaryotic system of bio- logicalcontainment for recombinant DNA experiments. Gene 8 : 17-24.

Botstein, D. and R. W. Davis (1982). Principles and practice of recombinant DNA research with yeast. In: Molecular Biology the Yeast Saccharomyces. and Gene Ex- pression. J. N. Strathern, E. W. Jones & J. R. Broach eds. Cold Harbor Laboratory.

Cold Harbor, New York, 11 B: 607-636.

Breton, A. and Surdin-Kerjan, Y. (1977) Sulfate in Saccharomyces cerevisiae: bio- chemical genetic study. J. Bacteriol. : 3017-3021.

Briza, P., Winkler, G.,Kalchhauser, and Breitenbach, M. (1986) Dityrosine is a promi- nent component of the yeast ascospore wall. proof of its structure. J. Biol. Chem.

261: 4288-4294.

Casey, G. P. (1986) Cloning and analysis of two alleles ofthe ILV3 from Saccharo- myces carlsbergensis. Carlsberg Commun. 51 : 327-341.

Cherest, H., Davidian, J.-C., Thomas, D., Benes, V., Ansorge, W. and Surdin-Kerjan, Y.

(1997) Molecular of two high affinity sulfate in Saccharomy- ces cerevisiae. 145 (3) : 627-635.

Cherest, H. and Surdin-Kerjan, Y. (1992) Genetic analysis of a new mutation conferring cysteine auxotrophy in Saccharomyces cerevisiae: of the sulfur metabolism pathway. Genetics 130 : 51-58.

Cherest, H., Thao, N. N. and Surdin-Kerjan, Y. (1985) Transcriptional regulation the MET3of Saccharomyces cerevisiae. 34 : 269-281.

Dickenson, J. (1983) Cambridge Prize Lecture : Dimethyl sulphide-its and control in brewing. J. Inst. 89 : 41-46.

Dickenson, C. J. and Anderson, R. G. (1981) The relative of S- methylmethionine anddimethyl sulphoxide precursors of dimethyl sulphide in beer.

Proc. Eur. Brew. Conv. Congr., Copenhagen, pp. 413-420.

Dickinson,J. R., Lanterman, M. M. , D. J., Pearson, B. M., Sanz, P., Harrison, S. J. and Hewlins, J. E. (1997) A'3C nuclear resonance investigation of the me- tabolism ofleucine toisoamyl alcohol inSaccharomyces cerevisiae. J.Biol. Chem.

272: 26871-26878.

Drain, P. and Schimmel, (1986) Yeast LEU5 a PET-like that is not essential for leucine biosynthesisMol Genet. 204 : 397-403.

Drain, P. and Schimmel, P.(1988) Multiple that determine activity for the first step of leucine in Saccharomyces cerevisiae. 119 : 13-20.

E.B. C. Analytica Microbiologica (1977)J. Inst. 83 : 109-118.

Elskens, T., Jaspers, C. J. and Penninckx, M. (1991) Glutathione an endogenous sulphur in the yeast Saccharomyces cerevisiae. Gen. Microbiol. : 637-644.

Friden, P and Schimmel, (1988) LEU3 of Saccharomyces cerevisiae activatesmultiple genes for branched-chain amino acid biosynthesis by binding to a common decanucleo- tide core sequence. Mol. Cell. Biol. : 2690-2697.

Fujii, T., Iwamatsu, A.,Yoshimoto, Minetoki, T., Bogaki, T. and Nagasawa, N. (1993) European Patent Application 0 574 941 A3.

Fujii,T., Yoshimoto, H. Nagasawa, N., Bogaki, T., Tamai, Y. and Hamachi, M. (1996a) Nucleotide of alcohol acetyltransferase from lager yeast, Sac- charomyces carlsbergensis. 12 : 593-598.

Fujii, T., Yoshimoto, H. and Tamai, (1996b) Acetate ester production by Saccharomy- ces cerevisiae lacking theATF1 encoding the alcohol acetyltransferase. Ferment.

Bioeng. 81 : 538-542.

Gibson, R. and Large, P. J. (1985) The influence of assimilable compounds in wort on the ability of yeast to reduce dimethyl suiphoxide. J.Inst. 91 :401-405.

Gietz, D., St. Jean, A., Woods, R. A. and Schiestl, H. (1992) Improved for high efficiency transformation of intact yeast cells. Nucl. Res. 20 : 1425.

Gjermansen, C. and Sigsgaard, P. (1981) Construction of a hybrid brewing strain of Sac- charomyces carlsbergensis mating of meiotic segregants. Carlsberg Commun.

46: 1-11.

Hadfield, (1994) Construction of cloning and expression vectors, pp. 17-48 in : Molecu- lar of Yeast. A practical Johnston, J. R. (ed.). Oxford University Press, Oxford, UK.

Hadfield, Jordan, B. E., Mount, R. C., Pretorius, G. H. J. and Burbak, E. (1990) G418- resistance as a dominant marker and reporter for gene expression in Saccharomyces cerevisiae. Genet. 18 : 303-313.

Hansen, J. and Kielland-Brandt, M. (1994) Saccharomyces carlsbergensis two functional MET2 alleles similar tohomologues S. cerevisiae S. monacensis.

Gene 140 : 33-40.

Hansen, J., Cherest, H and Kielland-Brandt, M. (1994) Two divergent MET10 genes, one from Saccharomyces cerevisiae one from Saccharomyces carlsbergensis, en- code the a subunit of sulfite and specify potential sites for FAD and NADPH. J. Bacteriol. : 6050-6058.

Hansen, J. and Kielland-Brandt, M. (1996a) Inactivation ofMET2 brewer's yeast in- creases the level of sulfite beer. J. Biotechnol. : 75-87.

Hansen, J. and Kielland-Brandt, C. (1996b) Inactivation ofMET10 in yeast specifically increases SO2 during beer production. Nature Biotechnol. : 1587- 1591.

Hoffman, C. S. and Winston, F. (1987) A ten-minute DNA preparation from yeast efficiently releases autonomousplasmids transformation of Escherichia coli. 57 :267-272.

Hu, Y., Cooper, T. G. and Kohlhaw, B. (1995) The Saccharomyces cerevisiae Leu3 protein activates expression of GDH1, key gene in nitrogen assimilation. Mol. Cell. Biol.

15: 52-57.

Johannesen, P. F. 1994. Increasing theflux the sulfur assimilatory Over- production of MET3, MET14 andMET16 inSaccharomyces cerevisiae. University of Copenhagen, Copenhagen.

Kielland-Brandt, M. C.,Nilsson-Tiligren, Gjermansen, C., Holmberg, and Pedersen, M. B. (1995) Genetics of brewing yeasts. In: Wheals, E., Rose, A. H. and Harrison, J. S., eds. : The Yeasts, 2ndedn. , Vol. Academic Press, London, pp. 223-254.

Kispal Steiner H., Court D. A., Rolinski, and Lill, (1996) Mitochondrial cytoso- licamino acid transaminases from yeast, homologs the myc oncogene- regulated protein J. Biol. 271 : 24458-24464.

Kohihaw, B. (1988) Methods in Enzymology : 414-435, Academic Press, Inc.

Korch, Mountain, H. A. and Byström, S. (1991) Cloning, nucleotide and regulation ofMET14, gene encoding the APS kinase of Saccharomyces cerevisiae.

Mol.Genet. 229 : 96-108.

Langin, T., Faugeron, G., Goyon, C., Nicolas, and Rossignol, (1986) The MET2 gene of Saccharomyces cerevisiae: cloning and nucleotide Gene 49: 283-293.

Lee S., Villa, and Patino H. (1995) Yeast strain development for enhanced production ofdesirable alcohols/esters beer. J. Am. Soc. Brew. Chem. 53 : 153-156.

Leemans, C., Dupire, S. and Macron, J.-Y. (1993) Relation wort DMSO and DMS concentration in beer. Proc. Eur. Brew. Conv. Congr., Oslo, 709-716.

Lowe, L. E. and Dreyer, T. 1997. Quantitation of total sulphur in beer by auto- mated headspace gas chromatography. Proc. Eur. Brew. Conv. Congr., Maastricht, pp.

649-654.

Malcorps, and Dufour, J.-P. (1992) Short-chain and medium-chain aliphatic-ester syn- thesis in Saccharomyces cerevisiae. J. Biochem. 210 : 1015-1022.

Meilgaard, (1975) Flavor chemistry of beer : Part II: Flavor threshold of 239 aroma volatiles. MBAA Tech. 12 : 151-168.

Minetoki,T., Bogaki, T., Iwamatsu, Fujii, T. and Hamachi, M. (1993) The purification, properties and internal peptide sequences of alcohol acetyltransferase from Sac- charomyces cerevisiae No. 7. Biosci. Biotech. Biochem. 57 : 2094-2098.

Moskovitz,J., Weissbach, H. and Brot, N. (1996) Cloning expression of a mammalian gene involved in the reduction of methionine sulfoxide in proteins. Proc. Natl.

Acad. Sci. U. S. A. 93 : 2095-2099.

Moskowitz,J., Berlett, S., Poston, J. M. and Stadtman, E. R. (1997) The yeast peptide- methionine sulfoxide functions as an antioxidant in vivo. Proc. Natl. Acad.

U.S. A. 94 : 9585-9589.

Mountain, A., Byström, S., Larsen, J. T. and Korch, (1991) Four major tran- scriptional in the methionine/threonine biosynthetic pathway of Saccharomyces cerevisiae. 7 : 781-803.

Nagami, K., Takahashi, T., Nakatani, K. and Kumada, J. (1980) Hydrogen sulfide in brewing. MBAA Tech. 17 : 64-68.

Nagasawa, N., Bogaki, T., Iwamatsu, and Kumagai, C. (1996) NCBI Acces- sion No. D86325.

Nishimura, Kawasaki, Y., Kaneko, Y., Nosaka, K. and Iwashima, (1992) A positive regulatory gene, THI3, required for thiamine metabolism inSaccharomyces cerevisiae.

J. Bacteriol. : 4701-4706.

Omura, F. and Shibano, Y. (1995) Reduction of hydrogen sufide production in brewing yeast by constitutive expression of MET25 gene. J. ASBC 53 : 58-62.

Orr-Weaver, T. L. , J. W. and Rothstein, R. J. (1981) Yeast transformation : A model system for the study of recombination. Proc. Natl. Sci. USA 78 : 6354-6358.

Pedersen, M. (1986) DNA sequence polymorphisms in the genus Saccharomyces. IV.

Homoeologous chromosomesI I I ofSaccharomyces bayanus, S.carlsbergensis, S. uvarum. Carlsberg Commun. 51 : 185-202.

Pedersen, M. (1988) The use of nucleotide polymorphisms and DNA karyo- typingin the identication of brewer's yeast strains and in microbiological control.:In Lin- skens, H. F. and Jackson, J. F. (eds.) Modern methods of plant analysis 7 : 180-194 Springer-Verlag. Berlin.

Porque, G., Baldesten, and Reichard, P. (1970) The involvement of the thioredoxin system in the reduction of methionine sulfoxide andsulfate. J.Biol. 245 : 2371- 2374.

Rahman, M. A., Nelson, Weissbach, H. and Brot, N. (1992) Cloning, sequencing, and expression of the Escherichia coli peptide methioninesulfoxide gene. J. Biol.

Chem. 267 : 15549-15551.

Ramos, C. and Calderon, I. L. Overproduction of threonine by Saccharomyces cerevisiae resistant to hydroxynorvaline. Appl. Env.Microbiol. : 1677-1682.

Ramos-Jeunehomme, C., de Keyser, L. and Masschelein, A. (1979) Formation de sub- stances aromatiques et cinetiques d'absorption des acides amines do mout. Proc. Eur.

Brew. Conv. Congr., Berlin, 505-519.

Rikkerink,H. A., Magee, B. B. and Magee, P. T. (1988) Opaque-white phenotype transi- tion: a programmed morphological in Candida albicans. J.Bacteriol. : 895- 899.

Rothstein, R. (1991) Targeting, Disruption, Replacement, and allele : Integrative DNAtransformation in yeast. Meth. Enzymol. :281-301.

Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989) Molecular cloning. laboratory man- ual. Cold Harbor Laboratory Press, Cold Harbor, New York.

Sangsoda, S., Cherest, H. and Surdin-Kerjan, Y. (1985) The expression of the MET25 gene of Saccharomyces cerevisiae isregulated transcriptionally. Mol. Genet.

200: 407-414.

Satoh, T. and Kurihara, F. N. (1987) Purification and properties of dimethylsulfoxide re- ductase containing a molybdenom from a photodenitrifier, Rhodopseudomonas sphaeroides f. s.denitrificans. Biochem. (Tokyo) 102 : 191-197.

Schaeffer, A. B. and Fulton, (1933) A simplifie of staining endospores. Sci- ence 77 : 194.

Scherer, S. and Davis, R. W. (1979) Replacement chromosome segments wih altered DNAsequences constructed in vitro. Proc. Natl. Sci. USA 76 : 4951-4955.

Schiestl, H. and Gietz, R. D. (1989) High efficiency transformation of intact yeast cells using single stranded nucleic acids as a carrier. Curr. Genet. 16 : 339-346.

Schuler, D., Altschul, F. and Lipman, D. J. (1991) A workbench for multiple alignment construction and analysis. Proteins : Structure, Function and Genetics 9 : 180-190.

Schwartz, D. C. and Cantor, C. R. (1984) Separation of yeast chromosome-sized DNAs by pulsed gradient gel electrophoresis. Cell : 67-75.

Sherman, F. (1991) Getting started with yeast. In: C.Guthrie G. R. Fink (eds.), Meth- ods in Enzymology : Guide to Yeast Genetics and Molecular Academic Press Inc., Diego, California, 3-21.

Sherman, F., Fink, G. R. and Hicks, J. B. (1979) Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Harbor, N. Y. 11724, pp 61-64.

Sigsgaard, P. and Rasmussen, J. N. (1985) Screening of the brewing performance of new yeast strains. J. Am. Soc. Brew. Chem. 43 : 104-108.

Sikorski, R. S. and Hieter, P. (1989) A system of shuttle vectors and yeast host strain de- signed for efficient manipulation DNA in Saccharomyces cerevisiae. 122 : 19- 27.

Singer, R. A., Johnston, C. G. and Bedard, D. (1978) Methionine analogs and cell division regulation in the yeast Saccharomyces cerevisiae. Proc.Natl. USA 75 : 6083-6087 Takahashi, T., Nagami, K., Nakatani, K. and Kumada, J. (1980) Hydrogen sulfide in brewing-II. MBAA Tech. 17 : 210-214.

Tezuka, H., Mori, T., Okumara, Y., Kitabatake, K. and Tsumura, Y. (1992) Cloning of a gene suppressing hydrogen sulfide by Saccharomyces cerevisiae its ex- pression in a brewing yeast. J. Am. Soc. Brew. Chem. 50 : 130-133.

Thomas, D., Barbey, R. and Surdin-Kerjan, Y. (1990) Gene-enzyme relationship in the sulfate assimilation of Saccharomyces cerevisiae. of the 3-phos- phoadenylylsulfate structural gene. J. Biol. 265 : 15518-15524.

Thomas, D., Barbey, R., Henry, D. and Surdin-Kerjan, (1992) Physiological analysis of mutants of Saccharomyces Saccharomyces cerevisiae in sulphate assimilation. Gen. Microbiol.

138 : 2021-2028.

Thomas, D., Cherest, H. and Surdin-Kerjan, Y. (1989) Elements involved S-adeno- sylmethionine-mediated of the Saccharomyces cerevisiae MET25 gene. Mol.

Cell. Biol. : 3292-3298.

Weiner, J. H., Rothery, R. A., Sambasivarao, and Trieber, C. A. (1992) Molecular analy- sis ofdimethylsulfoxide : a complex iron-sulfur molybdoenzyme ofEscherichia coli. Biochim. Acta 1102 : 1-18.

Yamamoto, I. , N., Ujiiye, T., Tachibana, M., Matsuzaki, M., Kajiwara, H., Wata- nabe, Y., Hirano, Okubo, A., Satoh, T. and Yamazaki, (1995) Cloning andnucleo- tide of the gene encoding dimethyl sulfoxide from Rhodo- pseudomonas sphaeroides f. sp.denitrificans. Biosci.Biotech. Biochem. : 1850-1855.

Yanisch-Perron, C.,Vieria, Messing, J. (1985) Improved phage cloning vectors and host strains : nucleotide of the M13mp18 pUC19 vectors. Gene 33 : 103- 119.

Yoshimoto, H., Momma, T., Fujiwara, Nagasawa, N. and Tamai, Y. (1996a) Cloning and characterization of the ATF2 encoding alcohol acetyltransferase the bottom fermenting yeast Saccharomyces pastorianus. NCBI Accession No. D86480.

Yoshimoto, H., Momma, T., Fujiwara, D., Sone, H. and Tamai, Y. (1996b) Charac- terization of the ATF genes encoding alcohol acetyltransferase the bottom fermenting yeast Saccharomyces pastorianus. 1996 Yeast Genetics and Molecular Biology Meeting. University of Wisconsin. August 6-11. p. 257.

Zinder, S. H. and Brock, T. D. (1978a) Dimethyl sulphoxide by micro-organisms.

J. gen. Microbiol. : 335-342.

Zinder, S. H. and Brock, T. D. (1978b) Dimethyl sulfoxide an electron acceptor for an- aerobic growth. Arch.Microbiol. : 35-40. INDICATIONSRELATING TO A DEPOSITED MICROORGANISM (PCTRule 13bis) A. The indications made belowrelate tothe microorganism referred in the description on page 22 vine 12 B. IDENTIFICATION OFDEPOSIT Further depositsare identified an additional sheet Name of depositary institution American Culture Collection Address of depositary institution(tnclufingpostd code and counhy) 10801 University Blvd. Manassas, VA 20110-2209 USA Date deposit Accession Number 8 July 1998 ATCC 74454 C-ADDTTIONAL INDICATIONS(leave blank ijnot applicabk) This information is continued an additional } As regards the respective Patent Offices of the respective desig- nated states, applicants request that a sample of the deposi- ted microorganisms only be made available to an expert nominated by the requester until the date on which patent is granted or the date on which application has been refused or withdrawn or is deemed to be withdrawn. D. DESIGNATED STATESFOR WHICH INDICATIONS AREMADE (i f tha catiorcr arc not jor al ! dcsignatcd Stalrs)E. SEPARATE FURNISHING OF INDICATIONS (leave bLqnk if not applicable) 'Ibeindieatioaslistedbelowwillbesubmitted to the InternationalBureaulater (specify thegeneralnatureoflfeindcationrGg., Accession Number o {Deposi) For receiving Office use only For International use only Tris set wasreceived with thcintcrnationatappiication jj This sheet wasreceived bythe International Bureau : Authorized o£ficer Authorized officer

INDICATIONSTO DEPOSITED MICROORGANISMS (PCT Rule 12bis) Additional sheet Into the microorganism indicated on page 20 of the description, the following microorganisms have been deposited with American Type Culture Collection 10801 University Blvd., Manassas, VA 20110-2209 USA onthe dates and under the accession numbers as stated below: Accession Date of Description Description number deposit Page No. Line No.

ATCC74450 08 July 1998 89 16 ATCC74451 08 July 1998 22 12 ATCC74452 08 July 1998 46 27 ATCC74453 08 July 1998 18 ATCC74455 08 July 1999 5 For all the above-identified deposited microorganisms, the following additional indicationsapply: As regards the respective Patent Offices of the respective designated states, the appli- cants request that a sample of the deposited microorganisms stated above only be made availableto an expert nominated by the requester until the date on which the patent is granted or the date on which the application been refused or withdrawn or is deemed to be withdrawn.