Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
IMPROVED METHOD FOR FERMENTATION BY REPAIRING YEAST GENES
Document Type and Number:
WIPO Patent Application WO/2007/026049
Kind Code:
A3
Abstract:
This invention relates to a method for increasing the speed and extent of fermentations. According to the method in the fermentation is used a yeast strain transformed with a DNA molecule comprising a nucleotide sequence that encodes part of the amino acid sequence for an a-glucoside transporter. The transformation converts the transformed host to have increased capacity for transport of maltose or maltotriose, or both compared to an untransformed host. The invention relates also to a method for preparing a yeast strain for fermentation, a new yeast strain and a DNA integration cassette.

Inventors:
LONDESBOROUGH JOHN (FI)
VIDGREN VIRVE (FI)
RUOHONEN LAURA (FI)
Application Number:
PCT/FI2006/000292
Publication Date:
May 10, 2007
Filing Date:
August 31, 2006
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
PANIMOLABORATORIO BRYGGERILABO (FI)
LONDESBOROUGH JOHN (FI)
VIDGREN VIRVE (FI)
RUOHONEN LAURA (FI)
International Classes:
C12N15/81; C12C11/00; C12C12/00; C12N
Foreign References:
EP1217074A12002-06-26
Other References:
STEARMAN R ET AL: "YIpDCE1 - an integrating plasmid for dual constitutive expression in yeast", GENE: AN INTERNATIONAL JOURNAL ON GENES AND GENOMES, ELSEVIER, AMSTERDAM, NL, vol. 212, no. 2, 8 June 1998 (1998-06-08), pages 197 - 202, XP004122921, ISSN: 0378-1119
KOMADA Y ET AL: "Improvement of Maltose Fermentation Efficiency: Constitutive Expression of MAL Genes in Brewing Yeasts", JOURNAL OF THE AMERICAN SOCIETY OF BREWING CHEMISTS, AMERICAN SOCIETY OF BREWING CHEMISTS, ST.PAUL, MN,, US, vol. 53, no. 1, 1995, pages 24 - 29, XP002996798, ISSN: 0361-0470
RAUTIO JARI ET AL: "Maltose transport by brewer's yeasts in brewer's wort.", JOURNAL OF THE INSTITUTE OF BREWING, vol. 109, no. 3, 2003, pages 251 - 261, XP001249133, ISSN: 0046-9750
VIDGREN VIRVE ET AL: "Characterization and functional analysis of the MAL and MPH loci for maltose utilization in some ale and lager yeast strains", APPLIED AND ENVIRONMENTAL MICROBIOLOGY, vol. 71, no. 12, December 2005 (2005-12-01), pages 7846 - 7857, XP002419699, ISSN: 0099-2240
Attorney, Agent or Firm:
SEPPO LAINE OY (Helsinki, FI)
Download PDF:
Claims:

CLAIMS

1. A method to increase the speed and extent of fermentations, which comprises that in the fermentation is used a yeast strain transformed with a DNA molecule comprising a nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter, said part not comprising a functional α-glucoside transporter, and said transformation converting the transformed host to have increased capacity for transport of maltose or maltotriose or both compared to an untransformed host.

2. The method according to claim 1, wherein the α-glucoside transporter is selected from the group comprising AGTl, MALxI, MPHx and MTTl.

3. The method according to claim 1 or 2, wherein the yeast host strain that is transformed with the DNA molecule contains before said transformation a gene comprising a defect or defects in the coding sequence for an α-glucoside transporter and the transformation corrects said defect or defects.

4. The method according to claim 3, wherein the DNA molecule comprises a correct version of the part of the coding sequence containing a defect or defects in the said host gene.

5. The method according to any one of claims 1 to 4, wherein the nucleotide sequence encoding part of the amino acid sequence comprises the ATG start codon and is functionally fused to a yeast promoter.

6. The method according to claim 5, wherein the promoter is not or not strongly repressed by glucose and does not require induction by maltose or maltotriose

7. The method according to claim 5 or 6, wherein the 5 '-side of said promoter is joined to a flanking sequence homologous to a sequence in the promoter of the said host gene of claim 3 or 4.

8. The method according to claim 7, wherein the flanking sequence comprises at least

10 nucleotides.

9. The method according to any one of claims 3 to 8, wherein the DNA molecule which comprises the correct version of the part of the coding sequence comprises a sequence of at least 10 nucleotides after the position of the most 3 '-defect in the coding sequence of the host gene, said at least 10 nucleotides being homologous to the corresponding at least 10 nucleotides in the host gene after the position of the said most 3 '-defect.

10. The method according to any one of the preceding claims, wherein the transformed host does not contain any DNA from any non-Saccharomyces organism.

11. The method according to any one of the preceding claims, wherein the yeast host is a brewer's yeast.

12. The method according to any one of the preceding claims, wherein the defect or defects is/are selected from the group comprising a frame shift, a premature stop codon and a codon for a non-optimal amino acid.

13. The method according to claim 12, wherein the gene having a defect or defects does not encode a functional α-glucoside transporter.

14. The method according to claim 12, wherein the gene having a defect or defects encodes an α-glucoside transporter having lower activity or narrow substrate specificity compared to an α-glucoside transporter which does not have the same defect or defects.

15. The method according to any one of the preceding claims, wherein the gene comprising the defect is AGTl.

16. A method for preparing a yeast strain for fermentation, which comprises the steps:

-identifying in the target yeast a defective target α-glucoside transporter gene comprising a defect or defects, said defect or defects being selected from the

group comprising a frame shift, a premature stop codon and a codon for a non- optimal amino acid;

-identifying in the donor yeast the corresponding gene comprising the correct sequence; -constructing a DNA cassette that comprises the part of the donor gene, corresponding to the defect or defects of the target gene, and optionally a yeast- derived promoter;

-transforming the target yeast with the cassette, and

- screening for transformants having increased capacity to transport maltose or maltotriose or both compared to the capacity of the target strain.

17. The method according to claim 16, wherein the promoter is not or is not strongly repressed by glucose and does not require induction by maltose or maltotriose.

18. The method according to claim 16 or 17, wherein the cassette comprises a

Saccharomyces promoter.

19. The method according to any one of claims 16 to 18, wherein in the transformants one, two or more integrations have occurred in different copies of the chromosomes carrying the target gene.

20. The method according to any one of claims 16 to 19, wherein the cassette comprises flanking sequences that are homologous to the target gene.

21. The method according to any one of claims 16 to 20, wherein the cassette is integrated by homologous recombination between the flanking sequences of the cassette and the corresponding sequences in the target gene resulting in integrative replacement of the defect or defects in the target gene by the corresponding correct sequence of the donor gene.

22. The method according to any one of claims 16 to 21, wherein the method does not result in elimination of any functional gene from the genome of the target yeast.

23. The method according to any one of claims 16 to 22, wherein the gene having a defect or defects does not encode a functional α-glucoside transporter.

24. The method according to any one of claims 16 to 22, wherein the gene having a defect or defects encodes an α-glucoside transporter having lower activity or narrow substrate specificity compared to an α-glucoside transporter which does not have the same defect or defects.

25. A yeast strain for fermentation prepared by the method according to any one of claims 16 to 24 and having increased capacity for transport of maltose or maltotriose or both compared to the untransformed strain.

26. A DNA integration cassette comprising a Saccharomyces promoter functionally fused at its 3 '-end to a part of the coding sequence of a Saccharomyces gene for an α- glucoside transporter, said part comprising the ATG start codon and encoding an amino acid sequence that does not comprise a functional α-glucoside transporter and said promoter being fused at its 5 '-end to a flanking sequence that is homologous to the native promoter of a target gene or pseudogene for an α -glucoside transporter in the host yeast.

27. The DNA integration cassette according to claim 26, wherein the α-glucoside transporter is Agtl .

28. The DNA integration cassette according to claim 26 or 27, wherein the sequence of the DNA cassette is the sequence as shown in Figure 8 or 9.

29. A yeast strain transformed with the DNA integration cassette of any one of claims 26 to 28.

30. Use of the yeast strain according to claim or 25 or 29 for fermentation.

Description:

Improved method for fermentation by repairing yeast genes

Field of the Invention This invention relates to methods for changing the physiological behaviour of yeasts, in particular brewer's yeasts, by altering the arrangement of their genetic material. In particular it relates to methods to increase the fermentation capacity of brewer's yeast by increasing the activity of their maltose and maltotriose transporters.

Background

Brewer's yeast ferment the sugars in brewer's wort to produce mainly ethanol. At the same time they also produce various chemical side products that contribute to the flavour of the beer. The main sugars in wort are maltose (about 60 % of the total), maltotriose (about 20 %) and glucose (about 20 %). The sugars are fermented in the order glucose, maltose, maltotriose. It is believed that the rate at which maltose and maltotriose can enter the yeast cell (the uptake rate or transport rate) limits the speed of wort fermentations (see, e.g., Kodama et al [1995] J.Am. Soc. Brew. Chem. 53(l):24-29, Rautio and Londesborough [2003 ] J. Inst. Brew. 109(3):251-261). Faster fermentations would improve the economics of brewery fermentations. It also happens that at the end of brewery fermentations the beer still contains some sugar, so-called residual sugar. If there is too much residual sugar, the yield of ethanol and beer on wort carbohydrates is decreased and the flavour of the beer is altered. One cause of residual sugars appears to be that the capacity of the yeast to transport maltose and maltotriose into the cell decreases towards the end of fermentations.

Breweries now-a-days use increasingly strong (higher gravity) worts that contain up to 200 g of sugar/L, more than twice the traditional levels. Use of these very high gravity worts can improve the economics of beer production and also decrease the environmental costs (e.g., there are savings in energy consumption and the use of cleaning materials). To achieve these benefits it is important that the speed of fermentation is adequate and, especially, that the amount of residual sugars does not increase by more than the increase in wort gravity. However, residual sugars are a major problem with very high gravity fermentation, and unacceptably large amounts of maltose and, especially, maltotriose are frequently left unfermented.

Traditionally brewer's yeast collected from one fermentation was used to start the next fermentation. This process has been continued for hundreds of years. It is thus expected that present day strains of brewer's yeasts have been automatically selected to survive and reproduce well under brewery conditions. One example of this evolution is that strains of brewer's yeast contain multiple genes for α-glucoside transporters. Maltose and maltotriose are both α-glucosides. Some α-glucoside transporters have broad substrate specificity and can carry several α-glucosides including maltose and maltotriose. Some have narrow substrate specificity and can carry, e.g., only maltose. It is believed that these genes together endow the yeast cell with a number of different maltose and maltotriose transporters that give the cell a large capacity to transport these sugars into the cell, where they are then further metabolised to provide energy and cell material for the yeast and ethanol and flavour compounds for the brewer. Nevertheless, the evidence is that this transport step is not fast enough and limits the speed of wort fermentations (Kodama et al [1995] J.Am. Soc. Brew. Chem. 53(l):24-29, Rautio & Londesborough [2003 ] J. Inst. Brew. 109(3):251-261, J. Inst. Brew. 109(3):251-261).

Although brewer's yeasts have evolved for hundreds of years in brewer's wort, high and very high gravity worts have been introduced only during the last decades. By the time of their introduction, most breweries no longer re-cycled their yeast continuously. Instead pure cultures of the brewer's own strains are stored in a collection and at regular intervals fresh lots of yeast are grown up from these cultures and taken into use. This means that the natural evolutionary adaptation to brewery conditions has stopped. This in turn suggests that present day strains could be significantly improved: they have not reached an evolutionary limit. One way to improve brewer's yeast strains is to isolate spontaneous or mutagen-induced variants with improved performance. Another way is to combine the genetic potential of different yeast strains to create a new improved strain. This can be done by hybridisation between strains. However, hybridisation is a long and inefficient process that cannot be applied to all brewer's strains, and the genetic outcome is uncertain. Usually many other characters will be altered as well as the desired character. In the brewery context altered factors are likely to include flavour production and process behaviour. Another approach is to use genetic engineering techniques to rearrange the DNA of a particular yeast strain or to move DNA from one yeast strain to another. This

can be done in such a way that the new yeast strain does not contain DNA from any non-

Saccharomyces organism.

There are two main kinds of brewer's yeasts. Ale strains are Sacharomyces cerevisiae, but they are genetically more complicated than laboratory strains of S. cerevisiae: in particular they are usually aneuploid, i.e., each cell contains more than one copy of each chromosome but not necessarily the same number of copies of all chromosomes. Ale strains are probably closely related to baker's yeasts. Lager strains on the other hand are thought to be the result of one or more hybridisation events between a S. cerevisiae strain and some other, closely related, yeast species, possibly S. bayanus. Ale, lager, and baker's yeasts and laboratory strains of S. cerevisiae share much of their genetic information and provide a large pool of genetic variation that can be usefully exploited.

As discussed above the speed and extent of fermentation of brewer's wort by brewer's yeast is limited by the capacity of brewer's yeast strains to transport maltose and maltotriose across their cell membranes into the cytosol (Kodama et al [1995] J. Am. Soc. Brew Chem. 53, 24-29, Rautio & Londesborough [2003] J. Inst. Brew..109, 251-261). By "capacity to transport maltose" we mean the speed at which yeast cells can take up maltose from the medium across the cell membrane and into the cell, measured, e.g., as μmol of maltose/minute/g of dry yeast. In the prior art some attempts have been made to speed up the transport of maltose. For example, Kodama et al have described in a Japanese patent publication (Kodama etal , JP 6245750 A) and in the scientific literature (Kodama et al[1995] J. Am. Soc. Brew Chem. 53, 24-29) a brewer's yeast containing high-copy- number plasmids comprising a maltose transporter gene (MAL6T) under the control of a constitutive promoter from the yeast gene for glyceraldehyde 3-phosphate dehydrogenase. This transformed strain fermented VHG (24 0 P) wort faster than the control strain. The transformed strain contains non-yeast (bacterial) DNA, which is a disadvantage for applications in the food and drink industry because of public attitudes and official regulations. Furthermore, the high-copy-number plasmids are not genetically stable and selection pressure is required to maintain these plasmids in the yeast. Kodama et al, carried out their fermentations in wort containing the antibiotic G418 to provide selection pressure.

Patent publication EP 0306107 Bl describes baker's yeast strains in which the complete coding sequence of a maltose transporter gene has been provided with a new promoter and terminator (e.g. from the yeast ADHl gene) and this construct is then integrated into a yeast gene. This strategy results in the loss of one gene from the yeast. EP 0306107 Bl recognised this problem, and therefore chose to integrate the DNA construct into the SITl gene (EP 0306107 Bl, loc .cit see pl4, lines 17 et seq.), in the belief that the SITl gene is not needed by the baker's yeast during the industrial process. Nevertheless, the transformed yeast lacks at least one copy of the SITl gene. This could have unforeseen consequences, especially in the beer production process, where the exact physiological behaviour of the yeast impacts on flavour formation and yeast handling. Furthermore, the use of the complete ORF of the maltose transporter gene, plus new promoter and terminator and plus new flanking sequences to provide homology to the integration site (SITl) results in a large integration cassette. In general, transformation frequencies decrease when cassettes size increases. Of course the labour involved in constructing the cassette also increases.

Summary of Invention

It is an aim of the present invention to eliminate at least some problems associated with the prior art.

This invention provides an improved method to increase the speed and extent of fermentations. According to one preferred embodiment of the invention the speed and extent of fermentations is increased by using a yeast strain transformed with a DNA molecule comprising a nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter. More specifically, a yeast strain is transformed with a DNA molecule comprising a nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter, said transformation converting the transformed host to have increased capacity for transport of maltose or maltotriose or both compared to an untransformed host.

As discussed here earlier brewer's yeasts are believed to contain multiple genes encoding α-glucoside transporters. The present invention is based on the finding that in some strains of brewer's yeasts some of the genes encoding α-glucoside transporters contain defects in

their coding regions (i.e., sequence errors). In this invention we show that these defects or mistakes can be repaired using DNA molecules that do not encode functional proteins and comprise DNA sequences isolated from other brewer's yeast strains or other strains of Saccharomyces cerevisae. The repaired genes encode functional α-glucoside transporters and the transformed strains have increased capacities for transport of maltose or maltotriose or both and ferment brewer's worts faster and more completely compared to the untransformed strains.

Multiple genes encoding α-glucoside transporters comprise AGTl, MALxI, MPHx and MTTl (also known as MTYl). These genes are all similar to each other. Thus, different MALxI genes share 97 % identity, AGTl is 57 % identical to MAL61, MTTl is 90 % identical to MAL31 and 54 % identical to AGTl, and MPHx is 75 % identical to MAL61 and 53 % identical to AGTl. Furthermore, these genes are characteristically found in sub- telomeric regions of chromosomes. Thus, they form a closely related family. We have found that some of these genes are defective in some strains of brewer's yeasts. For example some lager yeast strains contain AGTl genes that cannot encode full length permeases, because their coding sequences are interrupted by premature stop codons and frame shifts. If these genes produce protein products, they are truncated proteins, lacking at least one transmembrane domain and unable to transport maltose. We disclose a simple and effective method to correct the sequence of these genes so that they encode full length transporters. Transformed strains with this correction have increased maltose/maltotriose transport activity and ferment brewer's worts faster and more completely.

As well as repairing a defective gene, the same method can also be used simultaneously to replace the promoter of the gene by a new promoter with desired characteristics. For example, the expression of many genes for maltose and maltotriose transporters is strongly repressed by glucose and the expression of some of these genes requires induction by maltose or maltotriose. The promoters can be replaced by promoters that are not or are not strongly repressed by glucose and do not require induction by maltose or maltotriose.

As well as repairing genes with mistakes such as premature stop codons and frame shifts, the same method can also be used to replace non-optimal amino acids. Here non-optimal amino acids includes both amino acids that prevent the protein from functioning as a transporter and also amino acids that limit the function of the transporter, for example by

limiting its substrate specificity. We disclose an example of a codon change in an AGTl gene that prevents the transporter from functioning. It is well known to those skilled in the art that the substrate specificity of catalytic proteins is determined by the nature of critical amino acids in the substrate binding sites.

The present invention thus provides a method to increase the speed and extent of fermentations, which comprises that in the fermentation is used a yeast strain transformed with a DNA molecule comprising a nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter, said part not comprising a functional α- glucoside transporter, and said transformation converting the transformed host to have increased capacity for transport of maltose or maltotriose or both compared to an untransformed host.

More specifically, the method of the invention is characterized by what is stated in claim 1.

The present invention provides also a method for preparing from a yeast strain (the target yeast) an improved strain for fermentation, in particular for wort fermentation. The method comprises the steps: identifying in the target yeast a target α-glucoside transporter gene comprising a defect or defects;

- identifying in a donor yeast the corresponding gene comprising the correct sequence;

- constructing a DNA cassette that comprises the part of the donor gene, corresponding to the defect or defects of the target gene, - transforming the target yeast with the cassette, and screening for transformants having desired integration.

More specifically, the method for preparing a yeast strain is characterized by what is stated in claim 16.

By a defect or defects in the target gene is meant a frame shift, a premature stop codon or a codon for a non-optimal amino acid. A defective gene having a defect or defects may not encode a functional α-glucoside transporter or a defective gene may encode an α- glucoside transporter having lower activity or narrow substrate specificity compared to an

α-glucoside transporter which does not comprise the defect or defects, e.g. a naturally occurring α-glucoside transporter. The defect or defects may be corrected with a DNA cassette that comprises the part of the donor gene, corresponding to the defect or defects of the target gene. Alternatively or in addition the cassette may comprise a new promoter that is not or is not strongly repressed by glucose and does not require induction by maltose or maltotriose. Transformants having desired integration may be screened for. By desired integration is meant that one, two or more integrations have occurred in different chromosomes i.e. in one or more copies of the chromosomes carrying the target gene and that the transformants have increased capacity to transport maltose or maltotriose or both compared to the capacity of the target strain.

The present invention provides also a DNA integration cassette and improved yeast strains transformed according to the methods and with the DNA integration cassette of the invention. The yeast strains of the invention can be used for fermentation, in particular for wort fermentation.

More specifically, the DNA integration cassette is characterized by what is stated in the claim 26, and improved yeast strains by what is stated in claims 25 and 29.

A major advantage of the invention is that the fermentative capacity of a particular strain is increased by a minimal genetic change, which can be as little as the gain, loss or substitution of one nucleotide. This means that other properties of the strain, such as flavour production and process behaviour are not changed. Individual brewers can apply the invention to their own strain(s) without unwanted changes in the character of their own beers or without large changes in process conditions. However, the invention will shorten fermentation times, increase the conversion of sugars to ethanol and decrease the levels of residual sugars. Furthermore, the invention facilitates the use of higher gravity worts without leaving proportionately more residual sugars. A significant advantage of the invention is that the transformed strains do not contain any DNA from organisms other than Saccharomyces.

One novel aspect of the invention is that it is directed to repairing genes or pseudogenes that are already present in the target yeast strain, and to carrying out this repair in vivo. Repair includes providing such genes with promoters that are more appropriate to a

particular industrial application. Because the focus is on repair of existing genes, DNA integration cassettes of the invention are provided with flanking sequences that are very similar or identical to sequences in the gene to be repaired. This leads to another advantage of the invention, high frequencies of successful transformation with integration at the desired locus. This contrasts with the low transformation frequencies often obtained when industrial yeasts are transformed with integration cassettes having flanking sequencies that are obtained from a laboratory strain.

Improved yeast strains are obtained in cases where the cassette has integrated into one or more copies of the defective α-glucoside transporter gene in an aneuploid or alloploid target strain. In other words, the defect need not be corrected in all copies of the target genes. Similarly, it is not neccessary to replace the promoters of all copies of the target genes by promoters that are not or are not strongly repressed by glucose and do not require induction by maltose or maltotriose. By the screening methods disclosed here it is possible to find integrants, in which the defective gene has been corrected or the promoter replaced by another promoter. For example, successfull integrants grow faster on maltotriose or maltose in the presence of antimycin A than does the untransformed target strain. By the methods of the invention it is possible to find transformants that have increased capacity to transport maltose or maltotriose or both compared to the capacity of the target strain. Also, by Southern analysis of the transformants, the size of the cleavage product indicates the success of the integration. The nature of the screened transformant can be further studied and confirmed by sequencing using appropriate PCR primers. By the methods disclosed here a person skilled in the art can produce improved yeast strains in a repeatable manner.

Brief description of figures

Fig 1. Schematic diagram of two cassettes envisaged in the invention and the recombination events leading to their integration in the desired way. Target strain DNA is shown by narrow line, donor strain DNA by dotted line, other yeast DNA by thick line. Dashed lines indicate the parts of the target sequence containing one or more defects. The X symbols indicate desired cross-over events.

Fig 2. PFGE separation of chromosomes from laboratory and brewer's strains. The gels were loaded with 8 x 10 cells/lane of the brewer's strains and 30 x 10 6 cells/lane of the

laboratory strains. Strains were: lane 1, chromosome marker strain YNN295; lane 2, lager strain A15; lane 3, lager strain A24; lane 4, lager strain A64; lane 5, lager strain A72; lane 6, ale strain A60; lane 7, ale strain Al 79; lane 8, laboratory strain S288C; lane P 5 laboratory strain S150-2B; lane 10, laboratory strain CEN.PK2-1D; lane 11, laboratory strain RH144-3A; lane 12, chromosome marker strain YNN295. Chromosomes are identified on the left: chromosomes VII and XV are not resolved; chromosome II travels immediately above chromosome XIV.

Fig 3. Detection of MAL loci. A PFGE gel was loaded with about 30 x 10 6 cells/lane. Strains were: lane 1, Marker YNN 295; lane 2, A15; lane 3, A24; lane 4, A64; lane 5 A72; lane 6, Al 79; lane 7, Al 80; lane 8, Al 81; lane P, A60; lane 10, S150-2B; lanell, CEN.PK2-1D and Ianel2, RH144-3A. Panel A shows the separated chromosomes stained with ethidium bromide. Chromosome IV and the duplex VlIfKV are indicated. The gel was then blotted and the blot was hybridised with the following probes: B, AGTl; C, MALI 3 (AGTl); D, MAL61; E, MAL62 and F, MAL63. The MAL loci are identified on the left of panel D. The image for panel F was darkened to make the bands more visible.

Fig 4. Inhibition of maltose transport by trehalose. Rates of zero-trans uptake of maltose were measured at 5 mM 14 C-maltose in the presence of the indicated concentrations of trehalose. For each strain, the rate in the absence of trehalose was set at 1.00. The figure shows the reciprocal rates (1/ V) plotted against trehalose concentration for two ale yeasts, A60 (x) and A179 (•) and three lager yeasts, A15 (♦), A24 (■) and A64 (A). Error bars show the range between two or three independent experiments.

Fig 5. Multicopy plasmid used to express normal and potentially defective maltose transporter genes in a laboratory strain. This example plasmid contains the AGTl gene from the lager strain Al 5 inserted between PGKl promoter and terminator sequences. The plasmid also contains the URA3 gene to allow selection in an ura3-laboratory strain.

Fig 6. Alignment of the AGTl promoter sequences of Al 5 and A24 lager strains, A60 ale strain and the Saccharomyces Genome Data Base (SGDB) sequence.

Fig 7. The integration cassettes for Repair and New Promoter made in Example 6.

Fig 8. Vector sequence containing the long form of the Repair and New Promoter cassette disclosed in Example 6. The two MspAlI restriction sites used to cut out the cassette and the ATG start codon of the AGTl gene are shown bold and underlined. The PGKl promoter sequence is shown in italics.

Fig 9. Vector sequence containing the short form of the Repair and New Promoter cassette disclosed in Example 6. The two MspAlI restriction sites used to cut out the cassette and the ATG start codon of the A GTl gene are shown bold and underlined. The PGKl promoter sequence is shown in italics.

Fig 10. Attenuation profiles during fermentations of 15 0 P wort by duplicate growths of the untransformed target strain (Al 5 A and Al 5 B), duplicate growths of Integrant 1 (Int 1 A and Int 1 B) and single growths of Integrants 2 and 14 (Int 2, Int 14). The apparent extract (a measure of the remaining unfermented carbohydrate) was calculated from daily measurements of the specific gravity.

Fig 11. Attenuation profiles during fermentations of 25 0 P wort by duplicate growths of the untransformed target strain (Al 5 A and Al 5 B), duplicate growths of Integrant 1 (Int 1 A and Int 1 B) and single growths of Integrants 2 and 14 (Int 2, Int 14). The apparent extract (a measure of the remaining unfermented carbohydrate) was calculated from daily measurements of the specific gravity.

Fig 12. Growth and flocculation of the target strain (Al 5) and three integrants during the wort fermentations shown in Figs 10 and 11. Solids (mainly yeast) in suspension were collected by centrifugation, washed with distilled water and dried at 105 0 C to constant mass.

Fig 13. Yeast-derived volatile aroma compounds in beers produced from 25 0 P wort by the untransformed yeast (Al 5) and three integrants.

Detailed Description.

Definitions

By "a nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter " is meant such part of a gene encoding an α-glucoside transporter that is capable of converting the transformed host to have increased capacity for transport of maltose or maltotriose or both compared to an untransformed host. In particular, "a nucleotide sequence that encodes the part of the amino acid sequence for an α-glucoside transporter" means the nucleotide sequence capable of converting a target gene comprising a defect to be functional or to function more effectively. Preferably the " nucleotide sequence that encodes the part " comprises a corresponding correct sequence of the target gene, but does not comprise the complete gene, in other words "a nucleotide sequence that encodes part of the amino acid sequence" does not encode full length amino acid sequence for an α-glucoside transporter. Optionally, the "nucleotide sequence that encodes part of the amino acid sequence for an α-glucoside transporter " is functionally fused to a promoter with desired characteristics.

By "defective α-glucoside transporter gene" or an α-glucoside transporter gene having a defect or defects is meant any α-glucoside transporter gene that because of a premature stop codon or ύ frame shift or both does not encode a full length functional transporter or because of encoding a non-optimal amino acid at one or more positions in the amino acid sequence of the transporter produces a transporter with undesired low activity or narrow substrate specificity compared to an α-glucoside transporter that has other amino acids at the position or positions of the non-optimal amino acid or amino acids, for example a naturally occurring α-glucoside transporter. The premature stop codon, frame shift and codon or codons encoding said non-optimal amino acid or amino acids are each included also by the term "mistake" in a gene. If a gene contains more than one defect, then by the last defect or most 3 '-defect we mean the one that is furthest from the ATG start codon and closest to the terminator.

By "α-glucoside transporter" is here meant transporters which carry one or more α- glucosides including maltose or maltotriose or both. Maltose and maltotriose are both α- glucosides. Some α-glucoside transporters have broad substrate specificity and can carry several α-glucosides including maltose and maltotriose. Some have narrow substrate specificity and can carry, e.g., only maltose.

By "capacity to transport maltose" we mean the speed at which yeast cells can take up maltose from the medium across the cell membrane and into the cell, measured, e.g., as μmol of maltose/minute/g of dry yeast.

By the expression "to have increased capacity for transport of maltose or maltotriose or both " is meant that a given mass of the transformed yeast is capable of transporting maltose or maltotriose or both into the yeast cell faster than the same mass of an untransformed host after growth under the same conditions. The increase is at least 10 %, preferably it is at least 30 % compared to the level transported to an untransformed host. More preferably the increase is at least 50 %.

The maltose and maltotriose transport capacities of the transformants can be measured as described in Example 2.

By "(pseudo)genes" is meant here both functional and non-functional genes.

"Flanking sequences" mean here DNA sequences of each end of the DNA cassette, which are identical or very similar (or, as is often said, homologous) to the host DNA. They are used to direct the cassette to the desired site of the genome. The flanking sequences can be of any length, preferably they are at least 10 nucleotides, more preferably 10 - 1000 nucleotides, typically 40 - 400 nucleotides.

According to one aspect of the invention a target brewer's yeast can be repaired by the steps:

1. A defective gene for a maltose transporter is identified in the target yeast and sequenced to reveal the nature of the defect or defects. This is now called the target gene.

2. The corresponding functional gene is identified in another brewer's yeast strain, or baker's yeast strain or some other S . cerevisiae strain and sequeneed. This is now called the donor gene.

3. A DNA cassette is made that comprises the part of the donor gene corresponding to the defective part of the target gene, this part being flanked in the cassette by DNA sequences homologous to the target gene sequences on either side of the defective part of the target gene. Optionally, if the DNA sequence used includes the ATG start codon, a new (yeast- derived) promoter can be functionally attached before the ATG start codon. In this case the 5 '-homologous flanking sequence is taken from the promoter of the target gene and is attached to the upstream end of the new promoter in the cassette. This has the advantage that not only will the defective gene in the host be corrected, but it will also now be driven by a new promoter, whose properties can be chosen as desired.

4. The target strain is then transformed with the cassette. Homologous recombination between the flanking sequences of the cassette and the corresponding sequences in the target gene results in integrative replacement of the defective part of the target gene by the corresponding functional part of the donor gene. If a new promoter was included in the cassette, this will also be integrated into the host genome and drive the corrected gene. These integration events are shown in Fig 1. Transformants are screened to identify those in which the integration has occurred as planned. Some convenient screening procedures are described below and in the Examples.

6. Because brewer's strains are aneuploid (and lager strains are alloploid) transformants may be recovered in which one, two or more integrations have occurred in different copies of the chromosomes carrying the target gene. Individual transformants can be characterised by Southern and PCR analyses to show how the integration has occurred. The maltose and maltotriose transport capacities of the transformants can be measured to identify those with desired properties. Fermentation tests are performed on small and pilot scales to identify those transformants with improved performance. Examples 8 and 11 disclose that the desired effects on α-glucoside transport and wort fermentation are obtained when either all or less than all copies of the target gene are repaired by integration of the cassette.

7. The transforniants chosen by the above tests are then used in the brewery in essentially the same way as the target strain was used. The transformants are integrants so that no special handling is required to ensure their genetic stability. Apart from short stretches (2 to 20 nucleotides) of synthetic DNA, all their DNA is derived from brewer's or baker's yeast strains or laboratory strains of S. cerevisiae, so that no precautions appropriate to genetically modified organisms are required. Fermentations will be faster, so that yeast should be cropped and chill-back applied sooner than with the host strain. It is expected that the flavour characteristics of the produced beer will be the same as that produced by the target yeast under the same conditions. Because residual sugars at the end of fermentation are lower than with the target strain, either a drier beer can be produced, or the fermentation can be stopped earlier (by chilling). However, this property of the transformants advantageously facilitates the use of higher gravity worts: usually with higher gravity worts proportionally more sugars, especially maltotriose, remain unfermented, and this problem is minimised by using the transformed strains of the invention.

This present invention is now described as a number of steps, some of which are optional in certain cases. For example, the information disclosed in this specification suggests that many lager strains contain defective AGTl genes similar to those of lager strains Al 5 and A24, so that the DNA cassette disclosed in Example 6 is likely to be a successful tool to carry out the invention with many lager strains. In such cases the invention can be practised by transforming the target lager strain with the cassette of Example 6 and selecting and screening transformants as described in Steps 7, 8 and 9. At least some ale strains (possibly all) and possibly some lager strains contain functional AGTl genes without the premature stop codon and frame shift found in lager strains Al 5 and A24.

However, we disclose in Example 5 that the promoters of Hie AGTl genes of ale strain A60 and lager strain Al 5 are almost identical from nucleotides -1 to -564. The cassette of Example 5 will work also with many ale strains and with many lager strains that do not contain in their AGTl genes the premature stop codons found in Al 5 and A24. In these cases transformation with the cassette of Example 6 will replace the resident promoter of these yeasts' AGTl genes by the strong constitutive PGKl promoter and at the same time can replace the amino acid sequence of the encoded transporter by that of the Agtl transporter from strain A60 over the region of amino acids 1 to at least 395. Because the Agtl transporter of A60 is an efficient carrier of maltose and maltotriose (see Examples 2

and 4) and because the Agtl proteins of A60 and other brewer's yeasts exhibit different sequences in this region (see Table 5) such transformation will often repair the target gene, by changing its amino acid sequence and improving the catalytic properties of its encoded transporter.

Thus, for a large number of both lager and ale strains the invention can be practised by transforming the target strain with the DNA cassette of Example 6 as described in Step 7 and then selecting and screening transformants as described in Steps 8 and 9. These procedures are quickly and easily performed by a person skilled in the art. If they do not yield transformants with the desired properties, then the invention is practised by first investigating the α-glucoside transporter genes of the target strain according to some or all of Steps 1 -5, and then designing and constructing a suitable DNA cassette according to Step 6 before carrying out the transformation and selection and screening of transformants.

Summary of Steps:

Step 1: Identification of (pseudo)genes for α-glucoside transporters in the target strain.

Step 2: Kinetic characterisation of maltose and maltotriose transport by the target strain.

Step 3: Sequencing a potentially defective α-glucoside transporter gene in the target yeast.

Step 4: Confirmation of the non-functionality of the potentially defective gene. Step 5: Finding from another yeast a functional version of the target yeast's defective gene.

Step 6: Construction of the repair cassette.

Step 7: Transformation and selection

Step 8: Analysis of the biochemical and fermentation characteristics of the transformants.

The teaching of steps 1 to 8 can be applied by a person skilled in the art to any α- glucosidase and to any Saccharomyces yeast strain. In particular, the teaching can be applied to any brewer's yeast strain to have increased capacity for transport of maltose or maltotriose or both.

Step 1. Identification of (pseudo)genes for α-glucoside transporters in the target strain

Chromosomal locations of many yeast genes for α-glucoside transporters are known, including those of AGTl, MPHx and the different MALxI (x = 1-4 or 6). The sequences of these genes are available, e.g. in the Saccharomyces Genome Data Base (SGBD). This makes it easy to design hybridisation probes that distinguish between AGTl, MPHx and

MALxI genes. MTTl genes for maltose/maltotriose transporters and suitable probes and PCR primers for these have also been described (Diervorst et al [2005] Yeast 22, 775-788). The chromosomes of the target brewer's yeast are separated by pulsed field gel electrophoresis (PFGE), e.g., as described in Example 1. Blots of the gels are hybridised with specific probes to reveal which genes for α-glucoside transporters are present in the yeast. The method reveals both functional genes and also genes with small sequence errors that render their coding regions non-functional. For example, single nucleotide changes that introduce stop codons, frame shifts or both can render the gene non-functional. Sequence changes that alter the amino sequence of the encoded protein can also decrease or abolish its activity as an α-glucoside transporter, rendering the gene defective. These changes can be so small that they do not prevent the hybridisation of specific probes.

Example 1 shows how this step was carried out for several industrial brewer's strains. The ale and lager strains analysed apparently contained several MALxI genes (expected to encode maltose transporters that cannot carry trehalose) and also AGTl genes (expected to encode maltose transporters that can carry trehalose). This agrees with Jespersen et al, (1999) Appl. Environ. Microbiol. 65:450-456. The lager yeasts also contained MPHx genes, expected to encode maltose transporters that cannot carry trehalose.

Step 2. Kinetic characterisation of maltose and maltotriose transport by the target strain

As an optional step to help identify a defective gene or gene(s), the characteristics of maltose and maltotriose transport by the target strain can be determined. There is literature information about the substrate specificity of maltose transporters of yeast (e.g., Day et al, 2002, Yeast 19:1015-27; Day et al, 2002, Appl. Environ. Microbiol. 68: 5326-5335; Chang et al, 1989, J. Bacteriol. 171:6148-6154). Maltose transporters encoded by AGTl can carry also trehalose, maltotriose and α-methylglucoside, those encoded by MPHx genes can carry maltose and maltotriose but not trehalose, those encoded by MAL31 and MAL61 genes can carry maltose and probably maltotriose but not trehalose. Studies of the substrate specificity of maltose transport and in particular the inhibition of maltose uptake by other α-glucosides indicate which transporters function in the target strain. In Example 2 we show that maltose transport by several lager strains is only weakly inhibited by trehalose, maltotriose and α-methylglucoside whereas transport of maltose by ale strains was strongly inhibited by these other α-glucosides. This indicated that transporters encoded by AGTl

were not important in the lager strains. It also indicates that transporters encoded by

MALxI genes are not important in ale strains, so that it is likely some of the MALxI genes found in ale strains will prove to be defective as (see below) the AGTl genes of many lager strains are defective.

Step 3. Sequencing of a potentially defective α-glucoside transporter gene in the target yeast

The sequences of AGTl, MALxI, MPHx and MTTl genes are available, for example in the

SGDB and NCBI data bases. PCR primers are designed using this sequence information and the genes are amplified using appropriate forward and reverse primers. The amplified fragment can include the whole open reading frame (coding sequence) of the gene or also the promoter. In some cases whole genomic DNA of the target yeast is used as template. For the MALxI genes, however, there can be several different copies on different chromosomes (MALIl on chromosome VII, MAL21 on chromosome II, MAL31 on chromosome III and so on). It is therefore an advantage to use chromosomes isolated by PFGE (as described in Example 1) as templates, because the sequence information is then immediately related to a particular MALxI gene on a particular chromosome. The PCR fragments are then sequenced. For example, they can be cloned using the TOPO TA cloning kit (Invitrogen) and propagated in E. coli. Plasmid DNA is isolated from independent E. coli clones and sequenced using universal Ml 3 Forward and Reverse primers to sequence the start and end of the pCO-TOPO plasmid ligated gene. Internal sequencing primers are then designed on the basis of the first sequence information.

Because brewer's yeasts are aneuploid and lager strains alloploid, the same strain can contain different versions of a particular maltose transporter gene carried on different copies of the same chromosome. For example, if a yeast cell contains several different copies of chromosome VII it can also contain two or more versions of AGTl each on a different copy of chromosome VII. Mistakes can occur in PCR and sequencing reactions. Therefore several independent clones should be sequenced to identify possibly defective genes. If a sequence error is found, e.g. a premature stop codon, a frame shift or an amino acid change that is likely to be unacceptable (e.g. the exchange of a proline for any other amino acid, since prolines are usually strongly conserved), a single DNA fragment containing that part of the sequence can be amplified by several independent PCR reactions and each independent fragment sequenced to confirm that the error exists in the

target yeast DNA and is not a PCR or sequencing error. This approach will also indicate whether all copies of the gene contain the error. Example 3 discloses how this approach was used to show that the AGTl genes of two lager strains both contain a premature stop codon. It also discloses that in two ale yeasts the AGTl genes encoded full length genes that are likely to encode functional transporters. The ale genes encoded proteins with a few amino acid changes compared to the SGDB sequence for AGTl (this sequence is entered as MALxI in the SGDB because the laboratory strain used to give the SGDB sequence has the AGTl allele in its MALI locus), but none of these changes was of the non-conservative type likely to cause loss of catalytic activity.

Step 4. Confirmation of the non-functionality of the potentially defective gene When a potentially defective gene has been identified, it may be obvious that it cannot encode a functional transporter. For example, if it has a very premature stop codon so that the encoded protein is very much shorter than known sugar transporters, it can be assumed to be non-functional. In other cases, e.g. if it has a stop codon only slightly before the normal full length stop, it may be unclear whether or not the encoded protein can be a functional transporter. The functionality of the gene is tested by cloning it and expressing it (e.g., from a multicopy plasmid with a KanMx marker and G418 selection as described in Example 4) in a yeast strain that is devoid of maltose transporters. The transformed yeast is then tested to see whether or not it has acquired maltose transport activity. It is convenient, but not essential, to clone the gene under test behind a constitutive promoter, e.g. a PGKl promoter. In this way, the maltose transport activity of the transformed strain can be measured after growth on glucose. Example 4 illustrates how cloning AGTl genes from different brewer's strains and over-expression of the cloned genes in a laboratory yeast was used to show that the potentially defective AGTl gene from a lager strain did not encode a functional maltose transporter but the apparently full length and normal A GTl genes from two ale strains did encode functional maltose transporters. In the same way, genes that encode functional transporters with undesired low activity or narrow substrate specificity as a result of particular amino acid residues in their sequences can be identified.

Step 5. Finding from another yeast a functional version of the target yeast's defective gene

Functional examples of AGTl, MALxI, MPHx, and MTTl genes have been described in the literature and sequences that encode functional transporters are known. Yeast strains

containing these functional genes are available and the desired genes can be amplified by PCR. Alternatively, one of these genes can be amplified by PCR from any convenient yeast strain and the amplified gene sequenced or overexpressed (e.g., from a multicopy plasmid with a KanMx marker and G418 selection as described in Example 4) in a yeast strain without maltose transport activity to confirm that it encodes a functional transporter. This latter approach is exemplified in Examples 3 and 4, which disclose how the AGTl genes from two ale strains were amplified by PCR, sequenced and over-expressed in a laboratory strain lacking maltose transport activity. Both these AGTl genes encoded full length, 616 amino acid proteins, with no obvious defects. Over-expression of each gene in the laboratory strain caused a large increase in maltose transport activity. The new maltose transport activity was strongly inhibited by maltotriose, which indicates that the transporter encoded by each of the ale-derived AGTl genes is also able to transport maltotriose. The ability to carry maltotriose effectively is a desirable property for a maltose transporter in brewer's yeast. The ability to carry maltotriose can also be directly confirmed by uptake experiments performed as described in Example 2 but using 14 C-maltotriose instead of 14 C- maltose. For reliable results, commercial l4 C-maltotriose should be re-purified as described (Dietvorst et al [2005] Yeast 22, 775-788), Maltotriose transport activity can also be determined by measuring the change in extracellular pH caused by the uptake of protons that accompanies uptake of maltotriose catalysed by the transporters, which are H + - symports.

The experiments with AGTl genes described in Examples 3 and 4 showed that the AGTl genes from these ale strains were suitable DNA molecules for repairing the defective AGTl genes in the lager strains Al 5 and A24. Analogous experiments with other α- glucoside transporter genes are readily performed by a person skilled in the art.

Step 6. Construct a repair cassette

This step involves some of the novel aspects of the invention. We have disclosed above and in Example 3 that strains of brewer's yeasts contain genes for α-glucoside transporters with defects in their sequences, but that nevertheless comprise DNA sequences of hundreds of nucleotides that are extremely similar to or even identical to homologous stretches in functional genes. This situation is surprising, because current strains of brewer's yeasts have evolved for hundreds of years in brewer's wort where maltose and

maltotriose are the main sugars. It is widely believed that their multiple genes for α- glucoside transporters have evolved in response to this environment. It is therefore surprising that mutations that apparently destroy the functionality of the encoded transporters have survived and even spread through the yeast population, so that two apparently unrelated lager strains (A 15 and A24) contain the same defect.

This surprising situation means that the defective genes can be repaired by homologous recombination between the defective gene and a DNA molecule containing a correct sequence. Such a DNA molecule can be obtained, as described above and in Examples 3 and 4 by PCR from yeast strains that contain functional copies of the gene in question. The invention includes two kinds of DNA cassettes:

1. Simple Repair Cassette

The defect(s) in the maltose transporter gene of the target yeast may consist of only one nucleotide or may extend over a few hundred nucleotides. A simple repair cassette is a DNA molecule cloned from the donor gene (by PCR or restriction enzymes or any other method) and comprising the correct version of the defective sequence flanked on each side by enough homologous sequence to facilitate recombination. This cassette and the desired integration event are shown schematically in the upper part of Fig 1. The cassette comprises between about 20 and a few hundred nucleotides of homologous sequence. Recombination often occurs by cross-over between the ends of the cassette and the corresponding homologous sequences in the target gene, with the result that the defective region of the target gene is replaced by the correct DNA sequence. The flanking sequences may be of any length except that the cassette advantageously does not encode a functional transporter. Preferably the present invention is concerned with the repair of defective genes and (see "2. Cassette for Repair and New Promoter) optionally replacement of promoters by more advantageous promoters. Typically, the method of the present invention does not result in elimination of any functional gene in the genome of the target host. Maltose transporters from yeast comprise eight transmembrane domains. Absence of one or more of these domains is likely to destroy the function of the transporter. Also, if the cassette lacks at least 150 bp from either the 5 '- or 3 '-terminus, the encoded protein is unlikely to be a functional transporter. To determine whether the cassette encodes a functional transporter the cassette can be joined in frame to a promoter and over-expressed in a yeast strain lacking maltose transport and the transport activity of the transformant measured

(methods are outlined in Step 5). However, usually the cassette is so short compared to the foil length gene that it is clear it does not encode a functional transporter.

2. Cassette for Repair and New Promoter This cassette and the desired integration event are shown schematically in the lower part of Fig 1. The cassette comprises a DNA sequence cloned from the donor gene (by PCR or restriction enzymes or any other method) and comprising the correct version of the defective sequence from the ATG start to at least 10 nucleotides, preferably 50 to 1000 nucleotides and more preferably 50 to 400 after the position of the last defective nucleotide of the target gene (or pseudogene). The 3 '-end of this sequence provides for homologous recombination with the target gene, the desired cross-over occuring 3 '- to the last defect in the target gene. Before the ATG start the cassette contains a suitable yeast promoter. At the 5'- end of this promoter sequence is a stretch of flanking sequence that is homologous to the promoter region of the target gene. This stretch of flanking sequence is preferably between 50 and 1000 nucleotides, more preferably 50 and 400 nucleotides . It is obtained by cloning the promoter region of the target gene. We disclose that in some brewer's strains the promoter regions of some maltose transporter genes differ significantly from the promoter sequences provided in, e.g., the SGDB. For example, the AGTl promoters of some lager and ale strains are almost identical to each other from nucleotides -1 to -564, but differ markedly from the SGDB sequence. Because the sequences of the ORFs of these genes closely resemble the sequences in the SGDB, a person skilled in the art can clone and sequence their promoters by using, e.g., one of the chromosome walking techniques (e.g. Mueller, P.R. and Wold, B. 1989. Science 246: 780-786). An illustration of how to do this is provided in Example 5.

The Cassette for Repair and New Promoter can integrate into the target gene as shown in the lower part of Fig 1. The result is to correct the defective part of the target gene and simultaneously put the ORF (open reading frame, the part of a gene that encodes a protein) under the control of the new promoter. The new promoter is chosen to give the desired expression characteristics. In the example of this type of cassette provided in Example 6, the promoter chosen was the PGKl promoter from a laboratory strain of S. cerevisiae. This promoter is relatively strong, does not require maltose for induction and is not repressed by glucose. The result (Example 8) is that although the target strain, A 15, contains little or no maltose and maltotriose transport activity after growth on glucose, the transformed strain

contains a powerful maltose/maltotriose transport activity. Because of this powerful maltose/maltotriose transport activity, the transformed strain ferments brewer's wort faster and more completely than the untransformed target strain (Example 9).

Step 7. Transformation and selection

The target strain can be transformed with the integration cassette and successful transformants selected by any suitable method known in the art. A convenient method is described in Example 7. The target strain (A 15) was co-transformed with an integration cassette of the Repair and New Promoter type and a self replicating plasmid containing the KanMx marker, giving improved resistance to the antibiotic G418. The particular integration cassettes used (Example 6) comprised the PGKl promoter functionally linked to the first 1482 bp of the AGTl ORF from ale strain, A60. At the 5'- end of the PGKl promoter was flanking sequence homologous to the promoter of the defective AGTl gene of strain Al 5. In one cassette this flanking sequence was nucleotides -1 to -387 and in another nucleotides -1 to -703. After transformation the cells were allowed to recover by growth on YP/glucose overnight. They were then plated onto YP/maltotriose containing antimycin A and G418. The first colonies to appear (in about 3- 4 days) were selected. These are expected to contain the KanMx-plasmid (because they can grow in the presence of G418). It is also expected that the cassette has integrated as desired, giving a functional AGTl gene driven by a PGKl promoter so that an efficient maltotriose transporter is produced in the glucose-grown cells before they are plated onto maltotriose. The presence of this transporter enables the cells to use the maltotriose carbon source immediately. Cells without a constitutive maltotriose transporter exhibit a long lag phase on maltotriose/antimycin A (Londesborough, [2001] Biotechnol. Letts. 23, 1995-2000). AU the first 55 transformants selected in this way produced the expected PCR product when their genomic DNA was used as a template with PCR primers recognising the AGTl ORF and PGKl promoter. This shows that the transformation, integration and selection proceeded with high efficiency. A high frequency of successful integration is expected with cassettes of this invention, because the flanking sequences are known to be very similar or identical to the homologous sequences at the desired integration site. This contrasts with the situation when, e.g., ORFs of interest are provided with, say, PGKl promoters and terminators derived from laboratory strains and attempts made to integrate them into the PGKl loci of brewer's yeasts. Evidence is accumulating for considerable

sequence differences between the homologous genes of laboratory and brewer's yeasts, which may account for the frequent failure of this approach. The novel approach of the present invention avoids this potential difficulty. The transformants are then grown on media (e.g. YP/glucose) lacking G418. This causes loss of the self-replicating plasmid containing KanMx, and provides transformants that contain no non-yeast DNA. It is notable that selection of integrants with a constitutive maltotriose transporter by plating on maltotriose/antimycin A is so powerful that it is probably unneccessary to cotransform with the marker plasmid encoding resistance to G418. The invention can be practised with or without co-transformation by a marker plasmid.

Step 8. Analysis of the biochemical and brewery characteristics of the transformants

Appropriate Southern and PCR analysis of the transformants 1 genomic DNA gives basic information about the integration event(s), including an indication of how many copies of the repaired gene have been produced in each cell. Because the target brewer's yeasts are aneuploid (and alloploid in the case of lager strains), more than one integration event may occur in each genome. This means that Southern and PCR analyses may reveal from the same clone both the pattern expected from the untransformed target strain and the pattern expected after integration of the cassette.

To asses the transport activities quantitatively, transformants are grown on YP/glucose, YP/maltose or other media, harvested and their maltose and maltotriose transport capacities measured. This is illustrated in Example 8.

To assess the brewery characteristics, transformants are grown, e.g., in brewer's wort, to get enough cell mass to carry out wort fermentations. No selection pressure is needed, because the transformants are integrants. Brewer's wort is then pitched with the transformed yeast and with the untransformed target yeast. Transformants cause faster fermentations and also leave less residual maltose and maltotriose (especially important when high gravity worts are used). The production of aroma compounds and the organoleptic qualities of the produced beer are measured using standard methods. Most transformants produce beer with the same character as that made by the target strain. These fermentation analyses are illustrated in Example 9.

The above Detailed Description explains how the invention is used to repair maltose transporter genes in brewer's strains and replace their promoters by new promoters. The transformed brewer's strains ferment worts faster and more completely than the untransformed strains. Slow and incomplete fermentation of worts is a major problem as brewers try to use stronger worts in production. Especially the so-called very high gravity (VHG) worts with original extracts above 16 0 P frequently cause problems with very slow fermentations and large amounts of residual maltotriose and maltose. The invention has advantages compared to earlier approaches. For example, EP 0 306 107 Bl describes baker's yeast strains (not brewer's yeast) in which the complete coding sequence of a maltose transporter gene has been provided with a new promoter and terminator (e.g. from the yeast ADHl gene) and this construct is then integrated into a yeast gene. This strategy results in the loss of one gene from the yeast. EP 0 306 107 recognised this problem, and therefore chose to integrate the DNA construct into the SITl gene (EP 0 306 107 , loc .cit see p 14, lines 17 et seq.), in the belief that the SITl gene is not needed by the baker's yeast during the industrial process. Nevertheless, the transformed yeast lacks at least one copy of the SITl gene This could have unforeseen consequences, especially in the beer production process, where the exact physiological behaviour of the yeast impacts on flavour formation and yeast handling. Furthermore, the use of the complete ORF of the maltose transporter gene, plus new promoter and terminator and plus new flanking sequences to provide homology to the integration site (SITl) results in a large integration cassette. In contrast the integration cassettes of our invention are relatively small molecules: the cassettes disclosed in Example 6 are only 3.7 and 3.4 kbp and integrated with high frequencies. In general, transformation frequencies decrease when cassettes size increases. Of course the labour involved in constructing the cassette also increases.

Kodama et al ([1995] J. Am. Soc. Brew Chem. 53, 24-29) have described a brewer's yeast containing high-copy-number plasmids comprising a maltose transporter gene (MAL6T) under the control of a constitutive promoter. This transformed strain fermented VHG (24 0 P) wort faster than the uncontrolled strain. This transformed strain contains non-yeast (bacterial) DNA, which is a disadvantage for applications in the food and drink industry because of public attitudes and official regulations. Furthermore, the high-copy-number plasmids are not genetically stable and selection pressure is required to maintain these plasmids in the yeast. Kodama et al, carried out their fermentations in wort containing the

antibiotic G418 to provide selection pressure. This is unacceptable on the industrial scale and with a product for human consumption.

The following examples are for illustration of the present invention and should not be construed as limiting the present invention in any manner.

Example 1. Determination of the apparent MAL, AGTl and MPH genotype of the target strain

This example illustrates how to identify in a target strain the genes for maltose/maltotriose transporters, including pseudogenes that hybridise with gene-specific reporters but contain small sequence errors that render them non-functional.

PFGE (Pulsed Field Gel Electrophoresis). Yeast strains were propagated in YP/2% glucose for 2 days at 30 0 C and then harvested by centrifugation (3000 x g, 5 min, 4°C). Supernatants were decanted and cells resuspended in 10 ml of 4 0 C 50 mM EDTA, pH 8. Cell concentrations were determined by OD 6 oo measurements and 8 to 60 x 10 6 cells were placed in each sample plug. Sample plugs were prepared using the CHEF Genomic DNA Plug Kit for Yeast (Bio Rad). Cells were centrifuged (3 min, 5000 * g, 4°C) and resuspended in the kit's Cell Suspension Buffer. Lyticase was added to 150 U/ml final concentration followed immediately by melted 2% Clean Cut agarose to a final concentration of 0.75%. These mixtures were dispensed into molds and allowed to solidify to produce plugs that could be loaded into the sample wells of the PFGE apparatus. After the agarose solidified the sample plugs were pushed out of the molds into the kit's Lyticase buffer containing 170 U lyticase/ml. Plugs were incubated in the lyticase solution for 2 h at 37°C. Lyticase buffer was removed and the plugs rinsed with sterile water. Plugs were incubated in the kit's Proteinase K reaction buffer containing 240 U of proteinase K/ml for 18 h at 50 0 C. After that the plugs were washed four times in the kit's washing buffer (1 h per wash) with gentle agitation and stored at 4°C in the same buffer. Sample plugs were loaded into the wells of a 1.0% Ultra Pure agarose (Bio Rad) gel. PFGE was performed at 14°C in 0.5 x TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA, pH 8). A CHEF Mapper XA Pulsed Field Electrophoresis System (Bio Rad) was used with the following settings: 6V/cm in 120° angle; pulse length increasing linearly from 26 to 228 s; total running time, 22 h 52 min. A commercial chromosome marker preparation from S. cerevisiae strain YNN295 (Bio Rad) was used for molecular mass

calibration. Following electrophoresis, the gels were stained with ethidium bromide and scanned and quantified with a Typhoon imager (Amersham Biosciences, Espoo, Finland) to estimate the total amount of DNA in each lane.

Chromosome blotting and hybridization. Chromosomes separated by PFGE were partially depurinized by soaking the gels in 0.25 M HCl for 20 min. After that the gels were treated with 0.5 M NaOH/1.5 M NaCl for 30 min at room temperature and neutralized in 1.5 M NaCl/0.5 M Tris buffer, pH 7.5, for 30 min. The DNA was then transferred to a nylon filter (Hybond-N, Amersham Biosciences) by capillary blotting in 20 χ SSC (3 M NaCl/0.3 M sodium citrate, pH 7.0). After blotting, DNA was UV cross- linked to the membrane. Prehybridization was performed at 48°C for 2 h in a hybridization mixture containing 5* SSPE (3 M NaCl/0.2 M sodium phosphate/0.1 M EDTA, pH 7.4), 5 x Denhardt's solution, 50% formamide, 100 μg single strand salmon sperm DNA and 1 μg poly-A-DNA/ml. The hybridization mixture was then replaced with a fresh lot containing a labeled DNA probe at 0.5 x 10 6 cpm/ml. The probes were labeled with 32 P by the random primer protocol using Hexalabel Plus DNA Labeling Kit (Fermentas Life Sciences, Hanover, MD). Unless stated otherwise, hybridization was carried out at 42 °C overnight. After hybridization, filters were washed twice with 2 χ SSC at room temperature for 5 min and then in 0.1 x SSC/0.1 % sodium dodecyl sulfate (SDS) at 65°C for 40 min. Filters were exposed over night to a phosphoimager screen, which was then scanned and signals quantified with Image Quant software (Amersham Biosciences).

Results: MAL genotypes of the strains.

At low DNA loads, PFGE resolved 15 chromosome bands from the marker yeast strain, YNN295 (Fig 2). Chromosomes VII and XV were presumed to comigrate as a duplex. Chromosome FV, faint in Fig. 2, is clearly visible in the brewer's yeasts at higher DNA loads (Fig. 3A). The four laboratory strains all differed from each other and from the marker strain in the size of one or more chromosomes. For the brewer's strains, greater differences appeared, including bands not present in the marker strain. Bands at or near the positions of chromosomes that often carry MAL loci, i.e., chromosomes VII; III; II; XI and VIII, could be tentatively identified, although chromosome III was smaller in the lager strains than in the YNN295 marker strain. Chromosome III could not be identified in the

ale strains based on mobility alone. Blots of these PFGE gels were hybridized sequentially with specific probes for the transporter, maltase and activator genes at MAL loci (Table 1).

Table 1. PCR primers used to make probes.

Gene Direction a Primer sequence Sequence SEQ detected ID NO

AGTr F 5 '-TTGCTTTACAATGGATTTGGC-S ' 842-1828 1

R 5 '-CTCGCTGTTTTATGCTTGAGG-S ' 2

MALlS(AGTIf F 5 '-GACTTTAACTAAGCAAACATGC-S ' 3-580 3

R 5 '-CGTTCGATATTTGTGCAAAGCT-S 4

MAL61 C F 5 '-GGAGCCTTTCTATGCCCTGC-3 ' 361-1140 5

R 5 '-TAATGATGC ACCACAGGAGC-3 ' 6

MAL62 F 5 '-GCGTTGATGCTATTTGGGTT-S ' 155-933 7

R 5 '-GAAAAATGGCGAGGTACCAA-S ' 8

MAL63 F 5 -GTATTGCGAAAC AGTCTTGC-3 ' 5-886 9

R 5 '-CATCGACACAGTTAGTAGCC-3 10

MAL33 F 5 '-ATAGCTCCACCTCAGCCAGA-3 ' 772-1313 11

R 5'-TGATTGCAATGTTTCAGGGA-S ' 12

MPHx e F 5 '-TCCGTGGATCTTGTTGGAAA-S ' 505-1294 13

R 5 ' -GTCCAAAAGCGTAAAGGTCA-S ' 14

a F, forward; R, reverse. The numbering is from the first nucleotide of the translational start. 0 AGTl and MAL61 primers were the same as in (Jespersen et al, 1999, Appl. Environ. Microb. 65, 450-456). d Primers were designed with the Saccharomyces Genome Database primer design programme (http://seq.yeastgenome.org/cgi-bin/web-primer). e MPHx primers were the same as in (Day et al, 2002 Yeast, 19, 1015-1027).

Previously sequenced MALxI genes are 97% identical and the MAL61 probe is expected to hybridize to all of them. In some MALI loci the transporter gene is the AGTl allele instead of MALIl (Han et al, 1995, MoI. Microbiol. 17:1093-1107) and we used a specific AGTl probe to detect this allele. MALx2 genes are 97% identical and the MAL62 probe is expected to hybridize to all of them. There is greater variation amongst MALx3 genes:

MALI 3, MAL23 and MAL43 are 91-94% identical to MAL63, but MAL33 and an allele of MALI 3, found in MALI loci containing AGTl and referred to in this study as MALI 3(AGTl), share only about 75% identity to each other and to other MALx3 sequences. We used specific probes (MAL33 and MAL13(AGT1)) to detect these two genes.

The MAL61 and MAL62 probes hybridized in many lanes at the expected positions of the VII/XV duplex and chromosomes II, XI and III, which carry, respectively, MALI, MAL3, MAL4 and MAL2 loci (Fig. 3D). The AGTl and MALI 3(AGTl) probes hybridized to the VII/XV duplex of all the strains in this study, suggesting that they all carried (on chromosome VII) MALI loci containing the AGTl allele of the transporter gene and the MALI 3 (AGTl) allele of the activator gene. These probes did not hybridize to any other chromosomes. The MAL61, MAL62 and MAL63 probes also bound to the VII/XV duplex of all six brewer's strains (Figs 3D, E and F). This pattern suggests that the brewer's strains each contained at least two kinds of chromosome VII, one with a MALI locus containing the AGTl and MALI 3(AGTl) alleles and another with a MALI locus containing MALIl and MALI 3 alleles. In contrast, MAL61 and MAL63 did not bind at this duplex in the laboratory strains (MAL62 did; Fig 3E), suggesting that the single copy of chromosome VII in the (haploid) laboratory strains contained only the AGTl and AGTl (MALI 3) alleles.

Three lager strains (A24, A64 and A72), one ale strain (A60) and one laboratory strain (CEN.PK2-1D) contained complete MAL2 loci on chromosome III. Chromosome III is possibly missing from the other ale strain (see Al 79 in Fig 2).

The MAL61 and MAL62 probes both bound to chromosome II of all strains, suggesting that all strains carry MAL3 loci with transporter and maltase genes (Figs. 3D and E). As expected, the MAL63 probe did not bind to chromosome II (Fig 3F), which is consistent with the low identity (75%) between MAL63 and MAL33. However, the MAL33 probe bound to chromosome II of all strains (data not shown), indicating that these MAL3 loci are complete, i.e., they contain transporter, maltase and activator genes.

In all of the brewer's strains, but none of the laboratory strains, the MAL61, MAL62 and MAL63 probes hybridized to chromosome XI, suggesting that the brewer's strains all carried complete MAL4 loci.

A72 was the only strain where the MAL61, MAL62 and MAL63 probes all hybridized to chromosome VIII, which carries the MAL6 locus. For all of the other strains the hybridization of the MAL61 probe to chromosome VIII was very faint and hybridization of MAL62 and MAL63 probes could not be detected. None of the 30 brewer's yeasts studied by Jespersen et al (1999) Appl. Environ. Microbiol. 65:450-456 contained a MAL6 locus.

Weak hybridization was sometimes observed to other chromosomes. For example, with ale strain A60, the MAL61, MAL62 and MAL63 probes hybridized to a chromosome of ~1.3 Mbp (Figs 3D, E and F). Binding of & MAL61 probe to a ~1.3 Mbp chromosome also was observed by Jespersen et al. (1999) Appl. Environ. Microbiol. 65:450-456 in a lager yeast. For all brewer's strains, the MAL62 probe hybridized weakly to chromosome IX (immediately above the MAL22 bands in Fig 3E).

Deduced differences in the MAL genotypes (Table 2) between the five brewer's strains were few. In particular, ale strain A60 was identical to lager strains Al 5, A24 and A64 except for the weakly observed MAL hybridization to the ~1.3 Mbp chromosome of A60. Ale strain Al 79, and its derivatives, Al 80 and Al 81, were the only brewer's strains lacking a MAL2 locus. A72 was the only strain with a complete MAL6 locus. The maltose-negative laboratory strains S 150-2B, RH 144-3 A and YNN295 had complete

MALI (with AGTl and MAL31 (AGTl) alleles) and MAL3 loci. Evidently, their MALI and MAL3 loci, even collectively, are insufficient for a maltose positive phenotype. The maltose-positive laboratory strain, CEN.PK2-1D, had a genotype similar to the maltose negative strains plus a complete MAL2 locus.

Table 2. MAL and MPH genotypes of the studied strains 8

S150-2B, CEN.PK2- A60 A179, A15 A24, A72

RH144-3A, ID A 180, A64

YNN295 A181

AGTl AGTl AGTl AGTl AGTl AGTl AGTl

MALIl MALIl MALIl MALIl MALIl

MAL12 MAL12 MAL12 MAL12 MAL12 MAL12 MAL12

MAL13 MAL13 MAL13 MAL13 MAL13

MAL13(A) b MALl3(A) h MAL13(A) b MAL13(A) b MAL13(A) h MALU(Af MALU(Af

MAL2 C MAL21-23 MAL21-23 MAL21-23 MAL21-23 ML21-23

MAL3 C MAL31-33 MAL31-33 MAL31-33 MAL31-33 MAL31-33 MAL31-33 MAL31-33

MAL4 C MAL41-43 MAL41-43 MAL41-43 MAL41-43 MAL41-43

MAL6 (MALόlf (MALόlf (MALόlf (MALόlf MAL61

MAL62

MAL63

Other Chr IX e Chr IX e Chr IX 6 Chr IX e Chr IX e Chr IX e Chr IX e

MAL 1.3 Mbp f loci

MPHx (MPH2)" (MPH2f MPH2 MPH2 MPH3 MPH3 Other Chr MPH VII/XV loci

8 No entry means that the gene(s) were not detected in this yeast. b MAL13(A) indicates the MALI 3(AGTl) allele. c The MAL2, MAL3 and MAL4 loci either were completely absent or contained all three genes MALxI, MALx2 and MALx3, shown as, e.g., MAL21-23. d Parentheses indicate that the hybridization detected was very weak. e The MAL62 probe (but not other probes) bound weakly near chromosome EK in all strains, stronger in brewer's strains. f In A60 MAL61, MAL62 and MAL63 probes bound weakly to a chromosome ~1.3 Mbp in size.

Example 2. Kinetic characteristics of maltose transport by brewer's yeast strains

This example illustrates how to characterise kinetically the maltose and maltotriose uptake properties of the target yeast. Such characterisation can indicate which genes detected by hybridisation may be defective.

Materials. D-[U- 14 C]-maltose was from Amersham Biosciences (Espoo, Finland). Maltose for uptake experiments (minimum purity 99%), maltotriose (minimum purity 95%) and trehalose were from Sigma-Aldrich (Helsinki, Finland). Maltose for growth media and α- methylglucoside (methyl-α-glucopyranoside) were from Fluka (Helsinki, Finland).

Strains. Four industrial lager strains (Al 5, A24, A64 and A72) and two industrial ale strains (A60 and A 179) from VTTs collection were used as typical representatives of strains in current industrial use. Strains Al 80 and Al 81 were isolated as single cell clones from A179 and appeared identical to A179. The laboratory strains were S288C, S150-2B, CEN.PK2-1D (VW-IB), RH144-3A and the chromosome marker strain, YNN295, from Bio-Rad (Espoo, Finland). To determine the Mal-phenotypes, strains were grown in media with 1% yeast extract/2% Difco (Sparks, MD) Bacto-peptone (YP) containing 2% maltose for 2 days at 25 0 C. All strains reached an ODeoo of 10 - 12, except for the three laboratory strains S288C, S150-2B and RH144-3A, whose OD 600 remained below 1.4. These three laboratory strains were defined as Mal-negative.

Maltose transport. Yeasts were grown in 100 ml of YP/4% maltose in 250 ml flasks shaken at 150 rpm and 25°C to an OD 60 O of 6 to 12, corresponding to about 2 - 4 mg dry yeast/ml. Under these conditions residual maltose was between 2 and 0.5%. The yeast were harvested by centrifugation (10 min, 9000 x g, O 0 C), washed with ice-cold water, then with ice cold 0.1 M tartrate/Tris pH 4.2, and finally suspended in the same buffer to 200 mg fresh yeast/ml. Zero-trans 14 C-maltose uptake rates were immediately determined essentially as described by Lucero et al. ([1997] Appl. Environ. Microb., 63, 3831-3836) and 1 U catalyzes the uptake of 1 μmol maltose min "1 (5 mM maltose, pH 4.2, 2O 0 C).

Results. Inhibition of maltose transport by trehalose and maltotriose. Trehalose is the preferred substrate of the Agtl transporter but is not a substrate for the MaIx 1 and MPHx transporters. Therefore, trehalose is expected to inhibit competitively maltose transport by

the Agtl transporters, but not necessarily to affect transport by the Malxl and Mphl transporters. Maltose transport in both ale strains was strongly inhibited (50% at 10 mM trehalose and over 80% at 100 mM trehalose). For ale strain A60, inhibition leveled off at -82% (1/v = 5.5 in Fig 4), suggesting that inhibition is partial, with ~18% of the maltose transport capacity insensitive to trehalose. For ale strain A179, inhibition was 88% (1/v = 8.3) and still rising at 100 mM trehalose, suggesting that a smaller fraction (< 10%) of the maltose transport capacity is insensitive to trehalose. All three lager strains were less sensitive to trehalose (Fig. 4). For A24 and A64 inhibition reached 30 and 55% (1/v values of 1.4 and 1.8, respectively), at 100 mM trehalose, whereas maltose transport by lager strain Al 5 was not affected by up to 100 mM trehalose. This result suggests that maltose- grown Al 5 does not express maltose transporters capable of carrying trehalose, whereas in A24 and A64 these transporters can account for 30-50% of the maltose transport capacity at 5 mM maltose.

Also maltotriose and α-methylglucoside strongly inhibited maltose transport by both ale strains, but only weakly inhibited transport by the lager strains (Table 3). Maltotriose is a substrate of Agtl, Mphx and Mal31 transporters, and α-methylglucoside is a substrate of Agtl and MPHx, but not of the Mal31 transporter.

Table 3. Inhibition by maltotriose and α-methylglucoside of 14 C-maltose transport 3 .

Aϊ5 A24 A(H A60

Al 79

Inhibition by maltotriose (%) 12 ± 7 16 ± 11 24 ± 4 66 ± 3 59 ±

25

Inhibition by α-methyglucoside (%) 23 ± 3 lO ± lO 13 74 ± 3 41

a Transport of 5 mM 14 C-maltose was measured with the indicated strains (grown on maltose) in the presence and absence of 50 mM maltotriose or 170 mM α-methylglucoside. Where errors are shown, results for maltotriose are averages ± SDs (n = 3 - 6) and for α- methylglucoside averages of duplicates ± ranges.

These kinetic results suggest that when these brewer's strains are grown on maltose, high specificity maltose transporters, such as Malxlp, account for 0-15% of the maltose

transport capacity of ale strains but for 40-80% of the maltose transport capacity of lager strains. Correspondingly, low specificity transporters, such as Agtlp and Mphp, are predicted to be important in the ale strains. This suggests that the AGTl genes detected in lager yeasts detected by hybridisation are likely to be defective and that at least some of the MALxI detected in ale strains are likely to be defective.

Example 3. Sequencing of AGTl in two lager and two ale strains

AGTl gene sequences of each strain were amplified by PCR with the specific primers AGTl frw 5 '-ATGAAAAATATCATTTCATTGGT-S ' (SEQ ID NO: 15) and AGTl rev 5 '- TTAACATTTATCAGCTGCATTT-S' (SEQ ID NO.-16) from both ends of the gene. Genomic DNA of four brewer's strains (A 15, A24, A60 and A 179) was used as the template. The PCR-generated fragments were cloned using the TOPO TA cloning kit (Invitrogen, Carlsbad, CA) and propagated in E. coli. Plasmid DNA was isolated from independent E. coli clones and sequenced. Whole gene sequences were obtained with eight sequencing primers; universal M 13 Forward and Reverse primers to sequence the start and end of the pCR-TOPO plasmid ligated AGTl gene and 6 internal primers from the coding strand. In addition to these clones when the whole gene was sequenced, an additional twenty independent clones from each strain were sequenced with only one of the internal sequencing primers (Sekv4, 5'-AAAGCAGATTGAATTGAC-S') (SEQ ID NO: 17). This primer, starting at nucleotide 1011, gave a readable sequence from approximately nucleotides 1150 to 1500, which includes the region where the genes from some strains appeared to have a frame shift. The model 3100 Genetic Analyzer sequencer was used (Applied Biosystems, Foster, CA). Multiple alignments were performed with the Multalin programme at http://prodes.toulouse.irjra.fr/multalin/multalin.html.

Results. AGTl genes were amplified from four brewer's yeasts in three (Al 5), 2 (A24), 3 (A60) and 2 (Al 79) independent PCR reactions, ligated into the pCR-TOPO plasmid and independent clones isolated and sequenced (Tables 4 and 5). Of the eight whole gene sequences from the two ale strains (A60 and Al 79), seven encoded full length, 616 amino acid proteins, similar to the sequence in the Saccharomyces Genome Database (SGD; http://www.yeastgenome.org). AGTl is entered as MALIl because it is the allele of MALIl present in the S288C strain of the SGD. However, one sequence (from A179) had a stop codon (TGA) at nucleotides 1183-1185, leading to a truncated, 394 amino acid sequence. For the two lager strains, 7/8 whole gene sequences encoded this truncated

polypeptide, but one (from Al 5) encoded the full length protein. The nucleotide change leading to the stop codon was the insertion of an extra T at position 1183, which results in a stretch of eight consecutive Ts instead of seven, the TGA stop codon at positions 1183- 1185, and a frame-shift in the rest of the nucleotide sequence. Table 4. Sequencing of AGTl genes from ale and lager strains.

Strain PCR Independent 616 amino 394 amino

Reactions 8 Clones acids acids

Whole gene sequences

A15 3 5 1 4

A24 2 3 0 3

A60 3 5 5 0

Al 79 2 3 2 1

Nucleotides 1153- -1500

A15 3 22 1 21

A24 2 20 0 20

A60 4 18 17 1

Al 79 4 21 21 0

a The number of independent PCR reactions. b The number of independent clones sequenced. c The number of sequences encoding full length, 616 amino acid proteins. d The number of sequences encoding the truncated, 394 amino acid polypeptide.

Table 5. Amino acid changes between the proteins coded by the AGTl genes in the Saccharomyces genome database (SGD) and the ale and lager strains".

TMD b D SGD A60 A179 A15 A24

40 D N N N N

45 K R R

59 T I

74 V M M

78 M T T

80 A T T

83 D E E

102 K I I

128 S N N N N

I (164-184) 164 L Q Q Q Q

168 M T T

175 T P P

198 I V V

II (202-222) 215 I V V V V

219 I M M M M

226 S G G G G

228 A T T T T

111 (241-266)

276 G D D

IV (284-311) - -

333 V I I

359 N D D D D

375 T A A

381 S T T T T

385 V C C C C

392 Y C

395 E Stop Stop

Corrected frame 0

V (405-427) 409 L V V V V

VI (432-459) 448 I V/I d I/V d V V

459 S G G G

VII (461-489) 488 A T T

VIII (504-532) 509 L I

548 V A

556 T S

595 A D

a Amino acids are shown at positions where the SGD and brewer's yeast protein sequences differ; no entry means the amino acid is the same as in the SGD. b The positions of transmembrane domains (TMDs) indicated in SGD are shown. c The sequences of the lager strains' hypothetical proteins after the Stop codons were obtained by correcting the frame

shift. d Residue 448 was the same as in the SGD in 3 out of 13 A60 sequences and 10 out of

17 Al 79 sequences.

These results might mean that both Al 5 and A 179 contain both correct and frame-shifted versions of AGTl. Alternatively, Al 5 may contain only the frame-shifted version and A 179 only the correct version, but during PCR or sequencing reactions the sequences of seven or eight Ts were occasionally misread. To distinguish these hypotheses, we cloned and sequenced -20 independent pCR-TOPO clones containing the 1153-1500 portion of the AGTl gene from each strain. For the two ale strains, 38/39 sequences encoded full length proteins (Table 4), strongly suggesting that the single occurrence of the frame- shifted sequence represents a PCR or sequencing error rather than the presence of both correct and defective versions of AGTl in these ale strains. Conversely, 41/42 sequences from the two lager strains contain the extra T, strongly suggesting that all copies of AGTl in these strains are defective. The AGTl sequences from all four strains are deposited in the NCBI databases as DQ091763 (A60), DQ091764 (Al 79), DQ091765 (A15) and DQ091766 (A24).

Example 4. Overexpression of apparently normal and potentially defective AGTl genes to determine their functionality

AGTl (pseudo)genes were cloned from the lager strain Al 5 and from two ale strains, A60 and A 179. The genes were ligated between the PGKl promoter and terminator sequences of a multicopy plasmid, to give, e.g., YEplacl95PGKPT-AGTl(A15) shown in Fig. 5. This plasmid has an URA3 gene to allow for selection in ura3 -laboratory strains. Strains CEN-PK2- 1 D and S 150-2B were transformed with these plasmids and transformants selected on media lacking uracil. The transformants were grown up in glucose (this represses the endogenous maltose transport activity of CEN-PK-ID; S150-2B is a maltose- negative yeast, unable to grow with maltose as sole carbon source and lacking significant maltose transport activity), harvested and their maltose transport activity determined as described in Example 2 under standard conditions (5 mM 14 C-maltose, pH 4.2, 20 0 C) in the presence and absence of 50 mM maltotriose. Results are shown in Table 6.

Table 6. Maltose transport activity of laboratory strains carrying ,4 GTl (pseudo)genes from different brewer's yeasts. The yeast were grown on YP/2 % glucose. ND, not determined.

In the CEN-PK background, maltose transport activity of the untransformed host or host carrying the empty plasmid (YEplacl95PGKPT) was low and 50 mM maltotriose caused only moderate (8 - 18 %) inhibition. Transformation with the plasmid carrying the AGTl gene from the ale strain A60 caused a large increase in maltose transport activity, confirming that this ale gene encodes a functional transporter. This activity was strongly inhibited by maltotriose, which is expected if maltotriose is also a substrate of the encoded transporter. In contrast, transformation with plasmid comprising the AGTl pseudogene (with premature stop codon) from the lager strain Al 5 did not increase the maltose transport activity. In the background of the S150-2B strain (which is maltose negative) transformation with plasmids carrying AGTl genes from either ale strain A60 or ale strain A 179 caused large increases in maltose transport activity and again maltotriose was a strong inhibitor consistent with the encoded transporters being able to carry maltotriose as well as maltose.

During the cloning of the AGTl gene from strain Al 79 by PCR, one clone was found that had a single nucleotide change (possibly a PCR mistake) changing amino acid residue 117

from leucine to proline. When this clone was expressed in S150-2C on a multicopy plasmid (S150-2B/AGT1 (A179)-L117P in Table 6) no extra maltose transport activity was obtained. Presumably the change of amino acid to proline (associated with sharp turns in the direction of the amino acid chain) prevents the protein from taking up the proper conformation to form an active transporter.

This example shows how functionality of transporter genes can be examined by over- expressing the genes in laboratory yeast strains and measuring the maltose transport activity of the transformants. By this kind of experiment a person skilled in the art can detect defective and functional genes in brewer's yeast strains.

Example 5. Determination of the promoter sequences of AGTl genes by chromosome walking

This example illustrates how the promoter sequences of defective and functional genes can be determined. The determined sequences can be used to give homologous flanking sequences upstream of the new promoter used to practise the invention with the Repair and New Promoter cassette and the Cassette for New Promoter Alone described in Step 6 of the Detailed Description. Chromosome walking was applied to DNA from the two lager strains, Al 5 and A24, and the ale strain A60. The method used was that described by Mueller, P.R. and Wold, B. 1989. Science 246: 780-786.

In Fig 6 the sequences obtained are compared to the AGTl (MALIl) promoter sequence given in the SGDB. From nucleotides -1 to -316 the promoter sequences of Al 5 lager and A60 ale strains were both almost identical to the SGDB sequence (two changes at -8 and - 152 were shared by the two brewer's yeasts). From nucleotides -221 to -316 also the A24 lager sequence was identical to the other three (the A24 sequence from-1 to -221 is not yet available). From nucleotides -317 to -806, the three brewer's sequences were essentially identical to each other, apart from a deletion in the ale strain between -572 and -593.

However, these brewer's sequences deviated from the SGDB sequence: identity between the SGDB and the lager sequences was only 35 %.

Example 6. Typical Repair and New Promoter Cassettes

Integration cassettes were constructed in pBluescript II SK(-) vector (Stratagene, La Jolla, CA). A 1485 bp long PGKl promoter fragment and a 370 bp long PGKl terminator fragment were cloned from a laboratory strain of S. cerevisiae and ligated onto a YEplacl95 multicopy vector as a Hindlll fragment. There is a BgRl site located between the PGKl promoter and terminator. The promoter-terminator complex was now detached with Hindiϊl from the YEplacl95 vector and ligated into the pBluescript vector at its Hindlll site.

The gene encoding Agtl was amplified by PCR using brewer's yeast ale strain A60 as template. The following PCR primers for the sense and antisense directions were used 5'-CGAGATCTCGATGAAAAATATCATTTCATTGGT-S' (SEQ ID NO: 18) and 5'-GCAGATCTGCTTAACATTTATCAGCTGCATTT^' (SEQ ID NO: 19), respectively. BgHl restriction sites to facilitate the next cloning step were introduced by PCR and are underlined in the primer sequences. The PCR product was cloned into a pCR-TOPO vector and the sequence of the AGTl gene was verified by using eight sequencing primers; universal Ml 3 Forward and Reverse primers to sequence the start and end of the pCR- TOPO plasmid ligated AGTl gene and 6 internal primers from the coding strand. The AGTl gene was excised from the pCR-TOPO plasmid with BgRl enzyme and ligated between the PGKl promoter and terminator at the BgRl site in the pBluescript plasmid described above.

AGTl promoter fragments of two different lengths were amplified by PCR. Brewer's yeast lager strain Al 5 was used as template. Two different primer pairs were used: AGTl prom Cl -705 FRW and AGTl prom Cl -1 REV to amplify an AGTl promoter fragment from - 705 to -1 and the primer pair AGTlprom Cl -387 FRW and AGTl prom Cl -1 REV to amplify an AGTl promoter fragment from -387 to -1 of the AGTl promoter. EcoRI and Mspall restriction sites were introduced by PCR on the 5' ends of these fragments and an EcoRI site to their 3' ends to facilitate the next cloning steps and the final detachment of the integration cassette. Primers:

AGTlprom Cl -705 FRW

5'-CGGAATTCCAGCGGCAAGTCAAGAGAAGATGGAAC-S' (SεQ ID NO.20) AGTlprom Cl -387 FRW

5'-CGGAATTCCAGCGGGCGAGGAACAAGGTTTTTTTC-S' (SEQ ID NO:21)

AGTlprom Cl -1 REV 5'-CGGAATTCCGATTATAATATTTTTTTAGTTGT-S' (SEQ ID NO:22)

EcoRI sites are underlined and MspAlI sites are in bold in the primer sequences.

The AGTl promoter fragments were ligated to a pCR-TOPO plasmid and their sequences were verified. The promoter fragments were excised from the pCR-TOPO vector with EcoRI and ligated into the pBluescript vector already possessing the PGKl promoter- AGT1(A6O)-PGK1 terminator complex. Ligation was carried out to the EcoRI site, which is located next to the 5' end of the PGKl promoter. Thus two different kinds of integration cassettes (pBluescript-Short-Integration and pBluescript-Long Integration) were constructed with different lengths of AGTl promoter fragments flanking PGKl promoter at 5' side. The cassettes were excised from the pBluescript vector by using MspAlI. It detaches the 3.38 kbp (possessing 387 bp long AGTl promoter flank) and 3.70 kbp (possessing 705 bp long AGTl promoter flank) integration cassettes. Detachment takes place at the 5' end of the ligated AGTl promoter fragments at the newly introduced MspAlI site and on the other side of the cassette at the MspAlI site located in the AGTl gene at position 1482.

This example illustrates how to construct Repair and New Promoter Cassettes. The particular cassettes described in this example are suitable for repairing the defective AGTl genes in lager strains like Al 5 and A24 and can be used as such to practise the invention. The one with the shorter (387 bp) AGTl -derived flanking sequence can also be used to replace the resident promoter of the AGTl genes of at least all strains in which the sequence of the resident promoter from -1 to -387 is similar to that found in the lager strain Al 5 and ale strain A60 (which are virtually identical from -1 to -564). The longer one can also be used for this purpose in cases where the promoter sequences are similar between - 703 and about -650 to -600. The sequences of vectors comprising the long and short forms are provided in Figs 8 (SεQ ID NO:23) and 9 (SεQ ID NO:24). The sequences of the long and short cassettes are given as SεQ ID NO:25 and SεQ ID NO:26, respectively. Analogous cassettes can be made containing parts of the ORFs of other α-glucoside transporter genes.

Example 7. Transformation of A15 with the cassettes of Example 6

Al 5 brewer's yeast strain was transformed with the Repair and New Promoter cassettes (pBluescript-Short-Integration and pBluescript-Long Integration) constructed in Example 6. These cassettes contain part of the ale strain AGTl sequence and are expected to create constitutive expression of AGTl. Transformation was performed as co-transformation with pKX34 selection plasmid containing KanMx marker and giving resistance to the antibiotic G418.

Al 5 yeast cells were inoculated into liquid YPD medium and grown overnight at 3O 0 C to 2 x 10 7 cells/ml (OD600=1). Cells were diluted in fresh, warm YPD to OD600 0.1-0.2 (2-5 x 10 6 cells/ml) and regrown to OD600 0.5-0.6 (1-2 x 10 7 cells/ml). Cells were collected by centrifuging at 2000 rpm for 5 min at RT. Cells were then washed twice with 10 ml sterile water and once with 10 ml freshly made IxTE, 0.1 M LiAc. Cells were centrifuged at 2000 rpm for 5 min at room temperature (ca 20 0 C) after each wash. The pellet was resuspended in freshly made IxTE, 0.1 M LiAc to a concentration of 2 x 10 9 cells/ml and incubated for 15 min at 30 0 C without agitation. Cells were then divided into 50 μl aliquots (1 x 10 cells/aliquot). 300 μl freshly made 40% PEG, Ix TE, 0.1 M LiAc and 25 μl single-stranded salmon sperm carrier DNA (2 mg/ml; freshly denatured by heating 10 min at 100 0 C and immediately chilled on ice), were added to each aliquot as well as 6.4-8.0 μg integration cassette DNA and 0.6 μg selection plasmid pKX34 (volume of integration cassette DNA and selection plasmid was up to 5 μl in total). Suspensions were mixed by vortexing thoroughly and incubated without agitation for 30 min at 30 0 C. Cells were exposed to heat shock for 20 min at 42°C. Cells were centrifuged twice for 15 sec and supernatants decanted. Cells of each aliquot were resuspended into 1 ml YPD and incubated for 17 h at 3O 0 C with shaking (250 rpm). Next cells were centrifuged for 15 sec and supernatants decanted. Cells of each aliquot were resuspended into 1 ml sterile water and 100 μl aliquots were plated on selective plates YPD/G418 (200 μg/ml) or YP/2%maltotriose/Antimycin A (3 μg/ml)/G418(200 μg/ml). Plates were incubated at 30 0 C up to six days. Transformation frequencies were counted from colonies appearing on YPD/G418 (200μg/ml) plates. On YPD/G418 (200μg/ml) plates colonies appeared after two days at 30 0 C (Table 7). Transformation frequencies were around 50-110 transformants/ μg total transforming DNA (cassette + selection plasmid DNA).

Table 7. Colonies appearing on YPD/G418 (200μg/ml) plates after 2 days incubation at 3O 0 C.

Half (500 μl) of each transformation was plated to these plates. Short cassettes 2 and 5 are two similar cassettes digested from two independent clones of pBluescript-Short- Integration cassette construct (clones 2 and 5). Long cassettes 11 and 12 are two similar cassettes digested from two independent clones of pBluescript-Longlntegration cassette constructs (clones 11 and 12).

Colonies on

Amount of DNA in Day 2 Transformation frequencies transformation /μg of DNA transformation 1 6,4 μg cassette (short cassette 2) 0,6 μg pKX34 171 -50

transformation 2 8,0 μg cassette (short cassette 5) 0,6 μg pKX34 473 -110

transformation 3 8,0 μg cassette (long cassette 11) 0,6 μg pKX34 434 -100

transformation 4 8,0 μg cassette (long cassette 12) 0,6 μg pKX34 460 -110

On YP/2%maltotriose/Antimycin A (3mg/ml )/ G418 (200μg/ml) plates there was a marked delay before colonies appeared (Table 8). Even after 6 days the number of colonies able to grow in the presence of antimycin A was less than 10 % of those found in two days without antimycin A. Antimycin A is known to cause a long lag before growth on maltotriose or maltose for yeast cells lacking an induced or constitutive maltotriose transporter (Londesborough [2001] Biotechnol. Letts. 23, 1995-2000), so that growth on maltotriose/antimycin A or maltose/antimycin A is a good selection system to identify clones with effective transporters able to carry maltotriose, maltose or both. The transformants shown in Table 8 are therefore expected to have been transformed with the integration cassette as well as the marker plasmid pKX34 and this cassette to have integrated as planned, yielding a constitutive maltotriose transporter, expressed in the presence of glucose (before the cells were plated) and encoded by the repaired AGTl gene driven by the PGKl promoter created as outlined in the lower part of Fig 1. Table 8 shows that the long and short forms of the integration cassette performed equally well.

It is notable that selection of integrants with a constitutive maltotriose transporter by plating on maltotriose/antimycin A is so powerful that it is probably unneccessary to cotransform with the marker plasmid encoding resistance to G418. The invention can be practised with or without co-transformation by a marker plasmid.

Table 8. Colonies appearing on YP/2%maltotriose/Antimycin A (3μg/ml )/G418 (200μg/ml) Half (500 μl) of each transformation was plated to these plates. The table shows the number of new colonies appearing on Days 2-6 (no colonies were visible before Day 2).

Amount of DNA in Day 2 Day 3 Day 4 Day 5 Day 6 transformation

Transformation 1 6,4 μg cassette

(short cassette 2) 0,6 μg pKX34 2 5 15

Transformation 2 8,0 μg cassette

(short cassette 5) 0,6 μg pKX34 1 6 7 23

Transformation 3 8,0 μg cassette

(long cassette 11) 0,6 μg pKX34 1 1 . . 6 . 10 20

Transformation 4 8,0 μg cassette

(long cassette 12) 0,6 μg pKX34 1 3 18

Example 8. DNA analysis and maltose transport properties of the transformants

Of the 119 transformants able to grow on maltotriose/antimycin A shown in Table 8, so far 36 have been examined by Southern analyses to examine the nature of the integration events. After digestion with EcoRI, Southern assays probed with the AGTl probe described in Table 1 are expected to show a 6270 bp band from the untransformed Al 5 host and a 5344 band after integration of the PGKl-promoter-AGTl sequence into the host's AGTl gene. All 36 transformants analysed showed one of two patterns: in 7/36 the 6270 band was completely replaced by the smaller band and in 29/36 both the original and the smaller bands were seen. These patterns are consistent, respectively, with the planned integration occurring at all or some of the AGTl loci in Al 5.

Some of the Table 8 transformants were also tested by a PCR strategy using primers recognising the PGKl promoter and a sequence in the AGTl ORF distal to the part of this ORF that is in the integration cassette. From the host no PCR product is expected, whereas integration according to the plan (Fig 1 lower part) should yield a 2257 bp product. This product was seen in all 55 tested transformants.

Some of the Table 8 transformants were also tested by another PCR strategy using primers recognising the A GTl promoter and a sequence early in the AGTl ORF were used. A product of 684 bp is expected from the host and a larger product (2184 bp) from integrants where the PGKl promoter has been inserted between the AGTl promoter and the AGTl ORF. All 36 tested transformants showed the 2184 bp product.

These results show that the present integration and transformation strategy is highly efficient compared to results often obtained with industrial strains.

So far the maltose transport activity of three integrants has been measured in cells grown on YP/glucose. Under these conditions the maltose transport activity of the host (or target) strain, Al 5, is expected to be low, because glucose represses most maltose transporter genes and maltose is required to induce many of these genes. Results are shown in Table 9.

Table 9. Maltose transport activity of host and integrants after growth on glucose.

These results show that all three integrants had markedly increased maltose transport activity compared to the host, and that the new transport activity was strongly inhibited by maltotriose, implying that a transporter able to carry maltotriose as well as maltose is involved.

Example 9. Fermentation properties of the transformants

Integrants expressing constitutive maltose transport activity (e.g. those shown in Table 9) are first grown on media in the absence on G418. Single cell colonies recovered from these growths are tested for ability to grow in the presence of G418. Those that cannot grow have lost the pKX34 plasmid conferring resistance to G418, and are selected. These cured integrants contain no DNA from any non-Saccharomyces organism. The cured integrants are then grown in brewer's wort and pitched into brewer's wort in 2 L tall tubes. For example the methodology described by Rautio & Londesborough ([2003] J. Inst. Brew, 109, 251-261) can be used. The integrants cause faster fermentations and residual sugars are lower especially when the original gravity of the wort is very high (20 - 25 0 P). The yeast-derived aroma compounds of the green beers are measured by gas chromatographic analyses.

Example 10. Sequencing of the chimaeric repaired AGTl genes in several integrants

When the cassettes of Example 6 are used to transform a brewer's yeast with one or more mistakes in an AGTl gene, the desired integration event is as shown in the lower panel of Fig 1. In this scheme there is one cross-over event between the AGTl promoter sequences in the resident gene and in the cassette and another cross-over event between AGTl coding sequences 3 '- to the position of the sequence mistake(s). This is expected to happen, because these cross-over events involve the two ends of the cassette. However, because all the coding sequence in the cassette is extremely similar to the corresponding part of the coding sequence (open reading frame or ORF) in the resident gene, cross-over events could also occur 5'- to the position of the sequence mistakes. Such cross-over events would cause integration events that either do not repair the sequence mistakes in the target gene or do not introduced the new promoter, or both.

Altogether 20 independent clones of the about 2257 nucleotide PCR product described in Example 8 were isolated from seven integrants and sequenced. The size of these PCR products shows that they include the PGKl promoter inserted between the promoter and ORF of the AGTl gene, so that the left hand cross-over event in the lower panel of Fig 1 has occurred as desired. The DNA sequences of the AGTl genes in the target strain Al 5 and in the donor strain strain A60 differ at nucleotides 827, 997, 1123, 1183, 1465 and 1647 . Table 10 shows from which yeast (A60 donor or Al 5 target) the sequence was

derived at these positions. Table 10 shows results for Integrants 1-5, 14 and 17 (Integrants

1, 2 and 14 are identical, respectively, to Intl, Int2 and Int 14 in Table 9). For each of Integrants 1, 2 and 14 several independent clones (Al, A2 etc) were analysed. For 18 of the 20 independent clones, the nucleotide at position 1183 (the site of the premature STOP codon in the AGTl genes from the target strain, A15) was derived from the AGTl of the donor strain, A60, showing that, as expected, cross-over had occcurred 3 '- to the mistake in the A15-derived sequence. Only in 2 clones was the nucleotide in this position derived from the AGTl gene of the target strain.

Table 10. Origin of the chimaeric ORF sequences in AGTl genes from integrants

The table shows the origin of the nucleotides at positions 827, 997, 1123, 1183 (the site of the premature STOP codon) , 1465 and 1647 of the AGTl ORF for 20 independent clones from 7 integrants. Uncertain and unread sequence positions are indicated by a query (?). These results show that cassettes made as described in Example 6 integrate in the desired way with high efficiency so that about 90 % of the integrants selected by the strategy described in Example 7 contain chimaeric A GTl genes where the sequence mistake of the gene in the target strain has been repaired according to the plan shown in the lower panel of Fig 1.

Table 10

nt 827 nt 997 πt 1123 nt 1183 nt 1465 nt 1647

Integrant 1 A 1 A60 A15 A15 A15 A15 A15

Integrant 1 A 2 A60 A60 A60 A60 A15 A15

Integrant 1 A4 A60 A60 A60 A60 A60 A15

Integrant 1 A 5 A60 A60 A60 A60 A60 A15

Integrant 1 A6 A60 A60 A60 A60 ? ?

Integrant 1 A 7 A60 A60 A60 A60 A15 A15

Integrant 2 A 1 A60 A60 A60 A60 A60 A15 Integrant 2 A 2 A60 ? A60 A60 A60 A15 Integrant 2 A 3 A60 A60 A60 A60 A60 A15

Integrant 3 A A60 A60 A60 A60 A15 A15

Integrant4 A A60 A60 A60 A60 A60 A15

Integrant 5 A A60 A60 A60 A60 A15 A15

Integrant 14 A 1 A60 A60 A60 A60 A60 A15 Integrant 14 A 2 A60 A60 A60 A60 A60 A15 Integrant 14 A 3 A60 A60 A60 A15 ? ? Integrant 14 A 4 A60 A60 A60 A60 A60 A15 Integrant 14 A 5 A60 A60 A6Q A60 A60 A15 Integrant 14 A 6 A60 A60 A60 A60 ? ? Integrant 14 A 7 A60 A60 A60 A60 ? ?

Integrant 17 A A60 A60 A60 A60 A15 A15

Example 11. Two-litre scale brewer's wort fermentations with integrants

Three of the integrants isolated in Example 7 (Integrants 1, 2 and 14) were cured of the pKX34 plasmid by growth in the absence of G418 (see Example 9). These cured integrants and the untransformed target yeast (Al 5) were grown in 15 0 P brewer's wort at 18 0 C to give enough yeast mass for pitching of 16 0 P (high gravity) and 25 0 P (very high gravity, VHG) worts for 2-litre scale fermentations in stainless steel tall tubes. The worts used, fermentation conditions and analysis methods are typical for laboratory scale brewer's wort fermentation. They were essentially the same as described by Rautio and Londesborough, 2003 , Journal of the Institute of Brewing, 109, 251 -261 , hereby incorporated by reference. The 15 0 P worts were pitched with 5.0 g fresh yeast mass per litre wort and the 25 0 P worts with 8.0 g fresh yeast per litre. The fermentation temperature was raised to 14 0 C immediately after pitching, and for the 25 0 P fermentations was further raised to 18 0 C after 24 h.

Attenuation profiles for 15 and 25 0 P worts are shown in Figs 10 and 11, respectively. All three integrants attenuated both worts faster and jnore completely than the untransformed yeast. At 15 0 P, the lowest apparent extract reached by Al 5 (2.50 ± 0.02 0 P) was reached at least 2 days (42 h) earlier by all three integrants, and the apparent extracts reached by the integrants in 7 days (163 h) were 0.14 to 0.19 0 P lower than reached by A15 in the same time. At 25 0 P, the lowest apparent extract reached by Al 5 (4.57 ± 0.05 0 P) was reached at least 3 days (66 h) earlier by all three integrants, and the apparent extracts reached by the integrants in 8 days (187 h) were 0.40 to 0.53 0 P lower than reached by Al 5 in the same time. These savings in time or gains in apparent attenuation or both are of considerable economic value to a brewery, because they mean that fermentation tanks can be used more intensively, or more ethanol produced from the same amount of wort or both.

Growth and flocculation behaviour of the integrants were similar to those of the untransformed target strain, as shown in Fig 12. The final pH values of beers fermented by the integrants were lower than the pH of beers fermented by the untransformed target strain by up to 0.05 pH unit with the 15 0 P wort and 0.10 unit with the 25 0 P wort. This is in general a desired character, because beer pH sometimes rises towards the end of VHG fermentations because of release of cytoplasmic material from damaged yeast cells.

- ND, not determined.

Table 11 shows that residual maltose and, especially, maltotriose were lower in the final beers made by the integrants than in those made by the untransformed target strain, Al 5. Thus, the integrants have the desired property of being better able than the untransformed target strain to consume these sugars under the harsh conditions at the end VHG wort fermentations. Glucose was completely consumed by all the yeast strains much earlier in the fermentations. In the VHG beers, the increased sugar consumption resulted in markedly higher final concentrations of ethanol from integrants than from the untransformed yeast. Viabilities of yeast cropped at the end of the fermentations were at least as high for integrants as for the untransformed yeast.

Fig 13 compares the yeast-derived, volatile aroma compounds produced during the fermentations of the 25 0 P wort. There were no significant differences between Integrants 1 and 2 and the untransformed yeast, showing that the transformation of this invention does not interfere with flavour properties of the yeast strain. Integrant 14 also produced a similar

aroma profile as the untransformed parent, with the exception that in the green beer analysed there was still more acetaldehyde; however, this is likely to be reduced during the lagering of the green beer.

Integrants 1 and 14 were made with the long form of the cassette described in Example 6 and Integrant 2 with the short form. In Integrant 14, the cassette had integrated into all copies of the resident AGTl gene, whereas in Integrants 1 and 2 both repaired and original versions of the A GTl gene were present (see Table 9). Thus, both kinds of cassette perform as desired and it is not critical whether all or only some copies of the resident gene are repaired: in both cases fermentation ability is improved.

This example shows that the present invention provides genetically modified brewer's yeasts with improved fermentation speed and extent by repairing defective genes for α- glucoside transporters in industrial brewer's yeasts. However, the genetic modification did not introduce any DNA from non-Saccharomyces organisms: the new DNA sequences are from ale and laboratory strains of S. cerevisiae.

These examples illustrate but do not in any way limit the teaching of the Detailed Description. A person skilled in the art can apply this teaching using other yeast strains as hosts and donors and other α-glucoside genes.