FIELD OF THE INVENTION This invention is in the field of plant biotechnology. More specifically, the invention relates to a mutant barley lipoxygenase 1 gene (lox-1) that encodes an enzyme with severely reduced 9-hydroperoxy- octadecanoic acid forming activity. The invention also relates to the use of barley cultivars homozygous for lox-1 in brewing processes to reduce the formation of off-flavors in brewed products, such as beer, during storage.

BACKGROUND OF THE INVENTION Lipoxygenases are a family of enzymes (EC 1.13.11.12) that catalyze the dioxidation of free and esterified poly-unsaturated fatty acids containing a 1 (Z), 4 (Z)-pentadiene configuration. The products of lipoxygenase-catalyzed reactions have long been suspected as major culprits for the appearance of stale flavors in plant grain/seed and grain/seed derived food products (Robinson et al., 1995, Food Chem., 54: 33-43). Lipoxygenases have been implicated in the production of volatile hexanal aldehydes generated during soybean processing, which have an undesirable aroma, limiting the use of soybean proteins in food products. Three lipoxygenase isozymes expressed in soybean seed are believed to contribute to lipid oxidation and hexanal formation. Soybean mutants lacking one or more of these isozymes have been generated with the aim of reducing hexanal formation and improving their flavor stability. The success of this approach has been evaluated by Hildebrand et al., 1990, J. Agric. Food Chem. 38: 1934-1936. Mutants lacking soybean lipoxygenase 3 produced higher hexanal levels, suggesting that

this isozyme diverts 13-hydroxyperoxyoctadecanoids, produced by lipid oxidation, towards non-volatile products. The field performance of triple-null soybean lines, lacking all three seed lipoxygenases, has shown that these enzymes are not essential for normal agronomic and seed characteristics (Narvel et al., 1998, Crop Sci. 38: 926-928).

Lipoxygenases have also been implicated in the generation of off- flavors in rice, which can occur during grain storage. The release of free fatty acids can be detected in stored grain, which is indicative of the metabolism of the triglycerides reserves. The rice variety Daw Dam was found to accumulate lower levels of pentanals and hexanals giving a better flavor stability on storage (Susuki et al., 1999, J. Agric. Food Chem., 47: 1119-1124). This desirable phenotype was attributed to the absence of rice lipoxygenase-3, which oxidises unsaturated lipid acyl chains to form 9-hydroxyperoxyoctadecanoic positional isomers.

It is recognised that the lipoxygenase pathway is complex with many branches and its role in numerous aspects of plant growth and physiology are not fully understood. Modifications of the lipoxygenase pathway which alter 9-hydroperoxidation activity in seed crops are proposed to regulate their susceptibility to mycotoxin contamination by Aspergillus spp. (WO 9726364), which is consistent with the involvement of this pathway in plant pathogen resistance, but is not related to the aims of the invention herein described.

Among the many aroma volatiles which contribute to the flavor of beer, the higher unsaturated aldehydes with a 6-12 carbon chain have particularly low organoleptic flavor thresholds (Meilgaard 1975, MBAA Tech. Quart. 12: 151-168). Trans-2-nonenal, which is a member of this group, has both an extremely low flavor threshold of 0.11 ppb and contributes an unpleasant straw-like,"cardboard"flavor to the beer. The characteristic off-flavor caused by trans-2-nonenal is a common characteristic of beers stored for 1-3 months or more and is particularly detrimental to the flavor of lager beer, which is brewed with light malts and has a delicate flavor.

Sulfite has long been known to improve the flavor stability of beer, not only by binding oxygen and acting as an anti-oxidant, but also by forming volatile bisulfite addition compounds with aldehydes and ketones present in the beer. The two major sources of sulfite in beer are sulfite produced by yeast during fermentation via the sulfur assimilation pathway and sulfite added to the beer prior to packaging. Fermentation conditions that enhance yeast sulfite production and secretion will allow the formation of sulfite-carbonyl adducts from carbonyls present in the wort and prevent their further metabolism by the yeast (Dufour 1991, Proc. Eur. Brew. Conv. Congr., Lisbon, pp. 209-216). In this manner carbonyls such as acetaldehyde and diacetyl may be transferred to the beer. The ability of sulfite to prevent the appearance of the carbonyl compound trans-2-nonenal during beer aging has been demonstrated by brewing beer with a yeast strain blocked in the sulfur assimilation pathway (Johannesen et al., 1999, Proc. Eur. Brew. Conv. Congr., Nice, pp. 655-662). Following bottling, the beer was subjected to forced aging by storing it at 37°C for 7 days, after which trans-2-nonenal levels were found to be well above the taste-threshold. If 10 ppm sulfite was added to the low-sulfite beer just prior to bottling, the appearance of trans-2- nonenal during forced aging was significantly reduced. The reaction between sulfite and carbonyl compounds is reversible and under thermodynamic and kinetic control. The apparent equilibrium constants for bisulfite compounds ranges from 10-6M for carbonyl compounds such as acetaldehyde, hexanal, and decanal, to 10-3 for diacetyl and pyruvate (Dufour 1991, supra). During beer storage, gas exchange through the packaging will allow oxygen into the beer and sulfite will be lost, such that weaker bisulfite adducts will dissociate, allowing free carbonyls to appear in the beer. While sulfite unquestionably enhances the flavor- stability of beer, particularly in the short-term, its retention in packaged beer is strongly dependent on gas exchange through the packaging and temperature. In a finished beer the natural levels of sulfite produced during fermentation are variable and the addition of sulfite prior to

bottling is not a universally accepted practice. For these reasons sulfite alone does not provide a reliable method to enhance the long-term flavor- stability of beer under the different beer storage conditions used around the globe.

It is generally accepted that the trans-2-nonenal found in beer results from the oxidation of polyunsaturated fatty acids derived from barley grain lipids, where the 18-carbon chain fatty acid, linoleic acid [classified as an 18: 2, n-6 polyunsaturated fatty acid (Broun, Gettner and Sommerville 1999, Annu. Rev. Nutr. 19: 197-216)] is the most abundant.

However, there is little agreement in the literature as to the mechanism whereby trans-2-nonenal is formed. The presence of enzymatic pathways leading to trans-2-nonenal formation from poly-unsaturated fatty acids has been proposed, but the individual enzymatic steps have never been demonstrated experimentally in barley grain or during the malting process (Gardner 1988, Adv. Cereal Sci. Technol. 9: 161-215).

The concept of using anti-sense or co-suppression gene technology to reduce lipoxygenase-1 levels in barley grain, and thereby control 9- hydroperoxidation and reduce aldehyde and alcohol levels in the finished barley grain, has been proposed as a means to control off-flavor formation, but results of such an approach are not reported (McElroy and Jacobsen, 1995, BiolTechnology 13: 245-249).

A forcing test has been developed as a method for assessing the trans-2-nonenal potential of a beer, where trans-2-nonenal formation in wort or beer is induced by subjecting samples to elevated temperatures at reduced pH, (100°C, at pH 4.0 for 2 hours). Attempts to correlate the trans-2-nonenal potential in wort and finished beer with the total level of lipoxygenase activity in the kilned malt have indicated that lipoxygenase may contribute to the appearance of trans-2-nonenal in aged beer (Drost et al., 1990, J. Am. Soc. Brew. Chem. 48: 124-131). The conclusions that can be drawn from this study, however, are severely limited by the fact that the lipoxygenase activity in the barley malt was regulated at the end of the malting process by the degree of enzyme inactivation during kiln

drying. Thus, only the effect of the residual malt lipoxygenase activity on the trans-2-nonenal potential in the derived wort and finished beer was examined. The study failed to evaluate the lipoxygenases that catalyse the first step in the lipoxygenase enzymatic pathway in the barley grain during development and malting, and their role as determinants of trans-2-nonenal levels found in beer. Indeed, the absence of barley cultivars deficient in one or more lipoxygenase isoenzyme has made it impossible to provide convincing evidence for the role of the lipoxygenase pathway in barley malt in controlling the formation of trans-2-nonenal. Such experiments are needed to evaluate the contribution of enzymatic, as compared to auto-oxidative/chemical pathways, to the formation of trans-2-nonenal in beer. The brewing process involves a high temperature step of wort boiling where these non-enzymatic reactions are proposed to occur (Noel et al., 1999, J.

Agric. Food Chem. 47: 4323-4326).

SUMMARY OF THE INVENTION This invention provides a barley cultivar having greatly reduced lipoxygenase-1 activity. In one embodiment, the barley plants of the invention contain a mutant lox-1 gene expressing greatly reduced levels of the isoenzyme lipoxygenase-1. In an alternative embodiment, the barley plants contain a heterologous nucleic acid sequence expressing an antisense sequence to the wild-type lox-1, thereby reducing the enzyme's activity.

As shown herein, malt and wort produced from the reduced lipoxygenase barley of the invention, for example, from barley cultivars homozygous for a mutant lox-1 gene, are useful to produce beer with significantly enhanced flavor stability and reduced trans-2-nonenal levels, particularly under conditions known to promote the appearance of T2N. The invention demonstrates a correlation between the activity of barley malt lipoxygenase-1 to produce 9-hydroxyperoxy-octadecadienoic acids (9-HPOD), and the presence of trans-2-nonenal in beer. The

invention further demonstrates that the use of barley homozygous for the mutant lox-1 gene in the brewing process improves the flavor stability of the beer, both during storage and on exposure to elevated storage temperatures. These properties enhance the quality of the beer, and are useful to extend its shelf-life and reduce the need to cool beer during transport and storage.

The invention provides barley plants and portions thereof having reduced lipoxygenase-1 activity, including barley plants expressing mutant LOX-1 protein as described herein, as well as methods for producing such barley plants, plant portions, products of the plants, and particularly malt and beer products produced from the barley plants of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a graph showing the effect of the inhibitor nordihydroguaiaretic acid (NDGA) on immuno-affinity purified lipoxygenase 1 and 2 activity from embryos of 3 day germinated barley grain.

Figure 2 is a graph showing the fresh weight of developing grain of Line G and cv Vintage from 5 days after flowering to full-maturity (FM). Each determination is the mean single grain weight from 6 spikes.

Figure 3 is a graph showing the dry weight of developing grain of Line G and cv Vintage from 5 days after flowering to full-maturity (FM).

Each determination is the mean single grain weight from 3 samples of 5 grain.

Figure 4 is a graph showing total lipoxygenase activity in developing grain of Line G and cv Vintage from 5 days after flowering to full-maturity (FM).

Figure 5 is a graph showing 9-and 13-HPOD products of linoleic acid oxidation by lipoxygenase activity in developing grain of Line G.

Figure 6 is a graph showing total lipoxygenase activity in embryos of germinating grain of Line G and cv Vintage expressed as u. mol/min/10 embryos (U/10 embryos).

Figure 7 is a graph showing 9-HPOD and 13-HPOD products of linoleic acid oxidation by lipoxygenase activity in embryos of germinating grain of Line G and cv Vintage, showing levels of 9-HPOD and 13-HPOD.

Figure 8 is a Western blot showing immunodetection of lipoxygenase 1 in embryos of developing grain of Line G and cv Vintage [wt] from 5 days after flowering to full-maturity (FM).

Figure 9 is a Western blot showing immunodetection of lipoxygenase 1 in embryos of grain of Line G and cv Vintage [wild-type] germinated for 0-6 days.

Figure 10 is a Northern blot probed with the 3'non-transcribed region of the lox-1 cDNA and showing lipoxygenase 1 transcripts detected in developing grain of Line G and cv Vintage [wild-type] from 5 days after flowering to full-maturity (FM).

Figure 11 is a Northern blot probed with the 3'non-transcribed region of the lox-1 cDNA and showing lipoxygenase 1 transcripts detected in embryos of grain of Line G and cv Vintage [wt] germinated for 0-6 days.

Figures 12A-12G are a nucleotide sequence alignment of the promoter and transcribed region of the lox-1 wild-type cv Vintage allele (WT) and the Line G allele (LG). The transcription start site (+1), ATG start codon (+69) and translation stop codon (+4231) in the gene sequences are underlined. Nucleotide mutations identified in the Line G allele are shown in bold italics and indicated by an asterisk.

Figure 13 is a schematic presentation of the lox-1 gene of cv Vintage (wild-type) and the mutant lox-1 gene of Line G. The transcript from +1 to +4375 is composed of 7 exons (stippled boxes) and 6 introns (white boxes). Two mutations in the lox-1 gene are indicated.

Figure 14 is a schematic drawing of gene cassettes for transient expression of the wild-type lox-1 cDNA and lox-1 gene and the mutant lox-1 gene from Line G. The lipoxygenase coding sequences were cloned between the constitutive maize ubiquitin promoter with intron 1 (Ubi-1) and the nos terminator.

Figure 15 is a bar graph showing Lipoxygenase 1 activity in barley aleurone protoplasts transfected with gene cassettes containing the wild-type lox-1 cDNA ; the mutant lox-1 gene from Line G; WT lox-1 gene; and a control GUS reporter gene. Lipoxygenase activity in extracts of transfected protoplasts was assayed in microtiter plates by the oxidation of KI and quantitated spectrophotometrically. Lipoxygenase 1 activity was expressed as units per ug protein in the extract and is shown as the mean of 3 measurements from 2 replicate assays.

Figure 16 is a sequence alignment demonstrating that a RFLP between the wild-type and mutant lox-1 gene is due to a point mutation at nucleotide 2347, creating an additional AatII restriction site.

Figure 17 is a schematic presentation of the lox-1 PCR fragments amplified and cleaved in the polymerase chain reaction-cleavage amplified polymorphic site (PCR-CAPS) assay. The positions of PCR primers are indicated by arrows and the AatII sites are shown above the gene (sequence position). The exon and intron regions within the PCR product are distinguished by stippled and white boxes respectively, and the sizes of the AatII digestion fragments are given.

Figure 18 is an electrophoretic agarose gel showing lox-1 PCR fragments (652 bp) amplified in the first step of the PCR-CAPS assay from Line G and cv Vintage genomic DNA.

Figure 19 is an electrophoretic agarose gel showing RFLP detected by PCR-CAPS in the wild-type and mutant lox-1 gene. The AatII digestion fragments of the mutant gene include a unique 313 bp restriction fragment, indicated by an asterisk.

Figure 20 is a table showing a back-crossing program for the single recessive gene pair 11 (low lipoxygenase trait) of Line G to cv

Alexis. The LL genotype are plants expressing wild-type lipoxygenase activity (dominant allele), the 11 genotype are plants expressing the low- lipoxygenase (recessive allele). Ll are heterozygous plants containing both the wild-type and the low-lipoxygenase allele. Since the low- lipoxygenase trait is a recessive trait, Ll plants show wild-type lipoxygenase activity. After each round of back-crossing (including self- pollination), the 11 progeny is expected to represent 25% of the progeny.

The observed frequencies of low-lipoxygenase activity are indicated.

The calculated percentage of the cv Alexis genetic background having the homozygous low-lipoxygenase allele is indicated as % Alexis.

Figure 21 is an electrophoretic agarose gel showing PCR-CAPS detection of the mutant lox-1 gene in 11 progeny of the Line G-Alexis back-cross program. PCR-CAPS assay on genomic DNA of Line G (Lane 2), cv Vintage (Lane 3), 11 progeny of 3rd (Lane 4) and 4 back- cross (Lanes 5-9). DNA ladder (Lane 1). Control, backcrossed high lox line (lane 10).

Figures 22A-22B are a comparative alignment of amino acid sequences of soybean lipoxygenases LOX-1 (Gml), LOX-2 (Gm2), LOX-3 (Gm3), and barley lipoxygenases LOX-1 (Hvl) and LOX-2 (Hv2).

DETAILED DESCRIPTION OF THE INVENTION In accordance with the subject invention, plant materials, plant products, and methods are provided for producing a beverage, such as beer, the beverage having a reduced content of the off-flavor compound trans-2-nonenal, such that the flavor stability of the beverage, e. g., beer, during storage and on exposure to elevated temperatures is improved, relative to a control beverage. More particularly, the invention provides barley varieties whose developing and germinating grain produce greatly reduced activity levels of the enzyme lipoxygenase-1, denoted LOX-1, which, for example, when used in a beer brewing process, results in a

beer having reduced trans-2-nonenal levels, as compared with a control barley variety.

The methods used to generate, characterize, and validate a barley variety having greatly reduced LOX-1 activity, and use of this type of barley for the production of flavor-stable beer are described below.

1. Definitions As used herein, the following terms have the indicated definitions: "Plant portion"means a plant or specific part of a plant, such as the stem, leaves, roots, flowers, seeds, grains, fruits, or buds.

"LOX-1"means lipoxygenase-1 protein ;"lox-1"means the gene encoding LOX-1.

"Mutant barley lox-1"means a mutagenized barley gene encoding a mutant lipoxygenase 1 polypeptide.

"Non-mutated control"means a plant, nucleic acid, gene, polypeptide, plant portion, or plant product containing wild type gene or protein.

"Heterologous"means a non-native sequence, e. g., a sequence derived from another species, or a recombinantly engineered or synthetic sequence that differs from the native sequence.

"Plant product"means a product resulting from the processing of a plant or plant portion, and includes, for example, malt and wort.

"Acidic amino acid"means aspartic or glutamic acid.

"Basic amino acid"means histidine, lysine, or arginine.

"Polar amino acid"means threonine, serine, tyrosine, tryptophan, asparagine, or glutamine.

"Organoleptic properties"means properties appealing to the olfactory and taste senses that are analysed, for example, by a trained taste panel.

"Brewed product"means a product prepared by mashing, boiling, and fermenting, e. g., beer.

"Reduced trans-2-nonenal"means less than about 50%, as compared with wild-type (control) conditions.

2. Lipoxygenase Activity Lipoxygenase enzymes catalyze the oxidation of polyunsaturated fatty acids. In barley, the isoenzymes LOX-1 and LOX-2 are known.

LOX-1 primarily catalyzes 9-hydroperoxidation, whereas LOX-2 primarily catalyzes 13-hydroperoxidation of polyunsaturated octadecanoic fatty acids. The data shown in the Examples below demonstrates a correlation between barley LOX-1 9-hydroperoxidation activity and the presence of trans-2-nonenal in beer. Accordingly, barley having reduced LOX-1 activity is useful to produce beer having a reduced trans-2-nonenal level and/or potential as compared with a control.

3. Production of low lipoxygenase barley A variety of known genetic approaches can be used to produce the plants of the invention, that is, to reduce the level of lipoxygenase 1 enzyme activity expressed in a barley plant in a stable, inheritable manner. These approaches include, but are not restricted to antisense technology and mutagenesis, such as chemical and radiation induced mutagenesis, as well as site-directed mutagenesis.

Barley transformation. Barley can be transformed with various nucleic acid molecules designed to manipulate lox-1 gene expression or alter the architecture of the lox-1 gene. Various methods, for example, Agrobacterium tumofaciens-mediated transfer (Tingay et al, 1997, Plant J., 11: 1369-1376), particle bombardment (Wan and Lemaux, 1994, Plant Physio., 104: 37-48, or polyethylene glycol (PEG)-mediated DNA uptake (Funatsuki and Kihara, 1995, Theor. Appl. Genet., 91: 707-712), can be used to successfully introduce nucleic acids into a barley cell, for example into a protoplast, callus, or an embryo.

Various promoters can be used to drive expression of the gene of interest. For expression of lox-1-containing vectors, including antisense sequences, the native lox-1 promoter region can be used. The promoter sequence of lox-1 is contained in nucleotides 2602-3511, which includes the 5'UTR of EMBL accession no. U83904. Alternatively, promoters that drive expression of the gene of interest constitutively, for example the Ubi. 1 maize ubiquitin promoter, can be used (Wan and Lemaux, Supra ; Kjxrulff et al., in P. Mathis, Ed., 1995, Photosynthesis : from Light to Biosphere, Vol. II, 151-154). Expression vectors can also contain a transcription termination region, for example, the 3'terminator of the nopaline synthase gene (3'-nos) (Bevan, et al, 1983, Nucl. Acids Res., 11 : 369-385) has been fused to genes expressed in transgenic barley (Wan and Lemaux, Supra ; Funatsuki and Kihara, Supra).

Expression vectors can also contain a gene that allows for selection of transformed cells when the vector has been successfully integrated in the cell. These genes can encode antibiotic or herbicide resistance genes, for example the neomycin phosphotransferase (npt) or the phosphinothricin acetyl transferase (bar) gene. When expressed, such resistance genes allow for growth of the transformed cell in neomycin- or bialaphos-containing media, respectively (See, for example, Wan and Lemaux, Supra ; Funatsuki and Kihara, Supra ; Kjaerulff et al., in P.

Mathis, Supra).

Following transformation, cells can be grown in selective media for a period of time and then cultured to allow for the formation of shoots, followed by root systems, and then plantlets. A successful barley transformation procedure was developed by Funatsuki and Kihara, (Supra), where transformation of barley protoplasts by PEG with neomycin phosphotransferase-containing expression vectors and subsequent selection in neomycin yielded fertile plants containing the transgene. The transgene was shown to integrate into the genome and most of the transgenic plants expressed the protein encoded by the

transgene. These transgenic plants also were able to transmit and express the transgene following crosses.

It is understood that a variety of transformation methods, expression vectors, promoters, selectable markers, and the like are known and useful for transformation of barley.

Barley mutagenesis. The lox-1 gene can be targeted for site- specific mutagenesis using chimeric RNA/DNA oligonucleotides. These chimeric RNA/DNA oligonucleotides have been shown to successfully introduce mutations in plant cells (Zhu et al., 1999, Proc. Natl. Acad.

Sci. 96: 8768-8773; and Beetham et al., 1999, Proc. Natl. Acad. Sci. 96: 8774-8778) and mammalian cells (Yoon et al., 1999, Proc. Natl. Acad.

Sci. 93: 2071-2076) at desired locations. The chimeric RNA/DNA oligonucleotides can be transformed into the barley protoplasts or cells of interest in a variety of ways, for example using the PEG-mediated or particle bombardment-mediated transformation methods described above.

The individual protoplasts or cells can then regenerated by tissue culture to whole fertile plants, and the mutational event can be confirmed and followed, for example using a PCR-based approach as detailed in the Examples below.

This site-directed mutagenesis method can be applied to mutate specific residues in the lox-1 gene. The lox-1 gene can be mutated at one or more nucleotide position in the promoter region to downregulate or abolish lox-1 transcription. Specific mutagenesis can also be applied to introduce changes in the lox-1 coding region that, for example, reduce the enzyme's activity. Such mutations include, but are not limited to, insertions, deletions, and substitutions resulting in a frameshift, truncation of the LOX-1 protein, and/or alteration of the neutral and hydrophobic nature of the enzyme's substrate cavity.

Antisense expression. Reduction in lox-1 expression can also be accomplished by expression of a lox-1 antisense construct in the barley cells. Methods for the expression of antisense constructs in barley to reduce the expression of a targeted protein have been reported, for

example, in Gilpin, M. J. et al., 1998, In: Photosynthesis : Mechanisms and Effects, G. Garab, ed., Vol. IV, 2983-2986; Kjaerulff et al., 1995, In: Photosynthesis : from Light to Biosphere, P. Mathis, Ed., Vol. II, 151- 154.

Barley cells can be transformed with an expression construct containing an antisense nucleic acid sequence. The expression construct produces an antisense RNA molecule capable of specifically binding to at least a portion of the mRNA produced from the wild type lox-1 gene, through complimentary base pairing, and capable of disrupting the splicing of the pre-mRNA or translation of this mRNA. A constitutive or tissue/temporal specific promoter, for example, the barley lox-1 promoter described above, can drive expression of the antisense nucleic acid sequence.

Chemical mutagenesis. The chemical mutagen sodium azide (NaN3) has commonly been used for barley mutagenesis and is known to induce stable mutations in the DNA (deoxyribonucleic acid) sequence of the barley genome (Olsen et al., 1993, Proc. Natl. Acad. Sci. USA, 90: 8043-8047). Other chemical mutagens, for example, ethyl methanesulfonate (EMS), azidoglycerol (AG, 3-azido-1, 2-propanediol), methyl nitrosourea (MNU), and maleic hydrazide (MH) can also be used to induce DNA mutations (Rank, J. et al., 1997, Mutat. Res. 390: 121-7), as can UV irradiation.

As shown in the Examples below, the grain of the barley cultivars (cv) Vintage and Caruso were treated with sodium azide and propagated by self-fertilization through to the 3d generation (M3).

4. Identification and Selection of Low Lipoxygenase Barley Identification and selection of barley plants having reduced lipoxygenase isoenzyme activity in the grain can be achieved, for example, by analysis of lipoxygenase activity. Enzymatic assays can be used to determine the activity of the two major lipoxygenases known to be present in either mature or germinating grain, LOX-1 and LOX-2.

Such assays should distinguish LOX-1 activity from that of LOX-2.

One selective assay of LOX-1 and LOX-2 is based on the oxidation of a poly-unsaturated fatty acid by lipoxygenase and the spectrophotometric detection of the hydroperoxide product of such oxidation. The specificity of this assay for LOX-1 takes advantage of the comparative insensitivity of LOX-1 to an inhibitor, for example, NDGA, relative to LOX-2.

Selective assay can also be achieved using immunoprecipitation to selectively remove LOX-1 or LOX-2 from the assay. Specific anti- LOX-1 and anti-LOX-2 antibodies, for example, monoclonal antibodies, can be prepared from purified LOX-1 or LOX-2 as described in Holtman et. al, 1996, Supra.

These assay methods can be adapted for microtiter plate assay procedures, or other known repetitive, high throughput assay formats, allowing the rapid screening of many samples. These assays can be validated for screening leaf tips of germinating grain in a non-destructive manner, such that seedlings selected in the screen can be further propagated.

The loss of LOX-1 activity in putative mutants can be confirmed by assay of enzymatic activity. For example, grain extracts can be incubated with linoleic acid and the oxidation products of linoleic acid analyzed, for example, by reverse phase HPLC. The relative amounts of 9-HPOD and 13-HPOD formed from linoleic acid provides a measure of LOX-1 activity, whose major product is 9-HPOD.

As shown in the Examples below, approximately 20,000 grain of the M3 generation of mutagenized cv Vintage and cv Caruso were screened for LOX-1 and LOX-2 activity by oxidation assay in the presence of inhibitor and also by immunoprecipitation assays. Using these screening methods, a mutant in cv Vintage was found having a major reduction in LOX-1 activity, and was denoted Line G. The mutant phenotype was inherited in the M4 and M5 generations.

Seed produced from the Line G barley was deposited on January 4,2001, with the National Collections of Industrial, Food and Marine Bacteria (NCIMB), 23 St. Machar Drive, Aberdeen, AB243RY, Scotland, UK, under the terms of the Budapest Treaty, as Accession Number: NCIMB 41078.

5. Genetic Sequences A precise description of the genotypic alteration that accounts for the low-lipoxygenase phenotype in barley plants of the invention is useful for identifying plants having this genetic alteration and for crossing this genetic character into other barley cultivars in a breeding program. A variety of known molecular and biochemical methods can be used to determine the genetic basis for the low lipoxygenase phenotype.

It is generally recognized that both cis-acting and trans-acting genetic sequences can determine the expression of a given gene in the genome and the activity of the gene product. Control points in gene expression include the regulation of the timing, tissue-specificity and rate of gene transcription, the stability of the transcript and the rate of transcript translation. Both the level of gene expression and the stability and specific activity of the encoded enzyme will determine the level of enzyme activity detected in a tissue.

Alterations in a plant gene sequence can be determined by DNA sequencing of known relevant parts of the genome, while Northern analysis provides a tool to monitor stable transcript levels in a given plant

tissue. Enzyme expressed in plant tissue can be evaluated by extracting the enzyme from the tissue and measuring the enzymatic activity.

As shown in the Examples below, the identity of the genetic changes that determine the low-lipoxygenase phenotype of the Line G mutant induced in cv Vintage were determined in the following manner.

The structural gene encoding the LOX-1 protein, both in the parent cv Vintage and in the Line G, was amplified by the polymerase chain reaction (PCR), and the upstream promoter sequences, which regulate expression of the gene, as well as the entire coding sequence, comprising intron and exon sequences, were sequenced.

Comparison of the nucleotide sequences of the lox-1 gene from Line G and from wild-type cv Vintage revealed 2 nucleotide substitutions in 2 exons, of which one (at position + 2347) led to a non-conservative amino acid substitution (Glycine368# Aspartate) in the expressed protein.

Figure 22 shows an alignment of soybean (Gm: Glycine max L) lipoxygenases LOX1 (Acc. No. P08170), LOX2 (Acc. No. P08170), LOX3 (Acc. No. AAB41272) and barley (hv: Hordeum vulgare) lipoxygenases LOX1 (Acc. No. P29114) and LOX2 (Acc. No.

AAB70865.1). Conserved amino acid residues and conservative substitutions of charged residues are shown in bold. Secondary structure assignments for LOX3 of soybean Glycine max, where H=alpha helices and E=Beta strands, are shown above the alignment, and residues relevant to enzyme function (identified by an asterix or filled circle) are shown, as described in Skrzypczak-Jankun et. al., 1997, Proteins 29: 15- 31.

Amino acid residues that participate in non-heme iron binding or essential for catalysis (*) in soybean LOX3 include: H518, H523, H709 [3 N atoms]; N7, 3 I857. The equivalent residues in barley LOX 1 are HS", Hs22, H708, N712 and I862. Residues in soybean LOX3 with a predicted role in catalysis (@) are: H266, H513, H776, F264, F272, F714, W519, R552, R726, D766, D779, K237. The equivalent residues in barley LOX1 are: H26"H512, H775 F2sg, F267, F713, W518, R551, R725, D778, and K273.

Proline (P86, 109 167, 171, 223, 234, 291, 311, 324, 343, 345, 371, 381, 382, 486, 541, 548, 600, 616, 627, 685, 726, 734, 788, 829, 833, 839, 857) and glycine (G49, 67, 68, 70, 91, 107, 137, 187, 192, 210, 217, 218, 260, 306, 307, 336, 392, 409, 458, 474, 490, 569, 607, 674, 676, 720, 736, 783, 828, 850, 855) residues (+) located in loops and helix-capping positions in protein secondary structures, may facilitate sharp turns and folding of the peptide backbone.

Alignment of related plant lipoxygenases indicated that the Glycine-368 in barley LOX-1 is strongly conserved. Furthermore, this residue, which corresponds to Glycine-353 in soybean LOX-1, is one of 35 highly-conserved residues out a total of 58 residues that line the substrate cavity II of the enzyme, as seen from its crystal structure.

These conserved residues are highlighted (boxes) in the alignment of plant lipoxygenase sequences shown in Figure 22 (Minor et. al., 1996, Biochemistry 35: 10687-10701), and include the following barley LOX-1 residues: Y224, L268, W3559 E364 G368, V369, N370, I374, L424, L499, K501, A502, V504, D508, S509, H512, Q513, L514, H517, W518, H522, I556, L559, A560, L564, I565, I570, T574, S585, Q715, Y718, N724, R725, P726, T727, L772, and I862. All but 7 of the 35 conserved residues are neutral or hydrophobic residues. The substitution of a charged residue at position Glycine-368 in barley or at another conserved neutral or hydrophobic residue lining the substrate cavity II, is likely is likely to disturb the structural and functional properties of the enzyme. The G#D368 mutation in barley Line G LOX1 ( *) is located between alpha-helix H6 and beta-strand E12.

As shown in Figure 22, the lipoxygenase family of enzymes shares a high degree of sequence conservation, which is reflected in their conserved secondary structure, determined for several members of the plant lipoxygenase family including soybean LOX1 and LOX3 (Skrzypczak-Jankun et al., 1997, supra). Barley LOX1 shares 56.9% sequence identity and 67.8% sequence similarity with soybean LOX3.

Several amino acid residues in the soybean LOX3 isoenzyme have been identified as ligands for the non-heme iron, or are suggested to be essential for its activity (denoted by * ). In view of the high sequence

conservation between the barley LOX1 and the soybean LOX3, it is reasonable to predict that residues in the barley LOX1 sequence that are homologous to those identified as important for the function of LOX3 may also be essential for enzymatic activity. Thus, non-conservative amino acid substitutions at any of these positions, including substitutions of those residues in barley LOX1 corresponding to the 35 highly conserved residues of soybean LOX3 that line the substrate cavity and in other positions essential for enzyme activity, are likely to reduce lipoxygenase activity.

The amino acid residues proline and glycine are known to facilitate turns in a peptide backbone when they are located between secondary structural elements, which allow a protein to assume a folded tertiary structure. Proline and glycine residues are also common in helix capping motifs (Parker and Hefford, 1997, Protein Eng., 10: 487-496, http://www. expasy. ch). The single non-conservative substitution in Line G LOX1, where a glycine located between two predicted structural elements was replaced by aspartate, led to a significant loss of enzyme activity. It is thus predicted that mutation in the LOX-1 gene causing a non-conservative amino acid substitutions at one or more of the proline or glycine residues in the barley LOX1, located in regions outside the structural elements, may similarly prevent folding of the native protein and consequently reduce the activity of the encoded enzyme.

Thus, in one embodiment, a useful mutant barley plant of the invention having reduced lipoxygenase 1 activity contains a mutated nucleic acid sequence that alters the neutral or hydrophobic nature of the substrate cavity of the enzyme by insertion of one or more acidic, basic, or polar amino acids. For example, a useful nucleic acid sequence [SEQ ID NO: 11 encodes a barley LOX-1 protein [SEQ ID NO: 12] having a substitution at amino acid 368 from Glycine to Xaa, where Xaa is an acidic, basic, or polar amino acid. One specific amino acid sequence of the barley mutant LOX-1 of the invention is that where Xaa is aspartic acid, e. g., Line G.

As shown in the Examples below, the genotypic changes in Line G had no detectable influence on lox-1 gene expression, but the LOX-1 activity detected in mature and germinating grain of Line G were approximately 9% of that detected in grain of the parent line, cv Vintage.

In order to provide direct evidence that the amino acid mutation in LOX- 1 of Line G was responsible for the low-LOX-1 phenotype, the coding sequence of Line G lox-1 and cv Vintage lox-1 were expressed transiently in protoplasts from barley aleurone, and the activity of the mutant LOX-1 enzyme was shown to be. strongly reduced in comparison to the wild-type LOX-1 enzyme.

6. Transfer between breeding lines The detection of alterations in genetic character of the barley plants of the invention genotype is useful to identify the presence of a specific genetic character in a barley line, and to facilitate the transfer of this character between breeding lines in a breeding program. A variety of molecular tools are available for the detection of alterations in genomic sequence. Such methods include, but are not restricted to, detection of restriction fragment length polymorphisms (Gebhardt and Salamini 1992, Int. Rev. Cytology., 135: 201-237) and quantitative PCR based detection methods such as amplification using fluorescent primers, e. g. the TaqMan primer probe systems (Ibraham et al., 1998, Anal. hem 70,2013-2017). The choice of detection method will depend on the specific genetic character but should preferably be rapid and provide clearly interpretable data.

As shown in the Examples below, a PCR-Cleavage Amplified Polymorphic Site assay (PCR-CAPS) was provided for the detection of the mutant lipoxygenase-1 gene of Line G. The nucleotide substitution in the lox-1 gene in Line G at position + 2347 introduced an additional site of recognition by the AatII restriction endonuclease that can be detected by the PCR-CAPS assay. Suitable detection methods for lox-1

are not restricted to this assay, but can equally well be based on TaqMan technology, and other known detection methods.

Also shown in the Examples below, the PCR-CAPS assay was applied to 4 generations of breeding material from a back-cross program, where the low-lipoxygenase phenotype in Line G was systematically back-crossed into cv Alexis. Inheritance of the low-lipoxygenase phentoype was shown to follow the inheritance of the lox-1 gene, and the phenotype was identified as recessive and only seen in lines homozygous for the lox-1 gene.

Accordingly, plant progeny of the invention includes breeding lines, for example, derived in a back-crossing program, that contain mutant lox-1 and express a low lipoxygenase phenotype.

7. Brewing The barley plants of the invention, including plant parts, plant progeny, grain, and plant products such as malt and wort, having low lipoxygenase 1 activity, are demonstrated herein to be useful for the manufacture of a beverage having reduced levels of free trans-2-nonenal over a measured period of time, or under conditions of elevated storage temperature, as compared to a beverage produced from a wild-type control barley variety. For the purpose of these comparisons the sulfite content of the beer is controlled to 5ppm or below, since it is recognised that higher sulfite levels at the time of bottling will temporarily delay the appearance of free trans-2-nonenal. For example, beer brewed from malt derived from the mutated barley Line G described herein, possessed stabilized organoleptic properties over a measured period of time as compared with beer brewed from malt derived from a control, non- mutated barley.

Brewing trials and evaluation of bottled beer provide the best method for evaluating the influence of different ingredients on the quality and stability of the finished beer. In order to test the influence of different barley malts, sufficient barley grain is needed to perform the

malting and brewing trials on a pilot scale and semi-industrial scale.

During the period of barley propagation, the field performance of the barley line can be evaluated. The malting properties of a barley line can be evaluated during pilot or industrial scale malting, and should preferably lie within national malting quality recommendations eg. the European Brewing Convention recommendations for malting quality (Analytica-EBC/European Brewing Convention, 1998, Publ. Hans Carl Getranke-Fachverlag, Nürnberg, Germany). Following pilot or semi- industrial scale brewing, the beer is packaged in brown bottles and cooled to 5°C for optimal storage. At this stage the fresh beer can be analysed by trained taste panels able to detect specific beer flavors, including the off-flavor compound trans-2-nonenal. Additionally, the beer is chemically analysed for major flavor components including trans-2- nonenal. These methods of beer quality analysis are then repeated on the beer following various storage conditions known to reveal the long-term storage stability of the beer, for example, forced aging treatments.

As shown in the Examples below, Line G barley was propagated in the field over several seasons in order to malt 10 tons of this line in an industrial malthouse. The control barley varieties cv Vintage and cv Nevada, both having the wild-type LOX-1 phenotype, were malted under similar conditions. The kilned malt from Line G and the control barley cultivars lay within the specifications required for the semi-industrial brewing trials.

Brewing trials were performed on a 30-hl scale and evaluation of the freshly bottled beers revealed that beers brewed from malt of both Line G and the control cultivars had a trans-2-nonenal content below the taste-threshold and were deemed satisfactory by a taste-panel. Two forced-aging treatments, either storage at 37°C for 7 days or storage for 6 to 12 weeks at 30°C, were used to evaluate the flavor-stability of the beer.

The flavor-stability of beer brewed from Line G malt were found to be superior to that of control malt, both with respect to taste panel

evaluation as well as the level of free trans-2-nonenal, and the improvement was found to be statistically significant.

EXAMPLES The present invention is further defined in the Examples below.

It should be understood that the Examples, while indicating preferred embodiments, are given by way of illustration only.

Example 1 Screening and Selection of Lipoxygenase Isoenzyme Mutants from Mutagenised Barley 1. Barley mutagenesis Grains of barley, Hordeum vulgare cv Vintage and cv Caruso, were mutagenised with sodium azide according to a published procedure (Kleinhofs et al., 1978 Mutation Research 51 : 29-35). The mutagenesis introduces point mutations in the genomic DNA that, for example, may result in amino acid changes in encoded proteins. The mutated M1 grains were propagated in the greenhouse through two generations, and the M3 grain collected for screening. The observed frequency of single gene trait mutants in the M2 generation, according to Kleinhofs et al., 1978, supra, are 1.0-2.7 mutants per 10,000 grain from the M2 generation. Since most gene mutations are recessive and only detectable in the homozygous state, the mutagenized population was screened at the M3 generation where the expected proportion of homozygous mutant grain would be higher. A mutation frequency of 0.9-2.3 per 10,000 grain was expected in the mutagenized material at M3.

2. A non-destructive assay of lipoxygenase 1 (LOX-1) and lipoxygenase-2 (LOX-2) activity in M3 mutagenized grain A rapid screening procedure for detection of mutant barley grain with reduced LOX-1 activity was developed with the following criteria:

The screening procedure should not prevent propagation of the grain/seedling; the selected grain/seedling tissue should express quantifiable levels of lipoxygenase activity; the assay should distinguish LOX-1 activity from that of LOX-2 ; and the assay procedure should encompass multiple samples.

The levels of total lipoxygenase activity in different tissues of the germinating grain, namely the shoot, root, and scutellum tissue of embryo and the endosperm were assayed as follows: Extracts of barley seedling tissue were prepared by homogenising the tissue in ice-cold 20 mM Tris-HCl, pH 7.5, containing 2 mM NaN3 and 0.5 mM phenylmethylsulfonyl fluoride (PMSF), followed by removal of insoluble material by centrifugation at 1000 g for 10 minutes. Lipoxygenase activity in 100 al extract was assayed at 25°C, by addition of 2.9 ml of 20 mM linoleic acid substrate, prepared by dispersing 35 u. l linoleic acid (free acid, L-1376, Sigma, USA) in 5 ml H, O containing 1% Tween 20.

The reaction was followed spectrophotometrically, where the rate of increase in absorbance at 234 nm (A234nm), due to the formation of conjugated diene in the hydroperoxide product, is proportional to the enzyme activity present. One unit of lipoxygenase activity is defined as A A234 = 0.001 per minute in a 3-ml reaction, equivalent to the oxidation of 0.12 umole linoleic acid.

The leaf tissue of grain germinated for 4 days in the dark had the highest detected levels of lipoxygenase activity (Holtman et al., 1996, Plant Physiology 111: 569-576). Leaf tips from 4-day seedlings were thus selected for the non-destructive lipoxygenase screening assay. The pH optimum of total barley lipoxygenase activity was tested between pH 4.5 and pH 9.0 and found to be pH 6.5. Hence a 25 mM HEPES buffer (pH 6.5) containing 0.2 M boric acid was selected for the screening assay.

Since both LOX-1 and LOX-2 enzymes were immuno-detected in shoots of 4-day seedlings (Holtman et al., 1996, supra), a LOX-1 and LOX-2 specific assay was used. The lipoxygenase inhibitor

nordihydroguaiaretic acid (NDGA), identified by Eskin et al., 1977, Crit.

Rev. Food, Science and Nutrition 9: 1-40, was found to be a selective inhibitor of barley lipoxygenases. NDGA at 1x10-5 M strongly inhibited purified barley LOX-2, while LOX-1 retained 47 % activity (Figure 1).

The selectivity of this inhibitor was tested in the leaf tip assay, by determining the ratio of 9-hydroperoxyoctadecanoid (9-HPOD) to 13- hydroperoxyoctadecanoid (13-HPOD), which result from linoleic acid oxidation by LOX-1 and LOX-2, respectively. In the lipoxygenase assay of cv Vintage leaf tips, the proportion of 13-HPOD formed fell from 24.5% to 9.5% on addition of 1'10-5 M NDGA.

A selective assay for LOX-2 activity in leaf tip extracts was based on the use of LOX-1-specific monoclonal antibody (5D2) (Holtman et al., 1996, supra) to immunoprecipitate LOX-1 present in the extracts.

The residual lipoxygenase activity detected in the extracts after LOX-1 precipitation provided a measure of LOX-2 activity. The efficiency of this immunoprecipitation (described below) was evaluated by quantifying the residual LOX-1 and LOX-2 in the extract supernatant by ELISA assay, using specific monoclonal antibodies against LOX-1 (denoted 5D2) and LOX-2 (denoted 5.8) (Holtman et al., 1996, supra). LOX-1 immunoprecipitation from extracts of cv Vintage leaf tips removed 85% of (LOX-1) protein and 15% of LOX-2 protein.

Immunoprecipitation was performed in a V-bottom 96-well plate by adding 5 ul 5D2-coated Dynabeads (Dynal) and 75 1ll buffer [20 mM Tris-HCl pH 7.5,1% v/v Bovine Calf Serum (HyClone)] to 20 al of each leaf tip extract. The plate was incubated on a titerplate shaker (MTS4, IKA, Labor Technik) for 1 hour at 4°C. The immunoprecipitate was pelleted by centrifugation at 4°C in a Sigma 302-K centrifuge for 10 minutes at 2000 rpm. The supernatant (70 1ll) from each sample was assayed for lipoxygenase activity in a flat bottom 96 well plate, as described below, but with addition of 100 pl assay buffer (25 mM HEPES, 0.2 M boric acid, pH 6.5).

The LOX-1 and LOX-2 assays were adapted for a high- throughput screening method. Leaf tips (1 cm) from eight 4 day- germinated grains were individually homogenised in 150 al ice-cold buffer (20 mM Tris-HCl, pH 7.5) for 2 x 30 seconds in a multi-well homogeniser (Berg et al., 1992, Electrophoresis 13: 76-81). After centrifugation for 15 minutes at 3000 rpm, 40 ul ofthe supernatant of each extract was transferred to a flat bottom 96 well plate. To each well, 170 ul buffer (25 mM HEPES, 0.2 M boric acid pH 6.5,1-10-5 M NDGA) and 10 pl substrate (20 mM linoleic acid) were added and then incubated for 20 minutes at 25°C. The reaction was terminated by the addition of 20 ul saturated potassium iodide solution (KI) and incubated for a further 8 minutes at 25°C. The redox reaction between hydroperoxydienes and KI yields Riz, which was monitored by its extinction maximum at 350 nm in a microplate reader (Multiskan MCC/340).

3. Identification of potential lipoxygenase 1 mutants in the M3 and M4 grain of mutagenised barley Grain of the M3 generation of cv Vintage and cv Caruso was stored at 45°C for 6.5 days to break dormancy, ensuring a 95% germination frequency. M3 grain of cv Vintage (9318) and cv Caruso (9633) was germinated and screened for lines whose LOX-1 activity was 15% or less of wild-type grain. The putative mutant lines (50 cv Vintage and 42 cv Caruso lines) were propagated to the M4 generation, harvested, and the germinated grain re-screened. The mutant LOX-1 phenotype was confirmed in one cv Vintage line and six cv Caruso lines, after measuring the lipoxygenase activity in extracts of 5 leaf-tips from each line. When the LOX-1 and LOX-2 activities in germinating embryos of these 7 putative mutants were examined, only the cv Vintage mutant (denoted Line G) showed a major reduction in LOX-1. In mature quiescent grain, lipoxygenase activity present in the embryo is almost exclusively LOX-1 activity, due to the differential expression pattern of the two isoenzymes (Schmitt and van Mechelen, 1997, Plant Sci. 128: 141-150). The total

lipoxygenase activity in extracts of embryos from Line G mature dry grain (M5 generation) was 0.060.04 U/mg protein in comparison to 0.74 0. 44 U/mg protein in cv Vintage embryo extracts, as determined by the spectrophotometric lipoxygenase assay described in section 2 of Example 1. The residual lipoxygenase activity in mature embryos of Line G in both the M4 and M5 generations was found to be approximately 9 % of the parental line.

Example 2 Line G is a cv Vintage Mutant with a Low-Lipoxygenase Phenotype The agronomic properties and mutant phenotype of Line G were analysed in material of the M5 generation. Initial analyses were conducted to confirm that the analysed M5 material was homozygous for the mutant phenotype. The low LOX-1 phenotype in Line G, detected in the M3 generation, could result from a dominant or a recessive mutation.

If the Line G selected at the M3 generation was heterozygous for a dominant mutation, then subsequent generations would show segregation for the phenotype. The lipoxygenase activity in 26 individual Line G embryos from quiescent grain of the M5 generation was measured and compared to cv Vintage wild type embryos. The lipoxygenase activity in all Line G embryos was very low, with an average of 0.06 0. 04 U lipoxygenase per mg protein, compared to 0.740.44 U lipoxygenase per mg protein in wild type cv Vintage embryos. These data confirmed that Line G in the M5 generation was homozygous for the low lipoxygenase trait.

1. Line G has a wild type plant growth physiology and grain development.

Line G and cv Vintage grain were germinated and grown in a climate chamber under 16 hours light at 15°C and 8 hours dark at 12°C at a relative humidity of 80%. The growth characteristics of Line G and cv Vintage plants were similar with regard to plant height, number of tillers per plant, the onset of flowering and number of grains per spike. The fresh weight (Figure 2) and dry weight (Figure 3) of grain of Line G and wild type cv Vintage during development from 5 days after flowering (DAF) until full maturity, approximately 90 DAF, were very similar.

2. Line G grain have a low-lipoxygenase 1 phenotype throughout development Lipoxygenase activity was measured in extracts of developing barley grain of Line G (M5 generation) and wild type cv Vintage. Grain was homogenised in ice-cold 20 mM Tris-HCl buffer pH 7.5 containing 0.1 % (v/v) Nonidet P-40, a non-ionic detergent that enhances lipoxygenase extraction, and centrifuged at 15,000 g for 20 minutes to remove insoluble material. Lipoxygenase activity in the extracts was measured polarographically in 200 ul oxygen-saturated buffer (0.2 M boric acid, 25 mM HEPES, pH 6.5) containing 1.2 mM linoleic acid at 25°C, using a Clark-type electrode to measure oxygen consumption.

Lipoxygenase activity increased during the first 20 days of grain development in both Line G and wild-type grain, but only in Line G did the activity level fall during grain maturation (Figure 4).

The relative amounts of 9-HPOD and 13-HPOD formed during linoleic acid oxidation provides a measure of the levels of LOX-1 and LOX-2 activity in the grain extracts. In this case Nonidet P-40 was omitted from the grain extraction buffer to avoid the co-extraction of hydroperoxide-consuming enzymes. The extracts (100 1), mixed with 10 ml 50 mM phosphate buffer pH 6.5 containing 200 uM linoleic acid, were incubated for 20 minutes. The reaction was terminated by adjusting

the pH to 3.5, and an internal standard was added. The hydroperoxides formed in the assay were bound on an octadecyl solid phase column (Bakerbond, Baker) and eluted with methanol. The 9-HPOD and 13- HPOD were then separated by reverse phase HPLC on a C-18 column with an isocratic elution solvent (tetrahydrofuran : Methanol : H20 : acetic acid; 25: 30: 44.9: 0.1 (v/v) adjusted to pH 5.5 with concentrated ammonia) at a flow rate of 0.5 ml/minute as described by Aarle et al., 1991, FEBS Letters 280: 159-162. Hydroperoxides were detected at 234 nm and the HPOD peaks were corrected against the internal standard, prostaglandin B2.

Figure 5 shows that 13-HPOD was the major product of lipoxygenase activity present in grain during the first 20 DAF, while 9- HPOD was formed by lipoxygenases active during grain maturation.

While both Line G and wild-type grain extracts shared a similar profile of 13-HPOD synthesising activity, Line G did not show the wild-type rise in 9-HPOD synthesising activity. These data are consistent with a loss of LOX-1 activity in maturing Line G barley grain.

3. Line G grain have a low-lipoxygenase 1 phenotype on germination Total lipoxygenase activity in extracts of embryos of grain germinated at 15°C was assayed as described in Example 1. The lipoxygenase activity present in quiescent wild-type grain declined during the first 4 days of germination and then increased (Figure 6). In Line G, lipoxygenase activity in quiescent grain was very low but increased after 4 days.

Analysis of the HPODs formed by the lipoxygenase activity in germinating embryos showed that 9-HPOD was the major product of lipoxygenases present in quiescent wild-type grain (Figure 7). The level of 9-HPOD formation fell with the decline in lipoxygenase activity in the extracts. The rise in lipoxygenase activity after 4 days was accompanied by the formation of both 9-HPOD and 13-HPOD. The low lipoxygenase

activity in Line G quiescent grain was associated with an absence of HPOD formation, while the rise in activity after 4 days mainly produced 13-HPOD. These data provide evidence that LOX-1 activity leading to the formation of 9-HPOD is greatly reduced in the embryos of both developing, quiescent and germinating barley grain of Line G, while LOX-2 activity leading to formation of 13-HPOD is unchanged in Line G.

Example 3 Line G has a Mutant Lipoxygenase 1 Gene (lox-1) Causing a Low Lipoxygenase Phenotype The molecular basis for the low-LOX-1 phenotype of Line G was investigated in order to provide a complete description of the mutant. The following analyses were performed to provide a complete characterization of the phenotype: 1. Lipoxygenase-1 is synthesised in the developing and germinating grain of Line G Western blot analysis of extracts of embryos from developing and germinating barley grain were performed in parallel with the measurement of lipoxygenase activity, as described in Example 2. The crude extracts were separated by sodium dodecyl sulphate- polyacrylamide gel electrophoresis (SDS-PAGE) according to Laemmli, 1970, Nature 227: 680-685. The separated proteins were transferred to nitrocellulose by semi-dry blotting, according to Towbin et al., (1979) Proc. Natl. Acad. Sci. USA 76: 4350-4354. The blot was probed with the LOX-1 specific monoclonal antibody, 5D2, as described Holtman et al., 1996, Plant Physiology 111 : 569-576, at 500x dilution, followed by incubation with goat anti-mouse antibody coupled to alkaline phosphatase, and detected with the alkaline phosphatase substrates nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl-phosphate as described

by Holtman et al., 1996, Plant Physiol. 111 : 569-576. The Western analyses revealed that LOX-1 protein was detected in developing grain from 10 DAF in cv Vintage embryos and the level increased during grain maturation (Figure 8). The protein was also present in the embryo of cv Vintage quiescent grain but declined slowly during germination (Figure 9). Although LOX-1 is recognised in extracts of Line G embryos and migrates in SDS-PAGE as a protein of similar size to cv Vintage LOX-1, the immunodetectable levels of the protein in Line G were slightly lower than in cv Vintage.

2. The lox-1 gene is expressed in the developing and germinating grain of Line G Total RNA was isolated from embryos of developing and germinating barley grain, according to the procedure of Hensgens and van Os-Ruygrok, 1989, Rice Genet. Newslett. 6: 163-168, in parallel with the measurement of lipoxygenase activity, described in Example 2. The RNA samples (7.5 ig) were separated on denaturing agarose gels and Northern blotted as described by Sambrook et al., 1989 in Molecular Cloning, a Laboratory Manual, Cold Spring Harbour Laboratory Press, NY. The blots were hybridised with a 32P-labelled probe generated from the barley 3'untranslated region, nucleotides 2659-2801 [SEQ ID NO : 1], of the lox 1 cDNA (EMBL Accession no. L35931) as described by Holtman et al., 1996 Plant Physiol. 111 : 569-576, using the Amersham Random Prime Kit.

Lox-1 transcripts encoding LOX-1 were detected in embryos of developing and mature cv Vintage and Line G grain from 30 DAF (Figure 10). The level of lox-1 trancripts increased during germination in both cv Vintage and Line G embryos, indicating de novo expression of the lox-1 gene (Figure 11). Since the detectable levels of lox-1 transcripts were similar in Line G and cv Vintage embryos, neither reduced lox-1 transcription or transcript stability can account for the low- lipoxygenase phenotype of Line G.

3. The lox-1 gene of Line G encodes a mutant form of lipoxygenase-1 The nucleotide sequence of the lox-1 gene of Line G and cv Vintage were analysed and compared in order to determine the molecular basis for the low-LOX-1 phenotype of Line G, which is characterised by normal transcription of the lox-1 gene, but reduced accumulation and activity in the expressed lipoxygenase enzyme in grain.

Genomic DNA from Line G and wild-type cv Vintage was isolated from seedling leaf tissue according to a method described by Pich and Schubert 1993, Nucleic Acids Res. 21: 3328. The lox-1 gene in the genomic DNA preparations was amplified by polymerase chain reaction (PCR) using primers based on the sequence of the barley lox-1 gene (van Mechelen et al. 1995, BBA 1254: 221-225; Rouster et al., 1997, Plant J. 11 : 513-523). The position and sequence of the oligonucleotide primers used to amplify the lox-1 promoter and coding regions, indicated in Figure 12 were as follows: Forward primer 5'-GAA AAG CTT GGA GGT AGA CGC TGC- 3' [SEQ ID NO : 2] and reverse primer 5'-TAT AGG ATC CTT GTT CTT GGC CTC CTC TCC TCG-3' [SEQ ID NO : 3] were used to PCR amplify the lox-1 promoter domain (-361 to +68) of Line G and cv Vintage lox-1.

Forward primer 5'-AGT GAA AAA CAG TGT GCT GGT G-3' [SEQ ID NO : 4] and reverse primer 5'-GGC TTA AAG AGC AAC TGC TGA-3' [SEQ ID NO : 5] were used to PCR amplify the Line G lox-1 coding region.

Forward primer 5'-CAA GAT GCA TAT GCT GCT GGG AG-3' [SEQ ID NO : 6] and reverse primer 5'-CGA TGG TTT AAA TTA GAT GGA GAT GCT GT-3' [SEQ ID NO : 7] PCR amplified the cv Vintage lox-1 coding region.

The PCR reactions consisted of 250 ng genomic DNA in a 50 il volume containing 50 pmol primer and 2 U Pfu DNA polymerase (Promega) according to the enzyme suppliers instructions. The PCR

amplifications were carried out in a Stratagene Robocycler: 1 minute at 94°C, 1 cycle; 1 minute at 94°C, 2 minutes at 62°C, and 5 minutes at 72°C, 30 cycles; 10 minutes at 72°C, 1 cycle. The PCR products were separated on 1.2 % agarose gels. DNA fragments, corresponding in length to the amplified region, were purified using Qiax II Gel extraction kit (Qiagen) and cloned into the plasmid pcDNA2.1 (Invitrogen). The nucleotide sequence of both strands of the cloned lox-1 promoter and coding regions was determined using the dideoxynucleotide chain termination reaction with specific oligonucleotide primers and analysed on an ABI PRISM 310 Genetic Analyzer (PE Biosystems). Sequence comparisons were performed using the DNA STAR sequence analysis software package (DNA STAR Inc., USA).

The promoter region and intron-exon structure of the barley lox-1 coding region are shown in Figure 13, and were deduced from a comparison of the nucleotide sequence of the wild-type lox-1 genomic and cDNA sequences (Figure 12). The sequenced region of the lox-1 promoter region from-363 to +68, (numbered relative to the determined transcription start site; van Mechelen et al., 1995, BBA 1254: 221-225), is sufficient to direct embryo-specific and temporally-regulated gene expression characteristic of the native gene (Rouster et al, 1998, Plant J 15: 435-440). The promoter and transcribed region of the wild-type lox-1 gene [SEQ ID NO : 8] is 4663 nt in length and contains 6 introns of between 82 nt and 634 nt in length, which are absent from the respective cDNA [SEQ ID NO : 10] and must therefore be removed during RNA transcript splicing.

Comparison of the nucleotide sequence of lox-1 of Line G with that of wild-type (Figure 12) showed that the Line G lox-1 allele has two point mutations. One is a silent C-+T substitution at position 221 in exon 1, and the second is a G-A substitution at position 2347 in exon 3 (Figure 13). The wild-type barley lox-1 gene encodes a protein of 862 amino acid residues [SEQ ID NO : 9], while the mutation at position 2347

in Line G lox-1 allele causes an amino acid substitution of glycine to aspartic acid at residue 368 in the encoded protein.

Alignment of related plant lipoxygenases indicated that the glycine-368 in barley LOX-1, is strongly conserved. Furthermore this residue, which corresponds to glycine-353 in soybean LOX-1, is one of 51 neutral or hydrophobic residues which line the substrate cavity of the enzyme, as seen from its crystal structure (Minor et al., 1996, Biochemistry 35: 10687-10701) and is shown in (Figure 22). The insertion of a charged amino acid residue at this position is thus likely to disturb the structural and functional properties of the enzyme.

4. The mutated LOX-1 protein encoded by the Line G lox-1 allele has low enzymic activity and is responsible for the low lipoxygenase phenotype of Line G.

The sodium azide mutagenesis of cv Vintage grain, which induced the mutated lox-1 allele in Line G, may have induced additional mutations in the Line G genome. Two experimental approaches have been taken to demonstrate that the mutant lox-1 allele in Line G is responsible for its low lipoxygenase phenotype, rather than other mutations in the genome. The enzymic activity of the LOX-1 encoded by the mutant and wild-type lox-1 allele have been determined in order to prove that the glycineoaspartic acid substitution in the mutant enzyme causes reduced stability and activity. The two lox-1 genes were transiently expressed in aleurone protoplasts isolated from imbibed mature grain, since the level of endogenous lipoxygenase expression in these cells was expected to be below detection limits. None of the identified barley lipoxygenase genes, which are expressed in germinating barley, are detected in the aleurone tissue (van Mechelen et al., 1999 supra). In order to direct transient expression of the lox-1 gene in aleurone protoplasts, their coding regions were translationally fused to a constitutive promoter known to be active in these protoplasts.

The coding regions of the mutant (sequence positions +1 to +4350) and the wild-type lox-1 gene (sequence positions +69 to +4230) and the wild-type lox-1 cDNA (sequence positions +69 to +2654) each cloned in plasmid pcDNA2.1 (see section 3), were excised by digestion of the KpnI and EcoRV sites in vector polylinker. The coding regions were cloned in the pUBARN plasmid (Jensen et al., 1998, Hereditas 129: 215-225) between the constitutively active maize ubiquitin Ubi promoter (as described in U. S. patent No. 005510474A) and the Nos terminator, in place of the bar gene which encodes phosphinotricin acetyl transferase (Figure 14).

Protoplasts were isolated from aleurone tissue of imbibed Hordeum vulgare cv Himalaya according to the protocol of Skriver et al.

1991, Proc. Natl. Acad. Sci. USA 88: 7266-7270. Aliquots of 2. 105 protoplasts were transfected at 0°C with ~100 pg plasmid DNA (equimolar amounts of each plasmid) by polyethylene glycol (PEG) mediated DNA uptake (Lee et al., 1997, Plant Mol. Biol. 13: 21-29), and then incubated in aleurone protoplast culture media at 25°C as described previously (Skriver et al., 1991 supra). After 48 hours incubation, the culture medium was carefully removed and the protoplasts were re- suspended and homogenised in 300 pLl lipoxygenase assay buffer (0.2 mM boric acid, 25 mM HEPES, pH 6.5). The homogenates were centrifuged at 15,000 g for 5 minutes to pellet insoluble material, and the supernatants (10 1ll) were subsequently assayed for total lipoxygenase activity using the rapid screening assay described in Example 1, section 1, but with omission of the NDGA inhibitor. The protein content of the protoplast extracts was measured by a Bradford dye-binding assay (Bradford 1976, Anal. Biochem., 72: 248) supplied by Bio-Rad Laboratories, Hercules, California, USA, and lipoxygenase activity was expressed per mg protein in the extract.

Protoplasts transfected with the control plasmid, pUBI-GUS, where the maize ubiquitin-1 promoter directs expression of the ß- glucuronidase reporter gene, gave no detectable lipoxygenase activity.

Transient expression of the wildtype lox-1 gene and cDNA in protoplasts both gave high levels of lipoxygenase activity in the protoplast extracts (Figure 15). The higher expression of the wild-type lox-l cDNA in comparison to the genomic sequence may be due to a higher transfection frequency for the smaller lox-1 cDNA expression plasmid (4929 bp versus the 6505 bp lox-l gene construct). Transient expression of the mutant lox-1 gene gave low levels of lipoxygenase activity,-10 % of wild-type lipoxygenase activity. These data clearly demonstrate that the mutant lox-1 gene in Line G encodes a lipoxygenase with greatly reduced activity, which accounts for the low lipoxygenase phenotype.

Example 4 PCR-Cleavage Amplified Polymorphic Site (PCR-CAPS) Assay: A Method Used for Identification of the Mutant lox-1 Gene An analytical method allowing the identification of the Line G mutant lox-1 gene in any genetic background was developed based on the PCR-CAPS assay. The assay involves PCR amplification of genomic DNA fragments, followed by digestion of the amplified sequences with a specific restriction endonuclease to display a restriction fragment length polymorphism (RFLP).

The coding sequence of the mutant lox-1 gene harbours two point mutations (see Example 3), where the mutation at position 2347 (Figure 12) introduces an additional Aat II restriction endonuclease cleavage site, not found in the wild-type lox-1 gene (Figure 16). The following PCR- CAPS assay, based on the polymorphism created by the presence of this restriction site in the lox-1 gene, is shown to descriminate between a wild-type lox-1 gene and a mutated lox-1 gene.

Genomic DNA was isolated from young leaves of M6 seedlings of Hordeum vulgare, L. cv Vintage and Line G according to the procedure of Pich and Schubert (1993, supra). The DNA sequence encompassing position 2347 (Line G lox-1 gene mutation site) was

amplified by PCR, using primers specific for the lox-1 gene [SEQ ID NO : 8]. The DNA fragments amplified by the selected forward primer 5'-CGCTACGACGTCTACAACGA-3' [SEQ ID NO : 13] and reverse primer 5'-CAGACTACTTTTTGGCGGGA-3' [SEQ ID NO : 14] are shown in Figure 17. PCR reactions were carried out with 250 ng genomic DNA in a 50-pl volume containing 50 pmol of each primer and 1 unit Taq DNA polymerase (Promega) according to the suppliers instructions. PCR amplifications were carried out on a Stratagene Robocycler as follows: 1 minute at 94°C, 1 cycle; 1 minute at 94°C, 1.5 minutes at 60°C, and 2 minutes at 72°C, 30 cycles; 10 minutes at 72°C, 1 cycle. The amplified fragments of the mutant and wild-type lox-1 gene were-650 bp (Figure 18), corresponding to the expected size (Figure 17). The PCR products, purified on a spin column (Qiagen), were digested with 25 unit Aat II restriction endonuclease for 24 hours at 37°C and analyzed on a 1.2% agarose gel.

Digestion of the wild-type lox-1 PCR product yielded DNA fragments of 10,179, and 462 bp, and the fragments from the mutant lox- 1 PCR product were 10,149,179, and 313 bp, where additional DNA fragments were due to partial digestion of the lox-1 PCR product (Figure 19). The fragment pattern corresponds to the expected RFLP resulting from this mutation, where the 313 bp fragment is unique to the mutant lox-1. This PCR-CAPS assay provides a reproducible and specific tool for identification of the lox-1 allele in barley and can thus be exploited in barley breeding programs aimed at introducing this gene in new barley varieties.

Example 5 Back-Crossing the Low Lipoxygenase Phenotype of Line G to cv Alexis Demonstrates a Genetic Linkage to the Mutant lox-1 Gene Repeated back-crossing was used to transfer the low-lipoxygenase phenotype from line G into a recurrent parent (in this case the cv. Alexis).

The back-crossing program shown in Figure 20, combined with selection

for the low-lipoxygenase phenotype, progressively substitutes the Line G genome by the recurrent parent genome. Furthermore, other mutations introduced into the Line G genome during the sodium azide mutagenesis treatment will be eliminated. In the first back-cross of the homozygous low-lipoxygenase Line G (denoted genotype 11) to cv Alexis (denoted genotype LL) the progeny lines will be heterozygous (denoted genotype L0. A low-lipoxygenase phenotype due to a recessive mutation will not be detectable in lines heterozygous for the mutation. The progeny are self-pollinated and will give a normal Mendelian segregating population, namely 1LL : 2 Ll : 1 Il. The low lox homozygous 11 genotype resulting from the first back-cross will have 50 % cv Alexis genetic background.

After ten rounds of back-crossing, the recurrent parent background will be approximately 99.9 %.

Hordeum vulgare, L. cv Alexis and Line G were propagated in a greenhouse throughout the back-crossing program. Back-crossed progeny grains were germinated in petri dishes on filterpaper, soaked with 4 ml H2O, for 3 days at 22°C in the dark. The low-lipoxygenase lines were screened by measuring total lipoxygenase activity in extracts of the coleoptile (top 7 mm) from the germinating seedlings, as described in Example 1. Progeny of the 3 and 4"back-cross were also analysed for inheritance of the mutant lox-1 gene using the PCR-CAPS assay described in Example 4.

The expected frequency of the low-lipoxygenase phenotype in the segregating progeny of the four back-cross generations was 25% for a recessive mutation. The observed frequency of low-lipoxygenase activity in the progeny (24 grains) of the four back-cross generations is in agreement with the expected frequency (Figure 20). When the 3d and 4"' back-cross progeny having the low lox homozygous 11 genotype were analysed with the PCR-CAPS assay, they were all found to have the diagnostic 313 bp fragment, while progeny having wild-type lipoxygenase activity lacked this fragment (Figure 21).

The back-crossing program demonstrates that the mutant lox-1 allele can be transferred to a new genetic background and is inherited in a recessive monofactorial manner following Mendelian segregation. Since the recurrent parent background is 93.8% in the 4"'back-cross progeny, the co-inheritance of the mutant lox-1 gene and the low-lipoxygenase phenotype provides confirmation of their genetic linkage.

Example 6 Beer Brewed From Line G Barley Malt Accumulates Less trans-2- nonenal During Storage, Giving an Improved Flavour Stability Hordeum vulgare L cv Vintage and Line G were propagated in the field over several seasons in order to provide sufficient grain for industrial malting. The following industrial scale malting and brewing trials as well as analyses of the finished beer were performed to demonstrate the value of the Line G low-lipoxygenase barley for improved flavour stability.

1. Industrial malting and kilning of Line G and cv Vintage Malting was performed on a 10-ton scale in an industrial malthouse in two trials as follows: Trial 1: Line G barley grain (1996 harvest) Steeping conditions: 8 hours wet; 14 hours dry; 8 hours wet; 10 hours dry; 4 hours wet in 16°C steeping water. Malting conditions: 12 hours at 18°C ; 24 hours at 16°C ; 24 hours at 14°C ; 60 hours at 12°C.

Kilning conditions: 12 hours at 60°C ; 3 hours at 68°C ; 4 hours at 74°C ; 3 hours at 80°C.

Trial 2: cv Vintage and Line G (1996/1997 harvest) Steeping conditions: 8 hours wet; 10 hours dry; 6 hours wet; 15 hours dry; 4 hours wet in 15°C steeping water. Malting conditions: 5 days with inlet air at 15°C and spraying to maintain moisture level.

Kilning conditions: 10 hours at 50°C ; 2 hours at 60°C ; 2.5 hours at 80°C.

Malting analyses of 2 samples of the Line G malt from Trial 1 compared to the control malt, cv Nevada (Table 1) and from Trial 2 compared to cv Vintage (Table 2) confirmed that Line G malt was suitable for brewing trials.

TABLE 1 MALTING TRIAL 1 Barley variety cv Nevada LineG LineG Crop year 1996 1996 1996 Malt analyses Moisture content % 4.7 4.3 4.4 Extract fine as is. % ai. 76.9 76. 1 75.3 Extract fine d. m. % dm. 81.4 79. 5 78.7 Saccharification time Min <10 <10 <10 Diastatic power % WK 252 373 365 Color EBC 2. 8 4. 4 3. 8 pH 6. 16 5. 97 5. 99 Turbidity EBC 9. 0 2. 5 2. 4 Total protein d. m. % 10. 35 10. 74 12. 03 Soluble nitrogen mg/1 647 787 765 Sol. Protein % malt d. m. % dm. 3. 7 4. 4 4. 3 Kolbach 35. 3 40. 8 35. 4 Free Amino Nitrogen mg/1 97 125 118 Friability % 89.5 85. 6 89. 5 Whole unmodified grains % 1. 1 0.6 0.5 Partly unmodified grains % 2. 3 1. 0 0. 6 P-glucaninwort mg/1 114 66 36 P-glucan in malt % w/w 0.24 0.11 0.05 S-methylmethionine/DMS eq. µg/g 2.4 6.4 8.4 Free DMS pg/g 1. 0 6. 6 4. 7 NDMA µg/kg n.d. 0. 3/0.6 0.3/0.3 TABLE 2 MALTING TRIAL 2 Barley variety Vintage Line G Line G Crop year 1996 1996 1997 Moisture content % 4.1 4.1 4.3 Extract fine as is. % ai. 77. 0 75. 6 77. 5 Extract fine d. m. % dm. 80. 3 78. 8 80. 9 Fine/coarse difference % dm. 0.7 1.6 1.7 Saccharification time min--< 10 Diastatic power % WK 343 342 268 Color EBC 2. 5 2. 8 3. 4 PH 6. 05 6. 01 6. 12 Turbidity EBC 1. 5 1. 3 2. 5 Total protein d. m. % 10.98 12. 22 9. 82 Soluble nitrogen mg/l 696 741 610 Sol. Protein % malt d. m. % dm. 3. 9 4. 1 3. 4 Kolbach 35. 2 33. 7 34. 6 Free Amino Nitrogen mg/l 110 117 100 Friability % 91. 3 81. 8 89. 5 Whole unmodified grains % 0. 7 1. 1 1. 3 Partly unmodified grains % 1. 0 2. 7 2. 7 (3-glucan in wort mg/l 97 172 117 Beta-glucan in malt % w/w--0. 3 S-methylmethionine/DMS eq. pg/g Free DMS µg/g - - 2. Industrial brewing with Line G, cv Vintage malt, and control malt cv Nevada

Two brewing trials were performed, using wort prepared from Line G and control malt cv Nevada malt in Trial 1 and from Line G and control malt cv Vintage in Trial 2.

Beer was brewed on a 30-hl industrial scale with 475 kg malt according to the following scheme: Mashing in at 50°C ; 30 minutes at 50°C ; 30 minutes heating from 50-70°C ; 15 minutes at 70°C. A portion of the wort was heated for 20 minutes from 70-100°C and 5 minutes at 100°C, while the main mash was kept at 70°C for another 25 minutes and then the two mashes was combined and kept for 10 minutes at 76°C. The brewing steps of wort boiling, whirlpool separation of spent grain, cooling, fermentation, lagering and packaging in brown glass bottles were according to standard brewing practise.

3. Flavor stability and T2N content of beer brewed from Line G, cv Vintage malt and control malt cv Nevada The freshly bottled beer was stored at 5°C and analysed within 2 months of production. The flavor-stability of the fresh and stored beer was evaluated in two independent laboratories following two different types of beer storage conditions. In laboratory A the beer was subjected to a forced aging process, where the beer was stored at 37°C for a period of 7 days, while in laboratory B the beer was stored at 30°C for 6 and 12 weeks. Trans-2-nonenal levels in beer were determined by gas chromatography and mass spectrometric detection following derivatisation of carbonyls with 0- (2, 3,4,5,6-pentafluorobenzyl)- hydroxylamine, essentially as described by Gronqvist et al. 1993 Proceedings of the 24th EBC Congress, Oslo, 421-428. A trained beer taste panel evaluated the overall flavor score of the beer, which includes detection of a cardboard flavor, indicative of free trans-2-nonenal in the beer.

Laboratory A: Forced-Aging Tolerance Comparison of beer, brewed from Line G and the control malt, cv Nevada, in the first brewing trial (Table 3) demonstrated that beer from

Line G had a greater flavor stability and a lower trans-2-nonenal content following forced-aging as compared to the controls. The second trial, comparing beer brewed from Line G malt with beer brewed from cv Vintage malt, the parental cultivar, confirmed the initial data (Table 4).

TABLE 3 BREWING TRIAL 1 Barley Malt cv Nevada Line G SO2 Content (mg/ml) 1 1 T2N** (ppb)-Fresh Beer 0. 009 0. 005 T2N** (ppb)-Aged Beer (37°C/7 DAY) 0.117 0.025 Flavor*-Fresh Beer 5. 9 5. 3 Flavor*-Aged Beer (370C/7 DAY) 1. 3 5. 1 * Flavor evaluation scale 1-10 of increasing quality; **trans-2-nonenal.

TABLE 4 BREWING TRIAL 2 Barley Malt cv Vintage Line G SO2 Content (mg/ml) 2 2. 5 T2N** (ppb)-FreshBeer 0. 023 0. 019 T2N** (ppb)-Aged Beer (37°C/7 Day) 0. 078 0.035 FLAVOR*-Fresh Beer 5. 5 6. 1 FLAVOR*-Aged Beer (37°C/7 Day) 2. 9 5. 9 * Flavor evaluation scale 1-10 of increasing quality; **trans-2-nonenal.

Laboratory B: 30°C Storage Tolerance Beer brewed from Line G malt had lower trans-2-nonenal levels following 6 and 12 weeks at the elevated storage temperature of 30°C, when compared to beer brewed from either of the reference malts (Table 5 and 6) and had a better flavor-stability as judged by a taste panel. The taste-threshold for trans-2-nonenal in these analysed beers lies close to 0.08 ppb.

TABLE 5 BREWING TRIAL 1 Barley Malt cv Nevada Line G trans-2-nonenal (ppb)-Fresh Beer 0.050 0.044 trans-2-nonenal (ppb)-30°C/6 weeks 0.072 0.037 trans-2-nonenal (ppb)-300C/12 weeks 0.095 0.046 trans-2-nonenal flavor *-Fresh Beer 0.6 0.3 trans-2-nonenal flavor *-30°C/6 weeks 3.7 1.4 trans-2-nonenal flavor*-30°C/12 weeks 2. 5 0.6 * trans-2-nonenal flavor detection score on a scale of 1-10

TABLE 6 BREWING TRIAL 2 Barley Malt cv Vintage Line G trans-2-nonenal (ppb)-Fresh Beer 0. 070 0. 062 trans-2-nonenal (ppb)-30°C/6 weeks 0.093 0.070 trans-2-nonenal (ppb)-30°C/12 weeks 0. 133 0.080 trans-2-nonenal flavor*-Fresh Beer 0.3 0.9 trans-2-nonenal flavor*-30°C/6 weeks 2. 5 1. 7 trans-2-nonenal flavor*-30 () C/12 weeks 2. 2 1.3 * trans-2-nonenal flavor detection score on a scale of 1-10

The improved flavor-stability of beer brewed from Line G malt, as measured by the levels of trans-2-nonenal detected in the beer following storage at 30°C from the combined brewing trial data, is shown to be statistically significant (Table 7).

TABLE 7 TRANS-2-NONENAL IN STORED BEER fresh Mean Stdev Difference 2-tailed p Significant (p < 0.05) reference 0. 060 0. 012 0. 007 0.34 no line-G 0.053 0.011 6 weeks, 30 °C Mean Stdev Difference 2-tailed p Significant (p < 0.05) reference 0. 083 0. 013 0.029 0.031 yes line-G 0. 054 0.021 12 weeks, 30 °C Mean Stdev Difference 2-tailed p Significant (p < 0.05) reference 0.114 0.023 0.051 0.003 yes line-G 0. 063 0. 020

Since the natural sulfite levels were low in both brewing trials, the free trans-2-nonenal levels in the aged beer would closely reflect the trans-2-nonenal potential of the different beers, namely the level of trans-2-nonenal adducts present in the fresh beer. Addition of sulfite can temporarily delay the staling process, by complexing free-trans-2- nonenal, until sulfite levels are reduced by oxidation due to gaseous exchange through the packaging.

The described brewing trials with low-LOX-1 barley malt provide the first unequivocal evidence that LOX-1 activity in barley during the malting and brewing process is a key determinant of the appearance of the off-flavor compound trans-2-nonenal in aged beer.

The above specification includes citations to numerous publications. Each publication is hereby incorporated by reference for all purposes, as if fully set forth. INDICATIONS RELATING TO DEPOSITED MICROORGANISM OR OTHER BIOLOGICAL MATERIAL (PCT Rule 13bis) A. The indications made below relate to the deposited microorganism or other biological material referred to in the description on page 16 line B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional sheet Name of depositary institution National Collections of Industrial, Food and Marine Bacteria (NCIMB) Address of depositary institution (including postal code and country) 23 St. Machar Drive, Aberdeen, AB243RY, Scotland, United Kingdom Date of deposit Accession Number 04 January 2001 NCIMB 41078 C. ADDITIONAL INDICATIONS (leave blank if not applicable) This information is continued on an additional sheet D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (iflhe indications are noifor all designated States) E. SEPARATE FURNISHING OF INDICATIONS (leave blank ifnot applicable) The indications listed below will be submitted to the International Bureau later (speclhegeneralnature ofthe indications e. g.,'Accession Number oJDeposit'J For receiving Office use only For Intemational Bureau use only g This sheet was received with the international application This sheet was received by the Intemational Bureau on: Authorized officer Authorized officer 1 W \J_ oC