WO2000043529A1

WO2000043529A1 - Method of increasing the water soluble antioxidant content in mechanically harvested peas

Info

Publication number: WO2000043529A1
Application number: PCT/EP2000/000331
Authority: WO
Inventors: Christine Davies; Karl John Hunter; Heike Dornenburg
Original assignee: Unilever Plc; Unilever Nv; Hindustan Lever Limited
Priority date: 1999-01-19
Filing date: 2000-01-17
Publication date: 2000-07-27
Also published as: AU3149400A; EP1144663A1

Abstract

This invention related to a method of improving the water soluble antioxidant content in mechanically harvested peas by way of one or more genetic modifications that are capable of reducing the activity of the lipoxygenase type-2 isoenzyme.

Description

METHOD OF INCRE-SING THE WATER SOLUBLE ANTIOXIDANT CONTENT IN

MECHANICALLY HARVESTED PEAS

FIELD OF THE INVENTION

The invention relates to a method for improving the water soluble antioxidant content of peas, products derived therefrom and methods for genetically modifying peas, in particular for affecting the water soluble antioxidant content.

BACKGROUND TO THE INVENTION

Antioxidants vitamins are essential to man, however these cannot be synthesised or stored in sufficient quantities within the human body. Other molecules also have antioxidant properties. It is therefore necessary that the diet contains a regular, adequate supply of these compounds. Antioxidants can be subdivided into those which are lipid soluble, such as polyenes e.g. retinoic acid and those which are only soluble in aqueous solution such as ascorbate, glutathione, hydroxycinnamates, and some plant derived flavonoids.

Antioxidants are integrally involved in many cellular systems where they can be preferentially oxidised to allow regeneration of other metabolites and prevent toxin build up. For example, ascorbate reacts rapidly with superoxide, singlet oxygen, ozone and hydrogen peroxide thus removing these reactive and potentially harmful forms of oxygen generated during aerobic metabolism from cellular systems . Such a capacity for preferential oxidation makes antioxidants susceptible to degradation in plant systems, particularly in the ripening of fruit and vegetables where antioxidants levels can fall dramatically as the plant material ages . Given that plants are the major source of antioxidants compounds in the human diet, it is clearly desirable to reduce or slow the antioxidant degradation post-harvest in fruit and vegetables destined for human consumption, thereby providing more nutritious food products.

One particular group of enzymes that are generally associated with the post harvest spoilage of fruits and vegetables are the lipoxygenases . Lipoxygenases are a widespread group of enzymes that are of importance due to their involvement in both animal processes, leading to hypersensitivity and inflammatory responses, and in plant processes, leading to flavour and aroma/odour formation. These enzymes have been studied quite intensively in plants and are relatively abundant in legume seeds.

The lipoxygenase family comprises a group of non-heme iron- containing enzymes that catalyse the stereospecific dioxygenation of polyunsaturated fatty acids containing a 1,4- cis, cis-pentadiene moiety, to a pentadienyl radical intermediate. The overall process involves the redox cycling of the iron present in the enzyme molecule. Fatty acid hydroperoxides undergo one electron reduction to generate active ferric lipoxygenases from the inactive ferrous form and fatty acid alkoxyl radicals.

Much of the knowledge of lipoxygenase in the scientific literature has come from work on soybean lipoxygenase, which has emphasised the importance of these enzymes in flavour development but provided little understanding in respect of nutritional components . Research with the soybean has indicated that there are five different known isoforms of lipoxygenase in soybean. Flavour improvement of soybean preparations has been achieved through the identification of a mutant line lacking lipoxygenase-2 ( lox2-) .

Soybean lipoxygenases have also been shown to co-oxidise polyenes, including the lipid soluble antioxidants such as retinoic acid; which can be co-oxidised by soybean LOX-2 and to a lesser extent by LOX-3.

Recent research (Roy et al, 1996 Food Chem. Tox. 1996 34: 563- 570) has given some evidence of the lipoxygenase isoenzyme type 5 from soybean, being capable of ascorbic acid co-oxidation which has been generalised as property of all soybean lipoxygenases .

Pea seeds primarily contain lipoxygenases corresponding to only two of the soybean enzymes (LOX-2 and LOX-3) . lox2- peas lack the LOX-2 polypeptide, and on oxidation of linoleic acid give 13-hydroperoxide: 9-hydroperoxide in the ratio 1:2; whereas standard peas produce approximately equal amounts of 13- hydroperoxide and 9-hydroperoxide (Z. u & D.S. Robinson, J. Agric. Food Chem. 43 1995) . Peas with a low activity of lipoxygenase-3 isoenzyme were found to produce more of the 13- hydroperoxide .

The Applicants have found a higher total lipoxygenase activity for lox 2- peas as compared to standard, illustrated in Figure 1. Though unexpected these results are taken to be indicative of over expression of the LOX-3 isoenzyme in the lox2- peas. Lipoxygenase activity as determined by an oxygen electrode assay was strongly dependent on the developmental stage of the pea seeds . Figure 1 illustrates lipoxygenase activity in Iox2+ and lox2- peas during different development times (measured by oxygen uptake) . Pea development stages can be described as immature

(15 and 20 DAA) , mature (22 and 25 DAA) , and overmature (28 and 30 DAA) .

Lipoxygenase activities were higher in Iox2- compared to lox2+ peas presumably due to overexpression of the LOX-3 isoenzyme.

Cellular damage during the harvesting and subsequent processing prior to sale and consumption of a fruit or vegetable crop will often cause the release of the polyunsaturated fatty acids and monoglycerides that are the substrate of the lipoxygenases and in this way the spoilage reactions in the fruit or vegetables are initiated. This is particularly relevant to the mechanical harvesting of peas which requires the pea seeds to be threshed from their pods, commonly inflicting significant cellular damage on the peas and making fatty acids available for lipoxygenase. It would therefore be beneficial to control these undesirable effects of lipoxygenases to preserve the nutritional quality in the peas for a prolonged time period.

It has been shown in research into peas that an index for lipid peroxidation correlates with other deterioration indices, including ascorbate loss (Furata et al, 1995 Biosci. Biotech. & Biochem. 59(1), 111-112). Furata teaches that lipid oxidation occurring in fats, oils etc. is chemically driven, with no suggestion of any enzyme involvement .

Further interest has been paid in the context of peas to increasing the content of the water soluble antioxidant ascorbate, by genetic modification to the pathway of biosynthesis between sucrose and starch, EP 98302744.2 unpublished. Earlier findings of WO 98/01574 have demonstrated that a mutation at the rug 3 locus encoding plastidial phosphoglucomutase is able to provide an increase in sucrose content over a longer ripening period.

The applicants have identified a clear need to establish the means by which water soluble antioxidants are degraded in pea seeds so that such degradation can be controlled. The present invention is therefore directed towards a method of improving the water soluble antioxidant content of mechanically harvested pea seeds.

SUMMARY OF INVENTION

The present invention is based on the surprising finding that a genetically modified pea plant in which the gene encoding lipoxygenase type-2 isoform is at least substantially inactive can be used to improve post harvest levels of water soluble antioxidants above those seen in wildtype or lox3- type peas which have undergone similar mechanical harvesting. The applicants have surprisingly found that the oxidation of lipids, specifically mediated by LOX-2 isoenzyme is responsible for the degradation of water soluble antioxidants in peas. Whereas, lipid oxidation by LOX-3 isoenzyme does not lead to antioxidant co-oxidation. In particular, the applicants have found that the oxidation of lipid by LOX-2 isoenzyme is largely responsible for the post harvest degradation of ascorbate in mechanically harvested peas.

Accordingly, the invention is based on the finding that genetic modifications that reduce the biosynthesis of LOX-2 isoenzyme, result in a higher level of water soluble antioxidants in mechanically harvested peas as compared to similarly harvested unmodified peas.

The present invention in the first aspect therefore comprises a method of improving water soluble antioxidant content in mechanically harvested peas, characterised by a genetic modification capable of reducing the activity of lipoxygenase type-2 isoenzyme (LOX-2) in said peas.

These and other aspects of the invention will be described in further detail in the description which will follow hereinafter.

DETAILED DESCRIPTION OF THE INVENTION

For the purpose of the invention the term genetic modification is taken to include any modification of oligonucleotide coding sequence or function e.g. mutation, antisense and co- suppression; use or alteration of promoters, terminators or enhancers etc. Co-oxidation describes the consumption of antioxidant during the process of lipoxygenase catalysed lipid oxidation.

In order to achieve an effective reduction in the activity of the lipoxygenase-2 isoenzyme it is preferable that the genetic modification is at the LOX-2 locus.

The lox2- lines used in the present invention were derived from a mutant line of Pisum fulvum, lacking seed lipoxygenase-2. The lox2- was bred into a round seeded variety of pea called "Birte" (available from John Innes Centre, Norwich Research Park, Colney, Norwich.) Subsequently, these homozygous lox2-lox2- lines have undergone crosses into vining pea varieties such as Harrier, Novella etc. to remove P. fulvum and Birte background. Here progeny were selected on the basis of wrinkled seed, and as such are closer in their phenotype to the vining pea parent than to the round seeded parent. These were later screened using standard immunochemical techniques to confirm the presence of the lox2- trait.

The skilled person would recognise that there are alternative methods for developing pea plants that show the lox2- trait, i.e. plants showing a reduction in LOX-2 isoenzyme activity. Moreover, it is clear that the scope of the present invention is not restricted to improving the content of water soluble antioxidant in mechanically harvested peas by the introduction of such a trait by a particular breeding programme or genetic modification route.

Mutant pea lines with the lox 2- trait could be developed by a process for introgression of the trait into commercial pea lines e.g. using molecular markers defining the region of the chromosome encoding the LOX-2 gene and thereby restricting the amount of carry-over of the undesirable background from the parent line.

Further lox2- lines would be achievable by conducting a mutagenesis programme and screening the progeny for mutations affecting the LOX-2 gene, including introns and promoter regions . Such a programme could be carried out at the level of mutagenising seeds using appropriate chemicals or radiation, or by site directed mutagenesis of the LOX-2 gene.

Alternatively there are a number of methods for producing transgenic plants that are available to those skilled in the art that are known to be suitable for the process of the invention. Pea plants lacking lipoxygenase-2 activity could be produced by transformation with a vector carrying the LOX-2 gene or part thereof, either in a sense or anti-sense orientation. Construction of such vectors as known to those skilled in the art typically comprises a promoter and terminator sequence separated by the coding region of interest.

A second aspect of the invention therefore comprises a method of improving water soluble antioxidant content in mechanically harvested peas, as stated above, wherein the genetic modification is a mutation at the LOX-2 locus.

Although the Applicants believe that the method of the invention is applicable to increasing the content of all water soluble antioxidants in mechanically harvested peas, it is preferable that the antioxidants are selected from the group comprising, ascorbate, glutathione, hydroxycinnamates, and plant derived flavonoids . Ascorbate degradation was determined during lipid oxidation and a clear reduction in the capacity for degradation in lox2- pea extracts in comparison to unmodified peas has been demonstrated (Figure 2) . Extracts of the modified peas showed a constant level of ascorbate degradation that was independent of lipoxygenase activity.

The Applicants have also found that superoxide anion production as well as 13-linoleic acid hydroperoxides were not responsible for the ascorbate degradation.

Whilst not wishing to be bound by any particular theory the Applicants believe that the link between lipid oxidation and ascorbate degradation is co-oxidation, which must be a property of the lipoxygenase-2 isoenzyme present in standard peas and not present in the lox2- peas. Furthermore, the Applicants believe that the elevated level of ascorbate that are maintained in the lox2- peas, compared to that in unmodified peas, has a more general inhibitory effect on further ascorbate degradation as well as that of other water soluble antioxidants.

Pea lipoxygenase-2 isoenzyme is therefore capable of ascorbate co-oxidation on oxidation of fatty acids containing a 1,4 cis,cis pentadiene moiety. The Applicants believe that ascorbyl free radicals produced by the co-oxidation reaction disproportionate to form dehydroascorbate and ascorbate rather than reacting with molecular oxygen to generate partially reduced reactive oxygen species.

Nutritionally ascorbate is particularly important to human health and it is therefore particularly preferred that this water soluble antioxidant is maintained post harvest by the method of the present invention. Accordingly a further aspect of the invention provides a method of improving water soluble antioxidant content in mechanically harvested peas as discussed above, wherein said water soluble antioxidant content preferably comprises ascorbate.

In a related aspect of this invention the Applicants have also surprisingly found that a reduction in the activity of lipoxygenase-2 isoenzyme not only maintains ascorbate levels post harvest, through a reduction of its co-oxidation during lipid oxidation, but also as a consequence of a differing response within the pea to conditions of stress. The Applicants have therefore also found that the pea seed lipoxygenase-2 isoenzyme is a stress related enzyme.

During the vining process of commercial pea harvesting where the peas are threshed from the pod using a mechanical harvester, it is inevitable that an amount of wounding or stress to the pea crop occurs . Such wounding causes a general cellular increase in non-specific peroxidase activity which is believed by the Applicants to cause stress driven ascorbate degradation.

The Applicants have found that peas lacking the lipoxygenase-2 isoenzyme do not show a significant increase in non-specific peroxidase activity on wounding and therefore are less susceptible to stress driven ascorbate degradation post harvest than equivalent unmodified peas. This is of importance in the harvesting and packaging of the peas allowing ascorbate levels to be maintained for longer post harvest, thus providing a larger temporal window for processing of the peas prior to blanching, freezing or other appropriate process steps. Ascorbate was found to inhibit lipid oxidation whereas α- tocopherol, a non-water soluble antioxidant stimulated the reaction. The inhibition and stimulation effect of these antioxidants was dependent on the maturity and type of pea (standard and lox2-) and is believed to reflect the presence of different lipoxygenase isoenzymes. Total lipoxygenase activity increased during seed development in both standard and lox2- pea extracts with the ratio of the formation of oxodienes to hydroperoxides also changing.

In standard peas the inhibitory effect of ascorbate decreased and the stimulatory effect of α-tocopherol increased with maturity. In contrast to this, the inhibitory effect of ascorbate increased with maturity in lox2- pea extracts and stimulating effect of α-tocopherol was not as dramatic as seen in the more mature standard peas .

Activation of lipoxygenases by α-tocopherol in lox2- peas was only 2-fold in seeds 15 days after anthesis, and even lower in seeds 22 and 28 days after anthesis, compared to 28 day lox2+ pea extracts where a 6.5-fold activation was seen.

It is recognised that the lox2- trait would be particularly suitable for application in combination with additional genetic modifications that are capable of increasing pea ascorbate content. In this way the combined effect of not only improving post harvest pea ascorbate content by maintaining levels that conventionally exist for a longer period, but, more desirable further elevated ascorbate levels can be achieved.

The Applicants have shown in the earlier co-pending application EP 98302744.2 (unpublished at the time of filing) that genetic modifications affecting the biosynthetic pathway from sucrose to starch provide an increased level of ascorbate in plant tissues. It is now appreciated that it would be highly desirable to combine a genetic modification capable of increasing the sucrose content of peas with a genetic modification capable of reducing the activity of the lipoxygenase-2 isoenzyme.

For the purpose of the present invention genetic modification yielding an increase of one or more intermediates in the sucrose starch synthesis pathway selected from the group comprising; UDP glucose, glucose-1-phosphate, glucose-6-phosphate, fructose-6- phophate and ADP glucose can be used to provide an increase in the content of ascorbate. Preferably the Applicants seek to increase the content of glucose-6-phosphate.

A further aspect of the present invention therefore relates to a method of improving the ascorbate content in mechanically harvested peas, characterised by the combination of a genetic modification capable of reducing the activity of the lipoxygenase-2 isoenzyme with a further genetic modification at any locus capable of reducing the activity of one or more enzymes involved in the sucrose to starch biosynthetic pathway.

Moreover the genetic modifications for reduction of the particular enzyme activity in sucrose/starch synthesis pathway may take place at one or more of the group of loci comprising r, rb, rug3, rug4 , rug5 encoding the enzymes; starch branching enzyme, ADP glucose pyrophosphorylase, plastidial phosphoglucomutase, sucrose synthase and starch synthase respectively.

WO 98/01574, describes the importance of plastidial phosphoglucomutase as a key enzyme in the sucrose/starch biosynthetic pathway, moreover, that a reduction in the activity of this enzyme can give rise to pea seeds which have particularly high levels of sucrose in comparison to the wildtype. This enzyme is encoded at the rug3 locus.

In view of the above mentioned positive correlation between sucrose and ascorbate levels in pea seeds the Applicants believe that an increase in the content of ascorbate in a lox2- pea plant can be achieved by reducing plastidial phosphoglucomutase activity preferably by means of a modification at the rug3 locus. The Applicants have accordingly undertaken to combine the lox2- trait with a further genetic modification that provides reduced plastidial phosphoglucomutase activity.

As outlined above in the context of introducing the lox2- trait into a pea line, the skilled person would be aware of a number of techniques that could be used to provide the desired genetic modifications such as molecular marker assisted selection, molecular genetic transformation techniques etc. By way of illustration of this aspect of the present invention the Applicants have chosen to cross peas as prepared in the mutagenesis programme described in detail in WO 98/01574 and containing a mutation at the rug3 locus with pea lines showing the lox2- trait.

This embodiment of the present invention therefore resides in a method of improving the ascorbate content in mechanically harvested peas, wherein said genetic modifications are capable of reducing lipoxygenase type-2 isoenzyme and plastidial phosphoglucomutase enzyme activities. A further embodiment hereof being further characterised in that the reduction in lipoxygenase type-2 isoenzyme activity is achieved by a genetic modification at the LOX-2 locus and the reduction in plastidial phosphoglucomutase activity is achieved by a genetic modification at the rug 3 locus .

Peas obtained according to the method outlined above would also be encompassed by the present invention, in particular where said peas are blanched/processed and in the frozen state.

The invention also comprises a modified pea plant or part thereof, characterised by reduced activity of lipoxygenase type- 2 isoenzyme and plastidial phosphoglucomutase compared to an equivalent unmodified plant. An equivalent unmodified plant is a plant which has a substantially identical genotype to a modified plant of the invention excepting the genetic sequence modifications present in those plants of the invention.

Similarly the invention comprises a modified pea plant or part thereof showing reduced activity of lipoxygenase-2 isoenzyme and plastidial phosphoglucomutase as described above wherein said modified pea plant or part thereof comprises a genetic modifications at the lox2 and rug3 loci.

The invention will be further described by way of detailed examples .

Example 1

The effect of the water soluble antioxidants on lipoxygenase activity.

The effect of ascorbate and its associated degradation on lipoxygenase activity was examined in pea (Pisum sativum) extracts using standard peas { lox2+) and peas lacking the lipoxygenase-2 isoenzyme { lox2-) . Standard peas ( Pisum sativum cv. Harrier) and lox2- peas { Pisum sativum cv. Harrier x lox2-, lox2- derived from Pisum fulvum and originally from John Innes Centre, Norwich, UK) were grown under greenhouse conditions. Emerging pea flowers (1st and 2nd node) were tagged to enable peas at different stages in development to be selected for study. Peas were collected at different days after anthesis (DAA) . Raw material was frozen in liquid nitrogen prior to storage at -70°C.

5g of frozen pea seeds were homogenised on ice in 20ml 0.1M Tris/HCl buffer, pH 6.8 with an homogeniser (Ultra-Turrax) for 30 seconds at 23,000 rpm. The cell homogenate was centrifuged for 30 minutes at 35,000g. 2.5ml supernatant were gel-filtered through a PD-10 Column (Pharmacia) to remove salts and endogenous antioxidants, into 0.1M Tris/HCl buffer, pH 6.8 and used in enzyme assays.

Ascorbate concentration was monitored at 298nm during 10 minutes of incubation at 25°C. The incubation medium consisted of 800μM linoleic acid sonicated with Tween 20, 1:1 w/w in 50mM Tris/HCl buffer, pH 8.3. Ascorbate degradation was also determined in 50mM Tris/HCl buffer at pH 6.0, 7.0, 8.0 and 9.0. lOOμl of enzyme extract was used.

Figure 2 illustrates the correlation between ascorbate degradation and lipoxygenase activity in different lox2+ and lox2- peas.

Ascorbate added to cell free pea extracts was degraded concomitantly with increasing lipoxygenase activity in lox2+ pea extracts as assayed by the oxidation of linoleic acid performed at a pH of 8.0, 8.3 and 9.0. The amount of ascorbate (500μM) used in this assay is known to inhibit lipoxygenase activity by 18% (Roy and Kulkarni, 1996) . Chemical degradation of ascorbate was excluded by using a linoleic acid/ascorbate mixture as a blank.

The rate of ascorbate degradation was nearly constant in the experiments performed with lox2- pea extracts (Figure 2) . Lipoxygenase activities in lox2- pea extracts were significant higher than in lox2+ peas at the same stage of development (Figure 1 and 2) , but the behaviour of lipoxygenase action regarding oxidation of ascorbate is completely different.

As the lox2- peas lack the lipoxygenase-2 isoenzyme, it is concluded that LOX-2 is predominantly responsible for ascorbate degradation observed in the experiments in the presence of linoleic acid with lox2+ pea extracts. As ascorbate degradation did not increase with lipoxygenase activity in lox2- pea extracts, Figure 2, the LOX-3 isoenzyme is not responsible for ascorbate degradation.

Superoxide anions

The production of superoxide radicals was measured by monitoring the reduction of cytochrome c as an increase of absorbance at 500nm. The standard assay mixture contained 22μM cytochrome c, 500μM ascorbate, 800μM linoleic acid emulsion with Tween 20, 1:1 w/w in 50mM Tris buffer, pH 8.3. The reference cuvette contained all the components except enzyme extract and served as the control. The inhibition of cytochrome c reduction by 20- 150U/ml superoxide dismutase was used as a test for superoxide anion radical production.

Cytochrome c reduction was monitored at 550nm in the presence and absence of superoxide dismutase (SOD) to identify the potential generation of superoxide anion radicals concomitant with ascorbate co-oxidation.

The failure to block the reduction of cytochrome c using superoxide dismutase concentrations as high as 150U/ml indicates a lack of generation of superoxide anion radicals during ascorbate oxidation by the lipoxygenases. With respect to this, it can be presumed that the SOD uninhibitable cytochrome c reduction was due to the production of ascorbyl free radicals (otherwise forming dehydroascorbate and ascorbate) rather than the formation of superoxide radicals .

The use of high-purity water in our study reduced the autoxidation of ascorbate to an undetectable level over the incubation time used. It has been documented in the literature that ascorbate is an excellent chain-breaking antioxidant and free radical scavenger (Rose and Bode, 1993 FASEB J. 7, 1135- 1142) . One of the most important reasons for this property is the fact that ascorbyl free radical reacts with oxygen very slowly (Bielski et al, 1975 Annals N.Y. Acad. Sci. 258, 231-237) and hence the possibility of formation of reactive oxygen species is minimal .

Hydroperoxide intermediates

500μM ascorbate was incubated with different concentrations of 13 (S) -linoleic acid hydroperoxide (0, 5, 10, 25, lOOμM per assay) . This resulted in a low rate of ascorbate degradation (6.15 nmoles/min per g fresh weight). Furthermore, when cell free pea extract was added (lox2+ and lox2 ) ascorbate degradation was slightly higher (8.96 nmoles/min per g fresh weight) , but still low and independent of level of lipoxygenase present.

This is evidence that ascorbate degradation is due to co- oxidation of fatty acids by lipoxygenase and not due to a chemical reaction involving the hydroperoxide. Other ascorbate degrading enzymes

The role of lipoxygenases in degradation of ascorbate and the apparent non-involvement of other enzymes is further supported by investigation of the other ascorbate degrading enzymes found in peas (Table 1) . Although significant activities were observed for ascorbate peroxidase and guaiacol peroxidase, both enzymes require the presence of hydrogen peroxide to oxidise ascorbate. As this was not present in the ascorbate degradation experiments, co-oxidation by lipoxygenase is indicated. Although ascorbate oxidase activity was detected in both lox2- and standard peas, the activity does not correlate with ascorbate degradation (Figure 2; Table 1) .

Table 1. Activities of ascorbate oxidising enzymes and peroxidases found in cell free pea extracts

AOX: ascorbate oxidase (μmol/min/g FW) ; APOX: ascorbate peroxidase (μmol/min/g FW) ; POX: peroxidase (ΔAbs₄oo/min g FW) ; DAA: Days After Anthesis. Example 2

Lipoxygenase-2 isoenzyme and stress responses in peas.

Standard peas (Pisum sativum cv. Harrier) and Iox2- peas ( Pisum sativum cv. Harrier x lox2-; lox2- parent from John Innes Centre, Norwich, UK) were grown under greenhouse conditions.

Cell free extract preparation and enzyme assays (guaiacol peroxidase, ascorbate oxidase, and lipoxygenase activity) are as described above. Phosphatidylcholine from soybean (70.6% linoleic acid, 9.2% linolenic acid; Sigma P7443) and phosphatidic acid from egg yolk (11.2% linoleic acid; Sigma P9511) were used in emulsion.

On wounding caused by vining analogous to commercial pea harvesting (threshing the peas from the pod using a mechanical harvester) , a significant increase was detected for unspecific

(guaiacol) peroxidase activity in standard peas of nearly 30% which was not seen in lox2- peas.

Peroxidase is known to be an enzyme which is induced in plant cells exposed to stress (e.g. wounding, drought and salt; Huh et al, 1997 Mol. Gen. Genet. 255(4): 382-391). Cells with higher levels of peroxidase are not as sensitive to stress as these with lower levels of peroxidase. This is because hydrogen peroxide which is produced as part of the stress response will be removed quickly by peroxidases (Dδrnenburg & Knorr, 1997 J. Agri. Food Chem. 45(10): 4173-4177). Lack of induction of this enzyme in lox2- peas suggest that these seeds do not respond in the same way to stress. The LOX-2 isoenzyme is therefore implicated as a stress enzyme. Ascorbate peroxidase activity was also determined but did not change as a result of wounding. Ascorbate peroxidase is an antioxidant enzyme in peas which removes hydrogen peroxide produced during aerobic metabolism and stress reactions (Hernandez et al, 1995 Plant Sci. 105(2): 151-167).

The combined effects of a differing stress response and a reduction of lipoxygenase-2 isoenzyme co-oxidation of ascorbate produces lox2- peas with significantly reduced post harvest losses in ascorbate.

Figure 3 illustrates the ascorbate loss for two standard varieties (Harrier and Novella) and these varieties with the lox2- trait introgressed, at 21 and 32°C at 2 hours post harvest.

Ascorbate oxidases

Ascorbate oxidase activity increased in lox2- peas by up to 90% on wounding, whereas ascorbate oxidase decreased in standard peas by 30% . The increase in levels for Iox2- peas is unexpected, but indicates that LOX-2 might have a role in suppression of this enzyme.

Lipid hydrolysis

Although the concentration of free choline in standard peas was lower than in Iox2- peas immediately after wounding, free choline released by the action of phospholipase D increased by 14.0% per hour during the first four hours after wounding. In lox2- peas an increase in free choline of only 0.5% per hour was seen over 6 hours. Again this suggests that Iox2- peas do not respond to stress in the same way as standard peas.

Corresponding data from the same peas show that the release of free fatty acids occurred sooner with time after wounding in standard compared to Iox2- peas. The loss of ascorbate in standard peas (2.14% per hour) was greater than in lox2- peas

(1.7%) .

Figure 4 illustrates the effect of wounding on marker metabolites in (a) standard and (b) lox2- peas after vining. Standard peas and lox2- peas differ in their response to stress caused by wounding. Standard peas show an increase in peroxidase activity and a decrease in ascorbate oxidase activity whilst lox2- peas show an increase in ascorbate oxidase activity. An immediate release of choline and free fatty acids seen on wounding in standard peas is not seen in lox2- peas, indicating LOX-2 isoenzyme has some role in mediating this response.

Example 3

Peas that are substantially lacking in plastidial phosphoglucomutase and lipoxygenase-2 isoenzyme activity.

(I)

To develop peas that are substantially lacking in the expression of plastidial phosphoglucomutase and lipoxygenase isoenzyme-2 a breeding programme was set up with the crossing of breeding lines expressing one of these two traits. In the starting lines each of the traits was expressed in a background of Harrier and Novella.

rug3 lines lox2- lines

(A) Harrier x sim 41 (C) Harrier x lox2-

(B) Novella x sim 43 (D) Novella X lox2-

Lines A and B have a mutation at the rug3 locus and C and D have a mutation at the lox-2 locus. Rug 3 lines were produced in accordance with the methodology outlined in WO 98/01574 whereas lox 2- lines were produced as outlined above, hereby further summarised. Original lox2- pea material was obtained from John Innes Centre, Norwich Research Park, Colney Lane, Norwich, England as lox21ox2 backcrossed in the round pea variety BIRTE. The following breeding programme was therefore conducted to put the lox2- trait in commercial pea varieties Harrier and Novella.

1994 Harrier x Birte Novella x Birte

FI Selfed

Winter 1994/95 F2 (test for lox 2-)

F3, F4, in glass house 1995 i

F5 glass house multiplication

Winter 1995/96

F5 seed on microplot

1997 Vined selection in field + F6 seed multiplication

Quality selection and multiplication

The lines A, B C and D could then be crossed according to the following table to produce peas showing the lox2- and rug 3 traits:

The following flow diagram outlines the further breeding programme that was undertaken:

98452 98453 98454 98455

1998 FI Selfed

I

F2

Winter 1998/99 (test for lox2- by DNA assay and Rug3 by visible seed characteristics)

F3, F4 in glass house

Winter 1999/2000 F5 glass house multiplication

(II)

Alternatively peas substantially lacking in the expression of lipoxygenase isoenzyme-2 and plastidial phosphoglucomutase may be developed by molecular genetic technique wherein transformation vectors are used to introduce the LOX-2 and/or RUG3 gene sequences or parts thereof, in an antisense orientation: Peas substantially lacking in expression of lipoxygenase type-2 isoenzyme.

(i) Isolation of LOX-2 polynucleotide sequence from pea:

In a technique based on the method of Edwards et al (Nucleic Acids Research (1991) 19 1349) , a small piece of pea leaf is removed and transferred to a homogenisation tube on ice. The leaf material is briefly ground using a pestle. 400μl of Edward's buffer (200mM Tris.HCl (pH 7.5), 250mM sodium chloride, 25mM ethylenediaminetetraacetic acid and 0.5% (w/v) sodium dodecyl sulphate) is added to the tube. This is centrifuged for 5 minutes at room temperature to pellet the plant cell debris.

300μl of the supernatant is added to 300μl of isopropanol, mixed and left for 5 minutes at room temperature to precipitate the nucleic acids. The sample is centrifuged again and the supernatant discarded. The tube is inverted to drain, and the pellet rinsed in 500μl of 70% aqueous ethanol. The tube is centrifuged as before, the supernatant is discarded and the tube inverted to drain. The DNA pellet is left to air dry for 60 minutes, then resuspended in lOOμl of TE buffer (lOmM Tris.HCl (pH 8.0) and ImM ethylenediaminetetraacetic acid).

The LOX2 fragment (shown in its sense orientation in Figure 6) is amplified by the polymerase chain reaction (PCR) as follows: lOμl of lOx PCR buffer, 3μl of 50mM magnesium chloride, lμl of nucleotide triphosphates (lOμM each), lμl of primer LI (5'- AACAGCTAGCACAAGATAAGAGGGACAGTG-3' (seq. I.D. no. 1)), lμl of primer L2 (5' -TTCACTCGAGCAAAATTTTCATCTCTTGGA-3' (seq. I.D. no.2)), lμl of pea DNA, 0.5μl of Taq DNA polymerase (5 U/μl) , 0.05μl of pfu DNA polymerase (5 U/μl) , 83.5μl of sterile water are put into a PCR tube and overlaid with 2 drops mineral oil.

The PCR is performed using the following conditions: 2.00 minutes at 94°C (denature), 0.50 minutes at 55°C (anneal), 3.00 minutes at 72°C (extend) and 0.75 minutes at 94°C (denature, remaining 35 cycles) .

Agarose gel electrophoresis shows that a 800 base pair fragment is produced, and this can be purified using a QIA quick PCR purification kit. Primers LI and L2 are based on the LOX2 sequence published by Casey (Genebank Accession N°: X17061) .

(ii) Ligation into vector pT7 :

The fragment can then be ligated into the pT7 vector by taking 15μl of the fragment, lμl of pT7 vector, 2μl of lOx ligation buffer and 2μl T4 DNA ligase (1 U/μl) and incubating at 4°C overnight .

(iii) Transformation into Escherichia coli cells and purification of transformed plasmid:

The resulting ligation is then transformed into competent XL1- blue Escherichia coli cells, prepared from lOOμl of an E. coli cell culture added to 25ml LB medium containing Tetracycline (12.5μg/ml) and incubated at 37°C overnight with shaking. 1ml of this culture is added to 100ml Lennox broth (LB) medium (0.5% (w/v) sodium chloride, 1% (w/v) yeast extract and 1% (w/v) Bactotryptone) without tetracycline and incubated for a further 2 - 3 hours. Bacteria are centrifuged and washed in lOOmM CaCl₂ then resuspended in 5ml CaCl? and left on ice for 60 minutes. 4μl of the ligation is added to 200μl competent E. coli cells and left on ice for 30 minutes.

The cells are then heat shocked at 42°C for exactly 40 seconds and 300μl LB medium added. The cells then undergo incubated at

37°C for 30 minutes with shaking and are subsequently plated onto LB agar plates containing tetracycline and Ampicillin

(50μg/ml) and incubated overnight at 37°C.

The presence of the recombinant plasmid in the bacterial colonies is checked by PCR using T7 and U19 as primers, and positive clones grown in 2xTY medium containing Ampicillin. Colonies demonstrating the correct fragment are grown in 50ml 2xTY medium with Ampicillin overnight at 37°C with shaking.

The plasmid is purified from the cell culture as follows: bacteria are pelleted by centrifugation, resuspended in 4ml solution 1 (50mM glucose, 25mM Tris.HCl (pH 8.0), lOmM EDTA (pH 8.0)). 8ml of solution 2 (0.2M sodium hydroxide, 1% sodium dodecyl sulphate) is added and the mixture incubated at room temperature for 5 minutes. 6ml buffer solution 3 (5M potassium acetate, 11.5ml glacial acetic acid, 28.5ml water) is added and the mixture incubated at 4°C for 15 minutes. The mixture is then strained through 2 layers of Miracloth and 10ml isopropanol added. The solution is centrifuged at and the pellet resuspended in 1ml TE buffer containing lOμl/ml RNase.

This mixture is incubated at 50°C for 20 minutes and chloroform/ phenol extracted once and chloroform extracted once. 0.7 volumes of isopropanol and 0.1 volumes of sodium acetate (3M, pH

5.2) is added and mixed. The suspension is then centrifuged at 13,000 rpm, room temperature for exactly 5 minutes and the precipitate air-dried.

The resultant DNA is diluted to 3μg/ml and restriction digests are set up. This requires 2μl DNA, 3μl lOx buffer, 3μl 0.1% bovine serum albumin, lμl of each restriction enzyme (10 U/μl) and sterile water added to 30 μl. The mixtures are incubated at 37°C for 45 minutes.

Treatments are as follows (1) Eco Rl, buffer H, (2) Nde 1/Nhe 1, buffer M, (3) Sal 1/Mun 1, buffer H. From the size of the resultant fragments on agarose gels, the insertion and orientation of the DNA are determined.

Restriction digests are set up as follows: Restriction digests set up as follows: (1) 2μl DNA, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Xho 1, lμl Nhe 1 and 20μl sterile water

(2) 0.5μl pP5LN vector, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Xho 1, lμl Nhe 1 and 21.5μl sterile water. The mixtures are digested for 90 minutes at 37°C and run on an agarose gel for 2 hours. Bands at approximately 800 base pairs are cut out from digestion 1 preparations and the band for preparation 2 is cut out.

The DNA is purified using a QIAquick gel extraction kit and fragments (1) and (2) ligated using a mixture of 2μl lOx ligation buffer, 2μl T4 DNA ligase, 5μl (2) and llμl (1) . The mixture is incubated overnight at 4°C. The ligations are inserted into competent E. coli cells as before. After checking for insertion of the fragment by PCR as before, bacteria containing the recombinant plasmid are grown up overnight as before.

Plasmid DNA is again purified, diluted to 0.5μg/μl and restriction digests set up using a mixtures of 7μl DNA, 3μl lOx buffer, 3μl 0.1% bovine serum albumin, lμl of each restriction enzyme and sterile water to 30μl. These mixtures are incubated at 37°C for 45 minutes.

Treatments are as follows (1) Eco Rl, buffer H, (2) Hin DIII/Nco 1, buffer H and (3) Hin DIII/Eco RV, buffer B. The resulting solutions are run on an agarose gel to determine correct insertion of the LOX2 antisense fragment.

The fragments are sequenced using the following primers: (L3) 5'-ACTTGTCAAGTATAGAGAAG-3' (seq. I.D. no. 3), (L4) 5'- GCTGGTGAATCTGCATTCAA -3' (seq. I.D. no. 4), T7 and U19. The fragment is then digested and ligated into the pGPTV vector as follows: (LOX2 antisense fragment) 2μl of DNA, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Hin Dili, lμl Eco Rl and 14μl sterile water, and (vector) 2μl pGPTV vector, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Hin Dili, lμl Eco Rl and 20μl sterile water are incubated at 37°C for 60 minutes.

After running on an agarose gel, the appropriate bands are cut out and purified using a QIAquick gel purification kit. The DNA is ligated into the vector as follows: 3μl 10 x ligation buffer, 3μl T4 DNA ligase, 4μl vector and 20μl of the LOX2 antisense DNA. This is incubated overnight at 4°C and transformed into E. coli cells as before. The cells are plated on LB agar plates containing Tetracycline and Kanamycin (50μg/ml) and incubated overnight at 37°C. (The pGPTV vector contains the Kanamycin resistance gene necessary for clonal selection) . Colonies are checked for the presence of the plasmid containing the LOX2 antisense fragment by PCR as before, except 30035S and nos-as are used as primers, and the time for extension is 1 minute. Colonies containing the L0X2 antisense fragment are grown in 50ml 2xTY medium with Kanamycin, overnight at 37°C with shaking and the plasmid purified as before. The plasmid is checked for correct insertion of the gene by restriction digests as follows:

Plasmid DNA is purified as before, diluted to and restriction digests set up as follows: 7μl DNA (3μg/μl) , 3μl lOx buffer, 3μl 0.1% bovine serum albumin, lμl of each restriction enzyme and sterile water to 30μl. The mixtures are incubated at 37°C for 45 minutes. Treatments are as follows, (1) Xmn 1, buffer B, (2) Hin DIII/Eco Rl, buffer B, (3) Hin Dili/Xho 1, buffer L and (4) Nhe 1/Sna Bl, buffer M.

(iv) Agrobacterium transformation:

The recombinant plasmid is then transformed into competent Agrobacterium tumefaciens (strain LBA 4404) and the bacteria grown on LB agar plates containing Kanamycin. Bacteria containing the LOX2 antisense fragment are grown overnight in LB medium at 28°C with shaking. The overnight culture is centrifuged and the cell pellet is resuspended in liquid MS basal medium supplemented with 3% sucrose. (v) Genetic modification of pea plant tissue and the propagation thereof:

By way of example the applicants herein provide a detailed account of a process that can be followed for genetic modification, by agrobacterium transformation, of pea as published by Grant et al (Plant Science (1998) 139 159-164, Plant Cell Reports (1995) 15 254-258). It is recognised that it would be well within the capacity of the skilled person in the art to make any minor modifications to this process, that may be specific to the transformation of a particular plant or plant tissue.

Briefly, immature pea pods are harvested at the eating stage and the pods surface sterilised in 0.75% (w/v) sodium hypochlorite with Tween 80 for 20 minutes. Seeds are removed and cut in half from the seed attachment scar. Half the seed, distal from to the embryonic axis, was discarded. The testa is removed and the embryonic axis is cut out adjacent to the cotyledonary stalk and discarded. The remaining cotyledon segments were immersed for one hour in an overnight culture of A. tumefaciens.

Cotyledons were then plated on B5 medium (Gamborg et al Experimental Cell Research (1968) 50 151-158) containing 1.3mg/l 6-benzylaminopurine, 30g/l sucrose, 8g/l agar (pH 5.5) and 19.6mg/l acetosyringone and are placed so that the cotyledonary stalk is perpendicular to the medium. After six days, cotyledons are washed three times with sterile water and once with 400mg/l timentin. Cotyledons are blotted and transferred at two to four-weekly intervals to B5 medium containing 1.3mg/l benzylaminopurine, 30g/l sucrose, 8g/l agar (pH 5.8), 75mg/l kanamycin sulphate and 200mg/l timentin. After three to four transfers, the cotyledon tissue is cut away from the growing callus and shoots. When shoots reach more than 10mm, they are excised and transferred to root initiation medium (B5 with lmg/1 indole-3-butyric acid, 30g/l sucrose, 8g/l agar and 200mg/l timentin (pH 5.8).

After 7 days, shoots are transferred to B5 medium with no plant growth regulators, 200mg/l timentin and 75mg/l kanamycin sulphate for reselection and root elongation. Where axillary shoots elongate at the base of the main stem, they are cut off and placed in rooting medium. In this way some multiplication occurs.

Shoots approximately 4cm in height with roots actively growing on selection medium are transferred to soil and grown in a glasshouse 16 hours 18 - 23°C day/8 hours 12 - 16°C night) .

Leaves are tested by PCR for the LOX2 antisense fragment when the plants are large enough to remove a small sample. The sample is transferred to a homogenisation tube on ice and the material is ground using a pestle. 400μl of Edward's buffer (0.2M Tris.HCl (pH 7.5), 0.25M sodium chloride, 25mM EDTA, 0.5% sodium dodecyl sulphate) is added to the tube and the sample is centrifuged at room temperature to pellet the plant cell debris. 300μl of the supernatant is added to 300μl of isopropanol to precipitate the nucleic acids. The suspension is left for 5 minutes at room temperature and the sample is centrifuged again. The supernatant is discarded and the tube invert tube to drain. The pellet is rinsed in 500μl of 70% ethanol and the suspension centrifuged as before. The supernatant is discarded and the tube inverted drain. The DNA pellet is air dried for 60 minutes and the DNA re-dissolved in lOOμl of TE buffer. PCR is performed as before, using 30035S and L4 as primers. Figure 5 shows LOX2 sense fragment sequence from Pisum sativum (Reverse orientation, 5' -3') seq. I.D. no. 5;

Peas substantially lacking in expression of plastidial phosphoglucomutase .

(i) Isolation of RUG3 polynucleotide sequence from pea:

The RUG3 fragment (shown in its sense orientation in Figure 7) is isolated in a similar manner to the LOX-2 above modified such that the RUG3 fragment is amplified by the polymerase chain reaction (PCR) as follows: lOμl of lOx PCR buffer, 3μl of 50mM magnesium chloride, lμl of nucleotide triphosphates (lOμM each), lμl of primer Rl (5' -TACAGCTAGCACAACTTCATCATCTCTGCG-3' (Seq. I.D. no.6)), lμl of primer R2 (5'- CTTTCTCGAGGCTGAAAATCGAACACTGTC-3', (Seq. I.D. no.7)), lμl of pea DNA, 0.5μl of Taq DNA polymerase (5 U/μl), 0.05μl of pfu DNA

polymerase (5 U/μl), 83.5μl of sterile water are put into a PCR tube and overlaid with 2 drops mineral oil.

The PCR is again performed using the following conditions: 2.00 minutes at 94°C (denature), 0.50 minutes at 55°C (anneal), 3.00 minutes at 72°C (extend) and 0.75 minutes at 94°C (denature, remaining 35 cycles) .

Agarose gel electrophoresis shows that a 800 base pair fragment is produced, and this can be purified using a QIA quick PCR purification kit.

(ii) Ligation into vector pT7 :

The fragment can then be ligated into the pT7 vector by taking 15μl of the fragment, lμl of pT7 vector, 2μl of lOx ligation buffer and 2μl T4 DNA ligase (1 U/μl) and incubating at 4°C overnight.

The resulting ligation is then transformed into competent XL1- blue Escherichia coli cells as disclosed above.

The resultant DNA is diluted to 3μg/ml and restriction digests are set up. This requires 2μl DNA, 3μl lOx buffer, 3μl 0.1% bovine serum albumin, lμl of each restriction enzyme (10 U/μl) and sterile water added to 30μl. The mixtures are incubated at 37°C for 45 minutes. Treatments are as follows (1) Hin Dili, buffer B, (2) Nde 1/Nhe 1, buffer M, (3) Sal 1/Dra 1, buffer H. From the size of the resultant fragments on agarose gels, the insertion and orientation of the DNA are determined.

(2) 0.5μl pP5LN vector, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Xho 1, lμl Nhe 1 and 21.5μl sterile water. The mixtures are digested for 45 minutes at 37°C and run on an agarose gel for 2 hours. Bands at approximately 800 base pairs are cut out from digestion 1 preparations and the band for preparation 2 is cut out.

The DNA is purified as above and the ligations are once again inserted into competent E. coli cells. After checking for insertion of the fragment by PCR as before, bacteria containing the recombinant plasmid are grown up overnight as before.

Treatments are as follows (1) Eco Rl/Hin Dill, buffer B, (2) Hin DIII/Nco 1, buffer B and (3) Hin DIII/Dra 1, buffer B. The resulting solutions are run on an agarose gel to determine correct insertion of the RUG3 antisense fragment. The fragments are sequenced using the following primers: (R3) 5'-GAAAAAAGGTGAAAGTGTTT-3' (Seq. I.D. no. 8), (R4) 5'- GTTTATCATGAGTGCGAGCC-3' (Seq. I.D. no. 9), T7 and U19. The fragment is then digested and ligated into the pGPTV vector as follows: (RUG3 antisense fragment) 2μl of DNA, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Hin Dili, lμl Eco Rl and 14μl sterile water, and (vector) 2μl pGPTV vector, 3μl lOx B buffer, 3μl 0.1% bovine serum albumin lμl Hin Dili, lμl Eco Rl and 20μl sterile water are incubated at 37°C for 60 minutes.

After running on an agarose gel, the appropriate bands are cut out and purified using a QIAquick gel purification kit. The DNA is ligated into the vector as follows: 3μl 10 x ligation buffer, 3μl T4 DNA ligase, 4μl vector and 20μl of the RUG3 antisense DNA. This is incubated overnight at 4°C and transformed into E. coli cells as before.

The cells are plated on LB agar plates containing Tetracycline and Kanamycin (50μg/ml) and incubated overnight at 37°C. Colonies are checked for the presence of the plasmid containing the RUG3 antisense fragment by PCR as before, except 30035S and nos-as are used as primers, and the time for extension is 1 minute. Colonies containing the RUG3 antisense fragment are grown in 50ml 2xTY medium with Kanamycin, overnight at 37°C with shaking and the plasmid purified as before. The plasmid is checked for correct insertion of the gene by restriction digests as follows:

Plasmid DNA is purified as before, diluted to and restriction digests set up as follows: 7μl DNA (3μg/μl) , 3μl lOx buffer, 3μl

0.1% bovine serum albumin, lμl of each restriction enzyme and sterile water to 30μl. The mixtures are incubated at 37°C for 45 minutes. Treatments are as follows, (1) Xmn 1, buffer B, (2) Hin DIII/Eco Rl, buffer B, (3) Hin DHI/Xho 1, buffer L and (4) Nhe 1/Hin Dili, buffer M.

(iv) Agrobacterium transformation:

The recombinant plasmid is then transformed into competent Agrobacterium tumefaciens (strain LBA 4404) and the bacteria grown on LB agar plates containing Kanamycin. Bacteria containing the RUG3 antisense fragment are grown overnight in LB medium at 28°C with shaking. The overnight culture is centrifuged and the cell pellet is resuspended in liquid MS basal medium supplemented with 3% sucrose.

(v) Genetic modification of pea plant tissue and the propagation thereof:

Immature pea embryos can then be transformed and propagated as described above to introgress the rug3 trait.

Figure 6 shows RUG3 sense fragment sequence from Pisum sativum (Reverse orientation, 5' -3') Seq. I.D. No. 10.

It would be recognised by the person of average skill in the art that either the approach of conventional breeding, genetic transformation or a combination of the two could be followed to introduce the desired lox2- and rug3 traits into a commercial pea line.

Claims

1. Method of improving the water soluble antioxidant content in mechanically harvested peas, characterised by the use of a genetic modification capable of reducing the activity of ipoxygenase type-2 isoenzyme in said peas.

2. Method according to claim 1, wherein the genetic modification is at the LOX-2 locus.

3. Method according to claim 1 to 2, wherein the water soluble antioxidant content comprises ascorbate.

4. Method of improving the ascorbate content in mechanically harvested peas by the combination of a genetic modification capable of reducing the activity of lipoxygenase type-2 isoenzyme with a further genetic modification at any locus capable of reducing the activity of one or more enzymes involved in the sucrose to starch biosynthetic pathway.

Method according to claim 4, wherein said genetic modifications are capable of reducing lipoxygenase type-2 isoenzyme and plastidial phosphoglucomutase enzyme activities.

6. A method according to claim 5, further characterised in that the reduction in lipoxygenase type-2 isoenzyme activity is achieved by a genetic modofication at the LOX-2 locus and the reduction in plastidial phosphoglucomutase activity is achieved by a genetic modification at the rug3 locus.

7. Peas obtained by the method according to any one of claims 4 to 6.

8. Peas obtained to claim 7 wherein said peas are in a frozen state.

9. A modified pea plant or part thereof with an improved ascorbate content, characterised by reduced activity of lipoxygenase type-2 isoenzyme and plastidial phosphoglucomutase compared to an equivalent unmodified plant .

10. A modified pea plant or part thereof according to claim 9, wherein said modified pea plant or part thereof comprises genetic modifications at the lox2 and rug3 loci.